Probing the conformational changes in the enzymatic hydrolysis of 2-acetamido-2-deoxy-β-D-glucopyranosides

Article abstract of DOI:10.1002/hlca.200490229

The isoquinuclidines 7 and 8 were synthesised and tested as inhibitors of hexosaminidases from jack beans and from bovine kidney. These isoquinuclidines mimick the ^1,4β-conformer of a N-acetyl-glucosamine-derived β-D-glucopyranoside; they are c

Full text of DOI:10.1002/hlca.200490229

2566
Helvetica Chimica Acta Vol. 87 (2004)
Probing the Conformational Changes in the Enzymatic Hydrolysis of
2-Acetamido-2-deoxy-b-d-glucopyranosides
by Matthias Bˆhm and Andrea Vasella*
Laboratorium f¸r Organische Chemie, ETH-Hˆnggerberg, Wolfgang-Pauli Strasse 10, CH-8093 Z¸rich
Dedicated to Prof. K. K. Balasubramaniam (K.K.B.) on the occasion of his 65th birthday
The isoquinuclidines 7 and 8 were synthesised and tested as inhibitors of hexosaminidases from jack beans
and from bovine kidney. These isoquinuclidines mimick the 1,4B-conformer of a N-acetyl-glucosamine-derived
b-d-glucopyranoside; they are competitive inhibitors with Ki values from 0.014 to 0.30 mm. The strong inhibition
of these enzymes agrees with the hypothesis that the enzymatic hydrolysis of 2-acetamido-2-deoxy-b-d-
glucopyranosides proceeds via a boat-like conformer with a pseudo-axial scissile glycosidic bond and a
pseudo-axial acetamido substituent optimally oriented to effect an intramolecular substitution of the aglycon.
Introduction.
According to the principles of stereoelectronic control, the
enzymatic hydrolysis of b-d-glycopyranosides1) requires coplanar orientation of the
scissile bond and a non-bonding, doubly occupied orbital of the endocyclic O-atom
[3][4]. This implies that the reaction coordinate is characterised by an initial
conformational change of the substrate. Several conformations of the substrate satisfy
the principles of stereoelectronic control. It has been proposed, on the basis of
crystallographic evidence, that all of them are induced by individual glycosidases [1].
1
Thus, a 1,4B- or S3-conformation of the À 1unit has been deduced from the crystal
structure of cellulases belonging to families2) 5 [6] and 7 [7] in complexes with
substrate analogues and, similarly, for hexosaminidases belonging to families 18 [8][9]
and 20 [10][11] in complexes with substrates.
We have synthesised the isoquinuclidines 1 6 (Fig.) and tested them as glycosidase
inhibitors to evidence the conformational changes induced on their substrate by b-
mannosidases and b-glucosidases [12 14]. These isoquinuclidines mimick the 1,4B-
conformer, of a b-d-manno- and b-d-glucopyranoside, and possess a basic N-atom that
can interact with the catalytic acid of the glycosidase. A comparison of the inhibition of
snail b-mannosidase by the d-manno-isoquinuclidines 1 and 2, and of the b-glucosidase
from Caldocellum saccharolyticum (family 1) by the d-gluco-configured isoquinucli-
dines 3 and 4 evidenced that the b-mannosidase induces a conformational change of its
substrate towards a 1,4B-conformer, while the b-glucosidase induces a different
conformational change [14]. A further comparison of the inhibition of these
1
)
)
For recent review articles on the mechanism of glycoside hydrolysis, see [1][2].
Glycosidases have been classified according to their amino acid sequence [5]; enzymes belonging to the
same family should have a similar tertiary structure and hydrolyse their substrates according to a similar
mechanisms. A regularly updated database is available on the internet (http://afmb.cnrs-mrs.fr/CAZY/
index.html).
2
¹ 2004 Verlag Helvetica Chimica Acta AG, Z¸rich

Helvetica Chimica Acta Vol. 87 (2004)
2567
glycosidases by d-gluco- and d-manno-configured tetrahydropyridoimidazoles, i.e.,
inhibitors mimicking the reactive intermediates, suggests that there is no similar
difference between the transition states for glycoside hydrolysis by these enzymes [15].
These observations imply that inhibitors mimicking an early part of the reaction
coordinate have a particularly high chance of being selective.
Figure. The isoquinuclidines 1 and 2 mimicking the 1,4B-conformer of b-d-mannopyranosides, the isoquinucli-
dines 3 6 mimicking b-d-gluco- and 2-deoxy-b-d-arabinohexopyranosides, and the acetamido-isoquinuclidines
7 and 8 mimicking 2-acetamido-2-deoxy-b-d-glucopyranosides
The above-mentioned crystal structures of hexosaminidases of families 18 and 20
suggest that the isoquinuclidine-derived mimicks 7 and 8 of a 1,4B-conformer of a 2-
acetamido-2-deoxy-b-d-hexopyranoside should be good inhibitors of these glycosi-
dases. We report the synthesis and inhibitory activity of 7 and 8.
