Light-inducible flux control of triosephosphate isomerase on glycolysis in Escherichia coli.

An engineering tool for controlling flux distribution on metabolic pathways to an appropriate state is highly desirable in bioproduction. An optogenetic switch, which regulates gene expression by light illumination is an attractive on/off switchable system, and is a promising way for flux control with an external stimulus. We demonstrated a light-inducible flux control between glycolysis and the methylglyoxal (MGO) pathway in Escherichia coli using a CcaS/CcaR system. CcaR is phosphorylated by green light and is dephosphorylated by red light. Phosphorylated CcaR induces gene expression under the cpcG2 promoter. The tpiA gene was expressed under the cpcG2 promoter in a genomic tpiA deletion strain. The strain was then cultured with glucose minimum medium under green or red light. We found that tpiA messenger RNA level under green light was four times higher than that under red light. The repression of tpiA expression led to a decrease in glycolytic flux, resulting in slower growth under red light (0.25 hr -1 ) when compared to green light (0.37 hr -1 ). The maximum extracellular MGO concentration under red light (0.2 mM) was higher than that under green light (0.05 mM). These phenotypes confirm that the MGO pathway flux was enhanced under red light.

Biochemical characterisation of triose phosphate isomerase from the liver fluke Fasciola hepatica.

Triose phosphate isomerase (TPI) catalyses the interconversion of dihydroxyacetone phosphate and glyceraldehyde 3-phosphate, a reaction in the glycolytic pathway. TPI from the common liver fluke, Fasciola hepatica, has been cloned, sequenced and recombinantly expressed in Escherichia coli. The protein has a monomeric molecular mass of approximately 28 kDa. Crosslinking and gel filtration experiments demonstrated that the enzyme exists predominantly as a dimer in solution. F. hepatica TPI is predicted to have a ß-barrel structure and key active site residues (Lys-14, His-95 and Glu-165) are conserved. The enzyme shows remarkable stability to both proteolytic degradation and thermal denaturation. The melting temperature, estimated by thermal scanning fluorimetry, was 67 °C and this temperature was increased in the presence of either dihydroxyacetone phosphate or glyceraldehyde 3-phosphate. Kinetic studies showed that F. hepatica TPI demonstrates Michaelis-Menten kinetics in both directions, with Km values for dihydroxyacetone phosphate and glyceraldehyde 3-phosphate of 2.3 mM and 0.66 mM respectively. Turnover numbers were estimated at 25,000 s(-1) for the conversion of dihydroxyacetone phosphate and 1900 s(-1) for the conversion of glyceraldehyde 3-phosphate. Phosphoenolpyruvate acts as a weak inhibitor of the enzyme. F. hepatica TPI has many features in common with mammalian TPI enzymes (e.g. ß-barrel structure, homodimeric nature, high stability and rapid kinetic turnover). Nevertheless, recent successful identification of specific inhibitors of TPI from other parasites, suggests that small differences in structure and biochemical properties could be exploited in the development of novel, species-specific inhibitors.

Triosephosphate isomerase is a common crystallization contaminant of soluble His-tagged proteins produced in Escherichia coli.

Attempts to crystallize several mammalian proteins overexpressed in Escherichia coli revealed a common contaminant, triosephosphate isomerase, a protein involved in glucose metabolism. Even with triosephosphate isomerase present in very small amounts, similarly shaped crystals appeared in the crystallization drops in a number of polyethylene glycol-containing conditions. All of the target proteins were His-tagged and their purification involved immobilized metal-affinity chromatography (IMAC), a step that was likely to lead to triosephosphate isomerase contamination. Analysis of the triosephosphate isomerase crystals led to the structure of E. coli triosephosphate isomerase at 1.85 Šresolution, which is a significant improvement over the previous structure.

Unravelling the bull fertility proteome.

