key: cord-034363-6uscua0y
authors: Cerda-Cavieres, Christopher; Quiroz, Gabriel; Iturriaga-Vásquez, Patricio; Rodríguez-Lavado, Julio; Alarcón-Espósito, Jazmín; Saitz, Claudio; Pessoa-Mahana, Carlos D.; Chung, Hery; Araya-Maturana, Ramiro; Mella-Raipán, Jaime; Cabezas, David; Ojeda-Gómez, Claudia; Reyes-Parada, Miguel; Pessoa-Mahana, Hernán
title: Synthesis, Docking, 3-D-Qsar, and Biological Assays of Novel Indole Derivatives Targeting Serotonin Transporter, Dopamine D2 Receptor, and Mao-A Enzyme: In the Pursuit for Potential Multitarget Directed Ligands
date: 2020-10-10
journal: Molecules
DOI: 10.3390/molecules25204614
sha: 
doc_id: 34363
cord_uid: 6uscua0y

A series of 27 compounds of general structure 2,3-dihydro-benzo[1,4]oxazin-4-yl)-2-{4-[3-(1H-3indolyl)-propyl]-1-piperazinyl}-ethanamides, Series I: 7(a–o) and (2-{4-[3-(1H-3-indolyl)-propyl]-1-piperazinyl}-acetylamine)-N-(2-morfolin-4-yl-ethyl)-fluorinated benzamides Series II: 13(a–l) were synthesized and evaluated as novel multitarget ligands towards dopamine D(2) receptor, serotonin transporter (SERT), and monoamine oxidase-A (MAO-A) directed to the management of major depressive disorder (MDD). All the assayed compounds showed affinity for SERT in the nanomolar range, with five of them displaying Ki values from 5 to 10 nM. Compounds 7k, Ki = 5.63 ± 0.82 nM, and 13c, Ki = 6.85 ± 0.19 nM, showed the highest potencies. The affinities for D(2) ranged from micro to nanomolar, while MAO-A inhibition was more discrete. Nevertheless, compounds 7m and 7n showed affinities for the D(2) receptor in the nanomolar range (7n: Ki = 307 ± 6 nM and 7m: Ki = 593 ± 62 nM). Compound 7n was the only derivative displaying comparable affinities for SERT and D(2) receptor (D(2)/SERT ratio = 3.6) and could be considered as a multitarget lead for further optimization. In addition, docking studies aimed to rationalize the molecular interactions and binding modes of the designed compounds in the most relevant protein targets were carried out. Furthermore, in order to obtain information on the structure–activity relationship of the synthesized series, a 3-D-QSAR CoMFA and CoMSIA study was conducted and validated internally and externally (q(2) = 0.625, 0.523 for CoMFA and CoMSIA and r(2)(ncv) = 0.967, 0.959 for CoMFA and CoMSIA, respectively).

The fluorinated benzamides derivatives 12a-d were finally connected to indolylpropylpiperazines 9a-c to achieve the expected compounds 13a-l, with yields ranging from 47% to 85% (Scheme 5). In summary, 15 compounds were synthesized for Series I in yields ranging from 42% to 91%. The synthetic pathway of this series involved the fluorinated benzamide derivatives 12a-d, which were obtained from commercially available isomeric fluoro nitrobenzoic acids in a three-step sequence with good to excellent yields as shown in Scheme 4. In summary, 15 compounds were synthesized for Series I in yields ranging from 42% to 91%. The synthetic pathway of this series involved the fluorinated benzamide derivatives 12a-d, which were obtained from commercially available isomeric fluoro nitrobenzoic acids in a three-step sequence with good to excellent yields as shown in Scheme 4. The fluorinated benzamides derivatives 12a-d were finally connected to indolylpropylpiperazines 9a-c to achieve the expected compounds 13a-l, with yields ranging from 47% to 85% (Scheme 5). The fluorinated benzamides derivatives 12a-d were finally connected to indolylpropylpiperazines 9a-c to achieve the expected compounds 13a-l, with yields ranging from 47% to 85% (Scheme 5).

