Chester Clinic

 

Pii: s0167-4838(00)00232-6

Biochimica et Biophysica Acta 1543 (2000) 275 293 Glucoamylase: structure/function relationships, and protein engineering JÖrgen Sauer a, Bent W. Sigurskjold b, Ulla Christensen c, Torben P. Frandsen d, Ekaterina Mirgorodskaya e, Matt Harrison e, Peter Roepstor¡ e, Birte Svensson a;* a Department of Chemistry, Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-2500 Copenhagen, Valby, Denmark b Department of Biochemistry, August Krogh Institute, University of Copenhagen, Universitetsparken 13, DK-2100 Copenhagen Ò, Denmark c Chemical Laboratory 4, Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen Ò, Denmark d Novo Nordisk, Novo Alle¨, DK-2880 Bagsv×rd, Denmark e Department of Molecular Biology, University of Southern Denmark, Odense University, Campusvej 55, DK-5230 Odense M, Denmark Received 15 March 2000; received in revised form 31 August 2000; accepted 28 September 2000 Glucoamylases are inverting exo-acting starch hydrolases releasing L-glucose from the non-reducing ends of starch and related substrates. The majority of glucoamylases are multidomain enzymes consisting of a catalytic domain connected to a starch-binding domain by an O-glycosylated linker region. Three-dimensional structures have been determined of free and inhibitor complexed glucoamylases from Aspergillus awamori var. X100, Aspergillus niger, and Saccharomycopsis fibuligera.
The catalytic domain folds as a twisted (K/K)6-barrel with a central funnel-shaped active site, while the starch-binding domain folds as an antiparallel L-barrel and has two binding sites for starch or L-cyclodextrin. Certain glucoamylases are widely applied industrially in the manufacture of glucose and fructose syrups. For more than a decade mutational investigations of glucoamylase have addressed fundamental structure/function relationships in the binding and catalytic mechanisms. In parallel, issues of relevance for application have been pursued using protein engineering to improve the industrial properties.
The present review focuses on recent findings on the catalytic site, mechanism of action, substrate recognition, the linker region, the multidomain architecture, the engineering of specificity and stability, and roles of individual substrate binding subsites. ß 2000 Elsevier Science B.V. All rights reserved.
Keywords: Catalytic base; Binding loop; O-Glycosylated linker; Site-directed mutagenesis; Sequence replacement variant; Mass spectrometry; Bifunctional inhibitor; Isothermal titration calorimetry; Molecular recognition; Pre-steady-state kinetics; lase, EC 3.2.1.3) catalyse hydrolysis of K-1,4 and K-1,6 glucosidic linkages to release L-D-glucose Glucoamylases (GAs) (1,4-K-D-glucan glucohydro- from the non-reducing ends of starch and related poly- and oligosaccharides ([1,2]; for reviews see [3,4]). Fungal GA is widely used in the manufacture of glucose and fructose syrups. Although activity Abbreviations: CD, catalytic domain; GA, glucoamylase; (kcat/Km) towards the K-1,6 linkage is only 0.2% of ITC, isothermal titration calorimetry; SBD, starch-binding do- that for the K-1,4 linkage [1,5 7] this su¤ces to ad- * Corresponding author. Fax: +45-3327-4708; versely a¡ect the yield in industrial sacchari¢cation.
E-mail: [email protected] This property together with a need for other develop- 0167-4838 / 00 / $ see front matter ß 2000 Elsevier Science B.V. All rights reserved.
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 ments motivated both fundamental and goal-ori- e¢t from elevated thermostability and enhancement ented protein engineering of GA.
of activity in the neutral pH range [3,4]. An increase The GAs constitute glycoside hydrolase family 15 of the glucose yield in sacchari¢cation beyond the [8,9] and at least 23 primary structures are known current 96% level might be achieved by suppressing from ¢lamentous fungi, yeast, eubacteria and archae the activity of GA on K-1,6 linkages. A very di¡erent [10,11]. GA has frequently served as a prototype in way to exploit GA is to utilise the O-glycosylated investigations on glycoside hydrolases as will be ap- linker to connect di¡erent domains by generation parent from the present review. The three-dimension- of fusion proteins with new combinations of multiple al structure of the catalytic domain (CD; aa 1 471) functionalities. Variations on this approach include of GA (aa 1 616) from Aspergillus awamori var.
fusions of SBD from GA to L-galactosidase for pu- X100 has been described in detail for native and li- ri¢cation purposes [37], or to the C-terminus of an gand-complexed forms [12 17]. Furthermore, pre- K-amylase to increase its capacity to bind onto and liminary structure determination was made of wild- degrade starch granules and other recalcitrant forms type and mutants of A. niger GA [18] which has 94% of starch (N. Juge, J.N. Larsen, C.S.M. Furniss, V.
sequence identity to GA from A. awamori var. X100.
Planchot, M.-F. Le Gal-Coe«¡et, D.B. Archer, B.
In addition, the crystal structure of the yeast GA Svensson, G. Williamson, unpublished). Di¡erent from Saccharomycopsis ¢buligera, which lacks a binding polypeptide tails were also added to GA to starch-binding domain (SBD), has been published facilitate puri¢cation [38]. Recently GA has been recently [19]. GA from A. awamori var. X100 folds used in fusions as a vehicle for the production of into an (K/K)6-barrel and the C-terminal part (aa recombinant proteins in A. niger [39].
440 471) of the CD wraps around the (K/K)6-motif and constitutes the N-terminal part of an O-glycosy- lated linker (aa 440 508) that connects to a C-termi- nal SBD (aa 509 616). The conformation of the most highly O-glycosylated part of the linker (aa 472 508) 2.1. Catalytic domain is unknown. The structure of SBD from A. niger GA has been determined by NMR in free form [20] and The catalytic domain (CD) of GA from A. awa- bound to L-cyclodextrin, a well-known starch mimic mori var. X100 contains 13 K-helixes of which 12 [21]. The structure of the entire GA is thus not avail- form an (K/K)6-barrel. In this fold, six outer and able. The domain-level organisation of GA has been six inner K-helixes surround the funnel-shaped active addressed, however, using di¡erent biophysical tech- site, constituted by the six highly conserved KCK niques [22 24].
segments [10,11] that connect the N-termini of the GA catalyses hydrolysis of glucosidic linkages with inner with the C-termini of the outer helixes [12 inversion of the anomeric con¢guration [25 29]. Sev- 15] (Fig. 1). The catalytic site includes the general en subsites were identi¢ed kinetically to participate in acid and base catalysts Glu179 and Glu400 situated substrate recognition [30]. The general acid catalyst at the bottom of a pocket [13,31,32,40]. CDs of GAs and proton donor Glu179 and the catalytic base from A. awamori var. X100, A. niger and S. ¢buligera Glu400 in GA of A. niger are characterised by muta- share a very similar fold. The S. ¢buligera GA con- tional analysis [31,32]. Moreover, an array of amino tains 14 K-helices, 12 of which makes up the (K/K)6- acid residues that bind directly or via a network of motif in an organisation identical to that of A. awa- interactions with substrate at di¡erent subsites has mori var. X100 and A. niger CD [19]. Two extra been subjected to site-directed mutagenesis (for re- short helices protrude from the K-helix connecting views see [3,4]). Mutational analysis combined with loops in the ¢rst and the last pair of antiparallel biophysical techniques gave information on individ- helices in the fold [19]. The most pronounced di¡er- ual subsites and has described the impact of the pro- ence between these GAs, however, is the lack of SBD tein on transition state stabilisation and di¡erent in the S. ¢buligera enzyme [10,41]. A single Ser re- steps in the mechanism of action [33 36].
placement between two very closely related S. ¢buli- Application of GA in starch industries would ben- gera GAs is responsible for activity di¡erences [42].
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Fig. 1. Stereoview of the catalytic domain (aa 1 471) of Aspergillus awamori var. X100 GA complexed with the pseudotetrasaccharide acarbose (rings A, B, C, D are indicated). The C- and N-termini are indicated together with the side chains of the two catalytic resi- dues E179 and E400 (from [15]).
2.2. Starch-binding domain actions with L-cyclodextrin. Mutagenesis studies re- vealed, however, that they vary only little in a¤nity, The C-terminal SBD of A. niger was prepared Ka being 3.6U104 and 1.6U105 M31, respectively both by proteolysis and in recombinant form, and [47]. Also the enthalpies and entropies of binding solution structures of the free and the L-cyclodex- are similar as analysis of the binding thermodynam- trin-complexed SBD were determined by NMR spec- ics of L-cyclodextrin and SBD by isothermal titration troscopy [20,21]. SBD consists of eight L-strands or- calorimetry (ITC) did not resolve the two sites ganised in two L-sheets forming a twisted L-barrel structure [20,43]. Two starch-binding sites, seen to accommodate the starch mimic L-cyclodextrin [44 2.3. Linker region 46], are located on opposite sides of the top' of the domain, i.e., away from the linker attachment The serine- and threonine-rich O-glycosylated re- point as seen in Fig. 2 [20,21]. These sites display gion of A. niger GA (aa 440 508) contains a very distinctly di¡erent structure and non-covalent inter- highly O-glycosylated C-terminal segment of about Fig. 2. Stereoview of the starch binding domain (SBD) from A. niger GA complexed with the starch mimic L-cyclodextrin at the two binding sites. The C- and N-termini are indicated (from [21]).