In the last few decades a negative association between the level of milk production and fertility has been observed. Currently, the most utilized method of measuring male fertility employed by the livestock industry is related to the Non-Return Rate (NRR). Through differential proteome analysis, this study evaluated changes in the expression of the protein profile of spermatozoa collected from 16 bulls with different levels of field fertility expressed as an estimated relative conception rate (ERCR). The main aim is to identify putative protein markers to be used as putative indices of fertility. Two dimensional electrophoresis coupled with mass spectrometry analysis was used for protein separation and identification. To improve differential proteome analysis among experimental groups, a part of shotgun MS analysis was also performed. Three protein spots showed a differential expression pattern among all ERCR classes. Alpha enolase was significantly down-regulated in the ERCR- group, while two other proteins, isocitrate dehydrogenase and triosephosphate isomerase, were up-regulated in ERCR- in comparison to ERCR+. Alpha-enolase and isocitrate dehydrogenase subunit alpha (IDH-alpha) have been described in the literature for having a potential role in bull fertility. The possibility of determining protein biomarkers for fertility is more useful and less expensive than ERCR for acquiring rapid estimation of fertility because it does not require the use of field insemination trials. Shotgun MS analysis conducted on the same samples revealed 7 proteins down-regulated in the ERCR- group and 1 protein up-regulated. Among these proteins, calmodulin, ATP synthase mitochondrial subunits alpha and delta, malate dehydrogenase and sperm equatorial segment protein 1 were shown to be linked with sperm fertility.

Characterization of stress and methylglyoxal inducible triose phosphate isomerase (OscTPI) from rice.

As compared with plant system, triose phosphate isomerase (TPI), a crucial enzyme of glycolysis, has been well studied in animals. In order to characterize TPI in plants, a full-length cDNA encoding OscTPI was cloned from rice and expressed in E. coli. The recombinant OscTPI was purified to homogeneity and it showed Km value of 0.1281 ± 0.025 µM, and the Vmax value of 138.7 ± 16 µmol min (-1) mg (-1) which is comparable to the kinetic values studied in other plants. The OscTPI was found to be exclusively present in the cytoplasm when checked with the various methods. Functional assay showed that OscTPI could complement a TPI mutation in yeast. Real time PCR analysis revealed that OscTPI transcript level was regulated in response to various abiotic stresses. Interestingly, it was highly induced under different concentration of methylglyoxal (MG) stress in a concentration dependent manner. There was also a corresponding increase in the protein and the enzyme activity of OscTPI both in shoot and root tissues under MG stress. Our result shows that increases in MG leads to the increase in TPI which results in decrease of DHAP and consequently decrease in the level of toxic MG.

G-CSF downregulates natural killer cell-mediated cytotoxicity in donors for hematopoietic SCT.

In G-CSF-mobilized hematopoietic SCT (HSCT), natural killer (NK) cells have a critical role in GVHD and GVL effects. However, regulation of NK cell response to G-CSF remains unclear. This study assayed G-CSF effects in both HSCT donors and NK-92MI cells. The donors who received G-CSF had significantly decreased NK cell cytotoxicity. Levels of phosphatidylinositol 3-kinase (PI3K) and phosphorylated (p)-Akt, but not mammalian target of rapamycin (mTOR), were downregulated in NK cells from G-CSF-injected donors. G-CSF also decreased cytotoxicity without affecting viability and NF-κB of NK-92MI cells. PI3K and p-ERK expression were also decreased in G-CSF-treated NK-92MI cells, and their inhibitors, wortmannin and PD98059, respectively, both enhanced the downregulation of cytotoxicity. These effects were accompanied by decreased expression of a cytotoxicity-related gene, triosephosphate isomerase (TPI). Wortmannin, but not PD98059, enhanced the downregulation of TPI in G-CSF-treated NK-92MI cells, indicating a correlation between PI3K and TPI. We conclude that G-CSF-impaired NK cell cytotoxicity may accompany PI3K/Akt signaling. The effect is transient and NK cells may recover after G-CSF clearance, suggesting that G-CSF-mobilized HSCT may benefit both acute GVHD prevention and late-phase GVL promotion in HSCT recipients.

