Piperazinyl Bicyclic Derivatives as Selective Ligands of the α2δ-1 Subunit of Voltage-Gated Calcium Channels

Abstract

The synthesis and pharmacological activities of a new series of piperazinyl quinazolin-4-(3H)-one derivatives acting toward the α2δ-1 subunit of voltage-gated calcium channels (Ca_vα2δ-1) are reported. Different positions of a micromolar HTS hit were explored, and best activities were obtained for compounds containing a small alkyl group in position 3 of the quinazolin-4-(3H)-one scaffold and a 3-methyl-piperazin-1-yl- or 3,5-dimethyl-piperazin-1-yl-butyl group in position 2. The activity was shown to reside in the R enantiomer of the chain in position 2, and several eutomers reached single digit nanomolar affinities. Final modification of the central scaffold to reduce lipophilicity provided the pyrido[4,3-d]pyrimidin-4(3H)-one 16RR, which showed high selectivity for Ca_vα2δ-1 versus Ca_vα2δ-2, probably linked to its improved analgesic efficacy-safety ratio in mice over pregabalin.

Gabapentinoids, represented by pregabalin (1, Figure 1), are close analogues of the inhibitory neurotransmitter γ-aminobutyric acid (GABA) and are postulated to exert their action by blocking the α2δ subunit of voltage-gated calcium channels (VGCC).¹ This subunit is one of the four constituents of VGCC, namely, α₁ (Ca_vα₁), β (Ca_vβ), α2δ (Ca_vα2δ), and γ (Ca_vγ), and although it exerts an auxiliary function, it is key for increasing the expression of the central pore forming α₁ subunits in the plasma membrane and modulating their function. This provides a reduction in the calcium-dependent release of multiple neurotransmitters and calcium upregulation in the spinal dorsal horn and dorsal root ganglia (DRG). As a result, gabapentinoids are useful drugs in the treatment of several neuropathies, such as neuropathic pain² and fibromyalgia³ and other nonpain indications, such as epilepsy⁴ or restless leg syndrome.⁵

Figure 1.

Structures of pregabalin (1) and HTS hit 2 and summary of SAR study leading to 16RR.

The overall function of the α2δ subunit is still a matter of intense research.⁶ The mechanism of calcium current increase is unknown, and it can only be partially explained by an increase in the trafficking of the channels to augment their amount on the cell surface. In addition, there is a pool of α2δ not associated with calcium channels but to other proteins that may account for some of its actions. These proteins include trombospondins, low-density lipoprotein receptor-related protein 1, large conductance potassium channels, and N-methyl-d-aspartate (NMDA) receptors.^6,7 In addition, emerging roles of the α2δ subunit in synaptic connectivity⁸ or as key organizers of glutamatergic synapses⁹ are being identified and they will probably open the way to new therapeutic applications of Ca_vα2δ ligands.

Four Ca_vα2δ (Ca_vα2δ1–4) subunits of VGCC have been described with a differentiated distribution: while the Ca_vα2δ-1 subunit is quite ubiquitous, the Ca_vα2δ-2 and Ca_vα2δ-3 subunits are mainly concentrated in the brain and the Ca_vα2δ-4 subunit is largely non-neuronal. Pregabalin,⁷ gabapentin,¹⁰ and mirogabalin,¹¹ the three marketed gabapentinoids, bind to both the Ca_vα2δ-1 and the Ca_vα2δ-2 subunits (but not to the Ca_vα2δ-3 and Ca_vα2δ-4 subunits).

