Synthesis and evaluation of cyclic diamino benzamide based D3 receptor ligands

Dopamine (1) is a key neurotransmitter whose impact on pharmacological processes is mediated by a family of dopamine receptors designated D1, D2, D3, D4, and D5. Various diseases and conditions such as schizophrenia, drug abuse, depression, restless leg syndrome, Parkinson’s disease (PD), and inflammatory diseases have been linked to aberrant D3 activity. Herein, we report a series of novel D3 ligands with improved solubility over our previous lead compound, MC25-41 (2).


Introduction
The key neurotransmitter known as dopamine (1) was prepared synthetically for the first time in 1910 by George Barger and James Ewens [1] long before its pharmacological role was recognized. Over 45 years later, Katharine Montagu determined that dopamine was present in the human brain [2], and in 1958 Arvid Carlsson and Nils-Åke Hillarp demonstrated that this chemical acts as a neurotransmitter [3]. Over the next several decades, the pharmacological function and the means through which dopamine exerts its impact on biological systems has been elucidated. It is known, for example, that dopamine (1) is synthesized in the brain and the periphery, and it has been conclusively linked to a wide range of physiological functions. These functions include vasodilation, modulation of renal sodium excretion, altering urine output, learning, movement, and behavioral motivations [4].
At the cellular level, dopamine signaling is mediated by a family of G-protein coupled receptors (GPCRs) that are designated as D 1 , D 2 , D 3 , D 4 , and D 5 . The D 1 -like (D 1 and D 5 ) and D 2 -like (D 2 , D 3 and D 4 ) sub-families are based on genetic organization, amino acid homology, and pharmacological properties of the individual family members [5]. The D 3 receptor has been the subject of intense interest as a potential therapeutic target, as it has been linked to various disease states and conditions such as schizophrenia, drug abuse, depression, restless leg syndrome [6], Parkinson's disease (PD) [7,8], and various inflammatory diseases [9]. We recently described our effort to identify novel, selective D 3 ligands with potential utility for the treatment of cocaine use disorder. These studies led to the identification of MC25-41 (2, Fig. 1) as a potent D 3 ligand that (1) possesses a high level of selectivity for D 3 over other dopamine receptors and a range of other key CNS targets, (2) has a pharmacokinetic profile suitable for in vivo studies [10], and (3) attenuates motivation for cocaine in Sprague-Dawley rats [11]. While MC25-41 (2) has proven to be an effective tool molecule, its limited solubility (2 µM) may be lead to future problems in formulation and dosing [12]. As part of an effort to address this issue, we have developed a new series of novel, selective D 3 ligands whose solubility is significantly improved over our original lead compound (2). Specifically, we have examined the impact of inserting an ether moiety in the linker chain. This provides additional opportunities to form hydrogen bonds with water and increases aqueous solubility. In addition, we have explored replacing the piperazine ring with two bioisosteres, homopiperazine and 2,6-diazaspiro [3.3]heptane. These alternative rings systems oriented their embedded nitrogen atoms in manners significantly different from the orientation of the corresponding piperazine nitrogen atoms. which leads to increased basicity and increased capability to form hydrogen bonds with water. Both of these factors led to increased aqueous solubility for the homopiperazine and 2,6-diazaspiro [3.3]heptane analogs.

