Structure-activity relationship and studies on the molecular mechanism of leishmanicidal N,V,C-coupled arylisoquinolinium salts

Article abstract of DOI:10.1021/jm801084u

Alternative drugs against leishmaniasis are desperately needed. Antimonials, the main chemotherapeutic tool, cause serious side effects and promote chemoresistance. We previously demonstrated that representatives of N,V,C-linked arylisoquinolines are prom

Full text of DOI:10.1021/jm801084u

626
J. Med. Chem. 2009, 52, 626–636
Structure-Activity Relationship and Studies on the Molecular Mechanism of Leishmanicidal
N,C-Coupled Arylisoquinolinium Salts
Alicia Ponte-Sucre,†,| Tanja Gulder,‡ Annemarie Wegehaupt,† Christoph Albert,§ Carina Rikanovic´,§ Leonhard Schaeflein,§
Andreas Frank,§ Martina Schultheis,† Matthias Unger,§ Ulrike Holzgrabe,§ Gerhard Bringmann,*,‡ and Heidrun Moll*,†
Institute of Molecular Infection Biology, UniVersity of Wu¨rzburg, Ro¨ntgenring 11, 97070 Wu¨rzburg, Germany, Institute of Organic Chemistry,
UniVersity of Wu¨rzburg, Am Hubland, 97074 Wu¨rzburg, Germany, Institute of Pharmacy and Food Chemistry, UniVersity of Wu¨rzburg, Am
Hubland, 97074 Wu¨rzburg, Germany, Laboratory of Molecular Physiology, UniVersidad Central de Venezuela, Caracas, Venezuela
ReceiVed May 14, 2008
Alternative drugs against leishmaniasis are desperately needed. Antimonials, the main chemotherapeutic
tool, cause serious side effects and promote chemoresistance. We previously demonstrated that representatives
of N,C-linked arylisoquinolines are promising leishmanicidal drug candidates. We now performed
structure-activity relationship studies varying the aryl portion of our lead substrate. The new series of
compounds show an enhanced selectivity against Leishmania major in comparison to their major host cell,
the macrophage. Our results suggest that the arylisoquinolinium salts decrease the macrophage infection
rate acting directly on the intracellular parasites. However, the activity of the 4′-i-propyl derivative might
also involve the modulation of cytokine and nitric oxide production by host macrophages. Additionally,
this isoquinoline acts synergistically with amphotericin B and does not interact with drug-metabolizing
cytochrome P450 enzymes involved in the metabolism of antileishmanial drugs. The results demonstrate
that the newly synthesized structurally simplified N,C-coupled arylisoquinolinium salts are promising
candidates to be considered as leishmanicidal pharmacophores.
Introduction
portion is connected with the isoquinoline moiety via a
nitrogen-carbon bond causing a positive charge at the nitrogen
atom in the isoquinoline portion, whereas usually it is linked
via a carbon-carbon biaryl axis.
Chemotherapy against leishmaniasis is unsatisfactory because
it is mainly based on antimony agents like sodium stibogluconate
and meglumine antimoniate, amphotericin B, miltefosine, and
paromomycin (for the chemical structures see Supporting
Information). The mode of action of these compounds is poorly
understood, and their toxicity causes serious side effects that
often result in patients deserting the treatment. According to
the World Health Organization (WHO), infections caused by
Leishmania parasites belong to the most hazardous infectious
diseases, the main reasons being the increasing number of cases,
where human immunodeficiency virus and Leishmania simul-
taneously infect the patient, and the worldwide escalating
frequency of chemoresistance to the standard therapy with
antimonial drugs.1-3 Thus, affordable alternative drugs against
leishmaniasis are desperately needed and the search for them
is a challenging task.
Naphthylisoquinoline alkaloids constitute an intriguing class
of natural products. They are axially chiral secondary metabo-
lites isolated from lianas of the small palaeotropic plant families
Ancistrocladaceae and Dioncophyllaceae.4 These compounds
exhibit a very good activity against pathogens causative of
several tropical infectious diseases, e.g., Plasmodium falciparum
and P. berghei,5-7 Trypanosoma brucei,8 and Leishmania
donoVani.9-11 The study of these pharmacologically interesting
natural products reached an essential stage with the discovery
of the N,C-linked arylisoquinolinium salts, a structurally new
alkaloid subtype.12,13 In these compounds, the naphthalene
In a recent study,14 we compared the efficacy of nine C,C-
coupled naphthylisoquinoline alkaloids to impair the growth of
L. major promastigotes and their host cells with that of five
N,C-linked substances, three of which were structurally simpli-
fied analogues of the natural naphthylisoquinolinium salts. We
also evaluated the effectiveness of selected substances like the
natural products ancistrocladinium A (1) and B (2) and the
synthetically prepared naphthylisoquinolinium salt 3 (Figure 1)
in decreasing the infection rate of peritoneal macrophages
containing L. major parasites. The results demonstrated that
representatives of this new coupling type together with its
synthetic derivatives are promising candidates to be considered
as lead compounds for leishmanicidal drugs. They are effective
against intracellular amastigotes at concentrations in the low
submicromolar range, while toxicity against various mammalian
cells is observed only at higher concentrations. Additionally,
we demonstrated that the leishmanicidal effect of the alkaloids
1 and 2 as well as that of the structurally simplified analogue 3
was not associated with the stimulation of cytokine secretion
or the production of nitric oxide by infected macrophages,
suggesting that the leishmanicidal activity of these compounds
is mainly targeted toward the parasite.14 Encouraged by these
promising results, we performed studies on the structure-activity
relationship (SAR) focusing on the aryl portion of our lead
structure 3.
In this report, we present the synthesis and an SAR study on
structurally simplified analogues of the naphthylisoquinolinium
alkaloids leading to the most active agents 4f, 4i, and 4m, which
exhibit an enhanced selectivity against L. major compared to
3. L. major promastigotes were sensitive to these derivatives in
the low micromolar range. The extent of plasma protein binding
of the compounds 4f, 4i, and 4m was in a range between 50
* To whom correspondence should be addressed. For H.M.: phone, +49-
931-312627; fax: +49-931-312578; E-mail: heidrun.moll@mail.uni-
wuerzburg.de. For G.B.: phone: +49-931-8885323; fax, +49-931-8884755;
E-mail: bringmann@chemie.uni-wuerzburg.de.
†
Institute of Molecular Infection Biology, University of Wu¨rzburg.
Laboratory of Molecular Physiology, Universidad Central de Venezuela.
Institute of Organic Chemistry, University of Wu¨rzburg.
|
‡
§
Institute of Pharmacy and Food Chemistry, University of Wu¨rzburg.
10.1021/jm801084u CCC: $40.75
2009 American Chemical Society
Published on Web 12/31/2008

N,C-Coupled Arylisoquinolines
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 627
Figure 1. The N,C-coupled isoquinolinium alkaloids ancistrocladinium A (1) and B (2) and their structurally simplified analogues 3, 4f, 4i, and
4m.
Scheme 1. Rational 1-Step Synthesis of the
and 70%. This is adequate from the pharmacokinetic point of
view for a new lead compound. However, for further develop-
N-Arylisoquinolinium Salts 4a
ments, it would be fundamental to decrease the extent of plasma
protein binding. On the other hand, the arylisoquinolines 4f,
4i, and 4m were toxic against mammalian cell lines at
concentrations (=30 µM, total concentration) that are signifi-
cantly higher than those effective against parasites. Interestingly,
when we assessed the interaction between amphotericin B and
a Reagents and conditions: (a) HOAc, room temperature.
the isoquinolines 2, 3, 4f, and 4m in vitro, using a modified
fixed-ratio isobologram method,15,16 we found that the i-propyl
derivative 4m and amphotericin B act in a synergistic manner.
We additionally investigated the drug interaction potential
of these substances using in vitro assays.17 Neither the naph-
thylisoquinoline 3 nor compounds 4f, 4i, or 4m inhibited
significantly the major human drug-metabolizing cytochrome
P450 (CYPa) enzymes 1A2, 2C8/9/19, or 3A4; however, they
were strong inhibitors of CYP2D6, an enzyme not involved in
the metabolism of antileishmanial drugs. These results suggest
that the potential drug interaction of the arylisoquinolinium salts
3, 4f, 4i, and 4m may not be of clinical significance if these
structurally simplified substances are applied in combination
with common leishmanicidal therapeutics.
Scheme 2. Reduction of Isoquinolinium Salts 4 and their
Dihydro Analogue 7a
Taken together, our results indicate that the arylisoquino-
linium salt 4m is a promising candidate for further investigation
of its potential usefulness as a leishmanicidal drug, especially
because combination therapy seems to be fundamental to
decrease the cytotoxicity of individual drugs as well as the
duration of the therapy and thus could have direct clinical
relevance to prevent the emergence of drug resistance.
a Reagents and conditions: (a) NaBH4, MeOH, H2O, 0 °C.
