Protein Engineering vol.5 no.2 pp. 185-188, 1992
Kinetic identification of a hydrogen bonding pair in the glucoamylase-maltose transition state complex
Michael R.Sierks1 and Birte Svensson2 Department of Chemistry, Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK 2500 Copenhagen Valby, Denmark 'Present address: Department of Chemical and Biochemical Engineering, University of Maryland, Baltimore County, Baltimore, MD 21228, USA 2
To whom correspondence should be addressed
Introduction Glucoamylase (GA) catalyzes the release of D-glucose by hydrolysis of a-1,4- and a-l,6-glucosidic linkages at the nonreducing ends of starch and related poly- and oligosaccharides (Hiromi et ai, 1983). The active site of fungal GA has been described by kinetics (Hiromi et ai, 1966a,b, 1983; Savel'ev etai, 1982), chemical modification (Clarke and Svensson, 1984b; Svensson et ai, 1990) and mutagenesis studies (Sierks et al., 1989, 1990). The substrate binding area is composed of seven consecutive subsites accommodating glucosyl residues of which subsite 2 possesses by far the strongest affinity, - 2 3 kJ/mol (Savel'ev et ai, 1982; Hiromi et ai, 1983; Sierks etai., 1989). The two catalytic carboxyl groups are situated between subsites 1 and 2 (Hiromi etai, 1966a,b, 1983). In Aspergillus niger GA, Asp 176, Glul79 and Glul80 constitute an essential acid cluster which is protected by the potent inhibitor acarbose (Svensson et al., 1990) and occurs invariantly in fungal glucoamylase (Itoh et al., 1987). Site-directed mutagenesis recently enabled assignment of Glul79 as the catalytic general acid and Aspl76 tentatively as the catalytic base. Differences between Glul80 — Gin and wild-type GA activity on a-1,4- and © Oxford University Press
Materials and methods Enzyme and substrate preparation The syntheses of 1-deoxy- and 1,2-dideoxy-D-maltose and 3-deoxy-, 6-deoxy- and parent methyl-j3-D-maltoside, have been described previously (Bock and Pedersen, 1987). D-Maltose was obtained from Merck. Construction of expression plasmids encoding the wild-type and Glul80 — Gin GA has been described elsewhere (Sierks etai., 1990). The recombinant proteins were produced in yeast (Innis et al., 1985) and isolated by affinity chromatography as published earlier (Clarke and Svensson, 1984b; Sierks et al., 1989). Activity measurements Enzyme activities were measured at 45 °C in 50 mM sodium acetate, pH 4.5. Initial rates of glucose release were determined from aliquots removed at appropriate time intervals using glucose oxidase as described elsewhere (Sierks et al., 1989). The various deoxy-D-glucose analogues produced by hydrolysis of the substrate analogues were not detected in this assay. As only limited amounts of substrate analogues were available the 185
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
Molecular recognition and site-directed mutagenesis are used in combination to identify kinetically, transition state interactions between glucoamylase (GA) and the substrate maltose. Earlier studies of mutant Glul80 — Gin GA had indicated a role in substrate binding for Glul80 (Sierks,M.R., Ford,C, Reilly,P.J. and Svensson,B. (1990) Protein Engng, 3,193-198). Here, changes in activation energies calculated from measured kcat/Km values for a series of deoxygenated maltose analogues indicate hydrogen bonding between the mutant enzyme and the 3-OH group of the reducing end sugar ring. Using the same substrate analogues and determining activation energies with wild-type GA an additional hydrogen bond with the 2-OH group of maltose is attributed to an interaction with the carboxylate Glul80. This novel combination of molecular recognition and site-directed mutagenesis enables an enzyme substrate transition state contact to be identified and characterized even without access to the three dimensional structure of the enzyme. Given the distant structural relationships between glucoamylases and several starch hydrolases (Svensson,B. (1988) FEBS Lett., 230, 72-76), such identified contacts may ultimately guide tailoring of the activity of these related enzymes. Key words: active site mutant/hydrogen-bond energy/ kinetics/molecular recognition
a-l,6-linked substrates indicated that the carboxylate form of Glul80 participates in binding at subsite 2 (Sierks et al., 1990). The highly conserved GA acid cluster region distantly resembles active site sequences in other starch hydrolases (Svensson, 1988). For example a proposed binding residue in both Taka-amylase A (Matsuura et al., 1984) and Bacillus subtilis a-amylase (Takase et al., 1990) aligns with Glul80 in GA (Svensson, 1988). Molecular recognition studies have in recent years mapped key polar groups of various carbohydrates in their respective interactions with specific antibodies and lectins (see for example Lemieux, 1984; Hindsgaul et al., 1985), the enzymes glycogen phosphorylase (Street et al., 1986) and glucoamylase (Bock and Pedersen, 1987), and a galactose binding protein (Vyas et al., 1988). In a parallel approach, site-directed mutagenesis enables mapping of functionally important amino acid residues in the protein. Here the novel combination of these two procedures is introduced to identify specific interactions between enzyme side chains and substrate structural elements. The specificity constants, kcJKm, for Glul80 — Gin and wild-type GA were determined for a series of substrate analogues where hydroxyl groups were substituted by hydrogen atoms in the reducing end ring of maltose. The change in free energy associated with each hydroxyl group can then be calculated from the relationship A(AG) = -RT MkcJKJJik^/KJy)] where x and v refer to substrates and analogues, respectively (Street etai, 1986). Assuming that the hydroxyl group is not directly involved in the catalytic step, the measured A(AG) values reflect transition state binding energy differences. Based on these results the 2-OH group in maltose was identified to interact with Glul80 of A. niger GA. While determination of both kC3t and Km might permit assignment of the binding energy difference to ground state specific or transition state specific interactions, this could not be done in the present study due to limiting substrate analogue concentrations.