Molecules 2020, 25 Table 1 summarizes the affinity of compounds 7a-o for SERT, D2 receptor, and MAO-A. Most compounds were potent and clearly selective as SERT ligands, showing in all cases affinities in the nanomolar range, whereas the affinities for D2 and MAO-A ranged from micromolar to much higher values, respectively. A detailed analysis of SERT activities indicates that a C-5 substitution of the indole ring with halogens (fluorine or bromine; compounds 7g or 7m) leads to more potent compounds than the unsubstituted derivative (7a). On the other hand, the presence of a halogen atom at C-6 of the benzoxazine ring increased the affinity (e.g., 7c, 7d, and 7e vs. 7a). Accordingly, the most potent compounds were those exhibiting a dual halogen substitution pattern (7i, 7j, and 7k), with Ki values below 10 nM. The C-7 halogen substitution on the benzoxazine ring gave no consistent effects, slight increases (7b and 7h) or decreases (7n) of affinity were observed, as compared with the corresponding C-7 unsubstituted compounds (7a, 7g, and 7m, respectively). The D2 receptor affinity for this series indicates that no conclusive structure-activity relationships can be extracted for these compounds. Nevertheless, it is apparent that dihalogenated derivatives, bearing one halogen atom at the C-5 of the indole ring and the other at either the C-6 or C-7 of the benzoxazine moiety (7h-7k, 7m-7o), resulted in more potent compounds than their corresponding monohalogenated or unsubstituted counterparts (7a-7e, 7g). Moreover, the presence of a methoxyl group at the C-6 of the benzoxazine ring has almost no effect on the affinity of the compounds for D2 receptor. It is worth mentioning that the dihalogenated compound 7n was the only derivative displaying comparable affinities for SERT and D2 receptor (D2/SERT ratio = 3.6) and could be considered as a potential leader in the search of more potent multitarget compounds. Scheme 5. Synthesis of Series II derivatives 13a-l. Reagents and conditions: K 2 CO 3 , CH 3 CN, 80 • C, Yield (47-85%). Table 1 summarizes the affinity of compounds 7a-o for SERT, D 2 receptor, and MAO-A. Most compounds were potent and clearly selective as SERT ligands, showing in all cases affinities in the nanomolar range, whereas the affinities for D 2 and MAO-A ranged from micromolar to much higher values, respectively. A detailed analysis of SERT activities indicates that a C-5 substitution of the indole ring with halogens (fluorine or bromine; compounds 7g or 7m) leads to more potent compounds than the unsubstituted derivative (7a). On the other hand, the presence of a halogen atom at C-6 of the benzoxazine ring increased the affinity (e.g., 7c, 7d, and 7e vs. 7a). Accordingly, the most potent compounds were those exhibiting a dual halogen substitution pattern (7i, 7j, and 7k), with Ki values below 10 nM. The C-7 halogen substitution on the benzoxazine ring gave no consistent effects, slight increases (7b and 7h) or decreases (7n) of affinity were observed, as compared with the corresponding C-7 unsubstituted compounds (7a, 7g, and 7m, respectively). The D 2 receptor affinity for this series indicates that no conclusive structure-activity relationships can be extracted for these compounds. Nevertheless, it is apparent that dihalogenated derivatives, bearing one halogen atom at the C-5 of the indole ring and the other at either the C-6 or C-7 of the benzoxazine moiety (7h-7k, 7m-7o), resulted in more potent compounds than their corresponding monohalogenated or unsubstituted counterparts (7a-7e, 7g). Moreover, the presence of a methoxyl group at the C-6 of the benzoxazine ring has almost no effect on the affinity of the compounds for D 2 receptor. It is worth mentioning that the dihalogenated compound 7n was the only derivative displaying comparable affinities for SERT and D 2 receptor (D 2 /SERT ratio = 3.6) and could be considered as a potential leader in the search of more potent multitarget compounds. Table 1 . Affinities, measured as Ki values at the serotonin transporter (SERT), D 2 receptor, and percent of monoamine oxidase-A (MAO-A) inhibition (at 100 µM) of indolepiperazinyl benzoxazine derivatives (Series I).

Molecules 2020, 25, x FOR PEER REVIEW 6 of 30 

Considering the pharmacological results, docking studies aimed to rationalize the molecular interactions and binding modes of the designed compounds were carried out only in the human SERT (hSERT) and in selected cases at the D2 receptor.

The most potent compounds 7g, 7h, 7i, and 7k showed a common docking pose (Figure 1 ), which favors the following stabilizing interactions: A π-π interaction between the indole ring and the πdonor aromatic residue Tyr176, a coulombic interaction between the protonated piperazine N-1 with the Asp98 residue, and a π-cation interaction for the protonated piperazine with Tyr95. Furthermore, aromatic interactions were also observed for the benzoxazine ring with the residues Phe341 and 

Considering the pharmacological results, docking studies aimed to rationalize the molecular interactions and binding modes of the designed compounds were carried out only in the human SERT (hSERT) and in selected cases at the D 2 receptor.

