Phenotype Screening of an Azole-bisindole Chemical Library Identifies URB1483 as a New Antileishmanial Agent Devoid of Toxicity on Human Cells.

We report the evaluation of a small library of azole-bisindoles for their antileishmanial potential, in terms of efficacy on Leishmania infantum promastigotes and intracellular amastigotes. Nine compounds showed good activity on L. infantum MHOM/TN/80/IPT1 promastigotes with IC50 values ranging from 4 to 10 μM. These active compounds were also tested on human (THP-1, HEPG2, HaCaT, and human primary fibroblasts) and canine (DH82) cell lines. URB1483 was selected as the best compound, with no quantifiable cytotoxicity in mammalian cells, to test the efficacy on intracellular amastigotes. URB1483 significantly reduced the infection index of both human and canine macrophages with an effect comparable to the clinically used drug pentamidine. URB1483 emerges as a new anti-infective agent with remarkable antileishmanial activity and no cytotoxic effects on human and canine cells.

Introduction

Leishmaniasis is a neglected disease caused by protozoan parasites transmitted by phlebotomine sandflies. More than 20 different Leishmania species all over the world cause a variety of clinical conditions broadly grouped in cutaneous (CL), mucosal (ML), and visceral leishmaniasis (VL).[1] The latter, being fatal if untreated, causes 20 000–40 000 deaths across the globe each year.[2,3] More than a million new cases are reported per year and 350 million people are at risk of contracting the infection.[1] Italy is an endemic country, with an increase in cases in the last two decades due to said disease spreading within traditionally endemic regions and to the appearance of autochthonous cases in previously nonendemic areas such as northern continental Italy.[2,4] The most widespread form of leishmaniasis endemic in southern Europe is zoonotic VL, involving humans and domestic dogs (which may serve as the main reservoir), sometimes associated with a few cases of CL. Both diseases are caused by L. infantum.[4]

Although in recent years three vaccine candidates have undergone clinical trials, and half a dozen more are in the pipeline, no efficacious vaccine against leishmaniasis is now available on the market.[5] Moreover, all currently available drugs are inadequate[1] (Figure ). Pentavalent antimonials (one example is meglumine antimoniate, Figure ), the first-line treatment, could not be efficacious due to widespread resistance to the drug. A few treatments against leishmaniasis have been introduced during the last decade, each of them having serious limitations. Amphotericin B, one of the most used second-line drugs, is effective against antimonial-resistant Leishmania strains, but it can induce acute and chronic toxicity. The Amphotericin B liposomal formulation (AmBisome) ameliorates the toxicity profile; yet its high formulation cost has limited its use.[6] Miltefosine is a highly potent oral drug against VL and CL; however, its use is limited due to high cost, teratogenicity, and long treatment. Paromomycin is the cheapest treatment, but its use has been associated with severe toxicity. Moreover, this drug requires parenteral administration, needs long treatment, and shows region-dependent efficacy. On the other hand, pentamidine is used in combination therapies at low dosages but shows several shortcomings (renal toxicity, myocarditis, diabetes mellitus, hypoglycemia, hypotension, and fever).[1]

Figure 1

Main commercial drugs against Leishmaniosis.

Due to the serious drawbacks of the available therapeutic options, shorter courses and inexpensive drugs, less toxic and more efficacious across all endemic regions, are urgently needed. Many research groups around the world have developed several new scaffolds as potential drugs against leishmaniasis.[1,7−10] Among them, a few bisindoles were reported (Figure ).

Figure 2

Potential antileishmanial agents sharing bis(indolyl) fragment.

The seminal work of Roy and his collaborators reported that 3,3′-diindolylmethane (DIM, Figure ) is an effective inhibitor of L. donovani topoisomerase IB.[11] More specifically, the prototypical member of the class, DIM, was reported to act as poison for topoisomerase IB, i.e., similar to the well-known drug camptothecin (CPT), it stabilizes the topoisomerase–DNA cleavage complex, thus blocking the relaxation process.[11,12] In subsequent work, the same authors reported three new DIM derivatives that were active against a DIM-resistant strain of L. donovani. In Figure , 2,2′-diphenyl-3,3′-diindolylmethane (DPDIM) is shown as the most promising example of the reported DIMs.

Moreover, the methylene linker of DIM represents an ideal point of chemical diversification that can be exploited to generate a library of compounds useful for phenotype screening. This strategy has been shown to be affordable and practical also by other authors[13−15] who used a six-membered aromatic ring as an additional side arm (i.e., compounds 4 and 6) or a spacer connecting the two indolyl moieties (compound 5) in different sets of DIM analogues obtaining new potent antileishmanial agents.

In detail, Bharate and co-workers reported in 2013 a new aryl-DIM potent class of antileishmanial agents. Among them, compound 4 (Figure ) was the most effective against L. donovani promastigotes and amastigotes, showing IC50 values lower than 10 μM.[13] However, they did not propose a biological target of the new aryl-DIMs.

