Synthesis of aryl 3-O-β-cellobiosyl-β-D-glucopyranosides for reactivity studies of 1,3-1,4-β-glucanases

Article abstract of DOI:10.1016/S0008-6215(98)00175-X

A series of substituted aryl β-glycosides derived from 3-O-β-cellobiosyl-D-glucopyranose with different phenol-leaving group abilities as measured by the pK(a) of the free phenol group upon enzymatic hydrolysis has been synthesised. Aryl β-glycosides with a pK(a) of the free phenol leaving group>5 were prepared by phase-transfer glycosidation of the per-O-acetylated α-glycosyl bromide with the corresponding phenol, whereas the 2,4-dinitrophenyl β-glycoside was obtained by condensation of 1-fluoro-2,4-dinitrobenzene with the partially acetylated trisaccharide followed by acid de-O-acetylation. The aryl β-glycosides have been used for reactivity studies of the wild-type Bacillus licheniformis 1,3-1,4-β-d-glucan 4-glucanohydrolase. The Hammett plot log k(cat) versus pK(a) is biphasic with an upward curvature at low pK(a) values suggesting a change in transition-state structure depending on the aglycon. Copyright (C) 1998 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(98)00175-X

Carbohydrate Research 310 (1998) 53±64
Synthesis of aryl 3-O-ꢀ-cellobiosyl-ꢀ-d-
glucopyranosides for reactivity studies of
1,3-1,4-ꢀ-glucanases
Antoni Planas*, Mireia Abel, Oscar Millet, Josep Palasõ, Cristina Pallares,
Josep-Lluõs Viladot
Â
Laboratory of Biochemistry, Department of Organic Chemistry, Institut Quõmic de Sarria, Universitat
Á
Ramon Llull, Via Augusta 390, 08017-Barcelona, Spain
Received 2 March 1998; accepted 27 June 1998
Abstract
A series of substituted aryl ꢀ-glycosides derived from 3-O-ꢀ-cellobiosyl-d-glucopyranose with
di€erent phenol-leaving group abilities as measured by the pKa of the free phenol group upon
enzymatic hydrolysis has been synthesised. Aryl ꢀ-glycosides with a pKa of the free phenol leaving
group>5 were prepared by phase-transfer glycosidation of the per-O-acetylated ꢁ-glycosyl bro-
mide with the corresponding phenol, whereas the 2,4-dinitrophenyl ꢀ-glycoside was obtained by
condensation of 1-¯uoro-2,4-dinitrobenzene with the partially acetylated trisaccharide followed by
acid de-O-acetylation. The aryl ꢀ-glycosides have been used for reactivity studies of the wild-type
Bacillus licheniformis 1,3-1,4-ꢀ-d-glucan 4-glucanohydrolase. The Hammett plot log kcat versus
pKa is biphasic with an upward curvature at low pKa values suggesting a change in transition-state
structure depending on the aglycon. # 1998 Elsevier Science Ltd. All rights reserved
Keywords: Aryl glycosides; 1,3-1,4-ꢀ-Glucanase; Enzyme kinetics; Hammett analysis
1. Introduction
mainly on monosaccharidases and exo-glycanases
(e.g. ꢀ-glucosidase from Agrobacterium faecalis [5],
ꢀ-galactosidase from Escherichia coli [6], exo-ꢀ-1,4-
glycanase from Cellulomonas ®mi [7], sialidases
from in¯uenza A virus and Salmonella typhimurium
[8]), and lysozyme [9,10] and xylanase from Bacil-
lus circulans [11,12] as representatives of endo-gly-
canases. Glycosidases are grouped in two families
according to their mechanism: retaining and
inverting enzymes, acting by double or single dis-
placement mechanisms, respectively. Although
members of one family share a common reaction
mechanism, the few reported examples show
Synthetic chromophoric oligosaccharides are of
widespread use as substrates and ligands to study
the kinetic properties of glycosidases. 4-Methyl-
umbelliferyl and aryl glycosides have been used to
analyse the substrate speci®city of a number of
glycosidases (see, for example, refs. [1±4]). However,
only few studies on linear free energy relationships
(LFER, namely Hammett and Brùnsted analyses)
have been performed, the reported examples being
* Corresponding author.
0008-6215/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII S0008-6215(98)00175-X

54
A. Planas et al./Carbohydrate Research 310 (1998) 53±64
remarkable di€erences in the shape and ꢂ (or ꢀ)
values of their Hammett plots, providing evidence
of subtle variations in the ®ne structure of their
catalytic machineries and reaction mechanisms.
Detailed mechanistic investigations for endo-acting
enzymes have proved to be extremely dicult due
in part to the complexity of the substrates and
therefore the enormous amount of synthetic chem-
istry required. A clear need exists for LFER studies
with endo-glycosidases, both acting with retention
and inversion of con®guration at the anomeric
center to better understand the ®ne details and
di€erences in their catalytic mechanisms.
1,3-1,4-ꢀ-d-glucan 4-glucanohydrolase (EC
3.2.1.73, 1,3-1,4-ꢀ-glucanase) is a retaining endo-
depolymerase that catalyzes the regio and stereo-
speci®c hydrolysis of ꢀ-glucans containing mixed
ꢀ-(1!3) and ꢀ-(1!4) glycosidic linkages such as
cereal ꢀ-glucans and lichenan [13±15]. The bacter-
ial 1,3-1,4-ꢀ-glucanases have signi®cant commer-
cial interest, particularly in the brewing and animal
feedstu€ industries [16,17]. The trisaccharide 3-O-
ꢀ-cellobiosyl-d-glucopyranose (1) and the tetra-
saccharide 3-O-ꢀ-cellotriosyl-d-glucopyranose are
the ®nal hydrolysis products of barley ꢀ-glucan,
thus de®ning the enzymatic speci®city of cleavage
on ꢀ-(1!4) glycosidic linkages in 3-O-substituted
glucopyranose units [15].
substrate representing a good balance between
synthetic complexity and reactivity; (b) the agly-
cons to be released as leaving groups in the enzy-
matic reactions are chosen by their phenol-leaving
group ability as measured by the pKa of the phenol
(pKa in parenthesis): phenol (10), 4-bromophenol
(9.34), 3-nitrophenol (8.39), 4-nitrophenol (7.18),
3,5-dinitrophenol (6.69), 3,4-dinitrophenol (5.36),
2,3-dinitrophenol (4.96), and 2,4-dinitrophenol
(3.96), which cover the pKa range from 4 to 10
[5,6].
The oligosaccharide 3-O-ꢀ-cellobiosyl-d-gluco-
pyranose (1) was obtained by enzymatic degrada-
tion of barley ꢀ-glucan (either as a puri®ed
polysaccharide or as a raw barley ¯our after treat-
ment with ethanol) with recombinant 1,3-1,4-ꢀ-
glucanase from Bacillus licheniformis [21]. After
full acetylation of the reaction mixture with acetic
anhydride±pyridine, chromatographic puri®cation
on silica gel a€orded 2.
