The present invention relates to a mutant serine protease having glutamyl endopeptidase activity, comprising an amino acid sequence that is at least 90 % identical to the amino acid sequence of SEQ ID NO: 1 and comprising an amino acid substitution at the postion corresponding to position G166 of SEQ ID NO: 1.

The present invention further relates to a polynucleotide encoding a mutant serine protease, to an expression vector comprising said polynucleotide, as well as to a host cell comprising the expression vector of the present invention. In yet another aspect, the present invention provides a method of determining the amount if HbAlc, and a use and a kit related thereto.

In a further aspect, the present invention pertains to a use of a mutant serine protease according to the invention for enzymatically cleaving a polypeptide.

Diabetes mellitus is defined as high blood glucose (hyperglycemia) in patients. Normally, the human body maintains blood glucose in a steady range: 70 mg/dl to 1 10 mg/dl (~ 3.9 to 6 mM). This optimal range can fluctuate throughout the day following diet, exercise, stress or illness.

Two main diabetes forms, i.e. types 1 and 2, were officially distinguished in 2003 by the ADA Expert Committee on the diagnosis and Classification of Diabetes Mellitus (Gavin et al., 2003). Type-1 diabetes, which was formerly called IDDM (Insulin Dependant Diabetus Mellitus) and appears most commonly in childhood, is characterized by little or no circulating insulin and can be controlled by regular injections of insulin. Type-2 diabetes was named NIDDM (Non Insulin Dependent Diabetes Mellitus) or adult-onset diabetes, which accounts for about 85 % to 95 % of all diagnosed cases of diabetes. In this form, the body does not respond to insulin. Other types of diabetes are defined separately besides these two diabetes categories: pre-diabetes, gestational diabetes and all other specific types of diabetes. Pre-diabetes is a high glucose level (Impaired Glucose tolerance: IGT) but not as high as normal diabetes. In most cases, people with pre-diabetes develop type-2 diabetes within 10 years. Gestational diabetes (Gestational Diabetes Mellitus) is developed by half of pregnant women and is similar to type-2 diabetes by a hyperglycemia diagnose. In this case, the pregnancy hormones block the action of insulin, thereby making the women less sensitive to their own insulin. It usually goes away after the child was born, but they have a 40 % to 60 % chance of developing diabetes, most of the time type-2 diabetes, within the next 5-10 years (IDF, 2007). The latter class of diabetes includes all the other specific types of diabetes. In this group are included people with beta cell function defects, insulin action defects and exocrine pancreatic disease or dysfunctions (Gavin et al., 2003). Diabetes is ranked among the leading causes of cardiovascular disease, blindness, renal failure and lower limb amputation. According to a survey of the World Health Organization (WHO), 171 million adults, older than 20 years old, were affected by Diabetes worldwide in 2000. At this time, the WHO estimates to reach 366 million diabetics in 2030. A recent study of the International Diabetes Federation (IDF) estimated in 2006 this number of 246 million, which represents 6 % of the world's population and has predicted that it would increase up to 380 millions of diabetics worldwide in 2025, which means 7.1 % of the adult population.

In diabetes management, glycemic control is correctly assessed by a combination between the self-monitoring blood glucose and the HbAlc test. The HbAlc test is a clinical blood glucose control from the previous 2-3 months (Nakamoto, 2004). By its measurement of the glycated hemoglobin (HbAlc), it is an official long-term key parameter for diabetes management. This test is necessary for the following-up of long-term complications of the diabetes disease as well as the short-term risk of life-threatening hypoglycemia according to the DCCT (DCCT/EDIC, 1993). This assay determines whether treatment is adequate and to guide treatment adjustments (Nathan et al., 2007).

In the blood, glucose binds reversibly to the hemoglobin in the red cells. The amount of glucose is directly proportional to the blood concentration. Blood glucose assessment is by this way directly related to Diabetes mellitus (Bunn et al., 1981). The glycated hemoglobin measurement, 'the HbAl c test', reflects the blood glucose mean during red cell life and is therefore a completely appropriate long-term parameter of blood glucose control, which prevents from diabetes long-term complications. In general, HbAlc is defined as hemoglobin that is irreversibly glycated at one or both amino terminal valines of the β- chains (β-Ν-1 -deoxyfructosyl-hemoglobin). It excludes hemoglobin, which is glycated on lysine residue of the β-chain. As the red blood cell life span is 120 days, the measurement of the glycated hemoglobin should provide an assessment of average blood glucose control during these 120 days prior to the test. Even if this measurement is heavily influenced by young red blood cells because of their greater number, the HbAl c test is an indicator of only the last 2-3 months, as blood glucose level assessment 30 days before the test determines roughly half of the HbAl c's test results (Nakamoto, 2004). In this case, a regular test of the glycated hemoglobin every three months gives a good estimation of the average blood sugar and thereby on treatment efficiency and patient ^' s compliance (Nathan et al., 2007). That is, the test is set as such that the percentage of HbAlc indicates the percentage of hemoglobin, which is linked to glucose molecules.

Proteases can be divided into five families: serine, threonine, cysteine, aspartic and metalloproteases. Serine, threonine and cysteine proteases differ by their catalytic action from the aspartic and metalloproteases. The nucleophile of the catalytic site is part of amino acids in the three first groups, whereas there is the interaction of an activated water molecule in the two other peptidase groups (Rawlings et al., 2004). Serine proteases are characterised by the nucleophilic attack of the hydroxyl group of the serine of the catalytic triad directly on the peptide bond.

To date, several proteases are known in the art which may be used for peptide mapping in hemoglobinopathies, including, e.g., proteases such as trypsin. Trypsin selectively hydrolyses peptide bonds at the carboxy-terminal side of lysine and arginine residues. In particular, trypsin can be used to selectively cleave the hemoglobin β-chain at the lysine residue at position 8. However, trypsin cannot be used to avoid the quantification of double or single glycated hemoglobins at the Lys-8 position.

Hence, there is always a need for the provision of an improved protease which can be used for the selective cleavage of hemoglobins. In the context of the present invention, it has surprisingly been found that a mutant serine protease with at least 90 % sequence identity to SEQ ID NO: 1 and with an amino acid substitution at position G166 has an increased specific activity towards a substrate relative to a corresponding wild-type enzyme.

In detail, directed evolution was performed on the wild-type V8-GluC enzyme having the amino acid sequence of SEQ ID NO: 1 to generate libraries of variants. Strategic amino acids to mutate were initially chosen with respect to the crystal structure of V8-GluC (Prasad et al., 2004). Some mutagenesis positions were decided after protein modelling of the V8-GluC with the hemoglobin substrate and some amino acid positions were defined according to the homology sequence comparisons of the recombinant wild type V8-GluC and glutamyl endopeptidases from other organisms. The site directed mutagenesis were focused on the differences between the V8-GluC and the recombinant V8-GluC (Carmona and Gray, 1987; Drapeau, 1978). Unexpectedly, it has been found that it is possible to generate serine protease mutants with improved substitute specificity, and even more unexpectedly it has been found that it is only a few well-defined amino acid positions which are of major relevance in that respect.

In particular, it was surprisingly found in the context of the present invention that the mutation of amino acid residue G166 of SEQ ID NO: 1 resulted in an increased specific activity of the enzyme, a biochemical feature which is reflected by a decreased Km-Value. The results of these experiments are, e.g., described in Example 2, item 2.5 and 2.6.

Accordingly, in a first aspect, the present invention provides a mutant serine protease having glutamyl endopeptidase activity, comprising an amino acid sequence that is at least 90 % identical to the amino acid sequence of SEQ ID NO: 1 and comprising an amino acid substitution at the position corresponding to position G166 of SEQ ID NO: 1.

The term "serine protease" (also referred to in the art as "serine endopeptidase") generally means an enzyme that cleaves peptide bonds in proteins in which one of the amino acids at the active site is a serine residue (S or Ser). Preferably, the serine protease of the invention has glutamyl endopeptidase activity, i.e. it cleaves at the carboxy-terminal side of glutamate and aspartate residues of proteins. Several organisms are able to produce glutamyl endopeptidase enzymes. The most well- known characterized glutamyl endopeptidases are derived from Staphylococcus or Bacillus strains, i.e. Staphylococcus aureus, Staphylococcus epidermis or Staphylococcus warnerii, and Bacillus sutilis or Bacillus licheniformis. The term "V8-Glutamyl endopeptidase" or "V8-GluC" as used herein generally means a glutamyl endopeptidase which is derived from the Staphylococcus aureus strain V8. Glutamyl endopeptidase are classified as EC 3.4.21.19 by the Enzyme Commission of the International Union of Biochemistry and belong to the serine protease family which is characterized by its common catalytic triad: histidine - serine - aspartic acid. Hence, preferably, a serine protease of the invention may be any of the proteases classified under EC 3.4.21.19.

