Synthesis of two lipopentasaccharides related to the nodulation factors of the nitrogen-fixing bacterium Rhizobium sp. NGR 234

Full text of DOI:10.1016/j.mencom.2007.01.004

Mendeleev
Communications
Mendeleev Commun., 2007, 17, 10–13
Synthesis of two lipopentasaccharides related to the nodulation
factors of the nitrogen-fixing bacterium Rhizobium sp. NGR 234
Sergey L. Sedinkin, Alexander I. Zinin, Nelli N. Malysheva, Alexander S. Shashkov,
Vladimir I. Torgov* and Vladimir N. Shibaev†
N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 119991 Moscow,
Russian Federation. Fax: +7 495 135 5328; e-mail: vladimir.torgov@mail.ru
DOI: 10.1016/j.mencom.2007.01.004
Two lipopentasaccharides related to the Nod factors of Rhizobium sp. NGR 234 have been synthesised.
The coordinated interactions of nitrogen-fixing bacteria of the
OAc
OH
genus Rhizobium and legumes become established as a result of
mutual exchange with signaling molecules between simbionts.1,2
So-called nodulation factors (Nod factors) play an important
role in the processes, which finally lead to the formation of
nodules containing nitrogen-fixing bacteria.3,4 Chemically, the
Nod factors represent lipochitooligosaccharides containing a
core constructed from four to five β-(1®4) bonded glucosamine
residues. The non-reducing glucosamine residue is N-acylated
by a long-chain fatty acid, and all other amino groups are
acetylated. A characteristic property of these compounds is an
additional modification of both reducing N-acetylglucosamine
and non-reducing N-acylglucosamine. These structural features
control the specificity of interactions between rhizobia and their
host plants.
OMe
O
Me
O
OH
OH
O
O
O
HO
HO
O
HO
O
HO
2
X
NH
NHAc
NHAc
O
1 X = OH
2 X = β-OMe
other protecting groups. Thus, protected pentasaccharide deriva-
tives 3 and 4, the precursors of target compounds 1 and 2,
should contain the following set of protecting groups: hydroxyl
groups of chitooligosaccharide core are protected with benzyl
groups, the amino group of the non-reducing glucosamine
residue is masked as an azido group, the amino groups of three
glucosamine residues are protected as phthalimido groups, the
α-L-fucose residue bears a methyl group at O-2, acetate at O-4
and a p-methoxybenzyl group at O-3.
Here we describe an approach that allows synthesising lipo-
oligosaccharides related to Nod factors from Rhizobium sp.
NGR 234, which enters into symbiosis with more than 110
genera of legumes.5
OY1
OY2
OMe
Me
O
OAc
MPM
O
OH
O
OH
O
OMe
O
O
O
Me
OBn
O
XO
XO
O
HO
O
HO
OH
NHAc
E
OBn
OBn
O
NMe
NHAc
3
D
C
B
A
O
O
O
O
R1
O
O
BnO
BnO
O
OR
NPhth
BnO
BnO
NPhth
BnO
NPhth
Y1 = H or Ac, Y2 = H or SO3H, X = H or CONH2
R1 = (CH2)9CH=CH(CH2)5Me or (CH2)14Me
N3
3 R = Bn
4 R = Me
This study deals with the synthesis of Rhizobium sp. NGR
234 Nod factors analogues, namely, pentasaccharide 1 and its
methyl glycoside 2 with a fixed anomeric configuration at the
reducing end.
Phth = o-C6H4(CO)2
MPM = p-MeOC6H4CH2
While planning the synthesis of protected derivatives 3 and
4, the second problem emerged: introduction of an α-L-fucose
residue at the 6-position of the reducing end of tetrasaccharide.
For this purpose, it is necessary to have a selectively removable
protecting group at that position. It is well known6 that fucosyla-
tion may lead to a mixture of α- and β-isomers, and their
separation could become an unsolvable problem for large mole-
cules. Thus, our decision to perform fucosylation at the di-
or trisaccharide step is strategically better justified than the
fucosylation of oligosaccharide acceptors with a chain length of
more than three sugar residues.
