Stereochemical differentiation in the Simmons-Smith reaction for cyclopropanated glucopyranose derivatives as molecular probes for glycosidases

Article abstract of DOI:10.1271/bbb.110120

Glucopyranose analogues carrying a bicyclo[4.1.0]- heptane framework (4) and the diastereomer of the cyclopropane moiety were synthesized as the unit for molecular probes to mimic the unstable transition state conformation of the glucopyranose ring in enz

Full text of DOI:10.1271/bbb.110120

Biosci. Biotechnol. Biochem., 75 (7), 1380–1382, 2011
Note
Stereochemical Differentiation in the Simmons-Smith Reaction
for Cyclopropanated Glucopyranose Derivatives
as Molecular Probes for Glycosidases
y
Nanako AKIYAMA, Shogo NOGUCHI, and Masaru HASHIMOTO
Faculty of Agriculture and Life Science, Hirosaki University, 3-Bunkyo-cho, Hirosaki 036-8561, Japan
Received February 10, 2011; Accepted April 2, 2011; Online Publication, July 7, 2011
[doi:10.1271/bbb.110120]
Glucopyranose analogues carrying a bicyclo[4.1.0]-
heptane framework (4) and the diastereomer of the
cyclopropane moiety were synthesized as the unit for
molecular probes to mimic the unstable transition state
conformation of the glucopyranose ring in enzymatic
hydrolysis. The synthesis features differentiation of the
ꢀ- and ꢁ-stereoselectivity in cyclopropanation of the
corresponding cyclohexene derivative (5).
ever, 3 was susceptible to the nucleophiles in the
enzyme to result in a stable chair conformation by
opening the epoxide, for example giving the structure
recorded in Protein Data Bank ID 2JAL.10) It would be
difficult to reproduce the transition state with 3. We
expected that replacing the epoxide moiety with a
cyclopropane group would provide stability against the
nucleophiles in glycosidases without changing the
conformation of the pyranose ring, which enables a
stable ligand-enzyme complex to be provided in the
desired form.
Key words: cyclopropane; glucopyranose; Simmons-
Smith reaction; stereoselectivity; transition
state analogue
We planned to introduce the cyclopropane ring by
Simmons-Smith cyclopropanation of corresponding cy-
clohexene 5 which was readily prepared from xylose by
following Kornienko’s procedure.11) The results are
summarized in Table 1. Since m-chloroperbenzoic acid
(mCPBA) oxidation was known to provide ꢀ-epoxide
stereoselectively (run 10),12) we expected that cyclo-
propanation would also proceed in the same stereo-
selective manner. However, the major product was ꢁ-
cyclopropanated 6ꢀ under the conventional conditions
(run 1). The stereochemistries of 6ꢀ and 6ꢁ were
determined by nuclear Overhauser effect (NOE) experi-
ments. Irradiation of the endo-proton in the cycropro-
pane ring of 6ꢁ resulted in signal enhancements at C2H
and C4H signals as shown in Fig. 2, while the similar
irradiation in 6ꢀ induced NOEs on the signals due to
C3H and C5H resonances. We then explored the
conditions which would reverse this selectivity. The
addition of 1,2-dimethoxyethane (DME) slightly in-
Understanding the detailed reaction mechanism for
cellulases is indispensable in designing its artificial
mutants with both higher activity and higher stability
than those of natural cellulases. Many oligo-cellulose
analogues with tolerance against cellulases have been
developed as inhibitors and provided stable complexes
with cellulases.1–3) However, it is difficult to insert the
substrate mimics in the reaction site (the ꢀ1 subsite) in a
desirable form, because the enzymes force the pyranose
ring at that site to become distorted.4)
The MolFeat program5) for molecular viewing has
given unbelievable 7-thio-1-oxa-bicyclo[4.1.0]heptane
structure 2 instead of the pyranose unit in 1 at the
expected reaction site in ꢀ-glucanase Cel6A from
Thermobifida fusca (Protein Data Bank ID 2BOD)6) as
shown in Fig. 1. This impossible bonding was caused by
the abnormally short distance between the pyranose
oxygen and the substituted glycosydyl sulfur atom,
because the program has a function to automatically
make chemical bonds in cases of the two atoms being
closer than a set distance. This phenomenon has also
been observed with RasMol,7) a popular public domain
molecular viewing application. This observation sug-
gested that the enzyme forced the pyranose ring to take a
strained half-chair conformation like 2. In other words,
the enzyme prefers that conformation.
.
creased the ꢀ-selectivity (run 2), while BF3 OEt2
accelerated the reaction to result in a higher yield, but
gave the undesired ꢁ-isomer exclusively (run 3).
