EP2294188A1

EP2294188A1 - Novel variants of pqq-dependent glucose dehydrogenase having improved substrate specificity

Info

Publication number: EP2294188A1
Application number: EP09768932A
Authority: EP
Inventors: Walter Weichel; Volker Moehrle; Markus Schindler; Rainhard Koch; Rolandas Meskys
Original assignee: Bayer Technology Services GmbH
Current assignee: Bayer Intellectual Property GmbH
Priority date: 2008-06-26
Filing date: 2009-06-17
Publication date: 2011-03-16
Also published as: JP2011525361A; CN103540572A; DE102008030435A1; CN102076846A; US8580547B2; WO2009156083A1; US20110236920A1

Abstract

The invention relates to novel PQQ-dependent soluble glucose dehydrogenases (sPQQGDH) which have an increased substrate specificity compared with the wild type, and also to methods for production and identification thereof.

Description

Novel variants of PQQ-dependent glucose dehydrogenase with improved substrate specificity

The present invention relates to PQQ-dependent glucose dehydrogenases (sPQQGDH), methods for their production and identification, genes encoding the sPQQGDH according to the invention, vectors for producing the genes, and glucose sensors comprising a glucose dehydrogenase according to the invention.

The enzymatic determination of glucose is of great importance in medical diagnostics. Here, the detection of glucose in the urine or in the blood to determine or follow-up of metabolic diseases is necessary. In diabetes mellitus, it may be necessary to determine the blood glucose level several times a day. This is done in the vast majority of measuring instruments on the market with an enzymatic process in which glucose is oxidized and the resulting hydrogen (2 H ⁺ and 2 e ^" ) is amperometrically quantified (P. Vadgama, J.Med.Eng Technol 5 [6], 293-298 (1981), KS Chua and LK Tan, Clin. Chem. 24 [1], 150-152 (1978)).

There are several different enzymes that are suitable for this purpose, but which work with different cofactors for the primary uptake of electrons and also have different biochemical properties such as specificity, stability or turnover rate. The conversion rate of an enzyme is usually expressed in units of activity (U); For glucose-oxidizing enzymes, the specific activity (U / mg) usually refers to the amount of glucose in micromol that is oxidized by one milligram of enzyme per minute under specified reaction conditions.

Frequently used enzymes are glucose dehydrogenases (GDH). In the case of NAD (P) ⁺ -dependent GDH, the cofactor is nicotinamide adenine dinucleotide (NAD ⁺ ) or nicotinamide adenine dinucleotide phosphate (NADP ⁺ ) (HE Pauly and G. Pösseiderer, Hoppe Seylers, Z. Physiol Chem., 356 [10] 1613-1623 (1975)), both quite unstable compounds that must be added to the reaction because they are not bound to the enzyme.

Another class of glucose dehydrogenases contains pyrroloquinoline quinone (PQQ) bound as an electron acceptor in the enzyme. Two structurally distinct glucose dehydrogenases of this type have been described, a membrane-anchored form mPQQGDH (AM Cleton-Jansen et al, J. Bacteriol 170 [5], 2121-2125 (1988)) and a smaller, soluble form sPQQGDH (AM Cleton-Jansen et al., Mol. Gen. Genet. 217 [2-3], 430-436 (1989)). Latter is due to their high spec. Activity, their insensitivity to oxygen and the tightly bound, stable cofactor enzyme, which is particularly suitable for the production of test systems for the determination of glucose. Therefore, this enzyme is used in a number of commercial blood glucose meters, despite a lower substrate specificity that this enzyme has compared to the other enzymes mentioned. Of course, due to the latter limitation, it would be desirable to have variants of this enzyme which have a higher substrate specificity.

In addition to the glucose substrate, the sPQQGDH also converts other sugars to the corresponding lactone, preferably disaccharides, whose reducing sugar is glucose, but also other reducing sugars. Since such sugars are not normally present in human blood, an enzyme assay based on sPQQGDH usually gives a result that corresponds to the glucose concentration. In the context of certain medical therapies, however, substances are administered whose degradation products include maltose, which is oxidized by the sPQQGDH and thus undesirably contributes to the measurement result (T. G. Schleis, Pharmacotherapy 27 [9], 1313-1321 (2007)). With the administration of xylose and galactose in diagnostic procedures or as an additive to medicines, these sugars are even recorded. It therefore makes sense to appropriately suppress these ancillary activities of the sPQQGDH.

Several attempts have been reported in the literature and in patents to improve the substrate specificity of sPQQGDH by introducing mutations (S. Igarashi et al., Biomol. Eng 21 [2], 81-89 (2004)). For this purpose, individual amino acids in the region of the catalytic center (EP1666586A1) but also elsewhere in the enzyme (US7132270B2) were exchanged for other amino acids by targeted mutagenesis or individual amino acids were added by insertion (EP1 367120A2). In addition, combinations of different mutations are described (EP1666586A1, US7132270B2). Furthermore, variants were described in which two amino acids were inserted at one site of sPQQGDH. However, no significant improvements in substrate specificity could be achieved thereby (WO2006085509).

Despite these results, there is still a need for variants of sPQQGDH, as no variants have been described so far that, with sufficiently high reactivity with glucose, allow sufficient discrimination between glucose and other sugars. Particularly important here is the ability of the enzyme to make the glucose concentration reliably determinable even in the presence of a mixture of glucose and other sugars. None of the variants described sufficiently demonstrates this property. Starting from the state of the art, the object thus is to provide glucose dehydrogenases which enable reliable determination of glucose even in the presence of a mixture of glucose and other sugars. In particular, the object is to provide glucose dehydrogenases, with which glucose can be reliably determined in the presence of maltose. A further object is to provide a method by which such glucose dehydrogenases can be prepared and identified which are particularly suitable for the determination of glucose in the presence of other sugars, in particular in the presence of maltose.

