Scaffold Hybridization Strategy Leads to the Discovery of Dopamine D₃ Receptor-Selective or Multitarget Bitopic Ligands Potentially Useful for Central Nervous System Disorders

Abstract

In the search for novel bitopic compounds targeting the dopamine D₃ receptor (D₃R), the N-(2,3-dichlorophenyl)piperazine nucleus (primary pharmacophore) has been linked to the 6,6- or 5,5-diphenyl-1,4-dioxane-2-carboxamide or the 1,4-benzodioxane-2-carboxamide scaffold (secondary pharmacophore) by an unsubstituted or 3-F-/3-OH-substituted butyl chain. This scaffold hybridization strategy led to the discovery of potent D₃R-selective or multitarget ligands potentially useful for central nervous system disorders. In particular, the 6,6-diphenyl-1,4-dioxane derivative 3 showed a D₃R-preferential profile, while an interesting multitarget behavior has been highlighted for the 5,5-diphenyl-1,4-dioxane and 1,4-benzodioxane derivatives 6 and 9, respectively, which displayed potent D₂R antagonism, 5-HT_1AR and D₄R agonism, as well as potent D₃R partial agonism. They also behaved as low-potency 5-HT_2AR antagonists and 5-HT_2CR partial agonists. Such a profile might be a promising starting point for the discovery of novel antipsychotic agents.

Introduction

Dopamine, a widely distributed neurotransmitter in both vertebrates and invertebrates, is a catecholamine produced in several areas of the brain. It plays several physiological roles by interacting with specific receptors, which are members of the G protein-coupled receptor (GPCR) superfamily and are classified into two main groups, namely, D₁-like, including D₁ and D₅ receptors (D₁Rs and D₅Rs), and D₂-like, comprising D₂, D₃, and D₄ receptors (D₂Rs, D₃Rs, and D₄Rs).¹ There is a high degree of amino acid homology within the binding sites of the D₂-like receptors, especially between the D₂R and D₃R subtypes.² Hence, the discovery of novel D₃R-selective compounds is still a challenge and, to date, no highly selective D₃R agonists or antagonists are available for therapeutic use in humans.³ Moreover, these receptors share high homology with other receptor systems, including serotonergic receptors.⁴

The discovery of D₃R-preferential compounds, the cloning of D₃R gene, and a better understanding of D₃R biology have made clear that D₃R pharmacology is deeply different from that of the other D₂-like receptor subtypes.⁵⁻⁷ The resolved crystal structure of D₃R in complex with the D₂R/D₃R-specific antagonist eticlopride has provided greater clarity on the molecular basis of ligand–receptor interactions and played an important role in the design of novel D₃R-selective ligands.^8,9

Altered D₃R signaling is associated with a number of pathological conditions, including substance use disorder, Parkinson’s disease (PD), schizophrenia, depression, and restless leg syndrome.¹⁰⁻¹⁷ Each of these conditions may benefit from pharmacological manipulation of D₃R signaling, and selective D₃R ligands may represent important pharmacological tools that might provide further information toward the D₃R (patho)physiological role and that lack motor side effects associated with D₂R blockade.¹⁷⁻²⁰

A promising strategy for improving D₃R selectivity involves the development of bitopic ligands that bear an aryl piperazine as the primary pharmacophore (PP) linked, via an alkyl chain of specified length and composition, to an arylcarboxamide as the secondary pharmacophore (SP) and has led to the discovery of potent and selective D₃R agents (Figure 1).²¹⁻²⁴

Figure 1.

General structure of selective D₃R bitopic ligands^10,24 and representative PPs (Ar₂ in red), SPs (Ar₁ in blue), and linkers.

Molecular modeling studies based on the D₃R crystal structure demonstrated that the PP recognizes the orthosteric binding site of the receptor, nicely overlapping with eticlopride, whereas the SP binds to the less conserved secondary binding pocket. Moreover, the unsubstituted or (3-OH or 3-F) substituted butyl linkers can favor (or disfavor) the correct binding pose of the ligand.²⁴

D₂-like multitarget ligands have also demonstrated their potential as therapeutically useful agents,²⁵ especially in PD and schizophrenia treatment.^26,27 In particular, combining D₂R/D₃R antagonism, 5-HT_1A receptor (5-HT_1AR) agonism, and 5-HT_2AR antagonism proved to be favorable in the management of schizophrenia.^5,28 Moreover, the simultaneous D₂-like receptor and 5-HT_1AR activation might be advantageous in PD therapy. In this case, the 5-HT_1AR stimulation might reduce the dyskinetic side effects induced by D₂-like receptor activation.^29,30

Over the last decade, we have demonstrated that the 1,4-dioxane nucleus represents a bioversatile carrier of ligands interacting with different receptor systems,³¹⁻⁴⁰ including D₂-like receptors.⁴¹ In particular, two properly substituted 1,4-dioxane compounds endowed with the fruitful multitarget combination of 5-HT_1AR/D₄R agonism and D₂R/D₃R/5-HT_2AR antagonism (compound 1) or D₂R/D₃R/D₄R/5-HT_1AR agonism (compound 2) have been discovered as potential starting points to develop new pharmacological tools for schizophrenia and PD therapy, respectively⁴¹ (Figure 2).

Figure 2.

Chemical structures of the multitarget 1,4-dioxane derivatives 1 and 2. Adapted from ref (41). Copyright 2019 ACS Chemical Neuroscience.

Based on these observations, the aim of the present investigation was to evaluate the utility of the substituted 1,4-dioxane scaffold as a SP of bitopic compounds targeting D₃R, in order to obtain new D₃R-selective or multitarget agents. Therefore, hybrid ligands bearing the N-(2,3-dichlorophenyl)piperazine nucleus, one of the most common PPs present in selective D₃R partial agonists and antagonists,²⁴ linked by an unsubstituted or 3-F-/3-OH-substituted butyl chain, to the 6,6- or 5,5-diphenyl-1,4-dioxane-2-carboxamide (compounds 3–8) or the 1,4-benzodioxane-2-carboxamide (compounds 9–11) scaffold as the SP, have been synthesized (Figure 3). All compounds were tested at human D₂-like receptor subtypes in radioligand competition binding assays. Moreover, the biological profiles of the most promising compounds 3, 6, and 9 were further evaluated in binding assays at other selected targets (D₁R, 5-HT_1AR, 5-HT_2AR, and 5-HT_2CR) and in functional assays at receptors in which they showed the highest affinities.

Figure 3.

Chemical structures of compounds 3–11 (SP = secondary pharmacophore, PP = primary pharmacophore).

Results and Discussion

Intermediate building blocks 12–16 were prepared according to the previously reported procedures.^32,42−44 The novel derivatives 3–11 were prepared according to Scheme 1 by amidation of acids 15 and 16(32) or the commercially available 17 with amines 12–14(42−44) in the presence of 1,1′-carbonyldiimidazole in THF. The diastereomers were separated by column chromatography (4a and 4b), fractional crystallization (10a and 10b), or preparative thin-layer chromatography (TLC) (11a and 11b) as described in the Methods section. Given the poor biological results of the separated diastereoisomers (Table 1), we did not attempt to assign their relative configuration.

Scheme 1. Reagents: (a) 1,1′-Carbonyldiimidazole, THF.

Table 1. Affinity Values (K_i)^a of 1–11 at Human D₂-like Receptor Subtypes and of 1–3, 6, 9, Clozapine, and Olanzapine at Human D₁R, 5-HT_1AR, 5-HT_2AR, and 5-HT_2CR.