The synthesis of aryl glycosides with a pKa of
the phenol-leaving group greater than 5 (6a±g) was
based on a phase-transfer glycosidation according
to the procedure reported by Dess et al. [24] for the
preparation of aryl glycosides of monosaccharides,
and previously used in our group for the synthesis
of the trisaccharide G4G3G-MU [21]. In that case,
the procedure improved substantially the con-
densation yield between the ꢁ-bromo derivative of
the peracetylated trisaccharide 2 and 4-methylum-
belliferone as compared to the classical procedure in
a homogeneous system (aqueous NaOH±acetone
mixture) [1,7]. As shown in Scheme 1, the ꢁ-glyco-
syl bromide 3Ðprepared from the peracetate 2 by
reaction with 33% hydrogen bromide in acetic
acidÐwas treated with the phenols 4a±g in a
biphasic system chloroform-1.25 M aqueous
NaOH in the presence of benzyltriethylammonium
bromide as phase-transfer catalyst to a€ord 5a±g in
25 to 65% yield (not optimized). Deacetylation of
compounds 5a±f with sodium methoxide in
methanol yielded the aryl glycosides 6a±f. Deace-
tylation of the base-sensitive 3,4- and 2,3-dini-
trophenyl glycosides 5f,g was performed with
hydrogen chloride in anhydrous methanol [25]. The
deprotected aryl glycosides 6a±g were recrystalized
from ethanol±water to give chromatographically
homegeneous materials for enzymological studies
(>99% purity by HPLC).
Following our structure/function studies of bac-
terial 1,3-1,4-ꢀ-glucanases [18±23], here we report
on the synthesis of a series of aryl glycosides of
3-O-ꢀ-cellobiosyl-d-glucopyranose as speci®c sub-
strates for reactivity studies. Their kinetic para-
meters for the enzyme-catalyzed hydrolysis with
1,3-1,4-ꢀ-glucanase from Bacillus licheniformis are
reported.
2. Results
Synthesis.ÐThe design of the target substrates
was based on two criteria: (a) the glyconic portion
should not be hydrolyzed by the enzyme, and
should contain the minimum structural require-
ments for binding to the active site. Earlier kinetic
work with low-molecular-weight ꢀ-glucan oligo-
saccharides having a 4-methylumbelliferyl agly-
con [20,23] ([G4]nG3G-MU1, n ˆ 0±3) showed
that the trisaccharide derivative (n ˆ 1) is a good
This strategy proved to be unsuccessful for the
preparation of the 2,4-dinitrophenyl glycoside due
to the lability of the glycosidic linkage with the
1
MU: 4-methylumbelliferyl, G4G3G-X: aryl 3-O-ꢀ-cel-
lobiosyl-d-glucopyranoside.

A. Planas et al./Carbohydrate Research 310 (1998) 53±64
55
Scheme 1.
aglycon (pKa of the phenol 3.96) under the basic
conditions used in the phase-transfer glycosidation
step. Introduction of the 2,4-dinitrophenyl group
at the anomeric center has been reported by react-
ing an acetylated glycosyl halide with 2,4-dini-
trophenol in the presence of potassium carbonate
[25]. A more ecient route, however, is based on
the electrophilic aromatic substitution of 1-¯uoro-
2,4-dinitrobenzene (FDNB) by the free anomeric
hydroxyl group of the saccharide in the presence of
1,4-diazabicyclo[2.2.2]octane (DABCO) in N,N-
dimethylformamide [7,26]. First, selective deacety-
lation of the anomeric hydroxyl group of the
peracetylated trisaccharide 2 with hydrazine acet-
ate led to 7 in 98% yield. Compound 7 was then
reacted with 1-¯uoro-2,4-dinitrobenzene in the
presence of 1,4-diazabicyclo[2.2.2]octane and
molecular sieves in anhydrous N,N-dimethylform-
Table 1
1H NMR chemical shifts (ꢃ), multiplicity, and coupling constants (Hz, in parenthesis) for compounds 6a±h (D2O, 300 MHz)
6a
6b
6c
6d
6e
6f
6g
6h
H-2 phenyl 7.17 d (8.5) 7.07 d (8.0) 7.83 dd (3.0) 7.26 d (10) 8.34 d (1.8)
H-6 phenyl 7.17 d (8.5) 7.07 d (8.0) 7.30 dd (8.0,3.0) 7.26 d (10) 8.34 d (1.8)
7.68 d (2.5)
7.51 dd (9.0, 2.5) 7.86 dd (1.2, 8.5) 7.61 d (9.6)
Ð
Ð
H-3 phenyl 7.42 dd (8.0) 7.54 d (8.0)
H-5 phenyl 7.42 dd (8.0) 7.54 d (8.0) 7.47 t (8.0)
H-4 phenyl 7.15 t (8.5) 7.95 dd (8.0,3.0)
5.13 d (7.5) 5.25 d (8.0)
Ð
8.27 d (10)
8.27 d (10)
Ð
5.31 d (7.5) 5.38 d (7.5)
4.86 d (8.0) 4.85 d (7.8)
4.52 d (8.0) 4.52 d (7.8)
Ð
Ð
Ð
8.21 d (9.0)
Ð
5.34 d (8.0)
4.85 d (8.0)
4.52 d (8.0)
Ð
8.90 d (3.0)
8.54 dd (9.3)
Ð
5.44 d (8.0)
4.84 d (8.0)
4.52 d (8.0)
4.05±3.30 m
7.80 dd (8.5)
8.06 dd (1.2, 8.5)
5.29 d (7.8)
4.84 d (8.0)
4.51 d (7.8)
4.03±3.30 m
Ð
8.82 d (1.8)
H-10
5.17 d (10)
H-1
4.84 d (8.0) 4.83 d (8.0) 4.83 d (8.0)
4.52 d (8.0) 4.52 d (8.0) 4.52 d (8.0)
4.03±3.30 m 4.03±3.30 m 4.04±3.30 m
H-100
other H
4.04±3.30 m 4.060±3.30 m 4.06±3.30 m

56
A. Planas et al./Carbohydrate Research 310 (1998) 53±64
Table 2
13C NMR chemical shifts (ꢃ) for compounds 6a±h (D2O, 75 MHz)
6a
6b
6c
6d
6e
160.2
6f
159.7
6g
151.6
6h
157.2
C-1 phenyl
C-2 phenyl
C-3 phenyl
C-4 phenyl
C-5 phenyl
C-6 phenyl
C-10,100
C-1
C-3
C-40
other Ca
C-6,60,600
159.4
158.6
157.0
164.5
119.5
132.8
126.3
132.8
119.5
105.5
102.8
86.8
117.9
121.4
135.4
121.4
117.9
105.4
102.9
86.7
110.8
149.1
118.1
130.2
123.8
105.4
101.9
86.7
119.4
129.0
145.5
129.0
119.4
105.4
102.1
86.5
120.6
151.7
116.1
151.7
120.6
105.5
102.9
86.4
112.2
145.0
136.7
127.1
120.4
104.0
101.9
86.7
135.2
142.7
122.4
137.8
126.7
105.4
103.7
86.1
141.8
125.0
144.4
132.7
120.6
105.4
102.9
86.3
81.4
81.4
81.4
81.4
81.5
81.4
81.5
81.5
79±71
63.4, 63.3,
62.9
79±71
63.4, 63.3,
62.9
79±70
63.4, 63.3,
62.9
79±70
63.4, 63.3,
62.9
79±70
63.4, 63.3,
62.9
79±70
63.7, 63.3,
61.9
79±70
63.4, 63.3,
62.9
79-70
63.5, 63.2,
62.9
C-2,20,200,30,300,4,400,5,50,500.
a
Table 3
Kinetic parameters of aryl glycosides 6a±d,f,h,i with B. licheniformis 1,3±1,4-ꢀ-glucanase at 55 C
ꢀ
kcat=KM Â 103
1
Compound aglycon
Concentration
range (mM)
Á" (M 1/cm)
/l (nm)
pKa phenol
kcat (s
)
KM (mM)
(M 1/s)
6a
6b
6c
6i
6d
6f
P
0.5±12
1.3±10
1.0±11
0.4±15
1.2±10
0.1±1.7
0.02±1.0
920/277
683/288
186/400
5020/365
6200/400
9496/425
6000/425
9.99
9.34
8.39
7.50
7.18
5.36
3.96
3.0‹0.4
3.8‹0.4
3.9‹0.1
4.2‹0.3
4.6‹0.5
39‹12
7.9‹0.6
2.9‹0.5
6.5‹0.4
3.3‹0.4
2.6‹0.5
1.8‹0.7
0.2‹0.03
0.38‹0.08
1.31‹0.14
0.61‹0.06
1.28‹0.27
1.72‹0.52
21.4‹14.7
(9‹2)Â103
4-BrP
3-NP
MU
4-NP
3,4-DNP
2,4-DNP
6h
1880‹90
ꢀ
Conditions: citrate±phosphate bu€er, pH 7.2, 0.1 mM CaCl2, 55 C.