The most potent compounds 7g, 7h, 7i, and 7k showed a common docking pose (Figure 1 ), which favors the following stabilizing interactions: A π-π interaction between the indole ring and the π-donor aromatic residue Tyr176, a coulombic interaction between the protonated piperazine N-1 with the Asp98 residue, and a π-cation interaction for the protonated piperazine with Tyr95. Furthermore, aromatic interactions were also observed for the benzoxazine ring with the residues Phe341 and Phe335. These drug-target interactions are in agreement with those described in the crystal structure of the hSERT in complex with the inhibitor (S)-citalopram [47, 48] . The relevance of the fluorinated substitution on the indole ring is clearly evidenced by comparison of compounds 7f and 7l. Both derivatives share the same substitution pattern in the benzoxazine ring, differing only by the presence of a fluorine atom at the indole moiety, which induces a different docking pose for 7f. Thus, the least potent compound of the series (7f) adopted a binding mode in which both indole and piperazine ring interactions are clearly less favored than the C-5 fluorinated counterpart ( Figure 2 ). Compounds showing intermediate affinities (7a-7e and 7m-7o) exhibited docking poses between the most and least favorable binding modes (not shown). The relevance of the fluorinated substitution on the indole ring is clearly evidenced by comparison of compounds 7f and 7l. Both derivatives share the same substitution pattern in the benzoxazine ring, differing only by the presence of a fluorine atom at the indole moiety, which induces a different docking pose for 7f. Thus, the least potent compound of the series (7f) adopted a binding mode in which both indole and piperazine ring interactions are clearly less favored than the C-5 fluorinated counterpart ( Figure 2 ). Compounds showing intermediate affinities (7a-7e and 7m-7o) exhibited docking poses between the most and least favorable binding modes (not shown). Phe335. These drug-target interactions are in agreement with those described in the crystal structure of the hSERT in complex with the inhibitor (S)-citalopram [47, 48] . The relevance of the fluorinated substitution on the indole ring is clearly evidenced by comparison of compounds 7f and 7l. Both derivatives share the same substitution pattern in the benzoxazine ring, differing only by the presence of a fluorine atom at the indole moiety, which induces a different docking pose for 7f. Thus, the least potent compound of the series (7f) adopted a binding mode in which both indole and piperazine ring interactions are clearly less favored than the C-5 fluorinated counterpart ( Figure 2 ). Compounds showing intermediate affinities (7a-7e and 7m-7o) exhibited docking poses between the most and least favorable binding modes (not shown). Docking simulations showed that compounds of this series adopt, at the D 2 receptor, a binding mode similar to that experimentally determined for the atypical antipsychotic risperidone [49] . Thus, the indole moiety appears located into the deep hydrophobic sub-pocket of the orthosteric site, lined by Cys118, Thr119, Ser197, Phe198, and Trp386, while the protonated piperazine N-1 locates in a favorable position to establish a coulombic interaction with Asp114 ( Figure 3) . Furthermore, the benzoxazine portion extends to the additional hydrophobic sub-pocket lined by Val91, Trp100, Phe110, and Tyr408, in a similar fashion to that observed in the crystal structure for the pyrimidinone moiety of risperidone. Interestingly, this general binding mode was observed for both the most and the least potent compounds of this series (7a, 7b, and 7m, 7n), respectively ( Figure 3A ,B). Therefore, it is tempting to speculate that the higher affinity showed by brominated derivatives (7m and 7n) is due to the formation of a halogen bond between the bromine and a hydroxyl group of an adjacent residue (e.g., Ser197). As observed (Figure 3 ), this could also change the position of the benzoxazine moiety, favoring its interactions at the more external hydrophobic sub-pocket. Docking simulations showed that compounds of this series adopt, at the D2 receptor, a binding mode similar to that experimentally determined for the atypical antipsychotic risperidone [49] . Thus, the indole moiety appears located into the deep hydrophobic sub-pocket of the orthosteric site, lined by Cys118, Thr119, Ser197, Phe198, and Trp386, while the protonated piperazine N-1 locates in a favorable position to establish a coulombic interaction with Asp114 ( Figure 3) . Furthermore, the benzoxazine portion extends to the additional hydrophobic sub-pocket lined by Val91, Trp100, Phe110, and Tyr408, in a similar fashion to that observed in the crystal structure for the pyrimidinone moiety of risperidone. Interestingly, this general binding mode was observed for both the most and the least potent compounds of this series (7a, 7b, and 7m, 7n), respectively ( Figures 3A and 3B) . Therefore, it is tempting to speculate that the higher affinity showed by brominated derivatives (7m and 7n) is due to the formation of a halogen bond between the bromine and a hydroxyl group of an adjacent residue (e.g., Ser197). As observed (Figure 3 ), this could also change the position of the benzoxazine moiety, favoring its interactions at the more external hydrophobic sub-pocket. Table 2 summarizes the affinity of Series II compounds 13a-13l for SERT, D2 receptor, and MAO-A. As in the case of indole benzoxazine derivatives (Series I), most indole morpholine ethylbenzamides (Series II) were potent SERT ligands, showing much lower affinities for D2 receptor and virtually no effect upon MAO-A activity. Regarding SERT activity, and in agreement with our previous studies, halogen substitution at C-5 of the indole ring with fluorine or bromine (compounds 13a-g) conducted an increase in affinity as compared with the unsubstituted analogues 13i-l, with the fluoro derivatives 13a-d being the most potent of the series. On the other hand, when the acetanilide portion, connected to the indolylpropylpiperazinyl fragment, was functionalized with a fluorine atom (at C-2) and a morpholino ethylcarboxamide, the best affinities were obtained when the bulkier substituent was located at meta position (compounds 13c, 13g, and 13k). Table 2 summarizes the affinity of Series II compounds 13a-13l for SERT, D 2 receptor, and MAO-A. As in the case of indole benzoxazine derivatives (Series I), most indole morpholine ethylbenzamides (Series II) were potent SERT ligands, showing much lower affinities for D 2 receptor and virtually no effect upon MAO-A activity. Regarding SERT activity, and in agreement with our previous studies, halogen substitution at C-5 of the indole ring with fluorine or bromine (compounds 13a-g) conducted an increase in affinity as compared with the unsubstituted analogues 13i-l, with the fluoro derivatives 13a-d being the most potent of the series. On the other hand, when the acetanilide portion, connected to the indolylpropylpiperazinyl fragment, was functionalized with a fluorine atom (at C-2) and a morpholino ethylcarboxamide, the best affinities were obtained when the bulkier substituent was located at meta position (compounds 13c, 13g, and 13k). 

Similar to the analysis of Series I and considering the pharmacological results, docking studies were carried out only in hSERT. In this series, seven compounds exhibited Ki values between 7 and 60 nM (13a, 13b, 13c, 13f, 13g, 13k, and 13l) . Docking simulations showed that compounds with the lowest Ki values (13c and 13g) share a common binding mode into the S1 site of the SERT, which is similar to that described for compounds of Series I ( Figure 4A ). Thus, the piperazine N-1 can establish ionic and π-cation interactions with Asp98 and Tyr176, respectively, while the indole moiety can participate in aromatic interactions with Tyr176 and Phe341. Interestingly, the ethylmorpholinic chain extends towards the extracellular vestibule (also known as the S2 site). On the other hand, for the compounds with the lowest affinities (13e and 13i), docking simulations showed that the piperazine N-1 was located farther away from Asp98 and Tyr95, making the possible ionic interactions with these residues unlikely or much weaker ( Figure 4B ). The analysis of the docking poses indicates that the most potent compounds, i.e., those having a 5,2-substitution pattern (13c, 13g, and 13k) exhibited an extended conformation at the binding site, while the least potent 

Similar to the analysis of Series I and considering the pharmacological results, docking studies were carried out only in hSERT. In this series, seven compounds exhibited Ki values between 7 and 60 nM (13a, 13b, 13c, 13f, 13g, 13k, and 13l). Docking simulations showed that compounds with the lowest Ki values (13c and 13g) share a common binding mode into the S1 site of the SERT, which is similar to that described for compounds of Series I ( Figure 4A ). Thus, the piperazine N-1 can establish ionic and π-cation interactions with Asp98 and Tyr176, respectively, while the indole moiety can participate in aromatic interactions with Tyr176 and Phe341. Interestingly, the ethylmorpholinic chain extends towards the extracellular vestibule (also known as the S2 site). On the other hand, for the compounds with the lowest affinities (13e and 13i), docking simulations showed that the piperazine N-1 was located farther away from Asp98 and Tyr95, making the possible ionic interactions with these residues unlikely or much weaker ( Figure 4B ). The analysis of the docking poses indicates that the most potent compounds, i.e., those having a 5,2-substitution pattern (13c, 13g, and 13k) exhibited an extended conformation at the binding site, while the least potent compounds (13e and 13i, showing a 2,4-substitution pattern) adopted a more constrained binding mode, impairing the most relevant interactions.