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Fig. 3. Transformed electrospray ionization mass spectrum of the heterogeneously-glycosylated glycopeptide linker (from Asn430 Phe519) prepared from A. niger GA [48] showing at least 17 di¡erent glycoforms. The average mass di¡erence between peaks in these spectra is 162.1 Da, corresponding to a single hexose residue. The raw mass/charge spectrum prior to transformation to true mass is shown as an inset, and the charge state of multiply-charged series of peaks is shown. Spectra are normalised to the most intense peak in each spectrum, and the transformed spectrum has been background-subtracted and smoothed.
30 aa that connects with SBD [10,48]. This particular in the structural model, but it has been speculated part of the linker has been attributed roles in stabil- that this part surrounds CD in a continuation from ity, secretion, and digestion of raw starch [49 52].
residue 471, to place SBD with one of the two bind- Mass spectrometric analysis of the peptide Asn430 ing sites near the active site. This resembles the ar- Phe519 shows a high degree of heterogeneity in the chitecture of cyclodextrin glucanotransferase in amount of attached sugars. At least 17 di¡erent gly- which a homologous C-terminal SBD is situated rel- coforms can be identi¢ed from the transformed elec- ative to CD to direct the substrate chain into the trospray ionisation mass spectra (Fig. 3; M. Harri- active site via one binding site of SBD and to be son, P. Roepstor¡, and B. Svensson; unpublished bound onto soluble or insoluble polymeric substrate results). Based on a calculated molecular mass of at the other [57]. The full-length linker is anticipated the peptide of 8562.28 Da and the experimentally to be conformationally £exible in accordance with determined mass of 18 991.3 Da for the glycoform the formation of 1:1 complexes between a bifunc- of the lowest molecular mass (Fig. 3), approximately tional inhibitor and GA [23,24].
63 moles of hexose are attached to the peptide. The O-glycosidically linked units range from single man- 2.4. Overall structure nosyl to branched mannotriosyl in wild-type A. niger GA [53 55]. Heterologous expression of A. niger GA SBD was earlier shown to be required for degra- results in large host-dependent variation in the con- dation of raw starch by GA, the natural G2 form (aa tent of sugars ranging from hypermannosylation by 1 512) without SBD having very low activity on raw Saccharomyces cerevisiae to modest over-glycosyla- starch [58]. Recently, isolated SBD acting on starch tion by both Pichia pastoris and a laboratory strain granules together with G2 showed a synergistic e¡ect of A. niger [50,56].
on the degradation of the insoluble substrate, sug- The ¢rst part (aa 440 471) of the O-glycosylated gesting that SBD binds onto starch as an individual region carries, as seen in the structural model of GA, entity and disrupts the compact structure of the about 10 exposed single mannosyl residues [12] which starch granule facilitating the hydrolysis by CD [59].
together with the two N-glycosidically linked units at The complete three-dimensional structure of intact Asp171 and Asp395 form a belt of carbohydrate GA comprising CD, the linker region, and SBD, is around the globular CD [12]. The highly O-glycosy- not known. In an attempt to delineate the relative lated part of the linker (aa 472 508) is not included position of CD and SBD, scanning tunnelling BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Fig. 4. Structures of the bifunctional ligands formed by acarbose and L-cyclodextrin joined by poly(ethyleneglycol) spacers of varying length. The ligands have either no spacer (L0) or spacers of 14, 36, and 73 Aî (L14, L36, L73), respectively.
microscopy indicated that the two domains are 3. Mechanism of action around 90 Aî apart [22]. Recently, a series of bifunc- tional inhibitors, in which the CD speci¢c pseudo- 3.1. Catalytic site tetrasaccharide inhibitor acarbose and the SBD- speci¢c ligand L-cyclodextrin were coupled via The widely accepted mechanism of hydrolysis in- thioglycoside linkages, was used to further analyse volves proton transfer to the glycosidic oxygen of the binding to the di¡erent domains. The bifunctional scissile bond from a general acid catalyst; formation molecules were synthesised without and with varying of an oxocarbenium ion; and a nucleophilic attack of lengths of poly(ethyleneglycol) spacers connected to water assisted by a general base catalyst [28,60 62].
the reducing end of acarbose and C6 of a glucose Glu179 and Glu400 in GA from A. niger have been ring in L-cyclodextrin shown in Fig. 4 [24]. Four identi¢ed as the general acid and the general base di¡erent heterobifunctional inhibitors were demon- catalyst, respectively, and pH-dependencies of strated by ITC to bind simultaneously at the active steady-state kinetic parameters are in accordance site of CD and one of the binding sites on SBD [23].
with a rate determining hydrolysis step involving It was thus concluded that in solution the two do- these two catalytic residues [13,31,32]. Also in accor- mains of the GA molecule either are in, or can be dance with this is the observation that mutation of brought into, close proximity. The sum of enthalpies Glu400 to Gln results in a reduction of kcat to 3% of in binding of acarbose and L-cyclodextrin gave essen- wild-type, showing the marked in£uence of this res- tially the same value as found for the enthalpy of the idue on the rate determining step [32].
bifunctional ligands. The binding a¤nities, however, The GA catalysis occurs with inversion of the were reduced approximately 105 times compared to anomeric con¢guration (Fig. 5) in a single displace- that of acarbose due to strong entropy penalties in ment mechanism and the gap between the catalytic binding of double-headed inhibitors [23]. Dynamic acids is 9.2 Aî as is typical for inverting glycoside light-scattering measurements on the binding of hydrolases [28,64 66]. In contrast the distance be- the bifunctional inhibitors suggested co-operation tween the catalytic acids in retaining glycoside hydro- between the domains [24]. The inhibitors were lases is only 4.8 5.5 Aî and hydrolysis occurs in a concluded to bind in a bimolecular complex with double displacement mechanism that includes a co- occupation of one site at each of the CD and valent intermediate [28,64]. In this mechanism the proposed covalent bond between substrate and pro- BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 However, as an important feature of the model, in agreement with the suggestions made for Hormoconis resinae GA [78] and in contrast to other Model I- type suggestions [30,73,76,77], it was found [33,35,36] that subsite +1 participates in the formation of the ES-complex and that the transformation of ES to E*S involves conformational changes, but not the ¢lling of a previously empty subsite 31 by relocation Fig. 5. The generally accepted catalytic mechanism of GA illus- of the substrate.
trating the action of the catalytic base E400 (top) and acid Tables 1 and 2 summarise the results. Further- E179 (bottom) in the water-assisted hydrolysis of substrate in- more, it has been shown that the pre-steady-state volving inversion of the con¢guration of the anomeric carbon.
kinetic results [35] are not in accordance with the classical model [30,79,80] of GA catalysed reactions, tein has the consequence that high precision of the which involves strong non-productive binding of spatial positioning of the two catalytic groups is nec- substrates and substrate length independent values essary for the nucleophilic attack on the glycosidic of the intrinsic catalytic constant. Since this model bond [65]. Such a strict geometrical requirement for is usually used in subsite energy calculations, it has the catalytic site seems not to apply for the inverting led to the false general acceptance of subsite +1 of GA, as it was best illustrated by the elevated activity GA as the one providing most of the substrate bind- of GA from A. niger in which the catalytic base was ing energies [6,30,32,69,81 83].
replaced by cysteine which was subsequently oxidised As seen from the three-dimensional structure of to cysteinesulphinic acid [29,66].
GA-inhibitor complexes shown in close-up in Fig. 6 [13,15 18] Trp52 and Trp120 are hydrogen bonded 3.2. Binding mechanism to the general acid catalyst, Glu179. The Trp52-bond is 3.04 Aî [16] and is therefore not shown in Fig. 6.
