Another paper was published ahead while they were in the process of revise.
Even though the first synthesis of the compound was published in 1988, when its bioactivity attracts attention, it is likely to compete for a practical synthetic method.
The advantage of this paper is that Late-Stage Diversification is possible. In fact, 12 derivatives were prepared and those with stronger activity than the natural product were found.
The trans-hydroindane skeleton was constructed in a short process using the classical three-component coupling method and RCM, and the quinone moiety was introduced by Negishi coupling and Friedel-Craft reaction. Finally, a carboxyl group was added by hydroboration-carboxylation to the trisubstituted double bond. The double bond of the late stage intermediate was convertible to the other functional groups to make various derivatives.
The transformation from a carbonyl and its α-position with three carbons to make cyclopentanone reminds me of Robinson cyclization. They made an enol triflate, extended the acetylene part for 3 carbons using Sonogashira, followed by Rautenstrauch cycloisomerization with gold catalyst. It works well without optimization in reasonable yield. The reaction mechanism is quite complicated, but it may become a standard cyclopentanone synthesis method. The use of acetylene and alkene is similar to Pauson-Khand.
The three reactions used for tetraquinane construction are Pauson-Khand, Rautenstrauch, and radical cyclization.
The Kulinkovich reaction is a seemingly strange reaction (a normal addition reaction once you know it is a reaction of dianion), so I was interested in how to use it in the total synthesis. The idea to use this reaction intramolecularly to construct 7 membered ring was not realized. Is it difficult to get the bicyclo[5,1,0]? I am curious if [4,1,0] or [3,1,0] can be obtained.
It was done well intermolecularly, so they combined it with ring-opening carbonylation and summarized as a synthetic method for β-keto esters from terminal olefins.
The ring construction is a royal aldol reaction, but with three carbonyls and four possible enolates. The bases are explored and the desired product was obtained in practical yield. The explanation of selectivity (SI) also makes sense.
The exo-methylene introduction did not work with the standard Eschenmoser’s salt, but was done in good yield with a simple method published in Agric.Biol.Chem.
Three photoreactions (dearomative-6π-electrocyclization, 2+2 addition, and decarboxylative borylation) are the key steps. The first attempt was a reductive radical desymmetrization cyclization with SET, which did not go well, but they succeeded in desymmetrization by dearmative electrocyclization, which led to the regioselective and anti-stereoselective reaction, and then to diastereoselective cyclizations with asymmetric induction. The final diastereoselectivity is not so good (50:31), but not so bad considering the position of the chiral center.
The ABC ring construction is convergent and practical enough.
The introduction of two carbon chains at the α- and β-positions of the enone is achieved by regioselective 2+2 addition, regioselective decarboxylative borylation, and retro-aldol reaction. The acetyl and aldehyde groups are introduced while creating a quaternary chiral center, so the methodology may be of general use. It is interesting that this kind of conversion became possible by inserting the decarboxylative borylation between the classical 2+2 addition and retro-aldol reactions. I would like to try using photochemical reactions.
In the retrosynthesis, the bond made by the quaternary center and the carbon next to the quaternary center was broken. Unfortunately, due to steric crowding, it did not work. They tried to overcome this problem by using less crowded precursors or intramolecular reactions. This also did not work, and they gave up on planned convergent synthesis.
In the end, the simple acrylic ester, was used as a radicophile practically to achieve 4-carbon , 3-carbon and 3-carbon 3-carbon enlongation, from the β-α and α-positions of the ester carbonyl, respectively. (Michael addition and two alkylations, all highly stereoselective) C and D rings are constructed by typical radical cyclization and RCM.
The reductive decarboxylative cyclization/reductive radical-polar crossover (RRPCO)/C-acylation cascade starting from a redox-active ester (RAE) is examined in detail. It is interesting that the carbon dioxide produced by decarboxylation acts as an acylating agent during the reaction.
So far, nine stereoisomers have been synthesized and none of them matched to the natural one.
Eventually, recent computational chemistry has (almost) established the leading isomer and the structure was confirmed by the total synthesis.
Although the four chiral center groups are so far apart that stereochemistry cannot be determined by relations to each other, they efficiently overcame this problem by using the latest computational chemistry.
The synthesis is done by full use of standard asymmetric synthetic reactions, cross-couplings, and RCM.
Synthesis of the model compound with simplified side chain moieties is more than a “preliminary experiment.
Finally, they prove that the side chain plays an important role in the bioactivity.
