Insights into the catalytic roles of the polypeptide regions in the active site of thermolysin and generation of the thermolysin variants with high activity and stability

The active site of thermolysin is composed of one zinc ion and five polypeptide regions [N-terminal sheet (Asn112-Trp115), alpha-helix 1 (Val139-Thr149), C-terminal loop 1 (Asp150-Gly162), alpha-helix 2 (Ala163-Val176) and C-terminal loop 2 (Gln225-Ser234)]. To explore their catalytic roles, we introduced single amino-acid substitutions into these regions by site-directed mutagenesis and examined their effects on the activity and stability. Seventy variants, in which one of the twelve residues (Ala113, Phe114, Trp115, Asp150, Tyr157, Gly162, Ile168, Ser169, Asp170, Asn227, Val230 and Ser234) was replaced, were produced in Escherichia coli. The hydrolytic activities of thermolysin for N-[3-(2-furyl)acryloyl]-Gly-l-Leu amide (FAGLA) and casein revealed that the N-terminal sheet and alpha-helix 2 were critical in catalysis and the C-terminal loops 1 and 2 were in substrate recognition. Twelve variants were active for both substrates. In the hydrolysis of FAGLA and N-carbobenzoxy-L-Asp-L-Phe methyl ester, the k(cat)/K(m) values of the D150E (in which Asp150 is replaced with Glu) and I168A variants were 2-3 times higher than those of the wild-type (WT) enzyme. Thermal inactivation of thermolysin at 80 degrees C was greatly suppressed with the D150H, D150W, I168A, I168H, N227A, N227H and S234A. The evidence might provide the insights into the activation and stabilization of thermolysin.     

Structural and functional characterization of rabbit and human L-gulonate 3-dehydrogenase

L-Gulonate 3-dehydrogenase (GDH) catalyzes the NAD(+)-linked dehydrogenation of L-gulonate into dehydro-L-gulonate in the uronate cycle. In this study, we isolated the enzyme and its cDNA from rabbit liver, and found that the cDNA is identical to that for rabbit lens lambda-crystallin except for lacking a codon for Glu(309). The same cDNA species, but not the lambda-crystallin cDNA with the codon for Glu(309), was detected in the lens, which showed the highest GDH activity among rabbit tissues. In addition, recombinant human lambda-crystallin that lacks Glu(309) displays enzymatic properties similar to rabbit GDH. These data indicate that GDH is recruited as lambda-crystallin without gene duplication. An outstanding feature of GDH is modulation of its activity by low concentrations of P(i), which decreases the catalytic efficiency in a dose dependent manner. P(i) also protects the enzyme against both thermal and urea denaturation. Kinetic analysis suggests that P(i) binds to both the free enzyme and its NAD(H)-complex in the sequential ordered mechanism. Furthermore, we examined the roles of Asp(36), Ser(124), His(145), Glu(157 )and Asn(196) in the catalytic function of rabbit GDH by site-directed mutagenesis. The D36R mutation leads to a switch in favor of NADP(H) specificity, suggesting an important role of Asp(36) in the coenzyme specificity. The S124A mutation decreases the catalytic efficiency 500-fold, and the H145Q, N196Q and N195D mutations result in inactive enzyme forms, although the E157Q mutation produces no large kinetic alteration. Thus, Ser(124), His(145) and Asn(196) may be critical for the catalytic function of GDH.     

Stabilization of flavobacterium meningosepticum glycerol kinase by introduction of a hydrogen bond

The thermostability of Flavobacterium meningosepticum glycerol kinase was increased by the change from Ser329 to Asp [Protein Eng., 14, 663-667 (2001)]. Based on a three-dimensional structure model of the mutant, we have postulated that a new charged-neutral hydrogen bond was formed between Asp329 and Ser414, and the formation of the hydrogen bond contributed to the stabilization of the tertiary structure and increased thermostability of the mutant enzyme. If the postulation is the case, FGK thermostabilization would be possible similarly by the single amino acid substitution from Ser414 to another amino acid which could form the hydrogen bond with Ser329. We did a single amino acid substitution of the wild-type enzyme from Ser414 to Asn. As we expected, S414N showed comparable thermostability to that of S329D. On the other hand, a difference in kinetic properties for ATP between S414N and S329D was observed.     

