Báo cáo khoa học: Substrate recognition by three family 13 yeast a-glucosidases Evaluation of deoxygenated and conformationally biased isomaltosides pptx

and Birte Svensson’ ' Department of Chemistry, Carlsberg Laboratory, Copenhagen Valby, Denmark; * Department of Chemistry, University of Alberta, Edmonton, Canada Important hydrogen bon

Substrate recognition by three family 13 yeast «-glucosidases

Evaluation of deoxygenated and conformationally biased isomaltosides

Torben P Frandsen'*, Monica M Palcic? and Birte Svensson’

' Department of Chemistry, Carlsberg Laboratory, Copenhagen Valby, Denmark; * Department of Chemistry,

University of Alberta, Edmonton, Canada

Important hydrogen bonding interactions between substrate

OH-groups in yeast «-glucosidases and oligo-1,6-glucosidase

from glycoside hydrolase family 13 have been identified by

measuring the rates of hydrolysis of methyl o-isomaltoside

and its seven monodeoxygenated analogs The transition-

state stabilization energy, AAGt, contributed by the indi-

vidual OH-groups was calculated from the activities for

the parent and the deoxy analogs, respectively, according

to AAGt = -kT In[(max/ Km )analog/ (Vmax! Km)parentl- This

analysis of the energetics gave AAG? values for all three

enzymes ranging from 16.1 to 24.0 kJ-‘mol™ for OH-2, -3’,

-4’, and -6’, i.e the OH-groups of the nonreducing sugar ring

These OH-groups interact with enzyme via charged hydro-

gen bonds In contrast, OH-2 and -3 of the reducing sugar

contribute to transition-state stabilization, by 5.8 and

4.1 kJ-mol™', respectively, suggesting that these groups

participate in neutral hydrogen bonds The OH-4 group is

found to be unimportant in this respect and very little or no contribution is indicated for all OH-groups of the reducing- end ring of the two o-glucosidases, probably reflecting their exposure to bulk solvent The stereochemical course of hydrolysis by these three members of the retaining family 13 was confirmed by directly monitoring isomaltose hydrolysis using 'H NMR spectroscopy Kinetic analysis of the hydrolysis of methyl 6-S-ethyl-o-isomaltoside and its 6-R- diastereoisomer indicates that a-glucosidase has 200-fold higher specificity for the S-isomer Substrate molecular rec- ognition by these a-glucosidases are compared to earlier findings for the inverting, exo-acting glucoamylase from Aspergillus niger and a retaining o-glucosidase of glycoside hydrolase family 31, respectively

Keywords: protein-carbohydrate interaction; NMR; glyco- sidase mechanism; substrate analogs; molecular recognition

Strong intermolecular hydrogen bonds are very important

in specificity of enzymes and other proteins that metabolize

or bind carbohydrates [1-6] Substrate analogs such as

deoxygenated sugars, facilitate identification of critical

contacts and enable quantification of the energetics of the

protein-carbohydrate binding at the level of individual

interacting sugar OH-groups and functional atoms or

groups in the protein [4,7-11] Alternatively, site-specific

mutants of a protein are useful in evaluation of specific

protein—carbohydrate interactions and further insight has

been gained by combining mutant enzymes and analogs

[7,9,10] The binding energy contributed by substrate

OH-groups has been determined for only a few carbohy-

drate active enzymes Of these, the starch hydrolase

glucoamylase from Aspergillus niger has been the most

intensively examined [7,9—13]

Correspondence to B Svensson, Department of Chemistry, Carlsberg

Laboratory, DK-2500 Copenhagen Valby, Denmark;

Fax: + 45 33 27 47 08; Tel: + 45 33 27 53 45;

E-mail: bis@)crc.dk

Enzymes: o-glucosidase (o-p-glucoside glucohydrolase, EC 3.2.1.20);

oligo-1,6-glucosidase (dextrin 6-o-glucanohydrolase, EC 3.2.1.10);

glucoamylase (o-p-glucan glucohydrolase, EC 3.2.1.3)

* Present address: Pantheco, Boge Allé, DK 2970 Horsholm Denmark

Dedication: this paper is dedicated to Prof Joachim Thiem on the

occasion of his 60" birthday

(Received 12 October 2001, revised 26 November 2001, accepted 30

November 2001)

Three-dimensional structures of protein—carbohydrate complexes can guide and support protein engineering and molecular recognition experiments For family 13 glycoside hydrolases, there are no crystal structures for a-glucosidases; however, the structure of free Bacillus oligo-1,6-glucosidase has been solved [14] Furthermore, only a few a-glucosidases are produced by heterologous gene expression, which is a prerequisite for structure—function relationship investiga- tions by site-directed mutagenesis [15—21] While the yeast genome is known and thus the primary structures of its a-glucosidases, the sequenced strain of Saccharomyces cerevisiae is not necessarily identical to the baker’s yeast used as a source of enzymes in the present study and sequences have not been reported for brewer’s yeast enzymes In view of this limited information, use of synthetic substrate analogs is particularly attractive for gaining knowledge of the nature and strength of substrate— a-glucosidase interactions Thus using deoxy-analogs key polar groups in maltose were identified for high pI barley a-glucosidase of glycoside hydrolase family 31 to be OH-4’ and -6’ with minor contributions for OH-3’, -2’, and -3 [13, 22]

