Communication

pubs.acs.org/JACS

Enantioselective Chemical Syntheses of the Furanosteroids

(−)-Viridin and (−)-Viridiol
Matthew Del Bel,# Alexander R. Abela,‡ Jeffrey D. Ng,‡ and Carlos A. Guerrero*,†
Department of Chemistry and Biochemistry, University of California, San Diego, 9500 Gilman Drive, La Jolla, California 92093-0358,
United States
*
S Supporting Information

ABSTRACT: Herein we describe concise enantioselective

chemical syntheses of (−)-viridin and (−)-viridiol. Our
convergent approach couples two achiral fragments of
similar complexity and employs an enantioselective
intramolecular Heck reaction to set the absolute stereo-
chemical configuration of an all-carbon quaternary stereo-
center. To complete the syntheses of these base- and
nucleophile-sensitive natural products, we conduct care-
fully orchestrated site- and diastereoselective oxidations
and other transformations. Our work is the first to
generate these targets as single enantiomers.

W ortmannin (1),1 viridin (3),2 and other furanosteroids

have been the subjects of structural and limited
biological studies for over seven decades (Figure 1).3 However, Figure 1. Furanosteroid natural products.
biological investigations involving wortmannin exploded in
1994, following three independent reports documenting its of (−)-viridin and (−)-viridiol. Our ultimate goal is to be able
ability to selectively and irreversibly inhibit phosphatidylinositol to generate viridin analogues with deep-seated changes to the
3-kinases (PI3Ks) at single-digit nanomolar concentrations.4 C- and D-rings (e.g., heterocyclic isosteres) because even subtle
Unsurprisingly, wortmannin became the reagent of choice in modification in this region of wortmannin leads to significant
fundamental biochemical studies of PI3Ks. Also, given the changes in potency6 and because this region embeds itself most
increasing awareness and understanding of PI3K’s role in deeply in the PI3K active site.13 Importantly, while medicinal
cancer development and progression,5 medicinal chemistry chemists have largely exhausted all reasonable synthetic
efforts ensued based on this steroid scaffold.6 Wortmannin’s modifications to wortmannin,6 our late-stage fragment coupling
direct use in therapeutic applications, however, is precluded by approach may permit broader SAR studies with the aim of
two problems: it is rapidly degraded in serum, and despite its addressing PI3K/target specificity. However, before such long-
high affinity for PI3Ks, its potency as an irreversible covalent term goals can be pursued, we wish to establish the feasibility of
inhibitor leads to general toxicity.7 In response, Wipf developed a fragment coupling strategy by completing a total synthesis of
PX-866 (2), a prodrug, from wortmannin, and this agent has viridin itself.
advanced to Phase II clinical trials.8 Alongside semisynthesis As shown in Scheme 1A, our synthesis begins with acrylate 5,
studies by Wipf and others, several groups have also pursued a generated in one step from a known aryl triflate14 via Heck
total synthesis of 1, with only Shibasaki’s efforts reaching alkenylation. Hydrogenation (3.4 atm of H2 over Pd/C)
fruition.9,10 reduces the alkene and removes the benzyl group, furnishing
In principle, viridin (3) may also be a viable candidate for dihydrocinnamic acid 6. This intermediate is then directly
PI3K inhibition-based chemotherapy. It, too, bears the hallmark dissolved in neat chlorosulfonic acid, stirred at ambient
2,4-diketofuran that imbues wortmannin with its potent, PI3K- temperature for 12 h, and quenched into ice-cold water,
alkylating activity. Remarkably, in a side-by-side comparison it enabling isolation of indanone 7 by simple trituration. Finally,
was found that wortmannin and demethoxyviridin, also a thioesterification and triflation yields functionalized indanone 8.
natural product, inhibited mammalian PI3K activity at single- The A-ring furan fragment is prepared starting from furan
digit nanomolar concentrations.11 Taken together, the bio- 9,15 previously prepared by Keay in his own studies9 toward
logical profile and untapped medicinal chemistry potential of viridin (see Scheme 1B). The hydroxyl group is converted to
viridin and its congeners provide strong motivation for the chloride using methanesulfonyl chloride and triethylamine,
synthesis studies. Sorensen duly completed the first total
syntheses of each viridin and viridiol (4)2c in racemic form in Received: March 22, 2017
2004.12 Herein, we report our own enantioselective syntheses Published: May 2, 2017

