Synthesis, physicochemical properties of allopurinol derivatives and their biological activity against Trypanosoma cruzi

Article abstract of DOI:10.1016/j.ejmech.2013.08.045

Chagas disease is caused by Trypanosoma cruzi (T. cruzi) leading to a huge number of infections and deaths per year, because in addition to many sufferers only having limited access to health services only an inefficient chemotherapy is available using drugs such as benznidazole and nifurtimox. Here, C6-alkyl (2a-c) and N1-acyl (3a-c) derivatives of Allopurinol (Allop, compound with activity against T. cruzi) were synthesized in good yields and their structures were unambiguously characterized. Only 2a, 2b and 3c showed inhibitory activity against the proliferative stages of the parasite when tested at 1 μg mL ^-1 with the 3c derivative exhibiting an IC50 value similar to that of Allop and not being toxic for mammalian cells. Relevant pharmaceutical physicochemical properties (pKa, stability, solubility, lipophilicity) were also determined as well by using Lipinski's rule, polar surface area and molecular rigidity. Taken together, the results demonstrated that the studied derivatives had optimal properties for bioavailability and oral absorption. For the stability studies, Micellar Liquid Chromatography was used as the analytical method which was fully validated according to the FDA guidelines and shown to be a suitable, sensitive and simple method for routine analysis of these Allop derivatives.

Full text of DOI:10.1016/j.ejmech.2013.08.045

European Journal of Medicinal Chemistry 69 (2013) 455e464
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis, physicochemical properties of allopurinol derivatives and
their biological activity against Trypanosoma cruzi
,
b
,
,
,
, 2
M.A. Raviolo a , M.E. Solana 2, M.M. Novoa b 2, M.S. Gualdesi a 1, C.D. Alba-Soto b
,
*
,
M.C. Briñón a
*
a Departamento de Farmacia, Facultad de Ciencias Químicas, Universidad Nacional de Córdoba, 5000 Córdoba, Argentina
b Instituto de Microbiología y Parasitología Médica (IMPaM-UBA-CONICET), Facultad de Medicina, Universidad de Buenos Aires,
1121 Ciudad Autónoma de Buenos Aires, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
Chagas disease is caused by Trypanosoma cruzi (T. cruzi) leading to a huge number of infections and
deaths per year, because in addition to many sufferers only having limited access to health services only
an inefficient chemotherapy is available using drugs such as benznidazole and nifurtimox. Here, C6-alkyl
(2aec) and N1-acyl (3aec) derivatives of Allopurinol (Allop, compound with activity against T. cruzi)
were synthesized in good yields and their structures were unambiguously characterized. Only 2a, 2b and
Received 30 May 2013
Received in revised form
28 August 2013
Accepted 31 August 2013
Available online 13 September 2013
3c showed inhibitory activity against the proliferative stages of the parasite when tested at 1 m
g mLꢀ1
with the 3c derivative exhibiting an IC50 value similar to that of Allop and not being toxic for mammalian
cells. Relevant pharmaceutical physicochemical properties (pKa, stability, solubility, lipophilicity) were
also determined as well by using Lipinski’s rule, polar surface area and molecular rigidity. Taken together,
the results demonstrated that the studied derivatives had optimal properties for bioavailability and oral
absorption. For the stability studies, Micellar Liquid Chromatography was used as the analytical method
which was fully validated according to the FDA guidelines and shown to be a suitable, sensitive and
simple method for routine analysis of these Allop derivatives.
Keywords:
Trypanosoma cruzi
Allopurinol derivatives
Physicochemical properties
Trypanocidal activity
Ó 2013 Elsevier Masson SAS. All rights reserved.
1. Introduction
ingestion of contaminated food as well as by blood transfusion
and congenital passage [1]. The global increase in human mi-
Chagas disease (Chagas) or American Trypanosomiasis is caused
by the protozoan Trypanosoma cruzi (T. cruzi), and remains a major
health problem in Central and South America that affects nearly 8e
10 million people throughout Latin America [1].
In endemic countries, T. cruzi is transmitted via haematopha-
gous triatomine insect vectors that release infective forms within
the feces and urine after every mammalian blood ingestion. Once
in the host, parasites invade cells and undergo differentiations
into intracellular amastigotes (AMAS), which after several rounds
of duplication transforms them into trypomastigotes (TRYP), the
form that disseminates the infection. Transmission may also
occur through laboratory accidents, organ transplantation,
grations is responsible for most of the cases reported in non-
endemic countries [2].
Paradoxically, 100 years after the first report describing the
morphology and the life cycle of the pathogen, neither vaccines nor
effective treatments for chronic cases are available. One reason for
this may be that Chagas disease belongs to a group of tropical in-
fections that are endemic mainly among low-income populations
of the developing regions of Africa, Asia, and America, which are
named neglected diseases [3]. The lack of interest of pharmaceutical
companies and the absence of effective social policies from the
affected states are responsible for the limited evolution toward an
improved pharmacotherapy.
Chagas disease initiates as an acute phase that is usually
asymptomatic. Then, after decades of chronic infection, 30e40% of
the infected people develop symptoms that can include cardiac
and/or digestive (megaesophagus or megacolon) forms. Despite
significant progress having been made in understanding the
biochemistry and physiology of the causative agent, the current
etiological treatment of Chagas disease is based on rather
* Corresponding authors. Tel.: þ54 351 5353865; fax: þ54 3515353364.
E-mail addresses: moraviolo@fcq.unc.edu.ar (M.A. Raviolo), macribri@
fcq.unc.edu.ar (M.C. Briñón).
1
Tel.: þ54 351 5353865x53355.
2
Tel.: þ54 11 59509500x2189; fax: þ54 11 59509577.
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.08.045

456
M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
responses, indicative of a reduction of the parasite burden, and
a)
b)
thereby sustained the feasibility of administration of two antipar-
asitic drugs in the chronic phase of Chagas disease. In addition,
these two drugs when administered together were reported to be
safe and effective in the treatment of Chagas disease reactivation
after heart transplantation [16].
The variable efficacy of Allop depends on the infective parasite
population, which varies among geographical areas [17]. In addi-
tion, in mammals, Allop is converted into oxypurinol by xanthine
oxidase (Fig. 1b) with a t1/2 ¼ 1e2 h, which is not a substrate for
HGPRT and has no anti-T. cruzi activity [10].
Although Allop presents trypanocidal activity, its failure in
avoiding Chagas disease progression may be due in part to inade-
quate blood levels caused by unfavorable physicochemical prop-
erties, thus leading to versatile responses in humans.
In an attempt to improve its performance, we have developed a
series of derivatives of Allop by chemical modification of their
specific functional groups, resulting in active anti-T. cruzi agents per
se or prodrug compounds. Then, their activities against T. cruzi as
well as their different relevant physicochemical properties, such as
their integrity in buffers, simulated fluids and human plasma were
determined, in addition to carrying out hydrolytic and oxidative
studies, and determining acid dissociation constants, lipophilicity
and solubility.
Fig. 1. Allopurinol (a) and its metabolite oxypurinol (b).
unspecific drugs developed more than four decades ago, such as
nifurtimox and benznidazole (Bz). Moreover, although the use of
these drugs to treat the acute phase of the disease is widely
accepted, etiological treatment at the chronic phase remains
controversial [4], with the undesirable side effects of both drugs
being major drawbacks and frequently forcing treatment to be
discontinued [5]. Research on anti-T. cruzi compounds has been to
date based on different strategies aimed at targeting specific
parasite enzymes or parasite DNA or at inducing oxidative stress
damage [6,7].
Nevertheless, the development of safer and more efficient try-
panocidal drugs remains a major goal in Chagas disease chemo-
therapy, with one possibility being to exploit differences in the
purine metabolism between T. cruzi and host cells. Although T. cruzi
does not synthesize purines de novo as in the case of as mammals,
the parasite is able to concentrate the pyrazolopyrimidines within
the cell and metabolize them as purines through a salvage pathway,
ultimately incorporating them into nucleic acids [8,9].