Triose phosphate isomerase, a novel enzyme-crystallin, and tau-crystallin in crocodile cornea. High accumulation of both proteins during late embryonic development.

Several enzymes are known to accumulate in the cornea in unusually high concentrations. Based on the analogy with lens crystallins, these enzymes are called corneal crystallins, which are diverse and species-specific. Examining crystallins in lens and cornea in multiple species provides great insight into their evolution. We report data on major proteins present in the crocodile cornea, an evolutionarily distant taxon. We demonstrate that tau-crystallin/alpha-enolase and triose phosphate isomerase (TIM) are among the major proteins expressed in the crocodile cornea as resolved by 2D gel electrophoresis and identified by MALDI-TOF. These proteins might be classified as putative corneal crystallins. tau-Crystallin, known to be present in turtle and crocodile lens, has earlier been identified in chicken and bovine cornea, whereas TIM has not been identified in the cornea of any species. Immunostaining showed that tau-crystallin and TIM are concentrated largely in the corneal epithelium. Using western blot, immunofluorescence and enzymatic activity, we demonstrate that high accumulation of tau-crystallin and TIM starts in the late embryonic development (after the 24th stage of embryonic development) with maximum expression in a two-week posthatched animal. The crocodile corneal extract exhibits significant alpha-enolase and TIM activities, which increases in the corneal extract with development. Our results establishing the presence of tau-crystallin in crocodile, in conjunction with similar reports for other species, suggest that it is a widely prevalent corneal crystallin. Identification of TIM in the crocodile cornea reported here adds to the growing list of corneal crystallins.

Hypoxic up-regulation of triosephosphate isomerase expression in mouse brain capillary endothelial cells.

A protein with a molecular mass of 27kDa was induced by hypoxia in a mouse brain capillary endothelial cell line and identified as triosephosphate isomerase (TPI) by amino-terminal sequencing. Hypoxia caused an elevation of the TPI protein level, concomitant with an increase of the TPI mRNA level. However, hypoxia resulted in an insufficient elevation of TPI activity level, compared to an increase of TPI protein level. When cells expressing the recombinant TPI protein with histidine tag were exposed to hypoxia and the TPI protein was affinity-purified, the catalytic activity (specific activity) of the TPI protein purified from hypoxic cells was substantially lower than that obtained from normoxic cells. In addition, three TPI isoforms with an electrophoretic multiplicity were found; two of the three isoforms were substantially increased in response to the hypoxia, but the level of the most acidic isoform was barely changed. The induction of TPI gene expression by hypoxia was suppressed by (1) a chelator of intracellular Ca(2+), (2) a blocker of non-selective cation channels, (3) a blocker of Na(+)/Ca(2+) exchangers, (4) an inhibitor of Ca(2+)/calmodulin-dependent protein kinases, and (5) an inhibitor of c-jun/AP-1 activation.

Conserved cysteine 126 in triosephosphate isomerase is required not for enzymatic activity but for proper folding and stability.

In triosephosphate isomerase, Cys126 is a conserved residue located close to the catalytic glutamate, Glu165. Although it has been mentioned that Cys126 and other nearby residues are required to maintain the active site geometry optimal for catalysis, no evidence supporting this idea has been reported to date. In this work, we studied the catalytic and stability properties of mutants C126A and C126S of Saccharomyces cerevisiae TIM (wtTIM). None of these amino acid replacements induced significant changes in the folding of wtTIM, as indicated by spectroscopic studies. C126S and C126A have K(M) and k(cat) values that are concomitantly reduced by only 4-fold and 1.5-fold, respectively, compared to those of wtTIM; in either case, however, the catalytic efficiency (k(cat)/K(M)) of the enzyme is barely affected. The affinity of mutated TIMs for the competitive inhibitor 2-phosphoglycolate augmented also slightly. In contrast, greater susceptibility to thermal denaturation resulted from mutation of Cys126, especially when it was changed to Ser. By using values of the rate constants for unfolding and refolding, we estimated that, at 25 degrees C, C126A and C126S are less stable than wtTIM by about 5.0 and 9.0 kcal mol(-)(1), respectively. Moreover, either of these mutations slows down the folding rate by a factor of 10 and decreases the recovery of the active enzyme after thermal unfolding. Thus, Cys126 is required for proper stability and efficient folding of TIM rather than for enzymatic catalysis.

