Asymmetric Synthesis of a 5,6,7,8-Tetrahydro-1,6-naphthyridine Scaffold Leading to Potent Retinoid-Related Orphan Receptor γt Inverse Agonist TAK-828F.

An asymmetric synthesis of the tetrahydronaphthyridine scaffold of TAK-828F as a RORγt inverse agonist has been developed. The synthesis features a newly discovered atom-economical protocol for Heck-type vinylation of chloropyridine using ethylene gas, an unprecedented formation of dihydronaphthyridine directly from 2-vinyl-3-acylpyridine mediated by ammonia, and a ruthenium-catalyzed enantioselective transfer hydrogenation as key steps. This represents the first example of the enantioselective synthesis of a 5,6,7,8-tetrahydro-1,6-naphthyridine compound. The new synthesis is also free of chromatography or distillation purification processes and therefore qualifies for extension to large-scale manufacture.

Introduction

Retinoid-related orphan receptor γt (RORγt), which is an orphan nuclear receptor, plays an important role in the differentiation of Th17 cells and production of IL-17A/IL-17F.[1] Th17 cells and inflammatory cytokines (such as IL-17A and IL-17F) result in a severe etiology accompanying the enhancement of a systemic new immune response in various autoimmune diseases, such as inflammatory bowel disease (IBD), rheumatoid arthritis, multiple sclerosis, and psoriasis.[2] RORγt has been reported to be mainly expressed in Th17 cells and functions as a transcription factor of IL-17A and IL-17F and a master regulator of Th17 cell differentiation.[3] Therefore, a medicament that inhibits the action of RORγt is expected to have a treatment effect on various immune diseases by suppressing the differentiation and activation of Th17 cells.

Through drug discovery, TAK-828F (1) has been identified by Takeda as a potent, selective, and orally available RORγt inverse agonist.[4] TAK-828F (1) is a tetrahydronaphthyridine ring-fused chiral amino acid bearing indane and cyclobutane moieties through two peptide bonds. The original synthetic route developed by the medicinal chemistry group is shown in Scheme .[4c] Pyridinylethylamine 5 was prepared from 2-methoxy-6-methylpyridine (2) via metalation and nucleophilic addition to paraformaldehyde, amination under Mitsunobu conditions, and finally deprotection using hydrazine. The Pictet–Spengler reaction with an ethyl glyoxylate polymer gave tetrahydronaphthyridine 6 as the HCl salt. After Boc protection of the secondary amine, silver-mediated O-selective methylation and hydrolysis of the ethyl ester afforded carboxylic acid 9, which was then condensed with aminoindane 10. The resulting racemate of 11 was subjected to chiral HPLC resolution to give optically pure (R)-11. After deprotection of the Boc group, the second amide bond formation with cyclobutanecarboxylic acid 13 and deprotection of the tert-butyl ester finally produced target compound 1.

Scheme 1

Original Medicinal Chemistry Synthesis of TAK-828F (1)

As the program advanced into the drug development stage, a synthetic process suitable for producing large quantities of the TAK-828F drug substance was needed. In this regard, the original synthesis described above had inherent issues, including (i) a poor overall yield (3.6% over 12 steps in the longest linear sequence); (ii) chromatographic purification; (iii) cryogenic reaction conditions; (iv) hazardous reagents, such as 1,1′-(azodicarbonyl)dipiperidine (ADDP) and hydrazine; (v) undesired methyl ether cleavage during the Pictet–Spengler reaction, resulting in the need for subsequent re-methylation using a stoichiometric amount of silver carbonate; and (vi) racemic synthesis with chiral HPLC resolution at a late stage of the synthesis. Based on these issues, an alternative synthetic route clearly needed to be pursued to develop a scalable synthetic process. However, after an extensive literature search, the synthesis of tetrahydro-1,6-naphthyridines was found to still be underdeveloped despite their significant value as a scaffold of biologically active molecules.[4h,5] Furthermore, to the best of our knowledge, no enantioselective synthesis of this particular ring system had been reported at that time, with only two other reports found on non-enantioselective methods.[6,7] In the original medicinal chemistry synthesis (Scheme ), the chiral center of target molecule 1 was generated by a Pictet–Spengler-type cyclization. However, enantioselective Pictet–Spengler reactions have been reported only for highly activated (hetero)aromatic substrates, such as pyrroles or indoles,[8] with no successful examples reported for inactivated aromatic rings, such as pyridines. Therefore, we aimed to evaluate a different chemical transformation to establish the chiral stereogenic center in an enantioselective fashion. Scheme outlines the retrosynthetic analysis of the projected synthesis. We envisaged that the chiral stereogenic center in the naphthyridine core could be established by asymmetric reduction of dihydronaphthyridine 17. The resulting chiral tetrahydronaphthyridine 16 could then be coupled with 15 to give 12, which is the same precursor in the existing route to target compound 1 (Scheme ). We expected that the synthesis of 17 would be achieved by the amination of 2-vinyl-3-acylpyridine 19 followed by intramolecular condensation, inspired by few literature precedents.[9−11] For an even more streamlined synthesis, we decided to pursue a tandem reaction to achieve these two transformations in one pot.

