Structures of D-xylose isomerase from Arthrobacter strain B3728 containing the inhibitors xylitol and D-sorbitol at 2.5 A and 2.3 A resolution, respectively

Xylose isomerase from Piromyces sp. E2 is a promiscuous enzyme with epimerase activity

Xylose isomerase catalyzes the isomerization of D-xylose to D-xylulose with promiscuous activity for other saccharides including D-glucose, D-allose, and L-arabinose. The xylose isomerase from the fungus Piromyces sp. E2 (PirE2_XI) is used to engineer xylose usage by the fermenting yeast Saccharomyces cerevisiae, but its biochemical characterization is poorly understood with divergent catalytic parameters reported. We have measured the kinetic parameters of the PirE2_XI and analyzed its thermostability and pH-dependence towards different substrates. The PirE2_XI shows promiscuous activity towards D-xylose, D-glucose, D-ribose and L-arabinose with variable effects depending on different divalent ions and epimerizes D-xylose at C3 to produce D-ribulose in a substrate/product dependent ratio. The enzyme follows Michaelis-Menten kinetics for the substrates used and although K_M values for D-xylose are comparable at 30 and 60 °C, the k_cat/K_M is three-fold greater at 60 °C. The purified PirE2_XI shows maximal activity at 65 °C in the pH range of 6.5-7.5 and is a thermostable enzyme, maintaining full activity over 48 h at 30 °C or 12 h at 60 °C. This is the first report demonstrating epimerase activity of the PirE2_XI and its ability to isomerize D-ribose and L-arabinose, and provides a comprehensive in vitro study of substrate specificity, effect of metal ions and temperature on enzyme activity and these findings advance the knowledge of the mechanism of action of this enzyme.

Room-Temperature Structure of Xylitol-Bound Glucose Isomerase by Serial Crystallography: Xylitol Binding in the M1 Site Induces Release of Metal Bound in the M2 Site

Glucose isomerase (GI) is an important enzyme that is widely used in industrial applications, such as in the production of high-fructose corn syrup or bioethanol. Studying inhibitor effects on GI is important to deciphering GI-specific molecular functions, as well as potential industrial applications. Analysis of the existing xylitol-bound GI structure revealed low metal occupancy at the M2 site; however, it remains unknown why this phenomenon occurs. This study reports the room-temperature structures of native and xylitol-bound GI from Streptomyces rubiginosus (SruGI) determined by serial millisecond crystallography. The M1 site of native SruGI exhibits distorted octahedral coordination; however, xylitol binding results in the M1 site exhibit geometrically stable octahedral coordination. This change results in the rearrangement of metal-binding residues for the M1 and M2 sites, the latter of which previously displayed distorted metal coordination, resulting in unstable coordination of Mg²⁺ at the M2 site and possibly explaining the inducement of low metal-binding affinity. These results enhance the understanding of the configuration of the xylitol-bound state of SruGI and provide insights into its future industrial application.

Metal Dependence of the Xylose Isomerase from Piromyces sp. E2 Explored by Activity Profiling and Protein Crystallography

Xylose isomerase from Piromyces sp. E2 (PirXI) can be used to equip Saccharomyces cerevisiae with the capacity to ferment xylose to ethanol. The biochemical properties and structure of the enzyme have not been described even though its metal content, catalytic parameters, and expression level are critical for rapid xylose utilization. We have isolated the enzyme after high-level expression in Escherichia coli, analyzed the metal dependence of its catalytic properties, and determined 12 crystal structures in the presence of different metals, substrates, and substrate analogues. The activity assays revealed that various bivalent metals can activate PirXI for xylose isomerization. Among these metals, Mn²⁺ is the most favorable for catalytic activity. Furthermore, the enzyme shows the highest affinity for Mn²⁺, which was established by measuring the activation constants (K_act) for different metals. Metal analysis of the purified enzyme showed that in vivo the enzyme binds a mixture of metals that is determined by metal availability as well as affinity, indicating that the native metal composition can influence activity. The crystal structures show the presence of an active site similar to that of other xylose isomerases, with a d-xylose binding site containing two tryptophans and a catalytic histidine, as well as two metal binding sites that are formed by carboxylate groups of conserved aspartates and glutamates. The binding positions and conformations of the metal-coordinating residues varied slightly for different metals, which is hypothesized to contribute to the observed metal dependence of the isomerase activity.

Crystal structure of glucose isomerase in complex with xylitol inhibitor in one metal binding mode

Glucose isomerase (GI) is an intramolecular oxidoreductase that interconverts aldoses and ketoses. These characteristics are widely used in the food, detergent, and pharmaceutical industries. In order to obtain an efficient GI, identification of novel GI genes and substrate binding/inhibition have been studied. Xylitol is a well-known inhibitor of GI. In Streptomyces rubiginosus, two crystal structures have been reported for GI in complex with xylitol inhibitor. However, a structural comparison showed that xylitol can have variable conformation at the substrate binding site, e.g., a nonspecific binding mode. In this study, we report the crystal structure of S. rubiginosus GI in a complex with xylitol and glycerol. Our crystal structure showed one metal binding mode in GI, which we presumed to represent the inactive form of the GI. The metal ion was found only at the M1 site, which was involved in substrate binding, and was not present at the M2 site, which was involved in catalytic function. The O₂ and O₄ atoms of xylitol molecules contributed to the stable octahedral coordination of the metal in M1. Although there was no metal at the M2 site, no large conformational change was observed for the conserved residues coordinating M2. Our structural analysis showed that the metal at the M2 site was not important when a xylitol inhibitor was bound to the M1 site in GI. Thus, these findings provided important information for elucidation or engineering of GI functions.

Expression, purification, crystallization and preliminary X-ray diffraction analysis of Bifidobacterium adolescentis xylose isomerase

Xylose isomerase (EC 5.3.1.5) is a key enzyme in xylose metabolism which is industrially important for the transformation of glucose and xylose into fructose and xylulose, respectively. The Bifidobacterium adolescentis xylA gene (NC_008618.1) encoding xylose isomerase (XI) was cloned and the enzyme was overexpressed in Escherichia coli. Purified recombinant XI was crystallized using the sitting-drop vapour-diffusion method with polyethylene glycol 3350 as the precipitating agent. A complete native data set was collected to 1.7 Å resolution using a synchrotron-radiation source. The crystals belonged to the orthorhombic space group P21212, with unit-cell parameters a = 88.78, b = 123.98, c = 78.63 Å.