	Ki ± SEM (nM), human cloned receptors
compd	D1R	D2R	D3R	D4R	D2/D3	D4/D3	5-HT1AR	5-HT2AR	5-HT2CR
1	22.9 ± 3.1b	12.2 ± 1.5b	13.1 ± 1.7b	8.34 ± 0.6b	0.9	0.6	0.66 ± 0.02b	52.5 ± 4.4b	1819 ± 165b
2	123 ± 11b	3.16 ± 1.0b	1.38 ± 0.07b	10.5 ± 0.1b	2	8	1.15 ± 0.03b	1412 ± 224b	9772 ± 501b
3	442 ± 52	342 ± 105	2.39 ± 0.69	1352 ± 449	143	566	288 ± 15	257 ± 38	642 ± 80
4a	NDd	39.5%c	164 ± 52	39.2%c	NDd	NDd	NDd	NDd	NDd
4b	NDd	41%c	77.1 ± 25.5	54.4%c	NDd	NDd	NDd	NDd	NDd
5	NDd	21.2%c	716 ± 178	19.1%c	NDd	NDd	NDd	NDd	NDd
6	700 ± 150	28.7 ± 5.4	1.58 ± 0.24	54.6 ± 7.3	18	35	9.1 ± 0.97	28 ± 6.0	90 ± 16
7	NDd	526 ± 28.2	23.4 ± 3.1	390 ± 46	23	17	NDd	NDd	NDd
8	NDd	1200 ± 210	295 ± 100	43.6%c	4	NDd	NDd	NDd	NDd
9	680 ± 160	46.4 ± 5.9	2.16 ± 0.63	74.7 ± 24.5	22	35	23.8 ± 5.2	62 ± 15	97 ± 25
10a	NDd	1097 ± 125	44.1 ± 8.2	282 ± 76	25	6	NDd	NDd	NDd
10b	NDd	783 ± 19	23.4 ± 2.4	244 ± 54	35	10	NDd	NDd	NDd
11a	NDd	10.6%c	54.8%c	29.4%c	NDd	NDd	NDd	NDd	NDd
11b	NDd	8.7%c	65.4%c	26.6%c	NDd	NDd	NDd	NDd	NDd
clozapinee	22.9e	135e	219e	46.8e	0.62e	0.21e	87e	4.07e	2.75e
olanzapinee	11.7e	21.4e	34.7e	17.8e	0.62e	0.51e	1514e	1.32e	3.89e

^aK_i values were determined by competitive inhibition of [³H]SCH23390 binding in mouse fibroblast cells stably expressing hD₁R; [³H]N-methylspiperone binding in HEK 293 cells stably expressing hD₂R, hD₃R, or hD₄R; [³H]8-OH-DPAT binding in HeLa cells stably expressing h5-HT_1AR; and [¹²⁵I]DOI binding in HEK cells stably expressing h5-HT_2AR or h5-HT_2CR.

^bTaken from ref (41).

^cFor low-affinity compounds, only the inhibition percentage of the radioligand binding at a test compound concentration of 10 μM is given. D₁R, 5-HT_1AR, 5-HT_2AR, and 5-HT_2cR data were obtained through the NIDA Addiction Treatment Discovery Program contract with Oregon Health & Science University.

^dND = not determined.

^eTaken from ref (48).

The novel hybrid derivatives 3–11 were tested at hD₂R, hD₃R, or hD₄R stably expressed in HEK293 cells by competition binding assays, using [³H]N-methylspiperone as the radioligand, to evaluate their D₂-like affinity and subtype selectivity, following the previously reported procedures.^41,45,46

The K_i values, calculated according to the Cheng–Prusoff equation,⁴⁷ are shown in Table 1 along with those of the previously reported compounds 1 and 2 and the antipsychotic agents clozapine and olanzapine.

The results reveal that all the new compounds 3–11 show higher affinity for D₃R with respect to D₂R and D₄R. The nature of both the linker and the SP plays a crucial role in the interaction with all the D₂-like receptors and in the subtype–selectivity profile of the ligands. In particular, the unsubstituted butyl chain confers to the ligands the highest affinities, while the presence of a 3-hydroxy or especially the inclusion of a 3-fluoro substituent is detrimental to D₂-like receptor binding. Indeed, compounds 3, 6, and 9, each bearing an unsubstituted butyl chain, display K_i values significantly lower than those of the 3-hydroxybutyl compounds 4, 7, and 10 and, especially, of the 3-fluorobutyl derivatives 5, 8, and 11. Minimal differences in affinity are observed between the 3-hydroxybutyl diastereomers 4a and 4b as well as 10a and 10b, nor between the 3-fluorobutyl diastereomers 11a and 11b, suggesting that the relative configuration between the stereocenters in the butyl chain and in position 2 of the 1,4-dioxane nucleus does not play a crucial role in the receptor interaction.

The nature of the SP also affects the D₃R selectivity profiles of the ligands. Compound 3, bearing the 6,6-diphenyl-1,4-dioxane scaffold, shows the best D₃R selectivity profile (D₂/D₃ = 143 and D₄/D₃ = 566) compared to the 5,5-diphenyl-1,4-dioxane and the 1,4-benzodioxane derivatives 6 (D₂/D₃ = 18 and D₄/D₃ = 35) and 9 (D₂/D₃ = 22 and D₄/D₃ = 35), respectively. Compared to the already published 6,6-diphenyl-1,4-dioxane multitarget compounds 1 and 2, the hybrid derivative 3 maintains high affinity only for D₃R, gaining in D₃R subtype selectivity.

The most promising hybrids (3, 6, and 9) were also evaluated for their binding affinity at D₁R ([³H]SCH23390, mouse fibroblast cells), 5-HT_1AR ([³H]8-OH-DPAT, HeLa cells), 5-HT_2AR, and 5-HT_2CR ([¹²⁵I]DOI, HEK cells) (data were obtained through the NIDA Addiction Treatment Discovery Program contract with Oregon Health & Science University) and the results are reported in Table 1.

Interestingly, compound 3 shows selectivity for D₃R not only over D₂R and D₄R but also over all other studied receptors (D₁/D₃ = 185, 5-HT_1A/D₃ = 121, 5-HT_2A/D₃ = 108, and 5-HT_2C/D₃ = 269). Compounds 6 and 9 also display negligible D₁R affinity but significantly lower K_i values at the serotoninergic 5-HT_1AR, 5-HT_2AR, and 5-HT_2CR subtypes and, therefore, are characterized by a more balanced 5-HT/D₂-like multitarget profile.

In vitro functional assays were also conducted for derivatives 3, 6, and 9 at all receptors at which they showed K_i values < 500 nM. The data, obtained through the NIDA Addiction Treatment Discovery Program contract with Oregon Health & Science University, are reported in Table 2. The results confirm the D₃R-preferential profile of 3, which behaves as a partial agonist (EC₅₀ = 9.8 nM), with low efficacy (36%) in the agonist mode; when tested as an antagonist, 3 shows an IC₅₀ value of 38 nM (% inhibition = 82.7%). This compound is also a weak D₂R antagonist and a 5-HT_1AR full agonist and exhibits very low potencies at D₁R and 5-HT_2AR.

Table 2. Potency (EC₅₀ or IC₅₀)^a and Efficacy (% Stimulation or % Inhibition)^b Values of 3, 6, and 9 at D₁R-D₄R, 5-HT_1AR, 5-HT_2AR, and 5-HT_2CR.

	functional profile of 3		functional profile of 6		functional profile of 9
receptor	EC50, nM (IC50, nM)	% stimul. (% inhib.)	EC50, nM (IC50, nM)	% stimul. (% inhib.)	EC50, nM (IC50, nM)	% stimul. (% inhib.)
D1 cAMP assay	(>10.000)	NDc	NDc	NDc	NDc	NDc
D2 mitogenesis assay	(139.2 ± 5.9)	(93.5)	(8.5 ± 1.3)	(95.2)	(40.0 ± 7.2)	(94.4)
D3 mitogenesis assay	9.8 ± 1.6	36.0	2.72	25.7	5.0 ± 1.7	34.0
D4 adenylate cyclase	NDc	NDc	21.7 ± 8.4	76.7	15.3 ± 1.9	73.1
5-HT1A [35S]GTPγS binding	232 ± 19	82.7	23.9 ± 8.8	90.8	22.7 ± 0.82	96.3
5-HT2A IP-1 formation	(7900 ± 9.8)	(45.4)	(650 ± 100)	(88.6)	(612 ± 58)	(86.8)
5-HT2C IP-1 formation	NDc	NDc	380 ± 130	33	750 ± 280	37

^aEC₅₀ or IC₅₀ values were from three experiments and data are presented as means ± SEM.