Scheme 2

Retrosynthetic Analysis for a New Asymmetric Synthesis of 1

Results and Discussion

Pyridinyl-2-oxoacetamide 23, a precursor to vinylpyridine key intermediate 19, was prepared via two different synthetic routes (Scheme ). In route A, the cyanation[12] of nicotinic acid chloride 21, followed by bromide-mediated hydration,[13] afforded 23 in good yield. In route B, an ethyl oxalyl group was introduced by metalation of 24 with a Grignard reagent, followed by mild-temperature treatment with diethyl oxalate. The resulting 25 was then treated with ammonia in ethanol to give 23 in high yield. As compounds 21 and 22 were susceptible to hydrolysis, route B was eventually selected for scale-up synthesis.

Scheme 3

Synthesis of Pyridinyloxoacetamide 23

The vinylation of chloropyridine 23 was initially conducted using potassium vinyltrifluoroborate (Scheme , method A)[10] to give 19 in good yield. As the trifluoroborate was glass-corrosive, not atom-economical, and an expensive vinyl source, its replacement with ethylene gas was attempted. Although the Heck reactions using ethylene gas had been reported for an aryl chloride[14] and aryl bromides,[15] no example was available for the conversion of chloropyridines. Nonetheless, we launched high-throughput screening (Table )[16] and successfully identified a new and effective set of conditions for the vinylation of 23 using DPEphos as the ligand (Scheme , method B).

Scheme 4

Vinylation of Chloropyridine 23

Table 1

Summary of High-Throughput Ligand Screening for the Vinylation of 23 with Ethylene Gas

		HPLC (area %)
entry	cat (mol %)	23	19
1	PdCl2 (20), (p-Tol)3P (40)	20.2	53.6
2	PdCl2 (20), (o-MeOC6H4)3P (40)	43.6	9.2
3	Pd(OAc)2 (20), Xantphos (20)a	12.3	70.7
4	Pd(dppf)Cl2·CH2Cl2 (10)	69.6	13.8
5	PdCl2 (20), DPEphos (20)	0.9	61.1
6	Scheme 4, method B	NDb	94.5

The reaction was conducted in DMF (50 v/w) in the absence of PTZ.

ND = not detected.

With 2-vinyl-3-acylpyridine 19 in hand, the next target was to develop a one-pot hydroamination/cyclization reaction to construct the dihydronaphthyridine ring (Scheme ). As projected in the retrosynthesis, dihydronaphthyridine 17 was obtained in good yield by heating 19 in NH3 solution in MeOH. A small amount of aromatized byproduct 26 was also observed, which was presumably generated from the oxidation of 17 by residual oxygen in the reaction mixture. Indeed, when previously isolated 17 was treated with aq. NaOH in MeOH under air, it was immediately oxidized and converted to 26. Owing to the air sensitivity of the product in solution, the formation of 26 in this step was difficult to completely prevent on a lab scale. However, the oxidized impurity was easily removed by an aqueous workup in the next step and caused no significant issue for the overall synthesis.

Scheme 5

Ammonia-Mediated One-Pot Hydroamination/Cyclization

Determined by 1H NMR.

