First total synthesis of the proposed structure of batatin VI

Article abstract of DOI:10.1021/ol4020255

The first total synthesis of batatin VI, an architecturally novel resin glycoside dimer, has been achieved via a convergent [5 + 3] glycosidic coupling approach. An improved protocol for the construction of the key 18-membered macrolactone core using a Keck macrolactonization method was introduced. However, the synthesized compound was not identical to the natural batatin VI.

Full text of DOI:10.1021/ol4020255

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
First Total Synthesis of the Proposed
Structure of Batatin VI
San-Yong Zhu,† Jia-Sheng Huang,† Shan-Shan Zheng,† Kai Zhu,† and Jin-Song Yang*,‡
Key Laboratory of Drug Targeting and Drug Delivery Systems of the Ministry of
Education, Department of Chemistry of Medicinal Natural Products, West China School
of Pharmacy, and State Key Laboratory of Biotherapy, West China Hospital,
Sichuan University, Chengdu 610041, China
yjs@scu.edu.cn
Received June 28, 2013
ABSTRACT
The first total synthesis of batatin VI, an architecturally novel resin glycoside dimer, has been achieved via a convergent [5 þ 3] glycosidic
coupling approach. An improved protocol for the construction of the key 18-membered macrolactone core using a Keck macrolactonization
method was introduced. However, the synthesized compound was not identical to the natural batatin VI.
Plants of the morning glory family (Convolvulaceae)
have been cultivated for over 2000 years and were widely
used in folk medicine for treating various diseases. Che-
mical studies on their glycosidic components have led to
isolation of a large number of resin glycosides, a class of
amphiphilic glycolipids with a characteristic macrolactone
backbone.1 The hydrophobic aglycons usually found in
these secondary metabolites are chiral hydroxyaliphatic
acids mainly including jalapinolic acid (11(S)-hydroxyhex-
adecanoic acid) and convolvulinolic acid (11(S)-hydroxy-
tetradecanoic acid).2 Yet the hydrophilic carbohydrate
sections typically consist of four to six monosaccharides
such as D-glucose, D-fucose, L-rhamnose, and D-quinovose.
These sugar residues are often modified with some fatty
acids. The macrocycle is formed through an intramolecular
esterification between the carboxyl group of aglycon and
one hydroxy group of the complex oligosaccharide chain.
Merremin, the first glycolipid ester-type dimer, was iso-
lated by Noda et al. in 1995.3 To date, several members of
this type of compound have been identified.3ꢀ7 They feature
intriguing structures that are composed of two units of the
same resin glycosidic acid. The carboxyl group of aglycon
in the acyclic unit esterifies with one hydroxy group of the
macrocyclic unit to form the dimer. Bioassays disclose that
these dimeric resin glycosides possess a broad spectrum of
activities such as increasing the release of γ-aminobutyric
acid (GABA) and glutamic acid,5 a high vasorelaxant
effect,5 and inhibition of multidrug efflux pumps.7b
† West China School of Pharmacy.
‡ West China Hospital.
(1) For reviews, see: (a) Lu, S.-F.; Guo, Z.-W.; O’Yang, Q.-Q.; Hui,
Y.-Z. Youji Huaxue 1997, 17, 488. (b) Pereda-Miranda, R.; Bah, M.
Curr. Top. Med. Chem. 2003, 3, 111. (c) Pereda-Miranda, R.; Rosas-
(3) Noda, N.; Tsuji, K.; Kawasaki, T.; Miyahara, K.; Hanazono, H.;
Yang, C.-R. Chem. Pharm. Bull. 1995, 43, 1061.
(4) Bah, M.; Pereda-Miranda, R. Tetrahedron 1997, 53, 9007.
~ ꢀ
Ramırez, D.; Castaneda-Gomez, J. In Progress in the Chemistry of
´
Organic Natural Products, Vol. 92, Chapter 2; Kinghorn, A. D., Falk,
H., Kobayashi, J., Eds.; Springer-Verlag: New York, 2010; p 77 and
references cited therein.