Identification of critical residues for the activity and thermostability of Streptomyces sp. SK glucose isomerase

The role of residue 219 in the physicochemical properties of D-glucose isomerase from Streptomyces sp. SK strain (SKGI) was investigated by site-directed mutagenesis and structural studies. Mutants G219A, G219N, and G219F were generated and characterized. Comparative studies of their physicochemical properties with those of the wild-type enzyme highlighted that mutant G219A displayed increased specific activity and thermal stability compared to that of the wild-type enzyme, while for G219N and G219F, these properties were considerably decreased. A double mutant, SKGI F53L/G219A, displayed a higher optimal temperature and a higher catalytic efficiency than both the G219A mutant and the wild-type enzyme and showed a half-life time of about 150 min at 85 °C as compared to 50 min for wild-type SKGI. Crystal structures of SKGI wild-type and G219A enzymes were solved to 1.73 and 2.15 Å, respectively, and showed that the polypeptide chain folds into two structural domains. The larger domain consists of a (β/α)8 unit, and the smaller domain forms a loop of α helices. Detailed analyses of the three-dimensional structures highlighted minor but important changes in the active site region as compared to that of the wild-type enzyme leading to a displacement of both metal ions, and in particular that in site M2. The structural analyses moreover revealed how the substitution of G219 by an alanine plays a crucial role in improving the thermostability of the mutant enzyme.

D-xylose isomerase from a marine bacterium, Vibrio sp. strain XY-214, and D-xylulose production from β-1,3-xylan

The xylA gene from a marine bacterium, Vibrio sp. strain XY-214, encoding D-xylose isomerase (XylA) was cloned and expressed in Escherichia coli. The xylA gene consisted of 1,320-bp nucleotides encoding a protein of 439 amino acids with a predicted molecular weight of 49,264. XylA was classified into group II xylose isomerases. The native XylA was estimated to be a homotetramer with a molecular mass of 190 kDa. The purified recombinant XylA exhibited maximal activity at 60°C and pH 7.5. Its apparent K (m) values for D-xylose and D-glucose were 7.93 and 187 mM, respectively. Furthermore, we carried out D-xylulose production from β-1,3-xylan, a major cell wall polysaccharide component of the killer alga Caulerpa taxifolia. The synergistic action of β-1,3-xylanase (TxyA) and β-1,3-xylosidase (XloA) from Vibrio sp. strain XY-214 enabled efficient saccharification of β-1,3-xylan to D-xylose. D-xylose was then converted to D-xylulose by using XylA from the strain XY-214. The conversion rate of D-xylose to D-xylulose by XylA was found to be approximately 40% in the presence of 4 mM sodium tetraborate after 2 h of incubation. These results demonstrated that TxyA, XloA, and XylA from Vibrio sp. strain XY-214 are useful tools for D-xylulose production from β-1,3-xylan. Because D-xylulose can be used as a source for ethanol fermentation by yeast Saccharomyces cerevisiae, the present study will provide a basis for ethanol production from β-1,3-xylan.

Metal ion roles and the movement of hydrogen during reaction catalyzed by D-xylose isomerase: a joint x-ray and neutron diffraction study

Conversion of aldo to keto sugars by the metalloenzyme D-xylose isomerase (XI) is a multistep reaction that involves hydrogen transfer. We have determined the structure of this enzyme by neutron diffraction in order to locate H atoms (or their isotope D). Two studies are presented, one of XI containing cadmium and cyclic D-glucose (before sugar ring opening has occurred), and the other containing nickel and linear D-glucose (after ring opening has occurred but before isomerization). Previously we reported the neutron structures of ligand-free enzyme and enzyme with bound product. The data show that His54 is doubly protonated on the ring N in all four structures. Lys289 is neutral before ring opening and gains a proton after this; the catalytic metal-bound water is deprotonated to hydroxyl during isomerization and O5 is deprotonated. These results lead to new suggestions as to how changes might take place over the course of the reaction.

Multi-frequency high-field EPR studies on metal-substituted xylose isomerase

The bacterial enzyme D-xylose isomerase (XI) catalyses the conversion of D-xylose to D-xylulose. Each subunit of the homotetrameric protein contains a bimetallic active centre requiring divalent metal ions such as Mg2+, Mn2+ or Co2+ for catalytic activity. We report here on XI in which the metal binding site 1 is specifically loaded with EPR active Mn2+, while binding site 2 is occupied by Co2+ or Cd2+, rendering a catalytically active or inactive species respectively. The Q-band (34 GHz) EPR spectra of these mixed-metal samples (Co2+/Mn2+ and Cd2+/Mn2+ XI) show a clear influence of the metal in site 2 on the Mn2+ EPR parameters. Likewise, a systematic increase of the zero field splitting parameters (zfs) of Mn2+ in site 1 upon incubation with the inhibitor xylitol or substrates for both mixed-metal samples is found. For Co2+/Mn2+ XI complexed with substrate, a drastic line broadening of the central -1/2 <--> +1/2 transition is observed in Q-band EPR, which was not amenable to analysis so far. By means of a multi-frequency approach at frequencies beyond Q-band, the relevant zfs parameters were derived from spectral simulations of EPR spectra measured at 94, 285 and 670 GHz. It is shown that parallel to the increase of the D-value its distribution also grows considerably in going from free Co2+/Mn2+ XI to the species complexed with inhibitor or substrate. For XI with bound substrate, D-values in the range of 70-90 mT and a distribution of about 30 mT were found from simulation trials. The large distribution in zfs values is thought to be correlated to the structural disorder induced by the shift of the metal ion of site 2 into a location necessary for the isomerisation reaction. The results are discussed with respect to high-resolution crystal data.

Influence of divalent cations on the structural thermostability and thermal inactivation kinetics of class II xylose isomerases

The effects of divalent metal cations on structural thermostability and the inactivation kinetics of homologous class II d-xylose isomerases (XI; EC 5.3.1.5) from mesophilic (Escherichia coli and Bacillus licheniformis), thermophilic (Thermoanaerobacterium thermosulfurigenes), and hyperthermophilic (Thermotoga neapolitana) bacteria were examined. Unlike the three less thermophilic XIs that were substantially structurally stabilized in the presence of Co2+ or Mn2+ (and Mg2+ to a lesser extent), the melting temperature [(Tm) approximately 100 degrees C] of T. neapolitana XI (TNXI) varied little in the presence or absence of a single type of metal. In the presence of any two of these metals, TNXI exhibited a second melting transition between 110 degrees C and 114 degrees C. TNXI kinetic inactivation, which was non-first order, could be modeled as a two-step sequential process. TNXI inactivation in the presence of 5 mm metal at 99-100 degrees C was slowest in the presence of Mn2+[half-life (t(1/2)) of 84 min], compared to Co2+ (t(1/2) of 14 min) and Mg2+ (t(1/2) of 2 min). While adding Co2+ to Mg2+ increased TNXI's t(1/2) at 99-100 degrees C from 2 to 7.5 min, TNXI showed no significant activity at temperatures above the first melting transition. The results reported here suggest that, unlike the other class II XIs examined, single metals are required for TNXI activity, but are not essential for its structural thermostability. The structural form corresponding to the second melting transition of TNXI in the presence of two metals is not known, but likely results from cooperative interactions between dissimilar metals in the two metal binding sites.

