Design , Antileishmanial Activity , and QSAR Studies of a Series of Piplartine Analogues

Laboratory of Pharmaceutical Chemistry, Universidade Federal da Paraı́ba, 58051-900 João Pessoa, PB, Brazil Department of Cellular and Molecular Biology, Universidade Federal da Paraı́ba, 58051-900 João Pessoa, PB, Brazil Department of Pharmacy, Federal University of Sergipe, 49100-000 São Cristóvão, SE, Brazil Escuela de Ciencias Fı́sicas y Matemáticas, Universidad de Las Américas, Quito, Ecuador Gonçalo Moniz Institute, Oswaldo Cruz Foundation (IGM-FIOCRUZ/BA), Salvador, Bahia 40296-710, Brazil


Introduction
Leishmaniasis is a tropical disease prevalent in 98 countries and has high impact on South America, Africa, and Asia (especially India). is illness is caused by protozoan parasites of the Leishmania genus and is transmitted to humans by the bite of infected female phlebotomine sandflies.ere are at least twenty species of Leishmania that cause clinical manifestations in humans, and they can occur in three different forms: (i) visceral leishmaniasis, (ii) cutaneous leishmaniasis, and (iii) mucocutaneous leishmaniasis.e most severe form is visceral leishmaniasis, characterized by prolonged fever, hepatomegaly, splenomegaly, substantial weight loss, and progressive anemia, and if left untreated, it can be fatal in 95% of cases [1][2][3].
e drugs currently available for treatment of leishmaniasis are pentavalent antimonials glucantime, bisamidines (pentamidine and stilbamidine), miltefosine, and amphotericin B. Unfortunately, all these drugs have several limitations, such as low efficacy, toxicity to the liver and heart, elevated cost, and parasite resistance.For these reasons, new and more efficient therapeutic approaches are needed for prevention and treatment of leishmaniasis [1].Natural products provide a virtually unlimited source of inspiration for new, powerful and selective drug leads.It is estimated that 20,000 plant species have antiparasitic activities.A wide variety of secondary metabolites have antileishmanial activity, namely, alkaloids, triterpenes, terpenoids, chalcones, saponins, glycosides, acetogenins, and avonoids [4][5][6].
Piperaceae family, also known as pepper family, comprises over 1,000 species distributed pantropically.e Piper genus is the most diverse of the Piperaceae family and produces a number of structural classes (lignans, phenylpropanoids in the form of amides, and other derivatives of phenylpropanoids) with several biological activities, including antileishmanial activity [7][8][9].In a study [7], antileishmanial activity of n-hexane, ethyl acetate, acetone, and methanol extracts from Piper cubeta fruits and Piper retrofractum stem bark was investigated against Leishmania donovani promastigotes.Among these substances, piplartine demonstrated the highest leishmanicidal activity, with an IC 50 value of 7.5 μM.Furthermore, this molecule was 3 times more potent than positive control pentamidine (IC 50 25 μM) [7,8].
Piplartine (1) (Figure 1), also known as piperlongumine, is an alkamide found in large quantities in long pepper (Piper longum L.).In addition to antileishmanial and trypanocidal activities, this alkamide has been also reported as having other pharmacological activities, including antitumor, cytotoxic, antinociceptive, antiplatelet aggregation, and antimetastatic activities [10].us, the goals of the present study were to synthesize a series of 32 piplartine analogues and evaluate the structure-activity relationship among these derivatives against Leishmania amazonensis promastigotes.
e experiments were performed against promastigote forms, that is the proliferative form found in the invertebrate host, and the tests compounds were assayed at following concentrations: 3.125, 6.25, 12.5, 25, 50, 100, 200, and 400 µg/ml.e antileishmanial activity was assessed as IC 50 and expressed in µM.IC 50 values of the piplartine analogues are summarized in Table 1.
e statistical parameters of the obtained 3D-QSAR model as well as superposition of the 3D structures from which it was derived are summarized in Figure 2. In this gure are also presented the general sca old of the compounds under study, the plot of the predicted vs. observed pIC 50 values for the training set, and the results of LOO cross-validation procedures.
e obtained statistical parameters show that the developed model is accurate and robust.is is supported by the high values of q 2 obtained during the LOO and LMO cross-validation experiments.Regarding the in uence of the steric and electrostatic factors on bioactivity, it was found that both contribute almost equally to explain the observed bioactivity.
In general, compounds containing O in the X position (see Figure 2) were more active than those containing NH at the same position.is e ect can be clearly observed for compounds 17 and 23 whose only di erence is the presence of NH (17) vs. O (23) at position X, compound 23 being 6.5fold more active than 17. is can be due to electrostatic elds surrounding this position, as shown in Figure 3(a).ese elds indicate that negatively charged substituents are favored at position X, which is the case of the compounds with O at this position.is observation can be generalized to all the compounds in the dataset.A special case is compound 29 that lacks the double bond spacer.In terms of electrostatic potential, the O atom at the X position remains close to the electronegative eld while the lack of the spacer 2 Journal of Chemistry places the trimethoxyphenyl group in a region where an electronegative MIF is present.us, compounds without the spacer will have higher bioactivity compared to those containing the double bond spacer.Finally, the electropositive potential observed around position R correlates with the favorable influence of aromatic substitutions at this position.
We also analyzed the influence of the steric molecular interaction fields.ese fields are summarized in Figure 3(b), and it can be seen that the substituents attached Cinnamic esters (R)               From Figure 3, it can be seen that compound 27 lies entirely within the steric favorable region, while compounds 10 and 21 bear substituents which protrude into the unfavorable steric region.In addition to 27, the other two most active compounds (29 and 25) are also substituted at R with groups completely falling within the steric favorable region.On the other hand, large substituents such as those present in compounds 9 and 26 reach the steric unfavorable region, which reduce their bioactivity.To avoid falling within the steric unfavorable region, if rings are used at R, no more than one methyl group should link them to X. Furthermore, nonbulky groups at R such as aliphatic substitutions lead to nonoptimal bioactivity because of their inability to completely occupy the steric favored region.
e structureactivity relationship (SAR) derived from these analyses is summarized in Figure 3(c).