With the successful development of the ring-closure reaction, our attention was turned to enantioselective reduction of the resulting carbon–nitrogen double bond. High-throughput screening was conducted under more than 100 sets of conditions, including Ru-catalyzed transfer hydrogenation reactions and Ru, Rh, and Ir-catalyzed hydrogenation reactions (Table ),[16] based on previous reports on the asymmetric reduction of dihydroisoquinolines.[17,18] As a result, transfer hydrogenation using catalyst 30(19) was found to be optimal (entry 4, Table ). The reaction was further optimized to afford 16 with excellent conversion and high enantioselectivity (Scheme ). Compound 16 was then Boc-protected and isolated as compound 31 by crystallization with effective upgrade of the enantiomeric purity.

Table 2

Summary of High-Throughput Catalyst Screening for the Asymmetric Reduction of 17

entry	reductant	cat	conditions	HPLC area %	% ee
1	H2 (3 MPa)	27	tBuOK (10 equiv), MeOH, 40 °C	25.9	100
2	H2 (3 MPa)	28	nBu4NI (10 equiv), AcOH/toluene 40 °C	93.5	34.2
3	H2 (3 MPa)	29	MeOH/THF, 40 °C	82.3	85.5
4	HCOOH (6 equiv)	30	Et3N (2.4 equiv), DMF, rt	89.1	82.7

Scheme 6

Ru-Catalyzed Enantioselective Transfer Hydrogenation and Product Isolation

Determined by HPLC.

The coupling reaction of 16 or 31 with haloindane 15 was then examined to obtain the corresponding amide 12 or 11 (Scheme ) as the precursor to target compound TAK-828F (1). Initially, common Pd-catalyzed conditions[20] and copper-mediated methods[21] for amidation were examined using substrate 16. Although the Pd-catalyzed conditions were not effective, the copper-mediated conditions afforded the desired coupling product 12, albeit in a low yield with oxidized byproducts 32 and 33 (Scheme ). However, for the copper-mediated reactions, significant erosion of the optical purity was observed, even under mildly basic conditions. This was an unexpected result because the stereogenic center had proven to be stable under strongly basic conditions, as shown in Table (entry 1). Therefore, the undesired racemization was hypothesized to occur mainly through a redox-based pathway between 12 and 32, which might be promoted in the presence of copper. To prevent the undesired redox-based side reactions, N-Boc-protected dihydronaphthyridine 31 was employed as the substrate for the reaction with aryl iodide 15a or bromide 15b as coupling partners (Table ). Although the use of a substoichiometric amount of copper iodide afforded good conversion with a slight loss in enantioselectivity, the reactivity was only moderate (entry 1). In contrast, the reaction using a stoichiometric amount of copper iodide gave a much better yield with reasonable retention of the stereochemical integrity (entry 2). However, further racemization was observed after a prolonged reaction (entry 3). To our delight, deterioration of the enantiomeric purity was effectively suppressed by lowering the reaction temperature and using a slight excess of aryl iodide 15a (entry 4). The reaction with aryl bromide 15b gave a lower conversion, even at higher temperatures, with significant racemization observed (entry 5) (Table ).

Scheme 7

Copper-Mediated N-Arylation of 16 and the Plausible Redox-Based Racemization Pathway

HPLC area %.

Table 3

Copper-Mediated N-Arylation of 31(16)

entry	Ar-X	CuI (equiv)	DMEDA (equiv)	T (°C)	time (h)	HPLC area % (% ee)
1a	15a	0.5	1.0	40	48	64 (99.2)
2b	15a	1.0	2.0	40	7	84 (97.4)
3b	15a	1.0	2.0	40	24	89 (88.7)
4a	15ac	1.1	2.2	rt	30	82d (99.9)
5a	15b	1.0	2.0	100	7.5	33 (44.0)

Initial optical purity of 31: >99.9% ee.

Initial optical purity of 31: 98.8% ee.

1.1 equiv was used.

Isolated yield: 87%.

