Facile anomer-oriented syntheses of 4-methylumbelliferyl sialic acid glycosides

Article abstract of DOI:10.1039/d1ob00877c

As part of a program to find new sialidases and determine their enzymatic specificity and catalytic activity, a library of 4-methylumbelliferyl sialic acid glycosides derivatised at the C-5 position were prepared fromN-acetylneuraminic acid. Both α- and β-4-methylumbelliferyl sialic acid glycosides were prepared in high yields and stereoselectivity. α-Anomers were accessedviareagent control by utilising additive CH₃CN and TBAI, whereas the β-anomers were synthesised through a diastereoselective addition reaction of iodine and the aglycone to the corresponding glycal followed by reduction of the resulting 3-iodo compounds. Both anomer-oriented synthetic pathways allow for gram-scale stereoselective syntheses of the desired C-5 modified neuraminic acid derivatives for use as tools to quantify the enzymatic activity and substrate specificity of known sialidases, and potential detection and investigation of novel sialidases.

Full text of DOI:10.1039/d1ob00877c

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Facile anomer-oriented syntheses of
4-methylumbelliferyl sialic acid glycosides†
Cite this: Org. Biomol. Chem., 2021,
19, 6644
Abdullah A. Hassan
and Stefan Oscarson
*
As part of a program to find new sialidases and determine their enzymatic specificity and catalytic activity,
a library of 4-methylumbelliferyl sialic acid glycosides derivatised at the C-5 position were prepared from
N-acetylneuraminic acid. Both α- and β-4-methylumbelliferyl sialic acid glycosides were prepared in high
yields and stereoselectivity. α-Anomers were accessed via reagent control by utilising additive CH3CN and
TBAI, whereas the β-anomers were synthesised through a diastereoselective addition reaction of iodine
and the aglycone to the corresponding glycal followed by reduction of the resulting 3-iodo compounds.
Both anomer-oriented synthetic pathways allow for gram-scale stereoselective syntheses of the desired
C-5 modified neuraminic acid derivatives for use as tools to quantify the enzymatic activity and substrate
specificity of known sialidases, and potential detection and investigation of novel sialidases.
Received 5th May 2021,
Accepted 23rd June 2021
DOI: 10.1039/d1ob00877c
rsc.li/obc
differentiation,10 cell signalling11 and apoptosis.12 Recently,
Bertozzi and co-workers designed an αHER2 antibody-sialidase
Introduction
Sialic acids are a family of nonulosonic acids that are found as complex that selectively cleaves α-linked sialoglycans from
terminal residues in a wide variety of mammalian and prokar- breast cancer cells in order to induce an immune response
yotic cell lines. As a result of their terminal position in glyco- against breast cancer in mice.13
proteins and glycolipids, they have been shown to be impli-
As the amount of sialic acid molecules incorporated into
cated in numerous biologically important processes such as the cell is regulated by sialidase enzymes, methods to identify
cellular recognition, signalling and adhesion.1 Their presence and quantify sialidase activity is of paramount interest.7
on cell surfaces are critical to normal cellular function in Generally, the hydrolytic activity of sialidases can be quantified
mammals.2 Similarly, in bacterial and viral species, sialic acids by using synthetic fluorogenic substrates, such as 4-methyl-
are vital components for pathogenesis and bacterial nutrition.3,4 umbelliferylneuraminic acid, or other chromogenic or radio-
In nature, there are currently over 80 differently modified labelled substrates.14–17 While the α-4-methylumbelliferyl
members of the sialic acid family. As a starting point, we have N-acetylneuraminic acid (17α) is commercially available, the
selected to work on the most common C-5 modified sialic acid corresponding β-anomers (to probe the potential existence of
derivatives found in mammalian and bacterial cells (Fig. 1). β-sialidases) of common C-5 derivatives of sialic acid are not
The exact biological consequence of these modifications is still easily accessible. Additionally, the chemistry of 4-methyl-
an ongoing endeavour in glycobiology.5
umbelliferyl sialic glycosides has been left unexplored for
The amount of sialic residues incorporated into glycoconju-
gates within a given cell is typically regulated by the expression
levels and activity of sialyltransferase and sialidase enzymes.6
While sialyltransferases catalyse the addition of sialic acid
monomers to their corresponding glycoconjugate moieties,
sialidases (also referred to as neuraminidases, EC 3.2.1.18) cat-
alyse the cleavage of sialic acid residues from their component
glycan and peptide chains.7,8 Consequently, sialidases have
been shown to be critical components for a myriad of cellular
phenomena such as modulation of protein recruitment,9 cell
Centre for Synthesis and Chemical Biology, University College Dublin,
BelfieldDublin, Ireland. E-mail: stefan.oscarson@ucd.ie
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
d1ob00877c
Fig. 1 Selected C-5 derivatised neuraminic acid derivatives found in
nature.