Structural analysis of the target molecules demonstrated that
the amino group of the non-reducing glucosamine residue should
be acylated independently of other amino groups. One of the
main tasks is the careful selection of protecting groups for
the amino groups of glucosamine residues, which will allow
making a chemical differentiation of the two types of amino
groups in the molecule. To solve the problem, we decided to
utilise the orthogonality of azido and phthalimido protecting
groups together with mild deprotecting conditions of all the
†
Deceased.
– 10 –

Mendeleev Commun., 2007, 17, 10–13
We chose a convergent synthetic scheme for constructing the
Glycosylation of acceptors 10 and 11 with thioglycoside 9a8
effected by N-iodosuccinimide (NIS) and Et3SiOSO2CF3 (TESOTf)
at –20 °C led to disaccharide derivatives 13 and 14 in 86% and
95% yields, respectively. Their structures were proved by NMR
data. The spin–spin coupling constant J1',2' 8.1 Hz for 13 or 8.0 Hz
for 14 confirmed the formation of β-(1®4)-disaccharides.
carbohydrate backbone of pentasaccharides 3 and 4, due to the
expectation of increasing yield in comparison with the linear
scheme. Moreover, the separation of β-L-fucoside, which is
formed in the course of α-L-fucosylation in the trisaccharide
synthesis, would be much easier than in the case of longer
oligosaccharides. Disaccharide glycosyl donor 5 and two tri-
saccharide acceptors 6 and 7 were chosen for the synthesis.
OBn
OMP
O
O
OBn
OBn
ClH2CCOO
BnO
HO
+
SEt
NPhth
OR
NPhth
BnO
O
O
BnO
BnO
O
SEt
NPhth
BnO
9a
10 R = Bn
11 R = Me
N3
5
DCM, MS 4A
NIS, TESOTf
OBn
OAc
O
–20 °C
MPM
OMe
OMP
Me
O
OBn
O
O
O
ClH2CCOO
BnO
O
OR
NPhth
BnO
O
O
NPhth
O
HO
BnO
OR
NPhth
BnO
NPhth
13 R = Bn
14 R = Me
6 R = Bn
7 R = Me
To proceed to fucosylation at the 6-position of the ‘reducing’
glucosamine residue, it was necessary to remove selectively the
methoxyphenyl protecting group from the primary hydroxyl
group. This was accomplished by 15 min treatment of a water–
acetonitrile solution of 13 at 0 °C with a double excess of
ammonium cerium(IV) nitrate followed by column chromato-
graphy, which provided derivative 15 in 87% yield. The absence
of signals corresponding to the methoxyphenyl protecting group
in the NMR spectra proved the structure of 15.
These compounds were synthesised from previously prepared
protected monosaccharide derivatives 8,7 9,8 10, 119 and 12.10
Compound 9 was used as a common precursor for the synthesis of
disaccharide 5, as well as for trisaccharide derivatives 6 and 7.
OBn
OBn
OMP
O
O
O
BnO
BnO
HO
BnO
HO
BnO
SEt
OBn
NPhth
OH
N3
NPhth
13, 14
8
9
10
(NH4)2Ce(NO2)6 MeCN–H2O (9:1), 0 °C
OMP
OAc
O
MPM
OBn
OH
O
OMe
SEt
HO
OMe Me
O
BnO
O
O
NPhth
ClH2CCOO
BnO
O
BnO
OR
NPhth
11
12
MP = p-MeOC6H4
NPhth
Glycosyl donor 8a was prepared by the reaction of hemi-
acetal derivative 8 with trichloroacetonitrile in the presence of
DBU7 in 81% yield. According to NMR data, mainly α-trichloro-
acetimidate 8a was formed with only traces of the β-isomer.
15 R = Bn
16 R = Me
Disaccharide alcohol 16 was obtained analogously in a yield
of 93%. Its structure was also proved by NMR spectra.