Table 4
Kinetic parameters of aryl glycosides 6a,c±i with B. licheniformis 1,3±1,4-ꢀ-glucanase at 30 C
ꢀ
kcat=KM Â 103
1
Compound aglycon
Concentration
range (mM)
Á" (M 1/cm)
/l (nm)
pKa phenol
kcat (s
)
KM (mM)
(M 1/s)
6a
6c
6i
6d
6e
6f
P
3-NP
MU
4-NP
3,5-DNP
3,4-DNP
2,3-DNP
2,4-DNP
0.7±20
1019/277
221/400
9.99
8.39
7.50
7.18
6.69
5.36
4.96
3.96
0.49‹0.02
0.85‹0.01
0.75‹0.02
0.81‹0.03
4.4‹0.2
14.2‹0.2
272‹6
590‹15
7.6‹0.4
0.065‹0.006
0.67‹0.03
0.80‹0.05
1.1‹0.1
9.6‹0.9
2.0‹1.5
0.3±6.5
0.3±6.0
0.1±6.0
0.1±4.1
0.07±1.8
0.1±4.0
0.02±0.8
1.28‹0.04
0.93‹0.04
0.73‹0.05
0.46‹0.03
0.44‹0.023
1.28‹0.07
0.18‹0.02
5136/365
9867/402
1671/400
9616/400
2359/418
8954/400
6g
6h
211‹15
3340‹300
ꢀ
Conditions: citrate±phosphate bu€er, pH 7.2, 0.1 mM CaCl2, 30 C.
amide to give 5h in 50% yield after chromato-
graphic puri®cation. Deprotection of 5h was
achieved with hydrogen chloride in anhydrous
methanol [25] to a€ord the target compound 6h in
67% yield. The ꢀ con®guration of the glycosidic
bond with the aryl aglycons of compounds 6a±h
NMR data for 6a±h are summarized in Tables 1
and 2.
Enzyme kinetics.ÐThe aryl glycosides were used
as substrates of Bacillus licheniformis 1,3±1,4-ꢀ-
glucanase (recombinant wild-type enzyme expressed
in E. coli [18]). Michaelis±Menten parameters for
the hydrolysis of 6a±h, as well as of the previously
synthesized 4-methylumbelliferyl substrate 6i [19],
were determined by monitoring the release of the
chromophoric aglycon by UV spectrophotometry
1
was ascertained by H NMR spectroscopy, which
showed a characteristic doublet at ꢃ 5.11±5.44 ppm
with a coupling constant of 8±10 Hz assigned to the
anomeric protons of the ꢀ-glycosides. Complete

A. Planas et al./Carbohydrate Research 310 (1998) 53±64
kcat‰SŠ‰EoŠ
57
at the appropriate wavelength as described in the
Experimental section. Kinetic results are pre-
ꢄo ˆ
ꢀ1†
2
‰SŠ
KM ‡ ‰SŠ ‡
KI
ꢀ
sented in Table 3 (reactions at 55 C) and Table 4
ꢀ
(reactions at 30 C). As previously shown for the
4-methylumbelliferyl glycoside 6i, substrate inhibi-
tion was observed at high substrate concentration.
Since no transglycosidation was detected over the
entire concentration range (as shown by TLC), the
inhibition is likely due to binding of a second
molecule of substrate to the binding site (composed
of 6 or 7 subsites) [20,23]. Accordingly, data were
®tted to an uncompetitive substrate inhibition
model [eq (1)] where a second molecule of substrate
binds with formation of unproductive ternary
complexes ES2.
All the aryl glycosides proved to be good substrates
of Bacillus licheniformis 1,3-1,4-ꢀ-glucanase. They
undergo a single glycosidic bond cleavage with
release of the chromophoric aglycon and the tri-
saccharide 1 (TLC). Initially, the enzymatic
hydrolyses of substrates 6a±d, 6f, and 6h,i were
studied at 55 ꢀC (Table 3), the temperature of
maximum activity of the wild type enzyme with
natural polymeric substrates [27]. The 2,4-dini-
trophenyl glycoside 6f is highly reactive; it has the
highest kcat and the lowest KM values. However,
the spontaneous hydrolysis (non-enzymatic) at the
assay temperature (55 ꢀC) is high. The enzyme
concentration in the enzymatic assay had to be
raised to obtain initial rates at least one order of
magnitude faster than the spontaneous hydrolysis,
thus necessitating great care in monitoring of
initial reaction rates.
The steady-state kinetic parameters for the 1,3-
1,4-ꢀ-glucanase-catalyzed hydrolysis of the aryl
ꢀ
glycosides at 55 C show no simple dependence
on aglycon acidity. The Hammett plot (log kcat
versus pKa) in Fig. 1(a) is biphasic with an
upward curvature at low pKa values. Only two
substrates de®ne the plot in the low pKa region
(6f and 6h), so the slope de®ning the ꢀ value in
the Hammett analysis can not be accurately
determined.
Fig. 1. Hammett relationship for the 1,3±1,4-ꢀ-glucanase-cat-
alyzed hydrolysis of aryl glycosides 6a±h. Data are presented
in the form of a Brùnsted plot (log kcat versus pKa of the
aglycon phenol). Enzymatic reactions were carried outꢀ in
citrate±phosphate bu€er pH 7.2, 0.1 mM CaCl2, (A) at 55 C,
and (B) at 30 ꢀC.
Fig. 2. Hammett plot log ꢁkcat=KM† versus pKa for the 1,3±
1,4-ꢀ-glucanase-catalyzed hydrolysis of aryl glycosides 6a±h at
30 C.
ꢀ

58
A. Planas et al./Carbohydrate Research 310 (1998) 53±64
Scheme 2.
To assess this behaviour and be able to evaluate
The electronic e€ect of substituents on rates (and
equilibria) of organic reactions is formally expres-
sed by the Hammett equation:
the dependence of kcat on aglycon acidity for aryl
glycosides with good leaving groups, the kinetic
parameteꢀrs were determined at a lower tempera-
ture (30 C) to reduce spontaneous hydrolysis of
the reactive substrates, and two new aryl glycosides
with a pKa of the aglycon <7 (6e and 6g) were
included in the analysis. Kinetic data are summar-
ized in Table 4, and the corresponding Hammett
plot (log kcat versus pKa) is shown in Fig. 1(b).
Again, the same biphasic plot is observed. Sub-
strates with an aryl aglycon whose pKa of the free
phenol is higher than 7 show essentially no depen-
dence of kcat on aglycon acidity, whereas substrates
with good leaving groups (pKa of the free phe-
nol<7) have increasing kcat values as the pKa
decreases. Linear regression of the data log kcat
versus pKa for the substrates 6d±h gives a slope of
0.9. The Hammett plot for kcat=KM is also shown
in Fig. 2. Although correlation is poor, data could
be ®tted to a single line with a slope of 0.8
approximately.
log k ˆ constant ‡ ꢂÁꢅ
where k is the rate constant, ꢂ is the reaction con-
stant, and ꢅ the substituent constant [32,33]. For
the process being studied, enzymatic hydrolysis of
a series of substituted aryl glycosides, the sub-
stituent constant is often expressed by the pKa of
the free phenol, and the Hammett equation is then
written in the form of the Brùnsted equation:
log k ˆ constant ‡ ꢀÁpKa
where ꢀ is now the reaction constant re¯ecting the
sensitivity of the reaction to the electronic e€ect of
the substituents.