High-level functional expression of a fungal xylose isomerase: the key to efficient ethanolic fermentation of xylose by Saccharomyces cerevisiae?

Evidence is presented that xylose metabolism in the anaerobic cellulolytic fungus Piromyces sp. E2 proceeds via a xylose isomerase rather than via the xylose reductase/xylitol-dehydrogenase pathway found in xylose-metabolising yeasts. The XylA gene encoding the Piromyces xylose isomerase was functionally expressed in Saccharomyces cerevisiae. Heterologous isomerase activities in cell extracts, assayed at 30 degrees C, were 0.3-1.1 micromol min(-1) (mg protein)(-1), with a Km for xylose of 20 mM. The engineered S. cerevisiae strain grew very slowly on xylose. It co-consumed xylose in aerobic and anaerobic glucose-limited chemostat cultures at rates of 0.33 and 0.73 mmol (g biomass)(-1) h(-1), respectively.

Isolation and characterization of a xylose-glucose isomerase from a new strain Streptomyces thermovulgaris 127, var. 7-86

A thermostable D-xylose–glucose isomerase was isolated from the thermophilic strain Streptomyces thermovulgaris 127, var. 7-86, as a result of mutagenic treatment by γ-irradiation of the parent strain, by precipitation and sequential chromatographies on DEAE–Sephadex A50, TSK-gel, FPLC-Mono Q/HR, and Superose 12™ columns. The N-terminal amino acid sequence and amino acid analysis shows 73–92% homology with xylose–glucose isomerases from other sources. The native molecular mass, determined by gel filtration on a Superose 12™ column, is 180 kDa, and 44.6 and 45 kDa were calculated, based on amino acid analysis and 10% SDS-PAGE, respectively. Both, the activity and stability of the enzyme were investigated toward pH, temperature, and denaturation with guanidine hydrochloride. The enzyme activity showed a clear pH optimum between pH 7.2 and 9.0 with D-glucose and 7.4 and 8.3 with D-xylose as substrates, respectively. The enzyme is active up to 60–85°C at pH 7.0, using D-glucose, and up to 50–60°C at pH 7.6, using D-xylose as substrates. The activation energy (Ea = 46 kJ·mol–1) and the critical temperature (Tc = 60°C) were determined by fluorescence spectroscopy. Tc is in close coincidence with the melting temperature of denaturation (Tm = 59°C), determined by circular dichroism (CD) spectroscopy. The free energy of stabilization in water after denaturation with Gdn.HCl was calculated to be 12 kJ·mol–1. The specific activity (km values) for D-xylose-glucose isomerase at 70°C toward different substrates, D-xylose, D-glucose, and D-ribose, were determined to be 4.4, 55.5, and 13.3 mM, recpectively.Key words: D-xylose-glucose isomerase, protein sequencing, protein stability, protein denaturation.

Glucose isomerase: insights into protein engineering for increased thermostability

Thermostable glucose isomerases are desirable for production of 55% fructose syrups at >90 degrees C. Current commercial enzymes operate only at 60 degrees C to produce 45% fructose syrups. Protein engineering to construct more stable enzymes has so far been relatively unsuccessful, so this review focuses on elucidation of the thermal inactivation pathway as a future guide. The primary and tertiary structures of 11 Class 1 and 20 Class 2 enzymes are compared. Within each class the structures are almost identical and sequence differences are few. Structural differences between Class 1 and Class 2 are less than previously surmised. The thermostabilities of Class 1 enzymes are essentially identical, in contrast to previous reports, but in Class 2 they vary widely. In each class, thermal inactivation proceeds via the tetrameric apoenzyme, so metal ion affinity dominates thermostability. In Class 1 enzymes, subunit dissociation is not involved, but there is an irreversible conformational change in the apoenzyme leading to a more thermostable inactive tetramer. This may be linked to reversible conformational changes in the apoenzyme at alkaline pH arising from electrostatic repulsions in the active site, which break a buried Arg-30-Asp-299 salt bridge and bring Arg-30 to the surface. There is a different salt bridge in Class 2 enzymes, which might explain their varying thermostability. Previous protein engineering results are reviewed in light of these insights.

Genomic and expression analyses of alternatively spliced transcripts of the MLL septin-like fusion gene (MSF) that map to a 17q25 region of loss in breast and ovarian tumors

We previously defined a common region of 17q25 loss in breast and ovarian tumors, suggesting localization of at least one putative tumor suppressor gene. Genomic clones from the interval were used to isolate candidate transcripts. One novel transcript had strong homology to a septin family of GTPase genes involved in cytokinesis. This gene was recently identified as a myeloid/lymphoid leukemia (MLL) fusion protein partner in acute myeloid leukemia and was named MSF (MLL septin-like fusion). As this gene may play roles in both leukemogenesis and tumorigenesis, it is essential to understand its structure and normal expression. We cloned two human alternative transcripts and identified a third database variant of MSF. RNA expression studies with a probe common to the three novel sequences showed differential expression of 4.0- and 3.0-kb transcripts in all adult and fetal tissues tested. A probe spanning sequence unique to one MSF variant detected specific expression of the 4.0-kb transcript in all tissues. Another probe unique to a different MSF variant detected a 4.0-kb transcript only in skeletal muscle. Proteins of 422 and 586 amino acids were predicted from the novel alternate transcripts and included both a xylose isomerase 1 domain and a GTPase domain. Nine common exons, three alternatively spliced exons, and six polymorphisms were identified.