Synthesis. Treatment of the stable trifluoromethanesulfonate 9 [12 14] with
NaN3 yielded 96% of the azide 10 (Scheme). Reduction of 10 with LiAlH4, followed by
acetylation of the crude amino alcohol 12, gave the acetamido-isoquinuclidine 13
(96%). Acetal cleavage with Me3SiBr yielded 76% of the tertiary alcohol 14 that was
deacetylated to the N-benzylated isoquinuclidine 7 (85%). Hydrogenolysis of the
benzyl derivative 7 followed by ion-exchange chromatography yielded 93% of the N-
unsubstituted isoquinuclidine 8 that was, thus, obtained in six steps and 55% yield from
the trifluoromethanesulfonate 9. To exclude a contamination of 8 by the N-benzylated
precursor 7 [12][13] or by Pd3), we converted the isoquinuclidine 8 to the carbamate
15, which was chromatographed on silica gel and deprotected with HCl in MeOH/H2O
to again provide 8.
The gluco-configuration of the azide 10 was deduced by comparison of its NMR
spectra with those of its manno-configured precursor 9 and with those of the d-gluco-
configured acetate 11 [14]4). The 1H-NMR spectra of the manno-configured
isoquinuclidine 9 and the gluco-configured isoquinuclidines 10 and 11 show character-
istic differences (see Table 2 in the Exper. Part and [12 14]); e.g., the coupling constant
J(1,7)ˆ 1.7 Hz of 9 vs. 4.0 of 10 and 4.4 Hz of 11. All the gluco-configured
isoquinuclidines 10 14, 7, and 8 show long-range W coupling (J(6exo,7) ˆ 1.9 Hz for
10 and 11, and 2.0 2.3 Hz for 13, 14, 7, and 8) that is not visible in the 1H-NMR spectra
of manno-configured isoquinuclidines.
3
)
)
Pd and its complexes inhibit glycosidases; for examples, see [16][17].
The d-gluco-configured acetate 11 was obtained by treatment of 9 with CsOAc. Its structure was
established by X-ray crystal-structure analysis [14].
4

2568
Helvetica Chimica Acta Vol. 87 (2004)
Scheme
a) NaN3, DMF; 96%. b) LiAlH4, THF. c) 4-(Dimethylamino)pyridine (DMAP), pyridine/Ac2O; 96% from 10.
d) Me3SiBr, mol. sieves 3-ä, CH2Cl2; 76%. e) NH3, MeOH; 85%. f) Pd(OH)2/C, H2, MeOH/H2O/HCl; 93%.
g) 1 . (t-BuO)2C(O), MeOH/Et3N. 2. pyridine/Ac2O; 75%. h) HCl, MeOH/H2O; 90%.
Inhibition Studies.
The acetamido-isoquinuclidines 7 and 8, the d-gluco-
isoquinuclidine 4, and the 2-deoxy-d-arabino-isoquinuclidine 6 are competitive
inhibitors of the N-acetyl-b-hexosaminidases from jack beans and from bovine kidney
(Table 1)5). The same inhibition constants were obtained for samples of the N-
unsubstituted isoquinuclidine 8 prepared by hydrogenolysis of the N-benzylated
isoquinuclidine 7 and for samples prepared by deacylation of the carbamate 15. A
comparison of the inhibition of these enzymes by the isoquinuclidines 7 and 8 that takes
both their pKHA values and the pH of the enzyme assays into account evidences that the
glycosidases are inhibited by the free amines 7 and 8 rather than by the corresponding
‡
‡
ammonium salts 1 ¥ H and 2 ¥ H .
Table 1. Inhibition Constants Ki [mM] for the Inhibition of the N-Acetyl-b-hexosaminidases from Jack Beans and
from Bovine Kidney by the Isoquinuclidines 4, 6, 7, and 8
Inhibitor
KHA
N-Acetyl-b-hexosaminidase
N-Acetyl-b-hexosaminidase
from jack beans (pH 5.0)
from bovine kidney (pH 4.1)
4
6
7
8
7.8
11
25
0.0810.30
0.014
210
31
0
5.9
7.7
0.067
5
)
The N-acetyl-b-hexosaminidase from jack beans belongs to the glycosyl hydrolase family 18 [18]. Its 3D
structure was determined by X-ray crystal-structure analysis [19]. The amino acid sequence and the 3D
structure of the N-acetyl-b-hexosaminidase from bovine kidney are not known.

Helvetica Chimica Acta Vol. 87 (2004)
2569
The stronger inhibition of the hexosaminidases by the N-unsubstituted isoquinu-
clidine 8 than by the N-benzylated isoquinuclidine 7 evidences that the inhibition is due
to specific interactions between these enzymes and the isoquinuclidines and not to
unspecific hydrophobic interactions (with the N-Bn substituent)6). The relative
strength of the inhibition of these N-acetyl-b-hexosaminidases by the acetamido-
isoquinuclines 7 and 8, the gluco-configured isoquinuclidine 4, and the arabino-
configured isoquinuclidine 6 confirms the strong interactions of these enzymes with the
acetamido substituent7).
The inhibition of the N-acetyl-b-hexosaminidases from jack beans and from bovine
kidney by the isoquinuclidines 4 and 6 8 evidences that the enzymatic hydrolysis of 2-
acetamido-2-deoxy-b-d-glucopyranosides proceeds via a 1,4B- (or closely related)
conformer with a pseudo-axial scissile glycosidic bond, and also with a pseudo-axial
C(2)-acetamido subsitutent. The pseudo-axial orientation of the AcNH group is
optimal for a nucleophilic attack at the anomeric center, displacing the aglycon, as
suggested on the basis of crystal structures [8][10].