Previous studies with retaining glycosidases have
shown that Hammett relationships for the enzyme-
catalyzed hydrolyses of aryl glycosides of di€erent
leaving group ability fall into the following three
categories:
3. Discussion
Retaining glycosidases operate by a double-dis-
placement mechanism [28±31] as shown in Scheme
2. In the ®rst step (glycosylation) the amino acid
residue acting as a general acid protonates the gly-
cosidic oxygen with concomitant C±O cleavage of
the scissile glycosidic bond to give a glycosyl-
enzyme intermediate stabilized by the catalytic
nucleophile. The second deglycosylation step
involves the attack of a water molecule assisted by
the conjugated base of the general acid to render
the hydrolysis product with overall retention of
con®guration at the anomeric center. Both steps
proceed through transition states with oxocarbe-
nium ion character.
(a) Linear Hammett relationship with negative ꢀ
value for a wide range of leaving-group pKa
values (e.g. Cellulomonas ®mi exo-ꢀ-1,4-gly-
canase [7], neuraminidase from in¯uenza A
virus [8], Agrobacterium faecalis ꢀ-glucosi-
dase mutants [34] and Bacillus circulans
xylanase [12]). The glycosylation step invol-
ving glycosidic bond cleavage is then the rate
limiting step. A large negative ꢀ value re¯ects
a large degree of negative charge accumula-
tion on the phenolate oxygen. It indicates a
late transition state with almost complete
C±O bond breaking and relatively little pro-
ton donation. Conversely, a ꢀ value close to

A. Planas et al./Carbohydrate Research 310 (1998) 53±64
59
zero may indicate an early transition state at
the beginning of C±O bond breaking and
signi®cant proton donation.
(ꢀ ˆ 0:9), thus suggesting they monitor the
same process.
2. If substrate binding to the active site is
equivalent for all the aryl glycosides, e.g.
binding constants within the same order of
magnitude and the same binding mode in
(b) Biphasic Hammett relationship with a con-
cave-downward shape (e.g. Agrobacterium
faecalis ꢀ-glucosidase [5,35]) that reveals a
change in the rate determining step. Less
reactive substrates (high pKa of the phenol)
are leaving-group dependent with a negative
ꢀ value, for which the glycosylation step is
rate limiting as in case (a). Log kcat (or
log kcat=KM) levels o€ for substrates with
good leaving groups, and no signi®cant
dependence of their reactivity on pKa is
observed. The glycosylation step is not rate
determining any more but rather deglyco-
sylation (that is aglycon-independent)
becomes rate limiting for these more reac-
tive substrates.
(c) Complex Hammett relationships where low
correlations are obtained, the points being
badly scattered (e.g. ꢀ-galactosidase from E.
coli [6]). It indicates that electronic e€ects
alone cannot provide an explanation for the
variation of kcat with aglycone structure,
revealing a dependence on aglycone shape
rather than acidity. Also a complex beha-
viour was obtained for lysozyme, which yiel-
ded a strong biphasic convex-upward shape
where di€erent mechanisms must obtain for
electron-withdrawing and electron-donating
substituents [9,10,36].
.
terms of E S complex structure, the curved
plot log kcat versus pKa then suggests a
change in the transition state structure
depending on the aglycone. For less reactive
aryl glycosides (pKa of the phenol leaving
group>7) ꢀ is approximately zero, consistent
with an early transition state with little C±O
bond breaking and little charge development.
On the other hand, substrates with a pKa of
the phenol leaving group<7 are aglycon-
dependent and ꢀ becomes 0.9. It re¯ects a
large degree of negative charge accumulation
on the phenolate oxygen that corresponds to a
late transition state with extensive C±O bond
cleavage. In terms of proton donation to the
leaving group, however, the curved plot may
also re¯ect an increase in the degree of proton
transfer as the leaving group becomes more
basic, with the slope varying from about 0.9
with acidic leaving groups (2,4-dinitrophenyl
to 3,5-dinitrophenyl, as would be expected for
reaction with little proton donation) to zero
or even slightly positive (as is seen with the
speci®c acid-catalyzed hydrolysis of glyco-
sides). Then the situation may be really com-
plicated since both e€ects, extend of C±O
bond cleavage and proton donation may
overlap and vary themselves as leaving group
acidity changes.
The observed Hammett relationship for 1,3-1,4-
ꢀ-glucanase (Fig. 1) does not match any of the
previous categories. Based on the reported exam-
ples summarized above, the mechanistic inter-
pretation of this unusual behavior is obscure, but
several working hypothesis can be proposed for
further analysis:
3. The premise of similar binding modes for all
the aryl glycosides as stated in (ii) may not
hold, rather the substrates may position the
aryl aglycons in a di€erent conformation into
the active site cleft. Inspection of the structure
of the aglycons reveals that those having a
pKa of the free phenol<7 have a disubstituted
phenyl ring (2,4-, 2,3-, 3,4-, and 3,5-dinitro
substitution), whereas those having a pKa>7
are monosubstituted, except for the 4-methyl-
umbelliferone which also has a disubstituted
phenyl ring but as a part of a bicyclic system.
It might be possible that steric constrains give
rise to two di€erent binding modes in subsite
+I for both groups of substrates, the aglycon
being accommodated in di€erent conforma-
tions, thus giving rise to two independent
1. The glycosylation step is rate determining for
all the studied aryl glycosides; for deglycosy-
lation to become rate limiting kcat would level
o€ for activated substrates. On the other
hand, the plot logꢁkcat=KM† versus pKa (Fig. 2)
can be ®tted approximately to a single line of
slope 0.8, indicating that the ®rst irreversible
step, glycosylation, is rate-determining in
kcat=KM. This slope is very similar to that
obtained from the leaving group-dependent
part of the log kcat versus pKa plot

60
A. Planas et al./Carbohydrate Research 310 (1998) 53±64
structure±activity relationships. This is a
80% EtOH (600 mL) was vigorously stirred under
re¯ux for 12 h. The suspension was ®ltered, the
solid washed with aq 80% EtOH and dried at
room temperature. The solid thus obtained was
suspended in H2O (300 mL) and a fresh extra-
cellular medium of a recombinant E. coli culture
expressing Bacillus licheniformis 1,3-1,4-ꢀ-gluca-
nase (100 mL) was added (®nal concentration of
0.3 U/mL). After incubation in an orbital shaker
for 17 h at 55 ꢀC, the supernatant was collected and
concentrated up to 150 mL. Oligosaccharides were
precipitated stepwise with aq 95% EtOH
(2Â150 mL), and acetone (450 mL). The solids
were collected by centrifugation and dried. The
®nal crude mixture (23 g) was dissolved in 1:1
Ac2O±pyridine (350 mL) and stirred for 18 h at
room temperature. After pouring the reaction
mixture on ice-water (1.5 L), the yellow precipitate
was collected by ®ltration, washed with H2O, and
puri®ed by ¯ash chromatography (1:2!1:1
EtOAc±CHCl3), to yield 2 (6.78 g) and perace-
tylated 3-O-ꢀ-cellotriosyl-d-glucopyranose (1.24 g).
