Erythrinane skeleton synthesis

A completely different route devised by Prelog and co-workers (14) not only afforded a new synthesis of the erythrinane skeleton but also achieved a method of introducing an oxygen function at C-3, the site of the aliphatic methoxyl in the alkaloids. The synthesis is outlined in Fig. 5. The dihydroisoquinolinium salt XXXIV was prepared by Bischler-Napieralski ring closure of the lactam XXXIII. Hydrolysis of the vinyl chloride of XXXIV gave the methyl ketone XXXV. When this salt was made alkaline, addition of the carbanion from the acidic methyl to the C=N double bond created the spiro link. Sodium borohydride reduction of XXXVI gave a mixture of epimeric alcohols. One of them had an IR-spectrum identical with that of a transformation product (XXXVII) of erysonine (If) and on resolution w ith tartaric acid its (— )-enantiomer proved to be identical with XXXVII. 

Two minor by-products formed in the phosphoric acid catalyzed cyclization of LXI have different carbon skeletons than the main products. One is the hydroxylactam LXV, shown to arise by acid-catalyzed rearrangement of LXIII. In acid solution it exists as the colored imonium ion LXVI 33). The second by-product is LXVII. Its formation is explained by the generation of two different cations from LXII in acid, one leading to the erythrinane skeleton (LXIV), the other to the isomeric apo skeleton of LXVII. Bromination of LXI with X-bromosuccinimide afforded the bromo derivative VIII, whose cyclization led almost exclusively to the apo skeleton and a synthesis of apoerysopine (see Fig. 2). 

When these approaches proved unsuccessful, a total synthesis of erysotrine was finally achieved beginning with the oxalyl derivative (LIT) of 4-methoxycyclohexanone. The synthetic scheme has so far been reported only in preliminary communications 27, 36). The condensation with -(3,4-dimethoxyphenyl)-ethylamine leading to the tetracyclic erythrinane skeleton and the further conversion to LV have been discussed in Section III, C (Fig. 8) the remainder of the synthesis is outlined in Fig. 10. 

The heart of these biosynthetic proposals is the oxidation of LXXXVIII to XCI, and the plausibility of this conversion has received dramatic support in two laboratories. In vitro oxidation of LXXXVIII (Ri = R2 = CH3) with alkaline ferricyanide was found by both Scott and coworkers 40) and by Mondon and Ehrhardt 40a) to afford XCI (Ri = R2 = CH3) in 35% yield. Mondon and Ehrhardt were then able to convert XCI to ( + )-dihydroerysodine, using the reactions shown in Fig. 12. This synthesis incidentally confirms the substitution pattern in ring D of erysodine. The facile formation of the erythrinane skeleton in this manner supports the basic biogenetic scheme, and the results of incorporation experiments with isotopically labelled LXXXVIII and XCI will be awaited with great interest. 

The greater electrophilicity of the A -acyliminium ion as compared with an ordinary iminium ion was nicely illustrated as early as 1957. In experiments directed at the total synthesis of erythrina alkaloids, cyclization of iminium ion (60) to the erythrinane skeleton (61) fails (equation 34). However, A(-acylim-inium ions (62) and (63) can both be converted into the desired skeleton in good yields. A recent il-... 

Besides this widespread pathway based on the intermediate N-acyliminium ion, several other established methods have been applied to construct the C5/C13 bond of the erythrinane skeleton. Thus, the Heck reaction has proved to be an attractive approach to the target compounds. The synthesis reported by Rigby (Scheme 10) starts with a smooth [1 -I- 4] cycloaddition of certain isocyanides to the vinyl isocyanate 79 affording the required hydroindolone 80. Then the iodoarene moiety has been installed by AT-alkylation with the phenethylmesylat 81 giving the N-alkylated precursor 82. Cyclization of 82 under Heck conditions yields the expected 7,8-dioxoerythrinane 83 as a single diastereomer, which then has been converted to ( )-2-cpi-erythrinitol (84) in twelve additional steps (64). 

Wasserman HH, Amici RM (1989) The Chemistry of Vicinal Tricarbonyls. A Total Synthesis of ( )-3-Demethoxyerythratidinone. J Oig Chem 54 5843 Padwa A, Hennig R, Kappe CO, Reger TS (1998) A Triple Cascade Sequence as a Strategy for the Construction of the Erythrinane Skeleton. J Oig Chem 63 1144, and lit cited therein... 

Ishibashi H, Harada S, Sato K, Ikeda M, Akai S, Tamura Y (1985) Synthesis of the Erythrinan Skeleton by Acid-Promoted Cyclization of iV-(3-Oxo-l-cyclohexen-l-yl)-iV-[2-(3,4-dimethoxyphenyl)ethyl]-Q -(methylsulfinyl)acetamide. Chem Pharm Bull 33 5278... 

Westling, M., Smith, R. and Livinghouse, T. 1986. A convergent approach to heterocycle synthesis via silver ion mediated a-ketoimidoyl hahde-arene cychzations. An application to the synthesis of the erythrinane skeleton. J. Org. Chem. 51 1159-1165. 

These results led to the postulation of the spirocyclic erythrinane skeleton and this was confirmed by synthesis of the parent nucleus by Belleau (1953). The structures of the two lactonic alkaloids (3) were also elucidated by Boekelheide (1960) and coworkers, who recognised their close structural identity to the other aromatic erythrina alkaloids (1) the lactonic alkaloids also underwent an apo -type rearrangement under drastic acidic conditions. Subsequently the structures of both the aromatic and lactonic alkaloids were confirmed by X-ray crystallography of erythra-line hydrobromide (Nowacki and Bonsma 1958) and of the erythroidines (Hanson 1963). The spiro centre was shown to have the same configuration in all the erythrina alkaloids by use of optical rotatory measurements (Weiss and Ziffer 1963, Beecham 1971). 

---