The agreement between the inhibition of the two hexosaminidases by 7 and 8 with
the above-mentioned crystal-structure analyses lends additional weight to the
significance of the strong difference between the inhibition of snail b-mannosidase
by the manno-configured isoquinuclidines 1 and 2, and the inhibition of the b-
glucosidase from C. saccharolyticum by the gluco-configured isoquinuclidines 3 and 4.
While snail b-mannosidase and the hexosaminidases of families 18 and 20 appear to
distort their substrates towards a 1,4B (or related)-conformer, the b-glucosidase from C.
saccharolyticum (and presumably all glucosidases of the same family of glycohydro-
lases) choose a different conformation to comply with the principle of stereoelectronic
control.
We thank D. Manser and M. Schneider for the pKHA determinations, Dr. B. Bernet for checking the Exper.
Part, and the Swiss National Science Foundation and Oxford Glycosciences Ltd., Abingdon (UK), for generous
support.
Experimental Part
General. Solvents were distilled before use. If not specified otherwise, all reactions were carried under a N2
atmosphere. −Normal workup× implies pouring the reaction mixture into the indicated sat. aq. soln., extracting
into the mentioned org. solvent, if necessary washing with the indicated sat. aq. soln., drying of the org. layer
(Na2SO4), filtration, and evaporation of the volatiles. TLC: Merck silica gel 60F-254 plates; detection with I2, by
heating with Mostain (400 ml of 10% H2SO4 soln., 20 g of (NH4)6Mo7O24 ¥ 6 H2O, 0.4 g of Ce(SO4)2) or with
KMnO4 soln. (3 g of KMnO4, 20 g of K2CO3, 0.25 ml of AcOH in 300 ml of H2O). Flash chromatography (FC):
silica gel Fluka 60 (0.04 0.063 mm). IR Spectra: 0.5% KBr suspension or 3% CHCl3 soln. 1H- (300 MHz) and
13C-NMR (75 MHz spectra): chemical shifts d in ppm and coupling constants J in Hz.
(1R,2S,5R)-5-Methyl-2-(1-methyl-1-phenylethyl)cyclohexyl (1R,4S,5S,7S,8R)-8-Acetoxy-7-azido-2-benzyl-
4-(methoxymethoxy)-3-oxo-2-azabicyclo[2.2.2]octane-5-carboxylate (10). A soln. of 9 (700 mg, 0.946 mmol)
and NaN3 (92 mg, 1.4 mmol) in DMF (12 ml) was stirred at 808 for 3 h. Normal workup (AcOEt/H2O) and FC
(cyclohexane/AcOEt 3 :1) gave 10 (576 mg, 96%). Colourless foam. Rf (cyclohexane/AcOEt 3 :2) 0.56. M.p.:
6
)
Most glycosidases possess a hydrophobic platform in the active site. Interactions with this hydrophobic site
may lead to unspecific binding of hydrophobic molecules [20].
Unlike most other retaining glycosidases, the N-acetyl-b-hexosaminidases belonging to families 18 and 20
lack a catalytic nucleophile; the C(2)-acetamido substituent of the substrate acts instead as intramolecular
catalytic nucleophile. For a review see [21].
7
)

2570
Helvetica Chimica Acta Vol. 87 (2004)
80 828 (Et2O/pentane). [a]2D5 ˆ ‡40.9 (c ˆ 1.0, CHCl3). IR (CHCl3): 2966m, 2928s, 2852m, 2113s, 1729s, 1693s,
1600w, 1496m, 1452m, 1372m, 1327m, 1158m, 1130m, 1054m, 1023m, 976m, 957m, 925w, 852w. 1H-NMR
(CDCl3): see Table 2; additionally, 7.37 7.19 (m, 9 arom. H); 7.15 (tt, J ˆ 6.5, 2.5, 1arom. H); 4.95, 4.62 (2 d, J ꢀ
7.7, OCH2O); 4.85, 4.34 (2d, J ꢀ 14.8, PhCH2); 4.77 (irrad. at 3.34 ! d, J ˆ 0.6, HÀC(8)); 4.75 (td, J ˆ 10.6, 4.4,
HÀC(1')); 3.31( s, MeO); 2.11 2.01 (m, irrad. at 4.75 ! change, HÀC(2')); 2.06 (s, AcO); 2.03 1.97 (m,
HaxÀC(3')); 1.93 (irrad. at 3.34 ! t, J ˆ 11.6, HendoÀC(6)); 1.82 (dq, J ꢀ 13.4, 3.7, HeqÀC(4')); 1.70 (irrad. at
3.34 ! ddd, J ˆ 12.5, 5.7, 1.8, HexoÀC(6)); 1.69 1.62 (m, irrad. at 4.75 ! change, HeqÀC(6')); 1.52 1.35 (m,
HÀC(5')); 1.23, 1.17 (2s, Me2PhC); 1.14 (qd, J ꢀ 13.2, 3.4, HaxÀC(3')); 0.90 (qd, J ꢀ 13.1, 3.4, HaxÀC(4')); 0.88 (q,
J ꢀ 11.7, irrad. at 4.75 ! m, HaxÀC(6')); 0.87 (d, J ˆ 6.5, MeÀC(5')). 13C-NMR (CDCl3): 171.03, 169.87, 168.56
(3s, 3 CˆO); 152.25, 136.21 (2s); 128.65 (2d); 128.56 (2d); 127.96 (d); 127.77 (2d); 125.07 (2d); 124.90 (d); 93.03
(t, OCH2O); 78.07 (d, C(7)); 76.58 (s, C(4)); 75.59 (d, C(1')); 65.29 (d, C(8)); 55.93 (q, MeO); 52.10 (d, C(1));
50.32 (d, C(2')); 48.10 (t, PhCH2); 41.44 (d, C(5)); 41.36 (t, C(6')); 39.33 (s, Me2PhC); 34.50 (t, C(4')); 31.15 (d,
C(5')); 29.32, 22.87, 21.67, 20.67 (4q, MeÀC(5'), Me2PhC, MeCˆO); 26.64, 26.13 (2t, C(6), C(3')). ESI-MS
‡
‡
‡
(632.321): 1287 (3, [2 M ‡ Na] ), 655 (100, [M ‡ Na] ), 633 (4, [M ‡ H] ), 132 (14), 116 (33), 87 (16). Anal.
