Rhamnogalacturonan II: Chemical Synthesis of a Substructure Including α-2,3-Linked Kdo**

Article abstract of DOI:10.1002/chem.202100837

The synthesis of a fully deprotected Kdo-containing rhamnogalacturonan II pentasaccharide is described. The strategy relies on the preparation of a suitably protected homogalacturonan tetrasaccharide backbone, through a post-glycosylation oxidation approach, and its stereoselective glycosylation with a Kdo fluoride donor.

Full text of DOI:10.1002/chem.202100837

Communication
doi.org/10.1002/chem.202100837
Chemistry—A European Journal
Rhamnogalacturonan II: Chemical Synthesis of a
Substructure Including α-2,3-Linked Kdo**
Enzo Mancuso,[a] Cecilia Romanò,*[a] Nino Trattnig,[b] Philipp Gritsch,[c] Paul Kosma,[b] and
Mads H. Clausen*[a]
Abstract: The synthesis of a fully deprotected Kdo-contain-
ing rhamnogalacturonan II pentasaccharide is described.
The strategy relies on the preparation of a suitably
protected homogalacturonan tetrasaccharide backbone,
through a post-glycosylation oxidation approach, and its
stereoselective glycosylation with a Kdo fluoride donor.
Rhamnogalacturonan II (RG-II) is a complex polysaccharide
representing around 10% of pectin in the cell wall of higher
plants.[1] Its structure is highly conserved among plant species
and it is constituted by a homogalacturonan (HG, GalA-α-(1!
4)-GalA) backbone decorated with side branches, named AÀ F,
consisting of 13 different types of monosaccharides intercon-
nected by over 20 different glycosidic linkages (Figure 1).[2,3]
Interestingly, the monosaccharides composing the side-chains
include a number of rare sugars such as 3-deoxy-d-manno-oct-
2-ulosonic acid (Kdo), l-aceric acid (Ace), d-apiose (Api), 3-
deoxy-d-lyxo-hept-2-ulosaric acid (Dha), l-galactose, 2-O-meth-
yl-d-xylose, and 2-O-methyl-l-fucose. Unlike rhamnogalactur-
onan I (RG-I), the RG-II structure has been found to be
evolutionarily conserved in all vascular plants, with minor
variations limited to methylation and/or acetylation of the side
Figure 1. RG-II polysaccharide structure and target pentasaccharide 1.
chains.[3] RG-II exists predominantly as a dimer that is covalently
cross-linked by a 1:2 borate-diol diester between apiofuranosyl
residues belonging to side chain A of two different RG-II
monomers.[4,5] This important structural characteristic contrib-
utes to the generation of networks of pectic polysaccharides by
favouring cell adhesion and wall mechanical strength. The
fundamental role of RG-II in the life of plants is further
supported by the observation that genetic modifications
affecting RG-II structure dramatically decrease the formation of
the dimer, in turn affecting plant growth and development.[6,7]
Furthermore, mutations preventing the synthesis of sugar
nucleotides UDP-Api and CMP-Kdo are lethal and provide
further evidence for the essential role of RG-II in plant
growth.[8–10] Though the complexity of the RG-II structure
requires the involvement of multiple glycosyltransferases (GTs),
activated donors, and additional methylating/acetylating en-
zymes for its biosynthesis, to date only rhamnogalacturonan
xylosyl transferases RGXT1-4 (CAZy family GT77) have been
characterized for the synthesis of the A side chain (transferring
α-(1!3)-d-xylose on the internal l-fucose).[11–13] Additional 26
putative GTs have been listed by a bioinformatic approach.[14]
The growing interest towards the plant biomass as an
abundant reservoir for the production of next generation
biofuels and value-added products requires in-depth under-
standing of the structure of plant cell wall components,
specifically their biosynthesis and degradation. In light of this,
[a] Dr. E. Mancuso, Dr. C. Romanò, Prof. Dr. M. H. Clausen
Department of Chemistry, Center for Nanomedicine and Theranostics
Technical University of Denmark
Kemitorvet 207, 2800, Kgs. Lyngby (Denmark)
E-mail: ceroma@kemi.dtu.dk
mhc@kemi.dtu.dk
[b] Dr. N. Trattnig, Dr. P. Kosma
Department of Chemistry
University of Natural Resources and Life Sciences
18 Muthgasse, 1190 Vienna (Austria)
[c] Dr. P. Gritsch
Institute of Applied Synthetic Chemistry TU Wien
Getreidemarkt 9, 1060 Vienna (Austria)
[**] Kdo=3-deoxy-d-manno-oct-2-ulosonic acid.