Surprisingly, it has been found that sPQQ glucose dehydrogenases, in which three to five amino acids are inserted in the region of the substrate binding region, have a glucose specificity which is better than the starting form of the enzyme.

An object of the present invention are therefore sPQQ glucose dehydrogenases having an improved glucose specificity compared to the starting form of the enzyme, characterized in that three to five amino acids are inserted in the region of the substrate binding region.

The sPQQGDH according to the invention can be encoded by genes which are produced by mutagenesis from wild-type strains naturally occurring, such as Acinetobacter calcoaceticus LMD 79.41 (K.Kojima K. et al, Biotechnology Letters 22, 1343-1347 (2000)) and other strains of the genus Acinetobacter (PJ Bouvet PJ and OM Bouvet, Res. Microbiol., 140 [8], 531-540 (1989)), in particular those which are characterized by increased Thermostabi quality as described in EP 1623025 Al Acinetobacter sp. Tribes, to be won. Preferred strains are those described in EP 1623025 Al Acinetobacter sp. Strains PT16, KOZ62, KOZ65, PTN69, KG106, PTN26, PT15, KGN80, KG140, KGN34, KGN25, and KGN100; the strain KOZ65 is particularly preferred.

In order to be able to obtain the sPQQGDH according to the invention, an insertion of codons at the position corresponding to the substrate binding site must take place in a sPQQGDH gene which is suitable in this way. The spatial structure of some sPQQGDH is known and deposited in public databases (A. Oubrie et al., J. Mol. Biol. 289 [2], 319-333 (1999); A. Oubrie et al, EMBO J. 18 [19] , 5187-5194 (1999)). Thus, the removal of individual amino acid positions to the bound substrate can be estimated. Such estimates are possible, for example, with visualization programs for proteins such as the Cn3D software program available on the Internet (http://www.ncbi.nlm.nih.gov/Structure/CN3D/cn3d.shtml). The sequence of the published X-ray structure is based on the sPQQGDH from Acinetobacter calcoaceticus, which is 455 amino acids long. The starting form of the particularly preferred sPQQGDH from strain KOZ65 has 456 amino acids, as does the enzyme from Acinetobacter baumannii. In both cases an additional glutamine residue is responsible for the position 289. In the following, however, statements on the position of individual amino acids still refer to the shorter sequence of sPQQGDH from the strain Acinetobacter calcoaceticus LMD 79.41 used in the X-ray structure. The amino acid located there is given before the position in the one-letter code for amino acids.

Preferred for mutagenesis are amino acid regions whose distances from the substrate are not more than 5 angstroms, e.g. the positions Q76, D143, Q168, Y343 and the adjacent amino acid residues. For example, before or after one of the specified

Positions are made an insertion of three or more amino acids to change the spatial nature of the substrate binding site. Particularly preferred is a

Insertion of additional amino acids between positions Q168 and L1 69. Variants of KOZ65 with insertions at this site surprisingly showed to a large extent significantly reduced activity on maltose as compared to glucose. Particularly preferred are sPQQGDH with the insertion sequences listed in Table 1 between the positions

Q168 and L169. For comparison, the wild type is indicated in the first line. The sPQQGDH according to the invention are characterized, as shown in Table 1, by increased substrate specificity (lower maltose / glucose ratio and lower level of interference of glucose in a glucose / maltose mixture).

Table 1:

The present invention further provides a method for the production and identification of sPQQGDH, which show increased substrate specificity compared to the wild type. The method according to the invention allows the generation of a multiplicity of variants by means of local random mutagenesis, with three, four, five or more inserted amino acids at a desired site of the sPQQGDH sequence. The enzymes produced are screened for substrate specificity and the best variants selected.

The inventive method is characterized in that in a first step, a vector is prepared in which a suitable sequence area in front of and behind the desired

Insertion site was deleted and replaced by a unique for this vector restriction site. In vector DNA which is opened with this restriction enzyme, in a second step double-stranded, oligonucleotide sequences produced by PCR amplification are inserted, which contain the desired insertion in addition to the sequence deleted in the vector, so that a gene product is produced which has no further amino acid sequence

Changes than the insertion of three or more amino acids in the desired location. The PCR produced from synthetic oligonucleotides to be inserted

Double-stranded fragment contains the three or more additional, randomized codons for

Change in substrate specificity, but also the codons deleted in the vector. For this purpose, the codons were selected which express themselves particularly well in E. coli. The

Nucleotide sequence therefore deviates from the sPQQGDH gene next to the insertion region

KOZ65, but not the amino acid sequence.

For the insertion of three additional amino acids 8,000 enzyme variants are conceivable, with four additional amino acids it is 160,000 and with five additional amino acids it is 3.2 Millions of variants. In the synthesis of the mutagenic oligonucleotides, therefore, it is possible to use, in addition to completely randomized sequences, those which contain degenerate codons at some or all positions for the particular amino acids to be inserted. This can be used to reduce the number of possibilities overall, to avoid stop codons, or to favor single amino acids at specific positions.

In a third step of the method according to the invention, recombinant vectors which have been prepared in the manner described above are transformed into suitable host bacteria and the host bacteria are propagated in a suitable nutrient medium. The methods for cloning genes from microorganisms and their expression in another host, as well as the techniques for isolating and altering nucleic acids, are known to the person skilled in the art (see, for example, Molecular Cloning - A Laboratory Manual, Vol 1, Sambrook, J. and Russel, DW, 3rd edition, CSHL Press 2001). Likewise, the cultivation of microorganisms and the purification of recombinant enzymes thereof are state of the art (see ibid, Vol. 3).

In a fourth step of the method according to the invention, the bacterial colonies are tested for their activity against various sugars and the best ones identified. According to the invention, not only a test is carried out with respect to the activity towards the individual sugars but also with respect to the interference of different sugars.

