Design, Synthesis and Biological Evaluation of Novel 1,3,5-Triazines: Effect of Aromatic Ring Decoration on Affinity to 5-HT7 Receptor

Considering the key functions of the 5-HT7 receptor, especially in psychiatry, and the fact that effective and selective 5-HT7 receptor ligands are yet to be available, in this work, we designed and synthesized novel 1,3,5-triazine derivatives particularly based on the evaluation of the effect of substituents at aromatic rings on biological activity. The tested compounds showed high affinity to the 5-HT7 receptor, particularly ligands N2-(2-(5-fluoro-1H-indol-3-yl)ethyl)-N4-phenethyl-1,3,5-triazine-2,4,6-triamine 2 (Ki = 8 nM) and N2-(2-(1H-indol-3-yl)ethyl)-N4-(2-((4-fluorophenyl)amino)ethyl)-1,3,5-triazine-2,4,6-triamine 12 (Ki = 18 nM) which showed moderate metabolic stability, and affinity to the CYP3A4 isoenzyme. As for the hepatotoxicity evaluation, the tested compounds showed moderate cytotoxicity only at concentrations above 50 µM. Compound 12 exhibited less cardiotoxic effect than 2 on Danio rerio in vivo model.


Introduction
The 5-HT 7 receptor (5-HT 7 R) is one of seven types of protein G-coupled aminergic serotonin receptors [1]. It is found in the central nervous system, mainly in the brain, in which it has regulatory functions, for example (the day-night cycle), and may affect behavior, mood, emotions, or memory [2,3]. The receptor is also expressed outside the central nervous system, e.g., in the intestines [4], mammary glands [5], lungs [6], or prostate [7].
Three human splicing variants (h5-HT 7(a) , h5HT 7(b) , and h5-HT 7(d) ) have been reported so far. As long as the h5-HT 7(a) and h5-HT 7(b) variants do not differ significantly in terms of the number of amino acids in the C-terminal tail, for example, and maintain similar pharmacological properties, the h5-HT 7(d) variant shows the greatest differences in the C-terminal tail, which may lead to slightly different functionality [1]. The primary signaling pathway for the 5-HT 7 receptor involves the receptor binding the ligand, followed by phosphorylation of subunit G combined with its dissociation into subunit G S and heterodimer G 12 . In the subsequent stage, protein G S (canonical signaling) is activated, which triggers isoform CA (adenyl cyclase) and leads to an intrinsic increase in cAMP (cyclic adenosine monophosphate) levels. cAMP induces PKA (protein kinase A) expression, which in turn induces further phosphorylation of other proteins, for example, on Ras, ERK, and Akt pathways. It has also been shown that protein Gs and protein G 12 (non-canonical In our previous paper [15], we also investigated the effect of substituents (ortho-OMe and 2,3-Cl2) at the phenyl ring attached to the aminoethyl chain on affinity toward 5-HT7R. The substituents were found to reduce receptor affinity, but the effect of a specific substituent position on activity was not tested. To investigate this aspect, we decided to synthesize a group of ligands 7-15 (Scheme 1, type 2) containing Cl, F, and OMe at positions ortho-, meta-, and para-and to evaluate their effects on affinity toward 5-HT7R. We also decided to incorporate more complex substituents into the structures of studied compounds (phthalimide 16 and benzimidazole 17-19 cores, Scheme 1, type 3) due to the potential formation of hydrogen bonds to stabilize the ligand-receptor additionally. Unexpectedly, we prepared compounds 20-22 without the aminoethyl chain in our synthesis experiments (Scheme 1, type 4). To determine the effect of the lack of an alkyl linker between the triazine and the aromatic system on affinity to the receptors in question, we also evaluated these compounds in in vitro tests. All the resulting compounds were tested in an extended receptor panel, including affinity toward 5-HT1A, 5-HT2A, 5-HT6, and D2 receptors, to determine the selectivity of the compounds. Bioconformation and key interactions involved in the forming the ligand-receptor complex were proposed for active structures. The two best compounds were evaluated in terms of safety and bioavailability in in vitro ADME-Tox tests.
particularly the fluorine atom [18,19]. We showed [20] that the compounds tested without any ring substituent ( Figure 1A) had low in vitro metabolic stability. By incorporating chlorine in the structure [15] (Figure 1B), metabolic stability increased with a slight decrease in cytotoxicity in the HepG2 cell line. This correlation is also confirmed by the original report by Mattson et al. [21], in which the incorporation of fluorine atoms in the molecule increased stability almost five times compared to the unsubstituted compound (Figure 2).

Figure 1.
Effect of the chlorine atom on in vitro metabolic stability-two additional chlorine atoms (B) increase metabolic stability versus unsubstituted compound (A). % remaining: quantity of the compound that remained after incubation with mouse liver microsomes (MLMs) [15].

Figure 2.
Effect of the fluorine atom on metabolic stability. % remaining: quantity of the compound that remained after incubation with human liver microsomes (HLMs) [21].  Figure 1. Effect of the chlorine atom on in vitro metabolic stability-two additional chlorine atoms (B) increase metabolic stability versus unsubstituted compound (A). % remaining: quantity of the compound that remained after incubation with mouse liver microsomes (MLMs) [15].
particularly the fluorine atom [18,19]. We showed [20] that the compounds tested without any ring substituent ( Figure 1A) had low in vitro metabolic stability. By incorporating chlorine in the structure [15] ( Figure 1B), metabolic stability increased with a slight decrease in cytotoxicity in the HepG2 cell line. This correlation is also confirmed by the original report by Mattson et al. [21], in which the incorporation of fluorine atoms in the molecule increased stability almost five times compared to the unsubstituted compound (Figure 2).

