Design and Synthesis of Conformationally Flexible Scaffold as Bitopic Ligands for Potent D3-Selective Antagonists

Previous studies have confirmed that the binding of D3 receptor antagonists is competitively inhibited by endogenous dopamine despite excellent binding affinity for D3 receptors. This result urges the development of an alternative scaffold that is capable of competing with dopamine for binding to the D3 receptor. Herein, an SAR study was conducted on metoclopramide that incorporated a flexible scaffold for interaction with the secondary binding site of the D3 receptor. The alteration of benzamide substituents and secondary binding fragments with aryl carboxamides resulted in excellent D3 receptor affinities (Ki = 0.8–13.2 nM) with subtype selectivity to the D2 receptor ranging from 22- to 180-fold. The β-arrestin recruitment assay revealed that 21c with 4-(pyridine-4-yl)benzamide can compete well against dopamine with the highest potency (IC50 = 1.3 nM). Computational studies demonstrated that the high potency of 21c and its analogs was the result of interactions with the secondary binding site of the D3 receptor. These compounds also displayed minimal effects for other GPCRs except moderate affinity for 5-HT3 receptors and TSPO. The results of this study revealed that a new class of selective D3 receptor antagonists should be useful in behavioral pharmacology studies and as lead compounds for PET radiotracer development.


Introduction
Targeting D 2 and D 3 receptors has been studied for the treatment of neuropsychiatric disorders such as schizophrenia, and substance use disorders and addiction [1][2][3][4]. However, preferential localization of D 3 receptors in limbic regions of the human brain suggested that D 3 receptors may be a suitable target for developing therapeutics for treating neuropsychiatric disorders [5,6]. Other studies have demonstrated that this receptor plays a role in mediating the motivational actions of psychostimulants such as cocaine and amphetamine, and D 3 antagonists have shown great promise in blocking cocaine self-administration in rodents and nonhuman primates [7,8]. The recent observation in the treatment of opioid use disorder has accelerated the need for the clinical evaluation of drugs targeting D 3 receptors [9][10][11].
The development of dopamine D 3 -selective ligands continues to be a challenging area of medicinal chemistry research due to the high sequence homology of D 2 and D 3 receptor within the transmembrane (TM) domains (~79%) [12]. For developing receptor subtype selectivity, a "bitopic ligand" design has proven to be effective in the development of D 3 -selective compounds [13,14]. In this approach, a protonated basic amine in different scaffolds forms a salt bridge with Asp110 3.32 of the D 3 receptor in the orthosteric binding site (OBS), which is important for high binding affinity and the potency [15]. A secondary pharmacophore 2 of 27 having an aromatic ring and appropriate linker group can result in high selectivity for the D 3 receptor by the interaction with the secondary binding site (SBS) [16][17][18].
The first D 3 -selective scaffold contained an N-aryl piperazine moiety as the orthosteric binding fragment and an aryl carboxamide moiety with an alkyl linker as a secondary binding fragment [16,[19][20][21][22][23][24][25]. This scaffold exhibited a sub nM binding affinity and good subtype selectivity for D 3 receptors versus D 2 receptors. However, these ligands also exhibited high binding affinity for other GPCRs (e.g., 5-HT or adrenergic receptors), which may lead the unwanted side effects [26][27][28]. Moreover, the in vivo properties of radiolabeled versions of this scaffold were not useful as PET radiotracers since they could not compete with endogenous dopamine for binding to the D 3 receptor in vivo [21,29]. Since the replacement of substituents on benzamide or secondary binding fragments did not result in a significant change in properties of the N-aryl piperazine congeners, many groups have pursued other scaffolds, including azabicyclo [3.1.0]hexane [30][31][32], azaspiro alkane [33], diazaspiro alkane [34], tranylcypromine [35], or phenylcyclopropylmethylamine (PCPMA) [36]). However, these can be limited for clinical use under certain circumstances due to the poor bioavailability or toxicity [37][38][39] or are still under investigation. Recently, D 2 /D 3 receptor agonist-and antagonist-modified bitopic ligands were developed based on (+)-PD128,907 or PF-592379 for selective agonist [40] and eticlopride for D 2 /D 3 receptor ligands [41]. These compounds had comparatively low selectivity for D 3

versus D 2 receptors.
In the current study, we designed a new class of D 3 receptor antagonist having the conformationally flexible scaffold of metoclopramide and the eticlopride-based benzamides (e.g., [ 18 F]fallypride, Ki D 2 = 0.02 nM and D 3 = 0.19 nM [42,43], IC 50 = 1.7 nM [29]; [ 11 C]FLB457, Ki = 0.02 nM for D 2 /D 3 receptors [44]) as lead compounds. Metoclopramide is largely used as an antiemetic; however, this compound also exhibited the low affinity for mixed D 2 /D 3 receptors with the orthosteric binding fragment [45]. A combination of this flexible scaffold with well-established primary pharmocophore of the eticlopride-based benzamides was expected to achieve the high binding affinity and the potency for D 3 receptors. Since the basic amine in this scaffold is structurally flexible without ring strain, the secondary binding fragment can be extended to strongly interact with the SBS while the orhosteric binding fragment remains bound to the OBS. Comprehensive screening was investigated for off-target interactions with other GPCRs; computational studies were also performed to provide the rational for the excellent potency of developed D 3 receptor antagonists.
