2-(Fluoromethoxy)-4'-(S-methanesulfonimidoyl)-1,1'-biphenyl (UCM-1306), an Orally Bioavailable Positive Allosteric Modulator of the Human Dopamine D1 Receptor for Parkinson's Disease.

Tolerance development caused by dopamine replacement with l-DOPA and therapeutic drawbacks upon activation of dopaminergic receptors with orthosteric agonists reveal a significant unmet need for safe and effective treatment of Parkinson's disease. In search for selective modulators of the D1 receptor, the screening of a chemical library and subsequent medicinal chemistry program around an identified hit resulted in new synthetic compound 26 [UCM-1306, 2-(fluoromethoxy)-4'-(S-methanesulfonimidoyl)-1,1'-biphenyl] that increases the dopamine maximal effect in a dose-dependent manner in human and mouse D1 receptors, is inactive in the absence of dopamine, modulates dopamine affinity for the receptor, exhibits subtype selectivity, and displays low binding competition with orthosteric ligands. The new allosteric modulator potentiates cocaine-induced locomotion and enhances l-DOPA recovery of decreased locomotor activity in reserpinized mice after oral administration. The behavior of compound 26 supports the interest of a positive allosteric modulator of the D1 receptor as a promising therapeutic approach for Parkinson's disease.

Introduction

Parkinson’s disease (PD) is a chronic, progressive, neurodegenerative disorder characterized by the loss of dopaminergic neurons in the brain. In addition to the main symptoms derived from motor dysfunction, PD is associated with memory loss and eventually dementia, due to the disruptive effect on dopaminergic transmission in the hippocampus and the prefrontal cortex.[1,2] Dopamine (DA) replacement with l-DOPA was proposed about 60 years ago[3] and is still the primarily used therapy and most effective treatment against the debilitating motor symptoms of PD. However, in the long term, the therapeutic index of l-DOPA decreases and its anti-Parkinsonian action is very often associated with progressive decline in symptomatic benefit (motor fluctuations), development of dyskinesia, and other adverse effects, including hypotension, hallucinations, and gastrointestinal disturbances.[4,5] Due to l-DOPA tolerance development, an alternative approach for PD treatment is based on the dopaminergic stimulation using agonists that activate the DA receptors.[6] These are members of a family belonging to the class A of the G-protein coupled receptors (GPCRs) and are subdivided into two groups: D1-like subtypes (D1 and D5) couple to Gs and Golf proteins stimulating the production of the second messenger cAMP, and D2-like subtypes (D2, D3, and D4) couple to Gi inhibiting cAMP production.[7,8] The D1 and D2 receptors are the most abundant subtypes in the brain and have thus been the most studied. In particular, the D1 receptor (D1R) has not only well-documented roles in motor activity but also in memory function.[9,10] Hence, activation of the D1R with orthosteric agonists has been a target for drug discovery efforts to develop improved therapies for movement disorders in PD and cognitive decline associated with PD and other neuropsychiatric pathologies such as schizophrenia, Alzheimer’s disease, and other forms of dementia.[11−14] A number of these agonists are endowed with anti-Parkinsonian properties and have been strongly validated at the preclinical stage.[6,11] However, D1R orthosteric agonists suffer from numerous therapeutic drawbacks including: low selectivity over other DA receptors; poor pharmacokinetic profile; inverted U-shaped dose response, probably due to overstimulation at higher dose; rapid onset of tolerance and desensitization caused by a constant activation of the receptor; and tolerability issues, such as hypotension and dyskinesia.[12] Hence, their clinical development has been very challenging and largely unsuccessful so far, and there remains a significant unmet need for safe and effective treatment of PD.

In recent years, the allosteric modulation approach has emerged as a new and alternative strategy to regulate GPCR functions, and the development of allosteric ligands has received widespread attention in drug discovery programs targeting numerous receptors.[15−25] The novel compounds bind at allosteric sites that are spatially distinct from the orthosteric site where the endogenous, natural ligand does. Binding of a ligand to the allosteric site induces conformational changes of the receptor, which results in the modulation of the affinity, potency, and/or signal transduction efficacy of the endogenous/orthosteric ligand response. The allosteric ligand can be classified as a positive allosteric modulator (PAM, enhancing signaling), a negative allosteric modulator (NAM, reducing signaling), or a silent allosteric modulator (SAM, no effect on signaling).[26] GPCR allosteric modulators present unique advantages over orthosteric classical ligands including: a higher pharmacological selectivity, due to the greater structural diversity hypothesized for allosteric sites in contrast to the conserved nature of the orthosteric site among related GPCRs; and a more physiological effect because they exert their indirect action only in the presence of the endogenous ligands, which limits the action of the allosteric ligand producing a saturability of the effect (“ceiling effect”) and protecting against a potential overdose of an orthosteric ligand.[27−30] Hence, allosteric modulation has become a promising approach toward the discovery of safer drugs that offer the maximum benefit while minimizing side effects. However, the development of allosteric modulators for GPCRs has been challenging and has afforded few FDA-approved drugs: the calcium-sensing receptor PAMs cinacalcet and etelcalcetide, the CCR5 receptor NAM maraviroc, the CXCR4 NAM/CXCR7 allosteric agonist plerixafor, the smoothened receptor NAM vismodegib, and the GABAA receptor PAM brexanolone.[31−36] In this context, the positive allosteric modulation approach to upregulate D1R activity has been recently proposed as a novel strategy toward improved dopaminergic therapies for PD. Rather than directly activating the D1R, a PAM will potentiate the action of endogenous DA, thus inducing a more physiological mode of action and avoiding the pitfalls displayed by orthosteric ligands such as overstimulation, tolerance development, and low tolerability.[11,37] However, this therapeutic potential has not been clinically assessed for disclosed D1R PAMs.[11] Among them, LY3154207 and ASP4345 (Figure ) have reached phase 2 development for the treatment of PD and schizophrenia, respectively, which will hopefully validate the clinical use of a D1R PAM.[38−40]

Figure 1

Structures of positive allosteric modulators of the dopamine D1R tested in clinical trials.

In the present work, we contribute to the highly desired search for selectively acting D1R allosteric modulators. The screening of an in-house chemical library using a potentiatior-mode cAMP assay in a cell line stably expressing the human D1R, followed by structural modification of identified hit 4′-methoxy[1,1′-biphenyl]-4-carbaldehyde, has led to new biphenyl derivative 3 characterized in vitro as a D1R PAM (Figure ). A subsequent medicinal chemistry program resulted in compound 26 (UCM-1306, Figure ), a non-covalent PAM that increases the endogenous DA maximal effect both in human and mouse D1 receptors, does not induce receptor desensitization, exhibits no agonist activity and subtype selectivity, and when orally administered it potentiates l-DOPA recovery of decreased locomotor activity in a preclinical model of PD. The compound enhances memory in a novel object recognition test (NORT), supporting its additional use in complicated PD patients with cognitive impairment.

Figure 2

Search of new synthetic modulators of the D1R starting from the initial identified hit.

Results and Discussion

Identification of Compound 3

An in-house chemical library of 250 compounds was screened using an in vitro assay measuring accumulation of cAMP in a human neuroblastoma SK-N-MC cell line stably expressing physiological levels of the D1R but not the D5R.[41] To assess allosteric modulation of the D1R, cells were treated with a fixed concentration of the test compounds (10 μM) for 15 min and then co-incubated with increasing concentrations of the endogenous agonist DA. The cAMP concentration was quantified by homogeneous time-resolved fluorescence energy transfer (HTRF) and the effect of each compound on the DA concentration–response curve was measured. A compound inducing a potentiation >20% in the DA maximal effect (measured as Emax) was considered a potential PAM of the receptor. Using this potentiatior-mode cAMP assay, 4′-methoxy[1,1′-biphenyl]-4-carbaldehyde, exhibiting the highest increase in the DA Emax (55%, Figure ), was identified as an initial hit for the search of new synthetic allosteric modulators of the D1R.

