Red Algal Molecules - Synthesis of Methyl Neo-β-carrabioside and Its S-Linked Variant via Two Synthetic Routes: A Late Stage Ring Closure and Using a 3,6-Anhydro- d -galactosyl Donor

Article abstract of DOI:10.1021/acs.joc.0c02339

Methyl neo-β-carrabioside has been synthesized for the first time, employing either a late stage ring closure to install the required 3,6-anhydro-bridge or a suitable 3,6-anhydro-galactosyl donor to form the unfavored 1,2-cis-equatorial α-linkage. Using the late stage ring closure approach, an S-linked analogue of methyl neo-β-carrabioside was also realized. These compounds have applications in the identification and characterization of marine bacterial exo-α-3,6-anhydro-d-galactosidases that have specific activity on red algal neo-carrageenan oligosaccharides, such as those found in both family 127 and 129 of the glycoside hydrolases. In addition a biochemical assay using the synthesized methyl neo-β-carrabioside and the marine bacterial exo-α-3,6-anhydro-d-galactosidase ZgGH129 demonstrates that the minimum substrate unit for the enzyme is neo-β-carrabiose.

Full text of DOI:10.1021/acs.joc.0c02339

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
pubs.acs.org/joc
Article
Red Algal Molecules - Synthesis of Methyl Neo-β-carrabioside and Its
S‑Linked Variant via Two Synthetic Routes: A Late Stage Ring
Closure and Using a 3,6-Anhydro‑D‑galactosyl Donor
Michael D. Wallace, Elizabeth Ficko-Blean,* and Keith A. Stubbs*
Cite This: https://dx.doi.org/10.1021/acs.joc.0c02339
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Methyl neo-β-carrabioside has been synthesized for the first
time, employing either a late stage ring closure to install the required 3,6-
anhydro-bridge or a suitable 3,6-anhydro-galactosyl donor to form the
unfavored 1,2-cis-equatorial α-linkage. Using the late stage ring closure
approach, an S-linked analogue of methyl neo-β-carrabioside was also
realized. These compounds have applications in the identification and
characterization of marine bacterial exo-α-3,6-anhydro-D-galactosidases that
have specific activity on red algal neo-carrageenan oligosaccharides, such as
those found in both family 127 and 129 of the glycoside hydrolases. In
addition a biochemical assay using the synthesized methyl neo-β-
carrabioside and the marine bacterial exo-α-3,6-anhydro-D-galactosidase ZgGH129 demonstrates that the minimum substrate unit
for the enzyme is neo-β-carrabiose.
Carrageenans are hydrocolloids, highly hydrated molecules
with gelling capabilities and the ability to increase viscosity.3
Both the 3,6-anhydro-bridge, which locks the corresponding
INTRODUCTION
■
Carrageenans are complex sulfated polysaccharides and one of
the major cell wall components of carrageenophyte red
macroalgae (Rhodophyta).1 The structure is comprised
primarily of repeating units of D-galactose and the bicyclic
carbohydrate 3,6-anhydro-D-galactose (some variants contain
6-O-sulfo-D-galactose in place of 3,6-anhydro-D-galactose),
containing alternating β-(1,4) and α-(1,3) linkages (Figure
1). The core carrageenan backbone is further decorated with
1
pyranose ring in the C4 conformation, and the sulfate groups
are responsible for carrageenans’ rheological properties. This is
also observed with agars, the other major red macroalgal
polysaccharides, which encompass agarose (unsulfated) and
sulfated derivatives. Agars have a similar repetitive disaccharide
motif to carrageenans but contain 3,6-anhydro-L-galactosyl
residues rather than 3,6-anhydro-D-galactosyl residues. Due to
the rheological properties and natural abundance of carra-
geenans, they have been utilized in food, personal care, and
cosmetic products. Some carrageenans have also been shown
to exhibit bioactivities, such as antitumor or antiviral activity3
presumably due to the sulfations that mimic the sulfations on
animal glycans.2
Some marine heterotrophic bacteria use carrageenans as a
carbon source.4,5 Recently a polysaccharide utilization locus
and regulon from Zobellia galactanivorans dedicated to the
catabolism of carrageenans was discovered.4 As part of the
study, four genes encoding exo-α-3,6-anhydro-D-galactosidase
activity were described, an enzyme class which had been
predicted to exist in Nature but had not been elucidated. The
Figure 1. Major variants of 3,6-anhydro-D-galactose containing
carrageenan units, β-, κ-, and ι-carrageenan, with different levels of
sulfation.
