Total Synthesis of Alvaradoins e and F, Uveoside, and 10-epi-Uveoside

Article abstract of DOI:10.1021/acs.orglett.9b03546

Concise total syntheses of the anthracenone C-glycosides alvaradoins E and F, uveoside, and 10-epi-uveoside (1-4) have been accomplished from chrysophanic acid 8 and bromosugar 9. Key steps in the syntheses include the DBU-induced coupling of 8 and 9 to p

Full text of DOI:10.1021/acs.orglett.9b03546

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Total Synthesis of Alvaradoins E and F, Uveoside, and 10-epi-
Uveoside
Kevin Ng, Ryan Shaktah, Laura Vardanyan, and Thomas G. Minehan*
Department of Chemistry and Biochemistry, California State University, Northridge, 18111 Nordhoff Street, Northridge, California
91330-8262, United States
S
* Supporting Information
ABSTRACT: Concise total syntheses of the anthracenone C-glycosides alvaradoins E and F, uveoside, and 10-epi-uveoside
(1−4) have been accomplished from chrysophanic acid 8 and bromosugar 9. Key steps in the syntheses include the DBU-
induced coupling of 8 and 9 to produce β-C-glycoside 11, and a Pb(OAc)4-mediated Kochi reaction to introduce the C-1′
oxygen atom of the natural products. Isothermal titration calorimetry and fluorescence binding studies reveal that compounds 1
and 2 have good affinity for the plasma protein HSA.
he anthracenone C-glycosides alvaradoin E (1, Figure 1)
and alvaradoin F (2, Figure 1) were isolated from the
the investigators to evaluate the in vivo activity of 1 and 2 in
the P388 murine lymphocytic leukemia model. Alvaradoin E
showed antileukemic activity (125% T/C) at a dose of 0.2 mg/
kg per injection when administered intraperitoneally. Uveoside
(3, Figure 1) was isolated in 1998 from the chloroform extract
of the roots of Picramnia antidesma by Hernandez-Medel and
co-workers;2 further work on the root bark of Picramnia
antidesma by the same research group resulted in the isolation
(in 2002) of 10-epi-uveoside (4, Figure 1), a substance also
displaying elevated cytotoxicity toward KB cells.3 Given the
heightened biological profile of this family of C-glycoside
natural products, together with the fact that there is only a
single previous total synthesis of a related anthrone C-
glycoside,4 we decided to undertake a synthetic study of
compounds 1−4.
T
One of the synthetic challenges anticipated en route to
compounds 1−4 was the installation of the acid-labile
anomeric C-1′ acetate and benzoate esters, a structural feature
absent in the previously prepared anthrone glycoside
cassialoin.4 We envisioned that a Hunsdiecker-type reaction5
on a carboxylic acid precursor would allow us to install the C-
1′ oxygen atom in the form of a more stable acetal moiety,
which could be transformed into the requisite C-1′ ester in the
penultimate step of the synthesis. The realization of this plan is
detailed in the present Letter.
Figure 1. Chemical structures of alvaradoins E and F, uveoside, and
10-epi-uveoside.
Over the last two decades, there have been numerous
advances in C-glycoside synthesis.6 Transition-metal catalyzed
cross-coupling7 and metathesis8 reactions, [3,3]-sigmatropic
leaves of the tropical tree Alvaradoa haitiensis in 2005 and
2007.1 Both substances exhibited pronounced cytotoxicities
toward human oral epidermoid carcinoma (KB) cell lines
(EC50(1) = 0.050 μM; EC50(2) = 0.065 μM) among others;
furthermore, alvaradoin E was demonstrated to induce
apoptosis of cultured LNCaP cells. These results prompted
Received: October 7, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b03546
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
rearrangements,9 and radical-olefin couplings10 have been used
to prepare arene- and C(sp3)-linked C-glycosides. Since it is
known that anthracenones undergo base-induced alkylation
reactions at C-10,11 we decided to attempt a direct C-glycoside
synthesis by combining anthracenones with C-1 bromosugars
under basic conditions. Our initial studies employed anthralin
(5, Scheme 1) as the nucleophile and bromoglucoside 6b12 as
direction, late-stage removal of these esters under acidic
conditions proved to be problematic (vide infra) and
ultimately motivated a switch to the less bulky isobutyrate
esters, which were installed by treatment of 11 with isobutyryl
chloride and triethylamine in CH2Cl2 (Scheme 2). Meth-
Scheme 2. Transformation of Glycoside 11 into Acetate 14
Scheme 1. C-Glycosylation of Anthralin (5) and
Chrysophanol (8)
anolysis of the primary acetate was then achieved by treatment
with 5% HCl in methanol, affording the primary alcohol 12 in
77% overall yield. TEMPO-catalyzed oxidation to the
carboxylic acid in 76% yield was then achieved under Zhao’s
conditions, affording 13.15
After an extensive survey of conditions for achieving a
Hunsdiecker-type conversion5 of acid 13 to the corresponding
glycosyl halide, we discovered that Kochi’s protocol16a
involving treatment of 13 with lead tetraacetate in acetic acid
and THF leads to a clean and stereoselective conversion to α-
acetate ester 14,16b albeit in moderate yields (40%). However,
the starting material could be efficiently recovered in good
yields (51%) from this reaction and recycled to increase overall
throughput to compound 14.