The low activity of noncharged C,C-coupled naphthyliso-
quinoline alkaloids against L. major promastigotes prompted
us to evaluate the influence of the oxidation level of the
isoquinoline portion and thus the positive charge at the nitrogen
atom in the isoquinoline portion. To this end, selected substrates
were further transformed to N,C-coupled tetrahydroisoquinolines
by treatment of the corresponding isoquinolines or their dihydro
analogues with NaBH4 (Scheme 2). The reduction of the fully
dehydrogenated isoquinolinium salts 4 with sodium borohydride
resulted in the cis-configured compounds 8 in high diastereo-
selectivities (ds > 90). If the dihydro substrates 7 were applied
as the educt, by contrast, the attack of the hydride transfer
reagent occurred mainly from the opposite side, delivering the
trans isomer of 8 as the major product (dr ) 2:1).
Results and Discussion
Chemistry. In addition to the previously reported naphthyl-
isoquinolinium salt 3,14 we prepared a series of dehydroiso-
quinolines of type 4 that varied in the aryl portion. These
substrates were synthesized in a convergent way by cyclocon-
densation of the benzopyrylium salt 518 and differently substi-
tuted aromatic amines 6 in good to excellent yields (Scheme
1). This synthetic pathway allowed us to rapidly generate SAR
and to efficiently scale up key compounds needed for initial
studies on the mode of action of this novel class of anti-infective
compounds.
a Abbreviations: CYP, drug-metabolizing cytochrome P450 enzymes;
DC, dendritic cells; DMSO, dimethylsulfoxide; EC50, concentration that
decreases the macrophage infection rate by 50%; FIC, fractional inhibitory
concentration; IC50, concentration that inhibits cell proliferation by 50%;
IL, interleukin; LPS, lipopolysaccharide; NADPH, nicotinamide adenine
dinucleotidephosphate;NO,nitricoxide;QSAR,quantitativestructure-activity
relationship; SAR, structure-activity relationship; ΣFICs, sum FICs; TGF-
ꢀ, transforming growth factor-ꢀ, TNF-R, tumor necrosis factor-R.
Structure-Activity Relationship. Compounds of type 4 and
8 were tested for their ability to inhibit the proliferation of L.
major promastigotes in vitro using the Alamar Blue assay.14,19
In parallel, the toxicity of the compounds was tested using the
macrophage cell line J774.1 (Table 1).

628 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Ponte-Sucre et al.
Table 1. In Vitro Activities of Synthetic Analogues of Arylisoquinolines 4 and the Standard Amphotericin B against L. major Promastigotes and J774.1
Macrophages
a The data represent IC50 values ( standard deviations and are expressed in µM. Experiments with parasites and macrophages were performed in parallel.
b The index was calculated as the ratio of the IC50 value for macrophages to the IC50 to decrease the proliferation of L. major promastigotes. n.d.: not
determined.1-3 The activity values described for compounds 4f, 4i, and 4m against L. major and macrophages are significantly different with p < 0.05.
At first the SAR of the aryl portion was explored by
maintaining the isoquinoline portion. The results in Table 1 show
that compounds 4a-g, 4j, and 4i-p exhibited a very good
activity against the pathogen L. major with submicromolar to
micromolar IC50 (the concentration that inhibits 50% cell
proliferation) values, which were comparable to that of the
parent compound 3 (IC50 ) 2.91 µM)14 and that of amphotericin
B evaluated under the same conditions (IC50 ) 5.07 µM). The
sensitivity of Leishmania promastigotes to amphotericin B was
10- to 25-fold lower than previously reported in the literature.19
However, it is important to note that intracellular amastigotes,
the parasite stage residing in host macrophages, were sensitive
to amphotericin B within the same range as previously de-
scribed14 (see the section on the macrophage infection rate
below). Notably, all active substances described here bear a
lipophilic, weakly electron-donating aryl portion displayed either
by a nonsubstituted aromatic ring system or by alkyl groups
that are located ortho or para to the heterobiaryl axis at the
phenyl ring. The importance of this structural feature was further
demonstrated by the synthesis and the testing of the correspond-
ing derivatives 4q-y equipped with electron-withdrawing or
hydrophilic electron-donating substituents, e.g., halogen, car-
boxy, or methoxy. They displayed a very poor or even no
inhibition of the parasite growth.
The sterical demand of the aryl portion, e.g., naphthyl vs
anthracenyl, or of the size of the alkyl groups, e.g. methyl vs
propyl, seems not to influence the activity against L. major
promastigotes significantly but plays an important role for the
toxicity of the compounds. In comparison, compound 4d having
an anthracenyl portion was about 5-fold more active against
the parasite than its naphthalene analogue 4b but at the same
time up to 35-fold more toxic against macrophages, while the
phenyl analogue, tested in our recent study,14 inhibited the
growth of the macrophages at concentrations even 190-fold

N,C-Coupled Arylisoquinolines
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 629
Table 2. In Vitro Activity of Tetrahydroisoquinolines 8 against L. major Promastigotes and J774.1 Macrophages
a The data represent IC50 values ( standard deviations and are expressed in µM. Experiments with parasites and macrophages were performed in parallel.
Table 3. Cytotoxicity of Selected Synthetic Analogues of the
Arylisoquinolines and Amphotericin B against Dendritic cells (DC),
NIH 3T3 Fibroblasts, and Peritoneal Macrophages
higher than that of 4d. Interestingly, the substrates 4w-y,
possessing huge but more hydrophilic aryl moieties, e.g.,
quinolinyl, benzthiazolyl, or anthraquinoyl, were not cytotoxic,
compound
DCa
NIH 3T3a
peritoneal macrophages
indicating that the lipophilicity but not the bulkiness of the aryl
portion is the decisive factor for the toxicity of the N,C-coupled
arylisoquinolinium salts. This result was further corroborated
by the analogues having either a higher number of substituents,
e.g., trimethyl, or bearing a longer carbon side chain at the
phenyl ring, leading to more lipophilic and thus more toxic
compounds than the monosubstituted arylisoquinolinium salts.
The N,C-linked tetrahydroisoquinolines 8, having a tertiary
nitrogen atom at the heterocycle, displayed weak or no leish-
manicidal activities (Table 2), while their corresponding dehydro
or dihydro14 analogues showed good to very good activities
against proliferation of L. major promastigotes (Table 1). This
finding can most likely be explained by the difference in the
hybridization of the nitrogen atoms in 4 as compared to 8 (sp2
vs sp3). This does not only lead to an electrostatically diverse
scaffold of the substrates due to the positively charged com-
pounds 4 in contrast to 8 but also to large differences in the
molecular shapes of 4 and 8. Their conformational arrays are
mainly induced by the different orientations of the two molecular
portions. In the case of the positively charged salts 4, the
isoquinolinium and the aryl portions are linked via a heterobiaryl
axis leading to an almost orthogonal alignment of the two halves,
while in the tetrahydroisoquinolines 8, the nitrogen atom adopts
a tetrahedral conformation.
Because the C,C-coupled naphthylisoquinoline alkaloids,
which also feature an almost orthogonal assembly of the two
molecular halves and whose molecular shapes are thus com-
parable to those of compounds 4, exhibit only low effectiveness
against L. major, as was demonstrated in our recent study,14
the activities of compounds 4 against Leishmania parasites
herein seem to depend on their cationic structure.
Altogether, the most promising substrates evaluated in the
SAR studies are the methyl and i-propyl isoquinolinium
derivatives 4f, 4i, and 4m (Table 1). They are effective against
L. major promastigotes at concentrations in the low micromolar
range, while they impair the growth of J774.1 macrophages at
concentrations that are at least 6-fold higher. The data also
suggest that the efficacy of these compounds against L. major
amphotericin B
4f
4i
4m
nd
nd
>70
24.0 ( 8.00 49.8 ( 1.10
8.76 ( 2.58 37.6 ( 13.8
7.60 ( 2.78 11.1 ( 5.83
12.77 ( 3.07
9.67 ( 2.21
8.81 ( 2.87
a The data represent IC50 values ( standard deviations and are expressed
in µM. Experiments with parasites and macrophages were performed in
parallel. nd: not determined.
is similar to that reported for amphotericin B in the experiments
described herein, although they are more toxic against the
macrophage cell line J774.1.
