Three-Step Synthesis of the Antiepileptic Drug Candidate Pynegabine

Abstract

Pynegabine, an antiepileptic drug candidate in phase I clinical trials, is a structural analog of the marketed drug retigabine with improved chemical stability, strong efficacy, and a better safety margin. The reported shortest synthetic route for pynegabine contains six steps and involves the manipulation of highly toxic methyl chloroformate and dangerous hydrogen gas. To improve the feasibility of drug production, we developed a concise, three-step process using unconventional methoxycarbonylation and highly efficient Buchwald–Hartwig cross coupling. The new synthetic route generated pynegabine at the decagram scale without column chromatographic purification and avoided the dangerous manipulation of hazardous reagents.

1. Introduction

Epilepsy, a chronic disorder of the brain, is one of the most common neurological diseases, affecting around 50 million people globally [1]. Clinical antiepileptic drugs (AEDs) are only effective in about 70% of patients with epilepsy, making the discovery of new AEDs with novel mechanisms and/or low side effects critically important [2,3]. Retigabine (Figure 1), an AED approved by the U.S. Food and Drug Administration (FDA) in 2011, is the first and only marketed drug functioning as an agonist of KCNQ2 voltage-gated potassium ion channels [4]. However, in 2013, the FDA added warnings to the drug label emphasizing the serious risks associated with the use of retigabine, which included skin discoloration and retinal pigment abnormalities [5]. Subsequently, the production of retigabine was permanently discontinued in 2017 due to the declining market for the medication [6]. Several pigmented dimers of retigabine have been identified in an in vivo study, suggesting the correlation between the discoloration and the triaminobenzene segment of retigabine [7]. Therefore, the above adverse events of retigabine seem to originate from its chemical structure but not from the KCNQ2 target [8]. Accordingly, great efforts are made to search for new KCNQ2 agonists for AED candidates based on the optimization of retigabine structure [9,10,11,12,13,14,15,16].

Pynegabine, also known as HN37, is a structural analog of retigabine (Figure 1) [17]. Its ratios of the brain to blood exposure are 10 times more than those of retigabine in mice and rats. The EC₅₀ of pynegabine for KCNQ2 channels and KCNQ2/KCNQ3 channels are also 55- and 125-fold more potent than those of retigabine. In a maximal electroshock seizure model test, the TD₅₀/ED₅₀ ratio of pynegabine was at least eight-fold bigger than those of retigabine and two other AEDs (levetiracetam and topiramate) which show a better safety margin. More importantly, by change of the triaminobenzene segment in retigabine, pynegabine exhibited much better stability for temperature, humidity, and light, implying a lower risk of discoloration [18]. Pynegabine has progressed into phase I clinical trials for epilepsy treatment in China (ChinaDrugTrials Identifier: CTR20201676, CTR20222616).

There are two synthetic routes reported for pynegabine, both starting from 2,6-dimethylaniline (1, Scheme 1). The first-generation route prepared pynegabine at the milligram scale with a 30% total yield in 10 steps, out of which four steps applied column chromatography for purification [17,18]. Since this route is not feasible for large-scale production, a second-generation route was developed, achieving the synthesis of 10 g of pynegabine with a 30% total yield over six steps [17]. Although the latter synthetic route eliminated the unfavorable chromatographic purification procedure, the requirements for hazardous reagents and conditions such as highly toxic methyl chloroformate [19] for the preparation of compound 11 and the dangerous manipulation of hydrogen gas [20] still remained. Therefore, the development of a more efficient and safer synthetic route for the manufacture of pynegabine is highly desirable. Herein, we describe a concise, three-step synthetic route for the preparation of pynegabine at the 10 g scale with easily handling reagents and conditions.

2. Results and Discussions

Our proposed new route is shown in Scheme 2. The core reaction of this three-step strategy is the palladium-catalyzed Buchwald–Hartwig cross coupling of bromide 17 and commercially available p-fluorobenzylamine (18). The Buchwald–Hartwig reaction is a highly versatile protocol used to forge aryl C–N bonds [21]. In the present case, this reaction would considerably shorten the synthesis route by replacing the two-step reductive amination (from 13 to 15) and omitting the nitrogen-installation steps (nitration and hydrogenation from 11 to 13). Bromide 17 can be prepared from commercially available 2,6-dimethyl-4-bromoaniline (16), and the known compound 15 can be transformed to pynegabine by a process similar to that applied in the original route (Scheme 1).