Figure 1.
Effect of the chlorine atom on in vitro metabolic stability-two additional chlorine atoms (B) increase metabolic stability versus unsubstituted compound (A). % remaining: quantity of the compound that remained after incubation with mouse liver microsomes (MLMs) [15].

Figure 2.
Effect of the fluorine atom on metabolic stability. % remaining: quantity of the compound that remained after incubation with human liver microsomes (HLMs) [21]. Effect of the fluorine atom on metabolic stability. % remaining: quantity of the compound that remained after incubation with human liver microsomes (HLMs) [21].
In our previous paper [15], we also investigated the effect of substituents (ortho-OMe and 2,3-Cl 2 ) at the phenyl ring attached to the aminoethyl chain on affinity toward 5-HT 7 R. The substituents were found to reduce receptor affinity, but the effect of a specific substituent position on activity was not tested. To investigate this aspect, we decided to synthesize a group of ligands 7-15 (Scheme 1, type 2) containing Cl, F, and OMe at positions ortho-, meta-, and paraand to evaluate their effects on affinity toward 5-HT 7 R. We also decided to incorporate more complex substituents into the structures of studied compounds (phthalimide 16 and benzimidazole 17-19 cores, Scheme 1, type 3) due to the potential formation of hydrogen bonds to stabilize the ligand-receptor additionally. Unexpectedly, we prepared compounds 20-22 without the aminoethyl chain in our synthesis experiments (Scheme 1, type 4). To determine the effect of the lack of an alkyl linker between the triazine and the aromatic system on affinity to the receptors in question, we also evaluated these compounds in in vitro tests. All the resulting compounds were tested in an extended receptor panel, including affinity toward 5-HT 1A , 5-HT 2A , 5-HT 6 , and D 2 receptors, to determine the selectivity of the compounds. Bioconformation and key interactions involved in the forming the ligand-receptor complex were proposed for active structures. The two best compounds were evaluated in terms of safety and bioavailability in in vitro ADME-Tox tests.

Chemistry
Final type 1 compounds were synthesized according to Scheme 2, starting from commercially available 5-substituted indoles 23-28, converted to 3-substituted aldehydes 29-34 via Vilsmeier-Haack formylation with 90-100% yields. In a subsequent stage, the aldehydes were subjected to the Henry reaction to obtain nitrovinyl derivatives 35-40. The synthesis of the derivatives initially followed patent [22] under conventional reflux of the reaction mixture. As long as yields of more than 90% were obtained in a small scale of 0.5-1 g, side products formed in the reaction mixture (according to TLC) with low yields of 40-56% when the scale was increased to 5 g. However, we found that performing the reactions under microwave irradiation (P = 85 W) for 20 min allows obtaining desired products in more than 90% yields, irrespective of the scale. Crude 35-40 were used in the next stage without any further purification. As for compound 35, another two stages involved reduction: first of the double bond using a mild reducing agent (NaBH 4 ) to give 41 and subsequently reduction of the nitro group with zinc in boiling 36% HCl solution, to finally give compound 47. As compounds 36-40 lacked the nitrile substituent, which could also be reduced in harsher conditions, simultaneous reduction of the double bond and the nitro group could be performed in the presence of LiAlH 4 to give final 5-substituted tryptamines 42-46, respectively. The subsequent stage was the synthesis of core compounds 51. Cyanuric chloride 48 was reacted with phenylethylamine 49 at 0 • C, and the resulting product 50 was treated with ammonia water. Synthesis was conducted at room temperature for 5 h, and core compound 51 was obtained in 86% yield, which further reacted with 5-substituted tryptamines 42-47 to give final type 1 products with a yield of more than 50%. The synthesis of compounds 1-6 was conducted under microwave irradiation, similar to our previous reports [15].
40-56% when the scale was increased to 5 g. However, we found that performing the reactions under microwave irradiation (P = 85 W) for 20 min allows obtaining desired products in more than 90% yields, irrespective of the scale. Crude 35-40 were used in the next stage without any further purification. As for compound 35, another two stages involved reduction: first of the double bond using a mild reducing agent (NaBH4) to give 41 and subsequently reduction of the nitro group with zinc in boiling 36% HCl solution, to finally give compound 47. As compounds 36-40 lacked the nitrile substituent, which could also be reduced in harsher conditions, simultaneous reduction of the double bond and the nitro group could be performed in the presence of LiAlH4 to give final 5-substituted tryptamines 42-46, respectively. The subsequent stage was the synthesis of core compounds 51. Cyanuric chloride 48 was reacted with phenylethylamine 49 at 0 °C, and the resulting product 50 was treated with ammonia water. Synthesis was conducted at room temperature for 5 h, and core compound 51 was obtained in 86% yield, which further reacted with 5-substituted tryptamines 42-47 to give final type 1 products with a yield of more than 50%. The synthesis of compounds 1-6 was conducted under microwave irradiation, similar to our previous reports [15].