Bis(indolyl)-pyridine 5 (Figure ) is the most interesting compound reported by Kalam Khan et al. against Leishmania parasites,[14] but it is less potent than previously described DIMs and no information about its mechanism of action is available. Very recently, the group of Taha has reported several phenyl-aminothiazole-DIM derivatives as potent antileishmanial agents.[15] In particular, compound 6 (Figure ) has shown an outstanding effect on the protozoa with a submicromolar IC50 value. The authors claimed that this new class of compounds may inhibit the pteridine reductase. Yet, no inhibition studies on the isolate enzyme were reported.

As part of our ongoing investigations on the biological activities and applications of bisindole derivatives[16−20] and leishmaniasis,[21−24] in the present study, a focused library of selected published azole-bisindole derivatives 1–3 (Schemes and 2)[20,25] sharing the bis(indolyl) motif with the above-mentioned antileishmanial agents and presenting an additional azole side arm was screened against several human and canine L. infantum promastigote strains. The most active compounds (IC50 < 20 μM) were then tested on human and canine macrophages, as well as other human cell lines to check for their potential toxicity. Pyrrole-bisindole 1b, named URB1483, was found to be the most specific compound against parasites and was tested for its efficacy on L. infantum-infected human and canine macrophage-like cell lines.

Scheme 1

Reagents and Conditions: (i) Methyl or Ethyl (R5) Propiolate, DCM, at Room Temperature (rt), Overnight; (ii) 1,2-Diaza-1,3-diene 8, Toluene, Reflux, 2 h; (iii) TFA, Reflux, 2–4 h; (iv) 1,2-Diaza-1,3-diene 8, ACN, rt, 1 h; (v) Paraformaldehyde, Reflux, 4 h

Scheme 2

Reagents and Conditions: (i) DIPEA, Ethanol, 9 in Acetonitrile, 0 °C to rt, 2 h

Results

Chemistry

Pyrrole-bisindoles 1a–d and imidazole-bisindoles 2a–f were synthesized according to previously reported methods[19] starting from the opportune bisindoles 7a–g and 1,2-diaza-1,3-diene 8a–c[16,20,25] (Scheme ).

The treatment of bisindole 7 with the opportune propiolate in dichloromethane (DCM) followed by the addition of 1,2-diaza-1,3-diene 8 in toluene and in the presence of trifluoroacetic acid (TFA) at reflux furnished the corresponding pyrrole-indole 1. The mechanism of this sequential three-component reaction involves the preliminary formation of enamine intermediate, subsequent Michael addition to azoene system, and final intramolecular heterocyclization with loss of the carbazate residue.

On the other hand, the indole-imidazole scaffold 2 was synthesized by a conjugate addition of bisindole 7 to 8 in acetonitrile (ACN) at room temperature (rt), followed by the condensation with the paraformaldehyde at reflux. For this three-component reaction, a thermal-assisted 1,5-electrocyclization of 1,2-diaza-1,3-diene-derived azavinyl azomethine ylides appears to be operative.[20]

Triazole-bisindoles 3 were instead synthesized in a single step (Sakai’s protocol) from α,α-dichlorotosylhydrazone 9 and bisindole 7,[16,20] respectively, in the presence of N,N-diisopropylethylamine (DIPEA) as a base in an ethanol/acetonitrile mixture (Scheme ).[20]

Effect of Bisindoles on L. infantum Promastigote Viability

The in vitro antileishmanial activity evaluation of bisindole derivatives was carried out by treating L. infantum MHOM/TN/80/IPT1 promastigotes for 72 h with scalar dilution 1:2 or 2:3 (from 20 to 0.31 μM) of each molecule. As positive controls, Leishmania parasites were also treated with pentamidine, nonliposomal amphotericin B, and miltefosine. The compounds 1a,b, 2a–f, and 3b showed IC50 values between 3.7 and 10 μM. Compounds 1c, 1d, 3a, and DIM showed IC50 > 20 μM and were not taken into consideration for subsequent experiments (Table ). Next, the activity of bisindoles was evaluated on three L. infantum clinical isolates (two canines and one human), confirming the activity of all tested compounds (Table and Supporting Information).