6644 | Org. Biomol. Chem., 2021, 19, 6644–6649
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
many years despite existing synthetic routes suffering from solvent control methodology previously investigated by
poor yields, low stereoselectivity and formation of the unde- others.23
sired glycal side product.18–21 To this end, we report herein
Our synthesis towards the equatorial α-glycosides of the
synthetic strategies to access C-5 derivatised neuraminic acid desired C-5 functionalised Neu5Ac derivatives began with
substrates with a fluorogenic 4-methylumbelliferyl (4-Mu) agly- preparation of the common Neu5Ac thiophenol donor 7,
cone functionality in both anomeric configurations as poten- which was prepared in 2 steps from commercially available
tial tools to probe sialidase activity and specificity.
Neu5Ac.24 Initial efforts were then focused on derivatising this
shared intermediate at the C-5 position to subsequently intro-
duce our requisite fluorogenic glycosides. Accordingly, we con-
verted 7 into its corresponding N-glycolyl neuraminic acid by
condensation of free amine 8 with activated ester 9 in a
Results and discussion
Our
syntheses
starts
from commercially available mixture of CH3CN and water affording Neu5Gc thioglycoside
N-acetylneuraminic acid (1) (Scheme 1). The synthetic pathway 11 in 61% yield over three steps (Scheme 2).25 Simple replace-
is divided into two modular anomer-oriented approaches uti- ment of the benzylated activated ester into its acetylated form
lising glycal 5 and thiophenol 6 to stereoselectively install the 10 allowed for the gram scale synthesis of Neu5Gc donor 12 in
fluorogenic aglycone at the anomeric position. Our previously 89% yield over three steps. Donors 11 and 12 were prepared to
developed PhI(OAc)2/I2 mediated diastereoselective olefin determine if the electronic differences in the glycolyl side
addition reaction would give rise to the β-anomers,22 while we chain affected the stereochemical outcome of the glycosylation
envisioned that the α-anomers could be obtained by utilising reaction. Similarly, the free amine complex 8 was also treated
Scheme 1 Retrosynthetic analysis of C-5 derivatised sialic acid glycosides.
Scheme 2 Synthesis of acetyl and benzyl N-glycolyl, N-Troc, and Kdn thioglycoside derivatives.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 6644–6649 | 6645

View Article Online
Paper
Organic & Biomolecular Chemistry
with 2,2,2-trichloroethoxycarbonyl chloride (TrocCl) to chemo- acetonitrile (CH3CN), as an additive solvent known to increase
selectively mask the amine functionality. Acetylation of the α-selectivity in sialidations,23,29–32 significantly improved the
Troc protected derivative gave Neu5Troc thioglycoside 13 in α-selectivity (α/β 8.3 : 1) in addition to giving a higher isolated
78% yield over three steps (Scheme 2). Next, using a modified yield (entry 6). Interestingly, additive tetra-N-butylammonium
oxidative deamination methodology originally developed by iodide (TBAI, 0.8 eq.) in conjunction with our improved reac-
Zbiral26 and Ogura,27 and later applied to thioglycosides by tion conditions of NIS/AgOTf in the presence of CH3CN furn-
Crich et al.,28 the requisite Kdn thiophenyl sialoside was pre- ished complete α-selectivity (see ESI† for plausible mecha-
pared via treatment of 7 with nitrosonium tetrafluoroborate nism) with an excellent yield of 91% (entry 8). The addition of
(NOBF4) to furnish the corresponding nitrosyl amide inter- TBAI to our reaction mixture was inspired by seminal work on
mediate 14 in an almost quantitative yield (Scheme 2). glycosyl iodides in the research group of Gervay-Hague.34–36
Subsequent stepwise addition of sodium isopropoxide, tri- However, to the best of our knowledge, the anomeric selectivity
fluoroethanol and acetic acid to the N-nitrosoamide species of sialic acids was not investigated in these studies. Compound
gave the desired oxidatively deaminated Kdn thioglycoside 17α was straightforwardly deprotected, as described in the lit-
donor 15 in an overall 38% yield.28
erature,20 in an 80% yield (see ESI† for details).