With acetate 14 in hand, we next attempted hydrolysis of the
isobutyrate and acetate esters. Exposure of 14 to basic
conditions (NaOMe in MeOH; cat. NaCN, MeOH; RNH2,
MeOH) resulted in substrate decomposition with the
production of anthraquinone byproducts; in addition, the
recovered C-glycosides possessed a mixture of stereoisomers at
C-5′ (Scheme 3). Exposure of 14 to acidic conditions (5−20%
HCl in MeOH) instead resulted in elimination of the C-1′
acetate group and the formation of alkene-containing by-
products. After extensive experimentation, it was found that
solvolysis of the acetate with allyl alcohol could be achieved in
the presence of BF3·OEt2; subsequent treatment of the
resulting allyl glycoside with 20% HCl in MeOH at 40 °C
gave a 73% overall yield of diol 15. Exposure of 15 to 10 mol %
PdCl2 in MeOH and THF then resulted in the clean formation
of α-configured hemiacetal 16 in 86% yield.17
the carbohydrate electrophile for the substitution reaction.
Treatment of an equimolar CH2Cl2 solution of 6b and 5 with 1
equiv of DBU resulted in a rapid disappearance of 6b, along
with the formation of C-glycoside 7 on TLC. Upon isolation, it
was found that compound 7 was formed in 42% yield as a 7:1
mixture of β and α stereoisomers at C-5′. Encouraged by this
result, we prepared the bromogalactoside 9b13 by HBr/AcOH
treatment of 1,6-di-O-acetyl-2,3,4-tribenzyl-galactose 9a. Com-
bination of 9b with 5 in the presence of one equivalent of DBU
gave rise to a 65% yield (from 6a) of C-glycoside 10 as a single
β-stereoisomer (>20:1 β:α) at C-5′. Analogously, treatment of
a mixture of 9b and chrysophanol (8) with DBU provided a
51% yield (from 9a) of C-glycoside 11, again with the selective
production (>20:1 β:α) of the C-5′ β stereoisomer in excess;
in addition, a 1:1 mixture of diastereomers was formed at C-
10.14 TLC also showed the formation of oxidized aromatics
(anthraquinones) as well as hydrolyzed carbohydrates (6 or 9,
R2 = OH), accounting for the balance of the material in these
reactions. All attempts to thwart their production by
performing the reaction under vigorously anhydrous and
anaerobic conditions led to no significant improvement in the
yields of 7, 10, or 11 obtained.