2.1. Methoxycarbonylation of Aniline 16

To avoid the toxic methyl chloroformate (from 10 to pynegabine and from 1 to 11 in Scheme 1), we applied an unusual method [22] for the methoxycarbonylation of aniline 16 (

Table 1. Methoxycarbonylation of aniline 16.

Entry	16 (g)	Boc2O (eq)	DMAP (eq)	MeCN (mL)	MeOH (mL)	Time 1 (h)	Yield 2 (%)
1	5.0	1.4	1.0	100	5	20	63
2	5.0	1.4	0.1	100	5	20	66
3	20.0	1.4	0.1	200	50	8	65
4 3	2.0	1.4	0.1	0	20	80	<20 4

¹ Reflux reaction time following MeOH addition. ² Yield of 17 after recrystallization. ³ MeCN was not used and MeOH was added from beginning. ⁴ Calculated according to the ¹H NMR ratio of the crude mixture. rt = room temperature; min = minute; eq = equivalent; h = hour.

). Treatment of 16 with di-tert-butyl dicarbonate (Boc₂O) and 4-(dimethylamino)pyridine (DMAP) in MeCN afforded isocyanate 19 within 10 min at room temperature. The addition of MeOH to the reaction mixture, followed by heating to reflux, provided carbamate 17 in full conversion. The purification of 17 was simply achieved without column chromatography through acidic aqueous washing to remove DMAP, followed by recrystallization. Using the 0.1 equivalent of DMAP instead of the stoichiometric amount in the literature [22] did not affect the yield (Table 1, entry 2) and the reaction could be scaled up to 20 g scale (Table 1, entry 3). Notably, in this reaction, MeOH should be added immediately after the complete formation of 19 because of the easy oligomerization property of isocyanate [23]. We observed a bulk of precipitate (presumably polyisocyanate) when extending the isocyanation step to 30 min before MeOH addition in a preliminary experiment. Additionally, the use of MeOH instead of MeCN at the beginning of synthesis (Table 1, entry 4) resulted in low conversion, thereby confirming the importance of MeCN as the initial solvent.

2.2. Buchwald–Hartwig Cross Coupling of the Compounds 17 and 18

For the key Buchwald–Hartwig cross coupling between the compounds 17 and 18, we initially screened the reaction conditions using allylpalladium (II) chloride dimer as the precatalyst and BrettPhos as the ligand [24]. With Cs₂CO₃ as the base (

Table 2. Screening of the reaction conditions for Buchwald–Hartwig cross coupling of 17 and 18.

Entry	18 (eq)	Ligand	Base	Solvent	Temp (°C)	Time (h)	Yield (%)
1	1.1	BrettPhos	Cs2CO3	CPME	95	16	66 1
2	1.1	BrettPhos	Cs2CO3	PhMe	95	16	25 1
3	1.1	BrettPhos	Cs2CO3	tAmOH	95	16	22 1
4	1.1	BrettPhos	Cs2CO3	DMAc	95	16	18 1
5	1.1	BrettPhos	tBuONa	THF	40	16	86 1
6	1.1	XPhos	tBuONa	THF	40	16	86 1
7	1.1	tBu-XPhos	tBuONa	THF	40	16	89 1
8	1.1	Me4tBu-XPhos	tBuONa	THF	40	16	96 1
9	1.1	JohnPhos	tBuONa	THF	40	16	83 1
10	1.1	DavePhos	tBuONa	THF	40	16	80 1
11	1.1	CPhos	tBuONa	THF	40	16	86 1
12	1.1	BippyPhos	tBuONa	THF	40	16	98 1
13 4	1.1	BippyPhos	tBuONa	THF	40	23	73 2
14 4	0.95	BippyPhos	tBuONa	THF	40	23	63 3
15 4	0.95	BippyPhos	tBuONa	THF	50	5	86 3
16 5	0.95	BippyPhos	tBuONa	THF	50	3	72 3
17 6	0.95	BippyPhos	tBuONa	THF	50	2	78 3

¹ Determined by HPLC using naphthalene as an internal standard. ² Isolated yield after column chromatography. ³ Isolated yield after recrystallization. ⁴ Using 0.2 g of 17. ⁵ Using 2 g of 17. ⁶ Using 12.5 g of 17. CPME = cyclopentyl methyl ether; DMAc = N,N-dimethylacetamide; THF = tetrahydrofuran.