General synthesis of aryl (2,3,4,6-tetra-O-acetyl-
ꢀ-d-glucopyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-
d-glucopyranosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-
d-glucopyranosides (5a±g).ÐThe ꢁ-glycosyl bro-
mide 3 was prepared as described in [21] by treat-
ment of 2 with 4.1 M HBr in HOAc for 30 min at
room temperature. After usual work-up, the pro-
duct (98% yield) was used immediately in the next
glycosidation steps without further puri®cation.
The purity of the crude product was controlled by
TLC and 1H NMR (CDCl3; ꢃ 6.5, d, 1 H, J1,2 4 Hz
, H-1). A solution of 3 (9.86 g, 10.0 mmol) in
CH2Cl2 (100 mL) was added to the corresponding
phenol (20.0 mmol) and triethylbenzylammonium
bromide (2.30 g, 8.46 mmol) in aq 5% NaOH
(20 mL). The biphasic system was heated under
re¯ux with vigorous stirring for 4 to 5 h. The mix-
ture was diluted with H2O (40 mL), and the aqu-
eous layer extracted with CH2Cl2. The combined
organic phases were washed with aq 5% NaOH,
neutralized with diluted HCl, washed with H2O,
and dried (MgSO4). After evaporation of the sol-
vent, chromatographic separation gave 5a±g in 25±
65% yields.
highly speculative interpretation since no
structural evidence on how the leaving group
binds into subsite +IÐeither a glucose unit in
the natural substrates or an aryl aglycon in
these synthetic substratesÐhas been obtained
yet. Work is in progress to solve the 3D-
structure of a mutant enzyme±substrate com-
plex by X-ray crystallography.
The Hammett relationship here obtained repre-
sents a novel mechanistic situation for retaining
glycosidases. Whether these e€ects are a con-
sequence of the endo action of the enzyme, as endo-
glycosidases have been studied only sparingly,
remains to be evaluated. Nevertheless other
mechanistic probes (kinetic isotope e€ects, pre-
steady state kinetics, etc.) should be used to unra-
vel the ®ne details of the mechanistic pathways of
endo-glycosidases.
4. Experimental
General.ÐNMR spectra were recorded on a
Varian Gemini 300 spectrometer. 1H chemical
shifts (ꢃ) were referenced to internal Me4Si for
solutions in CDCl3, and to internal TSPNa for
solutions in D2O; 13C chemical shifts were refer-
1
enced to the solvent signal. H and 13C NMR data
for compounds 6a±h are listed in Tables 1 and 2.
TLC was performed on precoated silica gel 60 F254
plates with detection by UV light or by charring
ꢀ
with aq 50% H2SO4 and heating at 125 C for
15 min. 1,3-1,4-ꢀ-glucanase was obtained from
recombinant Escherichia coli HB101 harboring a
Bacillus licheniformis ꢀ-glucanase expression plas-
mid as reported in refs. [18,27]. The protein was
puri®ed from the culture supernatant according to
ref. [22]. Enzyme concentration was determined by
UV spectrophotometry ("280 ˆ 0:0145 ꢆg=mL=cm
[37]). Absorbance measurements and kinetics were
performed on a Varian Cary 4E spectrophotometer
with a thermostated cell holder.
(2,3,4,6-Tetra-O-acetyl-ꢀ-d-glucopyranosyl)-(1!4)-
(2,3,6-tri-O-acetyl-ꢀ-d-glucopyranosyl)-(1!3)-1,2,
4,6-tetra-O-acetyl-d-glucopyranose (2).ÐThis com-
pound was obtained by enzymatic hydrolysis of
barley ꢀ-glucan and further acetylation following a
modi®cation of the method described by Malet et
al. [21]. A suspension of barley ¯our (100 g) and aq
Phenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-glucopyrano-
syl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyranosyl)-
(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyranoside (5a).Ð
NMR: 1H (300 MHz, CDCl3), ꢃ 1.98±2.14 (10 s, 30
H, 10 Ac), 3.62±4.42 and 4.82±5.15 (H-2,20,200,30,300,
4,40,400,5,50,500, 6a,60a,600a,6b,60b,600b), 4.47 (d, 1 H,

A. Planas et al./Carbohydrate Research 310 (1998) 53±64
61
J1,2 8.0 Hz, H-100), 4.59 (d, 1 H, J1,2 8.0 Hz, H-10),
5.05 (d, 1 H, J1,2 10.0 Hz, H-1), 5.26 (dd, 1 H,
J2,3=J3,4=8.0 Hz, H-3), 6.95 (d, 2 H, J2,3 8.5 Hz,
phenyl H-2,6), 7.06 (t, 1 H, J3,4=J4,5=7.5 Hz,
phenyl H-4), 7.29 (dd, 2 H, J2,3=J3,4=8.0 Hz,
phenyl H-3,5); 13C (75 MHz, CDCl3), ꢃ 20.4±20.9
(COCH3), 61.5 and 62.2 (C-6,60,600), 67.7, 68.2,
71.1, 71.6, 72.0, 72.5, 72.6, 72.7, and 72.8 (C-
2,20,200,30,300,4,400,5,50,500), 76.2 and 78.8 (C-3,40),
99.1 (C-1), 100.8 and 100.9 (C-10,100), 116.8 (phenyl
C-2,6), 123.2 (phenyl C-4), 129.5 (phenyl C-3,5),
156.9 (phenyl C-1), 168.8±170.7 (COCH3).
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
1
side (5d).ÐNMR: H (300 MHz, CDCl3), ꢃ 1.99±
2.16 (10 s, 30 H, 10 Ac), 3.60±4.35 and 4.63±5.17
(H-2,20,200,30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b),
4.48 (d, 1 H, J1,2 8.0 Hz, H-100), 4.60 (d, 1 H, J1,2
8.0 Hz, H-10), 5.09 (d, 1 H, J1,2 7.5 Hz, H-1), 5.29
(dd, 1 H, J2,3=J3,4=8.0 Hz, H-3), 7.04 (d, 2 H, J2,3
9.6 Hz, phenyl H-2), 8.20 (d, 2 H, J2,3 9.6 Hz, phe-
nyl H-3); 13C (75 MHz, CDCl3), ꢃ 20.3±20.8
(COCH3), 61.5, 62.0, and 62.1 (C-6,60,600), 67.7,
67.9, 71.1, 71.6, 72.0, 72.1, 72.4, 72.7, 72.8, and
72.9 (C-2,20,200,30,300,4,400,5,50,500), 76.2 and 78.6 (C-
3,40), 98.0 (C-1), 100.8 and 100.9, (C-10,100), 116.5
(phenyl C-2,6), 125.7 (phenyl C-3,5), 143.2 (phenyl
C-4), 161.2 (phenyl C-1), 168.1±170.4 (COCH3).
3,5-Dinitrophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
4-Bromophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
1
side (5b).ÐNMR: H (300 MHz, CDCl3), ꢃ 1.96±
2.19 (10 s, 30 H, 10 Ac), 3.61±4.42 and 4.82±5.16
(H-2,20,200,30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b),
4.47 (d, 1 H, J1,2 8.0 Hz, H-100), 4.59 (d, 1 H, J1,2
8.0 Hz, H-10), 5.07 (d, 1 H, J1,2 10.0 Hz, H-1), 5.24
(dd, 1 H, J2,3=J3,4=8.0 Hz, H-3), 6.84 (d, 1 H, J2,3
9.0 Hz, phenyl H-2), 7.39 (d, 2 H, J2,3 9.0 Hz, phenyl
H-3); 13C (75 MHz, CDCl3), ꢃ 20.5±20.9 (COCH3),
61.5 and 62.1 (C-6,60,600), 67.7, 68.1, 71.1, 71.6, 72.0,
72.1, 72.4, 72.6, 72.7, and 72.8 (C-2,20,200,30,300,
4,400,5,50,500), 76.2 and 78.8 (C-3,40), 99.1 (C-1), 100.8
and 100.9 (C-10,100), 115.8 (phenyl C-2,6), 118.7
(phenyl C-4), 132.4 (phenyl C-3,5), 155.9 (phenyl C-
1), 168.8±170.6 (COCH3).