Table 1

Potency of Azole-bisindole Derivatives on L. infantum Promastigote Strainsa

compound	L. infantum MHOM/TN/80/IPT1 IC50 (μM) (95% CI)	L. infantum canine clinical isolate 1 IC50 (μM) (95% CI)	L. infantum canine clinical isolate 2 IC50 (μM) (95% CI)	L. infantum human clinical isolate IC50 (μM) (95% CI)
1a	6.6 (6.1–7.2)	7.0 (6.5–7.5)	5.7 (5.3–6.1)	6.4 (5.4–7.5)
1b (URB1483)	3.7 (3.2–4.2)	4.1 (3.8–4.5)	3.7 (3.5–3.9)	7.2 (5.8–8.9)
1c	>20 (43%)bc	n.t.c	n.t.c	n.t.c
1d	>20 (11%)b	n.t.c	n.t.c	n.t.c
2a	10 (9.2–11)	6.8 (6.2–7.5)	4.8 (4.6–5.0)	8.1 (7.2–9.1)
2b	7.7 (7.0–8.4)	6.4 (6.0–6.8)	5.0 (4.7–5.3)	6.1 (5.3–7.0)
2c	10 (9.4–11)	8.7 (7.9–9.6)	6.5 (6.2–6.9)	8.6 (7.8–9.4)
2d	8.5 (7.8–9.2)	9.7 (9.2–10.1)	7.0 (6.5–7.6)	7.5 (6.8–8.3)
2e	4.9 (4.3–5.6)	4.8 (4.5–5.1)	3.6 (3.3–3.8)	5.5 (5.0–6.0)
2f	4.9 (4.4–5.4)	6.8 (6.2–7.4)	5.6 (5.3–6.0)	5.3 (4.9–5.7)
3a	>20 (22%)b	n.t.c	n.t.c	n.t.c
3b	8.1 (7.2–9.2)	5.7 (5.3–6.1)	5.3 (5.1–5.5)	7.4 (6.6–8.3)
DIM	>20 (10%)b	>20 (6%)b	>20 (12%)b	>20 (22%)b
Pent	2.6 (2.1–3.2)	1.7 (1.6–1.9)	1.4 (1.3–1.5)	1.5 (1.4–1.6)
Amph B	0.064 (0.054–0.075)	0.12 (0.12–0.13)	0.11 (0.11–0.12)	0.084 (0.075–0.092)
Milt	2.9 (2.5–3.3)	6.6 (6.0–7.2)	6.1 (5.6–6.5)	1.2 (1.0–1.5)

IC50 values for all of the strains are reported as mean and 95% confidence interval, from at least three independent experiments. Each experimental condition was conducted at least in duplicate.

Percentage of inhibition at 20 μM.

Not tested; Pent: Pentamidine; Amph B: nonliposomal amphotericin B; Milt: miltefosine.

Cytotoxic Effect of Bisindoles in Human and Canine Cell Lines

The cytotoxicity of the most active compounds 1a, 1b (URB1483), 2a–f, and 3b was evaluated in THP-1, DH82, HEPG2, HaCaT, and human primary fibroblasts (HPF) cells, in terms of concentrations of drug required to reduce cell viability by 50% (CC50). First, to test the viability of THP-1 and DH82 cells following bisindoles treatment, the cells were treated for 72 h with each compound at five different concentrations (2, 10, 20, 80, and 200 μM). Notably, URB1483 did not show quantifiable toxicity in both cell lines. The other compounds showed a cytotoxicity >20 μM in both cell lines, with the exception of compounds 2d (15.4 and 18.9 μM in THP-1 and DH82 cells, respectively) and 1a (11.1 μM in DH82 cells) (Table and Supporting Information).

Table 2

CC50 Values of Azole-bisindole Derivatives on THP-1 and DH82 Cells

	THP-1		DH82
compund	CC50 (μM) (95% CI)	selectivity index (CC50/IC50a)	CC50 (μM) (95% CI)	selectivity index (CC50/IC50a)
1a	124 (91–193)	18.8	11.1 (8.1–14.0)	1.7
1b (URB1483)	>200 (2%)b	>55	>200 (15%)b	>55
2a	37.9 (25.2–58.6)	3.8	35.3 (29.3–42.2)	3.5
2b	26.9 (16.7–36.9)	3.5	24.8 (21.0–29.0)	3.2
2c	38.6 (26.4–59.1)	3.8	23.5 (20.0–28.9)	2.3
2d	15.4 (12.8–19.0)	1.8	18.9 (15.4–24.0)	2.2
2e	55.4 (24.6–79.6)	11.3	24.3 (20.7–29.3)	5.0
2f	35.1 (21.5–49.9)	7.2	>200 (43%)b	>41
3b	>200 (33%)b	>25	31.9 (22.6–45.5)	3.9
Pent	103 (73–179)	40.3	10.6 (7.9–12.9)	4.1
Amph B	2.8 (1.5–3.8)	45.8	>200 (42%)b	>3333
Milt	27.8 (22.5–35.5)	9.8	95.9 (86.2–109)	33.6

Selectivity index calculated considering IC50 on L. infantum MHOM/TN/80/IPT1.

Percentage of inhibition at 200 μM; Pent: Pentamidine; Amph B: nonliposomal amphotericin B. Milt: miltefosine.

Next, CC50 was evaluated in DH82, HEPG2, and HaCaT cell lines and HPF cells after 24 h treatment with serially diluted bisindole compounds (2, 10, 20 μM). In this case, incubations were carried out for 24 h to evaluate the cytotoxicity in actively dividing cells. In these experiments, all treatments with bisindole derivatives compounds did not show cytotoxic effects at the highest dose used (20 μM), and therefore, it was not possible to calculate the CC50 values. Interestingly, in these cell lines, only the reference compound nonliposomal amphotericin B showed cytotoxicity after 24 h treatment between 4.6 and 17.3 μM (Table S1, Supporting Information).

Efficacy of Compound URB1483 on L. infantum Intracellular Amastigotes

Based on the analysis of the IC50 and CC50 obtained on L. infantum promastigotes and THP-1 or DH82 cells, respectively, URB1483 was found to be the most effective and yet selective compound against the parasites and, therefore, it was selected for further experiments.