The Ca_vα2δ-1 subunit is present in many neuronal cell types,¹² including DRG neurons, where it is upregulated in neuropathic pain conditions^13,14 and modulates spinal hyperexcitability.¹⁵ Point mutations on the binding site on Ca_vα2δ-1 in a transgenic mouse line diminished binding and antihyperalgesic activity of gabapentinoids,^15,16 and recent findings suggest that injury-induced maladaptation of the Ca_vα2δ-1 subunit is responsible for the chronic development of neuropathic pain.¹⁷ In addition, the Ca_vα2δ-1 subunit is the one forming heterodimeric complexes with NMDA receptors to potentiate their synaptic trafficking and excitatory synaptic transmission. Interrupting this interaction reverses the synaptic activity associated with neuropathic pain.¹⁸

On the other hand, the Ca_vα2δ2 expression is not upregulated in neuropathic pain models^13,14 and is restricted to some areas, such as the cerebellum, a key player in coordinating voluntary movements.¹² In fact, the deletion of the Ca_vα2δ2 subunit causes cerebellar ataxia and absence seizures in mice and humans,^19,20 via a mechanism that probably does not involve calcium channels.²¹

The previous findings, among others, led to the hypothesis that the Ca_vα2δ-1 contributes to the analgesic effects of pregabalin and gabapentin, whereas the Ca_vα2δ-2 contributes to their central nervous system (CNS) side effects (such as sedation, dizziness, ataxia and fatigue). In the case of mirogabalin, its slower dissociation rate from Ca_vα2δ1 versus Ca_vα2δ2 is postulated to be the reason behind its wider safety margin in relation to pregabalin, which shows the same dissociation rate for both of them.¹¹

In concordance with the previously mentioned distribution, behavioral, and neurochemical studies, we set up a program devoted to the finding of compounds binding to the Ca_vα2δ-1 with selectivity toward the Ca_vα2δ-2 subunit. The final aim was to find safer drugs to be used as analgesic agents or in the treatment of other neuropathies. A high throughput screening (HTS) with our internal library (ca. 80 000 compounds) provided compound 2 (Figure 1, K_i Ca_vα2δ-1 = 1810 ± 82 nM) as an unprecedented scaffold in the field. It is worth mentioning that most of the described Ca_vα2δ-1 ligands are amino acidic structures closely related to 1, with only a few non-amino acidic scaffolds published. These have been summarized elsewhere²² and in our recent work describing a new family of bicyclic diazepinones as dual ligands of the Ca_vα2δ-1 and the norepinephrine transporter.²³

We describe here the structure–activity relationship (SAR) study that led from the HTS hit 2 to the identification of the highly selective derivative 16RR, through a process summarized in Figure 1 and outlined below.

The synthesis of the compounds was accomplished as depicted in Schemes 1 and 2.²⁴ Anthranilic acids 3 were coupled to amines 4 using 1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium-3-oxid hexafluorophosphate (HATU) and the resulting amides were reacted with acyl chlorides 6 in DCM in the presence of TEA. Cyclization of diamides 7 using iodine and hexamethyldisilazane (HMDS) provided quinazolin-4-(3H)-one derivatives 8 in good overall yields (around 65% for the three steps). Bromination of the α position of the alkyl chain in position 2 was accomplished in good yield (ca. 80%) with bromine in acetic acid. Finally, displacement of the bromide leaving group of 9 with the corresponding amines 10 provided compounds 11 in good yields (60–80%). Most of the amines 10 contained a second free NH group, but no protection was used. Instead, an excess of amine directly provided the target compounds. The potential dimeric derivatives were not observed at all or were detected in the crude mixtures in very small proportion and never isolated. In the case of unsymmetrical amines, addition always occurred by the less sterically hindered nitrogen, and a decrease in reactivity was observed when α-substituted amines were used. All the compounds 11a–x were prepared following this process, except 11n, 11p, and 11r that were obtained from the corresponding bromo derivatives 11m and 11q. As described in the experimental part, compound 11m was Boc-protected and treated with Zn(CN)₂ under catalysis of Pd(PPh₃)₄ or with pyridin-4-ylboronic acid under Suzuki conditions, followed in both cases by deprotection with trifluoroacetic acid in DCM, to provide 11n and 11p, respectively. Analogously, 11r was obtained from 11q.

Scheme 1.

Reagents and conditions: (a) DMF, TEA, HATU, 0 °C to rt, overnight; (b) DCM, TEA, 0 °C to rt, overnight; (c) I₂, HMDS, DCM, 0 °C to rt, overnight; (d) Br₂, AcOH, NaOAc, 50 °C, 3 h; (e) ACN, TEA, KI, 90 °C, overnight.

