US 20030104595 A1
The present invention relates to improved variants of soluble pyrroloquinoline quinone (PQQ)-dependent glucose dehydrogenases (s-GDH), to genes encoding mutated s-GDH, to mutant proteins of s-GDH with improved substrate specificity for glucose, and to different applications of these s-GDH variants, particularly for determining concentrations of sugar, especially of glucose in a sample.
1. A mutant of the soluble form of EC 220.127.116.11 also known as PQQ-dependent soluble glucose dehydrogenase (s-GDH) said mutant characterized in that it has an at least two-fold increased substrate specificity for glucose, as compared to at least one other selected sugar substrate.
2. The mutant according to
3. The mutant according to
4. The mutant of PQQ-dependent s-GDH according to
5. The mutant of PQQ-dependent s-GDH according to
6. A mutant of the soluble form of EC 18.104.22.168 also known as PQQ-dependent soluble glucose dehydrogenase (s-GDH) said mutant characterized in that
a) the substrate specific reactivity towards glucose is essentially comparable to that of the wild-type enzyme, and
b) the substrate specific reactivity towards maltose is 30% or less as compared to the wild-type enzyme.
7. The mutant according to
8. The mutant of a PQQ-dependent s-GDH according to
9. A mutant protein of PQQ-dependent s-GDH according to
10. The mutant protein of
11. The mutant of
12. The mutant of
13. A mutant protein of PQQ-dependent s-GDH according to
14. The mutant protein of
15. The mutant protein of
16. A mutant protein of PQQ-dependent s-GDH comprising the amino acid sequence of WPXaaVAPS (SEQ ID NO: 1), wherein said Xaa residue is an amino acid residue other than threonine.
17. The mutant protein of
18. A mutant protein of PQQ-dependent s-GDH comprising the amino acid sequence of TAGXaaVQK (SEQ ID NO: 2), wherein said Xaa residue is an amino acid residue other than asparagine.
19. The mutant protein of
20. A mutant protein of PQQ-dependent s-GDH comprising the amino acid sequence of ADGXaaNGL (SEQ ID NO: 3), wherein said Xaa residue is an amino acid residue other than glutamine.
21. The mutant of
22. An isolated polynudeotide encoding the s-GDH mutant protein according to any of
23. An expression vector comprising an isolated polynucleotide as defined in
24. A host cell comprising the expression vector of
25. A process for producing s-GDH variants comprising culturing the host cell of
26. An expression vector comprising an isolated polynucleotide as defined in
27. A process for producing s-GDH variants with the construct of
28. An improved method of detecting, determining or measuring glucose in a sample using a s-GDH mutant according to any of the proceeding claims, said improvement comprising a more specific detection of glucose.
29. The method of
30. A device for the detection or measurement of glucose in a sample comprising a s-GDH mutant according to any of claims 1-29 and other reagents required for said measurement.
 This is a continatuation-in-part of U.S. application Ser. No. 09/710,197 filed Nov. 9, 2000, the complete disclosure of which is incorporated herein by reference.
 The present invention relates to improved variants of soluble pyrroloquinoline quinone (PQQ)-dependent glucose dehydrogenases (s-GDH), to genes encoding mutated s-GDH, to mutant proteins of s-GDH with improved substrate specificity for glucose, and to different applications of these s-GDH variants, particularly for determining concentrations of sugar, especially of glucose in a sample.
 Two types of PQQ-dependent glucose dehydrogenase (EC 22.214.171.124) have been characterized: One is membrane-bound (m-GDH), the other is soluble (s-GDH). Both types do not share any significant sequence homology (Cleton-Jansen et al., (1989); Oubrie, et al. (1999b)). They are also different regarding both their kinetic as well as their immunological properties (Matsushita et al., 1995).
 Quinoproteins use quinone as cofactor to oxidize alcohols, amines and aldoses to their corresponding lactones, aldehydes and aldolic acids (Duine 1991a,b; Davidson 1993; Anthony 1996; Anthony and Ghosh 1997; Anthony 1998; Goodwin and Anthony 1998). Among quinoproteins, those containing the noncovalently bound cofactor 2,7,9-tricarboxy-1H-pyrrolo [2,3-f]quinoline-4,5-dione (PQQ) constitute the largest sub-group (Duine 1991). All bacterial glucose dehydrogenases known so far belong to this sub-group with PQQ as the prosthetic group (Anthony and Ghosh 1997, Goodwin and Anthony 1998).
 In bacteria, there are two completely different types of PQQ-dependent glucose dehydrogenases (EC126.96.36.199): the soluble type (s-GDH) and the membrane-bound type (m-GDH) (Duine et al., 1982; Matsushita et al., 1989a,b). The m-GDHs are widespread in Gram-negative bacteria, s-GDHs, however, have been found only in the periplasmatic space of Acinetobacter strains, like A. calcoaceticus (Duine, 1991a; Cleton-Jansen et al., 1988; Matsushita and Adachi, 1993). Through searching sequence databases, two sequences homologous to the full-length A. calcoaceticus s-GDH have been identified in E.coli K-12 and Synechocystis sp. Additionally, two incomplete sequences homologous to A. calcoaceticus s-GDH were also found in the genome of P.aeruginosa and Bordetella pertussis (Oubrie et al. 1999a), respectively. The deduced amino acid sequences of these four uncharacterized proteins are closely related to A. calcoaceticus s-GDH with many residues in the putative active site absolutely conserved. These homologous proteins are likely to have a similar structure and to catalyze similar PQQ-dependent reactions (Oubrie et al., 1999a).