Molecules 2020, 25, x FOR PEER REVIEW 10 of 30 compounds (13e and 13i, showing a 2,4-substitution pattern) adopted a more constrained binding mode, impairing the most relevant interactions. 

To systematize the structure-activity relationship of the synthesized molecules, we carried out a 3-D-QSAR study of the CoMFA and CoMSIA type. The complete series of 27 molecules was divided into training (19 compounds) and test sets (8 compounds) in a ratio of 70:30, selecting the test set compounds at random to avoid bias. The q 2 values for the best models were 0.625 and 0.523 for CoMFA and CoMSIA, respectively while the r 2 ncv values were 0.967 and 0.959 for CoMFA and CoMSIA, respectively. The statistical summary, as well as the tables of affinities for both models and their respective graphs, are incorporated in the Supplementary Material.

The steric contour map of CoMFA ( Figure 5A ) shows a green polyhedron on the bromine atom of compound 7k, the most active of the series. This means that the insertion of bulky atoms or groups in this position is favorable for biological activity. This is consistent with docking studies showing that compounds of series I place halogen into the void space close to lipophilic residues like Trp100 and Tyr408. In the case of compounds of series II, the meta-substituted benzamides placed the chain towards the green region, not the ortho-substituted ones, so it is preferable that the chains are in the meta-position. This is confirmed in the docking of these compounds, in which better accommodation is observed in the SERT binding site. On the other hand, the electrostatic contour map ( Figure 5B ) shows three blue polyhedra of significant size. This means that the presence of positively charged atoms in these positions would be favorable for affinity. Such polyhedra are located on the carbon atom bonded to the halogen in the case of series I, suggesting that the presence of electronegative atoms bonded to the aforementioned carbon is favorable. The second blue polyhedron is localized on the oxygen atom of the carbonyl group belonging to the ortho-substituted series II amidecompounds. Therefore, oxygen atom remotion would be favorable for affinity. Finally, the third polyhedron is observed on the oxygen atom of the morpholine ring in the ortho-substituted compounds for series II, indicating that changing the morpholine by a piperazine or piperidine ring should lead to better affinities. Furthermore, alkyl chains substitutions at the ortho-position in the benzamide ring resulted in less favorable affinities compared to meta substitutions as was experimentally corroborated.

The hydrophobic contour map of CoMSIA ( Figure 5C ) showed a gray polyhedron at position C5 of the indole ring, meaning that the presence of hydrophilic groups is favorable for affinity. In fact, the C5 fluorine-substituted indoles displayed the best affinities of the series. Other polar groups like A B Figure 4 . Docking poses in SERT obtained for compounds 13c in cyan and 13g in purple (A), and 13e in light blue, and 13i in orange (B). Nearby residues < 5 Å (grey sticks) and Na + atoms (pink spheres) are shown. Dotted lines represent ionic interactions, orange lines represent π-cation interactions, and aromatic interactions are shown with green lines.

To systematize the structure-activity relationship of the synthesized molecules, we carried out a 3-D-QSAR study of the CoMFA and CoMSIA type. The complete series of 27 molecules was divided into training (19 compounds) and test sets (8 compounds) in a ratio of 70:30, selecting the test set compounds at random to avoid bias. The q 2 values for the best models were 0.625 and 0.523 for CoMFA and CoMSIA, respectively while the r 2 ncv values were 0.967 and 0.959 for CoMFA and CoMSIA, respectively. The statistical summary, as well as the tables of affinities for both models and their respective graphs, are incorporated in the Supplementary Material.

The steric contour map of CoMFA ( Figure 5A ) shows a green polyhedron on the bromine atom of compound 7k, the most active of the series. This means that the insertion of bulky atoms or groups in this position is favorable for biological activity. This is consistent with docking studies showing that compounds of series I place halogen into the void space close to lipophilic residues like Trp100 and Tyr408. In the case of compounds of series II, the meta-substituted benzamides placed the chain towards the green region, not the ortho-substituted ones, so it is preferable that the chains are in the meta-position. This is confirmed in the docking of these compounds, in which better accommodation is observed in the SERT binding site. On the other hand, the electrostatic contour map ( Figure 5B ) shows three blue polyhedra of significant size. This means that the presence of positively charged atoms in these positions would be favorable for affinity. Such polyhedra are located on the carbon atom bonded to the halogen in the case of series I, suggesting that the presence of electronegative atoms bonded to the aforementioned carbon is favorable. The second blue polyhedron is localized on the oxygen atom of the carbonyl group belonging to the ortho-substituted series II amide-compounds. Therefore, oxygen atom remotion would be favorable for affinity. Finally, the third polyhedron is observed on the oxygen atom of the morpholine ring in the ortho-substituted compounds for series II, indicating that changing the morpholine by a piperazine or piperidine ring should lead to better affinities. Furthermore, alkyl chains substitutions at the ortho-position in the benzamide ring resulted in less favorable affinities compared to meta substitutions as was experimentally corroborated. red polyhedron on the halogen atom at position 5 of the indole ring means that the presence of electron-rich atoms is favorable for affinity. It is interesting to note that the blue polyhedron intersecting the carbon atom of indole at position 5 is complementary to the red polyhedron. In consequence, the presence of a positive charge on the indole ring is favorable for activity. Other potential electron-withdrawing groups to be explored are CN, NO2, and COR. Docking studies showed π-stacking interaction between the π-deficient indole ring with Tyr176, Phe341, and Trp386 residues. The hydrophobic contour map of CoMSIA ( Figure 5C ) showed a gray polyhedron at position C5 of the indole ring, meaning that the presence of hydrophilic groups is favorable for affinity. In fact, the C5 fluorine-substituted indoles displayed the best affinities of the series. Other polar groups like OH, NH 2 , or NR 2 would also be interesting to evaluate at this position. Similarly, a yellow polyhedron located on the bromine atom of the benzoxazine framework (compound 7k) means that the presence of lipophilic groups is favorable for activity. In concordance, halogens like Cl, Br, and I would be the most appropriated substituents and groups, such as aromatic rings, alkyl, and/or alkoxy chains, could also be explored. In the case of compounds of series II, a yellow polyhedron is located on the amide group of the meta-substituted compounds; therefore, the replacement of the amide by a less-polar function, such as a ketone or ester, would be an interesting option to explore. On the other hand, the electrostatic contour map of CoMSIA ( Figure 5D ) showed two polyhedra around compound 7k. A red polyhedron on the halogen atom at position 5 of the indole ring means that the presence of electron-rich atoms is favorable for affinity. It is interesting to note that the blue polyhedron intersecting the carbon atom of indole at position 5 is complementary to the red polyhedron. In consequence, the presence of a positive charge on the indole ring is favorable for activity. Other potential electron-withdrawing groups to be explored are CN, NO 2 , and COR. Docking studies showed π-stacking interaction between the π-deficient indole ring with Tyr176, Phe341, and Trp386 residues.