Hypoxia up-regulates triosephosphate isomerase expression via a HIF-dependent pathway.

The glycolytic enzyme triosephosphate isomerase (TPI) catalyses the reversible conversion of dihydroxyacetone phosphate into glyceraldehyde-3-phosphate. We report here that the expression of TPI at both the mRNA and protein levels is increased by hypoxia in vivo and in vitro. The temporal pattern of hypoxic TPI induction is very similar to that of genes triggered by the hypoxia-inducible transcription factor (HIF) and is mimicked characteristically by cobalt and by deferoxamine, but is absent in cells with a defective aryl hydrocarbon receptor nuclear translocator (ARNT, here HIF-1beta) and in cells lacking HIF-1alpha protein. We conclude from these findings that the expression of TPI is regulated via the HIF pathway and thus belongs to the family of classic oxygen-regulated genes. The physiological meaning of an increased expression of TPI in hypoxygenated tissues is probably to increase the flow of triosephosphates through the glycolytic cascade thus leading to an increase of anaerobic energy generation.

Identification of metabolic enzymes in renal cell carcinoma utilizing PROTEOMEX analyses.

PROTEOMEX, an approach which combines conventional proteome analysis with serological screening, is a powerful tool to separate proteins and identify immunogenic components in malignant diseases. By applying this approach, we characterized nine metabolic enzymes which were differentially expressed in renal cell carcinoma (RCC) cell lines and compared their expression profiles to that of normal kidney epithelium cells. Four of these proteins, superoxide dismutase (SODC), triosephosphatase isomerase (TPIS), thioredoxin (THIO) and ubiquitin carboxyl-terminal hydrolase (UBL1) were further analysed for both their constitutive and interferon (IFN)-gamma inducible protein expression pattern in cell lines or tissue specimens derived from RCC or normal kidney epithelium using Western blot analysis and immunohistochemistry, respectively. With the exception of the RCC cell line MZ1940RC, which completely lacks the expression of UBL1, a heterogeneous and variable expression pattern of the different metabolic enzymes was detected in RCC and normal renal epithelium. The highest differences in the expression levels were found for THIO in the RCC cell lines, which was 2-fold upregulated when compared to autologous normal kidney epithelium. Moreover, IFN-gamma treatment did not influence the constitutive expression of these metabolic enzymes. Thus, PROTEOMEX represents a valuable approach for the identification of metabolic enzymes which might be used as markers for the diagnosis of RCC.

"Early" protein synthesis of Lactobacillus delbrueckii ssp. bulgaricus in milk revealed by [35S] methionine labeling and two-dimensional gel electrophoresis.

