Studies related to Norway spruce galactoglucomannans: Chemical synthesis, conformation analysis, NMR spectroscopic characterization, and molecular recognition of model compounds

Article abstract of DOI:10.1002/chem.201200510

Galactoglucomannan (GGM) is a polysaccharide mainly consisting of mannose, glucose, and galactose. GGM is the most abundant hemicellulose in the Norway spruce (Picea abies), but is also found in the cell wall of flax seeds, tobacco plants, and kiwifruit. Although several applications for GGM polysaccharides have been developed in pulp and paper manufacturing and the food and medical industries, attempts to synthesize and study distinct fragments of this polysaccharide have not been reported previously. Herein, the synthesis of one of the core trisaccharide units of GGM together with a less-abundant tetrasaccharide fragment is described. In addition, detailed NMR spectroscopic characterization of the model compounds, comparison of the spectral data with natural GGM, investigation of the acetyl-group migration phenomena that takes place in the polysaccharide by using small model compounds, and a binding study between the tetrasaccharide model fragment and a galactose-binding protein (the toxin viscumin) are reported. Copyright

Full text of DOI:10.1002/chem.201200510

FULL PAPER
DOI: 10.1002/chem.201200510
Studies Related to Norway Spruce Galactoglucomannans: Chemical
Synthesis, Conformation Analysis, NMR Spectroscopic Characterization, and
Molecular Recognition of Model Compounds
Filip S. Ekholm,[a] Ana Ardꢀ,[b] Patrik Eklund,[a] Sabine Andrꢁ,[c]
Hans-Joachim Gabius,[c] Jesffls Jimꢁnez-Barbero,[b] and Reko Leino*[a]
Abstract:
Galactoglucomannan
distinct fragments of this polysacchar-
ide have not been reported previously.
Herein, the synthesis of one of the core
trisaccharide units of GGM together
with a less-abundant tetrasaccharide
fragment is described. In addition, de-
tailed NMR spectroscopic characteriza-
tion of the model compounds, compari-
son of the spectral data with natural
GGM, investigation of the acetyl-group
migration phenomena that takes place
in the polysaccharide by using small
model compounds, and a binding study
between the tetrasaccharide model
fragment and a galactose-binding pro-
tein (the toxin viscumin) are reported.
(GGM) is a polysaccharide mainly con-
sisting of mannose, glucose, and galac-
tose. GGM is the most abundant hemi-
cellulose in the Norway spruce (Picea
abies), but is also found in the cell wall
of flax seeds, tobacco plants, and kiwi-
fruit. Although several applications for
GGM polysaccharides have been de-
veloped in pulp and paper manufactur-
ing and the food and medical indus-
tries, attempts to synthesize and study
Keywords: conformation analysis ·
galactoglucomannan
·
molecular
recognition · NMR spectroscopy ·
oligosaccharide synthesis
Introduction
pounds.[1a,4] Consequently, a deeper understanding of the
role of GGM in pulp and paper manufacturing has been
achieved together with emerging applications of GGM in
immunology.[5] GGM has also been shown to be a promising
substitute of the widely utilized gum arabic (arabinogalac-
tan) as an emulsifier of hydrophobic beverages.[1a]
Galactoglucomannan (GGM) is the most abundant hemicel-
lulose in the Norway spruce (Picea abies).[1] GGM has also
been found in several other plants and fruits including club
mosses, horsetails, whisk ferns, kiwifruit, and tobacco
plants.[2] The utilization of water-soluble polysaccharides, in
particular GGM, in mechanical pulping and papermaking
has traditionally been of great importance.[3] Recently, meth-
ods for the large-scale isolation of acetylated GGM
(AcGGM) from P. abies have been developed, thus simulta-
neously sparking new interest in the applications and chemi-
cal properties of these compounds. Polysaccharide material
suited for such work can now be isolated on the kilogram
scale, thus providing sufficient amounts of material for
screening and extending the applications of these com-
Several earlier studies have focused on the structural elu-
cidation of water-soluble AcGGM from P. abies.[6] It has
been suggested that the main chain consists of randomly dis-
tributed b-(1!4)-linked mannopyranosyl and b-(1!4)-
linked glucopyranosyl units. Furthermore, the presence of a-
(1!6)-linked galactopyranosyl residues has been confirmed
in several studies.[6a,7] In all previous investigations, the gal-
actopyranosyl units were found to be linked to the manno-
pyranoside residues of the linear backbone of GGM. These
monosaccharides have been suggested to be present in a
mannose/glucose/galactose (Man/Glc/Gal) ratio of 4:1:0.1,
although this ratio is strongly influenced by the isolation
conditions.[8] In addition, the acetyl-group pattern of
AcGGM from P. abies with an acetylation degree of 0.3 has
been analyzed, and the acetyl groups are, according to the
currently accepted view, mainly present at the O2 and O3
atoms of the mannopyranosyl units (Figure 1).[1a,6a]
Although AcGGM can be readily isolated on a large
scale, several ambiguities concerning the structure and ace-
tylation pattern of this material remain. In fact, determina-
tion of structure and size of AcGGM has been very difficult
due to the heterogeneous nature of the native material. This
behavior is reflected in the wide range of molecular weights
(i.e., 20–78 kDa) reported for these polysaccharides.[1a,4] In
[a] Dr. F. S. Ekholm, Dr. P. Eklund, Prof. Dr. R. Leino
Laboratory of Organic Chemistry
ꢀbo Akademi University
Piispankatu 8, 20500 ꢀbo (Finland)
E-mail: reko.leino@abo.fi
[b] Dr. A. Ardꢁ, Prof. Dr. J. Jimꢂnez-Barbero
Chemical and Physical Biology
Centro de Investigathiones Biolꢃgicas, CSIC
Ramiro de Maeztu 9, 28040 Madrid (Spain)
[c] Priv.-Doz. Dr. S. Andrꢂ, Prof. Dr. H.-J. Gabius
Institute of Physiological Chemistry
Faculty of Veterinary Medicine
Ludwig-Maximilians-University Munich
Veterinꢄrstr. 13, 80539 Munich (Germany)
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

Figure 1. One of the possible chemical structures of the AcGGM backbone.
addition, acetyl-group migration has been proposed to occur
from O2 and O3 to even the O6 position in the mannopyra-
nosyl residues.[1a] An investigation of such phenomena is dif-
ficult to perform on the heterogeneous polysaccharide sam-
ples available. Therefore, we became interested in the syn-
thesis and possible utilization of model compounds. Herein,
the synthesis of one of the trisaccharide repeating units of
AcGGM, a less-abundant tetrasaccharide fragment contain-
ing an a-(1!6)-linked galactopyranosyl residue and some
model substrates for acetyl-group migration studies are re-
ported. In addition, a complete NMR spectroscopic charac-
terization and conformation analysis of the model com-
pounds and a comparison of the NMR spectroscopic data of
model compounds with natural GGM polysaccharide are
presented. The tetrasaccharide fragment bearing an a-(1!
6)-linked galactopyranosyl residue was further utilized in a
binding study with the galactose-binding protein viscumin,
an AB toxin, to test bioaffinity.
Figure 2. The synthetic targets representing one of the trisaccharide re-
peating units of GGM 1 (top) and a less-abundant tetrasaccharide frag-
ment 2 (bottom).
Results and Discussion
Figure 3. Building blocks prepared for the synthesis of trisasaccharide 1.
Synthesis of model compounds: The applications of
AcGGM have in recent years rapidly expanded from its tra-
ditional roles in pulp- and paper-making processes to appli-
cations in the food and medicinal industries.[1a,3,5] From a bi-
ochemical standpoint, the roles of the individual fragments
of the GGM polysaccharide would be interesting to study as
such, particularly in view of immunological applications.
Previously, this investigation would not have been possible
due to the heterogeneous nature of the native samples. In
the case of AcGGM, the individual fragments contain chal-
lenging glycosidic linkages, including b-mannopyranosyl-
and a-galactopyranosyl-linked sugar units, which for a long
time were considered difficult to construct by chemical
methods. In the present study, the synthesis of two oligosac-
charide fragments found in AcGGM was planned to provide
material for such studies (Figure 2).
For the initial attempt on the synthesis of trisaccharide 1,
a linear strategy was devised that started from the three
monosaccharide building blocks 3–5 (Figure 3). The synthe-
sis of donor 3 has been addressed previously and was based
on previously reported procedures.[9] Acceptor 4 was pre-
pared by the route shown in Scheme 1. The commercially
available methyl mannopyranoside 6 was converted into 7
Scheme 1. Synthesis of acceptor 4. i) 1. PTSA, C6H5OCHACTHUNTGRNE(UNG OMe)2, 2 h,
608C, 200 mbar (60%); 2. BnBr, NaH, DMF, 3 h, RT (88%); ii) TFA,
Et3SiH, CH2Cl2, 2 h, RT (50%). PTSA=para-toluenesulfonic acid,
TFA=trifluoroacetic acid.
by utilizing one equivalent of benzaldehyde dimethyl acetal
under slightly acidic conditions, in analogy to the procedures
described by McGowan and Berchtold in 1982.[10a] The re-
maining hydroxyl groups were benzylated by using the
standard protocol involving NaH and BnBr. Next, the ben-
zylidene acetal was partially cleaved by the reductive meth-
odology of DeNinno et al.[11a] and Arasappan and Fraser-
Reid[11b] to give the 4-OH/6-OBn-containing acceptor 4 in
moderate yield. A similar efficiency was also witnessed pre-
viously in the partial cleavage of benzylidene-protected
mannopyranosides under these conditions.[12]
&
2
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
With building blocks 3 and 4 in hand, attention was
turned to the preparation of building-block donor 5, which
was synthesized according to the synthetic route shown in
Scheme 2. Peracetylated d-glucopyranose was subjected to
tallization of disaccharide 10 in the presence of donor 5 was
impossible due to the small scale of the synthesis. In addi-
tion, the partial reductive cleavage, following the protocol
described by Arasappan and Fraser-Reid, resulted in a poor
yield of the desired disaccharide 11.[11b]
Consequently, the synthetic strategy required revision. As
the selective ring-opening reaction of the benzylidene acetal
moiety in disaccharide 10 turned out to be difficult, the syn-
thetic approach was changed from linear to convergent. The
convergent route would add two potential benefits to the
synthetic pathway: Importantly, the b-mannopyranosyl link-
age could be constructed first, followed by convergent cou-
pling of disaccharide to acceptor 4.
Scheme 2. Synthesis of donor 5. i) 1. SHPh, BF3·OEt2, CH2Cl2, 24 h, RT
(96%; a/b 1:9); 2. NaOMe, MeOH/THF 3:1, 24 h, RT (95%); 3. PTSA,
C6H5OCHACHTUNGTRENNUNG(OMe)2, DMF, 608C, 200 mbar (78%; pure b anomer);
ii) BzCl, pyridine, 5 h, RT (90%). BzCl=benzoyl chloride.
In addition, the selective ring-opening reaction of the ben-
zylidene acetal unit could now be performed on the glucose
building block instead of the disaccharide. To avoid the
problematic crystalline nature displayed by the dibenzoylat-
ed glucoside, the benzoyl groups were simultaneously re-
placed by acetyl groups. The synthesis of trisaccharide 1 is
shown in Scheme 4. Acceptor 12 was synthesized from 9 by
acetylation of the free hydroxyl group followed by a selec-
tive ring-opening reaction of the benzylidene acetal moiety
by the use of methods described previously.[11] In this ap-
proach, the yield proved to be excellent. The large variation
in the efficiency of the methodology described by Arasap-
pan and Fraser-Reid was highly dependent on the stereo-
chemistry of the individual sugars involved in the reac-
tions.[11b] Similar results have been reported previously.[12]
Next, donor 3 was activated following the b-mannosylation
protocol of Crich et al. and then coupled with acceptor
12.[15] The glycosylation reaction proceeded with moderate
efficiency and high selectivity (b/a=10:1). The moderate ef-
ficiency observed is most likely due to the sluggish reaction
and the small amounts (~15% each) of starting material/hy-
drolyzed donor present in the reaction mixture. Although
not optimized further and explored in the present study, the
yield of this individual glycosylation reaction could possibly
be enhanced by the addition of 1-octene to the reaction mix-
ture after the activation of donor 3. In earlier reports, 1-
octene was shown to be a mild reagent capable of trapping
the thioadducts formed as side products during the activa-
tion and conversion of donor 3 into the activated glycosyl
triflate.[16] Donor 13 was then coupled with acceptor 4 under
the reaction conditions optimized in the earlier attempt dis-
played in Scheme 4 (0.06 equivalents of TMSOTf). The con-
vergent glycosylation proceeded smoothly to give the pro-
tected trisaccharide 14 in 77% yield of isolated product. A
BF3·OEt2-promoted glycosylation followed by deacetylation
under conditions developed by Zemplꢂn et al.[13] The benzy-
lidene-protected thioglycoside 9 was synthesized according
to previously described procedures.[10] As expected, the
yields were significantly higher in the benzylidene formation
with the glucopyranoside relative to those of the mannopyr-
anosides. This outcome is mainly due to the trans relation-
ship of the 2- and 3-hydroxyl groups in the glucose moiety,
which significantly decreases formation of the dibenzylidene
side product formed with mannopyranosides. The remaining
hydroxyl groups were benzoylated by means of standard
methods by using BzCl in pyridine. With all of the building
blocks prepared, the focus was next shifted to assembly of
trisaccharide 1.
Following a linear pathway, donor 5 was first coupled with
acceptor 4 in the presence of the conventional NIS/TMSOTf
promoter/activator system (Scheme 3).[14] Several difficulties
were encountered during the first attempts of this glycosyla-
tion reaction. In our initial approach, by following the stand-
ard activation protocol with 0.12 equivalents of TMSOTf, a
moderate 47% yield of the desired product was obtained
with the formation of several side products. When the
amount of TMSOTf was decreased to 0.06 equivalents, the
yield of the desired disaccharide 10 increased. The main dif-
ficulty encountered during the synthesis was the highly crys-
talline nature of the benzoylated glucose donor 5 and disac-
charide 10. The characteristic feature of these compounds,
namely, the formation of solids/crystals, posed a major con-
cern during chromatographic purification because the prod-
ucts partially solidified in the column (despite being highly
soluble in the solvent system) and were lost, thereby lower-
ing the amount of isolable material. Furthermore, the crys-
two-step
deprotection
se-
quence that consists of deace-
tylation[13b] followed by hydro-
genolysis gave the trisacchar-
ide 1 in 84% overall yield.
With the trisaccharide re-
peating unit synthesized, the
focus was shifted toward the
assembly of the less-abundant
tetrasaccharide fragment 2.
