Design, synthesis, and enzymatic property of a sulfur-substituted analogue of trigalacturonic acid

Article abstract of DOI:10.1016/j.bmcl.2005.08.019

A sulfur-substituted analogue of trigalacturonic acid (3) was synthesized. The synthesis features the application of 3-cyano-3-(tert-butyldimethylsilyl) oxypropylthioether (CSP) as a novel protective group for thiols. This analogue was designed with the expectation that it would be a stable analogous substrate for endo-polygalacturonase isolated from Stereum purpureum based on computer modeling experiments. Surface plasmon resonance experiments revealed that 3 forms a stable complex with the target enzyme.

Full text of DOI:10.1016/j.bmcl.2005.08.019

Bioorganic & Medicinal Chemistry Letters 15 (2005) 4932–4935
Design, synthesis, and enzymatic property of a
sulfur-substituted analogue of trigalacturonic acid
Kazunori Yamamoto,a Naoki Watanabe,a Hiroko Matsuda,a Keiichiro Oohara,a
Tomoyuki Araya,a Masaru Hashimoto,a,* Kazuo Miyairi,a Isao Okazaki,b Minoru Saito,b
Tetsuya Shimizu,c Hiroaki Katoc and Toshikatsu Okunoa
aFaculty of Agriculture and Lifescience, Hirosaki University, 3-Bunkyo-cho, Hirosaki 036-8561, Japan
bFaculty of Science and Technology, Hirosaki University, 3-Bunkyo-cho, Hirosaki 036-8561, Japan
cGraduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-Shimo-Adachi-cho, Sakyo-ku, Kyoto 606-8501, Japan
Received 24 June 2005; revised 18 July 2005; accepted 6 August 2005
Available online 16 September 2005
Abstract—A sulfur-substituted analogue of trigalacturonic acid (3) was synthesized. The synthesis features the application of
3-cyano-3-(tert-butyldimethylsilyl)oxypropylthioether (CSP) as a novel protective group for thiols. This analogue was designed with
the expectation that it would be a stable analogous substrate for endo-polygalacturonase isolated from Stereum purpureum based on
computer modeling experiments. Surface plasmon resonance experiments revealed that 3 forms a stable complex with the target
enzyme.
Ó 2005 Elsevier Ltd. All rights reserved.
Endo-polygalacturonase (endo-PGs) hydrolyzes the
internal glycosidyl linkages of polygalacturonic acids
(pectic acids) into oligomers. Because oligogalacturonic
acids (low-molecular-sized pectic acids) have recently
attracted attention as functional supplemental foods,1
such as Ca2+ carriers,2 dietary fiber,3 and so on, investi-
gating in detail that the mechanism of endo-PG has
gained importance in the last decade. However, elucida-
tion of the details of the reaction for endo-type glycosi-
dases including endo-PGs generally lags behind those
of exo-glycosidases due to difficulty in designing suitable
molecular probes.4
complex of endo-PG 1 with the oligogalacturonate as a
reactant, because enzymatic hydrolysis occurred in the
crystalline to afford only the complex with the hydroly-
sates. Thus, we planned to investigate in detail the reac-
tion mechanism of this enzyme employing stable
oligogalacturanate mimics. We report here the design
and synthesis of sulfur-substituted analogue 3, designed
as a stable mimic of natural substrate 2 against
endo-PG 1.
We expected that replacing one acetal oxygen cleaved by
the enzyme would result in tolerance to the enzyme with
minimum structural alteration.7 On the other hand,
Miyairi studied the kinetic properties of endo-PG 1
revealing that the minimum size as a substrate is trimer
2.6 As shown in Figure 1, the enzyme hydrolyzes 2 site-
specifically at the glycoside bond between C4O and C10.
Tetramer 1 was hydrolyzed more easily by this enzyme,
but, site specificity decreased, also cleaving the bond
between C40O and C100 at a ratio of 10%. This suggests
that the enzyme can slide on the sugar chain. In con-
trast, in these experiments, the terminal pyranose unit
of 2 binds selectively with the subsite +1 in endo-PG 1.
