Enantioselective Syntheses of Aeruginosin 298-A and its Analogs.

The synthesis of both natural and unnatural organic compounds in opticaly pure form is a central challenge in chemistry, especially in relation to the study of biologically active compounds. Aeruginosin 298-A was isolated from the freshwater cyanobacterium Microcystis aeruginosa （NIES-298） is an equipotent thrombin and trypsin inhibitor. It has a tetrapeptide-like structure including nonstandard α-amino acids such as 3-（4-hydroxyphenyl）lactic acid （Hpla） and 2-carboxy-6-hydroxyoctahydroindole （Choi）. I developed a versatile synthetic process focusing on the preparation of 14 for the synthesis of aeruginosin 298-A as well as their several attractive analogs, which were synthesized to gain insight into the structure-activity relations. In the process all stereo centers were controlled by catalytic asymmetric phase-transfer reaction promoted by two-center asymmetric catalysts and catalytic asymmetric epoxidation promoted by a lanthanide-BINOL complex.1, 2 Furthermore, serine protease inhibitory activities of aeruginosin 298-A and its analogs were also examined. 2-carboxy-6-hydroxyoctahydro indole （Choi） is a unnatural amino acid and most important key intermediate in the total synthesis of aeruginosin 298-A. I prepared 14 via phase-transfer alkylation of 5 with 7, which was prepared from panisyl alcohol, as a electrophile. When using （R,R）-6a as a catalyst,the asymmetric PTC of 5 and 7 proceeded smoothly to afford the desired product L-8 in 80% yield and 88% ee. Treatment of the product L-8 with 4 N HCl in methanol promoted deprotection of the benzophenone imine and ketal,transesterification, migration of the C-C double bond, and then 1,4-addition of the resulting amine to enone, leading to the bicyclic compound 10 in 72% yield （one-pot, five reactions）. After benzylation,the key intermediate was obtained in 84% yield as a diastereomixture （11:12 = 2:1）. The undesired isomer 11 was transformed to the desired 12 under acidic conditions （78%, 11:12 = 1:8） and 12 was successfully converted to the key intermediate 14 （71% in 2 steps） by following Bonjoch's procedure.3In addition, optically pure 14 was obtained by recrystallization （>99% ee, 77%）.

Furthermore this unnatural amino acid 15 was successfully used in the total synthesis of aeruginosin 298-A. As shown in Scheme 3, coupling reaction with 16 proceeded smoothly using EDC, HOBt condition to afford 17 with 68% overall yield （dr=12:1）.Compound 17 after Boc deprotection was coupled with compound 18 under HATU condition. Reduction using LiBH4 in THF, subsequent deprotection of TIPS and Cbz groups provided aeruginosin 298-A 1.4 Moreover, six additional analogues were synthesized in a similar way and screened for inhibitory activities against serine protease trypsin.5

Direct Catalytic Asymmetric Aldol-Tishchenko Reaction of Methylene Ketones.6

Since the first successful intermolecular direct catalytic asymmetric aldol reaction of aldehydes with unmodified ketones using heterobimetallic asymmetric catalyst devoleped by Shibasaki's group, Shibasaki's and other groups have attempted this type of direct reaction with great success. In almost all of these direct asymmetric catalyses, however, only limited donors, such as methyl ketones, α-hydroxy ketones, and easily enolizable aliphatic aldehydes, are utilized. Thus, despite the high demand for the development of a direct aldol reaction of ethyl ketones, direct aldol reaction of ethyl ketones are viewed as a formidable synthetic challenge due to poor participation of the resulting aldolates in catalyst turnover and a strong tendency towards retro-aldol reactions. I recently accomplished this elusive aldol reaction using the aldol-Tishchenko reaction. 6By coupling an irreversible Tishchenko reaction to a reversible aldol reaction, the catalytic aldol-Tishchenko reaction provides high product yields and highenantioselectivities.

Preliminary studies using propiophenone（15a）with 4-cnlorobenzaldenyde（16） in the presence of 10 mol % of LLB, afforded the anticipated sequential aldol-Tishchenko product with excellent diastereoselectivity and moderate enantiocontrol （>98/2 dr and 64% ee for 18）; however,the catalytic efficiency was unsatisfactory （Table 1, entry 1）. I next examined the use of a metal salt additive. While a variety of lithium salts were productive in this context, 30mol % LiOTf provided the optimal reaction efficiency and high selectivity （entry 2）. Keeping the 1:3:3 ratio oflanthanum, BINOL, and LiOTf in mind, I investigated whether the same catalytic system would beaccomplished by mixing 6 equiv of BuLi to the mixture of the 1:3 ratio of La（OTf）3 and BINOL. The commercial availability of La（OTf）3, high tolerance to air and water, and superior levels of asymmetric induction and efficiency （entry 3） exhibited by this modified procedure prompted me to further explore this catalytic condition. Switching to ketone 15b had a significant effect on the reactivity and Tishchenko selectivity, while maintaining the similar enantiocontrol （entry 4）. Superior levels of asymmetric induction were realized by decreasing the amount of BuLi. Thus, 1:3:5.6 La（OTf）3:BINOL:BuLi was the appropriate ratio for a broad range of substrates （entry 5）.

