Synthesis of O- and N-alkylated products of 1,2,3,4-tetrahydrobenzo[c][2,7]naphthyrin-5(6H)-one

Abstract

Efficient syntheses of O- and N-alkylated products of 1,2,3,4-tetrahydrobenzo[c][2,7]naphthyrin-5(6H)-one are presented. The O-alkylated analogues were synthesized through a reduction-cyclization cascade and a selective O-alkylation reaction; whereas the N-alkylated analogues were obtained through a key Buchwald coupling.

Graphical abstract

Tetrahedron

Muscarinic acetylcholine receptors (mAChRs), consisting of five distinct subtypes (denoted M₁ to M₅), are important drug targets with various potential clinical applications. ^1-3 These G-protein coupled receptors are involved in many critical physiological functions, with each subtype having distinct physiological roles.^4,5 Thus, developing subtype-selective ligands for these receptors has been a major focus in drug discovery.¹ Specifically, the M₅ mAChR has been suggested to be a potential therapeutic target for a number of diseases,^1-3 including cerebrovascular insufficiency,^6,7 Alzheimer's disease and cognitive impairment,⁷ schizophrenia,⁸ and drug dependence.⁹ Recent studies from our laboratory, employing a scaffold of 4-phenyl-1,2,5,6-tetrahydropyridine-3-carboxylic acid, have led to the discovery of a series of M₅-preferring orthosteric antagonists, e.g. 1 and 2 (Figure 1).¹⁰ One important strategy of lead optimization is to reduce conformational flexibility. Compounds 3 and 4 were designed to mimic the cis- and trans-rotamers, respectively, of 4-phenyl-1,2,5,6-tetrahydropyridine-3-carboxylates in which the rotatable bond between the carboxylate moiety and C3 of tetrahydropyridine ring is locked by forming a ring structure (Figure 2). In the present study, we investigate the synthesis of 3 and 4.

Figure 1.

Structures of compounds 1 and 2.

Figure 2.

General structures of the target analogues.

For the synthesis of 3 and 4 with variations on the O- or N-alkyl group, an efficient route would incorporate the alkyl groups on a common intermediate 1,2,3,4-tetrahydrobenzo[c][2,7]naphthyrin-5(6H)-one (5) at the last step (Figure 3). This strategy is based on the assumption that selective alkylation of either the O or N atom of 5 can be achieved. We planned to synthesize intermediate 6 through the triflate intermediate (7), using a well-established Suzuki coupling reaction employed in our previous work.¹⁰ Conversion of 6 to 5 could be achieved via a reduction-cyclization cascade (Figure 3).

Figure 3.

Retrosynthetic plan.

The preparation of 5 (R = Me) was relatively straightforward. Compound 7 was prepared from 8 (Figure 3) as reported.¹⁰ Suzuki coupling of 7 with 2-nitrophenylboronic acid in refluxing THF in the presence of Pd(PPh₃)₄ afforded compound 6 in 90% yield. Treatment of 6 with TFA in DCM at room temperature, followed by N-methylation using formalin and NaCNBH₃ in ethanol at 0 °C afforded compound 9 in 66% yield. Finally, a reduction-cyclization cascade was achieved by treating 9 with Na₂S₂O₆ and K₂CO₃ in ethanol and water (1:1 ratio) at room temperature to afford 5 in 69%) yield. Alternatively, the same reduction-cyclization conditions were applied to compound 6 to form the tricyclic compound 10 in 74% yield. However, attempts to remove the Boc group from 10 resulted in complex mixtures (Scheme 1).

Scheme 1.

Synthesis of intermediate 5.

We then carried out a limited study to determine if regioselective N- versus O-alkylation of 5 could be achieved. Alkylation of 2-pyridones¹¹ and 2(1H)-quinolinones¹² usually results in formation of both N- and O- alkylated products. It was reported that alkylation of alkali salt of 2-pyridone in DMF occurred predominately at N atom whereas using the silver salt in benzene afforded exclusively the O-alkylated product.¹³ However, treatment of 5 with various bases, including NaH in THF or DMF, CaH₂¹⁴ in THF or DMF, LiHMDS in THF, nBuLi in THF, and K₂CO₃ in DMF, followed by reacting with 3,4-dimethoxyphenethyl bromide, afforded exclusively the O-alkylated product 3a (Scheme 2). Among the reaction conditions studied, K₂CO₃ in DMF at 80 °C gave the best yields (75-82%, Scheme 2).

Scheme 2.

Synthesis of O-alkylated analogues 3a-3d.

The selectivity between N- and O-alkylation can be affected by a number of factors.^13,15 Under the same reaction conditions, selectivity is mainly affected by the structure features in the ambident nucleophiles and the alkylating agents. While the high regioselective alkylation of 5 is unexpected, a combination of the steric effect from the relatively large phenylalkylbromide and the electronic effect from the basic amino group in 5 that favors the phenolic formation could result in the observed exclusive O-alkylation.

Since direct alkylation of 5 is overwhelmingly favorable towards O-alkylation, an alternative synthetic route was designed to obtain N-alkylated analogues 4. As depicted in Scheme 3, Suzuki coupling of triflate 7 with 2-bromophenylboronic acid yielded compound 12 in 80% yield. Compound 12 was subjected to ester hydrolysis in the presence of NaOH to give the carboxylic acid 13. Compound 13 was subsequently coupled with various amines in the presence of EDCI to form the corresponding amide 14 in 65-75% yield. Intramolecular Buchwald coupling of 14 in the presence of Pd(OAc)₂, BINAP and Cs₂CO₃ in dry toluene at 90 °C for 12 h provided the cyclization product 15 in 78-95%) yield.¹⁶ Removal of the Boc group in 15 by TFA afforded 16 (87-92% yield), which was then treated with formalin and NaCNBH₃ in ethanol at 0 °C for 1 h to generate the desired product 4 in 71-90% yield.

Scheme 3.

Synthesis of N-alkylated analogues 4a-4e.

Structural assignments for O-alkylated and N-alkylated isomers were made by 1-D and 2-D NMR experiments.¹⁷ The most distinct NMR signals between 3a and 4a ¹⁸ are the ¹³C chemical shifts of C11 and C5 (Figure 4). As observed in the HSQC data, in compound 3a, the ¹H signal of H11 (triplet) at 4.45 ppm exhibited a cross peak to C11 at 75.8 ppm, such a significant downfield shift is typical for a carbon that connects to an oxygen atom. On the other hand, for compound 4a, the ¹³C chemical shift of C11 appeared at 46.1 ppm, which is common for a nitrogen bearing carbon. Long-range carbon–hydrogen couplings observed in the HMBC experiments further supported these assignments. For example, the carbonyl carbon in 4a exhibited a peak at 160.6 ppm, as indicated by the HMBC cross peak between H11 and C5, while the corresponding C5 in 3a was found at 130.4 ppm.

Figure 4.

Characteristic carbon–hydrogen couplings (HSQC and HMBC) observed for 3a and 4a.

In summary, we have developed efficient methods for the synthesis of O- and N-alkylated products of 1,2,3,4-tetrahydrobenzo[c][2,7]naphthyrin-5(6H)-one. These compounds represent conformational restricted analogues of 4-phenyl-1,2,5,6-tetrahydropyridine-3-carboxylates. Biological evaluation of these novel analogues is currently underway.