Macrolide Core Synthesis of Calysolin IX Using an Intramolecular Glycosylation Approach

Article abstract of DOI:10.1002/ejoc.201901480

The utility of intramolecular glycosylation for the synthesis of the 27-membered macrocyclic ring is highlighted in this first total synthesis of the most complex resin glycoside isolated to date – Calysolin IX. Oligosaccharide-containing macrolides core

Full text of DOI:10.1002/ejoc.201901480

A Journal of
Accepted Article
Title: Macrolide Core Synthesis of Calysolin IX via Intramolecular
Glycosylation Approach
Authors: Mirosław Nawój, Artur Grobelny, and Jacek Mlynarski
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.201901480
Link to VoR: http://dx.doi.org/10.1002/ejoc.201901480
Supported by

10.1002/ejoc.201901480
European Journal of Organic Chemistry
COMMUNICATION
Macrolide Core Synthesis of Calysolin IX via Intramolecular
Glycosylation Approach
Mirosław Nawój,[a] Artur Grobelny[a] and Jacek Mlynarski*[b]
Abstract: The utility of intramolecular glycosylation for the
synthesis of 27-membered macrocyclic ring is highlighted in this first
total synthesis of the most complex resin glycoside isolated to date -
Calysolin IX. Oligosaccharide-containing macrolides core was
effectively constructed by TfOH/NIS-promoted intramolecular
glycosylation of thioglycosyl donor. As the glycosidic bond must be
created en route to target structure, we show that this unusual yet
efficient approach can effectively reduce the number of steps in total
synthesis of complex natural macrolides. This attempt is
documented as an efficient tool in the synthesis of gigantic macrolide
rings thus proving their practical utility in the total synthesis of sugar-
containing targets.
employing ring closure metathesis (RCM) for macrolide
cyclization, which was applied to the synthesis of various resin
glycosides.[7] This strategy was also utilized in the total
synthesis of Woodrosin I (1, Figure 1) - one of the most complex
resin glycosides known to date - isolated from Convovulacaeae
plant family.[8] The molecule comprises of 11-(S)-
hydroxyhexadecanoic acid (jalapinolic acid) as an aglycon,
which is tied back to form a characteristic macrolide ring that
spans two sugar units of their tetrasaccharide backbones.
Introduction
Achieving high efficiency and stereocontrol in glycosylation
reactions is arguably still the major effort in the synthesis of
widely distributed carbohydrate containing oligomers.[1] While
intermolecular formation of glycosidic bond has reached a
satisfactory level of development, the intramolecular
glycosylation strategy is still not broadly used and accepted in
the synthesis.[2] As a result, construction of the macrocyclic
rings by means of glycosylation is not seen as an efficient tool in
the synthesis of macrocyclic compounds[3] and therefore
practically neglected in the literature.[4] This is particularly
inconvenient in the synthesis of various structurally complex
macrocyclic glycolipids being a central part of resin glycosides.
Taking these compounds as an excellent example, resin
glycosides are conjugates between complex oligosaccharide
part and hydroxy fatty acid as an aglycon moiety. The latter
chain is attached to sugars to form a specific macrolactone ring
linking two or more sugar units of the backbone. Not surprisingly,
known studies revealed that the construction of the cyclic
macrolide core of resin glycosides is a crucial step eventually
determining synthetic strategy and overall efficiency.[5] In most
cases, this was carried out by intramolecular esterification
between the aglycon carboxylic acid and a sugar hydroxy group
under high-dilution conditions.[6]
Figure 1. Structure of Woodrosin I (1) and Calysolin IX (2).
It is obvious that such strategy requires the use of
additional steps, particularly the ring closing metathesis followed
by hydrogenation of thus created double bond, not to mention
multistep synthesis of two separated macrolide-precursors
containing orthogonally protected sugar moieties. This caused
the lower overall yield, and require application of Ru-catalyst.