In order to manipulate the hydroxymethyl group of the
carbohydrate moiety to achieve installation of the C-1′ oxygen
atom, it was necessary to protect the C-1 and C-8 hydroxyl
groups as sterically bulky esters that would be resistant to
conditions for hydrolysis of the acetate ester of 11. While the
bis-pivaloate derivative initially showed promise in this
For the preparation of alvaradoins E and F, installation of
the acetate moiety at C-1′ was required. Compound 16 was
B
DOI: 10.1021/acs.orglett.9b03546
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
benzoic acid in CH2Cl2, gave rise to α-benzoate 18 in 76%
yield. Hydrogenation then provided uveoside and 10-epi-
uveoside in 90% yield as a 1:1 mixture of diastereomers, which
could be separated by radial chromatography. Once again,
NMR (1H and 13C), MS, and optical rotation data recorded for
synthetic uveoside and 10-epi-uveoside were in accord with
those reported for the natural compounds.29
Scheme 3. Preparation of Hemiacetal 16
Given that numerous C-aryl glycosides are known to form
strong complexes with duplex nucleic acids,20 we undertook
DNA binding studies with synthetic compounds 1−4 (see
Supporting Information). Thermal denaturation studies21 with
CT DNA were precluded by the fact that compounds 1−4
underwent decomposition in aqueous phosphate buffer
solution (pH = 7.21) at elevated temperatures, as evidenced
by UV spectroscopy. Therefore, utilizing fluorescence spec-
troscopy, we investigated the displacement of ethidium
bromide (10 μM) from CT DNA (10 μM) by increasing
concentrations of synthetic alvaradoins E and F, and found
only a small effect on ethidium fluorescence emission intensity
(at 590 nm) over the 0.01−100 μM concentration range of 1
and 2 (pH = 7.21, NaH2PO4/Na2HPO4 buffer); indeed, 85%
of the initial fluorescence intensity of ethidium was measured
at 50 μM of the ligands, and extrapolation of the data gave a
C
50 value of 240 μM. Similar results were obtained in ethidium
displacement studies employing uveoside and 10-epi-uveoside
(extrapolated C50 = 430 μM).22 Under otherwise identical
experimental conditions, positive control netropsin produced a
significant decrease in ethidium emission fluorescence intensity
with a directly measurable C50 value of 10 2.5 μM. These
results indicate that, despite structural similarities to the C-aryl
glycoside family of natural products, the anthracenone C-
glycosides associate only weakly with duplex nucleic acids.
In light of these data, we explored the possibility that this
family of C-glycosides may form complexes with proteins;
indeed, synthetic C-glycosides have been previously shown to
associate with plant-derived carbohydrate binding proteins.23
Titration of the abundantly available plasma protein human
serum albumin24 (HSA; N form at pH 7) with alvaradoins E
and F showed a marked quenching of the Trp 214 fluorescence
emission at 338 nm (λex = 280 nm; Ksv = 2.79 0.33 × 104
exposed to HBr in acetic acid to form the C-1′ glycosyl
bromide, which was immediately combined with silver acetate
in acetic acid to provide α-acetate ester 17 in 73% yield
(Scheme 4).18 Hydrogenation of 17 over Pearlman’s catalyst19
Scheme 4. Completion of the Syntheses of 1−4
M−1; Kb = 6.60
1.42 × 104 M−1; see Supporting
Information).25,26 A similar experiment performed with
uveoside and 10-epi-uveoside revealed significantly less
quenching of the Trp 214 fluorescence emission as compared
to the alvaradoins and a notably weaker binding (Ksv = 6.11
0.55 × 103 M−1; Kb = 0.88 0.36 × 103 M−1). Isothermal
titration calorimetry27 (ITC) of HSA (57 μM) with
alvaradoins E and F (500 μM solution, 25 °C, pH = 7.21,
NaH2PO4/Na2HPO4 buffer) gave a binding constant Kb of
3.01 0.58 × 104 M−1, along with an enthalpy of binding, ΔH,
of −11.7 0.5 kcal/mol and an entropy of binding, ΔS, of
−18.8 2.2 cal/mol·K; similarly, ITC experiments with the
uveosides showed only weak binding (Kb < 1 × 103 M−1). HSA
competition binding experiments employing warfarin (site I
marker) or ibuprofen (site II marker) and the alvaradoins are
currently underway to clarify which binding site (Sudlow site I
or Sudlow site 2) on the protein is preferred by the
anthacenone-C-glycosides.28
then afforded a 95% yield of alvaradoins E and F as a 1:1
mixture of diastereomers, which could be separated by careful
and repeated column chromatography (SiO2, 98:2 → 94:6
CHCl3/MeOH). NMR (1H and 13C),29 mass spectroscopy
(MS), and optical rotation data recorded for synthetic
alvaradoins E and F were in accord with those reported for
the natural compounds. Similarly, exposure of 16 to HBr in
acetic acid, followed by treatment with silver benzoate and
In summary, we have developed short total syntheses of the
anthracenone-C-glycosides alvaradoins E and F, uveoside, and
10-epi-uveoside. Although these compounds display poor
binding to duplex DNA, alvaradoins E and F show significant
affinity for the plasma protein human serum albumin.