With our C-5 derivatised neuraminic acid thioglycosides in
Thereafter, we applied the developed α-selective glycosyla-
hand, we began glycosylation studies towards the desired tion protocol to the other C-5 functionalised sialic acid donors
4-methylumbelliferyl α-glycosides. Control glycosylation reac- (Table 2). In all cases, high to excellent diastereoselectivity
tions between donor 7 and our 4-Mu acceptor 16 were carried towards the α anomer was observed, proving the optimised
out, using mainly N-iodosuccinimide (NIS) as thiophilic pro- glycosylation conditions to be quite general. Complete
moter, to optimize the reaction (Table 1).29–32
α-selectivity and excellent yield (89%, 21α) was observed with
Initially N-iodosuccinimide (NIS) and triflic acid (TfOH) in the benzylated Neu5Gc(Bn) donor 11. While good α-selectivity
CH2Cl2 were tested. Unsurprisingly, at ambient room tempera- (α/β = 4.3 : 1) was obtained with the acetylated Neu5Gc donor
ture, large quantities of the competing glycal side product 18 12, the isolated yield of the α -anomer product 22α was slightly
was observed in addition to the hydrolysed donor; 18% of 17α/ lower (72%) than its benzylated counterpart. However, in con-
β in an anomeric mixture of 1 : 2.2 (α/β) was also recorded trast to the Neu5Ac analogue (17α/β), chromatographic separ-
(entry 1).33 Lowering the temperature to either −40 °C (entry 2) ation of the anomers was not an issue. For the Troc-protected
or −78 °C (entry 3) gave similar stereoselectivity but higher donor 13, an α/β ratio of 19 : 1 of product 23α was obtained,
yield because of less elimination, as expected. The anomers affording a 75% yield of the α-anomer 23α after silica gel
were quite difficult to separate by column chromatography as column chromatography. Complete α-selectivity was again
their retention times on silica gel are very similar. Changing obtained for the acetyl protected Kdn glycosyl donor 15, giving
the promotor to dimethyl(methylthio)sulfonium trifluoro- 24α in 67% isolated yield. All reactions were performed on
methanesulfonate (DMTST, entry 4), resulted in the formation gram scale (1–4 g).
of the glycal side product 18 with only trace quantities of 17
Having developed a robust protocol to stereoselectively
observable by NMR. However, the glycosylation reaction using prepare fluorogenic sialic acid α-glycosides, we began our
NIS/AgOTf at −40 °C gave more promising results (entry 5) studies towards the corresponding β-configured 4-methyl-
with a yield of 72% and a slightly better preference for the umbelliferyl glycosides. To determine the feasibility of the
α-anomer 17α. Expanding on these conditions, introduction of diastereoselective glycal addition reaction previously developed
Table 1 Test glycosylation reaction
Yield of
Entry
1
Promoter (equivalents)
NIS/TfOH (1 : 1 eq.)
Solvent
CH2Cl2
Temp.
rt
Time
12 h
Product & anomeric ratioa
17b
Elimination (18), hydrolysis and
17 α/β = 1 : 2.2
18%
2
3
4
5
6
7
NIS/TfOH (1 : 1 eq.)
NIS/TfOH (1 : 1 eq.)
DMTST (1 eq.)
NIS/AgOTf (1 : 1 eq.)
NIS/AgOTf (1 : 1 eq.)
NIS/AgOTf/TBAI (1 : 1 : 0.4 eq.)
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2 : CH3CN (1 : 1)
CH2Cl2 : CH3CN (1 : 1)
−40 °C
−78 °C
0 °C
−40 °C → rt
−78 °C
−78 °C
12 h
24 h
16 h
12 h
16 h
16 h
18 & 17 α/β = 1 : 1.6
17α/β = 1.3 : 1
46%
68%
—
72%
87%
91%, α only
18 & trace of 17α/β
18, hydrolysis & 17α/β = 1.7 : 1
17α/β = 8.3 : 1
17α
a Determined by 1H NMR analysis of the crude reaction mixture. b Isolated yields.