The proteomes of exponentially growing and stationary cells of Lactobacillus delbrueckii ssp. bulgaricus grown in rich medium (MRS) were separated by two-dimensional polyacrylamide gel electrophoresis (2-DE) and quantified after Coomassie staining. Stationary cells grown in MRS were inoculated in reconstituted skim milk, and "early" protein synthesis during the first 30 min of fermentation in milk was monitored by [35S]methionine labeling and 2-DE. In contrast to exponentially growing or stationary cells, the predominant "early" proteins were small (< 15 kDa) and of low pI (< 5.3). Quantification of the proteome of the "early" lag phase based on 47 "spots" revealed that only three "early" proteins accounted for more than 80% of the total label. They were identified as pI 4.7 and 4.9 isoforms of the heat-stable phosphoryl carrier protein (HPr) with 45.2 and 9.4% of total label, respectively, and an unknown protein called EPr1 ("early" protein 1) with 26.6% of total label. Although an N-terminal sequence of 19 amino acids was obtained, no homologs to EPr1 could be found. De novo synthesis of the 10 and 60 kDa heat shock proteins (GroES and GroEL) was considerably lower (0.04 and 0.9% of total label, respectively), indicating only low levels of stress. Synthesis of triosephosphate isomerase (Tpi) as marker for glycolytic enzymes reached only 0.08% of total label. Our results demonstrate that inoculation in milk, resulting in a change from glucose to lactose as carbon source, imposes only little need for synthesis of stress or glycolytic enzymes, as sufficient proteins are present in the stationary, MRS-grown cells. The high level of expression of the pI 4.7 isoform of HPr suggests a regulatory function of the presumed Ser-46 phosphorylated form of HPr.

Understanding protein lids: kinetic analysis of active hinge mutants in triosephosphate isomerase.

In previous work we tested what three amino acid sequences could serve as a protein hinge in triosephosphate isomerase [Sun, J., and Sampson, N. S. (1998) Protein Sci. 7, 1495-1505]. We generated a genetic library encoding all 8000 possible 3 amino acid combinations at the C-terminal hinge and selected for those combinations of amino acids that formed active mutants. These mutants were classified into six phylogenetic families. Two families resembled wild-type hinges, and four families represented new types of hinges. In this work, the kinetic characteristics and thermal stabilities of mutants representing each of these families were determined in order to understand what properties make an efficient protein hinge, and why all of the families are not observed in nature. From a steady-state kinetic analysis of our mutants, it is clear that the partitioning between protonation of intermediate to form product and intermediate release from the enzyme surface to form methylglyoxal (a decomposition product) is not affected. The two most impaired mutants undergo a change in rate-limiting step from enediol formation to dihydroxyacetone phosphate binding. Thus, it appears that k(cat)/K(m)'s are reduced relative to wild type as a result of slower Michaelis complex formation and dissociation, rather than increased loop opening speed.

At least one intron is required for the nonsense-mediated decay of triosephosphate isomerase mRNA: a possible link between nuclear splicing and cytoplasmic translation.

Mammalian cells have established mechanisms to reduce the abundance of mRNAs that harbor a nonsense codon and prematurely terminate translation. In the case of the human triosephosphate isomerase (TPI gene), nonsense codons located less than 50 to 55 bp upstream of intron 6, the 3'-most intron, fail to mediate mRNA decay. With the aim of understanding the feature(s) of TPI intron 6 that confer function in positioning the boundary between nonsense codons that do and do not mediate decay, the effects of deleting or duplicating introns have been assessed. The results demonstrate that TPI intron 6 functions to position the boundary because it is the 3'-most intron. Since decay takes place after pre-mRNA splicing, it is conceivable that removal of the 3'-most intron from pre-mRNA "marks" the 3'-most exon-exon junction of product mRNA so that only nonsense codons located more than 50 to 55 nucleotides upstream of the "mark" mediate mRNA decay. Decay may be elicited by the failure of translating ribosomes to translate sufficiently close to the mark or, more likely, the scanning or looping out of some component(s) of the translation termination complex to the mark. In support of scanning, a nonsense codon does not elicit decay if some of the introns that normally reside downstream of the nonsense codon are deleted so the nonsense codon is located (i) too far away from a downstream intron, suggesting that all exon-exon junctions may be marked, and (ii) too far away from a downstream failsafe sequence that appears to function on behalf of intron 6, i.e., when intron 6 fails to leave a mark. Notably, the proposed scanning complex may have a greater unwinding capability than the complex that scans for a translation initiation codon since a hairpin structure strong enough to block translation initiation when inserted into the 5' untranslated region does not block nonsense-mediated decay when inserted into exon 6 between a nonsense codon residing in exon 6 and intron 6.

