A Convergent Synthesis of the Tetrasaccharide Fragment of the Purported Structure of Durantanin I

Article abstract of DOI:10.1002/bkcs.12265

A convergent synthesis of the tetrasaccharide subunit in the proposed structure of durantanin I is reported. The signature step is represented by the unique assembly of apiofuranoside ring by the sequential Pd-Ru metal catalysis. Per-dihydroxylation at the late stage delivered the target compound in a highly efficient manner. In addition, a tetrasaccharide derivative possessing unnatural apiose unit was also synthesized with comparable efficiency to that for the natural form.

Full text of DOI:10.1002/bkcs.12265

Communication
DOI: 10.1002/bkcs.12265
BULLETIN OF THE
K. Lee et al.
KOREAN CHEMICAL SOCIETY
A Convergent Synthesis of the Tetrasaccharide Fragment
of the Purported Structure of Durantanin I
*
Keehwan Lee, Mijin Kim, and Young Ho Rhee
Department of Chemistry, Pohang University of Science and Technology, Kyungbuk 37673, Republic of
Korea. *E-mail: yhrhee@postech.ac.kr
Received January 21, 2021, Accepted March 1, 2021, Published online March 26, 2021
A convergent synthesis of the tetrasaccharide subunit in the proposed structure of durantanin I is reported.
The signature step is represented by the unique assembly of apiofuranoside ring by the sequential Pd Ru
metal catalysis. Per-dihydroxylation at the late stage delivered the target compound in a highly efficient
manner. In addition, a tetrasaccharide derivative possessing unnatural apiose unit was also synthesized
with comparable efficiency to that for the natural form.
Keywords: Durantanin, Oligosaccharide, Apiose, Total synthesis, Hydroalkoxylation
Triterpenoid saponins are found in a wide variety of dicoty-
ledonous plants as a constituents of the cell membrane. Their
diverse structures and unique biological/pharmacological
activities have drawn considerable attention not only from
the field of natural product chemistry but also from that of
synthetic organic chemistry.1 Durantanin I, a member of
triterpenoid-type saponin, was isolated from the leaves of
Duranta repens by Hiradate and co-workers.2,3 This com-
pound exhibits significant plant growth inhibitory activities.
Its structure was elucidated as polygalacic acid-3-O-β-D-
glucopyranoside combined with unique tetrasaccharide
subunit consisting of 28-O-[α-L-rhamnopyranosyl-(1 ! 30)-
β-D-apiofuranosyl-(1 ! 4)-α-L-rhamnopyranosyl-(1 ! 2)-
α-L-arabinopyranoside] (Scheme 1). This tetrasaccharide
moiety bearing an apiofuranose residue poses a great syn-
thetic challenge. However, access to this fragment has
remained unknown, despite previous studies on the prepa-
ration of various apiofuranose glycosides.4 Here, we wish
to report a first synthesis of the proposed structure for this
tetrasaccharide moiety. A key event involves palladium-
catalyzed asymmetric intermolecular hydroalkoxylation of
alcohol nucleophile in combination with the ring-closing-
metathesis that assembles the apiofuranose unit in a highly
efficient manner. 5–7
On the basis of our own experience in the de novo
β-apiofuranoside synthesis,6 we envisaged that the
tetrasaccharide unit can be constructed from the tetraene pre-
cursor 1 by the late-stage per-dihydroxylation.8 The proposed
substrate-driven stereoselectivity of the dihydroxylation of
BCD ring is verified by the related studies.6,8 In addition, we
were confident that the stereoselectivity of the unprecedented
dihydroxylation of the ring A may be effectively controlled
by the cis-1,2-bis-alkoxy groups installed in the ring A. The
intermediate 1 can be prepared in a convergent manner from
the allylic alcohol 2 and alkoxyallene 3 by way of the Pd Ru
sequential metal catalysis discussed above.6 Notably, a skele-
ton of the β-D-apiofuranoside (ring C) is welded by this trans-
formation. Both of these intermediates can be readily derived
from commercially available 3,4-di-O-acetyl-6-deoxy-L-
glucal 4 by Lewis acid-mediated coupling reaction (Ferrier
reaction).9 Of the two reactions, construction of the AB ring
seemed to be more challenging because it generates a disac-
charide component. Thus, we decided to explore first the
assembly of 3.
The initial stage of the work commenced with the asym-
metric synthesis of the cyclic benzylic acetal intermediate
10 (Scheme 2). For example, commercially available alco-
hol 7 (1 equiv) was combined with benzyloxyallene
8 (2 equiv) in the presence of Pd2(dba)3 (2.5 mol %), ligand
(S,S)-L (5 mol %) and catalytic amount of Et3N (0.1 equiv)
in CH2Cl2 at 20 ꢀC. As depicted in Scheme 1, this reaction
successfully produced acyclic acetal 9 in ~99% yield. Sub-
sequent ring-closing-metathesis (RCM) reaction employing
first generation Grubbs catalyst at 40 ꢀC gave cyclic acetal
10 in 86% yield and 95% ee (for the determination of ee
and absolute configuration, see the SI). At this point, we
reasoned that compound 6 could be obtained from 10 by
way of syn-epoxide formation (compound 11) and the ensu-
ing base-mediated opening reaction. As anticipated, initial
efforts toward direct epoxidation (such as m-
chloroperbenzoic acid or dimethyldioxirane) provided the
isomeric anti-product as the major product. This unsuccess-
ful preliminary result led us to consider a well-known pro-
tocol mediated by the bromohydrin formation.10 Indeed,
reaction of 10 with N-bromosuccinimide (2.5 equiv) in
dimethylsulfoxide/H2O and the subsequent addition of NaH
(1.5 equiv) generated the desired syn-epoxide 11 in 69%
(over two steps). Treatment of this compound with strong
base (such as t-BuLi) failed to produce 6 in a reproducible
manner. Thus, we decided to rely on an indirect method
†These authors contributed equally to this work.
Bull. Korean Chem. Soc. 2021, Vol. 42, 679–682
© 2021 Korean Chemical Society, Seoul & Wiley-VCH GmbH
Wiley Online Library
679

