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.

中文翻译：

---

三环化合物的有效合成，(±)-(4aβ,8aβ,10aα)-1,2,3,4,4a,6,7,8,8a,9,10,10a-十二氢-1,1,4a -TRIMETHYL-2-OXOPHENANTHRENE-8a-羧酸，其甲酯，和 (±)-(4aβ,8aβ,10aα)-3,4,4a,6,7,8,8a,9,10,10a-十氢-8a-羟甲基-1,1,4a-TRIMETHYLPPHENANTHREN-2(1H)-ONE

我们将注意力集中在 C-8a 位置的修饰上，因为一些具有生物活性的天然产物在同一位置具有功能（例如，抗肿瘤类 quassinoids4）。对于我们计划合成的 C-8a 官能化 TBE 化合物，简单的三环化合物 3-5 可能是非常理想的中间体。我们设想通过标准还原甲基化从已知的酸 65,6 制备 3-5。7 然而，尝试在不含质子供体的液氨中用 5-7 当量的锂还原甲基化酸 6，然后用重氮甲烷酯化得到 4 in 30% 的产率（7 次实验的平均值，产率波动）以及许多副产品。这些副产物对 4 的纯化造成了严重的困难。尝试使用一当量的叔丁醇得到与不使用质子供体相似的结果。尝试将甲基酯 7,6 用重氮甲烷还原甲基化，使用 10 当量锂在不含质子供体的液氨中得到所需化合物 3-5 以及几种副产物，包括烯酮 6 和 8 . 经过大量实验，我们发现加入一当量的水会显着改善这种还原性甲基化反应。因此，使用 7.2 当量锂和 1 当量水对 7 进行还原甲基化，然后用异戊二烯淬灭过量的锂，然后在 -78 °C 下使用碘甲烷清洁地产生 3-5，分别为 38%、15% 和 36%产率（总计 89%）。产量是可重复的，我们已经通过这个程序准备了 3-5 次。这些化合物可以很容易地通过用碱水溶液萃取酸来分离，然后是柱层析（见实验部分）。此外，它们很容易转化为单一化合物。例如，醇5的氧化（例如，琼斯试剂和RuO2-NaIO4等）得到酸3，并且酸3和甲酯4通过三个步骤（缩酮化、用LiAlH4还原和脱缩酮化）转化为醇5。酸3可能是合成松香烷和totarane二萜的重要中间体。