M.R.Sierks and B.Svensson
a) E-O—HOH + HOH—A-S b) E-O—HOH + HOH + S c) E X — O H 2
+ HOH—B-S ;
[E-O—A-S] + H O H — O H 2 t [E-O—HOH S] + HOH [E-XB-S] + H O H — O H 2
Fig. 1. Hydrogen bonding representation according to Jencks (1969) and Hine (1972) where O and X represent binding groups of the wild-type GA = E, and A and B are binding groups from the substrate S. (a) Hydrogen bonding between the carboxylate of Glu 180 (O~), and the 2-OH group (A) of D-maltose. (b) Retained hydrogen binding of a water molecule to the carboxylate of Glu 180 (CT) in the enzyme -1,2-dideoxy-D-maltose complex, (c) Formation of a weak hydrogen bond between the 1- or 6-OH group (B) of D-maltose and group X of the enzyme resulting in loss of a good hydrogen bond (Fersht etal., 1985; Street etal., 1986).
Results and discussion The specificity constants, kcJKm, as determined for wild-type and Glu 180 — Gin GA towards maltose and maltose analogues are presented in Table I. While the specificity of wild-type enzyme actually increased for the 1-deoxy and 6-deoxy compounds relative to the parent compounds it decreased 4-fold with the 1,2-dideoxy analogue and 70-fold with the 3-deoxy analogue. The kcJKm values for Glu 180 — Gin GA decreased from 90 to 1200-fold compared with wild-type (Table I), presumably due to higher Km and lower £cat values as seen earlier in hydrolysis of D-maltose by this mutant enzyme (Sierks et al., 1990). The specificity of both mutant and wild-type GA increased 3-fold with the 6-deoxy and decreased 70-fold with the 3-deoxy substrates. Removal of the 1- and 2-OH groups, in contrast, had no significant effect on the specificity of the mutant (Table I). Clearly, while the 2-OH group is only important for substrate hydrolysis by wild-type GA, both mutant and wild-type GA depend strongly on the 3-OH group. The contribution of an individual maltose hydroxyl group to transition state stabilization can be expressed by the difference in transition state activation energy, A(AG), between the parent substrate and the deoxy analogue in question, assuming both adopt essentially the same preferred conformation and do not alter the catalytic mechanism (Street et al., 1986). The present substrate analogues and parent substrates as based on NMR spectroscopy studies on the free compounds have the same conformation except 186
Table I. Specificity constants, kcJKm (s ' mM ') for glucoamylase catalyzed hydrolysis of a series of maltose analogues Substrate
Enzyme Glu 180 - Gin
D-Maltose Methyl-/3-D-maltoside 1-Deoxy-D-maltose 1,2-Dideoxy-D-maltose Methyl-3-deoxy-|3-D-maltoside Methyl-6-deoxy-0-D-maltoside
Wild-type
2
2.14 x io6.6 x lO" 3 1.68 x io-2 2.16 x lO" 2 9.4 x io-5 2.32 x io- 2
7.3 7.6 13.8 1.9 0.11 15.0
The reactions were performed at 45°C in 50 mM sodium acetate, pH 4.5. The variations in kcJKm was < 10%.
Table II. Binding energy contribution A(AG) (kJ/mol) of individual substrate hydroxyl groups Group
Substrate set
1-OH
x = 1-deoxy-D-maltose y = D-maltose
2-OH
x = 1,2-dideoxy-D-maltose y = 1-deoxy-D-maltose
3-OH
6-OH
Enzyme Glu 180 - Gin 0.6
Wild-type -
1.7
- 0.7
5.2
x = methyl-3-deoxy-0-Dmaltoside y = methyl-|3-D-maltoside
11.2
11.3
x = methyI-6-deoxy-|3-Dmaltoside y = methyl-/3-D-maltoside
- 3.4
-
1.8
Differences in activation energies, A(AG), between substrates (JT and y) were calculated according to A(AG) = -RT]n[(kcJKm)x/(kcm/Km)y] (Wilkinson etal., 1983).
for the methyl-3-deoxy-(3-D-maltoside where a small deviation has been observed (K.Bock, personal communication). Since both wild-type and mutant GA were purified by affinity chromatography on acarbose—Sepharose (Clarke and Svensson, 1984b; Sierks et al., 1990) important active site structural features are retained in the mutant enzyme. An increased activation energy of 2 - 6 or 15 — 19 kJ/mol is expected when a hydrogen bonding group that interacts with an uncharged or a charged group respectively, is removed, in the enzyme transition state complex (Fersht et al., 1985; Street et al., 1986). The present A(AG) values (Table II) thus indicate hydrogen bonding only for the 3-OH group in the transition state complex with Glu 180 — Gin, but for both the 2- and 3-OH groups in that of wild type GA. If the Glul80 side chain is hydrogen bonding to a substrate hydroxyl group of maltose, the A(AG) calculated from comparison of the wild-type—maltose
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
specificity constant, kQJKm, was measured using low substrate concentrations where the individual fccal and Km could not be determined. The reactions obeyed the relationship kcJKm = VQ/EOSQ, where v0 is the initial rate of hydrolysis and EQ and SQ are initial concentrations of enzyme and substrate, respectively. The reported kcJKm values are averaged results using repeated samples of So concentrations differing by a factor of two. If the rates were not within 10% of being linearly related, lower substrate concentrations were used. The SQ values were typically 0.1 mM and below for wild-type GA and 0.5 mM and below for Glu 180 — Gin GA, which are well below the reported Km values for maltose obtained with these enzymes, of 1.73 and 41.4 mM, respectively (Sierks etal., 1990). Km values for substrate analogues would be expected to be similar or higher as a potential binding group has been removed from the substrate. Depending on the substrate analogue, Eo varied from 0.24 to 10.8 jtg/ml for wild-type GA and from 4.8 to 215 jtg/ml for Glu 180 — Gin GA. Total reaction times ranged from 25 to 300 min. The GA concentration was determined spectrophotometrically using e2go = 1-37 X 105 M~' cm" 1 (Clarke and Svensson, 1984a). The small difference between the effective concentrations at pH 4.5 of wild-type and mutant GA due to their different pKa values for the general acid catalyst (Sierks et al., 1990) has been ignored.