C
DOI: 10.1021/acs.orglett.9b03546
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
coupling constants of 1.6 and 9.4 Hz; the C-10 proton resonance of
10 at 3.69 ppm (doublet) displayed a coupling constant of 2.0 Hz.
These data are indicative of the β-configuration at C-5′; see ref 16b
and Procko, K. J.; Li, H.; Martin, S. F. Org. Lett. 2010, 12, 5632.
(b) We attempted chromatographic separation (SiO2) of the C-10
diastereomers of compound 11 (as well as later compounds 17 and
18) but were unsuccessful. Since the diastereomers lead to different
natural products, we proceeded with the mixtures until adequate
chromatographic resolution could be achieved at the natural product
stage, as indicated in the original isolation papers (ref 1).
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.9b03546.
■
S
Detailed experimental procedures including spectro-
scopic and analytical data (PDF)
AUTHOR INFORMATION
(15) Zhao, M.; Li, J.; Mano, E.; Song, Z.; Tschaen, D. M.;
Grabowski, E. J. J.; Reider, P. J. J. Org. Chem. 1999, 64, 2564.
(16) (a) Sheldon, R. A.; Kochi, J. K. Org. React. 1972, 19, 279. (b)
■
Corresponding Author
*E-mail: thomas.minehan@csun.edu.
ORCID
1
The C-1′ proton resonance in the H NMR spectrum of compound
14 (CDCl3) at 5.98 ppm (doublet) displayed a coupling constant of
2.4 Hz; C.1′ coupling constants of a similar magnitude were measured
for compounds 17 and 18. These data are indicative of the α-
configuration at C.1′. See: Karplus, M. J. Am. Chem. Soc. 1963, 85,
2870.
Kevin Ng: 0000-0002-2317-9555
Thomas G. Minehan: 0000-0002-7791-0091
Notes
(17) Lohman, G. J. S.; Seeberger, P. H. J. Org. Chem. 2004, 69, 4081.
(18) (a) Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503.
(b) Schmidt, R. R.; Castro-Palomino, J. J.; Retz, O. Pure Appl. Chem.
1999, 71, 729.
(19) Conway, S. J.; Gardiner, J.; Grove, S. J. A.; Johns, M. K.; Lim,
Z.-Y.; Painter, G. F.; Robinson, D. E. J. E.; Schreiber, C.; Thuring, J.
W.; Wong, L. S.-M.; Yin, M.-X.; Burgess, A. W.; Catimel, B.; Hawkins,
P. T.; Ktistakis, N. T.; Stephens, L. R.; Holmes, A. B. Org. Biomol.
Chem. 2010, 8, 66.
(20) Hacksell, U.; Daves, G. D. Prog. Med. Chem. 1985, 22, 1.
(21) Wilson, W. D.; Tanious, F. A.; Fernandez-Saiz, M.; Rigl, C. T.
Methods in Mol. Biol. Drug-DNA Interact. Protoc. 1997, 90, 219.
(22) Morgan, A. R.; Lee, J. S.; Pulleyblank, D. E.; Murray, N. L.;
Evans, D. H. Nucleic Acids Res. 1979, 7, 547.
(23) Wei, A.; Boy, K. M.; Kishi, Y. J. Am. Chem. Soc. 1995, 117,
9432.
(24) Liu, C.; Liu, Z.; Wang, J. PLoS One 2017, 12, No. e0176208.
(25) Chatterjee, T.; Pal, A.; Dey, S.; Chatterjee, B. K.; Chakrabati, P.
PLoS One 2012, 7, No. e37468.
(26) Na, N.; Zhao, D.-Q.; Li, H.; Jiang, N.; Wen, J.-Y.; Liu, H.-Y.
Molecules 2016, 21, 54.
(27) Callies, O.; Hernandez Daranas, A. Nat. Prod. Rep. 2016, 33,
881.
(28) Sudlow, G. D. J. B.; Birkett, D. J.; Wade, D. N. Mol. Pharmacol.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the National Science Foundation (CHE-
1508070) and the donors of the American Chemical Society
Petroleum Research Fund (53693-URI) for their generous
support of this research. We thank the UCR mass spectrometry
facility for accurate mass determinations.