Crystal structures of thermostable xylose isomerases from Thermus caldophilus and Thermus thermophilus: possible structural determinants of thermostability

The crystal structures of highly thermostable xylose isomerases from Thermus thermophilus (TthXI) and Thermus caldophilus (TcaXI), both with the optimum reaction temperature of 90 degrees C, have been determined by X-ray crystallography. The model of TcaXI has been refined to an R-factor of 17.8 % for data extending to 2.3 A and that of TthXI to 17.1 % for data extending to 2.2 A. The tetrameric arrangement of subunits characterized by the 222-symmetry and the tertiary fold of each subunit in both TcaXI and TthXI are basically the same as in other xylose isomerases. Each monomer is composed of two domains. Domain I (residues 1 to 321) folds into the (beta/alpha)8-barrel. Domain II (residues 322 to 387), lacking beta-strands, makes extensive contacts with domain I of an adjacent subunit. Each monomer of TcaXI contains ten beta-strands, 15 alpha-helices, and six 310-helices, while that of TthXI contains ten beta-strands, 16 alpha-helices, and five 310-helices. Although the electron density does not indicate the presence of bound metal ions in the present models of both TcaXI and TthXI, the active site residues show the conserved structural features. In order to understand the structural basis for thermostability of these enzymes, their structures have been compared with less thermostable XIs from Arthrobacter B3728 and Actinoplanes missouriensis (AXI and AmiXI), with the optimum reaction temperatures of 80 degrees C and 75 degrees C, respectively. Analyses of various factors that may affect protein thermostability indicate that the possible structural determinants of the enhanced thermostability of TcaXI/TthXI over AXI/AmiXI are (i) an increase in ion pairs and ion-pair networks, (ii) a decrease in the large inter-subunit cavities, (iii) a removal of potential deamidation/isoaspartate formation sites, and (iv) a shortened loop.

Thermozymes

Enzymes synthesized by thermophiles (organisms with optimal growth temperatures > 60 degrees C) and hyperthermophiles (optimal growth temperatures > 80 degrees C) are typically thermostable (resistant to irreversible inactivation at high temperatures) and thermophilic (optimally active at high temperatures, i.e., > 60 degrees C). These enzymes, called thermozymes, share catalytic mechanisms with their mesophilic counterparts. When cloned and expressed in mesophilic hosts, thermozymes usually retain their thermal properties, suggesting that these properties are genetically encoded. Sequence alignments, amino acid content comparisons, and crystal structure comparisons indicate that thermozymes are, indeed, very similar to mesophilic enzymes. No obvious sequence or structural features account for enzyme thermostability and thermophilicity. Thermostability and thermophilicity molecular mechanisms are varied, differing from enzyme to enzyme. Thermostability and thermophilicity are usually caused by the accumulation of numerous subtle sequence differences. This review concentrates on the mechanisms involved in enzyme thermostability and thermophilicity. Their relationships with protein rigidity and flexibility and with protein folding and unfolding are discussed. Intrinsic stabilizing forces (e.g., salt bridges, hydrogen bonds, hydrophobic interactions) and extrinsic stabilizing factors are examined. Finally, thermozymes' potential as catalysts for industrial processes and specialty uses are discussed, and lines of development (through new applications, and protein engineering) are also proposed.

Thermotoga neapolitana homotetrameric xylose isomerase is expressed as a catalytically active and thermostable dimer in Escherichia coli

The xylA gene from Thermotoga neapolitana 5068 was expressed in Escherichia coli. Gel filtration chromatography showed that the recombinant enzyme was both a homodimer and a homotetramer, with the dimer being the more abundant form. The purified native enzyme, however, has been shown to be exclusively tetrameric. The two enzyme forms had comparable stabilities when they were thermoinactivated at 95 degrees C. Differential scanning calorimetry revealed thermal transitions at 99 and 109.5 degrees C for both forms, with an additional shoulder at 91 degrees C for the tetramer. These results suggest that the association of the subunits into the tetrameric form may have little impact on the stability and biocatalytic properties of the enzyme.

Sequence similarity between xylose isomerase and replicase: another TIM-barrel in the replicase structure?

The BLAST search using the strand beta 2 (46_GAHGVTFHDDDLIP) of the (alpha/beta)8-barrel of xylose isomerase from Streptomyces olivochromogenes resulted in retrieving the sequentially similar segment of replicase from garlic latent virus (692_GGHGIGFHRDD). The detailed analysis of the entire amino acid sequences of both xylose isomerase and replicase suggested that the polypeptide segment 644-1046 of replicase (the entire length of this enzyme is 1924 residues) could share the structure of xylose isomerase (20.7% identity using the entire sequence of xylose isomerase). The relatedness of replicase and xylose isomerase is supported by the fact that the sequence similarity can be observed along the whole sequence of xylose isomerase (386 amino acid residues). The sequence of replicase exhibits moreover the similarity with that of lycopene cyclase, an enzyme implicated in the beta-carotene biosynthesis, that was previously found to share similarity with xylose isomerase. Thus the relevant segment of replicase is predicted to adopt an (alpha/beta)8-barrel topology similar to that of xylose isomerase.

Role of electrostatics at the catalytic metal binding site in xylose isomerase action: Ca(2+)-inhibition and metal competence in the double mutant D254E/D256E

The catalytic metal binding site of xylose isomerase from Arthrobacter B3728 was modified by protein engineering to diminish the inhibitory effect of Ca2+ and to study the competence of metals on catalysis. To exclude Ca2+ from Site 2 a double mutant D254E/D256E was designed with reduced space available for binding. In order to elucidate structural consequences of the mutation the binary complex of the mutant with Mg2+ as well as ternary complexes with bivalent metal ions and the open-chain inhibitor xylitol were crystallized for x-ray studies. We determined the crystal structures of the ternary complexes containing Mg2+, Mn2+, and Ca2+ at 2.2 to 2.5 A resolutions, and refined them to R factors of 16.3, 16.6, and 19.1, respectively. We found that all metals are liganded by both engineered glutamates as well as by atoms O1 and O2 of the inhibitor. The similarity of the coordination of Ca2+ to that of the cofactors as well as results with Be2+ weaken the assumption that geometry differences should account for the catalytic noncompetence of this ion. Kinetic results of the D254E/D256E mutant enzyme showed that the significant decrease in Ca2+ inhibition was accompanied by a similar reduction in the enzymatic activity. Qualitative argumentation, based on the protein electrostatic potential, indicates that the proximity of the negative side chains to the substrate significantly reduces the electrostatic stabilization of the transition state. Furthermore, due to the smaller size of the catalytic metal site, no water molecule, coordinating the metal, could be observed in ternary complexes of the double mutant. Consequently, the proton shuttle step in the overall mechanism should differ from that in the wild type. These effects can account for the observed decrease in catalytic efficiency of the D254E/D256E mutant enzyme.