Nevertheless, it has been more than 15 years since the first publishment. Congratulations!
This is a reinforced version of the synthesis of Zygadenine (via intramolecular DA, radical cyclization, and late stage redox) published in JACS in 2023. The intermediates synthesized there are regioselectively and stereoselectively functionalized to four alkaloids. Specifically, oxidation at positions 3 and 4 of the A ring, oxidation at positions 6 and 7 of the B ring, and reduction at position 15 of the D ring.
The most interesting point is that the iodonium ion reacted with the adjacent benzoyl and benzylidene acetal groups to generate benzoyl orthoester and remove benzylidene group simultaneously. The reaction mechanism is quite plausible.
The key reaction creates five chiral centers (two of which are quaternary centers) in a single step, where the asymmetric induction is as follows
catalyst > next to boron > next to secondary hydroxyl group > [hexacyclic transition state] (desymmetrization) > remaining chiral centers.
Except for the first reaction, the rest of the reactions are normal organic chemistry.
Although the price of the catalyst is a concern, the reaction is practical because 30 g of the initial synthetic intermediate with the important chiral centers can be prepared.
Aside from the key step, there is useful information during the derivation to the natural product.
Improved Kabalka reduction method, which can reduce two carbonyl groups to two methylenes in good yield.
Selective PCC oxidation of axial secondary alcohols in the presence of primary alcohols.
The RCM, which only works well with conjugated dienes.
The conversion of exomethylene to hydroxymethyl, which does not work well with hydroboration, using radical addition of thiophenol.
It may be useful to keep in mind these points.
The dynamic kinetic resolution of racemic alenylzinc is new. The method of making organometallics in advance and adding imine did not work. Barbier’s reaction conditions were essential.
According to SI procedure,
First, zinc was added and activated with iodine (brown color disappears), then imine was added and heated to 90°C, then finally propargyl halide (1.2 equivalents) was added dropwise.
Where was the chirality at the amine derived from? Maybe chiral Sulfone, sugar part, allenyl zinc worked cooperatively. Since it is a great selectivity. I would like to see how kinetic resolution would occur if only chiral sulfone (Ellman Sulfinimin) were used.
C-NH-SO2-NH-C -> (-SO2，-H2 )-> C-N=N-C -> (-N2) -> C-C
to form adjacent quaternary asymmetric centers
C-N(OMs)-NH-C ->(-MsOH) ->C-N=N-C- ->(-N2)-> C-C
to form a bond between the SP2 carbon and the quaternary chiral center.
This is the basic transformation.
Some optimization of the conditions was necessary from time to time, because slight structural differences cause reactivity differences. Finally they succeeded in synthesizing four [n + 1] oligocyclotryptamine alkaloids using a basically unified method. Excellent!
The absolute structure of the head cyclotryptamine is opposite. The C-C bond formation was achieved at 3-3′ (biosynthetically, oxidative coupling interestingly gives meso dimer), then at the 7-position of one benzene ring and the 3-position of another cyclotryptamine. The process was repeated from one bond after another.
Starting from the dimer headcap, a bifunctional cyclotriptamine with a good leaving group at the 3-position and a mesyl amide at the 7-position is connected to make an N=N bond, and a C-C bond is made by N2 elimination. After connecting the required number of cyclotriptamine, the endcap was attached using the same technique. Removal of all protecting groups and reduction of all methyl carbamates to methyl groups gave the final product.
This method of C-C bond formation works with retention of stereochemistry in a good yield, so it might be applicable to other natural product synthesis.
uiploaded (+)-Dihydropleurotinic Acid and (−)-Pleurotin
https://www.ohira-sum.com/wp-content/uploads/2024/12/jacsau24-4206.pdf
論文を直してる最中に先行発表されたとのこと。
https://www.ohira-sum.com/wp-content/uploads/2024/07/jacs24-18230.pdf
初合成が1988年に発表されているような化合物でも，生理活性が注目されると，実用的合成法を目指して競合することになるんでしょう。
本論文のウリはLate-Stage Diversificationが可能であること。実際に12個の誘導体を作って，天然物より強い活性をもつものを見つけています。
古典的な３成分連結法とRCMでトランスヒドロインダン骨格を短工程で構築し，キノン部分は根岸カップリングとFriedel-Craft で導入。最後に3置換二重結合へのhydroboration-carboxylationでカルボキシル基を加えたのがポイントで，その二重結合を種々官能基変換して，いろんな誘導体を作ることができたわけです。
Another paper was published ahead while they were in the process of revise.