Glucoamylase Glul80-maltose-2-OH specific binding
Wild-typemaltose AAG Glu 180 - Glnmaltose
k
Wild-type 1-deoxy-Dmaltose I AAG
Wild-type1,2-dideoxyD-maltose I 17.7
AAG ^ Glu 180 - Gln0.6 1-deoxyD-maltose
AAG
AAG -0.7
11.8 V Glu 180 - Gln1,2-dideoxyD-maltose
This implies that the energy losses from removal of the bonds to the 1-OH and 2-OH groups are independent of the order in which they are removed. The interactions of GA with the 2-OH group are independent of replacement of the 1-OH group and
therefore the 1,2-dideoxy-D-maltose analogue can be used to identify conclusively the Glul80-2-OH hydrogen bond. It is interesting to compare the results obtained with the two substrates maltose and methyl-/3-D-maltoside. Wild-type GA showed essentially no difference in k^l^ values with these two substrates, whereas Glu 180 — Gin GA had decreased activity toward the methylated substrate (Table I). A calculated A(AG) of 3.1 kJ/mol for methyl-/3-D-maltoside, as compared with maltose indicated that the Glu 180 — Gin GA is sensitive to /3-linked methylation while the wild-type GA is not. It is not known whether this is a direct or indirect interaction. The 1-, 3- and 6-OH groups at the reducing end glucosyl residue are not indicated to interact directly with Glu 180. However, their roles in transition state stabilization can be illustrated by changes in activation energies for the corresponding substrate analogues (Table II). First, the large A(AG) of ~ 11 kJ/mol for the methyl-3-deoxy-/?-D-maltoside compared with the methyl-/3-D-maltoside determined for the mutant as well as wild-type GA, indicates that the hydrogen bonding pattern of the 3-OH group is both crucial for activity and insensitive to the Glu 180 — Gin mutation. Second, the size of the A(AG) suggests that the 3-OH group binds with either one charged or two uncharged side chains (Fersht et al., 1985; Street et al., 1986). This is also consistent with the earlier assignment of the 3-OH in methyl-0-D-maltoside as a key polar group (Bock and Pedersen, 1987). In addition, the small increase in binding energy with wild-type enzyme caused by removal of the 1- or the 6-OH group signifies that these groups adversely affect maltose transition state stabilization, perhaps due to either their interference with protein-carbohydrate hydrophobic interactions or participation in unfavorably long hydrogen bonds (Fersht etal., 1985) with enzyme side chains, or both (Figure lc). Neither the 1- nor the 6-OH group are influenced strongly by the Glu 180 — Gin mutation. Because maltotriose and higher oligomers lack the 1-OH group in question and are preferred substrates over maltose (Hiromi et al., 1983), it is not surprising that this group would not contribute to transition state binding. Recent work showed that the Glu 180 — Gin mutant, as compared with wild-type GA had decreased relative specificity for maltose over isomaltose by 32-fold (Sierks et al., 1990). This implies that the Glul80—2-OH hydrogen bond plays a major role in the ability of GA to select for a-1,4- over a-l,6-linked substrates. Since Glul80 aligns with proposed binding residues of other starch hydrolases (Svensson, 1988), this function in substrate bond selectivity therefore may guide rational protein engineering of GA and the related enzymes. In conclusion, a hydrogen bond between the A.niger GA Glu 180 carboxylate side chain and the 2-OH group of the reducing ring in the substrate D-maltose has been identified by a novel combination of site-directed mutagenesis and molecular recognition. Moreover, the 2- and 3-OH groups of that glucosyl residue account for the high binding affinity of subsite 2 of GA, whereas the 1- and 6-OH groups slightly counteract activity. Other mutants of GA are currently being explored using Dmaltose analogues to depict additional pairs of enzyme-substrate binding atoms or atom groups. This procedure reveals structural and energetic details of actual enzyme—substrate transition state contacts, which can very elegantly supplement crystallographic studies from which such insight is rarely obtained. Even in the absence of a three dimensional structure, identification of this type of interaction provides information on active site substrate contacts which can be utilized for enzyme design or in modelling of related enzyme-substrate complexes in known as well as predicted structures. 187
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