Special Issue: A New Era of Discovery in Carbohy-
drate Chemistry
different levels of sulfation and may also be substituted with
methyl and pyruvate groups. The major 3,6-anhydro-D-
galactose containing variants of carrageenans are κ- and ι-
carrageenan, and the desulfated is β-carrageenan (Figure 1);
however, natural carrageenans are hybrid structures that
contain diverse carrabiose motifs within the polymer.2
Received: October 2, 2020
https://dx.doi.org/10.1021/acs.joc.0c02339
© XXXX American Chemical Society
J. Org. Chem. XXXX, XXX, XXX−XXX
A

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Organic Chemistry
pubs.acs.org/joc
Article
Figure 2. Two approaches for the synthesis of methyl neo-β-carrabioside 1 and the S-linked variant 2. PG = protecting group, LG = leaving group.
a
Scheme 1
a
(a) i. NaOMe, MeOH, rt; ii. TBDMSCl, imidazole, DMF, rt; iii. Ac2O, pyridine, CH2Cl2, rt, 73% over 3 steps; (b) TBAF, AcOH, THF, rt 95%;
(c) ClAc2O, pyridine, CH2Cl2, 0 °C, 95%; (d) i. PdCl2, NaOAc, AcOH/H2O 9:1, EtOAc, rt; ii. Cl3CCN, DBU, CH2Cl2, rt; iii. 8, TMSOTf, 4 Å
MS, Et2O, CH2Cl2, −20 °C, 20% over three steps; (e) 8, TMSOTf, 4 Å MS, Et2O, CH2Cl2, −20 °C, 38%.
biochemical function of these enzymes is to cleave the 3,6-
anhydro-D-galactosyl residue from the non-reducing end of
neo-carrageenan oligosaccharides (“Neo” denotes that the
nonreducing end residue of the carrageenan oligosaccharide is
either a 3,6-anhydro-D-galactosyl or 6-O-sulfo-D-galactosyl
residue). Three of the enzymes are classified as belonging to
family GH127 of the glycoside hydrolases, whereas the fourth
falls into family GH129 (http://www.cazy.org/).6 Gene
deletion experiments abolished growth for the bacterium
when the ZgGH127-3 and ZgGH129 double mutant was
grown on carrageenan substrates demonstrating their im-
portance in carrageenan catabolism and to the biology of the
marine bacterium.4 An in-depth biochemical investigation into
ZgGH129 found it to be inactive on neo-κ-carrageenan
oligosaccharide substrates4 but active on oligosaccharides
with the neo-β-carrageenan motif at the nonreducing end.7
Furthermore, the enzyme demonstrated activity on a novel
synthetic substrate allowing a fine kinetic characterization of its
enzymatic properties.7
synthesis of two new chemical tools to aid in the further study
of this class of enzymes, the disaccharide methyl neo-β-
carrabioside (methyl 3-O-(3,6-anhydro-α-D-galactopyranosyl)-
β-D-galactopyranoside) 1 and the non-hydrolyzable S-linked
variant 2 as a potential inhibitor of exo-α-3,6-anhydro-D-
galactosidases (Figure 2). As part of our investigation into the
synthesis, we also examined the use of a 3,6-anhydro-galactosyl
donor capable of directly forming the required 1,2-cis-
equatorial α-linkage. The exploration and development of the
successful synthetic routes described will also aid in the
synthesis of related analogues and putative inhibitors.
RESULTS AND DISCUSSION
■
Two synthetic strategies were considered for the preparation of
both 1 and 2 (Figure 2). The first strategy was to form the
desired α-(1,3) glycoside and then install a suitable leaving
group (e.g., sulfonate) at the C6 position of the nonreducing
galactosyl residue, which would allow for late-stage 3,6-
anhydro-bridge formation. Interestingly this method has a
biosynthetic inspiration as galactose-sulfurylases, which are
unique red algal enzymes, utilize a 6-O-sulfo-D-galactosyl
residue as a substrate to form the 3,6-anhydro-bridge and
release sulfate8−11 (Despite being known in Nature bio-
Despite the biological and industrial importance of
carrageenans there are currently only limited molecular tools
to study the biochemistry of the enzymes that possess exo-α-
3,6-anhydro-D-galactosidase activity.4,7 Here, we describe the
B
https://dx.doi.org/10.1021/acs.joc.0c02339
J. Org. Chem. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Organic Chemistry
pubs.acs.org/joc
Article
a
Scheme 2
a
(a) 8, TMSOTf, 4 Å MS, CH2Cl2, 0 °C, 88%; (b) 70% w/w HF-pyridine, THF, rt, 90%; (c) TsCl, DMAP, pyridine, CH2Cl2, rt, 87%; (d) 80% aq.