Anti-Leishmania Activity of Compounds 5-36.
Pentavalent antimonials, in use for more than six decades, are still the first-line of treatment for visceral leishmaniasis, also known as kala-azar black fever.However, this therapeutic option presents several limitations, including side effects, the need for daily injections, and drug resistance, and requires a long treatment regimen.us, developing and testing of new compounds with leishmanicidal activity is of paramount importance to improve disease treatment and control.Several medicinal plant secondary metabolites have shown interesting antileishmanial activities, including alkaloids, steroids, terpenoids, phenolic acids, and phenylpropanoids [13][14][15].
e chloroform extract of Valeriana wallichii rhizomes was investigated to identify the structures responsible for its antileishmanial activity.A bioassay-guided fractionation was undertaken and resulted in (−)-bornyl caffeate, α-kessyl alcohol, two cinnamic acid derivatives, and four valtrates.All these compounds exhibited antileishmanial activity against L. major promastigotes, with IC 50 values varying between 0.8 and 48.8 µM.Furthermore, some compounds were more potent than the positive control miltefosine (IC 50 � 36.2 µM) [16].
A phytochemical study guided by the antileishmanial activity of the crude extract from the stem bark of Tecoma mollis yielded seven phenylpropanoid glycosides.ese secondary metabolites were tested against the promastigote forms of L. donovani, and pentamidine and amphotericin B were used as positive controls.e compounds acteoside, luteoside A, and luteoside B shared the highest IC 50 values: 30.08, 15.07, and 6.71 µg/ml, respectively.e IC 50 values of the positive controls were 0.34 µg/ml to amphotericin B and 1.31 µg/ml to pentamidine.
e obtained results also revealed that ortho-dihydroxy phenyl group (catechol moiety) was important to the antileishmanial activity [17].
Analysis on the structure-activity relationship revealed that, in general, an acrylate moiety on cinnamic esters appears be more important for improving antileishmanial activity when compared to the acrylamide moiety on amides.is effect can be clearly observed comparing compounds 13 (IC 50 � 0.101 ± 0.031 μM) and 29 (IC 50 � 0.662 ± 0.188 μM), whose only difference is the replacement of the OH group 13 by NH 29 at this position, being compound 13 6.6-fold more active than 29.
e ester 20 (IC 50 � 0.007 ± 0.008 µM) containing the monoterpenic substructure bornyl attached on the (E)-3-(3,4,5-trimethoxyphenyl)-acryloyl moiety yielded the best IC 50 value, near the positive control amphotericin B (IC 50 � 0.0015 µM). is result demonstrates the importance of this radical on antileishmanial activity and can be a prototype for the planning of new derivatives with higher antileishmanial activities.In an earlier study, a series of compounds based on caffeic acid bornyl ester was synthesized and tested in vitro against L. major, L. donovani promastigotes, and L. major amastigotes to investigate the SAR.
e introduction of methyl groups contributes to increase lipophilicity.It was observed that the methyl group of the ester 15 (IC 50 � 0.125 ± 0.035 μM) in the para-position on the phenylethyl moiety was the determinant change to improve the antileishmanial activity when compared with ester 14 (IC 50 � 0.975 ± 0.24 μM) lacking this group [9,12].
All used reagents were of reagent grade and purchased from Sigma Aldrich.e following equipments were used: Irprestige-21 Shimadzu Fourier Transform spectrophotometer (Shimadzu, Kyoto, Japan) for IR, and Varian Mercury spectrometer (Palo Alto, CA, USA) 200 MHz for 1 H NMR and 50 MHz for 13 C NMR.
e standard for chemical shifts was the CDCl 3 or TMS solvent peak.For HR-MS, the LS-MALDI TOF/TOF apparatus was used equipped with a high-performance laser (λ � 355 nm) and a reflector operated by the software FlexControl 2.4 (Bruker Daltonics G, bsH, Bremen, Germany), and microOTOF operated by software Bruker Compass DataAnalysis 4.0 (Bruker Daltonics G, bsH, Bremen, Germany).Melting points were measured on an equipment Tecnal PFM-II, 220 V. To measure the optical activity of 20 and of starting material (−)-borneol a digital polarimeter P-2000 (Jasco) was used after standardization with 98% anhydrous methanol.

