Protein Engineering, Vol. 12, No. 10, 819-824,
October 1999
© 1999 Oxford University Press
Roles of catalytic residues in 
-amylases as evidenced by the structures of the product-complexed mutants of a maltotetraose-forming amylase
Institute for Protein Research, Osaka University, Suita, Osaka 565-0871 and 2 Hayashibara Biochemical Laboratories, Inc., Amaseminami, Okayama 700-0834, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionReferences |
|---|
The crystal structures of the four product-complexed single mutants of the catalytic residues of Pseudomonas stutzeri maltotetraose-forming
-amylase, E219G, D193N, D193G and D294N, have been determined. Possible roles of the catalytic residues Glu219, Asp193 and Asp294 have been discussed by comparing the structures among the previously determined complexed mutant E219Q and the present mutant enzymes. The results suggested that Asp193 predominantly works as the base catalyst (nucleophile), whose side chain atom lies in close proximity to the C1-atom of Glc4, being involved in the intermediate formation in the hydrolysis reaction. While Asp294 works for tightly binding the substrate to give a twisted and a deformed conformation of the glucose ring at position –1 (Glc4). The hydrogen bond between the side chain atom of Glu219 and the O1-atom of Glc4, that implies the possibility of interaction via hydrogen, consistently present throughout these analyses, supports the generally accepted role of this residue as the acid catalyst (proton donor).
Keywords: -amylase/maltotetraose-forming amylase/product-complexed mutant/reaction mechanism/X-ray structure
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionReferences |
|---|
Maltotetraose-forming
-amylase (G4-amylase) from Pseudomonas stutzeri is an exo-amylase of which the crystal structure has been determined for the uncomplexed wild-type G4-2 (the catalytic part of the G4-1 enzyme). Previous studies (Morishita et al., 1997
; Yoshioka et al., 1997) revealed that this enzyme contains a (ß/)8 barrel structure that is frequently found in other endo-type -amylases (Matsuura et al., 1984; Kadziola et al., 1994; Machius et al., 1995; Qian et al., 1995; Fujimoto et al., 1998; Strobl et al., 1998), isoamylase (Katsuya et al., 1998), cyclodextrin glucanotransferases (Kubota et al., 1991; Strokopytov et al., 1996) and also in ß-amylase (Mikami et al., 1994). Further, these studies revealed a highly conserved three-dimensional arrangement of the three catalytic residues (Katsuya et al., 1998) that are located at the C-terminal end of the ß-strands in the barrel, as well as in the amino acid sequences in various -amylases (Svensson, 1994). Thus, it now appears possible to discuss the general catalytic mechanism in the -amylase family of enzymes (Henrissat, 1991; Takata et al., 1992; Svensson, 1994).
The product-complexed two crystal forms (type 1 and 2) of the E219Q mutant of G4-amylase, where the catalytic glutamic acid has been converted to glutamine, were obtained by co-crystallization with maltopentaose. Their structure analyses (Yoshioka et al., 1997) revealed the following three points: (i) the binding mode of each of the non-reducing end four glucose residues (the product maltotetraose) after substrate cleavage; (ii) the structural mode of the recognition of the non-reducing end of the substrate amylose, that determines the exo-acting specificity of this enzyme and (iii) a deformation of a sugar ring at the cleaved reducing end. In other substrate analogue or inhibitor-complexed amylase structures (Kadziola et al., 1994; Machius et al., 1995; Qian et al., 1995; Fujimoto et al., 1998), this kind of deformation has not yet been observed, except in the case of glucoamylase complexed with D-gluco-dihydro-acabose (Stoffer et al., 1995).
Further, the fact that the catalytic residue mutated enzyme cleaved the maltopentaose and a bound maltotetraose was observed in the crystal (Yoshioka et al., 1997), implies that even such a mutant enzyme possesses non-zero activity. In G4-amylase, it has been shown that maltotetraose could also be a substrate for a transglycosylation reaction in the synthesis of a maltooligosaccharide derivative (Usui and Murata, 1988). This fact suggests the possibility of observing the near transition state of this enzyme with bound maltotetraose by using mutant enzymes. Deformation of the reducing end glucose unit in the E219Q mutant of G4-amylase may provide experimental support for this view.
We have now crystallized the product complexes of four different single mutants of the three catalytic residues Glu219, Asp193 and Asp294—E219G, D193N, D193G and D294N—and determined the structures of these complexes by the single crystal X-ray diffraction method to elucidate the detailed structural modes of sugar binding at the active site. These structures are compared and the possible roles of mutated residues in the catalytic and substrate binding mechanism are discussed. These findings may be generalized to other enzymes in the -amylase family.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionReferences |
|---|
Activity assay
The activity of each mutant G4-amylase was measured as previously described (Nakada et al., 1990), using 5 ml of 1% soluble starch in 20 mM phosphate buffer (pH 7.0) as substrate, and 0.2 ml of an enzyme solution. The reaction was carried out for 20 min at 40°C and the reducing power of maltotetraose produced was measured by the Somogyi–Nelson method. One unit of enzyme activity is defined as the amount of enzyme which liberates 1 µmol of maltotetraose per min.
Crystallization
All complexed mutants of G4-amylase were crystallized under the same conditions as those employed for the previously reported E219Q complex, using 14 mg/ml protein in 20 mM Tris–HCl buffer pH 7.5, and 0.7 to 0.9 M ammonium sulfate in the presence of 10 mg/ml maltopentaose by the hanging drop vapor diffusion technique at 20°C. The crystals were grown to a size suitable for X-ray diffraction within 1–2 weeks. All crystals obtained belonged to the orthorhombic complex type 2 form (Yoshioka et al., 1997) and had essentially the same cell dimensions, except for small but significant differences in the D193G crystal, as shown in Table I.
|
Data collection
All X-ray intensity data were collected with an imaging plate diffractometer, Rigaku RAXIS-IIC, up to 2 Å resolution by using graphite monochromatized Cu-K radiation from a rotating anode generator operated at 40 kV and 100 mA with a focus size of 0.3x0.3 mm2. A summary of data collection is given in Table I
. As shown in this table, we collected two kinds of data sets (D294N0 and D294N) for the Asp294Asn
mutant complex crystal for the reasons described in the Results and discussion.
Structure determination and refinements
Because the crystals of E219G, D193N and D294N mutants co-crystallized with maltopentaose were isomorphous with the previously reported E219Q type 2 complexed crystal (Yoshioka et al., 1997), the refinements carried out were straightforward, starting from the E219Q structure, excluding bound maltotetraose and water molecules and with a replaced mutated residue. Because the crystal of the D193G mutant was not isomorphous with the type 2 crystal, the molecular replacement method was applied by using the program AUTOMR (Matsuura, 1991). The solution of the molecular replacement was unambiguous and the refinement was carried out using the coordinates derived from the MR solution as the initial model. The molecular model of bound maltotetraose was obtained by fitting the initial model into the (2Fo – Fc) map calculated for each complexed crystal, by omitting maltotetraose using the program CHAIN (Sack, 1988). The bound water molecules were checked against the (Fo – Fc) map calculated after incorporating maltotetraose by using the program SEAWAT (Y.Matsuura, unpublished program) to obtain the starting coordinates list of water molecules. We used the restrained least-squares program PROFFT (Hendrickson and Konnert, 1980; Finzel, 1987) for refinement calculations of every complexed mutant structure. The manual model corrections for the polypeptide chain were made by using the program CHAIN with interactive computer graphics. After manual model corrections, water molecules were added or removed from the list at points of convergence of refinement based on peak height and hydrogen bond geometry, which were checked using the same system. Reflections in the lowest resolution range (d > 10 Å) were omitted from all calculations. The final refinement statistics are summarized in Table II. The average positional errors were estimated from the Luzzati plots (Luzzati, 1952) to be approximately 0.2 Å for every structure (plots not shown).
|
Coordinates
The coordinates and the structure factors of E219G, D193N, D193G and D294N complexes have been deposited in the Protein Data Bank (Bernstein et al., 1977) (accession code: 1qi4, r1qi4sf; 1qi3, r1qi3sf; 1qpk, r1qpksf; 1qi5, r1qi5sf, respectively).
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionReferences |
|---|
Mutant enzyme activities
The results of the activity measurements of the mutant enzymes are shown in Table III. All the mutants tested showed a significant decrease in activity, especially for the mutants of the three catalytic residues D193N, E219Q and D294N, where the activities were non-detectable (shown as <0.1 U/mg). These catalytic residue mutants were further checked for residual activity by increasing the concentration of enzyme >1000-fold, which resulted in a significant amount of activity. This fact agrees with the findings in the present structure analyses where maltopentaose used for co-crystallization was hydrolysed into maltotetraose that bound in the active site cleft.
|
Structural deviations from E219Q
The overall structural deviations of the polypeptide structure of each of the present mutants from that of the E219Q mutant were calculated by the least-squares method based on the main chain atoms, as shown in Table IV. The overall deviations are small, but significant positional differences (>2 Å) were observed for the side chain atoms, especially in the case of arginine residues, although these mutant proteins form nearly isomorphous crystals.
|
Structures of mutated enzyme active sites and interactions with carbohydrate
For the present structure analyses, single mutations, E219G, D193N, D193G and D294N, have been introduced at the catalytic residues which surround the reducing end glucose unit (Glc4) of the bound product maltotetraose. Major structural changes have occurred in the enzyme active site region and in Glc4. The r.m.s. deviations of the atoms from the E219Q complex in Glc4 were significantly larger (except for E219G) than those in the three non-reducing end glucose units (Glc1–Glc3), as shown in Table IV. A superimposed picture of the five complexed mutant enzymes at the active site is shown in Figure 1, depicting only the reducing end glucose unit of the bound maltotetraose and the three catalytic amino acid residues. As shown in Figure 1, the Glc4 ring is deformed, adopting a half-chair conformation in all mutant complexes except in D294N. The average position of Glc4 with respect to the enzyme active site is similar to that of the E219Q complex (Yoshioka et al., 1997) in the present E219G and D294N complexes. However, in D193G the position moved downward to the bottom of the cleft, and in D193N it moved upward from the bottom. The torsion angles about the glucosidic linkages, 
and 
, of maltotetraose for each complex structure are shown in Table V. These values are more dispersed among mutants at the Glc3–4 as compared with those at 1–2 and 2–3, but all lie in a reasonable region of conformational energy space (Tran et al., 1989).
|
|
The side chain at position 219 in the D193N, D193G and D294N mutants rotated about the C
–Cß bond relative to that in E219Q, probably because of structurally different side chains. In E219G, a water molecule Wat518 was incorporated in the space of the side chain of Glu219 in the wild type, and forms hydrogen bonds to Arg233NH1, Ile241O and Asp244OD1. Similarly, in D193G, the water molecule Wat730 occupied the space of the side chain of Asp193, and formed hydrogen bonds to O6 of Glc4. Hydrogen bonds and close contacts surrounding the Glc4 residue in the complexed mutants are summarized in Table VI. Here, the atom named O1 is used for the oxygen atom attached to C1 of the reducing end glucose residue, and O4 for that involved in glucosidic linkage formation or for the non-reducing end hydroxyl oxygen attached to C4. As shown in Table VI and Figure 2, in all complexes except D193G, water molecules (Wat591 in E219Q, Wat733 in E219G, Wat731 in D193N and Wat755 in D294N), being hydrogen-bonded to O1 of Glc4, may play a certain role. These water molecules further form hydrogen bonds to other water molecules or protein atoms, with a possibility of participating in the reaction as the water for hydrolysis in G4-amylase. On the other hand, a well conserved water molecule has been found in uncomplexed forms of various -amylases (Katsuya et al., 1998), also with a possibility of participating in the catalysis. The following three hydrogen bonds always exist: between O1 and the side chain atoms of the residues at position 219; that between O2, O3 (Glc4) and the side chain atoms at position 294; and that between O6 (Glc4) and the side chain atoms of position 193. In addition, O6 (Glc4) forms hydrogen bonds with His117NE2 in the E219Q, E219G and D193G mutants, while the distances in D193N and D294N are a little too long (3.5 and 3.6 Å, respectively) for hydrogen bond formation. The O5-atom in the ring of Glc4 seems to interact with the side chain atoms at position 193, except in D193G, where Wat730 is located closer (3.0 Å). It is uncertain, however, whether these atoms form hydrogen bonds as found in the crystals of carbohydrate molecules (Jeffrey, 1990).
|
|
In analyzing the D294N complex structure, we first used the crystal mounting technique as for the other crystals to collect intensity data (indicated as D294N0 in Table I
). However, by using these data we observed high B-values in the atoms composing maltotetraose, probably reflecting their low occupancies. Subsequently, the intensity data were collected for this complex by soaking the crystal in the capillary in a solution containing 10 mg/ml maltopentaose during data collection (D294N in Table I). After refinement of the structure of D294N, lower B-values were obtained for the maltotetraose atoms with an average value of 27.6 Å2, as compared with the value of 38.8 Å2 as found in the unsoaked crystals. The corresponding values were 20.2, 19.9, 22.1 and 17.6 Å2, for E219Q, E219G, D193N and D193G mutant crystals, respectively. For the purpose of comparison with other complexes in the present discussion, we therefore used the structure of D294N. Possible roles of the catalytic residues
The three catalytic residues Glu219, Asp193 and Asp294 in the present enzyme, and in all -amylase family enzymes (Henrissat, 1991; Takata et al., 1992; Svensson, 1994) are fully conserved in amino acid sequences and in the three-dimensional structures so far known (Katsuya et al., 1998). This suggests that these catalytic residues share their individual roles and possibly cooperate with each other in the catalytic process in the same way in all family enzymes. Possible mechanism and roles of the catalytic residues deduced from the discussion below are schematically shown in Figure 3.
|
Of these residues, Glu219 in this enzyme is currently considered to work as the general acid catalyst that donates a proton to O4 (glucosidic bond oxygen) in the first step of the general acid–base catalysis. In the present analyses, one of the side chain atoms of this residue is always hydrogen bonded to O1 of Glc4 (Table VI
), which supports this supposition. Further, this residue is supposed to be protonated at the optimum pH for catalysis (Matsuura et al., 1984), and to perform the protonation of the glucosidic O4-atom prior to cleavage of the C1–O4 bond at the position between subsites –1 and +1 (Davies et al., 1997). The second step of the catalysis is the formation of the reaction intermediate. The side chain of Asp193 is pointing toward the O1–C1–O5 triad of the cleaved reducing end (position –1) glucose unit (Glc4), being hydrogen bonded to O1 and possibly also to O5. In particular, in the present complex structure analyses (Table VI), OD1 of Asp193 is unusually close to the C1 (Glc4), the distances between them ranging from 2.6 to 3.1 Å, considerably shorter than those of van der Waals contact. It is very difficult to imagine that the C1-atom has the potential to participate in the formation of a usual hydrogen bond, even though it has an attached hydrogen atom. The Glc4 ring is distorted centered on the atom C1 in all complexed structures except in D294N. This distortion may cause the hydrogen atom attached to C1 to displace somehow in a stereochemically unnatural position, thus making it possible for the C1 and Asp193OD1 atoms to form a close contact with each other. One possible interpretation is that C1 is to some extent positively charged in the product complexed state, and that the contact between C1 and OD1 atoms demonstrate partly ionic character. On the other hand, it can also be said that this contact has in part a covalent character. Whichever interpretation is favored, the unusually close contact of these atoms and the position of the OD1 atom relative to the point of cleavage, identifies Asp193 as the most likely candidate for involvement in the reaction intermediate in the catalytic process. With respect to the nature of the chemical bond (Pauling, 1960), all bonds have an intermediate character between the two extremes, ionic or covalent, and it remains difficult to declare whether the intermediate formation proceeds in a single displacement via an oxocarbonium ion (Phillips, 1967) or in a double displacement mechanism via a covalent bond formation (Koshland, 1953) between C1 and OD1 of Asp193. The isolation of a covalently bound adduct (Kuroki et al., 1993; Uitdehaag et al., 1999) may not be valid proof of the formation of a covalent intermediate, because once isolated, a covalent bond will be formed whatever was the true intermediate. The sofa-formed glucose oxocarbonium ion intermediate could exist only under a highly stereochemically restricted reaction field. Furthermore, there is also a possibility that the intermediate may in itself cause a transition in bond character from ionic to covalent. It is implied that the intermediate formation involving an amino acid residue is not necessarily or absolutely indispensable, from the fact that even in the D193G mutant complex crystal, which has no side chain atoms to form a reaction intermediate, the maltopentaose present during the crystallization was hydrolysed to maltotetraose. A probable explanation for this mechanism is that some other component, like an ion or a water molecule, from the solvent replaces the group to form the intermediate.
As described in the previous section, in the D294N mutant complex structure, maltotetraose showed a low occupancy, as demonstrated by high B-values, in binding onto the cleft. This fact suggests that the mutation of Asp294Asn lowered the binding constant of maltotetraose, even though the side chain atoms of Asn294 seem to maintain hydrogen bonds between both O2 and O3 of Glc4. This suggestion is supported by the experiment in which the occupancy was raised (B-values lowered) by soaking the crystal in the substrate solution during data collection. As shown in Table V, in the D294N complex the torsion angle (Glc3–4) adopts a small value compared with the others, detwisting the Glc4 ring. This means that Asp294 compels to twist this ring, with the glucosidic oxygen to be cleaved (between –1 and +1) orienting toward an optimum position for the proton donation by Glu219. Furthermore, the most remarkable fact in this mutant complex is that the deformation of Glc4 was not observed. These findings, taken together, make it highly probable that Asp294 is working predominantly to tightly bind the substrate in cooperation with other amino acid residues involved in binding the substrate (Yoshioka et al., 1997). This tightening and twisting might induce the ring distortion of the reducing end glucose unit (Glc4) in order to lower the reaction potential of the hydrolysis. Thus it may be possible to define the role of Asp294 as fixing the substrate in the catalytic process to help accelerate the acid–base hydrolysis reaction (Figure 3).
|
Notes |
|---|
1 Present address: Protonic NanoMachine Project, ERATO, 3-4 Hikaridai, Seika, 619-0237, Japan
3 To whom correspondence should be addressed
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionReferences |
|---|
Bernstein,F.C., Koetzle,T.F., Williams,G.J.B., Meyer,E.F., Brice,M.D., Rodgers,J.R., Kennard,O., Shimanouchi,T. and Tasumi,M. (1977) J. Mol. Biol., 112, 535–542.[Web of Science][Medline]
Davies,G.J., Wilson,K.S. and Henrissat,B. (1997) Biochem. J., 321, 557–559.
Finzel,B.C. (1987) J. Appl. Crystallogr., 20, 53–55.
Fujimoto,Z., Takase,K. Doui,N., Momma,M., Matsumoto,T. and Mizuno,H. (1998) J. Mol. Biol., 277, 393–407.[Web of Science][Medline]
Hendrickson,W.A. and Konnert,J.H. (1980) In Diamond,R., Ramaseshan,S. and Venkatesan,K. (eds) Computing in Crystallography. Indian Academy of Science, Bangalore, pp. 13.01–25.
Henrissat,B. (1991) Biochem. J., 280, 309–316.
Jeffrey,G.A. (1990) Acta Crystallogr., B46, 89–103.
Kadziola,A., Abe,J., Svensson,B. and Haser,R. (1994) J. Mol. Biol., 239, 104–121.[Web of Science][Medline]
Katsuya,Y., Mezaki,Y., Kubota,M. and Matsuura,Y. (1998) J. Mol. Biol., 281, 885–897.[Web of Science][Medline]
Koshland,D.E. (1953) Biol. Rev., 28, 416–420.