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/chem.202100837
Chem. Eur. J. 2021, 27, 7099–7102
7099
© 2021 Wiley-VCH GmbH

Communication
doi.org/10.1002/chem.202100837
Chemistry—A European Journal
synthetic chemistry can help by furnishing well-defined oligo-
saccharide fragments that can be used for probing the activity
of newly isolated hydrolytic enzymes, GTs involved in the
biosynthesis of plant cell wall components, or in the production
of novel monoclonal antibodies (mAbs) for elucidation of
differences in plant cell wall components in mutants, organ and
tissue types, and different developmental stages.[15,16] Notably,
the plant/algal glycome differs in its composition from the
mammalian and bacterial ones, with an higher incidence of
uronic acids (d-GalpA, d-GlcpA), furanoses (d/l-Araf), and rare
sugars such as d-Api and l-Gal.[17,18] The essential role of RG-II in
pectin structure has prompted several synthetic efforts towards
the preparation of small oligosaccharide fragments of this
complex structure. In particular, much attention has been
focused to the preparation of portions of side chains A and
B,[19–24] especially containing the rare sugar d-Api. Conversely,
no synthesis has been reported of side chain B fragments
containing l-Ace residues, possibly due to its tedious synthesis
from l-xylose or d-arabinose.[25,26] Additionally, to date no RG-II
fragments containing the smaller side chains C and D have
been synthetically targeted, although several procedures have
been described to access Kdo,[27–29] a common monosaccharide
found in bacterial polysaccharides, and Dha.[30] In this context,
we have been interested in the development of a synthetic
strategy to access the HG backbone of RG-II and its further
functionalization with a branching Kdo moiety, in order to
obtain a unique RG-II pentasaccharide fragment including the
unexplored C side chain (Figure 1, compound 1). The synthesis
of target pentasaccharide 1 was envisioned to proceed via the
preparation of the linear tetrasaccharide HG backbone utilizing
four orthogonally protected d-galactose building blocks and a
post-glycosylation oxidation approach (Scheme 1). Following, a
[4+1] glycosylation was planned to insert the branching Kdo
pyranose moiety. After preparation of the planned building
blocks 2–5 (Supporting Information), the assembly started with
a NIS/TESOTf promoted glycosylation with glycosyl acceptor 2
and donor 3 yielding disaccharide 6 in 72% yield. Careful
tyl ester by Zemplén conditions (!7, 96%) and glycosylation of
the newly formed disaccharide acceptor 7 with glycosyl donor
4, under the established conditions of the [1+1] glycosylation,
yielded trisaccharide 8 in 85% yield and as a single α-anomer.
