Synthesis of the branched-chain sugar aceric acid: A unique component of the pectic polysaccharide rhamnogalacturonan-II

Article abstract of DOI:10.1021/jo051012b

Described herein is the synthesis of 3-C-carboxy-5-deoxy-L-xylose (aceric acid), a rare branched-chain sugar found in the complex pectic polysaccharide rhamnogalacturonan-II. The key synthetic step in the construction of aceric acid was the stereoselectiv

Full text of DOI:10.1021/jo051012b

SCHEME 1. Retrosynthetic Analysis for Synthesis
of Aceric Acid
Synthesis of the Branched-Chain Sugar
Aceric Acid: A Unique Component of the
Pectic Polysaccharide
Rhamnogalacturonan-II†
Nigel A. Jones,‡ Sergey A. Nepogodiev,*,‡
Colin J. MacDonald,‡ David L. Hughes,‡ and
Robert A. Field*,‡,§
School of Chemical Sciences and Pharmacy, University of
East Anglia, Norwich, NR4 7TJ, U.K., and Department of
Biological Chemistry, John Innes Centre, Norwich Research
Park, Norwich, NR4 7UH, U.K.
chains can cause severe plant growth defects.3 These
defects have been attributed to changes in primary cell
wall architecture, which is highly dependent on the
ability of RG-II to exist in a dimeric form that is
covalently cross-linked through a borate diester.3 Only
one particular monosaccharide residue of each RG-II
molecule, namely â-D-apiofuranose of side chain A, is
thought to be involved in borate diester formation.4 Side
chain B of RG-II also contains an apiofuranose residue
which is, however, not thought to be involved in borate
cross-linking.
Fragments of chains A and B will be useful for
investigating the borate complexation phenomenon; syn-
thesis of several fragments of side chain A has recently
been reported.5 Construction of side chain B requires
access to a unique branched-chain acidic monosaccharide
3-C-carboxy-5-deoxy-L-xylose, referred to as aceric acid
(Acef 1, Scheme 1).6 This branched sugar has only been
isolated6 in analytical quantities from RG-II itself and
has not previously been prepared by chemical synthesis.
We describe herein approaches to the synthesis of aceric
acid and its C-2 epimer from readily available monosac-
charide precursors.
s.nepogodiev@uea.ac.uk; r.a.field@uea.ac.uk
Received May 20, 2005
Described herein is the synthesis of 3-C-carboxy-5-deoxy-L-
xylose (aceric acid), a rare branched-chain sugar found in
the complex pectic polysaccharide rhamnogalacturonan-II.
The key synthetic step in the construction of aceric acid was
the stereoselective addition of 2-trimethylsilyl thiazole to
5-deoxy-1,2-O-isopropylidene-R-L-erythro-pentofuran-3-
ulose (2), which was prepared from L-xylose. The thiazole
group was efficiently converted into the required carboxyl
group via conventional transformations. Aceric acid was also
synthesized by dihydroxylation of a 3-C-methylene deriva-
tive of 2 followed by oxidation of the resulting hydroxyl-
methyl group. The C-2 epimer of aceric acid was also
synthesized using thiazole addition chemistry, starting from
L-arabinose.
The construction of aceric acid was investigated using
two different pathways, according to the retrosynthetic
analysis outlined in Scheme 1. In both cases L-xylose,
which has the correct stereochemistry at C-2 and C-4,
can serve as a convenient starting monosaccharide.
L-Xylose was trapped in the furanose form as a 1,2-O-
isopropylidene derivative7 and then transformed into
5-deoxy-3-ulose derivative 2, as reported8 for the D-
enantiomer. One commonly used approach in the syn-
thesis of branched-chain sugars with a branching ar-
The complex pectic polysaccharide rhamnogalactur-
onan-II (RG-II), which is found in the cell walls of all
higher plants, plays a key role in plant growth, develop-
ment, and resistance to disease.1 The backbone of RG-II
contains at least eight R-(1 f 4)-linked D-GalpA residues
to which are attached four structurally distinct oligosac-
charide side chains (designated A-D).2 It has been shown
that relatively minor changes in the structure of the side
† Dedicated to Professor Steve Ley on the occasion of his 60th
birthday.