Allopurinol (1, Allop 1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-
4-one, Fig. 1a) employed for the treatment of hyperuricemia, is an
hypoxantine analog used by the T. cruzi’s hypoxanthineeguanine
phosphoribosyltransferase (HGPRT) as an alternative substrate.
This enzyme can incorporate Allop into the parasite’s ribonucleic
acid as a non-physiological nucleotide; thus blocking the synthesis
of new purine nucleotides [10,11]. The activity of Allop in chemo-
therapy for Chagas disease has been extensively investigated, but
the results are somewhat contradictory [12e14]. In a recent pilot
study, Perez-Mazliah et al. [15] showed that the combination of
Allop and Bz induced significant modifications of the T and B cell
2. Results and discussion
2.1. Chemistry
Taking into account that C6 of Allop is a target of metabolism for
the xanthine oxidase enzyme, a series of derivatives with alkyl
groups on this position (2aec) were synthesized in two steps,
obtaining 6-methyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-
one (2a), 6-ethyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(2b) and 6-propyl-1,5-di hydro-4H-pyrazolo[3,4-d]pyrimidin-4-
one (2c) by employing the method of Cheng and Robins [18]
(Scheme 1a). It is important to point out that Biagi et al. have
found that these compounds to be ineffective inhibitors of this
enzyme, and that neither physicochemical properties nor activity
against T. cruzi were previously studied [19].
N1-acyl prodrugs of Allop were also synthesized (3aec) to
improve their physicochemical properties; whose design was based
a)
b)
Scheme 1. Synthesis procedures used for the development of derivatives. a) C6-alkyl derivatives of Allop (2aec), b) N1-acyl derivatives of Allop (3aec).

M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
457
100
80
60
40
20
0
with the load of AMAS in cell monolayers being reduced by treat-
ment with 5 or 50 M of each drug (p < 0.001) (Fig. 3b). Both the
Allop and 3c compounds displayed a dose-dependent trypanocidal
effect, showing IC50 values of 9.62 and 11.30 M, respectively.
Neither Allop nor the 3c derivatives were toxic for mammalian
cells at the concentration tested (Fig. 4).
m
m
2.3. Physicochemical properties
Bz
2.3.1. Acid dissociation constants (pKa)
2a-a
2a-c 3a-a
3a-c
2a-b
3a-b
There are different experimental and theoretical methods for
determining the pKa [24], with software packages usually being
employed to estimate this parameter. Meloun and Bordovská [25]
evaluated the accuracy of the pKa data generated by Advanced
Chemistry Design (ACDpKa) packages, and concluded that this
software provided the most accurate pKa prediction. In order to
validate this theoretical method, we obtained the pH-solubility
profile of Allop (data not shown), which gave a pKa value of 9.7,
being similar to that obtained by ACDpKa software. All these values
are reported in Table 1.
Although the Allop drug molecule has two acidic groups, for all
the derivatives studied only the ionization of the NH-5 atom was
observed, with compounds 2aec showing similar pKa values (pKa
9.3) to Allop (pKa 9.7). However, compounds 3aec presented lower
pKa values than the parent compound (pKa 7.6), due to the eC(O)e
R substituent in the N ꢀ 1 atom providing more acidic character-
istics to the molecules. The pKa values of 3aec are in agreement
with those reported by Bundgaard and Falch [20].
Fig. 2. Growth inhibition of T. cruzi amastigotes by C6-Alkyl (2aec) and N1-Acyl (3ae
c) derivatives of Allopurinol. AMAS were cultured in Vero cell monolayers in the
presence of 1 m
g mLꢀ1 of each compound. The graph represents the rate of growth
inhibition related to untreated T. cruzi-infected cells (0%). Data are representative of
two independent experiments. The dashed line shows the effect of Benznidazole (Bz)
as the reference drug. Derivatives were selected on the basis of their behavior
compared to Bz.
on the removal of the NH-acidic proton. Thus, N1-acyl derivatives of
Allop,
1-acetyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(3a), 1-propanoyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(3b) and 1-butanoyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-
one (3c) were prepared according to the Bundgaard procedures
(Scheme 1b) [20]. In this way, one hydrogen bond donor group was
eliminated, thereby having a significant effect on relevant proper-
ties such as crystal lattice energy, solubility, dissolution rate, and
permeability [21].
The structures of compounds 2aec and 3aec were character-
ized by high resolution mass spectrometry and spectroscopic
methods. 1H NMR, 13C NMR, HSQC-DEPT (Heteronuclear Single
Quantum Correlation-Distortionless Enhancement by Polarization
Transfer) and HMBC (Heteronuclear Multiple Bond Correlation) of
2aec and 3aec were performed on DMSO-d6. Although these
compounds have been previously reported, the NMR bidimensional
techniques allowed us to unequivocally reassign the chemical shifts
of 3aec to those reported by Bundgaard and Falch [20]. In this way,
in the only two protons of Allop bonded to carbon atoms (CH) (H-6,
2.3.2. Integrity
Stability indicating methods have become an important aspect
of Food and Drug Administration (FDA) requirements. In addition,
there is an increasing tendency toward the development of stability
indicating assays, using the approach of stress testing as mentioned
in the International Conference on Harmonisation (ICH, guidelines
Q1A) [26] and Guidance for Industry of FDA [27] concerning an
in vivo bioavailability that included a discussion on gastrointestinal
(GI) stability. Another important aspect to consider is the stability
of drug molecules in plasma, which provides useful information for
in vivo studies and potential structural modifications.
Stress testing also encompasses the influence of, among others,
oxidizing agents, light, temperature and susceptibility over a wide
range of pH values. In contrast, the FDA has recommended stability
studies of 1 h in simulated gastric fluid (SGF; pH 1.2 with pepsin)
and 3 h in simulated intestinal fluid (SIF; pH 6.8 with pancreatin) in
order to mimic GI conditions. Thus, a significant degradation (>5%)
of a drug molecule could suggest potential instability in the
gastrointestinal tract (GIT) [27]. For this reason, the aim of the
current study was also to determine the inherent stability of 2aec
and 3aec under these conditions, using a validated Micellar Liquid
Chromatography (MLC) as the analytical method.
d
8.40 and H-3,
observed in 3aec derivatives, leading to signals for H-3 and H-6 of
8.28, respectively. In addition, the structure of
d 8.34), an inversion of their chemical shifts was
d
8.35 ꢁ 0.01 and
d
2aec was confirmed by the HSQC-DEPT technique, thereby
showing that C-6 is a quaternary carbon. All the 1H and 13C NMR
data of the compounds given in the Experimental section are in full
agreement with the proposed structures, with the representative
HSQC-DEPT and HMBC for Allop, 2c and 3c being shown in the
Supplementary Materials.
2.2. Activity
The inhibitory effect of each C6-alkyl- and N1-acyl derivative of
Allop against TRYP was measured, but no compound displayed any
consistent trypanocidal activity in the range of concentrations
assayed (data not shown). To test whether the derivatives would
display any in vitro activity against the intracellular forms of T. cruzi,
2.3.2.1. Validation method. The FDA guidelines [28] were followed
to validate the MLC method, by evaluating linearity, limit of
detection, limit of quantification, intra and inter-day precision,
selectivity and recovery parameters. This method was validated for
all compounds according to their chemical structures; Method A for
2aec and Method B for 3aec (see Experimental section). Fig. 5a,b
shows chromatograms obtained from a mixture of 2aec and 3aec
in their respective matrices, showing that there was no interference
between Allop and their derivatives. These results revealed that
2aec and 3aec could be quantified by MLC in the presence of Allop
as a degradation product.