In the Staphylococcus aureus strain V8, two isoforms of the V8-Glutamyl endopeptidase exist, the first was described by Drapeau in 1978 (Drapeau, 1978), while the second was discovered by Carmona in 1987 (Carmona and Gray, 1987). These two protein sequences slightly differ in length. The V8-GluC endopeptidase, which is encoded from the sspA gene, is known in the art under several common synonyms: endoproteinase Glu-C, V8 Protease, V8 proteinase, staphylococcal serine proteinase, Staphylococcus strain V8 serine endopeptidase. V-GluC specifically cleaves peptide bonds on the carboxyl-terminal side of aspartic acid and more preferentially of glutamic acids.

The term "SEQ ID NO: 1 " as referred to herein denotes the amino acid sequence as shown in SEQ ID NO: 1 (see also Figure 4), and represents the amino acid sequence of V8-GluC from Staphylococcus aureus in its active form as described by Carmona and Grey in 1987. In particular, the term "SEQ ID NO.: 1" refers to the amino acid sequence as set forth below (not including the N-terminal prosequence of 39 amino acids):

1 VILPNNDRHQ ITDTTNGHYA PVTYIQVEAP TGTFIASGVV VGKDTLLTNK 51 HVVDATHGDP HALKAFPSAI NQDNYPNGGF TAEQITKYSG EGDLAIVKFS

101 PNEQNKHIGE VVKPATMSNN AETQVNQNIT VTGYPGDKPV ATMWESKGKI 151 TYLKGEAMQY DLSTTGGNSG SPVFNEKNEV IGIHWGGVPN EFNGAVFINE 201 NVRNFLKQNI EDIHFANDDQ PNNPDNPDNP NNPDNPNNPD EPNNPDNPNN 251 PDNPDNGDNN NSDNPDAA

A "mutant serine protease" as used herein generally means a serine protease as detailed above, but with at least one amino acid substitution as compared to the corresponding wild-type enzyme. A mutant serine protease of the invention, however, may have one or more amino acid additions or may comprise one or more amino acid deletion(s). The introduction of single and/or multiple amino acid mutation into a protein of interest as well as the deletion of amino acid residues is a standard procedure known in the art as site- directed mutagenesis which is routinely used in the field of molecular biology. The principles of site-directed mutagenesis are, e.g., described in the Instruction Manual of the QuickChange™ Site-directed Mutagenesis Kit (http ://www. stratagene . com/manual sf), and are, e.g., illustrated in paragraph 1.3 of the Material & Methods section.

The term "at least 90 % identical" or "at least 90 % sequence identity" as used herein means that the sequence of the mutant serine protease according to the present invention has an amino acid sequence characterized in that, whithin a stretch of 100 amino acids, at least 90 amino acids residues are identical to the sequence of the corresponding wild-type sequence. In one embodiment, the sequence of the mutant serine protease according to the present invention may comprise, in addition to the substitution at position G 166 of SEQ ID NO.: 1 , up to 10 % conservative amino acid substitutions. Examples of conservative amino acid substitutions include those listed below:

Original Residue Conservative Substitutions

Ala Ser

Arg Lys

Asn Gin; His

Asp Glu

Cys Ser

Gin Asn

Glu Asp

His Asn; Gin

He Leu, Val

Leu lie; Val

Lys Arg; Gin; Asn

Met Leu; He

Phe Met; Leu; Tyr

Ser Thr Thr Ser

Tip Tyr

Tyr Trp; Phe

Val He; Leu Sequence identitiy according to the present invention can, e.g., be detemined by methods of sequence alignment in form of sequence comparision. Methods of sequence alignment are well known in the art and include various programs and alignment algorithms which have been described in, e.g., Pearson and Lipman (1988). Moreover, the NCBI Basic Local Alignment Search Tool (BLAST) is available from several sources, including the National Center for Biotechnology Information (NCBI, Bethesda, MD).

In the context of the present invention, it has surprisingly been found that the amino acid exchange of G166 of SEQ ID NO: 1 to either G166I (glycine to isoleucine), G166L (glycine to leucine), or G166R (glycine to arginine) is particularly suitable to provide a mutant serine protease with glutamyl endopeptidase activity. Accordingly, in a preferred embodiment, the mutant serine protease of the invention is characterized in that is has an amino acid substitution selected from the group consisting of G166I, G166L, and G166R.

Preferably, this amino acid substitution is G166I.

In an alternatively preferred embodiment, a mutant serine protease of the invention comprises or consists of the amino acid sequence as defined by SEQ ID NO.: 2 (see Figure 5). SEQ ID NO.: 2 denotes the V8-GluC endopeptidase in its active form as described by Drapeau (1978), and has the amino acid sequence as detailed below (not including the N- terminal prosequence of 39 amino acids):

1 VILPNNDRHQ ITDTTNGHYA PVTYIQVEAP TGTFIASGVV GKDTLLTNKH

51 VVDATGDPHA LKAFPSAINQ DNYPDGGFTA EQITKYSGEG DLAIVKFSPN

101 EQNKHIGEVV KPATMSNNAE TQVDQNITVT GYPGDKPVAT MWESKGKITY

151 LKGEAMQYNL STTGGNSGSP VFNEKNEVIG IHWGGVPNQF NGAVFINNEN

201 VRNFLKQNIE DIHFANDDQP NNPDNPDNPN NPDNPNNPDE PNNPDNPNNP

251 DNPDNGDNNN SDNPDAA Preferably, a mutant serine protease may have the sequence of SEQ ID NO.: 2 with a point mutation at the position equivalent to G166 of SEQ ID NO: 1 (i.e. position G164 of SEQ ID NO: 2). More in particular, the amino acid residue of SEQ ID NO.: 2 corresponding to G166 of SEQ ID NO: 1 is substituted by either isoleucine, leucine or arginine, preferably by isoleucine.

A mutant serine protease according to the present invention has an increased specificity activity as compared to a corresponding wild-type enzyme, a biochemical feature which can be experimentally measured, determined and/or quantified by the appropriate enzymatic test system. Particularly, suitable procedures for analyzing an increased specific activity are, e.g., outlined in paragraph 2 of the Materilal & Method section and Example 1 , item 1.1. The expert in the field will appreciate that comparison of the specific activities is best carried out at equimolar concentrations of the substrate molecules investigated using well-defined assay conditions. Otherwise, corrections for differences in concentrations have to be made. Accordingly, in a preferred embodiment, the mutant serine protease of the invention is characterized in that is has an at least 2-fold increased specific activity relative to the corresponding wild type enzyme, preferably an at least 2.5-fold increased specific activity, preferably an at least 3 -fold increased specific activity, more preferably an at least 3.5 -fold increased specific activity, and most preferably an at least 4-fold increased specific activity. The term "specific activity" is well known to the skilled person, it is preferably used to describe the enzymatic activity per amount of protein. Various methods are known in the art to determine the specific activity of an enzyme, one of which, e.g., includes the analysis of the selective cleavage of a peptide substrate at a position carboxy-terminal to a glutamic or aspartic acid. Such enzymatic assays can be carried out in an in vitro set up, and aim at determining the enzymatic activity of a protein by analyzing the catalytic cleavage of a synthetic substrate, e.g., a paranitroanilin substrate such as, for example, Z- Phe-Leu-Glu-pNA. Paranitroaniline substrates are commonly used for the detection and measurement of serine protease activities by spectrophotomer. One unit of enzyme activity is thereby defined as the amount of enzyme required to cleave 1 μιτιοΐε of substrate per minute in the presence of a certain amount of substrate in a defined volume. The enzymatic activity is then calculated by determining the specific activity based on complete cleavage of the substrate in the reaction volume. In case of a peptide-pNP substrate, for example, the formation of p-Nitro-Phenol (pNP) is quantified at a wavelength of 405 nm. A test for analyzing the enzymatic activity of a mutant serine protease according to the present invention is, for example, described in Example 1, 1.1. Specific activity according to the present invention may further include calculation of the m-value. In particular, enzymes catalyse substrate by reducing the activation energy and by altering the steric constant in the Arrhenius equation. In a equilibrium, an enzyme (E) binds a substrate (S) to form an enzyme-substrate complex (E-S). The E-S complex can dissociate or irreversibly convert the substrate to a product. The Michaelis-Menten equation describes the relationship between the rate of substrate conversion by an enzyme to the concentration of the substrate. In this equation, V is the rate of conversion, Vmax is the maximum rate of conversion, [S] is the substrate concentration, and Km is the Michaelis constant, the substrate concentration at which a catalized reaction proceeds at half of its maximum velocity Vmax. It enables to estimate how well the enzyme binds its substrate. Km is an apparent dissociation constant (Kd) and represents the substrate concentration at half of the maximum enzyme velocity Vmax. Therefore, a lower Km value indicates a higher affinity for the substrate. The Lineweaver-Burk equation affords a line with a slope of Km/Vmax and y-intercept of 1/Vmax. The Km-value and Vmax can be easily determined by plotting the Lineweaver-Burk equation, which is the inverse of the Michaelis-Menten equation.