The next series focused on preparing analogs having a spacer group with an aromatic ring system for interacting with the SBS. For the aromatic ring moiety, we tested 4-(thiophen-2-yl)benzamide or 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol. The 4-(Thiophen-2-yl)benzamide fragment was chosen from our previous results. This aromatic ring system was observed in LS-3-134 and other structural congeners having a high D 3 affinity and excellent selectivity versus the D 2 receptor [46][47][48]. 3b was N-alkylated with N- (3-bromopropyl) or N-(4-bromobutyl)phthalimide and then the protecting phthalimide 7a or 7b was hydrolyzed using hydrazine hydrate by heating for 3 h to give primary amine 8a or 8b (Scheme 2). 4-(Thiophen-2-yl)benzoic acid was converted to the corresponding acyl chlo-ride using thionyl chloride at RT followed by treatment with 8a or 8b to give 9a or 9b in 50% or 20% yield, respectively. The triazole-thiol ether analogs were prepared by reduction of 5c and 5d to give alcohols 10a and 10b, which were converted to 11a and 11b using Mitsunobu reaction. The desired products 11a or 11b were obtained in 16% and 28% yield, respectively (Scheme 2). The next series focused on preparing analogs having a spacer group with an aromatic ring system for interacting with the SBS. For the aromatic ring moiety, we tested 4-(thiophen-2-yl)benzamide or 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol. The 4-(Thiophen-2yl)benzamide fragment was chosen from our previous results. This aromatic ring system was observed in LS-3-134 and other structural congeners having a high D3 affinity and excellent selectivity versus the D2 receptor [46][47][48]. 3b was N-alkylated with N- (3-bromopropyl) or N- (4-bromobutyl)phthalimide and then the protecting phthalimide 7a or 7b was hydrolyzed using hydrazine hydrate by heating for 3 h to give primary amine 8a or 8b (Scheme 2). 4-(Thiophen-2-yl)benzoic acid was converted to the corresponding acyl chloride using thionyl chloride at RT followed by treatment with 8a or 8b to give 9a or 9b in 50% or 20% yield, respectively. The triazole-thiol ether analogs were prepared by reduction of 5c and 5d to give alcohols 10a and 10b, which were converted to 11a and 11b using Mitsunobu reaction. The desired products 11a or 11b were obtained in 16% and 28% yield, respectively (Scheme 2).
Inspection of the structure of 9a reveals that two different benzamide fragments which share the tert-amine are capable of interacting with the OBS of the D2 and D3 receptors. Therefore, fragments 12 and 14 were synthesized for evaluation in in vitro binding The next series focused on preparing analogs having a spacer group with an aromatic ring system for interacting with the SBS. For the aromatic ring moiety, we tested 4-(thiophen-2-yl)benzamide or 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol. The 4-(Thiophen-2yl)benzamide fragment was chosen from our previous results. This aromatic ring system was observed in LS-3-134 and other structural congeners having a high D3 affinity and excellent selectivity versus the D2 receptor [46][47][48]. 3b was N-alkylated with N- (3-bromopropyl) or N- (4-bromobutyl)phthalimide and then the protecting phthalimide 7a or 7b was hydrolyzed using hydrazine hydrate by heating for 3 h to give primary amine 8a or 8b (Scheme 2). 4-(Thiophen-2-yl)benzoic acid was converted to the corresponding acyl chloride using thionyl chloride at RT followed by treatment with 8a or 8b to give 9a or 9b in 50% or 20% yield, respectively. The triazole-thiol ether analogs were prepared by reduction of 5c and 5d to give alcohols 10a and 10b, which were converted to 11a and 11b using Mitsunobu reaction. The desired products 11a or 11b were obtained in 16% and 28% yield, respectively (Scheme 2).
Inspection of the structure of 9a reveals that two different benzamide fragments which share the tert-amine are capable of interacting with the OBS of the D2 and D3 receptors. Therefore, fragments 12 and 14 were synthesized for evaluation in in vitro binding studies (Scheme 3). 12 was synthesized by N-methylation from the secondary amine 3b in Inspection of the structure of 9a reveals that two different benzamide fragments which share the tert-amine are capable of interacting with the OBS of the D 2 and D 3 receptors. Therefore, fragments 12 and 14 were synthesized for evaluation in in vitro binding studies (Scheme 3). 12 was synthesized by N-methylation from the secondary amine 3b in 19% yield. For 14, 4-(thiophen-2-yl)benzoic acid was conjugated with 3-bromopropylamine through acyl chlorination followed by N-alkylated with N-methylethanamine.