Scheme 2.

Reagents and conditions: (a) H₂SO₄, NaNO₂, 0 °C to rt, 16 h; (b) HATU, TEA, DCM, rt, 16 h; (c) I₂, HMDS, DCM, 50 °C, 16 h; (d) TBAF, THF, 0 °C, 30 min; (e) 2,6-lutidine, Tf₂O, DCM, −78 °C, 2 h; (f) DMF/DCM, −78 °C to rt, 4 h (; (g) MsCl, DIPEA, DCM, −78 °C, 5 min; (h) ACN, 80 °C, 24 h.

The final compounds contain an asymmetric center in the alkyl chain of position 2 and were obtained as racemic mixtures, except when indicated with the suffix R or S, corresponding to the R or S configuration, respectively. Several enantiomers were separated by chiral preparative HPLC, as exemplified here by compound 11m (Table 2). The activity was always shown to reside in one of the enantiomers, while the other exhibited only residual affinity. The four individual diastereoisomers of compound 11x were obtained by reaction with 2-methylpiperazine with the bromo derivative 9a followed by chiral HPLC preparative separation.

Table 2. Modification of R¹ over Compound 11k.

compd	R1	Cavα2δ-1 (Ki, nM)a	CYP3A4 inhibition (%)b
11m	6-Br	16 ± 28	95
11mR	6-Br	8 ± 2	94
11mS	6-Br	2903 ± 703	88
11n	6-CN	288 ± 139	70
11o	6-NHEt	44 ± 23	69
11p	6-(4-pyridinyl)	38 ± 13	92
11q	7-Br	15 ± 24	97
11r	7-(4-pyridinyl)	18 ± 24	97
11s	5-Br	66 ± 16	95
11t	8-Br	>1000c	NDd

^aBinding affinity (K_i, nM) in human α2δ-1 enriched membranes from hamster tumor CHO-K1 cells (Human Cav2.2/β3/α2δ1 Calcium Channel Cell Line, ChanTest) using [³H]-gabapentin as the radioligand. Each value is the mean ± SD of two determinations.

^bPercentage of inhibition after incubation at 1 μM with recombinant human CYP3A4 using BFC (7-benzyloxy-4-trifluoromethylcoumarin) as the CYP probe substrate and fluorescence detection.

^cLess than 50% inhibition at 1 μM.

^dNot determined.

The pyrido[4,3-d]pyrimidin-4(3H)-one derivatives 16 (Table 4) were synthesized in an analogous way to that described in Scheme 1, but using alternative conditions. As described in the experimental part, the initial 4-aminonicotinic acid was converted to 4-aminonicotinoyl chloride and reacted with ethylamine (instead of HATU-coupling), followed by reaction with pentanoyl chloride and cyclization with TMSCl (instead of iodine). Bromination was performed with NBS in acetonitrile and final displacement with (R)- or (S)-2-methylpiperazine provided the two pairs of diastereoisomers, that were separated using chiral HPLC.

Table 4. Diastereoisomers of Compound 16.

compd	Cavα2δ-1 (Ki, nM)a	CYP3A4 inhibition (%)b
16RR	6 ± 9	46
16RS	27 ± 8	56
16SR	>1000c	15
16SS	>1000c	44

^aBinding affinity (K_i, nM) in human α2δ-1 enriched membranes from hamster tumor CHO-K1 cells (Human Cav2.2/β3/α2δ1 Calcium Channel Cell Line, ChanTest) using [³H]-gabapentin as the radioligand. Each value is the mean ± SD of two determinations.

^bPercentage of inhibition after incubation at 1 μM with recombinant human CYP3A4 using BFC (7-benzyloxy-4-trifluoromethylcoumarin) as the CYP probe substrate and fluorescence detection.

^cLess than 50% inhibition at 1 μM.