© 2017 American Chemical Society 6819 DOI: 10.1021/jacs.7b02829

J. Am. Chem. Soc. 2017, 139, 6819−6822
Journal of the American Chemical Society Communication

Scheme 1. Preparation of Coupling Partners 8 (A) and 12 (B)

Scheme 2. Synthesis of Advanced, Diastereomeric Hydroxy Ketals 19 and 20

which is then displaced by allylmagnesium chloride, furnishing molecular Heck reaction catalyzed by a t-Bu-PHOX-ligated
unconjugated diene 10. Ring-closing metathesis catalyzed by Pd(0) complex.17 In line with previous observations, the use of
Grubbs’s second-generation Ru-alkylidene catalyst gives 11, and this ligand with 1,2,2,6,6-pentamethylpiperidine is crucial in
subsequent lithiation-stannylation provides furan 12, which was achieving complete reactant consumption and in minimizing
used directly as obtained in subsequent chemistry due to slow alkene isomerization (observed with diphosphine ligands; see
destannylation upon storage. the Supporting Information for additional discussion of this
Although we were initially interested in using lithiated 11 in a reaction).17a,b Heck cyclization thus furnishes 14 in 75% yield
ketone synthesis with indanone 8, the superiority of a and high enantioselectivity (>99% ee).
Liebeskind stannane-thioester coupling16 quickly became Unconjugated alkene 14 bears the complete carbon skeleton
evident (see Scheme 2). At first, the site of oxidative addition of viridin but resides at a much lower oxidation state. We
on indanone 8 was uncertain (aryl triflate versus thioester), believe this dichotomy of complexity, manifest in the work of
though in practice only reaction of the thioester was observed, others as well, is why no synthesis of viridin has followed
furnishing diketone 13 in high yield (83% based on 8) on a Sorensen’s work.9,12 The first step to ameliorate this oxidation
multigram scale. In the stereochemically defining event of our state difference is a diastereoselective Upjohn dihydroxylation,
synthesis, ketone 13 undergoes highly enantioselective intra- furnishing diol 15; the relative flatness of the A-, B-, and C-
6820 DOI: 10.1021/jacs.7b02829
J. Am. Chem. Soc. 2017, 139, 6819−6822
Journal of the American Chemical Society Communication