system with either the Glul80 — Gin-maltose system, where the protein part of the bond is removed, or the wild-type-deoxymaltose system, where the substrate part of the bond is removed, should yield a value reflecting the loss of a hydrogen bond. The A(AG) resulting from comparing the Glul80 — Gin-maltose system with the Glul80 — Gin—deoxy-maltose system, however, is expected to be around zero because the hydrogen bond has already been removed. Following this reasoning the 2-OH group of the reducing end glucosyl unit of D-maltose is concluded to donate a hydrogen bond to the carboxylate group of GA Glul80. Glul80 is negatively charged (Sierks etal, 1990), therefore the measured A(AG) of 5.2 kJ/mol (Table II) caused by loss of the hydrogen bond to the 2-OH group is lower than expected. This may, however, be explained by a maintained hydrogen bond with water as discussed below. The substitution of Glul80 by Gin, raised the activation energy for 1-deoxy-D-maltose hydrolysis by 17.7 kJ/mol, as calculated from values in Table I using A(G) = -/?71n[A:eal/A:jmut/(/tca,/ #m)wt] (Wilkinson et al., 1983) while removal of the 2-OH from this substrate increased A(AG) by only 5.2 kJ/mol (Table II; see also the scheme). The difference between these values accounts for retention of hydrogen bonded water with the enzyme. The entropy gained when bound water is expelled into bulk water, which is generally accepted as a driving force for enzyme—substrate complex formation (Fersht et al., 1985; Street et al., 1986), has been estimated to be ~ 13 kJ per mole of water released at 45°C (Fersht, 1985). Substitution of Glul80 by Gin removes a charged residue from the enzyme leaving an unsolvated dipole. Removal of the 2-OH group, however, would leave an unsolvated charged group. The remaining group, Glul80, should retain a hydrogen bond to water decreasing the change in activation energy by —13 kJ/ml. The increase in activation energy caused by substitution of the Glul80 side chain is 5.2 kJ/mol, which results from the loss of the hydrogen bond with the 2-OH group, plus 13 kJ/mol since there is no longer release of a water molecule to the bulk solvent. This accounts for the A(AG) of 17.7 kJ/mol realized by replacement of Glul80 by Gin and the 5.2 kJ/mol realized by removal of the 2-OH. It is therefore reasonable to suggest that a water molecule remains bound to Glul80 in the complex formed between enzyme and 1,2-dideoxy-D-maltose by occupying the larger space available when the 1- and 2-OH groups are substituted by hydrogens (Figure la and b). Admittedly, the best way to measure the strength of the interaction with the 2-OH group would be to use the 2-deoxy substrate analogue, but this compound is not presently available. However, as shown in the following scheme depicting A(AG) values (kJ/mol), the total free energy change in going from the wild-type—maltose system to the Glu 180 — Gin—1,2dideoxy-D-maltose system is independent of the path taken.
M.R.Sierks and B.Svensson
Acknowledgements The authors are grateful to Klaus Bock for the gift of substrate analogues and to Cetus Corporation for the gift of the glucoamylase gene, the yeast expression vector pGAC9, and the Saccharomyces cerevisiae strain, C468.
References
Received on August 28, 1991; revised and accepted on December 18, 1991
188
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
Bock.K. and Pedersen.H. (1987) Acta Chem. Scand. Ser., B41, 617-628. Clarke.A.J. and Svensson.B. (1984a) Carlsberg Res. Commun., 49, 111 - 122. Clarke.A.J. and Svensson.B. (1984b) Carlsberg Res. Commun., 49, 559-566. Fersht,A.R. (1985) Enzyme Structure and Mechanism. W.H.Freeman and Company, New York. Fersht.A.R. Shi.J., Knill-Jones,J., Lowe,D.M., Wilkinson,A.J. Blow.D.M., Brick,P., Carter.P., Waye,M.M.Y. and Winter,G. (1985) Nature, 314, 235-238. Hindsgaul,O., Khare.D.P., Bach.M. and Lemieux,R.U. (1985) Can. J. Chem., 63, 2653-2658. Hine,J. (1972)/ Am. Chem. Soc, 94, 5766-5771. Hiromi.K., Hamauzu.Z., Takahashi,K. and Ono,S. (1966a) J. Biochem., 59, 411-418. Hiromi,K., Takahashi.K., Hamauzu.Z. and Ono.S. (1966b) J. Biochem., 59, 469-475. Hiromi.K., Ohnishi.M. and Tanaka.A. (1983) Mol. Cell. Biochem., 51, 79-95. Innis.M.A., Holland.M.J., McCabe.P.C, Cole.G.E., Wittmann.V.P., Tal.R., Watt.K.W.K., Gelfand.D.H., Holland.J.P. and Meade.J.H. (1985) Science, 228, 21-26. Itoh,T., Ohtsuki.I., Yamashita,I. and Fukui,S. (1987) J. Bacterioi, 169, 4171-4176. Jencks,W.P. (1969) Catalysis in Chemistry and Enzymology. McGraw-Hill, New York. Lemieux.R.U. (1984) Proceedings from the Vlllth International Symposium on Medicinal Chemistry, 1, 329-351. Matsuura.Y., Kusunoki,M., Harada,W. and Kakudo.M. (1984)7. Biochem., 95, 697-702. Savel'ev.A.N., Sergeev.V.R. and Firsov.L.M. (1982) Biochemistry (USSR), 47, 330-336. Sierks.M.R., Ford,C, Reilly.P.J. and Svensson.B. (1989) Protein Engng, 2, 621-625. Sierks.M.R., Ford.C, Reilly.P.J. and Svensson.B. (1990) Protein Engng, 3, 193-198. Street,I.P., Armstrong,C.R. and Withers,S.G. (1986) Biochemistry, 25, 6021-6027. Svensson.B. (1988) FEBS Lett., 230, 7 2 - 7 6 . Svensson.B., Clarke.A.J., Svendsen.I. and Mpller.H. (1990) Eur. J. Biochem., 188, 2 9 - 3 8 . Takase.K., Matsumoto.T., Mizuno.H. and Yamane.K. (1990) Protein Engng, 3, 351-352. Vyas.N.K., Vyas.M.N. and Quiocho.F.A. (1988) Science, 242, 1290-1295. Wilkinson.A.J., Fersht.A.R., Blow.D.M. and Winter.G. (1983) Biochemistry, 22, 3581-3586.
Kinetic identification of a hydrogen bonding pair in the glucoamylase-maltose transition state complex
Michael R.Sierks1 and Birte Svensson2 Department of Chemistry, Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK 2500 Copenhagen Valby, Denmark 'Present address: Department of Chemical and Biochemical Engineering, University of Maryland, Baltimore County, Baltimore, MD 21228, USA 2
To whom correspondence should be addressed
Introduction Glucoamylase (GA) catalyzes the release of D-glucose by hydrolysis of a-1,4- and a-l,6-glucosidic linkages at the nonreducing ends of starch and related poly- and oligosaccharides (Hiromi et ai, 1983). The active site of fungal GA has been described by kinetics (Hiromi et ai, 1966a,b, 1983; Savel'ev etai, 1982), chemical modification (Clarke and Svensson, 1984b; Svensson et ai, 1990) and mutagenesis studies (Sierks et al., 1989, 1990). The substrate binding area is composed of seven consecutive subsites accommodating glucosyl residues of which subsite 2 possesses by far the strongest affinity, - 2 3 kJ/mol (Savel'ev et ai, 1982; Hiromi et ai, 1983; Sierks etai., 1989). The two catalytic carboxyl groups are situated between subsites 1 and 2 (Hiromi etai, 1966a,b, 1983). In Aspergillus niger GA, Asp 176, Glul79 and Glul80 constitute an essential acid cluster which is protected by the potent inhibitor acarbose (Svensson et al., 1990) and occurs invariantly in fungal glucoamylase (Itoh et al., 1987). Site-directed mutagenesis recently enabled assignment of Glul79 as the catalytic general acid and Aspl76 tentatively as the catalytic base. Differences between Glul80 — Gin and wild-type GA activity on a-1,4- and © Oxford University Press
Materials and methods Enzyme and substrate preparation The syntheses of 1-deoxy- and 1,2-dideoxy-D-maltose and 3-deoxy-, 6-deoxy- and parent methyl-j3-D-maltoside, have been described previously (Bock and Pedersen, 1987). D-Maltose was obtained from Merck. Construction of expression plasmids encoding the wild-type and Glul80 — Gin GA has been described elsewhere (Sierks etai., 1990). The recombinant proteins were produced in yeast (Innis et al., 1985) and isolated by affinity chromatography as published earlier (Clarke and Svensson, 1984b; Sierks et al., 1989). Activity measurements Enzyme activities were measured at 45 °C in 50 mM sodium acetate, pH 4.5. Initial rates of glucose release were determined from aliquots removed at appropriate time intervals using glucose oxidase as described elsewhere (Sierks et al., 1989). The various deoxy-D-glucose analogues produced by hydrolysis of the substrate analogues were not detected in this assay. As only limited amounts of substrate analogues were available the 185
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
Molecular recognition and site-directed mutagenesis are used in combination to identify kinetically, transition state interactions between glucoamylase (GA) and the substrate maltose. Earlier studies of mutant Glul80 — Gin GA had indicated a role in substrate binding for Glul80 (Sierks,M.R., Ford,C, Reilly,P.J. and Svensson,B. (1990) Protein Engng, 3,193-198). Here, changes in activation energies calculated from measured kcat/Km values for a series of deoxygenated maltose analogues indicate hydrogen bonding between the mutant enzyme and the 3-OH group of the reducing end sugar ring. Using the same substrate analogues and determining activation energies with wild-type GA an additional hydrogen bond with the 2-OH group of maltose is attributed to an interaction with the carboxylate Glul80. This novel combination of molecular recognition and site-directed mutagenesis enables an enzyme substrate transition state contact to be identified and characterized even without access to the three dimensional structure of the enzyme. Given the distant structural relationships between glucoamylases and several starch hydrolases (Svensson,B. (1988) FEBS Lett., 230, 72-76), such identified contacts may ultimately guide tailoring of the activity of these related enzymes. Key words: active site mutant/hydrogen-bond energy/ kinetics/molecular recognition
a-l,6-linked substrates indicated that the carboxylate form of Glul80 participates in binding at subsite 2 (Sierks et al., 1990). The highly conserved GA acid cluster region distantly resembles active site sequences in other starch hydrolases (Svensson, 1988). For example a proposed binding residue in both Taka-amylase A (Matsuura et al., 1984) and Bacillus subtilis a-amylase (Takase et al., 1990) aligns with Glul80 in GA (Svensson, 1988). Molecular recognition studies have in recent years mapped key polar groups of various carbohydrates in their respective interactions with specific antibodies and lectins (see for example Lemieux, 1984; Hindsgaul et al., 1985), the enzymes glycogen phosphorylase (Street et al., 1986) and glucoamylase (Bock and Pedersen, 1987), and a galactose binding protein (Vyas et al., 1988). In a parallel approach, site-directed mutagenesis enables mapping of functionally important amino acid residues in the protein. Here the novel combination of these two procedures is introduced to identify specific interactions between enzyme side chains and substrate structural elements. The specificity constants, kcJKm, for Glul80 — Gin and wild-type GA were determined for a series of substrate analogues where hydroxyl groups were substituted by hydrogen atoms in the reducing end ring of maltose. The change in free energy associated with each hydroxyl group can then be calculated from the relationship A(AG) = -RT MkcJKJJik^/KJy)] where x and v refer to substrates and analogues, respectively (Street etai, 1986). Assuming that the hydroxyl group is not directly involved in the catalytic step, the measured A(AG) values reflect transition state binding energy differences. Based on these results the 2-OH group in maltose was identified to interact with Glul80 of A. niger GA. While determination of both kC3t and Km might permit assignment of the binding energy difference to ground state specific or transition state specific interactions, this could not be done in the present study due to limiting substrate analogue concentrations.
M.R.Sierks and B.Svensson
a) E-O—HOH + HOH—A-S b) E-O—HOH + HOH + S c) E X — O H 2
+ HOH—B-S ;
[E-O—A-S] + H O H — O H 2 t [E-O—HOH S] + HOH [E-XB-S] + H O H — O H 2
Fig. 1. Hydrogen bonding representation according to Jencks (1969) and Hine (1972) where O and X represent binding groups of the wild-type GA = E, and A and B are binding groups from the substrate S. (a) Hydrogen bonding between the carboxylate of Glu 180 (O~), and the 2-OH group (A) of D-maltose. (b) Retained hydrogen binding of a water molecule to the carboxylate of Glu 180 (CT) in the enzyme -1,2-dideoxy-D-maltose complex, (c) Formation of a weak hydrogen bond between the 1- or 6-OH group (B) of D-maltose and group X of the enzyme resulting in loss of a good hydrogen bond (Fersht etal., 1985; Street etal., 1986).
Results and discussion The specificity constants, kcJKm, as determined for wild-type and Glu 180 — Gin GA towards maltose and maltose analogues are presented in Table I. While the specificity of wild-type enzyme actually increased for the 1-deoxy and 6-deoxy compounds relative to the parent compounds it decreased 4-fold with the 1,2-dideoxy analogue and 70-fold with the 3-deoxy analogue. The kcJKm values for Glu 180 — Gin GA decreased from 90 to 1200-fold compared with wild-type (Table I), presumably due to higher Km and lower £cat values as seen earlier in hydrolysis of D-maltose by this mutant enzyme (Sierks et al., 1990). The specificity of both mutant and wild-type GA increased 3-fold with the 6-deoxy and decreased 70-fold with the 3-deoxy substrates. Removal of the 1- and 2-OH groups, in contrast, had no significant effect on the specificity of the mutant (Table I). Clearly, while the 2-OH group is only important for substrate hydrolysis by wild-type GA, both mutant and wild-type GA depend strongly on the 3-OH group. The contribution of an individual maltose hydroxyl group to transition state stabilization can be expressed by the difference in transition state activation energy, A(AG), between the parent substrate and the deoxy analogue in question, assuming both adopt essentially the same preferred conformation and do not alter the catalytic mechanism (Street et al., 1986). The present substrate analogues and parent substrates as based on NMR spectroscopy studies on the free compounds have the same conformation except 186
Table I. Specificity constants, kcJKm (s ' mM ') for glucoamylase catalyzed hydrolysis of a series of maltose analogues Substrate
Enzyme Glu 180 - Gin
D-Maltose Methyl-/3-D-maltoside 1-Deoxy-D-maltose 1,2-Dideoxy-D-maltose Methyl-3-deoxy-|3-D-maltoside Methyl-6-deoxy-0-D-maltoside
Wild-type
2
2.14 x io6.6 x lO" 3 1.68 x io-2 2.16 x lO" 2 9.4 x io-5 2.32 x io- 2
7.3 7.6 13.8 1.9 0.11 15.0
The reactions were performed at 45°C in 50 mM sodium acetate, pH 4.5. The variations in kcJKm was < 10%.
Table II. Binding energy contribution A(AG) (kJ/mol) of individual substrate hydroxyl groups Group
Substrate set
1-OH
x = 1-deoxy-D-maltose y = D-maltose
2-OH
x = 1,2-dideoxy-D-maltose y = 1-deoxy-D-maltose
3-OH
6-OH
Enzyme Glu 180 - Gin 0.6
Wild-type -
1.7
- 0.7
5.2
x = methyl-3-deoxy-0-Dmaltoside y = methyl-|3-D-maltoside
11.2
11.3
x = methyI-6-deoxy-|3-Dmaltoside y = methyl-/3-D-maltoside
- 3.4
-
1.8
Differences in activation energies, A(AG), between substrates (JT and y) were calculated according to A(AG) = -RT]n[(kcJKm)x/(kcm/Km)y] (Wilkinson etal., 1983).
for the methyl-3-deoxy-(3-D-maltoside where a small deviation has been observed (K.Bock, personal communication). Since both wild-type and mutant GA were purified by affinity chromatography on acarbose—Sepharose (Clarke and Svensson, 1984b; Sierks et al., 1990) important active site structural features are retained in the mutant enzyme. An increased activation energy of 2 - 6 or 15 — 19 kJ/mol is expected when a hydrogen bonding group that interacts with an uncharged or a charged group respectively, is removed, in the enzyme transition state complex (Fersht et al., 1985; Street et al., 1986). The present A(AG) values (Table II) thus indicate hydrogen bonding only for the 3-OH group in the transition state complex with Glu 180 — Gin, but for both the 2- and 3-OH groups in that of wild type GA. If the Glul80 side chain is hydrogen bonding to a substrate hydroxyl group of maltose, the A(AG) calculated from comparison of the wild-type—maltose
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
specificity constant, kQJKm, was measured using low substrate concentrations where the individual fccal and Km could not be determined. The reactions obeyed the relationship kcJKm = VQ/EOSQ, where v0 is the initial rate of hydrolysis and EQ and SQ are initial concentrations of enzyme and substrate, respectively. The reported kcJKm values are averaged results using repeated samples of So concentrations differing by a factor of two. If the rates were not within 10% of being linearly related, lower substrate concentrations were used. The SQ values were typically 0.1 mM and below for wild-type GA and 0.5 mM and below for Glu 180 — Gin GA, which are well below the reported Km values for maltose obtained with these enzymes, of 1.73 and 41.4 mM, respectively (Sierks etal., 1990). Km values for substrate analogues would be expected to be similar or higher as a potential binding group has been removed from the substrate. Depending on the substrate analogue, Eo varied from 0.24 to 10.8 jtg/ml for wild-type GA and from 4.8 to 215 jtg/ml for Glu 180 — Gin GA. Total reaction times ranged from 25 to 300 min. The GA concentration was determined spectrophotometrically using e2go = 1-37 X 105 M~' cm" 1 (Clarke and Svensson, 1984a). The small difference between the effective concentrations at pH 4.5 of wild-type and mutant GA due to their different pKa values for the general acid catalyst (Sierks et al., 1990) has been ignored.
Glucoamylase Glul80-maltose-2-OH specific binding
Wild-typemaltose AAG Glu 180 - Glnmaltose
k
Wild-type 1-deoxy-Dmaltose I AAG
Wild-type1,2-dideoxyD-maltose I 17.7
AAG ^ Glu 180 - Gln0.6 1-deoxyD-maltose
AAG
AAG -0.7
11.8 V Glu 180 - Gln1,2-dideoxyD-maltose
This implies that the energy losses from removal of the bonds to the 1-OH and 2-OH groups are independent of the order in which they are removed. The interactions of GA with the 2-OH group are independent of replacement of the 1-OH group and
therefore the 1,2-dideoxy-D-maltose analogue can be used to identify conclusively the Glul80-2-OH hydrogen bond. It is interesting to compare the results obtained with the two substrates maltose and methyl-/3-D-maltoside. Wild-type GA showed essentially no difference in k^l^ values with these two substrates, whereas Glu 180 — Gin GA had decreased activity toward the methylated substrate (Table I). A calculated A(AG) of 3.1 kJ/mol for methyl-/3-D-maltoside, as compared with maltose indicated that the Glu 180 — Gin GA is sensitive to /3-linked methylation while the wild-type GA is not. It is not known whether this is a direct or indirect interaction. The 1-, 3- and 6-OH groups at the reducing end glucosyl residue are not indicated to interact directly with Glu 180. However, their roles in transition state stabilization can be illustrated by changes in activation energies for the corresponding substrate analogues (Table II). First, the large A(AG) of ~ 11 kJ/mol for the methyl-3-deoxy-/?-D-maltoside compared with the methyl-/3-D-maltoside determined for the mutant as well as wild-type GA, indicates that the hydrogen bonding pattern of the 3-OH group is both crucial for activity and insensitive to the Glu 180 — Gin mutation. Second, the size of the A(AG) suggests that the 3-OH group binds with either one charged or two uncharged side chains (Fersht et al., 1985; Street et al., 1986). This is also consistent with the earlier assignment of the 3-OH in methyl-0-D-maltoside as a key polar group (Bock and Pedersen, 1987). In addition, the small increase in binding energy with wild-type enzyme caused by removal of the 1- or the 6-OH group signifies that these groups adversely affect maltose transition state stabilization, perhaps due to either their interference with protein-carbohydrate hydrophobic interactions or participation in unfavorably long hydrogen bonds (Fersht etal., 1985) with enzyme side chains, or both (Figure lc). Neither the 1- nor the 6-OH group are influenced strongly by the Glu 180 — Gin mutation. Because maltotriose and higher oligomers lack the 1-OH group in question and are preferred substrates over maltose (Hiromi et al., 1983), it is not surprising that this group would not contribute to transition state binding. Recent work showed that the Glu 180 — Gin mutant, as compared with wild-type GA had decreased relative specificity for maltose over isomaltose by 32-fold (Sierks et al., 1990). This implies that the Glul80—2-OH hydrogen bond plays a major role in the ability of GA to select for a-1,4- over a-l,6-linked substrates. Since Glul80 aligns with proposed binding residues of other starch hydrolases (Svensson, 1988), this function in substrate bond selectivity therefore may guide rational protein engineering of GA and the related enzymes. In conclusion, a hydrogen bond between the A.niger GA Glu 180 carboxylate side chain and the 2-OH group of the reducing ring in the substrate D-maltose has been identified by a novel combination of site-directed mutagenesis and molecular recognition. Moreover, the 2- and 3-OH groups of that glucosyl residue account for the high binding affinity of subsite 2 of GA, whereas the 1- and 6-OH groups slightly counteract activity. Other mutants of GA are currently being explored using Dmaltose analogues to depict additional pairs of enzyme-substrate binding atoms or atom groups. This procedure reveals structural and energetic details of actual enzyme—substrate transition state contacts, which can very elegantly supplement crystallographic studies from which such insight is rarely obtained. Even in the absence of a three dimensional structure, identification of this type of interaction provides information on active site substrate contacts which can be utilized for enzyme design or in modelling of related enzyme-substrate complexes in known as well as predicted structures. 187
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
system with either the Glul80 — Gin-maltose system, where the protein part of the bond is removed, or the wild-type-deoxymaltose system, where the substrate part of the bond is removed, should yield a value reflecting the loss of a hydrogen bond. The A(AG) resulting from comparing the Glul80 — Gin-maltose system with the Glul80 — Gin—deoxy-maltose system, however, is expected to be around zero because the hydrogen bond has already been removed. Following this reasoning the 2-OH group of the reducing end glucosyl unit of D-maltose is concluded to donate a hydrogen bond to the carboxylate group of GA Glul80. Glul80 is negatively charged (Sierks etal, 1990), therefore the measured A(AG) of 5.2 kJ/mol (Table II) caused by loss of the hydrogen bond to the 2-OH group is lower than expected. This may, however, be explained by a maintained hydrogen bond with water as discussed below. The substitution of Glul80 by Gin, raised the activation energy for 1-deoxy-D-maltose hydrolysis by 17.7 kJ/mol, as calculated from values in Table I using A(G) = -/?71n[A:eal/A:jmut/(/tca,/ #m)wt] (Wilkinson et al., 1983) while removal of the 2-OH from this substrate increased A(AG) by only 5.2 kJ/mol (Table II; see also the scheme). The difference between these values accounts for retention of hydrogen bonded water with the enzyme. The entropy gained when bound water is expelled into bulk water, which is generally accepted as a driving force for enzyme—substrate complex formation (Fersht et al., 1985; Street et al., 1986), has been estimated to be ~ 13 kJ per mole of water released at 45°C (Fersht, 1985). Substitution of Glul80 by Gin removes a charged residue from the enzyme leaving an unsolvated dipole. Removal of the 2-OH group, however, would leave an unsolvated charged group. The remaining group, Glul80, should retain a hydrogen bond to water decreasing the change in activation energy by —13 kJ/ml. The increase in activation energy caused by substitution of the Glul80 side chain is 5.2 kJ/mol, which results from the loss of the hydrogen bond with the 2-OH group, plus 13 kJ/mol since there is no longer release of a water molecule to the bulk solvent. This accounts for the A(AG) of 17.7 kJ/mol realized by replacement of Glul80 by Gin and the 5.2 kJ/mol realized by removal of the 2-OH. It is therefore reasonable to suggest that a water molecule remains bound to Glul80 in the complex formed between enzyme and 1,2-dideoxy-D-maltose by occupying the larger space available when the 1- and 2-OH groups are substituted by hydrogens (Figure la and b). Admittedly, the best way to measure the strength of the interaction with the 2-OH group would be to use the 2-deoxy substrate analogue, but this compound is not presently available. However, as shown in the following scheme depicting A(AG) values (kJ/mol), the total free energy change in going from the wild-type—maltose system to the Glu 180 — Gin—1,2dideoxy-D-maltose system is independent of the path taken.
M.R.Sierks and B.Svensson
Acknowledgements The authors are grateful to Klaus Bock for the gift of substrate analogues and to Cetus Corporation for the gift of the glucoamylase gene, the yeast expression vector pGAC9, and the Saccharomyces cerevisiae strain, C468.
References
Received on August 28, 1991; revised and accepted on December 18, 1991
188
Downloaded from http://peds.oxfordjournals.org/ at University of Bath Library & Learning Centre on July 21, 2015
Bock.K. and Pedersen.H. (1987) Acta Chem. Scand. Ser., B41, 617-628. Clarke.A.J. and Svensson.B. (1984a) Carlsberg Res. Commun., 49, 111 - 122. Clarke.A.J. and Svensson.B. (1984b) Carlsberg Res. Commun., 49, 559-566. Fersht,A.R. (1985) Enzyme Structure and Mechanism. W.H.Freeman and Company, New York. Fersht.A.R. Shi.J., Knill-Jones,J., Lowe,D.M., Wilkinson,A.J. Blow.D.M., Brick,P., Carter.P., Waye,M.M.Y. and Winter,G. (1985) Nature, 314, 235-238. Hindsgaul,O., Khare.D.P., Bach.M. and Lemieux,R.U. (1985) Can. J. Chem., 63, 2653-2658. Hine,J. (1972)/ Am. Chem. Soc, 94, 5766-5771. Hiromi.K., Hamauzu.Z., Takahashi,K. and Ono,S. (1966a) J. Biochem., 59, 411-418. Hiromi,K., Takahashi.K., Hamauzu.Z. and Ono.S. (1966b) J. Biochem., 59, 469-475. Hiromi.K., Ohnishi.M. and Tanaka.A. (1983) Mol. Cell. Biochem., 51, 79-95. Innis.M.A., Holland.M.J., McCabe.P.C, Cole.G.E., Wittmann.V.P., Tal.R., Watt.K.W.K., Gelfand.D.H., Holland.J.P. and Meade.J.H. (1985) Science, 228, 21-26. Itoh,T., Ohtsuki.I., Yamashita,I. and Fukui,S. (1987) J. Bacterioi, 169, 4171-4176. Jencks,W.P. (1969) Catalysis in Chemistry and Enzymology. McGraw-Hill, New York. Lemieux.R.U. (1984) Proceedings from the Vlllth International Symposium on Medicinal Chemistry, 1, 329-351. Matsuura.Y., Kusunoki,M., Harada,W. and Kakudo.M. (1984)7. Biochem., 95, 697-702. Savel'ev.A.N., Sergeev.V.R. and Firsov.L.M. (1982) Biochemistry (USSR), 47, 330-336. Sierks.M.R., Ford,C, Reilly.P.J. and Svensson.B. (1989) Protein Engng, 2, 621-625. Sierks.M.R., Ford.C, Reilly.P.J. and Svensson.B. (1990) Protein Engng, 3, 193-198. Street,I.P., Armstrong,C.R. and Withers,S.G. (1986) Biochemistry, 25, 6021-6027. Svensson.B. (1988) FEBS Lett., 230, 7 2 - 7 6 . Svensson.B., Clarke.A.J., Svendsen.I. and Mpller.H. (1990) Eur. J. Biochem., 188, 2 9 - 3 8 . Takase.K., Matsumoto.T., Mizuno.H. and Yamane.K. (1990) Protein Engng, 3, 351-352. Vyas.N.K., Vyas.M.N. and Quiocho.F.A. (1988) Science, 242, 1290-1295. Wilkinson.A.J., Fersht.A.R., Blow.D.M. and Winter.G. (1983) Biochemistry, 22, 3581-3586.