The eutomers of compounds 11 were assigned to the R configuration, based on the enantioselective route applied to different compounds and exemplified in Scheme 2 for the 6-bromo derivative. The route involved the formation of compound 14 by reaction of 5a with (S)-2-hydroxypentanoic acid 13, that was obtained from L-norvaline 12 using a described method that proceeds with retention of configuration.²⁵ Cyclization of 14 in the conditions described above provided a mixture of compound 15 with some trimethylsilyl protected alcohol, which was fully deprotected using tetrabutylammonium fluoride (TBAF). Transformation of the hydroxyl group to a triflate and its displacement with the corresponding amines 10 occurred with inversion of configuration, to provide the final derivatives 11xRS and 11zR. Alternatively, conversion of the hydroxyl group to a mesylate and subsequent displacement provided the final derivative 11yR. This allowed the assignment of the R configuration to the eutomers, since the three compounds were identical to the corresponding active isomers isolated by chiral preparative HPLC from the racemic mixtures obtained using Scheme 1. The enantiomeric ratios were good in all cases, as determined by chiral HPLC analysis (enantiomeric ratios of 97.0(+):3.0(−), 96.5(+):3.5(−), and 95.7(+):4.3(−) for 11xRS, 11yR, and 11zR, respectively). The configuration of the pyridyl derivatives 16 was assigned by analogy to the 6-bromo analogues, with the R enantiomer as the eutomer.

The biological activity was evaluated at 10 μM in human α2δ-1 and α2δ-2 enriched membranes using [³H]-gabapentin as the radioligand. For those compounds showing inhibition higher than 50%, the K_i (nM) was calculated. The results are shown in Tables 1-4, where pregabalin (1) and the HTS hit 2 are displayed for comparison purposes. Most of the compounds described were devoid of Ca_vα2δ-2 activity. Therefore, these results are not tabulated and they are only described in the text in the few cases where some affinity was found.

Table 1. Initial Modifications of Hit Compound 2.

compd	sc	R3	R2	Cavα2δ-1 (Ki, nM)a	CYP3A4 inhibition (%)b
1				19 ± 12	43
2	A	Et	2-methoxyethyl	1810 ± 82	37
11a	A	Pr	2-methoxyethyl	638 ± 345	67
11b	A	Me	2-methoxyethyl	4896 ± 1307	0
11c	A	H	2-methoxyethyl	>10 000c	6
11d	B	Pr	2-methoxyethyl	241 ± 246	94
11e	B	Et	2-methoxyethyl	761 ± 235	86
11f	B	Bu	2-methoxyethyl	1340 ± 856	97
11g	B	CyPrMe	2-methoxyethyl	413 ± 188	94
11h	A	Et	3-methoxypropyl	>10 000c	NDd
11i	B	Pr	benzyl	4816 ± 45	98
11j	B	Pr	2-furylmethyl	230 ± 101	99
11k	B	Pr	ethyl	36 ± 28	85
11l	B	Pr	methyl	73 ± 20	80

^aBinding affinity (K_i, nM) in human α2δ-1 enriched membranes from hamster tumor CHO-K1 cells (Human Cav2.2/β3/α2δ1 Calcium Channel Cell Line, ChanTest) using [³H]-gabapentin as the radioligand. Each value is the mean ± SD of two determinations.

^bPercentage of inhibition after incubation at 1 μM with recombinant human CYP3A4 using BFC (7-benzyloxy-4-trifluoromethylcoumarin) as the CYP probe substrate and fluorescence detection.

^cLess than 50% inhibition at 10 μM.

^dNot determined.

No functional activity of the Ca_vα2δ-1 ligands identified was evaluated at this stage, since there is no assay available for screening purposes. Extensive studies are needed to determine suitable approaches (culture conditions, subunit expression, etc.) to provide meaningful information on downstream signaling for each Ca_vα2δ-1 ligand, as shown in the literature for gabapentin.^26,27

To identify from the beginning compounds with good potential regarding their absorption, distribution, metabolism, and excretion (ADME) properties, the active compounds were tested in two additional tests. Metabolic stability was evaluated in human, mouse, and rat liver microsomes,²⁸ and all the compounds tested showed acceptable clearances (generally below 15 μL/min/mg protein). Inhibition in recombinant human cytochrome P450 isoforms (rhCYP1A2, 2C9, 2C19, 2D6, and 3A4) was evaluated to identify potential drug–drug interaction issues.²⁹ In this case, it was found that several derivatives inhibited CYP3A4, and their inhibition values (% at 1 μM) are reported in Tables 1–4.