ꢀ
ꢀ
ꢀ
(5) Leon-Rivera, I.; Miron-Lopez, G.; Estra-Soto, S.; Aguirre-Crespo,
ꢀ
F.; Gutierrez, M. C.; Molina-Salinas, G. M.; Hurtado, G.; Navarrete-
(2) Ono, M.; Yamada, F.; Noda, N.; Kawasaki, T.; Miyahara, K.
Chem. Pharm. Bull. 1993, 41, 1023.
ꢀ
Vazquez, G.; Montiel, E. Bioorg. Med. Chem. Lett. 2009, 19, 4652.
r
10.1021/ol4020255
XXXX American Chemical Society

Scheme 1. Retrosynthetic Analysis
Scheme 2. Preparation of Macrolactone Acceptor 4
manipulation, only three types of protecting groups, i.e.
p-methoxybenzyl (PMB), levulinoyl (Lev), and isopropyli-
dene groups, were employed.
Our studies on the total synthesis of batatin VI (1) began
with the preparation of the macrolactone alcohol 4(Scheme2).
By use of the Schmidt glycosylation conditions,11 the
known 912c was reacted with monosaccharide 10 (see
Supporting Information) in CH2Cl2 promoted by cataly-
tic amounts of TMSOTf to furnish 11 in quantitative
yield. Glycoside 11 was then saponified with KOH in a
THF/H2O (9:1) cosolvent at 55 °C to release the 20-OH and
the aglycon carboxyl groups, giving seco-acid 12 in 92% yield.
With the precursor 12 in hand, efforts were focused
on the key ring-closing process. Producing the unique
macrolactone core in these natural products has been
considered a major synthetic challenge, in which two
methodologies have been adopted so far.1c,8 One is a
macrolactonization approach.12 For instance, the Coreyꢀ
Nicolaou protocol was used, respectively, by the Schmidt,12a
Yu,12b,c and our9b groups for the total synthesis of calo-
nyctin A1, tricolorin A, and batatoside L. While the
Yamaguchi method was utilized by Heathcock12dꢀf and
However, studies on their systematic structureꢀactivity
relationship are still very limited.
Despite that several total syntheses of resin glycosides
have been accomplished,1c,8 no synthetic work on their
dimers has been reported to date. In a continuing investiga-
tion on the synthesis of resin glycosides,9 wehavedeveloped
an efficient route to optically active methyl 11(S)-jalapino-
late starting from commercially available (R)-glycidol.9a
Then, by use of a macrolactonization strategy, we realized
the total synthesis of batatoside L9b and the construction of
the 20- and 21-membered macrolactone rings of merremo-
side-type resin glycosides.9a Here, we describe the first total
synthesis of the proposed structure of dimeric batatin VI (1,
Scheme 1). This natural product along with three analogues
was isolated by Pereda-Miranda et al. from the tuberous
roots of sweet potato (Ipomoea batatas).6b Structural char-
acterization revealed that 1 is the first example of an ester-
type dimer of heterotetrasaccharide operculinic acid C.10
The ester linkage between the macrocycle-containing unit
A and the acyclic unit B is located at C-3 of the rhamnose
moiety c in unit A. Moreover, each monomeric unit is
acylated with fatty acids including n-dodecanoic (Dodeca),
trans-cinnamic (Cna), and 2(S)-methylbutyric (Mba) acids.
As retrosynthetically outlined in Scheme 1, we envi-
sioned that 1 could be obtained in a highly convergent
manner via a [5 þ 3] glycosidic coupling between penta-
saccharide alcohol 2 and trirhamnosyl trichloroacetimi-
date3. The assemblyof 2 wouldinturninvolvea sequential
[2 þ 2 þ 1] glycosylation of three building blocks: the
heterodisaccharide acceptor 4, the dirhamnosyl donor 5,
and the fucopyranosyl donor 6 (see Supporting Information).
