Abstruct

A polycyclic ether (-)-brevisin (1, Figure 1) was isolated from the red tide dinoflagellate Karenia brevis, which produces a variety of polycyclic ethers such as brevetoxins, brevenal, and a monocyclic ether amide brevisamide. Its unique structure consists of two fused tricyclic ether ring assemblies bridged by a methylene carbon and a conjugated aldehyde side chain, which is similar to brevenal and brevisamide. Interestingly, in spite of its unique skeletal structure which is divided into two tricyclic ether units by the methylene, 1 inhibits the binding of tritiated 42-dihydrobrevetoxin B (PbTx-3) to the voltage sensitive sodium channels (VSSC). However, similarly to the other marine polycyclic ethers, the biological activities have not been fully investigated due to the extremely small amount of supply from natural sources. In order to elucidate its interaction with VSSC and evaluate its other biological activities, chemical synthesis to supply material was essential.

Synthesis of a BC/DE ring model of brevisin for diastereomeric confirmation through the acyclic junction

The separated skeletal structure in the center of the molecule is quite unusual among marine polycyclic ethers, and the methylene-linked structure is presumed to generate molecular flexibility, which plays an important role in an interaction with VSSC. In order to elucidate the interaction of 1 with VSSC, structural confirmation was essential. Firstly, the author synthesized a BC/DE ring model (2, Figure 2) for diastereomeric confirmation through the acyclic junction. The NMR chemical shifts from CH-14 to CH-25 of 2 were compared with those of 1, as listed in Table 1. The observed chemical shifts of 2 around methylene junction agreed well with those of 1, while relatively large difference in the B ring and the E ring regions are considered to be due to the absence of the A and the F rings. The synthetic BC/DE ring model bear a strong similarity to the same region of brevisin in its NMR data, supporting the diastereomeric relation of 1 around methylene juncture. The unique structure of brevisin was confirmed as depicted in 1 by the partial synthesis. Then, the author began the total synthesis of 1 based on this confirmation.

Synthetic plan

The employed synthetic plan is summarized in Scheme 1. Side chain fragment 3 and iodide fragment 4 would be connected by means of Suzuki-Miyaura cross coupling. Polycyclic ether core would be synthesized from the ABC ring methyl ketone 5 and the EF ring aldehyde 6 by aldol addition and subsequent construction of the D ring. The tricyclic ether 5 was synthesized from the A ring exocyclic enol ether 7 and the C ring ketene acetal phosphate 8 by a Suzuki-Miyaura cross coupling-based strategy invented in this laboratory.

Synthetic of the ABC ring

The synthesis of A ring fragment 7 started from 3-benzyloxy-1-propanol (9), which was converted to acrylate 10 in 3 steps. Ring-closing metathesis of 10 by Grubbs' second-generation catalyst furnished α,β-unsaturated lactone 11 in 94% yield. Dihydroxylation of 11 by OsO4 followed by protection with TES groups gave lactone 12. Treatment of 12 with Petasis reagent led to the A ring exocyclic enol ether 7 (Scheme 2).

The A ring fragment 7 was connected by Suzuki-Miyaura cross coupling to the C ring ketene acetal phosphate 8, which was prepared in 8 steps from commercially available 2-deoxy-D-ribose. Cross-coupled product 13 was converted to tricycles 14 by the construction of the B ring. Then the ABC ring methyl ketone 5 was synthesized in 3 steps from alcohol 14 (Scheme 3).

Synthetic of the EF ring

The synthesis of the EF ring fragment 6 started from acid catalyzed 6-endo cyclization of hydroxy epoxide 16 to afford pyran 17. Pyran 17 was converted to allylstannane 18 in 3 steps. This ester 18 was reduced to the corresponding aldehyde by DIBALH, and then treated with BF3・OEt2 to furnish the EF ring compound 19. Then the EF ring aldehyde 6 was synthesized in 4 steps from 19 (Scheme 4).

Total synthesis of (-)-brevisin

Connection of 5 and 6 by aldol addition is illustrated in Scheme 5. Treatment of the lithium enolate derived from 5 with aldehyde 6 furnished coupled product 20 as a separable mixture of C-23 diastereomers. Silyloxyketone 20 was converted to pentakis-TES ether 21 by construction of the D ring.