Scheme 2.
Synthesis pathway for compounds of type 1 1-6. i-DMF, POCl3, O °C then rt., 1 h; iiammonia acetate, nitromethane, MW, 20 min; iii-LiAlH4, THF, rt, 3 days; iv-NaBH4, MeOH/DMF, rt, 3 h; v-Zn, HCl, reflux, 5 h; vi-DIPEA, THF, 0 °C, 2 h; vii-ammonia solution 20%, rt, 5 h; viii-K2CO3, TBAB, DMF, MW, 2.5 min. Type 2 compounds were synthesized according to Scheme 3. First, amines 61-68 were obtained in the reaction of 2-chloroethylamine hydrochloride 52 and appropriately substituted aniline 53-60. Subsequently, the resulting compounds reacted with readily available [20] core compound 69 in the presence of K2CO3 and microwave irradiation (p = 50 W) for 2.5 min [15]. Final type 2 compounds were isolated with yields of 43-75%. In spite of the complete conversion of substrate 69 (according to TLC), compounds 8, 12, and 15 were not obtained. However, it was found that the ethyl bridge was probably eliminated during the reaction (see the mass spectra, Supplementary Materials) to give compounds 20-22 (Scheme 1, type 4) in which aniline derivatives were attached directly to the triazine system (Scheme 4). The resulting products were easily isolated during work-up as white precipitates, which did not require purification. Their structures were confirmed using spectroscopy: 1 H NMR, 13  Type 2 compounds were synthesized according to Scheme 3. First, amines 61-68 were obtained in the reaction of 2-chloroethylamine hydrochloride 52 and appropriately substituted aniline 53-60. Subsequently, the resulting compounds reacted with readily available [20] core compound 69 in the presence of K 2 CO 3 and microwave irradiation (p = 50 W) for 2.5 min [15]. Final type 2 compounds were isolated with yields of 43-75%. In spite of the complete conversion of substrate 69 (according to TLC), compounds 8, 12, and 15 were not obtained. However, it was found that the ethyl bridge was probably eliminated during the reaction (see the mass spectra, Supplementary Materials) to give compounds 20-22 (Scheme 1, type 4) in which aniline derivatives were attached directly to the triazine system (Scheme 4). The resulting products were easily isolated during work-up as white precipitates, which did not require purification. Their structures were confirmed using spectroscopy: 1 H NMR, 13 C NMR, and MS. To confirm our hypothesis, we decided to synthesize the selected compound 21 starting from intermediate 69 and para-fluoroaniline (Scheme 5). The reaction was conducted similarly to the previous one in the presence of potassium carbonate (3 eq.) and sodium carbonate (3 eq.) to give title compound 21. According to HPLC-MS analysis, the content of the desired product in the reaction mixture was 23% for potassium carbonate and 73% for sodium carbonate (see Supplementary Materials). Desired compounds 8, 12, and 15 were obtained with a yield of more or equal to 60% using the same synthesis method while only changing the base to sodium carbonate. potassium carbonate (3 eq.) and sodium carbonate (3 eq.) to give title compound 21. According to HPLC-MS analysis, the content of the desired product in the reaction mixture was 23% for potassium carbonate and 73% for sodium carbonate (see Supplementary Materials). Desired compounds 8, 12, and 15 were obtained with a yield of more or equal to 60% using the same synthesis method while only changing the base to sodium carbonate.    potassium carbonate (3 eq.) and sodium carbonate (3 eq.) to give title compound 21. According to HPLC-MS analysis, the content of the desired product in the reaction mixture was 23% for potassium carbonate and 73% for sodium carbonate (see Supplementary Materials). Desired compounds 8, 12, and 15 were obtained with a yield of more or equal to 60% using the same synthesis method while only changing the base to sodium carbonate.  The synthesis of type 3 compounds (Scheme 6) started with protecting 2-chloroethylamine hydrochloride 52 with a Boc group, followed by coupling of the resulting product 70 with phthalimide 71 or benzimidazole 72 and its derivatives 73 and 74. The reactions were performed under microwave irradiation (P = 60 W) for 60 s in the presence of sodium hydroxide and TBAB (tetrabutylammonium bromide). Isolated products 75-78 reacted with 69 under microwave irradiation (P = 50 W) in the presence of K2CO3 and TBAB to give final compounds 15-18 with yields of 25-56%.  potassium carbonate (3 eq.) and sodium carbonate (3 eq.) to give title compound 21. According to HPLC-MS analysis, the content of the desired product in the reaction mixture was 23% for potassium carbonate and 73% for sodium carbonate (see Supplementary Materials). Desired compounds 8, 12, and 15 were obtained with a yield of more or equal to 60% using the same synthesis method while only changing the base to sodium carbonate.  The synthesis of type 3 compounds (Scheme 6) started with protecting 2-chloroethylamine hydrochloride 52 with a Boc group, followed by coupling of the resulting product 70 with phthalimide 71 or benzimidazole 72 and its derivatives 73 and 74. The reactions were performed under microwave irradiation (P = 60 W) for 60 s in the presence of sodium hydroxide and TBAB (tetrabutylammonium bromide). Isolated products 75-78 reacted with 69 under microwave irradiation (P = 50 W) in the presence of K2CO3 and TBAB to give final compounds 15-18 with yields of 25-56%.  potassium carbonate (3 eq.) and sodium carbonate (3 eq.) to give title compound 21. According to HPLC-MS analysis, the content of the desired product in the reaction mixture was 23% for potassium carbonate and 73% for sodium carbonate (see Supplementary Materials). Desired compounds 8, 12, and 15 were obtained with a yield of more or equal to 60% using the same synthesis method while only changing the base to sodium carbonate. The synthesis of type 3 compounds (Scheme 6) started with protecting 2-chloroethylamine hydrochloride 52 with a Boc group, followed by coupling of the resulting product 70 with phthalimide 71 or benzimidazole 72 and its derivatives 73 and 74. The reactions were performed under microwave irradiation (P = 60 W) for 60 s in the presence of sodium hydroxide and TBAB (tetrabutylammonium bromide). Isolated products 75-78 reacted with 69 under microwave irradiation (P = 50 W) in the presence of K2CO3 and TBAB to give final compounds 15-18 with yields of 25-56%.
The substituent at indole position 5 in the A ring (type 1 compound) had no significant effect on increased binding to the 5-HT 7 receptor. The ligand with a weakly deactivating fluorine substituent 2 was found to be the most active among the tested type 1 compounds and showed an affinity of K i = 8 nM toward 5-HT 7 R. In addition, in terms of the effect of the halogen atom, ligand 3 with bromine (5-HT 7 R K i = 126 nM) had moderate activity, followed by ligand 4 with chlorine (5-HT 7 R, K i = 481 nM). Indole substitution at position 5 with CN, a strongly deactivating substituent (compound 1), resulted in activity toward the 5-HT 7 receptor being lost. As for two activating substituents: Me (weakly activating) and OMe (strongly activating), moderate activity toward the 5-HT 7 receptor was shown for ligand 6 (K i = 100 nM), while ligand 5 had a low activity with K i = 629 nM. As for type 1 compounds, a halogen substituent at the B ring resulted in active (9,11,12) or moderately active ligands (7,8,10) with respect to 5-HT 7 R, while a slightly larger substituent (OMe) resulted in the loss of activity. When analyzing the effect of the substituted position, ligands in the para position had the highest activity, and the ortho position was the least active. It is concluded based on analysis of the data that the fluorine substituent (para > meta > ortho) is the strongest, followed by chlorine (para > meta > ortho) and finally methoxy (para > meta > ortho). For example, para-F (18) and para-Cl (9), with K i values of 18 nM and 19 nM, respectively, were found to be the most active ligands. Isomeric meta-F (11) and meta-Cl (8) compounds had slightly lower activity compared to the previous ones, with K i values of 24 nM and 131 nM, respectively. Even though the compounds with the OMe substituent were inactive, a tendency for a relatively stronger para position compared to the weaker meta and, finally, the weakest ortho position was also seen in this group. As for type 3 ligands, incorporation of larger substituents R 3 resulted in most cases in the loss of activity (K i > 1000 nM) toward 5-HT 7 R (ligands 16, 18, and 19). Only the ligand with an unsubstituted benzimidazole ring (17) had a moderate activity with K i = 227 nM. The type 4 compounds are also found to be inactive (20,21) or weakly active (22), but they provide further evidence of the significant effect of the linker between the triazine core and the aromatic system [20]. When analyzing affinity to the other receptors tested (5-HT 1A , 5-HT 2A , 5-HT 6, and D 2 ), most of the designed compounds have no activity (K i > 1000 nM) or low activity (500 nM < K i < 1000 nM).