Conserved tryptophan residues are involved in in- Trp317 and Glu180 are situated on the opposite teractions of the GAs with substrates and inhibitors £ank of the active site (Fig. 6), but are not in close [6,30,67 71] and changes in intrinsic enzyme £uores- contact. Kinetic results (Tables 1 and 2) show that cence result as binding occurs. These changes were the Trp317- and Glu180-mutants react almost iden- earlier assumed to involve only a tryptophan in sub- tically, and structure energy minimisation calcula- site +1 [30], but recent structure and function studies tions further show that the same loss of Arg305 have shown that in addition to this tryptophan a and Glu180 hydrogen bonds to the substrate occurs number of other tryptophans are involved [15 in each of these mutants [36]. The Trp52 Trp120 £ank of the active site has been designated the Pre-steady-state kinetics analysis of the binding K-£ank, and the £ank with Trp317 and Glu180 has mechanism of wild-type and mutant A. niger GA been designated the L-£ank [36].
has been based on the intrinsic protein £uorescence Mutations on the L-£ank primarily a¡ect the sec- changes that occur when substrates and inhibitors ond reaction step, assumed to be a conformational bind [33,35,36,72 75]. Formation of complexes with change, where the substrate obtains the correct posi- single exponential kinetics was seen in all cases and tion for catalysis after the initial association. Appar- analysis of their concentration dependencies all ently, this step involves the formation of hydrogen showed results in accordance with a three-step reac- bonds to Arg305 and to Glu180. This is in excellent tion mechanism (Model I) of catalysis involving two agreement with the critical role of Glu180 for the intermediates: ES, the initial association complex, induction of a productive conformation of isomal- and E*S, the Michaelis complex (i.e., the most stable tose [63]. In spite of the interaction of Trp317 with enzyme substrate intermediate) [30,33,35,36,72 77].
Glu400, apparently this tryptophan plays no role in the catalytic step, but exerts its e¡ect in the binding E ‡ S01ES02ESÿ! of the substrate, particularly assisting in the confor- BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Results for the reaction of maltose with wild-type forms and mutant glucoamylases (pH 4.5, 8³C)Enzyme G1, amino acid residues 1 616; G2, amino acid residues 1 512 [58].
mational change bringing the substrate in place for The Trp52- and the Trp120-mutants at the K-£ank catalysis. Interestingly, the resulting kinetic e¡ects show similar changes of the kinetics, since both mu- obtained when Trp317 is changed to Phe closely par- tations lead to almost total loss of catalytic turnover.
allels that obtained when Glu180 was changed to Apparently the correct position of Glu179 for catal- Gln [35,36]. Furthermore, the changed pattern of ysis is not obtained in the second reaction step here.
hydrogen bonds between Arg305, Asp309, Tyr306 The mutants, further show stronger substrate bind- and Glu180, obtained when Trp317 is mutated to ing than wild-type GA [33,36]. The kinetic parame- Phe is the same as when Glu180 is mutated to Gln ters (Tables 1 and 2) are very similar for the wild- [36]. The mutation of Trp317 clearly a¡ects the posi- type G1 and G2 forms of GA. Trp52 is situated at tion of Glu400, but kc is not changed. This supports the bottom of the active site, it stacks with the the generally accepted view that the rate determining C5 C6 part of substrate glucose moiety in subsite step is in the actual hydrolysis. In the wild-type en- 31 (ring a in Fig. 6) and makes a hydrogen bond zyme this is the formation of an oxocarbenium ion, [16] to the catalytic acid, Glu179 (Fig. 6). Therefore whereas when the assistance of the general base cat- this tryptophan is in a position where it is expected alyst is lacking at low pH and in Glu400-mutants it to be of great importance in the catalytic mechanism.
may change to the nucleophilic attack of water. It is The k2 values decrease generally more than in the thus indicated that the catalytic base, Glu400, is not Trp317- and Glu180-mutants. But this does not re- involved in that elementary step of the catalysis, sult in weaker binding, since the k32 values also which is rate determining.
greatly decrease. The Michaelis complexes with the Results for the reaction of maltotetraose with wild-type forms and mutant glucoamylases (pH 4.5, 8³C)Enzyme vFmax([S]) (%) Kd (mM) G1, amino acid residues 1 616; G2, amino acid residues 1 512 [58].
d From [35].
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Fig. 6. Stereoview of the active site of glucoamylase from A. awamori var. X100 with bound D-gluco-dihydroacarbose [16]. The four rings are marked a, b, c and d. Hydrogen bond interactions less than 3.0 Aî are represented by dashed lines, that of Trp52 to Glu179 is 3.04 Aî and therefore does not appear. We de¢ne the K-£ank of the active site as that on the right side of the ligand and the L-£ankas that on the left side in this representation.
Trp52CPhe mutant show Km values (within exper- The results show that Trp52 plays a role in destabi- imental error, Km = Kd) approximately one order of lisation of the Michaelis complex as well as of the magnitude less that those of the wild-type. This is association complex. The catalytic rates, as is seen similar to results obtained on the binding of the in- from the kc values, are extremely low, and the power hibitor 1-deoxynojirimycin [34].
of attraction lost here is greater than that gained An important feature of the Michaelis complex from the lowering of the Km values. The result is most probably is the presence of a hydrogen bond an absolute increase of the transition state energy between the substrate oxygen of the scissile bond and barrier. This increase is substantially larger than it Glu179. The Trp52CPhe and Trp120CPhe muta- would be if only a compensation of the stronger tions perturb Glu179 and the observed e¡ects of the binding in the Michaelis complex was involved. It mutations on the rearrangement step in which the is clearly indicated that Trp52 plays important roles ES-complex transforms into the Michaelis complex, in binding as well as in catalysis. The di¡erences E*S, indicate that it is the formation of this bond in between maltose and longer substrates (Tables 1 the second reaction step, which is impeded. Since the and 2) appear to be a general phenomenon complexes formed are only one order of magnitude weaker than wild-type Michaelis complexes, and the All in all Model I with a rate determining hydro- catalytic activities are almost totally lost, it seems as lysis step is supported by these ¢ndings. It has re- if in these mutant GAs Glu179 does not make the cently been suggested [75,85] that product dissocia- right interactions with the substrate for catalysis. As tion and not hydrolysis should be rate determining.
seen from Tables 1 and 2, the G2 form of the This seems highly unlikely, however, in the light of Trp52CPhe mutant shows K1 values less than or the known pH-dependency of kc [2,5,32], the weak equal to those of the wild-type, maltotriose and mal- binding of glucose [30], and the results showing more totetraose binding stronger in the ¢rst association slow, but nevertheless fast second reaction step of all complex, whereas maltose shows no signi¢cant of the mutants. A study of the pH-dependence of the change. The Km values decrease, which means stron- pre-steady-state kinetic parameters of the interaction ger binding. This is in accordance with the classical of wild-type GA and maltose further has shown that theory of enzyme function [84], which points out the the k2 value is slightly increasing in the range pH 5 bene¢t of a maximal power of attraction of the tran- 7, while it is not decreasing with the pK of the cata- sition state, a transition state stabilisation', at the lytic acid as would be the case, if this step was the expense of the attraction of the Michaelis complex.
actual hydrolysis step (U. Christensen, unpublished).
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 modi¢cation of the pseudotetrasaccharide exerted considerable impact on the a¤nity which was re- The pseudotetrasaccharide acarbose (Fig. 7) binds duced to Ka = 3.2U107 M31. The determination of with high a¤nity (Ka = 1012 M31) to GA [71,86] at the bound conformation of D-gluco-dihydroacarbose subsites 31 through +3 [15,17]. This is a substan- by transferred NOE NMR experiments indicates that tially stronger interaction than what is usually ob- the inhibitor is bound in a conformation that is sim- served between carbohydrates and proteins [87].
ilar to the conformation in the crystal structure of The free energy of binding is composed of approx- the complex, but di¡erent from the predominant so- imately two-thirds enthalpy and one-third entropy.
lution conformation of the free inhibitor [88]. L-ido- The pseudodisaccharide acarviosine (Fig. 7), which Dihydroacarbose (Fig. 7) was also obtained in the comprises the ¢rst two units of acarbose at the preparation of the D-gluco isomer of reduced acar- non-reducing end, binds with a much lower a¤nity bose and this inhibitor with an inverted chair con- of Ka = 7.8U106 M31 [86]. This means that the two formation of the hydrogenated valeinamine ring glucose units of acarbose are responsible for a con- showed an even weaker binding of 2.2U105 M31 siderable amount of the total binding free energy.
D-gluco-Dihydroacarbose (Fig. 7) prepared by hydro- A study of acarbose and 1-deoxynojirimycin (Fig.
genation of the valeinamine ring in acarbose showed 7) binding to a number of GA mutants with single a similar binding as acarbose except for a subsite 31 amino acid substitutions in CD has been reported distorted chair conformation [16,17]. The structural [34]. There are vast changes in the a¤nity for acar- bose ranging from a slight increase (Ka = 1013 M31) down to a¤nities of KaW103 M31. The large reduc- tions in a¤nity occurred for mutations in residues directly involved in hydrogen bonds with the sub- strate or in stacking interactions and also for muta- tions in groups involved in stabilisation of either substrate binding residues or catalytic residues. Other alterations had little or no e¡ect on acarbose bind- ing. Most of the mutants had almost wild-type a¤n- ity for 1-deoxynojirimycin, while only two seem to have abolished binding of this inhibitor completely Novel thioglucoside disaccharide analogue inhibi- tors of GA were synthesised and characterised in- volving kinetic measurements, molecular modelling, and detailed NMR conformational analysis which revealed important structural details underlying e¤- cient GA-oligosaccharide complexation [89,90].