The activities may be considered specific activities, i. Conversion rates, which are implemented by a defined amount of enzyme per unit time, can be determined. For glucose oxidizing enzymes, the specific activity (U / mg) usually refers to the amount of glucose in micromoles that is oxidized by one milligram of enzyme per minute under specified reaction conditions. As a rule, glucose is present in excess of the enzyme in order to eliminate the influence of the glucose concentration on the reaction rate. The conversion rate is determined by monitoring the degradation of a reactant or the formation of a product over time. The time tracking can e.g. done photometrically. For determination of reaction rates and turnover rates, reference may be made to textbooks of physical chemistry (e.g., Peter Atkins, Physical Chemistry, W.H. Freeman & Co., 7th Edition, 2002).

First, the individual colonies are divided and an enzyme test (determination of the specific activity) with glucose as substrate and a second enzyme test with a second sugar, e.g.

Maltose performed as a substrate. Opposed to the colonies with the highest activities Glucose with at the same time least activity towards the second sugar (eg maltose) also determines the activity against a mixture of glucose and the second sugar (eg maltose). It turned out, surprisingly, in the study of some variants with very low relative reactivity to maltose as the sole substrate, that this property of one variant is not necessarily associated with an independent property, namely the reliable discrimination of glucose and maltose, when both sugars are present as a mixture , This can lead both to the overestimation and to the underestimation of the glucose concentration present in this mixture. In the search for enzymes to be used for the determination of glucose concentrations in test samples, therefore, a high glucose specificity on mixtures of glucose and other sugars is of crucial importance, or in other words, the lowest possible interference of the sugar in question. The interference of a sugar is calculated from the enzyme activity measured on a mixture of glucose and interfering sugar, minus the activity measured at the same concentration of glucose alone, normalized to this glucose activity (see Eq.

j V mixture - V reference, f ^ <. >.

^~ V reference

In equation 1 / interference of glucose with another sugar, V _R φ _ren: the rate at which glucose degraded by the observed enzyme and V _Miscnung the rate at which glucose degraded in a mixture with each other sugar , The rates can be determined, for example, photometrically (see examples 3, 4, 7 and 9).

Positive values in this context mean that the enzyme activity is higher on a mixture than on glucose alone; So apparently both sugars are oxidized. Negative values may be interpreted as oxidizing the interfering sugar, but decreasing the activity of the enzyme for glucose. Both effects are undesirable for an enzyme to be used for the specific quantification of glucose.

The method according to the invention makes it possible to produce and identify variants of sPQQGDH whose substrate specificity has changed compared with the starting form of the enzyme. The method according to the invention can also be modified and / or expanded in such a way that instead of the substrate specificity or in addition to this, other biochemical properties such as eg thermostability are tested and thus variants improved with respect to the starting form the other biochemical properties are identified. The method according to the invention thus makes it possible to select variants of sPQQGDH with particularly advantageous properties.

The present invention also relates to genes which encode the sQQQGDH according to the invention and vectors for generating the genes.

The present invention furthermore relates to a glucose sensor comprising an sPQQGDH according to the invention for the determination of glucose. The construction and operation of glucose-based dehydrogenase-based glucose sensors are known to those skilled in the art (see, for example, EP146332A1). Thereafter, such a glucose sensor comprises an electrode system of a working electrode, a counter electrode and a reference electrode on an insulating plate, on which an enzymatic reaction layer is applied, which includes a glucose dehydrogenase and an electron acceptor in contact with the electrode system.

Suitable working electrodes include carbon, gold, platinum and similar electrodes on which an enzyme according to the invention is immobilized by cross-linking agent: embedding in a polymer matrix, jacketing with a dialysis membrane, using a photocrosslinkable polymer, an electrically conductive polymer or a redox polymer, fixing the enzyme in a polymer or adsorbing on the electrode with an electron mediator including ferrocene or its derivatives or any combination thereof. Preferably, sPQQGDH according to the invention are immobilized on the electrode in the form of a holoenzyme, although they may also be immobilized as an apoenzyme and provided with PQQ as a separate layer or in a solution. Usually, sPQQGDHs of the present invention are immobilized on a carbon electrode with glutaraldehyde and then treated with an amine-containing reagent to block the glutaraldehyde. As the counter electrode, e.g. a platinum electrode and as a reference electrode e.g. an Ag / AgCl electrode can be used.

The amount of glucose can be measured as follows: PQQ, CaCl ₂ and a mediator are placed in a thermostat cell containing a buffer and left at a constant temperature. Suitable mediators include, for example, potassium ferricyanide and phenazine methasulfate. An electrode on which an sQQQGDH of the present invention has been immobilized is used as a working electrode in combination with a counter electrode (eg, a platinum electrode) and a reference electrode (eg, Ag / AgCl electrode). After to When a steady state current has been applied to the working electrode, a glucose-containing sample is added to measure the increase in current. The glucose level of the sample can be read from a calibration curve made with glucose solutions at standard concentrations.

In the following, the invention will be described in more detail by way of examples without, however, restricting them to the examples.

Examples

The chemicals listed in the following examples are commercially available, e.g. at the company Sigma-Aldrich, or at the respective specified companies.