Initial modification of hit compound 2 involved exploration of positions R² and R³ and the amine moiety NR⁴R^4’. Over the parent compound (R¹ = H, R² = methoxyethyl, R³ = ethyl), the amine was changed and only the (3S,5R)-3,5-dimethylpiperazin-1-yl derivative 11d provided Ca_vα2δ-1 activity, improving that of 2. Many other amines were explored (such as the methylated analogues of 11d and 11k, the piperidin-1-yl, 4-aminopiperidin-1-yl, 2-oxopiperazin-1-yl, or the linear 2-methylaminoethyl derivatives, results not shown), and all of them were devoid of activity. A change of R³ over the piperazin-1-yl or (3S,5R)-3,5-dimethylpiperazin-1-yl derivatives (Table 1) indicated that propyl was the optimal length in both cases (11a and 11d), with shorter (11b–e), longer (11f), or ramified (11g) derivatives being less active. A change of the methoxyethyl substituent in R² showed that increasing length or bulkiness was detrimental (11h and 11i). Activity was somehow restored by introduction of 2-furylmethyl instead of benzyl (11j), but a substantial improvement was seen when reducing the chain length to provide the ethyl or methyl derivatives, 11k and 11l, respectively.

The nature of R¹ was explored while keeping the remaining substituents of 11k (R² = ethyl, R³ = propyl, NR⁴R^4’ = (3S,5R)-3,5-dimethylpiperazin-1-yl). Many different analogues were prepared²⁴ and only a handful of them are shown in Table 2. Introduction of bromine in position 6 or 7 provided 11m and 11q with a 2-fold improvement in affinity versus the parent hydrogenated, 11k. Isolation of the two enantiomers of 11m showed that there was clearly an eutomer, 11mR, assigned to the R configuration via the enantioselective synthesis described above. The distomer 11mS retained a residual affinity. Changing the bromine by cyano in 11n provided a loss in affinity, while the ethylamino derivative 11o retained affinity, as did other electron-donating groups explored (results not shown). Larger groups were also tolerated but led to a worsening of other properties. For instance, compounds 11p and 11r showed 53% and 95% inhibition of Ca_vα2δ-2 at 10 μM, respectively. Finally, substitution in position 5 led to a small impairment in affinity versus the parent hydrogen (see 11s versus 11k), while the 8-bromo derivative 11t was much less active.

An additional round of exploration of the amine substituent was carried out over the 6-bromo-2-butyl-3-ethyl derivatives (Table 3). The naked piperazin-1-yl derivative, 11u, was shown to be less active than the dimethylated 11m, confirming the results of the hydrogenated counterparts. The methylaminopiperidine 11v and bicyclic compound 11w displayed a diminished activity, but the 2-methylpiperazin-4-yl derivatives 11x were more interesting. In this case, the individual diastereoisomers were isolated as described above and both the 11xRS and 11xRR derivatives were highly active. The corresponding enantiomers 11xSR and 11xSS were poorly active, as expected. The (S)-2-ethyl and 2,2-dimethyl derivatives 11yR and 11zR, prepared only via the enantioselective route described in Scheme 2, showed a diminished affinity versus the monomethyl 11xRS.