6646 | Org. Biomol. Chem., 2021, 19, 6644–6649
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 Optimised conditions for glycosylation of C-5 functionalised sialic acid derivatives
R
Promoter (equivalents)
Solvent
Temp.
Time
Product & anomeric ratioa
Yield of α-anomerb
NH(BnGc), 11
NH(AcGc), 12
NHTroc, 13
OAc, 15
NIS/AgOTf/TBAI (1 : 1 : 0.4 eq.)
NIS/AgOTf/TBAI (1 : 1 : 0.4 eq.)
NIS/AgOTf/TBAI (1 : 1 : 0.4 eq.)
NIS/AgOTf/TBAI (1 : 1 : 0.4 eq.)
CH2Cl2 : CH3CN (1 : 1)
CH2Cl2 : CH3CN (1 : 1)
CH2Cl2 : CH3CN (1 : 1)
CH2Cl2 : CH3CN (1 : 1)
−78 °C
−78 °C
−78 °C
−78 °C
16 h
24 h
12 h
24 h
21α only
89%
72%
75%
67%
22α/β 4.3 : 1
23α/β 19 : 1
24α only
a Determined by 1H NMR analysis of the crude reaction mixture. b Isolated yields. 4 Å molecular sieves were used in all the reactions.
Scheme 3 Diastereoselective addition of 16 to 25 to furnish 27.
in our laboratory,22 we carried out an initial test reaction because of their easier accessibility, especially on a large scale,
with 9-O-trityl protected per-benzylated Neu5Ac glycal 25 and facile deprotection (no risk of reducing the 4-methyl-
(Scheme 3). Treatment of 25 with PhI(OAc)2, molecular iodine umbelliferyl moiety). Their respective glycals, 28–30, were
and 4-Mu acceptor 16 gave the desired β anomer 26 in 95% easily accessed from the previously prepared thioglycosides 12,
selectivity (α/β = 1 : 19). Importantly, since in enzyme reactions 13, and 15. The sialyl thioglycosides were initially converted
even small contaminations of the wrong anomer can give false into the corresponding bromides and in situ elimination of the
positives, it was found that the 3-iodo anomers were easily sep- resulting bromides by addition of Et3N furnished our desired
arated by silica gel column chromatography. Tributyl tin 2,3-anhydro derivatives 28–30 in yields ranging from 85–96%
hydride (Bu3SnH) reduction of the iodo function then afforded (see ESI† for synthesis of glycals). Glycal formation was con-
compound 27 in 89% overall yield. The anomeric configur- firmed by high resolution mass spectrometry and by the pres-
ation was determined by measuring 3JC-1, H-3 coupling constant ence of the diagnostic alkene H-3 proton peaks in 1H NMR
of a selectively decoupled coupled 13C NMR experiment. spectra at around 5.98 ppm. With the C-5 derivatised glycals in
Additionally, the relative anomeric ratio was measured by inte- hand, their conversion into the desired β analogues was inves-
grating the H-3eq proton after reduction of the iodo species.37 tigated (Scheme 4). Accordingly, the acetylated form of Neu5Gc
Degassing the reaction mixture via ‘freeze–pump–thaw’ (28) was treated with our optimised reaction conditions of PhI
method was critical for the successful reduction of the iodide (OAc)2 and I2 in a mixture of CH3CN and CH2Cl2 at room
species in this reaction.
temperature. Excess equivalents of 4-Mu (16, 5 eq.) were
Performing the same reaction but starting from the acetyl- necessary for higher yields. Completion of the reaction
ated Neu5Ac glycal 18 resulted in significantly lower diastereo- (between 20 and 30 min) was monitored by TLC and rapidly
selectity (1 : 2.1) and a 60% yield of 17β. However, since also worked up. It should be noted that leaving the reaction on for
these stereoisomers were easily separated, it was decided to longer than required results in a complex mixture. Again, a
extend this methodology to the other C-5 functionalized sialic relatively low α/β-ratio of 1 : 3 was observed. After anomer sep-
acid derivatives using the acetylated glycals as precursors aration followed by reduction of the iodide functionality using
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 6644–6649 | 6647

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 4 Reagents & conditions: (a) PhI(OAc)2, I2, 16, CH3CN : CH2Cl2, rt, 20–30 min; (b) Bu3SnH, AIBN, degassed toluene, reflux, 18 h.