1 m
g mLꢀ1 of each compound was assayed as previously stated [22],
with Fig. 2 showing that Allop as well as the 2a, 2b and 3c, de-
rivatives inhibited T. cruzi amastigotes (AMAS) growth in a better, or
at least similar way to Bz.
In order to estimate their IC50 values, T. cruzi infected cell
monolayers were subjected to 0.5e50 mM of each derivative and
the percentage of infection was calculated. As can be seen in Fig. 3a,
only the Allop and 3c compounds were able to reduce the prolif-
eration of intracellular AMAS of virulent RA strain in Vero cells [23],
2.3.2.1.1. Calibration parameters, limit of detection (LOD) and
limit of quantification (LOQ). The calibration graphs for all

458
M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
a)
ALLOP
2a-a
100
50
0
2a-b
3a-c
0
5
0
5
0
5
0
5
5.
0.
5.
0.
5.
0.
5.
0.
50.0
50.0
50.0
50.0
M
b)
200
150
100
50
ALLOP
3a-c
RPMI-10% FCS
*
**
***
0
0
0.5
5
50
M
Fig. 3. Effect of 2a, 2b and 3c derivatives of Allopurinol on the growth of T. cruzi amastigotes. AMAS were cultured in triplicate in the presence of 0.5e50 mM of each compound. a)
The rate of infection was determined as the percentage of the treated T. cruzi-infected cells over the basal untreated T. cruzi-infected cells. b) The number of AMAS/cell was counted
over a total of 100 infected cells, as described in Experimental section. Results are expressed as means ꢁ SE from three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001
from the Bonferroni post-test.
compounds were constructed using the areas of the chromato-
graphic peaks (triplicate injections) obtained at seven different
concentrations equally distributed in the range 5 ꢂ 10ꢀ9 to
5 ꢂ 10ꢀ7 mol mLꢀ1. The adjusted parameters obtained are shown in
Table 1S, and the regression coefficients (r2) were always higher
than 0.9998.
The LOD for Allop, 2aec and 3aec was calculated using the 3s
criterion (three times the standard deviation of the lowest con-
centration solution included in the calibration divided by the slope
of the calibration graph) for a series of 10 solutions containing low
concentrations of each compound (Table 1S). The LOQ was then
100
50
0
Table 1
Acid dissociation constants (pKa), intrinsic solubility (S0) and melting point (mp, ꢃC)
of Allop and its derivatives.
Comp
pKa
S0
mp
25 ꢃC
37 ꢃC
Allop
2a
2b
2c
3a
9.7a; 9.3 ꢁ (0.4)b
9.3 ꢁ (0.4)b
0.49 ꢁ (0.05)d
1.48 ꢁ (0.10)d
0.71 ꢁ (0.06)d
2.07 ꢁ (0.05)d
>65e
0.90 ꢁ (0.04)d
2.02 ꢁ (0.30)d
0.94 ꢁ (0.03)d
2.56 ꢁ (0.15)d
>65e
>350
304
325
292
267
275
233
9.3 ꢁ (0.4)b
9.3 ꢁ (0.4)b
7.6 ꢁ (0.4)b; 7.7c
,
c
c
3b
3c
7.6 ꢁ (0.4)b
>65e
>65e
Fig. 4. Effect of 3c derivative on murine splenocyte viability. Cultures were carried out
on 96-well plates with 2 ꢂ 106 cells mLꢀ1 for 32 h in the presence of compound 3c at a
concentration of 50 ꢂ IC50. Cell viability was determined by the Alamar BlueÔ method
according to the instructions of the manufacturer. Data are expressed as the per-
centage of viability compared to control (non-treated cells). Bz and Allop were
considered as reference drugs. Bars represent the means ꢁ SE of two experiments
carried out in duplicate.
,
7.6 ꢁ (0.4)b
>65e
>65e
a
Calculated from the pH-solubility profile (25 ꢃC).
ACDpKa.
b
c
Bundgaard et al.
d
e
mg mLꢀ1
.
m
g mLꢀ1
.

M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
459
b)
a)
Fig. 5. MLC chromatograms showing: (a) Allop and 2aec derivatives in simulated intestinal fluid, (b) Allop and 3aec derivatives in simulated gastric fluid. For details of the
chromatographic conditions, see Section 4.5.
selected for the lowest concentrations used in the calibration
graphs.
measurements of the intra-day precision values taken on ten days
over a 3-month period, performed by different analysts and using
various apparatus at the same concentrations. In Table 3S, the re-
sults of intra- and inter-day precisions are expressed as the per-
centage of the relative standard deviation (RSD, %) for the intra- and
inter-day values. In all cases, the RSD values were lower than 1.85%.
These results showed that the proposed MLC method was
suitable for the analysis of these compounds in buffer solutions and
also in the SGF and SIF samples. Therefore, the procedure devel-
oped here can be used for the quality control, routine analyses and
pharmacokinetic studies of all the Allop derivatives.
2.3.2.1.2. Recovery and precision (intra- and inter-day). For the
recovery assays, three different concentrations (c1 ¼ 7.5 ꢂ 10ꢀ9
,
c2 ¼ 6 ꢂ 10ꢀ8 and c3 ¼ 3.5 ꢂ 10ꢀ7 mol mLꢀ1, ten replicates for each
standard) for all compounds were spiked in the buffers (pH 1.2 and
pH 6.8) and in the SGF and SIF matrices, and then analyzed in the
proposed mobile phases for each series of compounds. The stan-
dard solutions were processed and analyzed following the above-
described procedure and relative (analytical) recovery was calcu-
lated by comparing the obtained concentration from the standard
solutions with those of the working ones. Table 2S lists an average
of the percentage of recovery in all matrices, which revealed
satisfactory recoveries and no significant deviations for all the
studied compounds.
2.3.2.2. Degradation studies of Allop derivatives: hydrolytic, oxidative
and buffer solutions (pH 1.2 and pH 6.8), SGF, SIF and human plasma.
The accelerated degradation studies for the six derivatives of Allop
were assayed under the ICH and FDA recommendation procedures
in order to identify their degradation products. In all cases, how-
ever, only Allop was detected. Table 2 shows the percentages of
degradation (%), t1/2 and kinetic constants (kobs, where appropriate)
for Allop, 2aec and 3aec compounds under-hydrolytic and
oxidative conditions as well as in buffers (pH 1.2 and 6.8), SGF, SIF
and human plasma matrices.
Under stress conditions, only up to 10% of the degradation
process was observed for derivatives 2aec in NaOH (0.1 M) at 90 ꢃC
after 16 h (3.7e10.4%), whereas for 3aec an immediate degradation
took place not only in NaOH (0.1 M) but also in the HCl (0.1 M) and
H2O2 (3.5%) solutions. Although, under FDA conditions, the 2aec
derivative was stable in all media, 3aec suggested instability in the
GIT. For 3aec in SGF and pH 1.2, up to 32% of degradation was
observed during 1 h, with no significant differences observed in
their kobs for pepsin action. In contrast, in SIF and pH 6.8 conditions,
the 3aec derivatives underwent a 100% degradation after 3 h of
reaction, and pancreatin may have been responsible for this pro-
cess. In human plasma, Allop and 2aec were stable at 100%, while
the 3aec derivatives were very unstable and interacted in all the
stability studies. It was also observed that all compounds followed a
pseudo-first-order degradation kinetics.
The intra- and inter-day precisions were determined by
analyzing all compounds at the same concentrations of the recov-
ery. The intra-day precision was evaluated by injecting these three
test solutions ten times on the same day whereas the inter-day
precision determination involved calculating the average of ten
Table 2
Stability studies of Allop and its derivatives in sodium hydroxide (NaOH, 0.1 M,
90 ꢃC), hydrochloric acid (HCl, 0.1 M, 90 ꢃC), hydrogen peroxide (H2O2) 30% for 2aec
and 3.5% for 3aec, simulated gastric and intestinal fluids (SGF, SIF, respectively),
buffers pH 1.2 and 6.8, and human plasma.