, entries 1–4), ether-type solvent gave the optimal conversion at 95 °C, and a higher yield was obtained with a stronger base (tBuONa) in tetrahydrofuran (THF) at a lower temperature (Table 2, entry 5). Further ligand screening (Table 2, entries 6–12) revealed that BippyPhos [25] was most suitable for this coupling reaction (Table 2, entry 12). However, since substrate 18 and product 15 were both amines, the excess 18 which could not be fully consumed need to be removed by column chromatography (Table 2, entry 13). Therefore, using a slight excess of 17 was tested. Under this condition, 18 was completely consumed and pure 15 was successfully obtained through acid–base extraction and recrystallization without column chromatography (Table 2, entry 14). Further optimization of the reaction temperature and time (Table 2, entry 15) showed that a slightly higher temperature (50 °C) gave better yield and reduced the reaction time to a few hours. Scale-up of the reaction to decagram levels maintained the yield level (Table 2, entries 16 and 17).

Interestingly, ¹H-NMR spectra of compound 15 revealed two sets of peaks (integration ratio: ~2:1) for OMe and amide NH groups in CDCl₃ (Figure 2a). This might be due to the slow rotation of the hindered carbamate C–N bond [26]. Changing the solvent to d₆-DMSO altered the integration ratio to ~4:1, confirming its origination from the cis-trans isomerization of the amide group. The ¹³C-NMR spectra also showed a similar peak pattern (see Supporting Information).

2.3. Propargylation of Compound 15

The third step in pynegabine synthesis involves the propargylation of compound 15. We reproduced this reaction successfully at a 10 g scale with an 85% yield using a similar procedure as that reported previously (Scheme 1b). Additionally, we found that the target pynegabine also showed amide-isomeric peaks in the NMR spectra using CDCl₃ and d₆-DMSO as the solvents (Figure 2b).

3. Materials and Methods

3.1. General Information

Reagents were used as received from commercial suppliers unless otherwise indicated. Reaction solvents (MeCN, MeOH, THF, DMF) were purchased as anhydrous solvents (H₂O ≤ 50 ppm) and stored with active molecular sieves 3A or 4A under Ar. All solvents for work-up procedures were used as received. Analytical thin-layer chromatography (TLC) was performed using glass TLC plates (silica gel HSGF254 plates, Yantai Huayang New Material Technology Co., Ltd., Yantai, China). Visualization was accomplished with UV light or staining with 5% (w/v) phosphomolybdic acid/EtOH followed by heating.

Melting points were uncorrected and measured with a Yanaco MP-J3 melting point apparatus. ¹H- and ¹³C-NMR spectrum were acquired on a Bruker AVANCE III-400 spectrometer. Chemical shifts are indicated in parts per million (ppm) downfield from tetramethylsilane (TMS, δ = 0.00) with residual undeuterated solvent peaks used as internal references for ¹H-NMR and deuterated solvent peak shifts for ¹³C-NMR. Multiplicities are reported as s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), br (broad), or combinations of those. Mass spectra were generated using electrospray ionization (ESI) and measured on a Thermo-Fisher Accela liquid chromatography system coupled with an Exactive Plus Orbitrap mass spectrometer.