Finally, the validity of the new synthetic route was confirmed by converting the resulting compound (R)-11 into the target molecule 1 according to the original route (Scheme ). Compared with the original synthesis of (R)-11, the new synthetic route had successfully decreased the number of steps in the longest linear sequence from nine to six, drastically improved the overall yield (from approx. 4 to 25%, Scheme ), and eliminated the need for hazardous or expensive reagents employed in the original synthesis (Scheme ). Furthermore, the new intermediate, 15a, was readily prepared from the indane fragment 15b through single-step iodination,[22] while the original route required three steps for the conversion of fragment 15b to aminoindane 10 (Scheme ).[4a]

Scheme 9

Summary of the New Synthetic Route to (R)-11

Scheme 8

Synthesis of Iodoindane 15a and Comparison with That of the Original Indane Intermediate 10

Conclusions

A highly efficient asymmetric synthesis of RORγt inverse agonist TAK-828F (1) has been achieved by developing a new synthetic route to the chiral tetrahydronaphthyridine core scaffold. The new synthesis features several key transformations, namely, the Heck reaction of 2-chloropyridine 23 with ethylene gas, the unprecedented one-pot cyclization and amination of 3-acyl-2-vinylpyridine 19, and the enantioselective transfer hydrogenation of dihydronaphthyridine 17. The new synthetic route is also free of chromatographic purification, making it suitable for scale-up.[23] We expect this method to be extendable for the synthesis of various other chiral tetrahydronaphthyridine compounds.

Experimental Section

General Experimental Methods

All reactions were conducted under an inert gas atmosphere using commercially available reagents and solvents without further purification unless otherwise noted. All reactions that required heating were heated using an oil bath. NMR chemical shifts were recorded in ppm relative to tetramethylsilane (0 ppm) as s (singlet), bs (broad singlet), d (doublet), t (triplet), or m (multiplet).

2-Chloro-6-methoxynicotinoyl Chloride (21)

A 100 mL round-bottom flask was charged with 2-chloro-6-methoxynicotinic acid 20 (6.0 g, 32.0 mmol) and SOCl2 (12 mL). The mixture was heated to 60 °C for 3 h with stirring. Volatiles were removed using a rotary evaporator to give a slightly yellowish white solid (6.6 g). The product was used in the next reaction without further purification owing to moisture sensitivity. 1H NMR (500 MHz, CDCl3) δ 8.40 (d, J = 8.5 Hz, 1H), 6.79 (d, J = 8.5 Hz, 1H), 4.05 (s, 3H); 13C{1H} NMR (126 MHz, CDCl3) δ 165.9, 163.4, 150.3, 145.2, 121.2, 109.8, 55.2.

2-Chloro-6-methoxynicotinoyl Cyanide (22)

To a mixture of CuCN (2.9 g, 32.0 mmol) and CH3CN (24 mL) in a 100mL four-neck round-bottom flask was added 21 (6.0 g, 29.1 mmol). The mixture was heated to 70 °C for 30 min with stirring, followed by cooling to rt. The solvent was exchanged with toluene (30 mL) through repeated concentration using a rotary evaporator and toluene addition. Insoluble materials were filtered out through a Celite pad and the filtrate was concentrated to give a white solid (5.8 g). The product was used in the next reaction without further purification owing to moisture sensitivity. 1H NMR (500 MHz, CDCl3) δ 8.30 (d, J = 8.8 Hz, 1H), 6.86 (d, J = 8.5 Hz, 1H), 4.09 (s, 3H); 13C{1H} NMR (126 MHz, CDCl3) δ 168.7, 163.7, 151.8, 144.1, 121.2, 113.0, 110.6, 55.5.

Ethyl 2-(2-Chloro-6-methoxypyridin-3-yl)-2-oxoacetate (25)