Comp Stress conditions
NaOH HCl H2O2 SGF
Allop No degradation
FDA conditions
Human
plasma
pH 1.2
SIF
pH 6.8
2a
2b
2c
10.4a
9.3a
No degradation
3.7a
3a
3b
3c
Immediate
degradation
18.9b
17.1b
100c
100c
8.4e
9.5e
6.1e
0.0052d 0.0051d 0.0503d 0.0331d
32b
0.0074d 0.0071d 0.0380d 0.0186d
26.5b 27.0b 100c 100c
0.0062d 0.0063d 0.0455d 0.0299d
30.3b
100c
100c
These results show that Allop was stable for all matrices and
studied conditions, indicating that C6-derivatives acted as drugs
whereas the N1-ones behaved as prodrugs and yielded Allop in a
short time period under physiological conditions, but immediately
under stress conditions.
a
% degradation (16 h).
b
c
% degradation (37 ꢃC, 1 h).
% degradation (37 ꢃC, 3 h).
d
e
kobs minꢀ1
.
t
½
(min).

460
M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
2.3.3. Solubility
This equation may be regarded as a special case of the Collander
one, which relates the partition coefficients measured in two
different partitioning systems. In fact, the slope of the Collander
equation is a measure of the solvent system’s sensitivity to changes
in the hydrophobicity of the solutes relative to n-octanol, (i.e a
slope of 1.0 indicates isodiscriminative behavior). Eq. (1) has a slope
of 1.35, thus indicating a good system study [36]. The log k0w values
of compounds 2aec and 3aec were then used to determine their
log Po/w values from this standard curve with Table 3 showing these
obtained values and Fig. 6 demonstrating the correlation between
the standard curve and the analyzed derivatives.
It is important to point out that Bundgaard et al reported log Po/
w values in pure water (pH y 6.5) for N1-derivatives of ꢀ0.35; 0.30
and 0.85 (acetyl-, n-propyl- and n-butyl-, respectively) [20]. In this
medium, these compounds were partially ionized (pKa ¼ 7.6)
whereas in our study (buffer pH 5.5) they behaved as unionized
ones. For this reason, the log P values reported by Bundgaard
seemed to be more hydrophilic than those we obtained here
(log P ¼ 0.40, 1.22 and 1.99, for 3a, 3b and 3c, respectively).
Many protocols have been described in the literature for
measuring solubility, with classical approaches being based on the
saturation shake-flask method [29,30], which was used to deter-
mine the intrinsic solubility equilibrium constant, S0, of Allop and
2aec, in the buffer pH 6.8 at 25 ꢃC (of interest in pharmaceutical
operations) and at 37 ꢃC. These experimental solubility data are
shown in Table 1, with the Allop value being similar to that of
literature [31].
Taking into account that the 3aec derivatives can degrade over
the time of the study, we used the turbidity detection, which was
popularized by Lipinski et al. [21]. In all cases, the solubility in
buffer pH 6.8 at 25 ꢃC and 37 ꢃC was greater than 65 g mLꢀ1
m ,
suggesting that the solubility was not responsible for the poor
bioavailability of these compounds [21].
Taking into account that the melting point of solids is an indi-
cator of molecular cohesion, it can be used as a guide to the solu-
bility value in a closely related series of compounds. [32]. Therefore,
as we observed that 2aec and 3aec melted at lower temperatures
than Allop (Table 1) accompanied by decomposition-volatizing
processes, this may indicate a lower lattice energy that might also
be associated with the improved solubility of the Allop derivatives
[32].
2.3.5. Other relevant properties
For the majority of drugs, the preferred route of administration
is by oral ingestion, and therefore when considering the potential
pharmacological use of 2aec and 3aec, they must demonstrate an
adequate in vivo behavior. The “rule of five”, developed by Lipinski
et al. [21] [log P < 5, H-bond donor ꢄ 5 (expressed as the sum of OH
and NH groups), H-bond acceptor ꢄ 10 (expressed as the sum of N
and O atoms), and molecular weight < 500] is the most widely
accepted method used. In recent years, this rule has been expanded
to the study of other parameters, such as polar surface area (PSA)
and molecular rigidity as indicated by the number of rotatable
bonds (NRB) [37]. Considering that Lipinski thought that certain
properties were important for drug bioavailability and absorption,
if two of these parameters are out of range, a poor absorption or
permeability is possible.
The PSA, defined as the sum of surfaces of polar atoms in a
molecule, provides a good correlation with experimental transport
data. This has been successfully applied for the prediction of in-
testinal absorption, among other parameters [38], with the physical
explanation being that polar groups are involved in the desolvation
process when they move from an aqueous extracellular environ-
ment to the more lipophilic interior of the membranes. In addition,
it has been shown that PSA, calculated according to the method
proposed by Ertl et al. [39], in most cases produces a sigmoid
relationship with intestinal absorption, showing that poorly
2.3.4. Lipophilicity
Although the octanolewater distribution ratio, Po/w, is generally
accepted to be the best physicochemical property to measure the
hydrophobicity of chemicals, it has been demonstrated that the
capacity factor (k0) of a compound obtained by RP-HPLC is an in-
direct reliable descriptor of its lipophilicity. Moreover, the log k0
water ðlog k0wÞ is an extrapolated parameter from the binary phase
to 100% of water, and is an even better descriptor of lipophilicity
because it is independent of any organic modifier effects [33].
Here, we used the method proposed by the Organization for
Economic Cooperation and Development (OECD), who prepared
guidelines that should be followed in order to produce reliable Po/w
values using RP-HPLC [34]. According to OECD, the log Po/w value of
an analyte can be evaluated from a calibration graph constructed
from the log Po/w values of literature versus experimental log k0w of
reference compounds. In this way, seven compounds (Allop [31],
didanosine [35], aniline [34], phenol [34], benzoic acid [34], toluene
[34] and naphthalene [34]) were used as standard compounds, and
the following good correlation was found (eq. (1)) between log kw0
and the log Po/w values of the reference compounds.
Log Po=w ¼ 1:348 ꢁ ð0:106Þlog k0w ꢀ 0:011 ꢁ ð0:028Þ
(1)
n ¼ 7; r2 ¼ 0:9953; SD ¼ 0:280
Table 3
Molecular properties of Allop and its derivatives.
Comp
Lipinski rules (LR)
Lipophilicity H-
PSAa
NRBb
H-bond
MW
(Log Pow
)
bond
acceptors
donor
Allop
2a
2b
2c
3a
ꢀ0.55
ꢀ0.05
0.52
1.14
0.40
1.22
1.99
<5
2
2
2
2
1
1
1
ꢄ5
3
3
3
3
5
5
5
ꢄ10
136.1
150.1
164.2
178.2
178.2
192.2
206.2
<500
74.4
74.4
74.4
74.4
80.6
80.6
80.6
ꢄ140
0
0
1
2
0
1
2
ꢄ10
3b
3c
OPc
a
Polar surface area (PSA).
Number of rotatable bonds (NRB).
Optimal properties for bioavailability and oral absorption (OP).
b
c
Fig. 6. Relationship between log k0w and log Po/w values for standards and Allop
derivatives.

M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
461
2
ꢀ
absorbed compounds are identified as those with PSA ꢅ 140 A
were of commercial quality (>97%) from freshly opened containers,
and solvents were of analytical grade.
[38].
The last parameter studied was the molecular rigidity, which
suggests that if NRB is >10 a poor oral bioavailability could occur
[37]. Summing up, when the developed derivatives 2aec and 3aec
were subjected to Lipinski rules, PSA and NRB analyses, it can be
observed in Table 3, that all of them exhibited acceptable
characteristics.