The next series probed the size of substituents on the tert-amine group. The pendent synthons for allyl (15a) and 4-fluorobenzyl (15b) were prepared from ethylenediamine (Scheme 4). For the synthesis of 15a, one of the primary amines was protected with a trifluoroacetyl group and the other primary amine alkylated with allyl bromide. The secondary amine was protected as a N-Boc and the trifluoroacetyl group was removed. 15b was synthesized in a similar method with 15a except a reductive amination with 4fluorobenzaldehyde was used. The prepared synthon 15a or 15b was conjugated with 1b, and the N-Boc was removed to give intermediates 17a,b. These intermediates were treated with N-propylphthalimide to give 18a,b. Removal of the phthalimide group with hydrazine hydrate gave corresponding N-propyl intermediate 19a (via reduction of the The next series probed the size of substituents on the tert-amine group. The pendent synthons for allyl (15a) and 4-fluorobenzyl (15b) were prepared from ethylenediamine (Scheme 4). For the synthesis of 15a, one of the primary amines was protected with a trifluoroacetyl group and the other primary amine alkylated with allyl bromide. The secondary amine was protected as a N-Boc and the trifluoroacetyl group was removed. 15b was synthesized in a similar method with 15a except a reductive amination with 4-fluorobenzaldehyde was used. The prepared synthon 15a or 15b was conjugated with 1b, and the N-Boc was removed to give intermediates 17a,b. These intermediates were treated with N-propylphthalimide to give 18a,b. Removal of the phthalimide group with hydrazine hydrate gave corresponding N-propyl intermediate 19a (via reduction of the N-allyl group) and the 4-fluorobenzyl analog 19b. The intermediates 19a and 19b were conjugated with 4-(thiophen-2-yl)benzoic acid to give the desired products 20a and 20c in 71% or 32% yield, respectively. For the N-allyl analog, 17a was directly N-alkylated with 13 to give 20b in 21% yield.  To investigate the nature of the aromatic moiety for binding to the SBS, aryl carboxamides 21b-j were synthesized using the same method described for the synthesis of 9a but using different aryl carboxylic acids and the naphthamide 21a was synthesized using 2-naphthoyl chloride using in the basic condition (Scheme 5). The desired benzamide analogs were obtained in yields ranging from 20 to 86%, respectively. The purity of all investigated compounds was confirmed prior to analysis and was greater than 95% on a 2695 Alliance LC-MS (Supplemental Table S1).
To investigate the nature of the aromatic moiety for binding to the SBS, aryl carboxamides 21b-j were synthesized using the same method described for the synthesis of 9a but using different aryl carboxylic acids and the naphthamide 21a was synthesized using 2-naphthoyl chloride using in the basic condition (Scheme 5). The desired benzamide analogs were obtained in yields ranging from 20 to 86%, respectively. The purity of all investigated compounds was confirmed prior to analysis and was greater than 95% on a 2695 Alliance LC-MS (Supplemental Table S1).

SAR Study
Two different assays were used to evaluate the properties of the analogs described above. The receptor binding affinity was measured by radioligand binding assays using [ 125 I]IABN with D2 or D3 receptors highly expressed HEK293 cells [49]. The functional activity of the analogs was determined using a β-arrestin recruitment assay. The assay was initially conducted in agonist binding mode to confirm that they function as antagonists at the D3 receptor. Once this efficacy was confirmed, the assay was conducted in antagonist mode to determine the ability of the antagonist to compete with dopamine at the D3 receptor. The results of the antagonist mode assay are reported as IC50 values [50][51][52]. Imax values were individually calculated from the assay and reliable with over 50% inhibition.
The first series of compounds evaluated were those synthesized in Schemes 1 and 2 ( Table 1). The dimethyl amino analogs 6a-d displayed a relatively low binding affinity for both D2 and D3 receptors. These data suggest that a basic amine moiety in the spacer group reduces affinity at both receptors. The observation that 6d had a 10-fold higher affinity than its structural congener 6b indicates that the Br-substituent is more preferred in the OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D2 and D3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D3 receptor for this scaffold. It is of interest to note that 9a had ~170-fold higher affinity at the D3 versus the D2 receptor.

SAR Study
Two different assays were used to evaluate the properties of the analogs described above. The receptor binding affinity was measured by radioligand binding assays using [ 125 I]IABN with D 2 or D 3 receptors highly expressed HEK293 cells [49]. The functional activity of the analogs was determined using a β-arrestin recruitment assay. The assay was initially conducted in agonist binding mode to confirm that they function as antagonists at the D 3 receptor. Once this efficacy was confirmed, the assay was conducted in antagonist mode to determine the ability of the antagonist to compete with dopamine at the D 3 receptor. The results of the antagonist mode assay are reported as IC 50 values [50][51][52]. Imax values were individually calculated from the assay and reliable with over 50% inhibition.
The first series of compounds evaluated were those synthesized in Schemes 1 and 2 ( Table 1). The dimethyl amino analogs 6a-d displayed a relatively low binding affinity for both D 2 and D 3 receptors. These data suggest that a basic amine moiety in the spacer group reduces affinity at both receptors. The observation that 6d had a 10-fold higher affinity than its structural congener 6b indicates that the Br-substituent is more preferred in the OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D 2 and D 3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D 3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D 3 receptor for this scaffold. It is of interest to note that 9a had~170-fold higher affinity at the D 3 versus the D 2 receptor.
