Synthesis and evaluation of arylpiperazines derivatives of 3,5-dioxo-(2H,4H)-1,2,4-triazine as 5-HT_1AR ligands

Abstract

5-HT_1AR agonist or partial agonists are established drug candidates for psychiatric and neurological disorders. We have reported the synthesis and evaluation of a series of high affinity 5-HT_1AR partial agonist PET imaging agents with greater selectivity over α-1AR. The characteristic of these molecules are 3,5-dioxo-(2H,4H)-1,2,4-triazine skeleton tethered to an arylpiperazine unit through an alkyl side chain. The most potent 5-HT_1AR agonistic properties were found to be associated with the molecules bearing C-4 alkyl group as the linker. Therefore development of 3,5-dioxo-(2H,4H)-1,2,4-triazine bearing arylpiperazine derivatives may provide high affinity selective 5-HT_1AR ligands. Herein we describe the synthesis and evaluation of the binding properties of a series of arylpiperazine analogues of 3,5-dioxo-(2H,4H)-1,2,4-triazine.

The serotonin receptor, 5-HT_1AR, is a G protein-coupled receptor (GPCR).¹ It is widespread and heterogeneously distributed in CNS as somatodendritic autoreceptors in raphe nuclei and postsynaptically in cortical and subcortical brain regions.² In addition to 5-HT, 5-HT_1AR affects release and signaling by a variety of neurotransmitters, hormones and growth factors in brain.^3–5 Alteration of 5-HT_1AR binding has been implicated in the pathophysiology of a variety of neuropsychiatric diseases and neurodegenerative disorders.^6–9 5-HT_1AR agonists (eg: buspirone and tandospirone) are currently approved drugs for anxiety and depression.^{10, 11} 5-HT_1AR agonists and partial agonists are being evaluated as antipsychotic drugs with fewer side effects for the treatment of schizophrenia.^12–14 Furthermore, 5-HT_1AR desensitization and increased 5-HT_1AR postsynaptic activation has also been postulated as part of the antidepressant action of SSRIs and perhaps other antidepressants.^{2, 5}

Many 5-HT_1AR agonists and partial agonists belong to aminotetraline and arylpiperazine structural skeletons (Figure 1).¹⁵ Recent additions of compounds to these prominent structural skeletons include CUMI-101, S14506 (1), benzo[d]oxazole 2, alkylthiobenzimidazole, benzothiazole, and benzofuro[7,6-g]quinoline 3.^16–18 Compounds belonging to several novel structural classes such as derivatives of 4-(aminomethyl)-4-fluoropiperidin-1-ones 4, 5-spiroimidazolidine-2,4-dione 5, indolylalkyl-amines, ergolines, dihydrofuoroaporphines, 1,3-oxathiolane, 1,3-dioxolanes and pyrroloimidazoles have also been identified recently as agonist ligands at 5-HT_1AR (Figure 1).^16–23 Although several ligands with high affinity for 5-HT_1AR have been developed, a major drawback of these ligands is their significant affinity for alpha1-adregenic receptor (α-1AR).

Figure 1.

Represenatative examples of recent 5-HT_1AR agonist ligands

Positron Emission Tomography (PET) is a noninvasive method to measure 5-HT_1AR binding in living human brain.^{24, 25} Estimation of in vivo occupancy of agonist drugs at 5-HT_1AR is a valuable tool for drug development and monitoring the therapeutic intervention. Additionally, in vivo PET estimation of the sensitivity of binding of 5-HT_1AR ligands to competition by intrasynaptic serotonin would make it possible to compare serotonergic transmission in healthy and diseased populations.²⁴ A number of 5-HT_1AR antagonist radioligands such as [carbonyl-¹¹C] WAY-100635, [¹⁸F]MPPF and [¹⁸F]FCWAY have been routinely used for research PET studies in man.^24,25 However, an agonist ligand would bind preferentially to the G-protein coupled high affinity (HA) state of the receptor and thereby offers advantages to an antagonist ligand.²⁴