Communication
ISSN (Print) 0253-2964 | (Online) 1229-5949
BULLETIN OF THE
KOREAN CHEMICAL SOCIETY
producing allylic alcohol from the epoxide precursor.11 Fol-
lowing a related protocol, epoxide 11 was first treated with
(PhSe)2 (0.8 equiv) and NaBH4 (1.6 equiv) in EtOH. Sub-
sequent oxidation with H2O2 under reflux successfully gave
allylic alcohol 6 in 75% yield (over two steps).12 Conver-
sion of the alcohol into the benzyl ether and the subsequent
Os-catalyzed dihydroxylation reaction produced cis-1,-
2-arabinopyranoside 60 in 81% yield (over two steps) as a
single diastereomer, as anticipated. The relative and abso-
lute stereochemistry of this compound were rigorously
established by the comparison of the spectral data with the
literature values (see the SI).13
With allylic alcohol 6 in hand, preparation of disaccha-
ride (AB ring) was then investigated. Initial efforts using
strong Lewis acid-mediated Ferrier-type glycosylation
showed only formation of untractable mixture of com-
pounds, presumably due to the instability of the anomeric
centers in 12 (Scheme 3). Notably, employing
Pd(MeCN)2Cl2 (10 mol %) catalyst based upon Galan’s
work14 generated 12, albeit in low ~20% yield. Increasing
the catalyst loading (30 mol %) under diluted condition in
CH2Cl2 delivered 12 in significantly higher 84% yield as
an inseparable mixture of two anomers (α:β = 8:1).15 Sub-
sequent deacetylation using catalytic K2CO3 (0.2 equiv)
proceeded uneventfully to produce 13 in diastereomerically
pure form in 81% yield after column chromatography.
From this allylic alcohol, alkoxyallene 3 was prepared via a
two-step event combining propargylation and base-
catalyzed isomerization16 in 68% yield (over two steps).
Unlike 6, achiral allylic alcohol 5 was smoothly converted
to the corresponding acetal 14 in 80% yield by the Ferrier-
type glycosylation using sub-stoichiometric amount of
BF3ꢁEt2O (0.4 equiv). Deacetylation of this compound pro-
vided alcohol 15 in 87% yield. Introduction of O-benzyl
group using NaH (1.2 equiv) and BnBr (1.2 equiv) followed
by desilylation reaction employing TBAF (1.2 equiv) gener-
ated alcohol substrate 2 in 80% yield (over two steps,
Scheme 4)
Scheme 1. Retrosynthetic analysis of durantanin I.
Scheme 2. Synthesis of intermediate 6.
Having secured the coupling partners 2 and 3, we exam-
ined the key sequential catalysis that assembles
furanoglycoside (ring C). As depicted in Scheme 5, acyclic
acetal 16a was obtained from 2 and 3 in 94% yield by
employing Pd2(dba)3 (2.5 mol %) and (R,R)-L (5 mol%) at
40 ꢀC. The subsequent RCM reaction employing first genera-
tion Grubbs catalyst (10 mol %) at 40 ꢀC gave the desired
tetrasaccharide 1 as the exclusive diastereomer in 95% yield
(dr = 24:1).17 Final per-dihydroxylation reaction of this com-
pound using OsO4 (5 mol %) and N-methylmorpholine oxide
(NMO) (13 equiv) produced octa-ol 17a in 81% yield as a
single diastereomer. Due to the troublesome separation from
residual NMO, this compound was converted into the per-
acetyl derivative 18a. This task was accomplished in 75%
yield by the treatment of 17a with Et3N (16 equiv), Ac2O
(10 equiv) in the presence of N,N-dimethylaminopyridine
(1 equiv). Thus, the target compound 18a was obtained from
benzyloxyallene 2 in 14 (longest linear) steps. Remarkably,
Scheme 3. Preparation of the alkoxyallene 3.
Bull. Korean Chem. Soc. 2021, Vol. 42, 679–682
© 2021 Korean Chemical Society, Seoul & Wiley-VCH GmbH
www.bkcs.wiley-vch.de
680