calc. for C35H45NO7 (632.75): C 66.44, H 7.01, N 8.85; found: C 66.50, H 6.99, N 8.75.
Table 2. Selected 1H-NMR Chemical Shifts [ppm] and Coupling Constants [Hz] of the Isoquinuclidines 7 and 8
in CD3OD, and 9 11, 13, and 14 in CDCl3
7
8a)
9 [14]
10b)
11 [14]b)
13
2.91
2.76
2.76
4.79
4.23
1.82
1.68
2.18
14
2.84
HÀC(1)
2.57
2.64
2.76
3.46
4.19
1.84
1.51
1.80
2.77
3.62
3.34
3.53
HaÀC(3)
HbÀC(3)
HÀC(5)
HÀC(6)
HendoÀC(7)
HexoÀC(7)
HÀC(8)
2.71
3.06
3.46
3.95
1.98
1.45
1.84
2.88
2.70
4.67
4.32
1.84
1.57
2.05
5.20
4.80
1.91
1.46
1.91
4.77
3.28
1.93
1.70
2.01
4.83
4.47
1.78
1.66
2.06
J(1,6)
2.2
1.8
3.13.5
10.0
1.6
1.6
2.2
1.7
2.2
3.7
4.4
2.2
3.5
4.4
2.5
4.0
2.2
1.9
3.7
9.0
2.5
2.6
4.5
6.3
2.1
2.2
1.9
3.9
J(1,7
)
)
endo
c
J(1,7
)
exo
J(3a,3b)
J(3a,8)
J(3b,5)
J(5,6)
J(6,NH)
J(6,7exo
J(7endo,8)
J(7exo,8)
J(7endo,7exo
10.9
10.9
1.6
1.6
3.7
1.5c)
1.5c)
4.4
1.8
5.0
7.8
0.9
0.6
1.9
10.0
6.7
6.5
2.0
11.2
3.7
14.0
7.5
)
2.3
10.9
4.0
10.9
6.2
2.0c)
11.3
5.5c)
13.7
6.5
5.9
1.9
10.3
5.7
10.3
5.9
14.0
10.9
3.7
14.0
4.4
)
12.1
13.6
J(8,CHaÀC(8))
J(8,CHbÀC(8))
5.9
10.0
8.1
a) Assignment based on a DQFCOSY.GP and a HSQC.GP spectrum. b) Numbering according to the
numbering of 7. c) Broad signals; coupling constants deduced from the width of the signals at the half of their
height.