A 1 L four-neck round-bottom flask was charged with 24 (120.0 g, 539.4 mmol) and dry THF (240 mL), and the resulting solution was cooled to 12 °C. A THF solution of isopropylmagnesium chloride (269.7 mL, 2 M, 1.2 equiv) was added dropwise over 40 min while keeping the reaction stirred at rt for 2 h. A separate 2 L round-bottom flask was charged with diethyl oxalate (87.6 mL, 1.2 equiv) and dry THF (240 mL) and cooled to −8 °C. The arenemagnesium solution prepared as mentioned above was added dropwise to the diethyl oxalate solution over 75 min while keeping the reaction temperature below 1 °C. The reaction was stirred at 0–3 °C for 1 h and quenched by adding 1 M aq HCl (600 mL). After stirring at rt for 10 min, the organic layer was separated. The solvent was exchanged with EtOH through repeated concentration using a rotary evaporator and EtOH addition. The net solution volume was adjusted to 480 mL by adding EtOH, and the solution was seeded with 25 and stirred at rt for 20 min to give a suspension. H2O (480 mL) was added over 1 h, and the resulting suspension was stirred at rt for 15 h. The solids were collected by filtration, washed with 1:2 EtOH/H2O (450 mL), and dried in a vacuum oven at 40 °C for 5 h to afford 25 (107.8 g) as a pale purple solid. The purity was determined to be 94.0 wt % by HPLC assay; 77% yield (corrected according to wt % purity). An analytically pure sample was prepared by recrystallization from EtOH. 1H NMR (500 MHz, CDCl3) δ 8.07 (d, J = 8.5 Hz, 1H), 6.80 (d, J = 8.5 Hz, 1H), 4.43 (q, J = 7.1 Hz, 2H), 1.42 (t, J = 7.1 Hz, 3H); 13C{1H} NMR (126 MHz, CDCl3) δ 184.8, 166.1, 163.6, 149.7, 142.6, 122.3, 110.4, 62.8, 55.0, 13.9; HRMS m/z [M + H]+ calcd. for C10H10ClNO4 244.0350, found 244.0371.

2-(2-Chloro-6-methoxypyridin-3-yl)-2-oxoacetamide (23)

(From 22) To a mixture of H2SO4 (51.0 mL), NaBr (523.4 mg, 5.1 mmol), and Ac2O (4.8 mL, 50.9 mmol) in a 200 mL four-neck round-bottom flask was added 22 (10.0 g, 50.9 mmol) at rt. The mixture was stirred at rt for 2 h and then poured into 8 M aq NaOH (144 mL) with crushed ice. The precipitate was filtered and washed with 1 M aq NaOH. 1 M HCl (30 mL) was then added to the mixture to generate a precipitate, which was collected by filtration, washed sequentially with 5% aq NaHCO3 (40 mL) and H2O (20 mL), and dried in a vacuum oven at 50 °C to give 23 as a colorless solid (8.3 g); 76% yield for three steps from 20.

(From 25) A 2 L four-neck round-bottom flask equipped with a mechanical stirrer was charged with 25 (100.0 g, 94.0 wt %, 385.8 mmol) and 2 M NH3 in EtOH (600 mL). The reaction initially became a homogeneous solution and then turned into a thick slurry after stirring at rt for 5 min. The resulting slurry was stirred at rt for a total of 24 h. The solids were collected by filtration, washed with EtOH (200 mL), and dried in a vacuum oven at 40 °C for 2 h to give 23 as a colorless solid (77.6 g); 94% yield. 1H NMR (500 MHz, DMSO-d6) δ 8.37 (bs, 1H), 8.13 (d, J = 8.5 Hz, 1H), 8.02 (bs, 1H), 7.00 (d, J = 8.5 Hz, 1H), 3.96 (s, 3H); 13C{1H} NMR (126 MHz, DMSO-d6) δ 226.3, 202.7, 202.2, 184.7, 181.1, 160.2, 147.1, 92.2; HRMS m/z [M + H]+ calcd. for C8H7ClN2O3 215.0202, found 215.0218.

2-(6-Methoxy-2-vinylpyridin-3-yl)-2-oxoacetamide (19)

Method A (Suzuki–Miyaura Coupling)