Besides, imidate 3 could be easily derived from glycosyla-
tion of compound 79b with 8. To guarantee the expected
1,2-trans stereochemistry of each glycosylation, an acyl-
type functionality acting as a neighboring participating
group was incorporated at the 2-position of each donor.
Furthermore, in order to simplify the protection/deprotection
ꢀ
(6) (a) Escalante-Sanchez, E.; Pereda-Miranda, R. J. Nat. Prod.
ꢀ
´
rez, D.; Escalante-Sanchez, E.; Pereda-
2007, 70, 1029. (b) Rosas-Ramı
Miranda, R. Phytochemistry 2011, 72, 773.
~
ꢀ
(7) (a) Castaneda-Gomez, J.; Pereda-Miranda, R. J. Nat. Prod. 2011,
~
ꢀ
ꢀ
74, 1148. (b) Castaneda-Gomez, J.; Figueroa-Gonzalez, G.; Jacobo, N.;
Pereda-Miranda, R. J. Nat. Prod. 2013, 76, 64.
(8) For reviews on the synthesis of resin glycosides, see: (a) Furukawa,
€
J.; Sakairi, N. Trends Glycosci. Glycotechnol. 2001, 13, 1. (b) Furstner, A.
Eur. J. Org. Chem. 2004, 943.
(9) (a) Zhu, X.-M.; He, L.-L.; Yang, G.-L.; Lei, M.; Yang, J.-S.
Synlett 2006, 3510. (b) Xie, L.; Zhu, S.-Y.; Shen, X.-Q.; He, L.-L.; Yang,
J.-S. J. Org. Chem. 2010, 75, 5764. (c) Shen, X.-Q.; Xie, L.; Gao, L.; He,
L.-L.; Yang, Q.; Yang, J.-S. Carbohydr. Res. 2009, 344, 2063.
(10) Ono, M.; Yoda, S.; Kawasaki, T.; Miyakara, K. Chem. Pharm.
Bull. 1989, 37, 3209.
(11) (a) Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1986, 25, 212. (b)
Schmidt, R. R.; Jung, K.-H. In Preparative Carbohydrate Chemistry;
Hanessian, S., Ed.; Dekker: New York, 1997; p 283.
(12) For the synthesis of resin glycosides by a macrolactonization
approach, see: (a) Jiang, Z.-H.; Geyer, A.; Schmidt, R. R. Angew.
Chem., Int. Ed. Engl. 1995, 34, 2520. (b) Lu, S.-F.; O’Yang, Q.-Q.; Guo,
Z.-W.; Yu, B.; Hui, Y.-Z. Angew. Chem., Int. Ed. Engl. 1997, 36, 2344. (c)
Lu, S.-F.; O’Yang, Q.-Q.; Guo, Z.-W.; Yu, B.; Hui, Y.-Z. J. Org. Chem.
1997, 62, 8400. (d) Larson, D. P.; Heathcock, C. H. J. Org. Chem. 1996,
61, 5208. (e) Larson, D. P.; Heathcock, C. H. J. Org. Chem. 1997, 62,
8406. (f) Brito-Arias, M.; Pereda-Miranda, R.; Heathcock, C. H. J. Org.
Chem. 2004, 69, 4567. (g) Furukawa, J.; Kobayashi, S.; Nomizu, M.;
Nishi, N.; Sakairi, N. Tetrahedron Lett. 2000, 41, 3453.
B
Org. Lett., Vol. XX, No. XX, XXXX

Table 1. Optimization of Keck Macrolactonization of 12a
Scheme 3. Synthesis of Pentasaccharide Acceptor 2
entry
reagent
solvent time (h) yield of 13b
1
2
3
4
5
6
DCC, pyridine, PPTS
DCC, pyridine, PPTS
DCC, Et3N, PPTS
DCE
DCE
DCE
DCE
2
45%
36%
25%
92%
77%
63%
12
3
DCC, DMAP, PPTS
3
DCC, DMAP, DMAP HCl DCE
5
3
DCC, DMAP, PPTS
toluene
3
a All reactions wererunwith12(1 equiv, 1 mM), DCC (10equiv), base
(pyridine or Et3N or DMAP, 100 equiv), and a proton source (PPTS or
DMAP HCl, 10 equiv) in DCE or toluene under reflux. b Isolated yields.