Considering the number of steps, the most efficient
strategy would be however to close the ring by intramolecular
glycosylation reaction between sugar units, which in fact, must
be done during the synthesis, anyway. Such a general concept
assuming intramolecular glycosylation to create 27-membered
ring as a crucial step was, however, perilous due to the lack of
literature reports on this subject.[10] To prove this hypothesis we
present herein such approach for the synthesis of the most
complex resin glycosides isolated to date – Calysolin IX (2,
Figure 1). This new resin glycoside having 27-membered ring
macrolactone structure (jalapins) was isolated from the leaves,
stems, and roots of Calystegia soldanella (Convolvulaceae).[9]
Alternatively, Fürstner developed an elegant approach
[a]
[b]
Mirosław Nawój, Artur Grobelny
Faculty of Chemistry
Jagiellonian University
Gronostajowa 2, Cracow, Poland
Prof. dr. Jacek Mlynarski
Institute of Organic Chemistry
Polish Academy of Sciences
Kasprzaka 44/52, Warsaw, Poland
jacek.mlynarski@gmail.com
jacekmlynarski.pl
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201901480
European Journal of Organic Chemistry
COMMUNICATION
The sugar moiety of 2 was found to exist in partially acylated
forms comprising (2S)-methylbutyric acid and tiglic acid. All
those feature make Calysolin IX an extremely exciting synthetic
target. Efficient macrolactonization, but also proper planning of
orthogonally protected monosaccharides would be an exciting
trial by fire for our concept.
Leaving disaccharide fragment 5 to be attached at the very
end, we envisaged that the macrolactone precursor itself can be
assembled in a highly convergent manner from the disaccharide
building blocks 3 and 4 (Scheme 1), which were obtained as
described in Schemes 2 and 3. A detailed description of the
synthesis of these fragments can be found in the Supporting
Information material.
Results and Discussion
The target Calysolin IX (2) differs from formerly
synthesized Woodrosin I (1) by number and kind of sugar units.
Particular synthetic challenges arise from the 27-membered
macrolide ring spanning three glucose and one 6-deoxyglucose
(Quinovose). Peripheral disaccharide unit comprises of glucose
and rhamnose (5) can be attached to central tetrasaccharide
core at the late stage of synthesis, according to our
retrosynthetic plan outlined at Scheme 1. This risky approach
allows however to focus on the construction of macrolide core at
the first steps. A key disconnection is obviously the glycosidic
linkage between the Glc and Glc’ rings. This bond would be
formed by intramolecular coupling of glucose-glucose
disaccharide glycosyl donor
4 with a glucose-quinovose
Scheme 2. Synthesis of disaccharide 3. [a] i) Oxone®, acetone,
DCM/NaHCO3(aq), 0 °C to rt, 3 h; ii) 1M ZnCl2∙OEt2, MS 4Å , THF, -78 °C to rt,
48 h, 71 %. [b] 0.1 M TMSOTf (cat.), MS 4Å, DCM, -20 °C, 3 h, 83 %.
disaccharide glycosyl acceptor 3. To execute this key step, both
disaccharide fragments must be linked by using 11-(S)-
hydroxyhexadecanoic acid (jalapinolic acid). This requires
efficient method for the formation of ester bond between
carboxyl function in 3 and 2-OH of glucose ring. While it looks
straightforward on paper, the implementation of this plan
requires more careful choice of protecting groups in sugar
molecules as two kind of ester-type residues and lactone ring
are inherently present in final molecule. The occurrence of
double bond in tiglic acid residue made the global application of
benzyl groups unsuitable for this synthesis, of course. For this
reason, after several simulations, we found to use the PMB
ethers as the most promising candidates for the hydroxyl groups
protection and final, last-stage deprotection.
The most efficient approach to disaccharide part 3 began
with the synthesis of jalapinolic ester 6 from undecyne and
hexanal. Enantioselective reduction of alkynone with (S,S)-
Noyori catalyst resulted in the alcohol with high ee.[11] Thus
obtained aglycone was used for the one-pot formation of -
glycoside from 6-deoxyglucal (7). This protocol not being widely
used in the glycosylation turned out to be method of choice for
the preparation of target -glycoside. Efficient, stereoselective
synthesis of epoxide ring by using Oxone,[11] followed by zinc
chloride-promoted opening of oxirane ring[12] resulted in desired
glycoside 8 with free OH group at the C-2 position suitable for
next glycosylation step. This was done by using
trichloroacetimidate 9 in the presence of a catalytic amounts of
TMSOTf to give disaccharide 3 in 83% yield (Scheme 2).