Example 1: Preparation of a vector for insertion of additional amino acids between Q168 and L169

For the production of variants according to the invention, a gene for an sPQQGDH must be available in an expression plasmid which can be propagated in microorganisms which permits induced expression of sPQQGDH and which can be used to carry out the desired mutations. A variety of plasmids and microorganisms are known in the art, which can be used for this purpose. In this and all other examples, a construct based on the commercially available vector pASK-IBA3 (IBA GmbH, Göttingen) was used. In pASK-IBA3, the sPQQGDH gene from the strain Acinetobacter sp. KOZ65 cloned in the following way. From genomic DNA of Acinetobacter sp. KOZ65 was probed with the two oligonucleotides GDH-U3 (5'-TGGTAGGTCTCAAATGAATAAACATTTATTGGC TAAAATTAC-3 '; SEQ ID 19) and GDH-L5 (5'TTCAGCTCTGAGCTTTATATGTAAATCTAATC-S'; SEQ ID 20) by means of a PCR reaction (Phusion DNA Polymerase, Finnzymes Oy) according to the manufacturer's instructions generated a 1, 46 kb long DNA fragment. The fragment was purified and then incubated with the restriction enzyme Bsal. The vector pASK-IBA3 was opened with the restriction enzyme HindIII. The restriction enzyme was denatured by heat treatment and the resulting DNA overhangs were filled in with T4 DNA polymerase in the presence of dATP, dCTP, dGTP and dTTP. Subsequently, the DNA was purified and incubated with the restriction enzyme Bsal. The thus released, 1 14 bp fragment was discarded; into the 3.1 kb vector fragment, the PCR amplicon was ligated with the sOQ65GDH gene from KOZ65. The ligation product was transformed into transformation competent cells of the E. coli strain DH5α. The vector pAI3B-KOZ65 prepared in this way is suitable for the inducible expression of KOZ65 sPQQGDH. The nucleotide sequence of the vector pAI3B-KOZ65 is given under SEQ ID 21, the amino acid sequence of the coded sPQQGDH under SEQ ID 22.

To be able to insert three or more amino acids between Q 168 and L 169, the vector pAI3B-KOZ65-U was first prepared. For this, plasmid DNA of the vector pAI3B-KOZ65 was used as template for a PCR amplification A with the primers IBA-For (5'-TAGAGTTATTTTACCACTCCCT-3 ', SEQ ID 23) and Ecol09-Up (5'-GGTTAGGTCCCTGATCACCAATCG-3'; SEQ ID 24) and a PCR amplification B with the primers BfuA-US 1 (5 '-CACAGGTACACCTGCCGCCACT-S', SEQ ID 25) and Ecol 09-Down (5'-TACGAGGACCCAACAGGAACTGAGC-S '; SEQ ID 26). The amplificate formed from A was purified and cut with the restriction enzymes XbaI and EcoO109I. The resulting 588 base pair (bp) fragment was isolated. The B-derived amplicon was purified and cut with the restriction enzymes BfuAI and EcoO109I. The resulting 353 bp fragment was also isolated. The vector pAI3B-KOZ65 was cut with the restriction enzymes BfuAI and XbaI. The resulting, 3567 bp fragment was isolated on an agarose gel. Kits and reagents from New England Biolabs GmbH (Frankfurt), Invitrogen (Karlsruhe), Qiagen (Hilden) and Roche Diagnostics (Mannheim) were used according to the manufacturer's instructions for the work steps described. Additional methods were taken from the set of methods of Sambrook and Russell (Molecular Cloning - A Laboratory Manual, Sambrook, J. and Russel, DW, 3rd edition, CSHL Press 2001).

The three fragments were then ligated and transferred by chemical transformation into E. coli DH5α. From 12 colonies plasmid DNA was prepared and checked by appropriate restriction digests to see if the desired plasmid was formed, which was the case in 9 isolates. By sequencing four of these clones, it was confirmed that in all cases no further changes than those shown in Figure I e took place.

The vector pAI3B-KOZ65-U (SEQ ID 27) now contains a unique EcoO109I restriction site (RG'GNCCY). EcoO109I generates 3 bp long 5 'overhangs in which the middle nucleotide is arbitrary; in pAI3B-KOZ65-U it is an A on the coding strand. The DNA coding for the amino acids R166 to Pl82 and another nucleotide are deleted in pAI3B-KOZ65-U, which leads to a frame shift. The attempt to express this gene therefore leads to a truncated polypeptide without function (SEQ ID 28). The vector is suitable for the uptake of synthetic, double-stranded DNA fragments that result in the recovery of a functional sPQQGDH gene with insertions between positions Q 168 and L 169, as further described in Example 2.

Example 2 - Preparation of sPQQGDH Insertional Mutants Four synthetic oligonucleotides were used to prepare insertion mutants having four additional amino acids between positions Q 168 and L 169: EarV2 random (5'-CGACGTAACCAGNNKNNNNKKNTKCTGGCTTACCTG-3 ', SEQ ID 29), EarV2-Lower (5'-GTGTGCTGTGCCTGGTTCGGCAGGAACAGGTAAGCCAG-3 '; SEQ ID 30), EarV2-UpAmp (5'-CCTACCTACGACTCTTCCGACGTAACCAG-S'; SEQ ID 31) and EarV2-LoAmp (5'-CCATGCTCTTCAGTCGGAGTGTGCTGTGCC-3 '; SEQ ID 32). In Fig. 2, the strategy for preparing a double-stranded insertion fragment from these oligonucleotides is schematic shown. The sequence of the four amino acids to be inserted is randomized in this case: NNK NNK NNK NNK. N stands for a mixture of all four nucleotides (dATP (A), dCTP (C), dGTP (G), dTTP (T)) in the synthesis of the oligonucleotide at the respective position, while K stands for G or T. The sequence NNK allows codons for all 20 amino acids but only for one of the three stop codons, reducing the likelihood of incomplete gene products. Other combinations of nucleotides at the individual positions are freely selectable, for example to restrict the possible amino acids at a codon position. Oligonucleotide EarV2 random contains the coding sequence from Rl 66 to Ll 72, the opposite strand oligonucleotide EarV2-Lower overlaps in the range Ll 69 to Ll 72 and extends to Tl 81. Oligonucleotide EarV2-UpAmp overlaps for 12 bp at the 3 'end with the 5 'end of EarV2 random oligonucleotide while EarV2-LoAmp oligonucleotide overlaps at the 3' end with the 5 'end of EarV2-Lower oligonucleotide for also 12 bp. The latter two oligonucleotides each contain an Earl interface (CTCTTCN 'NNN).