 Bacterial s-GDHs and m-GDHs have been found to possess quite different sequences and different substrate specificity. For example, A. calcoaceticus contains two different PQQ-dependent glucose dehydrogenases, one m-GDH which is active in vivo, and the other designated s-GDH for which only in vitro activity can be shown. Cleton-Jansen et al., (1988;1989a,b) cloned the genes coding for the two GDH enzymes and determined the DNA sequences of both these GDH genes. There is no obvious homology between m-GDH and s-GDH corroborating the fact that m-GDH and s-GDH represent two completely different molecules.
 The gene of s-GDH from A. calcoaceticus has been cloned in E. coli behind a leader sequence and a strong promoter. After being synthesized in the cell, the s-GDH is translocated through the cytoplamic membrane into the periplasmic space (Duine 1991a, Matsushita and Adachi 1993). Like the native s-GDH from A. calcoaceticus, s-GDH expressed in E.coli is also a homodimer, with one PQQ molecule and three calcium ions per monomer (Dokter et al., 1986a,b; 1988; Olsthoorn and Duine, 1996; Oubrie et al., 1999a,b,c). s-GDH oxidizes a wide range of mono- and disaccharides to the corresponding ketones which further hydrolyze to the aldonic acids, and it is also able to donate electrons to PMS (Phenazine metosulfate), DCPIP (2,6-Dichlorophenolindophenol), WB (Wurster's blue) and short-chain ubiquinones such as ubiquinone Q1 and ubiquinone Q2 (Matsushita et al., 1989a,b), several artificial electron acceptors such as N-methylphenazonium methyl sulfate (Olsthoorn and Duine 1996; 1998) and electroconducting polymers (Ye et al., 1993).
 In view of s-GDH's high specific activity towards glucose (Olsthoorn and Duine 1996) and its broad artificial electron acceptor specificity, the enzyme is well suited for analytical applications, particularly for being used in (bio-)sensor or test strips for glucose determination in diagnostic applications (Kaufmann et al., 1997). Glucose oxidation can be catalyzed by at least three quite distinct groups of enzymes, i.e., by NAD-dependent, dye-linked glucose dehydrogenases, by flavoprotein glucose oxidase or by quinoprotein GDHs (Duine 1995). A rather slow autooxidation of reduced s-GDH has been observed, demonstrating that oxygen is a very poor electron acceptor for s-GDH (Olsthoorn and Duine 1996). s-GDH can efficiently donate electrons to PMS, DCPIP, WB and short-chain ubiquinones such as Q1 and Q2, but it can not efficiently donate electrons directly to oxygen.
 Traditional test strips and sensors for monitoring glucose level in blood, serum and urine. e.g. from diabetic patients use glucose oxidase. However, since glucose oxidase transfers its electrons to oxygen, it is known that oxygen may have a negative impact on glucose measurements which are based on this enzyme. The major advantage of PQQ-dependent glucose dehydrogenases is their independence from oxygen. This important feature is e.g., discussed in U.S. Pat. No. 6,103,509, in which some features of membrane-bound GDH have been investigated.
 An important contribution to the field has been the use of s-GDH together with appropriate substrates. Assay methods and test strip devices based on s-GDH are disclosed in detail in U.S. Pat. No. 5,484,708. This patent also contains detailed information on the set-up of assays and the production of s-GDH-based test strips for measurement of glucose. The methods described there as well in the cited documents is herewith included by reference.
 Other patents or applications relating to the field and comprising specific information on various modes of applications for enzymes with glucose dehydrogenase activity are U.S. Pat. No. 5,997,817; U.S. Pat. No. 6,057,120; EP 620 283; and JP 11-243949-A.
 A commercial system which utilizes s-GDH and an indicator that produces a color change when the reaction occurs (Kaufmann et al. 1997) is the Glucotrend® system distributed by Roche Diagnostics GmbH.
 Despite the above discussed important advantages there also is a major inherent problem of s-GDH. S-GDH has rather a broad substrate spectrum as compared to m-GDH. That is, s-GDH oxidizes not only glucose but also several other sugars including maltose, galactose, lactose, mannose, xylose and ribose (Dokter et al. 1986a). The reactivity towards sugars other than glucose may in certain cases impair the accuracy of determining blood glucose levels, in some diabetic patients. In particular patients on peritoneal dialysis treated with icodextrin (a glucose polymer) may contain in their body fluids, e.g., in blood, high levels of other sugars, specially maltose (M. Dratwa, et al., Perit. Dial. Int. 1998 Nov-Dec;18(6):603-9).
 Therefore clinical samples, as e.g. obtained from diabetic patients, especially from patients with renal complications and especially from patients under dialysis may contain significant levels of other sugars, especially maltose. Glucose determinations in samples obtained from such critical patients may be impaired by maltose.
 There are scarce reports in the literature on attempts to produce modified PQQ-dependent s-GDHs which exhibit altered substrate specificity. Due to a negative outcome most of these efforts have not been published. Igarashi et al., (1999) report that introducing a point mutation at position Glu277 leads to mutants with altered substrate specificity profile. However, none of these mutants, lead to an at least two-fold increased improved specificity for glucose as e.g., compared to xylose, galactose or maltose.
 It can be summarized that the attempts known in the art aiming at improvements of properties of s-GDH, especially its specificity towards glucose, have not been successful to the extend required for accurate monitoring of glucose levels in patients having high levels of sugars other than glucose.
 A great demand and clinical need therefore exists for mutant forms of s-GDH which feature an improved specificity for glucose as substrate.
 It was the task of the present invention to provide new mutants or variants of s-GDH with significantly improved substrate specificity for glucose as compared to other selected sugar molecules, e.g., like galactose or maltose.