Melting points were determined on a hot-stage apparatus and were uncorrected. The 1 H and 13 C-NMR spectra were obtained on a Bruker DRX-300 spectrometer (300 and 75 MHz, respectively) in CDCl 3 , DMSO-d 6 , and CD 3 COCD 3 -d 6 . Chemical shifts were recorded in ppm (δ) relative to TMS as an internal standard. J values are given in Hz. Micro-analyses were carried out on a Fisons EA 1108 analyzer. High-resolution mass spectra were recorded on a DSA-TOFAxION 2 TOF MS (Perkin Elmer, Shelton, CT, USA), positive mode. Silica gel Merck 60 (70-230 mesh) and aluminum sheets coated with silica gel 60 F254 were used for column and TLC chromatography, respectively.

To a solution containing 2-Chloro-1-(2,3-dihydrobenzo[b] [1, 4] oxazin-4-yl) ethanamide 4a (1.5 g; 7.09 mmol) in dry CH 3 CN (60 mL), N-Boc-piperazine (1321 mg; 7.09 mmol) and anhydrous K 2 CO 3 (980 mg; 7.09 mmol) were added. The mixture was stirred at 80 • C for 24 h. After this time, the mixture was diluted with water (100 mL) and the solution extracted with EtOAc (100 mL × 3), dried over anhydrous Na 2 SO 4 , and concentrated under reduced pressure. The organic crude was purified by silica gel column chromatography with EtOAc as eluent, to provide 5a (2152 mg; 84% yield) as a white solid. m. [1, 4] oxazin-4-yl)-ethanamide 4c (1.5 g; 6.53 mmol), N-Boc-piperazine (1216 mg; 6.53 mmol), and anhydrous K 2 CO 3 (902 mg; 6.53 mmol), to afford 5c (2033 mg; 82% yield) as a white solid. m. [1, 4] oxazin-4-yl)-2-oxo-ethyl]-1-piperazinyl] tert-butylcarbamate 5a (2 g; 5.53 mmol) in dry CH 2 Cl 2 (20 mL) and trifluoroacetic acid (12 mL) was stirred at 0 • C, for 4 h. After this time, dry CH 2 Cl 2 (200 mL) was added and neutralized with solid NaHCO 3 (10 g) to later filter on celite. The mixture was finally diluted with a saturated solution of NaHCO 3 (200 mL), extracted with EtOAc (8 × 50 mL), dried over anhydrous Na 2 SO 4 , and concentrated under vacuum to obtain pure 6a (867 mg; 82% yield) as an unstable yellow light solid, highly hygroscopic; 1 

To a solution of 3-(5-Fluoro-1H-3-indolyl)-propyl-4-methylbencensulfonate 1b (268 mg; 0.72 mmol) in CH 3 CN (50 mL), 1-(7-Fluoro-2,3-dihydro-benzo[b] [1, 4] [1, 4] oxazin-4-yl) ethanamide 4e (227 mg; 0.92 mmol), and anhydrous K 2 CO 3 (127 mg; 0.92 mmol) were added. The mixture was heated at 80 • C for 24 h. After this time, the resulting mixture was poured into water (100 mL) and extracted with EtOAc (4 × 50 mL), dried over anhydrous Na 2 SO 4 , and concentrated under reduced pressure. The organic crude was purified by column chromatography EtOAc/MeOH 111.1, 111.8, 114.8, 118.5, 118.7, 119.1, 121.2, 122.6, 125.4, 126.3, 127.6, 128.1, 136.7, 146.0 To a mixture containing water-acetic acid-ethanol (1:1:1), 4-Fluoro-N-(2-morpholin-4-yl-ethyl)-2nitro-benzamide 10a (1 g; 3.36 mmol) and iron powder (734 mg; 13.1 mmol) were added. The resulting mixture was heated and stirred for 3 h at 70 • C. After this time, the mixture was filtered to remove excess metallic iron, transferred to a flask containing a mixture of EtOAc/H 2 O (400 mL, 1:1), and neutralized with NaHCO 3 (10 gr). The aqueous phase was extracted with EtOAc (50 mL × 3). The organic layer was dried over anhydrous Na 2 SO 4 and concentrated under vacuum to give a crude, which was purified by column chromatography with EtOAc/MeOH (6:1) to give 11a (891 mg; 98% yield) as a yellow light solid. m.p.: 120.3-121. 