First, the oligonucleotides EarV2 random and EarV2-Lower in the presence of 10 mM Tris-HCl pH 7.9 at 25 ⁰ C, 10 mM MgCl _2, 50 mM NaCl, 1 mM dithiothreitol (Buffer NEB2) (in a PCR machine GeneAmp 9600, Perkin-Elmer) heated to 90 ⁰ C and cooled to 10 ⁰ C for 5 min. The thus formed, partially double-stranded pieces of DNA were synthesized by addition of all four DNA nucleotides and the Klenow fragment of DNA polymerase from E. coli by incubation for 15 min at room temperature to complete double strands. Thereafter, the reaction mixture was used in a PCR reaction with the oligonucleotides EarV2-UpAmp and EarV2-LoAmp to amplify the resulting duplex. Since the overlap of the amplifying oligonucleotides EarV2-UpAmp and EarV2-LoAmp to the newly formed double strand is only 12 bp each, the first two amplification cycles were not with the thermostable polymerase of the PCR kit but with the Klenow fragment of the DNA polymerase from E. coli to obtain the binding of the oligonucleotides EarV2-UpAmp and EarV2-LoAmp used as primers. For this purpose, the mixture was brought to 94 ⁰ C for 1 min and then placed on ice. Then 1 μL of DNA polymerase (Klenow fragment) was added and incubated for 3 min at 18 ° C and 1 min at 25 ° C. Thereafter, the batch was brought again for 1 min at 94 ° C and then placed on ice, 1 ul DNA polymerase (Klenow fragment) was added and incubated for 3 min at 18 ° C and 1 min at 25 ° C. Only then the following amplification program was in the PCR machine carried out (20 cycles): 1 min 94 ⁰ C, 1 min 52 ⁰ C, 1 min 72 ° C.

After PCR amplification, the reaction was purified by column chromatography (PCR Purification Kit, Qiagen) and cut with Earl. The reaction mixture was then added on an agarose gel (4% agarose) and the desired 64 bp fragment (SEQ ID 33) was isolated from the gel and purified.

The vector pAI3B-KOZ65-U described in Example 1 was cut with EcoO109I and subsequently dephosphorylated with alkaline phosphatase according to the manufacturer's instructions (New England Biolabs) in order to prevent recircularization of the vector. Subsequently, the 64 bp fragment and the dephosphorylated, EcoO109I linearized vector in a ratio of 5 to 1 overnight at 16 ⁰ C were ligated. The 5'-overhangs formed by EcoO109I and Earl are compatible in this way in such a way that the 64 bp fragment can be ligated in the desired orientation relative to the vector, but not in the other orientation with the vector. Thus, ligation products are excluded with misoriented insertions, but not the formation of multimers, each with the correct orientation. However, such genes do not lead to the expression of functional sPQQGDH proteins, because there is a frame shift. The resulting ligation product was transformed into E. coli OH5a. From the transformation mixture, 1/25, 4/25 and 20/25 were plated on a 20 cm × 20 cm agar plate to obtain a colony density of about 1000-5000 colonies per plate at least at one dilution.

This example shows the preparation of a library whose clones express sPQQGDH variants with four additional amino acids each. The preparation of libraries with different length of insertion has been carried out in the same way.

Example 3 - Screening of a Variant Library

To select suitable variants from a library prepared according to the method described in Example 2, 2200 colonies were taken from two agar plates with a total of about 4000 colonies using a colony-picking machine QPix (from Genetix) and in microtiter plates (MTP) Transfer 200 μL LB medium with ampicillin ad 100 μg / mL to each well. In addition, four wells were inoculated on 4 MTP with cultures of the parent strain £ .co // - DH5α :: pAI3B-KOZ65. The MTP were sealed with a gas-permeable film (Airpore Sheets, Qiagen) and shaken for 5 h at 37 ⁰ C. Subsequently, 50 μL of each individual culture was transferred to a new culture vessel, also in an MTP, in which 150 μL LB medium with ampicillin was present. For this purpose, a pipetting robot was used, which performs 96 pipetting operations in parallel. In addition, this medium contained 200 ng / mL anhydrotetracycline to induce the expression of the sPQQGDH gene. These MTPs were shaken at 28 ^° C. overnight. In order to detect the enzyme activity of a large number of colonies, a test was used in which the digestion of the cells is not required. After activation of the enzyme by addition of PQQ, the oxidation of the sugar can be directly due to the decolorization of dichlorophenolindophenol (DCPIP) whose reduction will be pursued. For screening, 40 μL of the induced culture were mixed with 60 μL glucose test solution and another 40 μL with 60 μL maltose test solution from each well, with one MTP containing either only glucose test solution or only maltose test solution. Again, the pipetting robot was used. In this way, each individual culture could always be assigned a consistent position on all culture and measuring plates. The glucose test solution contained the following components (in parentheses: final concentration in the assay): 50 mM glucose (30 mM), 750 μM DCPIP (450 μM), 1.5 μM PQQ (0.9 μM), 1 mM CaCl ₂ (0 , 6 mM), 0.1% NaN ₃ (0.06%), silicone defoamer (Fluka # 85390) 62.5 ppm (37.5 ppm), 50 mM tris- (hydroxymethyl) -aminomethane-HCl (TRIS-HCl ) pH 7.6 (30 mM). In the case of the maltose test solution, no glucose but maltose was present in the same concentration.