 Surprisingly it has been found that it is possible to significantly improve the substrate specificity specific activity of s-GDH for glucose, as compared to other sugars, and to at least partially overcome the above described problems known in the art.
 The substrate specificity for glucose as compared to other selected sugar molecules has been significantly improved by providing mutant s-GHD according to the present invention as described herein below and in the appending claims. Due to the improved substrate specificity of the new forms of s-GDH significant technical progress for glucose determinations in various fields of applications is possible.
 It has now surprisingly been found that it is possible to provide s-GDH mutants with improved substrate specificity towards glucose as a substrate as compared to other selected sugars. New s-GDH variants are disclosed which exhibit a significant higher substrate specificity for glucose, especially as compared to maltose.
 Also disclosed are mutant s-GDH molecules, which, compared to the wild-type enzyme, exhibit essentially the same specific activity for glucose as substrate but markedly reduced activity for other selected sugar molecules.
 Such comparison of specific activities for various substrate molecules is based on and calculated in relation to the original enzymatic activities of a wild-type enzyme, e.g., as isolated from Acinetobacter calcoaceticus.
 Mutated s-GDH proteins as well as polynucleotide-sequences coding for such proteins exhibiting improved properties, especially increased specificity for glucose are also provided.
 s-GDH mutants comprising at least one amino acid substitutions at a position selected from the group consisting of positions 348 and 428 (numbered according to the mature s-GDH protein of A. calcoaceticus) have been found to provide for s-GDH enzymes with improved properties, especially with improved specificity for glucose.
 The improved s-GDH mutants can be used with great advantage for the specific detection or measurement of glucose in biological samples, especially in tests strip devices or in biosensors.
 The following examples, references, sequence listing and figures are provided to aid the understanding of the present invention, the true scope of which is set forth in the appended claims. It is understood that modifications can be made in the procedures set forth without departing from the spirit of the invention.
FIG. 1: Nucleotide (DNA) sequence of the Acinetobacter calcoaceticus PQQ dependent soluble glucose dehydrogenase gene and the corresponding amino acid sequence.
FIG. 2: Protein sequences of A. calcoaceticus PQQ-dependent s-GDH and A. baumanni s-GDH aligned according to sequence homology.
FIG. 3: Illustration of pACSGDH vector referred to in Example 1 containing the wild-type or mutated DNA sequences of soluble PQQ-dependent glucose dehydrogenase.
FIG. 4: Nucleotide (DNA) sequence of the pACSGDH vector referred to in Example 1 containing the wild-type DNA sequence of soluble PQQ-dependent glucose dehydrogenase.
 In a first embodiment a mutant of the soluble form of EC188.8.131.52 also known as PQQ-dependent soluble glucose dehydrogenase (s-GDH), said mutant characterized in that it has an at least two-fold improved substrate specificity for glucose as compared to at least one other selected sugar substrate is described.
 As discussed above, two completely different enzyme families with glucose dehydrogenase activity, grouped together under EC184.108.40.206, are characterized to date. These two enzyme families, however, appear not be related to each other.
 For the purpose of this invention only the soluble form of GDH is relevant and discussed herein below.
 It is known in the art that the wild-type DNA-sequence of such soluble PQQ-dependent glucose dehydrogenase can be isolated from strains of Acinetobacter. Most preferred is the isolation from Acinetobacter calcoaceticus-type strand LMD 79.41. The sequence of this wild-type s-GDH (the mature protein) is given in FIG.: 1. Other LMD strains of Acinetobacter may also be used as source of wild-type s-GDH. Such sequences can be aligned to the sequence obtained from A. calcoaceticus and sequence comparisons be made (see FIG.: 2). It also appears feasible to screen DNA-libraries of other bacterial strains, as for example described for E.coli K-12 (Oubrie et al., 1999a) above and to identify sequences related to s-GDH in such genomes. Such sequences might also be used to generate s-GDH mutants with improved substrate specificity for glucose.
 The term “mutant” or “variant” in the sense of the present invention relates to a s-GDH protein which compared to a corresponding wild-type sequence exhibits at least one amino acid substitution. The expert in the field will appreciate that there are various possibilities to produce polynucleotides encoding for such a polypeptide sequence of mutated s-GDH. It is of course also possible to generate mutants comprising one or more additions or deletions of amino acids.
 A mutant according to the present invention is characterized in that it has at least a two-fold improved substrate specificity for glucose as compared to at least one other selected sugar substrate.
 In order to calculate the substrate specificity or cross-reactivity one easy way is to set the activity measured with glucose as substrate to 100% and to compare the activity measured with the other selected sugar to the glucose value. Sometimes, in order not to be redundant, simply the term specificity is used without making special reference to a selected other sugar substrate.
 The expert in the field will appreciate that comparison of (re-)activities is best made at equimolar concentrations of the substrate molecules investigated using well-defined assay conditions. Otherwise corrections for differences in concentrations have to be made.
 Standardized and well-defined assay conditions have to be chosen in order to assess (improvements in) specificity. The enzymatic activity of s-GDH for glucose as substrate as well as for other selected sugar substrates is measured as described in Example 6.
 Based on these measurements cross-reactivity and (improvements in) specificity are assessed.
 The s-GDH (cross-)reactivity for a selected sugar in percent is calculated as Cross-reactivity=(activity selected sugar/activity glucose)×100%. (Cross-)reactivity for maltose of wild-type s-GDH according to the above formula has been determined as about 105%. Wild-type s-GDH (cross-)reactivity for galactose has been measured as about 50% (cf. Table 1).