BULLETIN OF THE
Communication
KOREAN CHEMICAL SOCIETY
Scheme 6. Convergent synthesis of diastereomeric tetrasaccharide.
Scheme 4. Preparation of the allylic alcohol 2.
In summary, the first synthetic route to the tetra-
saccharide subunit of durantanin I and its diastereomer was
devised. The key reaction utilizes the Pd-catalyzed hydro-
alkoxylation of alcohol followed by Ru-catalyzed ring-closing-
metathesis reaction. Currently, we are working on the total
synthesis of durantanin I as well as expanding the utility of the
sequential metal catalysis for the synthesis of other saponin
oligosaccharides.
Acknowledgments. Financial support for this work was
provided by the National Research Foundation of Korea,
which is funded by the Korean Government (NRF-
2015R1A2A1A15056116 and NRF-2012M3A7B4049645).
Supporting Information. Additional supporting informa-
tion may be found online in the Supporting Information
section at the end of the article.
References
Scheme 5. Convergent synthesis of the tetrasaccharide.
1. For selected reviews on triterpenoid saponin, see(a)B. Dinda,
S. Debnath, B. C. Mohanta, Y. Harigaya, Chem. Biodiversity
2010, 7, 2327. (b)V. R. Netala, S. B. Ghosh, P. Bobbu,
D. Anitha, V. Tartte, Int. J. Pharm. Pharm. Sci. 2015, 7, 24.
2. For more information about Durantanin I and related natural
products, see(a)S. Hiradate, H. Yada, T. Ishii, N. Nakajima,
M. Ohnishi-Kameyama, H. Sugie, S. Zungsontiporn, Y. Fujii,
Phytochemistry 1999, 52, 1223. (b)C. Furusawa, S. Yasuda,
H. Tsuji, S. Ito, H. Miyashita, H. Yoshimitsu, T. Nohara,
M. Ono, Nat. Prod. Res. 2016, 30, 246.
3. For selected examples of other apiose-containing oligosaccha-
ride natural products, see(a)Y.-L. Song, G.-S. Zhou, S.-X.
Zhou, Y. Jiang, P.-F. Tu, D.-G. Polygalins, Nat. Prod. Res
2013, 27, 1220. (b) J. W. Chung, E. J. Noh, H. L. Zhao, J.-S.
Sim, Y. W. Ha, E. M. Shin, E. B. Lee, C. S. Cheong, Y. S.
Kim, Biol. Pharm. Bull. 2008, 31, 2114. (c) G. Ragupathi,
P. Damani, K. Deng, M. M. Adams, J. Hang, C. George,
P. O. Livingstone, D. Y. Gin, Vaccine 2010, 28, 4260.
the chemoselective metal catalysis allowed for the facile syn-
thesis of potentially labile tetraene 1, which was per-
dihydroxylated at the late stage to deliver the desired target
octa-ol 17a in a highly efficient manner.
In addition to the natural form 17a, its unnatural analog
17b possessing β-L-apiofuranoglycoside was also easily
synthesized by simply changing the ligand to the enantio-
meric (S,S)-L form for the palladium-catalyzed hydro-
alkoxylation reaction (Scheme 6). This reaction required
somewhat higher catalyst loading (10 mol %) for the com-
plete conversion. Nevertheless, the desired compound 10
was obtained in pure form after RCM reaction in 90% yield
over two steps (dr = 1:13).17 The subsequent per-
dihydroxylation with OsO4 (5 mol %) and NMO (13 equiv)
gave octa-ol 17b in 83% yield, which was also converted
into the per-acetyl derivative 18b in 70% yield. This facile
synthesis of the diastereomeric form illustrates another ben-
eficial feature of the proposed synthesis, which should find
further use for the derivatization of the natural products.
4. For selected examples for the synthesis of apiose-containing
oligosaccharides, see (a) S. A. Nepogodiev, M. Fais, D. L.
Hughes, R. A. Field, Org. Biomol. Chem 2011, 9, 6670.
(b) M. Kojima, Y. Nakamura, K. Komori, S. Akai, K. Sato,
S. Takeuchi, Tetrahedron 2011, 67, 8276. (c) Y. Zhang,
Bull. Korean Chem. Soc. 2021, Vol. 42, 679–682
© 2021 Korean Chemical Society, Seoul & Wiley-VCH GmbH
www.bkcs.wiley-vch.de
681