Starting from this hit, related compounds 1–10 (Figure ) were proposed to explore the substitution pattern on the biphenyl scaffold as well as to produce structurally novel compounds by the modification of the methoxy group.[42] For the synthesis of biphenyl analogues 1–10, a strategy based on Suzuki–Miyaura coupling and fluoroalkylation reaction was followed (Scheme ). Thus, coupling between the appropriate arylboronic acids and bromobenzene derivatives was carried out using Pd(PPh3)4 as the catalyst under microwave (MW) irradiation or thermal conditions to afford intermediates 11–13. Next, alkylation of these intermediates or commercial 4′-hydroxy[1,1′biphenyl]-4-carbaldehyde with chlorofluoromethane allowed to obtain monofluoromethoxy derivatives 1–4. Difluoromethoxy analogues 5–7 were readily synthesized by the coupling reaction between (4-formylphenyl)boronic acid and the corresponding commercially available bromodifluoromethoxybenzene, whereas compound 8 was prepared following the Suzuki–Miyaura-fluoroalkylation sequence, using diethyl [bromo(difluoro)methyl]phosphonate as an alkylating reagent. It should be noted that compounds bearing formyl and alkoxy groups in 2- and 2′-position were not considered because both 2′-hydroxy[1,1′-biphenyl]-2-carbaldehyde and 2′-alkoxy[1,1′-biphenyl]-2-carbaldehyde are reported to undergo ring-closure to form the corresponding hemiacetal or acetal, respectively.[43] In the case of alkylsulfanyl derivatives 9 and 10, direct Suzuki–Miyaura reaction using 2-bromobenzenethiol as a starting material did not work. Hence, mono- or difluoroalkylation of 2-bromobenzenethiol was first carried out to obtain intermediates 14 and 15, respectively, which afforded final compounds 9 and 10 by coupling reaction with (4-formylphenyl)boronic acid (Scheme ).

Scheme 1

Synthesis of Compounds 1–10

Reagents and conditions: (a) Pd(PPh3)4, Na2CO3, toluene/EtOH/H2O or THF/H2O, MW, 100/120/130 °C, 20 min or Δ, on, 49–90%; (b) ClCH2F (2.0 M in DMF), Cs2CO3, DMF, −78 °C to rt, on, 59–84%; (c) BrF2CP(O)(OEt)2, KOH, ACN/H2O, −78 °C to rt, on, 40–81%.

Potential PAM activities of synthesized compounds 1–10 were tested in the potentiator-mode cAMP production assay, and the results of D1R activation are summarized in Table . The effect observed for the compounds at a fixed concentration of 10 μM over DA concentration–response curves revealed a potentiation of the DA Emax. Based on the observation that the compounds did not modify DA EC50 but they increased DA Emax, a concentration–response curve in the presence of DA EC70 concentration was built for those compounds that enhanced more than 30% the DA Emax. The PAM efficacy was measured as the percentage of increase over DA EC70 and the potency as the EC50 observed in the concentration–response curve. Compounds 1–3 and 5–7 revealed that the 2′-position is the most favorable for the alkoxy group (3 and 7, with DA potentiation of 82 and 45% at 10 μM, respectively). Hence, the position of the formyl group was explored in 2′-fluoroalkoxy analogues 4 and 8, revealing a marked drop of activity when the carbonyl is situated in the 3-position (DA potentiation of 23 and 12% at 10 μM, respectively). Compounds 3 and 7, bearing a formyl group in the 4-position and a fluoroalkoxy moiety in the 2′-position, were also the best potentiators when tested in the presence of an EC70 concentration of DA (91 and 62%, respectively, Table ). Replacement of oxygen for a sulfur atom in the alkoxy moiety did not improve the allosteric modulation in derivative 10 (44% potentiation at 10 μM, Table ), and a marked drop was observed in analogue 9 (14% potentiation at 10 μM, Table ).

Table 1

Effect of Compounds 1–10 in DA-induced cAMP Production in Human D1R Endogenously Expressed in a Neuroblastoma Cell Line

compd	CHO position	XCHYZ position	X	Y, Z	potentiation of DA Emax (%)a	% maximum increaseb	EC50 (μM)c
1	4	4′	O	H, F	29 ± 6	nd	nd
2	4	3′	O	H, F	25 ± 9	nd	nd
3	4	2′	O	H, F	82 ± 8	91 ± 6	12.7 ± 2.4
4	3	2′	O	H, F	23 ± 5	nd	nd
5	4	4′	O	F, F	35 ± 8	37 ± 5	2.3 ± 0.4
6	4	3′	O	F, F	37 ± 6	35 ± 7	10.8 ± 1.6
7	4	2′	O	F, F	45 ± 5	62 ± 10	18.5 ± 2.7
8	3	2′	O	F, F	12 ± 2	nd	nd
9	4	2′	S	H, F	14 ± 3	nd	nd
10	4	2′	S	F, F	44 ± 5	101 ± 10	21.8 ± 3.6

Effect over the DA concentration–response curve at a fixed concentration of compound = 10 μM.

Efficacy (measured as % of maximum increase over DA EC70) in the concentration–response curves of the compounds over DA EC70.

Potency (measured as EC50) of the compounds at the DA EC70 concentration; values are the mean ± SEM of three independent experiments with duplicate determinations; nd = not determined.

Clearly, among the newly identified modulators of the D1R, compound 3 is the most efficient PAM, potentiating DA Emax in 82% at 10 μM. Further pharmacological evaluation in the potentiator-mode cAMP assay at different fixed concentrations (0.1–10 μM) resulted in a concentration-dependent potentiation of DA effect in human SK-N-MC and mouse CAD neuroblastoma cell lines (Figure S1A,B), and it enhanced EC70 DA effect in 91%, with an EC50 of 12.7 μM (Table and Figure S1C). No cAMP response to compound 3 was observed in the absence of DA, indicating that the modulator behaved mostly as a potentiator with no agonist activity (Figure S1D). In order to ascertain if this allosteric potentiation was due to a modulation on DA affinity, radioligand binding assays in the presence of DA IC50 (4 μM) were carried out. The calculation of the affinity ratio of [3H]SCH-23390[44] showed a concentration-dependent effect of compound 3 by increasing the affinity ratio values, which evidences a positive allosteric behavior of this compound over the DA effect (Figure S2A). When tested in transfected cells expressing human D2, D3, D4, or D5 receptors, compound 3 (10 μM) did not enhance DA Emax in the potentiator-mode cAMP assay (Figure S3). Results in competitive binding assays revealed that 3 displayed marginal displacement (12% at 10 μM) of the radioligand [3H]SCH-23390, whereas full displacement was observed for orthosteric agonist haloperidol (Ki = 5.5 nM, Figure S4). This confirms that modulator 3 does not bind to the high-sequence homology orthosteric site but should be located in an allosteric one, in agreement with the observed subtype selectivity.

Based on the pharmacological characterization of compound 3 as a specific PAM of the human D1R, the next step was an in vitro ADMET profiling (Table S1). Overall, the compound showed good permeability and low hERG inhibition, but its low solubility, high HSA binding, low serum stability, and moderate inhibition in a panel of CYP450 hampered in vivo validation of the therapeutic interest of compound 3 and prompted us to search for a new compound with improved pharmacokinetic properties.

D1R PAM Optimization: from Compound 3 to Compound 26

A medicinal chemistry program, including 2-fluoromethoxy analogues 16–27, was conducted for the optimization of compound 3 (Figure ). The synthesis of target compounds 16–21 was accomplished using a Suzuki–Miyaura coupling followed by monofluoroalkylation (Scheme ). Thus, the adequate bromobenzene derivative and the corresponding arylboronic acid were coupled according to the previously described conditions, to obtain intermediates 28–32, which afforded final compounds 16–20 by a reaction with chlorofluoromethane. In the case of derivative 20, intermediate 32 was oxidized to sulfone 33 before the fluoroalkylation step. Carboxylic acid 21 was readily accessible through basic hydrolysis of methyl ester 17. Regarding amide 22, the coupling reaction between (4-carbamoylphenyl)boronic acid and 2-bromophenol failed to give the expected biphenyl derivative. In this case, 2-bromophenol was transformed into fluorinated derivative 34, which did provide final compound 22 by Suzuki–Miyaura coupling. Sulfoxide analogues 23 and 24 were prepared from 1-bromo-4-(methylsulfanyl)benzene and 1-bromo-4-(ethylsulfanyl)benzene (Scheme ), respectively. Thus, oxidation to arylsulfoxides 35 and 36 and subsequent coupling with (2-hydroxyphenyl)boronic acid afforded intermediates 37 and 38, leading to the final target compounds by standard fluoroalkylation. NH-sulfoximines 25 and 26 were prepared through two different strategies (Scheme ). Ethylsulfoximine derivative 25 was synthesized via classical methodology, starting with an imination reaction with hydrazoic acid of previously synthesized sulfoxide 36 to obtain the corresponding sulfoximine 39. Next, coupling with (2-hydroxyphenyl)boronic acid under MW standard conditions and reaction with chlorofluoromethane yielded desired compound 25. For the synthesis of methylsulfoximine 26, a safer procedure was used for the initial imination step. Thus, the reaction of commercial 1-bromo-4-(methylsulfanyl)benzene with cyanamide and N-bromosuccinimide afforded an intermediate N-cyanosulfylimine, which was directly transformed into NH-sulfoximine 41 through oxidation and hydrolysis. Then, Suzuki–Miyaura followed by fluoroalkylation yielded final compound 26. N-Methylsulfoximine 27 was obtained using the same Suzuki–Miyaura-fluoroalkylation sequence but starting from N-methylated intermediate 43, which was synthesized by reductive alkylation of 41 employing formaldehyde and formic acid.