4.2. Equipment
High Resolution Mass Spectrometry (HRMS) was performed in a
Bruker micrOTOF-Q II, using the electrospray ionization (ESI) pos-
itive mode and atmospheric pressure photoionization (APPI). Nu-
clear Magnetic Resonance (NMR) experiments were performed on
Bruker Avance II 400, Ultrashield, Frequency 1H NMR 400.16 MHz
3. Conclusions
13
and C NMR 100.62 Hz, Dual BBI Probe, at 25 ꢃC, using DMSO-d6
(99.8%, Merck) as solvent. The assignment of all exchangeable
protons (OH, NH) was confirmed by the addition of D2O (99.9%,
Sigma), and chemical shift values are reported in parts per million
Different strategies were developed for the synthesis of six Allop
derivatives (2aec and 3aec), which were unequivocally charac-
terized by different spectroscopic methods. Regarding these ana-
(d) relative to tetramethylsilane (TMS). The splitting pattern ab-
lyses, the results showed that all derivatives had
a better
breviations are as follows: s, singlet; d, doublet; t, triplet; qua,
quartet; quin, quintet. Ultraviolet spectrophotometric (UV) studies
were carried out using a Shimadzu Model UV-160A spectropho-
tometer, with 1 cm quartz cells. Melting point (mp) and thermal
analysis were carried out using differential scanning calorimetry
(DSC) 2920 modulated DSC (TA Instrument); the temperature axis
and the cell constant of the DSC were calibrated with indium, and
thermogravimetric analysis (TG) measurements were performed
with a TA Instrument Hi-res TG 2950 (TA Instrument). Weighed
samples (1.5e3 mg, C-33 Microbalance, Cahan) were scanned in
covered aluminum (Al) pans under a dynamic dry nitrogen atmo-
sphere (50 mL minꢀ1). The pH values were measured with a CRI-
SON GLP-21 pH meter equipped with an Ag/AgCl glass electrode.
physicochemical profile than Allop, as well as that all met Lipinski’s
rules and had suitable characteristics of PSA and NRB.
When the inhibitory effects of the C6-alkyl- and N1-acyl de-
rivatives of Allop against T. cruzi were studied, none of the com-
pounds showed any significant trypanocidal activity in vitro against
TRYP. In contrast, the AMAS resulted susceptible to the 2a, 2b and
3c derivatives of Allop when tested at 1 m
g mLꢀ1. As this hypo-
xantine analog is used by the T. cruzi’s HGPRT as an alternative
substrate during the duplicative process [10], it is certain that the
AMAS forms were involved. On testing different concentrations of
the derivatives, only the 3c compound displayed an inhibitory ef-
fect, probably by being incorporated into the parasite ribonucleic
acid as no physiological nucleotides or blocking of the synthesis of
new purine nucleotide occurred. The 3c derivative exhibited an
IC50 value similar to Allop, which was not toxic for mammalian
cells. The development and maintenance of chronic lesions are due
to persistent intracellular parasites in targeted tissues, which the
immune response is unable to eradicate.
Hence, the possible role of 3c in Chagas disease may become a
practical tool in the improvement of the prognosis of these patients.
It is important to point out that modifications on C-6 (2aec) led to
chemical entities that could not be transformed into Allop in bio-
logical media such as plasma and SIF. Although these compounds
did not have parasiticidal activity per se, they presented better
physicochemical properties than the prototype (Allop).
4.3. Chemistry
4.3.1. Characterization of C6-alkyl series (2aec)
4.3.1.1. 6-Methyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(2a). Light yellow solid (82% yield), Rf ¼ 0.45 (CH2Cl2/EtOH 10:1);
mp: 304 ꢃC; HPLC tr ¼ 3.7 min, MLC tr ¼ 3.0 min. 1H NMR (DMSO-
d6):
d 13.55 (s, 1H, NH-1), 11.92 (s, 1H, NH-5), 8.00 (s, 1H, H-3), 2.33
(s, 3H, H-10). 13C NMR (DMSO-d6):
d
159.3 (CO, C-4), 157.5 (C, C-6),
155.8 (C, C-7a), 134.2 (CH, C-3), 104.1 (C, C-3a), 21.9 (CH3, C-10).
HRMS (APPI) calcd for C6H7N4O (M þ H)þ m/z: 151.0614; found:
151.0622; UV lmax/nm (MeOH, pH 5) 250.
The 3aec series were promising candidates, especially com-
pound 3c, since it demonstrated not only excellent activity against
T. cruzi, but also no cytotoxicity and gave Allop in an adequate time.
These features led to a better lipophilicity and solubility than Allop,
as well as to a lower crystalline lattice, with the possibility of better
pharmacokinetic profiles of the regenerated active compound.
4.3.1.2. 6-Ethyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(2b). Light yellow solid (83% yield): Rf ¼ 0.64 (CH2Cl2/EtOH 10:1);
mp: 325 ꢃC; HPLC tr ¼ 5.3 min, MLC tr ¼ 4.7 min. 1H NMR (DMSO-
d6): d 13.58 (s, 1H, NH-1), 11.92 (s, 1H, NH-5), 7.98 (s, 1H, H-3), 2.63
(qua, 2H, H-10), 1.22 (t, 3H, H-20). 13C NMR (DMSO-d6):
d 162.1 (C, C-
6), 158.8 (CO, C-4), 154.7 (C, C-7a), 135.4 (CH, C-3), 104.1 (C, C-3a),
28.0 (CH2, C-10), 11.7 (CH3, C-20). HRMS (APPI) calcd for C7H9N4O
(M þ H)þ m/z: 165.0771; found: 165.0778. UV lmax/nm (MeOH, pH
5) 250.
4. Experimental section
4.1. Chemicals and reagents
4.3.1.3. 6-Propyl-1,5-di hydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(2c). Light yellow solid (81% yield): Rf ¼ 0.72 (CH2Cl2/EtOH 10:1);
mp: 292 ꢃC; HPLC tr ¼ 8.8 min, MLC tr ¼ 7.8 min. 1H NMR (DMSO-
Allopurinol, propionic anhydride, butyric anhydride, acetyl
chloride, propionyl chrolide and butyryl chloride were purchased
from Sigma, and acetic anhydride was purchased from Merck. The
buffer materials of pH 1.2 (hydrochloric acid/sodium chloride), pH
5.5 (acetic acid/sodium acetate) and pH 6.8 (sodium dihydrogen
phosphate/phosphoric acid) were of analytical reagent grade.
The SGF and SIF were prepared according to USP specifications
[40]. Sodium dodecyl sulfate (SDS) was purchased from Biopack, 1-
propanol (analytical grade) was purchased from Cicarelli and
methanol (MeOH, HPLC grade) was acquired from Sintorgan. The
water used in all studies was of Milli-Q grade (MilliporeÒ), and
solutions and mobile phases were filtered through Millipore filters
d6):
d 13.55 (s, 1H, NH-1), 11.98 (s, 1H, NH-5), 7.99 (s, 1H, H-3), 2.57
(t, 2H, H-10), 1.71 (sex, 2H, H-20), 0.91 (t, 3H, H-30). 13C NMR (DMSO-
d6): d 160.5 (C, C-6), 159.3 (CO, C-4), 155.6 (C, C-7a), 134.2 (CH, C-3),
104.2 (C, C-3a), 36.5 (CH2, C-10), 20.7 (CH2, C-20), 14.0 (CH3, C-30).