Yeast o-glucosidase and oligo-1,6-glucosidase are exo- acting glycoside hydrolases catalyzing release of a-p-glucose from nonreducing ends of various o-linked substrates The enzymes are further subclassified into type I, hydrolysing heterogeneous substrates like aryl glucosides and sucrose more efficiently than maltose; type II being highly active on maltose and isomaltose but of low activity toward aryl glucosides; and type III resembling type I, but hydrolysing

di- and oligo-saccharides and starch at comparable rates

[23,24] The sequence classifies a-glucosidases in glycoside

hydrolase families 13 and 31 [25-27] Yeast o-glucosidases

and oligo-1,6-glucosidase belong to family 13 and are

of type I that prefers p-nitrophenyl-o-p-glucopyranoside

[28]

Glycoside hydrolase family 13 (or ‘the a-amylase family’)

currently comprises 28 specificities of amylolytic and related

enzymes Several crystal structures of enzyme-inhibitor

complexes highlight active sites created by B — « segments

in catalytic (B/a)g barrel domains (reviewed in [29-31))

Because no ligand complex is available of oligo-1,6-

glucosidase, the only structure-determined exo-acting

a-glucosidase [14], side-chains participating in substrate

binding and catalysis are solely identified by sequence

comparison Clearly o-glucosidases lack the sequence motif

in B > loop 4 of family 13 [30] containing residues

binding substrate at subsite + 2 (nomenclature as in [32]) in

a-amylases, cyclodextrin glycosyltransferases, and related

enzymes [30,33-37]

In this study, seven monodeoxygenated isomaltosides are

used to map substrate OH-groups required by yeast

a-glucosidases and oligo-1,6-glucosidase in hydrolysis of

the o-1,6-glucosidic bond The energy contributed by each

OH-group for transition-state stabilization reflects the

strength of a specific protein—-carbohydrate contact and

energy profiles for the o-glucosidases and oligo-1,6-glucosi-

dase are compared 'H-NMR spectroscopy was used to

confirm that all enzymes hydrolyse isomaltose with reten-

tion of anomeric configuration characteristic of family 13

(reviewed in [38))

The exo-acting glucoamylase similarly to the o-glucosid-

ases catalyses the releases of glucose from the nonreducing

ends of substrates, but with inversion of the anomeric

configuration [39] While glucoamylase prefers the R-isomer

of isomaltose diastereoisomeric analogs [40], a-glucosidase

in the present study selects methyl 6-S-ethyl-o-isomaltoside

in preference to the R-isomer Molecular recognition of

isomaltosides is more similar for the yeast «-glucosidase and

oligo-1,6-glucosidase of glycoside of hydrolase family 13

when compared to that of glucoamylase of glycoside

hydrolase family 15 or of a type I a-glucosidase from the

retaining glycoside hydrolase family 31 [9,10,40,41]

MATERIALS AND METHODS

Enzymes and substrates

Oligo-1,6-glucosidase from baker’s yeast (EC 3.2.1.10;

Lot no 23H8080), and oglucosidases from brewer’s

(EC 3.2.1.20; Type VI; Lot no 21F8105) and baker’s

(EC 3.2.1.20; Type I; Lot no 122H8000) yeast were

obtained from Sigma After dissolution in 50 mm phosphate

pH 6.8 (a-glucosidases) or 50 mm sodium maleate pH 6.8

(oligo-1,6-glucosidase) followed by extensive dialysis at 4 °C

against these buffers, the different enzymes (oligo-1,6-

glucosidase, 30 UmL™'; brewer’s yeast o-glucosidase,

200 U-mL"!; baker’s yeast o-glucosidase, 61 U-mL™') were

used without further purification in the kinetic and stereo-

chemical studies One unit is defined as the amount of

enzyme required to liberate | umol of glucose from

p-nitrophenyl o-p-glucoside (Sigma) per min at 30 °C The

synthesized methyl o-isomaltoside, seven monodeoxy-

genated methyl o-isomaltosides [42], methyl 6-R-C-ethyl- and methyl 6-S-C-ethyl-o-isomaltoside [41] were generous gifts of U Spohr and the late R U Lemieux, University of Alberta, Edmonton, Canada

Enzyme assays a-Glucosidase activity was determined at 30 °C in 0.1 m sodium maleate, pH 6.8 (oligo-1,6-glucosidase) or 50 mm phosphate, pH 6.8 (a-glucosidases) Glucose [7,10,40,41] was analysed for analogs at the reducing end ring (reaction volume 100 uL), aliquots (15 wL) being transferred at regular time intervals to microtiter plate wells already containing quench solution (200 nL 1™ Tris, pH 7.6,

5 UmL" glucose oxidase (A niger), 1 U-mL"' peroxidase (horseradish), and 0.21 mgmL™! o-dianisidine) Absor- bances were read at 450 nm after 1 h incubation at room temperature using a microtiter plate reader (Ceres UV900Hdi, Bio-Tek), and quantified using p-glucose as a standard [22,40] Deoxygenated glucose analogs were ana- lysed essentially as described [10,40,41] with substrate analogs at the nonreducing end sugar (reaction volume