The infection of human monocytic THP-1 cell line was conducted with L. infantum MHOM/TN/80/IPT1 as described in the Materials and Methods section. Infected macrophages were treated with compound URB1483 or pentamidine (used as the positive control) for 72 h. The infection index was significantly reduced following the treatment with URB1483, in a dose-dependent manner (one-way ANOVA p < 0.001) (Figure a), evocative of specific mechanism of action.

Figure 3

Effect of URB1483 and Pent on intracellular L. infantum amastigotes. (A) THP-1 cells infected with L. infantum MHOM/TN/80/IPT1; (B) THP-1 cells infected with L. infantum human clinical isolate; (C) DH82 cells infected with L. infantum MHOM/TN/80/IPT1. In all cases, cells were infected for 24 h at 37 °C; the drugs were added, and the infection index was calculated after 72 h of treatment. Data are expressed as mean ± SEM of three independent experiments. Each experimental condition was conducted at least in duplicate. *p < 0.05, **p < 0.01, ***p < 0.001.

Interestingly, the percentage of infected cells and the average number of amastigotes per infected cell decrease in a dose-dependent manner, in both URB1483 and pentamidine treatments. The THP-1 cells were also infected with the L. infantum human clinical isolate. In this case, the parasite resulted to be less susceptible to both URB1483 and Pent, compared to the reference strain MHOM/TN/80/IPT1 (Figure b).

The same infection and treatment protocols were also performed with the DH82 cell line. In DH82 cells infected with MHOM/TN/80/IPT1 strain, the infection index decreased significantly at the higher dose (20 μM) in both URB1483 (p < 0.01) and pentamidine (p < 0.05) treated cells (Figure c). It is noteworthy that Pent had a cytotoxic effect on DH82 cells after 72 h treatment (Table ) (not observed after 24 h treatment), in contrast to URB1483, which did not significantly affect the cell viability (Figure b).

Figure 4

Cytotoxic effect of treatment with pentamidine (Pent), in contrast to URB1483, on DH82 cells after 72 h treatment. (A) Uninfected DH82 cells were treated with URB1483 or Pent for 72h at concentrations of 0, 2, 10, and 20 μM and stained with Hoechst dye for fluorescence microscope observation. (B) Effect of treatment was monitored by calculating the number of cells, considering at least five images per treatment condition. Data are mean ± SEM; ***p < 0.001.

Two-way ANOVA followed by Bonferroni post hoc test did not show significant differences in infection indexes between treatments with compound URB1483 and pentamidine in all experiments (see Supporting Information Table S2), accounting for similar effects of the two molecules.

L. donovani Topoisomerase IB as Potential Target

Prompted by the encouraging results obtained on both L. infantum promastigote and L. infantum intracellular amastigotes, we searched for molecular target responsible for the antileishmanial activity displayed by compound URB1483. Literature data point to topoisomerase IB as the most likely target for bisindole compounds.[11,12] The antileishmanial activity displayed by the bis-indolyl derivative DIM has been linked to the ability of this compound to block DNA relaxation with a mechanism similar to that of topotecan,[26] i.e., stabilizing the formation of a ternary complex composed of the inhibitor itself, the leishmanial enzyme topoisomerase IB, and double-strand DNA. We thus performed molecular modeling investigations assuming that URB1483 could interact with the topoisomerase IB–DNA complex by targeting the same binding site recognized by topotecan. First, a three-dimensional model of L. donovani topoisomerase I bound to DNA-topotecan complex was built using available structural information in the PDB, i.e., the human form of topoisomerase I bound to DNA and topotecan (PDB ID 1K4T), and the L. donovani form of topoisomerase I bound to nicked DNA (PDB ID 2BS9), by following the computational protocol reported by Roy et al. (see the Materials and Methods section for details).[12] This strategy appeared reasonable as the comparison between these two topoisomerase structures reveals that, despite a diverse architecture (monomeric the human, dimeric the L. donovani isoform), all of the amino acid residues that line the topotecan-binding pocket are entirely conserved between the two species.[27]

The resulting model of L. donovani topoisomerase IB bound to topotecan was employed to identify the binding mode for URB1483 that could account for its specific mechanism of action. Docking simulations point to a pose for URB1483 to some extent resembling the one experimentally observed for topotecan (Figure ), with one indolyl fragment well superposed on the A-ring of topotecan and one terminal carboxylic acid ethyl ester installed on the pyrrole nucleus occupying nearly the same space of the lactone E-ring of topotecan. The second indolyl fragment of URB1483 protruded in a broad cavity of topoisomerase IB, normally occupied by solvent bulk. This additional cavity is targeted by other classes of topoisomerase poisons such as indolocarbazole and indenoisoquinoline derivatives,[28] which place a bulky group in this region (Figure S29, Supporting Information).

Figure 5

Molecular model of URB1483 (green carbons) docked in the CPT binding site present in the topoisomerase IB–DNA complex (dark gray carbon atoms). The structure of topotecan (pink carbon atoms) is also displayed for comparison. The secondary structure of topoisomerase IB is represented by red (α-helices) or cyan (β-strands) cartoons, while black cartoons represent the secondary structure of the DNA (namely, 22-base pair duplex oligonucleotide).