General Procedure for Preparation of Compound 35.
To a solution of 0.25 g of carboxylic acid in 250 ml of ROH under stirring, 0.5 ml of 96% (v/v) H 2 SO 4 were added and solution was refluxed for 3 h.After removing of half ROH under reduced pressure, the solution was diluted in 10 ml of water, and product was extracted with ethyl acetate.en, the product was successively washed with 5% (w/v) NaHCO 3 and water and dried with anhydrous Na 2 SO 4 .Finally, the solvent was evaporated to yield the pure compound [27][28][29].(35)

General Procedure for Preparation of Compound 13.
In a 50 ml round-bottom flask, 0.42 mmol (0.1 g) of 3,4,5trimethoxycinnamic acid were dissolved in 5.0 ml of acetone.en, 1.68 mmol (0.22 ml) of Et 3 N and 0.43 mmol of alkyl halide were added in the same solution, and the reaction mixture was refluxed for 48 h.e acetone was removed under reduced pressure, and the content of the reactions was diluted in 10 ml of ethyl acetate and 10 ml of water.e products were extracted thrice with 10 ml of ethyl acetate.Afterwards, organic phases were successively washed with 5% (w/v) NaHCO 3 and water, dried over anhydrous Na 2 SO 4 , and filtered.e product was purified via a silica gel 60 column chromatography (mobile phase: hexaneethyl acetate (1 : 1)) [30].(13)

General Procedure for Preparation of Compounds 18-21.
To a solution of 0.1 g (0.42 mmol) of 3,4,5-trimethoxycinnamic acid dissolved in 6.3 ml of CH 2 Cl 2 , 0.54 mmol of ROH and 0.015 g of DMAP were added.After 5 min of stirring, 0.1 g (0.54 mmol) of DCC was also added.Stirring was continued overnight at room temperature.e solvent was removed under vacuum, and the remaining content was diluted in 10 ml of ethyl acetate and 10 ml of water.e product was extracted with 10 ml ethyl acetate (three times), and the organic phases were successively washed with 5% (w/v) NaHCO 3 and water, dried over anhydrous Na 2 SO 4 , and filtered.After removing ethyl acetate under vacuum, the products were purified on a silica gel 60 column chromatography (mobile phase: hexane-ethyl acetate).

