Synthesis and structure–activity relationships of ionizable 1,3,4-oxadiazol-2(3H)-ones as peripherally selective FAAH inhibitors with improved aqueous solubility

Introduction

Fatty acid amide hydrolase (FAAH), an integral membrane bound enzyme, is a member of the extensive family of serine hydrolases [1] that catalyses the degradation of lipid signalling fatty acid amides including oleamide and anandamide (AEA) [2]. AEA binds and activates both central (CB1) and peripheral (CB2) receptors of the endocannabinoid system (ECS) and potentially has clinical relevance in a wide range of diseases and pathological conditions. Therefore, both the ECS and FAAH have become recognised as promising therapeutic targets for the treatment of a range of central and peripheral disorders [3]. Due to increasing interest over the last two decades, numerous small molecule inhibitors of FAAH belonging to various chemical classes have been reported (for recent reviews see [4–7] and references cited therein).

Recently, our group described a series of 3-phenolic and 3-catecholic 5-phenoxy-1,3,4-oxadiazol-2(3H)-ones possessing in vivo FAAH inhibition in mice [8]. Several of the disclosed compounds showed peripherally selective inhibition of FAAH that may be particularly beneficial for the treatment of certain cardiovascular diseases such as hypertension [9] and heart failure [10]. More recently, other research groups have also disclosed alternative 1,3,4-oxadiazol-2(3H)-one analogues demonstrating good selectivity and high affinity for FAAH [11–13]. Within the frame of the same project, we prepared in parallel a series of peripherally selective inhibitors designed to have physicochemical properties eventually suitable for the treatment of respiratory [14] and eye disorders [15] via topical, local or systemic administration.

As an early lead, catechol 1 (Fig. 1) showing potent in vivo inhibition of FAAH at 10 and 30 mg/kg p.o in the mouse was selected. It was hypothesised that substitution of the phenolic hydroxyls with ionisable amine-containing residues could enhance aqueous solubility whilst maintaining the favourable in vivo inhibitory profiles of the parent phenols 2 and 3 (Fig. 1).

Fig. 1:

Chemical structures of lead compounds 1–3.

Chemistry

The synthetic route to obtain the target compounds from phenols 2 or 3 (Fig. 1) and alcohols 4a–p is briefly outlined in Scheme 1. Commercially available (4b–d, 4l) or in-house prepared alcohols (according to published procedures; 4a, 4e–k, 4m–p [16–25]) were condensed with the corresponding phenols 2–3 under standard Mitsunobu reaction conditions in tetrahydrofuran (THF) to provide the N-protected or non-protonated adducts 5–23 in moderate to good yield [26].

Scheme 1:

(a) Reagents and conditions: Ph₃P, DEAD, THF, 0–5°C, then 20–25°C; (b) Neat TFA, 20–25°C, then 2 M ethereal HCl solution in EtOAc; *represents point of attachment.

Cleavage of the N-Boc protecting group was performed with neat trifluoroacetic acid (TFA) at 20–25°C. For most of the compounds (except 28, 29 and 41) the TFA salts thus obtained were converted to corresponding hydrochlorides by treatment with 2 M ethereal hydrochloric acid in inrt solvent such as ethyl acetate [27].

In a similar fashion, the non-protonated tertiary amines were converted to their hydrochloride salts. Structures of all prepared target amino derivatives 24–43 are shown in Table 1.

Table 1:

In vivo (in homogenates of mouse liver and brain) FAAH inhibition by synthesised 5-(2,4-difluorophenoxy)-3-phenyl-1,3,4-oxadiazol-2(3H)-ones 24–43.

No.	R5	R6	Salt	Braina, b, c	Livera, b, c	Solubilityd
URB597	–	–	–	8.9±2	5.9±1	<0.05
2	OH	H	–	86±11	12±3	<0.055
3	H	OH	–	94±5	12±2	<0.055
24		H	2×HCl	74±15	5±1	>10
25		H	HCl	20±7	10±4	13.0
26		H	HCl	50±11	14±2	1.95
27		H	HCl	10±5	9±2	>71
28	H		TFA	108±7	69±20	4.6
29		H	TFA	74±5	7±3	>32
30		H	2×HCl	85±10	8±3	>75
31		H	HCl	63±6	17±3	0.37
32	H		HCl	57±10	37±4	0.1
33		H	HCl	86±9	22±2	0.35
34	H		HCl	87±17	48±2	0.46
35		H	HCl	65±10	12±8	0.1
36		H	HCl	87±5	53±9	0.1
37		H	–	102±13	87±12	<0.006
38		H	HCl	78±11	24±3	1.6
39		H	HCl	82±10	31±12	0.5
40		H	HCl	93±4	39±14	>35.5
41		H	HCl	62±19	7±1	2.65
42		H	HCl	88±4	29±16	0.47
43		H	HCl	90±13	11±3	>2

^a% of Control; ^b30 mg/kg, p.o, FAAH activity was determined 1 h after administration [28]; ^cresults are mean SEMs of 4–8 experiments; ^dmg/mL in water.

