INTRODUCTION

Top Abstract Introduction Procedures Results & Discussion References

Quinoprotein GDH¹ bound to the inner membrane in Escherichia coli functions in a direct oxidation of D-glucose to D-gluconate and concomitantly transferring the electrons to cytochrome oxidase through ubiquinone in the respiratory chain (1, 2). Previous reports have indicated that GDH possesses the binding sites for PQQ (3, 4), metal ions such as Mg²⁺ or Ca²⁺ (5, 6), and ubiquinone (1, 8, 9) as well as substrate glucose. The enzyme from E. coli occurs as the apoenzyme (5) because the organism is unable to produce PQQ (6), but is readily reconstituted by incubation with PQQ and the metal ions (7). To elucidate the functional structure, topological analysis has been performed (10) which revealed that GDH possesses five membrane-spanning segments at the N-terminal one-sixth portion, and that the remaining C-terminal five-sixth portion occurs at the periplasmic side of the membrane. The large C-terminal portion is assumed to have the catalytic domain including the PQQ-binding site. The binding site of ubiquinone in GDH has been proposed to be at the N-terminal hydrophobic domain of the protein (11). The ubiquinone-binding site has been indicated to be at the region close to the periplasmic side by using reconstituted proteoliposomes (10), where no membrane potential is generated by the electron transfer from glucose to ubiquinone in the dehydrogenase.

Molecular genetic and biochemical analyses of the several quinoprotein dehydrogenases including glucose, methanol, and alcohol dehydrogenases have been performed, and their primary sequences are available (12-19). Their functional structures appear to be different from each other. Unlike the monomeric and membrane-bound GDH of E. coli (1), Acinetobacter calcoaceticus, or Gluconobacter oxydans (14, 17-20), the soluble MDH of Methylobacterium organophilum XX or Paracoccus denitrificans acts as a tetramer, 22, to transfer electrons to cytochrome c (13, 21, 22). Whereas, membrane-bound ADH of Acetobacter aceti occurs as a trimer, , to transfer electrons via intramolecular heme c moieties to ubiquinone (23). Alignment of the PQQ-binding protein or subunit in these dehydrogenases revealed that they have three homologous regions including the highly conserved region consisting of about 70 amino acid sequences at their C termini (20), which had been predicted as a common PQQ-binding site (16, 17). Therefore, the structure of the catalytic proteins of quinoprotein dehydrogenases may be partially similar to each other, but they transfer electrons to different components. Recently, three-dimensional structures of MDHs from three different sources of Methylophilus methylotrophus, Methylophilus W3A1, and Methylobacterium extorquens AM1 (24-26) have been determined, revealing that the amino acid residues interacting with PQQ and Ca²⁺ are dispersed in the whole -subunit, which forms a superbarrel structure made up of eight topologically identical four-stranded antiparallel -sheets. On the basis of the structure, a model structure of GDH was proposed except for its N terminus and several unique segments that are absent in MDH of M. extorquens (27). However, neither direct evidence supporting the structure nor x-ray crystallography has been reported yet.

In order to identify amino acid residues crucial for GDH function, we introduced random mutagenesis that may be useful for structurally unknown enzyme. A single protein GDH is a good model to elucidate the molecular mechanism of catalytic reaction or intramolecular electron transfer of the primary dehydrogenase in the respiratory chain. This is the first report on GDH mutant isolation and characterization except for one that showed a mutant with little effect on the activity (28). We obtained several E. coli mutant GDHs, and then characterized them in respect to their glucose dehydrogenase, ubiquinone reductase, and glucose oxidase activities. Mutant GDHs with a pronounced effect on the enzyme activities were purified and their kinetic parameters were determined. The mutation effect of some mutants were further examined by introducing different amino acid substitutions. We identified at least two crucial residues, Asp-730 and His-775, possibly related to proton transfer and to binding to PQQ, respectively, to which no corresponding residue was proposed in the GDH model (27).