General Procedure for Preparation of Amides.
To a solution of 0.1 g of 3,4,5-trimethoxycinnamic in 0.84 ml of DMF in a round-bottom flask, 0.06 ml (0.42 mmol) of trimethylamine was added.e solution was cooled in an ice water bath and 0.42 mmol of amine were added, followed by a solution of 0.42 mmol of BOP (0.84 ml) in CH 2 Cl 2 .After stirring for 30 min, the reaction continued at room temperature for 3 h.CH 2 Cl 2 was removed under vacuum, and the remaining solution was diluted in 10 ml of ethyl acetate and 10 ml of water.e product was extracted with 10 ml of ethyl acetate (three times), and the organic phases were washed successively with 1 N HCl, water, 1 M NaHCO 3 , and water, dried over anhydrous Na 2 SO 4 , filtered, and evaporated.e product was purified on a silica gel 60 column chromatography (mobile phase: hexane-ethyl acetate (1 : 1)).

General Procedure for Preparation of Compound 36.
To a 5 ml round-bottom flask with 0.1 g (0.47 mmol) of 3′,4′,5′-trimethoxyacetophenone in 0.1 ml (0.94 mmol) of benzylic alcohol, the solution of 0.007 g of CuBr (0.047 mmol) in 0.07 ml (0.94 mmol) of pyridine was added, and then the solution of 0.025 mmol BF 3 •Et 2 O in 0.5 ml of chlorobenzene was added dropwise.e mixture was then stirred at 130 °C open air for 10 h.After cooling at room temperature, chlorobenzene was removed under reduced pressure.e remaining content was diluted in 10 ml of ethyl acetate and 10 ml of water.e product was extracted with 10 ml of ethyl acetate (three times), and the organic phase was successively washed with 5% (w/v) NaHCO 3 and water, dried over anhydrous Na 2 SO 4 , and filtered.After the removal of ethyl acetate under vacuum, the product was purified on a silica gel 60-column chromatography (mobile phase: hexane-ethyl acetate) [35].Benzyl 3,4,.Yield 56%; yellow oil; MM: 307.12 g/mol; 1
e Schneider medium was used as negative control, and amphotericin B, a reference drug for treatment of visceral leishmaniasis, was used as positive control [37].e IC 50 values were defined as the concentration of each compound that reduced the absorbance of treated cells by 50% when compared with the cell control.All assays were performed in triplicate.

Alamar Blue Assay.
To assess the cytotoxicity against normal cells, the cell line MRC-5 (human lung fibroblast) was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and was cultured as recommended by the ATCC.e cell culture was tested for mycoplasma using Mycoplasma Stain Kit (Sigma-Aldrich) to validate the use of cells free from contamination.e cell viability was measured colorimetrically using the alamarBlue assay [38].Shortly, compounds 18, 20, and 36 (dissolved in DMSO and diluted in the RPMI-1640 medium at a range of eight different concentrations from 0.19 to 25 μg/ml) were added to each well and incubated for 72 h.Doxorubicin (Sigma-Aldrich) was used as the positive control.Four hours before the end of incubation, 20 μl of a stock solution (0.312 mg/ml) of resazurin (Sigma-Aldrich) was added to each well.Absorbance was measured at 570 nm and 600 nm using the SpectraMax 190 Microplate Reader (Molecular Devices, Sunnyvale, CA, USA).e half maximal inhibitory concentration (IC 50 ) values with 95% confidence intervals were obtained by nonlinear regression using GraphPad Prism (Intuitive Software for Science; San Diego, CA, USA).

Methods
e first step for modeling was obtaining the initial 3D structure of each compound using the OpenEye's Omega software [39].Fragments for 3D structures generation were obtained with MAKEFRAGLIB, which is a part of the OMEGA suite.e Merck Molecular Force Field (MMFF94s) excluding the Coulomb interactions (mmff94s_NoEstat) was employed for fragment generation.Fragments with 4.0 kcal above the global minimum as well as those with a RMS distance lower than 0.1 from any conformer in the database were discarded.Default parameters were used for generating one 3D conformation per compound with OMEGA.Conformational exploration and alignment of the compounds were performed with the Open3DALIGN software [40].A conformational library of each compound was obtained Journal of Chemistry through quenched molecular dynamics (QMD) conformational searches that employed TINKER with an implicit solvent model [41].Two hundred molecular dynamic simulations of length 100 ps were performed for each compound.
e remaining parameters for the QMD in Open3DALIGN were kept to their default values.Structural alignment took place was performed using a method that combines atomand pharmacophore-based alignments as implemented in Open3DALIGN.
3D-QSAR models were developed with the Open3-DQSAR suite [42], which performs partial least squares (PLS) regression models from molecular interaction fields (MIF).Unless otherwise noted, default parameters were employed for Open3DQSAR.e input to Open3DQSAR is a set of aligned conformers of the dataset with associated bioactivities.
e best alignment produced by Open3-DALIGN in the previous step was used for 3D-QSAR models generation.In this study, compounds 19, 26, 31, and 32 were removed from the modeling process because of their undetermined IC 50 .e bioactivity of the remaining compounds was transformed to pIC 50 values (pIC 50 � −log 10 IC 50 ), provided that IC 50 values are expressed in molar units.
A grid was constructed around the aligned molecules in such a way that its box exceeded in 5 Å the largest molecule and grid spacing was set to 0.5 Å. Steric and electrostatic molecular mechanics of MIFs were computed using the Merck force field (MMFF94).For the computation of both MIFs, an alkyl carbon (charge +1) was used as probe.Before model construction, variables were filtered to discard those with van der Waals energies above 10 4 kcal/mol.Corresponding points on the electrostatic MIF were also removed.For both MIFs, energies greater than 30 kcal/mol or lower than −30 kcal/mol were set to 30 kcal/mol and −30 kcal/mol, respectively.Also values of energy between −0.05 and 0.05 kcal/mol were set to 0 in the two MIFs, and all variables with standard deviation lower than 0.1 were discarded.All variables spanning up to four levels were removed from both fields.As a final filter, variables were scaled employing the AUTO option of Open3DQSAR.
Initial PLS models considered 10 principal components.e optimal number of principal components (PCs) was selected based on the value of the Leave One Out crossvalidation R 2 (q 2 ).Variables were then grouped using the Smart Region definition procedure implemented in Open3DQSAR taking into account the previously identified optimal number of PCs. e grouped variables were subject to a selection procedure according to the Fractional Factorial Design using Leave Many Out (LMO) cross validation with 20 runs and 5 groups.Only selected variables in previous step were kept on the dataset.After these variable selection procedures, the PLS model was recomputed considering the optimal number of PCs and validated using the LOO and LMO strategies.

Conclusion
We synthesized a series of 32 piplartine analogues.e preliminary bioassay revealed that the most prominent compounds had rigid rings as substituents.Compound 18 contains a furfuryl portion attached to the (E)-3-(3,4,5trimethoxyphenyl)-acryloyl moiety.is modification significantly improved the bioactivity.Molecular simplification of 36 in relation to other derivatives formed an analogue with better biological activity against Leishmania.Furthermore, bornyl radical appears to be important for the bioactivity, given that compound 20 revealed the strongest antileishmanial activity.Together, these results show that compounds 18, 20, and 36 are promising lead compounds in the development of new options of pharmacotherapeutic drugs for leishmaniasis.

Figure 1 :
Figure 1: Chemical structure of piplartine (1).Modi cations in the green and blue substructures of the molecule were investigated in the present study.