It is important to note that 9a has two different modes in which it can bind to the D 3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D 2 = 89.2 ± 5.6 nM, D 3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D 2 and D 3 receptors (Ki D 2 > 1000 nM and D 3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D 3 receptor (IC 50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2yl)benzamide binds to the SBS. OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D2 and D3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D3 receptor for this scaffold. It is of interest to note that 9a had ~170-fold higher affinity at the D3 versus the D2 receptor. than its structural congener 6b indicates that the Br-substituent is more preferred in the OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D2 and D3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D3 receptor for this scaffold. It is of interest to note that 9a had ~170-fold higher affinity at the D3 versus the D2 receptor. than its structural congener 6b indicates that the Br-substituent is more preferred in the OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D2 and D3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D3 receptor for this scaffold. It is of interest to note that 9a had ~170-fold higher affinity at the D3 versus the D2 receptor. than its structural congener 6b indicates that the Br-substituent is more preferred in the OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D2 and D3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D3 receptor for this scaffold. It is of interest to note that 9a had ~170-fold higher affinity at the D3 versus the D2 receptor. than its structural congener 6b indicates that the Br-substituent is more preferred in the OBS than the corresponding fluoropropyl substituent. Compounds 9a,b and 11a,b, which have aromatic groups in the SBS, displayed a higher affinity at both D2 and D3 receptors. The 4-(thiophen-2-yl)benzamide analogs were more potent at the D3 receptor than the corresponding 4-methyl-5-phenyl-4H-1,2,4-triazole-3-thiol analogs. These data suggest that benzamides are preferred in the SBS of the D3 receptor for this scaffold. It is of interest to note that 9a had ~170-fold higher affinity at the D3 versus the D2 receptor. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. It is important to note that 9a has two different modes in which it can bind to the D3 receptor. The first mode has the bromobenzamide moiety binding to the OBS and the 4-(thiophen-2-yl)benzamide binding to the SBS. The second mode has the 4-(thiophen-2yl)benzamide binding to the OBS and the bromobenzamide moiety binding to the SBS. In vitro binding studies revealed that fragment 12 showed non-selectively high Ki values at both of dopamine receptor subtypes (Ki D2 = 89.2 ± 5.6 nM, D3 = 21.8 ± 5.1 nM), whereas 14 did not show any binding affinity at D2 and D3 receptors (Ki D2 > 1000 nM and D3 > 1000 nM). Moreover, the β-arrestin recruitment assay indicated that compound 18 is very potent for the D3 receptor (IC50 = 4.6 ± 1.2 nM). These data are consistent with the first mode that the bromobenzamide moiety binds to the OBS and the 4-(thiophen-2-yl)benzamide binds to the SBS. Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D2 and D3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D3 receptor versus the D2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS.  Table 2 shows the effect of the size of the N-alkyl group in the tert-amine on the D 2 and D 3 receptor binding. Our results indicate that the N-ethyl substituent 9a showed the highest binding affinity and subtype selectivity at the D 3 receptor versus the D 2 receptor. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D 2 and D 3 receptors (Table 2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC 50 = 14.0 ± 7.4 nM) > 20a (IC 50 = 26.5 ± 12.9 nM) > 20b (IC 50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS.
A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D 3 affinity, but the overall D 3 versus D 2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D 3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D 2 affinities ranged between 107 and 525 nM, resulting in a D 3 selectivity ratio (i.e., D 2 /D 3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have~1 nM affinity for the D 3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10-fold higher than that of 21a (IC 50 = 1.3 vs. 16.4 nM). 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. There was a slight decrease in affinity in going from propyl to allyl groups, whereas the 4-fluorobenzyl group resulted in a large loss in affinity at both D2 and D3 receptors ( Table  2). When the size of substituents was increased, the binding affinity and subtype selectivity was decreased. This reduction in affinity also translated to the β-arrestin recruitment assay. That is, there was a trend of decreased potency in the order of 9a (IC50 = 14.0 ± 7.4 nM) > 20a (IC50 = 26.5 ± 12.9 nM) > 20b (IC50 = 51.6 ± 40.8 nM). Based on this SAR study, the N-ethyl group is the preferred alkyl group with respect to binding to the OBS. A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM).  A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM). A number of compounds were prepared to explore the nature of the interaction between the aromatic ring and the SBS. Previous studies with the N-aryl piperazine analogs revealed that a wide range of aromatic rings are tolerated in the SBS with respect to D3 affinity, but the overall D3 versus D2 selectivity can be influenced by the nature of this interaction. The results of this study are shown in Table 3. All compounds had good affinity at D3 receptors, with Ki values ranging between 0.8 and 13.2 nM. The D2 affinities ranged between 107 and 525 nM, resulting in a D3 selectivity ratio (i.e., D2/D3 ratio) ranging from 22.1-to 180-fold. The effect of the nature of the aromatic ring in the SBS on the ability of the antagonist to compete with dopamine in the β-arrestin assay was somewhat unexpected. For example, both 21a and 21c have ~1 nM affinity for the D3 receptor in the radioligand binding assay, but the potency of 21c in the β-arrestin recruitment assay was 10fold higher than that of 21a (IC50 = 1.3 vs. 16.4 nM).