In the context of the present invention, it has surprisingly been found that a mutant serine protease reveals a three-time lower Km-value as compared to its corresponding wild-type enzyme. The determination of Km-values according to the present invention is, e.g., described in Example 2, item 2.6. As used herein, the term „wild-type enzyme"generally means any naturally occuring enzyme with enzymatic activity of a serine protease. Preferably, a wild-type enzyme of the present invention may be any enzymes classified under EC 3.4.21.19.

In the context of the present invention, two different test assays were implemented in order to analyze the enzymatic activity of the mutant serine protease of the invention by cleavage of a chromophore substrate. At first, the six first amino acids of the hemoglobin β-chain without glycosylation were synthesized with a paranitroanilin bond at the carboxyl- terminal end (herein referred to as "pNA"). This substrate enables to check if the serine protease of the invention cuts correctly and specifically after the sixth amino acid of the hemoglobin beta-chain, i.e. carboxy-terminal to the glutamic acid. This substrate, which could not be glycated, was designated "hemoglobin hexapeptide substrate" and is also referred to as "HbA0(l-6)-pNA" herein. It represents the following amino acid pattern: (NH2-VHLTPE-pNA). Secondly, another substrate, which is glutamate specific was used to check the glutamic acid cleavage specificity of the mutant serine protease according to the present invention. It consists of a Carbobenzoxy-Phenylalanine-Leucine-Glutamate chain (Z-Phe-Leu-Glu-pNA), whose end is linked to paranitroanilin (also referred to herein as "pNA"). The substrate is termed "glutamate substrate" or "Z-Phe-Leu-Glu-pNA" herein. The term "Z" as used herein means carbobenzoxy. The use of these test systems for analyzing the specific activity of the mutant serine protease in the context of the invention is, e.g., described in Example 1 , item 1.1 and Example 2, item 2.5.

Accordingly, in a preferred embodiment, the mutant serine protease of the invention is characterized in that the increase in specific activity is measured using the substrate Z-Phe- Leu-Glu-pNA, wherein Z is carbobenzoxy and pNA is paranitroanilin.

A "substrate" according to the present invention may be any amino acid sequence of variable length which may represent a substrate of a serine protease. A substrate according to the present invention may be a peptide with a naturally occuring amino acid sequence (i.e. comprising the amino acid sequence of a naturally occuring enzyme substrate) or may be an artificially designed peptide (i.e. comprising the amino amino acid sequence of an experimentally defined enzyme substrate). A substrate of the invention may further comprise amino acid modifications, such as, e.g. modified amino acids, or N- or C- terminal non amino acid moieties, including, but not limited to, chromophores such as paranitroaniline. Particulary suitable peptide substrates of the present invention are, e.g., described in Example 1 , item 1.1. As also found in the context of the present invention, the mutant serine protease of the invention also reveals an increased thermostability as compared to the wild-type enzyme. The increased thermostability of the mutant serine protease of the invention is exemplified in Example 2, item 2.7 and Figure 3.

Accordingly, in yet another preferred embodiment, the mutant serine protease of the present invention is characterized in that it has an increased thermostability relative to the corresponding wild-type enzyme.

The term "thermostability" as used in the context of the present invention includes any kind of process which facilitates and/or enables a protein to be stable or to keep in solution, including any process which facilitates and/or enables the solubilisation of a protein of interest at any temperature, e.g. in a temperature range of from 4°C to 55 °C. The solubilisation and/or stabilization of a protein includes, but is not limited to, the absence of appearance of precipitation and/or aggregation product(s), the presence and/or absence of which can be analysed by standard methods known in the art including, e.g., the analysis by means of visual inspection, by means of optical density measurements including e.g. the determination of the total protein content of a composition containing the thermostabilized protein. The term "thermostability" further includes the feasibility of a protein to catalyze an enzymatic reaction. As already described above, said enzymatic activity can be determined and evaluated by means of biochemical standard assays known to the person skilled in the art and, e.g., described herein in paragraph 2 of the Material & Method section.

In another preferred embodiment, the mutant serine protease of the present invention is characterized in that it is capable of selectively cleaving off the N-terminal hexapeptide VHLTPE (SEQ ID NO: 3) from the N-terminus of the glycated hemoglobin (HbAlc) β- chain.

The term "selective cleavage" or "selectively cleaving" as used herein means that the enzyme specifically recognizes its target site. In particular, selective cleavage according to the present invention means that the enzyme selectively hydrolyzes a peptide bond at the carboxy-terminal position of either a glutamic or aspartic acid. The selective cleavage may further be influenced by the appropriate buffer conditions, i.e. the selective cleavage at the carboxy-terminal position of a glutamic acid may require different buffer conditions as the selective cleavage at the carboxy-terminal position of an apartic acid residue.

The term "glycated hemoglobin (HbAl c) β-chain" as referred to herein generally means glycated hemoglobin, also particularly refers to hemoglobin Ale, AI C or Hblc. In particular, in the context of the present invention, the term HbAlc β-chain means hemoglobin that is glycated (or glycosylated) at one or both amino terminal valines of the β-chains (β-Ν-1 -deoxyfructosyl-hemoglobin). In general, HbAlc is formed by a chemical reaction which simply depends on the presence of high plasma levels of glucose, known in the art as the Maillard reaction, and also called glycation (Nakamoto, 2004). In the case of the glycated hemoglobin, the Maillard reaction occurs under physiological conditions and at a specific site on the protein, i.e. at the N-terminal valine of the β-chain of hemoglobin. It can also take place at multiple sites at the amino-terminus of a chain and ε-amino group of several lysine residues on both a and β-chains.

The term "hexapeptide" as used herein generally means a peptide with six amino acid residues. A hexapeptide of the invention may be a peptide of natural origin (i.e. corresponding to a naturally occurring protein sequence), or may be a synthetic peptide with variable sequence. In particular, in the context of the present invention, a hexapeptide means a peptide corresponding to HbAlc of diverse organism. In particular, a hexapeptide of the present invention means a peptide with the amino acid sequence of VHLTPE (SEQ ID NO: 3), corresponding to the sequence of the first six N-terminal amino acid residues of the human HbAlc β-chain. The mutant serine protease of the invention may further be any functionally active fragment or derivative of either SEQ ID NO: 1 or SEQ ID NO: 2. The feature "functionally active fragment or derivative' ^" as used herein generally refers to any kind of protein revealing the enzymatic activity of V8-GluC in any partially, substituted or modified form. That is, a functionally active fragment of the present invention might be comprised of protein domain(s) originated from SEQ ID NO: 1 or SEQ ID NO: 2. A functionally active fragment may comprise any N-terminal, C-terminal or central protein domain of interest, or may be composed of any combination thereof. Furthermore, a functionally active derivative might comprise additional amino acid in form of, e.g., N- or C-terminal extensions and/or as part of inner protein domains.