 (Improved) specificity is calculated according to the following formula:
 As compared to the wild-type enzyme, a s-GDH form with an at least two-fold improvement in specificity for glucose versus maltose (maltose/glucose) accordingly with maltose as substrate has at most 52,5% of the activity as measured with glucose as substrate. Or, if, for example a mutant s-GDH has a cross-reactivity for maltose of 20% (determined and calculated as described above), this mutant as compared to the wild-type s-GDH therefore has a 5.25 fold improved substrate specificity (maltose/glucose).
 The term “specific activity” for a substrate is well known in the art, it is preferably used to describe the enzymatic activity per amount of protein. Various methods are known to the art to determine specific activity of GDH molecules, using glucose or other sugars as substrates (K. Sode et al., Biochemical and Biophysical Research Communications (1999) 264, 820-824). One of the methods available for such measurement is described in detail in the examples section.
 Whereas it is possible, to select many different sugar molecules and to investigate the specificity of s-GDH in comparison to any such selected sugar molecule, it is preferred to select a clinical relevant sugar molecule for such a comparison. Preferred selected sugars are selected from the group comprising mannose, allose, galactose, xylose, and maltose. Preferably, maltose or galactose are selected and mutant s-GDH is tested for improved substrate specificity compared to galactose or maltose. In a further preferred embodiment the selected sugar is maltose.
 It has surprisingly been found that the improvements in specificity of mutated s-GDH, e.g., for maltose vs. glucose, are quite considerable. It is therefore preferred that said substrate specificity for glucose as compared to the substrate specificity for at least one of the selected other sugar substrates is improved at least three-fold. Other preferred embodiments comprise s-GDH mutants characterized by an improved substrate specificity for glucose, which is at least 5 times or also preferred at least 10 times higher, as compared to the other sugar molecule(s) selected.
 Mutations in this enzyme lead in many cases to mutants with reduced specific activity for the substrate glucose. Such decrease in (absolute or overall) specific activity for the substrate glucose, however, may be critical for routine applications. Surprisingly it has been found that improved specificity for glucose must not go to the expense of reduced overall specific activity. It is therefore preferred that the s-GDH with improved specificity towards the substrate glucose exhibits at least 10% of the glucose specific activity as measured for the wild-type enzyme. It is of course more preferred that such mutated enzymes exhibit at least 20 or more preferred at least 50% of the respective glucose activity of wild-type s-GDH.
 In a further preferred embodiment, the invention relates to a mutant of the soluble form of EC220.127.116.11 also known as PQQ-dependent soluble glucose dehydrogenase (s-GDH), said mutant characterized in that
 a) the substrate specific reactivity towards glucose is essentially comparable to that of the wild-type enzyme and
 b) the substrate specific reactivity towards maltose is 30% or less as compared to the wild-type enzyme.
 The specific (re)activity towards glucose is considered to be essentially comparable to that of the wild-type enzyme if at least 50% of the original enzymatic activity for glucose of the wild-type enzyme is maintained. Also preferred are mutants exhibiting at least 80% or more preferred, at least 90% of the specific activity for glucose as measured for the wild-type enzyme.
 Quite surprisingly it has been found, that it is possible to obtain such s-GDH mutants or variants, which exhibit essentially the same enzymatic activity for glucose as the wild-type s-GDH but, nonetheless, significantly reduced substrate specific reactivity towards other selected sugars, especially towards maltose. Mutants characterized in that the substrate specific reactivity towards glucose is essentially comparable to that of the wild-type enzyme and in that the substrate specific reactivity towards maltose is 20% or less as compared to the wild-type enzyme are preferred. Further preferred are such mutants with which the maltose specific activity is 15% or even only 10% or less of the maltose specific activity as measured for the corresponding wild-type enzyme.
 Unexpectedly it has been found that it is possible to generate s-GDH mutants with improved substitute specificity and even more unexpected it has been found that it is only a few well-defined amino acid positions which are of major relevance in that respect.
 Mutants comprising only a single substitution of amino acid 253 did not lead to an at least two-fold improvement of substrate specificity for glucose, e.g., as compared to galactose or maltose.
 It has now surprisingly been found, that substitutions at amino acid positions 348 and 428 are important to generate s-GDH mutants or variants with significantly improved specificity for glucose.
 Residues 348 or 428 are not known to contribute to the substrate binding of s-GDH as identified and known in the art (BW. Dijkstra et al.,EMBO J 1999 Oct 1;18(19):5187-94; Oubrie, A., Dijkstra, B. W., 2000). No chemical or physical explanation is at hand, why substitutions of these amino acid residues alters the relative specific activity of s-GDH for glucose as compared to other sugar molecules of interest.
 As demonstrated in table 1, it has been found that a variety of s-GDH variants with improved specificity for glucose can be identified and generated, as long as at least one of the amino acids in positions threonine 348 or glutamine 428—as corresponding to wild-type s-GDH sequence positions from A. calcoaceticus—is substituted with appropriate other amino acids.
 It has been found in addition that substitution of amino acid 76 also has a positive effect on glucose specificity of s-GDH.
 In a preferred embodiment the mutant protein of a PQQ-dependent s-GDH according to the present invention comprises at least one amino acid residue substitution at a position selected from the group consisting of positions 348, and 428 of the corresponding wild-type sequence of A. calcoaceticus.
 s-GDH variants comprising a substitution of the amino acid threonine in position 348 are most preferred.