Two basic residues of the h-VPAC1 receptor second transmembrane helix are essential for ligand binding and signal transduction

We mutated the vasoactive intestinal peptide (VIP) Asp(3) residue and two VPAC(1) receptor second transmembrane helix basic residues (Arg(188) and Lys(195)). VIP had a lower affinity for R188Q, R188L, K195Q, and K195I VPAC(1) receptors than for VPAC(1) receptors. [Asn(3)] VIP and [Gln(3)] VIP had lower affinities than VIP for VPAC(1) receptors but higher affinities for the mutant receptors; the two basic amino acids facilitated the introduction of the negatively charged aspartate inside the transmembrane domain. The resulting interaction was necessary for receptor activation. 1/[Asn(3)] VIP and [Gln(3)] VIP were partial agonists at VPAC(1) receptors; 2/VIP did not fully activate the K195Q, K195I, R188Q, and R188L VPAC(1) receptors; a VIP analogue ([Arg(16)] VIP) was more efficient than VIP at the four mutated receptors; and [Asn(3)] VIP and [Gln(3)] VIP were more efficient than VIP at the R188Q and R188L VPAC(1) receptors; 3/the [Asp(3)] negative charge did not contribute to the recognition of the VIP(1) antagonist, [AcHis(1),D-Phe(2),Lys(15),Arg(16),Leu(27)] VIP ()/growth hormone releasing factor (8-27). This is the first demonstration that, to activate the VPAC(1) receptor, the Asp(3) side chain of VIP must penetrate within the transmembrane domain, in close proximity to two highly conserved basic amino acids from transmembrane 2.     

DNA polymerase beta: contributions of template-positioning and dNTP triphosphate-binding residues to catalysis and fidelity

The specific catalytic roles of two groups of DNA polymerase beta active site residues identified from crystal structures were investigated: residues possibly involved in DNA template positioning (Lys280, Asn294, and Glu295) and residues possibly involved in binding the triphosphate moiety of the incoming dNTP (Arg149, Ser180, Arg183, and Ser188). Eight site-specific mutants were constructed: K280A, N294A, N294Q, E295A, R149A, S180A, R183A, and S188A. Two-dimensional NMR analysis was employed to show that the global conformation of the mutants has not been perturbed significantly. Pre-steady-state kinetic analyses with single-nucleotide gapped DNA substrates were then performed to obtain the rate of catalysis at saturating dNTP (k(pol)), the apparent dissociation constant for dNTP (K(d)), catalytic efficiency k(pol)/K(d), and fidelity. Of the three template-positioning residues, Asn294 and Glu295 (but not Lys280) contribute significantly to k(pol). Taken together with other data, the results suggest that these two residues help to stabilize the transition state during catalysis even though they interact with the DNA template backbone rather than directly with the incoming dNTP or the opposite base on the template. Furthermore, the fidelity increases by up to 19-fold for N294Q due to differential k(pol) effects between correct and incorrect nucleotides. Of the four potential triphosphate-binding residues, Ser180 and Arg183 contribute significantly to k(pol) while the effects of R149A are relatively small and are primarily on K(d), and Ser188 appears to play a minimal role in the catalysis by Pol beta. These results identify several residues important for catalysis and quantitate the contributions of each of those residues. The functional data are discussed in relation to the prediction on the basis of available crystal structures.     

Stabilization of Flavobacterium meningosepticum Glycerol Kinase by Introduction of a Hydrogen Bond

The thermostability of Flavobacterium meningosepticum glycerol kinase was increased by the change from Ser329 to Asp [Protein Eng., 14, 663-667 (2001)]. Based on a three-dimensional structure model of the mutant, we have postulated that a new charged-neutral hydrogen bond was formed between Asp329 and Ser414, and the formation of the hydrogen bond contributed to the stabilization of the tertiary structure and increased thermostability of the mutant enzyme. If the postulation is the case, FGK thermostabilization would be possible similarly by the single amino acid substitution from Ser414 to another amino acid which could form the hydrogen bond with Ser329. We did a single amino acid substitution of the wild-type enzyme from Ser414 to Asn. As we expected, S414N showed comparable thermostability to that of S329D. On the other hand, a difference in kinetic properties for ATP between S414N and S329D was observed.     