1
side (5e).ÐNMR: H (300 MHz, CDCl3), ꢃ 1.97±
2.21 (10 s, 30 H, 10 Ac), 3.58±4.46 and 4.82±5.17
(H-2,20,200,30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b),
4.48 (d, 1 H, J1,2 7.8 Hz, H-100), 4.60 (d, 1 H, J1,2
8.1 Hz, H-10), 5.08 (d, 1 H, J1,2 9.3 Hz, H-1), 5.33 (dd,
1 H, J2,3=J3,4=9.3 Hz, H-3), 8.15 (d, 2 H, J2,4
2.1 Hz, phenyl H-2,6), 8.77 (dd, 1 H, J2,4=J4,6
=1.8 Hz, phenyl H-4); 13C (75 MHz, CDCl3), ꢃ 20.2±
20.9 (COCH3), 61.5, 62.1, and 62.2 (C-6,60,600), 67.7,
67.8, 71.0, 71.6, 71.9, 72.0, 72.6, 72.8, 72.8, and 73.0
(C-2,20,200,30,300,4,400,5,50,500), 76.2, and 78.5 (C-3,40),
98.4 (C-1), 100.8, and 101.0 (C-10,100), 113.0 (phenyl
C-4), 117.2 (phenyl C-2,6), 149.1 (phenyl C-3,5),
157.3 (phenyl C-1), 168.5±170.7 (COCH3).
3-Nitrophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
1
side (5c).ÐNMR: H (300 MHz, CDCl3), ꢃ 1.99±
3,4-Dinitrophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
2.17 (10 s, 30 H, 10 Ac), 3.62±4.43 and 4.63±5.16
(H-2,20,200,30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b),
4.48 (d, 1 H, J1,2 8.0 Hz, H-100), 4.59 (d, 1 H, J1,2
8.0, H-10), 5.03 (d, 1 H, J1,2 7.5 Hz, H-1), 5.30
(dd, 1 H, J2,3=J3,4=8.5 Hz, H-3), 7.30 (d, 1 H,
1
side (5f).ÐNMR: H (300 MHz, CDCl3), ꢃ 1.99±
2.17 (10 s, 30 H, 10 Ac), 3.62±4.36 and 4.83±5.17
(H-2,20,200,30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b),
4.47 (d, 1 H, J1,2 9.0 Hz, H-100), 4.59 (d, 1 H, J1,2
9.0 Hz, H-10), 5.09 (d, 1 H, J1,2 9.0 Hz, H-1), 5.29 (dd,
1 H, J2,3=J3,4=8.0 Hz, H-3), 7.23 (dd, 1 H, J2,6 3.5,
J5,6 9.0 Hz, phenyl H-6), 7.36 (d, 1 H, J2,6 3.5 Hz,
phenyl H-2), 7.99 (d, 1 H, J5,6 10.0 Hz, phenyl H-5);
13C (75 MHz, CDCl3), ꢃ 20.3±20.8 (COCH3), 61.5
and 62.1 (C-6,60,600), 67.7, 67.8, 71.1, 71.9, 72.0, 72.6,
72.8, and 72.9 (C-2,20,200,30,300,4,400,5,50,500), 76.2 and
78.3 (C-3,40), 98.0 (C-1), 100.8 and 101.0 (C-10,100),
112.2 (phenyl C-2), 120.4 (phenyl C-6), 127.1 (phenyl
C-5), 136.7 (phenyl C-4), 145.0 (phenyl C-3), 159.7
(phenyl C-1), 168.7±170.5 (COCH3).
J5,6 8.0 Hz, phenyl H-6), 7.47 (dd,
1
H,
J4,5=J5,6=8.0 Hz, phenyl H-5), 7.83 (dd, 1 H,
J2,4=J2,6=3.0 Hz, phenyl H-2), 7.95 (dd, 1 H, J3,4
8.0, J2,4 3.0 Hz, phenyl H-4); 13C (75 MHz, CDCl3),
ꢃ 20.3±20.9 (COCH3), 67.7, 68.0, and 68.2 (C-
6,60,600), 71.1, 71.4, 71.6, 72.0, 72.2, 72.4, 72.6,
72.7, 72.8, and 72.9 (C-2,20,200,30,300,4,400,5,50,500),
76.2, and 78.6 (C-3,40), 98.0 (C-1), 100.8 and 101.0
(C-10,100), 110.8 (phenyl C-2), 118.1 (phenyl C-4),
123.8 (phenyl C-6), 130.2 (phenyl C-5), 149.1
(phenyl C-3), 157.0 (phenyl C-1), 168.1±170.7
(COCH3).
4-Nitrophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-
2,3-Dinitrophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-

62
A. Planas et al./Carbohydrate Research 310 (1998) 53±64
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
10), 5.35 (dd, 1 H, J2,3=J3,4=9.5 Hz, H-3), 7.48 (d,
1 H, J5,6 9.6 Hz, phenyl H-6), 8.40 (dd, 1 H, J3,5
2.4, J5,6 9.6 Hz, phenyl H-5), 8.68 (d, 1 H, J3,5
2.4 Hz, phenyl H-3); 13C (75 MHz, CDCl3), ꢃ 20.4±
20.8 (CH3), 61.5, 61.7, and 62.1 (C-6,60,600), 67.7,
67.8, 71.0, 71.3, 71.6, 72.0, 72.6, and 72.8 (C-
2,20,200,30,300,4,400,5,50,500), 76.2, and 78.2 (C-3,40),
99.1 (C-1), 100.8, and 101.0 (C-10,100), 119.6 (phenyl
C-6), 121.2 (phenyl C-3), 128.4 (phenyl C-5), 140.7
(phenyl C-2), 142.5 (phenyl C-4), 153.4 (phenyl C-
1), 168.7±170.5 (CO).
General synthesis of aryl ꢀ-cellobiosyl-ꢀ-d-
glucopyranosides (6a±h) by deacetylation of 5a±
h.Ð(a) NaOMe in MeOH deacetylation: To a stir-
red solution of the speci®c chromophoric oligo-
saccharide 5a±f (3.37 mmol) in anhydrous MeOH
(114 mL) was added 0.1 M NaOMe in MeOH
(28 mL), and the mixture was stirred in the dark for
4.5±24 h at room temperature. The solid formed
was ®ltered and washed with cold MeOH to yield
6a±f. An additional fraction of 6a±f was obtained
from the mother liquors after neutralization with
excess Amberlite IR 120 (HÈ) followed by chro-
matography, with an overall yield of 60±94%,
except for 6f (30% yield). Deacetylation of 5f to
a€ord 6f was achieved in higher yield following
method (b) (see below).