The reaction pathway of the isomerization of D-xylose catalyzed by the enzyme D-xylose isomerase: a theoretical study

Different pathways of the metal-induced isomerization of D-xylose to D-xylulose are investigated and compared in detail using energy minimization and molecular dynamics simulation. Two theoretical models are constructed for the reaction: in vacuum and in the enzyme D-xylose isomerase. The vacuum model is constructed based on the X-ray structure of the active site of D-xylose isomerase. It contains the atoms directly involved in the reaction and is studied using a semi-empirical molecular orbital method (PM3). The model in the enzyme includes the effects of the enzyme environment on the reaction using a combined quantum mechanical and molecular mechanical potential. For both models, the structures of the reactants, products, and intermediate complexes along the isomerization pathway are optimized. The effects of the position of the "catalytic Mg2+ ion" on the energies of the reactions are studied. The results indicate: 1) in vacuum, the isomerization reaction is favored when the catalytic metal cation is at site A, which is remote from the substrate; 2) in the enzyme, the catalytic metal cation, starting from site A, moves and stays at site B, which is close to the substrate; analysis of the charge redistribution of the active site during the catalytic process shows that the metal ion acts as a Lewis acid to polarize the substrate and catalyze the hydride shift; these results are consistent with previous experimental observations; and 3) Lys183 plays an important role in the isomerization reaction. The epsilon-NH3+ group of its side chain can provide a proton to the carboxide ion of the substrate to form a hydroxyl group after the hydride shift step. This role of Lys183 has not been suggested before. Based on our calculations, we believe that this is a reasonable mechanism and consistent with site-directed mutation experiments.

Electron paramagnetic resonance of D-xylose isomerase: evidence for metal ion movement induced by binding of cyclic substrates and inhibitors

The interactions of substrates and inhibitors with the Mn2+ ions in the binuclear active center of D-xylose isomerase (XylI) were investigated by EPR spectroscopy at X- and Q-band frequencies. The metal binding site 1 (A site) was specifically occupied with Mn2+ ions by blocking the high-affinity metal binding site 2 (B-site) either with Co2+ ions, resulting in a catalytically active enzyme, or with Cd2+ or Pb2+ ions yielding an inactive enzyme species. Incubation of both the Co2+/Mn2+- and the Cd2+/Mn2+-XylI with the acyclic inhibitor xylitol revealed EPR spectra with well-resolved hyperfine patterns, but with increased zero field splitting (zfs) parameter D compared to the spectra without inhibitor. D was estimated by spectral simulation of the central --1/2<-->1/2 fine structure transition. D values of 33 and 50 mT were obtained for the Co2+/Mn2+-XylI and the Cd2+/Mn2+-XylI samples, respectively. These results indicate direct interaction of the xylitol with the Mn2+ in the A-site. More drastic changes are observed with the substrates D-xylose and D-glucose and with the cyclic inhibitors 5-thio-alpha-D-glucose and 2-desoxy-D-glucose. For Cd2+/Mn2+-XylI, the EPR spectra with substrates and cyclic inhibitors are similar to each other but different from the spectra with the acylic inhibitor xylitol. They exhibit well-resolved line patterns with a relative large zero field splitting, which was estimated to be in the range of D = 65-85 mT in the various complexes. Binding of substrates or of cyclic inhibitors to the Co2+/ Mn2+-XylI yields EPR spectra without resolved hyperfine interactions, indicative of dipolar interaction between the two paramagnetic metal ions. This can be explained with a decrease in the metal-metal distance. Furthermore, the EPR data strongly suggest that the corresponding metal ion movement is induced by binding of the cyclic conformation of either substrates or cyclic inhibitors and not by binding of the extended form of the sugars.

Exploring the interaction between D-xylose isomerase and D-xylose by free energy calculation

The numerical quadrature thermodynamic integration method is used to investigate enzyme-substrate interaction of D-xylose isomerase. A screening function for the coulombic interaction is introduced into the simulation to correct the effect of finite cut-off radius for the non-bonded interaction. The binding free energy difference for D-xylose with D-xylose isomerase and its N184D mutant has been calculated, and the result 3.9 +/- 1.2 kJ/mol agrees well with experimental data of 4.38 kJ/mol. In addition, the structure and dynamics of enzyme-substrate complex were simulated for mutant and wild-type enzyme, respectively. Analysis of the structures and intramolecular interactions of the complexes were found to be valuable for understanding the reaction mechanism of the enzyme D-xylose isomerase.

Purification and cloning of a thermostable xylose (glucose) isomerase with an acidic pH optimum from Thermoanaerobacterium strain JW/SL-YS 489

An unusual xylose isomerase produced by Thermoanaerobacterium strain JW/SL-YS 489 was purified 28-fold to gel electrophoretic homogeneity, and the biochemical properties were determined. Its pH optimum distinguishes this enzyme from all other previously described xylose isomerases. The purified enzyme had maximal activity at pH 6.4 (60 degrees C) or pH 6.8 (80 degrees C) in a 30-min assay, an isoelectric point at 4.7, and an estimated native molecular mass of 200 kDa, with four identical subunits of 50 kDa. Like other xylose isomerases, this enzyme required Mn2+, Co2+, or Mg2+ for thermal stability (stable for 1 h at 82 degrees C in the absence of substrate) and isomerase activity, and it preferred xylose as a substrate. The gene encoding the xylose isomerase was cloned and expressed in Escherichia coli, and the complete nucleotide sequence was determined. Analysis of the sequence revealed an open reading frame of 1,317 bp that encoded a protein of 439 amino acid residues with a calculated molecular mass of 50 kDa. The biochemical properties of the cloned enzyme were the same as those of the native enzyme. Comparison of the deduced amino acid sequence with sequences of other xylose isomerases in the database showed that the enzyme had 98% homology with a xylose isomerase from a closely related bacterium, Thermoanaerobacterium saccharolyticum B6A-RI. In fact, only seven amino acid differences were detected between the two sequences, and the biochemical properties of the two enzymes, except for the pH optimum, are quite similar. Both enzymes had a temperature optimum at 80 degrees C, very similar isoelectric points (pH 4.7 for strain JW/SL-YS 489 and pH 4.8 for T. saccharolyticum B6A-RI), and slightly different thermostabilities (stable for 1 h at 80 and 85 degrees C, respectively). The obvious difference was the pH optimum (6.4 to 6.8 and 7.0 to 7.5, respectively). The fact that the pH optimum of the enzyme from strain JW/SL-YS 489 was the property that differed significantly from the T. saccharolyticum B6A-RI xylose isomerase suggested that one or more of the observed amino acid changes was responsible for this observed difference.