Prompted by these computational results, the inhibitory activity of URB1483 was assessed on L. donovani topoisomerase IB by means of a plasmid relaxation assay (Figure S30, Supporting Information). This assay detects the different electrophoretic mobility of the DNA supercoiled plasmid converted, by the enzyme, to its relaxed forms, in the presence of increasing concentrations of the compound. Gathered data indicate that URB1483 does not inhibit DNA relaxation at all of the tested concentrations (up to 50 μM).

Therefore, further investigations, not necessarily based on the structural similarity between DIM and URB1483, to identify a molecular target accounting for the anti-Leishmania activity of pyrrole-bisindole derivatives will be needed.

Stability Studies of URB1483

Chemical stability data were measured by LC/MS approaches for URB1483 to preliminary assess its in vitro PK profile. Gratifyingly, our analyses revealed that URB1483 exhibited favorable chemical stability both in 0.90% w/v of NaCl water solution and in cellular medium. Under both conditions, the chemical stability was almost quantitative with more than 90% of URB1483 remaining unaltered after 72 h.

Discussion

Leishmaniasis is one of the most dangerous neglected tropical diseases, second only to malaria in parasitic causes of death.[5] Caused by global warming, the endemic regions of leishmaniasis are continuously spreading to nontropical areas, including Europe. So far, no vaccine against leishmaniasis is on the market. Approved drugs in second- and third-line treatments are currently limited and/or exorbitantly priced (i.e., amphothericin B; paromomycin; miltefosine; pentamidine) with a few of them being effective on antimonial-resistant Leishmania strains. Moreover, the use of these agents on infected patients is seriously hampered by the insurgence of acute and/or chronic toxicity. Therefore, there is an urgent need for developing safe, effective, and affordable drugs for the treatment of leishmaniasis. To fill this therapeutic gap, several research groups introduced new classes of active compounds against leishmaniasis, including bisindole derivatives.

In this context, phenotypic screening of our small library of previously reported azole-bisindoles 1–3 (Schemes and 2)[20,25] against four human and canine L. infantum strains was performed. With a few exceptions, all bisindoles, belonging to three different classes (pyrroles 1, imidazoles 2, and triazoles 3), showed good activity against the analyzed L. infantum promastigotes with IC50 values lower than 10 μM (Table ).

Conversely, pyrrole-bisindole 1c is less active than the other compounds, and 1d and triazole-bisindole 3a are nearly inactive. At this stage, it seems that an additional nitrogen-containing aryl substituent could increase the potency of the simple bisindole scaffold (DIM was almost inactive at 20 μM). In the pyrrole class 1, a double substitution of the indole rings is detrimental for their activity (compounds 1c and 1d showed low activity). All of the imidazole derivatives 2 are active in all four L. infantum promastigote strains, and in particular, 2e and 2f, having N-methylated indoles, are very potent. The methylation of the indole rings seems to play an important role also in the triazole-bisindole derivatives 3 (3a is inactive and 3b showed IC50 < 8 μM).

The most active azole-bisindoles 1a, URB1483, 2a–f, and 3b were then tested on human and canine macrophage-like cell lines, as well as other human cell lines to check for their potential toxicity. In general, the imidazole-bisindole derivatives 2 are more toxic than pyrroles 1 and triazole 3b in all of the tested cell lines. Pyrrole-bisindole URB1483 showed good activity against parasites and the best selectivity index (SI) on both human (THP-1) and canine (DH82) cells (SI > 55, Table ). Moreover, URB1483 did not show any appreciable toxicity against human hepatocytes, keratinocytes, and primary fibroblasts (Table S1, Supporting Information).

Although URB1483 shows high lipophilicity, this feature did not hamper its ability to inhibit leshmanial growth while sparing human cells. For this reason, it was chosen for further studies on other cell models.

Therefore, URB1483 was tested, in comparison with pentamidine, for its efficacy on L. infantum-infected human and canine macrophage-like cell lines. The efficacy of compound URB1483 and pentamidine was not significantly different in human and canine in vitro infection models, despite the fact that pentamidine showed lower IC50 compared to URB1483 in promastigotes (Table ). This may account for the better bioavailability of URB1483 inside the infected cells. Concerning the human infection model (THP-1 cells), the human clinical isolate appeared less susceptible to both drug treatments compared to the reference strain MHOM/TN/80/IPT1. This probably reflects its higher virulence (infection index was twice) and/or its lower susceptibility to treatment with compound URB1483 (IC50 value was about 2 times higher in the human clinical isolate compared to other strains) (Table ). As regards the canine infection model, it is noteworthy that pentamidine showed high toxicity in DH82 cells at 72 h (Table and Figure ), while compound URB1483 did not show any appreciable toxicity, underlying its potential use also in veterinary applications.