^bThe standard agonists SKF-38393 (D₁R), quinpirole (D₂-like subtypes), and serotonin (5-HT_1AR, 5-HT_2AR, and 5-HT_2CR) were used to determine the % stimulation; the standard antagonists SCH 23390 (D₁R), (+)-butaclamol (D₂R), NGB 2904 (D₃R), haloperidol (D₄R), WAY 100,635 (5-HT_1AR), ketanserin (5-HT_2AR), and SB242084 (5-HT_2CR) were used to determine the % inhibition.

^cND = not determined.

The results also highlight the interesting multitarget behavior of 6 and 9, which are characterized by similar functional profiles at all the studied receptors: they both are efficacious D₂R and 5-HT_2AR antagonists with high and low potency, respectively, efficacious 5-HT_1AR and D₄R agonists with high potency, as well as D₃R and 5-HT_2CR partial agonists with high and low potency, respectively. Moreover, when tested as D₃R antagonists, 6 and 9 show IC₅₀ values of 6.6 nM (64.0% efficacy) and 40.4 nM (64.2% efficacy), respectively.

Given that combinations of D₂R/D₃R antagonism or partial agonism with 5-HT_1AR agonism and 5-HT_2AR antagonism have been demonstrated to be favorable in the management of schizophrenia^28,49 and that D₄R activation might ameliorate cognitive impairment associated with schizophrenia,⁵⁰ the balanced multitarget profiles of 6 and 9 might be exploitable in the treatment of this disorder. Moreover, the partial activation of 5-HT_2CR might contribute to the potential antipsychotic activity of these agents.⁵¹

The here reported compounds underwent docking simulations on the resolved structure of the human D₃R in complex with eticlopride (PDB Id: 3PBL) to evaluate the main interactions stabilizing the computed complexes and to rationalize the observed differences in ligand affinity. The computed complexes of the compounds 3 and 9 also underwent 200 ns MD simulations to assess their stability and to investigate the resulting behavior of the human D₃R structure. Figures 4 and S1 display the overall putative complexes and the main stabilizing interactions for 3 (Figures 4A and S1A) and 9 (Figures 4B and S1B) as derived by the most representative cluster from MD runs. The reported complexes reveal that the dichlorophenylpiperazine moiety of 3 and 9 assumes comparable poses stabilized by similar key interactions. In detail, the ligand ammonium head elicits an ion pair with D3.32 (Asp110) further stabilized by the H-bond with Y7.43 (Tyr373). The 2,3-dichlorophenyl ring is engaged in π–π stacking interactions with F6.51 (Phe345) and F6.52 (Phe346), while the chlorine atoms can be involved in halogen bonds with S5.42 (Ser192), S5.42 (Ser193), and H6.55 (His349). C3.36 (Cys114) can also participate in this set of contacts through a π–S interaction. While showing a greater variability, the diphenyldioxane moiety in 3, as well as the benzodioxane ring in 9, are harbored within the same subpocket where they mostly stabilize hydrophobic contacts. In detail, they elicit similar π–π stacking interactions with Y7.35 (Tyr365) plus hydrophobic contacts with the surrounding apolar side chains. The dioxane oxygen and the amide linker are engaged in weak H-bonds with Y7.35 (Tyr365), S7.36 (Ser366), and S7.39 (Thr369).

Figure 4.

Key interactions stabilizing the putative complexes between D₃R and 3 (A) and 9 (B) within the binding site of D₃R.

Figure S2A analyzes the stability of these poses as assessed by the root-mean-square deviation (rmsd) values computed for the ligand’s atoms only. The obtained rmsd profile suggests the occurrence of two possible binding modes which differ for about 2.0 Å. A visual analysis of the two corresponding complexes reveals that the differences between the two poses are focused on the arrangement of the diphenyldioxane moiety in 3 and benzodioxane ring in 9, while the dichlorophenylpiperazine moiety of both ligands retains a significant stability during the MD simulations. Such a behavior is explainable by considering that these mobile moieties are engaged by weak interactions and thus are free to move to optimize the H-bonds stabilized by the amide linker and dioxane ring. Figure S2B shows that 3 assumes almost exclusively a binding mode which is that displayed in Figure 4 and is slightly shifted compared to that derived from docking simulations, while 9 exhibits an up and down profile with the starting pose being largely predominant. Figure S2B focuses on the rmsd profiles as computed for the backbone atoms of the human D₃R and provides evidence of substantial stability with rmsd values almost always lower than 6.0 Å. The complex with 3 reveals a slightly greater mobility than that with 9, which can be ascribed to its greater steric hindrance. Collectively, the overall stability of both ligands and receptor offers an encouraging confirmation for the reliability of the complexes shown in Figure 4.

Taken together, the reported computational results can explain why diphenyldioxane and benzodioxane analogues show similar affinity profiles. Indeed, these ligands stabilize an almost identical pattern of interactions with marginal differences involving the variable portion that however seems to be not so crucial for affinity, while having a key role in determining the observed selectivity profile. The D₃R selectivity of these ligands can be explained by considering the different wideness of the more external sub-pocket that accommodates the diphenyldioxane moiety. To verify this hypothesis, docking simulations were repeated for 3 on the resolved structure of D₄R in complex with nemonapride (PDB Id: 5WIV) by applying the same computational procedure based on the PLANTS program. In this case, the ligand is unable to assume acceptable poses in which the ammonium head contacts with D3.32 (Asp115). To further confirm the role of the pocket size in determining the selectivity of the simulated ligands, the void volume of the two binding sites was computed by using FPocket as implemented by the VEGA suite of programs⁵² and D₃R shows a larger pocket than D₄R (5043 vs. 4494 ų).

Docking results indicate that fluorine and hydroxyl derivatives can stabilize comparable sets of contacts. This finding is highly expected for the fluorine substitution since a marginal structural modification cannot have such a detrimental effect on the interaction capability of the resulting ligands. Similar considerations can be drawn for the hydroxyl analogues since the putative complexes do not reveal any detrimental roles of the hydroxyl function that can take part in the contacts stabilized by the ammonium head. Hence, docking results suggest that the negative effect played by these linker substitutions should not be ascribed to their interfering role on the interaction patterns with D₃R.

A possible explanation for the observed detrimental impact of these substitutions can be found in their effect on the basicity of the piperazine ring. An estimation of the effect of the fluorine atom on basicity can be derived by combining the available experimental pK_a value for aripiprazole, which is equal to 7.6 in 20% aqueous ethanol⁵³ with the average effect of the H/F exchange in β position to an amino group which decreases its basicity of around 2.0 pK_a unit.⁵⁴

A similar (albeit less pronounced) effect is also exerted by the hydroxyl function, which decreases the basicity of the vicinal amino groups due to the residual electronic delocalization from N to O atoms through the linking methylene groups.⁵⁵ For β hydroxyl functions, this effect is estimated to be equal to −1.0 in the corresponding pK_a values. Hence, fluorine and hydroxyl groups have a similar effect in decreasing the basicity of the ammonium head. In both cases, the resulting pK_a value should be less than 7.0, and this means that for these derivatives, the protonated form (which is involved in receptor recognition) is no longer the most probable state at physiological pH. As detailed above, such an effect is more marked for fluorine derivatives for which the predicted pK_a value should range around 5.5, thus indicating that the protonated state is virtually absent at physiological pH. The predicted value for the hydroxyl derivatives should be around 6.5, thus suggesting that the protonated state should represent about 10–15%. Notably, the different abundances of the protonated forms for the fluorine and hydroxyl analogues are in agreement with the measured affinity values.