Transferred NOE NMR measurements of methyl 5P-thio-4-N-K-maltoside, in complex with GA, showed that GA bound this analogue in a conforma- tion in the area close to the global energy minimum [91]. Kinetic evaluation showed e¤cient competitive inhibition of GA with a Ki value of only 4 WM by this analogue [89]. The importance of GA-oligosac- charide interactions at subsite +1 was emphasised by the higher Ki values of a series of 5-thio-D-glucopyr- Fig. 7. Competitive inhibitors of glucoamylase: acarbose (A),D-gluco-dihydroacarbose (B), L-ido-dihydroacarbose (C), methyl anosylarylamines compared to methyl 5P-thio-4-N-K- acarviosinide (D), and 1-deoxynojirimycin (E).
maltoside [89,90]. These compounds are, however, BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 more strongly bound by GA than the substrate 4.2. Replacement of the catalytic base by Cys-SO2H p-nitrophenyl-K-D-glucopyranoside [6,29], showing the importance of a ring sulphur and nitrogen in In order to enable engineering of the distance be- the interglycosidic linkage for e¤cient GA inhibition tween the catalytic general acid, Glu179, and base, Glu400, in GA, the catalytic base, which was previ- ously found to best tolerate substitution [32], was mutagenised to Cys. The side-chain was further at- 4. Protein engineering production tempted to be carboxyalkyl extended by modi¢cation of the SH group by reaction with various haloalkyl 4.1. Recombinant GA production carboxyl acids. This, however, failed to produce al- kylated Cys at position 400, but fortuitously, a GA E¤cient heterologous expression of GA encoding derivative was obtained with activity superior to that cDNA from A. awamori (identical to GA from of wild-type GA [29,66]. Subsequent chemical analy- A. niger) was established in the methylotrophic yeast sis involving HPLC-separation of peptide fragments P. pastoris [56]. To describe the in£uence of host prepared by treatment with the Endo-LysC protease dependent posttranslational modi¢cation on enzy- followed by matrix-assisted laser desorption/ionisa- matic and structural features, the recombinant pro- tion mass spectrometry combined with post-source tein produced in P. pastoris was compared to GA decay analysis enabled the unequivocal identi¢cation produced in the related hosts, S. cerevisiae and of the product as Cys400-SO2H [92]. Thus the thiol A. niger [56]. Recombinant GA produced in all three group had undergone spontaneous oxidation to the hosts showed essentially identical catalytic proper- sulphinic acid in the presence of the alkylating re- ties, but di¡ers in thermostability [56]. Molecular agent. Attempts to repeat the oxidation by a mixture mass determination using matrix-assisted laser de- of iodine and bromine successfully resulted in a GA sorption/ionisation mass spectrometry and neutral derivative with elevated activity [66]. Remarkably, sugar analysis revealed small, but signi¢cant varia- depending on the substrate, kcat increased up to tions in the glycosylation of the three recombinant 300% of the value for wild-type, an e¡ect most pro- GA forms. GA produced in S. cerevisiae thus has the nounced for longer K-1,4 maltooligosaccharides and highest content of carbohydrate with a measured K-1,6 isomaltooligosaccharides (Table 3). In contrast, molecular mass of 83.869 Da. GA produced in the a¤nity decreased (i.e., Km increased) with oligo- A. niger and P. pastoris had average molecular saccharide length. Similarly the a¤nity (Ka) for acar- mass of 82.839 Da and 82.327 Da, respectively [56].
bose decreased by a factor of approximately 103, Kinetic parameters for hydrolysis of malto- and isomaltooligosaccharides by wild-type and the Cys400-SO2H derivativeaSubstrate kcat/Km (s31 mM31) kcat/Km (s31 mM31) aData from [29].
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 while that of acarviosine decreased only by a factor loops simultaneously. If substitution was made at of 15, and the a¤nity for 1-deoxynojirimycin was either loop 3 or loop 5, the mutant GAs had very slightly raised by a factor of 3 (Table 4). Disaccha- low activity compared to wild-type and the double rides of varying bond type were also hydrolysed with loop mutant [7].
di¡erent retained activities, some having superior kcat (nigerose (K-1,3) and maltose), retained (kojibiose 4.4. Combination of protein engineering and substrate (K-1,2)) or inferior kcat (isomaltose). Only for niger- ose did the Km increase signi¢cantly [66]. Thus while the mutation of Glu400 to Cys dramatically de- The energetics of protein carbohydrate complexa- creased the activity this was with most substrates tion can, in principle, be described using two exper- more than restored by oxidation of the Cys to imental approaches: site-directed mutagenesis of se- Cys-SO2H. It is currently not known whether this lected active-site substrate binding residues and, behaviour is unique to GA, applies to inverting gly- molecular recognition of deoxygenated or otherwise coside hydrolases, or to other retaining and inverting chemically derivatised substrate analogues by the glycosidases in general.
wild-type enzyme. Both strategies have been exten- sively used to investigate protein substrate complex- 4.3. Replacement by homologue sequences at ation in GA from A. niger [3,63,93,94] and by com- bining the two approaches, direct identi¢cation of interacting pairs of atoms or groups of atoms and The architecture of the CD in GA has six loops support for enzyme induced substrate conformation- that connect the inner with the outer cylinder' of al changes have been achieved.
K-helices that create the substrate binding site and Mapping of the substrate key polar groups was carry the catalytic residues [12]. The clearly best con- done through wild-type GA recognition of deoxygen- served parts of the GA sequences are in these loop ated K-1,4- [93 95] and K-1,6-linked [63,96] substrate analogues. From these studies it became evident that GAs di¡er in stability and also show di¡erences in GA catalysed hydrolysis is strongly dependent on substrate speci¢city. The most extreme case of spe- charged protein-substrate hydrogen bonds from GA ci¢city variation is GA from Hormoconis resinae that to OH-3, OH-4P, and OH-6P as summarised in Fig. 8 has only 50-fold higher activity for K-1,4 compared for K-1,4-linked substrates [93 95] and to OH-4, to K-1,6 linkages as opposed to most GAs showing OH-4P, and OH-6P in K-1,6-linked substrates [63, 500 103 fold higher activity for the K-1,4-linked sub- 96]. Substitution of these particular OH groups by strates [78]. The H. resinae GA contains unusual se- hydrogen or methoxy groups are accompanied by quences in certain loops. Engineering mutation of loss in transition-state stabilisation of 11 19 kJ/mol loops 3 and 5 in A. niger GA to mimic the character- which is typical for groups involved in charged hy- istics of GA from H. resinae was possible without drogen bond interactions. Elimination of the corre- signi¢cant loss of activity and accompanied by de- sponding protein hydrogen-bond partners, Arg54, crease of the relative speci¢city for the K-1,4 over the Asp55, and Arg305, interacting with OH-4P, OH-6P, K-1,6 bond. However, mutation was required at both and OH-3/4, respectively, similarly resulted in dra- matic losses in transition-state stabilisation of up to 22 kJ/mol [6,82].
In addition, by coupling site-directed mutagenesis Ki for inhibitors of wild-type and Cys400-SO2H GAsa of GA and substrate molecular recognition structural details of transition-state interactions between GA and substrate can be detected [63,94]. Analysis of Glu180CGln GA using a series of deoxygenated maltose and isomaltose analogues thus demonstrated transition-state stabilising hydrogen bonds between aData from [29].
Glu180 and OH-2 in maltose [94] and OH-4 and BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Fig. 8. Representation of the energy contributions of the interactions between carbohydrate and protein as determined by mutagenesis and substrate analogue studies; the corresponding vvGV values are shown in italics and bold, respectively.
OH-3 in isomaltose [63]. The identi¢cation of these site +1 interactions which optimises transition-state interactions was shown prior to the acquisition of the stabilisation through charged hydrogen bonds to high resolution three-dimensional complexes demon- substrate OH-4P and 36P in subsite 31 [63]. GA strating the powerful combination of protein and substrate interactions at one subsite thus have critical substrate engineering. By using conformationally impact on crucial hydrogen bond formation at adja- biased substrate analogues such investigations have cent subsites.
been extended even further to obtain details on GA catalysed hydrolysis of an K-1,6-linked substrate [63].