400 pL) aliquots (100 tL) were transferred to quench buffer containing 60 UmL™ glucose oxidase, 1 UmL™! peroxi- dase, and 0.1 mgmL*! o-dianisidine, and the absorbances were read at 450 nm after 4 h incubation at room temperature, and quantified using the relevant deoxygenated D-glucose as standard The a-glucosidase catalyzed hydro- lysis was initiated by addition of 0.1-91 U enzyme The limited amounts of deoxygenated analogs available allowed only determination of second-order rate constants, Vinax/Km (s“*UTĐ = y,/E,S,, where v, is the initial rate of hydro- lysis, S, the initial substrate concentration, and E, the amount of enzyme in U Two S, concentrations of around 0.1 x Km were used to ensure that substrate hydrolysis was linear with time The increase in activation energy due

to substrate deoxygenation was calculated by AAGT =

—RTn[(Vmax/Km)al(Vmax/Km)pl [43], where a refers to ana- log and b to parent substrate For the two diastereoisomers, Vmax and K,, were determined by fitting initial rates at eight different substrate concentrations from 0.1 x K,, to4 x Kn

to the Michealis—Menten equation essentially as described previously [40]

Reaction stereochemistry Lyophilized enzymes were redissolved in 0.1 mM sodium phosphate pH 6.8 in D,O and the stereochemistry of isomaltose hydrolysis was determined by 'H NMR at

310 K using a Bruker AMX-600 spectrometer operated at

600 MHz After recording the substrate spectrum of

100 mm isomaltose (in 600 uL 0.1 M phosphate, pH 6.8,

in DO), enzyme was added (oligo-1,6-glucosidase, 40 U; baker’s, 135 U and brewer’s yeast o-glucosidase, 140 U) and reactions monitored by recording spectra at regular intervals

RESULTS AND DISCUSSION

Energetics of deoxy isomaltoside hydrolysis Vmax/Km Values for hydrolysis of methyl o-isomaltoside are comparable for the three enzymes, the œ-plucosidases from

Table 1 Specificity constants and AAGt* (kJ mol”) for a-glucosidase catalyzed hydrolysis of methyl a-isomaltoside and a series of mono-deoxy- genated analogs

Oligo-1,6-glucosidase? a-glucosidase (brewer’s yeast) a-Glucosidase (baker’s yeast)*

V max! Km (S*U~’) AAGt Vimax/Km (SU) AAGt Vinax/Km (SU) AAGt Methyl-œ-isomaltoside 14x 107 + 0.8 x 10°* 44x10°+429x10° - 23x107+3.2x10° -— 2-Deoxy-methyl-a-isomaltoside 16x10°+409x10° 58 3.7x10°4+49x10° O05 97x10°458x10° 243 3-Deoxy-methyl-a-isomaltoside 2.9x 10° +06x10° 41 60x10°+14x10° -08 1413x1074 13x10 15 4-Deoxy-methyl-œ-isomaltosde 114x107 +1.1x10° -— 57x10° 4£08x10° -0.7 24x10'2+ 20x10” -0.1 2’-Deoxy-methyl-a-isomaltoside 4.1 x 10°° 455x107 21.5 15x10°+ 12x107 21.2 26x10 8+ 30x107 24.0 3’-Deoxy-methyl-o-isomaltoside 1.1.x 1077+ 65x10 18.9 2.7x10%°+53x107 196 2.7x10%° +3110 23.9 4’-Deoxy-methyl-a-isomaltoside 10x10°7+1.1«107° 191 1.5x10°+44x107'° 21.2 32x10° 428x107? 23.5 6’-Deoxy-methyl-a-isomaltoside 3.2x1077 42.1x10° 161 79x10°+47x10? 167 114x107 41.5x10°% 19.6

4 AAGL = —RT In[(Vinax/Km)al(V max! Km)p] [43], where a and b refer to analog and parent substrate, respectively; ° At 30 °C using 0.1 mM sodium maleate, pH 6.8; © At 30 °C using 50 mm phosphate, pH 6.; ¢ At 30 °C using 50 mm phosphate, pH 6.8; © standard deviation

brewer’s and baker’s yeast showing 31 and 164% of the

activity of the oligo-1,6-glucosidase, respectively (Table 1)

Furthermore, the activity of the three enzymes was reduced

by roughly the same extent, 1.e 440—3400-, 560—2900-, and

1350—-8800-fold by substrate deoxygenation at OH-2’, -3’,

-4’, or -6’ (Table 1) The losses in activity compared to the

parent substrate for all enzymes were smallest for the

6’-deoxy analog and largest for the 2’-deoxy analog, while

intermediary losses in activity for 3’- and 4’-deoxy analogs

did show small variations among the enzymes (Table 1)

For the two a-glucosidases, deoxygenation at the reducing

end ring of the substrate had no effect or a very minor effect,

the activity varying relative to the parent substrate by

factors of 0.84-1.4 and 0.42-1.0 for the enzymes from

brewer’s and baker’s yeast, respectively In contrast, for

oligo-1,6-glucosidase, the deoxy-2, -3, and -4 analogs

showed ninefold, fivefold, and no reduction in Viy4,,/Kn,

respectively (Table 1)