(3) O’Neill, M. A.; Eberhand, S.; Albersheim, P.; Darvill, A. G.
Science 2001, 294, 846-849.
‡ University of East Anglia.
(4) Ishii, T.; Matsunaga, T.; Pellerin, P.; O’Neill, M. A.; Darvill, A.;
Albersheim, P. J. Biol. Chem. 1999, 274, 13098-13104.
(5) (a) Chauvin, A.-L.; Nepogodiev, S. A.; Field, R. A. Carbohydr.
Res. 2004, 339, 21-27. (b) Buffet, M. A. J.; Rich, J. R.; McGavin, R.
S.; Reimer, K. B. Carbohydr. Res. 2004, 339, 2507-2513. (c) Chauvin,
A.-L.; Nepogodiev, S. A.; Field, R. A. J. Org. Chem. 2005, 70, 960-
966.
§ John Innes Centre.
(1) (a) Ho¨fte, H. Science 2001, 294, 795-797. (b) Ridley, B. L.;
O’Neill, M. A.; Mohnen, D. A. Phytochemistry 2001, 57, 929-967. (c)
O’Neill, M. A.; Ishii, T.; Albersheim, P.; Darvill, A. G. Annu. Rev. Plant
Biol. 2004, 55, 109-139.
(2) (a) Darvill, A. G.; McNeil, M.; Albersheim, P. Plant Physiol. 1978,
62, 418-422. (b) Glushka, J. N.; Terrell, M.; York, W. S.; O’Neill, M.
A.; Gucwa, A.; Darvill, A. G.; Albersheim, P.; Prestegard, J. H.
Carbohydr. Res. 2003, 338, 341-352. (c) Rodr´ıguez-Carvajal, M. A.;
Herve´ du Penhoat, C.; Mazeau, K.; Doco, T.; Pe´rez, S. Carbohydr. Res.
2003, 338, 651-671. (d) Herve´ du Penhoat, C.; Gey, C.; Pellerin, P.;
Pe´rez, S. J. Biomol. NMR 1999, 14, 253-271.
(6) Spellman, M. W.; McNeil, M.; Darvill, A. G.; Albersheim, P.
Carbohydr. Res. 1983, 122, 115-129.
(7) Moravcova´, J.; Capkova´, J.; Stanek, J. Carbohydr. Res. 1994,
263, 61-66.
(8) Tronchet, J. M. J.; Graf, R. Helv. Chim. Acta 1972, 55, 613-
622.
10.1021/jo051012b CCC: $30.25 © 2005 American Chemical Society
Published on Web 09/09/2005
8556
J. Org. Chem. 2005, 70, 8556-8559

SCHEME 2
SCHEME 3
rangement similar to that found in aceric acid involves
addition of a nucleophile to an ulose.9 However, poten-
tially useful Grignard10 and organolithium11 reagents are
not likely to lead to the correct C-3 stereochemistry in
this example. The bicylo[3.3.0]octane ring system present
in 3-ulose 2 is expected to favor exo-addition of the
nucleophile to the keto group on steric grounds. An
example of highly stereoselective endo-addition to a
bicyclic 3-ketofuranose was reported12 in a reaction
utilizing 2-(trimethylsilyl)thiazole (2-TST) as a nucleo-
phile. Since the 2-TST group is a formyl group equiva-
lent,13 it is also a convenient precursor of the carboxylic
group, which can be easily generated by standard oxida-
tion of the formyl group.
afforded the same single stereoisomer 6, regardless of
whether AD mix R or AD mix â was used. One can
anticipate that the preferred face for dihydroxylation of
olefin 3 is exo with respect to the bicyclic ring system,
and therefore the reaction product 6 was tentatively
assigned as having the L-xylo-configuration. However, 2D
NOESY NMR spectroscopy data for 6, analyzed in the
absence of data for stereoisomeric compounds, were
inconclusive, leaving this assignment ambiguous. Fur-
thermore, a literature precedent15 shows that dihydroxy-
lation of a similar 3-C-alkene derivative of a 1,2-O-
isopropylidine derivative of a pentofuranose results in a
diol having the newly formed hydroxymethyl group anti
with respect to the isopropylidene group, indicative of
dihydroxylation of the more hindered endo-face of the
alkene. Due to the potential ambiguity in the structure
of 6, an alternative approach to installing the chiral
quaternary center was considered.