To determine the binding of all compounds at SERT, competitive binding assays were performed according to previously reported procedures with some modifications [36] . Briefly, assays were carried out in a total volume of 0.5 mL containing 9 µg protein of membrane from a clonal cell line HEK-293 that overexpresses SERT, 50 mM Tris buffer, pH 7.4, 120 mM NaCl, 5 mM KCl, 2 nM [ 3 H]-paroxetine (specific activity 20.8 Ci/mmol, PerkinElmer), and the compounds to be tested at different concentrations (10 −9 -10 −4 M). After 1 h at 27 • C, incubations were stopped by rapid filtration through Whatman GF/C filters presoaked in 0.5% polyethyleneimine, which were washed five times with 3 mL of ice-cold buffer, dried, and put in Eppendorf tubes with scintillation liquid. Radioactivity was counted by a liquid scintillation counter (MicroBeta 2450 microplate counter, PerkinElmer). Control curve was performed with fluoxetine in the same experimental conditions. Non-specific binding was determined with 10 µM fluoxetine.

To determine the binding of all compounds at D 2 receptor, competitive binding assays were performed according to provider indications with some modifications. Briefly, assays were carried out in a total volume of 0.5 mL containing 3 µg protein of membrane from a CHO-K1 clonal cell line that overexpresses D 2 receptor, 50 mM Tris buffer, pH 7.4, 120 mM NaCl, 5 mM KCl, 5 mM MgCl 2 , 1 mM EDTA, 0.5 nM [ 3 H]-methylspiperone (specific activity 64.1 Ci/mmol, PerkinElmer), and the compounds to be tested at different concentrations (10 −9 -10 −4 M). After 2 h at 27 • C, incubations were stopped by rapid filtration through Whatman GF/C filters presoaked in 0.5% polyethyleneimine, which were washed five times with 3 mL of ice-cold wash buffer (50 mM Tris buffer, pH 7.4, 154 mM NaCl), dried, and put in Eppendorf tubes with scintillation liquid. Radioactivity was counted as described before. Control curve was performed with haloperidol in the same conditions. Non-specific binding was determined with 10 µM haloperidol.

Analysis of data: All curves were fitted using the sigmoidal dose-response inhibition curve (variable slope) equation built into GraphPad PRISM 5.01 (GraphPad Software Inc., San Diego, CA, USA). The analysis gives the IC 50 value (i.e., the drug concentration inhibiting specific binding by 50%) to calculate Ki (affinity constant) by the Cheng-Prussof equation (Ki = IC 50 /(1 + ([radioligand]/Kd (radioligand))). The Kd values used correspond to 0.13 nM to [ 3 H] paroxetine on SERT [50] , and 0.1 nM to [ 3 H]-methylspiperone on D 2 (provided by the manufacturer). The IC 50 and Ki values correspond to the results of three independent experiments, each in triplicate. All data are expressed as the mean ± SEM.

All experimental procedures were approved by the Ethics Committee of the University of Santiago de Chile and the Science Council (FONDECYT) of Chile and followed internationally accepted guidelines (NIH Guide for the Care and Use of Laboratory Animals).

The effects of the compounds on rat MAO-A activity were studied following a previously reported methodology [50, 51] , using a crude rat brain mitochondrial suspension as a source of enzyme. Serotonin (100 µM) was used as the selective substrate for MAO-A. This compound and its metabolite were detected by HPLC with electrochemical detection. As an exploratory evaluation, the percentage of MAO-A inhibition in the presence of 100 µM of the different compounds was determined, with the idea of evaluating in detail those compounds showing an inhibitory activity in the range of 70-100%.

Molecular docking studies for the two families of compounds were performed on two different protein targets (SERT and D 2 receptor). All dockings were carried out at pH 7.4 in the crystal structures of human SERT (hSERT PDB: 5I73) [47] and human D 2 receptor [hD 2 PDB: 6CM4) [49] . All compounds were modelled using the Spartan'14 Software (Wavefunction, Inc. Irvine, CA) and geometry optimization calculations were carried out using the software package at the Hartree-Fock level using the 6-31G* basis set. Docking studies were performed using AutoDockv4.2 [52] software suite with Autodock Tools ADT 1.5.6 [52, 53] following the standard docking procedure for rigid proteins. Grid maps were calculated using the autogrid option with a grid volume of 70 × 70 × 70 points with a grid spacing of 0.375 Å and centered on the coordinates x, y, z: 33.3 184.5; 0.565 −9.397; and 37.019 28.136 for the SERT and D 2 receptor respectively. Docking simulations were performed with a Lamarckian genetic algorithm (LGA) and binding energies were estimated according to the internal scoring function implemented by the program; 250 independent runs per ligand were carried out with an initial population of 300 individuals. Default settings were used for all other parameters. The lowest free-energy resulting complexes were selected and further analyzed using the Visual Molecular Dynamic (VMD) visualization program [54] . Validation of the docking protocol was performed using the co-crystallized ligands (S)-citalopram and risperidone for SERT and D 2 , respectively.

CoMFA and CoMSIA studies were performed with Sybyl X-1.2 software [55] installed in a Windows 10 environment on a PC with an Intel Core i7 CPU. The geometric optimization, field calculation, and charges calculation were performed as previously reported [56] ( Figure S1 and Table S3 in Supplementary Material) [57] . The internal validation of the models was done by calculating the cross-validation coefficient q 2 [58] . The models with the highest value of q 2 were selected and then subjected to external validation [59] [60] [61] (Table S2 ). In all cases, the best models passed the validation limits [59] (Table S2 ). The regression graphs of each model and the tables of experimental versus calculated values are in the Supplementary Material (Table S3, Figure S2 ).