The AAG?T calculated from the V,,,,/K,, values deter-

mined for a given analog and the parent substrate,

respectively, indicated the energy contributed to transition-

state stabilization by corresponding the OH-group Because

AAG? for the four deoxy-analogs at the nonreducing sugar

ring, that binds to the enzymes at subsite —1, was in the

range 16.1-24.0 kJ-mol”' for the three enzymes (Table 1),

the removal of one of the OH-groups from this ring

dramatically affected substrate hydrolysis These

OH-groups can therefore be considered key polar groups

and most likely interact with charged residues on the

proteins [44] (Fig 1) At the reducing end ring, however,

AAG values of 4+6 kJ]mol ' for oligo-l,6-glucosidase

(Table 1) were obtained by replacement of the OH-2

and -3 groups, respectively, suggesting that these

OH-groups participate in neutral hydrogen bonds with the

enzyme (Fig 1) The OH-4 did not seem important in

substrate binding and hydrolysis

The three-dimensional structure of oligo-1,6-glucosidase

from Bacillus cereus [14], 1s currently the only available

structure of any type of a-glucosidases This enzyme has an

N-terminal (B/o)g barrel common to glycoside hydrolase

family 13 [30,31], a domain B that protrudes from the barrel

Bstrand 3, and a C-terminal Greek key motif Moreover

several extra-barrel secondary structure elements occur in

the segments that connect the B strands to the « helices of the

(B/o)g barrel fold [14] The catalytic site is located at the

bottom of a cleft between domain B and several of the

3 — « connecting segments [14,30] The molecular recog- nition of isomaltose analogs described above indicate very strong interaction of the nonreducing substrate ring at the enzyme subsite —1, most probably with charged side chains,

as a major driving force for stabilization of the enzyme— substrate transition-state Several of the side-chains inter- acting at subsite —1 will belong to the consensus sequence motifs containing catalytic acids, transition-state stabilizing histidines, and structurally important arginine and aspartate side chains [30]

While protein—substrate contacts at subsite —1 provide major binding energy, the distribution and strength of intermolecular hydrogen bonds involving the aglycon moiety and subsite +1, as well as subsites beyond subsite + 1 in type HI o-glucosidases, exhibit substrate specificity variation among the œ-elucosidases The yeast œ-plucosid- ases as reported here only show protein-carbohydrate hydrogen bonding involving subsite —1, and no sugar OH-groups associated stabilization energy was critical for

accommodation at subsite +1 As shown in Table 2, these

œ-ølucosidases that do not require hydrogen bonding to the

Charged

OH

Charged «#——— HO 0 Nguy I4

Charged «ầ———— HO

Neutral@

Charged

OH OMe

Fig 1 Schematic representation of proposed intermolecular hydrogen bond interactions between isomaltose and o-glucosidases from baker’s and brewer’s yeast and from oligo-1,6-glucosidase from baker’s yeast

“, only for oligo-1,6-glucosidase Invariant glycoside hydrolase family

13 side chain candidates of interaction with the four nonreducing substrate ring OH-groups are described in detail in a recent review [30].

Table 2 Kinetic parameters and AAG? for hydrolysis of isomaltose and p-nitrophenyl-c-p-glucopyranoside, and mono-deoxy analogs of methyl a-isomaltoside at binding subsite + 1 by a-glucosidases and glucoamylase

a-Glucosidase (Brewer’s yeast)?

p-Nitrophenyl-a-p-glucopyranoside 135 0.2 677

OH-2

OH-3

OH-4

Oligo-1,6-glucosidase*

p-Nitrophenyl-a-p-glucopyranoside 129

OH-2

OH-3

OH-4

Glucoamylase”

p-Nitrophenyl-a-p-glucopyranoside 0.50

OH-2

OH-3

OH-4

0.5

—0.8

—0.7

5.8 4.1

19.8 0.021

1.1 8.6 16.5

4 Data from Table 1; © [28]; © [7,51]

substrate aglycon also have much higher activity for

p-nitrophenyl-o-b-glucopyranoside, which lacks hydrogen

bonding groups in the aglycon, than for isomaltose Due to

effects on both k,,, and K,, yeast a-glucosidase thus has

4500-fold lower k ;/K,, for isomaltose than for p-nitrophe-

nyl-o-b-glucopyranoside, p-nitrophenol being also a better

leaving group than glucose Structural elements of the

nonsugar aglycon, however, were not explored It 1s

conceivable, however, that such specificity exists and could

be investigated using a series of synthetic substrates In

contrast, the activity of oligo-1,6-glucosidase significantly

depends on aglycon interactions at subsite +1 via neutral

hydrogen bonds with glucose OH-2 and -3 (Table 1) That

such protein interactions with sugar OH-groups are impotr-

tant for this enzyme 1n contrast to the two o-glucosidases 1s

also emphasized by the 225-fold difference in the value of

the relative specificity p-nitrophenyl o-glucoside/isomaltose,

(Keat/Km)/(Keat/Km), being 4500 for the brewer’s o-glucosi-

dase (which was chosen because it has the smallest

requirement for OH-groups at subsite +1; see Table 1),

and 20 for oligo-1,6-glucosidase (Table 2) Moreover, the

30-fold more favorable specificity constant (Keat/Km) for

isomaltose for the oligo-1,6-glucosidase over the a-glucosi-

dase indeed reflects the genuine specificity of the former

enzyme for exo-action on the o-1,6-linkage

Remarkably, barley o-glucosidase of glycoside hydrolase

family 31 has a completely different pattern for hydrolysis of

monodeoxy maltoside analogs which indicated strong

protein—substrate interactions at OH-4’ and -6’ and weaker,

probably neutral hydrogen-bonds with maltose OH-2’, -3’,

and -3 [22] An even stronger requirement for protein—

isomaltose aglycon interactions was found in 1somaltose

hydrolysis by glucoamylase, which depended on enzyme—

substrate transition-state interactions with OH-4 and -3 of

an energy of 16.5 and 8.6 kJ-‘mol", respectively (Table 2:

[7]) Glucoamylase thus has only sixfold lower k 4/K,,, for

isomaltose than for p-nitrophenyl-o-p-glucopyranoside

(Table 2) The substrate specificity differences and varia- tions in aglycon-protein contacts with the two o-glucosid- ases, the oligo-1,6-glucosidase, and glucoamylase emphasize that these enzymes display different geometry for the binding interactions with polar groups of substrates at subsite + 1 This will be investigated further 1n a study of the diastereo- isomer specificity of isomaltoside hydrolysis (see below) Catalytic mechanism

One feature of the disposition of substrate relative to enzyme during the various steps of the catalytic events directly relates to the mechanism of catalysis being funda- mentally different for retaining and inverting enzymes [38] The stereochemistry of isomaltose hydrolysis by yeast oligo- 1,6-glucosidase and a-glucosidases was confirmed to involve retention of the substrate anomeric configuration in the product This is illustrated for baker’s yeast o-glucosidase which shows 'H NMR spectra of isomaltose before (Fig 2A) and after (Fig 2B,C) addition of the enzyme

B

OM

ÁN c

| qT Ỉ J | 1 1 Ỉ | | { Ỉ | ị |

Fig 2 Hydrolysis of isomaltose by baker’s yeast «-glucosidase followed

by 1H NMR (A) before addition of enzyme; (B) 4 min; and (C) 16 h after addition of enzyme.

l Nu

On on OH |

H ————> H ————> H

No hr ys

Comparison of these spectra showed the appearance of a

doublet centered at 5.22 p.p.m This was assigned to H-1 of

free o-glucose while the resonance at 4.64 p.p.m., which

appeared later was assigned to H-1 of B-glucose which

stemmed from mutarotation of the initially released

a-glucose The anomer ratio of D-glucose (33% a: 67% 6)

was deduced from the 'H NMR spectrum after complete

hydrolysis of isomaltose (Fig 2C) and falls within the range

normally found for the equilibrrum mixture The stereo-

chemistry of the products thus confirmed that the three

a-glucosidases catalyze hydrolysis of 1isomaltose with reten-

tion of the anomeric configuration as 1s characteristic of

family 13 glycoside hydrolases Figure 3 shows the widely

accepted double displacement mechanism for retaining

hydrolases, which 1s believed to occur through oxacarbenr-

um ion transition-states and formation of a covalent

intermediate between the catalytic nucleophile and the C-1

of the substrate glycon [18,30,31,45—49] Further kinetics

analyses are not feasible due to the limited amounts of

analogs available; we therefore cannot determine the role of

a key polar group 1n the glycosylation or the deglycosylation

steps in the mechanism (Fig 3) However, one can conclude

that the discrimination of the diastereoisomer, as this is

associated with the Vy, and not the K,,, does not happen in

the initial reversible part of substrate complex formation,

but in subsequent steps of the catalytic mechanism [40]

Et +, HO

OH

O

OH OMe

HO -

OH é

o_ œ@$o

Fig 3 Catalytic mechanism for retaining gly-

HO 64 coside hydrolases including steps of protonation,

formation of a covalent intermediate, and HỌC, z product release, respectively, but not the inter-

mediate two transition-states (See text for AIB details and [30,31,39])

Recognition of diastereoisomeric isomaltoside derivatives

Isomaltose is flexible due to rotation around the C5—C6 bond It is possible to block this conformational flexibility

by alkylation of C6 (Fig 4) Previously, methyl 6-R- and methyl 6-S-methyl-o-isomaltoside were used to determine the preferred rotational conformer for glycoamylase [40] Hydrolysis catalyzed by baker’s yeast o-glucosidase (this enzyme was chosen as it has the highest activity of the two a-glucosidases; see Table 1) was similarly examined using methyl 6-R-ethyl- and methyl 6-S-ethyl-o-isomaltoside as the pair of conformationally biased substrate analogs (Table 3) While methyl 6-S-ethyl-o-isomaltoside was hydrolyzed with twofold lower V,,,,, but the same K,, as isomaltose (Table 3), the 6-R enantiomer was a poor substrate Vinax being 150-fold lower and K,, twofold higher than for isomaltose (Table 3) Baker’s yeast o-glucosidase thus preferred the 6-S isomer In contrast, glucoamylase from A niger hydrolyzed the 6-R enantiomer with 230-fold higher ka/Ky, Compared to the parent isomaltoside, the difference being essentially in the K,, and not in the rate of hydrolysis as for the a-glucosidase [40] This distinct preference for one of the two diastereoisomers of the C-6 alkyl isomaltose derivatives reflects the fact that one of the rotamers adopts a conformation with more favorable

Fig 4 Structure of the conformationally biased diastereosisomer substrates methyl 6-R- ethyl-c-isomaltoside (A) and methyl 6-S-ethyl- o-isomaltoside (B)

Table 3 Kinetic parameters for the hydrolysis of conformationally biased isomaltosides