 In a further preferred embodiment the mutated s-GDH is characterized in that the amino acid residue threonine at position 348 is substituted with an amino acid residue selected from the group consisting of alanine, glycine, and serine. In a most preferred embodiment glycine is used to substitute for threonine at position 348. One group of preferred s-GDH variants according to this invention comprises a substitution of the amino acid residue at position 348 and at at least one of the following positions 76, 143, 168, 169 and 428. In yet a further embodiment a mutant protein of PQQ-dependent s-GDH comprising an amino acid residue substitution at position 428 of the correspondent wild-type sequence from Acinetobacter calcoaceticus is disclosed, in which the asparagine residue of the wild-type sequence is replaced by other appropriate amino acid residues. Preferably, such amino acid residue is selected from the group consisting of leucine, proline and valine. It is preferred to substitute the asparagine at position 428 with proline .
 It further has proved that the amino acid glutamine at position 76 can be substituted to improve on the problems imposed by s-GDH cross-reactivity with other sugar molecules. In another preferred embodiment the mutant according to the present invention therefore comprises a substitution of glutamine at position 76 of the corresponding wild-type sequence from Acinetobacter calcoaceticus.
 It is preferred to select the amino acid used in such substitution from the group consisting of alanine, aspartic acid, glutamic acid, glycine, methionine, proline and serine.
 As described above, the substitution of at least one amino acid in positions 348, and/or 428 of the s-GDH sequence corresponding to the wild-type sequence isolated from A. calcoaceticus, can be used to significantly improve the glucose specificity of s-GDH. Further improved mutants are obtained by providing a mutant s-GDH protein comprising at least two amino acid substitutions, at least one of these being corresponding to amino acid position 76, 348, and/or 428. A further embodiment of the present invention therefore is a mutant protein of PQQ-dependent s-GDH with improved glucose specificity, comprising at least two amino acid residue substitutions, that substituted amino acid positions being selected from the group consisting of position 16, 22, 76, 116, 120, 127, 143, 168, 169, 171, 177, 227, 231, 255, 277, 295, 299, 308, 317, 348, 355, 422,428 and 438. of the corresponding mature A. calcoaceticus soluble PQQ-dependent s-GDH, characterized in that at least one of the amino acid residues, T348 or N428 is replaced.
 It is further preferred, that the at least two amino acid residues comprise a substitution in position 348 and at least one additional substitution selected from the group of positions comprising positions 76, 143, 168, 169, and 428.
 In a further preferred embodiment the at least two amino acid positions which are substituted in a mutant s-GDH are selected from the group consisting of amino acid position 76, 348, and/or 428.
 Mutants comprising substitutions at amino acid residues corresponding to positions 348 and 428 have been found as very advantageous for improving the specificity of s-GDH for glucose in comparison to other sugar substrates. It is especially preferred to design and select mutants of s-GDH, which comprise a substitution at both the positions 348 and 428. Most preferred are mutants comprising the preferred substitutions at both these positions as described above.
 The variant comprising T348G and N428P is most preferred. This terminology T348G and N428P is known to the art to indicate that threonine at position 348 is replaced by glycine and glutamine and position 428 is replaced by proline.
 Mutant s-GDH proteins comprising in addition to substitutions at positions 348 and 428 also substitutions at positions 76, 127 and 143, also represent preferred embodiments of the present invention.
 Further preferred examples of mutated s-GDH proteins according to the present invention comprise amino acid substitutions at positions 76 and 348 and also such mutants comprising substitutions at positions 76 and 428. In yet another preferred embodiment the mutated s-GDH protein according to the present invention comprises substitutions of the amino acid residues at positions 76, 348 and 428.
 Amino acid sequence analysis revealed that the sequence motives found in wild-type s-GDH from A. calcoaceticus on the one hand and A. baumannii on the other hand appear to be very conservative around the positions of major relevance to improve the specificity for glucose as identified in the present invention, i.e. positions 76, 348, and 428, as corresponding to wild-type s-GDH from A. calcoaceticus (c.f., FIG. 2).
 A preferred embodiment according to the present invention therefore is a mutant protein of PQQ-dependent s-GDH comprising the amino acid sequence of WPXaaVAPS (SEQ ID NO: 1). SEQ ID NO:1 corresponds to position 346-352 of A. calcoaceticus wild-type s-GDH or position 347-353 of A. baumannii wild-type s-GDH, wherein said Xaa residue is an amino acid residue other than threonine.
 It is further preferred that Xaa in SEQ ID NO: 1 represents glycine.
 A mutant of PQQ-dependent s-GDH, comprising the amino acid sequence of TAGXaaVQK (SEQ ID NO: 2) is another preferred embodiment of the invention. SEQ ID NO:2 corresponds to position 425-431 of A. calcoaceticus wild-type s-GDH or to position 426-432 of A. baumannii wild-type s-GDH, wherein said Xaa residue is an amino acid residue other than asparagine,.
 Most preferably the s-GDH mutant comprising SEQ ID NO:2 is characterized in that said Xaa residue is proline residue.
 Numerous possibilities are known in the art to produce mutant proteins. Based on the important findings of the present invention disclosing the critical importance of amino acid positions 348 and 428 and also the utility of position 76, the skilled artisan now can easily produce further appropriate variants of s-GDH. Such variants for example can be obtained by the methods known as random mutagenesis (Leung, D.W., et al., 1989) and/or directed mutagenesis (Struhl, K., et al. 1987). An alternative method to produce a protein with the desired properties is to provide chimaeric constructs, which contain sequence elements from at least two different sources. Such procedures known in the art may be used in combination with the information disclosed in the present invention to provide mutants or variants of s-GDH comprising at least one amino acid at sequence positions 348, and/or 428 of s-GDH.