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results & Discussion References

Materials-- Restriction enzymes, T4 DNA ligase, Taq DNA polymerase, and the DNA sequencing kit were purchased from Takara Shuzo (Kyoto, Japan). Oligonucleotide primers were made by ourselves on a DNA synthesizer, Gene Assembler Plus (Pharmasia, Uppsala, Sweden) or by Sawady Technology (Tokyo, Japan). All other chemicals were of analytical grade.

Bacterial Strains and Plasmids-- The bacterial strains used in this study were all derivatives of E. coli K12. Their relevant genotypes and plasmids are shown in Table I. Mutations gcd::cm, insertion of the cm gene into the gcd gene encoding GDH, and pts, lacking the ptsHI, from YU124 and LJ288, respectively, were transferred into W3110 by P1 transduction (32). The resulting YU245 was still able to grow on M9 glucose minimal medium (32). Thus, after N-methyl-N'-nitro-N-nitrosoguanidine mutagenesis (32) of YU245 and penicillin screening (33), YU312 that grew very slowly on the M9 glucose medium was isolated and used for screening gcd mutants. YU312 exhibited weak red colonies on tetrazolium plates (32) containing glucose and PQQ. While, YU312 cells harboring pACGCD1 or -2, which bears the gcd gene as described below, showed white colonies on the plates.

Subcloning the gcd Gene into a Low Copy Plasmid-- To certainly obtain gcd mutants, the gcd gene on a high copy plasmid, pUCGCD1 (19), was subcloned into a low copy plasmid of a pACYC177 derivative, pACYC177-322, in which the large BamHI-PstI fragment of pACYC177 had been connected with the small BamHI-PstI fragment of pBR322. The EcoRI-PstI fragment bearing the gcd gene from pUCGCD1 was inserted into the EcoRI-PstI site of pACYC177-322, generating pACGCD1. The DraI-PstI fragment bearing the gcd gene from pUCGCD1 was inserted into the HincII-PstI site of pUC118. From the resulting pUCGCD2, the EcoRI-PstI fragment bearing the gcd gene was inserted into the EcoRI-PstI site of pACYC177-322, generating pACGCD2. The gcd gene on pACGCD1 is about 200 base pairs larger at the 5'-upstream noncoding region than that on pACGCD2. We used these two plasmids for different mutagenesis because they have different restriction sites at the 5'-upstream region. After mutagenesis, each mutated region of the gcd gene was replaced into the wild type gcd gene of pUCGCD1 to facilitate enzyme assay.

Mutagenesis-- To mutagenize the conserved C terminus from the 734th to 796th amino acid in GDH, region-specific PCR mutagenesis was performed with two primers, 5'-TTGCATGCCTGCAGGTC-3' and 5'-TTGGATCCAACAGGGTACGCTGGTC-3' in an 0.5-ml microcentrifuge tube in a final volume of 100 µl in which 0.1 nM pUCGCD1, a dNTP mixture (125 µM each of dATP, dTTP, and dGTP, and 12.5 µM dCTP, or 125 µM each of dATP, dCTP and dGTP, and 12.5 µM dTTP), 100 pmol of each oligonucleotide primer, and 2.5 units of Taq DNA polymerase were added in 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 0.5 mM MnCl₂, 1.5 mM MgCl₂, and 0.001% gelatin (w/v). Twenty-five polymerase chain reactions, each of which consisted of denaturation at 94 °C for 1 min, annealing at 55 °C for 2 min, and extension at 72 °C for 2 min, were carried out by a DNA Thermal Cycler, PJ2000 (Perkin-Elmer) as described (10). The PCR products were electrophoresed by Sea-plaque gels (Takara Shuzo) and recovered by ethanol precipitation after extraction from the gel and phenol extraction. The PCR fragments were then digested with BssHII and PstI, and inserted into the BssHII-PstI site of pACGCD2.