Results and discussion
Our effort to improve the solubility of our lead series began with the incorporation of an oxygen atom in the central linker region (3a-3k). This addition adds an additional hydrogen bond acceptor, which could improve the solubility of our compounds, but also increases the length of the linker, which could alter D 3 binding affinity and selectivity. Target compounds were prepared as outlined in Scheme 1. DMT-MM mediated coupling of biaryl acid (4) with aminoalcohol (5) provided the corresponding amide, which was converted to the corresponding bromide (6) with carbon tetrabromide and triphenylphosphine. The bromine was displaced with an aryl piperazine (7) to provide the target compounds (3a-3k). Table 1 includes the in vitro binding (K i at D 3 and D 2 ) as well as the physicochemical properties (MW, TPSA, LogP) of target compounds (3a-3k). Table 2 provides a comparison of the solubility of the target compounds (3a-3k) with the corresponding MC25-41 analogs (8a-8k) The compounds prepared and tested have MW, TPSA, and cLogP values that are consistent with drug like properties, but all examples have either TPSA (3a, 3c) or cLogP (3b, 3d-3k) value outside of the range suggested of BBB penetration (TPSA < 90, cLogP = 2-4) [13]. It is noteworthy, however, that we have previously demonstrated that MC25-41 (2) is efficacious in rat models of cocaine addiction (10 mg/kg IP) [11] and a marmoset model of Parkinson's disease (10 mg/ kg PO) [14]. This indicates that MC25-41 (2) is able to penetrate the BBB despite the fact that its clogP (4.9) value are outside of the range suggestive of BBB penetration and Scheme. 1 Synthesis of (3a)-(3K) its and TPSA (87.6) is on the boarderline. It further suggests that related compounds may also be able to penetrate the BBB despite out-of-range values for cLogP and TPSA.
The structure-activity relationship analysis began with the 3-cyano analog (3a), which provided a direct comparison with our previous lead compound MC25-41 (2). A decrease in both D 3 binding affinity (K i = 128 nM) and selectivity over D 2 (13-fold) were observed. Similar D 3 potency was observed when the 3-CN (3a) was replaced with a 3-CF 3 (3b, D 3 K i = 175 nM), but selectivity decreased (3.2-fold). Relocation of the cyano substituent to the 2-position (3c) once again produced a compound with similar D 3 potency (K i = 140 nM) and selectivity (3.5-fold). Replacing the 2-CN (3c) with either a 2-Cl (3d) or 2-CF 3 (3e) led to an increase in both D 3 potency (K i = 27 nM and 37 nM respectively) and selectivity (11-fold and 17-fold respectively versus D 2 ). Addition of a second chlorine atom in either the 3-position (3f) or the 4-position (3g) caused a drop in D 3 potency (K i = 127 nM and 1653 nM respectively) and selectivity (3.6-fold and 1.3-fold respectively versus D 2 ). D 3 binding affinity was restored when the 2 chlorine atoms were relocated to the 3-and 5-positions (3h, D 3 K i = 22 nM), and selectivity over D 2 improved (D 2 K i = 332 nM, 14.9-fold). Incorporation of a methoxy group produced mixed results depending on the positioning of the substituent. The 2-methoxy analog (3i) is a potent D 3 ligand (K i = 24 nM), but selectivity was marginal (3.1-fold). The 4-methoxy analog (3j), on the other hand, showed little binding affinity for both D 3 (K i = 15265 nM) and D 2 (K i = 25435 nM). Finally, the 1-napthyl analog (3k) had moderate D 3 binding affinity (K i = 79 nM) and low D 2 selectivity (4fold).
We next turned our attention to replacing the piperazine ring of our lead compound MC25-41 (2) with either a homopiperazine (9a-9i) or a 2,6-diazaspiro [3.3]heptane (9j, 9k). Both of these moieties have been effectively used as piperazine bioisosteres [15], but there are differences that could impact D 3 binding and selectivity (Fig. 2). While the distance between the nitrogen atoms in a piperazine ring (10) and homopiperazine ring (11) is similar (2.86 Å versus 2.89 Å), this is not the case for 2,6-diazaspiro [3.3]heptane (12). The distance between nitrogen atoms is substantially larger (4.17 Å). In addition, the 3-dimensional shapes of the three rings systems are not the same. The shape of the piperazine ring (10) and homopiperazine (11) are similar in that the first exists in a standard chair confirmation, while the second exists in pseudo-chair conformation. The chair confirmation of the piperazine ring (10) allows for significant cross-ring interaction of the loan pairs of electrons of the two nitrogen atoms, but this interaction is substantially decreased in the homopiperazine ring (11) due to its pseudo-chair conformation. This difference makes the homopiperazine nitrogen atoms more basic and more available to participate in hydrogen bonding interactions that could increase solubility. In a similar but more pronounced manner, the nitrogen atoms of 2,6-diazaspiro[3.3]heptane ring system (12) are oriented very differently from the piperazine ring (10). In this instance, the nitrogen atoms are oriented so that their lone pairs of electrons are perpendicular to each other and therefore incapable of undergoing the interaction seen in the chair configuration of the piperazine ring (10). As a result the 2,6diazaspiro [3.3]heptane (12) nitrogen atoms are more basic and more available to participate in hydrogen bonding interactions that could increase solubility. The homopiperazine (9a-9i) and the 2,6-diazaspiro[3.3] heptane (9j, 9k) analogs were prepared by the methods described in Schemes 2 and 3. Buchwald coupling of homopiperazine (11) with an aryl bromide (13) provided the requisite aryl homopiperzines (14a-14i), which were reacted with alkyl bromide (15) under basic conditions to provide target compounds (9a-9g). The synthesis of pyridine analogs (9h) and (9i), on the other hand, began with a displacement reaction of homopiperazine (11) with the corresponding chloropyridine (16a or 16b) to provide aryl homopiperzines (14h) and (14i), This was followed by reaction with alkyl bromide (15) under basic conditions to provide target compounds (9j) and (9k). In a similar manner, the synthesis of 2,6-diazaspiro [3.3]heptane (9j) and (9k) began with the reaction of Boc-2,6-diazaspiro [3.3] heptane (17) with a chloropyridine (16a or 16b) followed by TFA mediated deprotection to provide (18a) and (18b). Reaction with alkyl bromide (15) under basic conditions to provide target compounds (9j) and (9k). Table 3 includes the in vitro binding (K i at D 3 and D 2 ) as well as the physicochemical properties (MW, TPSA, LogP) of target compounds (9a-9k). Table 4 provides a comparison of the kinetic solubility of the target compounds (9a-9k) with the corresponding analogs of MC25-41 (8a-8c, 8e, 8g, 8h, 8k-8o) The majority compounds prepared and tested have MW and TPSA values that are consistent with drug-like properties and suggestive of BBB penetration. The cyano-pyridine analogs (9h) and (9j) are notable exceptions, as their TPSA are 101. cLogP values of the majority of compounds are above the range suggested for orally delivered compounds. Notable exceptions include (9h), (9j), and (9k), and only one compound, (9j), has a cLogP value within the targeted range that suggests BBB Scheme. 2 Synthesis of (9a)-(9i) Scheme. 3 Synthesis of (9j) and (9k) Fig. 2 Structures of (10), (11), and (12) penetration. Overall, replacing the piperazine ring with a homopiperazine ring produced better results than those observed with 2,6-diazaspiro [3.3]heptane with respect to D 3 binding and selectivity over D 2 . Incorporation of a cyano substituents in either the 2-position (9a) or 3-positions (9c) produced compounds that were highly potent (K i = 5.7 nM and 6.3 nM) and highly selective (139-fold and 114-fold versus D 2 ). Replacing the cyano substituents with a CF 3 group in either the 2-postion (9b) or 3-position (9d) led to a small decrease in D 3 binding potency (K i = 30.2 nM and 18.6 nM), but selectivity over D 2 was substantially diminished (43-fold and 47-fold). The 2,4-di-chloro (9e) and 3,5di-chloro (9f) analogs were also less potent D 3 ligands (K i = 16.7 nM ad 33.1 nM) and less selective over D 2 (80fold and 27-fold) than the 3-CN analog (9a). Replacing the benzene rings of (9a) and (9b) with a pyridine ring led to a further reduction of D 3 binding potency (9h D 3 K i = 64.7 nM, 9i K i = 70.2 nM), but selectivity over D 2 increased (197-fold and 131-fold). Lastly, incorporation of the 2,6diazaspiro[3.3]heptane ring system (9j and 9k) led to a significant decrease in D 3 binding affinity (K i = 1093 nM ad 464 nM respectively).