REFERENCES
■
(1) (a) Mi, Q.; Lantvit, D.; Reyes-Lim, E.; Chai, H.; Phifer, S. S.;
Wani, M. C.; Wall, M. E.; Tan, G. T.; Cordell, G. A.; Farnsworth, N.
R.; Kinghorn, A. D.; Pezzuto, J. M. Anticancer Res. 2005, 25, 779.
(b) Phifer, S. S.; Lee, D.; Seo, E.-K.; Kin, N.-C.; Graf, T. N.; Kroll, D.
J.; Navarro, H. A.; Izydore, R. A.; Jimenez, F.; Garcia, R.; Rose, W. C.;
Fairchild, C. R.; Wild, R.; Soejart, D. D.; Farnsworth, N. R.; Kinghorn,
D.; Overlies, N. H.; Wall, M. E.; Wani, M. C. J. Nat. Prod. 2007, 70,
954.
(2) Hernandez-Medel, M.; Garcia-Salmones, I.; Snatillan, R.; Trigos,
A. Phytochemistry 1998, 49, 2599.
(3) Hernandez-Medel, M.; Pereda-Miranda, R. Planta Med. 2002,
68, 556.
(4) Koyama, Y.; Yamaguchi, R.; Suzuki, K. Angew. Chem., Int. Ed.
2008, 47, 1084.
(5) Crich, D.; Sasaki, K. The Hunsdiecker and Related Reactions. In
Comprehensive Organic Synthesis; Knochel, P., Molander, G. A., Eds.;
Pergamon Press: Oxford, 2014; Vol. 7, pp 818−836.
(6) (a) Kitamura, K.; Ando, Y.; Matsumoto, T.; Suzuki, K. Chem.
Rev. 2018, 118, 1495. (b) Yang, Y.; Yu, B. Chem. Rev. 2017, 117,
12281.
(7) (a) Zhu, F.; Rodriguez, J.; O’Neill, S.; Walczak, M. A. ACS Cent.
Sci. 2018, 4, 1652. (b) Yi, D.; Zhu, F.; Walczak, M. A. Org. Lett. 2018,
20, 4627. (c) Koester, D. C.; Leibeling, M.; Neufeld, R.; Werz, D. B.
Org. Lett. 2010, 12, 3934. (d) Koester, D. C.; Kriemen, E.; Werz, D. B.
Angew. Chem., Int. Ed. 2013, 52, 2985. (e) Mabit, T.; Siard, A.; Legros,
F.; Guillarme, S.; Martel, A.; Lebreton, J.; Carreaux, F.; Dujardin, G.;
Collet, S. Chem. - Eur. J. 2018, 24, 14069.
1976, 12, 1052.
(29) Though the NMR spectra of the natural products were
originally recorded in (CD3)2CO, we found that significant degrees of
epimerization of the separated compounds 1−4 occurred in this
solvent. In contrast, much less epimerization was observed in CDCl3.
For this reason, and to be consistent, 1H NMR and 13C NMR data for
the separated compounds 1−4 were recorded in CDCl3 (see
Supporting Information).
(8) Postema, M. H. D.; Piper, J. L.; Komandui, V.; Liu, L. Angew.
Chem., Int. Ed. 2004, 43, 2915.
(9) Sodeoka, M.; Hirai, G.; Watanabe, T.; Mayagi, T. Pure Appl.
Chem. 2009, 81, 205.
(10) Kiya, N.; Hidaka, Y.; Usui, K.; Hirai, G. Org. Lett. 2019, 21,
1588.
(11) Schaltegger, A.; Steiger, W. Arch. Pharm. 1986, 319, 575.
(12) Ciuffreda, P.; Ronchetti, F.; Lucio, T. J. Carbohydr. Chem. 1989,
8, 805.
(13) Nashed, M. A.; Laurens, A. Carbohydr. Res. 1976, 51, 65.
1
(14) (a) The C-5′ proton resonance in the H NMR spectrum of
compound 10 (CDCl3) at 3.50 ppm (doublet of doublets) displayed
D
DOI: 10.1021/acs.orglett.9b03546
Org. Lett. XXXX, XXX, XXX−XXX