Atlas Activity Analysis: 3D-SAR
To better understand SAR, we determined specific activity maps as a function of shape, hydrophobicity, and electrostatics using Activity Atlas (Flare, Cresset) [24]. According to the ligand shape, the incorporation of large substituents at indole position 5 (A ring) results in reduced activity toward 5-HT 7 R. A similar effect is found for substituents at the alkylaromatic ring (B ring). It can be concluded based on Figure 3B,D that large substituents (ligands 13-19, K i > 226 nM) cannot be accommodated in the receptor binding pocket, resulting in the loss of activity as shown by an increased number of steric clashes found between the ligand and the binding pocket. Ligands 7-12 with small substituents such as chlorine or fluorine show a much better fit to the binding pocket (lower number of steric clashes) as shown by their high or moderate activity to the 5-HT 7 receptor ( Figure 3A,C). It is noted that no steric clashes were found between the substituent and the receptor binding pocket for the most active type 2 ligand 12 (K i = 18 nM). This is due to the fact that ligands at the para positions are most preferred with the best fit to the receptor.
The substituent at indole position 5 in the A ring (type 1 compound) had no significant effect on increased binding to the 5-HT7 receptor. The ligand with a weakly deactivating fluorine substituent 2 was found to be the most active among the tested type 1 com-
The substituent at indole position 5 in the A ring (type 1 compound) had no significant effect on increased binding to the 5-HT7 receptor. The ligand with a weakly deactivating fluorine substituent 2 was found to be the most active among the tested type 1 com-
The substituent at indole position 5 in the A ring (type 1 compound) had no significant effect on increased binding to the 5-HT7 receptor. The ligand with a weakly deactivating fluorine substituent 2 was found to be the most active among the tested type 1 com-
The substituent at indole position 5 in the A ring (type 1 compound) had no significant effect on increased binding to the 5-HT7 receptor. The ligand with a weakly deactivating fluorine substituent 2 was found to be the most active among the tested type 1 com-
The substituent at indole position 5 in the A ring (type 1 compound) had no significant effect on increased binding to the 5-HT7 receptor. The ligand with a weakly deactivating fluorine substituent 2 was found to be the most active among the tested type 1 com-  According to ligand hydrophobicity, type 2 ligands (ligands 7-12) show the best fit to the favorable hydrophobic region (squared green) due to the presence of a hydrophobic aromatic ring which may contribute to π-π interactions (B ring, Figure 4A). The unfavorable hydrophobic region (squared magenta) is not occupied by hydrophobic substituents. Type 3 ligands (ligands 16-19) are a different case. The favorable hydrophobic region is occupied by hydrophilic parts of the substituents (imide and imidazole systems), while the unfavorable hydrophobic region is occupied by the hydrophobic aromatic ring ( Figure  4B). The effect of the hydrophobic/hydrophilic properties of substituents at indole position 5 (A ring) for type 1 ligands was difficult to determine based on our studies. According to ligand hydrophobicity, type 2 ligands (ligands 7-12) show the best fit to the favorable hydrophobic region (squared green) due to the presence of a hydrophobic aromatic ring which may contribute to π-π interactions (B ring, Figure 4A). The unfavorable hydrophobic region (squared magenta) is not occupied by hydrophobic substituents. Type 3 ligands (ligands 16-19) are a different case. The favorable hydrophobic region is occupied by hydrophilic parts of the substituents (imide and imidazole systems), while the unfavorable hydrophobic region is occupied by the hydrophobic aromatic ring ( Figure 4B). The effect of the hydrophobic/hydrophilic properties of substituents at indole position 5 (A ring) for type 1 ligands was difficult to determine based on our studies.