Accordingly, replacement of the glycosidic oxygen
(atom X in Fig. 1) with a sulfur atom would realize resis-
tance to endo-PG 1 with minimum structural alteration,
and this analogue would create a homogeneous complex
with the target enzyme. We expected that replacing only
Recently, Miyairi and Okuno, authors of the present pa-
per, found endo-PG 1 to be the substance responsible for
silver leaf disease on apples from the pathogenic Stere-
um purpureum.5 This glycosidase causes degradation of
pectin in the leaf cell wall, inducing the disease. Interest-
ingly, endo-PG 1 makes single crystallines that provide
an X-ray crystallographic structure of extremely high
6
˚
quality (less than 0.96 A resolution, PDB ID: 1K5C).
However, soaking experiments have not provided the
Keywords: Sulfur analogue of trigalacturonate; endo-PG; Molecular
dynamic calculation.
*
Corresponding author. Tel.: +81 172 39 3782; fax: +81 172 39
3782; e-mail: hmasaru@cc.hirosaki-u.ac.jp
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.08.019

K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4932–4935
4933
OH
O
HO
HO
OH
HO
-3
-2
COOH
O
COOH
HO
O
HO
HO
e
1"
ndo-
O
HO
4'
COOH
O
COOH
O
1'
10%
HO
O
HO
HO
X
100%
-1
HO
HO
COOH
O
4
COOH
O
90%
O
HO
HO
OR
+1
subsite
COOH
O
2: X = O, R = H
3: X = S, R = Me
OMe
1
Figure 1.
one oxygen would not cause serious structural alteration
in complexation.
Prior to the synthesis, we evaluated analogue 3 by
molecular dynamic simulations for the matrix of endo-
PG 1 and 3 including 8633 waters.8 These simulations
suggested that analogue 3 would bind to the enzyme in
almost the same hydrogen-bonding manner (Fig. 2) as
the presumed native complex. This native complex was
constructed based on two X-ray structures, a complex
with two mono-galacturonic acids that bind at the sub-
sites +1 and ꢀ1 individually (PDB ID: 1KCD) obtained
by using excess mono-galacturonic acid, and a complex
with a pentamer at subsite ꢀ1, ꢀ2, ꢀ3, ꢀ4 prepared by
soaking at higher pH solution in which the enzyme was
inactivated.9 Neither complex involved the glycoside
bonds between ꢀ1 and +1 (the reaction subsite).
Although the substituting glycosidyl oxygen to sulfur
did cause a minor conformational alteration, it did not
greatly disturb the network of hydrogen bondings
between the enzyme and the substance as shown in
Figure 2. Accordingly, analogue 3 would be an ideal
molecular probe for mechanistic investigation.
Figure 2. The matrix of endo-PG 1 with natural substrate 2 (upper)
and sulfur analogue 3 (lower) expected by the calculation using
extended COSMOS90. The initial geometries were constructed based
on the X-ray structure of endo-PG 1 with galacturonic acids at subsites
(+1, ꢀ1) and (ꢀ1, ꢀ2, ꢀ3, ꢀ4) obtained by the soaking experiments.
The thiol part was then synthesized. The a-thioacetyl
group was introduced stereoselectively at the anomeric
position via the double inversion of a-galactosyl bro-
mide 8. Treatment of 8 with Bu4NCl in DMPU at room
temperature provided b-chloride 9,13 which was further
treated with potassium thioacetate at 50 °C using a
one-pot procedure to give a-thioacetate 10 stereoselec-
tively in 73% overall yield. Prior to manipulating
2,3,4,6-O-functions, the 1-S-acetyl group was replaced
with a newly developed protective group, 3-cyano-3-
tert-butyldimethylsilylpropyl (abbreviated to CSP
group) thioether. The 1-S-acetyl group of 10 was re-
moved by basic methanolysis at ꢀ25 °C to give thiol
11 in 89% yield. Treatment of 11 with acrolein
(1.2 equiv) in DMF resulted in the Michael addition of
thiol group. Following aqueous work-up, the aldehyde
function in the resulting Michael adduct 12 was further
converted into O-silylated cyanohydrin by TBDMSCN
in the presence of KCN in CH3CN to provide 13 in
These computer modeling experiments motivated us to
conduct the synthesis. Our preliminary experiments sug-
gested that the glycosylation of thiols using glycosyl tri-
chloroacetimidates provides the desired glycoside only
in low yields despite some successful reports.10 Thus,
we focused on introducing thioglycoside linkage by cou-
pling 1-thioderivative 14 with glucopyranose derivative
7, carrying a trifluoro-methanesulfonate leaving group
at the C4 position.
Triflate 7 was first prepared as shown in Scheme 1.