rings7 ensures that the angular methyl group biases reactions Scheme 3. Completion of the Synthesis of Viridin (3):
occur on the α-face in this and several ensuing reactions. Next, Convergence of Alcohols 19 and 20 via Viridiol (4) and epi-
double Swern oxidation and O-methylation gives methoxy Viridiol (23)
enone 16 in 65% isolated yield over two steps. Of strategic
importance, taking advantage of the propensity of 1,2-diketones
to exist as their tautomeric forms is an indirect way to oxidize at
C3, an essential transformation given that we were unable to
oxidize at this position of intermediate 14 by other means in
other, unsuccessful approaches. Additionally, the acidity of the
diosphenol unit permitted site-selective deprotonation and
alkylation, sparing the D-ring.
The TBS group in 16 is removed by stirring this intermediate
in 1:1 TFA/CH3NO2 for 2 days; all fluoride-based desilylations
lead to decomposition. Site- and diastereoselective reduction of
17 is then accomplished using a three-stage, single-flask
sequence wherein: (1) the D-ring ketone is temporarily
protected as an enol silyl ether, (2) the A-ring ketone is
selectively reduced with BH3·THF; and (3) the intermediate
reduction product is desilylated with Et3N·3HF, delivering keto
alcohol 18 in 72% isolated yield.
At first glance, allylic alcohol 18 is only two steps away from
viridin, namely, hydroboration−oxidation and selective alcohol
oxidation. However, we were unable to perform the former
process, despite strong precedent.18 Instead, we executed an
alkene epoxidation−reduction sequence. Thus, treating enol 18
with AcOOH in methanol delivers an inconsequential 1:2
mixture of hydroxy ketal diastereomers, where methanol was
incorporated by design, mirroring chemistry reported in
Myers’s dynemicin synthesis.19 This trapping event was late-stage functionalization of the A-ring. Thereafter, several
incorporated deliberately for experiments that aimed to novel and strategic maneuvers are executed including (1) an
generate a C2-ketone, which failed due to the extreme intramolecular, highly enantioselective Heck reaction to set the
instability of the desired product, a compound that we were absolute stereochemical course of the synthesis; (2) double
never able to observe in pure form. Swern oxidation to indirectly oxidize C3; (3) an epoxidation-
Although alkene epoxidation seemed to solve our alkene trapping sequence to install a hydroxyl group at C3; and (4)
reactivity problem, ketals 19 and 20 presented a new, stern highly diastereoselective demethoxylation to address function-
challenge: diastereoselective demethoxylation, given that their ality at C2. We emphasize that although our syntheses are built
C2-ketone counterparts were intractable. Despite ample on the edifice of a highly enantioselective, intramolecular Heck
precedent for monodealkoxylation of ketals, particularly in reaction, the uncertainty of the path forward loomed large,
the chemical literature on (−)-oseltamivir,20 the complexity of especially in light of other studies where the viridin skeleton
our intermediates cast a dark shadow over the planned ketal was secured, but the natural product itself remained elusive.9
reduction, especially given the possibility for hydride shifts. Our syntheses of (−)-viridiol and (−)-viridin generate these
Nevertheless, as shown in Scheme 3, isolated diastereomers 19 targets as single enantiomers for the first time and proceed in
and 20 are demethoxylated using TMSOTf and a hydride 17 and 18 steps, respectively, from commercially available
donor. The low efficiency of these reactions is due to materials and thus compare favorably to prior art (for complete
decomposition of either reactants or products, and not to disclosure regarding step counting methodology, see the
poor diastereoselectivity.21 Moreover, both reducing agents, Supporting Information). More generally, by demonstrating
Et3SiH and BH3·S(CH3)2, are effective in the conversion of 19 the feasibility of a fragment coupling approach to access the
to 4 (viridiol) and of 20 to 21 (epi-viridiol), and the use of each viridin core, we have enabled the synthesis of analogues with
merely reflects the most detailed procedure available at the time deep-seated modifications to the C,D-ring system for medicinal
these studies were brought to a close. Finally, each isolated chemistry investigations.

■
diastereomer 4 and 21 is converted to viridin itself using
TEMPO-catalyzed oxidation, though a molar excess of each ASSOCIATED CONTENT
catalyst and terminal oxidant, PhI(OAc)2, was used to force the *
S Supporting Information
reaction to proceed at an acceptable rate. Synthetic viridin The Supporting Information is available free of charge on the
independently produced from diastereomers 4 and 21 was ACS Publications website at DOI: 10.1021/jacs.7b02829.
spectroscopically indistinguishable from each other as well as
from the natural product, although we noted a difference in Experimental procedures and spectroscopic data for all
specific optical rotation: synthetic (−)-3, [α]D = −145° (c = isolated intermediates, as well as additional references
0.23, CHCl3); natural (−)-3,2a [α]D = −222° (c = 1.00, and discussion (PDF)
CHCl3); see the Supporting Information for a discussion of this
discrepancy.
In summary, our synthesis of (−)-viridin up to fragments 8
■ AUTHOR INFORMATION
Corresponding Author
and 12 is direct and efficient, permitting rapid assembly and *[email protected]
6821 DOI: 10.1021/jacs.7b02829
J. Am. Chem. Soc. 2017, 139, 6819−6822
Journal of the American Chemical Society Communication