3
Sakairi,12g respectively, in the synthesis of tricolorin A and
other resin glycosides. But this strategy suffers from either
long reaction times (5ꢀ7 days)12aꢀc,9b or modest cycliza-
tion yields (60ꢀ71%).12dꢀg The other protocol is an olefin
ring closing metathesis (RCM) strategy,13 by which the
13aꢀg
€
Furstner group
and later a team from the Eisai
Research Institute13h built up the macrolidic systems and
succeeded in synthesizing a series of complex resin glyco-
sides. But in this approach, the need to selectively reduce
the formed disubstituted CdC double bond in the tether
without affecting the unsaturated ester substituents on the
sugar chain required careful planning.
In order to improve the efficiency of the macrocycle
formation, we decided to explore another protocol. Thus,
the Keck macrolactonization14,15 in which the activated
ester intermediate is formed in situ and does not need to be
isolated was chosen for this purpose (see Table 1). Initially,
the hydroxy acid 12 was subjected to the modified Keck
conditions (DCC, pyridine, PPTS, 1,2-dichloroethane
(DCE), reflux, 2 h),15a which provided the desired large
ring derivative 13 in 45% yield (Table 1, entry 1). Elonga-
tion of the reaction time gave a lower 36% yield (entry 2).
After a series of screenings (entries 1ꢀ4), it was found that
the base is crucial for this reaction. When 12 was macro-
lactonized with a combination of DCC, DMAP, and PPTS
in DCE under reflux for 3 h, the yield of the product 13 was
dramatically enhanced to 92% (entry 4). Switching the
macrocyclic framework of resin glycosides via the Keck
protocol has been established. This protocol offers a signifi-
cant advantage over the previous method in terms of a
shorter reaction time, simpler operation, and higher yield.
Subsequent desilylation of 13 with TBAF in THF gave
the required lactone alcohol 4 in 87% yield (Scheme 2).
The next task was to prepare the pentasaccharide ac-
ceptor 2 (Scheme 3). In the event, treatment of compound
149b with ceric ammonium nitrate (CAN) in aqueous
acetonitrile gave alcohol 15. Then, we attempted to couple
15 with chiral jalapinolic acid derivative 16 (see Supporting
Information). However, this was met with difficulty prob-
ably due to the steric hindrance around the secondary
3-OH of 15. A number of condensation reagents, such as
EDCI/DMAP, TBTU/DBU, and DCC/DMAP, were ex-
amined. At last, the DCC/DMAP system was found to be
the most effective, and the desired 17 was isolated in 65%
yield. Then, exposure of 17 to TBAF buffered with HOAc
in THF16 at 35 °C for 2 days generated an 80% yield of
acceptor 18. On activation with TMSOTf, this compound
was readily coupled to rhamnosyl imidate 199b at ꢀ78 f
0 °C in CH2Cl2 to form disaccharide 20. The anomeric
center of 20 was then unmasked by a PdCl2-catalyzed
deallylation. Activation of the resulting crude hemiacetal
was performed by treatment with CCl3CN and DBU11b to
give the desired imidate 5 in 79% yield over two steps.
Then, our attention was directed to the connection of
macrocycle acceptor 4 with the fully elaborated donor 5
proton source and the solvent to DMAP HCl and toluene,
respectively, resulted in decreased yields (entries 5ꢀ6).