Metabolic Stability
Highly active compounds 2 and 12 were submitted for metabolic stability evaluation tests using mouse liver microsomes (MLMs). UHPLC-MS analysis of the samples after compounds 2 and 12 were incubated for 2 h in the presence of MLMs showed that 20% and 29% of the original compound remained, respectively. Table 2 shows potential metabolic pathways for the tested compounds as well as Figure 6 shows possible main metab-  Figure 5. The left side represents the binding mode of ligand 2 (blue) and ligand 12 (yellow) in a homologous model of the 5-HT 7 receptor. Blue amino acids form hydrogen bonds, and purple amino acids represent π-π stacking hydrophobic interactions. The right side represents ligand-protein interactions.

Metabolic Stability
Highly active compounds 2 and 12 were submitted for metabolic stability evaluation tests using mouse liver microsomes (MLMs). UHPLC-MS analysis of the samples after compounds 2 and 12 were incubated for 2 h in the presence of MLMs showed that 20% and 29% of the original compound remained, respectively. Table 2 shows potential metabolic pathways for the tested compounds as well as Figure 6 shows possible main metabolites predicted by Metasite 6.0.1. It was found, based on the results, that the presence of fluorine at indole position C-5 (ligand 2) did not have a significant effect on increased metabolic stability compared to the unsubstituted compound [20]. It seems, however, that substituents at the B ring of the ligands have a more important role. Compound 12, containing a fluorine atom, was found to be more stable than compound 2 and the reference verapamil. The data correspond to our previous results [15], in which substituents at the B ring increased metabolic stability. Potential sites in the molecules sensitive to enzyme activity and leading to degradation were proposed using an in silico approach, also with MetaSite 6.0.1 (Figure 7).   [27].

CYP3A4 Interaction
Potential drug-drug interactions (DDIs) are an important aspect that should be considered when designing new compounds. Isoenzyme CYP3A4 is one of the varieties of enzymes responsible for xenobiotic metabolism [28]. We evaluated active compounds 2 and 12 in terms of affinity toward this isoenzyme (Figure 8). Both compounds at concentrations identical to that of ketoconazole, the reference compound (1 µM) or lower were shown to have no or slight inhibition activity toward CYP3A4. Compound 2 at 10 µM showed moderate (%CYP3A4 activity = 50) and compound 12 showed high (%CYP3A4 activity = 20) inhibition effect, respectively. A very high inhibition effect was observed for both compounds at 25 µM (%CYP3A4 activity < 15).

CYP3A4 Interaction
Potential drug-drug interactions (DDIs) are an important aspect that should be considered when designing new compounds. Isoenzyme CYP3A4 is one of the varieties of enzymes responsible for xenobiotic metabolism [28]. We evaluated active compounds 2 and 12 in terms of affinity toward this isoenzyme ( Figure 8). Both compounds at concentrations identical to that of ketoconazole, the reference compound (1 µM) or lower were shown to have no or slight inhibition activity toward CYP3A4. Compound 2 at 10 µM showed moderate (%CYP3A4 activity = 50) and compound 12 showed high (%CYP3A4 activity = 20) inhibition effect, respectively. A very high inhibition effect was observed for both compounds at 25 µM (%CYP3A4 activity < 15).

Hepatotoxicity
The tested compounds 2 and 12 were evaluated in terms of cytotoxicity against the HepG2 cell line to assess their hepatotoxic potential (Figure 9). It was an interesting finding that both compounds had proliferative activity in lower concentrations (<10 µM). A cytotoxic effect appeared only at 50 µM (2, % cell viability = 35; 12, % cell viability = 20); at 100 µM, the cells were practically no viable.