ORCID Tosaki, S.; Shibasaki, M. Angew. Chem., Int. Ed. 2002, 41, 4680. Formal
Carlos A. Guerrero: 0000-0002-2377-1102 enantioselective synthesis: Shigehisa, H.; Mizutani, T.; Tosaki, S.;
Ohshima, T.; Shibasaki, M. Tetrahedron 2005, 61, 5057.
Present Addresses (11) Woscholski, R.; Kodaki, T.; McKinnon, M.; Waterfield, M. D.;
#
Pfizer Worldwide Research and Development, 10770 Science Parker, P. J. FEBS Lett. 1994, 342, 109.
Center Drive, San Diego, California, 92121, United States. (12) Anderson, E. A.; Alexanian, E. J.; Sorensen, E. J. Angew. Chem.,
†
Bristol-Myers Squibb, One Squibb Drive, New Brunswick, Int. Ed. 2004, 43, 1998. Synthesis of (±)-viridin and (±)-viridiol:
New Jersey, 08903, United States. Alexanian, E. J. Ph.D. Thesis, Princeton University, 2006.
Author Contributions (13) Walker, E. H.; Pacold, M. E.; Perisic, O.; Stephens, L.; Hawkins,
‡ P. T.; Wymann, M. P.; Williams, R. L. Mol. Cell 2000, 6, 909.
A.R.A. and J.D.N. contributed equally.
(14) Fürstner, A.; Thiel, O.; Blanda, G. Org. Lett. 2000, 2, 3731.
Notes (15) Cristofoli, W. A.; Keay, B. A. Synlett 1994, 625.
The authors declare no competing financial interest. (16) Wittenberg, R.; Srogl, J.; Egi, M.; Liebeskind, L. S. Org. Lett.

■ ACKNOWLEDGMENTS
Dedicated to Prof. E. J. Corey, a true pioneer and inspiring
2003, 5, 3033.
(17) Reviews on Heck reactions using P,N-ligands: (a) Loiseleur, M.;
Hayashi, M.; Schmees, N.; Pfaltz, A. Synthesis 1997, 1997, 1338.
(b) Mc Cartney, D.; Guiry, P. J. Chem. Soc. Rev. 2011, 40, 5122. (c)
mentor. We thank the Molinski and Theodorakis Laboratories General review of the enantioselective intramolecular Heck reaction:
(both of UCSD) for use of analytical instruments used in the Dounay, A. B.; Overman, L. E. Chem. Rev. 2003, 103, 2945.
course of these studies; the Bertrand and Figueroa Laboratories (18) Selected examples: (a) Peterson, P. E.; Stepanian, M. J. Org.
for use of gloveboxes (both of UCSD); the Baran Laboratory Chem. 1988, 53, 1903. (b) Grieco, P. A.; Cowen, S. D.; Mohammadi,
(TSRI) for the use of a microwave reactor; Anthony Mrse F. Tetrahedron Lett. 1996, 37, 2699. (c) Cheng, X.; Khan, N.;
(UCSD) for NMR spectroscopic assistance; Dick Pederson Kumaran, G.; Mootoo, D. R. Org. Lett. 2001, 3, 1323.
(Materia) for a kind donation of alkene metathesis catalysts (19) Myers, A. G.; Tom, N. J.; Fraley, M. E.; Cohen, S. B.; Madar, D.
used in this study; Jeff Elleraas (Pfizer La Jolla) for J. J. Am. Chem. Soc. 1997, 119, 6072.
redetermination of the ee of 14, and Profs. Tadeusz Molinski (20) (a) Hunter, R.; Bartels, B.; Michael, J. P. Tetrahedron Lett. 1991,
32, 1095. (b) Bartels, B.; Hunter, R. J. Org. Chem. 1993, 58, 6756.
and James Hanson (University of Sussex, UK) for helpful
(c) Kim, S.; Do, J. Y.; Kim, S. H.; Kim, D. J. Chem. Soc., Perkin Trans. 1
discussion regarding optical rotation. We also thank Teresa Ng 1994, 2357. (d) Federspiel, M.; Fischer, R.; Hennig, M.; Mair, H.-J.;
(UCSD) for early technical assistance. We are also especially Oberhauser, T.; Rimmler, G.; Albiez, T.; Bruhin, J.; Estermann, H.;
indebted to Prof. Erik J. Alexanian (University of North Gandert, C.; Göckel, V.; Götzö, S.; Hoffmann, U.; Huber, G.; Janatsch,
Carolina, Chapel Hill) for granting us permission to use G.; Lauper, S.; Röckel-Stabler, O.; Trussardi, R.; Zwahlen, A. G. Org.
portions of his Ph.D. thesis for our Supporting Information. We Process Res. Dev. 1999, 3, 266.
acknowledge University of California, San Diego, start-up (21) As judged by crude reaction mixture analysis by 1H NMR.
funding.