Amino acid extensions may include, but are not limited to, one or more epitope tag(s) to which a substance, such as an antibody or an affinity matrix, can selectively bind. Such an epitope tag is generally placed at the amino- or carboxyl-terminus of the peptide but may also be incorporated as an internal insertion or substitution as the biological activity permits. The presence of an epitope-tag can, e.g., be detected by use of an antibody against said epitope. Also, provision of the epitope tag enables the peptide to be readily purified by affinity purification using an anti-tag antibody or another type of affinity matrix that binds to the epitope tag. Various tag polypeptides and their respective antibodies are well known in the art. Examples include poly-histidine (poly-his), poly-histidine-glycine (poly-his-gly) tags, the HA-tag polypeptide, the c-myc tag, the Strep-tag and the FLAG-tag. A functionally active fragment may also be a protein derived from the protein as defined by SEQ ID NO: 1 or SEQ ID NO: 2 by one or more amino acid deletions and/or substitutions. The deletions and/or substitutions may be C-terminally, N-terminally and/or internally. In one preferred embodiment of the invention the functionally active fragment consists of at least 60 %, preferably at least 70 %, more preferably of at least 80 %, still more preferably of at least 90 %, even more preferably of at least 95 %, and most preferably of 99 % of the amino acid sequence according to SEQ ID NO: 1 or SEQ ID NO: 2. The percentage of sequence identity can be determined, e.g., by sequence alignment. Methods of alignment of sequences for comparison are well known in the art and include various programs and alignment algorithms which have been described in, e.g., Pearson and Lipman (1988). Moreover, the NCBI Basic Local Alignment Search Tool (BLAST) is available from several sources, including the National Center for Biotechnology Information (NCBI, Bethesda, MD) as well as the Internet, for use in connection with the sequence analysis programs blastp, blastn, blastx, tblastn and tblastx.

A functionally active derivative of the present invention may also comprise sequence alterations including, but are not limited to, conservative substitutions, deletions, mutations and insertions. Moreover, a functionally active derivative of the present invention might comprise one or more modified amino acids including, but not limited to, e.g. phosphorylated, acetylated, ubiquitinated, and/or sumolylated residues. A functionally active derivative of the present invention might further encompass any sort of chemical label such as, e.g., fluorescence-labeled moieties. However, a functionally active derivative of the invention exhibits enzymatic activity of V8-GluC. As detailed above, a functionally active fragment or derivative can be identified and enzymatically characterized by means of the assay systems described herein and as, e.g., exemplified in paragraph 2 of the Material & Method section.

In the context of the present invention, a functionally active fragment or derivative of the invention means a protein which enzymatic activity amounts to at least 50 %, preferably to at least 70 %, more preferably to at least 80 %, especially to at least 90 %, particularly at least 95 %, most preferably at least 99 % of the activity of the protein as defined by SEQ ID NO: 1 or SEQ ID NO: 2 without sequence alterations.

Introducing point mutation(s) and/or deletion(s) into a protein of interest, thus thereby generating a functionally active fragment or derivative, is a standard procedure known to a person skilled in the art as site-directed mutagenesis, and product for this purpose are commercially available from a variety of suppliers including, e.g., Stratagene (USA).

In another aspect, the present invention relates to a polynucleotide encoding the mutant serine protease according to the present invention and as detailed above. The term "polynucleotide ^" as used herein generally refers to any nucleotide molecule which encodes the protein of the invention and which may be of variable length. Examples of a polynucleotide of the invention include, but are not limited to, plasmids, vectors, or any kind of DNA and/or RNA fragment(s) which can be isolated by standard molecular biology procedures, including, e.g. ion-exchange chromatography. A polynucleotide of the invention may be used for transfection or transduction of a particular cell or organism.

In a further aspect, the present invention pertains to an expression vector comprising a polynucleotide encoding a mutant serine protease as defined herein, characterized in that the polynucleotide is operably linked to a promoter sequence capable of promoting the expression of the polynucleotide in a host cell.

As used herein, the term "expression vector" generally refers to any kind of nucleic acid molecule that can be used to express a protein of interest in a cell. In particular, the expression vector of the invention can be any plasmid or vector known to the person skilled in the art which is suitable for expressing a protein in a particular host cell including, but not limited to, mammalian cells, bacterial cell, and yeast cells. In particular, expression vectors of the present invention include, but are not limited to, all vectors known in the art which are suitable for expressing a protein in Bacillus subtilis. More preferably, an expression vector of the invention may be, e.g., the pMSE3 vector (Silbersack et al., 2006). An expression construct of the present invention may also be a nucleic acid which encodes a serine protease of the invention, and which is used for subsequent cloning into a respective vector to ensure expression. Plasmids and vectors for protein expression are well known in the art, and can be commercially purchased from diverse suppliers including, e.g., Promega (Madison, WI, USA), Qiagen (Hilden, Germany), Invitrogen (Carlsbad, CA, USA), or MoBiTec (Germany). Methods of protein expression are well known to the person skilled in the art and are, e.g., described in Sambrook et al., 2000 (Molecular Cloning: A laboratory manual, Third Edition).

To ensure expression of the protein, the polynucleotide which encodes a mutant serine protease of the invention is operably linked to sequence which is suitable for driving the expression of a protein in a host cell. However, it is encompassed within the present invention that the claimed expression construct may represent an intermediate product, which is subsequently cloned into a suitable expression vector to ensure expression of the protein. The expression vector of the present invention may further comprise all kind of nucleic acid sequences, including, but not limited to, polyadenylation signals, splice donor and splice acceptor signals, intervening sequences, transcriptional enhancer sequences, translational enhancer sequences, drug resistance gene(s) or alike. Optionally, the drug resistance gene may be operably linked to an internal ribosome entry site (IRES), which might be either cell cycle-specific or cell cycle-independent. The term "operably linked" as used herein generally means that the gene elements are arranged as such that they function in concert for their intended purposes, e.g. in that transcription is initiated by the promoter and proceeds through the DNA sequence encoding the protein of the present invention. That is, RNA polymerase transcribes the sequence encoding the fusion protein into mRNA, which in then spliced and translated into a protein.

The term "promoter sequence" as used in the context of the present invention generally refers to any kind of regulatory DNA sequence operably linked to a downstream coding sequence, wherein said promoter is capable of binding RNA polymerase and initiating transcription of the encoded open reading frame in a cell, thereby driving the expression of said downstream coding sequence. The promoter sequence of the present invention can be any kind of promoter sequence known to the person skilled in the art, including, but not limited to, constitutive promoters, inducible promoters, cell cycle-specific promoters, and cell type-specific promoters.

In yet another aspect, the present invention pertains to a host cell comprising the expression vector of the present invention. A "host cell" of the present invention can be any kind of organism suitable for application in recombinant DNA technology, and includes, but is not limited to, all sorts of bacterial and yeast strain which are suitable for expressing one or more recombinant protein(s). Examples of host cells include, for example, various Bacillus subtilis or E. coli strains. A variety of E. coli bacterial host cells are known to a person skilled in the art and include, but are not limited to, strains such as DH5-alpha, HBl Ol , MV1 190, JM109, JMl Ol , or XL-1 blue which can be commercially purchased from diverse suppliers including, e.g., Stratagene (CA, USA), Promega (WI, USA) or Qiagen (Hilden, Germany). Bacillus subtilis strains which can be used as a host cell include, e.g., 1012 wild type: leuA8 metB5 trpC2 hsdRMl and 168 Marburg: trpC2 (Tip-), which are, e.g., commercially available from MoBiTec (Germany).

The cultivation of host cells according to the invention is a routine procedure known to the skilled person. That is, a polynucleotide encoding a mutant serine protease of the invention can be introduced into a suitable host cell(s) to produce the respective protein by recombinant means. These host cells can by any kind of suitable cells, preferably bacterial cells such as Bacillus subtillis, which can be cultivated in culture. At a first step, this approach may include the cloning of the respective gene into a suitable plasmid vector. Plasmid vectors are widely used for gene cloning, and can be easily introduced, i.e. transfected, into bacterial cells which have been made transiently permeable to DNA. After the protein has been expressed in the respective host cell, the cells can be harvested and serve as the starting material for the preparation of a cell extract containing the protein of interest. A cell extract containing the protein of interest is obtained by lysis of the cells. Methods of preparing a cell extract by means of either chemical or mechanical cell lysis are well known to the person skilled in the art, and include, but are not limited to, e.g. hypotonic salt treatment, homogenization, or ultrasonification. The cultivation of a host cell for expressing a mutant serine protease of the invention is described in Example 2, item 2. 1 .

As detailed above, the mutant serine protease according to the present invention selectively hydrolyzes a peptide bond at the carboxy-terminal position of either a glutamic or aspartic acid. This selective cleavage is of high value in the art, e.g., in the field of proteomics, where the generation of defined peptide fragments in a reproducible manner is of crucial importance. That is, in proteomics, peptides are usually generated upon digestion of polypeptides in a sample which are then characterized by mass spectroscopy.

