Nicolaou, Tang, Wang. ACIEE, 2009, EarlyView. DOI: 10.1002/anie.200900264.
The second of two papers published (almost) in succession in Angewandte from the KCN labs (also check out some more methodology in JACS – nice to see more methods from his group), sporolide B is the more challenging target, but devoid of any biological activity. Nicolaou justifies his synthesis by discussing the biosynthetic origins of the sporolides, which are potentially products of enediyne cyclisation. Convinced? Perhaps…
Anyway, the structure is certainly interesting, being one of relatively few natural products that needs to be drawn in 3D, as a 2D rendition would be a tough -draw. Perhaps the most interesting feature is the 1,4-dioxane-style bridging ring, which would be a tough disconnection in anyone’s book. The other intriguing disconnection is about the chloro-benzene ring, where KCN hoped to utilise a [2+2+2] cyclisation:
Those of you who have already seen the abstract for this paper will have noticed that crayon-o-vision is again reprised, but rather than yet again getting upset about it, I’ve decided to provide my own attempt at it. (Some readers apparently like it!)
Those will (potentially) be my last words on this subject, so let move promptly on to the chemistry. First up is construction of an intermediate distinctly reminiscent of the enediyne natural products mentioned earlier. To get to this point, the group already had 13 steps jotted-down in the lab-book, starting from a chiral cyclopentenone originally developed by Carl Johnson. The more interesting steps included an early palladium mediated carbonylation and later an exceptionally high-yielding Sonogashira coupling using only 2% palladium. The then used a rather nice bit of anion formation to install the chloroacetylene, in which pre-treatment of cis-dichloro ethylene with methyl lithium resulted in a pot of lithiochloroacetylene. Addition of the aldehyde resulted in a impressively clean reaction, almost completing the substrate for the [2+2+2].
The other partner was produced in a less ‘steppy’ manner, installing the asymmetry using a Sharpless dihydroxylation, and using the Yadav procedure to make the the chiral propargyl alcohol. Towards the end, though, I liked the usage of some old-school chemistry to provide an ortho-ester from the acetal protected catechol. The plumbic acetate is a reagent that always grabs my attention, as it featured on the retrosynthetic scheme in my PhD grant proposal. However, I didn’t quite get to the that point of my total synthesis, which tends to stick in my throat. In this case, I don’t quite get the point of the reaction – why do they need to do this chemistry at all? The whole group is removed after the [2+2+2], sits remote from the site of reaction, and surely the orthoester is less stable than the acetal…?
And now to that [2+2+2], and a regioselectivity issue, as there are two ways for the propargyl alcohol partner to approach the diyne. Nicolaou suggests that the perfect selectivity acheived results from steric (the chlorine atom) and chelation (the propargyl alcohol) control, exemplified in the scheme shown below (which I’ve cribbed directly from the paper). Damn nice preparation of a tetra-substituted benzene – something I’m appreciating more often now I’m med-chemming.
A bit of protecting group shennanigans, and it’s on to the macrocyclisation, using an approach gratifyingly different to the usual RCM or lactonisation events usually seen. To install the pyan-style bridge in a [4+2], they needed an equivalent of a diene – in this case, an ortho quinone, provided by silver-oxide oxidation of the diol. Then, in a remarkable feat of molecular flexibility, thermolysis of the intermediate led to a reasonable yield of the desired macrocycle (with a fair chunk of recovered starting material). The selectivity of this cycloaddition is remarkable – sure, the top face of the ‘dienophile’ is restricted by the bulk of the fused cyclpenadiol, but either face of the ortho-quinone should be accessible. Nicolaou describes as much, and admits that the group is none-the-wiser; must have been a damned good day in the lab, though!
The last reaction worth discussing is the delivery of oxygenation to the dioxane to give the ketal. This was neatly done with a bit of hypervalent iodine, using a ten-fold excess of PMB-alcohol. Delivery to the desired ether is presumably controlled by the para-phenol, and leaves only a few more steps to complete the synthesis; vis oxidation/reduction to correction the erroneous configuration of the cyclopentadiol, a round of deprotections and finally a selective epoxidation of the cyclohexadienone. This final reaction was done using a combiation of reagents that I hadn’t come across before – t-BuOOH /DBU; unfortunately there’s no discussion of why these conditions were chosen.
Mmm – another tasty synthesis from the Nicolaou labs. The man takes a lot of flak for this modus operandi, but I personally love this type of synthesis – small(ish), interesting molecules crafted by single-digit teams. However, this view has to be tempered by his ongoing conquest of maitotoxin – not a synthesis I feel is particularly necessary…