According to the literature, topoisomerase IB was identified as a likely biological target for this class of bisindoles.[11,12] Furthermore, molecular modeling investigations on topoisomerase IB point to a likely binding mode for URB1483 similar to the one experimentally observed for topotecan. Therefore, URB1483 was evaluated for its ability to inhibit on topoisomerase IB from L. donovani, which shared 98% sequence identity with L. infantum isoform. Unexpectedly, URB1483 failed to inhibit L. donovani topoisomerase IB activity, indicating that other targets are likely engaged by this pyrrole-bisindole. Efforts will be taken to discover the target of URB1483 in our ongoing research on antileishmanial agents.

Conclusions

Phenotypic screening of a small library of azole-bisindoles against several human and canine L. infantum strains was performed. Most of the tested compounds showed good activity against the promastigotes (IC50 values < 10 μM). URB1483, a pyrrole-bisindole derivative, showed good activity against parasites, and it did not affect the viability of canine and human cell lines, with a selectivity index >55. Moreover, the efficacy of URB1483 on human and canine in vitro infection models was comparable to that of the commercial drug pentamidine. Previous works on the bisindole prototype DIM demonstrated that its antileishmanial activity is due to its ability to block the DNA relaxation activity of Leishmania topoisomerase IB. This evidence along with computational studies supports the hypothesis that URB1483 should have worked as a topoisomerase IB inhibitor. Biochemical studies on the isolated enzyme ruled out the inhibition of topoisomerase IB as a mechanism of action for URB1483. Even if the search for the biological target is still ongoing, URB1483 may undoubtedly represent a promising lead compound for the generation of new anti-Leishmania agents with low toxicity on host cells.

Experimental Section

Materials and Methods

Chemistry

All organic solvents used in this study were purchased from Sigma-Aldrich (St. Louis, MO), Alfa Aesar (Haverhill, MA), or TCI (Tokyo, Japan). In particular, the antileishmanial drugs pentamidine isethionate salt, nonliposomal amphotericin B, and miltefosine, used as positive controls, were purchased from Sigma-Aldrich. Prior to use, acetonitrile, dichloromethane, and toluene were dried with molecular sieves with an effective pore diameter of 4 Å. Column chromatography purifications were performed under “flash” conditions using Merck (Darmstadt, Germany) 230–400 mesh silica gel. Analytical thin-layer chromatography (TLC) was carried out on Merck silica gel plates (silica gel 60 F254), which were visualized by exposure to ultraviolet light and an aqueous solution of cerium ammonium molybdate (CAM). Melting points were determined by Buchi (Gallen, Switzerland) B-540. 1H NMR and 13C NMR spectra were recorded on a Bruker (Billerica, MA) AC 400 and 100 spectrometer, respectively, and analyzed using the TopSpin 1.3 (2013) software package. Chemical shifts were measured using the central peak of the solvent. EI-MS and ESI-MS spectra were recorded with a Shimadzu (Kyoto, Japan) QP-5000 mass spectrometer and with a Waters (Milford, MA) Micromass ZQ spectrometer, respectively. The final compounds were analyzed on a ThermoQuest (Italia) FlashEA 1112 elemental analyzer for C, H, and N. The percentages found were within ±0.5% of the theoretical values. All of the tested compounds were >95% pure as determined by elemental analysis.

General Procedure for the Synthesis of Pyrrole-bisindole Derivatives 1

A mixture of the appropriate bisindole 7(16) (0.4 mmol) and methyl or ethyl propiolate (0.44 mmol) was stirred in DCM (1 mL) overnight at room temperature. A solution of the opportune azoalkene 8 (0.6 mmol) in toluene (4 mL) was added, and the reaction was refluxed for 2 h. A catalytic amount of TFA (two drops) was added, and the reaction was refluxed for an additional 2–4 h (TLC check). After removal of the solvent, the crude mixture was purified by column chromatography on silica gel to afford the corresponding pyrrole-bisindole 1.

The physicochemical data of compounds 1a–d, with purity >95% (determined by elemental analysis), are reported in the Supporting Information, and they are in agreement with those reported in the literature.[20]

General Procedure for the Synthesis of Imidazole-bisindole Derivatives 2

To a stirred solution of the appropriate bisindole 7(16) (0.4 mmol) in ACN (2 mL), the opportune azoalkene 8 (0.4 mmol) was added at room temperature. After the disappearance of the reagents, checked by TLC analysis (usually 1 h), paraformaldehyde (0.8 mmol) was added, and then the resulting mixture was refluxed for 4 h (TLC check). The solvent was evaporated under reduced pressure, and the crude residue was purified by column chromatography to give the corresponding imidazole-bisindole derivatives 2.

The physicochemical data of compounds 2a–d[20] and 2e–f,[25] with purity >95% (determined by elemental analysis), are reported in the Supporting Information, and they are in agreement with those reported in the literature.

General Procedure for the Synthesis of Triazole-bisindole Derivatives 3

To a cooled solution (0 °C) of the appropriate bisindole 7(16) (0.4 mmol) in ethanol (5 mL) was added DIPEA (2.4 mmol, 6 equiv). The solution was stirred for 10 min, after which hydrazine 9 (0.52 mmol, 1.3 equiv) dissolved in acetonitrile (4 mL) was added dropwise to the cooled solution, and stirring was continued at room temperature for 2 h (TLC check). After completion of the reaction, all volatiles were removed under reduced pressure and the residue was purified by column chromatography to give the corresponding triazole-bisindole 3.

The physicochemical data of compounds 3a,b, with purity >95% (determined by elemental analysis), are reported in the Supporting Information, and they are in agreement with those reported in the literature.[20]

Parasite Cultures

The reference strain L.infantum MHOM/TN/80/IPT1 (WHO international reference strain) was purchased from ATCC (ATCC 5013). Two L. infantum strains were isolated from lymph node aspirates of two infected symptomatic dogs, obtained from the veterinary clinic Santa Teresa (Fano, Italy), as previously described.[21] Moreover, one L. infantum strain was isolated from a skin biopsy taken during a routine diagnostic process from a patient with CL and previously treated with intralesional injections of glucantime.[29] All L. infantum promastigotes were cultured in Evans’ modified Tobie’s medium (EMTM) at 26–28 °C. To test bisindole compounds, the parasites were cultivated in RPMI-PY medium as described previously.[30]

Cell Cultures

The human monocytic cell line THP-1 (ECACC 88081201) was cultured in an RPMI-1640 medium. The canine macrophage-like cell line DH82 (ATCC CRL-10389) and the human hepatocellular carcinoma cell line HepG2 [HEPG2] (ATCC HB-8065) were cultured in Eagle’s minimum essential medium (EMEM). Human keratinocyte cell line HaCaT (CLS 300493) and human primary fibroblasts (HPF) from healthy subjects obtained from upper arm skin biopsies, kindly provided by Dr. Giosuè Annibalini (University of Urbino Carlo Bo), were grown in Dulbecco’s modified Eagle’s medium (DMEM). All media were supplemented with 10% (15% for DH82 cells) heat-inactivated fetal bovine serum (FBS), 2 mM l-glutamine, 10 g/L nonessential amino acid, 1 mM sodium pyruvate, 100 μg/mL streptomycin, and 100 U/L penicillin. All cell lines were maintained in a humidified incubator at 37 °C and 5% CO2. All cell culture reagents were purchased from Sigma-Aldrich (St. Louis, MO).

L. infantum Promastigotes Viability Assay

To investigate the bisindole activity on L. infantum strains, the late log/stationary phase promastigotes were resuspended in complete RPMI-PY medium at a density of 2.5 × 106 parasites/mL in 96-well plates (100 μL/well). The promastigotes were treated with scalar dilutions 1:2 or 2:3 of the 12 bisindole compounds (from 20 to 0.31 μM) for 72 h at 26 °C. As positive controls, the antileishmanial drugs pentamidine (Sigma-Aldrich) (from 10 to 0.16 μM), nonliposomal amphotericin B (Sigma-Aldrich) (from 1 to 0.0078 μM), and miltefosine (Sigma-Aldrich) (from 20 to 0.31 μM) were included. As the negative control, parasites were treated with the vehicle (DMSO). Each condition was carried out in duplicate. To evaluate the promastigotes viability, the CellTiter 96H Aqueous Non-Radioactive Cell Proliferation Assay (Promega), based on the ability of viable cells to convert a soluble tetrazolium salt [3-(4.5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2-(4-sulfophenyl)-2H-tetrazolium. MTS] to a formazan product, was conducted. Briefly, 20  μL of MTS/PMS (phenazine methosulfate, Sigma-Aldrich) was added to 100 μL of culture medium and incubated at 26 °C until formazan production. Absorbance was recorded on a Microplate Reader (Benchmark. Bio-Rad) at 492 nm. Results were shown as % of promastigotes growth inhibition compared to control (DMSO). The IC50 values were calculated using nonlinear regression curves in GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, CA). The equation used for data fitting was Y = 100/(1 + 10∧((Log IC50 – X)*HillSlope)) (hillslope not constrained), where X is the log of concentration and Y is the normalized response.

Cytotoxicity Assay

The cytotoxicity of bisindole compounds was evaluated in THP-1, DH82, HEPG2, HaCaT, and HPF cells. THP-1 cells were resuspended at a density of 5 × 106 cells/mL; 100 μL/well were seeded in a 96-well plate and treated with 20 ng/mL phorbol myristic acid (PMA) to induce differentiation into macrophages-like cells for 48 h. DH82 cells were seeded in a 96- well plate with a density of 2 × 105 cells/well and left to attach overnight. After cell adhesion to the plate, selected bisindole compounds were used at concentrations of 2, 10, 20, 80, and 200 μM, for 72 h at 37 °C. Moreover, to test cytotoxicity in actively proliferating cells, DH82, HEPG2, HaCaT, and HPF cells were seeded in a 96-well plate with a density of 2 × 105 cells/well and left to attach overnight; afterward, the cells were treated with selected bisindole compounds at concentrations of 2, 10, and 20 μM, for 24 h at 37 °C. The negative control, (DMSO), and the antileishmanial drugs pentamidine nonliposomal amphotericin B, and miltefosine, were included in each experiment. Each condition was carried out in duplicate. To evaluate the compound cytotoxicity, the CellTiter 96H Aqueous Non-Radioactive Cell Proliferation Assay (Promega) was conducted as described above. For each compound, the selectivity index (SI) was calculated as the ratio between cytotoxicity in THP-1 and DH82 cells (CC50, 72 h) and activity against L. infantum promastigotes (IC50, 72 h).