To investigate a thiazole addition approach to aceric
acid, 3-ulose 2 was treated with 2-TST in THF, under
thermodynamic control, followed by desilylation with
TBAF. This reaction sequence led to a 5:1 mixture of
endo- and exo-adducts 7 and 8, respectively, in an overall
yield of 70% (Scheme 3, Method A). In contrast, 2-thia-
zolyl lithium addition to compound 2 gave only the adduct
8 arising from sterically controlled exo-addition to the
trioxabicylo[3.3.0]octane ring system in an unoptimized
40% yield (Scheme 3, Method B). Both isomeric adducts
7 and 8 were benzylated to afford 3-O-benzyl derivatives
9 and 10, respectively. The stereochemistry of thiazole
addition to 3-ulose 2 was established by X-ray structural
analysis of compounds 7 and 10 (Figure 1), which
revealed that 10 has the required C-3 configuration for
aceric acid synthesis.
On the other hand, asymmetric dihydroxylation14 of
alkene 3, which can be prepared from 3-ulose 2 (Scheme
1) via a Wittig reaction, presents an alternative route to
aceric acid. In this approach, dihydroxylation should
proceed from the more accessible exo-face of the bicyclic
system, affording a branched-chain derivative with the
correct orientation of the 3-C-hydroxymethyl group rela-
tive to the furanose ring. Selective oxidation of the
primary OH group may then provide the required 3-C-
carboxy-derivative. Due to its simplicity, this approach
was investigated first.
1,2-O-Isopropylidene-L-xylofuranose (4), which is readily
accessible in 83% yield via a two-step procedure7 starting
from L-xylose, was converted into 5-deoxy sugar 5 in 56%
yield over two steps involving tosylation of the primary
OH group and reduction with LiAlH4 (Scheme 2). Com-
pound 5 was oxidized with pyridinium dichromate to
afford the unstable 3-ulose 2 in a 74% yield, which was
converted into the 3-C-methylene derivative 3, a low
boiling liquid, in 70% yield by reaction with H2CdPPh3.
Dihydroxylation of alkene 3 using Sharpless conditions14
(9) For reviews on the synthesis of branched-chain sugars, see: (a)
Yoshimura, J. Adv. Carbohydr. Chem. Biochem. 1984, 42, 69-134. (b)
Chapleur, Y.; Chre´tien, F. In Preparative Carbohydrate Chemistry;
Hanessian, S., Ed.; Marcel Dekker: New York, 1997; pp 207-262. (c)
Sviridov, A. F.; Shmyrina, A. Y.; Chizhov, O. S.; Kochetkov, N. K.
Bioorg. Khim. 1982, 8, 293-325.
(10) Dyer, J. R.; McGonigal, W. E.; Rice, K. C. J. Am. Chem. Soc.
1965, 87, 654-655.
(11) (a) Dahlman, O.; Garegg, P. J.; Mayer, H.; Schramek, S. Acta
Chem. Scand. 1986, 40, 15-20. (b) Sharma, G. V. M.; Reddy, V. G.;
Krishna, P. R.; Sankar, A. R.; Kunwar, A. C. Tetrahedron 2002, 58,
3801-3812. (c) Dondoni, A.; Marra, A.; Scherrmann, M. C. Tetrahedron
Lett. 1993, 34, 7323-7326.
(12) Carcano, M.; Vasella, A. Helv. Chim. Acta 1998, 81, 889-901.
(13) (a) Dondoni, A. Pure Appl. Chem. 1990, 62, 643-652. (b)
Dondoni, A.; Marra, A. Chem. Rev. 2004, 104, 2557-2599.
(14) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.;
Hartung, J.; Jeong, K.; Kwong, H.; Morikawa, K.; Wang, Z.; Xu, D.;
Zhang, X. J. Org. Chem. 1992, 57, 2768-2771.
The stereoselectivity of the reaction of 3-ketofuranose
2 with 2-thiazolyl lithium can be attributed to kinetic
control of the reaction and steric hindrance to addition
to the endo-face of the ketone. The mechanism of the
reaction of 2 with 2-TST is more complex: as noted for a
similar system,12 it is likely to involve several steps,
including equilibration of thiazolium anion intermedi-
ates, which allows stereodifferentiation between the endo
and exo-faces of the bicyclic molecule.