At this stage, in order to convert 21 to iodide 4, only primary TES ether had to be selectively removed in the presence of four secondary TES ethers. During this work, the author established the highly selective deprotection of silyl ethers using DIBALH, and its application on pentakis-TES ether 21 gave primary alcohol 22 in 88% yield. The primary alcohol 22 was converted to iodide 4 with I2, PPh3, and imidazole. Finally, connection of the fragments 3 and 4 by means of Suzuki-Miyaura cross coupling followed by the deprotection of all silyl groups and chemoselective oxidation of the allylic alcohol gave rise to 1 in 75% yield for three steps. The optical rotation and the other spectroscopic data of synthetic 1 were identical with those of natural 1.

In conclusion, the author accomplished the first total synthesis of (-)-brevisin (1). The polycyclic ether core was constructed in short steps by means of Suzuki-Miyaura cross coupling reaction and aldol addition as the key steps. It is noteworthy that the synthesis was accomplished in only 29 longest linear sequence from commercially available 2-deoxy-D-ribose. The present synthesis will be an important clue to elucidation of biological activities of marine polycyclic ether compounds.

Figure 1. Structures of brevisin (1), brevenal, and brevisamide

Figure 2. Sturucture of the BC/DE ring model(2)

Table 1. NMR chemical shifts of 1,2,and their differences (Δδ)in pyridine-d5

Scheme 1. Synthetic plan

Scheme 2. Synthesis of A ring

Scheme 3. Synthesis of the ABC ring

Scheme 4. Synthesis of the EF ring

Scheme 5. Synthesis of the ABC/DEF ring polycyclic ether core

Scheme 6. Completion of the total synthesis

本論文は6章からなり、海洋赤潮プランクトンより近年単離された天然物ブレビシン(下図)に関して、モデル化合物の化学合成によるその化学構造の推定と化合物本体の合成による化学構造の確認、および天然よりの供給が微細である本化合物の生物活性研究に供すべき試料の実践的かつ効率的な全合成について述べられている。

第1章は序論であり、本研究で述べられている縮合環型ポリエーテル天然有機化合物の背景として、これまでに魚類の大量斃死の原因となる赤潮プランクトンや、食物連鎖を経由した魚類の摂食による食中毒の原因となる海洋性植物プランクトンよりその原因化合物として単離・構造決定された同様の化学構造を有する一連の化合物が紹介されている。ここでこれらの毒性発現に関して報じられている生理学研究とその化学構造に基づく活性発現機構として提案されている作業仮説に関して、本論文提出者独自のものを含めて論じられている。さらに本論文で用いた合成研究の手段に関連して従来行われいるこうした化学構造の有機合成研究について詳細に述べられている。

第2章では本論文の合成標的としてのブレビシンの暫定的な構造決定の経緯と、前章に述べられている多くの縮合環型ポリエーテルと異なる特徴として、これが分子中央部に可動なメチレン連結部を含むためにこの部分を挟んだ相対立体配置(ジアステレオ異性)が不確定であることが冒頭に述べられ、これを解決すべく筆者が行ったこの部分のモデル化合物の合成によるこの2種の候補異性体の一方が天然物の化学構造であるという実質的な構造確認の経緯が述べられている。これにより本研究の位置付けが明確になっている。

第3章では前章で推定された化学構造に基づき、本天然物全体の化学合成を行いこれを達成した内容に関して述べられている。これによりこの化学構造を確定するとともに、天然のプランクトン由来からは困難な量的調達を克服する、有機合成による効率的かつ実践的な本化合物の調達を可能とし、結論である第4章にてこの意義が述べられている。

第5章は上記の実験で得られた結果の詳細、第6章では実験条件の詳細が記述されており、これにより追試が可能となっている。

本研究のうちこの背景となる化学構造解析の部分は佐竹真幸その他の印刷公表論文での共著者の寄与によるが、これを確定した有機合成部分はその計画立案と実行を含めて、本論文提出者の貢献寄与が大であると判断できる。

従って、本論文提出者である倉永健史は、博士(理学)の学位を授与できるものと認める。