Electrophysiological analysis of the mutated Na,K-ATPase cation binding pocket

Na,K-ATPase mediates net electrogenic transport by extruding three Na+ ions and importing two K+ ions across the plasma membrane during each reaction cycle. We mutated putative cation coordinating amino acids in transmembrane hairpin M5-M6 of rat Na,K-ATPase: Asp776 (Gln, Asp, Ala), Glu779 (Asp, Gln, Ala), Asp804 (Glu, Asn, Ala), and Asp808 (Glu, Asn, Ala). Electrogenic cation transport properties of these 12 mutants were analyzed in two-electrode voltage-clamp experiments on Xenopus laevis oocytes by measuring the voltage dependence of K+-stimulated stationary currents and pre-steady-state currents under electrogenic Na+/Na+ exchange conditions. Whereas mutants D804N, D804A, and D808A hardly showed any Na+/K+ pump currents, the other constructs could be classified according to the [K+] and voltage dependence of their stationary currents; mutants N776A and E779Q behaved similarly to the wild-type enzyme. Mutants E779D, E779A, D808E, and D808N had in common a decreased apparent affinity for extracellular K+. Mutants N776Q, N776D, and D804E showed large deviations from the wild-type behavior; the currents generated by mutant N776D showed weaker voltage dependence, and the current-voltage curves of mutants N776Q and D804E exhibited a negative slope. The apparent rate constants determined from transient Na+/Na+ exchange currents are rather voltage-independent and at potentials above -60 mV faster than the wild type. Thus, the characteristic voltage-dependent increase of the rate constants at hyperpolarizing potentials is almost absent in these mutants. Accordingly, dislocating the carboxamide or carboxyl group of Asn776 and Asp804, respectively, decreases the extracellular Na+ affinity.     

5'-methylthioadenosine nucleosidase from yellow lupine (Lupinus luteus): molecular characterization and mutational analysis

This is report of mutational analysis of higher plant 5'-methylthioadenosine nucleosidase (MTAN). We identified and characterized the gene encoding yellow lupine (Lupinus luteus) MTAN (LlMTAN). The role of active site amino acids residues Glu24, Phe134, Glu188 and Asp211 was analyzed by site-directed mutagenesis. The Glu24Gln and Asp211Asn substitutions completely abolished the enzyme activity. The Glu188Gln mutant showed only trace activity toward 5'-methylthioadenosine. These results indicate that these three amino acid residues are necessary for enzyme activity. Furthermore, as the result of replacement of Phe134 by less bulky leucine, LlMTAN acquired the ability to bind and hydrolyze S-adenosylhomocysteine. We also analyzed the sequence of the LlMTAN promoter region. It appeared that there may be a direct link between LlMTAN expression regulation and sulfate metabolism.     

Roles of Cys148 and Asp179 in catalysis by deoxycytidylate hydroxymethylase from bacteriophage T4 examined by site-directed mutagenesis.

The proposed roles of Cys148 and Asp179 in deoxycytidylate (dCMP) hydroxymethylase (CH) have been tested using site-directed mutagenesis. CH catalyzes the formation of 5-(hydroxymethyl)-dCMP, essential for DNA synthesis in phage T4, from dCMP and methylenetetrahydrofolate. CH resembles thymidylate synthase (TS), an enzyme of known three-dimensional structure, in both amino acid sequence and the reaction catalyzed. Conversion of Cys148 to Asp, Gly, or Ser decreases CH activity at least 10(5)-fold, consistent with a nucleophilic role for Cys148 (analogous to the catalytic Cys residue in TS). In crystalline TS, hydrogen bonds connect O4 and N3 of the substrate dUMP to the side-chain amide of an Asn; the corresponding residue in CH is Asp179. Conversion of Asp179 to Asn reduces the value of kcat/KM for dCMP by (1.5 x 10(4))-fold and increases the value of kcat/KM for dUMP by 60-fold; as a result, CH(D179N) has a slight preference for dUMP. Wild-type CH and CH(D179N) are covalently inactivated by 5-fluoro-dUMP, a mechanism-based inactivator of TS. Asp179 is proposed to stabilize covalent catalytic intermediates, by protonating N3 of the pyrimidine-CH adduct.     