In summary, the study on this novel class of antileishmanial
compounds described here implies that the good antileishmanial
properties of N,C-coupled isoquinolinium salts do not only
depend on the oxidation state of the isoquinoline portion, and
thus on the overall shape and on the electrostatic properties of
the substrates, but also strongly rely on the substitution pattern
of the aryl portion of these compounds. These further hints at
the structural features needed for antileishmanial activity help
to improve our model for the activity- and toxicity-guided
quantitative structure-activity relationship (QSAR) investiga-
tions that are in progress.
Cytotoxicity of Selected Arylisoquinolines (Compounds
4f, 4i, and 4m) against Different Cell Lines. The cytotoxicity
of the most promising substances 4f, 4i, and 4m (Figure 1) was
further evaluated with different cell lines. The compounds
inhibited the growth of dendritic cells as well as the survival of
NIH 3T3 fibroblasts and of peritoneal macrophages at concen-
trations at least 3-fold higher than the concentration needed for
a 50% inhibition of the growth of L. major promastigotes (Table
3). Together, these results demonstrated that the simplified
analogues 4f, 4i, and 4m of the N,C-coupled isoquinolinium
salts remained selective against Leishmania promastigotes
compared to different mammalian cells.
Effect of N,C-Linked Arylisoquinolines 4f, 4i, and 4m
on the Infection Rate of Macrophages. To further analyze
the antileishmanial activity of 4f, 4i, and 4m, freshly isolated
peritoneal macrophages were infected with L. major for 24 h

630 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Ponte-Sucre et al.
Table 4. Infection Rate of Selected N,C-Coupled Arylisoquinolines and
Amphotericin B
Modulation of Macrophage Cytokine Secretion by N,C-
Coupled Arylisoquinolinium Salts 3, 4f, 4i, and 4m. To further
analyze the leishmanicidal activity of the isoquinolines 3, 4f,
4i, and 4m, we evaluated the levels of interleukin-1 beta (IL-
1ꢀ), IL-6, IL-10, IL-12, transforming growth factor beta (TGF-
ꢀ) and tumor necrosis factor alpha (TNF-R) in culture super-
natants of noninfected or infected cells treated either with
compounds 3, 4f, 4i, or 4m in the absence or presence of the
potent macrophage activators IFN-γ or lipopolysaccharide
(LPS). Compound 3 was used as an internal control of the
experiment. As described previously,14 its effect is not associated
with the stimulation of cytokine secretion or the production of
nitric oxide by infected macrophages. The results demonstrate
that noninfected untreated cells secreted 0.035 ( 0.014 ng/mL
of IL-1ꢀ, 2.014 ( 1.159 ng/mL of IL- 6, 0.101 ( 0.080 ng/mL
of IL-10, 0.417 ( 0.014 ng/mL of IL-12, and 0.330 ( 0.108
ng/ml of TGF-ꢀ. Treatment with compounds 3, 4f, 4i, or 4m
did not cause significant changes in the levels of IL-1ꢀ, IL-6,
or TNF-R; these compounds strongly reduced the release of IL-6
and TNF-R in response to the macrophage stimulators LPS or
IFN-γ (data not shown). These are pro-inflammatory cytokines,
which are secreted in early stages of L. donoVani and L. major
infection23,24 and seem to be involved in the early recruitment
of inflammatory cells to the site of infection.25 Furthermore,
the host defense against L. major infection depends on the IL-
12-driven expansion of T helper 1 cells.26,27 However, aryliso-
quinolinium salts strongly inhibit IL-12 secretion by infected
macrophages. Indeed, these results indicate that the studied
isoquinolines do not stimulate macrophages to secrete pro-
inflammatory cytokines but rather interfere with their cytokine
response to potent stimulators.14
compound
infection ratea
indexc
amphotericin B 0.075 [0.070] (CIb: 0.040-0.139, r2 0.999) >1000
4f
4i
4m
0.092 [0.036] (CIb: 0.040-1.99, r2 0.923)
0.241 [0.104] (CIb: 0.125-0.461, r2 0.963)
0.158 [0.070] (CIb: 0.08-0.312, r2 0.978)
138
40
56
a The data represent EC50 values and are expressed in µM and in µg/mL
within rectangular brackets. b CI: confidence interval. c The index was
calculated as the ratio of the IC50 value for peritoneal macrophages (Table
3) to the EC50 to decrease the infection rate.
and then treated with increasing concentrations of the com-
pounds 4f, 4i, and 4m for 48 h. Table 4 presents a summary of
the data for macrophage survival and decrease in infection rate.
Untreated macrophages had infection rates of 20-30% (con-
sistent within each experiment), which were normalized to 100%
for further analysis of the results. Treatment of infected
macrophages with compounds 4f, 4i, and 4m resulted in a dose-
dependent decrease in infection rate. A 50% decrease (EC50)
was obtained at concentrations of 0.092 µM (confidence interval
0.040 to 1.99 µM, r2 0.923), 0.241 µM (confidence interval
0.125 to 0.461 µM, r2 0.963) and 0.158 µM (confidence interval
0.080 to 0.312 µM, r2 0.978) for 4f, 4i, and 4m, respectively
(Table 4). Similar to what was previously demonstrated,14
simultaneous addition of the arylisoquinolines 4f, 4i, or 4m and
the macrophage activator interferon-γ (IFN-γ) did not further
modify the EC50 (data not shown). The EC50 for amphotericin
B was 0.075 (confidence interval 0.0397 to 0.139 µM, r2 0.999),
similar to what had been previously described.14,20
These results indicate that a significant decrease in macroph-
age infection rate is obtained at concentrations of compounds
4f, 4i, and 4m that are 9- to 30-fold lower than those needed to
inhibit the growth of L. major promastigotes. The efficacy of
these active agents against intracellular amastigotes is compa-
rable to that of amphotericin B. Notably, although the EC50
against amastigotes was not significantly different from the one
obtained with the naphthylisoquinoline 3, the toxicity against
different cell lines and freshly isolated cells from BALB/c mice
is at least 40-fold, and in some cases even 500-fold, lower than
that against intracellular amastigotes. These data confirm our
previous finding that N,C-coupled isoquinolines are selective
against the parasites and in general do not stimulate the killing
mechanisms of macrophages.
The results also demonstrate that noninfected untreated cells
secreted 0.101 ( 0.080 ng/mL of IL-10, and 0.330 ( 0.108
ng/mL of TGF-ꢀ. However, the levels of TGF-ꢀ (Figure 2)
increased significantly (p < 0.001) after incubation with
compounds 3 or 4m (with or without IFN-γ), while the
substrates 4i and 4m, but not 3 and 4f, decreased IL-10 levels
(p < 0.001) significantly (Figure 2).
The results obtained with the 4′-i-propyl derivative 4m
strongly suggest that the action of some arylisoquinolines may
help to limit the extent of the inflammatory response and to
restore homeostasis of the host cell. Leishmania infection
induces TGF-ꢀ production and a delayed nitric oxide produc-
tion,25 consistent with a TGF-ꢀ inhibition of microbicidal action.
TGF-ꢀ is a cytokine with dual pro- and anti-inflammatory
functions.28 To play its role, TGF-ꢀ should act through the
induction of IL-10, which has been categorized as a T helper 2
cytokine.28 Infections with L. donoVani28,29 and L. major30
increase IL-10 levels, which in turn facilitates the intracellular
survival of the protozoa, orchestrates several immunomodulatory
pathways,25 and plays a significant role in disease initiation and
progression.31 The isoquinoline 4m increased the secretion of
TGF-ꢀ and decreased the secretion of the pro-inflammatory
cytokine IL-10, whereas it is very difficult to interpret the
leishmanicidal effect of compound 4f in view of the unusual
increase in IL-10 levels observed in the simultaneous presence
of compound 4f and IFN-γ. This result is even more interesting
with respect to the fact that compound 4f does not modulate
the secretion of TGF-ꢀ and, therefore, does not seem to play a
role in the feedback loop that exists between these two cytokines.
Because the pharmacokinetic behavior of a drug also governs
the efficacy, the extent of plasma protein binding was exemplar-
ily determined for the compounds 4f, 4i, 4k, 4m, 4n, and 4x,
using the method of automated continuous ultrafiltration de-
veloped and validated in our group.21 One of the compounds,
4x, did not induce a significant decrease in macrophage infection
rate. The percentages of binding were 55.9 ( 9.1% for 4f, 46.2
( 7.2% for 4i, 52.5 ( 7.4% for 4k, 70.3 ( 7.6% for 4m, 58.4
( 6.5% for 4n, and 32.9 ( 8.0% for 4x, indicating that only
half of the drug is available for the treatment of a leishmanial
infection and that there is no correlation between plasma protein
binding and the antileishmanial activity of the compounds.