To further confirm this hypothesis, the computed complex for 4a underwent a MD run with the same characteristics as described for 3 and 9. Figure S2 also reports the rmsd profiles for the hydroxyl derivative (4a) and reveals a remarkable stability of the corresponding complex. As described for the previous ligands, 4a also shows two possible binding modes even though the starting pose appears to be the most frequent one probably due to the stabilizing effect exerted by the hydroxyl group. Altogether, the performed simulations are not indicative of detrimental effects played by the hydroxyl function, and a similar consideration can also be extended to the fluorine atom and indirectly suggests that the drop in the affinity of these derivatives should be explained by considering factors that go beyond their interacting capacity with the human D₃R.

Conclusions

In conclusion, the biological and docking studies of the novel hybrid ligands 3–11 allowed us to get information about the role of the 1,4-benzodioxane or substituted 1,4-dioxane scaffold as the SP of bitopic compounds targeting D₃R and to obtain novel D₃R-selective or multitarget agents. An unsubstituted butyl chain between the pharmacophores characterizes the highest affinity compounds at D₃R: 3, 6, and 9. In particular, the 6,6-diphenyl-1,4-dioxane derivative 3 showed a D₃R preferential profile, behaving as a low-efficacy (36%) partial agonist/antagonist. An interesting multitarget profile has been highlighted for compounds 6 and 9, both behaving as efficacious D₂R and 5-HT_2AR antagonists with high and low potency, respectively, efficacious 5-HT_1AR and D₄R agonists with high potency, as well as D₃R and 5-HT_2CR partial agonists with high and low potency, respectively. Such a profile might be favorable for novel antipsychotic agents. However, further studies will be needed to support the therapeutic potential of these compounds for the treatment of schizophrenia.

Methods

Chemistry

General Procedures

Flash column chromatography was performed using silica gel (EMD Chemicals, Inc.; 230–400 mesh, 60 Å). Eluting solvents are described for each compound. ¹H NMR (400 MHz) and ¹³C NMR (100 MHz) spectra were acquired using a Varian Mercury Plus 400 spectrometer. Chemical shifts are reported in parts per million (ppm) and referenced according to a deuterated solvent for ¹H spectra (CDCl₃, 7.26) and ¹³C spectra (CDCl₃, 77.2). Combustion analysis was performed by Atlantic Microlab, Inc. (Norcross, GA), and the reported values agree within 0.4% of calculated values. Melting points were determined using a Thomas–Hoover melting point apparatus and are uncorrected. Anhydrous solvents were purchased from Aldrich and used without further purification, except for THF, which was freshly distilled from sodium-benzophenone ketyl. All other chemicals and reagents were purchased from Aldrich Chemical Co. On the basis of NMR, GC–MS, and combustion analysis data, all final compounds are >95% pure.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)butyl)-6,6-diphenyl-1,4-dioxane-2-carboxamide (3)

1,1′-Carbonyldiimidazole (0.18 g; 1.10 mmol) was added to a solution of 15(32) (0.30 g; 1.10 mmol) in THF (10 mL). The reaction was stirred at room temperature for 2 h. Amine 12(44) (0.33 g; 1.10 mmol) was added dropwise to the cooled solution (0 °C). The reaction was allowed to warm to room temperature and then stirred for 3 h. The solvent was removed under vacuum, and the residue was diluted in CHCl₃ and washed with NaHCO₃-saturated aqueous solution. The organic phase was dried over anhydrous Na₂SO₄, filtered, and concentrated under vacuum. The crude compound was purified by column chromatography eluting with EtOAc/CHCl₃ (5:5) to afford compound 3 as an oil in 80% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.66–6.85 (m, 13H, ArH), 4.63 (d, J = 32.49 Hz, 1H, dioxane), 4.18 (dd, J = 9.74, 9.75 Hz, 1H, dioxane), 3.79 (d, J = 12.07 Hz, 1H, dioxane), 3.62–3.43 (m, 2H, dioxane), 3.04 (m, 5H, CH₂N, CH₂NCO and NH), 2.65–2.39 (m, 8H, piperazine), 1.63 (m, 2H, CH₂), 1.51 (m, 2H, CH₂). ¹³C NMR (100 MHz, CDCl₃): δ 25.40, 27.73, 39.25, 40.90, 51.33, 51.83, 74.91, 84.23, 85.10, 90.21, 117.64, 123.98, 126.26, 127.18, 128.18, 129.25, 133.34, 139.62, 150.11, 172.71. The free base was transformed into the corresponding oxalate salt, which was crystallized from 2-PrOH (m.p. 107–108 °C). Anal. calcd for C₃₁H₃₅Cl₂N₃O₃·C₂H₂O₄: C, 60.19%, H, 5.66%, N, 6.38%, found; C, 60.34%, H, 5.51%, N, 6.52%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)-3-hydroxybutyl)-6,6-diphenyl-1,4-dioxane-2-carboxamide (4a and 4b)

These compounds were prepared following the procedure described for compound 3, starting from 15 and 13. The crude mixture of diastereomers was purified by column chromatography eluting with EtOAc/MeOH (99:1). The diastereomer 4a eluted first as an oil in 32% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.71–6.95 (m, 13H, ArH), 4.61 (d, J = 12.50 Hz, 1H, dioxane), 4.20 (m, 1H, dioxane), 3.95–3.25 (m, 4H, CHOH, dioxane), 3.10 (m, 7H, piperazine, NCH₂, CH₂N, NH and OH), 2.85 (m, 2H, piperazine), 2.62 (m, 3H, piperazine), 2.48 (m, 2H, piperazine), 1.79–1.60 (m, 2H, CH₂). The free base was transformed into the corresponding oxalate salt, which was crystallized from 2-PrOH (m.p. 146–147 °C). ¹³C NMR (100 MHz, DMSO): δ 36.43, 45.48, 51.15, 56.06, 61.68, 69.13, 71.91, 83.23, 86.62, 93.12, 117.64, 121.98, 126.22, 127.18, 128.18, 129.22, 133.34, 140.62, 150.01, 162.71. Anal. calcd for C₃₁H₃₅Cl₂N₃O₄·C₂H₂O₄: C, 58.76%, H, 5.53%, N, 6.23%, found C, 58.49%, H, 5.59%, N, 6.11%. The second fraction was the diastereomer 4b as an oil in 20% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.49–6.95 (m, 13H, ArH), 3.79 (d, J = 12.50 Hz, 1H, dioxane), 4.66 (d, J = 12.12 Hz, 1H, dioxane), 4.24 (m, 1H, dioxane), 4.09 (dd, J = 2.74, 3.51 Hz, 1H, dioxane), 3.48 (m, 2H, CHOH and dioxane), 3.08 (m, 8H, piperazine, CH₂N, CH₂N, NH and OH), 2.81 (m, 2H, piperazine), 2.58 (m, 2H, piperazine), 2.39 (m, 2H, piperazine), 1.63–1.51 (m, 2H, CH₂). The free base was transformed into the corresponding oxalate salt, which was crystallized from 2-PrOH (m.p. 122–123 °C). ¹³C NMR (100 MHz, DMSO): δ 36.38, 45.58, 51.12, 56.08, 61.48, 69.13, 75.91, 83.23, 86.62, 93.12, 117.64, 123.98, 126.22, 127.18, 128.18, 129.25, 133.34, 139.62, 150.11, 172.71. Anal. calcd for C₃₁H₃₅Cl₂N₃O₄·C₂H₂O₄: C, 58.76%, H, 5.53%, N, 6.23%, found; C, 58.55%, H, 5.63%, N, 6.04%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)-3-fluorobutyl)-6,6-diphenyl-1,4-dioxane-2-carboxamide (5)