Accordingly, in another aspect, the present invention relates to the use of a mutant serine protease as defined in the context of the present invention for enzymatically cleaving a polypeptide. Preferably, the mutant serine protease of the invention is used to enzymatically and specifically cleave polypeptides comprised in a sample and to analyze the peptides generated thereupon by mass spectroscopy.

The term "polypeptide" as used therein generally refers to any kind of amino acid sequence comprising least five amino acid residues, preferably at least ten amino acid residues, more preferably at least twenty amino acids, or at least 50 amino acid residues. A polypeptide according to the present the invention comprises amino acids that are covalently linked by peptide bonds, and may also include any kind of peptide or protein. As also detailed above, the mutant serine protease of the invention is particularly suitable for cleaving glycated hemoglobin with increased specific activity. That is, the mutant serine protease of the invention may be particularly suitable for the determination of HbAl c levels in a patient ^' s sample.

Therefore, in another aspect, the present invention provides a method of determining the amount of HbAlc, said method comprising the steps of a) providing a blood sample from a subject; b) incubating the blood sample of step a) with a mutant serine protease as defined in any of claims 1 to 8 under conditions conducive to the selective cleavage of the HbAlc β-chain, thus releasing an N-terminal hexapeptide; and c) determining the amount of hexapeptide released in step b), thereby determining the amount of HbAlc.

The feature "providing a sample" as used in the context of the present invention generally refers to all kind of procedures suitable to prepare blood sample from an individum. These procedures include, but are not limited to, standard biochemical and/or cell biological procedures suitable for the preparation of a lysed blood sample, i.e. a sample in which the hemoglobin is released from the erythrocytes. The preparation of a lysed blood sample is a standard procedure known to the person skilled in the art and may include, e.g., the use of hypotonic salt conditions.

The feature "conditions conducive to the selective cleavage" as used therein generally refers to the conditions under which the selective cleavage of a peptide substrate by a serine protease takes place. In particular, in the context of the present invention, conditions conducive to the selective cleavage refer to any parameter which may be necessary or sufficient in order to enable the selective cleavage of the HbAl β-chain at the carboxy- terminal to the amino acid at position 6, i.e. after residue E6 of the HbAl β-chain. Conditions conducive to the selective cleavage may include a variety of buffer and/or salt conditions, including the use of Tris/HCl buffer (e.g. Tris/HCl 0.1 M pH 7.8), potassium phosphate buffer (e.g. KH2P04 0.1M pH 7.8), or ammonium acetate buffer (e.g. NH4/Acetat 50 mM pH 4.3). Preferably, conditions conducive for the selective cleavage refers to the buffer conditions which are optimal for cleavage of the hemoglobin substrate at 37 °C at an enzyme-substrate ratio of 1 : 100 (enzyme: hemoglobin) and in the presence of 50 mM ammonium acetate buffer at pH 4.3. Conditions conducive for the selective cleavage of a substrate according to the present invention are, e.g., described in Example 2, e 2.12.

The feature "determining the amount of hexapeptide" as used herein generally means analyzing, visualizing and/or quantifying the amount of a peptide derived from HbAlc and generated by proteolytic cleavage by a mutant serine protease of the invention. In particular, it refers to the functional test of determining glycated hemoglobin (HbAlc) by means of ESI-MS (Electrospray-Ionisation-Mass spectrometry) as standardized by the National Group Standardization Program (NGSP) and the International Federation of Clinical Chemistry and laboratory medicine (IFCC), and for which the parameters are set as detailed below.

Both organizations standardize the HbAl C results to correlate the percentage of HbAlc to the plasma glucose. The plasma glucose concentration, designated "plasma glucose" in the equation, can be calculated in millimoles per liter (mmol/L) or in milligramms per deciliter (mg/dL) and depends on the percentage of HbAl C, referred to as HbAl C (%). The HbAl C standardization between the DCCT and NGSP can be calculated by the following way: Plasma glucose (mmol/L) = 1.98 x HbAlc (%) - 4.29; Plasma glucose (mg/dL) = 35.6 x HbAlc (%) - 77.3; HbAlc reference value = 4.8 -5.9 %.

The IFCC standardization is assessed according the following equations (not available in the US): Plasma glucose (mmol/L) = 1.73 x HbAlc (%) + 0.20; Plasma glucose (mg/dL) = 31.2 x HbAlc (%) + 3.51 ; HbAlc reference value = 2.9 - 4.2 %.

The Designated Comparison Methods (DCM) is the relationship between the IFCC and the NGSP ones. It can be calculated by an equation (Hoelzel et al., 2004). The HbAl c percentage defined by the NGSP is correlated to the HbAlC measurement of the IFCC by the following correcting factor: NGSP HbAlc = 0.915 (IFCC HbAlc) + 2.15 %. The HbAl c reference values show hyperglycemia in the 2-3 months before analysis. The HbAl c rates can increase up to 20 % in the case of a bad diabetes control. Rates under references are synonyms of hypoglycaemia, which can be caused by short red blood cell life span, hemoglobin variants or short-term hypoglycaemia. In the context of the present invention, selective cleavage of the HbAlc β-chain has been found particularly effective in the presence of an ammonium acetate buffer at pH 4.3.

Accordingly, in a preferred embodiment, the method of the invention is characterized in that the cleavage of step b) is carried out in an ammonium acetate buffer at acidic pH.

Moreover, as detailed above, the determination of the hexapeptide is preferably performed by mass spectrometry.

Hence, in yet another preferred embodiment, the method of the present invention is characterized in that the determining of step c) is carried out by means of mass spectrometry, preferably by means of Reverse Phase-HPLC/Electrospray lonisation - Mass Spectrometry (RP-HPLC/ESI-MS). The term "mass spectrometry" as used herein generally refers to any analytical technique suitable for the determination of the elemental composition of a sample or a molecule, including the analysis of the chemical structure of a molecule, such as a peptide or chemical compounds. In particular, mass spectrometry means an analytical technique which principle consists of ionizing chemical compounds to generate charged molecules or molecule fragments and measurement of their mass-to-charge ratios. In mass spectrometry, a sample is typically loaded onto a mass spectrometer instrument, where it undergoes vaporization, and where the components of the sample are ionized by applying an electronic beam, resulting in the formation of positively charged particles (i.e. ions). The positive ions are then accelerated by a magnetic field, and the mass-to-charge ratio of the particles is computed based on the details of motion of the ions as they transit through the electromagnetic fields. The ions are then detected and sorted according to m/z.

Mass spectrometry assesses the purity, sequence or molecular weight of biomolecules with the accuracy of millidaltons. The accuracy of the molecular weight is usually better than 0.01 % of the calculated mass whereas the SDS-PAGE method has an accuracy of 5 to 10 %, which can reach 50 % by the presence of lipid or carbohydrate in the initial solution. The accuracy of the measurement enables to observe posttranslational modifications and proteolytic processing. Electrospray ionization (ESI) as referred to herein means a technique used in mass spectrometry to produce ions, and which is especially useful in producing ions from macromolecules because it overcomes the propensity of these molecules to fragment when ionized. Mass spectrometry using ESI is also called electrospray ionization mass spectrometry (ESI-MS) or, less commonly, electrospray mass spectrometry (ES-MS). Electrospray ionization can be combined with reverse-phase HPLC, a technique which is referred to as RP-HPLC/ESI mass spectrometry. As used herein, the term "Reverse-Phase High Pressure Liquid Chromatography" or "RP-HPLC" means any chromatography method that uses a non-polar stationary phase.

Usually, the blood sample of the invention is subjected to a series of separation steps, each based on a different property to yield a pure hemoglobin preparation.

Accordingly, in another preferred embodiment, the method of the present invention may comprise one or more additional purification step(s) before and/or after step (a), and/or before and/or after step (b).

Additional purification step(s) may include any kind of biochemical method which supports and/or enables the purification of a protein of interest. Such standard methods are well known to a person skilled in the art, and include, but are not limited to, e.g., salting out, dialysis, (ultra)filtration, (ultra)centrifugation, gel-filtration chromatography, ion- exchange chromatography, and/or affinity chromatography.

In another aspect, the present invention also relates to the use of a mutant serine protease as defined in the context of the present invention for determining HbAlc.

