Convergent syntheses are characterized by the coupling of two or more synthetic intermediates of similar complexity, often late in a pathway. At its limit, a fully convergent synthesis is achieved when commercial or otherwise readily available intermediates are coupled to form the final target in a single step. Of course, in all but exceptional circumstances this level of convergence is purely hypothetical; in practice, additional steps are typically required to progress from fragment coupling to the target. Additionally, the length of the sequence required to access each target is a primary consideration in synthetic design.In this Account, we provide an overview of alkaloid, polyketide, and diterpene metabolites synthesized in our laboratory and present parameters that may be used to put the degree of convergence of each synthesis on quantitative footing. We begin with our syntheses of the antiproliferative, antimicrobial bacterial metabolite (-)-kinamycin F (1) and related dimeric structure (-)-lomaiviticin aglycon (2). These synthetic routes featured a three-step sequence to construct a complex diazocyclopentadiene found in both targets and an oxidative dimerization to unite the two halves of (-)-lomaiviticin aglycon (2). We then follow with our synthesis of the antineurodegenerative alkaloid (-)-huperzine A (3). Our route to (-)-huperzine A (3) employed a diastereoselective three-component coupling reaction, followed by the intramolecular α-arylation of a β-ketonitrile intermediate, to form the carbon skeleton of the target. We then present our syntheses of the hasubanan alkaloids (-)-hasubanonine (4), (-)-delavayine (5), (-)-runanine (6), (+)-periglaucine B (7), and (-)-acutumine (8). These alkaloids bear a 7-azatricyclo[4.3.3.01,6]dodecane (propellane) core and a highly oxidized cyclohexenone ring. The propellane structure was assembled by the addition of an aryl acetylide to a complex iminium ion, followed by intramolecular 1,4-addition. We then present our synthesis of the guanidinium alkaloid (+)-batzelladine B (9), which contains two complex polycyclic guanidine residues united by an ester linkage. This target was logically disconnected by an esterification to allow for the independent synthesis of each guanidine residue. A carefully orchestrated cascade reaction provided (+)-batzelladine B (9) in a single step following fragment coupling by esterification. We then discuss our synthesis of the diterpene fungal metabolite (+)-pleuromutilin (10). The synthesis of (+)-pleuromutilin (10) proceeded via a fragment coupling involving two neopentylic reagents and employed a nickel-catalyzed reductive cyclization reaction to close the eight-membered ring, ultimately providing access to (+)-pleuromutilin (10), (+)-12-epi-pleuromutilin (11), and (+)-12-epi-mutilin (12). Finally, we discuss our synthesis of (-)-myrocin G (13), a tricyclic pimarane diterpene that was assembled by a convergent annulation.In the final section of this Account, we present several paramaters to analyze and quantitatively assess the degree of convergence of each synthesis. These parameters include: (1) the number of steps required following the point of convergence, (2) the difference in the number of steps required to prepare each coupling partner, (3) the percentage of carbons (or, more broadly, atoms) present at the point of convergence, and (4) the complexity generated in the fragment coupling step. While not an exhaustive list, these parameters bring the strengths and weaknesses each synthetic strategy to light, emphasizing the key contributors to the degree of convergence of each route while also highlighting the nuances of these analyses.
Read full abstract