Scheme 3. Synthesis of disaccharides 4. [a] i) Oxone®, acetone,
DCM/NaHCO3(aq), 0 °C, 2 h; ii) 1M ZnCl2∙OEt2, MS 4Å , THF, -78 °C to rt, 36 h,
4a: 71 %; 4b: 78 %.
The same efficient protocol (Oxone, ZnCl2) was used for
the synthesis of second disaccharide fragment 4 (Scheme 3)
from glucal 10 and thioglycoside 11. The application of sugar
with SPh group, which will be needed in the macrolide ring
closing step reaction, does not interfere with the glycosylation
reaction at this stage. Two different protecting groups at 2-OH of
Scheme 1. Retrosynthetic analysis of Calysolin IX (2).
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201901480
European Journal of Organic Chemistry
COMMUNICATION
Glc’ ring have been used for testing two synthetic strategies,
which will be revealed later.
applied for the synthesis of PMB-protected 13b from 4b (83%,
entry 5).
Having both disaccharides in hand, we started to explore
the foremost, risky part of connection the macrolide ring. First,
TIPS-ester hydrolysis of 3 followed by deacetylation reaction by
using DIBAL liberated both OH and COOH groups (Scheme 4).
A standard deacetylation reaction under Zémplan protocol
(MeONa/MeOH) or other deprotection methods realized by
treatment with basic reagents caused a trans-3,2-TBS group
migration, leading to a mixture of 2’-O- and 3’-O-silyl protected
compounds in a (1:1) ratio. Thus obtained 2’-O-TBS derivative is
much less prone to react with the Olah’s reagent used in the
further step. This made carboxyl group in disaccharide 12 ready
for esterification with free OH in 4. However, the presence of a
hydroxyl group in the substrate 12 could enable intramolecular
esterification. Anticipating this, we applied sterically hindered
tert-butyldimethylsilyloxy protection at neighboring C-3 position
to prevent undesired side reaction i.e. the intramolecular
cyclization. This assumption turned out to be correct and
disaccharide 12 was smoothly esterified with alcohol 4 in the
presence of DIC and a catalytic amount of DMAP (Scheme 4)
The next stage of the synthesis required the efficient
assembly of the two terminus of each disaccharide to the
tetrasaccharide macrolide core of Calysolin IX. Deprotection of
the silyl group from 13a was easy and resulted in the
thioglycoside 14a with two free hydroxyl groups ready for
glycosidic bond formation. Scheme 5 presents required route to
macrolide core via stereo- and regioselective glycosylation.
Such a strategy, if efficient, would allow the formation of a
macrolide ring and leave a free hydroxyl group at C-2 of Glc
ready for final attachment of disaccharide 5 (Scheme 1). For
steric reasons, equatorially placed 3-OH group should be more
accessible allowing regioselective glycosylation, however, this
endeavor turned out to be far from trivial.
Table 1. Intermolecular esterification.
Equivalents of
Entry[a]
Substrate
Time
Yield[b]
substrate
1.25
1.5
1
2
3
4
5
4a
4a
4a
4a
4b
40 h
18 h
18 h
18 h
36 h
60 %
73 %
78 %
94 %
83 %
2.5
5.0
2.5
[a] Performed in DCM in the presence of DIC and catalytic amount of DMAP
at rt. [b] Isolated yields.
At the beginning we checked the reactivity of compound
4b where thioglycoside was additionally activated in the
presence of benzyl-type PMB ether attached to C-2 position of
Glc’ ring. Unfortunately, no attempt to activate SPh, including
triflic acid/NIS (Table 2, entry 1) was effective. Therefore the
alternative approach for the glycosidation steps was explored by
using ester-type levulinic protecting group at C-2 of Glc’ in the
substrate 14a allowing the orthoester formation at the crucial
step. This additional step forces the approximation of the
acceptor hydroxyl group to the activated glycosyl donor part.