Introduction of an α-L-fucose residue into the 6-position of
the reducing glucosamine residue in 15 and 16 was the next
step of the synthesis. We used compound 1210 as a fucose residue
donor, because the combination of protecting groups allows us
to get all the possible variants of fucose functionalisation.
OBn
CCl3CN, DBU
O
8
BnO
DCM, 4 °C, 16 h
BnO
N3
O
CCl3
8a
NH
12
OH
OH
OH
O
It is well known11 that, in the course of glycosylation of
thioglycoside acceptors by trichloroacetimidate donors, sulfur
nucleophilicity leads to side reactions, which may considerably
decrease the yield of the target glycoside. However, we
succeeded in finding the proper reaction conditions where
coupling 8a with acceptor 9 leads to disaccharide glycosyl
donor 5 in 52% yield. The structure of 5 was confirmed by
1H and 13C NMR spectroscopic data. The presence of the spin–
spin coupling constant J1',2' 9.8 Hz in its 1H NMR spectrum
proved the formation of a 1,2-trans glycoside bond.
SO3H
O
OAc
OH
OMe
OR
OMe
Me
O
Me
OMe
OR
OR
Me
O
Glycosylation of acceptors 15 and 16 with donor 12 in the
presence of CuBr2 gave trisaccharide derivatives 17 and 18 in
71 and 77% yields, respectively. The presence of characteristic
signals of the fucose methyl group at 1.1 ppm in the H NMR
spectra of both 17 and 18 proves the trisaccharide formation,
and the spin–spin coupling constant J1,2 3.0 Hz for H-1 of
fucose clearly shows the formation of α-glycosyl bond.
6
1
BF3·Et2O, DCM
8a + 9
5
MS 4A, –45 °C
– 11 –

Mendeleev Commun., 2007, 17, 10–13
For the conversion of pentasaccharide derivatives 3 and 4
12 + 15 or 16
into target compounds 1 and 2, it was necessary to make some
manipulations with protecting groups. Primarily, phthalimido
groups were replaced with acetamido ones by successive
treatment with hydrazine hydrate (without isolation of the
corresponding triamine derivative) followed by acetylation in
the presence of a catalytic amount of DMAP. Column chromato-
graphy of the reaction mixture gave rise to derivatives 19 and
20 in 89 and 87% yields, respectively.
DCM–DMF (5:1),
room temperature
CuBr2, Bu4NBr, MS 4A
OAc
O
MPM
OMe
Me
O
OBn
O
O
O
3, 4
HalH2CCOO
O
BnO
OR
NPhth
BnO
OAc
O
i, N2H4·H2O, PrOH, reflux
ii, Ac2O/Py, DMAP
NPhth
MPM
OMe
17 R = Bn
18 R = Me
Me
OBn
O
Hal = Cl, Br
OBn
OBn
O
Heating trisaccharides 17 and 18 in a pyridine–water mixture
led to the complete removal of the haloacetyl group. Chromato-
graphy of the reaction mixtures gave trisaccharidic glycosyl
acceptors 6 and 7 in 98 and 95% yields, respectively.
O
O
O
O
BnO
BnO
O
O
O
OR
NHAc
BnO
BnO
BnO
NHAc
N3
NHAc
19 R = Bn
20 R = Me
Py–H2O (4:1), 80 °C
17, 18
6, 7
The absence of signals corresponding to phthalimido groups
and the presence of three N-acetyl signals in the NMR spectra of
both 19 and 20 confirmed the completeness of transformation.
Simultaneous reduction of the azido group to the amino one
together with the removal of both benzyl and methoxybenzyl
groups by hydrogenolysis over a palladium catalyst could be an
attractive procedure for the conversion of protected derivatives
19 and 20 into unprotected oligosaccharides with a free amino
group at the non-reducing end. It was impossible to carry out
this transformation due to incomplete reaction. Probably, it was
accounted for by catalyst poisoning with the amine formed
because repetitive hydrogenolysis with fresh portions of the
catalyst led the reaction to completion.