(1R,4S,5R,6S,8R)-6-Acetamido-8-(acetoxymethyl)-2-benzyl-4-(methoxymethoxy)-2-azabicyclo[2.2.2]oct-
5-yl Acetate (13). A mixture of 10 (575 mg, 0.909 mmol) and LiAlH4 (533 mg, 13.6 mmol) in THF (30 ml) was
heated under reflux for 2 h, cooled to 08, and added dropwise to cold (08) MeOH (30 ml). After evaporation, the
residue was suspended in pyridine/Ac2O 2 :1(9 ml), treated with DMAP (4 mg, 0.032 mmol), and stirred at 24 8
for 12 h. The mixture was filtered through Celite, and the residue was washed with AcOEt (3 Â 10 ml). Normal
workup (AcOEt/NaHCO3 soln., brine) and FC (toluene/THF 1:0 ! 2 :1) gave (1R,2S,5R)-5-methyl-2-(1-

Helvetica Chimica Acta Vol. 87 (2004)
2571
methyl-1-phenylethyl)cyclohexyl acetate [22] (240 mg, 96%) as a slightly yellow oil and 13 (375 mg, 92%) as a
colourless foam. Rf (CH2Cl2/MeOH/Et3N 20 :1:0.1) 0.46. [ a]2D5 ˆ À2.8 (c ˆ 1.0, CHCl3). IR (CHCl3): 3437m,
3380w (br.), 2955m, 2826m, 1734s, 1674s, 1508s, 1453m, 1369s, 1150s, 1071m, 1046s, 909m, 829w. 1H-NMR
(CDCl3): see Table 2; additionally, 7.36 7.25 (m, 4 arom. H); 7.24 (tt, J ꢀ 6.8, 2.9, 1arom. H); 5.84 (br. d, J ˆ 6.2,
NH); 4.79 (irrad. at 4.23 ! d, J ˆ 2.6, HÀC(5)); 4.70, 4.58 (2d, J ˆ 7.5, OCH2O); 4.43 (dd, J ˆ 10.6, 4.4, irrad. at
2.18 ! d, J ˆ 11.5, CHaÀC(8)); 4.23 (irrad. at 4.79 ! br. d, J ꢀ 6.2, irrad. at 2.91 ! t, J ꢀ 5.0, HÀC(6)); 4.19 (t,
J ꢀ 10.0, irrad. at 2.18 ! d, J ˆ 10.0, CHbÀC(8)); 3.86, 3.76 (2d, J ˆ 13.4, PhCH2); 3.31( s, MeO); 2.91(irrad. at
4.23 ! change, HÀC(1)); 2.88 (irrad. at 2.18 ! d, J ˆ 9.0, HaÀC(3)); 2.76 (irrad. at 4.79 ! d, J ˆ 9.0,
HbÀC(3)); 2.13, 2.07, 1.94 (3s, 3 Ac); 1.82 (irrad. at 2.91 ! dd, J ˆ 13.7, 10.0, irrad. at 2.18 ! br. d, J ꢀ 14.0,
HendoÀC(7)); 1.68 (irrad. at 4.23 ! dt, J ꢀ 14.0, 3.7, irrad. at 2.91 ! br. d, J ꢀ 13.7, irrad. at 2.18 ! br. d, J ꢀ
13.7, HexoÀC(7)). 13C-NMR (CDCl3): 170.95, 170.83, 169.98 (3s, 3 CˆO); 138.34 (s); 128.33 (2d); 128.09 (2d);
126.86 (d); 91.30 (t, OCH2O); 76.68 (d, C(5)); 75.55 (s, C(4)); 65.58 (t, CH2ÀC(8)); 59.28 (t, PhCH2); 56.05 (q,
MeO); 54.11, 52.83 (2d, C(1), C(6)); 49.16 (t, C(3)); 35.52 (d, C(8)); 25.45 (t, C(7)); 23.20, 21.24, 21.04 (3q, 3
‡
‡
‡
MeCˆO). ESI-MS: 919 (6, [2 M ‡ Na] ), 471(38, [ M ‡ Na] ), 449 (100, [M ‡ H] ). Anal. calc. for
C23H32N2O7 (448.52): C 61.59, H 7.19, N 6.25; found: C 61.58, H 7.30, N 6.23.
(1R,4R,5R,6S,8R)-6-Acetamido-8-(acetoxymethyl)-2-benzyl-4-hydroxy-2-azabicyclo[2.2.2]oct-5-yl Acetate
(14). A cooled (08) mixture of 13 (100 mg, 0.223 mmol) and mol. sieves (3 ä, 20 mg) in CH2Cl2 (6 ml) was
treated dropwise with Me3SiBr (60 ml, 0.45 mmol) and heated to reflux for 2 h. Normal workup (AcOEt/
NaHCO3 soln., brine) and FC (CH2Cl2/i-PrOH 9 :1) gave 14 (69 mg, 76%). Colourless oil. Rf (CH2Cl2/MeOH
10 :1) 0.71. [a]D25 ˆ À 15.4 (c ˆ 1.0, CHCl3). IR (CHCl3): 3439w, 3388w, 2958w, 2837w, 1732s, 1673s, 1509m,
1453m, 1369s, 1248s, 1172w, 1126m, 1074m, 1030m, 977w, 909m, 852w. 1H-NMR (CDCl3): see Table 2;
additionally, 7.36 7.24 (m, 4 H); 7.21 (tt, J ꢀ 7.2, 2.2, 1H); 5.93 (br. d, J ˆ 6.5, NH); 4.46 (dd, J ˆ 10.9, 7.5,
CHaÀC(8)); 4.21( dd, J ˆ 10.9, 8.1, CHbÀC(8)); 3.86, 3.75 (2d, J ˆ 13.4, PhCH2); 2.54 (s, OH); 2.15, 2.05, 1.95
(3s, 3 Ac). 13C-NMR (CDCl3): 171.83, 170.98, 169.84 (3s, 3 CˆO); 138.45 (s); 128.39 (2d); 128.15 (2d); 126.94
(d); 79.18 (d, C(5)); 71.26 (s, C(4)); 65.76 (t, CH2ÀC(8)); 59.30 (t, PhCH2); 52.92, 52.46 (2d, C(1), C(6)); 50.52
‡
(t, C(3)); 37.70 (d, C(8)); 25.45 (t, C(7)); 23.35, 21.24, 21.16 (3q, 3 MeCˆO). ESI-MS: 831(4, [2 M ‡ Na] ), 427
‡
‡
(18, [M ‡ Na] ), 405 (100, [M ‡ H] ).
