Hetisine alkaloids

The genera of Aconitum (commonly known as Monkshood) and Delphinium, and to a lesser extent Rumex, Consolida, and Spiraea, have long been recognized as a rich source of alkaloid natural products [1], The diterpenoid alkaloids are generally classified into two major groups the Ci9-diterpenoid alkaloids (sometimes referred to as the Cig-norditerpenoid alkaloids) and the C2o-diterpenoid alkaloids. Within the C2o-diterpenoid alkaloids, at least 11 separate classes have been isolated, including the hetisine alkaloids (Chart 1.1). 

Among the first hetisine alkaloids isolated were nominine (1) [2], kobusine (2) [3], pseudokobusine (3) [4], hetisine (4) [5], and ignavine (5) [6] in the 1940s and 1950s (Chart 1.2). Since these early isolations, over 100 distinct hetisine alkaloids have... 

O-acyl derivatives thereof have shown potent vasodilatory activity [8]. A number of other hetisine alkaloids have shown diverse biological activities. These include nominine [9] (1) (local anesthetic, anti-inflammatory, and antiarrhythmic), hetisine [9a] (4) (hypotensive), ignavine [10] (5) (analgesic, anti-inflammatory, antipyretic, sedative, antidiuretic), zeravshanisine [9a] (8) (antiarrhythmic and local anesthetic), and tadzhaconine [9a, 11], (9) (antiarrhythmic) (Chart 1.3). 

The C2o-diterpene alkaloids have long served as classic targets within the field of natural product synthesis [14], Total syntheses of four C2o-diterpene alkaloids have thus far been reported atisine [15], veatchine [16], garryine [17], and napelline [18]. In spite of this progress, synthetic efforts toward the hetisine alkaloids have been relatively sparse. Prior to our work in the area, these efforts include one total synthesis and five synthetic studies. 

Scheme 1.2 Total synthesis of nominine, first total synthesis of a hetisine alkaloid...

In 1975, van der Baan and Bickelhaupt reported the synthesis of imide 37 from pyridone 34 as an approach to the hetisine alkaloids, using an intramolecular alkylation as the key step (Scheme 1.3) [23]. Beginning with pyridone 34, alkylation with sodium hydride/allyl bromide followed by a thermal [3,3] Claisen rearrangement gave alkene 35. Next, formation of the bromohydrin with A -bi omosuccinimide and subsequent protection of the resulting alcohol as the tetrahydropyranyl (THP) ether produced bromide 36, which was then cyclized in an intramolecular fashion to give tricylic 37. 

More recently in 2001, Winkler and Kwak reported methodology designed to access the pyrrolidine core of the hetisine alkaloids via a photochemical [2+2], retro-Mannich, Mannich sequence (Scheme 1.3) [26]. In a representative example of the methodology, vinylogous amide 42 was photo-irradiated to give the [2+2] cycloaddition product 43. Heating cyclobutane 43 in ethanol provided enamine 44 via a retro-Mannich reaction. Exposure of enamine 44 to acidic conditions then effected a Mannich reaction, resulting in pyrrolidine 45. 

In 2003, Williams and Mander reported a method designed to access the hetisine alkaloids (Scheme 1.3) [27]. This approach, based upon a previously disclosed strategy by Shimizu et al. [28], relied on arylation of a bridgehead carbon via a carbocation intermediate in the key step. Beginning with (1-keto ester 46, double Mannich reaction provided piperidine 47. Following a straightforward sequence, piperidine 47 was transformed to the pivotal bromide intermediate 48. In the key step, bromide 48 was treated with silver (I) 2,4,6-trinitrobenzenesulfonate in nitro-methane (optimized conditions) to provide 49 as the most advanced intermediate of the study, in 54 % yield. 

To investigate the feasibility of employing 3-oxidopyridinium betaines as stabilized 1,3-dipoles in an intramolecular dipolar cycloaddition to construct the hetisine alkaloid core (Scheme 1.8, 77 78), a series of model cycloaddition substrates were prepared. In the first (Scheme 1.9a), an ene-nitrile substrate (i.e., 83) was selected as an activated dipolarophile functionality. Nitrile 66 was subjected to reduction with DIBAL-H, affording aldehyde 79 in 79 % yield. This was followed by reductive amination of aldehyde x with furfurylamine (80) to afford the furan amine 81 in 80 % yield. The ene-nitrile was then readily accessed via palladium-catalyzed cyanation of the enol triflate with KCN, 18-crown-6, and Pd(PPh3)4 in refluxing benzene to provide ene-nitrile 82 in 75 % yield. Finally, bromine-mediated aza-Achmatowicz reaction [44] of 82 then delivered oxidopyridinium betaine 83 in 65 % yield. 

Clearly, the nitroalkene dipolarophile oxidoisoquinolinium betaine 123 is nonideal for the synthesis of the hetisine alkaloids, as mass throughput for the needed cycloadduct would be low, and conversion of the tertiary nitro group to carbon-based functionality, as would be required in the latter stages of the synthesis, could be problematic. On the other hand, an ene-nitrile dipolarophile has several potential advantages over nitroalkene dipolarophile. Most importantly, the ene-nitrile cycloadduct has carbon functionality installed at the C-10 position. Second, the conjugate addition byproduct pathway that occurs so readily for the nitroalkene oxidoisoquinolinium betaine 123 system (see Scheme 1.13) should be much slower... 

The photocycloaddition-retro-Mannich-Mannich methodology is featured in a concise synthesis of mesembrine. Irradiation of vinylogous amide 114 effects photocycloaddition-re/ro-Maimich sequence to give product 116 via the cyclobutane intermediate 115. Methylation with trimethyloxonium tetrafluoroborate followed by treatment with DMAP produces mesembrine in 84% yield. Other applications include construction of the bicyclic core of peduncularine and synthetic approaches to hetisine alkaloids and 8-substituted 6-azabicyclo[3.2.1]octan-3-ones. ... 

---