The method of the present invention is particulary suitable for determining the amount of glycated hemoglobin in a blood sample from an individum, in particular for determining the amount of glycated hemoglobin by an HbAl c test. Generally, an HbAlc test is performed by drawing blood from a vein, usually from the inside of the elbow or the back of the hand of an individum. An HbAl c of 6% or less is considered normal. If the HbAl c level is above 7%, it means that the diabetes control may not be as good as it should be. In case of diabetes, HbAl c level should be kept at or below 7%. Normal ranges may vary slightly among different laboratories. Abnormal results mean that the blood glucose levels have been above normal over a period of weeks to months.

In yet another aspect, the present invention provides a kit for the determination of HbAlc in a sample using a mutant serine protease as defined herein and at least one other agent required for said determination.

Other agents according to the present invention may include, but are not limited to, agents such as buffer(s) and/or substrate(s).

The following Figures and Examples are intended to illustrate various embodiments of the invention. As such, the specific modifications discussed are not to be construed as limitations on the scope of the invention. It will be apparent to the person skilled in the art that various equivalents, changes, and modifications may be made without departing from the scope of the invention, and it is thus to be understood that such equivalent embodiments are to be included herein.

FIGURES

Fig. 1 shows a comparison of the specific activity ratios between wild-type and mutant V8 GluC in different buffers. (A) E-Wt-H-GluC: wild type V8-Gluc (SEQ ID NO.: 1); E-M-H GluC: mutant V8-GluC (SEQ ID NO.: 1 with the amino acid exchange G166I). The E-Wt- H and E-M-H were compared to the Roche Lyo V8-GluC (SEQ ID NO.: 2) and NEB GluCs in Tris buffer pH 7.8 or potassium phosphate buffers at pH 7.8 or 5 with the Z-Phe- Leu-Glu-pNA as a substrate. Their activities were also assessed in ammonium acetate buffer at pH 4.3 with the HbA0(l -6)-pNA substrate. (B) Specific acitivity (SA) as calculated in U/mg. Fig. 2 shows a comparison of Km values of the wild-type (E-Wt-H GluC) and the mutant enzyme (E-M-H GluC) with respect to the Z-Phe-Leu-Glu-pNA substrate in Tris buffer

0.1 M pH 7.8. The E-M-H mutant Km value is about 3-fold decreased as compared to the Km valule of the enzymes by Roche, NEB and the E-Wt-H GluCs (Km valure of E-M-H GluC: 0.17; average value of three independent experiments).

Fig. 3 illustrates the temperature stability of the E-Wt-H and E-M-H GluC enzymes. The stability of the E-Wt-H and E-M-H GluCs were tested after 30 minutes of incubation in either the presence or absence of 5 mM of Glu-Glu dipeptide. The remaining activity is represented by the percentage of calculated activity in defined experiment conditions (buffers and pH) by referring to the V8-GluC activity at 25°C.

Fig. 4 shows the amino acid sequence of V8-GluC endoprotease as described by Carmona and Gray (1987; SEQ ID NO.: 1) not including the N-terminal prosequence encompassing 39 amino acids. This prosequence, which can be cleaved by another Staphylococus aureus protease, i.e. aureolysin, inactivates the protease after its synthesis in order to prevent it from its own autoproteolysis. Within the prosequence, the amino acid residue at position 39 may be substituted (N39E) in order to reduce autoproteolysis.

Fig. 5 shows the amino acid sequence of V8-GluC endoprotease as described by Drapeau, 1978 (SEQ ID NO: 2) not including the N-terminal prosequence encompassing additional 39 amino acids.

EXAMPLES

1. Material and Methods

1. V8-GIutamyI endopeptidase cloning and expression

1.1 Cloning of the V8-Glutamyl endopeptidase into the pMSE3 vector Shuttle The genomic DNA sequence of the V8-GluC was directly ordered at the German Collection of Microorganisms and Cell Cultures (DSMZ Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Braunschweig, Germany). This genomic DNA is composed of the complete sequence of the V8-GluC with the wild type Signal peptide, prosequence and active V8-GluC (see below). The V8-GluC was cloned with another wild type Signal peptide, which is efficiently recognized for the protein translocation in Bacillus subtilis (Brockmeier et al, 2006).

Signal peptide Prosequence T Active V8-GluC

1 29 68 336 aa PRE PRO PROTEIN

The active V8-GluC gene was extracted with its wild type prosequence by Polymerase Chain reaction. Flanking primers were designed with necessary restnction Sites, Fsel and MluNI, in order to hgate the V8-Glutamyl endopeptidase DNA The coding sequence was added with the Signal peptide between the Fsel and Mlunl restnction Sites. As a Tag can be added only on the carboxyl-terminal side, the reverse primer was designed with a polyhistidine Tag on the 5' side. The polyhistidine tag was designed with a triplet alternation of CAT and CAC, which encodes for histidine. The CAT codon presents a higher translation frequency than the CAC codon. The complete sequence with the promoter, signal peptide, prosequence of the V8-GluC, as well as the polyhistdine tag and the terminator, was then cloned in the multisite cloning of the pMSE3 vector (Silbersack et al., 2006).

1.2 Polymerase Chain Reaction (PCR)

The V8-GluC gene insert was cloned by PCR by addition of the AmyE signal peptide on the amino terminal side of the prosequence. During the PCR, primers hybridize to the DNA template at a specific temperature of annealing (Ta). This temperature can be calculated in all PCR following to this formula: Ta = Tm - 5°C. The Tm is the abbreviation of melting temperature and characterizes the stability of the DNA hybrid, which is formed between an oligonucleotide on the forward strand and its complementary on the reverse strand. It is defined as the temperature at which half of a given oligonucleotide is hybridized to its complementary on the reverse strand. It can be assessed by this way: Tm = 69.3°C + (0.4 x % GC) - 650/length of primer; the term " % GC" means the percentage of dGTP and dCTP as part of the primer sequence. The PCR was performed with an extension time of 1 min/kb. Following quantities were used:

- DNA template: 250 ng

- Pfu Turbo DNA polymerase: 2.5 U

- dNTPs concentrations: 200 μΜ each dNTP ( 0.8 n M total)

- Final reaction buffer IX.

- Primers: each 150 (= 0.3 μΜ). Primer design is shown in figure 32. 1.5 μί of each primer were mixed with 1 μΐ, of dNTP mix, 1 μΐ, of Pfu Polymerase (2.5 υ/μί), 5 μί of PCR reaction buffer 10X, 250 ng of template and sterile double distilled water qs to 50 μΐ,.

Tm: Primer melting temperature

Tm = 58°C

Ta: Annealing temperature of the primers

Ta= Tm-5°C = 53°C

Tm = 69.3°C + (0.4 x % GC) - 650/length of primer

Forward primer (SEQ ID NO: 4)

5' GCAA ggccggccgctgcgagcgctTTATCT TCAAAGGCTATGGACAATC 3'

Reverse primer (SEQ ID NO: 5)

5' GCAA tggccaCTAT TAAGCTGCAT CTGGATTG 3'

Reverse primer with the polvhistidine Tag (SEQ ID NO: 6)

5 GCAA tggccaCTA ATGATGATGATGATG ATGAGCTGCAT CTGGATTGTC 3' After the PCR-product purification, the DNA concentration was measured and cut by the restriction enzymes. The DNA was again cleaned by the PCR purification kit of Qiagen (Hilden, Germany). The vector was cut in the same way. 1 % agarose gel electrophoresis was run to separate the cut vector from the insert. The vector was extracted from the agarose gel by the Qiaquick gel extraction kit of Qiagen (Hilden, Germany). After measurement of the DNA concentration, the vector and the gene of interest were ligated.

1.3 Site directed mutagenesis PCR

The site directed mutagenesis PCR was carried out as follows:

- DNA template: 250 ng

- Pfu Turbo DNA polymerase: 2.5 U

- dNTPs concentrations: 200 μΜ each dNTP ( 0.8 rriM total)

- Final reaction buffer IX.

- Primers: each 100 ng^iL (= 0.2 μΜ).

After that the nucleotide codon was replaced by the wanted mutation according to the Bacillus subiilis usage codon, primers were designed by anchoring on the upstream and downstream 12-13 amino acids around the position to mutate.