A two-step variation of such conversion was reported,12
where an azido group was reduced to an amino group with
SmI2 followed by hydrogenolysis of benzyl groups over a
palladium catalyst.
After the addition of a THF solution of SmI2 to solutions
of protected 19 and 20 in aqueous THF, TLC of the reaction
mixtures showed complete reduction of azido groups into amines,
and column chromatography gave pure amino pentasaccharide
derivatives 21 and 22 in quantitative yields.
Glycosylation of trisaccharide derivatives 6 and 7 with a
15% molar excess of disaccharide donor 5 in the presence of
NIS and a catalytical amount of TESOTf was accomplished in
15 min according to TLC. After column chromatography of
the reaction mixtures, protected pentasaccharides 3 and 4 were
isolated in equal yields of 83%.
NIS, TESOTf
6, 7 + 5
3, 4
DCM, MS 4A, –20 °C
The presence of all characteristic signals in the 1H and 13C NMR
spectra of 3 and 4 together with the spin–spin coupling constants
(8.0–8.2 Hz) for H-1 of glucosamine residues proved the struc-
tures of the compounds. Additional confirmation of the structures
1
was obtained by two-dimensional H–1H ROESY spectroscopy
(Figure 1).
The presence of the following cross-peaks, namely, H-1 C/
H-4 B, H-1 C/H-6 B, H-1 B/H-6' A (Figure 1), and H-1 D/H-4
C, H-1 D/H-6'C in the ROESY spectrum of 3 unambiguously
showed both the presence of a chitotetraose core with β-(1®4)
glycoside bonds and α-(1®6) bond between L-fucose and reducing
glucosamine residues. Analogous spectral data confirming the
structure of 4 were obtained.
19, 20
OAc
MPM
SmI2, THF–H2O (10:1), 0 °C
O
OMe
Me
OBn
O
1A/5A
OBn
OBn
O
3.2
3.3
3.4
3.5
1E/6'A
1B/6'A
1C/5C
O
O
O
O
BnO
BnO
O
O
O
OR
NHAc
1B/5B
BnO
BnO
BnO
NHAc
NH2
NHAc
1C/6B
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
4.5
4.6
21 R = Bn
22 R = Me
Hydrogenolysis of pentasaccharide amines 21 and 22 in a
mixture of EtOH and AcOH with the addition of 0.5% formic
acid over 10% Pd(OH)2/C afforded pentasaccharides 23 and
24, which were isolated by gel filtration on Sephadex G-25 in
85 and 80% yields, respectively. NMR data for 23 and 24
confirmed the complete removal of benzyl and methoxybenzyl
groups. High-resolution mass spectra for 23 and 24 comply
with the calculated molecular weight (for 23 calculated 990.40,
detected 989.39 [M – 1]; for 24 calculated 1004.42, detected
1003.41 [M – 1]). Note that, in the mass spectrum of 23, a
minor signal at 1017.42 was present, which was interpreted as
1B/4A
1C/2C
1B/2,3B
1C/3C
1C/4B
5.3
5.2
5.1
5.0
4.9
4.8
d/ppm
Figure 1 1H–1H ROESY spectrum of 3.
– 12 –

Mendeleev Commun., 2007, 17, 10–13
an impurity of a derivative N-ethylated at the non-reducing
the deuterated solvents or their mixtures. Only general spectra
were recorded for 1 and 2, containing the signals present in the
spectra of starting compounds 23 and 24 together with those of
vaccenic acid, proving the formation of conjugates. The negative
ninhydrin reaction showed the absence of free amino groups.
The coincidence of calculated molecular masses for 1 and 2
(1254.65 and 1268.66, respectively) and those recorded by
high-resolution mass spectrometry (1253.66 and 1267.67 for
[M – 1], respectively) finally proved the structure of the target
compounds.
end. Due to a much lower reaction rate for the acylation of
secondary amines in comparison with primary ones, we
decided to use substance 23 without further purification.