N-(1R,4R,5R,6S,8R)-2-Benzyl-4,5-dihydroxy-8-(hydroxymethyl)-2-azabicyclo[2.2.2]oct-5-yl]acetamide
(7). A soln. of 14 (40 mg, 0.074 mmol) in MeOH (2 ml) was cooled to 08, saturated with NH3, stirred at 08 for
5 h, diluted with MeOH (10 ml), and evaporated. Ion-exchange chromatography (Amberlite CG-120, H form,
‡
0.1m aq. NH3) gave 7 (27 mg, 85%). Colourless solid. Rf (CH2Cl2/MeOH 10 :1) 0.26. [a]2D5 ˆ À23.9 (c ˆ 1.0,
EtOH). pKHA 5.90. IR (KBr): 3405s (br.), 3073m, 2926m, 1654s, 1560s, 1496m, 1420s, 1376m, 1316m, 1259w,
1154w, 1120m, 1077m, 1042m, 962w. 1H-NMR (CD3OD): see Table 2; additionally, 7.35 (br. dd, J ꢀ 7.8, 1.2, 2
arom. H); 7.27 (tt, J ꢀ 7.2, 1.7, 2 arom. H); 7.20 (tt, J ꢀ 7.2, 2.0, 1arom. H); 3.90 ( dd, J ˆ 10.3, 6.2, irrad. at 1.80 !
d, J ˆ 10.3, CHaÀC(8)); 3.85, 3.73 (2d, J ˆ 13.4, PhCH2); 3.72 (dd, J ˆ 10.3, 5.9, irrad. at 1.80 ! d, J ˆ 10.0,
CHbÀC(8)); 2.64 (irrad. at 1.80 ! d, J ˆ 10.0, HbÀC(3)); 1.95 (s, AcN); 1.51 (irrad. at 1.80 ! ddd, J ꢀ 10.6, 3.1,
2.2, HexoÀC(7)). 13C-NMR (CD3OD): 172.85 (s, CˆO); 139.87 (s); 128.75 (2d); 129.10 (2d); 127.93 (d); 78.69 (d,
C(5)); 73.33 (s, C(4)); 64.06 (t, CH2ÀC(8)); 60.21( t, PhCH2); 55.35, 54.27 (2d, C(1), C(6)); 51.20 (t, C(3)); 41.26
‡
‡
(d, C(8)); 26.11. (t, C(7)); 22.57 (q, MeCˆO). HR-MALDI-MS: 343.1622 (21, [M ‡ Na] , C17H24N2NaO4
;
‡
‡
calc.: 343.1634), 321.1811 (21, [M ‡ H] , C17H25N2O4 ; calc.: 321.1814). Anal. calc. for C17H24N2O4 ¥ H2O
(338.40): C 60.34, H 7.74, N 8.28; found: C 60.81, H 7.74, N 8.04.
For the enzyme tests, a sample of 7 was crystallized from MeOH. Anal. calc. for C17H24N2O4 ¥ MeOH
(336.60): C 61.34, H 8.01, N 7.95; found: C 61.50, H 8.01, N 7.95.
N-(1R,4R,5R,6S,8R)-4,5-Dihydroxy-8-(hydroxymethyl)-2-azabicyclo[2.2.2]oct-6-yl]acetamide (8). A sus-
pension of 7 (30 mg, 0.094 mmol) and 20% Pd(OH)/C (6 mg) in MeOH/H2O/conc. aq. HCl 1:1:0.1 (2.1ml)
was stirred at 248 for 12 h under H2 (5 bar). The suspension was filtered through Celite, and the residue was
washed with MeOH (3 Â 5 ml). The combined filtrate and washings were diluted with toluene (5 ml) and
evaporated. Ion-exchange chromatography (Amberlite CG-120, 0.1m aq. NH3) gave 8 (20 mg, 93%). Colourless
solid. Rf (MeOH/25% aq. NH3 soln. 3 :1) 0.44. [a]2D5 ˆ ‡50.7 (c ˆ 0.4, H2O). pKHA 7.7. IR (CHCl3): 3405s (br.),
2937m, 1636s, 1544s, 1378m, 1314m, 1174m, 1119m, 1084m, 1042m, 1011m, 925w. 1H-NMR (CD3OD, assignment
based on a DQFCOSY.GP and a HSQC.GP spectrum): see Table 2; additionally, 3.92 (dd, J ˆ 10.6, 6.5,
CHaÀC(8)); 3.64 (dd, J ˆ 10.3, 5.9, CHbÀC(8)). 13C-NMR (CD3OD, assignment based on a HSQC.GP
spectrum): 173.02 (CˆO); 78.69 (d, C(5)); 72.25 (s, C(4)); 63.82 (t, CH2ÀC(8)); 60.22 (d, C(6)); 48.68 (d, C(1));
‡
40.93 (t, C(3)); 40.16 (d, C(8)); 26.61( t, C(7)); 22.52 (q, MeCˆO). HR-ESI-MS: 253.1159 (20, [M ‡ Na] ,
‡
‡
‡
C10H18N2NaO4 ; calc.: 231.1345), 231.1339 (100, [M ‡ H] , C10H19N2O4 ; calc.: 231.1345).