Molecular and industrial aspects of glucose isomerase

Glucose isomerase (GI) (D-xylose ketol-isomerase; EC. 5.3.1.5) catalyzes the reversible isomerization of D-glucose and D-xylose to D-fructose and D-xylulose, respectively. The enzyme has the largest market in the food industry because of its application in the production of high-fructose corn syrup (HFCS). HFCS, an equilibrium mixture of glucose and fructose, is 1.3 times sweeter than sucrose and serves as a sweetener for use by diabetics. Interconversion of xylose to xylulose by GI serves a nutritional requirement in saprophytic bacteria and has a potential application in the bioconversion of hemicellulose to ethanol. The enzyme is widely distributed in prokaryotes. Intensive research efforts are directed toward improving its suitability for industrial application. Development of microbial strains capable of utilizing xylan-containing raw materials for growth or screening for constitutive mutants of GI is expected to lead to discontinuation of the use of xylose as an inducer for the production of the enzyme. Elimination of Co2+ from the fermentation medium is desirable for avoiding health problems arising from human consumption of HFCS. Immobilization of GI provides an efficient means for its easy recovery and reuse and lowers the cost of its use. X-ray crystallographic and genetic engineering studies support a hydride shift mechanism for the action of GI. Cloning of GI in homologous as well as heterologous hosts has been carried out, with the prime aim of overproducing the enzyme and deciphering the genetic organization of individual genes (xylA, xylB, and xylR) in the xyl operon of different microorganisms. The organization of xylA and xylB seems to be highly conserved in all bacteria. The two genes are transcribed from the same strand in Escherichia coli and Bacillus and Lactobacillus species, whereas they are transcribed divergently on different strands in Streptomyces species. A comparison of the xylA sequences from several bacterial sources revealed the presence of two signature sequences, VXW(GP)GREG(YSTAE)E and (LIVM)EPKPX(EQ)P. The use of an inexpensive inducer in the fermentation medium devoid of Co2+ and redesigning of a tailor-made GI with increased thermostability, higher affinity for glucose, and lower pH optimum will contribute significantly to the development of an economically feasible commercial process for enzymatic isomerization of glucose to fructose. Manipulation of the GI gene by site-directed mutagenesis holds promise that a GI suitable for biotechnological applications will be produced in the foreseeable future.

Protein purification, and cloning and characterization of the cDNA and gene for xylose isomerase of barley

The first eukaryotic xylose isomerase protein was purified from barley Hordeum vulgare. The enzyme requires Mn2+ for its activity and is fairly thermostable, with the optimum temperature being 60 degrees C. It showed maximum activity over a broad pH range (7.0-9.0). The molecular mass of the monomer was about 50,000 Da based on the SDS/PAGE, and the calculated value from the cDNA-deduced polypeptide sequence was 53,620 Da. A relative mass estimation of 100,000 Da was obtained from the Superose 12 chromatography, suggesting that the barley enzyme is a dimer. The cloned corresponding cDNA sequence of 1710 nucleotides encoded a polypeptide of 480 amino acids. The genomic sequence of 4473 nucleotides, revealed that the isomerase gene contained 20 introns, all starting with GT and ending with AG. One large intron was located in the 5'untranslated region. The barley isomerase has an insertion of about 40 residues at its amino terminus when compared to the prokaryotic cluster (family) II isomerases; cluster (family) I and cluster (family) II isomerases vary from the former in an insertion of around 50 residues at their amino termini. Comparison of the barley protein with the prokaryotic isomerases shows that the conserved catalytic and metal binding regions are also well conserved in barley.

Probing the roles of active site residues in D-xylose isomerase

The roles of active site residues His54, Phe94, Lys183, and His220 in the Streptomyces rubiginosus D-xylose isomerase were probed by site-directed mutagenesis. The kinetic properties and crystal structures of the mutant enzymes were characterized. The pH dependence of diethylpyrocarbonate modification of His54 suggests that His54 does not catalyze ring-opening as a general acid. His54 appears to be involved in anomeric selection and stabilization of the acyclic transition state by hydrogen bonding. Phe94 stabilizes the acyclic-extended transition state directly by hydrophobic interactions and/or indirectly by interactions with Trp137 and Phe26. Lys183 and His220 mutants have little or no activity and the structures of these mutants with D-xylose reveal cyclic alpha-D-xylopyranose. Lys183 functions structurally by maintaining the position of Pro187 and Glu186 and catalytically by interacting with acyclic-extended sugars. His220 provides structure for the M2-metal binding site with properties which are necessary for extension and isomerization of the substrate. A second M2 metal binding site (M2') is observed at a relatively lower occupancy when substrate is added consistent with the hypothesis that the metal moves as the hydride is shifted on the extended substrate.

Wild-type and mutant D-xylose isomerase from Actinoplanes missouriensis: metal-ion dissociation constants, kinetic parameters of deuterated and non-deuterated substrates and solvent-isotope effects

The metal-ion dissociation constants (Mg2+, Mn2+) of wild-type and mutant D-xylose isomerases from Actinoplanes missouriensis have been determined by titrating the metal-ion-free enzymes with Mg2+ and Mn2+ respectively. Substitution of amino acids co-ordinated to metal-ion 1 (E181D, D245N) dramatically affects the dissociation constants, pH-activity profiles and apparent substrate binding. Mutagenesis of groups ligated to metal-ion 2 is less drastic except for that of Asp-255: a decrease in metal-ion affinity, a change in metal-ion preference and an improved apparent substrate binding (at pH values above the optimum), especially in the presence of Mn2+, are observed for the D255N enzyme. Similar effects, except for a slightly increased metal-ion affinity, are obtained by mutagenesis of the adjacent Glu-186 to Gln and the unconserved Ala-25 to Lys. Moreover, the striking acidic-pH shifts observed for the D255N and E186Q enzymes support the crucial role of the water molecule, Wa-690, Asp-255 and the adjacent Glu-186 in proton transfer from 2-OH to O-1 of the open and extended aldose substrate. Mutations of other important groups scarcely affect the metal-ion dissociation constants and pH-activity profiles, although pronounced effects on the kinetic parameters may be observed.

The first solvation shell of magnesium ion in a model protein environment with formate, water, and X-NH3, H2S, imidazole, formaldehyde, and chloride as ligands: an Ab initio study

The first coordination shell of an Mg(II) ion in a model protein environment is studied. Complexes containing a model carboxylate, an Mg(II) ion, various ligands (NH3, H2S, imidazole, and formaldehyde) and water of hydration about the divalent metal ion were geometry optimized. We find that for complexes with the same coordination number, the unidentate carboxylate-Mg(II) ion is greater than 10 kcal mol-1 more stable than the bidentate orientation. Imidazole was found to be the most stable ligand, followed in order by NH3, formaldehyde, H2O, and H2S.