Even though the first synthesis of the compound was published in 1988, when its bioactivity attracts attention, it is likely to compete for a practical synthetic method.
The advantage of this paper is that Late-Stage Diversification is possible. In fact, 12 derivatives were prepared and those with stronger activity than the natural product were found.
The trans-hydroindane skeleton was constructed in a short process using the classical three-component coupling method and RCM, and the quinone moiety was introduced by Negishi coupling and Friedel-Craft reaction. Finally, a carboxyl group was added by hydroboration-carboxylation to the trisubstituted double bond. The double bond of the late stage intermediate was convertible to the other functional groups to make various derivatives.
(−)-Bipolarolide D uploaded
https://www.ohira-sum.com/wp-content/uploads/2024/12/jacsau24-4194.pdf
カルボニルとそのα位に炭素3個を足してシクロペンタノンを作る変換が，ロビンソン環化を思い出させる。エノールトリフラートを作り, 薗頭で炭素3個分のアセチレンパートを伸ばし，金触媒でRautenstrauch cycloisomerization。 optimizationなしでもそこそこの収率でうまくいってます。反応機構は結構複雑なんですが，定番のシクロペンタノン合成法になるかも。アセチレンとアルケンを使うのはのはPauson-Khandにも似ているし。
テトラキナン構築に使った反応はPauson-Khand, Rautenstrauch，ラジカル環化の３つ。
中国のグループに先行を許しています。
https://www.ohira-sum.com/wp-content/uploads/2024/07/jacs24-14427.pdf
こちらは[6+2]環化，分子内Heckの２つでテトラキナンを構築。ちょっと特殊？
側鎖の構築では同じような苦労をしています。
The transformation from a carbonyl and its α-position with three carbons to make cyclopentanone reminds me of Robinson cyclization. They made an enol triflate, extended the acetylene part for 3 carbons using Sonogashira, followed by Rautenstrauch cycloisomerization with gold catalyst. It works well without optimization in reasonable yield. The reaction mechanism is quite complicated, but it may become a standard cyclopentanone synthesis method. The use of acetylene and alkene is similar to Pauson-Khand.
The three reactions used for tetraquinane construction are Pauson-Khand, Rautenstrauch, and radical cyclization.
The Chinese group is ahead of them.
https://www.ohira-sum.com/wp-content/uploads/2024/07/jacs24-14427.pdf
This one uses two [6+2] cyclization and intramolecular Heck to construct tetraquinane. A bit special?
They have the same difficulty in constructing the side chain.
Uploaded (±)-Phaeocaulisin A
https://www.ohira-sum.com/wp-content/uploads/2024/12/jacs24-32276.pdf
Kulinkovich反応って一見不思議な反応だから（ジアニオンの反応とわかってしまえば普通の付加反応）、使ってみたい反応です。この反応を分子内でやって，β-ケトエステルに変換する経路がうまくいかなかった。
ビシクロ[5,1,0]は難しいのですかね？[4,1,0]とか[3,1,0]とかはできるのか個人的に興味あり。
分子間ではうまく行ったので、開環的カルボニル化と組み合わせて、末端オレフィンからβ-ケトエステルの合成法としてまとめています。
環構築は王道のアルドールですが、３つのカルボニルがあり、４種のエノラートができる可能性がある中で、塩基を探索し、実用的収率で目的物を得ているのでおもしろい。選択性の説明(SI)も納得できます。
エキソメチレン導入は定番のEschenmoser’s saltでうまくいかず、Agric.Biol.Chem.で発表された簡便な方法で収率よく合成されています。
The Kulinkovich reaction is a seemingly strange reaction (a normal addition reaction once you know it is a reaction of dianion), so I was interested in how to use it in the total synthesis. The idea to use this reaction intramolecularly to construct 7 membered ring was not realized. Is it difficult to get the bicyclo[5,1,0]? I am curious if [4,1,0] or [3,1,0] can be obtained.
It was done well intermolecularly, so they combined it with ring-opening carbonylation and summarized as a synthetic method for β-keto esters from terminal olefins.
The ring construction is a royal aldol reaction, but with three carbonyls and four possible enolates. The bases are explored and the desired product was obtained in practical yield. The explanation of selectivity (SI) also makes sense.