2572
Helvetica Chimica Acta Vol. 87 (2004)
(1R,4S,5R,6S,8R)-6-Acetamido-8-(acetoxymethyl)-2-[(tert-butoxy)carbonyl]-4-hydroxy-2-azabicyclo[2.2.2]-
oct-5-yl Acetate (15). A soln. of 8 (20 mg, 0.087 mmol) in MeOH/Et3N 10 :1 (1.1 ml) was treated with di(tert-
butyl) carbonate (49 mg, 0.175 mmol), stirred at 288 for 4 h, and evaporated. A soln. of the residue in pyridine/
Ac2O 2 :1(3 ml) was stirred at 5 8 for 24 h. Normal workup (AcOEt/NaHCO3 soln., brine) and FC (toluene/
AcOEt 2 :1 ! 0 :1) gave 15 (27 mg, 75%). Colourless foam. Rf (AcOEt) 0.16. M.p.: 103 1088. [a]2D5 ˆ ‡20.1
(c ˆ 1.0, CHCl3). IR (CHCl3): 3440w, 3332w (br.), 3031w, 3008w, 1735s, 1683s, 1503w, 1478w, 1458w, 1408m,
1368s, 1162m, 1102w, 1070w, 1032w, 978w, 91 0w, 870w. 1H-NMR (CDCl3, 6 :4 mixture of diastereoisomers): 6.81
(d, J ˆ 7.2, 0.6 H), 6.41( d, J ˆ 6.5, irrad. at 4.02 ! s, 0.4 H) (NH); 4.79 (d, J ˆ 2.4, 0.6 H), 4.70 (d, J ˆ 2.4, irrad.
at 4.02 ! s, 0.4 H) (HÀC(5)); 4.37 (br. dd, J ꢀ 10.8, 5.5, irrad. at 2.19 ! d, J ꢀ 10.8, CHaÀC(8)); 4.22 (dd, J ꢀ
11.5, 2.8, irrad. at 2.19 ! d, J ꢀ 10.8, CHbÀC(8)); 4.17 4.08 (m, irrad. at 4.02 ! change, irrad. at 2.19 !
change, irrad. at 1.57 ! change, HÀC(1), 0.6 HÀC(6)); 4.06 3.96 (m, 0.4 HÀC(6)); 3.82 (s, 0.6 H), 3.48 (s,
0.4 H) (OH); 3.49 (br. d, J ˆ 10.8, 0.6 H), 3.33 (br. d, J ˆ 11.2, 0.4 H) (HaÀC(3)); 3.44 (br. d, J ˆ 10.8, 0.6 H),
3.29 (br. d, J ˆ 11.2, 0.4 H) (HbÀC(3)); 2.26 2.12 (m, irrad. at 1.57 ! change, HendoÀC(7), HÀC(8)); 2.13 (s,
1.2 H), 2.12 (s, 1.8 H), 2.05 (s, 3 H), 1.98 (s, 1.8 H), 1.91 (s, 1.2 H), (3 Ac); 1.57 (br. d, J ꢀ 13.1, irrad. at 2.19 !
br. s, HexoÀC(7)); 1.48 (s, t-Bu). 13C-NMR (CDCl3, 6 :4 mixture of diastereoisomers): 171.44, 170.91 (2s,
MeCˆO); 170.11, 169.99 (2s, 2 MeCˆO); 153.98, 153.80 (2s, t-BuOCˆO); 80.44 (s, Me3C); 78.98, 78.89 (2d,
C(5)); 71.21, 70.11 (2s, C(4)); 64.59, 64.32 (2t, CH2ÀC(8)); 56.36, 55.73 (2d, C(6)); 46.83, 45.74 (2d, C(1)); 43.35,
42.75 (2t, C(3)); 36.91, 36.67 (2d, C(8)); 28.55, 28.47 (2q, Me3C); 25.94, 25.75 (2t, C(7)); 23.37, 23.26 (2q,
‡
‡
MeCˆO); 21.15, 21.08 (2q, 2 MeCˆO). HR-MALDI-MS: 437.1891 (56, [M ‡ Na] , C19H30N2NaO8 ; calc.:
‡
‡
437.1900), 381.1263 (100, [M À t-Bu ‡ H ‡ Na] , C15H22N2NaO8 ; calc.: 381.1274), 315.1540 (35, [M À t-
‡
‡
BuO2C ‡ 2 H] , C14H23N2O6 ; calc.: 315.1556). Anal. calc. for C19H30N2O8 (414.45): C 55.06, H 7.30, N 6.76;
found: C 55.31, H 7.43, N 6.61.
Hydrolysis of 15 to 8. A soln. of 15 (20 mg, 0.048 mmol) in MeOH/H2O/conc. aq. HCl 1:1:0.1(2.1ml) was
stirred at 258 for 12 h, diluted with toluene (3 Â 5 ml) and evaporated. Ion-exchange chromatography
‡
(Amberlite CG-120, H form, 0.1M aq. NH 3) gave 8 (10 mg, 90%). Colourless solid.
Inhibition of N-Acetyl-b-hexosaminidases. IC50 Values were determined at a substrate concentration
corresponding to the KM of each enzyme. The IC50 values were calculated by plotting the inhibitor concentration
vs. the rate of hydrolysis. Determination of the inhibition constants (Ki) was performed at five different
concentrations of the inhibitor bracketing the IC50 value. Ki Values and a values were determined from the
replot of the slopes and the replot of the 1/v axis intercepts of LineweawerÀBurk plots [23].
a) From Jack Beans. KM ˆ 0.72 mm ([24]: 0.62 mm). The inhibition studies were carried out at pH 5.0 (0.04m
citric acid/sodium citrate buffer, containing 0.8m NaCl) and 258. The reaction was started by the addition of 4-
nitrophenyl 2-acetamido-2-deoxy-b-d-glucopyranoside after pre-incubating the enzyme in the presence of the
inhibitor for 30 min at 258. After 10 min, the reaction was quenched by addition of borate buffer (pH 9.0, 0.02m).