3.2. Synthesis of Methyl (4-Bromo-2,6-dimethylphenyl)carbamate (17)

DMAP (1.22 g, 10.0 mmol) and 4-bromo-2,6-dimethylaniline (compound 16, 20.0 g, 100 mmol) were successively added to a solution of Boc₂O (30.5 g, 140 mmol) and MeCN (200 mL) in a 500-mL flask at room temperature under an Ar atmosphere. The reaction mixture was stirred for 10 min, followed by the addition of MeOH (50 mL) and reflux at 82 °C for 8 h. After cooling to room temperature, the reaction mixture was concentrated under reduced pressure. CH₂Cl₂ (40 mL) was then added to the solid residue, and the organic phase was washed with 1 mol/L hydrochloric acid (40 mL × 3), dried with Na₂SO₄, filtered, and concentrated under reduced pressure. The solid residue was then dissolved in CH₂Cl₂ (60 mL), and petroleum ether (150 mL) was added dropwise under stirring to induce precipitation. The precipitate was filtered to obtain compound 17 (16.90 g, 65% yield) as a white solid. m.p. 108.0–109.5 °C. ¹H-NMR (400 MHz, CDCl₃) δ 7.22 (s, 2H), 5.97 (s, 1H), 3.76 (s, 3H), 2.23 (s, 6H). ¹³C-NMR (101 MHz, CDCl₃) δ 154.7, 138.0 (2C), 132.7, 131.0 (2C), 120.7, 52.7, 18.2 (2C). HRMS (ESI): m/z 258.0126 [M+H]⁺ (calcd for C₁₀H₁₃BrNO₂⁺: 258.0124).

3.3. Synthesis of Methyl (4-(4-Fluorobenzylamino)-2,6-dimethylphenyl)carbamate (15)

THF (94 mL) and 4-fluorobenzylamine (compound 18, 5.26 mL, 5.76 g, 46.0 mmol) were successively added to the mixture of compound 17 (12.5 g, 48.4 mmol), [Pd(allyl)Cl]₂ (0.443 g, 1.21 mmol), and BippyPhos (2.45 g, 4.83 mmol) in a 350 mL flask at room temperature under an Ar atmosphere. The reaction mixture was stirred for 10 min, followed by the addition of tBuONa (1 mol/L in THF, 143 mL, 143 mmol) and stirring at 50 °C for 2 h. After cooling to room temperature, 36% hydrochloric acid (7 mL) was added to adjust the pH to 6, and the mixture was concentrated under reduced pressure. Methyl tert-butyl ether (100 mL) was added to the solid residue, and the organic phase was extracted with 0.1 mol/L hydrochloric acid (100 mL × 15). The combined aqueous layer was then adjusted to pH 13 with 1 mol/L aqueous NaOH and extracted with CH₂Cl₂ (750 mL × 2). The combined organic layer was then dried with Na₂SO₄, filtered, and concentrated under reduced pressure. Petroleum ether was added dropwise to the residue under stirring to induce precipitation. The precipitate was filtered to give compound 15 (10.92 g, 78% yield) as a white solid. m.p. 123.6–124.9 °C. ¹H-NMR (400 MHz, CDCl₃) δ 7.31 (dd, J = 8.8, 5.6 Hz, 2H), 7.02 (dd, J = 8.8, 8.8 Hz, 2H), 6.33 (s, 2H), 5.86 (s, 0.66H), 5.73 (s, 0.34H), 4.25 (s, 2H), 3.94 (s, 1H), 3.74 (s, 2H), 3.68 (s, 1H), 2.16 (s, 6H). ¹H-NMR (400 MHz, d₆-DMSO) δ 8.22 (s, 0.8H), 7.95 (s, 0.2H), 7.37 (dd, J = 8.4, 5.6 Hz, 2H), 7.13 (dd, J = 8.8, 8.4 Hz, 2H), 6.26 (s, 2H), 6.06 (t, J = 6.0 Hz, 1H), 4.22 (d, J = 6.0 Hz, 2H), 3.58 (s, 2.4H), 3.48 (s, 0.6H), 1.99 (s, 6H). ¹³C-NMR (101 MHz, CDCl₃) δ 162.0 (d, J = 245.1 Hz), 156.8 (minor), 155.6 (major), 147.0, 137.2 (2C), 135.1, 129.0 (d, J = 7.9 Hz, 2C), 124.3 (minor), 123.9 (major), 115.4 (d, J = 21.3 Hz, 2C), 112.3 (major, 2C), 112.0 (minor, 2C), 52.8 (minor), 52.4 (major), 47.6, 18.5 (2C). ¹³C-NMR (101 MHz, d₆-DMSO) δ 161.0 (d, J = 241.4 Hz), 155.8 (minor), 155.4 (major), 146.8, 136.5 (d, J = 2.9 Hz), 136.0 (major, 2C), 135.8 (minor, 2C), 128.9 (d, J = 7.9 Hz, 2C), 124.1 (minor), 123.8 (major), 115.0 (d, J = 21.2 Hz, 2C), 111.4 (2C), 51.6 (minor), 51.4 (major), 45.6, 18.3 (2C). HRMS (ESI): m/z 303.1508 [M + H]⁺ (calcd for C₁₇H₂₀FN₂O₂⁺: 303.1503).