The exo-methylene introduction did not work with the standard Eschenmoser’s salt, but was done in good yield with a simple method published in Agric.Biol.Chem.
uploaded Norzoanthamine
https://www.ohira-sum.com/wp-content/uploads/2024/12/jacs24-32305.pdf
３つの光反応（dearomative-6π-electrocyclization, 2+2 addition, decarboxylative borylation）を鍵段階としています。最初に試みたのはSETによる非対称化をともなう還元的ラジカル環化で，うまくいきませんでしたが，非芳香化電子環化による非対称化を成功させ，さらに位置選択的かつanti選択的反応，さらに不斉誘導をともなうジアステレオ選択的環化反応へと展開しています。最後のジアステレオ選択性はいまいち（50:31）ですが，不斉点の位置を考慮するとやむを得ないところでしょう。convergentで十分実用的なABC環構築法になっています。
エノンのα位とβ位への２つの炭素鎖導入を位置選択的な2+2付加と位置選択的脱炭酸ボリル化とレトロアルドールを利用して達成しています。アセチル基とアルデヒド基を４級不斉中心を作りながら導入した形で，一般にも使えそう。古典的な2+2付加とレトロアルドールの間に脱炭酸ボリル化反応を入れ込むと，こういう骨格構築ができるのですね。光反応使ってみたくなります。
Three photoreactions (dearomative-6π-electrocyclization, 2+2 addition, and decarboxylative borylation) are the key steps. The first attempt was a reductive radical desymmetrization cyclization with SET, which did not go well, but they succeeded in desymmetrization by dearmative electrocyclization, which led to the regioselective and anti-stereoselective reaction, and then to diastereoselective cyclizations with asymmetric induction. The final diastereoselectivity is not so good (50:31), but not so bad considering the position of the chiral center.
The ABC ring construction is convergent and practical enough.
The introduction of two carbon chains at the α- and β-positions of the enone is achieved by regioselective 2+2 addition, regioselective decarboxylative borylation, and retro-aldol reaction. The acetyl and aldehyde groups are introduced while creating a quaternary chiral center, so the methodology may be of general use. It is interesting that this kind of conversion became possible by inserting the decarboxylative borylation between the classical 2+2 addition and retro-aldol reactions. I would like to try using photochemical reactions.
(–)-macrocalyxoformins A and B and (–)-ludongnin C uploadef.
https://www.ohira-sum.com/wp-content/uploads/2024/12/natcom24-6052.pdf
逆合成で，４級中心と４級中心の隣の炭素が作る結合を切断しています。うまく行けばとても収束的できれいな合成になるところでしたが，残念なことに立体的な混み合いのため，うまくいきませんでした。混み合いの少ない前駆体や，分子内反応に持ち込んで克服することを試みましたが，これもうまくいかず，収束的合成は断念しています。
結局radicophileを最も単純なアクリル酸エステルとして，４炭素増炭し，エステルカルボニルのβ位α位α位からそれぞれ３炭素３炭素３炭素増炭し，（マイケル付加と２つのアルキル化，どれも高度に立体選択的）典型的なラジカル環化とRCMでC,D環をつくっています。
redox-active ester (RAE)を出発するreductive decarboxylative cyclization/reductive radical-polar crossover (RRPCO)/C-acylation cascade を詳しく検討しています。ラジカル発生の際に脱炭酸で生じる二酸化炭素がアシル化剤として働いているのがおもしろい。
In the retrosynthesis, the bond made by the quaternary center and the carbon next to the quaternary center was broken. Unfortunately, due to steric crowding, it did not work. They tried to overcome this problem by using less crowded precursors or intramolecular reactions. This also did not work, and they gave up on planned convergent synthesis.
In the end, the simple acrylic ester, was used as a radicophile practically to achieve 4-carbon , 3-carbon and 3-carbon 3-carbon enlongation, from the β-α and α-positions of the ester carbonyl, respectively. (Michael addition and two alkylations, all highly stereoselective) C and D rings are constructed by typical radical cyclization and RCM.