 According to procedures known in the state of the art or according to the procedures given in the examples section, it is possible to obtain polynucleotide sequences coding for any of the s-GDH mutants as discussed above. The invention therefore comprises also isolated polynucleotide sequences encoding s-GDH mutant proteins as described above.
 The present invention further includes an expression vector comprising a nucleic acid sequence according to the present invention operably linked a promoter sequence capable of directing its expression in a host cell.
 The present invention further includes an expression vector comprising a nucleic acid sequence according to the present invention operably linked to a promoter sequence capable of directing its expression in a host cell. Preferred vectors are plasmids such as pACSGDH shown in FIGS. 3 and 4.
 Expression vectors useful in the present invention typically contain an origin of replication, a promoter located upstream of the DNA sequence and are followed by the DNA sequence coding for all or part of s-GDH variants. The DNA sequence coding for all or part of the s-GDH variants is followed by transcription termination sequences and the remaining vector. The expression vectors may also include other DNA sequences known in the art, for example, stability leader sequences which provide for stability of the expression product, secretory leader sequences which provide for secretion of the expression product, sequences which allow expression of the structural gene to be modulated (e.g., by the presence or absence of nutrients or other inducers in the growth medium), marking sequences which are capable of providing phenotypic selection in transformed host cells, and the sequences which provide sites for cleavage by restriction endonucleases.
 The characteristics of the actual expression vector used must be compatible with the host cell, which is to be employed. For example, when cloning in an E. coli cell system, the expression vector should contain promoters isolated from the genome of E.coli cells (e.g., lac, or trp). Suitable origins of replication in E.coli various hosts include, for example, a ColE1 plasmid replication origin. Suitable promoters include, for example, lac and trp. It is also preferred that the expression vector include a sequence coding for a selectable marker. The selectable marker is preferably an antibiotic resistance gene. As selectable markers, ampicillin resistance, or canamycin resistance may be conveniently employed. All of these materials are known in the art and are commercially available.
 Suitable expression vectors containing the desired coding and control sequences may be constructed using standard recombinant DNA techniques known in the art, many of which are described in Sambrook et al. (1989).
 The present invention additionally concerns host cells containing an expression vector which comprises a DNA sequence coding for all or part of the mutant s-GDH. The host cells preferably contain an expression vector that comprises all or part of one of the DNA sequences having one or more mutations shown in Table 1. Further preferred are the host cells containing an expression vector comprising one or more regulatory DNA sequences capable of directing the replication and/or the expression of, and operatively linked to a DNA sequence coding for, all or part of mutant s-GDH. Suitable host cells include, for example, E.coli HB101 (ATCC 33694) available from Pomega (2800 Woods Hollow Road, Madison, Wis., USA), XL1-Blue MRF available from Stratagene (11011 North Torrey Pine Road, La Jolla, Calif., USA) and the like.
 Expression vectors may be introduced into host cells by various methods known in the art. For example, transformation of host cells with expression vectors can be carried out by polyethylene glycol mediated protoplast transformation method (Sambrook et al. 1989). However, other methods for introducing expression vectors into host cells, for example, electroporation, biolistic injection, or protoplast fusion, can also be employed.
 Once an expression vector containing s-GDH variants has been introduced into an appropriate host cell, the host cell may be cultured under conditions permitting expression of the desired s-GDH variants. Host cells containing an expression vector which contains a DNA sequence coding for all or part of the mutant s-GDH are, e.g., identified by one or more of the following general approaches: DNA hybridization, the presence or absence of marker gene functions, assessment of the level of transcription as measured by the production of s-GDH mRNA transcripts in the host cell, and detection of the gene product immunologically. Preferably transformed host cells are identified by enzyme assay, e.g., calorimetric detection.
 The present invention also teaches the generation and screening of s-GDH variants. Random mutagenesis and saturation mutagenesis is performed as known in the art. Variants are analyzed for substrate specificity for glucose, maltose as well as other sugars. The assay conditions chosen are adapted to ensure that the expected small enhancements brought about e.g., by a single amino acid substitution, can be measured. This has been accomplished by adjusting the assay conditions such that the wild type (or parent) enzyme activity is dose to the lower detection limit. One mode of selection or screening of appropriate mutants is given in Example 3. Any change or improvement as compared over the wild-type enzyme this way can be clearly detected.
 It should, of course, be understood that not all expression vectors and DNA regulatory sequences would function equally well to express the DNA sequences of the present invention. Neither will all host cells function equally well with the same expression system. However, one of ordinary skill in the art may make a selection among expression vectors, DNA regulatory sequences, and host cells using the guidance provided herein without undue experimentation and without departing from the scope of the present invention.
 The invention also relates to a process for producing s-GDH variants of the current invention comprising culturing a host cell of the invention under conditions suitable for production of the mutant s-GDH of the invention. For bacterial host cells, typical culture conditions are liquid medium containing the appropriate antibiotic and induction agent. Typical appropriate antibiotics include ampicillin, canamycin, chloroamphenicol, tetracyclin and the like. Typical induction agents include IPTG, glucose, lactose and the like.
 It is preferred that the polypeptides of the present invention are obtained by production in host cells expressing a DNA sequence coding the mutant s-GDH. The polypeptides of the present invention may also be obtained by in vitro translation of the mRNA encoded by a DNA sequence coding for the mutant s-GDH. For example, the DNA sequences may be synthesized as described above and inserted into a suitable expression vector, which in turn may be used in an in vitro transcription/translation system.
 An expression vector comprising an isolated polynucleotide as defined and described above operably linked to a promoter sequence capable of promoting its expression in a cell-free peptide synthesis system represents another preferred embodiment of the present invention.