Number of Cycles Temperature Duration

1 95°C 30 sec

18 95°C 30 sec

Ta= Tm-5°C 1 min

68°C 2 min per

kilobase

1 4°C to infinity 1.4 Electroporation in B. subtilis

New fresh competent cells were prepared every time. Three Falcon tubes, with 40 iL of WB600 Bacillus subtilis glycerol stock and 4 mL of Bacillus medium (15 g/L of Yeast extract, 10 g/L of NaCl, pH 7), were then incubated at 37°C at 230 RPM overnight. These precultures were poured into a 300-mL Erlenmeyer in 100 mL Bacillus medium the day after and cultivated at 37°C at 230 RPM. When the OD^m = 1.5 the culture was stopped and incubated at 4°C in ice for 1 hour. The culture was then centrifuged at 6000 RPM for 15 min at 4°C. The pellet was resuspended in 15 mL distilled and sterile cold double distilled water. It was centrifuged again at 6000 RPM for 15 min at 4°C. The pellet was resuspended in 1 mL 30 % PEG-4000 solution and put on ice. The electroporation cuvette was precooled on ice. 10 iL of plasmid preparation (~ 0.5 g) were mixed with 100 μί Bacillus subtilis suspension in an eppendorf. The suspension was pipetted into the electroporation cuvette. The electroporation cuvette was then dried and put between the two electrodes. 1 mL SOC Medium was prepared in a pipette and another 1 mL in a Falcon tube. Cells were electropored at 2500 V, 25 μΡ and 200 Ohms. As soon as possible (<ls), the 1 mL of the SOC Medium in the pipette was added into the cuvette after the electroporation. Then it was pipetted into the Falcon tube. The tube was then incubated for 1 hour at 37°C at 230 RPM. 150 μΐ, of the transformed solution were plated onto agar plates with 10 μg/mL of kanamycin sulfate. The other 850 μΐ, were centrifuged one min at 6000 RPM and the 600 μΐ, of supernatant were thrown away. The 150 μΐ. left were mixed with the pellet and plated onto agar plates with 10 μg/mL kanamycin sulfate. Agar plates were incubated overnight at 37°C.

7.5 Bacillus subtilis culture and expression of the V8-GluCs

3 Falcon tubes of 4 mL overnight culture of the WB600 Bacillus strain were incubated overnight at 37 °C and 230 RPM in Bacillus medium with 10 μg/μL kanamycin sulfate. Negative controls were taken with the WB600 strain without the plasmid, where the V8-GluC gene was inserted. These cultures were poured into 100 mL Bacillus medium in a 300 mL Erlenmeyer with 10 μg/μL of kanamycin sulphate and some antifoam drops the day after and cultivated at 37 °C and 230 RPM. When the OD600nm reaches 1 , the culture was induced with 0.5 % of acetoine (10 %). Once a mutant was found after the screening, cells were cultivated in 500 mL volume in 2 L flask.

The OD600 nm, the enzymatic activity and the protein concentration by Bradford were measured every two hours to follow the cell grow and the V8-GuC expression. The enzymatic activity and the protein concentration by Bradford were measured on 1 mL of the supernatant after 1 min centrifugation at 6000 RPM. Pellets were kept to check by DNA extraction and observation an agarose gel that the cells did not lose the plasmid. The rest of the pellet was lyzed to see if the enzyme was completely secreted or still in the cells. SDS gel was run to follow the expression of the GluC protease. After cultivation, the culture was centrifuged at 4 °C at 6000 RPM for 20 min. The pellet was discarded and the supernatant was stored at 4 °C before the purification and activation steps.

1.6 Activation of V8-GluC

The V8-GluC harbours a prosequence on its amino-terminal side to avoid its autodegradation. Once this prosequence is cleaved, the V8-GluC is active. 1.6.1 Activation of the V8-GluC by addition of GluC

The prosequence was tried to be cleaved by the addition of other GluC as the last amino acid of the prosequence was mutated to an aspartate. 0.5 U, 1 U and 3 U from a 1 mg/mL stock solution of Roche or NEB GluC were added to 50 mL of the wild type V8- GluC supernatant after cultivation. Samples were incubated for 1 , 2, 3, 4 hours or 2 days at 25 °C. Enzymatic activity was measured according to the chapter 5.1 in Tris-HCl buffer (0.1 M) at pH 7.8 with the Z-Phe-Leu-Glu-pNA substrate. Remaining activity was calculated by referring to the first measured activity before starting the GluC addition. A positive control was used without GluC addition. In this case, sterile double distilled water was added instead of the GluC stock solution. 1.6.2 Activation of the V8-GluC by temperature incubation

To activate the prosequence cutting, 1.5-mL dialyzed samples of the purification step were pretested and incubated for several days at different temperatures: 25 °C, 30 °C and 37 °C. Enzymatic activity was measured in Tris-HCl buffer, pH 7.8 with the Z-Phe- Leu-Glu-pNA substrate and samples were sent to the ESI-MS department to determine the molecular weight once the enzymatic activity measurement was stable.

2. Enzymatic test assays 2.1 Enzymatic test assay with the glutamate substrate at pH 7.8 in Tris-HCl buffer

6.62 mg of Z-Phe-Leu-Glu-pNA substrate (10 mM) were dissolved into 1 mL of ethanol and heated at 50°C by shaking up to complete dissolution. Substrate solutions were incubated at room temperature for 10 min before starting the enzymatic test assay. 1 mg/mL of Roche GluC or NEB GluC lyophilisates were prepared with double distilled water as positive control. V8-glutamyl endopeptidases (Roche, NEB, wild type or mutant V8-GluC) were diluted with double distilled water up to get an absorbance of 20-30 mA/min during the spectrophotometer measurement. In this case, the Roche GluC was diluted 1000 times. 625 μΐ. of Tris-HCl (0,1 M) at pH 7,8 buffer were pipetted into the cuvette with 75 μί of substrate Z-Phe-Leu-Glu-pNA (10 mM) and 100 of the V8- Glutamyl endopeptidase enzyme. After mixing, the absorption was measured at 405 nm for 5 min by a spectrophotometer at 25 °C.

2.2 Enzymatic test assay with the hemoglobin hexapeptide substrate at pH 7.8 in Tris-HCl buffer The synthetic hemoglobin substrate was taken instead of the glutamate substrate.

The hexapeptide was synthetized in the Chemistry department of Roche Diagnostics GmbH. 8.148 mg of hemoglobin hexapeptide substrate were weighted and were simply mixed with 1 mL of double distilled water up to complete dissolution.

2.3 Enzymatic test assay with the hemoglobin hexapeptide substrate at pH 4.3 in ammonium acetate buffer

The enzymatic test assay follows the same protocol as described in Material & Methods. Enzymatic test assay were carried out with the hemoglobin hexapeptide substrate at pH 7.8 in Tris-HCl buffer. The 100 mM Tris-HCl buffer was replaced by 50 mM ammonium acetate buffer at pH 4.3. The pH was adjusted with acetic acid.

2.4 Enzyme activity calculations: Units (U), Units per volume (U/mL), Units per milligram of lyophilisate (U/mg lyo) An activity unit is usually defined as the enzyme amount required for catalysing one micromole substrate amount per min under fixed conditions (temperature, buffer, pH). The V8-glutamyl endopeptidase activity unit is the needed amount of V8-glutamyl endopeptidase (lyophilisate of Roche, NEB, wild-type or mutant GluCs), which turns one micromole of HbA0(l -6)-pNA or Z-Phe-Leu-Glu-pNA into HbA0(l-6) + paranitroaniline or Z-Phe-Leu-Glu + paranitroaniline per min, measured at 25°C and 405 nm either with 0.1 M Tris-HCl, pH 7.8 buffer or 50 mM ammonium acetate, pH 4.3. The reaction has stoechiometric coefficients and different activities were calculated according to the Lambert-Beer law. Positive control was checked at each measurement with the Roche GluC lyophilisate. Measurements were repeated twice. 2.5 Km-value m-value was assessed with the glutamate substrate in Tris-HCl buffer, pH 7.8. The V8-GluC activity measurement protocol in Tris-HCl buffer at pH 7.8 at the cuvette scale was followed as it is described in the chapter 5.1.1. Instead of 10 mM glutamate substrate, different concentrations were tested. The enzyme dilution was searched for getting an absorbance with a dA/min of about 0.06 at the substrate saturation. The 10 mM glutamate substrate concentration was then diluted by ½ step dilution until about 10 mA/min were reached. The measurement was repeated three times. Km-values were assessed according to the Lineweaver-Burk linearization graph and the value average was calculated as final Km-value (Lineweaver and Burk, 1934). 2.6 Temperature stress

The V8-glutamyl endopeptidase was incubated at different temperatures for 30 min in waterbaths. Different temperatures were tested: 25-55°C in 5°C steps. Then, the volume activity was measured at in Tris buffer, pH 7.8 at the cuvette scale. Positive control was measured at 25°C. The remaining activity average was calculated after reproducing twice the experiment.