Sulfhydryl reagent susceptibility in proteins with high sequence similarity--triosephosphate isomerase from Trypanosoma brucei, Trypanosoma cruzi and Leishmania mexicana.

The amino acid sequence of triosephosphate isomerase from Trypanosoma brucei, Trypanosoma cruzi, and Leishmania mexicana have an identity of 68%. Using the numbering system for the T. brucei enzyme, in their aligned sequences, the T. cruzi and leishmanial enzymes have cysteine residues at positions 14, 40, 117 and 126. T. brucei triosephosphate isomerase has cysteine residues at positions 14, 40 and 126, and a valine residue at position 117. Dithionitrobenzoic acid and methylmethane thiosulfonate inhibited the three enzymes, but T. cruzi triosephosphate isomerase was more than 100-fold more sensitive. The sensitivity of wild type triosephosphate isomerase from T. cruzi and T. brucei to the reagents was equal to that of the Cys117Val and Val117Cys mutant enzymes, respectively. Triosephosphate isomerases that have cysteine residues at positions 40 and 126, but lack a cysteine residue at position 14 are insensitive to methylmethane thiosulfonate. Thus, sulfhydryl reagents act on Cys14. At stoichiometric concentrations, the reagents inhibited the three enzymes as a consequence of structural alterations as measured by binding of 8-anilino-1-napthalenesulfonic acid to previously buried hydrophobic regions. However, the times for half-maximal alterations were 10 min, 15 hours and over 30 hours for T. cruzi, T. brucei and L. mexicana triosephosphate isomerase, respectively. The effect of pH on the action of the sulfhydryl reagents and molecular modeling showed no differences in the solvent accessibility of Cys14. As Cys14 forms part of the dimer interface, the data indicate that, in the three enzymes, barriers of different magnitude hinder the interaction between the sulfhydryl reagents and Cys14. The barrier is lower in T. cruzi triosephosphate isomerase which makes its dimer interface more susceptible for perturbation.

Intron position affects expression from the tpi promoter in rice.

A series of promoter-GUS fusion constructs containing a portion of the rice triosephosphate isomerase (tpi) promoter, the first tpi intron, and the gene encoding bacterial beta-glucuronidase (GUS) were made. These constructs were electroporated into rice protoplasts and transient expression was monitored. Inclusion of the first intron from the rice tpi gene enhanced expression of the GUS gene from the tpi promoter when it was placed 5' of the GUS gene. When the tpi intron was placed in the 3'-untranslated region no enhancement of GUS gene expression was observed, indicating the importance of position in intron-mediated enhancement of gene expression.

The use of antibodies for characterization and quantification of a recombinant protein.

In this study, we characterized proteinase A secreted by recombinant Saccharomyces cerevisiae bearing a multicopy plasmid containing the encoding gene (PEP4). Polyclonal and monoclonal antibodies were raised to study the product heterogeneity. Characterization of proteinase A was performed by immunoelectrophoresis and immunoblotting techniques. None of the monoclonal antibodies raised against proteinase A was found to react with the glycosyl side chains; thus cross-reaction with other glycosylated proteins (e.g. carboxypeptidase Y) was very low. This study allowed us to develop an ELISA method for the quantification of proteinase A in culture supernatants as well as the evaluation of monoclonal antibodies for their use in immunoaffinity chromatography.

Stabilization of human triosephosphate isomerase by improvement of the stability of individual alpha-helices in dimeric as well as monomeric forms of the protein.