Kubota,M., Matsuura,Y., Sakai,S. and Katsube,Y. (1991) Denpun Kagaku, 38, 141–146.
Kuroki,R., Weaver,L.H. and Matthews,B.W. (1993) Science, 262, 2030–2033.
Luzzati,P.V. (1952) Acta Crystallogr., 5, 802–810.
Machius,G., Wiegand,G. and Huber,R. (1995) J. Mol. Biol., 246, 545–559.[Web of Science][Medline]
Matsuura,Y. (1991) J. Appl. Crystallogr., 24, 1063–1066.
Matsuura,Y., Kusunoki,M., Harada,W. and Kakudo,M. (1984) J. Biochem., 95, 697–702.
Mikami,B., Degano,M., Hehre,E.J. and Sacchettini,J.C. (1994) Biochemistry, 33, 7779–7787.[Medline]
Morishita,Y., Hasegawa,K., Matsuura,Y., Katsube,Y., Kubota,M. and Sakai,S. (1997) J. Mol. Biol., 267, 661–672.[Web of Science][Medline]
Nakada,T., Kubota,M., Sakai,S. and Tsujisaka,Y. (1990) Agric. Biol. Chem., 54, 737–743.[Web of Science][Medline]
Pauling,L. (1960) In The Nature of the Chemical Bond, 3rd Edn., Chapter 3, Cornell University Press, New York.
Phillips,D.C. (1967) Proc. Roy. Acad. Sci., 57, 484–495.
Qian,M., Haser,R. and Payan,F. (1995) Protein Sci., 4, 747–755.[Web of Science][Medline]
Sack,J.S. (1988) J. Mol. Graphics, 6, 224–225.
Stoffer,B., Aleshin,A.E., Firsov,L.M., Svensson,B. and Honzatko,R.B. (1995) FEBS Lett., 358, 57–61.[Web of Science][Medline]
Strobl,S., Maskos,K., Betz,M., Wiegand,G., Huber,H., Gomis-Rüth,F.X. and Glockshuber,R. (1998) J. Mol. Biol., 278, 617–628.[Web of Science][Medline]
Strokopytov,B., Knegtel,M.A., Penninga,D., Rozeboom,H.J., Kalk,K.H., Dijkuizen,L. and Dijkstra,B.W. (1996) Biochemistry, 35, 4241–4249.[Medline]
Svensson,B. (1994) Plant Mol. Biol., 25, 141–157.[Web of Science][Medline]
Takata,H., Kuriki,T., Okada,S., Takesada,Y., Iizuka,M., Minamiura,N. and Imanaka,T. (1992) J. Biol. Chem., 267, 18447–18452.
Tran,V., Buleon,A., Imberty,A. and Perez,S. (1989) Biopolymers, 28, 679–690.
Uitdehaag,J.C.M., Mosi,R., Kalk,K.H., van der Veen,B.A., Dijkhuizen,L., Withers,S.G. and Dijkstra,B.W. (1999) Nature Struct. Biol., 6, 432–436.[Web of Science][Medline]
Usui,T. and Murata,T. (1988) J. Biochem., 103, 969–972.
Yoshioka,Y., Hasegawa,K., Matsuura,Y., Katsube,Y. and Kubota,M. (1997) J. Mol. Biol., 271, 619–628.[Web of Science][Medline]
Received May 1, 1999; revised June 8, 1999; accepted June 21, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. E. Ciocchini, L. S. Guidolin, A. C. Casabuono, A. S. Couto, N. Inon de Iannino, and R. A. Ugalde A glycosyltransferase with a length-controlling activity as a mechanism to regulate the size of polysaccharides PNAS, October 16, 2007; 104(42): 16492 - 16497. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Tatara, B. R. Lee, T. Yoshida, K. Takahashi, and E. Ichishima Identification of Catalytic Residues of Ca2+-independent 1,2-{alpha}-D-Mannosidase from Aspergillus saitoi by Site-directed Mutagenesis J. Biol. Chem., July 3, 2003; 278(28): 25289 - 25294. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||