Substrate max (mM: s'-U7!) Km (mM) V max/Km (s"*UT} a-Glucosidase from baker’s yeast"

Isomaltose 2.8 x 10> 9.8 2.8 x 10”

Methyl 6-S-ethyl-z-isomaltoside 1.6 x 107 9.6 1.7 x 1077

Methyl 6-R-ethyl-c-isomaltoside 1.8 x 10 19.4 93x 107’

Glucoamylase from A niger? Keat (8 ') Km (mM) Kkeat|Km (S.-M ~')

Methyl 6-S-methyl-a-1somaltoside 1.1] 90.0 0.012

Methyl 6- R-methyl-œ-isomaltoside 0.68 0.71 0.96

* At 30 °C, using 50 mm phosphate, pH 6.8 ° [40]

spatial distribution of the groups that play an important

role in the enzyme recognition This finding stresses the

fundamentally different active site architecture that exists

for the inverting glucoamylase and the retaining o-gluco-

sidases Glucoamylase, in contrast to o-glucosidase, applies

a single displacement mechanism and belongs to a different

fold family, glycoside hydrolase 15 The specific activities

and substrate affinities are similar for these retaining and

inverting enzymes, all of which have reasonable capacity in

the glucose release from the nonreducing end of disac-

charides and small substrates However, the o-glucosidase

showed large variation in rate of hydrolysis between the

methyl 6-S- and 6-R-ethyl o-isomaltosides, with small

differences in affinity for the two distereoisomers, whereas

the discrimination by glucoamylase was associated with the

K,, [40] and not with the rate of hydrolysis (Table 3)

CONCLUSION

The enzyme preparations used in the present analysis are

considered valuable representatives of two categories of

yeast a-glucosidases The study strongly demonstrates the

advantage offered by enzymes with simple specificity for

application of substrate analogs in elucidation of the roles

of individual substrate groups or atoms in binding and

catalysis This is in contrast to other enzymes of the

glycoside hydrolase family 13 catalysing polysaccharide

degradation in an endo-like fashion, for which even model

substrates would typically be rather large and hence

extremely difficult, laborious and costly to synthesize In

addition, the option of several functional binding modes in

the active site cleft in these latter enzymes obscures

interpretation using analogs of the impact of specific

substrate groups on catalysis The fact that the enzymes

used in the present study possess a simple substrate

specificity and belong to the large family 13 representing

28 substrate specificities [30] suggests an application of the

present findings, ultimately, for rational protein engineering

of these and other family members with other specificities

Contacts with invariant Arg, His, and Asp residues involved

in charged hydrogen bonds to the glucose ring at subsite —1

in family 13 (reviewed in [30]) are thus proposed to be

responsible for the reported major role of hydroxyl groups

of this ring in transition-state stabilization While the

invariant Asp plays a role in catalysis [47] and mutation

leads to inactivation, the single mutation to Asn of each of

two His interacting at subsite —1 with OH-6 and OH-2 plus

OH-3, in case of barley o-amylase 1, affected transition-

state stabilization and reduced activity to 5 and 10%,

respectively [50] Structure guided sequence comparisons, in

contrast, do not yet allow tentative identification of specific

residues that are important for the interactions with the

substrate ring at subsite +1 in the oligo-1,6-glucosidase as

well as for controlling the exo-action at the level of the

nonreducing end ring at subsite —1 of all three «-glucosid-

ases included in the present comparison The findings on

substrate key polar groups and preferred isomaltoside

diasteroisomers, however, will be valuable in future mod-

eling of substrate complexes of the œ-plucosidases and

related enzymes if the structures become available The data

may thus guide protein engineering studies that address the

œ-l,4 and œ-l,6 bond specificity of these closely related

a-glycosidases

ACKNOWLEDGEMENTS

Bent O Petersen is gratefully acknowledged for performing the NMR spectroscopy experiments and Ulrike Spohr and Raymond U Lemieux are thanked for the synthetic substrate analogs This work was supported in part by funding from the Natural Sciences and Engineering Research Council of Canada (to M M P)

REFERENCES

1 Bourne, Y., van Tilbeurgh, H & Cambillau, C (1993) Protein— carbohydrate interactions Curr Opin Struct Biol 3, 681-686

2 Bundle, D.R & Young, N.M (1992) Carbohydrate—protein interactions in antibodies and lectins Curr Opin Struct Biol 2,

666-673

3 Lemieux, R.U (1996) How water provides the impetus for molecular recognition in aqueous solution Acc Chem Res 29,

373-380

4 Lis, H & Sharon, N (1998) Lectins: carbohydrate-specific pro- teins that mediate cellular recognition Chem Rev 98, 637-674

5 Toone, E.J (1994) Structure and energetics of protein-carbohy- drate complexes Curr Opin Struct Biol 4, 719-728

6 Vijayan, M & Chandra, N (1999) Lectins Curr Opin Struct

Biol 9, 707-714

7 Frandsen, T.P., Stoffer, B.B., Palcic, M.M., Hof, S & Svensson, B (1996) Structure and energetics of the glucoamylase-isomaltose transition-state complex probed by using modeling and deoxy- genated substrates coupled with site-directed mutagenesis J Mol

Biol 263, 79-89

8 Nikrad, P.V., Beierbeck, H & Lemieux, R.U (1992) Molecular recognition X A novel procedure for the detection of the inter- molecular hydrogen bonds present in a protein oligosaccharide complex Can J Chem 70, 241-253