Metabolism of some amino acids in relation to the photorespiratory nitrogen cycle of pea leaves

¹⁵N-labelled (amino group) asparagine (Asn), glutamate (Glu), alanine (Ala), aspartate (Asp) and serine (Ser) were used to study the metabolic role and the participation of each compound in the photorespiratory N cycle ofPisum sativum L. leaves. Asparagine was utilised as a nitrogen source by either deamidation or transamination, Glu was converted to Gln through NH₃ assimilation and was a major amino donor for transamination, and Ala was utilised by transamination to a range of amino acids. Transamination also provided a pathway for Asp utilisation, although Asp was also used as a substrate for Asn synthesis. In the photorespiratory synthesis of glycine (Gly), Ser, Ala, Glu and Asn acted as sources of amino-N, contributing, in the order given, 38, 28, 23, and 7% of the N for glycine synthesis; Asp provided less than 4% of the amino-N in glycine. Calculations based on the incorporation of¹⁵N into Gly indicated that about 60% (Ser), 20% (Ala), 12% (Glu) and 11% (Asn) of the N metabolised from each amino acid was utilised in the photorespiratory nitrogen cycle.Abbreviations Ala alamine - Asn asparagine - Asp aspartate - Glu glutamate - MOA methoxylamine - Ser serine     

Thermal denaturation of eukaryotic class 1 translation termination factor eRF1. Relationship between stability of the eRF1 molecule and alteration of functional activity of its mutants

Thermal denaturation of eukaryotic class-1 translation termination factor eRF1 and its mutants was examined using differential scanning microcalorimetry (DSK). Changes of free energy caused by mutants in the N domain of human eRF1 were calculated. Melting of eRF1 molecule composed of three individual domains is cooperative. Some amino acid substitutions did not affect protein thermostability and in some other cases even slightly stabilize the protein globule. These imply that these amino acid residues are not involved in maintenance of the 3D structure of human eRF1. Thus, in Glu55Asp, Tyr125Phe, Asn61Ser, Glu55Arg, Glu55A1a, Asn61Ser + Ser64Asp, Cys127Ala and Ser64Asp mutants selective inactivation of release activity is not caused by a destabilization of protein 3D structure and, most likely, is associated with local stereochemical changes introduced by substitutions of amino acid side chains in the functionally essential sites of N-domain molecule. Some residues (Asn129, Phe131) as shown by calorimetric measurements are essential for preservation of stable protein structure, but at the same time they affect selective stop codon recognition probably via their neighboring amino acids. Recognition of UAG and UAA stop codons in vitro is more sensitive to preservation of protein stability than the UGA recognition.     

Biosynthesis of Biotin-vitamers from Oleic Acid by Cryptococcus neoformans

Chitosanase (ChoA) from Mitsuaria chitosanitabida 3001 was successfully evolved with secretion efficiency and thermal stability. The inactive ChoA mutant (G151D) gene was used to mutate by an error-prone PCR technique and mutant genes that restored chitosanase activity were isolated. Two desirable mutants, designated M5S and M7T, were isolated. Two amino acids, Leu74 and Val75, in the signal peptide of ChoA were changed to Gln and Ile respectively in the M7T mutant, in addition to the G151D mutation. The L74Q/V75I double ChoA mutant was 1.5-fold higher in specific activity than wild-type ChoA due to efficient secretion of ChoA. One amino acid Asn222 was changed to Ser in the M5S mutant in addition to the G151D mutation. The N222S single ChoA mutant was 1.2-fold higher in specific activity and showed a 17% increase in thermal stability at 50 °C as compared with wild-type ChoA. This is the first study to achieve an evolutional increase in enzyme capability among chitosanses.     

Directed evolution to enhance secretion efficiency and thermostability of chitosanase from Mitsuaria chitosanitabida 3001

Chitosanase (ChoA) from Mitsuaria chitosanitabida 3001 was successfully evolved with secretion efficiency and thermal stability. The inactive ChoA mutant (G151D) gene was used to mutate by an error-prone PCR technique and mutant genes that restored chitosanase activity were isolated. Two desirable mutants, designated M5S and M7T, were isolated. Two amino acids, Leu74 and Val75, in the signal peptide of ChoA were changed to Gln and Ile respectively in the M7T mutant, in addition to the G151D mutation. The L74Q/V75I double ChoA mutant was 1.5-fold higher in specific activity than wild-type ChoA due to efficient secretion of ChoA. One amino acid Asn222 was changed to Ser in the M5S mutant in addition to the G151D mutation. The N222S single ChoA mutant was 1.2-fold higher in specific activity and showed a 17% increase in thermal stability at 50 degrees C as compared with wild-type ChoA. This is the first study to achieve an evolutional increase in enzyme capability among chitosanses.     