Chemoselective removal of the chloroacetyl ester was then
achieved in 79% yield by treatment of 8 with thiourea together
°
with NaHCO3 and tetrabutylammonium iodide (TBAI) at 55 C in
THF. Finally, the same glycosylation protocol was applied for
the reaction of trisaccharide acceptor 9 and donor 5 to discover
that tetrasaccharide 10 was only synthesized in a poor 34%
yield and as a 1:1 α:β mixture, possibly a consequence of the
decreased reactivity of the two coupling partners. Thus,
glycosylation yield and selectivity were improved by performing
°
the reaction in Et2O and by increasing the temperature to 0 C,
leading to the formation of tetrasaccharide 10 in 68% yield and
as a separable 5:1 α:β mixture. Removal of the 2-naphthylmeth-
yl (NAP) groups protecting the primary C-6 alcohols also proved
challenging, affording, at its best, the desired tetraol 11 in 52%
yield after treatment with DDQ in CH2Cl2:MeOH:H2O at room
temperature. Finally, the newly formed hydroxyl groups were
oxidized to the corresponding carboxylic acids via a two-step
protocol involving treatment with Dess-Martin periodinane
followed by sodium chlorite oxidation, and subsequently
protected as benzyl esters after reaction with phenyldiazo-
methane (!12, 72% over three steps). The obtained HG
backbone 12 was then reacted under standard Zemplén
conditions to remove the acetyl ester protecting the 3’’ À OH.
However, the reaction produced a number of byproducts that
were identified by LC-MS as a mixture of tetrasaccharides
without the 3’’ À OAc moiety and transesterified to the
corresponding methyl esters. Further investigations into a
viable deacetylation protocol were carried out (e.g. change in
equivalents of MeONa, DBU/MeOH, KCN/MeOH, AcCl/MeOH)
with noticeably poor results in the tested conditions. The
disappointing results prompted the search of a novel approach
towards the preparation of tetrasaccharide acceptor 13. Gratify-
ingly, when Zemplén removal of the acetyl ester was performed
after the two-step oxidation and before the benzylation of the
formed carboxylic acids, desired product 13 could be obtained
°
temperature control (À 40 C) led to the exclusive formation of
the desired α-product. Subsequently, removal of the chloroace-
°
Scheme 1. Reagents and conditions: (i) NIS, TESOTf, 4 Å MS, CH2Cl2, À 40 C, 1 h, 72%; (ii) MeONa, MeOH, THF, RT, 1.5 h, 96%; (iii) NIS, TESOTf, 4 Å MS, CH2Cl2,
°
°
°
À 40 C, 2 h, 85%; (iv) thiourea, NaHCO3, TBAI, THF, 55 C, 33 h, 79%; (v) NIS, TESOTf, 4 Å MS, Et2O, 0 C, 1 h, 68% (α:β=5:1); (vi) DDQ, CH2Cl2/MeOH/H2O, RT,
12 h, 52%; (vii) for 12: a. Dess-Martin periodinane, CH2Cl2, RT, 12 h, b. NaClO2, NaH2PO4, H2O, THF, t-BuOH, RT, 4 h, c. PhCHN2, AcOEt, RT, 3 h, 72% (over three
steps), for 13: a. Dess-Martin periodinane, CH2Cl2, RT, 12 h, b. NaClO2, NaH2PO4, H2O, THF, t-BuOH, RT, 4 h, c. MeONa, MeOH, THF, RT, on, d. PhCHN2, AcOEt, RT,
1.5 h, 68% (over four steps); (viii) MeONa, MeOH, THF, RT, on, 81%.