Bu3SnH/AIBN, the β-4-Mu-Neu5Gc derivative 22β was isolated
in a 68% yield over the two steps.
Conflicts of interest
The NeuTroc (29) and Kdn (30) glycals were also subjected There are no conflicts to declare.
to the same addition and reduction conditions and the
desired 4-methylumbelliferyl β-glycosides 23 and 24 were iso-
lated in 71% and 62% yield respectively, following tributyl tin
hydride mediated reduction (Scheme 4). The observed lower
Acknowledgements
Financial support from Research Council of Norway
(BIOTEK2021, Nr 254780) and Science Foundation Ireland (13/
IA/1959) is gratefully acknowledged. Drs Yannick Ortin and
Jimmy Muldoon in University College Dublin are also thanked
for performing NMR and mass spectrometry experiments.
β-selectivity for 18 (1 : 2.1, α : β), 28 (1 : 3, α : β), 29 (1 : 2.7, α : β),
and 30 (1 : 3.5, α : β) was attributed to our chosen ester protect-
ing groups, but all the diastereomeric mixtures were easily sep-
arated before the tin hydride reduction. Consequently, the
decrease in stereoselectivity is offset by the ability to generate
large quantities of pure β-4-methylumbelliferyl sialic glyco-
sides following chromatographic separation of the 3-iodo
species. It should be noted that no previous syntheses of the
β-umbelliferyl sialic glycosides for Kdn, Neu, and Neu5Gc have
been reported in literature.
References
1 R. Schauer, Curr. Opin. Struct. Biol., 2009, 19, 507–514.
2 A. Varki and P. Gagneux, Ann. N. Y. Acad. Sci., 2012, 1253,
16–36.
3 E. Severi, D. W. Hood and G. H. Thomas, Microbiology,
2007, 153, 2817–2822.
Conclusions
4 S. L. Newstead, J. A. Potter, J. C. Wilson, G. Xu, C. H. Chien,
A. G. Watts, S. G. Withers and G. L. Taylor, J. Biol. Chem.,
2008, 283, 9080–9088.
5 A. Varki, R. D. Cummings, J. D. Esko, P. Stanley,
G. W. Hart, M. Aebi, A. G. Darvill, T. Kinoshita,
N. H. Packer, J. H. Prestegard, R. L. Schnaar and
P. H. Seeberger, Essentials of Glycobiology, 3rd edn, 2017.
6 R. B. Parker and J. J. Kohler, ACS Chem. Biol., 2010, 5, 35–46.
7 M. Saito and R. K. Yu, Biology of the Sialic Acids, 1995, pp.
261–313.
8 K. Mehr and S. D. Withers, Glycobiology, 2016, 26, 353–359.
9 A. Hinek, A. V. Pshezhetsky, M. von Itzstein and
B. Starcher, J. Biol. Chem., 2006, 281, 3698–3710.
10 H. Yamanami, K. Shiozaki, T. Wada, K. Yamaguchi,
T. Uemura, Y. Kakugawa, T. Hujiya and T. Miyagi, Cancer
Sci., 2007, 98, 299–307.
11 T. Miyagi and K. Yamaguchi, Glycobiology, 2012, 22, 880–896.
12 C. Tringali, B. Lupo, L. Anastasia, N. Papini, E. Monti,
R. Bresciani, G. Tettamanti and B. Venerando, J. Biol.
Chem., 2007, 282, 14364–14372.
In summary, a library of C-5 derivatised 4-methylumbelliferone
sialic acid derivatives (Neu, Neu5Ac, Neu5Gc, Kdn) were syn-
thesised in high yields from N-acetylneuraminic acid. The α
anomers were stereoselectively accessed by synergistic aceto-
nitrile-mediated solvent control, aided by additive TBAI.