The reductive decarboxylative cyclization/reductive radical-polar crossover (RRPCO)/C-acylation cascade starting from a redox-active ester (RAE) is examined in detail. It is interesting that the carbon dioxide produced by decarboxylation acts as an acylating agent during the reaction.
uploaded Iriomoteolide-1a -1b
https://www.ohira-sum.com/wp-content/uploads/2024/12/jacs24-29836.pdf
これまで９つの立体異性体が合成されて，どれも一致しなかったとのこと。
結局最近の計算化学で有力な異性体を（ほぼ）確定し，全合成で確認しています。
４つの不斉中心群が離れているため，立体化学を相対的に決めていけないのですが，
最新の計算化学を効率よく使って克服しています。
合成は定番の不斉合成反応，クロスカップリング，RCMを駆使。
側鎖部分をシンプルにしたモデル化合物の合成は「予備実験」以上で，
最終的に，側鎖部分が生理活性に重要な役割があることを証明しています。
それにしても，単離構造決定から15年以上，congratulationsです。
So far, nine stereoisomers have been synthesized and none of them matched to the natural one.
Eventually, recent computational chemistry has (almost) established the leading isomer and the structure was confirmed by the total synthesis.
Although the four chiral center groups are so far apart that stereochemistry cannot be determined by relations to each other, they efficiently overcame this problem by using the latest computational chemistry.
The synthesis is done by full use of standard asymmetric synthetic reactions, cross-couplings, and RCM.
Synthesis of the model compound with simplified side chain moieties is more than a “preliminary experiment.
Finally, they prove that the side chain plays an important role in the bioactivity.
Nevertheless, it has been more than 15 years since the first publishment. Congratulations!
uploaded Veratrum alkaloids
https://www.ohira-sum.com/wp-content/uploads/2024/12/natcom24-7639.pdf
2023年にJACSで発表されたZygadenineの合成(分子内DA,ラジカル環化，late stage redoxを経由)の補強版です。そこで合成した中間体を位置および立体選択的に官能基変換し４種のアルカロイドに誘導しています。具体的にはA環の3,4位，B環の6,7位の酸化とD環の15位の還元です。
ヨードニウムイオンが隣接するベンゾイル，ベンジリデンアセタールと反応して，ベンゾイルのオルトエステル生成とベンジリデンの脱離が同時に起きるところが一番の見所で，反応機構はごもっともなものです。
This is a reinforced version of the synthesis of Zygadenine (via intramolecular DA, radical cyclization, and late stage redox) published in JACS in 2023. The intermediates synthesized there are regioselectively and stereoselectively functionalized to four alkaloids. Specifically, oxidation at positions 3 and 4 of the A ring, oxidation at positions 6 and 7 of the B ring, and reduction at position 15 of the D ring.
The most interesting point is that the iodonium ion reacted with the adjacent benzoyl and benzylidene acetal groups to generate benzoyl orthoester and remove benzylidene group simultaneously. The reaction mechanism is quite plausible.
uploaded (-)-cyathin B2
https://www.ohira-sum.com/wp-content/uploads/2024/11/jacs24-25078.pdf
鍵反応は一挙に５つの不斉中心（うち２つは四級中心）を一挙につくる反応ですが，不斉誘起としては
触媒>ホウ素の隣＞二級水酸基の隣＞[六員環遷移状態]（desymmetrization）>残りの不斉中心
で，最初の反応以外は普通の有機化学です。
触媒の価格が気になるものの，重要な不斉中心をもつ初期合成中間体を30g も用意できる実用的な反応とのこと。
鍵段階はともかく，天然物への誘導実験中に有用な情報があります。
２つのカルボニル基を好収率で２つのメチレンに還元できるimproved Kabalka reduction method,
一級アルコール存在下でのアキシアル二級アルコールの選択的PCC酸化，
共役ジエンでしかうまくいかないRCM，
ハイドロボレーションでうまくいかないエキソメチレンからヒドロキシメチルへの変換をチオフェノールのラジカル付加を使って実現，
など，覚えておくと役にたつかもしれません。
The key reaction creates five chiral centers (two of which are quaternary centers) in a single step, where the asymmetric induction is as follows
catalyst > next to boron > next to secondary hydroxyl group > [hexacyclic transition state] (desymmetrization) > remaining chiral centers.
Except for the first reaction, the rest of the reactions are normal organic chemistry.
Although the price of the catalyst is a concern, the reaction is practical because 30 g of the initial synthetic intermediate with the important chiral centers can be prepared.
Aside from the key step, there is useful information during the derivation to the natural product.
Improved Kabalka reduction method, which can reduce two carbonyl groups to two methylenes in good yield.
Selective PCC oxidation of axial secondary alcohols in the presence of primary alcohols.
The RCM, which only works well with conjugated dienes.
The conversion of exomethylene to hydroxymethyl, which does not work well with hydroboration, using radical addition of thiophenol.