Asp83, Glu113 and Glu134 are not specifically involved in Schiff base protonation or wavelength regulation in bovine rhodopsin

Site-specific mutagenesis was employed to investigate the proposed contribution of proton-donating residues (Glu, Asp) in the membrane domains of bovine rhodopsin to protonation of the Schiff base-linking protein and chromophore or to wavelength modulation of this visual pigment. Three point-mutations were introduced to replace the highly conserved residues Asp83 by Asn (D83N), Glu113 by Gln (E113 Q) or Glu134 by Asp (E134D), respectively. All 3 substitutions had only marginal effects on the spectral properties of the final pigment (less than or equal to 3 nm blue-shift relative to native rhodopsin). Hence, none of these residues by itself is specifically involved in Schiff base protonation or wavelength modulation of bovine rhodopsin.     

Dispensability of glutamic acid 48 and aspartic acid 134 for Mn2+-dependent activity of Escherichia coli ribonuclease HI

The activities of the eight mutant proteins of Escherichia coli RNase HI, in which the four carboxylic amino acids (Asp(10), Glu(48), Asp(70), and Asp(134)) involved in catalysis are changed to Asn (Gln) or Ala, were examined in the presence of Mn(2+). Of these proteins, the E48A, E48Q, D134A, and D134N proteins exhibited the activity, indicating that Glu(48) and Asp(134) are dispensable for Mn(2+)-dependent activity. The maximal activities of the E48A and D134A proteins were comparable to that of the wild-type protein. However, unlike the wild-type protein, these mutant proteins exhibited the maximal activities in the presence of >100 microM MnCl(2), and their activities were not inhibited at higher Mn(2+) concentrations (up to 10 mM). The wild-type protein contains two Mn(2+) binding sites and is activated upon binding of one Mn(2+) ion at site 1 at low ( approximately 1 microM) Mn(2+) concentrations. This activity is attenuated upon binding of a second Mn(2+) ion at site 2 at high (>10 microM) Mn(2+) concentrations. The cleavage specificities of the mutant proteins, which were examined using oligomeric substrates at high Mn(2+) concentrations, were identical to that of the wild-type protein at low Mn(2+) concentrations but were different from that of the wild-type protein at high Mn(2+) concentrations. These results suggest that one Mn(2+) ion binds to the E48A, E48Q, D134A, and D134N proteins at site 1 or a nearby site with weaker affinities. The binding analyses of the Mn(2+) ion to these proteins in the absence of the substrate support this hypothesis. When Mn(2+) ion is used as a metal cofactor, the Mn(2+) ion itself, instead of Glu(48) and Asp(134), probably holds water molecules required for activity.     

A catalytic defect in mitochondrial respiratory chain complex I due to a mutation in NDUFS2 in a patient with Leigh syndrome

In this study, we investigated the pathogenicity of a homozygous Asp446Asn mutation in the NDUFS2 gene of a patient with a mitochondrial respiratory chain complex I deficiency. The clinical, biochemical, and genetic features of the NDUFS2 patient were compared with those of 4 patients with previously identified NDUFS2 mutations. All 5 patients presented with Leigh syndrome. In addition, 3 out of 5 showed hypertrophic cardiomyopathy. Complex I amounts in the patient carrying the Asp446Asn mutation were normal, while the complex I activity was strongly reduced, showing that the NDUFS2 mutation affects complex I enzymatic function. By contrast, the 4 other NDUFS2 patients showed both a reduced amount and activity of complex I. The enzymatic defect in fibroblasts of the patient carrying the Asp446Asn mutation was rescued by transduction of wild type NDUFS2. A 3-D model of the catalytic core of complex I showed that the mutated amino acid residue resides near the coenzyme Q binding pocket. However, the K(M) of complex I for coenzyme Q analogs of the Asp446Asn mutated complex I was similar to the K(M) observed in other complex I defects and in controls. We propose that the mutation interferes with the reduction of coenzyme Q or with the coupling of coenzyme Q reduction with the conformational changes involved in proton pumping of complex I.     