(15) Houseknecht, J. B.; Lowary, T. L. J. Org. Chem. 2001, 67,
4150-4164.
J. Org. Chem, Vol. 70, No. 21, 2005 8557

SCHEME 4
FIGURE 1. X-ray crystal structures of endo-thiazolyl deriva-
tive 7 and 3-O-benzylated exo-thiazolyl derivative 10.
Conversion of the thiazolyl group of compounds 9 and
10 into the formyl group was achieved using essentially
a one-pot, three-step literature method16 involving N-
methylation (MeOSO2CF3), reduction (NaBH4), and metal-
promoted hydrolysis (AgNO3 in aqueous MeCN) (Scheme
4). Aldehydes 11 and 12 were prepared this way in 80
and 72% yield, respectively. Oxidation of aldehyde 11 to
carboxylic acid 13, followed by removal of protecting
groups by successive catalytic hydrogenolysis (compound
14) and acid hydrolysis, afforded target aceric acid 1 in
an overall 86% yield starting from 11. The structure of
synthetic aceric acid 1 was confirmed by 1H and 13C NMR
data, which were in good agreement with those reported
for natural aceric acid isolated from primary cell walls
of sycamore (Acer pseudoplantanus).6
Certainty about the stereochemistry of aldehydes 11
and 12 (arising from crystal structures of 7 and 10,
respectively) made it possible to confirm the C-3 config-
uration of compound 6 obtained by the dihydroxylation
approach (Scheme 2). To make this comparison, alde-
hydes 11 and 12 were converted into 3-C-hydroxymethyl
derivatives 6 and 15, respectively, via NaBH4 reduction
followed by removal of 3-O-benzyl group by catalytic
hydrogenolysis. Diol 6 synthesized from the endo-thiaz-
olyl derivative 9 proved to be identical to that obtained
earlier from the dihydroxylation route. Completion of
aceric acid synthesis route required selective oxidation
of the primary hydroxyl group which was successfully
carried out using a TEMPO-based procedure,17 but
purification of carboxylic acid 14 obtained by this route
proved difficult. Hence, although requiring fewer steps,
the dihydroxylation route to aceric acid proved less
practical than the 2-trimethylsilylthiazole addition route.
The aceric acid residue present in RG-II (Scheme 1)
has a synthetically challenging 1,2-cis-configuration. Due
to the particular difficulty in creating a 1,2-cis-glycofura-
nosidic bond,18 an indirect approach,19 involving initial
formation of 1,2-trans-glycosides followed by inversion of
the C-2 configuration of the glycosyl residue, was con-
sidered. In the current context, this requires the avail-
ability of the C-2 epimer of aceric acid. The 3-keto-
furanose derivative 16, prepared by known methods10
from L-arabinofuranose, served as a convenient starting
material for introduction of the C-3 branch point (Scheme
5). Application of 2-thiazolyllithium as a nucleophile
afforded a single isomer, 17, with the required C-3
stereochemistry in 70% yield. The same exo-addition
product, 17, was identified as the major component of a
2:1 mixture of stereoisomers obtained in a combined 48%
yield in the reaction of 16 with 2-TST.20 After protecting
the 3-OH group in 17 via benzylation, 3-C-thiazolyl
derivative 18 was transformed into aldehyde 19 in 80%
overall yield, using the three-step procedure described
for the synthesis of L-xylo-isomer 11. Sodium chlorite
oxidation of 19 afforded 3-C-carboxy derivative 20, which
was deprotected using catalytic hydrogenolysis (com-
pound 21) and subsequent acid hydrolysis leading to the
target C-2 epimer of aceric acid, 22, in 74% overall yield
from 19. Similar to the case of aceric acid 1, NMR spectra
of 22 in D2O showed two sets of signals that indicated
the presence of a mixture of anomers in which the isomer
with 1,2-trans-configuration (R-anomer) dominates (R-22/
â-22 ) 2.6:1). Assignment of the anomeric configuration
of R-22 and â-22 was based on anomeric 13C chemical
shifts, which showed a difference of 4.2 ppm, consistent
with literature data.21
The structure of aldehyde 19 was confirmed by conver-
sion into known 1,2-O-isopropylidene derivative 23,22
(20) The presence of the endo-isomer of 17 as the minor component
in the mixture of reaction products was shown by NMR spectroscopy
and mass spectrometry (see Supporting Information). 2-TST addition
to ketone 16 appears to give a “mismatched” reaction, with the exo-
addition product formed in modest excess. In contrast, the correspond-
ing addition to diastereomeric ketone 2 appears to be “matched” with
respect to generation of the desired endo-adduct required for aceric
acid synthesis.