This compound was prepared following the procedure described for compound 3, starting from 15 and 14. The crude compound was purified by column chromatography eluting with CHCl₃/acetone (8:2) to afford diastereomer 5 as an oil in 41% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.49–6.95 (m, 13H, ArH), 4.80 (m, 1H, CHF), 4.66 (d, J = 12.50 Hz, 1H, dioxane), 4.23 (dd, J = 3.52, 3.52 Hz, 1H, dioxane), 4.10 (dd, J = 3.12, 3.52 Hz, 1H, dioxane), 3.79 (d, J = 12.50 Hz, 1H, dioxane), 3.41 (m, 3H, CH₂N and dioxane), 3.10 (m, 5H, piperazine and NCH₂), 2.75 (m, 6H, piperazine and NH), 1.60–1.29 (m, 2H, CH₂). ¹⁹F NMR (376 MHz, CDCl₃/CFCl₃): δ −183.42 to −183.98 (m, 1F). The free base was transformed into the corresponding oxalate salt, which was crystallized from EtOH (m.p. 100–101 °C). ¹³C NMR (100 MHz, DMSO): δ 33.51, 34.80, 51.31, 51.88, 61.10, 74.90, 84.23, 85.00, 89.81, 90.20, 117.64, 121.98, 126.22, 127.18, 128.18, 129.22, 133.34, 140.62, 150.01, 162.71. Anal. calcd for C₃₁H₃₄Cl₂FN₃O₃·C₂H₂O₄: C, 58.58%, H, 5.36%, N, 6.21%, found; C, 58.28%, H, 5.44%, N, 6.10%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)butyl)-5,5-diphenyl-1,4-dioxane-2-carboxamide (6)

This compound was prepared following the procedure described for compound 3, starting from 16(32) and 12 to afford 6 as an oil in 65% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.66–6.85 (m, 13H, ArH)1.56 (m, 2H, CH₂), 4.87 (d, J = 29.31 Hz, 1H, dioxane), 3.64–3.24 (m, 4H, dioxane), 3.11 (m, 5H, NCH₂, CH₂N and NH), 2.69–2.35 (m, 8H, piperazine), 1.72 (m, 2H, CH₂). ¹³C NMR (100 MHz, CDCl₃): δ 25.40, 27.73, 39.25, 40.90, 51.33, 51.85, 69.13, 82.21, 90.04, 92.96, 117.64, 123.98, 126.22, 127.18, 128.26, 129.29, 133.34, 139.68, 150.11, 172.71. The free base was transformed into the corresponding oxalate salt, which was crystallized from EtOH (m.p. 94–95 °C). Anal. calcd for C₃₁H₃₅Cl₂N₃O₃·C₂H₂O₄: C, 60.19%, H, 5.66%, N, 6.38%, found; C, 59.99%, H, 5.48%, N, 6.60%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)-3-hydroxybutyl)-5,5-diphenyl-1,4-dioxane-2 Carboxamide (7)

This compound was prepared following the procedure described for compound 3, starting from 16 and 13.⁴³ The crude compound was purified by column chromatography eluting with EtOAc/CHCl₃/MeOH (5:5:1) to afford diastereomer 7 as an oil in 32% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.66–6.85 (m, 13H, ArH), 4.89 (m, 1H, dioxane), 3.93–3.59 (m, 4H, dioxane), 3.31 (m, 1H, CHOH), 3.08 (m, 6H, NCH₂, CH₂N, NH and OH), 2.92–2.31 (m, 8H, piperazine), 1.65–1.46 (m, 2H, CH₂). ¹³C NMR (100 MHz, CDCl₃): δ 36.33, 45.68, 51.82, 56.08, 61.48, 69.13, 75.91, 83.23, 86.62, 93.12, 117.64, 123.98, 126.22, 127.18, 128.26, 129.22, 133.34, 139.68, 150.11, 172.71. The free base was transformed into the corresponding oxalate salt, which was crystallized from MeOH (m.p. 132–133 °C). Anal. calcd for C₃₁H₃₅Cl₂N₃O₄·C₂H₂O₄: C, 58.76%, H, 5.53%, N, 6.23%, found, C, 59.01%, H, 5.40%, N, 6.35%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)-3-fluorobutyl)-5,5-diphenyl-1,4-dioxane-2-carboxamide (8)

This compound was prepared following the procedure described for compound 3, starting from 16 and 14.⁴² The crude compound was purified by column chromatography eluting with CHCl₃/acetone (8:2) to afford diastereomer 8 as an oil in 64% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.49–6.95 (m, 13H, ArH), 4.85 (m, 2H, CHF and dioxane), 4.30 (m, dioxane), 3.78 (m, 1H, dioxane), 3.58 (m, 2H, CH₂N), 3.01 (m, 4H, piperazine and NCH₂), 2.68 (m, 3H, piperazine and NH), 2.48 (m, 2H, piperazine), 2.29 (m, 2H, piperazine), 1.63–1.45 (m, 2H, CH₂). The free base was transformed into the corresponding oxalate salt, which was crystallized from 2-PrOH (m.p. 126–127 °C). ¹³C NMR (100 MHz, DMSO): δ ppm 33.43, 34.78, 51.29, 51.88, 61.10, 69.13, 82.23, 89.84, 90.08, 93.10, 117.64, 121.98, 126.22, 127.18, 128.18, 129.33, 133.34, 140.62, 150.01, 162.71. ¹⁹F NMR (376 MHz, DMSO): δ ppm −181.12 to −181.58 (m, 1F). Anal. calcd for C₃₁H₃₄Cl₂FN₃O₃·C₂H₂O₄: C, 58.58%, H, 5.36%, N, 6.21%, found; C, 58.33%, H, 5.53%, N, 6.06%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)butyl)-2,3-dihydrobenzo[b][1,4]dioxine-2-carboxamide (9)

This compound was prepared following the procedure described for compound 9, starting from 17 and 12. The crude compound was purified by column chromatography eluting with EtOAc/CHCl₃ (5:5) to afford compound 9 as an oil in 58% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.18–6.90 (m, 7H, ArH), 4.68 (dd, J = 7.43, 7.03 Hz, 1H, dioxane), 4.54 (dd, J = 2.73, 2.73 Hz, 1H, dioxane), 4.19 (dd, J = 7.03, 7.43 Hz, 1H, dioxane), 3.39 (m, 2H, NCH₂), 3.11 (m, 3H, CH₂N and NH), 2.63 (m, 4H, piperazine), 2.42 (m, 4H, piperazine), 1.60 (m, 4H, CH₂CH₂). The free base was transformed into the corresponding oxalate salt, which was crystallized from 2-PrOH (m.p. 167–168 °C). ¹³C NMR (100 MHz, DMSO): δ 21.29, 26.63, 38.24, 48.72, 51.73, 55.86, 65.29, 73.03, 117.44, 117.79, 120.22, 121.92, 122.01, 125.55, 126.51, 129.03, 133.14, 142.61, 143.48, 150.28, 164.73, 167.22. Anal. calcd for C₂₃H₂₇Cl₂N₃O₃·C₂H₂O₄: C, 54.16%, H, 5.27%, N, 7.58%, found; C, 54.01%, H, 5.34%, N, 7.32%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)-3-hydroxybutyl)-2,3-dihydrobenzo[b][1,4]dioxine-2-carboxamide (10a and 10b)