Scheme 2

Synthesis of Compounds 16–24

Reagents and conditions: (a) Pd(PPh3)4, Na2CO3, toluene/EtOH/H2O, MW, 120/170 °C, 10–20 min or Δ, on, 16–99%; (b) H2O2 (30%), (NH4)6Mo7O24·4H2O, methanol, 0 °C to rt, 1 h, 75%; (c) ClCH2F (2.0 M in DMF), Cs2CO3, DMF, −78 °C to rt, on, 31–97%; (d) NaOH, THF/H2O, rt, 12 h, quantitative; (e) mCPBA, DCM, 0 °C to rt, 4 h, 83–92%.

Scheme 3

Synthesis of Compounds 25–27

Reagents and conditions: (a) NaN3, H2SO4 (concd), CHCl3, 45 °C, on, 92%; (b) (2-hydroxyphenyl)boronic acid, Pd(PPh3)4, Na2CO3, toluene/EtOH/H2O, MW, 100/120 °C, 20 min, 45–90%; (c) ClCH2F (2.0 M in DMF), Cs2CO3, DMF, −78 °C to rt, on, 78–95%; (d) (i) cyanamide, NBS, tBuOK, methanol, rt, 1.5 h, quantitative; (ii) RuCl3, NaIO4 (aq, 0.15 M), ACN/DCM, rt, 3 h, 83%; (iii) H2SO4 (aq, 50%), Δ, 2 h, 77%; (e) HCO2H, HCHO (aq, 37%), 100 °C, 2 d, quantitative.

The assessment of the new synthesized compounds 16–27 as D1R PAMs was carried out in the potentiator-mode cAMP assay (Table ). Replacement of the aldehyde for other carbonyl-containing moieties (ketone, ester, carboxylic acid, and amide in compounds 16, 17, 21, and 22, respectively), sulfonamide (18 and 19), or sulfonyl (20) was detrimental for the activity. On the other hand, a sulfinyl or sulfoximine group seemed tolerable. In the case of sulfinyl derivatives, a marked drop in activity was observed from methyl to ethyl analogues (23, 60% vs 24, 16% at 10 μM), while potentiation was maintained for sulfoximine derivatives (25, 54% vs 26, 55% at 10 μM). Methylation of the nitrogen in the sulfoximine group produced a decrease in activity (27, 30% at 10 μM). In this series, sulfoximine analogue 26 was 2 orders of magnitude more potent than parent compound 3 (EC50 = 60 nM vs 12.7 μM, Table ) and exhibited high potentiation of DA Emax (55% at 10 μM). Hence, we next considered the synthesis of new sulfoximine derivatives 45–47, which were prepared using an NH-bromoarylsulfoximine as the coupling partner in a Suzuki–Miyaura reaction (Scheme ). For the synthesis of pyridine derivative 45, a one-pot borylation-Suzuki–Miyaura coupling strategy was carried out. Thus, NH-sulfoximine 41 was converted to the corresponding pinacol boronate derivative and coupled in situ to intermediate 48, obtained by fluoroalkylation of 2-bromopyridin-3-ol. Compounds 46 and 47 were synthesized by coupling of bromochloroarylsulfoximines 52 and 53 with (2-hydroxyphenyl)boronic acid followed by fluoroalkylation of the resulting biphenyl derivatives 54 and 55. Intermediates 52 and 53 were prepared by sequential imination, oxidation, and hydrolysis reactions of the corresponding bromochloroarylsulfides 49 and 50, according to the previously setup procedure. Sulfides 49 and 50, in turn, were synthesized by methylation of 4-bromo-2-chlorobenzenethiol and 4-bromo-3-chlorobenzenethiol (51), which was obtained from 4-bromo-3-chloroaniline via diazonium salt.

Table 2

Effect of Compounds 16–27 and 45–47 in DA-induced cAMP in Human D1R Endogenously Expressed in a Neuroblastoma Cell Line

compd	R	X	Y	potentiation of DA Emax (%)a	% maximum increaseb	EC50 (μM)c
3	CHO	H	CH	82 ± 8	91 ± 6	12.7 ± 2.4
16	COMe	H	CH	11 ± 3	nd	Nd
17	COOMe	H	CH	28 ± 3	nd	Nd
18	SO2NH2	H	CH	31 ± 2	42 ± 7	4.93 ± 1.2
19	SO2NHMe	H	CH	38 ± 5	28 ± 8	1.26 ± 0.7
20	SO2Me	H	CH	30 ± 7	30 ± 8	13.11 ± 2.8
21	COOH	H	CH	7 ± 3	nd	Nd
22	CONH2	H	CH	16 ± 4	nd	Nd
23	SOMe	H	CH	60 ± 9	73 ± 6	21.7 ± 5.8
24	SOEt	H	CH	16 ± 5	nd	Nd
25	SO(NH)Et	H	CH	54 ± 7	50 ± 7	30.2 ± 2.4
26	SO(NH)Me	H	CH	55 ± 7	45 ± 7	0.06 ± 0.01
27	SO(NMe)Me	H	CH	30 ± 5	33 ± 3	15.9 ± 2.7
45	SO(NH)Me	H	N	1 ± 2	nd	nd
46	SO(NH)Me	3′-Cl	CH	12 ± 4	nd	nd
47	SO(NH)Me	2-Cl	CH	20 ± 3	nd	nd

Effect over the DA concentration–response curve at a fixed concentration of compound = 10 μM.

Efficacy (measured as % of maximum increase over DA EC70) in the concentration–response curves of the compounds over DA EC70.

Potency (measured as EC50) of the compounds at DA EC70 concentration; values are the mean ± SEM of three independent experiments with duplicate determinations; nd = not determined.

Scheme 4

Synthesis of Compounds 45–47

Reagents and conditions: (a) ClCH2F (2.0 M in DMF), Cs2CO3, DMF, −78 °C to rt, on, 73–96%; (b) (i) B2pin2, Pd2(dba)3·CHCl3, SPhos, KOAc, 1,4-dioxane, 110 °C, on; (ii) Pd2(dba)3·CHCl3, K3PO4 (aq, 5 M), 110 °C, on, 24%; (c) (i) NaNO2, HCl (conc), potassium O-ethyl carbonodithioate, −5 to 75 °C, 1.5 h; (ii) KOH, EtOH, Δ, on, 80% (2 steps); (d) MeI, K2CO3, acetone, rt, 5 h, 56%-quantitative; (e) (i) cyanamide, NBS, KOtBu, methanol, rt, 1.5 h, quantitative; (ii) RuCl3, NaIO4 (aq, 0.15 M), ACN/DCM, rt, 3 h, 84–89%; (iii) H2SO4 (aq, 50%), Δ, 2 h, 46–51%; (f) (2-hydroxyphenyl)boronic acid, Pd(PPh3)4, Na2CO3, toluene/EtOH/H2O, MW, 100/120 °C, 20 min, 77–89%.

Structural modifications introduced in the new synthesized sulfoximine derivatives 45–47 produced an important depletion of D1R allosteric activity, as shown in the data from the potentiator-mode cAMP assay (Table ). Therefore, compound 26 exhibiting the best allosteric activity (high efficacy and the greatest potency) was selected for further pharmacological characterization and study of ADMET properties.

