Journal of the American Chemical Society
Communication
oxidation had already been demonstrated by Zhao and co-
workers in their synthesis of cephanolides B and C.20 Using
these conditions, along with a strong acid work-up, we realized
the C7 oxygenation along with TMS cleavage in one pot to
access cephanolide C (3) in 54% yield from 22 (eight steps
from 17).
with the C15 hydroxylated constitutional isomer), in 14 steps
from 17.
In summary, on the basis of a retrosynthesis guided by
chemical network analysis, we have developed highly concise
syntheses of cephanolide A (1, 14 steps), cephanolide B (2, 10
steps), cephanolide C (3, 8 steps), and cephanolide D (4, 14
steps) from a commercially available indanone (17). A key
design element of our synthesis plan was to identify a common,
versatile intermediate that could be applied to preparation of
all the cephanolide congeners. Our approach features rapid
construction of the core framework of the cephanolides by
employing an iterative Csp2−Csp3 cross-coupling, followed by
an enol ether/intramolecular inverse-demand Diels−Alder
reaction. We also showcased late-stage oxygenation tactics as
a powerful tool for achieving efficient peripheral structural
diversification. Our synthesis plan sets the stage for the
preparation of other structurally complex Cephalotaxus
norditerpenoids that involve scaffold modifications. These
efforts, as well as the development of an enantioselective
variant of the intramolecular Diels−Alder reaction applied
here, are the subject of ongoing studies in our laboratory.
To access cephanolide D, we effected the same benzylic
oxidation of 22 using PCC but left the tertiary hydroxy group
protected by using slightly modified conditions. We explored,
without success, many ketone62 and other carbonyl-based
auxiliaries in attempts to achieve directed C−C bond-forming
ortho C−H functionalization63−65 of 3 and its derivatives. Our
failure to install the methyl ester directly necessitated the
following effective, albeit indirect, approach. The ketone
installed at C7 of 22 was converted to the acetyl oxime by
condensation with hydroxylamine and subsequent acetylation
in the same pot to afford 24. Oxime-directed ortho C−H
acetoxylation following the precedent of Sanford et al.66
successfully functionalized the arene at C15. Global cleavage of
the acetyl groups, followed by oxidative removal of the
oxime,67 gave hydroxyketone 25 in 33% yield over the four
steps. Of note, while Sanford successfully employed free
oximes in ortho acetoxylations through in situ acetylation of
the oxime, in our case, the free oxime was not easily acetylated
under the acetoxylation conditions (AcOH/Ac2O, heating),
resulting in its oxidative cleavage to give the precursor ketone.
Finally, phenol 25 was treated with Tf2O in pyridine to give
the corresponding triflate (84% yield), which was then
subjected to Pd-catalyzed methoxy carbonylation and a
subsequent one-pot deprotection of the tertiary alcohol to
afford cephanolide D (4) in 93% yield (14 steps from 17).
Lastly, we addressed the synthesis of cephanolide A (1) from
common intermediate 21 (Scheme 2B). While the syntheses of
cephanolides B−D arose directly from reduction of the exo-
methylene group of 21, the synthesis of 1 required the
installation of a hydroxy group at C3. For this purpose, we
employed an allylic oxidation. Analyses of the crystal structures
of 13 or 20 indicated that the oxygenation was likely to occur
from the undesired convex face. Therefore, the allylic alcohol
resulting from SeO2 oxidation of 21 was oxidized to give an
enone (26) in one pot by using DMP in 76% yield.
Hydrogenation of 26 (Pd/C in MeOH), followed by
epimerization of the methyl-bearing stereocenter of 27, gave
the desired ketone (28) in 52% yield over two steps. Reduction
of the ketone group with NaBH4 delivered alcohol 29, which
was subjected to Suarez oxidation conditions68,69 employing I2
and PIDA to forge the desired THF ring without an event. The
free tertiary alcohol was obtained after TMS cleavage in the
same pot (98% yield over two steps), underlining the power of
the Suarez variant of the 1,5-HAT process for late-stage
oxygenation. Surprisingly, the resulting tertiary alcohol did not
undergo ionic deoxygenation under the same conditions59,60
that had worked in the cephanolide B synthesis. As a
consequence, we had to concede to a two-step procedure of
xanthate preparation followed by classical Barton−McCombie
deoxygenation to give 31 in 82% yield over two steps.
Oxygenation of the arene moiety of 31 using phthaloyl
peroxide61 as employed in the final step en route to 2 did not
afford the desired phenol in this case. Ultimately, we found
that 31 was oxygenated using the cyclopropane malonyl
peroxide,70 which was identified after an extensive survey of
reagents, to provide cephanolide A (1) in 39% yield (6:1 ratio
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
Experimental details and spectroscopic data (PDF)
Accession Codes
CCDC 2058197−2058198 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
■
Corresponding Author
Richmond Sarpong − Department of Chemistry, University of
California, Berkeley, Berkeley, California 94720, United
Authors
Maximilian Haider − Department of Chemistry, University of
California, Berkeley, Berkeley, California 94720, United
States
Goh Sennari − Department of Chemistry, University of
California, Berkeley, Berkeley, California 94720, United
States
Alina Eggert − Department of Chemistry, University of
California, Berkeley, Berkeley, California 94720, United
States
Complete contact information is available at:
Author Contributions
§M.H. and G.S. contributed equally to this work.
Notes
The authors declare no competing financial interest.
2713
J. Am. Chem. Soc. 2021, 143, 2710−2715