Screening of the gcd Mutants-- Mutagenized plasmid DNA was introduced into YU312, and weak red color colonies on tetrazolium plates (32) containing 0.5% glucose, 0.1 µM PQQ, and 50 µg/ml kanamycin were isolated as the gcd mutant. To avoid mutants producing immature GDHs, the isolates were then examined by Western blot using an antibody raised against GDH as described previously (35) and by measurement of GDH activity.

DNA Manipulations and Sequencing-- Conventional recombinant DNA techniques (36) were used. To determine the mutation sites, nucleotide sequencing was carried out by the dideoxy chain termination method (37) after subcloning DNA fragments into M13 mp18 or mp19 vector (38) or with the Thermo Sequenase cycle sequencing kit (Amersham, Brucks, United Kingdom). To determine the region including the mutation site of the mutants from hydroxylamine mutagenesis, recombination between the mutant gcd genes and the wild type gcd gene was performed. When the SalI-PstI fragment of the wild type gcd gene encoding the C-terminal five-sixth portion of the enzyme was replaced by corresponding mutant fragments, all recombinants showed the mutant phenotype. Further recombination experiments using two fragments produced from the SalI-PstI fragment by SmaI digestion indicated that two of them have mutation sites between the SalI and SmaI sites and the other two have between the SmaI and PstI sites. These limited regions were then subjected to nucleotide sequencing.

Site-specific Mutagenesis-- To obtain mutants, R687A, R687D, G689A, D693A, D730A, D730R, and H775A, site-specific mutagenesis was carried out using the Mutan-Super Express Km kit (Takara Shuzo). Mutagenic primers used: 5'-TGGAAGAAAGCTATTGGTAC-3'for R687A; 5'-TGGAAGAAAGATATTGGTAC-3' for R687D; 5'-AACGTATTGCTACGCCGCAG-3' for G689A; 5'-CGCCGCAGGCCAGTATGCCG-3' for D693A; 5'-GCTACGGCAGCTAACTACCT-3' for D730A; 5'-GCTACGGCACGTAACTACCT-3' for D730R; and 5'-GCAGGCGGTGCCGGTTCATT-3' for H775A. Introduced mutations were confirmed by nucleotide sequencing.

Enzyme Assay and Analytical Procedures-- Cells harboring wild type plasmid, pUCGCD1, or mutant plasmids, pUCGCDs, were grown in LB (1% Bacto-tryptone, 0.5% yeast extract, and 0.5% NaCl) containing ampicillin (100 µg/ml) for 14 h at 37 °C, harvested, and washed twice with 0.85% NaCl. The cells were then disrupted by passing twice through a French pressure cell press (16,000 p.s.i.). After removing unbroken cells by a low speed centrifuge, membrane fractions were recovered by centrifugation at 86,000 × g for 90 min. The membrane was suspended in 100 mM Tris-HCl, pH 7.0 (about 10 mg of protein/ml), for enzyme assay. PMS and Q-1 reductase activities of GDH, and glucose oxidase activity were measured as described previously (8, 10). Dehydrogenase (PMS reductase) activity was also measured with maltose, fucose, galactose, or xylose, instead of glucose, as a substrate as described previously (39). Succinate dehydrogenase (PMS reductase) activity was measured as a control by the same assay method used for GDH except that PQQ was omitted. Kinetic parameters were estimated from the Lineweaver-Burk plot drawn by using the program EnzymeKinetics (Trinity Software Technical Support, NH). Protein content was determined according to the Dulley and Grieve method (40) using bovine serum albumin as a standard. Fluorescence spectra were taken by using a Hitachi-650-10S of 1-ml samples (0.4 µM in enzyme) in 100 mM potassium phosphate, pH 7.0, containing 0.2% -octylgluoside and 3 mM MgCl₂ in the presence or absence of 1.5 µM PQQ. Fluorescence emission was scanned from 290 to 460 nm at 25 °C with an excitation at 280 nm.