 The polypeptides produced e.g. by procedures as describe above, may then be isolated and purified using various routine protein purification techniques. For example, chromatographic procedures such as ion exchange chromatography, gel filtration chromatography and affinity chromatography may be employed.
 One of the major applications of the improved s-GDH variants of this invention is for the use in test strips to monitor blood-glucose level in diabetic patients. Due to the insensitivity of PQQ-dependent glucose dehydrogenase towards oxygen, a system using the improved s-GDH variants is less prone to interference by oxygen than systems based on glucose oxidase. More important, since the s-GDH variants have improved specificity towards glucose and significantly decreased relative enzymatic activity towards other sugars, the interference due maltose, galactose, and/or other related sugars which may be present in a sample to be analyzed is significantly reduced. Of course many kinds of samples may be investigated. Bodily fluids like serum, plasma, intestinal fluid or urine are preferred sources for such samples.
 The invention also comprises a method of detecting, determining or measuring glucose in a sample using a s-GDH mutant according to the present invention. It is especially preferred that the improved method for detection of glucose in a sample is characterized in that said detection, determination or measurement of glucose is performed using a sensor or test strip device.
 Also within the scope of the present invention is a device for the detection or measurement of glucose in a sample comprising an s-GDH mutant according to this invention as well as other reagents required for said measurement.
 The s-GDH variants with improved substrate specificity of this invention can also be used to great advantage in biosensors (D'Costa et al., 1986, Laurinavicius et al., 1999a,b) for online monitoring of glucose in a sample or a reactor. For this purpose, the s-GDH variants can, for example, be used to coat an oxygen-insensitive glassy electrode with an osmium complex containing a redox conductive epoxy network (Ye et al., 1993) for more accurate determination of the glucose concentration.
 There are also other possible applications of the s-GDH variants with the improved substrate specificity according to this invention. For example, these s-GDH variants may be used in an aldonic acid production process. Wild-type s-GDH has a high turnover in substrate oxidation producing gluconic and other aldonic acids. By using the s-GDH variants, which are more specific for glucose, the production of gluconic acid would result in much less byproducts. With other s-GDH variants of different substrate specificity, it is possible to produce different aldonic acids as required.
 In the following examples, all reagents, restriction enzymes, and other materials were obtained from Roche Diagnostics Germany, unless other commercial sources are specified, and used according to the instructions given by the suppliers. Operations and methods employed for the purification, characterization and cloning of DNA are well known in the art (Current Protocols in Molecular Biology Volume 1-4, Edited by F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J A. Smith, K.Struhl; Massachusetts General Hospital and Harward Medical School by John Wiley & Sons, Inc.) and can be adapted as required by the skilled artisan.The following examples further illustrate the present invention. These examples are not intended to limit the scope of the present invention, but provide further understanding of the invention.
 Cloning and Expression of the wild-type A. calcoaceticus Soluble PQQ Dependent Glucose Dehydrogenase in E.coli
 The s-GDH gene was isolated from Acinetobacter calcoaceticus strain LMD 79.41 according to standard procedures. The wild-type s-GDH gene was subdoned into a plasmid containing the mgl promoter for adjustabel expression (cf. Patent application WO88/09373). The new construct was called pACSGDH (see FIGS. 3 and 4). The recombinant plasmids was introduced into a host organism selected from the E.coli group. These organisms were then cultivated under appropriate conditions and colonies showing s-GDH activity selected.
 The plasmid pACSGDH was isolated from a 200 ml over-night culture of the clone mentioned above using the QIAGEN Plasmid Maxi Kit (Qiagen) according to the manufacturers protocol. The plasmid was re-suspended in 1 ml bidest. water. The concentration of the plasmid was determined using a Beckman DU 7400 Photometer. The yield was 600 μg. Then the quality of the plasmid was determined by agarose gel electrophoresis.
 Mutagenic PCR
 To generate random mutations in the s-GDH-gene, mutagenic PCR (polymerase chain reaction) was performed. The pACSGDH plasmid and the DNA sequence encoding the mutated enzymes (PCR product from mutagenic PCR) were digested with the restriction enzymes Sph I and Eco RI. The products were gel purified. The digested DNA sequences were ligated and an aliquot of the ligation reaction mixture was used to transform competent E. coli cells. The transformants were subsequently selected on LB plates containing ampicillin.
 To assay, individual colonies were chosen, grown over night in LB medium containing ampicillin and subjected to screening (see Example 3).
 Mutagenic PCR reaction mixture:
 40 ng pACSGDH
 1× buffer without MgCl2 (Roche Diagnostics GmbH, Cat. 1699 105)
 dCTP, dTTP 1 mM
 DATP, dGTP 0.2 mM (Roche Diagnostics GmbH, Cat. 1969 064)
 The mutant colonies on agar plates described above where picked in microtiter plates (mtp) containing 200 μl LB-Ampicillin-media/well and incubated over night at 37° C. These plates are called master plates.
 From each master plate, 5 μl sample/well was transferred to a mtp containing 5 μl B-per/well (Bacterial Protein Extraction Reagent Pierce No.78248) for cell disruption and 240 μl of 0.0556 mM pyrollo-quinoline quinone(PQQ); 50 mM Hepes; 15 mM CaCl2 pH 7.0/well for activation of s-GDH were added. To complete the formation of the holoenzyme, the mtp was incubated at 25° C. for 2 hours and at 10° C. over night. This plate is called working plate.
 From the working plate 2×10 μl sample/hole were transferred to two empty mtps. After that, one was tested with glucose and the other with maltose or other selected sugar molecules as a substrate. All sugar molecules were used in equimolar concentrations.