Several studies attempted to develop a 5-HT_1AR agonist PET imaging agent with limited success in vivo.^{24, 25} We have recently identified several high affinity and selective 5-HT_1AR ligands as candidates for PET ligands.^26–29 Of these [¹¹C]CUMI-101, is the only successful partial agonist PET tracer available so far for the in vivo measurement of high affinity 5-HT_1AR agonist binding in nonhuman primates and human subjects with PET.^{26, 30–33} The characteristic feature of the ligands we identified is the presence of a 3,5-dioxo-(2H,4H)-1,2,4-triazine skeleton tethered to an arylpiperazine unit through a C-4 alkyl side chain. Prior to our studies 3,5-dioxo-(2H,4H)-1,2,4-triazines were reported as 5-HT_1AR ligands in a patent application with limited information.³⁴ Herein we describe the synthesis and evaluation of a series of 3,5-dioxo-(2H,4H)-1,2,4-triazine derivatives as 5-HT_1AR ligands.

The candidate 5-HT_1AR ligands characterized by an 3,5-dioxo-(2H,4H)-1,2,4-triazine tethered to arylpiperazines through an alkyl side chain were prepared by refluxing a solution of 2-(4-chlorobutyl)-4-methyl-1,2,4-triazine-3,5(2H,4H)-dione (6) preactivated with sodium iodide in acetonitrile with substituted arylpiperazine (7) in the presence of anhydrous K₂CO₃.²⁸ Thirty six analogues of 3,5-dioxo-(2H,4H)-1,2,4-triazine (compounds 8-41) were synthesized from the corresponding substituted arylpiperazines (Scheme 1).

Scheme 1.

Synthesis of azaurazil derivatives of aryl and heteroaryl peperazines

Most of the arylpiperazines were commercially available and used as such without further purification. 2,4-Dimethoxyphenylpiperazine was prepared by the treatment of 2,4-dimethoxyphenyl amine with bis(2-chloroethyl)amine hydrochloride in the presence of potassium carbonate in 2-butoxyethanol at 150 °C.³⁵ The naphthyl piperazines, for the synthesis of 3,5-dioxo-(2H,4H)-1,2,4-triazine derivatives (compounds 39-43) were prepared by the condensation of the corresponding napthylamines with bis(2-chloroethylamine)- hydrochloride in polyethylene glycol under microwave irradiation.²⁸ Binding affinity values (Ki) of the compounds at 5-HT_1AR were determined by competition binding studies with [³H]8-OH-DPAT using racemic 8-OHDPAT as the reference compound.^26–28 Ki of the compounds at α-IAR were determined by competition binding studies with [³H]-7-methoxyprazosin using phentolamine as the reference compound.^26–28 The agonist properties of selected ligands were examined on 5-HT_1AR using [³⁵S]GTPγS binding in membranes of Chinese hamster ovary cells stably expressing the human 5-HT_1AR (CHO-h5-HT_1A cells).^26–28

The unsubstituted phenyl analogue 18 has a 2.5 nM Ki for 5-HT_1AR and 25 times selectivity over α-1AR (Ki = 64 nM). In the case of substituted phenyl derivatives (Table 1), a variety of substituents such as methoxy (8), ethoxy (12), thiomethyl (17), cyano (20), fluoro (25), fluoroethyl (13), fluoropropyl (14), hydroxyl (24), thiohydroxy (31) groups were well tolerated at ortho-position and exhibited nanomolar affinity toward 5-HT_1AR. A few of these ligands have meaningful selectivity over α-1AR such as 45, 210, 42, >10,000, and >10000 times for compounds 8,13,14,24 and 31 respectively. The methoxy (9) and hydroxyl (25) substitution at meta-position of the phenyl group also resulted in high affinity ligands with excellent selectivity for 5-HT_1AR compared to α-1AR. The substitution of dimethylamino (15) and methylamino (16) analogues did not provide any 5-HT_1AR specificity or selectivity. Substitution on the phenyl moiety resulted in a sharp decrease in affinity for the 5-HT_1AR receptor for 4-methoxy, 1,4-dimethoxy, 1-cyano-4-methoxy, 4-iodo analogs 10, 11, 23 and 19 respectively. The introduction of a pyridyl ring results in high affinity 5-HT_1AR binding with excellent selectivity over α-1AR (27, 29, 32, 33).

Table 1.