According to these results, the design of hybrid or bifunctional compounds, i.e., molecules that incorporate two pharmacophores known to act at different receptors into a single chemical entity, is an attractive approach for the development of agents having a targeted polypharmacological profile [19, 62, 63] . In the present work, we attempted to combine SERT effects previously demonstrated for indolylalkylpiperazine derivatives, functionalizing the parent scaffold with structural fragments of drugs with known activity upon D 2 receptor or MAO-A. Unexpectedly, the synthesized compounds did not show, in most cases, a multitarget profile, since they exhibited a high affinity for SERT while showing almost no effect at D 2 receptor or MAO-A. This indicates that this strategy, although plausible, requires a very fine design of the fragments to be connected and how these are going to be linked. Beyond these considerations, our results highlight the remarkable stability of the indolylpropylpiperazine skeleton as SERT ligand, which exhibits a high affinity by this target, apparently regardless of the type of the associated moiety [34] [35] [36] . We think that this represents an important feature for the design of polypharmacological molecules, in which an effect upon SERT is pursued. Docking and QSAR results allowed us to rationalize the high SERT affinity observed for compounds in both studied families. Thus, the presence of a halogen at the C-5 position of the indole ring and fluorine atoms at the benzoxazine (Series I) or acetanilide (Series II) moieties probably induces electronic deprotection of the corresponding aromatic rings, favoring stronger π-π interactions of these frameworks with donor aromatic residues at the binding site.

Interestingly, one of the compounds (7n) showed a promissory multitarget profile, being the only derivative showing a relatively high and comparable affinity for SERT and D 2 receptor (Ki = 84.4 and 307 nM, respectively). Even though at this time it is difficult to determine the molecular aspects underlying this pharmacological promiscuity, it is clear that for polypharmacological drugs, a similar affinity for different receptors is the most relevant characteristic, and therefore 7n stands as a very attractive lead for further optimization. Figure S1 . The superimposed structures of all compounds used in the CoMFA/CoMSIA models. Figure S2 . Plots of experimental versus predicted pKi values for the training and test set molecules for CoMFA (A, B) and CoMSIA (C, D) models. Figure S3 . hSERT affinity curves for compounds of Series I (7a, 7b,  7c, 7d, 7e, 7f, 7g, 7h, 7i, 7j, 7k, 7l, 7m, 7n, 7o , and fluoxetine), displaying IC 50 values. Each determination was made in triplicate and the data were expressed as the mean ± SD. Figure S4 . D2 affinity curves for compounds of Series I (7a, 7b, 7c, 7d, 7e, 7f, 7g, 7h, 7i, 7j, 7k, 7l, 7m, 7n, 7o , and haloperidol), displaying IC 50 values. Each determination was made in triplicate and the data were expressed as the mean ± SD. Figure S5 . hSERT affinity curves for compounds of Series II (13a, 13b, 13c, 13d, 13e, 13f, 13g, 13h, 13i, 13j, 13k, 13l , and fluoxetine), displaying IC 50 values. Each determination was made in triplicate and the data were expressed as the mean ± SD. Figure S6 . D2 affinity curves for compounds of Series II (13a, 13b, 13c, 13d, 13e, 13f, 13g, 13h, 13i, 13j, 13k, 13l , and haloperidol), displaying IC 50 values. Each determination was made in triplicate and the data expressed as the mean ± SD.

HRMS: (EI) Calculated for C 30 H 38 BrFN 6 O 3 (M + ) = 629

-morpholin-4-ylethyl) Benzamide (13f) 5-Bromo-3-(3-piperazin-1-yl-propyl)-1H-indole 9c (187 mg; 0.58 mmol), 3-(2-chloro-acetylamino)-4-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12b (199 mg; 0.58 mmol), and anhydrous K 2 CO 3 (80 mg; 0.58 mmol), to afford 13f (193 mg

mmol), 5-(2-chloro-acetylamino)-2-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12d (199 mg; 0.58 mmol), and anhydrous K 2 CO 3 (80 mg; 0.58 mmol), to afford 13g (249 mg

53 (dd, 1H, H-3 , J o = 9

-morpholin-4-ylethyl) Benzamide (13h) 5-Bromo-3-(3-piperazin-1-yl-propyl)-1H-indole 9c (187 mg; 0.58 mmol), 2-(2-chloro-acetylamino)-5-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12c

14 (s, 2H, H-6 ), 3.40-3.43 (m, 2H, H-8 ), 3.57 (t, 4H, H-2

HRMS: (EI) Calculated for C 30 H 38 BrFN 6 O 3 (M + ) = 629

Piperazin-1-yl-propyl)-1H-indole 9a (150 mg; 0.62 mmol), 2-(2-chloro-acetylamino)-4-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12a (213 mg; 0.62 mmol), and anhydrous K 2 CO 3

DMSO-d 6 ): δ 1.86 (m, 2H, H-2 ), 2.25-2.49 (m, 12H, H-3 , H-4 , H-9 and H-1 ), 2.58 (m, 4H, H-5 )

5 Hz), 141.4, 163.8 (d, 1 J C-F = 245 Hz), 167.3, and 170.3 ppm. HRMS: (EI) Calculated for C 30 H

Piperazin-1-yl-propyl)-1H-indole 9a (150 mg; 0.62 mmol), 3-(2-chloro-acetylamino)-4-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12b (213 mg; 0.62 mmol), and anhydrous K 2 CO 3

19 (s, 2H, H-6 ), 3.38-3.40 (m, 2H, H-8 ), 3.57 (t, 4H, H-2 , J = 4.5 Hz), 6.96 (td, 1H, H-5 or H-6

3 Hz), 136.3, 154.7 (d, 1 J C-F = 249 Hz), 165.1, and 168.5 ppm. HRMS: (EI) Calculated for C 30 H

Piperazin-1-yl-propyl)-1H-indole 9a (150 mg; 0.62 mmol), 5-(2-chloro-acetylamino)-2-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12d (213 mg; 0.62 mmol), and anhydrous K 2 CO 3

13 (s, 2H, H-6 ), 3.39 (q, 2H, H-8 , J = 6.4 Hz), 3.58 (t, 4H, H-2 , J = 4.2 Hz), 6.96 (t, 1H, H-5 or H-6, J = 7.2 Hz), 7.06 (t, 1H, H-6 or H-5

Piperazin-1-yl-propyl)-1H-indole 9a (200 mg; 0.82 mmol), 2-(2-chloro-acetylamino)-5-fluoro-N-(2-morpholin-4-yl-ethyl)-benzamide 12c (281 mg