4.5. Linker region variants The analysis of wild-type and variant GA catalysed hydrolysis of conformationally biased isomaltoside The question of the distance between the two do- analogues demonstrated that Glu180 induces a con- mains in GA has been addressed through a series of formational change of the bound substrate via sub- linker mutants in which the highly O-glycosylated Fig. 9. Schematic representation of the constructed linker variants. The catalytic (CD; aa 1 466) and starch binding (SBD; aa 509 616) domains connected through the O-glycosylated (aa 467 508) linker region and sequence alignment of the linker regions of other GAs used to replace the A. niger linker. Asterisks above the alignment indicate known O-glycosylation sites in A. niger GA. Numbers in brackets refer to amino acid position in wild-type sequences.
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 sequence (aa 468 508) was replaced by shorter link- heterogeneously glycosylated as in the wild-type ers from related GAs [97] or by a proline rich un- GA when produced by various homologous and het- natural sequence (Fig. 9). Mass spectrometric analy- erologous hosts [56]. The rather high mass increase sis of the variants (Fig. 10) determined apparent for the variant with the Rhizopus oryzae linker sug- masses of 73 413 and 90 793 Da for the Humicola gests that N-linked glycosylation did occur on the grisea and Rhizopus oryzae GA linker variants, re- sequon ThrGlyAsn introduced in this variant at the spectively, and showed that the linker region was C-terminal end of the linker just prior to SBD and Fig. 10. MALDI-TOF spectra of wild-type and linker variant GAs. The mass of the singly charged species are indicated. Calculated values for the molecular mass of the polypeptide chain are shown in brackets.
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 that additional glycosylation may also have occurred cloning and expression of the target gene could alter- elsewhere (Figs. 9 and 10).
natively ¢nd a thermostable GA. Only a few thermo- Compared to wild-type GA these linker variant stable wild-type GAs have, however, been reported GAs have lower conformational stability in solutions including the enzymes of H. grisea var. thermoidea containing denaturants and also lower heat stability, [105], A. fumigatus [106], and Thermomyces lanugino- but normal catalytic activity towards raw starch and sus [107]. Furthermore, gene sequences are known of soluble oligo- and polysaccharide substrates (J. Sau- thermostable GAs from Clostridium sp. G0005 [108], er, T. Christensen, B.W. Sigurskjold, B. Svensson, Methanococcus jannaschii [109], and Thermoanaero- unpublished). A minimum length of 17 aa of the bacterium thermosaccharolyticum [110]. These en- replacing sequence seemed to be required for produc- zymes show approximately 40% sequence identity tion of functional GA variant protein in P. pastoris to GAs from Aspergilli. A common molecular fea- (J. Sauer, B. Svensson, unpublished). The essentially ture for the thermophilic enzymes seems to be the normal function of the variant with these short link- lack of helices 9, 10, and 11 of the (K/K)6-barrel ers as compared to the longer linker of 38 aa in wild- CD [110]. Information from these and forthcoming type GA, suggests that the natural linker £exibly sequences of thermostable GAs may guide future ra- positions CD relative to SBD. It would be desirable tional protein engineering towards a GA that exhib- to identify the structural elements that control the its activity and stability at elevated temperatures.
correct positioning, secondly, to ¢nd conditions under which a single structural conformer is predom- inant, and thirdly, to identify which factors can force 5. Conclusion and perspectives the association between the domains apart.
GA has been very thoroughly described using pro- 4.6. Engineering of industrial properties tein engineering techniques to study fundamental questions in the mechanism and speci¢city and for GA is an industrially extremely important enzyme, the further development of GA for industrial appli- used in the enzymatic conversion of starch into high cations. Whereas GA research in many respects have glucose and fructose syrups [98]. Although GAs from resulted in forefront discoveries and improved under- most sources are unstable at higher temperatures in- standing in the broad ¢eld of structure and function dustrial sacchari¢cation is currently performed at of glycoside hydrolases, the development of econom- 60³C. Development of a thermostable GA, capable ical industrial enzymes have bene¢ted less from the of performing industrial sacchari¢cation at elevated vast amount of protein engineering data. Desirable temperatures, would thus be of signi¢cant impor- improvements from an industrial point of view have tance to the starch processing industry. Small been mentioned above. The major questions to ad- achievements towards a thermostable GA were ful- dress on the basic knowledge include (i) the three- ¢lled through protein engineering of the enzymes dimensional structure of an intact GA with both from A. niger and A. awamori (for a review see SBD and CD, (ii) insight into the co-operation be- [99]). Several approaches, including replacement of tween CD and SBD and the possible relevance of glycines in K-helices [100], elimination of fragile variation in domain level organisation for degrada- Asp-X bonds [101] and substitution of asparagine tion of solid starches, (iii) further investigation of the in Asn Gly sequences [102] have been pursued using reaction mechanism and its dependence on the sub- site-directed mutagenesis. The most successful strat- strate to settle the discussion on under which condi- egy applied, however, seems to be engineering of tions hydrolysis or product release would be rate additional disulphide bonds into the molecule, in- limiting and to further understand the interplay be- creasing the Tm value of GA by up to 4³C tween enzyme and substrate/inhibitors, and (iv) time- resolved structural analysis of GA and substrates Screening of thermophilic bacteria and subsequent during the catalysis.
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Crystal structure of glucoamylase from Aspergillus awamori var. X100 to 2.2-Aî resolution, J. Biol. Chem. 267 (1992) The authors are grateful for support from the 19291 19298.
[13] E.M. Harris, A.E. Aleshin, L.M. Firsov, R.B. Honzatko, Danish Research Councils' Committee on Biotech- Re¢ned structure for the complex of 1-deoxynojirimycin nology (Grant no. 9502014).
with glucoamylase from Aspergillus awamori var. X100 to 2.4-Aî resolution, Biochemistry 32 (1993) 1618 1626.
[14] A.E. Aleshin, C. Ho¡man, L.M. Firsov, R.B. Honzatko, Re¢ned crystal structures of glucoamylase from Aspergillus awamori var. X100, J. Mol. Biol. 238 (1994) 575 591.
[15] A.E. Aleshin, L.M. Firsov, R.B. Honzatko, Re¢ned struc- [1] K. Hiromi, Z.I. Hamauzu, K. Takahashi, S. Ono, Kinetic ture for the complex of acarbose with glucoamylase from studies on gluc-amylase. II. Competition between two types Aspergillus awamori var. X100 to 2.4-Aî resolution, J. Biol.
of substrate having K-1,4 and K-1,6 glucosidic linkage, Chem. 269 (1994) 15631 15639.
J. Biochem. (Tokyo) 59 (1966) 411 418.
[16] B. Sto¡er, A.E. Aleshin, L.M. Firsov, B. Svensson, R.B.
[2] K. Hiromi, K. Takahashi, Z.I. Hamauzu, S. Ono, Kinetic Honzatko, Re¢ned structure for the complex of D-gluco-di- studies on gluc-amylase. 3. The in£uence of pH on the rates hydroacarbose with glucoamylase from Aspergillus awamori of hydrolysis of maltose and panose, J. Biochem. (Tokyo) 59 var. X100 to 2.2 Aî resolution: dual conformations for ex- (1966) 469 475.
tended inhibitors bound to the active site of glucoamylase, [3] T.P. Frandsen, H.-P. Fierobe, B. Svensson, Engineering spe- FEBS Lett. 358 (1995) 57 61.
ci¢city and stability in glucoamylase from Aspergillus niger, [17] A.E. Aleshin, B. Sto¡er, L.M. Firsov, B. Svensson, R.B.
in: L. Alberghina (Ed.), Protein Engineering in Industrial Honzatko, Crystallographic complexes of glucoamylase Biotechnology, Harwood Academic, Amsterdam, 1999, pp.
with maltooligosaccharide analogs: relationship of stereo- chemical distortions at the nonreducing end to the catalytic [4] P.J. Reilly, Protein engineering of glucoamylase to improve mechanism, Biochemistry 35 (1996) 8319 8328.
industrial properties. A Review, Starch/Sta«rke 51 (1999) [18] B. Sto¡er, T.P. Frandsen, B. Svensson, M. Gajhede, X-ray structure of an active site mutant glucoamylase, [5] M.R. Sierks, B. Svensson, Protein engineering of the relative Tyr48CTrp, from Aspergillus niger, in: The Carbohydrate speci¢city of glucoamylase from Aspergillus awamori based Bioengineering Meeting,, Elsinore, Denmark, 1995, Abstract on sequence similarities between starch-degrading enzymes, Protein Eng. 7 (1994) 1479 1484.
[19] J. Sevcik, A. Solovicova, E. Hostinova, J. Gasperik, K.S.
[6] T.P. Frandsen, T. Christensen, B. Sto¡er, J. Lehmbeck, C.