(16) (a) Altman, L. J.; Richheim, S. L. Tetrahedron Lett. 1971, 12,
4709-4711. (b) Dondoni, A.; Fogagnolo, M.; Medici, A.; Pedrini, P.
Tetrahedron Lett. 1985, 26, 5477-5480.
(17) de Nooy, A. E. J.; Besemer, A. C.; van Bekkum, H. Synthesis
1996, 1153-1174.
(18) Gadikota, R. R.; Callam, C. S.; Wagner, T.; Del Fraino, B.;
Lowary, T. L. J. Am. Chem. Soc. 2003, 125, 4155-4165.
(19) (a) David, S.; Malleron, A.; Dini, C. Carbohydr. Res. 1989, 188,
193-200. (b) Weiler, S.; Schmidt, R. R. Tetrahedron Lett. 1998, 39,
2299-2302.
(21) (a) Angyal, S. J.; Bethell, G. S. Aust. J. Chem. 1976, 29, 1249-
1265. (b) Bock, K.; Pedersen, C. Adv. Carbohydr. Chem. Biochem. 1983,
41, 27-66.
8558 J. Org. Chem., Vol. 70, No. 21, 2005

SCHEME 5
trans-orientation of these groups was established in
compounds 6 and 24 and a 1,2-cis-orientation in com-
pounds 15 and 25.23
In summary, we successfully synthesized aceric acid
1 from a readily available 1,2-O-isopropylidene derivative
of L-xylofuranose 2 using two different approaches. The
first approach was based on efficient, diastereoselective
2-trimethylsilylthiazole addition to 3-ulose 2 under ther-
modynamic control, followed by simple unmasking of the
3-C-formyl group and its oxidation to the 3-C-carboxy
group. In the second approach, dihydroxylation of a 3-C-
methylene derivative of L-xylofuranose 3 was applied,
followed by TEMPO-NaOCl oxidation of the primary OH
group. The overall yield starting from L-xylose was 13%
(over 10 steps) and 14% (over eight steps) in the first and
the second routes, respectively. In addition, the 2-epimer
of aceric acid, 22, was also synthesized starting from
L-arabinofuranose via derivative 16 (13 steps, 6% yield).
Experimental Section
5-Deoxy-1,2-O-isopropylidene-3-C-(thiazol-2-yl)-r-L-xy-
lofuranose (7) and 5-Deoxy-1,2-O-isopropylidene-3-C-
(thiazol-2-yl)-r-L-ribofuranose (8). 2-(Trimethylsilyl)thiazole
(1.46 mL, 9.13 mmol) was added to a stirred solution of
compound 2 (0.79 g, 4.57 mmol) in THF (10 mL), and the
resulting mixture was refluxed for 36 h. On cooling and
concentration in vacuo, the residue was dissolved in THF (10
mL) and treated with a 1 M solution of TBAF in THF (4.6 mL,
4.6 mmol). After being stirred for 30 min at room temperature,
the solution was concentrated in vacuo to afford a mixture of
isomeric thiazoles 7 and 8 in a 5:1 ratio (from integration of
anomeric signals in 1H NMR spectra). Individual products were
isolated by silica gel chromatography (hexane/EtOAc 8:1 f 4:1)
to give compound 7 as a crystalline solid (0.68 g, 58%) and
compound 8 as an oil (0.14 g, 12%). Compound 7: Rf ) 0.20
(hexane/EtOAc 1:1); mp 86-88 °C (EtOAc/hexane); [R]D -83 (c
1.0, CHCl3); IR (thin film): νmax 3100 (OH) cm-1; 1H NMR (400
MHz, CDCl3): δ 1.16 (d, J4,5 ) 6.4 Hz, 3 H, H-5), 1.36 (s, 3 H,
Me), 1.64 (d, 3 H, Me), 4.46 (q, J4,5 ) 6.4 Hz, 1 H, H-4), 4.51 (d,