HRMS (APPI) calcd for C8H11N4O (M þ H)þ m/z: 179.0927; found:
179.0938. UV lmax/nm (MeOH, pH 5) 250.
4.3.2. Characterization of N1-acyl series (3aec)
4.3.2.1. 1-Acetyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(3a). White solid, recrystallization from DMF/EtOH (60% yield);
Rf ¼ 0.73 (CH2Cl2/EtOH 8:2); mp: 267 ꢃC; HPLC tr ¼ 3.7 min, MLC
Type FH (4.5 mm) (Millipore) and degassed under vacuum. Reagents

462
M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
tr ¼ 3.0 min. 1H NMR (DMSO-d6):
d
12.74 (s, 1H, NH-5), 8.36 (s, 1H,
microscope. The percentage of host cells with more than one
H-3), 8.28 (d, 1H, H-6), 2.73 (s, 3H, H-20). 13C NMR (DMSO-d6):
amastigote (AMAS) in their cytoplasm was calculated by analyzing
a total of 400 host cells distributed in randomly chosen microscopic
fields. Data expressed as the percentage of inhibition of each
compound were compared with non-treated cells. Only those
drugs that displayed a similar or better behavior than Bz were used
for the IC50 calculation. In this case, T. cruzi infected monolayers
d
168.9 (CO, C-10), 157.4 (CO, C-4), 154.9 (C, C-7a), 150.7 (C, C-6),
138.5 (CH, C-3), 109.3 (C, C-3a), 24.6 (CH3, C-20). HRMS (ESI) calcd
for C7H6N4NaO2 (M þ Na)þ m/z: 201.0383; found: 201.0394. UV
lmax/nm (MeOH, pH 5) 276.
4.3.2.2. 1-Propanoyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-
one (3b). White solid, recrystallization from DMF/EtOH (62% yield),
Rf ¼ 0.77 (CH2Cl2/EtOH 8:2); mp: 275 ꢃC; HPLC tr ¼ 4.9 min, MLC
were subjected to 0.5e50 mM of selected compounds, and the rate
of infection as well as the mean number of AMAS per infected cell
was determined. The 50% inhibitory concentrations (IC50s) were
estimated by nonlinear regression analysis using Graph Pad Prism
5.03 software (La Jolla, USA).
tr ¼ 4.0 min. 1H NMR (DMSO-d6):
d 12.72 (s, 1H, NH-5), 8.35 (s, 1H,
H-3), 8.28 (d, 1H, H-6), 3.21 (qua, 2H, H-20), 1.17 (t, 3H, H-30). 13C
NMR (DMSO-d6): d
172.5 (CO, C-10), 157.3 (CO, C-4), 154.9 (C, C-7a),
150.7 (C, C-6), 138.4 (CH, C-3), 109.2 (C, C-3a), 29.8 (CH2, C-20), 8.8
(CH3, C-30). HRMS (ESI) calcd for C8H8N4NaO2 (M þ Na)þ m/z:
215.0539; found: 215.0655. UV lmax/nm (MeOH, pH 5) 276.
4.4.4. Mammalian cell toxicity assays
Only those compounds that attained an IC50 index in
mammalian cells 50 times greater than the IC50 of the parasite
were considered for further in vivo evaluation. Cultures were car-
ried out on 96-well plates with 2 ꢂ 106 murine splenocytes mLꢀ1
for 32 h in the presence of compound 3c, Bz and Allop at a con-
centration of 50 ꢂ IC50. Cell viability was determined by the Ala-
mar BlueÔ method according to instructions of the manufacturer.
Data were expressed as the percentage of viability related to control
(non-treated cells). An analysis of variance (ANOVA) with the
Bonferroni post-test was used to carry out the statistical analysis. A
p value < 0.05 was considered significant.
4.3.2.3. 1-Butanoyl-1,5-dihydro-4H-pyrazolo[3,4-d]pyrimidin-4-one
(3c). White solid, recrystallization from EtOH (65% yield), Rf ¼ 0.82
(CH2Cl2/EtOH 8:2); mp: 233 ꢃC; HPLC tr ¼ 7.0 min, MLC tr ¼ 7.7 min.
1H NMR (DMSO-d6):
d 12.71 (s, 1H, NH-5), 8.35 (s, 1H, H-3), 8.28 (s,
1H, H-6), 3.18 (t, 2H, H-20), 1.71 (sex, 2H, H-30), 0.97 (t, 3H, H-40). 13C
NMR (DMSO-d6):
171.6 (CO, C-10), 157.6 (CO, C-4), 155.1 (C, C-7a),
d
150.8 (C, C-6), 138.7 (CH, C-3), 109.5 (C, C-3a), 38.2 (CH2, C-20), 17.8
(CH2, C-30), 14.1 (CH3, C-40). HRMS (ESI) calcd for C9H10N4NaO2
(M þ Na)þ m/z: 229.0696; found: 229.0711. UV lmax/nm (MeOH, pH
5) 276.
4.5. Chromatographic conditions
4.4. Activity
Analytical thin layer chromatography (TLC) was performed on
percolated plates purchased from Merck (Silica Gel 60 F254). Sol-
vent system: 2aec: CH2Cl2/EtOH 10:1; 3aec: CH2Cl2/EtOH 8:2.
HPLC and MLC analyses were carried out using an AgilentÒ 1100
4.4.1. Parasites
Bloodstream trypomastigotes (TRYP) of RA strain of T. cruzi [23]
were obtained from T. cruzi infected mice at the peak of para-
sitemia. To enrich blood supernatants with TRYP, PBS-diluted blood
was centrifuged, incubated for 1 h a 37 ꢃC and the supernatant was
collected. Thus, parasites were pelleted by centrifugation for 30 min
at 10,000 ꢂ g, counted in a Neubauer hematocytometer and diluted
to 1.5 ꢂ 106 TRYP mLꢀ1 in RMPI 1640 medium supplemented with
10% fetal calf serum (FCS) for drug sensitivity assays.
apparatus equipped with a PhenomenexÒ column, Hypersil ODS 5
m
particle diameter (4.6 ꢂ 250 mm); HPLC conditions: 2aec: 80%
0.01 M NaC2H3O2/CH3COOH (pH 4.5), 20% MeOH, flow rate of
1.5 mL minꢀ1, 35 ꢃC,
l
¼ 250 nm; 3aec: 50% buffer pH 4.5
(NaC2H3O2/CH3COOH), 50% MeOH, flow rate of 1 mL minꢀ1, 35 ꢃC,
¼ 276 nm; MLC conditions: 2aec: SDS (0.05 M) e NaH2PO4/H3PO4
(0.01 M, pH 3.0), flow rate of 1.25 mL minꢀ1, 25 ꢃC,
¼ 250 nm; 3ae
l
l
c: SDS (0.1 M) e 7.5% (v/v) 1-propanol e NaH2PO4/H3PO4 (0.01 M,
4.4.2. In vitro drug activity against bloodstream trypomastigotes
Purified TRYP were incubated under an atmosphere of 5% CO2 at
37 ꢃC for 24 h with increasing doses of each synthetic compound
pH 3.0), flow rate of 1 mL minꢀ1, 25 ꢃC,
l
¼ 276 nm.
4.6. Stability studies
(0e70 m
g mLꢀ1) obtained from stock solutions of 40 mM dissolved
in DMSO. Benznidazole was considered as the reference drug. The
final concentration of DMSO in the culture medium remained
below 0.5%. A solution of 0.5% DMSO was used as negative control.
The following day, the remaining live parasites were counted using
a Neubauer chamber as previously described [41]. Three indepen-
dent experiments were performed in triplicate.
4.6.1. Hydrolytic and oxidative studies
Compounds 2aec and 3aec at a concentration of 1 mg mLꢀ1
were used in all stability assays. Acid and alkaline decomposition
studies were performed by heating the solution of each compound
in 0.1 M HCl or 0.1 M NaOH at 90 ꢃC for 16 h.
4.6.2. Buffers (pH 1.2 and pH 6.8), SGF, SIF and human plasma
4.4.3. In vitro drug activity against intracellular amastigotes
Irradiated Vero cells (2000 rads) seeded on round coverslips
were infected with TRYP. After 24 h of parasite-host cell interaction
(5:1 parasite:cell ratio), the infected cultures were washed to
remove free parasites and incubated for another 72 h with
The stock solutions (1.0 ꢂ 10ꢀ4 mol mLꢀ1) of each compound
were prepared in DMSO prior to use. Then, 100
mL of the stock
solution were added to a vial containing 1900 mL of buffer or matrix
(gastric or intestinal fluid) to obtain the work solutions. Afterward,
the vials containing the samples were placed in a water bath at
37 ꢃC throughout the experiment. In human plasma, the experi-
ments were performed under the experimental conditions sug-
gested by Di et al. [43]. Human plasma (unit Nꢃ 65261) was
generously supplied by Instituto de Hematología, Hemoterapia y
Banco de Sangre, Universidad Nacional de Córdoba, Argentina.