Excellent to good yields and α-stereoselectivity was observed
for all the C-5 functionalised sialic derivatives. Similarly, the
β-anomers were selectively synthesised via
a preferential
diastereoselective addition reaction to our C-5 derivatised neura-
minic acid glycal analogues. Both anomer-oriented synthetic
pathways allow for the gram scale stereoselective synthesis of
the desired C-5 modified neuraminic acid derivatives for further
chemoenzymatic studies and should be applicable to other
sialic acid analogues. It is our hope that the synthesis of these
derivatives will allow for the fluorometric quantification of the
enzymatic activity and substrate specificity of known sialidases,
and the detection of novel sialidases found in nature.
Experimental section
13 M. A. Gray, M. A. Stanczak, N. R. Mantuano, H. Xiao,
J. F. A. Pijnenborg, S. A. Malaker, C. L. Miller,
Full experimental details and procedures are disclosed in the ESI.†
6648 | Org. Biomol. Chem., 2021, 19, 6644–6649
This journal is © The Royal Society of Chemistry 2021

View Article Online
Organic & Biomolecular Chemistry
Paper
P. A. Weidenbacher, J. T. Tanzo, G. Ahn, E. C. Woods, 24 A. Marra and P. Sinaÿ, Carbohydr. Res., 1989, 187, 35–42.
H. Läubli and C. R. Bertozzi, Nat. Chem. Biol., 2020, 16, 25 T. Sugata and R. A. Higuchi, Tetrahedron Lett., 1996, 37,
1376–1384.
14 V. P. Bhavanandan, A. K. Yeh and R. Carubelli, Anal. 26 E. Schreiner and E. Zbiral, Liebigs Ann. Chem., 1990, 581–
Biochem., 1975, 69, 385–394. 586.
15 G. A. Schwarzmann, Biochim. Biophys. Acta, Lipids Lipid 27 R. Shirai, M. Nakamura, S. Hara, H. Takayanagi and
Metab., 1978, 529, 106–114. H. Ogura, Tetrahedron Lett., 1988, 29, 4449–4452.
16 E. Zbiral, H. H. Brandstetter, R. Christian and R. Schauer, 28 D. Crich and C. Navuluri, Angew. Chem., Int. Ed., 2010, 49,
Liebigs Ann. Chem., 1987, 781–786. 3049–3052.
17 C. D. Hunter, E. M. Porter and C. W. Cairo, Carbohydr. Res., 29 A. Demchenko, T. Stauchand and G. J. Boons, Synlett, 1997,
2020, 497. 818–820.
18 E. Zbiral, E. Schreiner, M. M. Salunkhe, G. Schulz, 30 K. K. T. Mong, T. Nokami, N. T. T. Tranand and P. B. Nhi,
2613–2614.
R. G. Kleineidam and R. Schauer, Liebigs Ann. Chem., 1989,
519–526.
Selective Glycosylation: Synthetic Methods and Catalysts,
2017, pp. 59–77.
19 A. Kuboki, T. Sekiguchi, T. Sugai and H. A. Ohta, Synlett, 31 D. Crich, Acc. Chem. Res., 2010, 43, 1144–1153.
1998, 479–482.
32 C. de Meo and B. T. Jones, Adv. Carbohydr. Chem. Biochem.,
2018, 75, 215–316 and references cited therein.
33 D. Crich and W. Li, J. Org. Chem., 2007, 72, 2387–2391.
20 C. Y. Zamora, M. D’Alarcao and K. Kumar, Bioorg. Med.
Chem. Lett., 2013, 23, 3406–3410.
21 N. Baggett and B. J. Marsden, Carbohydr. Res., 1982, 110, 34 M. J. Hadd and J. Gervay, Carbohydr. Res., 1999, 320, 61–69.
11–18.
35 J. Gervay, T. N. Nguyen and M. J. Hadd, Carbohydr. Res.,
1997, 300, 119–125.
36 J. Gervay-Hague, Acc. Chem. Res., 2016, 49, 35–47.
22 A. A. Hassan and S. Oscarson, Eur. J. Org. Chem., 2020,
6102–6108.
23 O. Kanie, M. Kiso and A. Hasegawa, J. Carbohydr. Chem., 37 H. Hori, T. Nakajima, Y. Nishida, H. Ohruiand and
1988, 7, 501–506.
H. A. Meguro, Tetrahedron Lett., 1988, 29, 6317–6320.
This journal is © The Royal Society of Chemistry 2021
Org. Biomol. Chem., 2021, 19, 6644–6649 | 6649