3
Thus, a more reliable approach to gain access to the
(13) For the RCM method in the synthesis of resin glycosides, see: (a)
€
€
€
Furstner, A.; Muller, T. J. Org. Chem. 1998, 63, 424. (b) Furstner, A.;
€
€
Muller, T. J. Am. Chem. Soc. 1999, 121, 7814. (c) Furstner, A.; Jeanjean,
F.; Razon, P.; Wirtz, C.; Mynott, R. Chem.;Eur. J. 2003, 9, 307. (d)
€
Furstner, A.; Jeanjean, F.; Razon, P.; Wirtz, C.; Mynott, R. Chem.;Eur.
€
J. 2003, 9, 320. (e) Furstner, A.; Jeanjean, F.; Razon, P. Angew. Chem.,
€
Int. Ed. Engl. 2002, 41, 2097. (f) Furstner, A.; Nagano, T. J. Am. Chem.
ꢁ
Soc. 2007, 129, 1906. (g) Nagano, T.; Pospısil, J.; Chollet, G.; Schulthoff,
´
€
€
S.; Hickmann, V.; Moulin, E.; Herrmann, J.; Muller, R.; Furstner, A.
Chem.;Eur. J. 2009, 15, 9697. (h) Postema, M. H. D.; TenDyke, K.;
Cutter, J.; Kuznetsov, G.; Xu, Q.-L. Org. Lett. 2009, 11, 1417.
(14) (a) Boden, E. P.; Keck, G. E. J. Org. Chem. 1985, 50, 2394. (b)
€
ꢀ
Kurti, L.; Czako, B. In Strategic Applications of Named Reactions in
ꢀ
Organic Synthesis;K€urti, L., Czako, B., Eds.; Elsevier: Burlington, 2005; p 238.
(15) For selected examples of the use of Keck macrolactonization in
the total synthesis of macrolides, see: (a) Kasai, Y.; Ito, T.; Sasaki, M.
Org. Lett. 2012, 14, 3186. (b) Hanessian, S.; Ma, J.; Wang, W. J. Am.
Chem. Soc. 2001, 123, 10200. (c) Mulzer, J.; Mantoulidis, A.; Oehler, E.
J. Org. Chem. 2000, 65, 7456.
(16) Smith, A, B., III; Ott, G. R. J. Am. Chem. Soc. 1996, 118, 13095.
TBAF buffered with HOAc has been screened as the best chioce for the
deprotection of the TBS ether. The use of other reagents such as TBAF
in THF or PPTS led to a 1:1 mixture of 17 and the acyl transfer isomer.
Org. Lett., Vol. XX, No. XX, XXXX
C

(Scheme 3). But this coupling reaction led to only a 45%
yield of the desired 21 under the same Schmidt glycosyla-
tion conditions. We reasoned that the unsatisfying out-
come might arise from the lability of the donor 5 in the
presence of the Lewis acid (TMSOTf). This prompted us to
survey another method, i.e., the inverse glycosylation
procedure to minimize the exposure of the imidate donor
to the Lewis acidic solutions.17 In such a glycosylation
process, the acceptor 4 was premixed with a catalytic
amount of TMSOTf (0.4 equiv) and 4 A molecular sieves
in dry CH2Cl2 prior to the gradual addition of a CH2Cl2
solution of the donor 5 (2.0 equiv) at ꢀ78 °C. As a result,
the unwanted donor hydrolysis was effectively prevented
and the yield of 21 was greatly improved to 85%. Subse-
quently, selective hydrolysis of the levulinate in 21 under
Scheme 4. Completion of Total Synthesis of Batatin VI (1)
mild basic conditions (N2H4 AcOH, CH2Cl2/CH3OH, rt)
yielded tetrasaccharide alcohol 22 in 82% yield.