1 m
g mLꢀ1 of each compound, as previously described [22]. Cell
cultures were maintained at 37 ꢃC in 5% CO2. Uninfected treated
cultures exposed to vehicle (0.5% DMSO) were used as controls. The
method for determining the rate of infection of host cells by T. cruzi
has previously been described in detail [42]. Briefly, host cells,
attached to the coverslip seeded in a 24-well plate, were gently
rinsed with PBS before being air dried, fixed in absolute methanol
and stained with Giemsa. The coverslip was then transferred to a
glass slide and the mounted cells were observed under a light
4.6.3. Chromatographic analysis
At an appropriate time, aliquots of 100
mL were removed and
then centrifugated. Then, 50 L of the resulting supernatant was
m

M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
463
added to a solution of 450
mL of the corresponding mobile phase.
Appendix A. Supplementary material
Each sample was immediately stored at ꢀ18 ꢃC until use. Upon
removal of the last samples, the stored solutions were allowed to
warm up to room temperature, and then the concentrations of
2aec and 3aec from the work solutions were monitored by MLC
to determine the rate at which the parent compounds dis-
appeared, with the formation of Allop being the only degradation
product.
Supplementary material associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.ejmech.
2013.08.045.
References
[1] WHO, Chagas Disease (American Trypanosomiasis) Fact Sheet 340, August
2012. http://www.who.int/mediacentre/factsheets/fs340/en/index.html.
[2] J. Gascon, C. Bern, M.J. Pinazo, Chagas disease in Spain, the United States and
other non-endemic countries, Acta Trop. 115 (2010) 22e27.
[3] C. Beyrer, J.C. Villar, V. Suwanvanichkij, S. Singh, S.D. Baral, E.J. Mills, Neglected
diseases, civil conflicts, and the right to health, Lancet 370 (2007) 619e627.
[4] J.A. Urbina, Specific chemotherapy of chagas disease: relevance, current lim-
itations and new approaches, Acta Trop. 115 (2010) 55e68.
[5] R. Viotti, C. Vigliano, B. Lococo, M.G. Alvarez, M. Petti, G. Bertocchi, A. Armenti,
Side effects of benznidazole as treatment in chronic chagas disease: fears and
realities, Expert Rev. Anti Infect. Ther. 7 (2009) 157e163.
[6] J.D. Maya, B.K. Cassels, P. Iturriaga-Vásquez, J. Ferreira, M. Faúndez, N. Galanti,
A. Ferreira, A. Morello, Mode of action of natural and synthetic drugs against
Trypanosoma cruzi and their interaction with the mammalian host, Comp.
Biochem. Physiol. A 146 (2007) 601e620.
4.7. Solubility
The solid compounds Allop and 2aec were added to a standard
phosphate buffer solution (pH 7.0, 80 mM) until saturation
occurred as indicated by undissolved excess drug. The thermo-
stated saturated solution (25 ꢃC) was shaken for 24 h to produce an
equilibrium between the two phases. At different times, the ali-
quots were removed and then centrifugated. The resulting super-
natant was analyzed by MLC to determine the integrity and
concentration of each compound.
For 3aec, the stock solutions (10
were prepared in DMSO, and 1 L of the corresponding solutions
mg mL
ꢀ1) of each compound
[7] F. Sánchez-Sancho, N.E. Campillo, J.A. Páez, Chagas disease: progress and new
perspectives, Curr. Med. Chem. 17 (2010) 423e452.
m
was added into a cuvette containing 2.5 mL pH 7.0 phosphate
buffer, with these steps being repeated up to 14 times. After each
step, an equilibrium time of 2 min was allowed before the
turbidity was analyzed. The precipitated particles led to an in-
crease in UV absorbance (600e800 nm) due to light scattering.
The system was kept under stirring at a controlled temperature of
25 ꢃC.
[8] W.E. Gutteridge, M.J. Davies, Enzymes of purine salvage in Trypanosoma cruzi,
FEBS Lett. 127 (1981) 211e214.
[9] W.N. Kelley, J.B. Wyngaarden, Effects of allopurinol and oxipurinol on purine
synthesis in cultured human cells, J. Clin. Invest. 49 (1970) 602e609.
[10] J.J. Marr, R.L. Berens, D.J. Nelson, Antitrypanosomal effect of allopurinol:
conversion in vivo to aminopyrazolopyrimidine nucleotides by Trypanosoma
cruzi, Science 201 (1978) 1018e1020.
[11] R.L. Berens, J.J. Marr, F. Steele Da Cruz, D.J. Nelson, Effect of allopurinol on
Trypanosoma cruzi: metabolism and biological activity in intracellular and
bloodstream forms, Antimicrob. Agents Chemother. 22 (1982) 657e661.
[12] R.H. Gallerano, J.J. Marr, R.R. Sosa, Therapeutic efficacy of allopurinol in pa-
tients with chronic chagas disease, Am. J. Trop. Med. Hyg. 43 (1990) 159e166.
[13] P. Gobbi, M.S. Lo Presti, A.R. Fernandez, J.E. Enders, R. Fretes, S. Gea,
P.A. Paglini-Oliva, H.W. Rivarola, Allopurinol is effective to modify the evo-
lution of Trypanosoma cruzi infection in mice, Parasitol. Res. 101 (2007) 1459e
1462.
4.8. Determination of lipophilicity by RP-HPLC ðlog ko0 w
Þ
The instrument conditions for all the analyses were as follows:
25 ꢃC, flow rate: 1 mL minꢀ1, injection volume: 20
L. Each analysis
m
[14] A. Rassi, A. Ostermayer Luquetti, A. Rassi Jr., G.G. Rassi, S.G. Rassi, I. Garcia Da
Silva, A.G. Rassi, Short report: specific treatment for Trypanosoma cruzi: lack of
efficacy of allopurinol in the human chronic phase of chagas disease, Am. J.
Trop. Med. Hyg. 76 (2007) 58e61.
[15] D.E. Perez-Mazliah, M.G. Alvarez, G. Cooley, B.E. Lococo, G. Bertocchi, M. Petti,
M.C. Albareda, A.H. Armenti, R.L. Tarleton, S.A. Laucella, R. Viotti, Sequential
combined treatment with allopurinol and benznidazole in the chronic phase
of Trypanosoma cruzi infection: a pilot study, J. Antimicrob. Chemother. 68
(2013) 424e437.
[16] R.B. Bestetti, T.A.D. Theodoropoulos, A systematic review of studies on heart
transplantation for patients with end-stage chagas’ heart disease, J. Card. Fail.
15 (2009) 249e255.
[17] N.L. Grosso, J. Bua, A.E. Perrone, M.N. Gonzalez, P.L. Bustos, M. Postan,
L.E. Fichera, Trypanosoma cruzi: biological characterization of a isolate from an
endemic area and its susceptibility to conventional drugs, Exp. Parasitol. 126
(2010) 239e244.
was performed isocratically using buffer pH 5.5eMeOH with a
methanol content of between 30% and 75% (v/v) at 5% increments.
Three injections were performed for each compound and five in-
jections for the standards at each MeOHebuffer ratio. The loga-
rithm of the capacity factor ðlog k0Þ for 2aec, 3aec and the standard
compounds was calculated for each MeOHebuffer ratio from the
retention times according to log k0 ¼ log(tR ꢀ t0)/t0; where tR is the
sample retention time, and t0 is the mobile phase hold time as
estimated by the retention time of MeOH used to dissolve each
sample. For each standard and 2aec and 3aec compounds, a plot of
log k0 versus percentage of organic modifier (i.e., MeOH) was
generated and extrapolated back to the 0% modifier in order to
determine the capacity factor of each compound at 100% of buffer
ðlog k0wÞ.
