Total Synthesis of Isokidamycin
B. Michael O’Keefe, Douglas M. Mans, David E. Kaelin, Jr, and Stephen F. Martin
 Carbohyd. Res., 19 (1971) 276-280
 J. Carbohydrate Chemistry, 2(2), 105-114 (1983)
 Aust. J. Chem, 2003, 56, 787 – 794
Dude, this needed some time… Maybe the longest review I ever wrote to date. But Isokidamycin is really a big beasty. Have a look at it:
The synthesis includes various intermediates and a long linear reaction sequence so I will start right off with the red fragment.
Benzylation of the benzoquinone promoted by silver oxide was followed by monobromination. The second bromine was added with the help of pyridine
tribromide, the quinone reduced to the dihydroquinone and selectively monoprotected as the methyl ether.
For identifying the route to the green fragment I had to dig out some really old papers which are cited by the group. The reconstruction was done as described in the paper but might not be 100% correct. I suppose they started with the dihydropyran shown which is commercially available. Monodeprotection and subsequent tosylation was followed by reductive cleavage of the tosylate and the acetyl protecting groups. Reprotection, hydration of the double bond and azidonation delivered after acetylation the green fragment shown. The hydration/azidonation step goes through a SN2’ reaction in which water (or even hydroxide) pushes out the acetate. Protonation and azide addition on the double bond produces the product.
The sugar fragment was then coupled under Lewis acid catalysis with furan (d.r. 72 : 28 with respect to the C3 stereocenter) and deprotected under standard conditions to give after chromatographic separation the major C3 epimer shown. Benzylation, azide reduction, Boc-protection and methylation gave the fully protected THP. Functionalization of the furan was accomplished by silylation followed by hydroboration/oxidation to give the terminal alcohol.
Another interesting THP fragment was isolated from vancomycin (!). By protecting the natural product and heating it in a methanolic solution of HCl, the aminosugar was set free. Some protecting group manipulations furnished the second sugar unit.
It seems to be odd to destroy such an important drug like vancomycin to produce this aminosugar but the author explain this sequence with the extremely long synthesis otherwise needed.
Now it’s getting interesting. The red and blue halves were combined through a Mitsonobu reaction in pretty high yield. In the presence of BuLi the 1,2-dibromonaphthalene produces an aryne intermediate which undergoes a Diels-Alder reaction with the furan ring. The ether handle was then cleaved off with TBAF and the resulting alcohol methylated.
A really nice way to build up the anthracene ring system.
The oxacyclic ring was opened with TMSOTf with concomitant cleavage of the Boc protecting group. Reductive amination, TIPS protection of the free alcohol, selective debenzylation, bromination and MOM protection furnished the fully protected anthracene ring system.
For the next step another intermediate was needed. A Corey-Fuchs alkynylation followed by in situ formylation gave the purple aldehyde shown.
This aldehyde was added to the anthracene ring through standard conditions and the resulting alcohol oxidized with barium manganate. Addition of diethylamine to the triple bond gave rise to a vinylogous amide which cyclizes under Lewis-acid catalysis which was followed by desilylation.
Ok, we are approaching the end. Glycosidation of the advanced intermediate with the vancomycin derived aminosugar was only possible in the presence of Sc(OTf)3. During the course the acetyl protecting group was cleaved off so it was reinstalled.
At last the protecting groups had to be cleaved off. Benzyl group removal was done first with BBr3, the Cbz group was removed in the presence of TMSI which was followed by reductive amination of the free monomethylated amine. The hindered acetate group was cleaved off under standard conditions and the two methyl groups removed oxidatively with cerium sulfate to give isokidamycin.
Wow… what a damn long and cool synthesis. Only two papers in JACS… I bet some generations of Ph.D. students were needed to get this done.