Table 3. Modification of Amine over 6-Bromo-2-butyl-3-ethyl Derivatives.

compd	NR4R4’	Cavα2δ-1 (Ki, nM)a	CYP3A4 inhibition (%)b
11u	piperazin-1-yl	76 ± 20	65
11v	4-methylaminopiperidin-1-yl	436 ± 60	83
11w	3,8-diazabicyclo[3.2.1]octan-1-yle	397 ± 161	74
11xRS	(S)-3-methylpiperazin-1-yl	91 ± 15	73
11xRR	(R)-3-methylpiperazin-1-yl	16 ± 15	51
11xSS	(S)-3-methylpiperazin-1-yl	479 ± 324	33
11xSR	(R)-3-methylpiperazin-1-yl	>1000c	21
11yR	(R)-3-ethylpiperazin-1-yl	123 ± 340	97
11zR	3,3-dimethylpiperazin-1-yl	1578 ± 1998	72

^aBinding affinity (K_i, nM) in human α2δ-1 enriched membranes from hamster tumor CHO-K1 cells (Human Cav2.2/β3/α2δ1 Calcium Channel Cell Line, ChanTest) using [³H]-gabapentin as the radioligand. Each value is the mean ± SD of two determinations.

^bPercentage of inhibition after incubation at 1 μM with recombinant human CYP3A4 using BFC (7-benzyloxy-4-trifluoromethylcoumarin) as the CYP probe substrate and fluorescence detection.

^cLess than 50% inhibition at 1 μM.

^eDerivative with 6-H instead of 6-Br.

Among the best compounds identified so far, derivatives 11mR and 11q showed high CYP3A4 inhibition. The methylpiperazines 11xRS and 11xRR were better in this regard, but they showed additional activities when tested versus a small panel of targets related to pain (σ₁R, 5-HT_1A receptor, 5-HT_2B receptor, α₁A and α₂A adrenoreceptors, dopamine transporter, μ-, δ-, and κ-opioid receptors, histamine H₁ receptor, norepinephrine transporter and serotonin transporter).³⁰ In fact, the compound 11xRS showed a K_i of 157 ± 67 nM for the μ-opioid receptor and was selected as a lead for a program aimed at the finding of dual compounds with Ca_vα2δ-1 and μ-opioid receptor activity, that will be reported in due course.

The introduction of nitrogen atoms in the central scaffold was then anticipated to decrease the cLogP, which could improve ADME properties and selectivity, since lipophilicity is a well-known parameter associated with suboptimal druglike properties³¹ and promiscuity.³² Several derivatives were prepared, but only the best set of diastereoisomers identified is shown in Table 4. Compound 16RR is the most potent derivative versus the Ca_vα2δ-1 identified so far in this compound class, while its diastereoisomer 16RS is also quite potent. As expected, the other enantiomeric pair (16SR and 16SS) is much less active.

Compound 16RR showed acceptable experimental lipophilicity (log P = 2.22, log D_7.4 = 0.49) and pKa (9.1 and 3.4) values and a good in vitro profile. It was devoid of CYP3A4 or 1A2, 2C9, 2C19, 2D6 inhibition (<50% at 1 μM), showed good metabolic stability in human, rat, and mouse liver microsomes (Cl_int = 2.6, 3.9, and 4.1 μL/min/mg protein, respectively), was highly selective versus the pain panel mentioned before, and was devoid of hERG inhibition (IC₅₀ > 10 μM). The only drawback identified in the ADME profile of compound 16RR was its suboptimal permeability in Caco-2 cells (43 nm/s) and elevated efflux ratio (5.4), that could be indicative of it being substrate of P-glycoprotein (P-gp).

To have a first indication of the series in vivo behavior, compound 16RR was evaluated in two mouse efficacy and safety models. Considering its expected poor oral absorption and brain distribution, it was administered using the ip route at the same dose in both models. Pregabalin was used for comparison in the same experimental paradigm. The analgesic activity was evaluated in the capsaicin test in CD1 male mice (n = 8–12 per group). After ip administration at 20 mg/kg, 16RR showed a 69% antiallodynic effect, that was superior to that shown by pregabalin (1, 34%) (Figure 2A). The plasma levels of 16RR were high at this dose (C_max 3071 ng/mL and AUC_0–t 8052 ng·h/mL). When tested in the rotarod test at 20 mg/kg ip, 16RR did not provide any sign of motor discoordination (Figure 2B), as no changes in the latency to fall-down at any time point were observed. In contrast, the same doses of pregabalin induced a marked motor impairment, suggesting that 16RR could have a superior safety margin compared to 1 in pain treatment.