Molecular Docking and Molecular Dynamics Simulations (MDS)
To understand the favorable binding profiles of the metoclopramide analogs, molecular docking and MDS studies were performed using different N-alkyl compounds (9a, 20a, 20b, 20c and 21c) with the D 3 receptor (PDB: 3PBL) ( Table 4). These compounds were chosen because they are close structural analogs and have a wide range in D 3 receptor affinity (1-300 nM). As reported in previous studies [13,29,53], the binding pose that formed a bridge hydrogen bond between the carboxylate of ASP110 3.32 and the protonated nitrogen was considered to be critical for high binding affinity for the D 3 receptor. The distance between the protonated nitrogen ranged between 2.6 and 2.9 Å, and 9a was found to have the closest interaction (2.6 Å). The estimated binding energies were not significantly different for each compound (−9.74 to −10.22 kcal/mol). Therefore, the difference in D 3 affinity of the five compounds cannot be explained by the distance between ASP110 3.32 and the protonated nitrogen atom, and the calculated binding energies from docking studies. In MDS studies, the root mean square distance (RMSD) was calculated over 50-200 ns in five copies of the MDS production ( Table 4). The first time frame (0 ns) of the production run was used as the reference position to determine the stability of each compound in the binding site. 21c presented the lowest standard deviation of RMSD (2.45 ± 0.49 Å) indicating the least amount of movement in the binding site. A relatively higher amount of motion (3.00 ± 0.82 Å) with 20c is consistent with the lower binding affinity for D 3 receptors. These results indicate the MDS studies correlate better with D 3 affinity than the results of docking studies.
The representative binding pose of the MDS production run is displayed in Figure 1. Within the OBS of the D 3 receptor, all the selected compounds were engaged in multiple interactions. The hydrogen bond with ASP110 3.32 and π staking interactions with PHE345 6.52 were observed with all five compounds. However, a halogen bond between VAL189 5.39 and the bromine of the 5-bromo-2,3-dimethoxybenzene moiety was observed for 9a, 21c, and 20c (Figure 1a,b,e, respectively). It is of interest to note the 21c, the most potent compound in the β-arrestin recruitment assay, which predicts the ability to compete with endogenous dopamine, had a cation-π interaction between the protonated nitrogen and PHE106 3.28 residue (Figure 1b). Mol. Sci. 2023, 23, x FOR PEER REVIEW 9 which exhibited the lowest binding affinity for D3 receptors, was lower than the hig affinity compounds. As mentioned above, 21c showed a high frequency of contacts PHE106 3.28 including approximately 95% of hydrophobic interactions and 10% of ca π interactions over the MDS production runs.  The summary of overall frequency of contacts from the MDS studies, including hydrophobic interactions, hydrogen bonds, the salt bridge, halogen bonds, and π-interactions, is shown in Figure 2. All five compounds formed stable interactions (frequency of contact > 0.6) with most of residues in the OBS (i.e., ASP110 3.32 , VAL111 3.33 , CYS114 3.36 , SER196 5.46 , PHE345 6.51 , and THR369 7.39 ). The frequency of all interactions in the OBS of 20c, which exhibited the lowest binding affinity for D 3 receptors, was lower than the higher-affinity compounds. As mentioned above, 21c showed a high frequency of contacts with PHE106 3.28 including approximately 95% of hydrophobic interactions and 10% of cation-π interactions over the MDS production runs.
Consistent with previous modeling studies, the formation of key interactions between ASP110 3.32 and the protonated nitrogen of the ligand stabilized the binding pose of 9a, 20a, 20b, and 20c (frequency of contact > 0.998) by 97.8% to 99.4% of the hydrogen bond formation. However, the frequency of contacts between ASP110 3.32 and 20c was relatively lower (frequency of contact = 0.990) and formed only 68.6% of hydrogen bonds over the MDS production runs.
In the SBS, 9a, 20a and 21c that exhibited high subtype selectivity, presented a moderate to high probability (frequency of contacts = 0.4-0.9) of interaction with VAL86 2.61 , LEU89 2.64 , GLY93 EL1 , and GLY94 EL1 . In addition, the pyridine of 21c formed a hydrophobic interaction with GLU90 2.65 (frequency of interaction = 0.563). In contrast to our expectations, 90% of the hydrophobic interactions that formed with VAL86 2.61 were from the 4-fluorobenzyl group whereas 10% of the interactions were from the 4-(thiophen-2-yl)benzamide moiety.
The average frequency of the overall interactions in the binding sites (i.e., OBS and SBS) was correlated with D 3 receptor binding affinity (r = −0.8756, and p = 0.0517). In addition, the average frequency of interaction in the OBS was significantly correlated with the IC 50 values from the β-arrestin recruitment assay (r = −0.9934, and p = 0.0066).  a) 9a, (b) 21c, (c) 20a, (d) 20b, and (e) 20c. The predicted interactions of each compound with residues in the OBS and the SBS of the D3 receptor were distinguished by the color. Red: hydrogen; cyan: π-interactions; green: halogen bond.