Chem. Eur. J. 2021, 27, 7099–7102
www.chemeurj.org
7100
© 2021 Wiley-VCH GmbH

Communication
doi.org/10.1002/chem.202100837
Chemistry—A European Journal
in 68% yield over four reaction steps. The key glycosylation
reaction for the introduction of the desired Kdo side-chain
involved the use of Kdo donors 15, 16, and 21 (Scheme 2). In
particular, donor 21 represents the ideal glycosylation candi-
date as its COOBn functionality can be removed at a late stage
of the synthesis during global deprotection and without any
additional saponification steps. While glycosyl donors 15[31] and
16[32] were prepared according to published procedures, Kdo
fluoride donor 21 was prepared in five steps from Kdo
ammonium salt, previously obtained from d-arabinose and
oxaloacetic acid following the Cornforth reaction.[33,34] Namely,
Kdo ammonium salt 17[34] was first converted into the
corresponding carboxylic acid by stirring in water in the
presence of Dowex H+ resin (!18, 94%), then the acid was
treated with an ethereal solution of phenyldiazomethane,
followed by the introduction of the 4,5-O- and 7,8-O-isopropy-
lidene rings (!19, 48% over two steps). Compound 19 was
then acetylated (!20, 64%) and reacted with 70% HF-pyridine
systems, such as TESOTf, TMSOTf, TfOH were tested in CH2Cl2
and at low temperatures. These results prompted the testing of
a new set of glycosylations employing di-O-isopropylidene Kdo
fluoride donor 16. Product formation was observed when the
°
reaction was first performed with BF3·Et2O in CH2Cl2 at À 40 C,
although in a complex mixture with hydrolysed donor and the
glycal ester by-product. Repeating the reaction at a higher
temperature showed a decrease in product formation, while
increasing the amount of promoter was beneficial. Other
promoting systems (TMSOTf or TfOH) were also tested, without
observing any product formation. The screened conditions with
model acceptor 14 and donors 15–16 indicated that fluoride
donor 16 tended to be more reactive, although more labile in
comparison to the acetylated 3-iodo-donor 15. In fact, possibly
due to the di-O-isopropylidene protecting moieties, fluoride 16
seemed more prone to decomposition and/or elimination side-
reactions, especially with little variations in terms of equivalents
of promoter or reaction temperature. In light of these
preliminary results, glycosylation of HG acceptor 13 was tested
°
at low temperature (À 70 C to RT) to afford the corresponding
°
fluoride donor 21 in 42% yield. To the best of our knowledge,
this route developed to access Kdo donor 21 represents a
shorter and higher-yielding protocol compared to other
synthetic pathways that have been previously reported. With
glycosyl donors 15, 16, and 21 in hand the glycosylation
reaction was first screened with model tetrasaccharide acceptor
14, obtained from the deacetylation of tetrasaccharide 10 (!
14, 81%). While arguably the reactivity of model acceptor 14
can be assumed to be different from the GlcA-containing
glycosyl acceptor 13, testing several glycosylation conditions
with the less precious material 13 allowed for a more systematic
search of optimal glycosylation conditions (Table S1 in Support-
ing Information; summary of [4+1] glycosylation attempts).
Acceptor 14 was initially tested in glycosylations with Kdo
fluoride donor 15. The glycosylation was first carried out with
BF3·Et2O in CH2Cl2. The role of the temperature was studied by
with donor 16 in the presence of BF3·Et2O (2 equiv.) at 0 C. The
reaction led to the formation of the corresponding pentasac-
charide product 22 in 26% yield and as a single α-isomer. This
result prompted the design of a protocol in which 1 equiv. of
promoter (BF3·Et2O) would be added every 3 hours at low
temperature to minimize the side-reactions leading to acetal
cleavage and to suppress donor elimination. Gratifyingly,
coupling between Kdo donor 21 and acceptor 13 following the
newly designed protocol afforded the desired pentasaccharide
product 23, although with loss of the 7,8-O-isopropylidene
acetal, in 31% yield and with complete α-stereoselectivity.
While not high yielding, the coupling proved successful and
enough material was produced to continue towards the
preparation of the desired final compound 1. To this end,
hydrolysis of the 4,5-O-isopropylidene acetal with 80% aq.
AcOH and global deprotection via hydrogenolysis produced the
desired RG-II fragment 1 in 61% yield over 2 steps (Scheme 3).
In conclusion, the synthesis of the first RG-II side-chain C
fragment was accomplished in 10 steps starting from four d-Gal
building blocks and a suitably protected Kdo fluoride donor.