9 Sierks, M.R & Svensson, B (1992) Kinetic identification of a hydrogen bonding pair in the glucoamylase/maltose transition state complex Prot Eng 5, 185-188

10 Sierks, M.R & Svensson, B (2000) Energetic and mechanistic studies of glucoamylase using molecular recognition of maltose

OH groups coupled with site-directed mutagenesis Biochemistry

39, 8585-8592

11 Spohr, U., Le, N., Ling, C.C & Lemieux, R.U (2001) The syn- theses of 6-C-alkyl derivatives of methyl «-isomaltoside for a study

of the mechanism of hydrolysis by amyloglucosidase Can J

Chem 79, 238-255

12 Bock, K & Pedersen, H (1988) Synthesis of 6-(S)deuterium- labelled derivatives of maltose and isomaltose Acta Chem Scand

Ser B 42, 190-195

13 Frandsen, T.P & Svensson, B (1998) Plant «-glucosidases of the glucoside hydrolase family 31 Molecular properties, substrate specificity, reaction mechanism, and comparison with family members of different origin Plant Mol Biol 37, 1-13

14 Watanabe, K., Hata, Y., Kizaki, H., Katsube, Y & Suzuki, Y (1997) The refined structure of Bacillus cereus oligo-1,6-glucosi- dase at 2.0 A resolution: structural characterization of proline- substitution sites for protein thermostabilization J Mol Biol 269,

142-153

15 Htilseweh, B., Dahlems, U.M., Dohmen, J., Strasser, A.W.M & Hollenberg, C.P (1997) Characterization of the active site of Schwanniomyces occidentalis ghucoamylase by in vitro mutagenesis Eur J Biochem 244, 128-133

16 Inohara-Ochiai, M., Nakayama, T., Goto, R., Nakao, M., Ueda,

T & Shibano, Y (1997) Altering substrate specificity of Bacillus

sp SAM1606 a-glucosidase by comparative site-specific muta- genesis J Biol Chem 272, 1601-1607

17 Inohara-Ochiai, M., Okada, M., Nakayama, T., Hemmi, H., Ueda, T., Iwashita, T., Kan, Y., Shibano, Y., Ashikari, T & Nishino, T (2000) An active-site mutation causes enhanced

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

39

36

reactivity and altered regiospecificity of transglucosylation cata-

lyzed by the Bacillus sp SAM 1606 a-glucosidase J Biosci Bioeng

89, 431-437

Liiv, L., Parn, P & Alamae, T (2001) Cloning of maltase gene

from a methylotrophic yeast, Hansenula polymorpha Gene 265,

77-85

Nakamura, A., Nishimura, I., Yokoyama, A., Lee, D.G., Hidaka, M.,

Masaki, H., Kimura, A., Chiba, S & Uozumi, T (1997) Cloning

and sequencing of an a-glucosidase gene from Aspergillus niger

and its expression in Aspergillus nidulans J Biotechn 53, 75-84

Okuyama, M., Okuno, A., Shimizu, N., Mori, H., Kimura, A &

Chiba, S (2001) Carboxyl group of residue Asp647 as possible

proton donor in catalytic reaction of a-glucosidase from Schizo-

saccharomyces pombe Eur J Biochem 268, 2270-2280

Taki, Y., Daimon, K & Suzuki, Y (1992) Cloning and expression

of a thermostable exo-a-1,4-glucosidase gene from Bacillus

Stearothermophilus ATCC12016 in Escherichia coli Appl Micro-

biol Biotechnol 38, 243-274

Frandsen, T.P., Olsen, F.L., Mirgorodskaya, E., Roepstorff, P &

Svensson, B (2000) High pI a-glucosidase from barley malt

Purification, enzymic characterization, and nucleotide sequence

Plant Physiol 123, 275-286

Chiba, S (1988) «-Glucosidases In Handbook of Amylases and

Related Enzymes (The Amylase Research Society of Japan, ed.)

pp 104-105 Pergamon Press, Oxford

Chiba, S (1997) Molecular mechanism in œ-glucosklase and glu-

coamylase Biosci Biotech Biochem 61, 1233-1239

Henrissat, B (1991) A classification of glycosyl hydrolases based

on amino acid sequence similarity Biochem J 280, 309-316

Henrissat, B & Bairoch, A (1993) New families in the classifica-

tion of glycosyl hydrolases based on amino acid similarities Bio-

chem J 293, 781-788

Coutinho, P.M & Henrissat, B The Carbohydrate-active Enzyme

Server, available from http://afmb.cnrs-mrs.fr/CAZY/

Yoshikawa, K., Yamamoto, K & Okada, S (1994) Classification

of some a-glucosidases and «-xylosidases on the basis of substrate

specificity Biosci Biotechn Biochem 58, 1392-1398

Janecek, Š (1997) «-Amylase family: molecular biology and evo-

lution Prog Biophys Mol Biol 67, 67-97

MacGregor, E.A., Janecek, S & Svensson, B (2001) Relationship

between structure and specificity in the «-amylase family Biochim

Biophys Acta 1546, 1-20

Svensson, B (1994) Protein engineering in the o-amylase family:

catalytic mechanism, substrate specificity, and stability Plant

Mol Biol 25, 141-157

Davies, G.J., Wilson, K.S & Henrissat, B (1997) Nomenclature

for sugar-binding subsites in glucosyl hydrolases Biochem J 321,

557-559

Brzozowski, A.M & Davies, G.J (1997) Structure of the

Aspergillus oryzae a-amylase complexed with the inhibitor acar-

bose at 2.0 A resolution Biochemistry 36, 10837-10845

Brzozowski, A.M., Lawson, D.M., Turkenberg, J.P., Bisgard-

Frantzen, H., Svendsen, A., Borchert, T.V., Dauter, Z., Wilson,

K.S & Davies, G.J (2000) Biochemistry 39, 9099-9107

Dauter, Z., Dauter, M., Brzozowski, A.M., Christensen, S.,

Borchert, T.V., Beier, L., Wilson, K.S & Davies, G.J (1999)

X-ray structure of Novamyl, the five-domain ‘maltogenic’

a-amylase from Bacillus stearothermophilus: maltose and acarbose

complexes at 1.7 A resolution Biochemistry 38, 8385-8392

Kadziola, A., Sogaard, M., Svensson, B & Haser, R (1998)

Molecular structure of a barley o-amylase-inhibitor complex:

37

38

39

40

4I

42

43

44

45

46

47

48

49

50

51

implications for starch binding and catalysis J Mol Biol 278,

205-217

Qian, M., Haser, R., Buisson, G., Duée, E & Payan, F (1994) The active center of a mammalian o-amylase Structure of the complex

of a pancreatic o-amylase with a carbohydrate inhibitor refined to 2.2 A resolution Biochemistry 33, 6284-6294

Ly, H.D & Withers, S.G (1999) Mutagenesis of glycosidases Ann Rey Biochem 68, 487-522

Weill, C.E., Burch, R.J & van Dyk, J.W (1954) An c-amylo- glucosidase that produces f-glucose Cereal Chem 31, 150-

158

Palcic, M.M., Skrydstrup, T., Bock, K., Le, N & Lemieux, R.U (1993) Substrate recognition by amyloglucosidase: evaluation of conformationally biased isomaltosides Carbohydr Res 250, 87-

92

Sierks, M.R., Bock, K., Refin, S & Svensson, B (1992) Active site similarities of glucose dehydrogenase, glucose oxidase and gluco- amylase probed by deoxygenated substrates Biochemistry 31,

8972-8977

Lemieux, R.U., Spohr, U., Bach, M., Cameron, D.R., Frandsen, T.P., Stoffer, B.B., Svensson, B & Palcic, M.M (1996) Chemical mapping of the active site of glucoamylase of Aspergillus niger Can J Chem 74, 319-335

Wilkinson, A.J., Fersht, A.R., Blow, D.M & Winther, G (1983) Site-directed mutagenesis as a probe of enzyme structure and catalysis: tyrosyl-tRNA synthetase cysteine-35 to glycine-35 mutation Biochemistry 22, 3581-3586

Fersht, A.R., Shi, J.-P., Knill-Jones, J., Lowe, D.M., Wilkinson, A.J., Blow, D.M., Brick, P., Carter, P., Waye, M.M.Y & Winter,

G (1985) Hydrogen bonding and biological specificity analyzed

by protein engineering Nature 314, 235-238

Davies, G.J., Mackenzie, L., Varrot, A., Dauter, M., Brzozowski, A.M., Schiilein, M & Withers, $.G (1998) Snapshots along an enzymatic-reaction coordinate — analysis of a retaining B-glucoside hydrolase Biochemistry 37, 11707-11713

McCarter, J.D & Withers, $.G (1994) Mechanisms of enzymatic glycoside hydrolysis Curr Opin Struct Biol 4, 885-892 Uitdehaag, J.C.M., Mosi, R., Kalk, K.H., van der Veen, B.A., Dikhuizen, L., Withers, $.G & Dijkstra, B.W (1999) X-ray structures along the reaction pathway of cyclodextrin glycosyl- transferase elucidate catalysis in the «-amylase family Nat Struct

Biol 6, 432-436

van der Veen, B., Uitdehaag, J.C.M., Penninga, D., van Alebeek, G.-J., Smith, L.M., Dijkstra, B.W & Dikhuizen, L (2000) Rational design of cyclodextrin glycosyltransferase from Bacillus circulans strain 251 to increase œ-cyclodextrin production J Biol

Chem 296, 1027-1038

Vocadlo, D.J., Davies, G.J., Laine, R & Withers, S.G (2001) Catalysis by hen egg-white lysozyme proceeds via a covalent intermediate Nature 412, 835-838

Sogaard, M., Kadziola, A., Haser, R & Svensson, B (1993) Site- directed mutagenesis of histidine 93, aspartic acid 180, glutamic acid 205, histidine 290, and aspartic acid 291 at the active site and tryptophan 279 at the raw starch binding site in barley «amylase

1 J Biol Chem 268, 22480-22484

Frandsen, T.P., Christensen, T., Stoffer, B., Lehmbeck, J., Dupont, C., Honzatko, R.B & Svensson, B (1995) Mutational analysis of the roles in catalysis and substrate recognition of arginines 54 and 305, aspartic acid 309, and tryptophan 317 located at subsites | and 2 in glucoamylase from Aspergillus niger Biochemistry 34, 10162-10169.