3.4. Synthesis of Pynegabine

Diisopropylethylamine (12.1 mL, 8.98 g, 69.5 mmol) and propargyl bromide (3.59 mL, 4.95 g, 41.6 mmol) were successively added to a solution of compound 15 (10.5 g, 34.7 mmol) and DMF (84 mL) in a 500 mL flask at room temperature under an Ar atmosphere. The reaction mixture was then stirred at 60 °C for 14 h and after cooling to room temperature, water (350 mL) was added to induce precipitation. Filtration gave a solid residue that was dried and dissolved in CH₂Cl₂ (30 mL). Petroleum ether (90 mL) was added dropwise under stirring to induce precipitation, and the precipitate was filtered to give pynegabine (10.02 g, 85% yield) as a white solid. m.p. 136.2–137.6 °C. ¹H-NMR (400 MHz, CDCl₃) δ 7.27 (dd, J = 8.8, 5.2 Hz, 2H), 7.01 (dd, J = 8.8, 8.8 Hz, 2H), 6.59 (s, 2H), 5.92 (br, 1H), 4.47 (s, 2H), 3.95 (s, 2H), 3.75 (s, 2H), 3.68 (s, 1H), 2.20 (brs, 7H). ¹H-NMR (400 MHz, d₆-DMSO) δ 8.36 (s, 0.8H), 8.08 (s, 0.2H), 7.33 (dd, J = 8.8, 6.0 Hz, 2H), 7.15 (dd, J = 8.8, 8.8 Hz, 2H), 6.54 (s, 2H), 4.48 (s, 2H), 4.09 (s, 2H), 3.60 (s, 2.4H), 3.50 (s, 0.6H), 3.14 (s, 1H), 2.06 (s, 6H). ¹³C-NMR (101 MHz, CDCl₃) δ 162.0 (d, J = 244.9 Hz), 156.7 (minor), 155.5 (major), 147.6, 136.9 (2C), 134.0, 128.8 (d, J = 7.9 Hz, 2C), 125.2 (minor), 124.9 (major), 115.4 (d, J = 21.3 Hz, 2C), 113.9 (major, 2C), 113.6 (minor, 2C), 79.5, 72.3, 54.3, 52.8 (minor), 52.4 (major), 39.7, 18.8 (2C). ¹³C-NMR (101 MHz, d6-DMSO) δ 161.2 (d, J = 242.1 Hz), 155.7 (minor), 155.2 (major), 146.1, 136.0 (major, 2C), 135.9 (minor, 2C), 134.9 (d, J = 2.6 Hz), 129.0 (d, J = 8.0 Hz, 2C), 125.8 (minor), 125.5 (major), 115.1 (d, J = 21.2 Hz, 2C), 113.2 (2C), 80.5, 74.6, 53.4, 51.7 (minor), 51.5 (major), 40.1, 18.5 (2C). Elemental analysis calcd for C₂₀H₂₁FN₂O₂: C, 70.57; H, 6.22; N, 8.23. Found: C, 69.61; H, 6.09; N, 8.23. HRMS (ESI): m/z 341.1667 [M + H]⁺ (calcd for C₂₀H₂₂FN₂O₂⁺: 341.1660).

4. Conclusions

We successfully developed a concise, three-step route for the decagram-scale synthesis of the AED candidate pynegabine (Scheme 3). This route provided a 41% total yield without column chromatographic purification and avoided the original dangerous manipulation of methyl chloroformate and hydrogen gas, thereby making this method more suitable for pynegabine manufacture in the subsequent phases of clinical trials and future marketing.

5. Patents

S.X., Y.-J.S., S.-P.Z., Y.-L.G. and S.-C.L. are inventors on patent application 202310142629.9 submitted by the Institute of Materia Medica, Chinese Academy of Medical Sciences, that covers a part of the work of this manuscript.