Functional importance of calcium binding sites in outer membrane phospholipase A

Outer membrane phospholipase A (OMPLA) is an integral membrane enzyme that hydrolyses phospholipids requiring Ca(2+) as cofactor. In vitro studies have shown that OMPLA is only active as a dimer. The structures of monomeric and dimeric OMPLA provided possible clues to the activation process. In the inhibited dimeric species calcium ions are located at the dimer interface ideally suited to stabilise the oxyanion intermediates formed during catalysis. The side chain hydroxyl function of Ser152 is one of the ligands of this interfacial calcium. In the crystal structure of monomeric OMPLA the interfacial calcium site is lacking, but calcium was found to bind at a site involving the carboxylates of Asp149 and Asp184. In the current study the relevance of the identified calcium sites has been studied by site-directed mutagenesis. The Ser152Asn variant confirmed the importance of the interfacial calcium site for catalysis, and also demonstrated that this site is essentially involved in the dimerisation process. Replacements of the ligands in monomeric OMPLA, i.e. Asp149Asn, Asp149Ala and Asp184Asn, only showed minor effects on catalytic activity and dimerisation. A stronger effect observed for the variant Asp184Ala was explained by the proximity of Asp184 to the catalytically important Ser152 residue. We propose that Asp149 and Asp184 provide an electronegative funnel that may facilitate Ca(2+) transfer to the interfacial calcium site.     

Functional importance of calcium binding sites in outer membrane phospholipase A

Outer membrane phospholipase A (OMPLA) is an integral membrane enzyme that hydrolyses phospholipids requiring Ca²⁺ as cofactor. In vitro studies have shown that OMPLA is only active as a dimer. The structures of monomeric and dimeric OMPLA provided possible clues to the activation process. In the inhibited dimeric species calcium ions are located at the dimer interface ideally suited to stabilise the oxyanion intermediates formed during catalysis. The side chain hydroxyl function of Ser152 is one of the ligands of this interfacial calcium. In the crystal structure of monomeric OMPLA the interfacial calcium site is lacking, but calcium was found to bind at a site involving the carboxylates of Asp149 and Asp184. In the current study the relevance of the identified calcium sites has been studied by site-directed mutagenesis. The Ser152Asn variant confirmed the importance of the interfacial calcium site for catalysis, and also demonstrated that this site is essentially involved in the dimerisation process. Replacements of the ligands in monomeric OMPLA, i.e. Asp149Asn, Asp149Ala and Asp184Asn, only showed minor effects on catalytic activity and dimerisation. A stronger effect observed for the variant Asp184Ala was explained by the proximity of Asp184 to the catalytically important Ser152 residue. We propose that Asp149 and Asp184 provide an electronegative funnel that may facilitate Ca²⁺ transfer to the interfacial calcium site.     

Mass spectrometric characterization of isoform variants of peanut (Arachis hypogaea) stem lectin (SL-I)

Matrix assisted laser desorption/ionization–time-of-flight (MALDI–TOF) mass spectrometric (MS) analysis of purified Arachis hypogaea stem lectin (SL-I) and its tryptic digests suggested it to be an isoformic glucose/mannose binding lectin. Two-dimensional gel electrophoresis of SL-I indicated six isoforms (A1–A6), which were confirmed by Western blotting and MALDI–TOF MS analysis. Comparative analysis of peptide mass spectra of the isoforms matched with A. hypogaea lectins with three different accession numbers (Q43376_ARAHY, Q43377_ARAHY, Q70DJ5_ARAHY). Tandem mass spectrometric (MS/MS) analysis of tryptic peptides revealed these to be isoformic variants with altered amino acid sequences. Among the peptides, the peptide T12 showed major variation. The ¹⁹⁹Val–Ser–Tyr–Asn²⁰² sequence in peptide T12 of A1 and A2 was replaced by ¹⁹⁹Leu–Ser–His–Glu²⁰² in A3 and A4 (T12′) while in A5 and A6 this sequence was ¹⁹⁹Val–Ser–Tyr–Val²⁰² (T12″). Peptide T1 showed the presence of ¹⁰Asn in the isoforms A1–A5 while in A6 this amino acid was replaced by ¹⁰Lys (T1′). Overall amino acid sequence as identified by MS/MS showed a high degree of similarity between A1, A2 and among A3, A4, A5. Carbohydrate binding domain and adenine binding site seem to be conserved.