3
The glycosylation of 22 with the fucose donor 6 also
proved to be challenging (Scheme 3). When the glycosylation
was carried out under the same conditions as those used for
9 f 11, only small amounts of product were detected. This is
likely due to the steric hindrance of the flexible jalapinolic
acid chain. After extensive experimentation, we were pleased
to find that, by increasing the amount of 6to 8 equiv18 and by
conducting this reaction at ambient temperature for 20 h,
the expected 23 could be successfully obtained in 93% yield
and with complete β-stereoselectivity. This compound was
occurring batatin VI.19 The most significant deviation in
1
the H NMR is observed at H-3 of the sugar f in unit B
(ΔδH = 1.4 ppm). Moreover, clear differences (ΔδC g 1
ppm) in most anomeric C-1 signals are also observed. The
structures of the synthetic intermediates 23 and 26 were
evidently determined through the use of NMR and ESI-
MS analysis. Complete analysis of 1D (1H, 13C, 400 MHz)
and 2D (gCOSY, HMQC, and HMBC) NMR spectroscopy
of 1 confirmed that all the anomeric centers, the linkages of
sugar residues, and the substitution positions of ester groups
were installed exactly as assigned in the published structure
of batatin VI by the Pereda-Miranda group6b (see Support-
ing Information for details). Consequently, a structural
revision of batatin VI may be required.20
In summary, the first total synthesis of the proposed
structure of batatin VI has been accomplished in a highly
convergent manner. The key steps include an improved
protocol for the closure of the 18-membered lactone via a
Keck macrolactonization and two inverse glycosylation
procedures for the prevention of hydrolysis of the donors.
The present work provides a facile entry into the natural
ester-type resin glycoside dimers. Structural reconsidera-
tion of batatin VI is currently underway.
subsequently treated again with N2H4 AcOH to remove the
3
Lev group, affording the corresponding 2 in excellent yield.
Meanwhile, the preparation of trisaccharide intermedi-
ate 3 was investigated (Scheme 4). Reaction of monosac-
charide 249b with levulinic acid (LevOH) afforded 25 that
underwent a clean 4-O-desilylation with TBAF to give
alcohol 8. Glycosidic coupling of 8 withthe known donor 7
resulted in a 94% yield of trirhamnoside 26 which was
easily converted into 3 utilizing the same two-step trans-
formation procedure (58% yield over two steps).
Finally, the stage was set to bring the fragments 2 and 3
together via O-glycosylation (Scheme 4). Under the similar
TMSOTf-activated inverse glycosylation conditions, the
coupling between 2 and 3 (1.6 equiv) proceededverywell to
furnish protected octasaccharide 27 in 96% yield as the
sole product. Global deprotection of 27 involved (i) clea-
vage of the Lev group with N2H4 AcOH, (ii) removal of
3
the isopropylidene acetals under acidic conditions (AcOH,
H2O), and (iii) oxidation deprotection of the PMB ethers
with DDQ, which gave the target 1 in 62% yield over three
steps after chromatographic purification.
Acknowledgment. The financial support from the NSFC
(21172156, 21021001), the 973 program (2010CB833202),
and the Ministry of Education (NCET-08-0377, 2010018-
1110082) is highly appreciated.
Unfortunately, the NMR data of our synthetic material
was not in full agreement with those of the naturally
(17) Schmidt, R. R.; Behrendt, M.; Toepfer, A. Synlett 1990, 694. The
€
inverse glycosylation technique has also played a key role in Furstner and
Yang’s resin glycoside syntheses; see refs 13d, 13e, and 9b, respectively.
(18) Donor 6 was unable to be recovered as it decomposed to the
corresponding hemiacetal during the glycosylation.
(19) The NMR spectra of both the natural (ref 6b) and the synthetic
samples were recorded in pyridine-d5.
(20) For a recent total synthesis of the proposed structure of a natural
Supporting Information Available. Experimental de-
1
tails, H and 13C NMR spectra for all new compounds,
and 2D NMR spectra for 1, 23, and 26. This material is
available free of charge via the Internet at http://pubs.acs.org.
€
glycolipid, see: Kondoh, A.; Arlt, A.; Gabor, B.; Furstner, A. Chem.;
Eur. J. 2013, 19, 7731.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX