AN EFFICIENT SYNTHESIS OF TRICYCLIC COMPOUNDS, (±)-(4aβ,8aβ,10aα)-1,2,3,4,4a,6,7,8,8a,9,10,10a-DODECAHYDRO-1,1,4a-TRIMETHYL-2-OXOPHENANTHRENE-8a-CARBOXYLIC ACID, ITS METHYL ESTER, AND (±)-(4aβ,8aβ,10aα)-3,4,4a,6,7,8,8a,9,10,10a-DECAHYDRO-8a-HYDROXYMETHYL-1,1,4a-TRIMETHYLPHENANTHREN-2(1H)-ONE

Abstract

Our ongoing efforts for the improvement of anti-inflammatory and antiproliferative activity of oleanolic acid analogues led us to discover 2-cyano-3,12-dioxooleana-1,9(11)-dien-28-oic acid (CDDO, 1) and related compounds.1 In connection with these investigations, we have found that tricyclic compounds with similar enone functionalities in rings A and C are also a novel class of highly active inhibitors of nitric oxide (NO) production in mouse macrophages.2 In particular, bis-cyano enone (±)-2 is orally active in a preliminary in vivo inflammation model.2 In addition, we have found that (+)-2 having the opposite configuration to that of CDDO shows 10 times higher inhibitory activity than (-)-2 on NO production in mouse macrophages.3 These results encouraged us to design and synthesize analogues of 2. Thus, we focused our attention on the modifications of the C-8a position, because some biologically active natural products have functionalities at the same position (e.g., anti-tumor quassinoids4). For our projected synthesis of C-8a functionalized TBE compounds, the simple tricycles 3-5 are potentially very desirable intermediates. We envisioned preparing 3-5 from the known acid 65,6 by standard reductive methylation.7 However, attempts to reductively methylate acid 6 with 5-7 equivalents of lithium in liquid ammonia containing no proton donor, followed by esterification with diazomethane gave 4 in 30% yield (average of 7 experiments, the yield fluctuates) along with many by-products. These by-products caused serious difficulty for the purification of 4. An attempt with one equivalent of tert-butanol gave similar results as without a proton donor. Attempts to reductively methylate methyl ester 7,6 which is prepared from 6 with diazomethane, using 10 equivalents of lithium in liquid ammonia containing no proton donor gave the desired compounds 3-5 in low yield along with several by-products including enones 6 and 8. After much experimentation, we have found that the addition of one equivalent of water dramatically improves this reductive methylation reaction. Thus, the reductive methylation of 7 using 7.2 equivalents of lithium and one equivalent of water followed by quenching the excess lithium with isoprene, and then methyl iodide at -78 °C cleanly produced 3-5 in 38%, 15%, and 36% yields (total 89%), respectively. The yields are reproducible and we have prepared 3-5 several times by this procedure. These compounds can be easily separated by extracting the acid with aqueous base, followed by column chromatography (see Experimental Section). Also, they were easily converted to a single compound. For example, oxidation (e.g., Jones reagent and RuO2-NaIO4 etc.) of alcohol 5 gave acid 3, and both acid 3 and methyl ester 4 were converted to alcohol 5 in three steps (ketalization, reduction with LiAlH4, and deketalization). Acid 3 may be an important intermediate for the synthesis of abietane and totarane diterpenoids.