21, 22
OAc
OH
H2, Pd(OH)2/C,
EtOH–AcOH (10:1), 36 °C
OMe
O
Me
O
O
OH
OH
OH
O
Et3N, DMSO
O
O
23, 24 + 26
1, 2
HO
HO
O
O
O
HO
HO
NHAc
HO
X
NH2
NHAc
NHAc
As a conclusion, two lipooligosaccharides 1 and 2 related to
Nod factors from Rhizobium sp. NGR 234 have been synthesised.
23 X = OH
24 X = β-OMe
Online Supplementary Materials
Selective acylation of the amino group in 23 and 24 is the
last step in the synthesis of target lipooligosaccharides 1 and 2.
There are many acylation methods, but the most convenient
procedure for the acylation of amino groups in the presence of
hydroxyls uses the N-hydroxysuccinimide esters of corresponding
carboxylic acids.
Experimental part including both complete synthetic methods
and spectral data is presented in Online Supplementary Materials
which can be found in the online version at doi:10.1016/
j.mencom.2007.01.004.
References
Activated ester 26 was obtained from (Z)-11-octadecenoic
1
acid 25. The H and 13C NMR spectra of 26 contained signals
1
2
R. F. Fisher and S. R. Long, Nature, 1992, 357, 655.
J. Denarie, F. Debelle and C. Rosenberg, Ann. Rev. Microbiol., 1992,
46, 497.
at 2.82 and 25.5 ppm, respectively, related to two methylene
groups of succinimide; the signals of protons at the double bond
remained unchanged,13 indicating conservation of (Z)-configura-
tion of the double bond.
3
N. L. Price, B. Relic, F. Talmont, A. Lewin, D. Prome, S. G. Pueppke,
J. Denarie, J.-C. Prome and W. J. Broughton, Mol. Microbiol., 1992, 6,
3575.
4
5
6
N. L. Price, F. Talmont, J. M. Wieruszeski, D. Prome and J. C. Prome,
Carbohydr. Res., 1996, 289, 115.
X. Perret, C. Staehelin and W. J. Broughton, Microbiol. Mol. Biol. Rev.,
O
HO
2000, 64, 180.
25
S. Sato, M. Mori, Y. Ito and T. Ogawa, Carbohydr. Res., 1986, 155,
C6–C10.
NHS, DCC, DMAP, DMF
7
8
9
W. Kinzy and R. R. Schmidt, Liebigs Ann. Chem., 1985, 1537.
C. Unverzagt, Chem. Eur. J., 2003, 6, 1369.
S. Ikeshita, A. Sakamoto, Y. Nakahara, Y. Nakahara and T. Ogawa,
Tetrahedron Lett., 1994, 3124.
O
O
N
O
10 A. I. Zinin, V. I. Torgov, V. N. Shibaev, A. S. Shashkov and W. J.
Broughton, Izv. Akad. Nauk, Ser. Khim., 1998, 513 (Russ. Chem. Bull.,
1998, 47, 496).
11 F. Belot and J.-C. Jacquinet, Carbohydr. Res., 1996, 290, 79.
12 D. Tailler, J.-C. Jacquinet and J.-M. Beau, J. Chem. Soc., Chem.
Commun., 1994, 1827.
26
O
Coupling of 26 with pentasaccharide derivatives 23 and 24
was carried out at room temperature in a DMSO solution in the
presence of a catalytic amount of triethylamine. TLC showed
complete conversion in 24 h. After column chromatography,
compounds 1 and 2 were isolated in 70% yield. At this
step, compound 1 was separated from the admixture present
in starting compound 23.
13 J. G. Batchelor, J. H. Prestegard, R. J. Cushley and S. R. Lipsky, J. Am.
Chem. Soc., 1973, 95, 6358.
The complete characterisation of target lipooligosaccharides
1 and 2 by NMR spectroscopy was complicated by the impos-
sibility to obtain a solution of sufficient concentration in any of
Received: 13th July 2006; Com. 06/2753
– 13 –