■ REFERENCES
(1) Brian, P. W.; Curtis, P. J.; Hemming, H. G.; Norris, G. L. F.
Trans. Br. Mycol. Soc. 1957, 40, 365.
(2) (a) Viridin isolation: Brian, P. W.; McGowan, J. C. Nature 1945,
156, 144. (b) Grove, J. F.; McCloskey, P.; Moffatt, J. S. J. Chem. Soc. C
1966, 743, 743. (c) Viridiol isolation and structure elucidation:
Moffatt, J. S.; Bu’Lock, J. D.; Yuen, T. H. J. Chem. Soc. D 1969, 839.
(3) Review: Hanson, J. R. Nat. Prod. Rep. 1995, 12, 381.
(4) (a) Arcaro, A.; Wymann, M. P. Biochem. J. 1993, 296, 297.
(b) Yano, H.; Nakanishi, S.; Kimura, K.; Hanai, N.; Saitoh, Y.; Fukui,
Y.; Nonomura, Y.; Matsuda, Y. J. Biol. Chem. 1993, 268, 25846.
(c) Powis, G.; Bonjouklian, R.; Berggren, M. M.; Gallegos, A.;
Abraham, R.; Ashendel, C.; Zalkow, L.; Matter, W. F.; Dodge, J.;
Grindey, G.; Vlahos, C. J. Cancer Res. 1994, 54, 2419.
(5) Review: Bader, A. G.; Kang, S.; Zhao, L.; Vogt, P. K. Nat. Rev.
Cancer 2005, 5, 921.
(6) (a) Norman, B. H.; Shih, C.; Toth, J. E.; Ray, J. E.; Dodge, J. A.;
Johnson, D. W.; Rutherford, P. G.; Schultz, R. M.; Worzalla, J. F.;
Vlahos, C. J. J. Med. Chem. 1996, 39, 1106. (b) Creemer, L. C.; Kirst,
H. A.; Vlahos, C. J.; Schultz, R. M. J. Med. Chem. 1996, 39, 5021. (c)
For a much lengthier and inclusive list of references on the medicinal
chemistry of wortmannin, see the Supporting Information..
(7) Wipf, P.; Halter, R. J. Org. Biomol. Chem. 2005, 3, 2053.
(8) (a) Wipf, P.; Minion, D. J.; Halter, R. J.; Berggren, M. I.; Ho, C.
B.; Chiang, G. G.; Kirkpatrick, L.; Abraham, R.; Powis, G. Org. Biomol.
Chem. 2004, 2, 1911. (b) Ihle, N. T.; Williams, R.; Chow, S.; Chew,
W.; Berggren, M. I.; Paine-Murrieta, G.; Minion, D. J.; Halter, R. J.;
Wipf, P.; Abraham, R.; Kirkpatrick, L.; Powis, G. Mol. Cancer Ther.
2004, 3, 763.
(9) For a full listing of synthesis studies toward furanosteroids
beyond these references, see the Supporting Information.
(10) Semisynthesis: Sato, S.; Nakada, M.; Shibasaki, M. Tetrahedron
Lett. 1996, 37, 6141. Total synthesis: Mizutani, T.; Honzawa, S.;

6822 DOI: 10.1021/jacs.7b02829

J. Am. Chem. Soc. 2017, 139, 6819−6822