Figure 2.

(A) Effects of pregabalin (1) and 16RR on mechanical hypersensitivity (allodynia) induced by intraplantar injection of capsaicin (1 μg) to mouse hind paw. The results represent the percentage reduction in capsaicin-induced mechanical hypersensitivity. (B) Effect of 1 and 16RR on motor coordination using the rotarod test. The results represent the latencies to fall-down from the elevated rotating drum. Rotarod latencies were measured at 30, 120, and 180 min after administration of vehicle or compounds 1 and 16RR at 20 mg/kg ip.

In summary, exploration of the different positions of a micromolar piperazinyl quinazolin-4-(3H)-one HTS hit provided a series of derivatives with single digit nanomolar affinities versus the Ca_vα2δ-1 subunit of VGCC. Modification of the central quinazolinone to reduce lipophilicity, provided the pyrido[4,3-d]pyrimidin-4(3H)-one lead compound 16RR, an unprecedented scaffold in the field, that shows an excellent selectivity over the Ca_vα2δ-2, the subunit postulated to be linked to the undesired effects of gabapentinoids. Compound 16RR showed a good ADME profile, except for its moderate permeability in Caco-2 cells and high efflux ratio, properties that have been addressed in a subsequent lead optimization program that will be reported in due course. As an initial indication of the in vivo behavior of the series, compound 16RR assessed in the rotarod model in mice at a dose eliciting analgesia (i.e., capsaicin induced mechanical allodynia) did not induce motor discoordination. In contrast, pregabalin did produce motor discoordination with evident overlapping between the motor impairment and the antiallodynic effect. These results suggest that selectively targeting the Ca_vα2δ-1 in relation to the Ca_vα2δ-2 may provide an interesting approach for the treatment of pain and other neuropathies. Compound 16RR may also serve as pharmacological tool to provide further insight on the roles played by the two Ca_vα2δ-1/2 subunits, in view of the novel functions that are being identified for these proteins, both dependent and independent of calcium channels.

Glossary

Abbreviations

ADME

absorption, distribution, metabolism and excretion

BFC

7-benzyloxy-4-trifluoromethylcoumarin

Ca_vα2δ-1

α2δ-1 subunit of voltage-gated calcium channels

Ca_vα2δ-2

α2δ-2 subunit of voltage-gated calcium channels

CNS

central nervous system

DRG

dorsal root ganglia

HATU

1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium3-oxid hexafluorophosphate

HMDS

hexamethyldisilazane

hERG

human ether-a-go-go-related gene

NMDA

N-methyl-d-aspartate

P-gp

P-glycoprotein

rhCYP

recombinant human cytochrome P450

SAR

structure–activity relationships

VGCC

voltage-gated calcium channels

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.1c00416.

• Detailed synthetic procedures of representative compounds; experimental procedures for the determination of α2δ-1 and α2δ-2 activity and CYP3A4 inhibition; experimental procedures for determination of metabolic stability of 16RR; experimental procedures for in vivo testing: plasma levels determination and efficacy and safety testing in mice; analytical data of final compounds (purity, ¹H NMR, HRMS); ¹³C NMR data for selected final compounds; enantiomeric ratios by chiral HPLC and optical rotation of enantiopure compounds; ¹H NMR characterization of intermediates; HPLC traces of final compounds (PDF)

Author Contributions

The manuscript was written by C.A. and contributions of the rest of the authors, mainly M.G. and J.L.D. A.F., C.A., and J.L.D. contributed to the design of the new compounds. A.F., A.L., and S.R. synthesized the compounds, R.E. analyzed the compounds, A.D. did the binding experiments, E.P.-S. and M. P. did the in vivo pharmacology experiments, R.F.R. did the ADME and pharmacokinetics experiments, B.F. and M.G. coordinated part of the work, and J.M.V. supervised the work. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.

Special Issue

Published as part of the ACS Medicinal Chemistry Letters virtual special issue “Medicinal Chemistry in Portugal and Spain: A Strong Iberian Alliance”.