Purification of Mutant GDHs-- Four mutant GDHs with significant effect on the enzyme activity were purified according to the procedure as described previously (10). In S357L and G689D, all purification steps were performed in the presence of 10 nM PQQ to stabilize enzyme activity. Purified mutant proteins as well as wild type were analyzed by SDS-10% polyacrylamide gel electrophoresis, revealing that their purity was more than 95% homogeneity. To take fluorescence spectra, the purified protein was readsorbed to DEAE column equilibrated with 10 mM potassium phosphate, pH 7.0, containing 0.1% Triton X-100, and after washing with the same buffer in the absence of Triton, the enzyme was eluted with the same buffer containing 0.2% -octylgluoside.

	RESULTS AND DISCUSSION

Top Abstract Introduction Procedures Results & Discussion References

Isolation of the gcd Mutants-- Out of 30 Gcd mutants isolated from region-specific mutagenesis, targeting the conserved C terminus of GDH, four were found to exhibit less than 10% PMS reductase activity of the wild type. In the mutant plasmids, pACGCD2Ms, from the four mutants, the BssHII-PstI fragments corresponding to those prepared by the PCR mutagenesis were sequenced. Three of them were found to have a single mutation in the region and the remaining one has double, and all mutations except for one of the double mutants occurred at the conserved amino acid residues among GDHs of E. coli, A. calcoaceticus, and G. oxydans (Table I). Of these mutants, three mutant GDHs from pACGCD2M9, M13, and M15 were detected by Western blot analysis using an antibody against E. coli GDH but not one from pACGCD2M5 (data not shown). pACGCD2M5 has a nonsense mutation at the 765th codon in the gcd gene, resulting in producing an immature GDH presumably susceptible to intracellular proteolytic degradation. The mutated protein from pACGCD2M9 is expected to have an extra 18 amino acid residues because of mutation of the intrinsic stop codon, which was indicated as a slightly slow migration in SDS-polyacrylamide gel electrophoresis.

Enzyme Activities of Mutant GDHs-- To compare the mutant GDHs with the wild type, activities of PMS reductase, Q-1 reductase, and glucose oxidase were measured with the membrane fractions (Table II). Relative GDH protein contents in the fractions were estimated by Western blot analysis (Fig. 2), and then relative PMS reductase activity was calculated based on the relative GDH contents. The results revealed that all mutant GDHs have PMS reductase activity less than 10% of the wild type, with comparable Q-1 reductase and glucose oxidase activities except for G741S and -797K (where - means a stop codon). The latter two mutant GDHs seem to retain equivalent activities of PMS reductase and glucose oxidase to those of the wild type, so their apparent reduced activities may be due to lower content of GDH in the membrane than that of the wild type. The mutant -797K reduced Q-1 reductase activity without the significant effect on K_m values for PQQ and Mg²⁺ or on glucose oxidase activity, which reflects the normal electron transfer from GDH via intrinsic Q-8 to cytochrome oxidase. Thus, it seems that the additional C-terminal 18 amino acid residues of the mutant GDH may hamper access of the artificial Q-1, or change the conformation of the Q-1 reacting site but not of the Q-8 reacting site.

Kinetic Parameters of Mutant GDHs-- To define characteristics of the mutant GDHs, their kinetic parameters were compared with those of the wild type in the membrane fractions (Table II). S357L, G689D, and H775R showed significantly increased K_m values for PQQ, and especially, H775R had 230-fold higher K_m values for PQQ than wild type. P326L, G689D, G741S, and the double mutant E742G/P757L showed slightly higher K_m values for Mg²⁺. All showed nearly the same K_m value for glucose as that of wild type, indicating that no mutant appears to be affected at the substrate-binding site. Since GDH has some activities for fucose, galactose, xylose, mannose, and maltose (39), substrate specificity of the mutant GDHs was tested but no mutant showed different substrate specificity from the wild type (data not shown). Therefore, it is likely that no mutation occurs in amino acid residues related to the substrate binding, and that the mutations in these mutant GDHs may not cause total conformational changes of the protein.