These compounds were prepared following the procedure described for compound 3, starting from 17 and 13. The crude compound was purified by column chromatography eluting with CHCl₃/CH₃OH (95:5) to afford the diastereomeric mixture 10a/10b as an oil in 45% yield. ¹H NMR (400 MHz, CDCl₃): δ ppm 1.51–1.70 (m, 2H, CH₂), 2.39 (m, 2H, piperazine), 2.56 (m, 2H, piperazine), 2.78 (m, 2H, piperazine), 3.04 (m, 4H, piperazine, NH and OH), 3.37 (m, 2H, CH₂N), 3.63 (m, 2H, NCH₂), 3.80 (m, 1H, CHOH), 4.22 (m, 1H, dioxane), 4.49 (m, 1H, dioxane), 4.67 (dd, J = 6.64, 7.03, 1H, dioxane), 6.90–7.12 (m, 7H, ArH). The free bases were transformed into the corresponding oxalate salts, and the diastereomers were separated by crystallization from 2-PrOH. Diastereomer 10a (m.p. 88–89 °C): ¹H NMR (400 MHz, DMSO): δ 8.20 (m, 1H, ArH), 7.31 (m, 2H, ArH), 7.16 (m, 1H, ArH), 6.95 (m, 1H, ArH), 6.81 (m, 2H, ArH), 4.77 (dd, J = 5.85, 5.86 Hz, 1H, dioxane), 4.31 (dd, J = 3.10, 2.25 Hz, 1H, dioxane), 4.20 (dd, J = 6.30, 5.86 Hz, 1H, dioxane), 3.82 (m, 1H, CHOH), 2.43 (m, 2H, piperazine), 3.19 (m, 10H, piperazine, NCH₂, CH₂N, NH and OH), 3.01 (m, 1H, piperazine), 2.90 (m, 1H, piperazine), 1.45–1.55 (m, 2H, CH₂). ¹³C NMR (100 MHz, DMSO): δ 25.92, 35.13, 35.72, 48.64, 52.46, 54.19, 61.79, 67.76, 73.04, 90.79, 105.00, 117.74, 120.22, 121.99, 122.06, 125.14, 126.52, 129.05, 133.49, 142.60, 143.48, 150.37, 163.82, 167.34. Anal. calcd for C₂₃H₂₇Cl₂N₃O₄·C₂H₂O₄: C, 52.64%, H, 5.12%, N, 7.37%, found; C, 52.37%, H, 5.30%, N, 7.51%. Diastereomer 10b (m.p. 99–100 °C). ¹H NMR (400 MHz, DMSO): δ 8.20 (m, 1H, ArH), 7.37 (m, 2H, ArH), 7.20 (m, 1H, ArH), 6.95 (m, 1H, ArH), 6.81 (m, 2H, ArH), 4.80 (dd, J = 5.85, 5.86 Hz, 1H, dioxane), 4.31 (d, J = 11.26 Hz, 1H, dioxane), 4.22 (dd, J = 5.41, 5.86 Hz, 1H, dioxane), 3.84 (m, 1H, CHOH), 3.10 (m, 11H, piperazine, NCH₂, CH₂N, NH and OH), 2.95 (m, 1H, piperazine), 2.43 (m, 2H, piperazine), 1.53 (m, 2H, CH₂). ¹³C NMR (100 MHz, DMSO): δ 15.57, 35.08, 35.62, 48.28, 52.56, 61.35, 63.10, 65.32, 73.01, 102.34, 106.67, 117.45, 117.75, 120.22, 122.03, 125.63, 126.48, 129.05, 133.14, 142.60, 143.48, 150.17, 163.10, 167.35. Anal. calcd for C₂₃H₂₇Cl₂N₃O₄·C₂H₂O₄: C, 52.64%, H, 5.12%, N, 7.37%, found; C, 52.45%, H, 5.28%, N, 7.56%.

N-(4-(4-(2,3-Dichlorophenyl)piperazin-1-yl)-3-fluorobutyl)-2,3-dihydrobenzo[b][1,4]dioxine-2-carboxamide (11a and 11b)

These compounds were prepared following the procedure described for compound 3, starting from 17 and 14 to afford the diastereomeric mixture 11a/11b as an oil in 36% yield. ¹H NMR (400 MHz, CDCl₃): δ 7.18 (m, 2H, ArH), 6.78–6.99 (m, 5H, ArH), 4.92 (m, 1H, CHF), 4.70 (m, 1H, 2-CH dioxane), 4.49 (m, 1H, 3-CH dioxane), 4.21 (m, 1H, 3-CH dioxane, diastereomeric ratio 50:50), 3.49 (m, 2H, CH₂N), 3.04 (m, 4H, piperazine and NCH₂), 2.63–2.88 (m, 6H, piperazine and NH), 2.18 (m, 1H, piperazine), 1.90 (m, 2H, CH₂). The diastereomers 11a and 11b were separated by preparative TLC eluting with CHCl₃/CH₃OH (95:5). ¹⁹F NMR (376 MHz, CDCl₃/CFCl₃): δ ppm −182.18 to −182.43 (m, 1F, diastereomer 11a), −182.60 to −182.98 (m, 1F, diastereomer 11b). The free bases were transformed into the corresponding oxalate salts, which were crystallized from 2-PrOH. Diastereomer 11a (m.p. 200–201 °C). Anal. calcd for C₂₃H₂₆Cl₂FN₃O₃·C₂H₂O₄: C, 52.46%, H, 4.93%, N, 7.34%, found; C, 52.67%, H, 5.08%, N, 7.25%. C, H, N. Diastereomer 11b (m.p. 180–182 °C). Anal. calcd for C₂₃H₂₆Cl₂FN₃O₃·C₂H₂O₄: C, 52.46%, H, 4.93%, N, 7.34%, found; C, 52.22%, H, 5.03%, N, 7.22%.

Receptor Binding Studies

Radioligand Binding Assays at Human D₂R, D₃R, and D₄R

Binding at dopamine D₂-like receptors was determined using previously described methods.^41,45,46 Membranes were prepared from HEK293 cells stably expressing human D₂, D₃, or D₄, grown in a 50:50 mix of Dulbecco’s minimal essential medium (DMEM) and Ham’s F12 culture media, supplemented with 20 mM HEPES, 2 mM l-glutamine, 0.1 mM nonessential amino acids, 1× antibiotic/antimycotic, 10% heat-inactivated fetal bovine serum, and 200 μg/mL hygromycin (Life Technologies, Grand Island, NY) and kept in an incubator at 37 °C and 5% CO₂. Upon reaching 80–90% confluency, the cells were harvested using premixed Earle’s Balanced Salt Solution (EBSS) with 5 μM EDTA (Life Technologies) and centrifuged at 3000 rpm for 10 min at 21 °C. The supernatant was removed, and the pellet was resuspended in 10 mL hypotonic lysis buffer (5 mM MgCl₂, 5 mM Tris, pH 7.4 at 4 °C) and centrifuged at 20,000 rpm for 30 min at 4 °C. The pellet was then resuspended in fresh EBSS buffer made from 8.7 g/L Earle’s Balanced Salts without phenol red (US Biological, Salem, MA), 2.2 g/L sodium bicarbonate, pH to 7.4. A Bradford protein assay (Bio-Rad, Hercules, CA) was used to determine the protein concentration, and membranes were diluted to 500 μg/mL and stored at −80 °C for later use. Radioligand competition binding experiments were conducted using thawed membranes. Test compounds were freshly dissolved in 30% DMSO and 70% H₂O to a stock concentration of 100 μM. To assist the solubilization of free-base compounds, 10 μL of glacial acetic acid was added along with DMSO. Each test compound was then diluted into 13 half-log serial dilutions using 30% DMSO vehicle; the final test concentrations ranged from 10 μM to 10 pM. The previously frozen membranes were diluted in fresh EBSS to a 100 μg/mL (for D₂ or D₃) or 200 μg/mL (D₄) stock for binding. Radioligand competition experiments were conducted in glass tubes containing 300 μL of fresh EBSS buffer, 50 μL of diluted test compound, 100 μL of membranes (10 μg of total protein for D₂ or D₃ and 20 μg of total protein for D₄), and 50 μL of [³H]N-methylspiperone (0.4 nM final concentration; PerkinElmer). Nonspecific binding was determined using 10 μM butaclamol (Sigma-Aldrich, St. Louis, MO), and total binding was determined with the 30% DMSO vehicle. All compound dilutions were tested in triplicate and the reaction was incubated for 1 h at room temperature. The reaction was terminated by filtration through Whatman GF/B filters, presoaked for 1 h in 0.5% polyethylenimine, using a Brandel R48 filtering manifold (Brandel Instruments, Gaithersburg, MD). The filters were washed three times with 3 mL of ice-cold EBSS buffer and transferred to scintillation vials. CytoScint liquid scintillation cocktail (3 mL, MP Biomedicals, Solon, OH) was added, and vials were counted using a PerkinElmer Tri-Carb 2910 TR liquid scintillation counter (Waltham, MA).