In Vitro Pharmacological and Pharmacokinetic Characterization of D1R PAM 26

Compound 26 was tested in the potentiator-mode cAMP assay at different fixed concentrations (1, 5, and 10 μM). Figure A shows a potentiation of the DA Emax in a concentration-dependent manner, with a maximum value of 55% at 10 μM, and it was demonstrated to be a reversible effect (Figure B). A PAM behavior was also observed in mouse D1R (Figure C). When tested with EC70 DA, 26 increased cAMP in a concentration-response manner with high potency (EC50 = 60 nM, Figure D). Importantly, no cAMP response to compound 26 was determined in the absence of DA, so the new D1R PAM does not display agonist activity (Figure E). Therefore, it should have negligible direct activating effects on the human D1R in the absence of the endogenous ligand. It was also confirmed that the compound does not induce receptor desensitization and it does not induce further increase in DA-induced desensitization (Figure F). Hence, efficacy should be retained over time upon repeated dosing of the D1R PAM, avoiding the development of tolerance. Compound 26 showed a concentration-dependent increase of the affinity ratio values in the presence of 4 μM DA, evidencing a positive allosteric behavior of this compound over DA effect (Figure S2B). As expected for a PAM, subtype selectivity over human D2, D3, D4, and D5 receptors was achieved with compound 26 (Figure S5), whereas selective activation of the D1R with agonists has historically been very challenging due to the high sequence homology at the orthosteric site.

Figure 3

(A) Concentration–response curves in human D1R for cAMP production of DA alone (black) and in the presence of different concentrations of compound 26. (B) Concentration–response curves in human D1R of DA (black), 10 μM 26 for 10 min (red), and 10 μM 26 for 10 min then washing with assay buffer (blue). (C) Concentration–response curves in mouse D1R of DA alone (black) and in the presence of 10 μM 26 (red). (D) Concentration–response curve in human D1R of compound 26 in the presence of DA EC70 concentration. (E) Concentration–response curves in human D1R of DA (black) and compound 26 (red). (F) Concentration–response curves in human D1R of DA (red), after preincubation for 1 h with 10 μM DA (black), with 10 μM 26 (green), and with both 10 μM DA and 10 μM 26 (blue).

In view of the allosteric efficacy of compound 26, in vitro ADMET properties were determined, and they are summarized in Table . In nephelometry assay, a solubility more than 3 times higher than parent compound 3 (50 vs 15 μM, Table S1) was observed. Interaction with HSA proteins revealed 79% binding with moderate dissociation constant (Kd = 1.56 × 10–4 M), indicating a significant reduction compared to 3 (>99% binding, Table S1). PAMPA showed a decent permeability value (P) of 24 × 10–6 cm/s. Compound 26 displayed a low hERG inhibition (21% at 10 μM). In mouse serum, 93% of the compound remained after 4 h, whereas 50% had been observed for remaining compound 3 (Table S1). To assess first-pass metabolism, compound 26 was incubated in liver homogenates and 65% of remaining compound was detected after 24 h. Hence, the compound displays a good stability profile overall. In a panel of CYP450, low inhibition was observed for CYP2D6 and CYP34A (3 and 9% of inhibition, respectively), while moderate blockade was obtained for CYP1A2, CYP2C9, and CYP2C19 with the percentage of inhibition values in the range of 37–49% (Table ).

Table 3

ADMET Profile for Compound 26

solubility (μM)a	50
HSA binding (%)b	79 (Kd = 1.56 × 10–4 M)
P (cm/s)c	24 × 10–6
hERG inhibition (%)d	21
mouse serum stability (%)e	93
liver homogenate stability (%)e	65
CYP2D6 (%)f	3 ± 1
CYP34A (%)f	9 ± 1
CYP1A2 (%)f	49 ± 2
CYP2C9 (%)f	38 ± 3
CYP2C19 (%)f	37 ± 2

Maximum solubility measured by nephelometry.

Binding to human serum albumin (HSA) determined at a concentration of 5 μM.

Permeability in the parallel artificial membrane permeability assay (PAMPA).

Blockade of the K+ channel current at a concentration of 10 μM.

Remaining compound quantified after 4 and 24 h, respectively.

Percentage of inhibition of CYP450 activity measured at a concentration of 10 μM by fluorescence assays.

Altogether, new compound 26 behaves as a potent and efficient D1R PAM and exhibits an improved ADMET profile than parent compound 3. As the sulfoximine group provides chirality to the molecule, it was mandatory to evaluate both enantiomers of compound 26. Therefore, each enantiomer was prepared separately and tested for allosteric activity at the D1R in the potentiator-mode cAMP assay. For the enantioselective synthesis of the sulfoximines, a route in which chirality is introduced through asymmetric oxidation of an intermediate sulfide was used (Scheme ). Thus, previously synthesized sulfide 32 was transformed into monofluorinated analogue 56 by reaction with chlorofluoromethane. Next, asymmetric oxidation of 56 catalyzed by vanadyl acetylacetonate in the presence of the appropriate chiral tert-leucinol derivative,[45,46] yielded the corresponding sulfoxides (R)- and (S)-23 with enantiomeric ratio (er) values higher than 97:3. The reaction of these intermediates with ammonium carbamate and (diacetoxyiodo)benzene provided the highly enantioenriched sulfoximines (R)- and (S)-26 (er > 97:3) under very mild conditions[47] (Scheme ).

Scheme 5

Synthesis of Enantiomers of Compound 26

Reagents and conditions: (a) ClCH2F (2.0 M in DMF), Cs2CO3, DMF, −78 °C to rt, on, 84%; (b) H2O2 (30%), (S)- or (R)-2-((E)-{[1-(hydroxymethyl)-2,2-dimethylpropyl]imino}methyl)-4,6-diiodophenol, [VO(acac)2], CHCl3, 0 °C, 20 h, 59–64%, er > 97:3; (c) ammonium carbamate, PhI(OAc)2, MeOH, rt, 30–60 min (open flask), 61–67%, er > 97:3.

Both enantiomers (R)- and (S)-26 exhibited similar pharmacological activity—D1R PAM efficacy (42 and 43% potentiation at 10 μM) and no agonist activity (Figure S6)—to that of racemic compound 26, and the latter was selected as a drug candidate for in vivo validation of therapeutic interest.

Binding Site for the New D1R PAM 26

In silico docking calculations were performed to predict the binding mode between compound 26 (S- and R-stereoisomers) and the active form of human D1R. The results in the receptor homology model (Figure , top panel) show that both enantiomers bind in a pocket formed by TM3, TM4, and intracellular loop 2 (ICL2), a region previously identified as an allosteric binding site of the D1R.[38,48] Both (R)- and (S)-26 bind with an almost equal pose establishing π–π stacking interactions with Trp123 as well as polar interactions with Val116 (H-bond) and Lys134 (halogen bond). Additionally, they form hydrophobic interactions with the following residues: Val119, Trp123, Tyr131, Lys134, Met135, and Leu143. The difference in the binding of the two enantiomers is that (S)-26 forms an additional polar interaction (H-bond) with Asp120 (Figure , top left panel) that (R)-26 cannot establish. In order to support the binding poses, short molecular dynamics (MD) simulations were performed, which reveal that the docking poses are quite stable and compounds (R)- and (S)-26 remain bound to ICL2. Therefore, the residues interacting with the new compound are similar to those previously proposed for PAMs LY3154207 and DETQ docked in the D1R homology model.[38,48]

Figure 4

Homology model (top panel, yellow cartoon representation) and electron microscopy model (bottom panel, cyan cartoon representation) of the ICL2 region of human D1R bound to (S)-26 (left panel, orange stick representation) and (R)-26 (right panel, pink stick representation) as predicted by docking calculations. The key interactions between the ligand and residues are highlighted: blue lines indicate favorable H-bond interactions, grey dashed lines hydrophobic interactions, turquoise lines halogen bonds, and green dashed lines π–π stacking interactions.

Very recently, Teng et al. have published an electron microscopy (EM) structure of the D1R bound to LY3154207 (PDB code: 7X2F).[49] In this structure, mostly similar to the homology model used, the most relevant difference is that the conformation of Trp123 is flipped. This causes a shift in the rotation of the LY3154207 pose at the binding site, compared to that previously proposed in the homology model. To understand how this would affect the binding mode of compound 26, we performed a new docking study using the EM structure. The results in Figure (bottom panel) show that both enantiomers bind in the same region as in the homology model, but adopting a rotated pose forced by the new conformation of Trp123, as in the case of LY3154207. (R)- and (S)-26 can establish π–π stacking interactions with Trp123 and polar interactions with Ala139 (H-bond), Arg130 (halogen bond), and Lys134 (halogen bond). They can also form hydrophobic interactions with the following residues: Trp123, Arg130, Tyr131, Lys134, Met135, and Leu143. Both enantiomers bind without major differences, but (R)-26 can establish an additional H-bond with Leu143 and an extra hydrophobic interaction with Ala139 (Figure , bottom right panel).

Overall, docking studies in Figure —using the homology model or the EM structure of the receptor—reveal that the new PAM 26 interacts with the human D1R in a similar way to that reported for other known PAMs of the receptor.[38,48,49]

In Vivo Efficacy of the New D1R PAM 26 in Locomotor Activity Models

Prior to the evaluation of the new D1R PAM 26 in animal models, the in vivo cerebrospinal fluid (CSF) and brain permeability were studied. Following a single oral dose administration of 26 at 5 mg/kg to male BALB/c mice, plasma concentrations were quantifiable up to 8 h with Tmax at 0.5 h (Figure ). Concentrations in the CSF and brain were quantifiable up to 4 h. Brain-to-plasma concentration ratios were in the range of 0.55–1.15 between 0.5 and 4 h and CSF-to-plasma concentration ratio in the range of 0.13–0.21 between 0.5 and 4 h. In addition, CSF-to-plasma free concentration ratio of approximately 1 indicated that the compound does not show efflux transport from the brain.

Figure 5

Compound levels in the plasma, brain, and CSF of mouse administered with 5 mg/kg (po) of compound 26. Horizontal dotted line indicates the EC50 value of the compound.

In view of the good brain penetration and oral availability, compound efficacy in vivo was tested in two different locomotor activity assays: a cocaine-induced hyperactivity model in normal animals and a PD animal model based on the modulation of l-DOPA effect in reserpinized mice. In the first experiment, the administration of compound 26 (1 mg/kg, ip) potentiated the hyperlocomotion induced by cocaine (20 mg/kg, Figure ). The cocaine-induced increase in extracellular DA concentration results in enhanced locomotion through activation of the D1R, which is the dopaminergic receptor responsible for this pharmacological action.[50] Administration of compound 26 increased this response, suggesting an in vivo potentiation of DA action at the D1R.

Figure 6

Effect of compound 26 in a cocaine-induced hyperactivity model. Pretreatment with 26 (1 mg/kg, ip) enhanced cocaine (20 mg/kg, sc)-induced hyperlocomotion in adult mice, with no intrinsic locomotor activity when 26 was administered alone [F(3.16) = 61.67, P < 0.0001, **P < 0.001, *P < 0.01 cocaine vs cocaine + 26 groups, N = 8 animals/group].

Regarding the PD model, reserpine is an alkaloid known to induce hypomobility/akinesia and rigidity, and reserpinized rodents are widely used as a preclinical model for PD. l-DOPA and all clinically active anti-Parkinsonian treatments show efficacy in the reserpine model, exhibiting a reversal of reserpine-induced akinesia. After pretreatment with a low dose of reserpine (2.5 mg/kg, sc), a partial depletion of DA was achieved in mice, as previously described, mimicking the low dopaminergic functionality in PD patients. As shown in Figure , reserpinized mice were treated with l-DOPA (100 or 400 mg/kg, ip) and/or PAM 26 (30 mg/kg, po). There was a small recovery of decreased locomotor activity with l-DOPA alone (as described),[3] a small recovery of decreased locomotor activity with 26 alone (as reported for other D1R PAMs),[37,38] and a moderate synergistic effect with l-DOPA plus compound 26, as expected for a D1R PAM. These results evidenced efficacy of the compound in ambulatory activity of mice with induced low DA levels, similarly to other D1R PAMs. For instance, DETQ and LY3154207 have displayed a synergistic effect on locomotion in human D1R transgenic mice; the larger effect observed for these PAMs was probably due to the higher level of receptor expression than that in wild-type reserpinized mice used in our model.[51] The study suggests that the new D1R PAM could be therapeutically beneficial to alleviate motor symptoms in PD patients as monotherapy by enhancing the effect of low levels of endogenous DA or the activity of co-administered l-DOPA. In the study, high levels of 26 in both the brain and plasma were confirmed (Table S2) and no sign of adverse effects was observed in the animals after oral dose administration of the compound.

Figure 7

In vivo effect of compound 26 in locomotor activity by measuring ambulatory activity and total distance travelled. Data are shown as the mean ± S.E.M (N = 9–10).

Effects of Compound 26 on Memory

Because PD has been linked to memory deficits and even dementia (alpha-synuclein aggregates that are a major feature of PD are also very prominent in Lewy Body Dementia),[52,53] we assessed whether compound 26 might have a beneficial effect on memory, by means of the NORT. Data in Figure suggest that the treatment with the compound increases memory trace, even 4 days after its administration and the presentation of the familiar object. This effect suggests that the compound helps consolidate long-term memory formation, probably through the described D1R-dependent modification of hippocampal interneurons functionality.[54] Therefore, the new D1R PAM is a good candidate not only for improving motor symptoms but also for addressing the key comorbid cognitive impairment associated with long-term PD.

Figure 8

Effect of compound 26 (1 mg/kg, ip) in the novel object recognition test. Enhanced memory, measured as the time spent exploring the novel object [upper panel, F(1.28) = 9.4, p < 0.005] and the preference over a familiar object [lower panel, F(1.28) = 7.1, p < 0.02] is observed at 24 and 96 h (2-way ANOVA). N = 8 animals/group.

Conclusions

In this work, we have developed compound 26 [2-(fluoromethoxy)-4′-(S-methanesulfonimidoyl)-1,1′-biphenyl] as an allosteric modulator of the dopaminergic D1R that increases the endogenous DA maximal effect in a dose-dependent manner in human and mouse receptors, is inactive in the absence of DA, modulates DA affinity for the receptor, exhibits subtype selectivity, shows a reversible effect, and does not show any desensitization effect avoiding tolerance. Competition assays and docking studies support the binding to an allosteric site of the receptor. The new PAM was found to be orally active in a reserpine-induced low-DA-level model, enhancing l-DOPA recovery of decreased locomotor activity with no sign of adverse effects, which suggests an allosteric modulation of DA effect in vivo. The compound has also memory enhancing effect that might improve cognitive impairment associated to PD. These results support the interest of a D1R PAM as a promising therapeutic approach for PD.

Experimental Section

Synthesis

Unless stated otherwise, starting materials, reagents, and solvents were purchased as high-grade commercial products from ABCR, Acros, Fluorochem, Scharlab, or Sigma-Aldrich, and were used without further purification. All non-aqueous reactions were performed under an argon atmosphere in oven-dried glassware. Tetrahydrofuran (THF) and dichloromethane (DCM) were dried using a Pure Solv Micro 100 L solvent purification system. Acetone was dried under K2CO3. Triethylamine was dried over KOH and distilled before use. Reactions under MW irradiation were performed in a Biotage Initiator 2.5 reactor. Reactions were monitored by analytical thin-layer chromatography (TLC) on silica gel plates supplied by Merck (Kieselgel 60 F-254) with detection by UV light (254 nm), 5% ninhydrin solution in ethanol, or 10% phosphomolybdic acid solution in ethanol. Products were purified by flash chromatography using a Varian 971-FP purification system using silica gel cartridges (Varian, particle size 50 μm). All compounds were obtained as oils, except for those whose melting points (mp) are indicated, which were solids. Mp (uncorrected) was determined on a Stuart Scientific electrothermal apparatus. Infrared (IR) spectra were measured on a Bruker Tensor 27 instrument equipped with a Specac ATR accessory of 5200–650 cm–1 transmission range; frequencies (ν) are expressed in cm–1. Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker AVANCE III 700 MHz (1H, 700 MHz; 13C, 175 MHz), Bruker AVANCE 500 MHz (1H, 500 MHz; 13C, 125 MHz), or Bruker DPX 300 MHz (1H, 300 MHz; 13C, 75 MHz; 19F, 300 MHz) instruments at the Universidad Complutense de Madrid (UCM) NMR core facilities. Bruker DPX 300 MHz equipment was used unless otherwise stated. Chemical shifts (δ) are expressed in parts per million relative to the residual solvent peak for 1H and 13C nucleus (CDCl3: δH = 7.26, δC = 77.16; DMSO-d6: δH = 2.50, δC = 39.52; acetone-d6: δH = 2.05, δC = 29.84; methanol-d4: δH = 3.31, δC = 49.00) and to internal (trifluoromethyl)benzene for the 19F nucleus; coupling constants (J) are in hertz (Hz). The following abbreviations are used to describe peak patterns when appropriate: s (singlet), d (doublet), t (triplet), q (quartet), qt (quintet), sext (sextet), m (multiplet), br (broad), and app (apparent). 2D NMR experiments (H,H-COSY, HMQC, and HMBC) of representative compounds were carried out to assign protons and carbons of the new structures. Numbered chemical structures for NMR assignation of final compounds 3, 8, 10, (R)- and (S)-23, (R)- and (S)-26 are shown in Figure S7.

For all final compounds, purity was determined by high-performance liquid chromatography (HPLC) coupled to mass spectrometry (MS) using an Agilent 1200LC-MSD VL instrument, and satisfactory chromatograms confirmed a purity of at least 95% for all tested compounds. LC separation was achieved with a Zorbax Eclipse XDB-C18 column (5 μm, 4.6 mm × 150 mm) or a Zorbax SB-C3 column (5 μm, 2.1 mm × 50 mm), both together with a guard column (5 μm, 4.6 mm × 12.5 mm). The gradient mobile phases consisted of A (95:5 water/acetonitrile) and B (5:95 water/acetonitrile) with 0.1% ammonium hydroxide and 0.1% formic acid as the solvent modifiers, and the gradients are indicated in Table S3. MS analysis was performed with an ESI source. The capillary voltage was set to 3.0 kV and the fragmentor voltage was set at 72 eV. The drying gas temperature was 350 °C, the drying gas flow was 10 L/min, and the nebulizer pressure was 20 psi. Spectra were acquired in positive or negative ionization modes from 100 to 1200 m/z and in the UV-mode at four different wavelengths (210, 230, 254, and 280 nm).

Optical rotation [α] was measured on an Anton Paar MCP 100 modular circular polarimeter using a sodium lamp (λ = 589 nm) with a 1 dm path length; concentrations (c) are given as g/100 mL. Enantiomeric ratios (er) were determined by HPLC using a Daicel Chiralpak IC column (5 μm, 4.6 mm × 150 mm), together with a guard column (5 μm, 4.6 mm × 12.5 mm). The gradient mobile phase consisted of a 1:9 mixture of water/methanol, using a flow of 0.9 mL/min for the injection of the sample and 0.5 mL/min during the run. HPLC traces were compared to racemic samples of sulfoxide 23 and sulfoximine 26, which were obtained in the absence of any chiral catalysts.

The following compounds were synthesized as previously described and their spectroscopic data correspond with those reported: 1-bromo-4-(methanesulfinyl)benzene (35),[55] 4-bromo-3-chlorobenzenethiol (51),[56] and (R)- and (S)-2-((E)-{[1-(hydroxymethyl)-2,2-dimethylpropyl]imino}methyl)-4,6-diiodophenol.[45]

General Procedure for Suzuki–Miyaura Reaction

A solution of the corresponding bromoaryl derivative (1.00 equiv), arylboronic acid (1.05–1.20 equiv), and Na2CO3 (2 or 4 equiv) in a 2:1.5:1 mixture of toluene/water/ethanol or a 1:1 mixture of THF/water (6.5 mL/mmol) was degassed by bubbling argon for 10 min. Pd(PPh3)4 (0.06 equiv) was then added and the reaction was refluxed overnight or heated under MW irradiation. After this time, the mixture was extracted with EtOAc (×3) and the organic phase was dried over Na2SO4, filtered, and evaporated under reduced pressure. The residue was purified by recrystallization, column chromatography, or preparative TLC to afford intermediates 11–13, 28–32, 37, 38, 40, 42, 44, 54, and 55 or final compounds 5–7, 9, 10, and 22.

2′-[(Difluoromethyl)sulfanyl][1,1′-biphenyl]-4-carbaldehyde (10)

Following the general procedure for the Suzuki–Miyaura reaction, compound 10 was obtained from 15 (104 mg, 0.433 mmol), (4-formylphenyl)boronic acid (70 mg, 0.455 mmol), and Na2CO3 (92 mg, 0.866 mmol) in a toluene/water/ethanol mixture by heating at 130 °C under MW irradiation for 20 min, as a solid (92 mg, 81%). Chromatography: hexane.

mp 46–47 °C. Rf: 0.46 (hexane/EtOAc 9:1). IR (ATR): ν 1702 (CHO), 1065, 1034 (C–F). 1H NMR (CDCl3): δ 6.66 (t, J = 56.6, 1H, CHF2), 7.39–7.48 (m, 3H, H3′, H4′, H5′), 7.52 (d, J = 8.1, 2H, H2, H6), 7.75 (dd, J = 7.5, 1.5, 1H, H6′), 7.95 (d, J = 8.2, 2H, H3, H5), 10.09 (s, 1H, CHO). 13C NMR (CDCl3): δ 120.7 (t, J = 274.1, CHF2), 124.8 (t, J = 2.6, C1′), 129.1 (C3′), 129.5 (C3, C5), 130.0 (C5′), 130.6 (C2, C6), 130.9 (C4′), 135.5 (C4), 136.4 (C6′), 145.7 (C2′), 146.6 (C1), 192.0 (CHO). 19F NMR (CDCl3): δ −91.6. HPLC (Gradient-I, column C3, tR, min): 12.27. MS (ESI, m/z, %): 265.0 ([M + H]+, 100).

General Procedure for the Monofluoroalkylation Reaction

To a mixture of the corresponding hydroxy- or sulfanylaryl derivative (1.00 equiv) and Cs2CO3 (1.60 equiv) in anhydrous DMF (7.0 mL/mmol) at −78 °C, a solution of chlorofluoromethane (2.0 M in DMF, 4.00 equiv) was added dropwise. The reaction was stirred and allowed to warm to rt overnight. Afterward, the mixture was diluted with water and extracted with Et2O (×3). The combined organic layers were dried over Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by column chromatography or preparative TLC to afford intermediates 14, 34, 48, and 56 or final compounds 1–4, 16–20, 23–27, 46, and 47.

2′-(Fluoromethoxy)[1,1′-biphenyl]-4-carbaldehyde (3)

Following the general procedure for the monofluoroalkylation reaction, compound 3 was obtained from 12 (1.01 g, 5.10 mmol) as oil (0.94 g, 84%). Chromatography: hexane.

Rf: 0.39 (hexane/DCM 7:3). IR (ATR): ν 1699 (CHO), 1215 (C–O–C), 1129, 1082 (C–F). 1H NMR (CDCl3): δ 5.68 (d, J = 54.3, 2H, CH2F), 7.19–7.28 (m, 2H, H3′, H5′), 7.37–7.44 (m, 2H, H4′, H6′), 7.68 (d, J = 8.2, 2H, H2, H6), 7.94 (d, J = 8.4, 2H, H3, H5), 10.06 (s, 1H, CHO). 13C NMR (CDCl3): δ 100.9 (d, J = 218.6, CH2F), 116.0 (d, J = 1.4, C3′), 124.1 (C5′), 129.6 (C3, C5), 130.0 (C4′), 130.4 (C2, C6), 130.8 (C1′), 131.1 (C6′), 135.2 (C4), 144.3 (C1), 153.7 (d, J = 3.1, C2′), 192.1 (CHO). 19F NMR (CDCl3): δ −150.9. HPLC (Gradient-I, column C3, tR, min): 11.42. MS (ESI, m/z, %): 231.1 ([M + H]+, 100).

2-(Fluoromethoxy)-4′-(S-methanesulfonimidoyl)-1,1′-biphenyl (26)

Following the general procedure for the monofluoroalkylation reaction, compound 26 was obtained from 42 (26 mg, 0.105 mmol) as a solid (23 mg, 78%). Chromatography: preparative TLC in toluene/methanol 95:5.

mp 144–145 °C. Rf: 0.40 (toluene/methanol 9:1). IR (ATR): ν 3341 (NH), 1220 (C–O–C), 1129, 1083 (SO), 994, 972 (C–F). 1H NMR (acetone-d6, 500 MHz): δ 3.10 (s, 3H, CH3), 5.84 (d, J = 54.4, 2H, CH2F), 7.26 (td, J = 7.5, 1.0, 1H, H5), 7.35 (d, J = 8.3, 1H, H3), 7.44–7.49 (m, 2H, H4, H6), 7.73 (d, J = 8.5, 2H, H2′, H6′), 8.04 (d, J = 8.5, 2H, H3′, H5′). 13C NMR (acetone-d6): δ 46.6 (CH3), 101.8 (d, J = 216.8, CH2F), 116.4 (d, J = 1.1, C3), 124.7 (C5), 128.3 (C3′, C5′), 130.8 (C4), 130.9 (C2′, C6′), 131.2 (d, J = 1.1, C1), 131.9 (C6), 143.1 (C1′), 144.2 (C4′), 154.5 (d, J = 3.0, C2). 19F NMR (acetone-d6): δ −151.0. HPLC (gradient-I, column C18, tR, min): 13.17. MS (ESI, m/z, %): 279.8 ([M + H]+, 100).