Experiments to probe the scope of the aldehyde substrates are summarized in Table 2. A variety of alkyl and heteroatom substituents can be incorporated on the phenyl ring at both the meta and para positions （Table 2,entries 1-7, 85-95% ee, 65-96% yield）. The aryl framework can be successfully extended to naphthalene and heteroaromatic derived systems （entries 8-10, 88-94% ee, 67-82% yield）. In addition, a number of aromatic ketones can also be used without loss of reaction efficiency or enantiocontrol （entries 11-15, 84-88% ee, 60-81%yield）. I next examined the capacity of the present catalytic system to catalyze asymmetric aldol-Tishchenko reactions of propyl and butyl ketones. As highlighted, the catalyst exhibited similar efficiency without Table 2: Direct aldol-Tishchenko Reactions: Substrate Scope. considerable deterioration of enantiocontrol（entries 16 and 17, 87-88% ee, 88-90% yield）.

Preliminary mechanistic studies were performed to inspect the relation between the aldol product and the Tishchenko product, as well as their stereoselectivities. The aldol byproduct 17 was obtained by the reaction of 15a and 16 was with no enantio-or diastereoselectivity. Attempted deliberate retro-aldolization of independently prepared racemic aldol adducts 17 （syn:anti = 7:3 or syn:anti = 3:7） under representative reaction conditions gave the same mixtures of 15a, 17（syn:anti = 4:6, racemic）, and 18 （>98/2 dr, 70%ee） starting with either a 7:3 or a 3:7 syn:anti ratio of aldol adducts 17. These results are consistent with the rapid retro-aldol cleavage of metal aldolate and confirm the essential role of the Tishchenko reaction in controlling the stereoselectivity.In summary, aldol- Tishchenko reaction was established as one of the useful methods for overcoming the retro-aldol reaction problem for a variety of aromatic donors and acceptors.

In addition, I and one of my coworker observed a dynamic structural change of LLB to a novel binuclear La2Li4pentakis（binaphthoxide） complex by the addition of LiOTf.7 Thus, I became interested in the role of LiOTf and the efficiency of these catalytic systems in other asymmetric reaction promoted by heterobimetalic bifunctional catalyst. In direct intermolecular aldol reaction of aldehydes and ketones, addition of LiOTf gradually decreased the enantioselectivity. On the other hand, in the case of nitro aldol reaction, LiOTf had no effect on the enantio or diastereoselectivities. In summary, heterobimetalic complex prepared from La（OTf）3:BINOL:BuLi （1:3:6） was successfully used for nitro aldol reaction without decrease in enantio and diastreoselectivities. This novel catalyst system can be efficient alternative for conventional LLB, Which was prepared from rather expensive La（OiPr）3.

アレギノシン298-Aの不斉合成

アレギノシン298-Aは淡水シアノバクテリアMicrocystis aeruginosa より単離され、セリンプロテアーゼ阻害活性を有する。ガナナデシカンはアレギノシン類の構造活性相関研究を指向し、すべての立体を不斉触媒により制御することで、柔軟で多様な類縁体合成の可能な合成ルートの確立を目指し研究に着手した。本化合物の合成戦略をScheme 1に示した。アミド結合部で切断することで四つのフラグメントへと逆合成できる。これらフラグメントのうち、ガナナデシカンは主としてChoi部を不斉相間移動触媒6 を用いることで効率的に構築する検討を行い、Scheme 2 に示す効率的なルートを確立した。さらに、2）Choi部と共同実験者らにより合成されたそれ以外のフクラグメントのカップリング反応によるアレギノシン298-Aの合成（Scheme 3）、そして3）類縁体の合成にも成功した。

メチレンケトン類の直接的触媒的不斉アルドールーティシェンコ反応の開発

非修飾ケトン、アルデヒド、エステル類を用いる直接的触媒的不斉アルドール反応が近年注目を集めているが、メチレンケトン類を求核剤とする直接的触媒的不斉アルドール反応は逆反応が優先するため高い化学収率とエナンチオ選択性を実現することが困難とされていた。

ガナナデシカンは可逆的な直接的アルドール反応と非可逆なティシェンコ反応を組み合わせることにより上記問題点の克服に成功した。初期検討の結果、希土類触媒が有効なことを見い出し、さらに詳細を検討した結果、La（OTf）3とBINOL、ブチルリチウム（1：3：5.6 の比率）より調製したランタン-リチウム-BINOL 触媒が最適であった。基質としては、電子求引基をもつ芳香族ケトンが特に有効であり、高いジアステレオ選択性、エナンチオ選択性、そして化学収率にてアルドールーティシェンコ体を得ることに成功した（Table 1）。予備的なメカニズム解析より、本反応において、a）アルドール反応は速やかに平衡に達しておりラセミのアルドール体を与えていること、そして、b）希土類触媒がアルドール体のキラル認識を経てティシェンコ反応を立体選択的に促進することが分かった。

以上の結果は創薬化学研究に対し重要な貢献をすると考え、博士（薬学）に十分相当する研究成果と判断した。

Scheme 1. Retrosynthetic Analysis.

Scheme 2. Synthesis of L-Choi

Scheme 3. Completion of total synthesis.

Table 1.Direct Aldol-Tishchenko Reactions.