Agelastatins

Scheme 7 Cyclopentene ring closure in the presence of urethane and sulfonamido groups in total synthesis of agelastatin (38) [30]...

The aziridination of alkenes catalysed by [CuCl(IPr)] complex 150 was used in a key step of the total synthesis of (+)-agelastatin 152 (Scheme 5.39) [44], The aziridination occurs in presence of 50 mol% of 150 in 52% yield. It is important to note that 150 was the only complex able to promote the aziridination of 149, an electron-deficient cyclopentene. 

The asymmetric allylic alkylation (AAA) reaction has been adapted for use with pyrrole nucleophiles <06JACS6054>. For example, treatment of pyrrole 55 and cyclopentene 56 with a palladium catalyst in the presence of a chiral additive gave pyrrole 57 in up to 92% ee. The latter was elaborated into piperazinone-pyrrole natural product, agelastatin A 94. 

The numerous excellent synthetic procedures published by A.C. Richardson also continue to be relied upon by contemporary organic chemists because of their great practicality and reproducibility. Indeed, a recent highly cited enantiospecific total synthesis of the antitumor agent, (—)-agelastatin A, justly testifies to this fact it exploits the remarkably useful Hough-Richardson aziridine as a key synthetic intermediate in the eventual published route to this complex natural product. 

SCHEME 8. Hale and Domostoj s use of die Hough-Richardson aziridine in their 2004 enantiospecific total synthesis of die antitumor alkaloid (—)-agelastatin A. 

Alkene Metathesis in Total Synthesis Valienamine, Agelastatin and Tonantzitlolone... 

Applications to natural product synthesis probe the limits of any synthetic method, as situations arise that would never have been considered in the course of first developing the method. Recent syntheses of valienamine 6, agelastatin 9 and tonantzitlolone 15 pressed the limits of ring-closing metathesis. 

The initial conceptualization of the agelastatin A problem took on the form shown below (Scheme 5).17 The key transform in this sequence features intramolecular addition of an amide-derived anion to a tethered alkynyliodonium salt within 33. The alkylidenecarbene generated from this nucleophilic addition, 32, then has a choice of two diastereotopic C-H bonds (Ha or Hb) for 1,5 insertion. Reaction with Ha would provide an advanced intermediate 31 en route to the target 28. Successful execution of this plan would extend alkynyliodonium salt chemistry in three new directions (1) use of an amine derivative as a nucleophile, (2) intramolecularity in the nucleophile addition step, and (3) diastereoselectivity upon alkylidenecarbene C-H insertion. At the initiation of this project, a lack of precedent on any of these topics suggested that focused scouting experiments to assess feasibility would be prudent before beginning work towards the natural product itself. 

SCHEME 5. A first generation approach to the agelastatin A skeleton through alkynyliodonium salt chemistry. 

Scheme 6. Preliminary results in support of the agelastatin A project...

Intramolecularity was the next issue to be probed within the context of alkynyliodonium salt/nucleophile addition reactions.53 1 No prior history was available to guide us, and so the prospects for success remained uncertain. Of primary concern was the potential for iodonium salt/base destructive interactions in competition with the desired N-H deprotonation reaction. A substrate that bore some resemblance to key portions of the agelastatin precursor 33 was prepared (Scheme 6), compound 39. This species duplicated the alkynyliodonium/"amide" pairing of the real system, but it lacked the complex piperazine carbene trap of 33. The tosylimide (pre)nucleophile was proposed as a compromise between what we really wanted (an N-methyl amide) and what would likely work (a tosylamide). Simple treatment of 39 with mild base effected the desired bicyclization to afford the tosylimide product 41 in decent yield. A transition state model 40 for C-H insertion that features an equatorial phenyl unit might rationalize the observed sense of diastereoselectivity. So, at least for 39, no evidence for possible interference by iodonium/base reactions was detected. 

Scheme7. Preliminary results which do not supportthe agelastatin A project.

The conclusion that this synthesis plan was no longer tenable, seemed inescapable. And so, wiser but no further along, we decided to salvage what we could from this exercise and export that information into a second-generation approach to agelastatin A that was designed to avoid this stereochemical pitfall. 

SCHEME 9. Another failed attempt to access the agelastatin A skeleton via alkynyliodonium salt chemistry. 

SCHEME 10. Thethird (and final) approach to agelastatin Athrough alkynyliodonium salt chemistry. 

The unexpected heteroatom lone pair reactivity exposed during the failed 2nd approach to agelastatin A mandated that sufficient attention be directed to the cyclization options of 64. As with carbene intermediate... 

---