Scheme 3. Synthesis of disaccharide 11. i) NIS (1.2 equiv), TMSOTf (0.12 equiv), CH2Cl2, 1.5 h, ꢀ258C (47%);
ii) NIS (1.2 equiv), TMSOTf (0.06 equiv), CH2Cl2, 1.5 h, ꢀ258C (76%); iii) TFA, Et3SiH, CH2Cl2, RT, 1–5 h
(<15%). NIS=N-iodosuccinimide, TMSOTf=trimethylsilyl trifluoromethanesulfonate.
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

R. Leino et al.
Scheme 4. Synthesis of GGM core trisaccharide 1. i) 1. Ac2O, pyridine, RT, 5 h (95%); 2. Et3SiH, TFA, 08C!RT, 3 h (85%); ii) 1. BSP, TTBP, Tf2O,
CH2Cl2, ꢀ608C, 0.5 h: 2. 12, ꢀ788C, 2 h (53%); iii) 4, NIS, TMSOTf, CH2Cl2, ꢀ408C, 1.5 h (77%); iv) 1. NaOMe, MeOH/THF (1:1), RT, 19 h (84%);
2. Pd/C, H2 (2.8 bar), MeOH, RT, 19 h (quant.). BSP = 1-(phenylsulfinyl)piperidine, TTBP = 2,4,6-tri-tert-butyl pyrimidine, Tf2O = triflic anhydride.
Trisaccharide 14 was chosen as a suitable starting material
for the synthesis of tetrasaccharide 2. Once again, selective
ring opening of the benzylidene acetal was required as a key
transformation in the preparation of trisaccharide acceptor
16. In this approach, however, the reverse selectivity in the
ring-opening reaction was desired (6-OH/4-OBn) as the gal-
actopyranosyl residue in 2 is a-(1!6)-linked to the sugar
backbone. For this purpose, several procedures have been
reported based on, for example, PhBCl2/Et3SiH,[17] LiAlH4/
coordinates to the 6-OH group of the sugar, thus providing
the 4-OBn derivative as the product.[21] Consequently, in the
present study, it became important to investigate different
Lewis acids under similar conditions. A procedure using Cu-
AHCTUNGRTEG(NNUN OTf)2 as the Lewis acid in the selective ring-opening reac-
tion of benzylidene acetals has been described recently by
Shie et al.[20] Initial attempts that utilized this protocol re-
sulted in complete cleavage of the benzylidene acetal
moiety. The solvent was identified as the major source of
the problems in this reaction, and monosaccharide 15 could
be isolated in high yield after switching from THF to
CH2Cl2. Although this procedure appeared to function well
for monosaccharides, only a few reports on the applications
of similar protocols to oligosaccharides have been described
previously.
ACTHNUTRGNEUNG
AlCl3,[18] TMSOTf/BH3·THF,[19] and Cu(OTf)2/BH3·THF.[20]
Instead of trisaccharide 14, the appropriately protected
monosaccharide 7 was utilized in the initial screening and
optimization of reaction conditions (the results are summar-
ized in Scheme 5). In our hands, several of the previously re-
When applying similar conditions to the protected trisac-
charide 14, the trisaccharide acceptor 16 was obtained in ex-
cellent yield (Scheme 6). This outcome is, to our knowledge,
Scheme 5. Screening of methodologies for the selective ring-opening re-
action of benzylidene acetals. i) PhBCl2, Et3SiH, ꢀ788C, CH2Cl2, 2–4 h
(35%); ii) LiAlH4, AlCl3, CH2Cl2/Et2O 4:1, RT!508C, 1 h (35%);
iii) TMSOTf, BH3·THF, CH2Cl2, RT, overnight (50%); iv) Cu
BH3·THF, CH2Cl2, RT, overnight (82%).
ACHTUNGERTN(NUNG OTf)2,
Scheme 6. Synthesis of the trisaccharide acceptor 16. i) CuACHTUNGTRENNUNG(OTf)2,
BH3·THF, CH2Cl2, RT, overnight (85%).
ported procedures did not work with the efficiencies report-
ed.[17–19] In the initial attempts, both the PhBCl2/Et3SiH[17]
[18]
and LiAlH4/AlCl3 protocols were quickly excluded from
one of the first times in which a similar procedure has been
applied to an oligosaccharide. The high efficiency and selec-
tivity observed under the conditions explored are encourag-
ing for future applications of this methodology.
With the trisaccharide acceptor prepared, the challenge
concerning a-galactosylation remained. For this purpose,
donors 17 and 18 were prepared according to reported pro-
cedures by using the routes shown in Scheme 7.[22] Both of
these donors have in recent years been applied to the a-gal-
actosylation of various acceptors and were therefore also
screened in the present study. Donor 17 has previously been
shown to result in reactions with high a-selectivity and effi-
further use due to poor efficiencies. By considering the mod-
erate efficiency obtained by the TMSOTf/BH3·THF proto-
col,[19] a continued investigation into borane-based systems
was launched. The selectivity in the partial cleavage of ben-
zylidene acetals is generally a combination of steric and
electronic factors. It has been suggested that 6-OBn deriva-
tives are formed when borane is activated by the Lewis acid,
thereby coordinating to the more nucleophilic 6-OH group.
In cases in which the borane moiety is not activated under
the reaction conditions employed, the Lewis acid becomes
the most electrophilic species in the reaction mixture and
&
4
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
oligosaccharides and aglycones.[22b,c,25] In the present study,
the methodology was applied to the a-selective galactosyla-
tion of 16, thus providing tetrasaccharide 19 in good yield as
shown in Scheme 8. Whereas the yield was 68% and the se-
lectivity excellent, the formation of a second tetrasaccharide
was also observed. Based on the JH1,H2 coupling constant in
1
the H NMR spectrum of this tetrasaccharide byproduct, it
appeared to be a-linked,[26] as it was most likely an orthoest-
er of the product. With the glycosylation efficiency already
in the desired range, further attempts to optimize the exper-
imental conditions were not made. The origin of the high a-
selectivity in the glycosylation reaction is puzzling because
ester-protecting groups are present at both the O2 and O3
positions of donor 18. A likely explanation for the high a-se-
lectivity may, at least in part, rest on the steric and electron-
ic properties induced by the 4,6-silylene acetal moiety, as
also suggested previously by Kiso and co-workers.[27] For the
synthesis of the deprotected tetrasaccharide 2, the silylene
acetal of 19 was cleaved in 68% yield by using a HF/pyri-
dine complex.[28] Next, the ester groups were demasked
under conditions similar to the ones developed by Zemplꢂn
et al.[13b] with NaOMe in a MeOH/THF mixture followed by
hydrogenolysis of the benzyl groups in 92% overall yield
over two steps.
Scheme 7. Synthesis of donors 17 and 18. i) 1. Ac2O, NaOAc, reflux, 2 h
(50%; pure b isomer); 2. SHEt, BF3·OEt2, CH2Cl2, 08C, 4 h (95%);
3. NaOMe, MeOH/THF (4:3), RT, overnight (quant.); 4. BnBr, NaH,
DMF, RT, 20 h (80%); ii) 1. Ac2O, NaOAc, reflux, 2 h (50%; pure b
isomer); 2. SHPh, BF3·OEt2, CH2Cl2, 08C, 4 h (95%); 3. NaOMe,
MeOH/THF (1:1), RT, overnight (quant.); 4. 1) Di-TBSACHTUNGTRENNUNG
RT, 3 h; 2) BzCl, RT, 2 h (57%). Di-TBS
flate.
ACHTUNGTRENNUNG
ciency in ethereal solvents including Et2O.[23] As a result,
these conditions were first screened in the a-galactosylation
of acceptor 16 by using the NIS/TMSOTf promoter
system.[14] However, only moderate-to-low selectivities (a/
b=4:1) were achieved under these conditions because the
separation of the two diastereomers by column chromatog-
raphy proved to be challenging. Although several variations
in solvent proportions (Et2O/CH2Cl2) and the activation/pro-
moter systems were tried, the selectivity of the glycosylation
could not be improved. Therefore, another methodology
using CuBr2, TBAB, AgOTf in dichloroethane and DMF,
utilized successfully earlier by Mukherjee and Misra for a-
galactosylation of a tetrasaccharide acceptor, was tested.[24a]
In the present study, however, this protocol failed to yield
the desired product.
With one of the core trisaccharide repeating units (i.e., 1)
and a less-abundant tetrasaccharide unit (i.e., 2) synthesized,
our attention could be turned to the investigation of the
acetyl-group migration phenomena that occurs during the
isolation of native GGM.
Acetyl-group migration in GGM: The acetylation degree of
GGM is approximately 0.3, which varies slightly depending
on the methods used in the isolation process.[1a,6] The acetyl
groups are mainly located at the O2 and O3 positions of the
mannopyranosyl residues. It has been proposed that acetyl
groups may migrate either directly from axial O2 to O6 or
from O3 to O6 in the mannopyranosyl units.[1a] Migration
between O2 and O3 is known to take place in substrates
The given experience caused us to turn our attention to
donor 18. This donor has been applied previously by Ando
and co-workers for the a-selective galactosylation of several
Scheme 8. Synthesis of tetrasaccharide 2. i) 18, NIS, TfOH (or TMSOTf), CH2Cl2, 08C, 1.5 h (68%); ii) HF/pyridine, THF, 08C!RT, 6 h (68%); iii) 1.
NaOMe, MeOH/THF (1:1), RT, overnight (96%); 2. Pd/C, H2 (2.8 bar), MeOH, RT, overnight (96%).
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

R. Leino et al.
that contain acetyl protecting groups and will not be dis-
cussed in more detail in the present context. The migration
from O2 and O3 to O6 is unique as it would require the
acetyl group to “leap” across the molecule in space. This
phenomenon is difficult to study on the native polysacchar-
ide due to a severe overlap of the signals in the NMR spec-
tra. Consequently, small molecular models were synthesized
to study this process. Mannopyranosides are known to exist
mainly in the energetically favored 4C1 conformation. On
the basis of this assumption, a few differently acetylated
mannopyranoside model compounds were synthesized
(Scheme 9). The synthesis of these compounds followed
standard protecting-group manipulations and is thus not
presented in more detail.[29]
Figure 4. High-energy conformations of mannopyranosides in which
acetyl-group migration could be possible. Left!right: 1C4, B1,4, 5H4.
similar conditions to explore the possibility of acetyl-group
migration comprehensively. In these conformers, the acetyl
groups are significantly closer to the O6 atom and migration
4
could be more favored than in the C1 conformation. Such
an approach should further be combined with molecular
modeling to create appropriate models to investigate the
phenomenon. The synthesis of such molecules for further
studies on the acetyl-migration behavior of GGM is current-
ly planned in our laboratories. The acetylated trisaccharide
fragment representing the second repeating unit in the
GGM backbone might prove to be an additional probe uti-
lizable in studies on the acetyl-group migration/cleavage
patterns of GGM.
NMR spectroscopic comparison of the model compounds
1
versus natural GGM: The complete H and 13C NMR spec-
troscopic characterization of both 1 and 2 was required to
compare the NMR spectroscopic data of the synthetic
model structures and natural GGM (Figure 2). The NMR
spectroscopic characterization of oligosaccharides can be
difficult due to severe signal overlap in the 1H NMR spectra.
Previously, we have described the complete NMR spectro-
scopic characterization of a b-(1!2)-linked mannotetraose
by using standard NMR spectroscopic experiments (1H; 13C,
1D-TOCSY, COSY, HSQC (coupled and decoupled), and
HMBC) in combination with spectral simulations with the
PERCH software.[30] A similar set of NMR spectroscopic ex-
periments was used for the spectral assignment reported
herein. The NMR spectra were recorded at 308 K (in D2O)
to shift the water peak, which would otherwise overlap with
Scheme 9. i) 1. Butanedione, trimethyl orthoformate, CSA, MeOH,
reflux, 24 h (79%); 2. TBDMSCl, imidazole, DMF, 08C!RT, 24 h, 90 %;
3. Ac2O, pyridine, dichloromethane, 48 h (92%); 4. 70% HF/pyridine,
THF, 08C!RT, 24 h (93%); ii) 1. PTSA, C6H5OCH
ACHTUNGRTEN(NGNU OMe)2, DMF, 608C,
200 mbar, 2 h (53%); 2. a) Bu2SnO, toluene, 1208C, 3 h; 2) TBAB, CsF,
BnBr, 1208C, 3 h (85%); 3. Ac2O, pyridine, RT, 3 h (93%); 4. 80%
AcOH, 708C, 1.5 h (87%). CSA=camphorsulfonic acid, TBAB=tetra-
butylammonium bromide, TBDMSCl=tert-butyldimethylsilyl chloride.
In this initial study, the possibilities of acetyl-group migra-
4
tion from O2 to O6 in C1-mannopyranosides were evaluat-
4
ed. Compound 21, locked in the C1-conformation by the di-
1
acetal protecting group between positions 3 and 4, was se-
lected as a suitable model compound to investigate the
O2!O6 acetyl-group migration. No migration was observed
when this compound was subjected to conditions similar to
the ones used for the isolation of native GGM (758C, D2O/
DMSO (dimethyl sulfoxide) 1:1). One possible reason for
the absence of migration could be the low degree of flexibil-
ity in this model molecule. Mannoside 22 was synthesized
for verification of the hypothesis. Although 22 exhibits a
much higher degree of flexibility, migration did not take
place in this compound either.
On the basis of these results, it seems quite unlikely that
migration would occur in the GGM polysaccharide from O2
to O6 as has been suggested previously.[1a] More model com-
pounds will, however, be needed to investigate the acetyl-
group migration from O3 to O6. Appropriately protected
and “locked” high-energy conformations (such as those de-
picted in Figure 4) should be synthesized and studied under
the anomeric protons. The H NMR spectrum of 1 displays
three well-separated anomeric protons at d=4.76 (d, H1),
4.74 (d, H1’), and 4.52 ppm (d, H1’). These signals were also
targeted by selective excitation with 1D-TOCSY[31] with a
spin-lock time of 300 ms. The results of these experiments
are visualized in Figure 5. By use of COSY, HSQC (coupled
and decoupled), and HMBC both the 1H and 13C NMR
spectra of 1 could be completely assigned.
This information, however, was not sufficient to define ac-
curate coupling constants. For this purpose, the NMR spec-
tral simulation software PERCH (Peak Research)[32] was
1
utilized, thus leading to accurate H NMR spectral simula-
tion, coupling constants, and coupling patterns (Figure 6).
The discussion of the NMR spectroscopic characterization
of oligosaccharides is exemplified and limited to the com-
plete assignment of 1. A similar set of NMR spectroscopic
techniques in combination with spectral simulations was ap-
plied to yield the NMR spectroscopic characterization of all
&
6
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
main signals for which large deviations have been reported
is the H2 signal of the b-(1!4)-linked mannopyranoside
residue. This signal has been reported to appear in natural
GGM both at d=3.08 and 4.08 ppm.[6a,8a,33] In both model
compounds 1 and 2, the chemical shift for H2 was close to
d=4.08 ppm, which suggests this value to be the correct
one. In fact, all the other signals from this residue are found
at similar frequencies in both the model compounds and
GGM. These data suggest either an assignment error or ty-
pographical mistake in the reports for the chemical shift of
H2 in GGM reported by Hannuksela and Hervꢂ du Penhoa-
t.[6a] The chemical shifts for H6a, H6b, and C6 of the b-(1!
4)-linked glucopyranoside residues have not been reported
for GGM, but differed by less than d=0.2 ppm in both
model substances and could, therefore, be assumed to be
similar in the polysaccharide structure. The H5 and C5
chemical shifts of the b-(1!4)-linked glucopyranoside resi-
dues of GGM appeared at a higher frequency than in model
compounds 1 and 2. As the values are almost exactly similar
for all the other signals in this residue, there might be some
uncertainties in the chemical shifts reported for GGM previ-
ously. Surprisingly, there is very little reported data available
for the b-(1!4)-linked mannopyranoside residue bearing an
a-(1!6)-linked galactopyranosyl unit. These chemical shifts
were easily identified in the model compounds, thus comple-
menting the available reported data. Interestingly, the H5
proton is shifted toward a higher frequency and the C5
carbon atom toward a lower frequency when the b-(1!4)-
linked mannopyranosyl unit is substituted by an a-(1!6)-
linked galactopyranosyl residue. A further notable differ-
ence in the chemical shifts of the a-(1!6)-galactosylated
mannopyranosyl unit is the large difference for the C6
carbon atom, which is located at d=66.8 ppm in the substi-
tuted mannopyranosyl residue versus d=61.5 ppm in the un-
substituted residue. These values should provide a conven-
ient reference point for future determinations of the amount
of 6-substituted b-(1!4)-linked mannopyranosyl units in the
GGM backbone.
Figure 5. 1D-TOCSY spectra of the different residues in 1. Top!bottom:
1
Manp’’, Glcp’, Manp, and the entire H NMR spectrum of 1.
Although slight deviations were found in the NMR spec-
troscopic characterization of GGM relative to results of re-
spective studies on model tri- and tetrasaccharides, most sig-
nals were found at positions that were in good agreement
with previously reported spectral data. These results indicate
that it is indeed possible to infer the composition of polysac-
charides by characterizing small molecular models of limited
size. To further evaluate and continue the study on the
GGM polymer, several other small molecular fragments
should be synthesized and analyzed. The most important
among these fragments should be the synthesis and charac-
terization of the second repeating unit bearing acetyl
groups.
Figure 6. Spectral simulation of the 1H NMR spectrum of 1 (d=4.08–
3.38 ppm region) with the PERCH NMR software. Top: simulated spec-
trum, bottom: observed spectrum.
the other compounds presented herein. With the NMR spec-
tra of model compounds 1 and 2 solved, our attention was
turned to the comparison of the chemical shifts obtained
1
with those of natural GGM. The H and 13C NMR chemical
shifts of 1 and 2 together with reported values of the chemi-
cal shifts of GGM[6a,8a,33] are summarized in Table 1. The
coupling constants and patterns observed in the 1H NMR
spectra of 1 and 2 and the JC,H coupling constants for the
anomeric centers are given in the experimental section.
Many of the chemical shifts for the individual signals in 1
and 2 correlate relatively well with the values reported for
natural GGM (Table 1).[6a,8a,33] For certain signals, significant
chemical-shift differences have been reported. One of the
Conformation analysis: Studies on the three-dimensional
conformation of polysaccharides are important for several
reasons. In the case of GGM, the natural shape and function
of the molecule plays an important role in any application.
Accordingly, the conformation analysis of model compound
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

R. Leino et al.
Table 1. Chemical shifts of natural GGM and model substances 1 and 2.[a]
Indeed, the measured small-
and medium-size
J
coupling
J6’’b,5’’ =
Sugar
H1
C1
H2
C2
H3
C3
H4
C4
H5
C5
H6a,b
C6
values
(J6’’a,5’’ =6.7,
ꢀ
Native GGM
b-Manp-(1!
2.1 Hz) for the C5 C6 linkage
indicate the simultaneous exis-
tence of both gg and gt rotam-
ers in equilibrium. Additionally,
for the two b-Manp-(1!4)-b-
Glcp and b-Glcp-(1!4)-a-
Manp linkages, the presence of
NOE interactions between the
anomeric proton of the substi-
tuted pyranose and H4 of its
corresponding aglycon strongly
argued for the presence of the
4.75
101.2
4.53
4.08 (or 3.08)[6a]
71.12
3.364
3.66
3.58
77.63
3.63
3.45
76.1
3.78
77.5
3.95, 3.74
61.9
–
72.50
3.702
74.98
!4)-b-Glcp-(1!
103.6
73.87
79.519
a-Galp-(1!
5.038
99.84
3.83
69.46
3.958
70.31
4.01
70.26
3.92
72.3
3.77
62.29
!6)-b-Manp-(1!
4.79
–
–
–
–
–
–
–
3.70
–
4.03–3.94,
3.84–3.79
Model substrate 1
b-Manp-(1!
4.73
100.6
4.05
71.1
3.65
73.4
3.58
67.3
3.41
77.0
3.93, 3.73
61.5
theoretical
global
minima.
!4)-b-Glcp-(1!
4.52
103.0
3.35
73.5
3.69
74.6
3.68
79.2
3.61
75.3
3.90, 3.73
60.8
Based on the coupling con-
stants obtained from the spec-
tral simulations, all of the resi-
dues adopt the expected 4C1
conformation. The NOE corre-
lations between noncontiguous
residues were not observed,
thus suggesting the backbone of
GGM to be linear. The a-(1!
6)-linked galactopyranosyl resi-
due does not influence the line-
arity of the structure and can,
therefore, be considered to be a
!4)-a-Manp
4.76
101.2
3.99
70.1
3.86
69.8
3.86
77.2
3.73
71.8
3.95, 3.84
60.9
Model substrate 2
!6)-b-Manp-(1!
4.75
100.8
4.07
71.1
3.65
73.5
3.68
67.2
3.61
75.1
3.96, 3.78
66.8
!4)-b-Glcp-(1!
!4)-a-Manp
a-Galp-(1!
4.53
103.0
3.35
73.5
3.69
74.7
3.67
79.6
3.62
75.1
3.89, 3.73
60.8
4.76
101.2
3.98
70.1
3.86
69.9
3.85
77.3
3.73
71.8
3.95, 3.84
60.9
5.00
98.9
3.81
69.1
3.95
71.6
3.99
69.9
3.95
70.0
3.74, 3.74
61.8
substituent.
Because
this
[a] Chemical shifts are expressed in d (ppm) with CD3OD/D2O as internal standards. The chemical shifts for
moiety is spatially readily acces-
sible and could thus serve as
contact point for molecular rec-
ognition with lectins,[37] we
GGM are reported values.[6a,8a,33]
2 was performed by using the MacroModel program as inte-
grated in the MAESTRO[34] graphical interface and the
MM3* force field.[35] By applying a combined molecular me-
chanics/NMR spectroscopic approach, the conformation
around every glycosidic linkage was elucidated.[36] The anal-
ysis of the potential-energy surfaces calculated for every F/
Y glycosidic torsion angle suggested the existence of one
well-defined minimum-energy conformer for the two b-(1!
4)-linkages (defined as H1-C1-O5-C4 and C1-O5-C4-H4),
thus corresponding to the typical syn F/Y geometry, in
agreement with the exo-anomeric effect. Two different po-
tential-energy maps were calculated for the a-(1!6)-link-
tested this assumption by using a model protein with specif-
icity to galactosides.[38]
Molecular recognition of tetrasaccharide 2 by a model
lectin: The applications of GGM have in recent years ex-
panded from the traditional roles in pulping and papermak-
ing to new areas including immunology. To delineate the po-
tential for interactions and the binding mode of GGM at
the molecular level, NMR spectroscopic experiments be-
tween tetrasaccharide 2 and a model lectin, the toxic Viscum
album agglutinin (VAA), were performed.[38] This lectin be-
longs to the AB-type family of toxin plant lectins and tar-
gets terminal galactose residues because mannose/glucose
have an IC50 value that is higher than galactose by more
than 100-fold.[39] With the complete NMR spectroscopic
characterization and conformation analysis of 2 done, the
prerequisites for such a study were available. First, evidence
of any interaction was deduced from the analysis of the
STD patterns[35,40] for a ligand/lectin molar ratio of 25:1,
which also permitted us to deduce the binding epitope of
the interacting ligand. Expectedly, VAA recognizes mainly
the galactopyranosyl unit in the tetrasaccharide, which
showed the highest saturation (Figure 7). In addition, signifi-
ꢀ
age, by considering either the gg or gt rotamers at the C5
C6 hydroxymethyl torsion angle. In this case, the corre-
sponding F angle was in the exo-anomeric region, whereas
the Y angle (defined as C1-O1-C6-C5) was predicted to be
in the anti region. Therefore, two model tetrasaccharides
were constructed with either gg or gt geometries at the O6-
substituted Man residue and submitted to molecular-dynam-
ics simulations with the same force field. No interconver-
sions to other energy regions were detected and the ob-
served NMR spectroscopic data (NOE interactions and J
couplings) were consistent with the found geometries.
&
8
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
Figure 8. A) Structural model of the complex of VAA (PDB code:
1PUM, tyrosine (Tyr) site) with the global-minimum conformation of 2
(gg rotamer around the a-Galp-(1!6)-b-Manp linkage) based on dock-
ing analysis and the experimental input from STD NMR spectroscopic
analysis. Only the galactose unit and the contiguous Man residue make
contact with the lectin. B) Structural model of the complex of VAA
(PDB code: 1PUM, Tyr site) with the global-minimum conformation of 2
(gt rotamer around the a-Galp-(1!6)-b-Manp linkage). Important steric
clashes between the ligand and lectin take place, thus precluding an inter-
action.
Figure 7. Bottom: saturation transfer difference (STD)-NMR spectrum of
the viscumin/tetrasaccharide 2 system. Top: the corresponding off-reso-
nance spectrum. The concentration of the lectin was 59 mm (pH 7.5, D2O,
PBS, 300 K) and the ligand/lectin=25:1 (mol/mol). PBS=phosphated-
buffered saline.
cantly smaller STD values were measured for the mannopyr-
anosyl residue closest to the galactose moiety. The STD sig-
nals observed were similar when the saturation was set in
the aromatic and aliphatic regions. These experimental ob-
servations suggest that the binding mode of 2 mainly in-
volves the terminal galactose unit. To translate this experi-
mental information into a 3D structural model, a docking
protocol was applied.[41]
and a less-abundant tetrasaccharide fragment has been de-
vised. During the development of the synthetic route, sever-
al methodologies for the selective ring-opening reaction of
benzylidene acetals[11,12,17–20] and a-galactosylations[23,24] were
evaluated. On the basis of our findings, several of the previ-
ously reported procedures for various reactions proceeded
with less than satisfactory efficiency in the current synthesis.
Nevertheless, high-yielding procedures were found, such as
the partial cleavage of benzylidene acetal moieties by using
the methodology reported by Shie et al.[20] and the a-galac-
tosylation protocol reported by Ando and co-workers.[22b,c,25]
In addition to the synthesis of the GGM backbone struc-
tures, a few model substrates were prepared to study the
acetyl-group migration patterns reported for GGM. The
acetyl-group migration from O2 to O6 was not observed in
these compounds, thus suggesting that similar migrations do
not take place in mannopyranosides in the 4C1 conforma-
tion.
With the backbone structures synthesized, a short compar-
ison between the NMR chemical shifts of the model com-
pounds and the GGM polymer was performed. Although
many of the chemical shifts were in good agreement with
each other, several uncertainties were also found by the use
of the small molecular models. The data presented herein
can prove valuable in the future as a reference when analyz-
ing the NMR spectra of the GGM polysaccharide.
New applications for the GGM polymer are being devel-
oped, including those from biomedical perspectives. Conse-
quently, a brief conformation analysis of the model com-
pounds and a binding study between a lectin (viscumin) and
tetrasaccharide 2 was conducted. Binding was confirmed by
STD and docking experiments. Furthermore, the existence
of a conformational-selection process, with exclusive binding
of only one of the two major conformers present in solution,
namely, the gg conformer, was deduced. Obviously, GGM
can have the potential to protect cells from the toxicity of
this potential biohazard.
A model of the molecular-recognition mode was then ob-
tained through a simple docking protocol.[39] The deposited
structure of VAA in the Protein Data Bank (PDB) was em-
ployed as a template for the protein, whereas both possible
conformers of the ligand, with either the gg or gt geometries
around the a-(1!6)-glycosidic linkage, were used for the
ligand. Because the STD experiments had unequivocally
demonstrated that the major epitope of 2 involved the gal-
actose residue, manual docking of the terminal galactose on
the preferred Tyr249 site[42] was performed. The nonreducing
galactose residue of the bound lactose in the crystallograph-
ic structure (PDB code: 1PUM) was employed as a template
for the manual docking. Interestingly, the gg rotamer led to
a complex (Figure 8A). Besides major contacts between the
galactose unit and lectin, proximity to VAA was also found
for the b-Manp residue. In contrast, the reducing end in the
gt rotamer will make major steric clashes with VAA (Fig-
ure 8B). Therefore, the modeling results suggest a confor-
mational-selection process, with exclusive binding of only
one of the two conformers present in solution (the gg rotam-
er around the (a-Galp-(1!6)-b-Manp) linkage on binding
of 2 to VAA).
Conclusion
Fundamental studies related to the nature of the GGM pol-
ysaccharide found in P. abies have been described. In detail,
a synthetic route to one of the trisaccharide repeating units
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

R. Leino et al.
tion, and the resulting mixture was stirred for 1.5 h. The reaction was
quenched by the addition of saturated aqueous NaHCO3 solution,
brought to RT, diluted with dichloromethane (30 mL), and washed with
saturated aqueous NaHCO3 solution (20 mL). The organic phase was
separated, dried over Na2SO4, filtered, and concentrated. The crude
product was purified by column chromatography (hexane/EtOAc 3:2),
thus providing 14 as a white foam (73 mg, 77%). Rf =0.39 (hexane/
EtOAc 3:2); [a]D =ꢀ11.0 (c=0.1 in CHCl3); 1H NMR (600.13 MHz,
CDCl3): d=7.47–7.13 (m, 35H; arom. H), 5.54 (s, 1H; CH’’Ph), 5.06 (dd,
Experimental Section
Reaction solvents were dried and distilled prior to use when necessary.
All the reactions containing moisture- or air-sensitive reagents were car-
ried out under argon. The NMR spectra were recorded on a Bruker
Avance NMR spectrometer operating at 600.13 MHz (1H: 600.13 MHz,
13C: 150.90 MHz). The probe temperature during the experiments was
kept at 258C, unless indicated otherwise. All the products were fully
characterized by using the following 1D techniques 1H, 13C, and TOCSY
in combination with the following 2D techniques DQF-COSY, NOESY,
HSQC, and HMBC by using pulse sequences provided by the manufac-
turer. Chemical shifts are expressed on the d scale (in ppm) using tetra-
methylsilane (TMS) or residual chloroform or methanol as internal
standards. Coupling constants are given in Hz and provided only once,
when first encountered. Coupling patterns are given as singlet (s), dou-
blet (d), triplet (t), and so forth. The computational analysis of the
1H NMR spectroscopic analysis of all the compounds was achieved by
using the PERCH NMR software with starting values and spectral pa-
rameters obtained from the various NMR spectroscopic techniques used.
HRMS was recorded using Bruker Micro Q-TOF with ESI operated in
the positive mode. Optical rotations were measured at 238C, unless oth-
erwise stated, with a PerkinElmer polarimeter equipped with a Na lamp
(l=589 nm). TLC analysis was performed on aluminum sheets precoated
with silica gel 60 F254 (Merck). Flash chromatography was carried out
on silica gel 60 (0.040–0.060 mm; Merck). Spots were visualized by UV
radiation followed by charring with H2SO4/MeOH (1:4) and heating.
J
3’,4’ =9.3, J3’,2’ =9.7 Hz, 1H; H3’), 4.86 (dd, J2’,1’ =8.1 Hz, 1H; H2’), 4.77
and 4.63 (each d, J=ꢀ12.1 Hz, each 1H; 2’’-CH2Ph), 4.77 and 4.51 (each
d, J=ꢀ11.9 Hz, each 1H; 6-CH2Ph), 4.76 and 4.57 (each d, J=ꢀ12.5 Hz,
each 1H; 3-CH2Ph), 4.73 (d, J1,2 =1.9 Hz, 1H; H1), 4.72 and 4.67 (each d,
J=ꢀ12.4 Hz, each 1H; 2-CH2Ph), 4.65 and 4.51 (each d, J=ꢀ12.4 Hz,
each 1H; 3’’-CH2Ph), 4.65 (d, 1H; H1’), 4.41 and 4.20 (each d, J=
ꢀ12.0 Hz, each 1H; 6’-CH2Ph), 4.33 (s, 1H; H1’’), 4.25 (dd, J6’’a,5’’ =4.9,
J
6’’a,6’’b =ꢀ10.3 Hz, 1H; H6’’a), 4.24 (dd, J4,3 =9.0, J4,5 =9.7 Hz, 1H; H4),
4.05 (dd, J4’’,5’’ =9.3, J4’’,3’’ =9.8 Hz, 1H; H4’’), 3.87 (dd, J4’,5’ =9.8 Hz, 1H;
H4’), 3.83 (dd, J3,2 =3.2 Hz, 1H; H3), 3.79 (dd, J6a,5 =4.6, J6a,6b =ꢀ11.0 Hz,
1H; H6a), 3.77 (dd, J6’’b,5’’ =10.0 Hz, 1H; H6’’b), 3.74 (dd, 1H; H2), 3.69
(dd, J6b,5 =1.9 Hz, 1H; H6b), 3.66 (ddd, 1H; H5), 3.59 (d, J2’’,3’’ =3.1 Hz,
1H; H2’’), 3.40 (dd, J6’a,5’ =2.1, J6’a,6’b =ꢀ11.3 Hz, 1H; H6’a), 3.33 (dd,
1H; H3’’), 3.31 (s, 3H, OCH3), 3.28 (dd, J6’b,5’ =3.2 Hz, 1H; H6’b), 3.11
(ddd, 1H; H5’), 3.09 (dd, 1H; H5’’), 1.99 (s, 3H; 3’-OCOCH3), 1.93 ppm
(s, 3H; 2’-OCOCH3); 13C NMR (150.9 MHz, CDCl3): d=170.0 (3’-
OCOCH3), 169.7 (2’-OCOCH3), 139.3–126.0 (arom. C), 102.1 (C1’’),
101.3 (C’’HPh), 100.6 (C1’), 99.2 (C1), 78.5 (C4’’), 78.4 (C3), 77.9 (C3’’),
76.3 (C2’’), 75.8 (C4’), 75.3 (C4), 75.2 (C2), 74.6 (2’’-CH2Ph), 74.4 (C5’),
73.6 (6-CH2Ph), 73.4 (6’-CH2Ph), 73.4 (C3’), 73.0 (2-CH2Ph), 72.5 (C2’),
72.3 (3’’-CH2Ph, 3-CH2Ph), 71.4 (C5), 68.6 (C6’’, C6), 68.0 (C6’), 67.4
(C5’’), 54.8 (OCH3), 21.0 (3’-OCOCH3), 20.7 ppm (2’-OCOCH3); HRMS:
m/z calcd for C72H78O18Na: 1253.5086 [M+Na]+; found: 1253.5080.
Thiophenyl 2,3-di-O-benzyl-4,6-O-benzylidene-b-d-mannopyranosyl-(1!
4)-2,3-di-O-acetyl-6-O-benzyl-b-d-glucopyranoside (13): BSP (54 mg,
0.26 mmol, 1.2 equiv), TTBP (80 mg, 0.32 mmol, 1.5 equiv), and Tf2O
(45 mL, 0.28 mmol, 1.3 equiv) were added to a solution of donor
3
(116 mg, 0.24 mmol) in dry dichloromethane (2.5 mL) at ꢀ608C (ace-
tone+dry ice). The resulting mixture was stirred for 0.5 h followed by
cooling to ꢀ788C and addition of acceptor 12 (110 mg, 0.25 mmol,
Methyl 2,3-di-O-benzyl-4,6-O-benzylidene-b-d-mannopyranosyl-(1!4)-6-
O-benzyl-b-d-glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-a-d-mannopyra-
noside: NaOMe (2 mg, 0.03 mmol,1 equiv) was added to a solution of 14
(33 mg, 0.027 mmol) in MeOH/THF (1:1, 3 mL), and the resulting mix-
ture was stirred at RT for 19 h. The reaction mixture was neutralized
with DOWEX 50 H+-form, filtered, and concentrated. The crude product
was purified by column chromatography (hexane/EtOAc 1:1) to give a
white foam (26 mg, 84%). Rf =0.40 (hexane/EtOAc 1:1); [a]D = +6.8
(c=1.0 in CHCl3); 1H NMR (600.13 MHz, CDCl3): d=7.48–7.12 (m,
35H; arom. H), 5.58 (s, 1H; CH’’Ph), 4.76 and 4.56 (each d, J=
ꢀ12.2 Hz, each 1H; 3’’-CH2Ph), 4.75 and 4.70 (each d, J=ꢀ12.2 Hz, each
1H; 2’’-CH2Ph), 4.75 and 4.61 (each d, J=ꢀ12.3 Hz, each 1H; 3-CH2Ph),
1.15 equiv) dissolved in dichloromethane (1.5 mL) over
a period of
15 min. The reaction mixture was stirred for 2 h and quenched by the ad-
dition of triethylphosphite (170 mL). The reaction mixture was stirred for
1 h at ꢀ788C, brought to RT, diluted with dichloromethane (30 mL), and
washed with saturated aqueous NaHCO3. The water phase was separated
and extracted with dichloromethane (3ꢆ20 mL). The combined organic
phases were washed with brine (30 mL), dried over Na2SO4, filtered, and
concentrated. The crude product was purified by column chromatography
(hexane/EtOAc 2:1), thus providing 13 as a white foam (100 mg, 53%).
Rf =0.54 (hexane/EtOAc 3:2); [a]D =ꢀ46.2 (c=0.2 in CHCl3); 1H NMR
(600.13 MHz, CDCl3): d=7.51–7.24 (m, 25H; arom. H), 5.55 (s, 1H;
CH’Ph), 5.23 (dd, J3,2 =9.3, J3,4 =9.4 Hz, 1H; H3), 4.92 (dd, J2,1 =10.1 Hz,
1H; H2), 4.78 and 4.67 (each d, J=ꢀ12.1 Hz, each 1H; 2’-CH2Ph), 4.71
and 4.57 (each d, J=ꢀ12.4 Hz, each 1H; 3’-CH2Ph), 4.68 (d, 1H; H1),
4.73 and 4.63 (each d, J=ꢀ12.3 Hz, each 1H; 6-CH2Ph), 4.72 (d, J1,2
=
1.9 Hz, 1H; H1), 4.71 and 4.65 (each d, J=ꢀ12.4 Hz, each 1H; 2-
CH2Ph), 4.59 (d, J1’2’ =7.9 Hz, 1H; H1’), 4.44 and 4.09 (each d, J=
ꢀ12.1 Hz, each 1H; 6’-CH2Ph), 4.36 (dd, J4,3 =9.2, J4,5 =9.6 Hz, 1H; H4),
4.28 (dd, J6’’a,5’’ =5.0, J6’’a,6’’b =ꢀ10.4 Hz, 1H; H6’’a), 4.26 (s, 1H; H1’’), 4.15
4.57 and 4.38 (each d, J=ꢀ11.9 Hz, each 1H; 6-CH2Ph), 4.38 (d, J1’,2’
=
1.0 Hz, 1H; H1’), 4.26 (dd, J6’a,5’ =4.8, J6’a,6’b =ꢀ10.2 Hz, 1H; H6’a), 4.07
(dd, J4’,5’ =9.3, J4’,3’ =9.8 Hz, 1H; H4’), 3.89 (dd, J4,5 =9.9 Hz, 1H; H4),
3.78 (dd, J6’b,5’ =10.0 Hz, 1H; H6’b), 3.68 (dd, J6a,5 =1.9, J6a,6b =ꢀ11.2 Hz,
1H; H6a), 3.65 (dd, J2’,3’ =3.1 Hz, 1H; H2’), 3.52 (dd, J6b,5 =3.5 Hz, 1H;
H6b), 3.46 (ddd, 1H; H5), 3.40 (dd, 1H; H3’), 3.13 (ddd, 1H; H5’), 2.08
(s, 3H; 2-OCOCH3), 2.08 ppm (s, 3H; 3-OCOCH3); 13C NMR
(150.9 MHz, CDCl3): d=170.0 (3-OCOCH3), 169.5 (2-OCOCH3), 138.6–
126.0 (arom. C), 101.9 (1JC1’,H1’ =157.4 Hz; C1’), 101.4 (C’HPh), 85.6 (1JC-
1,H-1 =157.1 Hz; C1), 78.9 (C5), 78.6 (C4’), 78.0 (C3’), 76.4 (C2’), 75.3
(C4), 74.6 (2’-CH2Ph), 74.2 (C3), 73.6 (6-CH2Ph), 72.4 (3’-CH2Ph), 70.2
(C2), 68.6 (C6’), 68.3 (C6), 67.4 (C5’), 20.9 (3-OCOCH3), 20.8 ppm (2-
OCOCH3); HRMS: m/z calcd for C50H52O12SNa: 899.3077 [M+Na]+;
found: 899.3080.
(dd,
J
4’’,5’’ =9.0,
J
4’’,3’’ =9.9 Hz, 1H; H4’’), 3.96 (dd,
J
6a,5 =4.2, J6a,6b
=
=
ꢀ11.5 Hz, 1H; H6a), 3.90 (dd, J3,2 =3.1 Hz, 1H; H3), 3.88 (dd, J6’’b,5’’
10.1 Hz, 1H; H6’’b), 3.78 (dd, J6b,5 =2.5 Hz, 1H; H6b), 3.77 (ddd, 1H;
H5), 3.73 (dd, 1H; H2), 3.54 (dd, J4’,3’ =8.7, J4’,5’ =9.6 Hz, 1H; H4’), 3.52
(dd, J3’,2’ =9.3 Hz, 1H; H3’), 3.48 (d, J2’’,3’’ =3.0 Hz, 1H; H2’’), 3.39 (dd,
1H; H3’’), 3.38 (dd, 1H; H2’), 3.29 (s, 3H, OCH3), 3.29 (ddd, 1H; H5’’),
3.19 (dd, J6’a,5’ =2.0, J6’a,6’b =ꢀ10.8 Hz, 1H; H6’a), 3.11 (dd, J6’b,5’ =3.4 Hz,
1H; H6’b), 3.11 ppm (dd, 1H; H5’); 13C NMR (150.9 MHz, CDCl3): d=
139.0–126.0 (arom. C), 103.4 (C1’), 102.7 (C1’’), 101.4 (C’’HPh), 99.1
(C1), 81.3 (C4’), 79.1 (C3), 78.4 (C4’’), 78.3 (C3’’), 75.7 (C2’’), 75.4 (C4),
74.9 (C2), 74.8 (2’’-OCH2Ph), 74.4 (C3’), 74.1 (C2’), 73.8 (C5’), 73.4 (6-
OCH2Ph), 73.3 (6’-OCH2Ph), 73.0 (3’’-OCH2Ph), 72.8ACTHNUGTRNE(UNG 2-OCH2Ph), 71.9
(3-OCH2Ph), 70.9 (C5), 69.2 (C6), 68.0 (C6’’), 67.9 (C6’), 67.4 (C5’’),
54.8 ppm (OCH3); HRMS: m/z calcd for C68H74O16Na: 1169.4875 [M+
Na]+; found: 1169.4862.
Methyl 2,3-di-O-benzyl-4,6-O-benzylidene-b-d-mannopyranosyl-(1!4)-
2,3-di-O-acetyl-6-O-benzyl-b-d-glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-
a-d-mannopyranoside (14): Preactivated molecular sieves (4 ꢀ) were
added to a solution of donor 13 (81 mg, 0.09 mmol, 1.2 equiv) and accept-
or 4 (36 mg, 0.08 mmol) in dry dichloromethane (2.5 mL), and the reac-
tion mixture was cooled to ꢀ408C. NIS (21 mg, 0.09 mmol, 1.2 equiv)
and TMSOTf (1.1 mL, 0.006 mmol, 0.06 equiv) were added to the solu-
Methyl b-d-mannopyranosyl-(1!4)-b-d-glucopyranosyl-(1!4)-a-d-man-
nopyranoside (1): Pd/C (10% Pd, 59 mg, 2.5 weight equiv) was added to
a solution of methyl 2,3-di-O-benzyl-4,6-O-benzylidene-b-d-mannopyra-
nosyl-(1!4)-6-O-benzyl-b-d-glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-a-
d-mannopyranoside (26 mg, 0.023 mmol) in dry MeOH (3 mL). The reac-
&
10
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
tion mixture was placed inside a reactor and the H2 pressure was set to
2.8 bar (40 psi). After 19 h, the reaction mixture was diluted with MeOH
(3 mL), filtered through celite, and concentrated to give 1 as a white
solid (11 mg, quant.). [a]D = +11.1 (c=1.0 in MeOH); 1H NMR
(600.13 MHz, D2O, 35 8C): d=4.76 (d, J1,2 =1.8 Hz, 1H; H1), 4.74 (d,
chloromethane (30 mL), and washed with saturated aqueous NaHCO3
solution (20 mL). The organic phase was separated, dried over Na2SO4,
filtered, and concentrated. The crude product was purified by column
chromatography (hexane/EtOAc 3:2), thus providing the product as a
colorless oil (20 mg, 69%). Rf =0.52 (hexane/EtOAc 3:2); [a]D = +24.9
(c=1.0 in CHCl3); 1H NMR (600.13 MHz, CDCl3): d=8.00–7.10 (m,
45H; arom. H), 5.73 (dd, J2’’’,1’’’ =3.7, J2’’’,3’’’ =10.6 Hz, 1H; H2’’’), 5.54 (dd,
J
1’’,2’’ =1.0 Hz, 1H; H1’’), 4.52 (d, J1’,2’ =8.0 Hz, 1H; H1’), 4.05 (dd, J2’’,3’’
3.3 Hz, 1H; H2’’), 3.99 (dd, J2,3 =3.3 Hz, 1H; H2), 3.95 (dd, J6a,5 =2.3,
6a,6b =ꢀ12.3 Hz, 1H; H6a), 3.93 (dd, J6a,5’’ =2.3, J6’’a,6’’b =ꢀ12.3 Hz, 1H;
H6’’a)), 3.90 (dd, J6’a,5’ =2.2, J6’a,6’b =ꢀ12.4 Hz, 1H; H6’a), 3.86 (dd, J3,4
=
J
J
3’’’,4’’’ =3.3 Hz, 1H; H3’’’), 5.28 (d, 1H; H1’’’), 5.06 (dd, J3’,4’ =9.3, J3’,2’ =
=
9.7 Hz, 1H; H3’), 4.96 and 4.66 (each d, J=ꢀ11.9 Hz, each 1H; 2’’-
CH2Ph), 4.85 (dd, J2’,1’ =8.0 Hz, 1H; H2’), 4.77 and 4.56 (each d, J=
ꢀ12.4 Hz, each 1H; 3-CH2Ph), 4.74 and 4.48 (each d, J=ꢀ11.9 Hz, each
1H; 6-CH2Ph), 4.72 (d, J1,2 =2.0 Hz, 1H; H1), 4.71 and 4.65 (each d, J=
ꢀ12.1 Hz, each 1H; 2-CH2Ph), 4.67 and 4.50 (each d, J=ꢀ11.8 Hz, each
1H; 4’’-CH2Ph), 4.67 (dd, J4’’’,5’’’ =1.0 Hz, 1H; H4’’’), 4.62 (d, 1H; H1’),
4.37 and 4.25 (each d, J=ꢀ12.1 Hz, each 1H; 6’-CH2Ph), 4.35 and 4.30
(each d, J=ꢀ11.8 Hz, each 1H; 3’’-CH2Ph), 4.25 (d, J1’’,2’’ =0.5 Hz, 1H;
H1’’), 4.22 (dd, J4,3 =8.9, J4,5 =9.6 Hz, 1H; H4), 4.09 (dd, J6’’’a,5’’’ =1.6,
9.2 Hz, 1H; H3), 3.85 (dd, J4,5 =9.7 Hz, 1H; H4), 3.85 (dd, J6b,5 =5.5 Hz,
1H; H6b), 3.73 (dd, J6’b,5’ =5.2 Hz, 1H; H6’b), 3.73 (dd, J6’’b,5’’ =6.4 Hz,
1H; H6’’b), 3.73 (ddd, 1H; H5), 3.69 (dd, J3’,4’ =8.8, J3’,2’ =9.7 Hz, 1H;
H3’), 3.68 (dd, J4’,5’ =9.7 Hz, 1H; H4’), 3.64 (dd, J3’’,4’’ =9.7 Hz, 1H; H3’’),
3.61 (ddd, 1H; H5’), 3.58 (dd, J4’’,5’’ =9.8 Hz, 1H; H4’’), 3.41 (ddd, 1H;
H5’’), 3.40 (s, 3H, OCH3), 3.35 ppm (dd, 1H; H2’); 13C NMR
(150.9 MHz, D2O, 35 8C): d=103.0 (1JC1’,H1’ =163.5 Hz; C1’), 101.2
(1JC1,H1 =173.4 Hz; C1), 100.6 (1JC1’’,H1’’ =160.8 Hz; C1’’), 79.2 (C4’), 77.2
(C4), 77.0 (C5’’), 75.3 (C5’), 74.6 (C3’), 73.5 (C2’), 73.4 (C3’’), 71.8 (C5),
71.1 (C2’’), 70.1 (C2), 69.8 (C3), 67.3 (C4’’), 61.5 (C6’’), 60.9 (C6), 60.8
(C6’), 55.4 ppm (OCH3); HRMS: m/z calcd for C19H34O16Na: 541.1745
[M+Na]+; found: 541.1744.
J
6’’’a,6’’’b =ꢀ12.9 Hz, 1H; H6’’’a), 3.89 (dd, J6a,5’’ =1.9, J6’’a,6’’b =ꢀ11.7 Hz,
1H; H6’’a), 3.86 (dd, J4’’,3’’ =9.1, J4’’,5’’ =9.5 Hz, 1H; H4’’), 3.85 (dd, J6’’’b,5’’’
1.4 Hz, 1H; H6’’’b), 3.83 (ddd, 1H; H5’’’), 3.82 (dd, J3,2 =3.1 Hz, 1H;
H3), 3.81 (dd, J6’’b,5’’ =4.3 Hz, 1H; H6’’b), 3.78 (dd, J6a,5 =4.6, J6a,6b
ꢀ10.9 Hz, 1H; H6a), 3.74 (dd, J4’,5’ =9.7 Hz, 1H; H4’), 3.72 (dd, 1H; H2),
3.69 (dd, J6b,5 =1.9 Hz, 1H; H6b), 3.66 (ddd, 1H; H5), 3.57 (dd, J2’’,3’’
=
=
Methyl 2,3,4-tri-O-benzyl-b-d-mannopyranosyl-(1!4)-2,3-di-O-acetyl-6-
O-benzyl-b-d-glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-a-d-mannopyra-
noside (16): Preactivated molecular sieves (4 ꢀ) and borane/THF com-
plex in THF (0.28 mL, 0.28 mmol, 5 equiv) were added to a solution of
14 (70 mg, 0.06 mmol) in dry dichloromethane (3 mL). The resulting mix-
=
3.0 Hz, 1H; H2’’), 3.42 (dd, J6’a,5’ =2.0, J6’a,6’b =ꢀ11.7 Hz, 1H; H6’a), 3.31
(s, 3H, OCH3), 3.30 (dd, J6’b,5’ =4.2 Hz, 1H; H6’b), 3.23 (dd, 1H; H3’’),
3.18 (ddd, 1H; H5’’), 3.14 (ddd, 1H; H5’), 2.10 (s, 3H; 3’-OCOCH3), 1.93
(s, 3H; 2’-OCOCH3), 1.10 and 0.92 ppm (each s, each 9H; CACHTUNGTRENNUNG(CH3)3);
ture was stirred for 10 min, and CuACHTUNRGTNEUNG(OTf)2 (1.4 mg, 0.004 mmol,
13C NMR (150.9 MHz, CDCl3): d=170.4 (3’-OCOCH3), 169.5 (2’-
OCOCH3), 166.1 (2’’’-OCOPh), 165.8 (3’’’-OCOPh), 139.3–126.9 (arom.
C), 101.6 (C1’’), 100.7 (C1’), 99.2 (C1), 98.0 (C1’’’), 82.2 (C3’’), 78.4 (C3),
75.8 (C4’), 75.4 (C2), 75.3 (C4, C5’’), 74.6 (2’’-OCH2Ph, C5’), 74.2 (C4’’),
73.9 (C2’’), 73.6 (4’’-OCH2Ph), 73.5 (6-OCH2Ph), 73.4 (6’-OCH2Ph), 73.1
(C3’’), 72.8 (2-OCH2Ph), 72.4 (C2’, 3-OCH2Ph), 71.4 (C5), 71.1 (C4’’’),
71.0 (3’’-OCH2Ph, C3’’’), 68.9 (C6’’), 68.8 (C2’’’), 68.6 (C6), 68.3 (C6’),
0.07 equiv) was added with further stirring for 3 h. The reaction mixture
was cooled to 08C, and the reaction was quenched by the addition of
Et3N and MeOH. The resulting mixture was concentrated to give the
crude product, which was purified by column chromatography (hexane/
EtOAc 3:2!1:1) to give the product as a colorless oil (60 mg, 85%).
Rf =0.52 (hexane/EtOAc 1:1); [a]D =ꢀ11.5 (c=0.1 in CHCl3); 1H NMR
(600.13 MHz, CDCl3): d=7.37–7.15 (m, 35H; arom. H), 5.08 (dd, J3’,4’
=
67.0 (C5’’’), 66.8 (C6’’’), 54.8 (OCH3), 27.5 and 27.3 (CACHTNUGTRNEUNG(CH3)3), 23.1 and
9.3, J3’,2’ =9.7 Hz, 1H; H3’), 4.88 (dd, J2’,1’ =8.0 Hz, 1H; H2’), 4.88 and
4.57 (each d, J=ꢀ11.6 Hz, each 1H; 3-CH2Ph), 4.77 and 4.51 (each d,
J=ꢀ11.9 Hz, each 1H; 6-CH2Ph), 4.76 and 4.62 (each d, J=ꢀ12.5 Hz,
each 1H; 2’’-CH2Ph), 4.76 and 4.56 (each d, J=ꢀ11.1 Hz, each 1H; 4’’-
CH2Ph), 4.72 (d, J1,2 =1.9 Hz, 1H; H1), 4.70 and 4.66 (each d, J=
ꢀ12.4 Hz, each 1H; 2-CH2Ph), 4.64 (d, 1H; H1’), 4.42 and 4.38 (each d,
J=ꢀ11.8 Hz, each 1H; 3’’-CH2Ph), 4.40 and 4.29 (each d, J=ꢀ12.1 Hz,
each 1H; 6’-CH2Ph), 4.38 (s, 1H; H1’’), 4.24 (dd, J4,5 =9.0, J4,3 =10.0 Hz,
20.8 (CACHTUNRTGENUNG(CH3)3), 20.8 (3’-OCOCH3), 20.7 ppm (2’-OCOCH3); HRMS: m/z
calcd for C100H114O25SiNa: 1765.7316 [M+Na]+; found: 1765.7266.
Methyl 2,3-di-O-benzoyl-a-d-galactopyranosyl-(1!6)-2,3,4-tri-O-benzyl-
b-d-mannopyranosyl-(1!4)-2,3-di-O-acetyl-6-O-benzyl-b-d-glucopyrano-
syl-(1!4)-2,3,6-tri-O-benzyl-a-d-mannopyranoside (20): The HF/pyridine
complex (18 mL) was added to a solution of 19 (50 mg, 0.03 mmol) in dry
THF (3 mL) at 08C, and the resulting mixture was brought to RT and
stirred for 19 h. The reaction mixture was diluted with dichloromethane
(30 mL) and quenched by the addition of saturated aqueous NaHCO3
solution. The organic phase was washed with brine, dried over Na2SO4,
filtered, and concentrated. The crude product was purified by column
chromatography (hexane/EtOAc 1:1) to give the product as a colorless
oil (31 mg, 68%). Rf =0.26 (hexane/EtOAc 1:2); [a]D = +20.8 (c=0.8 in
CHCl3); 1H NMR (600.13 MHz, CDCl3): d=8.02–7.18 (m, 45H; arom.
H), 5.72 (dd, J2’’’,1’’’ =3.8, J2’’’,3’’’ =10.5 Hz, 1H; H2’’’), 5.71 (d, 1H; H1’’’),
5.65 (dd, J3’’’,4’’’ =3.2 Hz, 1H; H3’’’), 5.12 (dd, J3’,4’ =9.5, J3’,2’ =9.8 Hz, 1H;
1H; H4), 3.86 (dd, J4’,5’ =9.8 Hz, 1H; H4’), 3.82 (dd, J6a,5 =4.3, J6a,6b
=
ꢀ11.0 Hz, 1H; H6a), 3.82 (dd, J6’’a,5’’ =2.9, J6’’a,6’’b =ꢀ9.9 Hz, 1H; H6’’a),
3.79 (dd, J3,2 =3.6 Hz, 1H; H3), 3.74 (dd, J4’’,3’’ =9.3, J4’’,5’’ =9.5 Hz, 1H;
H4’’), 3.72 (dd, 1H; H2), 3.68 (dd, J6b,5 =1.8 Hz, 1H; H6b), 3.66 (ddd,
1H; H5), 3.64 (dd, J6’’b,5’’ =5.8 Hz, 1H; H6’’b), 3.61 (d, J2’’,3’’ =2.9 Hz, 1H;
H2’’), 3.51 (dd, J6’a,5’ =2.1, J6’a,6’b =ꢀ11.4 Hz, 1H; H6’a), 3.34 (dd, J6’b,5’
=
3.5 Hz, 1H; H6’b), 3.31 (s, 3H, OCH3), 3.31 (dd, 1H; H3’’), 3.23 (ddd,
1H; H5’), 3.19 (ddd, 1H; H5’’), 2.00 (s, 3H; 3’-OCOCH3), 1.93 ppm (s,
3H; 2’-OCOCH3); 13C NMR (150.9 MHz, CDCl3): d=170.4 (3’-
OCOCH3), 169.7 (2’-OCOCH3), 139.3–127.1 (arom. C), 100.8 (C1’’),
100.7 (C1’), 99.4 (C1), 82.3 (C3’’), 78.5 (C3), 76.0 (C5’’), 75.4 (C4, C2),
75.3 (3-OCH2Ph), 75.2 (C4’), 74.9 (C4’’), 74.8 (C5’), 74.4 (C2’’), 74.0 (2’’-
OCH2Ph), 73.8 (6-OCH2Ph), 73.6 (6’-OCH2Ph), 73.5 (C3’), 72.9 (2-
OCH2Ph), 72.6 (C2’, 4’’-OCH2Ph), 71.6 (C5), 71.5 (3’’-OCH2Ph), 68.7
(C6), 68.5 (C6’), 62.8 (C6’’), 55.0 (OCH3), 21.0 (3’-OCOCH3), 20.9 ppm
(2’-OCOCH3); HRMS: m/z calcd for C72H80O18Na: 1255.5242 [M+Na]+;
found: 1255.5249.
H3’), 5.09 (dd,
J2’,1’ =8.2 Hz, 1H; H2’), 4.81 and 4.51 (each d, J=
ꢀ11.1 Hz, each 1H; 4’’-CH2Ph), 4.75 and 4.52 (each d, J=ꢀ11.9 Hz, each
1H; 6-CH2Ph), 4.73 (d, J1,2 =2.0 Hz, 1H; H1), 4.73 and 4.56 (each d, J=
ꢀ12.5 Hz, each 1H; 3-CH2Ph), 4.71 (d, 1H; H1’), 4.70 and 4.47 (each d,
J=ꢀ12.1 Hz, each 1H; 2’’-CH2Ph), 4.69 and 4.65 (each d, J=ꢀ12.4 Hz,
each 1H; 2-CH2Ph), 4.38 and 4.22 (each d, J=ꢀ12.1 Hz, each 1H; 6’-
CH2Ph), 4.34 and 4.30 (each d, J=ꢀ11.8 Hz, each 1H; 3’’-CH2Ph), 4.29
(dd, J4’’’,5’’ =1.2 Hz, 1H; H4’’’), 4.25 (s, 1H; H1’’), 4.21 (dd, J4,3 =8.8, J4,5
=
9.8 Hz, 1H; H4), 4.19 (dd, J4’,5’ =9.7 Hz, 1H; H4’), 4.10 (ddd, J5’’’,6’’’b =3.1,
Methyl 2,3-di-O-benzoyl-4,6-O-di-tertbutylsilylene-a-d-galactopyranosyl-
(1!6)-2,3,4-tri-O-benzyl-b-d-mannopyranosyl-(1!4)-2,3-di-O-acetyl-6-
O-benzyl-b-d-glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-a-d-mannopyra-
noside (19): Preactivated molecular sieves (4 ꢀ) were added to a mixture
containing acceptor 16 (21 mg, 0.017 mmol) and donor 18 (13 mg,
0.020 mmol, 1.2 equiv) in dry dichloromethane (2.5 mL), and the result-
ing mixture was cooled to 08C. NIS (4.5 mg, 0.020 mmol, 1.2 equiv) and
TMSOTf (0.2 mL, 0.001 mmol, 0.06 equiv) were added after 10 minutes to
the reaction mixture, which was stirred for 1.5 h, brought to RT,
quenched with saturated aqueous NaHCO3 solution, diluted with di-
J
5’’’,6’’’a =6.3 Hz, 1H; H5’’’), 4.05 (dd, J6’’a,5’’ =1.7, J6’’a,6’’b =ꢀ12.5 Hz, 1H;
H6’’a), 3.89 (dd, J6’’’a,6’’’b =ꢀ12.4 Hz, 1H; H6’’’a), 3.83 (dd, J3,2 =3.2 Hz,
1H; H3), 3.80 (dd, J6a,5 =4.6, J6a,6b =ꢀ11.1 Hz, 1H; H6a), 3.71 (dd, 1H;
H2), 3.70 (dd, 1H; H6’’’b), 3.69 (dd, J6b,5 =1.8 Hz, 1H; H6b), 3.66 (ddd,
1H; H5), 3.64 (dd, J4’’,3’’ =9.1, J4’’,5’’ =9.5 Hz, 1H; H4’’), 3.62 (dd, J6’’b,5’’
=
6.1 Hz, 1H; H6’’b), 3.55 (d, J2’’,3’’ =3.0 Hz, 1H; H2’’), 3.38 (dd, J6’a,5’ =2.4,
J
6’a,6’b =ꢀ11.0 Hz, 1H; H6’a), 3.33 (dd, J6’b,5’ =2.9 Hz, 1H; H6’b), 3.32 (s,
3H, OCH3), 3.31 (ddd, 1H; H5’’), 3.23 (dd, 1H; H3’’), 3.19 (ddd, 1H;
H5’), 2.08 (s, 3H; 3’-OCOCH3), 2.00 ppm (s, 3H; 2’-OCOCH3); 13C NMR
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

R. Leino et al.
(150.9 MHz, CDCl3): d=170.9 (3’-OCOCH3), 170.0 (2’-OCOCH3), 166.0
(3’’’-OCOPh), 165.8 (2’’’-OCOPh), 139.3–126.7 (arom. C), 101.0
(1JC1’’,H1’’ =155.8 Hz; C1’’; 1JC1’,H1’ =162.1 Hz; C1’), 99.1 (1JC1,H1 =168.0 Hz;
C1), 97.6 (1JC1’’’,H1’’’ =173.5 Hz; C1’’’), 82.1 (C3’’), 78.4 (C3), 76.5 (C5’’),
75.7 (C4), 74.8 (4’’-OCH2Ph), 74.3 (C4’’, C5’), 73.9 (C4), 73.6 (C2’’, 2’’-
OCH2Ph, 6-OCH2Ph), 73.4 (6’-OCH2Ph), 72.8 (C3’), 72.7 (2-OCH2Ph),
72.3 (C2’), 72.0 (3-OCH2Ph), 71.3 (C5, C3’’’), 71.1 (C5’’’), 71.0 (3’’-
OCH2Ph), 69.8 (C4’’’), 69.3 (C2’’’), 69.0 (C6’’), 68.7 (C6), 67.7 (C6’), 63.2
(C6’’’), 54.8 (OCH3), 20.9 (3’-OCOCH3), 20.8 ppm (2’-OCOCH3);
HRMS: m/z calcd for C92H98O25Na: [M+Na]+ 1625.6295; found:
1625.6266.
3.67 (dd, J3’,2’ =10.0 Hz, 1H; H3’), 3.65 (dd, 1H; H3’’), 3.62 (ddd, 1H;
H5’), 3.61 (ddd, 1H; H5’’), 3.40 (s, 3H, OCH3), 3.34 ppm (dd, 1H; H2’);
13C NMR (150.9 MHz, D2O, 35 8C): d=103.0 (1JC1’,H1’ =163.3 Hz; C1’),
101.2 (1JC1,H1 =172.5 Hz; C1), 100.8 (1JC1’’,H1’’ =159.4 Hz; C1’’), 98.9
(1JC1’’’,H1’’’ =170.6 Hz; C1’’’), 79.6 (C4’), 77.3 (C4), 75.1 (C5’, C5’’), 74.7
(C3’), 73.5 (C2’, C3’’), 71.8 (C5), 71.6 (C3’’’), 71.1 (C2’’), 70.1 (C2), 70.0
(C5’’’), 69.9 (C4’’’, C3), 69.1 (C2’’’), 67.2 (C4’’), 66.8 (C6’’), 61.8 (C6’’’),
60.9 (C6), 60.8 (C6’), 55.4 ppm (OCH3); HRMS: m/z calcd for
C25H44O21Na: 703.2273 [M+Na]+; found: 703.2273.
AHCTUNGTERG(NNUN 2’S,3’S)-Methyl 3,4-O-[2’,3’-dimethoxybutan-2’,3’-diyl]-a-d-mannopyra-
noside: The reaction was performed under previously described experi-
mental conditions[43] starting from commercially available methyl a-d-
mannopyranoside (0.46 g, 2.4 mmol), butan-2,3-dione (0.34 mL,
1.6 equiv), trimethylorthoformate (1.15 mL, 4.2 equiv) and CSA (50 mg,
0.1 equiv) to give, after purification by column chromatography with di-
ethyl ether as the eluent, the product as a white foam (0.59 g, 79%). Rf =
Methyl a-d-galactopyranosyl-(1!6)-2,3,4-tri-O-benzyl-b-d-mannopyra-
nosyl-(1!4)-6-O-benzyl-b-d-glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-a-
d-mannopyranoside: NaOMe (3 mg, 2.4 equiv) was added to a solution
of 20 (33 mg, 0.02 mmol) in MeOH/THF (1:1, 4 mL), and the resulting
mixture was stirred at RT for 19 h. The reaction mixture was neutralized
with DOWEX 50 H+-form, filtered, and concentrated. The crude product
was purified by column chromatography (MeOH/dichloromethane 1:10)
to give a white foam (26 mg, quant.). Rf =0.38 (MeOH/dichloromethane
1:10); 1H NMR (600.13 MHz, CD3OD+CDCl3): d=7.38–7.18 (m, 35H;
arom. H), 4.89 (d, J1’’’,2’’’ =4.4 Hz, 1H; H1’’’), 4.88 and 4.60 (each d, J=
ꢀ11.9 Hz, each 1H; 2-CH2Ph), 4.78 and 4.59 (each d, J=ꢀ11.7 Hz, each
1H; 3-CH2Ph), 4.76 and 4.69 (each d, J=ꢀ12.2 Hz, each 1H; 2’’-CH2Ph),
4.75 (d, J1,2 =2.0 Hz, 1H; H1), 4.73 and 4.60 (each d, J=ꢀ12.0 Hz, each
1H; 6-CH2Ph), 4.69 and 4.66 (each d, J=ꢀ12.2 Hz, each 1H; 4’’-CH2Ph),
4.50 (d, J1’,2’ =7.8 Hz, 1H; H1’), 4.46 (s, 2H; 3’’-CH2Ph), 4.44 and 4.23
(each d, J=ꢀ12.1 Hz, each 1H; 6’-CH2Ph), 4.34 (d, J1’’,2’’ =0.7 Hz, 1H;
0.10 (diethyl ether); 1H NMR (600.13 MHz, CDCl3): d=4.75 (d, J1,2
1.5 Hz, 1H; H1), 4.08 (dd, J4,5 =10.1, J4,3 =10.3 Hz, 1H; H4), 4.01 (dd,
=
J
J
J
3,2 =3.2 Hz, 1H; H3), 3.92 (ddd, J2,2-OH =2.0 Hz, 1H; H2), 3.84 (ddd,
6a,5 =3.0, J6a,6-OH =4.9, J6a,6b =ꢀ11.9 Hz, 1H; H6a), 3.78 (ddd, J6b,5 =4.7,
6,6-OH =7.7 Hz, 1H; H6b), 3.76 (ddd, 1H; H5), 3.37 (s, 3H; 1-OCH3),
3.28 and 3.26 (each s, 2’-OCH3, each 3H; 3’-OCH3), 2.37 (d, 1H; 2-OH),
1.98 (dd, 1H; 6-OH), 1.32 and 1.29 ppm (each s, each 3H, each CH3 (2’,
3’)); HRMS: m/z calcd for C13H24O8Na: 331.1363 [M+Na]+; found:
331.1373.
AHCTUNGTRENGN(UN 2’S,3’S)-Methyl 3,4-O-[2’,3’-dimethoxybutan-2’,3’-diyl]-6-O-tert-butyldi-
methylsilyl-a-d-mannopyranoside: The reaction was performed under
previously described experimental conditions[44] starting from 0.14 g
(0.4 mmol) of (2’S,3’S)-methyl 3,4-O[2’,3’-dimethoxybutan-2’,3’-diyl]-a-d-
mannopyranoside, 60 mg (2.0 equiv) imidazole and 74 mg (1.1 equiv) tert-
butyldimethylsilyl chloride to give, after column purification with hexane/
EtOAc 3:1 as an eluent, the title compound as a colorless syrup (0.17 g,
90%). Rf =0.63 (hexane/EtOAc 1:1); [a]D = +174.4 (c=0.5 in CHCl3);
1H NMR (600.13 MHz, CDCl3): d=4.70 (d, J1,2 =1.5 Hz, 1H; H1), 3.99
(dd, J3,2 =2.9, J3,4 =9.2 Hz, 1H; H3), 3.98 (dd, J4,5 =9.9 Hz, 1H; H4), 3.88
(ddd, J2,2-OH =3.3 Hz, 1H; H2), 3.85 (dd, J6a,5 =1.8, J6a,6b =ꢀ11.3 Hz, 1H;
H6a), 3.80 (dd, J6b,5 =5.4 Hz, 1H; H6b), 3.69 (ddd, 1H; H5), 3.35 (s, 3H;
1-OCH3), 3.27 and 3.24 (each s, each 3H; 2’-OCH3, 3’-OCH3), 2.21 (d,
1H; 2-OH), 1.31 and 1.28 (each s, each 3H; 2’-CH3, 3’-CH3), 0.89 (s, 9H;
H1’’), 4.28 (dd, J4,3 =8.8, J4,5 =9.6 Hz, 1H; H4), 4.01 (dd, J6a,5 =4.9, J6a,6b
=
ꢀ11.4 Hz, 1H; H6a), 3.95 (dd, J4’’’,5’’’ =1.2, J4’’’,3’’’ =3.3 Hz, 1H; H4’’’), 3.91
(dd, J6b,5 =2.1 Hz, 1H; H6b), 3.89 (dd, J3,2 =3.3 Hz, 1H; H3), 3.86 (dd,
J
J
3’’’,2’’’ =9.9 Hz, 1H; H3’’’), 3.81 (ddd, 1H; H5), 3.80 (dd, J6’’a,5’’ =2.7,
6’’a,6’’b =ꢀ11.6 Hz, 1H; H6’’a), 3.80 (ddd, J5’’’,6’’’b =5.5, J5’’’,6’’’a =6.3 Hz, 1H;
H5’’’), 3.77 (dd, 1H; H2), 3.75 (dd, J6’’’a,6’’’b =ꢀ11.1 Hz, 1H; H6’’’a), 3.74
(dd, 1H; H6’’’b), 3.74 (dd, J6b,5’’ =8.2 Hz, 1H; H6’’b), 3.74 (dd, 1H; H2’’’),
3.68 (dd, J4’’,3’’ =9.2, J4’’,5’’ =9.8 Hz, 1H; H4’’), 3.64 (dd, J2’’,3’’ =3.0 Hz, 1H;
H2’’), 3.52 (ddd, 1H; H5’’), 3.51 (dd, J4’,3’ =8.5, J4’,5’ =9.9 Hz, 1H; H4’),
3.48 (dd, J3’,2’ =9.4 Hz, 1H; H3’), 3.41 (dd, 1H; H3’’), 3.37 (dd, J6’a,5’ =2.2,
J
(dd,
6’a,6’b =ꢀ11.1 Hz, 1H; H6’a), 3.35 (s, 3H, OCH3), 3.29 (dd, 1H; H2’), 3.29
J
6’b,5’ =3.7 Hz, 1H; H6’b), 3.18 ppm (ddd, 1H; H5’); 13C NMR
(150.9 MHz, CD3OD+CDCl3): d = 139.5–127.5 (arom. C), 103.6 (C1’),
102.4 (C1’’), 99.6 (C1’’’), 99.6 (C1), 82.7 (C3’’), 82.0 (C4’), 78.8 (C3), 76.0
(C4), 75.7 (C3’, C2), 75.5 (2-OCH2Ph), 75.1 (C4’’), 74.7(2’’-OCH2Ph), 74.3
(C2’’, C5’’), 74.2 (C5’), 74.1 (C2’), 73.9 (6’-OCH2Ph), 73.7 (6-OCH2Ph),
73.2 (4’’-OCH2Ph), 72.6 (3-OCH2Ph), 72.4 (3’’-OCH2Ph), 71.8 (C5), 71.1
(C5’’’), 70.3 (C4’’’, C3’’’), 69.8 (C2’’’), 69.7 (C6), 68.7 (C6’), 67.2 (C6’’),
62.1 (C6’’’), 55.1 ppm (OCH3); HRMS: m/z calcd for C74H86O21Na:
1333.5559 [M+Na]+; found: 1333.5548.
SiCACHTNUTGRNEUNG
(CH3)3), 0.07 and 0.06 ppm (each s, each 3H, each SiCH3); 13C NMR
(150.9 MHz, CDCl3): d=100.9 (C1), 100.4 and 99.9 (C2’,C3’), 71.6 (C5),
69.9 (C2), 68.5 (C3), 63.0 (C4), 61.8 (C6), 54.7 (1-OCH3), 48.2 and 48.0
(2’-OCH3, 3’-OCH3), 26.0 (SiC
CH3), ꢀ5.0 and ꢀ5.9 ppm (Si
C19H38O8SiNa: 445.2228 [M+Na]+; found: 445.2237.
(2’S,3’S)-Methyl 2-O-acetyl-3,4-O-[2’,3’-dimethoxybutan-2’,3’-diyl]-6-O-
A
ACHTNUGTRENN(UGN CH3)3), 17.9 (2’-CH3, 3’-
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
tert-butyldimethylsilyl-a-d-mannopyranoside: The reaction was per-
formed under previously described experimental conditions[45] starting
from (2’S,3’S)-methyl 3,4-O-[2’,3’-dimethoxybutan-2’,3’-diyl]-6-O-tert-bu-
tyldimethylsilyl-a-d-mannopyranoside (0.13 g, 0.3 mmol), Ac2O (1.2 mL),
and pyridine (1.1 mL) to give, after purification by column chromatogra-
phy with hexane/EtOAc 4:1 as the eluent, the product as a colorless oil
(0.13 g, 92%). Rf =0.54 (hexane/EtOAc 1:1); [a]D = ++125.8 (c=1.2 in
Methyl a-d-galactopyranosyl-(1!6)-b-d-mannopyranosyl-(1!4)-b-d-glu-
copyranosyl-(1!4)-a-d-mannopyranoside (2): Pd/C (10% Pd, 59 mg, 2.5
weight equiv) was added to a solution of methyl a-d-galactopyranosyl-
(1!6)-2,3,4-tri-O-benzyl-b-d-mannopyranosyl-(1!4)-6-O-benzyl-b-d-
glucopyranosyl-(1!4)-2,3,6-tri-O-benzyl-a-d-mannopyranoside (24 mg,
0.018 mmol) in dry MeOH (2.3 mL). The reaction mixture was placed
inside a reactor and the H2 pressure was set to 2.8 bar (40 psi). After
19 h, the reaction mixture was diluted with MeOH (3 mL), filtered
through celite, and concentrated to give 2 as a colorless oil (12 mg,
quant.). [a]D = +31.8 (c=1.0 in MeOH); 1H NMR (600.13 MHz, D2O,
358C): d=5.00 (d, J1’’’,2’’’ =3.9 Hz, 1H; H1’’’), 4.76 (d, J1,2 =1.8 Hz, 1H;
H1), 4.75 (d, J1’’,2’’ =0.9 Hz, 1H; H1’’), 4.52 (d, J1’,2’ =8.0 Hz, 1H; H1’),
4.07 (dd, J2’’,3’’ =3.3 Hz, 1H; H2’’), 3.99 (dd, J4’’’,3’’’ =1.3, J4’’’,5’’’ =3.8 Hz, 1H;
CHCl3); 1H NMR (600.13 MHz, CDCl3): d=4.99 (dd, J2,1 =1.7, J2,3
=
3.2 Hz, 1H; H2), 4.67 (d, 1H; H1), 4.13 (dd, J3,4 =10.3 Hz, 1H; H3), 4.07
(dd, J4,5 =9.8 Hz, 1H; H4), 3.85 (dd, J6b,5 =4.5, J6b,6a =ꢀ11.8 Hz, 1H;
H6b), 3.81 (dd, J6a,5 =1.8 Hz, 1H; H6a), 3.66 (ddd, 1H; H5), 3.33 (s, 3H;
1-OCH3), 3.27 and 3.25 (each s, each 3H; 2’-OCH3, 3’-OCH3), 2.11 (s,
3H; 2-OCOCH3), 1.28 and 1.27 (each s, each 3H; 2’-CH3, 3’-CH3), 0.89
(s, 9H; SiCACTHNURGTNE(UNG CH3)3), 0.08 and 0.05 ppm (each s, each 3H; each SiCH3);
H4’’’), 3.98 (dd, J2,3 =3.4 Hz, 1H; H2), 3.96 (dd, J6’’a,5’’ =6.7, J6’’a,6’’b
=
13C NMR (150.9 MHz, CDCl3): d=170.9 (2-OCOCH3), 100.3 and 99.8
(C2’,C3’), 98.9 (C1), 71.5 (C5), 70.9 (C2), 66.2 (C3), 63.0 (C4), 61.5 (C6),
ꢀ10.6 Hz, 1H; H6’’a), 3.95 (dd, J6a,5 =2.7, J6a,6b =ꢀ12.6 Hz, 1H; H6a),
3.95 (dd, J5’’’,6’a =3.0, J5’’’,6’’’b =8.8 Hz, 1H; H5’’’), 3.95 (dd, J3’’’,2’’’ =10.2 Hz,
1H; H3’’’), 3.89 (dd, J6’a,5’ =2.2, J6’a,6’b =ꢀ12.5 Hz, 1H; H6’a), 3.86 (dd,
54.8 (1-OCH3), 48.2 and 48.0 (2’-OCH3, 3’-OCH3), 26.0 (SiC
(2-OCOCH3), 18.4 (SiC
(CH3)3), 18.0 and 17.9 (2’-CH3, 3’-CH3), ꢀ5.0 and
ꢀ5.2 ppm (Si(CH3)2); HRMS: m/z calcd for C21H40O9SiNa: 487.2334
[M+Na]+; found: 487.2317.
(2’S,3’S)-Methyl 2-O-acetyl-3,4-O-[2’,3’-dimethoxybutan-2’,3’-diyl]-a-d-
mannopyranoside (21): The reaction was performed under previously de-
ACHTUGNTRNEN(UNG CH3)3), 21.4
AHCTUNGTRENNUNG
J
3,4 =9.6 Hz, 1H; H3), 3.85 (dd, J6b,5 =5.3 Hz, 1H; H6b), 3.85 (dd, J4,5 =
AHCTUNGTRENNUNG
9.8 Hz, 1H; H4), 3.81 (dd, 1H; H2’’’), 3.78 (dd, J6’’b,5’’ =2.1 Hz, 1H;
H6’’b), 3.74 (dd, J6’’’a,6’’’b =ꢀ11.1 Hz, 1H; H6’’’a), 3.74 (dd, 1H; H6’’’b),
3.73 (dd, J6’b,5’ =5.1 Hz, 1H; H6’b), 3.73 (ddd, 1H; H5), 3.68 (dd, J4’,3’
=
ACHTUNGTRENNUNG
9.5, J4’,5’ =9.8 Hz, 1H; H4’), 3.68 (dd, J4’’,3’’ =9.7, J4’’,5’’ =9.9 Hz, 1H; H4’’),
&
12
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
scribed experimental conditions[46] starting from (2’S,3’S)-methyl 2-O-
acetyl-3,4-O-[2’,3’-dimethoxybutan-2’,3’-diyl]-6-O-tert-butyldimethylsilyl-
Acknowledgements
a-d-mannopyranoside (90 mg, 0.2 mmol) and HF/pyridine (80 mL) to
give, after purification by column chromatography with hexane/EtOAc
(1:2) as the eluent, the product as a white syrup (63 mg, 93%). Rf =0.21
(hexane/EtOAc 1:1); [a]D = +159.7 (c=1.2, CHCl3); 1H NMR
(600.13 MHz, CDCl3): d=5.04 (dd, J2,1 =1.6, J2,3 =3.3 Hz, 1H; H2), 4.69
(d, 1H; H1), 4.13 (dd, J3,4 =10.3 Hz, 1H; H3), 4.03 (dd, J4,5 =10.0 Hz,
1H; H4), 3.85 (ddd, J6a,5 =3.0, J6a,6-OH =5.2, J6a,6b =ꢀ11.8 Hz, 1H; H6a),
3.79 (ddd, J6b,5 =4.8, J6b,6-OH =7.7 Hz, 1H; H6b), 3.77 (ddd, 1H; H5), 3.36
(s, 3H; 1-OCH3), 3.264 and 3.260 (each s, each 3H; 2’-OCH3, 3’-OCH3),
2.14 (s, 3H; 2-OCOCH3), 1.88 (dd, 1H; 6-OH), 1.28 and 1.27 ppm (each
s, each 3H; 2’-CH3, 3’-CH3); 13C NMR (150.9 MHz, CDCl3): d=170.6 (2-
OCOCH3), 100.2 and 99.7 (C2’,C3’), 99.2 (C1), 70.5 (C2, C5), 65.8 (C3),
63.6 (C4), 61.6 (C6), 55.0 (1-OCH3), 48.1 and 47.9 (2’-OCH3, 3’-OCH3),
21.2 (2-OCOCH3), 17.8 and 17.6 ppm (2’-CH3, 3’-CH3); HRMS: m/z
calcd for C15H26O9Na: 373.1469 [M+Na]+; found: 373.1461.
Financial support from the Academy of Finland (projects nos 121334 and
121335) and Harry Elvings Legat is gratefully acknowledged. The group
at Madrid is grateful to MICINN for financial support through project
CTQ2009-08536 and also, together with the team from Munich, to the
EC for funding the GlycoHIT program. The authors thank Markku Reu-
nanen for HRMS analysis and Stefan Willfçr and Rainer Sjçholm for
fruitful discussions. Parts of this work are linked to our ongoing collabo-
rative activities within the COST CM1102 Action Multivalent Glycosys-
tems for Nanoscience–MultiGlycoNano.
[1] a) S. Willfçr, K. Sundberg, M. Tenkanen, B. Holmbom, Carbohydr.
Polym. 2008, 72, 197–210; b) A. Ebringerovꢁ, Z. Hromꢁdkovꢁ, T.
Heinze in Polysaccharides I: Structure, Characterization and Use
(Ed.: T. Heinze), Spinger, Berlin, 2005; c) S. Willfçr, A. Sundberg, J.
Hemming, B. Holmbom, Wood Sci. Technol. 2005, 39, 245–257.
[2] a) S. Eda, Y. Akiyama, K. Kato, Carbohydr. Res. 1984, 131, 105–
118; b) R. Schrçder, P. Nicolas, S. J. F. Vincent, M. Fischer, S. Rey-
mond, R. J. Redgewell, Carbohydr. Res. 2001, 331, 291–306; c) T. E.
Timell, Sven. Papperstidn. 1964, 67, 356–363.
[3] a) G. E. Annergren, S. A. Rydholm, Sven. Papperstidn. 1960, 63,
591–600; b) T. Hannuksela, P. Fardim, B. Holmbom, Cellulose 2003,
10, 317–324; c) A. Suurnakki, A. T. Oksanen, H. Kettunen, J. Bu-
chert, Nord. Pulp Pap. Res. J. 2003, 18, 429–435.
Phenyl 2-O-acetyl-3-O-benzyl-4,6-O-benzylidene-1-thio-a-d-mannopyra-
noside:[47] The reaction was performed according to previously described
procedures[45] starting from phenyl 3-O-benzyl-4,6-O-benzylidene-1-thio-
a-d-mannopyranoside (0.44 g, 0.98 mmol), Ac2O (3.6 mL), and pyridine
(3.2 mL) to give, after purification by column chromatography with
hexane/EtOAc (2:1) as the eluent, the product as a yellowish oil (0.44 g,
92%). Rf =0.57 (hexane/EtOAc 2:1); Selected analytical data: 1H NMR
(600.13 MHz, CDCl3): d=7.53–7.27 (m, 15H; arom. H), 5.65 (s, 1H;
CHPh), 5.62 (dd, J2,1 =1.4, J2,3 =3.4 Hz, 1H; H2), 5.46 (d, 1H; H1), 4.73
and 4.70 (each d, J=ꢀ12.3 Hz, each 1H; 3-OCH2Ph), 4.36 (ddd, J5,6a
=
[4] S. Willfçr, P. Rehn, A. Sundberg, K. Sundberg, B. Holmbom, Tappi
J. 2003, 2, 27–32.
[5] A. Ebringerovꢁ, Z. Hromꢁdkovꢁ, V. Hrꢇbalovꢁ, C. Xu, A. Sundberg,
S. Willfçr, Int. J. Biol. Macromol. 2008, 42, 1–5.
4.8, J5,6b =10.2, J5,4 =10.3 Hz, 1H; H5), 4.24 (dd, J6a,6b =ꢀ10.1 Hz, 1H;
H6a), 4.14 (dd, J4,3 =9.9 Hz, 1H; H4), 4.01 (dd, 1H; H3), 3.86 (dd, 1H;
H6b), 2.16 ppm (s, 3H; 2-OCOCH3); 13C NMR (150.9 MHz, CDCl3): d=
170.1 (2-OCOCH3), 137.8–125.8 (arom. C), 101.6 (CHPh), 87.2 (C1), 78.5
(C4), 74.1 (C3), 72.4 (3-OCH2Ph), 71.4 (C2), 68.4 (C6), 65.2 (C5),
21.0 ppm (2-OCOCH3).
[6] a) T. Hannuksela, C. Hervꢂ du Penhoat, Carbohydr. Res. 2004, 339,
ˇ
301–312; b) P. Capek, J. Alfçldi, D. LiSkovꢁ, Carbohydr. Res. 2002,
337, 1033–1037; c) S. Willfçr, R. Sjçholm, C. Laine, M. Roslund, J.
Hemming, B. Holmbom, Carbohydr. Polym. 2003, 52, 175–187.
[7] H. Meier, Acta Chem. Scand. 1960, 14, 749–756.
[8] a) J. Lundqvist, A. Teleman, L. Junel, G. Zacchi, O. Dahlman, F.
Tjerneld, H. Stꢈlbrand, Carbohydr. Polym. 2002, 48, 29–39; b) J.
Thornton, R. Ekman, B. Holmbom, F. ꢉrsꢈ, J. Wood Chem. Tech-
nol. 1994, 14, 159–175.
[9] See, for example: a) D. Crich, H. Li, Q. Yao, D. J. Wink, D. R.
Sommer, A. L. Rheingold, J. Am. Chem. Soc. 2001, 123, 5826–5828;
b) D. Crich, P. Jayalath, T. K. Hutton, J. Org. Chem. 2006, 71, 3064–
3070; c) M. Polꢁkovꢁ, M. U. Roslund, F. S. Ekholm, T. Saloranta, R.
Leino, Eur. J. Org. Chem. 2009, 870–888.
Phenyl 2-O-acetyl-3-O-benzyl-1-thio-a-d-mannopyranoside (22): Phenyl
2-O-acetyl-3-O-benzyl-4,6-O-benzylidene-1-thio-a-d-mannopyranoside
(58 mg, 0.11 mmol) was dissolved in AcOH (80%, 3 mL), and the result-
ing mixture was stirred for 3 h at 608C and concentrated. The crude
product was purified by column chromatography with hexane/EtOAc
(1:1) as the eluent to give the product as a colorless oil (43 mg, 89%).
Rf =0.18 (hexane/EtOAc 1:1). Selected analytical data: 1H NMR
(600.13 MHz, CDCl3): d=7.50–7.20 (m, 10H; arom. H), 5.60 (dd, J2,1
=
1.6, J2,3 =3.2 Hz, 1H; H2), 5.47 (d, 1H; H1), 4.74 and 4.48 (each d, J=
ꢀ11.2 Hz, each 1H; 3-OCH2Ph), 4.18 (ddd, J5,6a =3.4, J5,6b =4.8, J5,4
9.8 Hz, 1H; H5), 3.97 (dd, 4,3 =9.5 Hz, 1H; H4), 3.86 (dd, J6a,6b
=
=
J
ꢀ11.7 Hz, 1H; H6a), 3.83 (dd, 1H; H6b), 3.76 (dd, 1H; H3), 2.11 ppm (s,
[10] a) D. A. McGowan, G. A. Berchtold, J. Am. Chem. Soc. 1982, 104,
7036–7041; b) F. S. Ekholm, M. Polꢁkovꢁ, A. J. Pawłowicz, R.
Leino, Synthesis 2009, 4, 567–576.
[11] a) M. P. Deninno, J. B. Etienne, K. C. Duplantier, Tetrahedron Lett.
1995, 36, 669–672; b) A. Arasappan, B. Frasier-Reid, J. Org. Chem.
1996, 61, 2401–2406.
3H; 2-OCOCH3).
STD NMR spectroscopic experiments: STD experiments were performed
with a lectin concentration of 59 mm on a Bruker Avance 500 MHz NMR
spectrometer equipped with a 5 mm TXI inverse probe head. Protein sat-
uration was achieved by using a cascade of 40 selective 70 dB Gaussian
pulses of 50 ms in total duration and a total number of 360 scans[39] at a
ligand/lectin ratio of 25:1.[38b]
[12] S. D. Debenham, E. J. Toone, Tetrahedron: Asymmetry 2000, 11,
385–387.
[13] a) H. M. I. Osborn in Carbohydrates (Ed.: H. M. I. Osborn), Aca-
demic Press, Oxford, 2003; b) G. Zemplꢂn, ꢊ. Gerecs, I. Hadꢁcsy,
Ber. Dtsch. Chem. Ges. 1936, 69, 1827–1830.
[14] a) G. H. Veeneman, J. H. van Boom, Tetrahedron Lett. 1990, 31,
275–278; b) G.-J. Boons, S. Bowers, D. M. Coe, Tetrahedron Lett.
1997, 38, 3773–3776; c) H. Yin, T. L. Lowary, Tetrahedron Lett.
2001, 42, 5829–5832.
[15] a) D. Crich, S. Sun, Tetrahedron 1998, 54, 8321–8348; b) D. Crich,
A. Banerjee, Q. J. Yao, J. Am. Chem. Soc. 2004, 126, 14930–14934.
[16] D. Crich, B. Wu, Org. Lett. 2006, 8, 4879–4882.
[17] a) M. Kojima, Y. Nakamura, S. Takeuchi, Tetrahedron Lett. 2007, 48,
4431–4436; b) M. Sakagami, H. Hamana, Tetrahedron Lett. 2000,
41, 5547–5551.
Molecular modeling: Potential-energy maps were performed by using the
Coordinate Scan tool implemented in MAESTRO,[34] as described. The
generalized Born/surface area (GB/SA solvation model for water was
used. The calculations was performed with the AMBER force field.[35]
The same force field was employed for the MD simulations performed
on two models of tetrasaccharide 2, built from the global-minimum con-
formers found for every F/Y glycosidic linkage. The two models were
generated with either the gg or gt rotamers around the a-Galp-(1!6)-b-
Manp linkage of 2. MD simulations of 6 ns were carried out with equili-
bration of 100 ps, and an integration time step of 2 fs.
[18] a) F. G. De Las Heras, A. San Felix, A. Calvo-Mateo, P. Fernandez-
Resa, Tetrahedron 1985, 41, 3867–3873; b) P. Fꢋgedi, P. Nꢁnꢁsi, J.
Szejtli, Carbohydr. Res. 1982, 104, 55–67.
[19] K. Daragics, P. Fꢋgedi, Tetrahedron Lett. 2009, 50, 2914–2916.
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
13
&
ÞÞ
These are not the final page numbers!

R. Leino et al.
[20] C.-R. Shie, Z.-H. Tzeng, S. S. Kulkarni, B.-J. Uang, C.-Y. Hsu, S.-C.
Hung, Angew. Chem. 2005, 117, 1693–1696; Angew. Chem. Int. Ed.
2005, 44, 1665–1668.
Hoffmann, L. A. Russell, M. Debet, Carbohydr. Res. 1995, 271, 43–
54; c) I. M. Sims, D. J. Craik, A. Basic, Carbohydr. Res. 1997, 303,
79–92.
[21] a) R. Johnsson, D. Olsson, U. Ellervik, J. Org. Chem. 2008, 73,
5226–5232; b) M. Ohlin, R. Johnsson, U. Ellervik, Carbohydr. Res.
2011, 346, 1358–1370.
[34] Maestro, A powerful, all-purpose molecular-modeling environment,
Schrçdinger LLC, 2005.
[35] N. L. Allinger, Y. H. Yuh, J. H. Lii, J. Am. Chem. Soc. 1989, 111,
8551–8566.
[36] V. Roldꢃs, F. J. CaÇada, J. Jimꢂnez-Barbero, ChemBioChem 2011,
12, 990–1005.
[37] H.-J. Gabius, S. Andrꢂ, J. Jimꢂnez-Barbero, A. Romero, D. Solꢇs,
Trends Biochem. Sci. 2011, 36, 298–313.
[38] a) H.-J. Gabius, H. Walzel, S. S. Joshi, J. Kruip, S. Kojima, V. Gerke,
H. Kratzin, S. Gabius, Anticancer Res. 1992, 12, 669–676; b) M. Ko-
lympadi, M. Fontanella, C. Venturi, S. Andrꢂ, H.-J. Gabius, J. Jimꢂ-
nez-Barbero, P. Vogel, Chem. Eur. J. 2009, 15, 2861–2873.
[39] R. T. Lee, H.-J. Gabius, Y. C. Lee, J. Biol. Chem. 1992, 267, 23722–
23727.
[22] a) A. K. Sarkar, A. K. Ray, N. Roy, Carbohydr. Res. 1989, 190, 181–
190; b) A. Imamura, H. Ando, S. Korogi, G. Tanabe, O. Muraoka,
H. Ishida, M. Kiso, Tetrahedron Lett. 2003, 44, 6725–6728; c) A. Im-
amura, H. Ando, H. Ishida, M. Kiso, Org. Lett. 2005, 7, 4415–4418.
[23] See, for example: a) J. D. M. Olsson, J. Landstrçm, J. Rçnnols, S. Os-
carson, G. Widmalm, Org. Biomol. Chem. 2009, 7, 1612–1618; b) S.
Huang, H. Yu, X. Chen, Angew. Chem. 2007, 119, 2299–2303;
Angew. Chem. Int. Ed. 2007, 46, 2249–2253; c) O. N. Yudina, A. A.
Sherman, N. E. Nifantiev, Carbohydr. Res. 2001, 332, 363–371.
[24] a) C. Mukherjee, A. K. Misra, Glycoconjugate J. 2008, 25, 111–119;
see also: b) R. M. Owen, C. B. Carlson, J. Xu, P. Mowery, E. Fasella,
L. L. Kiessling, ChemBioChem 2007, 8, 68–82.
[25] A. Imamura, A. Kimura, H. Ando, H. Ishida, M. Kiso, Chem. Eur.
J. 2006, 12, 8862–8870.
[26] a) K. Bock, H. Thøgersen in Annual Reports on NMR Spectroscopy
(Ed.: G. A. Webb), Academic Press/Elsevier, London, 1982, pp. 2–
49; b) K. Bock, C. Pedersen, J. Chem. Soc. Perkin Trans. 2 1974,
293–299.
[40] B. Meyer, T. Peters, Angew. Chem. 2003, 115, 890–918; Angew.
Chem. Int. Ed. 2003, 42, 864–890.
[41] S. Martꢇn-Santamarꢇa, S. Andrꢂ, E. Buzamet, R. Caraballo, G.
Fernꢁndez-Cureses, M. Morando, J. P. Ribeiro, K. Ramꢇrez-Gualito,
B. de Pascual-Teresa, F. J. CaÇada, M. Menꢂndez, O. Ramstrçm, J.
Jimꢂnez-Barbero, D. Solꢇs, H. J. Gabius, Org. Biomol. Chem. 2011,
9, 5445–5455.
[27] A. Imamura, H. Ando, H. Ishida, M. Kiso, Heterocycles 2008, 76,
883–908.
[28] a) A. Kirschning, G.-W. Chen, G. Drꢄger, I. Schuberth, L. F. Tietze,
Bioorg. Med. Chem. 2000, 8, 2347–2354; b) K. C. Nicolaou, S. E.
Webber, Synthesis 1986, 453–461.
[42] a) M. Jimꢂnez, S. Andrꢂ, H.-C. Siebert, H.-J. Gabius, D. Solꢇs, Gly-
cobiology 2006, 16, 926–937; b) M. Jimꢂnez, S. Andrꢂ, C. Barillari,
A. Romero, D. Rognan, H.-J. Gabius, D. Solꢇs, FEBS Lett. 2008,
582, 2309–2312.
[43] For experimental details, see: D. K. Baeschlin, L. G. Green, M. G.
Hahn, B. Hinzen, S. J. Ince, S. V. Ley, Tetrahedron: Asymmetry 2000,
11, 173–197.
[29] See the Experimental Section for more details on the synthetic
methodologies utilized.
[30] F. S. Ekholm, J. Sinkkonen, R. Leino, New J. Chem. 2010, 34, 667–
675.
[31] a) A. Bax, D. G. Davis, J. Magn. Reson. 1985, 65, 355–360; b) H.
Kessler, H. Oschkinat, C. J. Griesinger, J. Magn. Reson. 1986, 70,
106–133; c) S. Berger, S. Braun, 200 and More NMR Experiments,
Wiley-VCH, Weinheim, 1998, 488–491.
[44] R. Alibꢂs, D. R. Bundle, J. Org. Chem. 1998, 63, 6288–6301.
[45] C. S. Hudson, J. K. Dale, J. Am. Chem. Soc. 1915, 37, 1264–1282.
[46] J. Robertson, P. M. Stafford in Carbohydrates (Ed.: H. M. I.
Osborn), Academic Press/Elsevier, Oxford, 2003, pp. 54–55.
[47] S. Cherif, J.-M. Clavel, C. Monneret, J. Carbohydr. Chem. 2002, 21,
123–130.
[32] a) R. Laatikainen, M. Niemitz, U. Weber, J. Sundelin, T. Hassinen, J.
Vepsꢄlꢄinen, J. Magn. Reson. Ser. A 1996, 120, 1–10; b) for detailed
information on the software, see: http://www.perchsolutions.com.
[33] a) J. M. van Hazendonk, E. J. M. Reinerrink, P. van Waard, J. E. G.
van Dam, Carbohydr. Res. 1996, 291, 141–154; b) L. A. Davis, R. A.
Received: February 16, 2012
Revised: May 16, 2012
Published online: && &&, 0000
&
14
&
www.chemeurj.org
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Synthesis and Study of Tri- and Tetrasaccharide Fragments of Galactoglucomannan
FULL PAPER
Modeling behavior: Detailed studies
on the chemical synthesis, NMR spec-
troscopic characterization, conforma-
tion analysis, and molecular recogni-
tion of molecular models of oligosac-
charide fragments present in the back-
bone of galactoglucomannan polysac-
charides found in natural Norway
spruce are reported (see figure). A tet-
rasaccharide fragment was further uti-
lized in a binding study with the galac-
tose-binding protein viscumin to test
bioaffinity.
Polysaccharides
F. S. Ekholm, A. Ardꢀ, P. Eklund,
S. Andrꢁ, H.-J. Gabius,
J. Jimꢁnez-Barbero,
R. Leino* ...................... &&&& — &&&&
Studies Related to Norway Spruce
Galactoglucomannans: Chemical
Synthesis, Conformation Analysis,
NMR Spectroscopic Characterization,
and Molecular Recognition of Model
Compounds
Chem. Eur. J. 2012, 00, 0 – 0
ꢅ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
15
&
ÞÞ
These are not the final page numbers!