However, the extent of plasma protein binding is within an
acceptable range. As an example, two new antibiotics, namely
moxifloxacin and telithromycin (see Supporting Information),
bind to a similar extent to human serum albumin.22 The most
active compound in the series of the N,C-coupled arylisoquino-
lines, however, showed a very high extent of plasma protein
binding. Further optimization of the compounds should aim at
a decrease of the plasma protein binding.
To analyze whether the decrease in macrophage infection rate
observed in the presence of compounds 3, 4f, 4i, or 4m
correlates with the capacity of macrophages to secrete nitric
oxide, we evaluated the nitrite production in noninfected and

N,C-Coupled Arylisoquinolines
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 631
Figure 2. TGF-ꢀ and IL-10 secretion by macrophages. Macrophages were infected with stationary-phase L. major promastigotes for 24 h and
treated with the compounds in the absence or presence of IFN-γ or LPS. n-Fold change (infected/noninfected) of TGF-ꢀ levels in supernatants of
infected macrophages cultured in the presence of compounds 3, 4f, 4i, or 4m in the absence or presence of IFN-γ or LPS (a); n-fold change
(infected/noninfected) of IL-10 levels in supernatants of infected macrophages cultured in the presence of compounds 3, 4f, 4i, or 4m in the
absence or presence of IFN-γ or LPS (b). The horizontal lines indicate changes in cytokine production induced by IFN-γ or LPS alone. p < 0.001,
p < 0.05. A ratio of 1 indicates no change.
Table 5. Mean ΣFICs of the Interaction between Amphotericin B and
the Arylisoquinolines 2, 3, 4f, or 4m towards Intracellular Amastigotes
In Vitro
fixed relationa
2
3
4f
4m
1:4
2:3
3:2
4:1
0.552
0.901
0.341
2.33
0.67
2.98
<0.5
<0.5
<0.5
<0.5
>4
>4
3.02
>4
2.76
>4
a The different fixed relations are established as amphotericin B:
compounds 2, 3, 4f, and 4m. FIC: fractional inhibitory concentrations of
compounds 2-4f and 4m; ΣFICs (sum FICs), [FIC amphotericin B + FIC
compounds 2-4f and 4m].
development of resistance.17 It could thus be of fundamental
impact for leishmaniasis therapy. In our studies, we have
explored the in vitro interaction between N,C-coupled aryliso-
quinolines and amphotericin B.
We assessed in vitro interactions using a modified fixed-ratio
isobologram method and analyzed it at the EC50 level.16 The
fractional inhibitory concentrations (ΣFICs, sum FICs) are
presented in Table 5. The results are the mean values of two
independent experiments. Representative isobolograms are
shown in Figure 4. Interactions were classified as synergistic
with values of ΣFICs g, as antagonistic with values of ΣFICs
g 4, and as indifferent with values of ΣFICs between >0.5
and g 4.17
The interaction of amphotericin B with the alkaloid ancis-
trocladinium B (2) was indifferent at all the fixed relations
evaluated with values of ΣFICs between 2.33 and 0.52. Notably,
the interaction of amphotericin B with structurally simplified
analogues 3 and 4f ranged from indifferent to antagonistic with
ΣFICs from 0.67 to >4. Finally, the interaction of amphotericin
B with compound 4m was synergistic with ΣFICs < 0.5 for all
the fixed ratios evaluated, reaching values as low as 0.003 when
a fixed relation (2:3) of amphotericin B to 4m was used. This
fixed relationship arises from the combination of amphotericin
B (0.08 µM) with compound 4m (0.15 µM); these concentra-
tions lie within their corresponding EC50 to decrease the
macrophage infection rate. Thus, the results suggest that the
interaction between amphotericin B and the 4′-i-propyl deriva-
tive 4m at their EC50 is extremely synergistic, indicating that
4m may be a promising candidate for further investigation of
its potential usefulness as an antiprotozoal drug.
Figure 3. Nitric oxide production by macrophages. Macrophages were
infected with stationary-phase L. major promastigotes for 24 h. Cells
were then washed and incubated for further 48 h with the compounds
in the absence or presence of IFN-γ or LPS. The culture supernatants
were then collected for determination of the nitrite concentrations.
n-Fold change (infected/noninfected) of the micromolar concentration
of nitrite in supernatants of infected macrophages cultured in the
presence of compounds 3, 4f, 4i, or 4m in the absence or presence of
IFN-γ or LPS. A ratio of 1 indicates no change.
infected cells. The nitric oxide release was measured after 48 h
of incubation of infected macrophages with the compounds
(Figure 3). Noninfected cells treated with compounds 3, 4f, 4i,
or 4m secreted nitric oxide levels that were similar to those of
noninfected untreated cells. These levels did not change in cells
incubated with IFN-γ or with compounds 3, 4f, 4i, or 4m plus
IFN-γ (data not shown). Infection with Leishmania promastig-
otes decreased the nitric oxide production (Figure 3). The
treatment with LPS, or compounds 3, 4f, 4i, or 4m and LPS
simultaneously, did not increase nitric oxide secretion. These
levels increased 3- to 4-fold in cells treated with IFN-γ and
2.7-fold in cells incubated only with compound 4m (p < 0.005).
These results strongly suggest that the mode of action of the
isoquinolinium salt 4m, with its sterically demanding i-propyl
substituent para to the heterobiaryl axis, may be different from
that of the other methyl-substituted N,C-coupled isoquinolines
tested and may help to limit the extent of the inflammatory
response and to restore homeostasis of the host cell.
Analysis of Interactions of N,C-Coupled Arylisoquinolinium
Salts 2, 3, 4f, and 4m with Amphotericin B against Intracel-
lular Amastigotes In Vitro. Combination therapy has been
addressed as a way to decrease (a) the cytotoxicity of individual
drugs, (b) the duration of therapy, and (c) the potential
Because in vitro data are based on an extended ratio and
concentration range, these results strongly underline the impor-
tance of testing the possible interaction between amphotericin

632 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Ponte-Sucre et al.
Figure 4. Representative isobolograms of in vitro interactions against intracellular amastigotes. Interactions are given at the EC50 levels. Numbers
at the axes represent normalized FIC values of amphotericin B (x axes) and the partner drug 2, 3, 4f, and 4m (y axes).
B and the arylisoquinolinium trifluoroacetate 4m in vivo. The
mechanism of interaction that leads to the enhancement of the
effects of these two agents remains to be elucidated. Ampho-
tericin B toxicity occurs through its binding to sterols in the
cell membrane, formation of aqueous pores, and the induction
of programmed cell death.32,33 Compound 4m induces the
appearance of intracellular vacuoles and its killing effect shares
characteristics with programmed cell death (Ponte-Sucre et al.,
unpublished).
at the tested concentrations of 1, 10, and 100 µM, compound
4m showed a higher selectivity toward CYP2D6 than the
prototypical inhibitor quinidine (Figure 5).
Because all tested concentrations of the aryl isoquinoline
reduced the activity of CYP2D6, we determined the IC50 values
of compounds 3, 4f, and 4m and of quinidine for the inhibition
of CYP2D6 to compare the inhibitory activities and to evaluate
the drug interaction potential of the substrates (Figure 6).
Quinidine had the highest inhibitory activity for CYP2D6, with
an IC50 value of 19.6 nM. Although compounds 3, 4f, and 4m
had higher IC50 values (0.9 µM, 56.1 nM and 109.4 nm; Figure
6), they can be considered as strong inhibitors of CYP2D6.
The inhibitory mechanism exerted by an inhibitor of drug-
metabolizing CYP enzymes strongly influences the pharmaco-
kinetics of concomitantly administered drugs metabolized via
the same CYP enzymes. Herein we demonstrate that compounds
4f and 4m are potent and highly selective inhibitors of CYP2D6
(Figures 5 and 6). Moreover, the results shown in Figure 7,
where the Lineweaver-Burk plot for the inhibition of CYP2D6
by compound 4m is illustrated, demonstrate that, similar to
quinidine (Ki 9.8 nM), compound 4m is a strong competitive
inhibitor of CYP2D6 with a Ki value of 54.7 nM. These results
suggest that 4m and the structurally closely related compounds
3 and 4f (Ki 0.45 µM, 28.1 nM) bind with high affinities to the
active site of the enzyme.