Wilson, Z. Dauter, Structure of glucoamylase from Saccha- Dupont, R.B. Honzatko, B. Svensson, Mutational analysis romycopsis ¢buligera at 1.7 Aî resolution, Acta Crystallogr. D of the roles in catalysis and substrate recognition of argi- Biol. Crystallogr. 54 (1998) 854 866.
nines 54 and 305, aspartic acid 309, and tryptophan 317 [20] K. Sorimachi, A.J. Jacks, M.-F. Le Gal-Coe«¡et, G. William- located at subsites 1 and 2 in glucoamylase from Aspergillus son, D.B. Archer, M.P. Williamson, Solution structure of niger, Biochemistry 34 (1995) 10162 10169.
the granular starch binding domain of glucoamylase from [7] H.-P. Fierobe, B.B. Sto¡er, T.P. Frandsen, B. Svensson, Aspergillus niger by nuclear magnetic resonance spectrosco- Mutational modulation of substrate bond-type speci¢city py, J. Mol. Biol. 259 (1996) 970 987.
and thermostability of glucoamylase from Aspergillus awa- [21] K. Sorimachi, M.-F. Le Gal-Coe«¡et, G. Williamson, D.B.
mori by replacement with short homologue active site se- Archer, M.P. Williamson, Solution structure of the granular quences and thiol/disul¢de engineering, Biochemistry 35 starch binding domain of Aspergillus niger glucoamylase (1996) 8696 8704.
bound to L-cyclodextrin, Structure 5 (1997) 647 661.
[8] B. Henrissat, A classi¢cation of glycosyl hydrolases based on [22] G.F.H. Kramer, A.P. Gunning, V.J. Morris, N.J. Belshaw, amino acid sequence similarities, Biochem. J. 280 (1991) G. Williamson, Scanning tunnelling microscopy of Aspergil- lus niger glucoamylase, J. Chem. Soc., Faraday Trans. 89 [9] B. Henrissat, A. Bairoch, New families in the classi¢cation (1993) 2595 2602.
of glycosyl hydrolases based on amino acid sequence simi- [23] B.W. Sigurskjold, T. Christensen, N. Payre, S. Cottaz, H.
larities, Biochem. J. 293 (1993) 781 788.
Driguez, B. Svensson, Thermodynamics of binding of heter- [10] P.M. Coutinho, P.J. Reilly, Structural similarities in gluco- obidentate ligands consisting of spacer-connected acarbose amylase by hydrophobic cluster analysis, Protein Eng. 7 and L-cyclodextrin to the catalytic and starch-binding do- (1994) 749 760.
mains of glucoamylase from Aspergillus niger shows that [11] P.M. Coutinho, P.J. Reilly, Glucoamylase structural, func- the catalytic and starch-binding sites are in close proximity tional and evolutionary relationships, Proteins 29 (1997) in space, Biochemistry 37 (1998) 10446 10452.
[24] N. Payre, S. Cottaz, C. Boisset, R. Borsali, B. Svensson, B.
[12] A.E. Aleshin, A. Golubev, L.M. Firsov, R.B. Honzatko, Henrissat, H. Driguez, Dynamic light scattering evidence for BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 a ligand-induced motion between the two domains of glu- revisiae by the addition of poly(aspartic acid) tails, Enzyme coamylase G1 of Aspergillus niger using heterobivalent sub- Microb. Technol. 15 (1993) 593 600.
strate analogues, Angew. Chem. Int. Ed. 38 (1999) 974 977.
[39] D.B. Archer, D.J. Jeenes, D.A. MacKenzie, Strategies for [25] C.E. Weill, R.J. Burch, J.W. Van Dyk, An K-amyloglucosi- improving heterologous protein production from ¢lamentous dase that produces L-glucose, Cereal Chem. 31 (1954) 150 fungi, Antonie Van Leeuwenhoek 65 (1994) 245 250.
[40] B. Svensson, A.J. Clarke, I. Svendsen, H. MÖller, Identi¢ca- [26] J.H. Pazur, T. Ando, The hydrolysis of glucosyl oligosaccha- tion of carboxylic acid residues in glucoamylase that partic- rides with K-D-(1,4) and K-D-(1,6) bonds by fungal amyloglu- ipate in catalysis and substrate binding, Eur. J. Biochem. 188 cosidases, J. Biol. Chem. 235 (1960) 297 302.
(1990) 29 38.
[27] D.E. Koshland Jr., Stereochemistry and the mechanism of [41] E. Hostinova, J. Balanova, J. Gasperik, The nucleotide se- enzymatic reactions, Biol. Rev. 28 (1953) 416 436.
quence of the glucoamylase gene GLA1 from Saccharomy- [28] J.D. McCarter, S.G. Withers, Mechanisms of enzymatic gly- copsis ¢buligera KZ, FEMS Microbiol. Lett. 67 (1991) 103 coside hydrolysis, Curr. Opin. Struct. Biol. 4 (1994) 885 [42] A. Solovicova, T. Christensen, E. Hostinova, J. Gasperik, J.
[29] H.-P. Fierobe, A.J. Clarke, D. Tull, B. Svensson, Enzymatic Sevcik, B. Svensson, Structure function relationships in glu- properties of the cysteinesul¢nic acid derivative of the cata- coamylases encoded by variant Saccharomycopsis ¢buligera lytic-base mutant Glu400CCys of Glucoamylase from As- genes, Eur. J. Biochem. 264 (1999) 756 764.
pergillus awamori, Biochemistry 37 (1998) 3753 3759.
[43] A.J. Jacks, K. Sorimachi, M.F. Le Gal-Coe¡et, G. William- [30] K. Hiromi, M. Ohnishi, A. Tanaka, Subsite structure and son, D.B. Archer, M.P. Williamson, 1H and 15N assignments ligand binding mechanism of glucoamylase, Mol. Cell. Bio- and secondary structure of the starch-binding domain of chem. 51 (1983) 79 95.
glucoamylase from Aspergillus niger, Eur. J. Biochem. 233 [31] M.R. Sierks, C. Ford, P.J. Reilly, B. Svensson, Catalytic (1995) 568 578.
mechanism of fungal glucoamylase as de¢ned by mutagene- [44] N.J. Belshaw, G. Williamson, Interaction of L-cyclodextrin sis of Asp176, Glu179 and Glu180 in the enzyme from As- with the granular starch binding domain of glucoamylase, pergillus awamori, Protein Eng. 3 (1990) 193 198.
Biochim. Biophys. Acta 1078 (1991) 117 120.
[32] T.P. Frandsen, C. Dupont, J. Lehmbeck, B. Sto¡er, M.R.
[45] N.J. Belshaw, G. Williamson, Speci¢city of the binding do- Sierks, R.B. Honzatko, B. Svensson, Site-directed mutagen- main of glucoamylase 1, Eur. J. Biochem. 211 (1993) 717 esis of the catalytic base glutamic acid 400 in glucoamylase from Aspergillus niger and of tyrosine 48 and glutamine 401, [46] B.W. Sigurskjold, B. Svensson, G. Williamson, H. Driguez, both hydrogen-bonded to the Q-carboxylate group of gluta- Thermodynamics of ligand binding to the starch-binding mic acid 400, Biochemistry 33 (1994) 13808 13816.
domain of glucoamylase from Aspergillus niger, Eur. J. Bio- [33] K. Olsen, U. Christensen, M.R. Sierks, B. Svensson, Reac- chem. 225 (1994) 133 141.
tion mechanisms of Trp120CPhe and wild-type glucoamy- [47] M.P. Williamson, M.-F. Le Gal-Coe«¡et, K. Sorimachi, lases from Aspergillus niger. Interactions with maltooligodex- C.S.M. Furniss, D.B. Archer, G. Williamson, Function of trins and acarbose, Biochemistry 32 (1993) 9686 9693.
conserved tryptophans in Aspergillus niger glucoamylase 1 [34] C.R. Berland, B.W. Sigurskjold, B. Sto¡er, T.P. Frandsen, starch binding domain, Biochemistry 36 (1997) 7535 7539.
B. Svensson, Thermodynamics of inhibitor binding to mu- [48] B. Svensson, K. Larsen, I. Svendsen, E. Boel, The complete tant forms of glucoamylase from Aspergillus niger deter- amino acid sequence of the glycoprotein, glucoamylase G1, mined by isothermal titration calorimetry, Biochemistry 34 from Aspergillus niger, Carlsberg Res. Commun. 48 (1983) (1995) 10153 10161.