The rate of hydrolysis was determined by measuring the absorption at l ˆ 405 nm and subsequently substracting
the absorption of a blank probe (H2O, buffer, substrate).
b) From Bovine Kidney. KM ˆ 1.24 mm ([25]: 1.87 mm). The inhibition studies were carried out in an
analogous way as for the inhibition of the N-acetyl-b-hexosaminidase from jack beans, but at pH 4.1(0.06 m citric
acid/sodium citrate buffer) and 378.
REFERENCES
[1] G. J. Davies, V. M.-A. Ducros, A. Varrot, D. L. Zechel, Biochem. Soc. Trans. 2003, 31, 523.
[2] A. Vasella, G. J. Davies, M. Bˆhm, Curr. Opin. Chem. Biol. 2002, 619; P. J. Berti, H. S. E. Tanaka, Adv.
Phys. Org. Chem. 2002, 37, 239; S. G. Withers, Carbohydr. Polym. 2001, 44, 325; D. L. Zechel, S. G. Withers,
Acc. Chem. Res. 2000, 33, 11; C. S. Rye, S. G. Withers, Curr. Opin. Chem. Biol. 2000, 4, 573.
[3] P. Deslongchamps, Tetrahedron 1975, 31, 2463.
[4] A. J. Kirby, −The Anomeric Effect and Related Stereoelectronic Effects at Oxygen×, Spinger-Verlag, Berlin,
1983.
[5] P. M. Coutinho, B. Henrissat, in −Recent Advances in Carbohydrate Bioengineering×, Eds. H. J. Gilbert, G.
Davies, B. Henrissat, B. Svensson, The Royal Society of Chemistry, Cambridge, 1999, p. 3.
[6] G. J. Davies, L. Mackenzie, A. Varrot, M. Dautes, A. M. Brzozowski, M. Sch¸lein, S. G. Withers,
Biochemistry 1998, 37, 11707; A. Varrot, G. J. Davies, Acta Crystallogr., Sect. D 2003, 59, 447.
[7] G. Sulzenbacher, H. Driguez, B. Henrissat, M. Sch¸lein, G. J. Davies, Biochemistry 1996, 35, 15280.

Helvetica Chimica Acta Vol. 87 (2004)
2573
[8] D. M. F. van Aalten, D. Komander, B. Synstad, S. Gaseidnes, M. G. Peter, V. G. H. Eijsink, Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 8979.
[9] Y. Papanikolaou, G. Prag, G. Tavlas, C. E. Vorgias, A. B. Oppenheim, K. Petratos, Biochemistry 2001, 40,
11338.
[10] I. Tews, A. Perrakis, A. Oppenheim, Z. Dauter, K. S. Wilson, C. E. Vorgias, Nat. Struct. Biol. 1996, 3, 638.
[11] G. Prag, Y. Papanikolaou, G. Tavlas, C. E. Vorgias, K. Petratos, A. B. Oppenheim, J. Mol. Biol. 2000, 300,
611.
[12] E. Lorthiois, M. Meyyappan, A. Vasella, Chem. Commun. 2000, 1829.
[13] M. Bˆhm, E. Lorthiois, M. Meyyappan, A. Vasella, Helv. Chim. Acta 2003, 86, 3787.
[14] M. Bˆhm, E. Lorthiois, M. Meyyappan, A. Vasella, Helv. Chim. Acta 2003, 86, 3818.
[15] M. Terinek, A. Vasella, Helv. Chim. Acta 2003, 86, 3482.
[16] M. D. Sulta, J. P. Lassig, M. G. Gooch, B. R. Evans, J. Woodward, Biochem. Biophys. Res. Commun. 1995,
209, 1046.
[17] C. Bl¸chel, C. V. Ramana, A. Vasella, Helv. Chim. Acta 2003, 86, 2998.
[18] B. Schlesier, V. H. Nong, C. Horstmann, M. Hennig, J. Plant Physiol. 1995, 147, 665.
[19] M. Hennig, J. N. Jansonius, A. C. Terwisscha van Scheltinga, B. W. Dijkstra, B. Schlesier, J. Mol. Biol. 1995,
254, 237.
[20] W. Nerinckx, T. Desmet, M. Claeyssens, FEBS Lett. 2003, 538, 1 .
[21] B. L. Mark, M. N. G. James, Can. J. Chem. 2002, 80, 1064.
[22] S. Aoyagi, R. Tanaka, M. Naruse, C. Kibayashi, J. Org. Chem. 1998, 63, 8397.
[23] I. H. Segel, −Enzyme Kinetics×, John Wiley & Sons, New York, 1975.
[24] E. C. K. Lai, S. G. Withers, Biochemistry 1994, 33, 14743.
[25] A. J. Pope, N. Mian, D. G. Herries, FEBS Lett. 1978, 93, 174.
Received June 18, 2004