Comprehensive Screening for Other GPCRs
Based on the results in the dopamine receptor radioligand binding assays, nine flexiblebased compounds were selected for further evaluation for off-target binding with other GPCRs through the Psychoactive Drug Screening Program (PDSP) (Supplemental Table S2) [54]. Previous studies with the N-aryl piperazine analogs showed high binding affinity for serotonin 5-HT 1A and 5-HT 2B receptors. For example, many of the N-aryl piperazine-based analogs that our group developed in the past for either the D 2 or D 3 receptor had high affinity for the 5-HT 1A receptor [55][56][57][58]. It is of interest to note that none of the panel submitted for evaluation had a high affinity for the 5-HT 1A receptor or any of the other GPCRs in the screening assay (Supplemental Table S2). Compounds 21a, 21c, and 21i had modest affinity for the 5-HT 3 receptor (Ki values 29-58 nM). Furthermore, a relatively high affinity of compounds 20a, 21a, 21e, 21g, and 21i for the peripheral benzodiazepine receptor (PBR) was observed. This mitochondrial-based protein is typically used as a target for imaging neuroinflammation. The results of the PDSP-binding assays also confirmed the data obtained in our lab for the binding of this panel of nine compounds to D 2 and D 3 receptors (Supplemental Table S2).

Discussion
The goal of the current study was to identify a new scaffold for D 3 -selective antagonists that must display a high affinity and selectivity for D 3 versus D 2 receptors in the radioligand binding assays, but also a high potency in a β-arrestin recruitment assay, which measures the ability of a compound to compete with dopamine in binding to the D 3 receptor [21,29]. Previous studies have shown that a PET radiotracer developed in our lab having a high affinity for the D 3 receptor (Kd~50 pM) and excellent selectivity versus the D 2 receptor (>150-fold) was not able to image D 3 receptors in vivo without pretreatment with drugs that reduce synaptic levels of dopamine [59].
For the current study, we chose metoclopramide as the lead compound for our SAR studies. Metoclopramide was chosen as the lead compound for this study because it has a modest affinity for both D 2 and D 3 receptors and it should be possible to make analogs of this compound having an improved D 3 binding affinity while minimizing D 2 receptor affinity by interacting with the SBS. The results of SAR indicated that 5-bromo-2,3-dimethoxybenzamide, the moiety from FLB457, was more favorable for binding to the OBS, which is important for determining affinity for both D 3 and D 2 receptors. The size of fragments in 9a,b or 11a,b that interact with the SBP residues of the D 3 receptor are important for high selectivity for D 3 versus D 2 selectivity [60]. It is of interest to note that the appropriate length of linker between the basic amine and the secondary binding fragment was one carbon shorter than other known D 3 receptor antagonists such as Narylpiperazine congeners. The D 3 receptor binding affinity was also affected by steric hindrance of the substituent on the basic amine.
A number of the compounds reported here exhibited excellent D 3 binding affinity (ranging from 0.8 to 13.2 nM) and excellent selectivity (22.1-to 180-fold) for D 3 vs. D 2 receptors. Although analogs such as 9a, 21a, 21d, and 21g exhibit high binding affinity and subtype selectivity for the D 3 receptor, 21c was identified as the best-in-series candidate because of its high D 3 affinity and selectivity, and excellent potency in the β-arrestin recruitment assay (IC 50 = 1.3 nM). This IC 50 value was comparable with fallypride that is widely used as a non-selective PET probe for D 2 /D 3 receptors and can bind to D 3 receptors in the presence of endogenous dopamine (fallypride, IC 50 = 1.7 nM) [29]. Moreover, the computational modeling studies demonstrated that the high potency of 21c may result from the short distance of the bridge-bond with ASP110 3.32 and the high-frequency contacts between 21c and residues in the OBS and SBS in the D 3 receptor.
Since metoclopramide was previously used in drug development and led to the identification of compounds having a diverse range of pharmacologic activity including mixed 5-HT 3 antagonists/5-HT 4 agonists (e.g., zacopride, BRL 24682) and D 2 antagonists (e.g., clebopride, BRL 25594) [45], there was a concern that the conformational flexibility of our compounds could result in significant off-target bindings to other G-proteins. By the comprehensive screening from PDSP, these compounds possess minimal affinity for other GPCRs except a moderate affinity for 5-HT 3 receptors (29-58 nM). Interestingly, 21d, which has an indole carboxamide as a secondary binding fragment, exhibited nM binding affinity for the histamine H 1 receptor (0.95 nM). Other compounds acquired affinity for the translocator protein (TSPO); however, it is not clear if this off-target binding would be problematic for using these compounds in D 3 receptor binding assays or behavioral studies. Further studies are ongoing in our lab to prepare radiolabeled versions of 21c for imaging D 3 receptors in the brain, and SAR studies are being conducted that aim to improve the properties of this new scaffold as a means of identifying potential D 3 receptor selective PET radiotracers.