The [4+1] synthetic strategy here presented allows for the easy
°
°
performing experiments at À 40 C, 0 C and room temperature
but in all three cases, a complex mixture of degradation
products was obtained. Donor elimination was observed when
the solvent was changed to CH3CN or different promoting
Scheme 2. Kdo donors 15, 16, 21. Reagents and conditions: (i) Dowex
50WX8 H+, 94%; (ii) a. PhCHN2, CH3CN, H2O, RT, 1 h, b. acetone, p-TsOH, RT,
on, 48% (over two steps), (iii) Ac2O, DMAP, pyridine, 2 h, 64%, (iv) 70% HF-
°
Scheme 3. Reagents and conditions: (i) for 22: 16, BF3·Et2O, 4 Å MS, 0 C, 1 h,
°
26%, for 23: 21, BF3·Et2O, 4 Å MS, À 40 C, 9 h, 31%; (ii) a. 80% aq. AcOH, 2 h,
°
pyridine, pyridine, À 70 C to RT. 2 h, 42%.
b. H2, 5% Pd/C, AcOEt/EtOH/H2O, o/n, 85% (over two steps).
Chem. Eur. J. 2021, 27, 7099–7102
www.chemeurj.org
7101
© 2021 Wiley-VCH GmbH

Communication
doi.org/10.1002/chem.202100837
Chemistry—A European Journal
[9] J. Petit, T. Paccalet, M. Hernould, P. Lerouge, A. Mouras, C. Chevalier,
assembly of
a suitably protected HG backbone and its
Plant Physiol. 2003, 133, 348–360.
subsequent branching. In particular, the modularity of the
approach makes it easily translatable to the synthesis of RG-II
fragments carrying other short side-chains such as D, E, and F.
In addition, while developing the synthesis of the RG-II side-
chain C fragment, a novel synthetic approach for the prepara-
tion of Kdo donor 21 was also developed with a shorter and
higher-yielding protocol (5 steps, 12% global yield).
[10] M. Mølhøj, R. Verma, W. D. Reiter, Plant J. 2003, 35, 693–703.
[11] J. Egelund, B. L. Petersen, M. S. Motawia, I. Damager, A. Faik, C. E. Olsen,
T. Ishii, H. Clausen, P. Ulvskov, N. Geshi, Plant Cell 2006, 18, 2593–2607.
[12] J. Egelund, I. Damager, K. Faber, C. E. Olsen, P. Ulvskov, B. L. Petersen,
FEBS Lett. 2008, 582, 3217–3222.
[13] X. L. Liu, L. Liu, Q. K. Niu, C. Xia, K. Z. Yang, R. Li, L. Q. Chen, X. Q. Zhang,
Y. Zhou, D. Ye, Plant J. 2011, 65, 647–660.
[14] A. Voxeur, A. André, C. Breton, P. Lerouge, PLoS One 2012, 7, 1–14.
[15] I. Moller, I. Sørensen, A. J. Bernal, C. Blaukopf, K. Lee, J. Øbro, F.
Pettolino, A. Roberts, J. D. Mikkelsen, J. P. Knox, et al., Plant J. 2007, 50,
1118–1128.
[16] C. Ruprecht, M. P. Bartetzko, D. Senf, P. Dallabernadina, I. Boos, M. C. F.
Andersen, T. Kotake, J. P. Knox, M. G. Hahn, M. H. Clausen, et al., Plant
Physiol. 2017, 175, 1094–1104.
Acknowledgements
[17] K. S. Egorova, A. N. Kondakova, P. V. Toukach, Database 2015, 2015, 1–
P.K. acknowledges financial support from the Austrian Science
Fund FWF (grant P 26919). M.H.C. and C.R. acknowledge
financial support from the Novo Nordisk Foundation (grant nos.
NNF18OC0053048 and NNF20OC0065094). M.H.C. and E.M.
acknowledge financial support from the Independent Research
Fund Denmark (grant no. 1335-00152). M.H.C. acknowledges
financial support from the Carlsberg Foundation (grant no.
CF19-0072).
22.
[18] A. Adibekian, P. Stallforth, M.-L. Hecht, D. B. Werz, P. Gagneux, P. H.