AcOH, rt, 91%; (e) NaOMe, MeOH, rt, 82%.
a
Scheme 3
a
(a) 12, TMSOTf, 4 Å MS, CH2Cl2, 0 °C, 23%; (b) i. 70% w/w HF−pyridine, THF, rt; ii. TsCl, DMAP, pyridine, CH2Cl2, rt, 61% over two steps;
(c) i. 80% aq. AcOH, rt; ii. NaOMe, MeOH, rt, 65% over two steps.
chemical studies on recombinant galactose-sulfurylase enzymes
have not yet been described in the literature and these enzymes
remain one of red algae’s most fascinating, yet least
understood, family of enzymes). In addition, this synthetic
methodology has been used in the synthesis of methyl β-
carrabioside and the unnatural derivative methyl 3-O-(3,6-
anhydro-β-D-galactopyranosyl)-α-D-galactopyranoside.12 The
second strategy was to explore the use of 3,6-anhydro-
galactosyl donors that would be suitable for α-glycosylation
(Figure 2). We were buoyed by previous studies which utilized
orthoester13 and cyanoethylidene14 3,6-anhydro-galactose
derivatives as glycosyl donors to prepare β-glycosides.
Moreover, Christina et al.15 while exploring galacturonic acid
lactone thioglycoside donors for use in making α-glycosides,
demonstrated the possible use of a 3,6-anhydro-galactosyl-type
donor in this regard. Overall though, the α-glycoside is
inherently more difficult to form as it is a 1,2-cis-equatorial
product, which is disfavored by a combination of both
neighboring group participation and the anomeric effect.
With the goal of synthesizing 1 and 2 the biosynthetic-inspired
route was first explored.
In the first instance the galactosyl trichloroacetimidate 3 was
trialled (Scheme 1), as a similar donor has been used with a
thiol acceptor,16 which was important in developing a shared
method to synthesize both 1 and 2. By including a chloroacetyl
protecting group at the C6 position, selective removal could be
achieved followed by tosylation for the subsequent ring
16
́
closure. Thus, Zemplen deacetylation of the triacetate 4,
followed by selective protection of the C6 hydroxy group with
the TBDMS group and acetylation of the remaining hydroxy
groups, yielded the diacetate 5. The TBDMS group was then
removed via treatment with TBAF, and the subsequent alcohol
6 was protected with a chloroacetyl protecting group to obtain
the triester 7. The allyl group was then removed via a Pd(II)-
mediated deprotection17 to give the presumed hemiacetal,
which was then treated with trichloroacetonitrile and DBU to
afford the trichloroactimidate 3. With 3 in hand glycosylation
with the known acceptor 818 was then attempted, using
TMSOTf as a catalyst and Et2O as an additive to promote α-
C
https://dx.doi.org/10.1021/acs.joc.0c02339
J. Org. Chem. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
The Journal of Organic Chemistry
pubs.acs.org/joc
Article
selectivity.19 Unfortunately, the only product isolated in low
yield (35%) was the disaccharide 9 which had the desired α-
(1,3)-linkage but had lost the 4-methoxybenzyl (PMB) group
at C2. Due to this unexpected result the reaction was
reattempted using the donor 10 which was previously
employed by Xia et al.16 (Scheme 1), to synthesize both O-
and S-linked isoglobotrihexosylceramides. However, in our
hands only the disaccharide 11 was obtained in a low yield
(38%), again with loss of the PMB group. Due to this result
another donor was sourced.
22. Protection of the C2 and C4 hydroxy groups with armed
non-participating benzyl groups28 yielded the allyl ether 23.
Removal of the allyl glycoside using a Pd(II)-mediated
deprotection17 to obtain the hemiacetal 24 looked successful
by TLC, but upon 1H NMR analysis the material was
determined to be a complex mixture of products with
distinctive aldehyde signals, although these did not seem to
relate to the major component (see Supporting Information).