Hepatotoxicity
The tested compounds 2 and 12 were evaluated in terms of cytotoxicity against the HepG2 cell line to assess their hepatotoxic potential (Figure 9). It was an interesting finding that both compounds had proliferative activity in lower concentrations (<10 µM). A cytotoxic effect appeared only at 50 µM (2, % cell viability = 35; 12, % cell viability = 20); at 100 µM, the cells were practically no viable.

In Vivo Cardiotoxicity
The compounds were evaluated for their ecotoxicity on the Danio rerio experimental model. OECD 236 test [29] with modifications was applied. Both compounds were found cardiotoxic within the non-toxic range, based on the heart rate measurement. For the compound 2, 5 µg/mL was found cardiotoxic ( Figure 10) and for 12, 7.5 µg/mL ( Figure 11). The results were confirmed by malformations observation. Pericardial edema (PE) was observed at 5 µg/mL for 2 ( Figure 12) and at 7.5 µg/mL for 12 ( Figure 13). Moreover, scoliosis (S) and tail autophagy (TA) were also noted.
way ANOVA and Bonferroni's Multiple Comparison Post Test in comparison with the negative control (100% of CYP3A4 activity). The compounds were examined in triplicate.

Hepatotoxicity
The tested compounds 2 and 12 were evaluated in terms of cytotoxicity against the HepG2 cell line to assess their hepatotoxic potential (Figure 9). It was an interesting finding that both compounds had proliferative activity in lower concentrations (<10 µM). A cytotoxic effect appeared only at 50 µM (2, % cell viability = 35; 12, % cell viability = 20); at 100 µM, the cells were practically no viable. Figure 9. The effect of cytostatic drug doxorubicin and 2 (left), 12 (right) on hepatoma HepG2 cell line viability after 72 h of incubation at 37°, 5% CO2. The statistical significance (GraphPad Prism 8.0.1) was evaluated by a one-way ANOVA, followed by Bonferroni's Comparison Test (* p < 0.05, ** p < 0.01, and **** p < 0.0001 compared with negative control DMSO 1% in growth media.

In Vivo Cardiotoxicity
The compounds were evaluated for their ecotoxicity on the Danio rerio experimental model. OECD 236 test [29] with modifications was applied. Both compounds were found cardiotoxic within the non-toxic range, based on the heart rate measurement. For the compound 2, 5 µg/mL was found cardiotoxic ( Figure 10) and for 12, 7.5 µg/mL ( Figure 11). The results were confirmed by malformations observation. Pericardial edema (PE) was observed at 5 µg/mL for 2 ( Figure 12) and at 7.5 µg/mL for 12 ( Figure 13). Moreover, scoliosis (S) and tail autophagy (TA) were also noted.

In Vivo Cardiotoxicity
The compounds were evaluated for their ecotoxicity on the Danio rerio experimental model. OECD 236 test [29] with modifications was applied. Both compounds were found cardiotoxic within the non-toxic range, based on the heart rate measurement. For the compound 2, 5 µg/mL was found cardiotoxic ( Figure 10) and for 12, 7.5 µg/mL ( Figure 11). The results were confirmed by malformations observation. Pericardial edema (PE) was observed at 5 µg/mL for 2 ( Figure 12) and at 7.5 µg/mL for 12 ( Figure 13). Moreover, scoliosis (S) and tail autophagy (TA) were also noted.

Discussion and Conclusions
In spite of the significant functions of the 5-HT 7 receptor [30] both in and outside the central nervous system and the recent progress in medicinal chemistry and pharmacology, a drug having selectivity toward 5-HT 7 R is yet to be fully developed [30]. Taking this into consideration, the objective of this paper was to design ligands showing high activity and selectivity toward the 5-HT 7 receptor without incorporating the arylpiperazine pharmacophore, which may increase affinity to other aminergic GPCRs.
All finally synthesized compounds were obtained by a condensation reaction supported by microwave irradiation for 2.5 min with a yield of more than 50%. An interesting fact turned out to be the synthesis of ligands 8, 12, and 15. So far, working on 1,3,5-triazines with indole motif [15,20], we have successfully used a mild alkaline agent, which was potassium carbonate, obtaining final products with medium or high yield. In the case of the mentioned ligands, the reactions did not proceed as it was expected, leading to the elimination of the ethyl bridge. The usage of a slightly weaker base, which was sodium carbonate, resulted in obtaining the desired products with good yield. It was found when studying type 1 compounds that incorporation of EWG or EDG substituents in most cases did not improve affinity toward the 5-HT 7 receptor in the majority of synthesized compounds. Compounds 1 and 3-6 had lower activity than the unsubstituted compound ( Figure 1A). Ligand 2, whose activity did not change compared to the unsubstituted compound, was an exception ( Figure 1A). The fact may be accounted for by the effect of bioisosterism of the fluorine atom (hydrogen bioisostere). When exploring the aromatic region of the B ring, ligands with chlorine, fluorine, and methoxy substituents in the ortho position showed the lowest activity; higher activity was shown for the substituents in the meta position and the highest for substituents in the para position (ligand 12, K i = 18 nM). It is noted that ligands with small substituents (Cl, F) were more active than those with slightly larger substituents (OMe). The incorporation of much larger heterocyclic compounds (type 3) resulted in the loss of activity on the 5-HT 7 receptor. Type 4 compounds showed once more that the distance between the triazine core and the aromatic system, which should be two or three atoms, was crucial [15,20]. The SARs for the resulting compounds were supported using 3D-QSAR computed methods. The bioconformation and the binding mode for the two best compounds (2 and 12) were determined using molecular modeling (docking), and the results were consistent with our previous reports [15,20]. The tested ligands (2 and 12) are 5-HT 7 receptor antagonists, have moderate metabolic activity (higher or similar to verapamil and still higher than unsubstituted ligands which were described in our publication [20]), and moderate or weak potential drug-drug interactions with respect to ketoconazole. As for hepatotoxicity, both compounds at more than 50 µM showed cytotoxicity against the HepG2 cell line. The ecotoxicity tests with the use of Daio rerio as a model organism turned out 2 to be more toxic than 12. Moreover, cardiotoxicity expressed as heart rate abnormalities was observed at higher doses for 2 than 12, and it suggests that cardiotoxic potential needs to be reduced in the future.