[35] U. Christensen, K. Olsen, B.B. Sto¡er, B. Svensson, Sub- [49] C.B. Libby, C.A. Cornett, P.J. Reilly, C. Ford, E¡ect of strate binding mechanism of Glu180CGln, Asp176CAsn, amino acid deletions in the O-glycosylated region of Asper- and wild-type glucoamylases from Aspergillus niger, Bio- gillus awamori glucoamylase, Protein Eng. 7 (1994) 1109 chemistry 35 (1996) 15009 15018.
[36] T. Christensen, B.B. Sto¡er, B. Svensson, U. Christensen, [50] T. Semimaru, M. Goto, K. Furukawa, S. Hayashida, Func- Some details of the reaction mechanism of glucoamylase tional analysis of the threonine- and serine-rich Gp-I domain from Aspergillus niger; kinetic and structural studies on of glucoamylase I from Aspergillus var. kawachi, Appl. En- Trp52CPhe and Trp317CPhe mutants, Eur. J. Biochem.
viron. Microbiol. 61 (1995) 2885 2890.
250 (1997) 638 645.
[51] T. Christensen, B. Svensson, B.W. Sigurskjold, Thermody- [37] L.J. Chen, C. Ford, A. Kusnadi, Z.L. Nikolov, Improved namics of reversible and irreversible unfolding and domain adsorption to starch of a L-galactosidase fusion protein con- interactions of glucoamylase from Aspergillus niger studied taining the starch-binding domain from Aspergillus gluco- by di¡erential scanning and isothermal titration calorimetry, amylase, Biotechnol. Prog. 7 (1991) 225 229.
Biochemistry 38 (1999) 6300 6310.
[38] I. Suominen, C. Ford, D. Stachon, H. Heimo, M. Nieder- [52] M. Goto, M. Tsukamoto, I. Kwon, K. Ekino, K. Furukawa, auer, H. Nurmela, C. Glatz, Enhanced recovery and puri¢- Functional analysis of O-linked oligosaccharides in threo- cation of Aspergillus glucoamylase from Saccharomyces ce- nine/serine-rich region of Aspergillus glucoamylase by ex- BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 pression in mannosyltransferase-disruptants of yeast, Eur. J.
lus awamori by oxidation of the Glu400CCys catalytic-base Biochem. 260 (1999) 596 602.
mutant to cysteinesul¢nic acid, Biochemistry 37 (1998) 3743 [53] A. Gunnarsson, B. Svensson, B. Nilsson, S. Svensson, Struc- tural studies on the O-glycosidically linked carbohydrate [67] A.J. Clarke, B. Svensson, Identi¢cation of an essential tryp- chains of glucoamylase G1 from Aspergillus niger, Eur. J.
tophanyl residue in the primary structure of glucoamylase Biochem. 145 (1984) 463 467.
G2 from Aspergillus niger, Carlsberg Res. Commun. 49 [54] B. Svensson, K. Larsen, A. Gunnarsson, Characterization of (1984) 559 566.
glucoamylase G2 from Aspergillus niger, Eur. J. Biochem.
[68] A.J. Clarke, B. Svensson, The Role of tryptophanyl residues 154 (1986) 497 502.
in the function of Aspergillus niger glucoamylase G1 and G2, [55] G. Williamson, N.J. Belshaw, M.P. Williamson, O-Glycosyl- Carlsberg Res. Commun. 49 (1984) 111 122.
ation in Aspergillus glucoamylase; conformation and role in [69] M.R. Sierks, C. Ford, P.J. Reilly, B. Svensson, Site-directed binding, Biochem. J. 282 (1992) 423 428.
mutagenesis at the active site Trp120 of Aspergillus awamori [56] H.-P. Fierobe, E. Mirgorodskaya, T.P. Frandsen, P. Roep- glucoamylase, Protein Eng. 2 (1989) 621 625.
stor¡, B. Svensson, Overexpression and characterization of [70] L.M. Firsov, K.N. Neustroev, A.E. Aleshin, C.M. Metzler, Aspergillus awamori wild-type and mutant glucoamylase se- D.E. Metzler, R.D. Scott, B. Sto¡er, T. Christensen, B.
creted by the methylotrophic yeast Pichia pastoris: compar- Svensson, NMR spectroscopy of exchangeable protons of ison with wild-type recombinant glucoamylase produced us- glucoamylase and of complexes with inhibitors in the 9 15- ing Saccharomyces cerevisiae and Aspergillus niger as hosts, ppm range, Eur. J. Biochem. 223 (1994) 293 302.
Protein Expr. Purif. 9 (1997) 159 170.
[71] B. Svensson, M.R. Sierks, Roles of the aromatic side chains [57] D. Penninga, B.A. van der Veen, R.M. Knegtel, S.A. van in the binding of substrates, inhibitors, and cyclomalto-oli- Hijum, H.J. Rozeboom, K.H. Kalk, B.W. Dijkstra, L.
gosaccharides to the glucoamylase from Aspergillus niger Dijkhuizen, The raw starch binding domain of cyclodextrin probed by perturbation di¡erence spectroscopy, chemical glycosyltransferase from Bacillus circulans strain 251, J. Biol.
modi¢cation, and mutagenesis, Carbohydr. Res. 227 (1992) Chem. 271 (1996) 32777 32784.
[58] B. Svensson, T.G. Pedersen, I. Svendsen, T. Sakai, M. Otte- [72] M.R. Sierks, C. Ford, P.J. Reilly, B. Svensson, Functional sen, Characterization of two forms of glucoamylase from roles and subsite locations of Leu177, Trp178 and Asn182 of Aspergillus niger, Carlsberg Res. Commun. 47 (1982) 55 Aspergillus awamori glucoamylase determined by site-di- rected mutagenesis, Protein Eng. 6 (1993) 75 79.
[59] S.M. Southall, P.J. Simpson, H.J. Gilbert, G. Williamson, [73] A. Tanaka, M. Ohnishi, K. Hiromi, Stopped-£ow kinetic M.P. Williamson, The starch-binding domain from gluco- studies on the binding of gluconolactone and maltose to amylase disrupts the structure of starch, FEBS Lett. 447 glucoamylase, Biochemistry 21 (1982) 107 113.
(1999) 58 60.
[74] K. Olsen, B. Svensson, U. Christensen, Stopped-£ow £uo- [60] M.L. Sinnott, Catalytic mechanisms of enzymatic glycosyl rescence and steady-state kinetic studies of ligand-binding transfer, Chem. Rev. 90 (1990) 1171 1202.
reactions of glucoamylase from Aspergillus niger, Eur. J.
[61] A. Konstantinidis, M.L. Sinnott, The interaction of 1-£uoro- Biochem. 209 (1992) 777 784.
D-glucopyranosyl £uoride with glucosidases, Biochem. J. 279 [75] S.K. Natarajan, M.R. Sierks, Identi¢cation of enzyme-sub- (1991) 587 593.
strate and enzyme-product complexes in the catalytic mech- [62] Y. Tanaka, W. Tao, J.S. Blanchard, E.J. Hehre, Transition anism of glucoamylase from Aspergillus awamori, Biochem- state structures for the hydrolysis of K-D-glucopyranosyl £u- istry 35 (1996) 15269 15279.
oride by retaining and inverting reactions of glycosylases, [76] M. Ohnishi, K. Hiromi, Kinetic studies on the interaction of J. Biol. Chem. 269 (1994) 32306 32312.
Rhizopus glucoamylase with maltodextrin and maltotriose, [63] T.P. Frandsen, B.B. Sto¡er, M.M. Palcic, S. Hof, B. Svens- utilizing the absorbance change near 300 nm, Carbohydr.
son, Structure and energetics of the glucoamylase isomaltose Res. 61 (1978) 335 341.
transition-state complex probed by using modeling and de- [77] M. Ohnishi, T. Matsumoto, T. Yamanaka, K. Hiromi, Bind- oxygenated substrates coupled with site-directed mutagene- ing of isomaltose and maltose to the glucoamylase from sis, J. Mol. Biol. 263 (1996) 79 89.
Aspergillus niger, as studied by £uorescence spectrophotom- [64] G. Davies, B. Henrissat, Structures and mechanisms of gly- etry and steady-state kinetics, Carbohydr. Res. 204 (1990) cosyl hydrolases, Structure 3 (1995) 853 859.
[65] S.L. Lawson, W.W. Wakarchuk, S.G. Withers, E¡ects of [78] R. Fagerstro«m, Subsite mapping of Hormoconis resinae glu- both shortening and lengthening the active site nucleophile coamylases and their inhibition by gluconolactone, J. Gen.
of Bacillus circulans xylanase on catalytic activity, Biochem- Microbiol. 137 (1991) 1001 1008.
istry 35 (1996) 10110 10118.