J1,2 ) 3.7 Hz, 1 H, H-2), 6.07 (d, J1,2 ) 3.7 Hz, 1 H, H-1), 7.45
(d, JA,B ) 3.2 Hz, 1 H, thiazole), 7.80 (d, JA,B ) 3.2 Hz, 1 H,
thiazole); 13C NMR (100 MHz, CDCl3): δ 11.6 (C-5), 26.8 (Me),
27.0 (Me), 81.2 (C-4), 82.8 (C-3), 86.2 (C-2), 105.3 (C-1), 113.0
(CMe2), 121.9, 141.2, 166.5 (thiazole); HR ESI-MS: found m/z
258.0795 [M + H]+, calcd for C11H16NO4S 258.0795. Compound
8: Rf ) 0.30 (hexane/EtOAc 1:1); [R]D -88 (c 1.0, CHCl3); IR
(thin film) νmax 3109 (OH) cm-1; 1H NMR (400 MHz, CDCl3): δ
0.99 (d, J4,5 ) 6.4 Hz, 3 H, H-5), 1.42 (s, 3 H, Me), 1.66 (s, 3 H,
Me), 3.40 (broad s, 1 H, OH), 4.27 (q, J4,5 ) 6.4 Hz, 1 H, H-4),
4.65 (d, J1,2 ) 3.9 Hz, 1 H, H-2), 6.16 (d, J1,2 ) 3.9 Hz, 1 H,
H-1), 7.36 (d, JA,B ) 3.3 Hz, 1 H, thiazole), 7.82 (d, JA,B ) 3.3
Hz, 1 H, thiazole); 13C NMR (100 MHz, CDCl3): δ 13.6 (C-5),
26.8 (Me), 26.9 (Me), 79.3, 82.7, 83.3 (C-2, C-3, C4), 105.0 (C-1),
113.2 (CMe2), 119.3, 143.1, 169.9 (thiazole); HR ESI-MS: found
m/z 258.0792 [M + H]+, calcd for C11H16NO4S 258.0795.
which represents a protected form of the well-known
branched-chain sugar streptose,10 a key component of the
antibiotic streptomycin. NaBH4 reduction and subse-
quent catalytic hydrogenolysis of aldehyde 19 gave diol
24, a diastereomer of diols 6 and 15. The fourth diaste-
reomer of this series of branched-chain sugar diols,
namely 25, was synthesized starting from 3-ketofuranose
16. The latter was first converted into its 3-C-methylene
derivative and then dihydroxylated, using the same
conditions as applied for transformation of 3-ketofuranose
2, giving diol 25.
In addition to crystallographic data for 3-C-thiazolyl
derivatives 7 and 10, assignment of the C-3 configuration
in the synthetic branched-chain sugars was accomplished
using selective 1D NOE experiments. Investigation of a
systematic series of diastereomeric 1,2-O-isopropylidene
derivatives 6, 15, 24, and 25 revealed NOEs indicative
of the relative orientation of the hydroxymethyl group
and the C-5 methyl group on the furanose ring. A 1,2-
Acknowledgment. This research was supported by
the U.K. Engineering and Physical Sciences Research
Council and the Weston Foundation. We are indebted
to the EPSRC Mass Spectrometry Service Centre,
Swansea for invaluable support.
Supporting Information Available: Experimental pro-
cedures, characterization data and NMR spectra for com-
pounds 1, 2, 6-15, 17-25, X-ray data for compounds 7 and
10, 1D NOE spectra for compounds 6, 15, 24, and 25. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(22) Paulsen, H.; Sinnwell, V.; Stadler, P. Chem. Ber. 1972, 105,
1978-1987.
(23) NOEs for compounds 6, 15, 24, and 25 were in the 2-3% range.
Details of the observed NOEs, which were obtained for these com-
pounds upon selective irradiation of the signals corresponding to H-1,
H-2, H-3, H-4, and H-3′, are summarized in Supporting Information.
JO051012B
J. Org. Chem, Vol. 70, No. 21, 2005 8559