General
5-(3-Fluoropropyl)-2,3-dimethoxybenzoic acid (1a) was prepared from methyl 5-allyl-3-methoxy salicylate via methylation for phenol, oxidation of the allyl group, fluorination, and hydrolysis of methyl ester [61]. 5-Bromo-2,3-dimethoxybenzoic acid (1b) was prepared from 5-bromo-2-hydroxy-3-methoxy benzoic acid via methylation for phenol and oxidation of aldehyde to carboxylic acid using silver catalyst [62]. For the OBS binding, tert-butyl (2-aminoethyl) ethylcarbamate was prepared from N-ethylethylenediamine via the primary amine protection, the secondary amine protection and the primary amine de-protection according to the literature [63]. 2-(2-Bromoethyl)-1,3-dioxolane and 2-(3-bromopropyl)-1,3dioxolane were prepared via reduction followed by cyclization [64]. The other reagents and solvents were purchased from Sigma-Aldrich, TCI, Matrix Scientific, Advanced chemtech, Fisher chemical, Ambeed, Chembridge corporation, Acros organics, and Decon laboratories and used as received (Supplemental Table S3). Reactions were monitored by thin layer chromatography (TLC) using TLC silica gel 60W F254S plates and the spots were detected under UV light (254 nm) or developed using ninhydrin. Flash column chromatography was carried out on a Biotage Isolera One with a dual wavelength UV-vis detector. 1 H and 13 C NMR spectra were obtained on a Bruker NEO-400 spectrometer (Bruker, Billerica, MA, USA). Chemical shifts (δ) were recorded in parts per million (ppm) relative to the deuterated solvent as an internal reference. Mass spectra (m/z) were recorded on a 2695 Alliance LC-MS (Waters Corporation, Milford, MA, USA) using positive electrospray ionization (ESI + ). High resolution mass spectra (HRMS, m/z) were acquired on a waters LCT premier mass spectrometer (Waters Corporation, Milford, MA, USA). PathHunter TM β-arrestin recruitment assay kit and the Chinese hamster ovary CHO-K1 cell line were purchased from DiscoverX (Fremont, CA, USA).  (15 mL, 196 mmol) was slowly added at 0 • C. The reaction mixture was warmed to RT and stirred for 1 h. The volatiles were removed followed by the residue was dissolved in CH 2 Cl 2 and the organic layer washed by aq saturated NaHCO 3 solution. The inorganic layer was extracted by CH 2 Cl 2 and the combined layer was washed by brine, dried over anhydrous MgSO 4 , filtered and concentrated in vacuo to afford 3a (850 mg, 98% yield) as a yellow oil. The crude product was used for the next step without further purification.  Bromo-N-(2-(ethylamino)ethyl)-2,3-dimethoxybenzamide (3b) 3b was synthesized using 2b (8 g, 18.55 mmol) in the same procedure as 3a and purified by flash chromatography on silica gel (CH 2 Cl 2 /7 N NH 3 in MeOH = 20:1) to afford 3b (6 g, 98% yield) as a yellow oil.      After completion of the reaction, the mixture was diluted with EtOAc and washed by aq saturated NaHCO 3 solution and brine. The organic layer was dried over Na 2 SO 4 , filtered and concentrated in vacuo. The crude product was purified by flash chromatography on silica gel (CH 2 Cl 2 /7 N NH 3 in MeOH = 20:1) to afford 10a (300 mg, 51% yield) as a colorless oil.    tert-Butyl allyl(2-aminoethyl)carbamate (15a) In a solution of ethylenediamine (2.8 mL, 41.6 mmol) in 100 mL of CH 2 Cl 2 , ethyl trifluoroacetate (4.9 mL, 41.6 mmol) in 100 mL of CH 2 Cl 2 was added dropwise at 0 • C. The mixture was warmed to RT and stirred for 1 h. The solvent was removed under the reduced pressure and the residue was dissolved with 210 mL of MeOH. Allyl bromide (3.6 mL, 41.6 mmol) and Et 3 N (6.4 mL, 46 mmol) were added slowly into the mixture and the mixture was stirred for 16 h. Then, (Boc) 2 O (9.6 mL, 41.6 mmol) was added and the mixture was stirred for another 4 h. The volatiles were removed under the reduced pressure and the residue was dissolved EtOAc. The organic layer was washed by aq 0.5 N HCl solution and brine, dried over MgSO 4 , filtered and concentrated in vacuo. Deprotection of trifluoroacetyl group was performed according to the reported method [63] and 2.8 g of 15a (34% yield) was obtained as a yellow oil. The crude product was used for the next step without further purification.