From each MTP with cultures, two plates were produced in this way, one for glucose, one for maltose. From each plate, the optical density was determined at a wavelength of 605 nm at various times after mixing culture and test solution (Eppendorf biophotometer): after 1 min, after 7 min, after 12 min and after 18 min. In this way, one enzyme kinetics for glucose and one for maltose as substrate could be taken from each originally isolated colony. From this it was possible to determine both the relative activity of the clones with each other with glucose as substrate and the ratio of the activities on glucose and maltose for each individual clone. Figure 3 shows the result of such a measurement for MTPs 13 and 22 from the screening of a library of five additional amino acids between positions Q 168 and L 169. It can be seen that about 2/3 of the colonies tested show a glucose turnover of less than 50 relative units, which is below the display accuracy of the test. In these colonies no significant expression of sPQQGDH is detectable, which is due to low growth of the culture, lack of induction of the sPQQGDH gene, unproductive double insertions or insertion of amino acid combinations that no longer allow enzymatic activity of the protein formed. Of the remaining colonies, the majority share less activity on maltose compared to glucose. Results from primary screenings as shown in FIG. 3 show a similarly high conversion rate of many variants in comparison with the starting clone ^I .co // - DH5α :: pAI3B-KOZ65. However, the relative activity is always to be verified, because this test, which is designed for a large number of individual measurements, no longer differentiates above a certain conversion rate can. Variants of sPQQGDH generally have a lower turnover rate compared to the sPQQGDH from KOZ65 than the screen shown would suggest.

Example 4 - Refined Screening of Individual Colonies

From a screening as described in Example 3, 188 colonies were selected in which both the activity on glucose and the ratio of activities on glucose and maltose appeared particularly promising, on two new MTPs together with cultures of the parent strain £ .co // - DH5α :: pAI3B-KOZ65 and, as described in Example 3, first grown at 37 ° C and then induced at 28 ° C overnight. For induction, two plates of each MTP were applied to get enough culture volume for the screening experiment. After induction, the parallel grown cultures of approximately 200 μL each were combined and mixed. Of each of these induced cultures, 40 μL each were transferred to plates each containing 60 μL of the test solutions described in Example 3, but with the following sugars: G - 50 mM glucose, M - 50 mM maltose, GM - 50 mM glucose and 50 mM maltose, GGaI - 50 mM glucose and 50 mM galactose or GXyI - 50 mM glucose and 50 mM xylose. These plates were measured immediately after mixing culture and measuring solution for 30 minutes at intervals of 30 seconds to obtain kinetics of each culture. From the data obtained for each culture, the rate of turnover on glucose and maltose alone, the ratio of activities on glucose and maltose, and above all the influence of maltose, galactose and xylose on the conversion of glucose (interference) can be estimated. FIG. 4 shows the result of this follow-up of selected colonies (second screening) for the screening described in example 3.

Two colonies of MTP 13 from the first screening (13-G3 and 13-B2, each marked with a rectangle in Figures 3 and 4) are shown by way of example to show colonies with a promising ratio of maltose to glucose activity (Figure 3). may show very different interference behavior: 13-G3 undergoes negative interference by maltose, 13-B2 positive interference. A total of 28 colonies were selected for further investigation on the basis of the two screenings described above.

Example 5 - Culture of strains expressing variant forms of sPQQGDH Initially, one of four selected colonies of a library with three additional amino acids between each of the positions Q168 and L169 was precultured as follows: 2 mL TB medium (at 100 μg / mL Ampicillin) were inoculated with the colony to be tested and shaken overnight at 37 ⁰ C and 225 rpm. For the main culture, 50 ml of TB medium (100 μg / ml ampicillin) were inoculated with the mature preculture. The culture was shaken at 37 ^° C. and 225 rpm until an optical density at 605 nm (OD ₆ Qs) of about 1 was reached. Then the expression of the enzyme was induced by the addition of an anhydrotetracycline (AHT) stock solution. The inductor was dissolved in DMF. The final concentration of AHT in the culture was 0.2 g / ml, the induction was carried out for 24 hours at 27 ⁰ C and 225 rpm. The harvesting of the cells was carried out by centrifugation of the main culture at 3220 xg for 15 min at 4 ° C. Depending on the amount, the cell sediments were resuspended in 10-20 ml of 50 mM 3-morpholino-propanesulfonic acid buffer (MOPS) pH 7.6 and 2.5 mM CaCl ₂ and disrupted by treatment with ultrasound until clear clarification of the cell suspension was detectable , Cell debris and any existing inclusion bodies were centrifuged at 48,745 xg for 10 min at 4 ⁰ C.

Example 6 - Purification of Variants sPQQGDH

The supernatants from Example 5 were diluted 1: 5 to 10 or 50 ml with column buffer (10 mM MOPS pH 7.6 + 2.5 mM CaCl ₂ ) and dried over a cation exchanger (Toyopearl CM-650M, TOSOH BIOSEP GmbH). separated by chromatography. For this purpose, a 10 x 1.42 cm column with about 16 mL column bed is used; the flow rate was 4 mL / min. The column was equilibrated with 10 column volumes of buffer followed by sample application. The column was washed with 4 column volumes of buffer, then the sPQQGDH was eluted by means of a linear salt gradient from 0 to 0.4 M NaCl in column buffer and the eluate was collected in fractions of 1 or 3 mL. sPQQGDH eluted under these conditions as a nearly homogeneous protein, so that no further purification was required. For protein determination, the OD _{280 of} the solution was determined. Using purified sPQQGDH protein from KOZ65, it had previously been determined by gravimetric analysis that the OD ₂₈ o of a 1 mg / mL sPQQGDH solution (PQQ-free apoenzyme) was 1, 27.