1
side (5g).ÐNMR: H (300 MHz, CDCl3), ꢃ 1.98±
2.24 (10 s, 30 H, 10 Ac), 3.56±4.42 and 4.83±5.29
(H-2,20,200,30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b),
4.47 (d, 1 H, J1,2 8.1 Hz, H-100), 4.58 (d, 1 H, J1,2
8.1 Hz, H-10), 5.00 (d, 1 H, J1,2 9.3 Hz, H-1), 5.26 (dd,
1 H, J2,3=J3,4=9.0 Hz, H-3), 7.62 (dd, 1 H,
J4,5=J5,6=8.1 Hz, phenyl H-5), 7.71 (d, 1 H, J5,6
7.5 Hz, phenyl H-6), 7.99 (d, 1 H, J4,5 8.1 Hz, phenyl
H-4); 13C (75 MHz, CDCl3), ꢃ 20.4±20.7 (COCH3),
61.5, 61.7, and 62.1 (C-6,60,600), 67.7, 67.8, 71.0, 71.5,
71.6, 72.0, 72.7, 72.7, 72.7, and 72.9 (C-2,20,
200,30,300,4,400,5,50,500), 76.1 and 78.2 (C-3,40), 100.8 (C-
1), 101.0 and 101.0 (C-10,100), 120.2 (phenyl C-4),
126.4 (phenyl C-6), 130.9 (phenyl C-2), 137.2 (phenyl
C-5), 140.0 (phenyl C-3), 148.6 (phenyl C-1), 168.9±
170.5 (COCH3).
(2,3,4,6-Tetra-O-acetyl-ꢀ-d-glucopyranosyl)-(1!4)-
(2,3,6-tri-O-acetyl-ꢀ-d-glucopyranosyl)-(1!3)-2,4,6-
tri-O-acetyl-d-glucopyranose (7).ÐA solution of per-
acetylated trisaccharide 2 (6.00 g, 6.21 mmol) in
DMF (25 mL) was treated ꢀwith hydrazine acetate
(720 mg, 7.82 mmol) at 50 C with stirring during
8 min. The reaction mixture was diluted with
EtOAc (200 mL), washed with aq satd NaCl and
H2O, and dried (MgSO4). Evaporation of the sol-
vent a€orded 7 as a white solid (4.83 g, 97%).
NMR: 1H (300 MHz, CDCl3), ꢃ 1.99±2.14 (10 s, 30
H, 10 Ac), 3.62±5.26 (H-2,20,200,30,300,4,40,400,5,50,
500,6a,60a,600a,6b,60b,600b), 4.54 (d, 1 H, J1,2 7.5 Hz,
H-100), 4.63 (d, 1 H, J1,2 7.5 Hz, H-10), 5.14 (d, 1/5
H, J1,2 9.3 Hz, H-1ꢀ), 5.37 (d, 4/5 H, J1,2 3.6 Hz, H-
1ꢁ), 5.51 (dd, 1 H, J2,3=J3,4=9.6 Hz, H-3).
(b) HCl in MeOH deacetylation: A solution of
compounds 5f±h (3.20 mmol) in anhydrous MeOH
(100 mL) was treated in the dark with 26% HCl in
anhydrous MeOH (75 mL) under Ar at 4 ꢀC. After
16 h, a new aliquot of HCl in MeOH (25 mL) was
added, and the system was stirred for 2 h. The sol-
ꢀ
2,4-Dinitrophenyl (2,3,4,6-tetra-O-acetyl-ꢀ-d-gluco-
pyranosyl)-(1!4)-(2,3,6-tri-O-acetyl-ꢀ-d-glucopyr-
anosyl)-(1!3)-2,4,6-tri-O-acetyl-ꢀ-d-glucopyrano-
side (5h).ÐCompound 7 (5.18 g, 6.44 mmol) and
diazabicyclo[2.2.2]octane (2.53 g, 22.43 mol) in
anhydrous DMF (75 mL) were stirred for 3 h under
Ar in the presence of 4 A molecular sieves. 1-
Fluoro-2,4-dinitrobenzene (1.59 g, 8.27 mmol) was
added, and the mixture was stirred for 18 h. The
mixture was ®ltered through Celite, CHCl3
(200 mL) was added, and the liquid was distilled
o€. The resulting syrup was dissolved into CHCl3,
washed with aq NaHCO3, dried (MgSO4), and the
solvent was evaporated. The solid was puri®ed by
column chromatography to yield 5h (3.54 g, 50%).
NMR: 1H (300 MHz, CDCl3), ꢃ 1.98±2.22 (10 s, 30
H, 10 Ac), 3.59±4.43, and 4.83±5.16 (H-2,20,200,
30,300,4,40,400,5,50,500,6a,60a,600a,6b,60b,600b), 4.48 (d, 1
H, J1,2 8.0 Hz, H-100), 4.60 (d, 1 H, J1,2 8.5 Hz, H-
vent was distilled o€ at low temperature (<20 C)
and the resulting syrup was treated with Et2O. The
solid was washed twice with cold acetone and
crystallized from MeOH-acetone to a€ord 6f±h
(50±70% yields).
1H and 13C NMR data for compounds 6a±h are
given in Tables 1 and 2, respectively.
Phenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-d-gluco-
pyranosyl-(1!3)-ꢀ-d-glucopyranoside (6a).Ðmp
ꢀ
(MeOH) 235±250 C (dec.); [ꢁ]2d0 35.5 (c 0.0984,
water). Anal. Calcd for C24H36O16Â2H2O: C,
46.75; H, 6.54. Found: C, 46.29; H, 6.49.
4-Bromophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-d-
glucopyranosyl-(1!3)-ꢀ-d-glucopyranoside (6b).Ð
ꢀ
mp (MeOH) 215±220 C; [ꢁ]2d0 n.d. Anal. Calcd
for C24H35O16Br1Â4H2O: C, 39.52; H, 5.67.
Found: C, 39.71; H, 5.32.
3-Nitrophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-d-
glucopyranosyl-(1!3)-ꢀ-d-glucopyranoside (6c).Ð

A. Planas et al./Carbohydrate Research 310 (1998) 53±64
63
ꢀ
mp (MeOH) 233±234 C; [ꢁ]2d0 43.4ꢀ (c 0.09675,
slope Abs versus time after correction for blank
rates (non-enzymatic hydrolysis). Kinetic para-
meters were calculated by ®tting the data to eq (1)
by non-linear regression.
water). Anal. Calcd for C24H35O18N1Â4H2O: C,
41.34; H, 6.21; N, 2.01. Found: C, 41.34; H, 6.08;
N, 1.90.
4-Nitrophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-d-
glucopyranosyl-(1!3)-ꢀ-d-glucopyranoside (6d).Ð
ꢀ
mp (EtOH±H2O) 187±189 C (dec.); [ꢁ]2d0 35.54ꢀ
Acknowledgements
(c 0.110, water). Anal. Calcd for C24H35O18N1-
Â4H2O: C, 41.34; H, 6.21; N, 2.01. Found: C,
40.92; H, 5.98; N, 1.95.
The authors gratefully acknowledge Professor
Michael Sinnott (UMIST, Manchester, UK) for
helpful comments on Hammett analysis. This work
was supported by Grant BIO97-0511-C02 from
Comision Interministerial de Ciencia y Tecnologõa
(CICYT), Madrid (Spain). M.A. acknowledges a
predoctoral fellowship from the Institut Quõmic
de Sarria, and J.-L.V. is a recipient of a pre-
doctoral fellowship from DGU, Generalitat de
Catalunya.
3,5-Dinitrophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-
d-glucopyranosyl-(1!3) - ꢀ-d-glucopyranoside (6e).Ð
ꢀ
mp (MeOH) 163±165 C (dec.); [ꢁ]2d0 51.43ꢀ (c
0.101, water). Anal. Calcd for C24H34O20N2-
Â3H2O: C, 39.78; H, 5.56; N, 3.87. Found: C,
39.34; H, 5.48; N, 3.63.
3,4-Dinitrophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-
d-glucopyranosyl-(1!3)-ꢀ-d-glucopyranoside (6f).Ð
ꢀ
mp (MeOH) 180±185 C (dec.); [ꢁ]2d0 36.8ꢀ (c
0.09785, water). Anal. Calcd for C24H34O20N2-
Â3H2O: C, 39.78; H, 5.56; N, 3.87. Found: C,
39.58; H, 5.09; N, 3.48.