Radioligand Binding Assay at Human D₁R

Mouse fibroblast cells expressing the human D₁R at high density (LhD₁ cells) are used. The cells are grown to confluence in DMEM containing 10% FetalClone1 serum (FCS, HyClone), 0.05% penicillin–streptomycin (pen-strep), and 400 μg/mL of Geneticin (G418). Three confluent 150 mm plates yield enough membranes for three assay plates with ∼10–15 μg protein per well. The cells from three 150 mm plates are scraped and centrifuged at 500×g for 5 min. The pellet is overlaid with 2 mL assay buffer (50 mM Tris–HCl containing 120 mM NaCl, 5 mM KCl, 2 mM CaCl₂, and 1 mM MgCl₂) and frozen at −70 °C. On the day of experiment, the pellet is homogenized in 30 mL assay buffer with a polytron. Cell homogenate (100 μL) is added to wells containing 800 μL of test drug or buffer. After a 10 min preincubation, 100 μL of [³H]SCH-23390 (0.18 nM final concentration) is added. The plates are incubated at 25 C for 60 min. The reaction is terminated by filtration using a Tomtec 96 well harvester, and the radioactivity on the filters is determined using a PerkinElmer microbeta scintillation counter. Nonspecific binding is determined with 1 μM SCH-23390.

Radioligand Binding Assay at Human 5-HT_1AR

A human cell line (HeLa) stably transfected with genomic clone G-21 coding for the human 5-HT_1AR was used. Cells were grown as monolayers in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum and gentamycin (100 μg/mL) under 5% CO₂ at 37 °C. Cells were detached from the growth flask at 95% confluency by a cell scraper and were lysed in ice-cold Tris (5 mM) and EDTA buffer (5 mM, pH 7.4). Homogenates were centrifuged for 20 min at 40000g, and pellets were resuspended in a small volume of ice-cold Tris/EDTA buffer (above) and immediately frozen and stored at -70 °C until use. On the day of experiment, cell membranes (80–90 μg of protein) were resuspended in binding buffer (50 mM Tris, 2.5 mM MgCl₂, and 10 mM pargyline, pH 7.4). Membranes were incubated in a final volume of 0.32 mL for 30 min at 30 °C with 1 nM [³H]8-OH-DPAT in the absence or presence of various concentrations of competing drugs (1 pM to 1 μM); each experimental condition was performed in triplicate. Nonspecific binding was determined in the presence of 10 μM 5-HT.

Radioligand Binding Assay at Human 5-HT_2AR and 5-HT_2CR

Human embryonic kidney cells expressing human 5HT_2AR (HEK-h5HT_2A) or human 5HT_2CR (HEK-h5HT_2C) are used. The cells are grown until confluent on 15 cm plates. The medium is removed, and the cells are washed with phosphate-buffered saline (PBS), scraped into 2 mL PBS, and frozen at −20 °C until needed. Cell suspension is thawed, 10 mL assay buffer (50 mM Tris, pH 7.4 at 37 °C, with 0.1% ascorbic acid and 5 mM CaCl₂) is added per plate of cells, and polytronned at setting 6 for 5 s. The homogenate is centrifuged at 15,500 rpm for 20 min. To minimize the residual 5HT concentration, the pellet is resuspended in buffer, polytronned, and centrifuged as mentioned above. The final pellet is resuspended in 2 mL buffer/plate of cells. The binding assay includes 50 μL drug, 5-HT or buffer, 50 μL cell homogenate, 50 μL [¹²⁵I]DOI (∼0.1 nM), and buffer in a final volume of 250 μL. Specific binding is defined as the difference between total binding and binding in the presence of 10 μM 5HT. The reaction is incubated for 1 h at 37 °C and terminated by filtration through Wallac A filter mats presoaked in 0.05% polyethylenimine using a Tomtec 96-well harvester. The radioactivity remaining on filters is determined using a Wallac betaplate reader.

Functional Assays

Adenylate Cyclase/cAMP Functional Assay

Mouse glioma cells expressing the monkey D₁R at low density (C6D1 low-D₁-density cells) are used. The cells are grown on 48-well plates in DMEM containing 10% FCS, 0.05% pen-strep, and 2 g/mL of puromycin. When the wells reach 80–90% confluency, DMEM is removed, and each well is rinsed once with 0.5 mL of EBSS buffer (116.4 mM NaCl, 5.4 mM KCl, 1.2 mM Na₂HPO₄, 1.3 mM CaCl₂, 1.2 mM MgSO₄, 15 mM HEPES, and 10 mM glucose, pH 7.4 at room temperature) containing 0.1% ascorbic acid, 2% bovine calf serum (BCS), and 0.11% 3-isobutyl-1-methyl-xanthine (IBMX), an inhibitor of cAMP phosphodiesterase. The cell number of C6D1 low-D₁-density cells per 48-well plate is determined. Two hundred thousand cells per well is optimal for the use of the cyclic AMP EIA kit (Cayman Chemical). The 48-well plates are preincubated for 20 min at 37 °C with EBSS buffer alone and then incubated for an additional 20 min with the test compound diluted in EBSS buffer (final volume 1 mL) in triplicate. The EBSS buffer is removed and 0.05 mL of 3% TCA is added to each well. After 1 h, the supernatant is diluted 1:50, and 50 μL of diluted supernatant is added to the cyclic AMP EIA test plate in a final assay volume of 200 μL/well. The assay is incubated for 18 h at 4 °C. The supernatant is aspirated, and each well is rinsed five times with 300 μL of EIA buffer. The developing reagent (200 μL, Ellman’s) is added to each well and incubated at room temperature for 2 h while gently rotating. The plate is read in a microplate spectrophotometer at 405 nM (BioRad Benchmark Plus). The linearity of signal is observed at cAMP amounts up to ∼100 pg.

Mitogenesis Functional Assay

CHOp-D2 and CHOp-D3 cells are maintained in α-MEM with 10% fetal bovine serum (FBS, Atlas Biologicals), 0.05% pen-strep, and 400 μg/mL of G418. To measure D₂R or D₃R stimulation of mitogenesis (agonist assay) or inhibition of quinpirole stimulation of mitogenesis (antagonist assay), CHOp-D2 or CHOp-D3 cells, respectively, are seeded in a 96-well plate at a concentration of 5000 cells/well. The cells are incubated at 37 °C in α-MEM with 10% FBS. After 48–72 h, the cells are rinsed twice with serum-free α-MEM and incubated for 24 h at 37 °C. Serial dilutions of test compounds are made by the Biomek robotics system in serum-free α-MEM. In the functional assay for agonists, the medium is removed and replaced with 100 μL of test compound in serum-free α-MEM. In the antagonist assay, the serial dilution of the putative antagonist test compound is added in 90 μL (1.1× of final concentration) and 300 nM quinpirole (30 nM final) is added in 10 μL. After another 24 h (D₂) or 16 h (D₃) incubation at 37 °C, 0.25 μCi of [³H]thymidine in αMEM supplemented with 10% FBS is added to each well, and the plates are further incubated for 2 h at 37 °C. The cells are trypsinized by addition of 10× trypsin solution (1% trypsin in calcium–magnesium-free phosphate-buffered saline), and the plates are filtered and counted as usual. Quinpirole is run each day as an internal control, and dopamine is included for comparative purposes.