Protecting the C6OH group as a tert-butyldimethylsilyl
(TBDMS) ether did not allow the reaction to proceed at
all. The chiral ligands reported by Takahashi13) were not
effective in this particular case. We experienced slightly
higher ꢀ-selectivity by using old DME, suggesting the
positive effect of moisture or some degradation products
from DME. Although H2O was not helpful, the addition
of a small amount of alcohol increased the ꢀ-selectivity
(runs 4–8). The amount of alcohol added affected the
yields greatly, but had less effect on the selectivity. We
achieved the best result by employing 0.4 eq. of MeOH
and DME (10 eq.) to so far provide the cyclopropanes in
54% yield with ꢀ:ꢁ ¼ 84:16 selectivity (run 6). Meth-
On the other hand, Umezawa et al. have isolated
cyclophellitol (3), a 7-oxa-bicyclo[4.1.0]heptane ana-
logue, from Phellinus sp. fungus.8) This molecule
potently inhibits ꢀ-glycosidases, probably due to repro-
duction of the half-chaired transition state structure.
Furthermore, such analogues as its diastereomers and its
aziridine derivatives have also been developed.9) How-
y
To whom correspondence should be addressed. Tel/Fax: +81-172-39-3782; E-mail: hmasaru@cc.hirosaki-u.ac.jp
Abbreviations: mCPBA, m-chloroperbenzoic acid; DME, 1,2-dimethoxyethane; NOE, nuclear Overhauser effect; TBDMS, tert-butyldimethylsilyl

Stereochemical Differentiation in the Simmons-Smith Reaction
1381
OH
OH
-glucanase
β
R
HO
HO
O
Y
O
R1
R2
X
X
Y
R
CH2OH
CH2OH
1
2
2
: X = O, Y = S, R = R = Glc
(in PDB:2BOD)
1: X = O, Y = S, R = Glc
3: X = CH, Y = O, R1 = lone pair electrons,
R2 = H (cyclophellitol)
1
2
4
: X = Y = CH, R = R =H
Fig. 1. Design of Cyclopropane Analogue 4.
Table 1. Cyclopropanation of Cyclohexene 5
X
X
6
Et2Zn (10 eq)
HO
HO
BnO
HO
BnO
CH2I2 (20 eq)
+
CH2Cl2
r.t
BnO
OBn
OBn
OBn
OBn
OBn
β-isomer
6:X = CH2, 7:X = O
OBn
α-isomer
5
Run
Additives
Time (h)
Products
Yield (%)
ꢁ : ꢀ
Recovery of 5 (%)
1
2
none
DME (10 eq.)
2
24
2
6
6
6
6
6
6
6
6
6
7
50
65
97
40
54
54
41
35
10
85
12:88
46:53
1:>99
90:10
84:16
84:16
72:28
71:29
40:60
92:8
8
0
3
DME (10 eq.), BF3OEt2 (5 eq.)
DME (10 eq.), MeOH (0.2 eq.)
DME (10 eq.), MeOH (0.3 eq.)
DME (10 eq.), MeOH (0.4 eq.)
DME (10 eq.), n-BuOH (0.2 eq.)
DME (10 eq.), n-BuOH (0.1 eq.)
0
4
6
59
36
33
41
31
0
5
3
6
5
7
3
8
5
9
1012Þ
DME (10 eq.), CH3SO2NH2 (0.1 eq.)
mCPBA epoxidation in CH2Cl2
24
3
—
HO
HO
H
Hendo
H
H
Hexo
5
H
BnO
BnO
H
4
H
BnO
BnO
2
3
H
Hexo
H
H
OBn
BnO
Hendo
H
6β
6α
Fig. 2. Stereochemistry of 6ꢀ and 6ꢁ.
anesulfonamide improved neither the selectivity nor
yield (run 9). Although we succeeded in reversing the
stereoselectivity as described, the role of the alcohols
added has remained unclear.
The benzyl groups in 6ꢁ were removed by hydro-
genolysis with 10% Pd(OH)2/C in MeOH to success-
fully provide 4 without destroying the cyclopropane
ring. Since its ꢁ-isomer was expected as an inhibitor of
the ꢁ-glucosidases, 6ꢀ was also deprotected in the same
manner to give the corresponding stereoisomer. The
1H-NMR spectra of these samples indicated sufficient
purity for further biological investigations.
Since we are focusing on ꢀ-glucanases, we are now
studying the incorporation of this unit into the middle of
the oligo-cellulose chain as the next stage. An enzymatic
assay of 4 and its isomer against exo-typed ꢀ-glyco-
sidases is also underway in our laboratory.
Typical procedure (run 6). Diethylzinc (1.05 mol/L in
hexane, 940 mL, 987 mmol) was added to a stirred
mixture of MeOH (1.6 mL, 39 mmol), DME (100 mL,
957 mmol), and CH2Cl2 (500 mL) at room temperature.