[18] C.C. Cheng, R.K. Robins, Potential purine antagonist VII. Synthesis of 6-
alkylpyrazolo-[3,4-d]pyrimidines, J. Org. Chem. 23 (1958) 191e200.
[19] G. Biagi, A. Costantini, L. Costantino, I. Giorgi, O. Livi, P. Pecorari, M. Rinaldi,
V. Scartoni, Synthesis and biological evaluation of new imidazole, pyrimidine,
and purine derivatives and analogs as inhibitors of xanthine oxidase, J. Med.
Chem. 39 (1996) 2529e2535.
Acknowledgments
[20] H. Bundgaard, E. Falch, Allopurinol prodrugs I. Synthesis, stability and phys-
icochemical properties of various N1-acyl allopurinol derivatives, Int. J.
Pharm. 23 (1985) 223e237.
This work was supported by grants from Secretaría de Ciencia
y Técnica 162/2012, Ministerio de Ciencia y Tecnología, Gobierno
de la Provincia de Córdoba 121/2008, Agencia Nacional de Pro-
moción Científica y Tecnológica (ANPCyT) PICT 2006-1325 and
2010-0866, Consejo Nacional de Investigaciones Científicas y
Técnicas (CONICET) PIP 0769 and Universidad de Buenos Aires
UBACYT K066. Both M.M.N. and M.S.G. acknowledge receipt of a
fellowship granted by Fondo Nacional de Promoción Científica y
Tecnológica (FONCyT) and CONICET, respectively. M.A.R. and
C.D.A.S. are members of the research career of CONICET. We thank
Dr. Gloria Bonetto for her assistance in NMR data recording, Lic.
MA Pino-Martinez and Lic. C. Miranda for the parasitemia de-
terminations, and Dr. Paul Hobson, native speaker, for revision of
the manuscript.
[21] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and
computational approaches to estimate solubility and permeability in drug
discovery and development settings, Adv. Drug Deliv. Rev. 46 (2001) 3e26.
[22] A.J. Romanha, S.L. Castro, M.N. Soeiro, J. Lannes-Vieira, I. Ribeiro, A. Talvani,
B. Bourdin, B. Blum, B. Olivieri, C. Zani, C. Spadafora, E. Chiari, E. Chatelain,
G. Chaves, J.E. Calzada, J.M. Bustamante, L.H. Freitas-Junior, L.I. Romero,
M.T. Bahia, M. Lotrowska, M. Soares, S.G. Andrade, T. Armstrong, W. Degrave,
Z.A. Andrade, In vitro and in vivo experimental models for drug screening and
development for chagas disease, Mem. Inst. Oswaldo Cruz 105 (2010) 233e
238.
[23] S.M. González-Cappa, A.T. Bijovsky, H. Freilij, L.A. Muller, A.M. Katzin, Isolation
of a Trypanosoma cruzi strain of predominantly slender form in Argentina
(Aislamiento de una cepa de Trypanosoma cruzi a predominio de formas
delgadas en la Argentina), Medicina 41 (1981) 119e120.
[24] G. Steele, T. Austin, Preformulation investigations using small amounts of
compound as an aid to candidate drug selection and early development, in:

464
M.A. Raviolo et al. / European Journal of Medicinal Chemistry 69 (2013) 455e464
M. Gibson (Ed.), Pharmaceutical Preformulation and Formulation, second ed.,
Informa Healthcare USA, Inc., New York, 2009, pp. 17e128.
[25] M. Meloun, S. Bordovská, Benchmarking and validating algorithms that esti-
mate pK(a) values of drugs based on their molecular structures, Anal. Bioanal.
Chem. 389 (2007) 1267e1281.
[26] ICH, Stability testing of new drug substances and products Q1A (R2), in:
Proceedings of the International Conference on Harmonisation, IFPMA,
Geneva, 2003.
[27] Guidance for Industry, Waiver of In Vivo Bioavailability and Bioequivalence
Studies for Immediate Release Solid Oral Dosage Forms Based on a Bio-
pharmaceutics Classification System, U.S. Department of Health and Human
Services, Food and Drug Administration, Rockville, MD, USA, 2000.
[28] Guidance for Industry, Bioanalytical, Method Validation, U.S. Department of
Health and Human Services, Food and Drug Administration, Rockville, MD,
USA, 2001.
[35] M.N. Nassar, T. Chen, M.J. Reff, S.N. Agharkar, Didanosine, in: K. Florey (Ed.),
Analytical Profiles of Drug Substances and Excipients V22, Acad. Press, New
York, 1993, pp. 185e227.
[36] T. Braumann, Determination of hydrophobic parameters by reversed-phase
liquid chromatography: theory, experimental techniques, and application in
studies on quantitative structureeactivity relationships, J. Chromatogr. 373
(1986) 191e225.
[37] J. Selbo, P.C. Chiang, Absorption and physicochemical properties of the NCE,
in: K. Tsaioun, S.A. Kates (Eds.), ADMET for Medicinal Chemists: a Practical
Guide, John Wiley & Sons, Inc., 2011, pp. pp125e144.
[38] H. van de Waterbeemd, Physico-chemical approaches to drug absorption, in:
H. van de Waterbeemd, H. Lennernäs, P. Artursson (Eds.), Drug Bioavailability:
Estimation of Solubility, Permeability, Absorption and Bioavailability, Wiley-
VCH, Weinheim, 2003, pp. 3e20.
[39] P. Ertl, B. Rohde, P. Selzer, Fast calculation of molecular polar surface area as a
sum of fragment-based contributions and its application to the prediction of
drug transport properties, J. Med. Chem. 43 (2000) 3714e3717.
[40] Test solutions, United States Pharmacopoeia 32/National Formulary 27, 2009.
[41] M.C. Fernández, S.M. González Cappa, M.E. Solana, Trypanosoma cruzi:
immunological predictors of benznidazole efficacy during experimental
infection, Exp. Parasitol. 124 (2010) 172e180.
[42] J. Nakajima-Shimada, Y. Hirota, T. Aoki, Inhibition of Trypanosoma cruzi
growth in mammalian cells by purine and pyrimidine analogs, Antimicrob.
Agents Chemother. 40 (1996) 2455e2458.
[29] A. Avdeef, Physicochemical profiling (solubility, permeability and charge
state), Curr. Top. Med. Chem. 1 (2001) 277e351.
[30] H. Tietgen, Physicochemical properties, in: H.G. Vogel, F.J. Hock, J. Maas,
D. Mayer (Eds.), Drug Discovery and Evaluation. Safety and Pharmacokinetic
Assay, Springer-Verlag, Berlin, Heidelberg, New York, 2006, pp. 399e408.
[31] S.A. Benezra, T.R. Bennett, Allopurinol, in: K. Florey (Ed.), Analytical Profiles of
Drug Substances V7, Acad. Press, New York, 1978, pp. 1e16.
[32] A.T. Florence, D. Attwood, Physicochemical Principles of Pharmacy, fourth ed.,
Pharmaceutical Press, London, 2006, pp. 139e176.
[33] A. Berthod, S. Carda-Broch, Determination of liquideliquid partition co-
efficients by separation methods, J. Chromatogr. A 1037 (2004) 3e14.
[34] Partition Coefficient (n-Octanol/Water), High Performance Liquid Chroma-
tography (HPLC) Method, OECD Guideline for Testing of Chemicals, Test Nꢃ
117, 2004.
[43] L. Di, E.H. Kerns, Y. Hong, H. Chen, Development and application of high
throughput plasma stability assay for drug discovery, Int. J. Pharm. 297 (2005)
110e119.