demonstrated improvements in
Follow-up studies have demonstrated that this compound is highly selective for D 3 over D 4 (D 4 K i = 1077 nM). Additional studies on this compound (9c) to assess its affinity for D 1 and D 5 , as well as full in vitro ADME profiling (e.g., Cyp450 inhibition, mouse and human liver microsome stability, permeability) are on-going.

Competitive radioligand-binding studies
For competitive binding studies, transfected HEK293 cell homogenates were suspended in homogenization buffer and incubated with radioligand [ 125 I]IABN, in the presence or absence of inhibitor at 37°C for 60 min with [ 125 I]IABN (total volume = 150 ul) as previously described [17]. Competitive radioligand studies were performed to determine the concentration of inhibitor that inhibits 50% of the specific binding of the radioligand (IC 50 value). The final radioligand concentration was approximately equal to the K d value for the binding of the radioligand. For each competition curve, triplicates were performed using two concentrations of inhibitor per decade over five orders of magnitude. Binding was terminated by the addition of cold wash buffer (10 mM Tris-HCl/150 mM NaCl, pH = 7.5) and filtration over a glass-fiber filter (Pall A/B filters, #66198). A Packard Cobra Gamma Counter was used to measure the radioactivity of [ 125 I]IABN.
The competition curves were modeled for a single binding site using where Bs is the amount of ligand bound to receptor and Bo is the amount of ligand bound to receptor in the absence of competitive inhibitor. L is the concentration of the competitive inhibitor. The IC 50 value is the concentration of competitive inhibitor that inhibits 50% of the total specific binding. IC 50 values were determined using nonlinear regression analysis with

Aqueous solubility assay
Compounds were assessed for their solubility at pH 7.4 using the commercially available Millipore MultiScreenTM Solubility filter system (Millipore, Billerica, MA). Analysis was performed by liquid chromatography tandem mass spectrometry (LC/MS/MS).