Proving this concept required additional evidence for the in situ
formation of orthoester intermediate from 14b. This was possible
by using NMR of reaction mixture. Most indicative was a doublet
in the 1H NMR at 5.52 ppm of anomeric proton and
corresponding signal in 13C NMR spectrum at 120.2 ppm which
confirm the newly formed orthoester junction (Scheme 5).
Scheme 4. Synthesis of 13. [a] KCN, MeOH/DCM (1:1), rt, overnight; [b] 1M
DIBAL, DCM, -78 oC, 2.5 h, 82 % (over both steps). [c] DIC, DMAP (cat.),
DCM, rt; conditions and yields see Table 1.
Only 1.25 equivalents of disaccharide 4a was enough to
obtained good yield of the desired tetrasaccharide 13a (Table 1,
entry 1). High regioselectivity and reitarability of the reaction
deserved further investigation and the results are collected in
Table 1. Although the application of 5 equivalents of 4a resulted
in the best reaction yield (94%, entry 4), we decided to use 2.5
equivalents of acid (78%, entry 3), mostly due to the highest
overall reaction economy. Thus, the same conditions have been
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201901480
European Journal of Organic Chemistry
COMMUNICATION
Scheme 5. Synthesis of 16. [a] HF∙Py, THF, 0 °C to rt, 48 h, 14a: 76 %; 14b: 72 %. [b] Conditions and yields see Table 2.
Table 2. Results of intramolecular glycosylation.
Concentration of
Entry[a]
1
Substrate
Conditions[b]
Time
40 h
Yield
substrate[c]
14b
TfOH (0.25 eq.)/NIS, DCM, -20 °C to 0 °C
4.5 mM
No reaction[b]
2
3
4
5
6
7
8
14a
14a
14a
14a
14a
14a
14a
TMSOTf (0.25 eq.)/NIS, Hx/DCM (2:1), -20 °C to rt
TMSOTf (0.65 eq.)/NIS, Hx/DCM (2:1), 0 °C
MeOTf (5.0 eq.), Hx/DCM (2:1), 0 °C
4.5 mM
4.5 mM
4.5 mM
10 mM
45 mM
10 mM
4.5 mM
24 h
18 h
2 h
No reaction
Deprotection of
PMB group
Deprotection of
PMB group
TfOH (0.1 eq.)/NIS, CHCl3, 0 °C
16 h
14 h
16 h
40 h
No reaction
13 %
TfOH (0.1 eq.)/NIS, DCM, -40 oC to 0 °C
TfOH (0.1 eq.)/NIS, DCM, -20 oC to 0 °C
TfOH (0.15 eq.)/NIS, DCM, -20 oC to 0 °C
45 %
58 %
[a] All reaction was performed with molecular sieves 4 Å under strictly anhydrous conditions. [b] Further addition of TfOH caused deprotection of
PMB group.
Nonetheless, the implementation of this strategy required
very careful optimization of the glycosylation conditions. Table 2
presents more informative steps showing the impact of the
applied conditions, kind of activators and substrate
concentration on the reaction outcome. Under most efficient
conditions described in entry 8 intramolecular glycosylation
of 14a using similar TfOH/NIS was found to proceed smoothly in
DCM at the substrate concentration of 4.5 mM, giving protected
macrolide core of Calysolin IX 16 in 58% yield. It is noteworthy,
that the transformation of orthoester into final macrolide required
prolonged reaction time at 0 °C. Another interesting observation
shows that a similar yield can be achieved under 10 and 4.5
concentration. This shows that intramolecular glycosylation is
tolerant to broader concentration and reaction conditions and
thus hypothetically useful for similar macrolide synthesis.