Results and discussion

As reported previously [8], within the 5-phenoxy-3-phenyl-1,3,4-oxadiazol-2(3H)-one scaffold few correlations could be established between results of in vitro and in vivo FAAH assays, and therefore, only in vivo experiments were used herein to evaluate the new series. FAAH inhibition by test compounds 24–43 was evaluated in mice at an oral dose of 30 mg/kg. Animals were sacrificed 1 h after administration and FAAH activity was determined in the liver and brain. The effects on FAAH inhibition of different substituents on the N-phenyl ring of the 1,3,4-oxadiazol-2(3H)-one core are shown in Table 1, which includes comparative data for reference compounds 2, 3 and URB597 [29]. The aqueous solubility of compounds 24–43 were also determined (Table 1).

Incorporation of a piperazine moiety at the para-position of the N-phenyl ring of oxadiazolone 2via an ethoxy linker resulted in a slightly more potent derivative 24, which was endowed with reasonable peripheral selectivity. Replacement of the secondary nitrogen atom in the piperazine ring of 24 either with a methylene group (25) or oxygen atom (26) led to a reduction in peripheral selectivity. Substitution of the piperidine moiety in 25 with the smaller pyrrolidine gave 27, which was highly potent but uninteresting in terms of peripheral selectivity. Shifting the ethoxy-piperazinyl moiety of 24 to the meta-position gave oxadiazolone 28, which exhibited only residual FAAH inhibition. Elongation of the ethoxy linker in 24 to the n-propoxy (29–30) made little difference in peripheral activity and selectivity. Incorporation of an N-phenyl ring in 24 at the secondary nitrogen atom of the piperidine ring led to a less potent derivative 31, while the same modification in 28 was found to be beneficial (32). A similar trend could be observed in the case of N-benzyl substituted analogues (cf. 28 and 34; 24 and 33). The presence of a more rigid bicyclic 1,2,3,4-tetrahydroisoquinoline substituent was well tolerated with an ethoxy linker in compound 35, however, the n-propoxy analogue 36 showed lower activity than 35. Surprisingly, when the size of the cycloaliphatic ring was reduced from 6- to 5-members as in 37, FAAH inhibition was practically abolished. Interestingly, an acyclic analogue of 35 (compound 38) was significantly less potent than the parent; whereas the acyclic counterpart of 36 (compound 39) was slightly more active. Replacement of the N-benzyl substituent in 38 with cyclohexyl group (40) appeared to be unfavourable. Attractive selectivity and potency was obtained when the piperidine residue was connected at position 4 or 3 of the ring, thus leaving the secondary nitrogen unsubstituted as in 41–43. For this particular series of compounds, a shorter methoxy linker proved superior (cf. 41 and 42), particularly in combination with the heterocyclic ring anchored at position 3 as in 43.

The time-dependent FAAH inhibition by selected compounds was also determined. Figures 2 and 3 highlight the inhibitory profile of oxadiazolones 24, 30 and 41 in peripheral and central tissues, respectively.

Fig. 2:

Liver FAAH activity after oral administration of 24 (closed black circle), 30 (open circle) and 41 (closed grey circle) (all at 30 mg/kg).

Fig. 3:

Brain FAAH activity after oral administration of 24 (closed black circle), 30 (open circle) and 41 (closed grey circle) (all at 30 mg/kg).

Compounds 24, 30 and 41 achieved the maximal inhibitory effect (92–95% inhibition) in the liver within 1 h after oral administration to mice at dose of 30 mg/kg. Thereafter, the inhibitory ability gradually diminished and returned to lower levels at 24 h post administration. The profile of central inhibition by 24, 30 and 41 was found to be different to that of liver inhibition. Oxadiazolones 24 and 30 produced only minimal inhibitory effect (15–26%) in the brain at 1 and 4 h, respectively, whereas compound 41 displayed slightly more marked inhibition (38%) up to 4 h post-dose. Compounds 24, 30 and 41 had no significant effect in the brain at time points later than 8 h.

Conclusions

A novel series of 5-(2,4-difluorophenoxy)-3-aryl-1,3,4-oxadiazol-2(3H)-ones was prepared. Several analogues were synthesised that provided significantly improved aqueous solubility over the phenols 2 and 3. Compounds 24 and 30 showed potent and peripherally selective in vivo FAAH inhibition after an oral dose of 30 mg/kg. Given that both failed to significantly modulate FAAH in the CNS and have improved physicochemical properties suitable for formulations for topical or inhaled administration, these molecules could prove beneficial for the treatment of certain respiratory and ocular disorders in which FAAH plays a role. 1,3,4-Oxadiazol-2(3H)-ones are known to inhibit other serine hydrolases such as monoacylglycerol lipase (MAGL) [12], α/β hydrolase domain 6 (ABHD6) [30], human lymphocyte antigen B-associated transcript 5 (BAT5) [31] and hormone-sensitive lipase (HSL) [11]. Derivatives 24 and 30 are currently being further evaluated to determine their target selectivity and therapeutic potential. Further pharmacological data on compounds 24 and 30 will be published elsewhere.