A 1 L four-neck round-bottom flask was charged with 23 (30.0 g, 139.8 mmol), potassium trifluorovinylborate (22.5 g, 1.2 equiv), Pd(dppf)Cl2 (0.7 g, 0.7 mol %), BHT (0.9 g), 1-propanol (150 mL), and N,N-diisopropylethylamine (60.9 mL, 2.5 equiv). The flask was evacuated and refilled with nitrogen five times. The resulting mixture was heated to 95 °C for 3.5 h. After cooling to 55 °C and diluting with THF (300 mL), the reaction mixture was stirred at 55 °C for 15 min and filtered to remove the insoluble materials, which were rinsed with warm THF (75 mL). The filtrate and washings were combined and concentrated to 169 g using a rotary evaporator. The residue was diluted with EtOH (150 mL) and concentrated to ∼150 g using a rotary evaporator, which was repeated a total of three times. The resulting slurry was chilled to 5 °C for 1 h with stirring. The solids were collected by filtration, washed with cold EtOH (60 mL), and dried in a vacuum oven at 45 °C for 2 h to afford 19 (22.3 g) as a pale yellow solid. The purity was determined to be 96.2 wt % by HPLC assay; 74% yield (corrected according to wt % purity). An analytically pure sample was prepared by further vacuum drying. 1H NMR (500 MHz, DMSO-d6) δ 8.32 (brs, 1H), 8.03 (d, J = 8.7 Hz, 1H), 7.99 (bs, 1H), 7.39 (dd, J = 10.6, 16.7 Hz, 1H), 6.87 (d, J = 8.7 Hz, 1H), 6.53 (dd, J = 2.2, 16.7 Hz, 1H), 5.65 (dd, J = 2.2, 10.6 Hz, 1H), 3.99 (s, 3H); 13C{1H} NMR (126 MHz, DMSO-d6) δ 191.2, 166.5, 164.2, 153. 9, 142.6, 132.9, 122.6, 120.9, 109.6, 53.5; HRMS m/z [M + H]+ calcd. for C10H10N2O3 207.0758, found 207.0764.

Method B (Heck Reaction)

A 120 mL autoclave vessel was charged with 23 (200.0 mg, 0.93 mmol), PdCl2 (8.3 mg, 5.0 mol %), DPEphos (25.0 mg, 5.0 mol %), LiCl (3.9 mg, 10.0 mol %), phenothiazine (10.0 mg), dry DMF (4.0 mL), and triethylamine (390.0 μL, 3.0 equiv). The resulting mixture was stirred at 80 °C under ethylene pressure (1.0 MPa) for 16 h. The reaction was allowed to cool to rt and the resulting mixture was purified by silica gel chromatography (20% EtOAc/hexane) to afford 19 (161.7 mg) as a pale yellow solid; 84% yield. The product was also isolated as crystals from the crude mixture using the same operation as in method A, affording an 80% yield.

2-Methoxy-7,8-dihydro-1,6-naphthyridine-5-carboxamide (17)

A 120 mL autoclave vessel was charged with 19 (2.0 g, 9.7 mmol), BHT (80 mg), and dry MeOH (80 mL). The resulting mixture was stirred at rt under NH3 pressure (0.30 MPa) for 2 h. The vessel was closed and heated to 60 °C (bath temperature) for 6 h. The pressure gauge indicated 0.65 MPa. The reaction was allowed to cool to rt and concentrated to 25 g using a rotary evaporator. The assay yield of the reaction solution was determined to be 79% by HPLC. The mixture was diluted with 2-propanol (20 mL) and concentrated to 25 g, which was repeated a total of four times. The resulting slurry was aged at rt for 1 h. The solids were collected by filtration, washed with 2-propanol (8 mL), and suction-dried at rt for 30 min to give 17 (1.24 g) as an off-white solid. 1H NMR indicated the presence of 26 (∼4 mol %); 60% yield (excluding impurity 26). 1H NMR (500 MHz, DMSO-d6) δ 8.20 (d, J = 8.7 Hz, 1H), 7.85 (bs, 1H), 7.51 (bs, 1H), 6.75 (d, J = 8.7 Hz, 1H), 3.91 (s, 3H), 3.83 (t, J = 7.9 Hz, 2H), 2.77 (t, J = 7.9 Hz, 2H); 13C{1H} NMR (126 MHz, DMSO-d6) δ 166.6, 164.3, 158.6, 157.6, 138.1, 115.8, 108.2, 53.5, 46.6, 27.8; HRMS m/z [M + H]+ calcd. for C10H11N3O2 206.0948, found 206.0924.

The use of a commercially available NH3 solution in MeOH instead of NH3 gas gave comparable results.

