Facile Syntheses of Substituted, Conformationally-Constrained Benzoxazocines and Benzazocines via Sequential Multicomponent Assembly and Cyclization

Abstract

A multicomponent assembly process (MCAP) was utilized to prepare versatile intermediates that are suitably functionalized for subsequent cyclizations via Ullmann and Heck reactions to efficiently construct substituted 2,6-methanobenzo[b][1,5]oxazocines and 1,6-methanobenzo[c]azocines, respectively. The intramolecular Ullmann cyclization was conducted in tandem with an intermolecular arylation that enabled the rapid syntheses of a number of O-functionalized methanobenzoxazocines.

An effective approach for discovering new lead compounds for drug development programs and for identifying molecular probes to study biological systems involves screening of chemical libraries based upon privileged scaffolds.¹ By varying the nature of peripheral substituents on these molecular frameworks, it is often possible to obtain hits across a wide range of biological targets. Toward developing a general approach to the synthesis of heterocyclic scaffolds comprising privileged substructures, we designed a novel strategy for diversity-oriented synthesis (DOS) that featured a multicomponent assembly process (MCAP) involving Mannich-type reactions to give substituted aryl methylamine derivatives.^2–4 These adducts can be subjected to various cyclization reactions that are enabled by selective functional group pairing to construct substituted heterocyclic ring systems. We have demonstrated the utility of this approach by applying it to syntheses of small libraries of diversely substituted benzodiazepines,⁵ norbenzomorphans,⁶ aryl piperidines⁷ and tetrahydroisoquinolines.⁸ We now report the extension of this useful methodology to the facile preparation of compounds having conformationally-constrained benzoxazocines 1 and benzazocines 2 as key structural subunits (Figure 1).

Figure 1.

Generic structures of conformationally-constrained benzoxazocines 1 and benzazocines 2.

Methylene-bridged benzoxazocines and benzazocines have received attention owing to their favorable biological profiles. For example, benzoxazocines related to 5, which were prepared in five steps from 4-methylpyridine (3) (eq 1), exhibit analgesic, hypotensive, and CNS stimulating activities,⁹ whereas benzazocines similar to 8, which have been prepared in a six-step sequence from indene (6) (eq 2), have shown moderate antinociceptive activity.¹⁰ Because routes to both of these scaffolds and their analogs are rare,¹¹ there is an unmet need for the development of flexible approaches for their synthesis that enable facile diversification for biological screening.

Although the routes depicted in eq 1 and 2 do provide access to analogs of 1 and 2, they are somewhat lengthy and not well suited for the facile syntheses of polysubstituted derivatives. Accordingly, we envisioned a unified approach to both scaffolds might be developed using an MCAP involving substituted bromobenzaldehydes 9 to assemble intermediates that could be readily transformed into substituted piperidines 10 and azepines 11 by a ring closing metathesis (RCM) (Scheme 1).^12,13 Stereoselective vic-dihydroxylation of 10 followed by an Ullmann cyclization would lead to compounds of the general structure 1, whereas cyclization of 11 via a Heck reaction followed by reduction would afford methanobenzazocines 2. Since a vast array of substituted bromobenzaldehydes are readily available, it is conceivable that a more diverse collection of compounds could be obtained. For example, we previously demonstrated that our MCAP/cyclization strategy may be applied to the preparation of scaffolds containing aryl chlorides, which can be easily derivatized through various cross-coupling reactions.⁶ We now report the reduction of the plan adumbrated in Scheme 1 to practice as exemplified by the syntheses of derivatives of 1 and 2 that incorporate functional handles that may be exploited for further diversification reactions.

Scheme 1.

General approaches to conformationally-constrained benzoxazocines 1 and benzazocines 2 from 2-bromobenzaldehydes 9.

The first step toward implementing the plan outlined in Scheme 1 involved condensation of o-bromobenzaldehyde (12) with allylamine, followed by treatment with methyl chloroformate and allylzinc bromide to give 13 in 94% yield (Scheme 2). In a similar fashion, 12 was treated sequentially with 1-amino-3-butene, benzyl chloroformate, and allylzinc bromide to furnish 14 in 86% yield. Although the stated yields for carbamates 13 and 14 are for purified materials, it is important from a practical standpoint that they are of sufficient purity to use directly in the next step. When the dienes 13 and 14 were subjected to ring closing metathesis (RCM) using Grubbs 2^nd generation catalyst, the tetrahydropyridine 15 and tetrahydroazepine 16 were produced in excellent overall yields from 12. Because these compounds serve as key intermediates, it is notable that they could be easily prepared on a multigram scale.

Scheme 2.

Multicomponent assembly processes followed by RCM reactions.