General
All primary substrates were purchased commercially from Sigma-Aldrich (St. Louis, MO, USA). The solvents used for column chromatography (purchased from Merck, Kenilworth, NJ, USA), thin layer chromatography (TLC), and preparative thin layer chromatography (pTLC) had purity above 99.5%. 1 H and 13 C NMR spectra were recorded using Bruker 400 MHz systems with TMS as an internal standard. Melting points were determined with the Böetius apparatus. HPLC-MS analyses were performed on the Shimadzu Nexera XR system equipped with PDA (SPD-M40) and LCMS-2020 detectors. Analyses were performed on Phenomenex XB-C18 1.7 µm (50 × 2.1 mm) (method A) column with gradient of solvents as a mobile phase: Solvent A (0.01% HCOOH in water) and B (0.01% HCOOH in methanol); t = 0 min, 10% of B, t = 4 min, 90% of B, t = 6 min, 90% of B, t = 6.1 min 10% of B, stop time 11 min or Phenomenex C18 1.7 µm (50 × 2.1 mm) (method B) column with gradient of solvents as a mobile phase: solvent A (0.01% HCOOH in water) and solvent B (0.01% HCOOH in MeOH): t = 0 min 5% of B, t = 3 min 90% of B, t = 4 min 90% of B, t = 4.5 min 5% of B stop time 7 min; flow rate 0.4 mL min −1 ; the UV-VIS detection was performed in a range of 240-700 nm, the MS data were collected in ESI + mode in a range of m/z 100-800 with scan speed 15,000 u/s and event time 0.1 s. Analytical thinlayer chromatography (TLC) was performed using 0.2 mm silica gel precoated aluminum sheets (60 F254, Merck), and UV light at 254 nm was used for visualization. Preparative thin-layer chromatography (pTLC) was performed using 2000 µm silica gel precoated glass backed (F254, Silicycle). A CEM Discover™ Focused Microwave System at 50 W power was used for all microwave-assisted reactions in order to obtain final compounds. Within 2.5 min of reaction with a power of 50 W, the temperature increased up to 120 • C, while the pressure increased up to 9 bar. Characterization of the intermediates and spectra for the final compounds can be found in Supporting Information.

General Procedure for the Synthesis of Compounds 29-34
12 mL of DMF was cooled to 0 • C, and phosphoryl chloride (38.5 mmol) was added dropwise. In this temperature solution of commercially available indoles 23-28 (35.5 mmol) in 3 mL of DMF was added dropwise, and the resulting mixture was stirred at room temperature for one hour. The reaction became a thick, pale suspension. Sodium hydroxide solution (10%, 40 mL) was added slowly to the reaction mixture (pH = 13-14), followed by precipitation of pale solid 29-34. The solid was filtered, rinsed with distilled water, and dried.

General Procedure for the Synthesis of Compounds 35-40
Intermediate 29-34 (23.5 mmol) was placed in a one-necked round bottom flask and dissolved in 60 mL nitromethane, and then, ammonia acetate (44.2 mmol) was added. The mixture was reacted in a microwave reactor at 80 W (100 • C) for 20 min. Reaction progress was monitored via TLC (hexane:EtOAc 1:1 v/v). After this time, TLC indicated full conversion of starting material, and the mixture was cooled to room temperature with precipitation of yellow solid 35-40. The solid was filtered, rinsed with distilled water, and dried. Intermediate 35 (4.1 g, 19.2 mmol) was dissolved in the 330 mL mixture of DMF:MeOH (1:1, v/v), followed by the addition of sodium borohydride (8 g, 21.1 mmol). The mixture was stirred at room temperature for 5 h. Reaction progress was monitored via TLC (hexane:EtOAc 1:1 v/v). After reaction completion reaction was diluted with 2M HCl to reach pH = 7. The solvent was reduced and extracted with EtOAc (3 × 150 mL). The crude product was triturated with a mixture of MeOH:chloroform to yield 2.17 g of the titled compound. The mother liquor was concentrated to dryness and purified at column chromatography eluted with hexane:EtOAc (v/v) 8:2 -> 6:4 to yield 0.8 g of the titled compound. Lithium aluminum hydride (25.9 mmol) was placed in a three-necked round bottom flask, followed by the addition of 20 mL dry THF. The resulting suspension was cooled to 0 • C, and mixture of intermediates 36-40 (4.7 mmol) in 20 mL THF was added dropwise. The mixture was stirred at room temperature for 72 h and then quenched with slow addition of mixture H 2 O:MeOH (9:1, v/v). The suspension was filtered by a Celite bed, and pHdependent extraction was performed: filtrate was acidified to reach pH 2-3 with 1 M HCl, followed by extraction with EtOAc (3 × 100 mL). The water layer was alkalized to reach pH 10 with 1 M NaOH and then extracted with EtOAc (3 × 100 mL). Organic layers were combined, dried over MgSO 4 , and concentrated to yield brown, sticky oil 42-46.