(R)-tert-Butyl-5-carbamoyl-2-methoxy-7,8-dihydro-1,6-naphthyridine-6-(5H)-carboxylate (31)

A 100 mL four-neck round-bottom flask was charged with 17 (2.0 g, 9.8 mmol), ammonium formate (1.8 g, 29.2 mmol), chloro(p-cymene)[(R,R)-N-(isobutanesulfonyl)-1,2-diphenylethylenediamine]ruthenium(II) (58.7 mg, 0.098 mmol), and CH3CN (50 mL). After stirring at 35 °C for 24 h under continuous N2 flow to remove CO2, 1 M aq citric acid (24 mL) and toluene (12 mL) were added at rt. The organic layer was separated and extracted with 1 M aq citric acid (12 mL) twice. The aqueous layers were combined and washed with toluene (12 mL), followed by the addition of K2CO3 (14.5 g). (Boc)2O (2.34 g, 10.7 mmol) in toluene (2 mL) was then added dropwise at rt, and the resulting mixture was stirred at rt for 1 h. The aqueous layer was separated and extracted with toluene (12 mL) twice, and the combined organic layer was washed with water (6 mL). The solvent was exchanged with MeOH (10 mL) through repeated concentration using a rotary evaporator and MeOH addition. The product was then precipitated by adding water (4 mL), followed by the slow addition of more water (6 mL). After stirring the slurry at 20 °C, the resulting precipitate was collected by filtration, washed with a mixture of MeOH (1.3 mL) and water (2.7 mL), and dried in a vacuum oven at 50 °C to give 31 (2.16 g) as a white solid; 72% yield, 98.9% ee. Although NMR spectra in CDCl3 showed complex patterns due to the presence of rotamers, the peaks were simplified when using DMSO-d6 as the solvent. 1H NMR (500 MHz, DMSO-d6) δ 7.74–7.84 (m, 1H), 7.71 (bs, 1H), 7.15 (br d, 1H), 6.70 (d, J = 8.5 Hz, 1H), 5.29 (s, 0.5H), 5.17 (s, 0.5H), 3.82 (s, 3H), 3.73–3.81 (m, 2H), 2.84–2.95 (m, 1H), 2.71–2.82 (m, 1H), 1.43 (br d, 9H); 13C{1H} NMR (126 MHz, DMSO-d6) δ 173.5, 173.1, 162.5, 154.8, 154.6, 152.8, 152.5, 138.8, 138.6, 121.1, 120.7, 108.8, 80.0, 57.7, 56.7, 53.5, 31.7, 28.5; HRMS m/z [M + H]+ calcd. for C15H21N3O4 308.1578, found 308.1605.

(R)-2-Methoxy-5,6,7,8-tetrahydro-1,6-naphthyridine-5-carboxamide (16)

A 30 mL round-bottom flask was charged with 31 (300.0 mg, 0.98 mmol), THF (1.5 mL), and 6 M HCl (1.2 mL). The mixture was stirred at rt for 6 h and basified by adding 4 M aq NaOH (2.4 mL). The aqueous layer was separated and extracted with THF (1.5 mL). The combined organic layer was concentrated using a rotary evaporator to give 16 as a colorless solid (181.0 mg); 89% yield. If necessary, the product could be further purified, including an improved ee, by recrystallization from EtOAc. 1H NMR (500 MHz, DMSO-d6) δ 7.59 (d, J = 8.5 Hz, 1H), 7.52 (bs, 1H), 7.19 (bs, 1H), 6.60 (d, J = 8.5 Hz, 1H), 4.29 (s, 1H), 3.79 (s, 3H), 3.03–3.12 (m, 1H), 2.83–2.97 (m, 2H), 2.58–2.75 (m, 2H); 13C{1H} NMR (126 MHz, DMSO-d6) δ 174.8, 162.0, 152.9, 138.7, 123.2, 108.1, 58.6, 53.3, 41.0, 32.6; HRMS m/z [M + H]+ calcd. for C10H13N3O2 208.1062, found 208.1081.