It may be useful to keep in mind these points.
uploaded thiolincosamines
https://www.ohira-sum.com/wp-content/uploads/2024/11/jacs24-29135.pdf
ラセミ体のアレニル亜鉛のdynamic kinetic resolutionが新規です。あらかじめ有機金属をつくっておいてイミンを加える方法ではうまくいかず，Barbierの条件が必須とのこと。SIを見てみると，
まず亜鉛をいれてヨウ素で活性化させ（茶色消滅），イミンを加えて90°Cに加熱，最後にハロゲン化プロパジル（1.2当量）を滴下するという手順です。
アミンの付け根の不斉はどこから誘導されたのでしょう。キラルスルフォン，糖部分，アレニル亜鉛の複合的な効果でしょうが，素晴らしい選択性なので，キラルスルフォン(Ellman Sulfinimin)だけだったら，どの程度のkinetic resolutionが起こるか見てみたい気がします。
The dynamic kinetic resolution of racemic alenylzinc is new. The method of making organometallics in advance and adding imine did not work. Barbier’s reaction conditions were essential.
According to SI procedure,
First, zinc was added and activated with iodine (brown color disappears), then imine was added and heated to 90°C, then finally propargyl halide (1.2 equivalents) was added dropwise.
Where was the chirality at the amine derived from? Maybe chiral Sulfone, sugar part, allenyl zinc worked cooperatively. Since it is a great selectivity. I would like to see how kinetic resolution would occur if only chiral sulfone (Ellman Sulfinimin) were used.
uploaded [n + 1] oligocyclotryptamine alkaloids
https://www.ohira-sum.com/wp-content/uploads/2024/11/jacs24-23574.pdf
C-NH-SO2-NH-C -> (-SO2，-H2 )-> C-N=N=C -> (-N2) -> C-C
で隣り合った４級不斉中心をつくり
C-N(OMs)-NH-C ->(-MsOH) ->C-N=N-C- ->(-N2)-> C-C
でSP2炭素と４級不正中心との結合をつくります。
これが基本
ちょっとした構造の違いで反応性に差があったりするので，時々条件の最適化は必要でしたが，，基本的に統一された手法で４種の [n + 1] oligocyclotryptamine alkaloids.の合成に成功しています。すばらしい。
頭のシクロトリプタミンの絶対構造が逆で，まず3-3’でC-C結合をつくり（生合成的には酸化的カップリングで，メソの２量体ができるとの説，おもしろい），つづいて，一方のベンゼン環の７位と別のシクロトリプタミンの３位が次々結合を作っていくことになります。
2量体のHeadcap から始まり，N=N結合を作るために，３位によい脱離基，７位にメシルアミドを有するbifunctional cyclotriptamineをつなぎ，N2脱離でC-C結合生成。必要な数をつなげた後，同様の手法でendcapをくっつけます。最後は全ての保護基を除き，すべてのメチルカーバメートをメチル基に還元。
このC-C結合の作り方，立体保持だし，結構収率もいいので他にも応用できそうです。
C-NH-SO2-NH-C -> (-SO2，-H2 )-> C-N=N-C -> (-N2) -> C-C
to form adjacent quaternary asymmetric centers
C-N(OMs)-NH-C ->(-MsOH) ->C-N=N-C- ->(-N2)-> C-C
to form a bond between the SP2 carbon and the quaternary chiral center.
This is the basic transformation.
Some optimization of the conditions was necessary from time to time, because slight structural differences cause reactivity differences. Finally they succeeded in synthesizing four [n + 1] oligocyclotryptamine alkaloids using a basically unified method. Excellent!
The absolute structure of the head cyclotryptamine is opposite. The C-C bond formation was achieved at 3-3′ (biosynthetically, oxidative coupling interestingly gives meso dimer), then at the 7-position of one benzene ring and the 3-position of another cyclotryptamine. The process was repeated from one bond after another.
Starting from the dimer headcap, a bifunctional cyclotriptamine with a good leaving group at the 3-position and a mesyl amide at the 7-position is connected to make an N=N bond, and a C-C bond is made by N2 elimination. After connecting the required number of cyclotriptamine, the endcap was attached using the same technique. Removal of all protecting groups and reduction of all methyl carbamates to methyl groups gave the final product.
This method of C-C bond formation works with retention of stereochemistry in a good yield, so it might be applicable to other natural product synthesis.