Synthesis of 3-(2-Aminoethyl)-1H-Indole-5-Carbonitrile (47)
A solution of 41 (1.5 g, 6.9 mmol) in 205 mL MeOH was added to the solution of zinc (10.4 g, 0.16 mol) in 205 mL 2M HCl and refluxed for 1.5 h. After this time mixture was cooled to room temperature and filtered. The filtrate was alkalized to pH 12, and MeOH was removed under reduced pressure. The resulting mixture was extracted with EtOAc (3 × 200 mL), and organic layers were combined, dried over MgSO 4 then evaporated to dryness. It obtained 0.87 g (67% yield) of titled compound 47, which was used in the next step without any further purification. To a solution of cynuric chloride 48 (8.35 g, 45.2 mmol) in 100 mL THF cooled to 0 • C, a solution of phenylethylamine 49 (5 g, 41.2 mmol) in 5 mL THF was added dropwise. The reaction was carried out at 0-3 • C for 2 h. The resulting precipitate was filtered, and the filtrate was diluted with 0.1 M HCl and extracted with chloroform (3 × 100 mL). Organic layers were combined, dried over MgSO 4 , and concentrated to yield a brown solid. The solid was triturated with acetone and then filtered. The black filtrate was purified using column chromatography eluted with hexane:EtOAc (v/v) 9:1-> 6:4 to yield 3.74 g of the titled compound. Creamy solid (34% yield); method B: ESI-MS calc. for C 11  Briefly, 50 (3.5 g, 13.0 mmol) was dissolved in 50 mL acetone, followed by the addition of 5.7 mL 25% ammonia solution. The reaction was carried out at room temperature for 5 h. The resulting precipitate was filtered, and the filtrate evaporated to dryness, yielding 2.8 g of titled compound 51. Creamy solid (86% yield); ESI-MS calc. for C 11  Briefly, 51 (0.25 g, 1.0 mmol), potassium carbonate (0.41 g, 3.0 mmol), and TBAB (0.032 g, 0.1 mmol) were ground in a mortar and transferred to a sealed tube which was previously charged with appropriate amine 42-47 (2.5 mmol). Subsequently, 5 wt % DMF was added. The mixture was reacted in a microwave reactor at 50 W for 2.5 min. Reaction progress was monitored via TLC (chloroform: MeOH 9:1 v/v). The mixture was cooled down and extracted with chloroform (3 × 20 mL). Organic layers were combined, dried over MgSO 4 , and concentrated. The crude product was purified via column chromatography with elution using chloroform then chloroform:MeOH (v/v) 99:1-> 97:3. The white or beige sticky oil was then dissolved in acetone and pH was adjusted to 2-3 with 4 M HCl in 1,4-dioxane. The resulting mixture was crushed by the addition of cold diethyl ether, then the white or beige powder was filtered and rinsed with cold diethyl ether and then dried to yield final product 1-6. In a round bottom flask, 2-chloroethanamine hydrochloride 52 (1.0 g, 8.6 mmol) was suspended in 8 mL of toluene. To the resulting mixture, appropriate aniline 53-60 (51.7 mmol) was added, and the mixture was refluxed for 20 h. After this period, toluene was evaporated, and residues were triturated with dichloromethane to yield solid as titled compounds 61-68 (optionally solid may be washed with diethyl ether). 4.1.17. General Procedure for the Synthesis of Final Compounds 7, 9-11, and 14 (microwave-assisted) Briefly, 69 [20] (0.25 g, 0.8 mmol), amines 61, 63-65, 67 (2.0 mmol) potassium carbonate (0.36 g, 2.5 mmol) and TBAB (0.032 g, 0.1 mmol) were ground in a mortar and transferred to a sealed tube. Subsequently, 5 wt % DMF was added. The mixture was reacted in a microwave reactor at 50 W for 2.5 min. Reaction progress was monitored via TLC (chloroform: MeOH 9:1 v/v). The mixture was cooled down and extracted with chloroform (3 × 20 mL). Organic layers were combined, dried over MgSO 4 , and concentrated. The crude product was purified via column chromatography with elution using chloroform then chloroform:MeOH (v/v) 99:1-> 97:3. Colorless sticky oil was then dissolved in acetone, and pH was adjusted to 2-3 with 4 M HCl in 1,4-dioxane. The resulting mixture was crushed by the addition of cold diethyl ether, and then the white or beige powder was filtered and rinsed with cold diethyl ether and then dried to yield final products 7, 9-11, and 14.