BULLETIN OF THE
Communication
KOREAN CHEMICAL SOCIETY
K. Wang, Z. Zhan, Y. Yang, Y. Zhao, Tetrahedron Lett.
2011, 52, 3154.
9. (a) R. J. Ferrier, J. Chem. Soc. 1964, 5443. (b) For a recent
review on Ferrier rearrangement, seeA. M. Gómez, F. Lobo,
C. Uriel, J. C. López, Eur. J. Org. Chem 2013, 7221.
10. M. V. Rao, M. Nagarajan, J. Org. Chem. 1988, 53, 1184.
11. K. S. Kim, C. W. Moon, J. I. Park, S.-H. Han, J. Chem. Soc.,
Perkin Trans. 2000, 1, 1341.
12. K. B. Sharpless, R. F. Lauer, J. Am. Chem. Soc. 1973, 95, 2697.
13. J. Rujirawanich, B. Kongkathip, N. Kongkathip, Carbohydr.
Res. 2011, 346, 927.
5. For selected recent examples on our contribution in this area,
see(a)J. Lee, J. Kang, Y. H. Rhee, Angew. Chem. Int. Ed 2020,
59, 2349. (b) J. Lee, S. Kang, J. Kim, D. Moon, Y. H. Rhee,
Angew. Chem. Int. Ed. 2019, 58, 628. (c) S. Lee, Y. H. Rhee,
J. Org. Chem. 2019, 84, 9353. (d) S. H. Jang, H. W. Kim, W.
Jeong, D. Moon, Y. H. Rhee, Org. Lett. 2018, 20, 1248. (e) W.
Lim, J. Kim, Y. H. Rhee, J. Am. Chem. Soc. 2014, 136, 13618.
6. For our own approach towards apiose-containing oligosaccha-
ride, seeM. Kim, S. Kang, Y. H. Rhee, Angew. Chem. Int. Ed
2016, 55, 9733.
7. For selected examples on the use of metal catalysed reactions
in the oligosaccharide synthesis(a)E. U. Sharif, H.-Y. L.
Wang, N. G. Akhmedov, G. A. O’Doherty, Org. Lett 2014,
16, 492. (b) B. Yu, Acc. Chem. Res. 2018, 51, 507. (c) H.
Kim, H. Men, C. Lee, J. Am. Chem. Soc. 2004, 126, 1336.
(d) H.-Y. Wang, K. Yang, D. Yin, C. Liu, D. A. Glazier,
W. Tang, Org. Lett. 2015, 17, 5272.
14. (a) A. Sau, M. C. Galan, Org. Lett. 2017, 19, 2857. (b) A.
Sau, R. Williams, C. Palo-Nieto, A. Franconetti, S. Medina,
M. C. Galan, Angew. Chem. Int. Ed. 2017, 56, 3640.
1
15. The dr of the compound 12 was determined by H NMR. See
SI for the detailed information.
16. S. Hoff, L. Brandsma, J. F. Arens, Recl. Trav. Chim. Pays-
Bas 1968, 87, 916.
17. The dr of the hydroalkoxylation reaction using (R,R)-L and
1
(S,S)-L was determined after RCM reaction by crude H NMR
of the compound 1 and 10. For the detailed analysis, see SI.
8. R. S. Babu, M. Zhou, G. A. O’Doherty, J. Am. Chem. Soc.
2004, 126, 3428.
Bull. Korean Chem. Soc. 2021, Vol. 42, 679–682
© 2021 Korean Chemical Society, Seoul & Wiley-VCH GmbH
www.bkcs.wiley-vch.de
682