Example 7 - Determination of the Enzyme Activity of sPQQGDH Variants

The following test was used to accurately determine the enzymatic activity. Each of the enzyme solutions to be investigated was diluted in the following buffer: 50 mM piperazine-1,4-bis (2-ethanesulfonic acid) (PIPES) pH 6.5, 0.1% Triton XI OO, 1 mM CaCl ₂ , 0.1% bovine serum albumin (BSA) and 3 μM PQQ. The substrate sugar, the electron mediator N-methylphenazonium methasulfate (PMS) and the detection reagent

Nitro blue tetrazolium chloride (NBT) was prepared separately: 1 1, 7 mL of a solution containing 50 mM PIPES pH 6.5 and 2% Triton X-100 in double distilled (bidist.) Water were added at 450 μL

0.9 M D-glucose, 450 μL 6 mM PMS and 450 μL 6.6 mM NBT (all substances also dissolved in bidistilled water), stored in the dark at 25 ⁰ C and consumed within one hour. If the enzyme activity is to be measured on sugars other than glucose, these are used in place of the glucose in a concentration of 0.9 M. The final concentrations of the individual reaction partners in the measurement solution were: Sugar substrate: 30 mM, PMS: 200 μM and NBT: 220 μM. 725 ul of the pre-heated test solution were mixed with 25 ul of diluted enzyme solution in a cuvette and the formation of formazan was monitored at 570 nm and 25 ⁰ C for 3 min. The reaction of 1 μmol glucose gives rise to 0.5 μmol formazan, whose molar extinction coefficient is ε = 40,200 IvT ¹ Cm ^-1 , so that the enzyme concentration in the sample can be determined as follows: 1 U / mL sPQQGDH corresponds to a change in OD ₅₇₀ at 1, 493 min ^"1 . A dilution series of the individual samples is necessary, because measurements in which values smaller than 0.05 U / mL or greater than 0.7 U / mL are determined must be discarded, since outside this range no sufficient linearity of the measurement is guaranteed. The concentration of the starting sample was then calculated taking into account the dilution step used. The results are shown in the following Table 2. Table 2:

Example 8 - Determination of the insertion sequence

In order to determine the amino acids inserted between the positions Q 168 and L 169 in an interesting variant, plasmid DNA was prepared from the respective strain using the QiaPrep Plasmid Isolation Kit (Qiagen, Hilden) according to the manufacturer's instructions and the DNA Sequence at the insertion site from both directions with the primers used oligonucleotides 4472-US1 (5 '-CACCGTTAAAGCTTGGATTATC ^', SEQ ID 34) and 4831 -DSl (5'-CCTTCATCGAAAGACCATCAG ^ 'SEQ ID 35) determined. The 3 'end of the sequence of 4472-US 1 is located 58 bp in front of the insertion site, the 3' end of the sequence of 4831 -DS1 is 82 bp behind the insertion site. The results of the sequence analysis are listed in the following Table 3. Table 3:

For orientation, the flanking amino acids 168Q and 172L (previously 169L) and their codons are indicated.

Example 9 - Determination of interference by sugars other than glucose

While the determination of the specific activity of sPQQGDH usually uses a substrate concentration of 30 mM in the assay, a lower concentration of the sugars in question is useful for determining interference; on the order of magnitude that occurs in potential later applications. The physiological concentration of glucose in the blood is about 100 mg / dL, which corresponds to about 5.6 mM. For this reason, the experiments described below were carried out with a glucose concentration of 100 mg / dL and 50 to 250 mg / dL of the interfering sugar. In some cases, however, other concentrations and ratios of the two sugars to each other were also examined.

FIG. 5 shows the result of an interference measurement of five different variants. 5381-06-C12, 5381-05-C4 and 5381-07-F10 are from a library of four additional amino acids, 5152-20-A7 and 5386-13-F10, from a library of five additional amino acids between positions Q168 and L169 , In all cases, only a small influence of the tested sugars is detectable: in the case of maltose, even with 250 mg / dL maltose (1.3 times the molar amount compared to glucose), the measurement in none of the clones by more than 8% affected. With galactose it is max. 16% and with xylose max. 14% interference.

Example 10 - Comparison of the Interferences of Various Sugars on KOZ65 and Clone 5152-20-A7

In order to make a direct comparison between the sPQQGDH without KOZ65 insertion mutation, purified variant sPQQGDH from clone 5152-20-A7 and purified enzyme from KOZ65 were assayed for mixtures of glucose and the sugars maltose, galactose, xylose, lactose, cellobiose or mannose. For each measurement point, duplicates were recorded. Figure 6 shows that the strong interference of maltose, cellobiose and lactose in the KOZ65 starting enzyme is suppressed in the enzyme expressed by variant 5152-20-A7 is. Especially in the case of maltose, the interference at approximately equimolar amounts (100 mg / dL glucose with 200 mg / dL maltose) has fallen from 63% for KOZ65 to 6% for 5152-20-A7, for lactose the interference is no longer significant.

Example 11 - Comparison of Variants from Different Libraries To demonstrate the generality of the method of the invention for producing sPQQGDH variants with improved substrate specificity, libraries of three, four and five additional amino acids were inserted between positions Q168 and L169 of wild-type sPQQGDH from KOZ65 and examines as described in the preceding examples. For comparison, a library with only two additional amino acids between positions Q 168 and L 169 was created and examined. However, only ten clones with weak substrate specificity were found (ratio of activities to maltose and glucose between 25 and 75%). Only one clone (5156-13-G12, see Table 4) had increased substrate specificity; but with a lack of spec. Activity on glucose. However, as shown in Table 4 below, all libraries with three to five additional amino acids contain clones that have a significantly higher substrate specificity than the wild-type.

Table 4:

(n.d .: not performed)

Figure 1 shows the preparation of a vector for insertion of synthetic DNA fragments. Shown are the positions and orientations (thick black arrows in Ib) of the primers for producing two fragments (PCR 1, PCR 2) from the sPQQGDH gene from KOZ65 (Ia) and the site at which later additional amino acids are to be inserted (" IP "in Ia and I b) Fig. Ic shows schematically the assembly of the vector pAI3B-KOZ65-U (1-2) from the amplified amplicon cleaved with XbaI and EcoO109I and purified from the anterior part of the coding Sequence, the EcoO109I and BfuAI cut and purified amplicon from the middle part of the coding sequence and the Xbal and BfuAI from the starting vector pAI3B-KOZ65 (1-1) cut and purified, 3567 bp vector fragment. The final vector pA13B-KOZ65-U (1-2) is shown in FIG. It can be seen from Fig. Id that translation of sPQQGDH into pAI3B-KOZ65-U stops four amino acid residues after the EcoO109I site when no insertion occurs.