Chemistry
tert-Butyl (2-aminoethyl)(4-fluorobenzyl)carbamate (15b) In a solution of N-(2-aminoethyl)-2,2,2-trifluoroacetamide (6.5 g, 41.6 mmol) in 200 mL of CH 2 Cl 2 , 4-fluorobenzaldehyde (4.46 mL, 41.6 mmol) and sodium triacetoxyborohydride (17.6 g, 83 mmol) were added. The mixture was stirred for 16 h at RT and washed by aq saturated NaHCO 3 solution and brine. The organic layer was dried over MgSO 4 , filtered and concentrated in vacuo. The crude product was purified by flash chromatography on silica gel (CH 2 Cl 2 /7 N NH 3 in MeOH = 20:1) to afford an intermediate (1.7 g, 6.5 mmol) as a colorless oil. Protection of Boc group and deprotection of trifluoroacetyl group were performed according to the reported method [63] and 1.6 g of 15b (15% yield) was obtained as a colorless oil. The crude product was used for the next step without further purification.

β-Arrestin Recruitment Assay
CHO-K1 cells which were overexpressed human D 3 receptors were cultured in assaycomplete TM cell culture kit 107. Cells were seeded at a density of 25,000 cells per well of 96-well plate, and incubated at 5% CO 2 , 37 • C. Two days later, test compounds were dissolved in DMSO, and diluted with 11-point series in phosphate-buffered saline (PBS). Prepared compounds were added to the cells, and it was incubated for 30 min at 5% CO 2 , 37 • C. Then, cells were treated with 30 nM (EC 80 ) of dopamine, and the plate was incubated another 90 min. PathHunter TM detection reagent was added to each well, and then plate was incubated for 80 min at RT in the dark. The chemiluminescent signal was measured by PerkinElmer Enspire plate reader (PerkinElmer, Boston, MA). Data were analyzed by Prism followed by non-linear regression.

Molecular Docking and Molecular Dynamics Simulations (MDS)
The 4 compounds with different N-alkyl groups (9a, 20a, 20b, and 20c) and the best candidate 21c were selected and performed for molecular docking and MDS studies on the D 3 receptor. The protonated status at physiological pH of each compound was predicted by using Open Babel v3.1.0 [65]. Then, the molecular docking studies and MDS were performed by following the previous protocols [29]. In brief, molecular docking studies were performed via the AutoDock 4.2 [66] plugin on PyMOL (pymol.org). The X-ray structure of the D 3 receptor (PDB: 3PBL, resolution: 2.89 Å) was obtained from the RCSB Protein Data Bank (www.rcsb.org (accessed on 19 May 2022)). Heteroatoms were removed from the crystal structure and polar hydrogens were added. Non-polar hydrogens were removed from selected compounds. A grid box with a dimension of 30 × 30 × 28.2 Å 3 was applied for covering OBS and SBS bindings. The Lamarckian Genetic Algorithm with a maximum of 2,500,000 energy evaluations was used to calculated 100 protein-ligand binding poses for each compound. The D 3 receptor−ligand complex that reproduced the crystallographic ligand binding pose with good docking score was subjected for the evaluation. The CHARMM-GUI web-sever [67] was used for MDS preparation. The topology and parameter files of protonated compounds were generated by the Ligand Reader and Modeler module [68,69]. The Bilayer Membrane Builder module [70,71] was used for building the MDS system with FF19SB force field. The protein-ligand complexes generated from docking studies were aligned to the D 3 receptor structure obtained from the Orientations of Protein in Membranes (OPM) database [72], and the POPC membrane were placed by using the OPM D 3 receptor model. The protein, ligand, and membrane complexes were solvated in a TIP3P water box, and then Monte-Carlo sampling was used to add 0.15 M NaCl for charge neutralization. The MDS studies were performed via Amber18 [73] on the high-performance computing (HPC) cluster at Center for Biomedical Image Computing and Analytics at the University of Pennsylvania. The input files of system minimization, 6 steps equilibration including 2 steps NVT ensemble and 4 steps NPT ensemble, and 5 copies of 200 ns production run for MDS were generated from the last step of Membrane Builder [70,71] on the CHARMM-GUI web-sever [67].
The 50 to 200 ns of production simulation with a total of 7500 frames (1500 frames of 5 production simulation copies) for each compound were used for further MDS analysis. The interactions between a ligand and protein in the production simulations were calculated by using the software BINANA v2.1 [74].

Conclusions
A new scaffold was designed based on metoclopramide and identified having high affinity and subtype selectivity for the D 3 receptor versus the D 2 receptor. Initially, 9a having 4-(thiophen-2-yl)benzamide was recognized as a lead compound showing high binding affinity and subtype selectivity for the D 3 receptor (Ki D 2 = 169 nM and D 3 = 1 nM). Although different aryl carboxamides exhibited excellent binding affinities preferring D 3 receptors, 21c was the most potent (IC 50 = 1.3 nM) for competing with dopamine in the β-arrestin recruitment assay. Furthermore, the comprehensive screening of 21c revealed the minimal off-target binding for other CNS targets. Molecular docking or MDS demonstrated that interactions between 21c and the D 3 receptor were comparable with fallypride that was known for potent D 2 /D 3 antagonists. These results suggested that 21c may have a greater potential for competing with synaptic dopamine for binding to the D 3 receptor. Overall, this novel scaffold can be developed as high-affinity D 3 receptor antagonists that bind with low affinity at D 2 receptors and other CNS receptors.