(m, 2H, H-8 ), 3.56 (t, 4H, H-2 , J = 4.5 Hz), 6.96 (td, 1H, H-5 or H-6

Ci/mmol; Code NET869)

Ci/mmol; NET856), membrane from clonal cell line HEK-293 that overexpresses SERT (Code: RBHSTM400UA), and membrane from CHO-K1 clonal cell line that overexpresses D 2 receptor (Code: RBHD2CM400UA) were purchased from Perkin-Elmer

Depression Fact Sheet

Depression, mania and self-reported creativity in bipolar disorder

Characterizing neurocognitive markers of familial risk for depression using multi-modal imaging, behavioral and self-report measures

Uncovering the Genetic Architecture of Major Depression

The Social Cost of Major Depression

Grand challenges in global mental health

History and evolution of the monoamine hypothesis of depression

New approaches to antidepressant drug discovery: Beyond monoamines

Pathophysiology of depression: Do we have any solid evidence of interest to clinicians? World Psychiatry

Chapter 34 Novel Therapeutic Targets for Major Depressive Disorder

The failure of the antidepressant drug discovery process is systemic

Developmental changes in serotonin signaling: Implications for early brain function, behavior and adaptation

The pharmacological basis of the serotonin system: Application to antidepressant response

Psychopharmacology of anxiety and depression: Historical aspects, current treatments and perspectives

Circadian regulation of depression: A role for serotonin

A brief history of the development of antidepressant drugs: From monoamines to glutamate

Perspectives in Designing Multifunctional Molecules in Antipsychotic Drug Discovery

Benzo-and thienobenzo-diazepines: Multi-target drugs for CNS disorders

The development of novel polypharmacological agents targeting the multiple binding sites of nicotinic acetylcholine receptors

Triple Reuptake Inhibitors as Potential Therapeutics for Depression and Other Disorders: Design Paradigm and Developmental Challenges

The Role of Dopamine and Its Dysfunction as a Consequence of Oxidative Stress

The debate over dopamine's role in reward: The case for incentive salience

learning and motivation

Dopamine System Dysregulation in Major Depressive Disorders

The Role of Dopamine in Schizophrenia from a Neurobiological and Evolutionary Perspective: Old Fashioned, but Still in Vogue

The role of dopamine in the brain-Lessons learned from Parkinson's disease

The Role of Dopamine in the Pathophysiology of Depression

Molecular role of dopamine in anhedonia linked to reward deficiency syndrome (RDS) and anti-reward systems

Major Depression as a Complex Dynamic System

Design, synthesis and evaluation of vilazodone-tacrine hybrids as multitarget-directed ligands against depression with cognitive impairment

Multitarget Selective Antidepressants Design: Latest Developments, Opportunities and Challenges

polypharmacy" and multi-target agents, complementary strategies for improving the treatment of depression: A comparative appraisal

Synthesis, in vitro evaluation and molecular docking of a new class of indolylpropyl benzamidopiperazines as dual AChE and SERT ligands for Alzheimer's disease

Synthesis and Docking of Novel 3-Indolylpropyl Derivatives as New Polypharmacological Agents Displaying Affinity for 5-HT(1A)

Synthesis, docking and pharmacological evaluation of novel homo-and hetero-bis 3-piperazinylpropylindole derivatives at SERT and 5-HT1A receptor

Synthesis and biological screening of novel indolalkyl arenes targeting the serotonine transporter

Synthesis, 5-hydroxytryptamine1A receptor affinity and docking studies of 3-[3-(4-aryl-1-piperazinyl)-propyl]-1H-indole derivatives

Inhibition of monoamine oxidase by indole and benzofuran derivatives

Synthesis and structure-activity relationship in a class of indolebutylpiperazines as dual 5-HT(1A) receptor agonists and serotonin reuptake inhibitors

Potent and highly selective dual-targeting monoamine oxidase-B inhibitors: Fluorinated chalcones of morpholine versus imidazole

Structure-guided development of dual β2 adrenergic/dopamine D2 receptor agonists

Gmeiner, P. β-Arrestin biased dopamine D2 receptor partial agonists: Synthesis and pharmacological evaluation

Proof of concept study for designed multiple ligands targeting the dopamine D2, serotonin 5-HT2A, and muscarinic M1 acetylcholine receptors

Synthesis, conformational analysis and antidepressant activity of moclobemide new analogues

Moclobemide: Therapeutic use and clinical studies

Synthesis and biological evaluation of potential acetylcholinesterase inhibitors based on a benzoxazine core

X-ray structures and mechanism of the human serotonin transporter

Structural basis for action by diverse antidepressants on biogenic amine transporters

Structure of the D2 dopamine receptor bound to the atypical antipsychotic drug risperidone

Monoamine Oxidase Inhibitory Properties of Some Methoxylated and Alkylthio Amphetamine Derivatives: Structure-Activity Relationships

Pharmacological profile of antidepressants and related compounds at human monoamine transporters

AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility

Python: A programming language for software integration and development

VMD: Visual molecular dynamics

1.2; Tripos International

Three-Dimensional Quantitative Structure-Activity Relationships (3D-QSAR) on a Series of Piperazine-Carboxamides Fatty Acid Amide Hydrolase (FAAH) Inhibitors as a Useful Tool for the Design of New Cannabinoid Ligands

Molecular similarity indices in a comparative analysis (CoMSIA) of drug molecules to correlate and predict their biological activity

Validation of the general purpose tripos 5.2 force field

Beware of q2!

Best Practices for QSAR Model Development, Validation, and Exploitation

Some case studies on application of "r(m)2" metrics for judging quality of quantitative structure-activity relationship predictions: Emphasis on scaling of response data

Polypharmacology of dopamine receptor ligands

Multitarget opioid ligands in pain relief: New players in an old game

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution

The authors declare no conflict of interest.