Despite this, the material was taken forward to attempt the
preparation of 20. Attempts at preparing the trichloro-
acetimidate using standard conditions did not result in any
observable formation of 20, but only again gave a complex
mixture of products. We presumed that inability to form 20
was due to 3,6-anhydro-galactose preferring to exist in the
open aldehyde form rather than the bicyclic pyranose form
(Scheme 4), which is required for the successful reaction. This
preference has been shown through the study of 3,6-anhydro-
D-galactose, which was found to have aldehydic character due
to the added ring strain caused by the 3,6-anhydro-bridge.29
Another possible reason is that this system could also be too
armed to be isolated with standard conditions, which
additionally would not make it a desirable glycosyl donor.
Based on these results, in order to avoid the 3,6-anhydro-
galactose hemiacetal, a glycosyl donor needed to be prepared
where the activatable leaving group was in place before
formation of the 3,6-anhydro-bridge. Indeed, this result
highlights the benefit of a thioglycoside, used previously in
this regard,15 where the activatable group is stable to many
different chemistries and this stability allows for manipulation
of the other hydroxy groups to generate molecules of
interest.30 Thus, the diol 25 was prepared from the 6-O-
tosylate 2631 via treatment with methanolic NaOMe (Scheme
5), and protection of 25 yielded the benzyl protected putative
The galactosyl donor 12 developed by Kiso and co-
workers20,21 has been shown to be an excellent glycosyl
donor that gives exclusively α-anomeric products, despite the
participating benzoyl group at C2 (Scheme 2). Gratifyingly,
glycosylation of 12 with the acceptor 8 resulted in formation of
the disaccharide 13 in excellent yield (82%) and with no
observable formation of the undesired β-glycoside. The di-tert-
butylsilylene group was then selectively removed with 70%
HF−pyridine to obtain the diol 14, which was selectively
tosylated to furnish the tosylate 15. Aqueous acid-mediated
hydrolysis of the 4,6-O-benzylidene acetal afforded the triol 16,
and finally treatment with methanolic NaOMe removed both
the benzoate protecting groups and concurrently formed the
desired 3,6-anhydro-bridge to give 1. With this successful route
now developed, it was then applied, using instead the thiol
acceptor 17,22 to the synthesis of 2 in good yield (Scheme 3).
Expediting the synthesis of 1 would be useful in the
synthesis of related compounds. Thus, with 1 and 2 in hand via
the late-stage ring closure method, we now looked into the
application of using a 3,6-anhydro-galactosyl donor that would
allow for the formation of α-glycosides (Figure 2). Indeed, it
has been suggested that molecules of this type would be highly
armed glycosyl donors.23,24 In the first instance we wanted to
explore the possible formation and use of a glycosyl imidate25
based donor, as these types of donors are common throughout
synthetic carbohydrate chemistry, so the synthesis of the 3,6-
anhydro-galactosyl-based trichloroacetimidate 20 was attemp-
ted (Scheme 4). Treatment of allyl β-D-galactopyranoside 21
using Appel reaction conditions, which have previously been
used in the synthesis of other 3,6-anhydro-galactosides,7,26,27
successfully installed the 3,6-anhydro-bridge yielding the diol
a
Scheme 5
a
Scheme 4
a
(a) NaOMe, MeOH, rt, 91%; (b) R−Br, NaH, DMF, rt, 95−99%.
donor 27. For α-selectivity, formation of the heavily disfavored
1,2-cis-equatorial bond was required. We were drawn to the
methodologies used in β-D-mannosyl-, β-L-rhamnosyl-, and
uronic acid 6,3-lactone-based glycosylations, as these have the
desired 1,2-cis-equatorial system. The preactivation strategy
pioneered by Crich and co-workers32 has been utilized for
these difficult glycosylations, which entails preactivation of an
appropriate thioglycoside with an activator and Tf2O.
Previously Christina et al.15 applied this system to a
comparable 3,6-anhydro-galactosyl donor to study the
reactivity and selectivity of a galacturonic acid 6,3-lactone
thioglycoside as a glycosyl donor. However, they did not
explore the utility of 3,6-anhydro-galactosyl donors in great
detail.15
In the first instance, the benzyl protected 27 was
glycosylated with a test acceptor 28, first using the common
NIS/TfOH promotor system for comparison. Pleasingly, the
disaccharide 29 was obtained in good yield (Table 1);
a
(a) CBr4, PPh3, pyridine, 60 °C, 91%; (b) BnBr, NaH, DMF, rt,
95%; (c) PdCl2, NaOAc, AcOH/H2O 9:1, EtOAc, rt.
D
https://dx.doi.org/10.1021/acs.joc.0c02339
J. Org. Chem. XXXX, XXX, XXX−XXX