Seeberger, Chem. Sci. 2011, 2, 337–344.
[19] A. L. Chauvin, S. A. Nepogodiev, R. A. Field, Carbohydr. Res. 2004, 339,
21–27.
[20] M. A. J. Buffet, J. R. Rich, R. S. McGavin, K. B. Reimer, Carbohydr. Res.
2004, 339, 2507–2513.
[21] S. A. Nepogodiev, R. A. Field, I. Damager, Annu. Plant Rev. 2010, 41, 65–
92.
[22] I. Backman, P. E. Jansson, L. Kenne, J. Chem. Soc. Perkin Trans. 1 1990,
50, 1383–1388.
[23] A. L. Chauvin, S. A. Nepogodiev, R. A. Field, J. Org. Chem. 2005, 70, 960–
966.
[24] Y. Rao, G. J. Boons, Angew. Chem. Int. Ed. 2007, 46, 6148–6151; Angew.
Conflict of Interest
Chem. 2007, 119, 6260–6263.
[25] N. A. Jones, S. A. Nepogodiev, C. J. MacDonald, D. L. Hughes, R. A. Field,
J. Org. Chem. 2005, 70, 8556–8559.
The authors declare no conflict of interest.
[26] M. S. M. Timmer, B. L. Stocker, P. H. Seeberger, J. Org. Chem. 2006, 71,
8294–8297.
Keywords: carbohydrates · glycosylation · Kdo · plant cell wall ·
[27] P. Kosma, Tetrahedron Lett. 2016, 57, 2133–2142.
[28] T. K. Pradhan, K. K. T. Mong, Isr. J. Chem. 2015, 55, 285–296.
[29] L.-S. Li, Y.-L. Wu, Curr. Org. Chem. 2003, 7, 447–475.
[30] A. Banaszek, Tetrahedron 1995, 51, 4231–4238.
[31] B. Pokorny, P. Kosma, Chem. Eur. J. 2015, 21, 305–313.
[32] N. Trattnig, J. B. Farcet, P. Gritsch, A. Christler, R. Pantophlet, P. Kosma, J.
Org. Chem. 2017, 82, 12346–12358.
rhamnogalacturonan
[1] D. Mohnen, Curr. Opin. Plant Biol. 2008, 11, 266–277.
[2] D. Ndeh, A. Rogowski, A. Cartmell, A. S. Luis, A. Baslé, J. Gray, I. Venditto,
J. Briggs, X. Zhang, A. Labourel, et al., Nature 2017, 544, 65–70.
[3] B. M. Yapo, Int. J. Carbohydr. Chem. 2011, 2011, 1–11.
[4] B. L. Ridley, M. A. O’Neill, D. Mohnen, Phytochemistry 2001, 57, 929–967.
[5] M. A. O’Neill, D. Warrenfeltz, K. Kates, P. Pellerin, T. Doco, A. G. Darvill, P.
Albersheim, J. Biol. Chem. 1996, 271, 22923–22930.
[33] J. W. Cornforth, M. E. Firth, A. Gottschalk, Biochem. J. 1958, 68, 57–61.
[34] R. Shirai, H. Ogura, Tetrahedron Lett. 1989, 30, 2263–2264.
[6] A. Fleischer, C. Titel, R. Ehwald, Plant Physiol. 1998, 117, 1401–1410.
[7] T. Ishii, T. Matsunaga, N. Hayashi, Plant Physiol. 2001, 126, 1698–1705.
[8] F. Delmas, M. Séveno, J. G. B. Northey, M. Hernould, P. Lerouge, P.
McCourt, C. Chevalier, J. Exp. Bot. 2008, 59, 2639–2647.
Manuscript received: March 6, 2021
Accepted manuscript online: March 26, 2021
Version of record online: April 9, 2021
Chem. Eur. J. 2021, 27, 7099–7102
www.chemeurj.org
7102
© 2021 Wiley-VCH GmbH