Human triosephosphate isomerase (hTIM) is a dimeric enzyme of identical subunits, adopting the alpha/beta-barrel fold. In a previous work, a monomeric mutant of hTIM was engineered in which Met14 and Arg98, two interface residues, were changed to glutamine. Analysis of equilibrium denaturation of this monomeric mutant, named M14Q/R98Q, revealed that its conformational stability, 2.5kcal/mol, is low as compared to the stability of dimeric hTIM (19.3 kcal/mol). The fact that this value is also lower than the conformational stabilities usually found for monomeric proteins suggests that the hTIM monomers are thermodynamically unstable. In the present work, we attempted to stabilize the M14Q/R98Q mutant by introducing stabilizing mutations in alpha-helices of the protein. Five mutations were proposed, designed to increase alpha-helix propensity by introducing alanines at solvent-exposed sites (Q179A, K193A), to introduce favorable interactions with helix dipoles (Q179D, S105D), or to reduce the conformational entropy of unfolding by introducing proline residues at the "N-cap" position of alpha-helices (A215P). Three replacements (Q179D, K193A, and A215P) were found to increase the stability of the native dimeric hTIM and the monomeric M14Q/R98Q. These results suggest that the monomeric hTIM mutant can be stabilized to a considerable extent by following well-established rules for protein stabilization. A comparison of the stabilizing effect performed by the mutations on the dimeric hTIM and the monomeric M14Q/R98Q allowed us to reinforce a model of equilibrium denaturation proposed for both proteins.

Cytoplasmic mRNA for human triosephosphate isomerase is immune to nonsense-mediated decay despite forming polysomes.

Nonsense codons between position 14 within the first exon and position 193 within the penultimate exon of the human gene for triosephosphate isomerase reduce mRNA abundance to 25% of normal. The reduction in abundance is due to the decay of newly synthesized mRNA that copurifies with nuclei. TPI mRNA that copurifies with cytoplasm is immune to decay. We show here that immunity is not due to the failure of nonsense-containing mRNA to form polysomes. This finding indicates that cytoplasmic mRNA, in contrast to nucleus-associated mRNA, may have lost one or more factors that are required for nonsense-mediated decay or gained one or more factors that confer immunity to nonsense-mediated decay.

The structural basis for pseudoreversion of the E165D lesion by the secondary S96P mutation in triosephosphate isomerase depends on the positions of active site water molecules.

The structural basis for the improvement in catalytic efficiency of the mutant E165D chicken triosephosphate isomerase by the secondary mutation, S96P, has been analyzed using a combination of X-ray crystallography and Fourier transform infrared spectroscopy. All X-ray structures were of the complex of phosphoglycolohydroxamate (PGH), an intermediate analog, with the isomerase, and each was solved to a resolution of 1.9 A. Comparison of the structure of the double mutant, E165D.S96P, with that of the single mutant, E165D, as well as with the wild-type isomerase shows only insignificant differences in the positions of the side chains in all of the mutants when compared with the wild-type isomerase, except that in both the E165D and E165D.S96P mutants, the aspartate side chain was approximately 0.7 A further away from the substrate analog than the glutamate side chain. Significant differences were observed in the crystal structure of the E165D.S96P double mutant in the positions of ordered water molecules bound at the active site. The loss of two water molecules located near the side chain at position 165 was observed in isomerases containing the S96P mutation. The resulting increase in hydrophobicity of the pocket probably causes an increase in the pKa of the catalytic base, D165, thereby improving its basicity. A new ordered water molecule was observed underneath the bound PGH in the E165D.S96P structure, which likely decreases the pKa's of the substrate protons, thereby increasing their acidity. An enzyme derived carbonyl stretch at 1746 cm-1 that is only observed in the IR spectrum of the E165D.S96P double mutant isomerase with bound substrates has been assigned to a stable ground state protonated D165-enediol(ate) intermediate complex. Thus, the gain in activity resulting from the S96P second site change probably results from a combination of improving the basicity of the enzyme, improving the acidity of the substrate protons, and stabilization of a reaction intermediate. All three of these effects seem to be caused by changes in bound water molecules.