References
2,3-Dinitrophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-
[1] H. van Tilbeurgh, M. Claeyssens, and C.K.
de Bruyne, FEBS Lett., 149 (1982) 152±156.
[2] M. Claeyssens, H. van Tilbeurgh, J.P. Kamerling,
J. Breg, M. Vrsanska, and P. Biely, Biochem. J.,
270 (1990) 251±256.
[3] C.G. Cupples, J.H. Miller, and R.E. Huber, J.
Biol. Chem., 265 (1990) 5512±5518.
[4] P. Tomme, S. Chavaux, P. Beguin, J. Millet,
J.P. Aubert, and M. Claeyssens, J. Biol. Chem., 266
(1991) 10313±10318.
[5] J.B. Kempton and S.G. Withers, Biochemistry, 31
(1992) 9961±9969.
[6] M.L. Sinnott and I.J.L. Souchard, Biochem. J., 133
(1973) 89±98.
[7] D. Tull and S.G. Withers, Biochemistry, 33 (1994)
6363±6370.
[8] X. Guo, G. Laver, E. Vimr, and M.L. Sinnott, J.
Am. Chem. Soc., 116 (1994) 5572±5578.
[9] G. Lowe, G. Sheppard, M.L. Sinnott, and
A. Williams, Biochem. J., 104 (1967) 893±899.
[10] C.S. Tsai, J.Y. Tang, and S.C. Subbarao, Biochem.
J., 114 (1969) 529±534.
[11] S.L. Lawson, W.W. Wakarchuk, and S.G.
Withers, Biochemistry, 35 (1996) 10110±10118.
[12] S.L. Lawson, W.W. Wakarchuk, and S.G.
Withers, Biochemistry, 36 (1997) 2257±2265.
[13] F.W. Parrish, A.S. Perlin, and E.T. Reese, Can. J.
Chem., 38 (1960) 2094±2104.
d-glucopyranosyl-(1!3)-ꢀ-d-glucopyranoside (6g).Ð
ꢀ
mp (MeOH) 169±171 C (dec.); [ꢁ]2d0 50.48ꢀ (c
0.105, water). Anal. Calcd for C24H34O20N2ÂH2O:
C, 41.87; H, 5.27; N, 4.07. Found: C, 41.43; H,
5.02; N, 4.13.
2,4-Dinitrophenyl ꢀ-d-glucopyranosyl-(1!4)-ꢀ-
d-glucopyranosyl-(1!3)-ꢀ-d-glucopyranoside (6h).Ð
ꢀ
mp (MeOH±MeCOMe) 225±228 C (dec.); [ꢁ]d20
65.4ꢀ (c 1.000, water). Anal. Calcd for C24H34
.
O20N2 1/2 H2O: C, 42.40; H, 5.19; N, 4.12. Found:
C, 42.41; H, 5.30; N, 4.01.
Enzyme kinetics.ÐAbsortivity coecients of the
chromophoric substrates and aglycons were deter-
mined at the appropriate wavelength under the
same conditions (temperature, bu€er, pH and ionic
strength) used in the enzyme kinetic experiments
[wavelength and Á" (ˆ "phenol
"substrate) for 6a±i
ꢀ
are listed in Table 3 (at 55 C) and Table 4 (at
30 ꢀC)]. Kinetic constants for the enzyme-catalyzed
hydrolysis of 6a±i were determined by measuring
the release of the aglycon [23]. All experiments
ꢀ
were run at 55 or 30 C in 6.5 mM citrate±87 mM
phosphate bu€er (pH 7.2) in the presence of
0.1 mM CaCl2 at substrate concentrations ranging
from 0.25ÂKM to 8.5ÂKM. The reactions were
started by addition of enzyme to a 5 min pre-
incubated mixture of bu€er, CaCl2 and substrate in
the case of 6a±f,i, or by addition of substrate in the
case of 6h. Initial velocities were derived from the
[14] M.A. Anderson and B.A. Stone, FEBS Lett., 52
(1975) 202±207.
[15] C. Malet, J. Jimenez-Barbero, M. Bernabe, C. Brosa,
and A. Planas, Biochem. J., 296 (1993) 753±758.

64
A. Planas et al./Carbohydrate Research 310 (1998) 53±64
[16] T. Godfrey, On comparison of key characteristics
of industrial enzymes by type and source. In T.
Godfrey and J. Reinchelt (Eds.), Industrial Enzy-
mology, Macmillan, London, 1983, p 466.
[17] N.A. Dierick, Arch. Animal Nutr. (Berlin), 39
(1989) 241±261.
[18] M. Juncosa, J. Pons, T. Dot, E. Querol, and
A. Planas, J. Biol. Chem., 269 (1994) 14530±14535.
[19] C. Malet, J.L. Viladot, A. Ochoa, B. Gallego,
C. Brosa, and A. Planas, Carbohydr. Res., 274
(1995) 285±301.
[26] H.J. Koeners, A.J. de Kok, C. Romers, and
J.H. van Boom, Recl. Trav. Chim. Pays-Bas, 99
(1980) 355±362.
[27] A. Planas, M. Juncosa, J. Lloberas, and E. Querol,
FEBS Lett., 308 (1992) 141±145.
[28] D.E. Koshland, Biol. Rev., 28 (1953) 416±436.
[29] M.L. Sinnott, Chem. Rev., 90 (1990) 1171±1202.
[30] B. Svensson and M. Sùgaard, J. Biotechnol., 29
(1993) 1±37.
[31] J.D. McCarter and S.G. Withers, Curr. Opin.
Struct. Biol., 4 (1994) 885±892.
[20] A. Planas and C. Malet, Contribution of subsites
to catalysis and speci®ty in the extended binding
deft of Bacillus 1,3,-1,4-ꢀ-d-glucan 4-glucano-
hydrolases. In S.B. Petersen, B. Svensson, and S.
Pedersen (Eds.), Carbohydrate Bioengineering,
Elsevier, Amsterdam, 1995, pp 85±95.
[21] C. Malet, J. Valles, J. Bou, and A. Planas, J. Bio-
technol., 48 (1996) 209±219.
[22] J. Pons, E. Querol, and A. Planas, J. Biol. Chem.,
272 (1997) 13006±13012.
[32] D.D. Johnson, The Hammett Equation, Cambridge
University Press, New York, 1973.
[33] C. Hansch, A. Leo, and R.W. Taft, Chem. Rev., 91
(1991) 165±195.
[34] S.G. Withers, K. Rupitz, D. Trimbur, and
R.A.J. Warren, Biochemistry, 31 (1992) 9979±
9985.
[35] J.C. Gebler, D.E. Trimbur, R.A.J. Warren,
R. Aebersold, M. Namchuk, and S.G. Withers,
Biochemistry, 34 (1995) 14547±14553.
[23] C. Malet and A. Planas, Biochemistry, 36 (1997)
13838±13848.
[24] D. Dess, H.P. Kleine, D.V. Weinberg, R.J.
Kaufman, and R.S. Sidhu, Synthesis, (1981) 883±885.
[25] F. Ballardie, B. Capon, and J.D.G. Sutherland, J.
Chem. Soc. Perkin I, (1973) 2418±2419.
[36] J.F. Kirsch, Linear free energy relationships in
enzymology, in N.B. Chapman and J. Shorter
(Eds.), Advances in Linear Free Energy Relation-
ships, Plenum, 1972, pp 369±400.
[37] J. Lloberas, E. Querol, and J. Bernues, Appl.
Microbiol. Biotechnol., 29 (1988) 32±38.