After acetylation of methyl 4,6-O-(4-methoxy)benzyli-
dene glucopyranoside (4),11 the benzylidene moiety
was removed by acetic acid to provide diol 5 in 90%
yield in two steps. Selective oxidation of the primary
alcohol of 5 was achieved by employing TEMPO
and PhI(OAc)2 under aqueous conditions.12 The fol-
lowing treatment with diazomethane gave methyl ester
6 in 91% in two steps. The C4-OH was then converted
into the sulfonate with Tf2O/pyridine, giving 7 in 89%
yield.
1
73% yield in two steps. The H NMR spectrum indicat-
ed that 13 is a 1:1 mixture of diastereomers regarding

4934
K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4932–4935
MP
OH
COOMe
O
COOMe
O
MP
O
a, b
O
c
O
d
HO
AcO
O
O
HO
AcO
TfO
AcO
HO
AcO
OH
O
HO
O
AcO
HO
AcO
AcO
OMe
O
OMe
OMe
OMe
AcO
4
5
6
7
AcO
AcO
m
l
MP
S
AcO
COOMe
O
S
AcO
AcO
OAc
O
COOMe
O
O
O
AcO
AcO
AcO
AcO
OAc
O
OAc
O
AcO
e,f
g,h
i-k
AcO
O
OMe
AcO
AcO
AcO
NC
OMe
MPMO
MPMO
R2
16
X
S
15
AcO
AcO
AcO
S
R1
OTr
O
SH
Y
8: R1= Br, R2 = H
9: R1= H, R2 = Cl
10: R1= SAc, R2 = H
12:X=Y=O
13:X=CN, Y=TBDMSO
11
OTBDMS
14
MPMO
MPMO
HO
O
CCl3
MPMO
MPMO
OTr
O
COOMe
O
COOMe
O
18
NH
MPMO
HO
HO
COOMe
MPMO
O
HO
MPMO
AcO
O
O
O
COOMe
O
COOMe
COOMe
O
AcO
O
r
s
p, q
AcO
AcO
o
AcO
AcO
n
S
COOMe
O
AcO
AcO
3
AcO
S
S
S
COOMe
O
COOMe
O
COOMe
O
AcO
AcO
AcO
AcO
OMe
AcO
AcO
17
MP=p-methoxybenzylidene
AcO
OMe
OMe
OMe
19
20
21
Scheme 1. Reagents and conditions; (a) Ac2O, Py (91%); (b) AcOH, H2O, 65 °C (99%); (c) TEMPO, PhI(OAc)2, acetone, H2O, then CH2N2, Et2O
(91%); (d) Tf2O, Py, CH2Cl2 (83%); (e) Bu4NCl, DMPU r.t., then KSAc 50 °C (73%); (f) NaOMe, MeOH, ꢀ25 °C (89%); (g) acrolein, DMF; (h)
TBDMSCN, KCN CH3CN (73%, two steps); (i) NaOMe, MeOH; (j) p-CH3O-PhCH(OCH3)2, CSA, DMF (69%, two steps); (k) Ac2O, Py (93%); (l)
TBAF, MS4A, THF (94%); (m) AcOH, H2O, 65 °C (86%); (n) TEMPO, PhI(OAc)2, CH2Cl2, H2O, then CH2N2, Et2O (53%); (o) 18, TESOTf,
CH2Cl2, ꢀ78 °C (73%); (p) AcOH, H2O, 60 °C (80%); (q) TEMPO, PhI(OAc)2, CH2Cl2, H2O, then CH2N2, Et2O (97%); (r) DDQ, CH2Cl2, H2O
(79%); (s) NaOH, THF, H2O (99%).
the cyanohydrin moiety. This protective function was
found to be stable during the following treatments.
After all acetyl groups in 13 were removed by basic
hydrolysis, the C4- and C6-OH were protected in the
form of p-methoxybenzylidene acetal under the usual
conditions. The remaining C2 and C3 alcohols were
acetylated again to provide 14 in good overall yield.
of pure 23b with TBAF/MS4A gave 15 in only a 47%
yield. Notably, the coupling reaction with pure 23b
proceeded more smoothly to give 15 in 81% yield by
employing sodium hydride as the base.
Glycoside 15 was then converted into glycosyl acceptor
17 by acidic cleavage of the 4-methoxybenzylidene
acetal (!16, 86% yield), oxidation of the primary
alcohol with TEMPO/PhI(OAc)2 followed by treatment
with CH2N2 (!17, 53% yield).
Treatment of 14 with TBAF in THF in the presence of
molecular sieves 4A and the following addition of tri-
flate 7 after 5 min using a one-pot procedure resulted
in a coupling reaction to provide thioglycoside 15
in 94% yield. The fluoride ion under aprotic conditions
induced the retro-Michael reaction of regenerated
aldehyde 22, giving 23a in a form of highly reactive
tetrabutylammonium mercaptide. This method was
more efficient than the following stepwise transforma-
tions. Aqueous work-up prior to the addition of 7 gave
thiol 23b in 52% yield (Scheme 2). The basic treatment
We failed in the glycosylation of 17 when galacturo-
nate 24 (Scheme 3)12,14 was employed as the donor
although the glucuronate derivative (the C4 equatorial
isomer) gave the adduct under the same conditions.
Carbocation 25 produced from 24 was likely deactivat-
ed by the C6-carbonyl group after flipping of the pyra-
nose ring. Thus, we decided to construct the carboxyl
group at the non-reducing end after glycosylation. As
expected, glycosylation of acceptor 17 with trityl ether
protected imidate 18 took place to produce trimer 19
stereoselectively in 73% yield by employing TESOTf
as the activator.15 Stereochemistry of 19 of the newly
furnished glycoside bond was confirmed by observing
a small coupling constant between C100H and C200H
1
(J = 1.5 Hz) in H NMR.
The trityl ether in 19 was selectively cleaved by aqueous
acetic acid at 60 °C in 80% yield. Oxidation of the result-
ing alcohol with TEMPO/PhI(OAc)2 was also effective
and was followed by treatment with diazomethane to
provide methyl ester 20 in 97% yield in three steps.
Scheme 2.

K. Yamamoto et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4932–4935
4935
3. Gidenne, T. Livest. Prod. Sci. 2003, 81, 105.
OMe
MPMO
MPM
4. Matsuda, H.; Oohara, K.; Morii, Y.; Hashimoto, M.;
Miyairi, K.; Okuno, T. Bioorg. Med. Chem. Lett. 2003, 13,
1063.
5. Shimizu, T.; Miyairi, K.; Okuno, T. Eur. J. Biochem. 2000,
267, 2380.
COOMe
O
δ
MPMO
MPMO
O
O
COOMe
O
flipping
TESOTf
MPMO
MPMO
MPMO
O
δ
O
NH
MPMO
MPMO
CCl3
24
25
26
6. Shimizu, T.; Nakatsu, T.; Miyairi, K.; Okuno, T.; Kato,
H. Acta Crystallogr. D 2001, 57, 1171.
Scheme 3.
7. Cao, H.; Yu, B. Tetrahedron Lett. 2005, 46, 4337.
8. (a) We used molecular dynamic (MD) simulation
program COSMOS90 developed by one of the authors
of this paper (M. Saito). To perform MD simulations of
endo-PG 1 with galacturonic acids, we extended COS-
MOS90 to make it possible to simulate sugar moieties
based on the glycam force field. However, glycam did
not contain the parameters for sulfur-substituted ana-
logues. They were prepared by performing ab initio
molecular orbital calculations (HF/6-31G*), employing
a-1,4-linked digalacturonic acid. MD simulation of
endo-PG 1 with the sulfur analogue was performed by
slowly changing the force field of the carbohydrate
analogue to the sulfur analogue for 40 ps and then
maintaining the force field for 140 ps. The details will be
reported elsewhere shortly by Saito et al. COSMOS90:
Saito, M. J. Chem. Phys. 1994, 101, 4055; Glycam:
(b) Kirschner, K. N.; Woods, R. J. Proc. Natl. Acad.
Sci. U.S.A. 2001, 98, 10541.
9. The detail will be reported by Kato and Shimizu shortly.
10. Driguez, H. Top. Curr. Chem. 1997, 86.
11. Evans, M. E. Carbohydr. Res. 1972, 21, 473.
12. Van den Bos, L. J.; Codee, J. D. C.; Van der Toorn, J. C.;
Boltje, T. J.; Van Boom, J. H.; Overkleeft, H. S.; Van der
Marel, G. A. Org. Lett. 2004, 6, 2165.
13. Blanc-Muasser, M.; Vigne, L.; Driguez, H. Tetrahedron
Lett. 1990, 31, 3869.
14. Magaud, D.; Grandjean, C.; Doutheau, A.; Anker, D.;
Shevchik, V.; Cotte-Pattatb, N.; Robert-Baudouy, J.
Tetrahedron Lett. 1997, 38, 241.
Finally, all protective groups of 20 were removed. The
MPM ethers were cleaved by DDQ oxidation without
affecting the sulfide function4 to give triol 21 in 88%
yield. The following basic treatment hydrolyzed all ester
groups. Ion exchange column chromatography (Dowex
50W, H+ form) after the reaction provided a pure sam-
ple of 3 as a white powder.16 Preliminary enzymatic
studies revealed that 3 inhibited hydrolysis of oligo-
galacturonic acid by endo-PG 1 and was stable under
these conditions. The experiments employing surface
plasmon resonance17 revealed the KD value of 3 to be
0.2 lmol/L.18
As described, we designed and synthesized a sulfur ana-
logue of trigalacturonic acid 3 as a stable mimic of the
natural substrate. Computer modeling calculations sug-
gested that 3 would be a stable analogue that can bind
with endo-PG 1 in the same manner as the natural sub-
strate. In fact, 3 was not only stable in the presence of
endo-PG 1 but also made a stable complex with the tar-
get enzyme. In these studies, we also demonstrated that
the CSP group is a useful protective group that can be
readily activated by TBAF under aprotic conditions.
These findings resulted in the successful preparation of
3 in sufficient quantity for the enzymatic studies
(>100 mg). Detailed structural and thermodynamic
studies on the complex using X-ray crystallographic
analysis, calorimetric experiments, and longer-time
MD simulations are under investigation in our
laboratories.
15. Schmidt, R. R.; Jung, K.-H. In Preparative Carbohydrate
Chemistry; Hanessian, S., Ed.; Marcel Dekker: New York,
1997, pp 283–312.
25
1
16. ½aꢁD ¼ þ133ꢂ (c 0.50, H2O), H NMR (400 MHz, 6.7 mg/
1.0 mL of D2O) d 3.28 (3H, s, OCH3), 3.44 (1H, dd,
J = 3.9, 10.4 Hz, C2H), 3.57 (1H, dd, J = 1.8, 4.6 Hz,
C4H), 3.61 (1H, dd, J = 3.9, 10.5 Hz, C20H), 3.78 (1H, dd,
J = 5.4, 11.4 Hz, C300H), 3.80 (1H, dd, J = 4.6, 10.4 Hz,
C30H), 3.98 (1H, dd, J = 5.4, 10.2 Hz, C200H), 4.02 (1H,
dd, J = 4.4, 10.2 Hz, C3H), 4.20 (1H, dd, J = 1.4, 3.0 Hz,
C40H), 4.30 (1H, d, J = 3.0 Hz, C400H), 4.75 (2H, d,
J = 3.9 Hz, C1H, C5H), 4.929 (1H, s, C50H), 4.933 (1H, d,
J = 3.9 Hz, C10H), 5.05 (1H, s, C500H), 5.39 (1H, d,
J = 5.4 Hz, C100H), 13C NMR (100 MHz, D2O) d 50.34,
55.51, 67.32, 67.68, 67.71, 68.60, 68.74, 69.26, 69.84, 69.86,
70.11, 70.84, 77.73, 87.29, 99.43, 99.94, 171.67, 172.18,
172.68 ppm. ESIMS (positive) m/z = 599 (M+Na)+, (neg-
ative), m/z = 575 (MꢀH)ꢀ.
Acknowledgments
Part of these studies was financially supported by Hiro-
saki University as Priority Studies. We thank President
Masahiko Endo of Hirosaki University for his fruitful
suggestions throughout these studies.
References and notes
1. Jensen, C. D.; Haskell, W.; Whittam, J. H. Am. J. Cardiol.
1997, 79, 34.
17. Biacore International AB, Rapsgatan 7, 754 50 Uppsala,
SE-754 50, Sweden, +46 18 67 57 00.
2. Zsivanovits, G.; MacDougall, A. J.; Smith, A. C.; Ring,
S. G. Carbohydr. Res. 2004, 339, 1317.
18. Shinohara, Y.; Kim, F.; Shimizu, M.; Goto, M.; Tosu, M.;
Hasegawa, Y. Euro. J. Biochem. 1994, 223, 189.