Figure 2 shows the preparation of a synthetic, double-stranded piece of DNA from the oligonucleotides described in the text (shown in the figure as arrows, the arrowhead indicates the 3 'end of the sequence, respectively). The lower part shows the oligonucleotides EarV2-Random and EarV2-Lower. In the first step, the 62 bp primary product is created from these overlapping single strands. In the second step, the oligonucleotides EarV2-UpAmp and EarV2-LoAmp are used to prepare the 97 bp PCR product whose DNA sequence is shown in the upper part. The encoded protein sequence is found at the bottom of Fig. 2. In the DNA sequence, the location of the two Earl interfaces is shown. The 64 bp fragment released with Earl from the PCR product is ligated into the EcoO109I opened vector.

FIG. 3 shows the summary of the data from the screening of a library of variant sPQQGDH clones on the example of two microtiter plates (a and b). On the abscissa (X-axis), for each colony of an MTP, the approximate velocity is given in arbitrary relative units, with which glucose is transferred from this colony. On the ordinate (Y-axis), the quotient of maltose and glucose conversion is plotted. This makes it possible to assess for each colony whether it is able to convert glucose and, if so, whether it allows a better distinction between glucose and maltose than the wild-type KOZ65. As a symbol of each colony their position on the respective microtiter plate was used, Al - H 12. The positive control KOZ65 is marked with arrows.

Figure 4 shows the results of the colonies selected from the screening described in Example 3. Three parameters are considered: the conversion rate to glucose, which is not shown on this picture, should be as high as possible; the interference (here for maltose as interfering sugar) plotted on the abscissa (X axis) should be as close as possible to zero, as is the ratio of maltose to glucose conversion plotted on the ordinate (Y axis). For comparison, some clones in Figs. 3 and 4 are marked in the same way.

Figure 5 shows interference studies of various clones with maltose (a), galactose (b) and xylose (c). Shown is the behavior for five purified, variant sPQQGDH preparations Mixtures of glucose and three other sugars. Plotted on the ordinate (Y axis) are the determined interference for mixtures of 100 mg / dL glucose in the measurement batch without further sugar and with 50, 100, 150, 200 or 250 mg / dL (X-axis) of the respective interfering sugar. Each measurement point was determined as a quadruplicate.

Figure 6 shows interference studies for KOZ65 (left) and clone 5152-20-A7 (right) on different sugars. For cellobiose (a), galactose (b), lactose (c), maltose (d), mannose (e) and xylose (f) is plotted on the ordinate (y-axis) how strong these sugars are at different concentrations of activity the sPQQGDH from KOZ65 or 5152-20-A7 in the presence of 100 mg / dL glucose (inhibitor concentration [mg / dL] on the X-axis). Each measurement series for an interfering sugar also includes a measurement without this sugar, which, like all the other measurements in the series, was based on a glucose value averaged from all measurements without interfering sugars. Due to the inaccuracy of measurement in enzymatic tests, therefore, the individual graphs do not necessarily start from zero, which theoretically should be the case.

Claims

claims

1. Soluble pyrroloquinoline quinone-dependent glucose dehydrogenase (sPQQGDH), characterized in that it is compared with a sPQQGDH one of the wild-type strains from the genus Acinetobacter sp. for example, Acinetobacter calcoaceticus LMD 79.41 has an insertion of three to five amino acids in the region of the substrate binding region.

2. sPQQGDH according to claim 1, characterized in that the insertion point has a distance from the substrate binding region of not more than 5 Ängstrom.

3. sPQQGDH according to claim 1 or 2, characterized in that the insertion site between the positions Q168 and L169 based on the sequence of the wild type Acinetobacter calcoaceticus LMD 79.41.

4. sPQQGDH according to one of claims 1 to 3, characterized in that it results by mutagenesis from a strain of the series PT 16, KOZ62, KOZ65, PTN69, KG 106, PTN26, PT15, KGN80, KG140, KGN34, KGN25, KGN100.

5. sPQQGDH according to one of claims 1 to 4 with an insertion of 3 amino acids, wherein the inserted amino acids are an A or G at position 1, a Y, F or W at position 2 and a Q, L, V or I at position 3 exhibit.

6. sPQQGDH according to one of claims 1 to 4 with an insertion of 4 amino acids, wherein the inserted amino acids an A or G at position 1, a G, D or L at position 2, an R or A at position 3 and an M or V at position 4.

7. sPQQGDH any one of claims 1 to 4 with an insertion of 5 amino acids, wherein the inserted amino acids an A, M, S or V at position 1, an E, G or S at position 2, an H, K, R, S or T at position 3, an F, H, I, L, N, V or Y at position 4 and an F, H, L, N, Q, T or Y at position 5.

8. sPQQGDH any one of claims 1 to 4 with an insertion sequence from the series SEQ ID 1 to SEQ ID 18.

9. gene encoding one of the proteins according to claims 1 to 8.

10. Vector, characterized in that it has a singular restriction site in the region of the insertion site, so that oligonucleotide sequences can be inserted at this position, which encode a sPQQGDH according to one of claims 1 to 6.

1 1. Vector according to claim 10 with the sequence SEQ ID 27.

12. A method for producing and identifying sPQQGDH according to one of claims 1 to 8, characterized in that in a first step, a vector is produced with a singular in the region of the insertion restriction restriction site, inserted in a second step, a plurality of different oligonucleotide sequences in the vector in a third step, the recombinant vectors so obtained are transformed into host bacteria and the bacteria are multiplied, in a fourth step the sPQQGDH produced by the bacteria are screened such that the activity of sPQQGDH is at least two different sugars individually as well the interference of at least one sugar with at least one second is determined, and in a fifth step those sPQQGDH are selected which have an improved substrate specificity compared to the unmodified sPQQGDH of the wild strain.

13. A glucose sensor comprising a sPQQGDH of claims 1 to 8.