 The dE/min was calculated and the value using glucose as substrate was set to 100% activity. The value obtained with the other sugar was compared to the glucose value and calculated in percent activity ((dE/min Maltose/dE Glucose)*100). This is equivalent to the cross-reactivity of the (mutant) enzyme.
 Sequencing of Mutant s-GDH Gene from Mutagenic PCR
 The plasmid containing the mutant s-GDH gene that leads to 50% maltose/glucose activity was isolated (High Pure Plasmid Isolation Kit, Roche 1754785) and sequenced using an ABI Prism Dye Terminator Sequencing Kit and ABI 3/73 and 3/77 sequencer (Amersham Pharmacia Biotech).
 Following primers were used:
 at position 340 (mature enzyme) change from E to G
 at position 348 (mature enzyme) change from T to S
 at position 369 (mature enzyme) change from N to H
 at position 413 (mature enzyme) change from S to N
 s-GDH Mutants Obtained by Saturation Mutagenesis
 The QuikChange Site-Directed Mutagenesis Kit (Stratagene, Cat. 200518) was used to substitude successively wild type amino acids at defined positions of the s-GDH-protein or of s-GDH-mutants (plasmide purification as discribed above) with other random amino acids.
 The 5′- and the 3′-primer used for mutagenesis were complementary to each other and contained NNN in a central position. These nucleotides were flanked by 12 to 16 nucleotides at each end. The sequences of the nucleotides were identical to the cDNA-strand or to the complementary cDNA-strand flanking the codon for the amino acid that had to be substituted. Instead of the codon, the primer contained NNN therefore the oligonucleotides code for every codon.
 For every defined position, one PCR reaction was performed.
 The PCR-reactions and the DpnI-restrictionendonuclease digestions were performed according to the manual.
 After that, 1 μl of each reaction was used for the electroporation of XL1F- cells. Cells were grown and the s-GDH-activities of the clones were determined as described above.
 To ensure statistically that all 20 amino acids variants were screened, 200 clones were tested for each position.
 The following primers where used:
 The amino acid changes at positions 340, 369 and 413 didn't change the substrate specificity. Only the wobble at position 348 did yield clones with a substrate specificity from 25-100% (maltose/glucose).
 Numerous rounds of mutagenic PCR and saturation mutagenesis were performed. It was found and confirmed that positions 348 and 428 are of major importance and that exchange of other amino acids may further improve the specificity for glucose of mutated s-GDH. Representative data and positions are given in table 1.
 5 Purification of mutant s-GDH T348G
 The grown cells (LB-Amp. 37° C.) were harvested and resuspended in potassium phosphate buffer pH 7.0. Cell disruption was performed by French Press passage (700-900 bar). After centrifugation the supernatant was applied to a S-Sepharose (Pharmacia) column equilibrated with 10 mM potassium phosphate buffer pH 7.0. After washing, the s-GDH was eluted using a salt gradient 0-1 M NaCl. The fractions showing GDH activity were pooled, dialysed against potassium phosphate buffer pH 7.0 and re-chromatographied on re-equilibrated S-sepharose column. The active fractions were pooled and subjected to a gel filtration using a Superdex® 200 column (Pharmacia). The active fractions were pooled and stored at −20° C.
 Enzyme Assay and Protein Determination of Mutant T348G and Wildtype GDH
 Protein determination was performed using the Protein Assay Reagent no. 23225 from Pierce (calibration curve with BSA, 30 Min. 37° C.).
 The GDH samples were diluted at 1 mg protein/ml with 0.0556 mM pyrollo-quinoline quinone(PQQ); 50 mM Hepes; 15 mM CaCl2 pH 7.0 and incubated at 25° C. for 30 minutes for reconstitution or activation.
 After activation 50 μl of sample were added to 1000 μl of a 0.2 M citrate buffer solution (pH 5.8; at 25° C.) containing 0.315 mg (4-(dimethylphosphinylmethyl)-2-methyl-pyrazolo-[1.5a]-imidazol-3-yl)-(4-nitrosophenyl)-amine (see patent U.S. Pat. No. 5,484,708)/ml and 33 mM sugar).
 Extinction at 620 nm is monitored during the first 5 minutes at 25° C.
 One Unit enzyme activity corresponds the conversion of 1 mMol mediator/min under the above assay conditions
 Calculation: Activity=(total volume*dE/min [U/ml]): (ε*sample volume*1)
(ε620 nm=30[1* mmol−1*cm−1]).
 The assay was performed with glucose, maltose and galactose (Merck, Germany).
 The wild-type and mutant T348G of s-GDH were applied for glucose determination. The reference samples contained 65 mg glucose/dl. The “test”-samples contained 65 mg glucose/dl and 130 mg/dl maltose. The same amounts of GDH activity (U/ml; see enzyme assay) were used for each assay.
 In a cuvette was mixed:
 1 ml 0.315 mg (4-(dimethylphosphinylmethyl)-2-methyl-pyrazolo-[1.5a]-imidazol-3-yl)-(4-nitrosophenyl)-amine ml/0.2 M citrate pH 5.8
 0.015 ml sample (glucose or glucose+maltose)
 0.045 ml H2O
 The assay was started adding 0.050 ml 90 U/ml s-GDH . The change of absorption at 620 nm was monitored. After 5 minutes constant values were observed and the dE/5 min calculated. The value obtained measuring the reference sample with wild-type s-GDH was set to 100%. The other values were compared to this reference value and calculated in %.
 It can be clearly seen that the ,,glucose-value” measured is markedly less impaired when the mutated s-GDH is used in this determination.