(R)-tert-Butyl-5-((7-fluoro-1,1-dimethyl-2,3-dihydro-1H-inden-5-yl)carbamoyl)-2-methoxy-7,8-dihydro-1,6-naphthyridine-6-(5H)-carboxylate ((R)-11)

A 30 mL Schlenk tube was charged with 31 (2.0 g, 6.5 mmol), K2CO3 (2.0 g, 14.1 mmol), CuI (1.4 g, 7.2 mmol), toluene (8.0 mL), 15a (2.2 g, 7.7 mmol), and N,N′-dimethylethylenediamine (1.6 mL, 14.5 mmol). The vessel was evacuated and refilled with argon five times, and the mixture was then stirred at rt for 30 h, followed by the addition of 25% aq NH3 (50 mL) and EtOAc (16 mL). The organic layer was separated, washed with saturated NH4Cl aq (16 mL) repeatedly until the blue color disappeared, and then rinsed with H2O (16 mL). The organic solvent was exchanged with EtOH (2 mL) through repeated concentration using a rotary evaporator and EtOH addition. AcOH (2 mL), H2O (2 mL), and a crystal seed of (R)-11 were added to form a seed bed. After slow addition of H2O (9 mL), the resulting precipitate was collected by filtration, washed with H2O (10 mL), and dried in a vacuum oven at 50 °C to give (R)-11 as a colorless crystalline solid (1.4 g). 91% yield. 1H NMR (500 MHz, CDCl3) δ 8.96 (bs, 1H), 7.52 (m, 1H), 7.08 (d, J = 11.6 Hz, 1H), 7.04 (s, 1H), 6.62 (d, J = 8.5 Hz, 1H), 5.62 (bs, 1H), 3.95–4.10 (m, 1H), 3.91 (s, 3H), 3.55 (bs, 1H), 2.87–2.99 (m, 2H), 2.85 (t, J = 7.3 Hz, 2H), 1.89 (t, J = 7.4 Hz, 2H), 1.53 (s, 9H), 1.33 (s, 6H); 13C{1H} NMR (126 MHz, CDCl3) δ 168.9, 163.1, 160.4, 158.4, 152.2, 146.9, 146.8, 138.8, 137.8, 133.4, 118.8, 111.6, 108.8, 105.8, 105.6, 81.8, 53.5, 44.4, 41.9, 40.5, 31.5, 31.1, 28.4, 27.5; HRMS m/z [M + H]+ calcd. for C26H32FN3O4 470.2434, found 470.2450.

5-Iodo-7-fluoro-1,1-dimethyl-2,3-dihydro-1H-indene (15a)

A 30 mL Schlenk tube was charged with CuI (95.2 mg, 0.5 mmol), NaI (3.0 g, 20.0 mmol), 1,4-dioxane (10 mL), N,N′-dimethylethylenediamine (107.5 μL, 1.0 mmol), and 15b (2.4 g, 10.0 mmol). The mixture was stirred overnight under reflux. The reaction was cooled to rt and filtered through a Celite pad. The filter cake was washed with EtOAc (20 mL), and the combined solution was washed successively with 10% aq NH3 (10 mL, twice), 20% aq citric acid (10 mL), and H2O (10 mL). The organic solvent was then removed using a rotary evaporator, and the solution was azeotropically dried with EtOH, affording the target product as a yellow oil (2.7 g); 94% yield. An analytically pure sample was prepared by distillation (106 °C, 5.0 mmHg). 1H NMR (500 MHz, CDCl3) δ 7.29 (s, 1H), 7.16 (d, J = 9.1 Hz, 1H), 2.89 (t, J = 7.3 Hz, 2H), 1.91 (t, J = 7.4 Hz, 2H), 1.34 (s, 6H); 13C{1H} NMR (126 MHz, CDCl3) δ 160.5, 158.5, 148.5 (d, J = 7.3 Hz), 137.6 (d, J = 15.4 Hz), 129.6 (d, J = 3.6 Hz), 122.8 (d, J = 23.6 Hz), 90.6 (d, J = 7.3 Hz), 44.6 (d, J = 1.8 Hz), 41.7, 30.7, 27.3 (d, J = 1.8 Hz); Anal. calcd. for C11H12FI: C, 45.54; H, 4.17; found: C, 45.16; H, 3.96.