Antiamastigote Assay on Infected Cells

The activity of URB1483 against intracellular amastigotes was evaluated in THP-1- and DH82-infected cells. Briefly, THP-1 cells were seeded in 35 mm dishes with a density of 6 × 105 cells/dish and treated with 20 ng/mL phorbol myristic acid (PMA) for 48 h to induce differentiation into macrophage-like cells. After differentiation, the cells were infected for 24 h with L. infantum MHOM/TN/80/IPT1 (or human clinical isolate) promastigotes with a parasite-to-cell ratio of 10:1. Noninternalized promastigotes were then removed and the cells were treated with URB1483 or with the positive control pentamidine at concentrations of 2, 10, and 20 μM, for 72 h. DH82 cells were seeded at a density of 2.5 × 105 cells/dish in 35 mm dishes for 24 h. The infection and treatment were performed as described above. Since the vehicle DMSO did not show any toxicity on L. infantum promastigotes or on mammalian cells, it was not included in the experiments of infection.

To monitor the infection, the cells were washed, formaldehyde/methanol fixed, stained with Hoechst dye, and observed with a fluorescence microscope. The infection index (percentage of infected macrophages × the average number of parasites per macrophage) was obtained by counting at least 300 cells for each condition.

Statistical Analysis

The evaluation of IC50 in promastigotes and CC50 in mammalian cells following bisindole treatment was performed by nonlinear regression analysis and expressed as mean and 95% confidence interval. Statistical analyses of infection indexes were performed using one-way ANOVA followed by Tukey’s multiple-comparison post hoc test and two-way ANOVA followed by Bonferroni correction for multiple comparisons. All tests were performed using GraphPad Prism version 8 (GraphPad Software, Inc., La Jolla, CA). A p value ≤ 0.05 was considered significant.

Molecular Modeling

Model Building

Differently from human topoisomerase I (hTopo I), which is produced as a monomeric enzyme composed of a single 765 residue polypeptide chain, L. donovani topoisomerase I is a heterodimeric protein composed of a large subunit (LdTOP1L) of 635 residues and a small subunit (LdTOP1S) of 262 residues. Despite this different organization, superposition of LdTOP1LS heterodimer (PDB ID 2B9S)[27] bound to DNA on the structure of hTopo I bound to DNA complexed with topotecan (PDB ID 1K4T)[26] reveals that the amino acids shaping the drug-binding cavity are conserved between the two species. Moreover, both human and L. donovani forms undertook similar interactions with the 22-bp DNA duplex oligonucleotide present in both the X-ray structures. The key difference between the two PDB complexes resides in the size of the major groove of the DNA double strand, which is slightly larger in human form, where the topotecan is accommodated. Using available structural information here summarized, a 3D model of L. donovani topoisomerase I bound to DNA-topotecan complex was built using Maestro 11.6[31] within the Schrodinger 2018–2 software, following the computational protocol reported by Roy et al.[12] In brief, after superposing the backbone atoms of the LdTOP1LS heterodimer on the backbone atoms hTopo I protein, the DNA double strand of Leishmania donovani isoform was replaced with the DNA double strand present in the hTopo I-DNA-topotecan complex. The resulting LdTOP1LS-DNA-topotecan ternary complex was submitted to a protocol of geometry optimization based on energy minimization using OPLS3e force field.[32] After deletion of topotecan from the binding site, the resulting structure was employed to perform docking simulation with Glide 7.9 software.[33]

Ligand Docking

Docking studies were performed with Glide included in the Schrodinger software package(34) starting from LdTOP1LS-DNA-topotecan ternary model described above, following a protocol successfully applied to predict the docking pose of indole-containing compounds.[35] The docking grid was centered on the position of topotecan ligand. Dimensions of enclosing and bounding boxes were set to 20 and 10 Å on each side, respectively, and van der Waals radii of protein atoms were not scaled during grid generation. The structure of URB1483 was built in Maestro and then energy-minimized in implicit solvent (water) with OPLS3e force field to an energy gradient of 0.01 kcal/(mol Å). The minimized ligands were docked within the topotecan-binding site (see above) using Glide software in Standard Precision mode with default settings. Poses were ranked according to the Gscore value.

Stability Studies of URB1483

An opportune aliquot of a stock solution of URB1483 in DMSO (10 mM) were added to a physiological solution (0.90% w/v of NaCl water solution) or EMEM or EMTM (URB1483 concentration = 50 μM) and maintained at 37 °C. At regular time points, aliquots of the described solutions were sampled, two volumes of ACN were added, and samples were centrifuged (8000 rpm, 10 min) and analyzed by HPLC-ESI-MS for a percentage of the remaining compound over incubation time.