Mutations Influencing on Affinity for PQQ-- Among the mutant GDHs, the altered amino acid residues of S357L and H775R are located close to the active site in the model. The S357L mutation influenced the affinity for PQQ but not for Mg²⁺, although Thr-353 beside Ser-357 hydrogen bonds to PQQ, and Asp-354 and Asn-355 to Ca²⁺ in the model. From these results and the evidence that S357L purification required the addition of PQQ, it is suggested that the Ser to Leu mutation in S357L may cause a local structural change and also influence the position of Thr-353 to reduce affinity for PQQ.

View larger version (53K): [in this window] [in a new window]   Fig. 1.   Alignment of the C-terminal amino acid sequences of quinoprotein dehydrogenases and mutation positions of some mutant GDHs. ECO-GDH (19), ACA-GDH (14), and GOX-GDH (18) are GDHs in E. coli, A. calcoaceticus, and G. oxydans, respectively, and MEX-MDH (27) is MDH in M. extorquens. This alignment and some structural features were shown according to Ref. 27. Only C-terminal alignment is presented and protein residue numbers were shown next to the protein names. Asterisks show identical residues between the sequences of GDH in E. coli and MDH in M. extorquens. Mutation sites of three mutant GDHs obtained here are shown by boxes.

View larger version (63K): [in this window] [in a new window]   Fig. 2.   Western blot analysis of the mutant GDHs obtained by region-specific and hydroxylamine mutageneses. The membrane fractions from YU312 containing pUCGCD1, pUC118, or pUCGCD mutants that were used for estimating various enzyme activities as shown in Table II were subjected to a SDS-10% polyacrylamide gel electrophoresis, and transferred to the blot membrane. YU312 cells containing pUCGCD1 or pUC118 were used as a positive and negative control, respectively. The wild type and mutant GDHs (GDH) were visualized using a polyclonal antibody against E. coli GDH as described previously (10). The relative amounts of the proteins were densitometrically estimated by using BIO-RAD Molecular Imager. Lanes 2-7 and 9-14 represent the membrane fractions from the positive control (wild type, 2.0 µg), the positive control (1.0 µg), G741S (12 µg), -797K (60 µg), H775R (2.0 µg), the double mutant of E742G/P757L (60 µg), the positive control (1.0 µg), the negative control (60 µg), G689D (2.0 µg), S357L (2.0 µg), D730N (2.0 µg), and P326L (2.0 µg), respectively. Lanes 1 and 8 are prestained markers, phosphorylase b (105,000) and bovine serum albumin (70, 800).

Possible Function of Asp-730-- D730N showed significantly decreased activities of PMS reductase, Q-1 reductase, and glucose oxidase, but no effect on affinity for PQQ or Mg²⁺ (Tables II and III). The substitution from Asp to Asn does not seem to cause a large conformational change. To further analyze the function of Asp-730, two mutants, D730A and D730R, were constructed and characterized with membrane fractions (Table IV, Fig. 3). Both mutants also showed low PMS reductase activities comparable to D730N.

View larger version (62K): [in this window] [in a new window]   Fig. 3.   Western blot analysis of the mutant GDHs obtained by site-specific mutagenesis and their stability in membrane fractions. The membrane fractions from YU312 containing pUCGCD1, pUC118, or pUCGCD mutants were used. Gel electrophoresis, blotting, and visualization were done as shown in Fig. 2. A, the mutant GDHs from site-specific mutagenesis in membrane fractions were analyzed after one freeze-thaw treatment. Lanes 1-10 represent the membrane fractions (1 µg) from the positive control (wild type) and negative control, R687A, R687D, D730A, D730R, H775A, G689A, D693A, and the positive control, respectively. B, R687A, R687D, and the wild type GDHs in membrane fractions which were freshly prepared in the presence or absence of 10 nM PQQ were analyzed. Lanes 1-6 represent the membrane fractions from the wild type (1 µg), the wild type (+ PQQ, 1 µg), R687A (10 µg), R687A (+ PQQ, 10 µg), R687D (10 µg), and R687D (+ PQQ, 10 µg), respectively. Lanes M are prestained markers.