Conclusions
We have succeeded in the first total synthesis of macrolide ring
of the most complex resin glycoside Calysolin IX known through
intramolecular glycosylation approach. Our work shows not only
the relevance of glycosylation to macrolide ring construction, but
also present usefulness of an orthoester-strategy to catch the
two ends of the ring, which supports closing the macrolide core.
With macrolide 16 in hand, the introduction of the missing
disaccharide Glc’’’-Rha moiety will be investigated in due course
and soon presented.
Acknowledgements
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201901480
European Journal of Organic Chemistry
COMMUNICATION
Financial support from the Polish National Science Centre
(MAESTRO Grant No. 2013/10/A/ST5/00233) is gratefully
acknowledged.
Keywords: natural products • macrolide • carbohydrates •
intramolecular glycosylation • total synthesis
[1]
[2]
Handbook of Chemical Glycosylation: Advances in Stereoselectivity
and Therapeutic Relevance, A. V. Demchenko Ed. Wiley-VCH,
Weinheim, 2008.
X. G. Jia, A. V. Demchenko, Beilstein J. Org. Chem. 2017, 13, 2028-
2048.
[3]
[4]
J. Xie, N. Bogliotti, Chem. Rev. 2014, 114, 157678-7739.
S.-H. Son, N. Yanagiya, J.-i. Furukawa, N. Sakairi, Syntlett, 2009, 18,
2957-2960.
[5]
[6]
(a) A. Fürstner, Eur. J. Org. Chem. 2004, 943. (b) J. Furukawa, N.
Sakairi, Trends Glycosci. Glycotechnol. 2001, 13, 1.
For the synthesis of Tricolorin A see: (a) D. P. Larson, C. H. Heathcock,
J. Org. Chem.1997, 62, 8406-8418. (b) S.-F. Lu, Q. O'yang, Z.-W. Guo,
B. Yu, Y.-Z. Hui, Angew. Chem. Int. Ed. 1997, 36, 2344-2346.
(a) A. Fürstner, T. Müller, J. Org. Chem. 1998, 63, 424. (b) A. Fürstner,
T. Müller, J. Am. Chem. Soc. 1999, 121, 7814.
[7]
[8]
(a) A. Fürstner, F. Jeanjean, P. Razon, Angew. Chem. Int. Ed. 2002, 41,
2097-2101. (b) A. Fürstner, F. Jeanjean, P. Razon, C. Wirtz, R. Mynott,
Chem. Eur. J. 2003, 9, 307-319. (c) A. Fürstner, F. Jeanjean, P. Razon,
C. Wirtz, R. Mynott, Chem. Eur. J. 2003, 9, 320-326.
M. Ono, A. Takigawa, Y. Kanemaru, G. Kawakami, K. Kabata, M.
Okawa, J. Kinjo, K. Yokomizo, H. Yoshimitsu, T. Nohara, Chem. Pharm.
Bull. 2014, 62, 97-105.
[9]
[10] To date only one example of intramolecular glycosylation to construct
less complex 22-membered macrolactone ring with trisaccharide
backbone was presented. For the synthesis of Tricolorin F see Ref. 4.
[11] E. U. Sharif, H.-Y. Leo Wang, N. G. Akhmedov, G. A. O’Doherty, Org.
Lett. 2014, 16, 492−495.
[12] P. Cheshev, A. Marra, A. Dondoni, Carbohydr. Res. 2006, 341, 2714-
2716.
[13] P. H. Seeberger, M. Eckhardt, C. E. Gutteridge, S. J. Danishefsky, J.
Am. Chem. Soc. 1997, 119, 10064-10072.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201901480
European Journal of Organic Chemistry
COMMUNICATION
Entry for the Table of Contents
COMMUNICATION
Intramolecular glycosylation as
an efficient tool for the synthesis of
27-membered macrocyclic ring is
highlighted in the first total
Mirosław Nawój, Artur Grobelny,
Jacek Mlynarski*
Page No. – Page No.
synthesis of the most complex
resin glycoside isolated to date -
Calysolin IX.
Macrolide Core Synthesis of
Calysolin IX via Intramolecular
Glycosylation Approach
This article is protected by copyright. All rights reserved.