Adenylate Cyclase/cAMP Functional Assay

HEK-D4.4-AC1 cells are grown to confluency on 150 mM plates. The cells are plated at a density of 375,000 cells per well in 48-well plates in DMEM supplemented with 5% FetalClone, 5% BCS, and pen-strep. After ∼36 h, the medium is changed to DMEM supplemented with 10% charcoal-stripped FetalClone and pen-strep. The medium is removed ∼18 h later. For agonist assays, 0.8 mL of EBSS (116 mM NaCl, 22 mM glucose, 15 mM HEPES, 8.7 mM NaH₂PO₄, 5.4 mM KCl, 1.3 mM CaCl₂, 1.2 mM MgSO₄, 1 mM ascorbic acid, 0.5 mM IBMX [3-isobutyl-1-methyl-xanthine], and 2% BCS, pH 7.4 at 37 °C) is added, cells are incubated for 20 min, agonists are added, and, after a 20 min incubation, 10 μM of forskolin is added in a final volume of 1 mL. For antagonists, 0.7 mL of EBSS is added, cells are incubated for 10 min, antagonists are added, cells are incubated for 10 min, 2 nM quinpirole is added, and after a 20 min incubation, 10 μM forskolin is added in a final volume of 1 mL. For all conditions, after a 20 min incubation with forskolin, the reaction is terminated by aspiration of the buffer, and 0.1 mL of trichloroacetic acid is added. The plates are incubated for 2 h on a rotator. Adenylate cyclase activity is measured using a cyclic AMP EIA kit (Cayman). Aliquots (9 μL) of each well are diluted to 200 μL with EIA buffer from the kit, and 50 μL of dilution is added to the EIA plate. After addition of tracer and monoclonal antibody, the EIA plates are incubated for 18 h at 4 °C. The reaction is aspirated, the plates are washed with 5 × 300 μL wash buffer, and Ellman’s reagent is added. After a 2 hour incubation in the dark on a rotator, the plates are read at 410 nm. Basal cAMP is subtracted from all values. D₄R agonists inhibit forskolin-stimulated cAMP formation, and maximal inhibition is defined with 1 μM quinpirole. Data normalized to percent forskolin stimulation are shown in the graph. The maximal effect is normalized to maximal effect of quinpirole in the tables. For antagonists, maximal reversal of inhibition of cAMP formation is defined with 10 μM haloperidol.

[³⁵S]GTPγS Binding Assay

The effects of the various compounds tested on [³⁵S]GTPγS binding in HeLa cells expressing the recombinant human 5-HT_1AR were evaluated according to the method of Stanton and Beer [41] with minor modifications. The stimulation experiments are as follows: Cell membranes (50–70 μg of protein) were resuspended in buffer containing 20 mM HEPES, 3 mM MgSO₄, and 120 mM NaCl (pH 7.4). The membranes were incubated with 30 μM GDP and various concentrations (from 0.1 nM to 10 μM) of test drugs or 8-OH-DPAT (reference curve) for 20 min at 30 °C in a final volume of 0.5 mL. The samples were transferred to ice, [³⁵S]GTPγS (200 pM) was added, and the samples were incubated for further 30 min at 30 °C. The preincubation with both agonist and antagonist, before initiating the [³⁵S]GTPγS binding, ensures that agonist and antagonist are at equilibrium. Nonspecific binding was determined in the presence of 10 mM GTPγS. Incubation was stopped by the addition of ice-cold HEPES buffer and rapid filtration on Unifilter B filters (PerkinElmer) using a Filtermate cell harvester (Packard). The filters were washed with ice-cold Hepes buffer, and the radioactivity retained on the filters was determined by a TopCount, PerkinElmer liquid scintillation counter at 90% efficiency.

Inositol-1-phosphate (IP-1) Formation

HEK-h5HT2A or HEK-h5HT2C cells are used as the tissue source. The day before the experiment, cells are plated in 24-well plates at a density of 400,000 cells per well using DMEM supplemented with charcoal-stripped FetalClone. Drugs are made up in stimulation buffer supplied in the kit. Medium is removed from the well, test compounds or serotonin or antagonist or buffer is added, and the cells are incubated for 1 h. The cells are lysed for 30 min, and 50 μL of cell lysate is added to the IP-1 plates. After the addition of appropriate antibodies, the plates are incubated for 3 h, washed six times, incubated with the substrate for 20 min, and, after termination of the reaction, the plate is read on a plate reader at 450 nm with a correction at 620 nm. Agonists are normalized to the maximal stimulation by serotonin, and antagonists are tested in the presence of 100 nM serotonin and normalized to the inhibition by 10 μM ketanserin (5-HT_2AR) or 1 μM SB 242084 (5-HT_2CR).

Computational Methods

The docking simulations were based on the resolved structure of D₃R in complex with eticlopride (PDB Id: 3PBL). The protein structure was checked and prepared as reported elsewhere.⁴⁰ The ligands were generated in their protonated state, and their structure was optimized by PM7 semi-empirical calculations as implemented by MOPAC2016 (keywords = PM7 CHARGE = 1.00 PRECISE GEO-OK).⁵⁶ The docking simulations were carried out using PLANTS⁵⁷ focusing the search within a 10 Å radius sphere around the bound eticlopride. For each ligand, 10 poses were generated and evaluated by the ChemPLP scoring function with a speed equal to 1. The same procedure was applied to perform docking simulations on the resolved structure of D₄R in complex with nemonapride (PDB Id: 5WIV).

For the performed MD runs, the membrane was added to the computed complexes using the CHARMM-GUI server.⁵⁸ The protein was oriented using the PPM server,⁵⁹ and the bilayer was made by phosphatidylcholine (POPC, 70%) and cholesterol (30%). TIP3P water molecules were added to both sides of the membrane, as well as Na⁺ and Cl^– ions to reach an ionic concentration of 0.15 M. Amber force fields ff14SB, GAFF, and Lipid17 were used for the proteins, ligands, and lipids, respectively. The systems underwent a three-step minimization: first, the hydrogen atoms were minimized, then the solvent molecules, and finally the whole system, applying restraints (5 kcal/mol·Å) in the α carbons. Then, the systems underwent heating where the temperature was brought to 300 K using the Langevin thermostat, followed by an equilibration phase first using the NPT ensemble with the Berendsen barostat (1 atm) and finally an NVT ensemble. The 200 ns production runs were performed with the NVT ensemble with a time step of 0.02 fs and the SHAKE algorithm. PME and PBC were applied.⁶⁰

Glossary

Abbreviations

DR

dopamine receptor

GPCR

G protein-coupled receptor

PD

Parkinson’s disease

PP

primary pharmacophore

SP

secondary pharmacophore

5-HT_1AR

5-HT_1A receptor

TLC

thin-layer chromatography

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acschemneuro.1c00368.

• Cartoon view of the computed complexes for 3 and 9 and rmsd profiles as derived from the performed MD runs for the simulated complexes for 3, 9, and 4a (PDF)

Author Contributions

A.B., A.H.N., F.D.B., G.G., P.P., W.Q., and A.P. designed, synthesized, and characterized the new compounds. They wrote the relative chemical experimental parts of the manuscript. A.B., A.H.N., and T.M.K. performed binding experiments at dopamine D₂, D₃, and D₄ receptors. They provided binding assays at D₁, 5-HT_1A, 5-HT_2A, and 5-HT_2C receptors and functional assays through the NIDA Addiction Treatment Discovery Program contract with Oregon Health & Science University. G.V. and S.G. performed the docking experiments and MD simulations. F.D.B. and W.Q. drafted the main text of the manuscript. All authors critically discussed and approved the final version of the manuscript.

The authors declare no competing financial interest.