Substituted phenyl and heterocyclic derivatives of 3,5-dioxo-(2H,4H)-1,2,4-triazine tethered piperazines

Compd	n	X	Y	R1	R	5-HT1AR Ki (nM)	α1_AR Ki (nM)
8 (CUMI-101)26	3	C	C	OCH3	H	0.15	6.75
927	3	C	C	H	5, OCH3	1.1	413
10	3	C	C	H	4,OCH3	414	>10,000
11	3	C	C	OCH3	4,OCH3	190	>10,000
12	3	C	C	OCH2CH3	H	0.66	1.8
1329	3	C	C	OCH2CH2F	H	0.1	21.4
14	3	C	C	OCH2CH2CH2F	H	0.8	280
15	3	C	C	N(CH3)2	H	180	350
16	3	C	C	NHCH3	H	220	450
17	3	C	C	SCH3	H	0.46	1.8
18	3	C	C	H	H	2.5	64
19	3	C	C	H	4, I	58.8	103
20	3	C	C	CN	H	0.835	7.3
21	3	C	C	CN	2, F	4.8	>10,000
22	3	C	C	CN	4, OMe	215	380
23	3	C	C	CN	2, Br	28	450
24	3	C	C	OH	H	1.7	>10,000
25	3	C	C	H	5, OH	2.8	>10,000
26	3	C	C	F	H	1.89	9.73
27	3	C	N	H	H	1.1	262
28	3	C	N	H	4, F	68	>10,000
29	3	C	N	H	2, F	1.2	>10,000
30	3	N	N	H	H	175	>10,000
31	3	C	C	SH	H	8	>10,000
32	3	N	C	H	3, F	0.2	>10,000
33	3	N	C	H	3, Br	1.4	>10,000
34	1	C	C	OCH3	H	240	>10,000
35	2	C	C	OCH3	H	180	>10,000
36	2	C	C	H	H	350	>10,000
37	2	C	C	F	H	250	>10,000
38	4	C	C	OCH3	H	150	>10,000

However, the pyridyl derivative with meta-substituted fluoro analog 28 had modest 5-HT_1AR binding affinity. Similarly replacement of pyrimidine ring instead of phenyl group results in a remarkable decrease in 5-HT_1AR binding affinity (30). Several substituted naphthyl derivatives also exhibited nanomolar affinity at 5-HT_1AR as well as possess reasonable selectivity over α-1AR (39-41, Table 2). The ideal length of alkyl linker was identified to be 4. When the chain length was decreased to 3 (35-37, 43), there was a dramatic drop in the binding affinity at 5-HT_1AR (Tables 1 and 2). Similarly we did not see considerable improvement in 5-HT_1AR binding affinity for linker with length 5 as shown for molecules 38 and 42.

Table 2.

Substituted naphthyl derivatives of 3,5-dioxo-(2H,4H)-1,2,4-triazine tethered piperazines.

Compd	n	R	5-HT1AR Ki (nM)	α1-AR Ki (nM)
826			0.15	6.75
39	3	OH	0.44	24.5
40[28,34]	3	OCH3	1.6	32.0
41	3	OCH2CH3	2.1	48
42	4	OCH3	240	>10,000
43	2	OCH3	175	>10,000

The promising compounds with higher ratios of 5-HT_1AR vs. α-1AR were tested for 5-HT_1AR functional measurements.³⁶ Agonist increased [³⁵]GTPγS binding over basal level in a concentration dependent manner. Maximal ligand stimulated [³⁵]GTPγS binding (E_max) for the selected compounds were in the range of 80–95% of that seen with 5-HT (Table 3).

Table 3.

Agonist properties of selected 5-HT_1AR ligands

Compd	GTPγS EC50, (nM)	GTPγS Emax(%)
8	0.1	80
9	0.7	95
13	0.65	77
18	0.3	93
27	0.08	85
29	0.8	80
32	0.2	95
40	0.05	95

In summary, a novel series of 3,5-dioxo-(2H,4H)-1,2,4-triazine tethered arylpiperazines has been identified as candidate agonist ligands with high affinity for 5-HT_1AR. Several members of the series such as 32 have nanomolar affinity for 5-HT_1AR, higher selectivity to 5-HT_1AR over α-1AR in comparison to CUMI-101 and robust agonist profile as measured by GTPγS binding.