(–)-Scabrolide B (again!)

Total synthesis is often likened to scaling an uncharted peak, so perhaps it’s not surprising that one of my favourite things to do outside the lab has long been rock climbing. In my preferred subdiscipline, bouldering, a climber attempts to struggle their way up 10–20 feet of rock with a lot of freedom (and minimal safety equipment). While this isn’t typically a team sport, it’s much more fun with a group of friends who can both assist with problem-solving and help soften falls.

Even after a couple of decades and thousands of climbs, I’m still surprised and delighted by how differently a group of climbers can approach the same problem. Comparing different tactics (‘beta’ in climberspeak) is often eye-opening and educational – plus, it’s just fun to debate the relative merits of sequences or moves.

As a lover of chemical history, I take similar joy in comparing how disparate research groups approach the synthesis of a common target. In the golden age of total synthesis that followed the second world war, popular natural products were made over and over again. Each decade had its classic test pieces, and research groups around the world would compete to find the shortest, most efficient or most elegant route to a hot molecule long after it had first been made.

Older readers might remember the taxol-mania of the 1990s, when over 30 research groups were said to be working on the same target, which now has been synthesised 11 times to date. Sadly, I suspect we’ll never see another molecule get taxol-level hype. A consequence of the waning number of total synthesis-focused research groups is that most natural products are now only synthesised once or twice. This deprives students of the opportunity to contrast and compare different routes to the same molecules, which can be highly educational.

Take a second look

This month, we are revisiting scabrolide B, which was also the subject of my November column that described a synthesis by Alois Fürstner and co-workers at the Max Planck Institute for Coal Research in Germany. David Sarlah and coworkers (now at Rice University in Texas, US) just published the second synthesis of this interesting target,¹ and I’d encourage students to compare this route to Fürstner’s.

As is common for this family of molecules, the Texas team disconnects through the central seven-membered ring. Unfortunately, the ring-closing metathesis approach they’d previously used to make scabrolide A leads to a dead end as introducing the sensitive enedione post-metathesis isn’t feasible (file this under foreshadowing). They instead switch to a Mukaiyama–Michael/Mukaiyama–aldol sequence that gives a more oxidised product.

Preparation of the right-hand side of the molecule begins with a nickel catalysed [3 + 2] annulation of a linalool-derived enone with methylene cyclopropane. A quartet of oxidations then give the required bicyclic lactone with great yield and diastereoselectivity (figure 1). As simple as this sequence looks on paper, it caused the team a lot of headaches and it’s worth diving into the supporting information to look at all the things that didn’t work here. Kudos to the team for including this detail.

Once the bicyclic lactone is coupled with norcarvone, it is time to introduce the tricky enedione. By following the same procedure as their previous synthesis of scabrolide A, the team is able to carry out γ-deprotonation followed by α-oxygenation (figure 2). However, they struggle to effect the 1,3-shift required to put the oxygen in the correct place, despite reports of success on very similar substrates. Eventually, the team makes the surprising discovery that the samarium dienolate resulting from the elimination of the alcohol they just installed does react with oxygen at the required γ-position, and a final Dess–Martin oxidation of the γ-alcohol completes the target.

Like the Fürstner group, the team is also able to convert scabrolide B into several other natural products, but you’ll have to check out the paper to get the whole story!