The N,C-coupled isoquinolines 3, 4f, and 4m are thus strong
and selective inhibitors of the drug-metabolizing CYP enzyme
2D6. This highly polymorphic isoenzyme is involved in the
metabolism of 20-25% of the drugs presently in clinical use.35
However, the currently applied medication against leishmaniasis
(e.g., pentavalent antimonials, amphotericin B, or miltefosine)
does not include drugs metabolized by CYP2D6. Interestingly,
the isoquinolines 3, 4f, and 4m are strong and highly selective
competitive inhibitors of CYP2D6; therefore, adverse drug
interactions due to the inhibition of metabolic pathways
catalyzed by CYP1A2, 2C8/9/19, or 3A4 can be excluded. The
inhibitory activity of compound 4i on CYP2C19 (Figure 4) is
Inhibition of Human Drug-Metabolizing CYP Enzymes
by N,C-Coupled Isoquinolines 3, 4f, 4i, and 4m. The most
important biotransformation pathway for drugs in mammals is
their oxidation by drug-metabolizing cytochrome P450 (CYP)
enzymes, which are expressed mainly in the liver but also in
the lung, kidney, or small intestine.34 Among CYP enzymes
involved in biotransformation reactions, the isoenzymes 1A2,
2C8, 2C9, 2C19, 2D6, and 3A4 are responsible for the
metabolism of about 90% of all known drugs.34 To assess their
relevance for drug metabolism, we tested whether the antile-
ishmanial N,C-coupled isoquinolines 3, 4f, 4i, and 4m interfere
with metabolic reactions catalyzed by the major human drug-
metabolizing CYP enzymes.
Because the medical treatment of parasitic infections often
requires the simultaneous application of various drugs, a low
inhibitory activity of such drugs on CYP enzymes is an
important prerequisite for their combined therapeutic use. In
Figure 5, we present the inhibitory activity of the isoquinolines
3, 4f, 4i, and 4m and of quinidine (see Supporting Information),
a strong and selective CYP2D6 inhibitor, on the applied CYP
enzymes. Except for 4i, all tested compounds are strong and
selective inhibitors of CYP2D6. The inhibitory activity of 3,
4f, and 4m is comparable to that of the CYP2D6 standard
inhibitor quinidine (Figure 5). It is important to note that
compound 4m in particular did not inhibit the activity of the
additional CYP enzymes at all, whereas compounds 3, 4f, and
4i, as well as quinidine also reduced the activity of CYP2C9,
CYP2C19, and CYP3A4 at a concentration of 100 µM. Thus,

N,C-Coupled Arylisoquinolines
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 633
Figure 5. Inhibition of CYP1A2, 2C8/9/19, 2D6, and 3A4 by 1, 10, and 100 µM of compound 3 (A), 4m (B), 4f (C), or 4i (D), or quinidine (E).
The results represent mean values of triplicate determinations ((standard deviation).
and the competitive inhibition of CYP2D6 points to a direct
interaction of these substrates with the catalytic center of the
enzyme. This high-affinity binding might be due to a coordina-
tion of the ferriprotoporphyrin IX (heme) of CYP2D6 as
demonstrated in the complexation of ferriprotoporphyrin IX by
antimalarial drugs like chloroquine and the antiplasmodial C,C-
linked naphthylisoquinoline alkaloid dioncophylline C.36
Conclusions
In summary, the presented study on the SAR of N,C-coupled
arylisoquinolines showed that, in addition to the decisive impact
of the biarylic structure for the activity, the aryl portion of these
compounds also plays an important role for both activity and
toxicity of this new class of antileishmanial agents. These
findings now provide the basis for a more specific QSAR-guided
design, selection, and synthesis of novel antileishmanial N,C-
linkedsyntheticarylisoquinolineswithanimprovedtoxicity-activity
ratio.
Furthermore, the results strongly suggest that the significant
decrease in macrophage infection rate induced by N,C-coupled
arylisoquinolinium salts is based on a direct effect of the
compounds on the intracellular parasites. However, the effect
Figure 6. Inhibition curves and IC50 values for the impairment of
CYP2D6 by compound 3 (A), 4m (B), and 4f (C) and of quinidine
of the 4′-i-propyl derivative 4m might involve both a direct
leishmanicidal effect and the modulation of the cytokine activity
and nitric oxide production by host macrophages.
(D). The results represent mean values of triplicate determinations ((
standard deviation).
too low to cause clinically relevant drug interactions with drugs
preferentially metabolized by this isoenzyme. This means that
compounds 3, 4i, 4f, or 4m could be safely prescribed together
with antileishmanial drugs in a combination therapy approach,
once the therapeutic index of these aryl isoquinolines is further
enhanced. The strong inhibitory activity of compounds 4f and
4m on CYP2D6 (Ki values 28.1 nM and 54.7 nM) is remarkable
Finally, because compound 4m acts synergistically with
amphotericin B in vitro, it is a promising candidate for further
investigations on its potential usefulness as an antiprotozoal
drug, although the toxicity of this newly synthesized structurally
simplified N,C-coupled arylisoquinolinium salt against the host
cells is still high. Additionally, because antileishmanial drugs
are not metabolized by the CYP2D6 isoenzyme, the risk of

634 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Ponte-Sucre et al.
J ) 7.0 Hz, 6H), 2.30 (s, 3H), 2.89 (s, 3H), 3.07 (q, J ) 6.8 Hz,
1H), 3.97 (s, 3H), 4.06 (s, 3H), 6.94 (d, J ) 2.1 Hz, 1H), 7.08 (d,
J ) 2.9 Hz, 1H), 7.25 (d, J ) 8.3 Hz, 2H), 7.53 (d, J ) 8.3 Hz,
2H), 8.03 (s, 1H). 13C NMR (CDCl3): δ ) 22.82, 24.09, 24.47,
34.67, 57.34, 57.46, 100.1, 103.3, 116.0, 120.6, 123.8, 126.8, 129.9,
137.6, 143.2, 145.1, 153.1, 159.4, 161.1, 162.0, 168.3; HRESIMS
adverse drug reactions with conventional antileishmanial drugs
and of drug interactions caused by inhibition of other drug-
metabolizing enzymes is excluded. In vivo experiments on the
combination of amphotericin B with 4m will be fundamental
to confirm the present data and to evaluate further the potency
of the arylisoquinolinium salt 4m in combination therapy.
calcd for C22H26NO2 (M)+, 336.19581; found, 336.19621.
+
Parasites. The cloned virulent L. major isolate MHOM/IL/81/
FE/BNI was maintained by passage in BALB/c mice (6-8 weeks
old). Promastigotes were grown in blood agar cultures at 26 °C,
5% CO2, 95% humidity. Prior to their use, promastigotes were
washed twice with phosphate-buffered saline and suspended at 1
× 108 cells/mL in Click RPMI 1640 medium supplemented with
10% fetal calf serum, 2 mM L-glutamine, 10 mM HEPES buffer
pH 7.2, 100 µg/mL penicillin, 160 µg/mL gentamicin, 7.5%
NaHCO3, and 5 × 10-5 M 2-mercaptoethanol (complete medium).
Cells and Cell Lines. The macrophage cell line J774.1 was
maintained in complete medium. Prior to their use, cells were
detached from the flasks with a rubber policeman, washed twice
with phosphate-buffered saline, and suspended at 2 × 106 cells/
mL in complete medium. Peritoneal macrophages were obtained
from BALB/c mice by a previously described protocol.14 Dendritic
cells were generated from bone marrow progenitors of BALB/c
mice as described.14 The mouse embryo fibroblast cell line NIH
3T3 was grown to 80-90% confluence using a previously described
protocol.14 For the experimental procedures, cells were detached
from the flasks with a rubber policeman, washed with phosphate-
buffered saline, and suspended in Dulbecco’s modified Eagle
medium (DMEM) at 2 × 106 cells/mL.
Experimental Section
General Information. All used solvents were distilled before
use. Commercially available material was used without further
purification. Reactions with NaBH4 were carried out in predried
glassware under argon. Melting points were determined on a
Reichert-Jung Thermovar hot plate and uncorrected NMR spectra
(1H NMR: 400 MHz; 13C NMR: 101 MHz) were recorded on a
Bruker AMX, using the solvents as the internal 1H and 13C standards
with coupling constants (J) in Hertz (Hz). HRESIMS was measured
on a Bruker Daltonik micrOTOF-focus. The combustion analyses
were performed on a Leco CHNS 932 analyzer. Thin-layer
chromatography was carried out using silica gel 60 F254 or C-18
F254s aluminum foil. Detection of the compounds was achieved
by fluorescence quenching at 254 nm or fluorescence at 356 nm.
Gel chromatography was performed using Sephadex LH-20.
Analytical HPLC was performed on a Jasco System (DG-1580,
LG-1580, PU-1580, CO-1560, AS-1555, MD-1510) using a Chro-
molith Performance RP-18e (100 mm × 4.6 mm) with a MeCN
(A)/H2O (B) solvent mixture complemented by 0.05% TFA: 0 min
5% A, 5 mL/min; 5 min 50% A, 5 mL/min; 9 min 100% A, 5 mL/
min. Preparative HPLC was carried out on a Jasco System (PU-
2087 Plus, MD-2010 Plus) with a Waters 600 Controller and a
Waters 996 Photodiode Array Detector. As a column a Chromolith
SemiPrep-RP-18e (100 mm × 100 mm) was used with a MeCN
(A)/H2O(B) solvent mixture with 0.05% TFA: 0 min 5% A, 10
mL/min; 5 min 50% A, 10 mL/min; 7 min 100% A, 10 mL/min.