With 15 in hand, our next objective was to stereoselectively introduce a hydroxyl group at C(4) of 15 cis to the aryl moiety for the planned etherification reaction. Because hydroxyl groups may be further derivatized by a number of refunctionalizations, we elected to dihydroxylate 15 so that the product benzoxazocine would contain a free hydroxyl group. Accordingly, tetrahydropyridine 15 was subjected to Woodward’s dihydroxylation conditions¹⁴ to form the vicinal diol 17 as a single diastereomer (Scheme 3). The stereochemical outcome of this transformation was presumably dictated by the pseudoaxial orientation of the aryl group that results from A^1,3-strain¹⁵ with the carbamate group. Syn-dihydroxylation from the more hindered face of the olefin, which is consistent with the mechanism of the Woodward dihydroxylation, then gave 17 in 60% yield. We envisioned that the cyclization of 17 by an intramolecular Ullmann reaction would also be facilitated by A^1,3-strain that would favorably position the aryl bromide proximal to the hydroxyl group at C(4). Indeed, when diol 17 was heated with CuI, 3,4,7,8-tetramethyl-1,10-phenanthroline (19) and Cs₂CO₃ in toluene according to the protocol of Buchwald,¹⁶ benzoxazocine 18 was obtained in 87% yield. The intramolecular Ullmann cyclization of 17 represents a novel entry to this bridged tricyclic scaffold.

Scheme 3.

Synthesis of benzoxazocine 18 via Ullmann reaction.

Although the C(3) hydroxyl group of 18 might be employed in bimolecular Ullmann coupling reactions, we were intrigued by the more attractive possibility of accessing 3-O-arylated benzoxazocines directly from diol 17. Specifically, we queried whether it might be feasible to develop a one-pot transformation that would feature both intra- and intermolecular Ullmann reactions. Indeed, we discovered that when 18 was heated with an aryl iodide, base, CuI and 19 in toluene (method A), a tandem double Ullmann reaction ensued to give aryl ethers 20 and 21 in 52% and 48% yields, respectively (Table 1, entries 1 and 2). This procedure was efficacious for aryl iodide coupling partners, but we found that such reactions with halopyridines proceeded to give 22 and 23 in only about 30% yields. This decrease in yield is attributed to competitive intermolecular bis-arylation of diol 17, a process that was observed to a much lesser extent with aryl iodide coupling partners. We soon discovered that the undesired bis-arylation could be avoided if the halopyridine component was introduced after the initial cyclization was complete. Accordingly, diol 17, CuI, 19 and base were heated in toluene until 17 had been consumed (TLC), whereupon the appropriate halopyridine was added (method B). In this manner 22 and 23 were obtained in 53% and 54% yields, respectively (entries 3 and 4).

Table 1.

One-pot synthesis of O-arylated benzoxazocines from diol 17.

entry	ArX	method	compound	yield
1		a	20	52%
2		a	21	48%
3		b	22	53%
4		b	23	54%

The benzoxazocine scaffold 18 was an excellent embarkation point for other O-derivatization processes as is exemplified by the syntheses of 24–27 (Scheme 4). For example, benzylation and acylation of the free OH group in 18 with 28 and 29, respectively, afforded the corresponding ether 24 and ester 25. Propargylation of 18 followed by a Huisgen [3+2] cycloaddition with the aryl azide 30 delivered triazole 26 in 97% overall yield from 18. Finally, an SN_Ar reaction between 18 and 2-chloropyrimidine (31) provided the O-arylated benzoxazocine 27.

Scheme 4.

Synthesis of benzoxazocine derivatives 24–27.

The relative orientation of the substituents on 17 is well suited to the purpose of appending additional fused rings as illustrated by the use of 17 in two different ring forming reactions to deliver novel heterocyclic scaffolds. Bis-O-allylation of 17 gave a diene intermediate that underwent a RCM reaction in the presence of Grubbs 2^nd generation catalyst to form the piperidino-1,4-dioxocine 32 (Scheme 5). Alternatively, when a mixture of 17 and 2,3-dichloropyrazine was treated with NaH in DMF, the fused tricycle 33 was obtained via a double SN_Ar reaction in 51% yield.

Scheme 5.

Fused scaffolds from diol 17.

Having developed a facile entry to several bridged benzoxazocines, we turned our attention to preparing the bridged benzazocine 36. Toward this goal, 16 was subjected to an intramolecular Heck cyclization in the presence of Bu₄NCl^4a,17 to provide a readily separable mixture (1.3:1.0) of benzazocine isomers 34 and 35 in 79% yield. Although attempts to isomerize 35 to the thermodynamic enecarbamate product 34 were unsuccessful,^18,6 the isomers were successfully converged to 36 via complementary reducing conditions. Namely, ionic reduction¹⁹ of 34 employing Et₃SiH and TFA and catalytic hydrogenation of 35 with Adam’s catalyst gave 36 in 59% and 96% yields, respectively (Scheme 6). Constrained benzazocine 36 is well suited for analog synthesis via N-derivatization. Moreover, we have shown that use of chlorinated bromobenzaldehyde inputs in the MCAP leads to benzoxazocines possessing aromatic functional handles, thereby enabling a broad range of diversification reactions.⁶

Scheme 6.

Bridged benzazocine 36 from an intramolecular Heck reaction.

In summary, we have extended our original MCAP/cyclization strategy to generate intermediates that can be quickly elaborated into conformationally constrained benzoxazocines and benzazocines. Our approach to O-arylated benzoxazocines was improved through the development of a one-pot double Ullmann reaction, in which functionalized benzoxazocines were obtained in only four steps from commercially available starting materials. Furthermore, diol 17 proved to be a versatile intermediate that could be diverted towards novel, fused heterocyclic ring systems. Further applications of this and related approaches to the syntheses of unique compound libraries are in progress, and the results of these investigations will be reported in due course.

Figure 2.

Previous syntheses of benzoxazocine 5 and benzazocine 8.