General Procedure for the Difluoroalkylation Reaction

Diethyl [bromo(difluoro)methyl]phosphonate (2.00 equiv) was added in one portion to a −78 °C cooled solution of the adequate hydroxy- or sulfanylaryl derivative (1.00 equiv) and KOH (20 equiv) in a 1:1 mixture of ACN/water (10 mL/mmol), and the reaction was stirred and allowed to warm to rt overnight. Next, the mixture was diluted with water and the aqueous layer was extracted with Et2O (×3). The combined organic layers were dried over Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by column chromatography or preparative TLC to afford intermediate 15 or final compound 8.

2′-(Difluoromethoxy)[1,1′-biphenyl]-3-carbaldehyde (8)

Following the general procedure for the difluoroalkylation reaction, compound 8 was obtained from 13 (19 mg, 0.096 mmol) as a solid (10 mg, 40%). Chromatography: preparative TLC in 98:2 toluene/methanol.

Rf: 0.34 (toluene/methanol 98:2). IR (ATR): ν 1703 (CHO), 1217 (C–O–C), 1138, 1052 (C–F). 1H NMR (CDCl3): δ 6.39 (t, J = 73.7, 1H, CHF2), 7.29–7.35 (m, 1H, H5′), 7.35 (dd, J = 7.4, 1.4, 1H, H3′), 7.39–7.47 (m, 2H, H4′, H6′), 7.63 (t, J = 7.6, 1H, H5), 7.79 (dt, J = 7.8, 1.5, 1H, H6), 7.91 (dt, J = 7.6, 1.4, 1H, H4), 8.00 (m, 1H, H2), 10.08 (s, 1H, CHO). 13C NMR (CDCl3, 125 MHz): δ 116.1 (t, J = 258.9, CHF2), 120.2 (C3′), 126.1 (C5′), 128.9 (C4), 129.1 (C6′), 129.6 (C5), 130.9 (C2), 131.5 (C4′), 132.8 (C1′), 135.6 (C6), 136.7 (C3), 138.2 (C1), 148.2 (t, J = 2.5, C2′), 192.3 (CHO). 19F NMR (CDCl3): δ −80.9. HPLC (Gradient-I, column C3, tR, min): 11.53. MS (ESI, m/z, %): 271.1 ([M + Na]+, 100).

Synthesis of (R)- and (S)-26

(R)- and (S)-2-(Fluoromethoxy)-4′-(methanesulfinyl)-1,1′-biphenyl [(R)- and (S)-23]

A solution of VO(acac)2 (0.010 equiv) in anhydrous CHCl3 (0.25 mL/mmol sulfide) was added to a solution of the appropriate R or S form of 2-((E)-{[1-(hydroxymethyl)-2,2-dimethylpropyl]imino}methyl)-4,6-diiodophenol (0.015 equiv) in anhydrous CHCl3 (0.25 mL/mmol sulfide) and the mixture was stirred at rt for 2 h. Next, a solution of 56 (1.00 equiv) in anhydrous CHCl3 (0.50 mL/mmol) was added and the reaction was stirred at rt for 30 min before cooling it to 0 °C (ice bath). After 30 min, hydrogen peroxide (30%, 1.20 equiv) was added and the mixture was stirred vigorously at 0 °C for 20 h. Then, the reaction was quenched with 10% aqueous solution of Na2S2O3 (3.30 mL/mmol sulfide) and extracted with DCM (×3). The combined organic layers were dried over Na2SO4, filtered, and the solvent was evaporated under reduced pressure. The crude residue was purified by column chromatography to afford the corresponding enantioenriched sulfoxide 23.

(R)-23. Following the previous procedure, compound (R)-23 was obtained from 56 (318 mg, 1.28 mmol) using chiral ligand (R)-2-((E)-{[1-(hydroxymethyl)-2,2-dimethylpropyl]imino}methyl)-4,6-diiodophenol (9.1 mg, 0.020 mmol), as oil (198 mg, 59%, er = 97.9:2.1). Chromatography: hexane to hexane/EtOAc 3:7. [α]D20 81.2 (c 0.24, acetone). Chiral HPLC (tR, min): 17.31. Spectroscopic data were in agreement with those described for racemate 23.

(S)-23. Following the previous procedure, compound (S)-23 was obtained from 56 (152 mg, 0.610 mmol) using chiral ligand (S)-2-((E)-{[1-(hydroxymethyl)-2,2-dimethylpropyl]imino}methyl)-4,6-diiodophenol (4.3 mg, 0.010 mmol), as oil (103 mg, 64%, er = 97.9:2.1). Chromatography: hexane to hexane/EtOAc 3:7. [α]D20 −80.3 (c 0.24, acetone). Chiral HPLC (tR, min): 15.66. Spectroscopic data were in agreement with those described for racemate 23.

(R)- and (S)-2-(Fluoromethoxy)-4′-(S-methanesulfonimidoyl)-1,1′-biphenyl [(R)- and (S)-26]

The adequate sulfoxide (R)- or (S)-23 (1.00 equiv), ammonium carbamate (4.00 equiv), and (diacetoxyiodo)benzene (3.00 equiv) were dissolved in methanol (2.0 mL/mmol) and the mixture was stirred at rt for 30–60 min (TLC) in an open flask. After completion, the solvent was evaporated under reduced pressure and the residue was purified by column chromatography to afford the corresponding enantioenriched sulfoximine 26.

(R)-26. Following the previous procedure, compound (R)-26 was obtained from (R)-23 (335 mg, 1.27 mmol) as a white solid (216 mg, 61%, er = 97.9:2.1). Chromatography: hexane to hexane/EtOAc 7:3. mp 95–96 °C. [α]D20 −21.5 (c 0.20, acetone). Chiral HPLC (tR, min): 17.17. Spectroscopic data were in agreement with those described for racemic compound 26.

(S)-26. Following the previous procedure, compound (S)-26 was obtained from (S)-23 (282 mg, 1.07 mmol) as a white solid (200 mg, 67%, er = 97.3:2.7). Chromatography: hexane to hexane/EtOAc 7:3. [α]D20 20.8 (c 0.20, acetone). Chiral HPLC (tR, min): 14.60. Spectroscopic data were in agreement with those described for racemic compound 26.

Pharmacology

Functional Activity at Human D1R and D2R

Functional studies at human D1 and D2 receptors were carried out in endogenously expressed human D1R in the SK-N-MC cell line (ATCC), and recombinant human D2R stably transfected in an in-house CHO cell line. Cells were thawed and seeded into a black 96-well plate (1 × 104 cells/well for SK-N-MC cell line, 5 × 103 cells/well for CHO cell line) in Opti-MEM containing 500 μM IBMX (3-isobutyl-1-methylxanthine). Test compounds were added to the cells and incubated for 15 min at 25 °C for D1R, and for 5 min at 37 °C for D2R. After this time, dopamine was added to the corresponding wells and incubated for additional 15 min at 25 °C for D1R, and 10 min at 37 °C for D2R. Next, treated cells were lysed (for the D2R assay, cells were previously treated with 10 μM forskolin for 5 min at 37 °C) and the cAMP concentration was measured by HTRF, using a kit from Cisbio. HTRF was read in a Tecan M1000 Genius Pro reader (λexcitation = 320 nm, λemission = 620 and 665 nm, 30 flashes), and data were normalized to the dopamine maximum effect and fitted to a 4-parameter logistic equation by using Prism v2.1 software (GraphPad Inc).

Functional Activity at Human D3R and D4R

Functional studies at both human D3 and D4 receptors were carried out in human D3 and D4 receptors, using a PatHunter Beta-arrestin eXpress GPCR assay kit from DiscoverX. Briefly, cells were plated and maintained for 48 h at 37 °C in a 5% CO2 atmosphere in cell plating reagent provided in the kit. After this time, test compounds were added to the cells and incubated for 30 min at 37 °C. Dopamine was added to the cells and incubated for an additional 90 min at 37 °C. Cells were then treated with the working detection solution provided with the kit for 1 h and beta-arrestin translocation was measured by luminescence detection (integration time = 500 ms) in a Tecan M1000 Genius Pro reader. Data were normalized to dopamine maximum effect and fitted to a 4-parameter logistic equation using Prism v2.1 software (GraphPad Inc).

Functional Activity at Human D5R

Functional studies at human D5R receptors were carried out in human D5R, using a cAMP Hunter eXpress DRD5 assay kit from DiscoverX. Briefly, cells were plated and maintained for 24 h at 37 °C in a 5% CO2 atmosphere in the cell plating reagent provided in the kit. After this time, test compounds were added to the cells and incubated for 15 min at 37 °C, followed by treatment with dopamine and incubation for an additional 30 min at 37 °C. Cells were then treated with the cAMP working detection solution provided with the kit for 1 h, and then cAMP solution A was added and incubated for 3 h at rt. cAMP formation was measured by luminescence detection (integration time = 500 ms) in a Tecan M1000 Genius Pro reader. Data were normalized to dopamine maximum effect and fitted to a 4-parameter logistic equation using Prism v2.1 software (GraphPad Inc).

Radioligand Binding Assays at Human D1R

Binding assays were carried out by using plasma membranes (12 μg/well) expressing human dopamine D1R (Perkin Elmer). Membranes were incubated in a Multiscreen FC 96-well plate (Millipore) with 0.7 nM [3H]SCH23390 (Perkin Elmer) and increasing concentrations of the test compound in assay buffer (50 mM Tris–HCl, 5 mM MgCl2; pH 7.4) for 1 h at 27 °C. After this time, the well content was filtered through a Millipore manifold and membranes were washed four times with assay buffer. The plate was dried and radioactivity was measured in a Microbeta Trilux liquid scintillation reader (Perkin Elmer). Non-specific binding was determined in the presence of 1 μM (+)-butaclamol. Data were normalized to the percentage of specific binding and fitted to a 4-parameter logistic equation using with Prism v2.1 software (GraphPad Inc). For the calculation of the affinity ratio of DA in the presence of the allosteric modulators, data were fitted according to the equations reported by Lazareno and Birdsall.[44]

Cocaine-induced Hyperlocomotion Assay

The experiment was conducted in adherence to the European Communities Council Directive (86/609/ECC) and Spanish regulations (BOE 252/34367-91, 2005) for the use of laboratory animals. The assay was performed according to a previously reported method[57] using 3 month old male mice with C57BL/6J background, and 8 mice were used for each experimental condition. All mice were handled and habituated to the injection procedures once per day for 5 days prior to behavioral testing. Experiments were carried out between 8:00 and 20:00, and the animals were acclimated to the experimental room for 30 min each day. Performance in the open field was recorded by a computer-based video tracking system (Smart v2.5, Panlab, Barcelona, Spain). Four open fields (50 × 50 × 50 cm, Panlab) with gray backgrounds were used, and the maximum light intensity in the center of the open field was 100 lux. Animals received a dose of 26 (1 mg/kg, ip) in a volume of 20 mL/kg of sterile saline as the vehicle, 15 min prior to a dose of cocaine (20 mg/kg, sc). Five min after cocaine injection, animals were placed in the open field and locomotion was measured at 5 blocks of 2 min along a 30-min period of analysis. Horizontal locomotion was measured as total distance traveled (cm).

Locomotor Activity in Reserpinized Mice

These studies were conducted in SAI Life Sciences Ltd. and all procedures were in accordance with the guidelines provided by the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA) as published in The Gazette of India, December 15, 1998. The study procedures and husbandry care of the study animals were performed in compliance with AAALAC (Unit no. 001384), OLAW (PHS Assurance No.A5937-01) and CPCSEA (Reg. no. 1240/PO/RcBi/S/08/CPCSEA) norms.

The study was carried out in male C57BL/6 wild-type (22–25 g) mice that were acclimatized for 5 days. Animals were randomized based on body weight into six groups (G1 to G6), ten animals per group (n = 10), being group G1 the naïve control group. Animals were fed with standard diet ad libitum. Animals were dosed with subcutaneous administration of reserpine (2.5 mg/kg in DMSO/saline) and after 20 h they were treated with a vehicle (10 mL/kg, po), compound 26 (30 mg/kg, po), l-DOPA (100 or 400 mg/kg, ip) + carbidopa (100 mg/kg, ip) and/or their combination, as indicated in Figure . The vehicle employed for compound 26 administration was 5% NMP + 5% solutol HS-15 + 90% normal saline. l-DOPA and carbidopa were administered 30 min prior to start of locomotor activity recording. Compound 26 was administered and 15 min later, mice were placed in the locomotor activity cage to measure the locomotor activity for 60 min (total distance travelled) using ALL Maze version 0.5 automated software enabled with video tracking system. After the activity sessions, raw data were reduced using ALL Maze 5.0 software and Microsoft Excel. Statistical analysis was done with GraphPad Prism-5 using one way ANOVA followed by Dunnett’s test/Tukey’s post hoc test and/or two-way ANOVA followed by Bonferroni post hoc test. Data were considered statistically significant, if P value was less than 0.05.

Novel Object Recognition Test

Procedures for animal experiments were conducted in adherence to the European Communities Council Directive (86/609/ECC) and Spanish regulations (BOE 252/34367-91, 2005) for the use of laboratory animals. Experiments were performed on 3 month old male mice with a C57BL/6J background. Eight mice were used for each experimental condition. Mice were maintained on a 12 h light/dark cycle (lights on at 08:00 am), with water and food provided ad libitum. Object recognition test was performed according to a previously reported method.[58] Briefly, mice were first habituated to an empty open field apparatus (40 × 40 × 40 cm) for 5 min. After the habituation session, mice received an ip injection of compound 26 (1 mg/kg) or vehicle (NaCl 0.9%). Twenty minutes later, mice were placed again in the open field, which now included two identical copies of an object (“familiar” object) located in two adjacent corners, and mice were allowed to explore for 10 min (acquisition session). In the first test session (24 h after the acquisition session), the objects were replaced by a copy of the familiar object and an unknown object (“new object 1”) located in the previous positions. Finally, in the second test session (96 h after the acquisition session), the objects were replaced by another copy of the familiar object and a new unknown object (“new object 2”). Mice were allowed to explore for 10 min in each test session. The total time of object exploration (defined as the mouse touching an object with its nose or forepaws) was scored, and object recognition memory was calculated by the following discrimination ratio: time exploring the new object—time exploring the familiar object)/total time exploring both objects.

Docking Studies

Docking calculations were performed using Autodock4[59] (using ga_num_evals = 2512000, ga_run = 100 and all the other parameters set to their default values). The D1R active state homology model was retrieved from a previously reported[38] while the EM structure was retrieved from Protein Data Bank (PDB code 7X2F).[49,60] Both structures were prepared for docking using pdb2pqr[61,62] with propka[63,64] protonation option at pH 7.4 and the peoepb force field.[65] (R)- and (S)-26 were modeled using RDKIT (Open-source cheminformatics) and their protonation state were adjusted at pH 7.4 by ChemAxon cxcalc module (command line version of ChemAxon’s Calculator Plugins, v16.10.24.0, 2016). Binding mode pictures were created using PyMOL v1.8 and PLIP v1.4.4.

Short MD simulations were performed using Gromacs v201814.[66] Previous to any simulation, the protein–ligand complex was embedded into a 128 POPC pre-equilibrated membrane. The Amber14SB,[67] the GAFF[68] and the Lipid14[69] force-field set of parameters were employed for the D1R active state model, the docking poses and the POPC (phosphatidylcholine) membrane, respectively. Simulations were carried out in explicit solvent using the SPC (simple point-charge) water model[70] with the imposition of periodic boundary conditions via a cubic box, at 0.15 M (NaCl). The temperature was maintained at 300 K using a V-rescale thermostat[71] and the pressure was maintained at 1 atm using a Berendsen barostat.[70] All bonds involving hydrogen atoms were constrained using the LINCS algorithm.[72,73] Before the production runs, the structure was energy minimized followed by a NVT (number of particles, volume, and temperature) equilibration, a slow heating-up phase (at constant temperature and pressure) and a NPT (number of particles, pressure, and temperature) equilibration, using harmonic position restraints on the heavy atoms of the protein. Then, the production run was performed without position restraints.