N-(4′-Methylphenyl)-6,8-dimethoxy-1,3-dimethylisoquinolini-
um Tetrafluoroborate (4f). A mixture of 288 mg (0.95 mmol) of
6,8-dimethoxy-1,3-dimethyl-2-benzopyrylium tetrafluoroborate (5)
and 100 mg (0.95 mmol) of 4-tolylamine in 10 mL glacial acetic
acid was stirred at room temperature for 8 h. The yellow solid was
filtered and washed with diethyl ether. The mother liquor was
diluted with methanol and purified by chromatography on Sepha-
dex-LH20 material with methanol as the eluent. The crude products
were combined and recrystallized from a mixture of methanol,
diethyl ether, and n-hexane to afford 364 mg (0.92 mmol, 97%) of
the isoquinolinium tetrafluoroborate 4f as a white solid, mp >250
Analysis of In Vitro Antiproliferative Activity. The analysis
of the in vitro antiproliferative activity of the compounds was done
using Alamar Blue.14,19 Amphotericin B was used as a reference
compound and positive control. For in vitro studies, the compounds
were dissolved in dimethyl sulfoxide and further diluted in medium.
The final concentration of dimethyl sulfoxide in the medium never
exceeded 1% (v/v) and had no effect on the proliferation of
extracellular or intracellular parasites. For each experiment, every
drug concentration was assayed in duplicate wells. Optical density
values at 48 h were used to calculate the concentration that inhibits
50% cell proliferation (IC50) via linear interpolation.37 The index
included in Table 1 was calculated as the ratio of the IC50 value
for macrophages to the IC50 for L. major.
Analysis of Macrophage Infection Rate. The activity of
compounds on the macrophage infection rate was analyzed as
described previously.14 Amphotericin B was used as a reference
compound and positive control. Intracellular parasites were quanti-
fied by staining with acridine orange and ethidium bromide and
analyzed by fluorescence microscopy at 495 nm as described
previously.38 Data on the infection rate are expressed as mean values
( standard errors of the mean of at least three experiments in which
300 macrophages were analyzed for each drug concentration. The
program Graph-pad was used to fit the data to nonlinear regression
and to determine the concentration that decreases the infection rate
to 50% (the 50% effective concentration [EC50]).
1
°C. H NMR (CDCl3) δ 2.27 (s, 3H), 2.50 (s, 3H), 2.92 (s, 3H),
4.00 (s, 3H), 4.05 (s, 3H), 6.91 (d, J ) 2.1 Hz, 1H), 7.03 (d, J )
2.1 Hz, 1H), 7.23 (d, J ) 8.3 Hz, 2H), 7.47 (d, J ) 8.1 Hz, 2H),
7.3 (s, 1H). 13C NMR (CDCl3) δ 22.05, 22.82, 24.05, 57.35, 57.46,
100.1, 103.2, 116.1, 123.8, 126.8, 132.4, 137.6, 142.2, 143.4, 145.1,
159.5, 162.1, 168.2. Anal. (C20H22NO2BF4) C, H, N.
N-(2′,6′-Dimethylphenyl)-6,8-dimethoxy-1,3-dimethylisoquino-
linium Tetrafluoroborate (4i). Yield ) 78%; brown needles, mp
1
198 °C. H NMR (CDCl3) δ 1.82 (s, 6H), 2.19 (s, 3H), 2.75 (s,
Analysis of Cytokine and Nitric Oxide Production. The
analysis of the effect of compounds on the cytokine production by
infected macrophages was done following a protocol described
previously.39 The potent macrophage activators IFN-γ (100 U/mL)
or LPS (15 µg/mL) were used as positive controls for the production
of cytokines. A concentration of 3 µM of compounds 3, 4f, 4i, and
4m, i.e., close to the IC50 against L. major promastigotes, was used
for this assay. After 24 h, the culture supernatants were collected
and the levels of IL-1ꢀ, IL-6, IL-10, IL-12, TGF-ꢀ, and TNF-R
were quantified by the sandwich enzyme-linked immunosorbent
assay. The developing color in the wells was read at a test
wavelength of 405 nm and a reference wavelength of 490 nm using
a Multiskan Ascent ELISA reader. Data were calculated as ng/
mL. Cytokines secreted by noninfected, untreated macrophages
were normalized to 100% for further analysis of the results. The
detection thresholds were 30.72 pg/mL for IL-1ꢀ, 39.00 pg/mL for
IL-4, 27.24 pg/mL for IL-6, 78.00 pg/mL for IL-10, 72.00 pg/mL
3H), 3.97 (s, 3H), 4.10 (s, 3H), 6.75 (d, J ) 2.3 Hz, 1H), 7.29 (d,
J ) 2.3 Hz, 1H), 7.33 (d, J ) 8.1 Hz, 2H), 7.42 (t, J ) 7.1 Hz,
1H), 8.38 (s, 1H). 13C NMR (CDCl3) δ 17.94, 21.39, 21.83, 57.44,
57.79, 100.8, 103.6, 115.9, 125.4, 131.1, 131.9, 133.7, 138.4, 143.5,
143.8, 157.9, 161.2, 161.5, 161.9, 168.7. HRESIMS calcd for
C21H24NO2 (M)+, 322.18016; found, 322.18011.
+
N-(4′-i-Propylphenyl)-6,8-dimethoxy-1,3-dimethylisoquinolini-
um Trifluoroacetate (4m). A mixture of 288 mg (0.94 mmol) of
6,8-dimethoxy-1,3-dimethyl-2-benzopyrylium tetrafluoroborate (5)
and 128 mg (0.95 mmol) of 4-i-propylphenyl amine in 10 mL
glacial acetic acid was stirred at room temperature for 8 h. The
solvent was evaporated; the brown oil was dissolved in 2 mL MeOH
and purified by chromatography on Sephadex-LH20 material with
methanol as the eluent. The crude product was submitted to
preparative HPLC. The obtained fractions were combined and the
solvent evaporated to yield 376 mg (0.84 mmol, 88%) of compound
1
4m as a yellow solid, mp 217 °C. H NMR (CDCl3) δ ) 1.34 (d,

N,C-Coupled Arylisoquinolines
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3 635
for IL-12, 222.00 pg/mL for IFN-γ, 118.00 pg/mL for TGF-ꢀ, and
40.00 pg/mL: for TNF-R.
The analysis of the effect of compounds on the nitric oxide
production by infected macrophages was done following a protocol
described in literature.39 The nitrite concentrations were determined
using sodium nitrite as a standard. Data are expressed as micromolar
concentrations of nitrite. Nitrite concentrations reflect the nitric
oxide levels released by macrophages.
Analysis of Plasma Protein Binding of Compounds 4f, 4i,
4k, 4m, 4n, and 4x. The binding of the compounds to albumin
was determined by means of a continuous ultrafiltration method.
In this procedure, an ultrafiltration cell with an ultrafiltration
membrane (molecular weight cutoff of 10 kDa) was connected to
a Bischoff DAD 3 L EU. Initially a buffered (3 × 10-2 M, pH 7.4
sodium-phosphate) solution of the substances (10 mg/L) was
pumped continuously through the ultrafiltration cell. During this
procedure the absorbance of the ultrafiltrate is measured at 264
nm and plotted versus time. The system is then rinsed with
phosphate buffer and 1 mL of bovine serum albumin 40 mg/mL
solution is injected into the system. The ultrafiltration membrane
retains the protein molecules in the cell. The buffered solution of
each compound is pumped again through the ultrafiltration cell,
and the absorbance of the ultrafiltrate is measured again. An
interaction between the compound and the proteins present in the
ultrafiltration cell leads to a shift of this second curve to the right
compared to the curve obtained initially. The extent of the shift to
the right of the curve corresponds to the protein avidity of the
compound. The data is evaluated with a special software written
by Prof. Nickel and Dr. Reyer (Bonn).
Analysis of the Interaction of Compounds 2, 3, 4f, and 4m
with Amphotericin B against Intracellular Amastigotes in
Peritoneal Macrophages In Vitro. In vitro drug interactions were
assessed using a modified fixed-ratio isobologram method.15 The
predetermined EC50 value of each drug was used to choose the top
concentrations of the individual drugs in the assay. In this way,
we ensured that the EC50 fell near the midpoint of a six-point 2-fold
dilution series. Top concentrations were 0.2 µM for amphotericin
B in a 72 h assay and 0.150, 0.08, 0.4, and 0.25 µM for each
compound. The top concentrations were used to prepare solutions
with fixed ratios (5:0, 4:1, 3:2, 2:3, 1:4, 0:5 of amphotericin B and
the corresponding compound). These solutions were five times
further diluted in a 2-fold dilution series. Peritoneal macrophages
were treated in the same way as in the macrophage infection rate
assay and the interaction assay mixtures were incubated at 37 °C,
5% CO2, and 95% air humidity for further 48 h. Drug activity was
determined from the percentage of infected cells in drug-treated
cultures in relation to nontreated cultures after acridine orange and
ethidium bromide staining. From the known concentration of
amphotericin B in the fixed-ratio solutions, EC50 values were
calculated by sigmoidal analysis using the program Graph-pad. For
each drug combination, an EC50 was obtained from the fixed-ratio
solutions at ratios 5:0 and 0:5. Solutions at ratios 4:1, 3:2, 2:3, and
1:4 yielded the EC50 of each of the drug combinations.17
Inhibitory Activity of Compounds 3, 4f, 4i, and 4m on
Human Drug-Metabolizing CYP Enzymes. Details of the herein
described methods have been published previously.15 Stock solu-
tions of CYP substrates were prepared in DMSO (4-50 mM). The
N,C-coupled isoquinolines 3, 4f, 4i, and 4m were also dissolved in
DMSO (stock solution 25 mM), and the final assay concentration
ranges were 0.01-100 µM.
The test compounds (3, 4f, and 4m) as well as quinidine were
incubated in a 100 mM Na/K-phosphate buffer containing 3 mM
MgCl2 (pH 7.4) with a commercially available mixture of CYP1A2/
2C8/9/19/2D6 and 3A4 (0.1 mg/mL; Supermix, NatuTec, Frankfurt,
Germany), NADPH (1 mM), and a substrate cocktail consisting of
tacrine (CYP1A2), paclitaxel (CYP2C8), tolbutamide (CYP2C9),
imipramine (CYP2C19), dextromethorphan (CYP2D6), and mida-
zolam (CYP3A4) at 37 °C for 30 min. Control incubations (100%
activity) were performed with pure DMSO. The reaction was
stopped by addition of 150 µL ice-cold MeOH containing the
internal standard reserpine. Finally, samples were centrifuged at
Figure 7. Lineweaver-Burk plot for the inhibition of CYP2D6 by
compound 4m (0, 50, 100, 250 nM). The results represent mean values
of triplicate determinations.
12000g for 6 min and the clear supernatants were used for liquid
chromatography/mass spectrometry (LC/ESI-MS) analysis. The
determination of the IC50 values was determined by incubating
compounds 3, 4f, or 4m or quinidine with 5 pmol/mL of CYP2D6
from baculovirus-infected insect cells (NatuTec) and the corre-
sponding substrate dextromethorphan as described above. By
incubation of various concentrations of dextromethorphan (0, 2.5,
5, 10, and 20 µM) with CYP2D6 (5 pmol/mL) and compound 4m
(0, 50, 100, 250 nM), a Lineweaver-Burk plot was constructed to
determine the inhibition mechanism. All incubations were per-
formed in triplicate.
The metabolites 1-hydroxytacrine, 6R-hydroxypaclitaxel, 4-hy-
droxytolbutamide, desipramine, dextrorphan, 1′-hydroxymidazolam,
and reserpine as the internal standard were analyzed with LC/ESI-
MS after automated online extraction on an Agilent (Palo Alto,
CA) LC/MSD system. The inhibitory activity was calculated by
dividing the metabolite to internal standard peak area ratios by the
peak area ratios obtained with control incubations and was expressed
as “% activity of control”. IC50 values were calculated by nonlinear
regression analysis using SigmaPlot version 10.0 (SPSS Inc.,
Chicago, IL). The Ki values of 3, 4f, and 4m, and of quinidine for
the inhibition of CYP2D6 were calculated using the Cheng-Prusoff
equation Ki ) IC50/(1 + [S]/Km).40
Statistical Analysis. Data on the antiproliferative activity of the
compounds (from at least two experiments) were analyzed with
Thermo Electron Ascent Software and Microsoft Excel. Differences
in IC50 and EC50 values of treated or untreated macrophages were
tested for statistical significance by the unpaired Student’s t test
using the Graph-pad program. FIC values and sum FICs (ΣFICs[FIC
amphotericin B + FIC compounds 2, 3, 4f, and 4m]) were
calculated as follows: FIC amphotericin B ) EC50 of drug in
combination/EC50 of drug alone. The same was applied to each
compound as the partner drug. FICs and ΣFICs were calculated
for all fixed-ratio solutions and FICs were used to construct
isobolograms. Mean ΣFICs were used to classify the nature of the
interaction.18
Acknowledgment. The authors gratefully acknowledge the
financial support from the Deutsche Forschungsgemeinschaft
(Sonderforschungsbereich 630), the Fonds der Chemischen
Industrie (fellowship to T. Gulder and supplies), and the
Hochschul- und Wissenschaftsprogramm of the University of
Wu¨rzburg (fellowship to T. Gulder). We thank Christina de Witt,
Christine Hambrecht, and Melanie Pavlov for technical as-
sistance, and Dr. Marcela Fajardo-Moser and Dr. Katharina
Remer for fruitful discussions.
Supporting Information Available: Chemical structures, ex-
perimental section, purity data, and HPLC analysis. This material
is available free of charge via the Internet at http://pubs.acs.org.

636 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 3
Ponte-Sucre et al.
(20) Escobar, P.; Matu, S.; Marques, C.; Croft, S. L. Sensitivities of
Leishmania species to hexadecylphosphocoline (miltefosine), ET-18-
OCH3 (edelfosine), and amphotericin B. Acta Trop. 2002, 81, 151–
157.
(21) Heinze, A.; Holzgrabe, U. Determination of the extent of protein
binding of antibiotics by means of an automated continuous ultrafil-
tration method. Int. J. Pharm. 2006, 10, 8–112.
(22) Brunton, L. L., Lazo, J. S., Parker, K. In Goodman & Gilman’s The
Pharmacological Basis of Therapeutics, 11th ed.; McGraw-Hill: New
York, 2005; pp 1851 and 1876.
(23) Matte, C.; Maion, G.; Mourad, W.; Olivier, M. Leishmania donoVani-
induced macrophages cyclooxygenase-2 and prostaglandin E2 syn-
thesis. Parasite Immunol. 2001, 23, 177–184.
(24) Gantt, K. R.; Schultz-Cherry, S.; Rodriguez, N.; Jeronimo, S. M.;
Nascimento, E. T.; Goldman, T. L.; Recker, T. J.; Miller, M. A.;
Wilson, M. E. Activation of TGF-ꢀ by Leishmania chagasi: importance
for parasite survival in macrophages. J. Immunol. 2003, 170, 2613–
2620.
(25) Olivier, M.; Gregory, D. J.; Forget, G. Subversion mechanisms by
which Leishmania parasites can escape the host immune response: a
signaling point of view. Clin. Microbiol. ReV. 2005, 8, 293–305.
(26) Moll, H.; Scharner, A.; Ka¨mpgen, E. Increased interleukin 4 (IL-4)
receptor expression and IL-4-induced decrease in IL-12 production
by Langerhans cells infected with Leishmania major. Infect. Immun.
2002, 70, 1627–1630.
(27) Berberich, C.; Ram´ırez-Pineda, J. R.; Hambrecht, C.; Alber, G.; Skeiky,
Y. A. W.; Moll, H. Dendritic cell (DC)-based protection against an
intracellular pathogen is dependent upon DC-derived IL-12 and can
be induced by molecularly defined antigens. J. Immunol. 2003, 170,
3171–3179.
(28) Omer, F. M.; Kurtzhals, J. A.; Riley, E. M. Maintaining the
immunological balance in parasitic infections: a role for TGF-ꢀ.
Parasitol. Today 2000, 16, 358–359.
(29) Murphy, M. L.; Wille, U.; Villegas, E. N.; Hunter, C. A.; Farrell,
J. P. IL-10 mediates susceptibility to Leishmania donoVani infection.
Eur. J. Immunol. 2001, 31, 2848–2856.
(30) Awasthi, A.; Mathur, R. K.; Saha, B. Immune response to Leishmania
infection. Indian J. Med. Res 2004, 119, 238–258.
(31) Padigel, U. M.; Farrell, J. P. Control of infection with Leishmania
major in susceptible BALB/c mice lacking the common gamma-chain
for FcR is associated with reduced production of IL-10 and TGF-ꢀ
by parasitized cells. J. Immunol. 2005, 174, 6340–6345.
(32) Ramos, H.; Valdivieso, E.; Gamargo, M.; Dagger, F.; Cohen, B. E.
Amphotericin B kills unicellular leishmanias by forming aqueous pores
permeable to small cations and anions. J. Membr. Biol. 1996, 152,
65–75.
(33) Lee, N.; Bertholet, S.; Debrabant, A.; Mu¨ller, J.; Duncan, R.; Nakhasi,
H. L. Programmed cell death in the unicellular protozoan parasite
Leishmania. Cell. Death Differ. 2002, 9, 53–64.
References
(1) Ponte-Sucre, A. Physiological consequences of drug resistance in
Leishmania and their relevance for chemotherapy. Kinetop. Biol. Dis.
2 2003, 14, http://www.kinetoplastids.com/home/.
(2) Ouellette, M.; Drummelsmith, J.; Papadopoulou, B. Leishmaniasis:
drugs in the clinic, resistance and new developments. Drug Resist.
Updates 2004, 7, 257–266.
(3) Croft, S. L.; Sundar, S.; Fairlamb, A. H. Drug resistance in leishma-
niasis. Clin. Microbiol. ReV. 2006, 19, 111–126.
(4) Bringmann, G., Gu¨nther, C., Ochse, M., Schupp, O., Tasler, S. Biaryls
in Nature: A Multi-facetted Class of Stereochemically, Biosyntheti-
cally, and Pharmacologically Intriguing Secondary Metabolites. In
Progress in the Chemistry of Organic Natural Products, Herz, W.,
Falk, H., Kirby, G. W., Moore, R. E., Eds.; Springer-Verlag: New
York, 2001; Vol. 82, pp 1-293.
(5) Franc¸ois, G.; Bringmann, G.; Dochez, C.; Schneider, C.; Timperman,
G.; Ake´ Assi, L. Activities of extracts and naphthylisoquinoline
alkaloids from Triphyophyllum peltatum, Ancistrocladus abbreViatus,
and Ancistrocladus barteri against Plasmodium berghei (Anka strain)
in vitro. J. Ethnopharmacol 1995, 46, 115–120.
(6) Franc¸ois, G.; Timperman, G.; Holenz, J.; Ake´ Assi, L.; Geuder, T.;
Maes, L.; Dubois, J.; Hanocq, M.; Bringmann, G. Naphthylisoquinoline
alkaloids exhibit strong growth-inhibiting activities against Plasmodium
falciparum and P. berghei in vitro s structure-activity relationships
of dioncophylline C. Ann. Trop. Med. Parasitol. 1996, 90, 115–123.
(7) Franc¸ois, G.; Timperman, G.; Eling, W.; Ake´ Assi, L.; Holenz, J.;
Bringmann, G. Naphthylisoquinoline alkaloids against malaria: evalu-
ation of the curative potentials of dioncophylline C and dioncopeltine
A against Plasmodium berghei in vivo. Antimicrob. Agents Chemother.
1997, 41, 2533–2539.
(8) Bringmann, G.; Ho¨rr, V.; Holzgrabe, U.; Stich, A. Antitrypanosomal
naphthylisoquinoline alkaloids and related compounds. Pharmazie
2003, 58, 343–346.
(9) Bringmann, G.; Hamm, A.; Gu¨nther, C.; Michel, M.; Brun, R.;
Mudogo, V. Ancistroealaines A and B, two new bioactive naphthyl-
isoquinolines, and related naphthoic acids from Ancistrocladus ealaen-
sis. J. Nat. Prod. 2000, 63, 1465–1470.
(10) Bringmann, G.; Messer, K.; Brun, R.; Mudogo, V. Ancistrocongolines
A-D, new naphthylisoquinoline alkaloids from Ancistrocladus con-
golensis. J. Nat. Prod 2002, 65, 1096–1101.
(11) Bringmann, G.; Dreyer, M.; Faber, J. H.; Dalsgaard, P. W.; Stærk,
D.; Jaroszewski, J.; Ndangalasi, H.; Mbago, F.; Brun, R.; Reichert,
M.; Maksimenka, K.; Christensen, S. B. Ancistrotanzanine A, the first
5,3′-coupled naphthylisoquinoline alkaloid, and two further, 5,8′-linked
related compounds from the newly described species Ancistrocladus
tanzaniensis. J. Nat. Prod 2003, 66, 1159–11165.
(12) Yang, L.-K.; Glover, R. P.; Yonganathan, K.; Sarnaik, J. P.; Godbole,
A. J.; Soejarto, D. D.; Buss, A. D.; Butler, M. S. Ancisheynine, a
novel naphthylisoquinolinium alkaloid from Ancistrocladus heyneanus.
Tetrahedron Lett. 2003, 44, 5827–5829.
(13) Bringmann, G.; Kajahn, I.; Pedersen, S. E. H.; Reichert, M.; Faber,
J. H.; Gulder, T.; Brun, R.; Christensen, S. B.; Ponte-Sucre, A.; Moll,
H.; Heubl, G.; Mudogo, V. Ancistrocladinium A and B, the first N,C-
coupled naphthyldihydroisoquinoline alkaloids from a Congolese
Ancistrocladus species. J. Org. Chem. 2006, 71, 9348–9356.
(14) Ponte-Sucre, A.; Faber, J. H.; Gulder, T.; Kajahn, I.; Pedersen, S. E. H.;
Schultheis, M.; Bringmann, G.; Moll, H. Activity of naphthyliso-
quinoline alkaloids and synthetic analogs against Leishmania major.
Antimicrob. Agents Chemother. 2007, 51, 188–194.
(15) Fivelman, Q. L.; Adagu, I. S.; Warhurst, D. C. Modified fixed-ratio
isobologram method for studying in vitro interactions between
atovaquone and proguanil or dihydroartemisinin against drug-resistant
strains of Plasmodium falciparum. Antimicrob. Agents Chemother.
2004, 48, 4097–4102.
(16) Bringmann, G.; Jansen, J. R. Einfache Synthesen nu¨tzlicher Diketo-
Bausteine fu¨r biomimetische Isochinolin- und Naphthalin-Synthesen.
Liebigs Ann. Chem. 1985, 2116–2125.
(17) Seifert, K.; Croft, S. L. In vitro and in vivo interactions between
miltefosine and other antileishmanial drugs. Antimicrob. Agents
Chemother. 2006, 50, 73–79.
(18) Unger, M.; Frank, A. Simultaneous determination of the inhibitory
potency of herbal extracts on the activity of six major cytochrome
P450 enzymes using liquid chromatography/mass spectrometry and
automated online extraction. Rapid Commun. Mass Spectrom. 2004,
18, 2273–2281.
(19) Mikus, J.; Steverding, D. A simple colorimetric method to screen drug
cytotoxicity against Leishmania using the dye Alamar Blue. Parasitol.
Int. 2000, 48, 265–259.
(34) Lewis, D. F. Human cytochromes P450 associated with the phase 1
metabolism of drugs and other xenobiotics: a compilation of substrates
and inhibitors of the CYP1, CYP2, and CYP3 families. Curr. Med.
Chem. 2003, 10, 955–972.
(35) Eichelbaum, M.; Ingelman-Sundberg, M.; Evans, W. E. Pharmaco-
genomics and individualized drug therapy. Annu. ReV. Med. 2006,
57, 119–137.
(36) Schwedhelm, K. F.; Horstmann, M.; Faber, J. H.; Reichert, Y.;
Bringmann, G.; Faber, C. The novel antimalarial compound dionco-
phylline C forms a complex with heme in solution. Chem. Med. Chem.
2007, 2, 541–548.
(37) Huber, W.; Koella, J. C. A comparison of three methods of estimating
EC50 in studies of drug resistance of malaria parasites. Acta Trop.
1993, 55, 257–261.
(38) Ponte-Sucre, A.; Campos, Y.; Ferna´ndez, M.; Moll, H.; Mendoza-
Leo´n, A. Growth and survival of Leishmania sp. are impaired by ion
channel blockers. Exp. Parasitol. 1998, 88, 11–16.
(39) Ponte-Sucre, A.; Vicik, R.; Schultheis, M.; Schirmeister, T.; Moll, H.
Aziridine-2,3-dicarboxylates: peptidomimetic cysteine protease inhibi-
tors with antileishmanial activity. Antimicrob. Agents Chemother. 2006,
50, 2439–2447.
(40) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition constant
(Ki) and the concentration of inhibitor which causes 50% inhibition
(IC50) of an enzymatic reaction. Biochem. Pharmacol. 1973, 22,
3099–3108.
JM801084U