Diiodomethane (163 mL, 1.97 mmol) was added after
5 min, and the mixture was further stirred at room
temperature for an additional 5 min. A solution of
cyclohexene 5 (42.4 mg, 98.4 mmol) in CH2Cl2 (500 mL)
was added at room temperature. After stirring for 5 h, a
saturated NH4Cl solution was added to decompose the
excess reagents. The mixture was then poured into H2O
(25 mL) and extracted with EtOAc (10 mL ꢁ 3). The
combined extracts were washed with brine, dried over
MgSO4, and concentrated under reduced pressure. The
residue was purified by silica gel column chromatog-
raphy (EtOAc:hexane ¼ 50:50) to give 6ꢀ and 6ꢁ
(23.5 mg, 54%) as a 16:84 mixture, together with

1382
N. AKIYAMA et al.
recovered 5 (14.0 mg, 33%). Pure 6ꢀ and 6ꢁ were
obtained by preparative TLC (Et2O:hexane ¼ 50:50).
m=z: 175.0898 (calcd. for C8H15O4 ½M þ Hꢂþ:
175.0970).
Physical data for 6ꢁ. ½ꢁꢂD þ67ꢃ (c 0.64, CHCl3);
The ꢁ-isomer was also obtained in a similar manner.
20
IR (film) cmꢀ1: 3395, 2920, 1025, 725; 1H-NMR ꢂ
(500 MHz, CDCl3): 0.17 (1H, dd, J ¼ 5:2, 11.0 Hz),
0.64 (1H, dt, J ¼ 5:2, 9.3 Hz), 1.14 (1H, ddd, J ¼ 5:8,
7.5, 9.3 Hz), 1.18 (1H, ddd, J ¼ 5:8, 9.6, 11.0 Hz), 2.31
(1H, dddd, J ¼ 4:4, 5.6, 9.6, 10.1 Hz), 3.16 (1H, t,
J ¼ 10:1 Hz), 3.59 (1H, dd, J ¼ 7:5, 10.1 Hz), 3.72 (1H,
dd, J ¼ 4:4, 10.7 Hz), 3.75 (1H, t, J ¼ 7:5 Hz), 3.79
(1H, dd, J ¼ 5:6, 10.7 Hz), 4.52 (1H, d, J ¼ 11:0 Hz),
4.66 (1H, d, J ¼ 11:0 Hz), 4.78 (1H, d, J ¼ 11:0 Hz),
4.79 (1H, d, J ¼ 11:0 Hz), 4.88 (1H, d, J ¼ 11:0 Hz),
4.92 (1H, d, J ¼ 11:0 Hz), 7.2–7.4 (15H, m); 13C-NMR
ꢂ (125 MHz, CDCl3): 6.16, 13.54, 14.61, 41.95, 65.85,
71.93, 74.97, 76.78, 78.62, 82.50, 86.35, 127.87, 127.91,
128.17, 128.37, 128.46, 128.55, 138.45; FAB-HRMS
m=z: 445.2373 (calcd. for C29H33O4 ½M þ Hꢂþ:
445.2379).
½ꢁꢂD 28ꢃ (c 0.22, H2O); H-NMR ꢂ (500 MHz, D2O):
0.48 (1H, dd, J ¼ 5:2, 10.8 Hz), 0.95 (1H, dt, J ¼ 5:2,
8.7 Hz), 1.18 (1H, ddd, J ¼ 5:2, 8.7, 10.8 Hz), 1.45 (1H,
dt, J ¼ 5:7, 8.7 Hz), 1.91 (1H, dddd, J ¼ 3:5, 6.4, 10.1,
10.8 Hz), 3.28 (1H, dd, J ¼ 8:7, 10.1 Hz), 3.38 (1H, t,
J ¼ 10:1 Hz), 3.88 (1H, dd, J ¼ 6:4, 10.8 Hz), 4.02 (1H,
dd, J ¼ 3:5, 10.8 Hz), 4.18 (1H, dd, J ¼ 5:7, 8.7 Hz);
13C-NMR ꢂ (125 MHz, D2O): 9.12, 12.60, 17.60, 45.05,
63.20, 71.20, 72.13, 74.42; FAB-HRMS m=z: 175.0908
(calcd. for C8H15O4 ½M þ Hꢂþ: 175.0970).
21
1
Acknowledgment
We are grateful to Professor Hideyo Takahashi of
Teikyo University for fruitful discussions on the cyclo-
propanation reaction and also for kindly presenting the
chiral ligands.
21
Physical data for 6ꢀ. ½ꢁꢂD þ79:9ꢃ (c 1.37, CHCl3);
IR (film) cmꢀ1: 3425, 2890, 1055, 735; 1H-NMR ꢂ
(500 MHz, CDCl3): 0.40 (1H, dd, J ¼ 5:0, 10.6 Hz),
0.79 (1H, dt, J ¼ 5:0, 8.7 Hz), 0.88 (1H, dddd, J ¼ 3:0,
5.3, 8.7, 10.6 Hz), 1.34 (1H, ddd, J ¼ 5:3, 6.3, 8.7 Hz),
1.78 (1H, br), 3.32 (2H, m), 3.71 (2H, m), 4.11 (1H, m),
4.57 (1H, d, J ¼ 11:0 Hz), 4.66 (1H, d, J ¼ 11:7 Hz),
4.78 (1H, d, J ¼ 11:0 Hz), 4.79 (1H, d, J ¼ 11:7 Hz),
4.88 (1H, d, J ¼ 11:0 Hz), 4.91 (1H, d, J ¼ 11:0 Hz),
7.2–7.4 (15H, m); 13C-NMR ꢂ (125 MHz, CDCl3):
11.13, 14.48, 16.41, 46.07, 65.91, 71.98, 75.72, 76.23,
81.10, 81.14, 85.28, 128.38, 128.72, 128.84, 129.15,
129.19, 129.21, 129.41, 139.17, 139.76, 139.80;
FAB-HRMS m=z: 467.2192 (calcd. for C29H32O4Na
½M þ Naꢂþ: 467.2198).
References
1) Moreau V, Viladot JL, Samain E, Planas A, and Driguez H,
Bioorg. Med. Chem., 4, 1849–1855 (1996).
2) Parsiegla G, Reverbel C, Tardif C, Driguez H, and Haser R,
J. Mol. Biol., 375, 499–510 (2008).
˚
3) Ubhayasekera W, Mun˜oz IG, Vasella A, Stahlberg J, and
Mowbray SL, FEBS J., 272, 1952–1964 (2005).
4) Liu Y, Yoshida M, Kurakata Y, Miyazaki T, Igarashi K,
Samejima M, Fukuda K, Nishikawa A, and Tonozuka T, FEBS
J., 277, 1532–1542 (2010).
5) MolFeat Version 4.5.2.17 by FiatLux, http://www.fiatlux.co.jp/
product/lifescience/molfeat/mol-index.html
6) Larsson AM, Bergfors T, Dultz E, Irwin DC, Roos A, Driguez
H, Wilson DB, and Jones TA, Biochemistry, 44, 12915–12922
(2005).
Deprotection giving 4. A solution of 6ꢁ (20.4 mg,
47.2 mmol) in MeOH (1.0 mL) was vigorously stirred for
12 h with 10% Pd(OH)2/C (10 mg) in H2 atmosphere
(1 atm). The suspension was filtered through a pad of
Celite and rinsed with MeOH. The filtrate was concen-
trated under reduced pressure to give 4 (6.4 mg,
36.8 mmol, 78%) as a white amorphous compound.
7) RasMol version 2.7.5. This program is available from the URL:
http://rasmol.org/
8) Atsumi S, Umezawa K, Iinuma H, Naganawa H, Nakamura H,
Iitaka Y, and Takeuchi T, J. Antibiot., 43, 49–53 (1990).
9) Tatsuta K, Niwata Y, Umezawa K, Toshima K, and Nakata M,
J. Antibiot., 44, 912–914 (1991).
22
1
½ꢁꢂD 47ꢃ (c 0.32, H2O); H-NMR ꢂ (500 MHz, D2O):
0.31 (1H, dd, J ¼ 5:3, 11.4 Hz), 0.86 (1H, dt, J ¼ 5:3,
9.1 Hz), 1.14 (1H, dt, J ¼ 5:3, 9.1 Hz), 1.47 (1H, dt,
J ¼ 5:3, 9.1 Hz), 2.30 (1H, dddd, J ¼ 4:4, 5.3, 8.3,
11.4 Hz), 3.14 (1H, t, J ¼ 10:2 Hz), 3.43 (1H, dd,
J ¼ 8:3, 10.2 Hz), 3.78 (1H, dd, J ¼ 4:4, 8.3 Hz), 3.79
(1H, dd, J ¼ 5:3, 8.3 Hz), 4.08 (1H, dd, J ¼ 4:4,
10.2 Hz); 13C-NMR ꢂ (125 MHz, D2O): 5.97, 13.95,
16.03, 42.16, 63.46, 68.41, 73.88, 77.99; FAB-HRMS
10) Gloster TM, Madsen R, and Davies GJ, Org. Biomol. Chem., 5,
444–446 (2007).
11) Kireev AS, Breithaupt AT, Collins W, Nadein ON, and
Kornienko A, J. Org. Chem., 70, 742 (2005).
12) Trost BM and Hembre EJ, Tetrahedon Lett., 40, 219–222
(1999).
13) Takahashi H, Yoshioka M, Ohno M, and Kobayashi S,
Tetrahedron Lett., 33, 2575–2578 (1992).