[79] K. Hiromi, Interpretation of dependency of rate parameters [66] H.-P. Fierobe, E. Mirgorodskaya, K.A. McGuire, P. Roep- on the degree of polymerization of substrate in enzyme-cat- stor¡, B. Svensson, A.J. Clarke, Restoration of catalytic ac- alyzed reactions. Evaluation of subsite a¤nities of exo-en- tivity beyond wild-type level in glucoamylase from Aspergil- zyme, Biochem. Biophys. Res. Commun. 40 (1970) 1 6.
BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 [80] K. Hiromi, Y. Nitta, C. Numata, S. Ono, Subsite a¤nities of glucoamylase probed by deoxygenated substrates, Biochem- glucoamylase: examination of the validity of the subsite istry 31 (1992) 8972 8977.
theory, Biochim. Biophys. Acta 302 (1973) 362 375.
[94] M.R. Sierks, B. Svensson, Kinetic identi¢cation of a hydro- [81] M.M. Meagher, Z.L. Nikolov, P.J. Reilly, Subsite mapping gen bonding pair in the glucoamylase-maltose transition of Aspergillus niger glucoamylase I and II with malto- and state complex, Protein Eng. 5 (1992) 185 188.
isomaltooligosaccharides, Biotechnol. Bioeng. 34 (1989) 681 [95] K. Bock, H. Pedersen, The substrate speci¢city of the en- zyme amyloglucosidase (AMG). Part I. Deoxy derivatives, [82] M.R. Sierks, B. Svensson, Functional roles of the invariant Acta Chem. Scand. 41 (1987) 617 628.
aspartic acid 55, tyrosine 306, and aspartic acid 309 in glu- [96] R.U. Lemieux, U. Spohr, M. Bach, D.R. Cameron, T.P.
coamylase from Aspergillus awamori studied by mutagenesis, Frandsen, B.B. Sto¡er, B. Svensson, M.M. Palcic, Chem- Biochemistry 32 (1993) 1113 1117.
ical mapping of the active site of glucoamylase of Aspergil- [83] K. Hiromi, Kinetic studies on the interaction of glucoamy- lus niger, Can. J. Chem. 74 (1996) 319 335.
lase with substrates and analogues, in: Y. Takehiko (Ed.), [97] B. Svensson, K.S. Bak-Jensen, M.T. Jensen, J. Sauer, K.W.
Enzyme Chemistry and Molecular Biology of Amylases and Rodenburg, Studies on structure, function, and protein en- Related Enzymes, CRC Press, Boca Raton, FL, 1995, pp.
gineering of starch-degrading enzymes, J. Appl. Glycosci.
46 (1999) 51 65.
[84] L. Pauling, Chemical achievement and hope for the future, [98] B.C. Saha, J.G. Zeikus, Microbial glucoamylases: Bio- Am. Scientist 36 (1948) 51 58.
chemical and biotechnological features, Sta«rke 41 (1989) [85] M.R. Sierks, B. Svensson, Catalytic mechanism of gluco- amylase probed by mutagenesis in conjunction with hydro- [99] C. Ford, Improving operating performance of glucoamy- lysis of K-D-glucopyranosyl £uoride and maltooligosaccha- lase by mutagenesis, Curr. Opin. Biotechnol. 10 (1999) rides, Biochemistry 35 (1996) 1865 1871.
[86] B.W. Sigurskjold, C.R. Berland, B. Svensson, Thermody- [100] H.M. Chen, Y. Li, T. Panda, F.U. Buehler, C. Ford, P.J.
namics of inhibitor binding to the catalytic site of glucoa- Reilly, E¡ect of replacing helical glycine residues with ala- mylase from Aspergillus niger determined by displacement nines on reversible and irreversible stability and production titration calorimetry, Biochemistry 33 (1994) 10191 10199.
of Aspergillus awamori glucoamylase, Protein Eng. 9 (1996) [87] E.J. Toone, Structure and energetics of protein-carbohydrate complexes, Curr. Opin. Struct. Biol. 4 (1994) 719 728.
[101] H.M. Chen, C. Ford, P.J. Reilly, Identi¢cation and elimi- [88] T. Weimar, B.O. Petersen, B. Svensson, B.M. Pinto, Deter- nation by site-directed mutagenesis of thermolabile aspartyl mination of the solution conformation of D-gluco-dihydro- bonds in Aspergillus awamori glucoamylase, Protein Eng. 8 acarbose, a high a¤nity inhibitor bound to glucoamylase by (1995) 575 582.
transferred NOE NMR spectroscopy, Carbohydr. Res. 326 [102] H.M. Chen, C. Ford, P.J. Reilly, Substitution of asparagine (2000) 50 55.
residues in Aspergillus awamori glucoamylase by site-di- [89] J.S. Andrews, T. Weimar, T.P. Frandsen, B. Svensson, B.M.
rected mutagenesis to eliminate N-glycosylation and inacti- Pinto, Novel disaccharides containing sulfur in the ring and vation by deamidation, Biochem. J. 301 (1994) 275 281.
nitrogen in the interglycosidic linkage. Conformation of 5P- [103] M.J. Allen, P.M. Coutinho, C.F. Ford, Stabilization of thio-4-N-maltoside bound to glucoamylase and its activity as Aspergillus awamori glucoamylase by proline substitution competitive inhibitor, J. Am. Chem. Soc. 117 (1995) 10799 and combining stabilizing mutations, Protein Eng. 11 (1998) 783 788.
[90] K.D. Randell, T.P. Frandsen, B. Sto¡er, M.A. Johnson, B.
[104] Y. Li, P.M. Coutinho, C. Ford, E¡ect on thermostability Svensson, B.M. Pinto, Synthesis and glycosidase inhibitory and catalytic activity of introducing disul¢de bonds into activity of 5-thioglucopyranosylamines. Molecular modeling Aspergillus awamori glucoamylase, Protein Eng. 11 (1998) of complexes with glucoamylase, Carbohydr. Res. 321 (1999) [105] L.R.O. Tosi, H.F. Francisco, J.A. Jorge, Puri¢cation an [91] E. Mirgorodskaya, H.-P. Fierobe, B. Svensson, P. Roep- characterization of an extracellular glucoamylase from the stor¡, Mass spectrometric identi¢cation of a stable catalytic thermophilic fungi Humicola grisea var. thermoidea, Can. J.
cysteinesul¢nic acid residue in an enzymatically active chemi- Microbiol. 39 (1993) 846 852.
cally modi¢ed glucoamylase mutant, J. Mass Spectrom. 34 [106] W. Brandani, R.M. Peralta, Puri¢cation and characteriza- (1999) 952 957.
tion of a thermostable glucoamylase from Aspergillus fumi- [92] T. Weimar, B. Sto¡er, B. Svensson, B.M. Pinto, Complexes gatus, Can. J. Microbiol. 44 (1998) 493 497.
of glucoamylase with maltoside heteroanalogues: Bound li- [107] D.-C. Li, Y.-J. Yang, Y.-L. Peng, C.-Y. Shen, Puri¢cation gand conformations by use of transferred NOE experiments and characterization of extracellular glucoamylase from the and molecular modeling, Biochemistry 39 (2000) 300 306.
thermophilic Thermomyces lanuginosus, Mycol. Res. 102 [93] M.R. Sierks, K. Bock, S. Refn, B. Svensson, Active site (1998) 568 572.
similarities of glucose dehydrogenase, glucose oxidase, and [108] H. Ohnishi, H. Kitamura, T. Minowa, H. Sakai, T. Ohta, BBAPRO 36304 12-12-00 J. Sauer et al. / Biochimica et Biophysica Acta 1543 (2000) 275 293 Molecular cloning of a glucoamylase gene from a thermo- Scott, N.S.M. Geoghagen, J.C. Venter, Complete genome philic Clostridium and kinetics of the cloned enzyme, Eur. J.
sequence of the methanogenic archaeon, Methanococcus Biochem. 207 (1992) 413 418.
jannaschii, Science 273 (1996) 1058 1073.
[109] C.J. Bult, O. White, G.J. Olsen, L. Zhou, R.D. Fleisch- [110] A. Ducki, O. Grundmann, L. Konermann, F. Mayer, M.
mann, G.G. Sutton, J.A. Blake, L.M. FitzGerald, R.A.
Hoppert, Glucoamylase from Thermoanaerobacterium ther- Clayton, J.D. Gocayne, A.R. Kerlavage, B.A. Dougherty, mosaccharolyticum: Sequence studies and analysis of mac- J.F. Tomb, M.D. Adams, C.I. Reich, R. Overbeek, E.F.
romolecular architecture of the enzyme, J. Gen. Appl. Mi- Kirkness, K.G. Weinstock, J.M. Merrick, A. Glodek, J.L.
crobiol. 44 (1998) 327 335.
BBAPRO 36304 12-12-00

Source: http://www.sut.ac.th/iat/biotech/Montarop/amylase/glucoamylase.pdf