Probing the catalytic roles of n2-site glutamate residues in Escherichia coli glutamine synthetase by mutagenesis

The contribution of metal ion ligand type and charge to catalysis and regulation at the lower affinity metal ion site (n2 site) of Escherichia coli glutamine synthetase (GS) was tested by mutagenesis and kinetic analysis. The 2 glutamate residues at the n2 site, E129 and E357, were changed to E129D, E129H, E357H, E357Q, and E357D, representing conservative and nonconservative alterations. Unadenylylated and fully adenylylated enzyme forms were studied. The Mn(2+)-KD values, UV-cis and fluorescence emission properties were similar for all mutants versus WTGS, except E129H. For kinetic determinations with both Mn2+ and Mg2+, nonconservative mutants (E357H, E129H, E357Q) showed lower biosynthetic activities than conservative mutants (E129D, E357D). Relative to WTGS, all the unadenylylated Mn(2+)-activated enzymes showed reduced kcat/Km values for ATP (> 7-fold) and for glutamate (> 10-fold). Of the unadenylylated Mg(2+)-activated enzymes, only E129D showed kinetic parameters competitive with WTGS, and adenylylated E129D was a 20-fold better catalyst than WTGS. We propose the n2-site metal ion activates ADP for departure in the phosphorylation of glutamate by ATP to generate gamma-glutamyl phosphate. Alteration of the charge density at this metal ion alters the transition-state energy for phosphoryl group transfer and may affect ATP binding and/or ADP release. Thus, the steady-state kinetic data suggest that modifying the charge density increases the transition-state energies for chemical steps. Importantly, the data demonstrate that each ligand position has a specialized spatial environment and the charge of the ligand modulates the catalytic steps occurring at the metal ion. The data are discussed in the context of the known X-ray structures of GS.

X-ray crystallographic structures of D-xylose isomerase-substrate complexes position the substrate and provide evidence for metal movement during catalysis

The X-ray crystallographic structures of the metal-activated enzyme xylose isomerase from Streptomyces olivochromogenes with the substrates D-glucose, 3-O-methyl-D-glucose and in the absence of substrate were determined to 1.96-, 2.19-, and 1.81-A resolution and refined to R-factors of 16.6%, 15.9%, and 16.1%, respectively. Xylose isomerase catalyzes the interconversion between glucose and fructose (xylose and xylulose under physiological conditions) by utilizing two metal cofactors to promote a hydride shift; the metals are bridged by a glutamate residue. This puts xylose isomerase in the small but rapidly growing family of enzymes with a bridged bimetallic active site, in which both metals are involved in the chemical transformation. The substrate 3-O-methylglucose was chosen in order to position the glucose molecule in the observed electron density unambiguously. Of the two essential magnesium ions per active site, Mg-2 was observed to occupy two alternate positions, separated by 1.8 A, in the substrate-soaked structures. The deduced movement was not observed in the structure without substrate present and is attributed to a step following substrate binding but prior to isomerization. The substrates glucose and 3-O-methylglucose are observed in their linear extended forms and make identical interactions with the enzyme by forming ligands to Mg-1 through O2 and O4 and by forming hydrogen bonds with His53 through O5 and Lys182 through O1. Mg-2 has a water ligand that is interpreted in the crystal structure in the absence of substrate as a hydroxide ion and in the presence of substrate as a water molecule. This hydroxide ion may act as a base to deprotonate the glucose O2 and subsequently protonate the product fructose O1 concomitant with hydride transfer. Calculations of the solvent-accessible surface of possible dimers, with and without the alpha-helical C-terminal domain, suggest that the tetramer is the active form of this xylose isomerase.

The xylose isomerase-encoding gene (xylA) of Clostridium thermosaccharolyticum: cloning, sequencing and phylogeny of XylA enzymes

The xylose isomerase (XylA)-encoding gene (xylA) of the thermophilic anaerobic bacterium, Clostridium thermosaccharolyticum NCIB 9385, was cloned as a 4.0-kb DNA fragment by complementation of the Escherichia coli xylA mutant strain, DS941. The open reading frame of 1317 bp encoded a protein of 439 amino acids (aa), with a calculated M(r) of 50,236. The gene was preceeded by a typical clostridial Shine-Dalgarno sequence, and was expressed constitutively in the cloning host. Downstream, the clone appeared to carry a xylB gene (encoding xylulokinase) in the same orientation as xylA. Comparison of the deduced aa sequence of the C. thermosaccharolyticum XylA with 18 other XylA showed that this family of proteins was separated into two clusters, one comprising proteins from organisms with G + C-rich DNA, and the other proteins from organisms with a lower G + C composition. Within the second cluster, the XylA of C. thermosaccharolyticum was most closely related to the enzymes from C. thermosulfurogenes (Thermoanaerobacterium thermosulfurigenes) and C. thermohydrosulfuricum (93 and 84% identity, respectively). Analysis of the aligned sequences indicated two signatures (VXW[GP]GREG[YSTA]E and [LIVM]EPKPX]EQ]P) which may be useful in isolation of novel XylA.

Arthrobacter D-xylose isomerase: chemical modification of carboxy groups and protein engineering of pH optimum

To try to lower the pH optimum, the carboxy groups of Arthrobacter D-xylose isomerase were coupled to glycinamide using a water-soluble carbodi-imide. In conditions that substituted all of the 59 carboxy groups in the denatured monomer, a maximum of 30 groups/monomer reacted in the native enzyme, whether in presence or absence of ligands, and the enzyme remained fully active and tetrameric throughout the coupling reaction. Purification by f.p.l.c. ion-exchange chromatography gave broad symmetrical peaks with increased pI, suggesting that the modified enzymes are essentially homogeneous. However, they are less stable than native enzyme in 8 M urea or on heating ('melting points' of 59 degrees versus 73 degrees C for the apoenzymes and 67 degrees versus 81.5 degrees C for the Mg(2+)-enzymes). Kinetic studies of the D-fructose isomerase activity at 30 degrees C showed that the glycinamidylated enzyme had unaltered activation constant for Mg2+, and Km was also similar to that of the native enzyme at pH 7.3, but increased rapidly at higher pH rather than remaining constant. Vmax. was constant from pH 6.2 to 8.0, suggesting a reduced pKa for His-219, which controls Vmax. in the native enzyme (normally 6.0). Three mutants were constructed by protein engineering with a view to reducing the pH optimum of enzyme activity. Two of these, Glu140-->Lys and Asp189-->Lys, could be detected in crude extracts of Escherichia coli by SDS/PAGE, but could not be purified, whereas mutant Trp136-->Glu was produced as a tetramer in amounts similar to the wild-type enzyme. However, it did not show any enzyme activity and was less stable in 0-9 M urea gradient PAGE.

Dominant role of local dipoles in stabilizing uncompensated charges on a sulfate sequestered in a periplasmic active transport protein

Electrostatic interactions are among the key factors determining the structure and function of proteins. Here we report experimental results that illuminate the functional importance of local dipoles to these interactions. The refined 1.7-A X-ray structure of the liganded form of the sulfate-binding protein, a primary sulfate active transport receptor of Salmonella typhimurium, shows that the sulfate dianion is completely buried and bound by hydrogen bonds (mostly main-chain peptide NH groups) and van der Waals forces. The sulfate is also closely linked, via one of these peptide units, to a His residue. It is also adjacent to the N-termini of three alpha-helices, of which the two shortest have their C-termini "capped" by Arg residues. Site-directed mutagenesis of the recombinant Escherichia coli sulfate receptor had no effect on sulfate-binding activity when an Asn residue was substituted for the positively charged His and the two Arg (changed singly and together) residues. These results, combined with other observations, further solidify the idea that stabilization of uncompensated charges in a protein is a highly localized process that involves a collection of local dipoles, including those of peptide units confined to the first turns of helices. The contribution of helix macrodipoles appears insignificant.

X- and Q-band EPR studies on the two Mn(2+)-substituted metal-binding sites of D-xylose isomerase

The two metal-binding sites (A and B)/subunit of the homotetrameric D-xylose isomerase (Xyl isomerase) from Streptomyces rubiginosus have been studied with Mn(2+)-EPR spectroscopy at X-band and Q-band frequencies and with electronic spectroscopy. Displacement studies in the visible absorbance range showed that Mn2+ have a higher affinity for the B site. With the low-affinity A site unoccupied, the coordination sphere of Mn2+ in the B site is quite distorted giving rise to a highly anisotropic X-band EPR spectrum. Simulation of the Q-band spectrum reveals a zero field splitting (zfs) D of about 45-48 mT and a rhombicity parameter E/D between 0.2 and 0.3. Occupation of both binding sites with Mn2+ induces a significant shift towards a higher symmetry in the coordination sphere of the B site resulting in similar zfs parameters for both binding sites. The change in A-site environment caused by B-site occupation was analysed in mixed Xyl isomerase derivatives, in which the B site is loaded with Co2+, Cd2+ or Pb2+ and the A site with Mn2+. In the Co2+/Mn2+ Xyl isomerase the Mn2+ has a relatively symmetric ligand environment with small zfs parameters (D = 12 mT, E/D < 0.15). Substituting Co2+ with Cd2+ or Pb2+ in the B site leads to a drastic increase in the zfs parameters of Mn2+ in the A site. The distortions are directly linked to the ionic radii of the ions bound to the B site and may be mediated by the carboxylate group of Glu216 that bridges the metal-binding sites. The EPR spectra also reflect the catalytic activity of the mixed metal samples. With the larger Cd2+ or Pb2+ in the B site, which are strongly influencing the stereochemistry of the A site, the catalytic activity is lost, whereas Co2+ and Mn2+ render the enzyme in an active state, so that the mutual influence on catalysis depends on the complex geometry of both metal-binding sites.

Arthrobacter D-xylose isomerase: protein-engineered subunit interfaces

Mutants of Arthrobacter D-xylose isomerase were constructed in which one or two disulphide bridges or additional salt bridges were introduced at the A-A* subunit interfaces. These showed no change in enzyme activity or stability compared with the wild-type enzyme. However, a Tyr253 mutant in which a disulphide bridge was introduced at the A-B* subunit interface showed reduced thermostability that was identical in both oxidized and reduced forms, and also reduced stability in urea. X-ray-crystallographic analysis of the Mn(2+)-xylitol form of oxidized Y253C (the Tyr253-->Cys mutant) showed a changed conformation of Glu185 and also alternative conformations for Asp254, which is a ligand to the Site-[2] metal ion. With fructose, Mg(2+)-Y253C has a similar Km to that of the wild-type, and its Vmax. is also similar below pH 6.4, but declined thereafter. In the presence of Co2+, Y253C has lower activity than wild-type at all pH values, but its activity also declines at alkaline pH. These results suggest that electrostatic repulsion from the new position of Glu185 causes Asp254 to move when His219 is unprotonated, thereby preventing M2+ binding at Site [2]. These results also suggest that subunit dissociation does not lie on the pathway of thermal inactivation of D-xylose isomerase, but that movements of active-site groups are a trigger for conformational changes that initiate the unfolding process.

Arthrobacter D-xylose isomerase: partial proteolysis with thermolysin

The pattern and kinetics of partial proteolysis of Arthrobacter D-xylose isomerase tetramer was studied in order to determine the flexibility of surface loops that may control its stability. It was completely resistant to trypsin, chymotrypsin and elastase at 37 degrees C, but thermolysin cleaved specifically and quantitatively at Thr-347-Leu-348 between helices 10 and 11 to remove 47 residues from the C-terminus of each 43.3 kDa subunit. At high temperatures, helices 9 and 10 were removed from each 38 kDa subunit to give a 36 kDa tetramer. The kinetics of nicking by thermolysin indicated that the Thr-347-Leu-348 loop is locked at low temperatures, but 'melts' at 25 degrees C and is fully flexible above 34 degrees C. The flexibility appears to be associated with binding of Ca2+ ions at the active site, since Co2+, Mg2+ and xylitol protect in proportion to their ability to displace Ca2+. The missing C-terminal helices make many intersubunit contacts that appear in the structure to stabilize the tetramer, but the properties of the purified nicked proteins are almost indistinguishable from the native enzyme. Both the 38 kDa tetramer and the 36 kDa tetramer are identically active and dissociate similarly in urea or SDS to fully active dimers, but the nicked dimers are slightly less stable to urea at 62 degrees C. In the Mg2+ form the thermostability of the 38 kDa tetramer is identical with that of the native enzyme, but the 36 kDa tetramer has a slightly lower 'melting point' (70 degrees C versus 80 degrees C), which may be due to unravelling from the end of helix 8. Since elimination of all the C-terminal helices and many intersubunit contacts has so little effect, one can conclude that the 'weak point' that controls the protein's thermostability lies within the N-terminal beta-barrel domain.