2. 7 Hemoglobin functional test by ESI-MS

The functional test of the IFCC by ESI-MS was experimented with a ratio hemoglobin/enzyme of 100. Hemoglobin was not glycated (HbAO). The positive control, the Roche GluC, was freshly dissolved in double distilled water to a 0.2 mg/mL concentration. 50 of the Roche GluC stock solution (0.2 mg/mL) were then mixed with 2 mg/mL of hemoglobin and 50 mM of ammonium acetate buffer at pH 4.3 to get a final volume of 500 μL·. Other GluC NEB, wild type or mutant V8- GluCs) were prepared in a volume to get a final concentration of 0.02 mg/mL. The V8-GluC was mixed with 2 mg/mL of hemoglobin (final concentration). Ammonium acetate (50 mM) at pH 4.3 was added to reach a total reaction volume of 500 μί. Samples were incubated at 37°C for 18 hours. The samples were then frozen at -20°C before being injected onto the ESI-MS device in the analytics department. II. Results

Example 1

1.1 Activity test assay set up and acticity measurement

The GluC activity unit is defined as the needed amount of V8-glutamyl endopeptidase (GluC lyophilisate of Roche from Drapeau, the NEB glutamyl endoproteinase and the wildtype or mutant V8-GluC from Carmona), which turns one micromole of HbAO or glutamate substrate (HbA0(l -6)-pNA or Z-Phe-Leu-Glu-pNA) into HbA0(l -6) or Z-Phe-Leu-Glu + paranitroaniline per minute at 25°C and 405 nm, either with Tris-HCl 0, 1 M at pH 7,8 buffer or ammonium acetate buffer 0.050 M at pH 4.3. As the reaction has stoechiometric coefficients, the activity could be deduced from the Beer- Lambert law. As the Roche GluC lyophilisate is usually taken for the HbAl c test, its activity must be determined to use as a reference. The Roche GluC, the lyophilisate glutamyl endopeptidase from Drapeau, has a lyophilisate activity of 40 U/mg Lyo and a specific activity of 40 U/mg in 0.1 M Tris-HCl buffer at pH 7.8 with the Z-Phe-Leu-Glu- pNA substrate. As a summary, the enzymatic test assay could be correctly set up but only with defined experimental conditions, Tris-HCl buffer at pH 7.8 and with Z-Phe-Leu-Glu- p A as a substrate. The Roche GluC lyophilisate was used as positive control with a specific activity of 40 U/mg in this enzymatic test assay for all further activity test measurements of this phD project. However, the experimental conditions of the activity test differ from the HbAl C functional test.

The main enzymatic parameter of the screening system is the activity detection. This led to the maor problem of the high substrate consumption for the screening system. As the production of the hemoglobin as substrate cost too much, it could not be used for screening. The Z-Phe-Leu-Glu-pNA substrate was taken for screening. Alternatively, the screening system was set up with the Tris buffer at pH 7.8 as the Z-Phe-Leu-Glu-pNA substrate precipitates at acidic pHs, as for example in ammonium acetate pH 4.3. It was decided that if a mutant was found with a better activity in Tris buffer at pH 7.8 with the Z- Phe-Leu-Glu-pNA substrate at the microtiterplate scale, a test with the hemoglobin substrate will be carried out afterwards at the cuvette scale in the conditions of the mass spectrometry measurement: 50 mM ammonium acetate at pH 4.3 with the hemoglobin substrate. The enzymatic assay at the cuvette scale was adjusted to read an absorbance at about 60 mA/min at the microtiterplate scale. This absorbance was chosen to get a flexible range of activity to detect mutants with a higher activity. An absorbance of 40 mA/min was measured after transfer of the activity assay with the Roche GluC from the cuvette scale to the microtiterplate scale with the wild type V8-GluC.

1.2 Km-value

Variant libraries were also screened according to their affinity, the Km-value, towards the substrate. For the Km-value screening establishment, the right dilution of substrate was chosen to get about 30 % of the Km-value at the microtiterplate scale. The Km-value was assessed with the Z-Phe-Leu-Glu-pNA as a substrate in Tris buffer at pH 7.8 with the Roche GluC at the cuvette scale. On the Km curve, the reaction is observed at first to be proportional to the quantity of substrate: the more substrate is added in the reaction, the faster the enzyme catalyzes the substrate. Then, it gets the saturation point where the enzyme works with a maximal speed, Vmax. The Km value was calculated around 0.40-0.50 mM for the Roche GluC in these experimental conditions according to the lineweaver Burk plot. As the initial substrate concentration of the cuvette scale protocol was around 1 mM, its dilution of 1/8 was supposed to reach the 30 % of the Km value for the microtiterplate scale-down. However, these 30 % could be observed by the microtiterplate reader after a 1/6 dilution of the cuvette scale substrate concentration. This difference is due to the difference of scale from the cuvette to the microtiterplate. The absorbance is measured through the width of the cuvette whereas it is through the well thickness at the microtiterplate scale.

1.3 Temperature stress The temperature stability of an enzyme belongs to the enzymatic features, which can be affected severely after mutagenesis. This parameter must be therefore implemented in the screening system to control its possible fluctuations. Temperature stress was modelled by measuring the Roche GluC activity in Tris buffer at pH 7.8 with the Z-Phe- Leu-Glu-pNA substrate after 30 minutes of incubation at different temperatures in a range of 25 to 60°C with 5°C steps. 1 °C step was tried between 50 and 55°C to get a detailed view of the remaining activity of the Roche GluC remaining activity under 40 %. The Roche GluC is no more stable after incubation at 40°C and a remaining activity of 20 % is observed at a temperature of 51°C. After transfer from the cuvette to the microtiterplate scale, a remaining activity around 5-10 % was observed after incubation at 51 °C for 30 minutes. This difference is due to the incubation way. Temperature stress experiment was achieved in waterbath at the cuvette scale whereas it was performed in metal plates at the microtiterplate scale. As this measurement was also not uniform on the whole microtiterplate, the measurement was repeated three times to confirm the measurement repeatability. The incubation was performed at 51 °C for the temperature stress screening. 1.4 Specificity

As the V8-GluC can cleave aspartate and glutamate substrates, the possible specificity change of the enzyme should have been also checked during the screening. The specificity of the V8-GluC cleavage, with a aspartic acid synthetic substrate instead of a glutamic acid one, could not be also implemented in the screening assay because of substrate costs. The specificity was chosen to be tested with the hemoglobin functional test by ESI-MS afterwards.

As a conclusion of the screening system implementation, it was established on three enzymatic parameters of the wild type V8-GluC: the enzymatic activity, the Km-value and the temperature stability. The enzymatic activity was set up with an absorbance of 40 mA/min, the Km-value on 30 % of its value and the temperature stability on 5-10 % of remaining activity. They were however implemented with different buffer, pH and substrate conditions than the HbAl C functional test. Furthermore, the specificity parameter was decided to be tested afterwards on the hemglobin functional test by ESI-MS. 1.5 The mutant G 1661

The mutant, G166L, was found with a higher activity and a slight improvement in the Km-value during the screening of these different positions. The mutant G166I was generated from the G166L by site directed mutagenesis to observe the hydrophobicity effect of the leucine and isoleucine on the mutation. Screening of the G166 position was continued meanwhile by saturated mutagenesis. The G166I and another mutant, G166R, were discovered with a higher activity and better Km-value during this second screening. These three mutants G166I, G166L and G166R showed a repeatable higher activity at the microtiterplate scale.

The enzymatic features of the variants had to be determined to compare the effect of each mutation at this position. Enzymatic parameters could not be calculated at the screening scale. The three mutants G1661, G166L and G166R were cultivated in a bigger scale (50 mL culture) to check their specific activity at the cuvette scale. After cultivation, the mutant G166L was 18 times more active than the wild type GluC, the mutant G166I 32 times more and the mutant G166R 34 times more after 31.5 hours of cultivation. This factor was changed to 4.2 times more for the mutant G166L, 12.8 more for the mutant G166I and 12.4 more for the mutant G166R after 70 hours of cultivation. Km-values in Tris buffer at pH 7.5 shows a slight Km-value improvement of a factor 3.8 for the mutant G166I, compared to the Roche lyophilisate V8-GluC, whereas it is quite similar for the mutants G166I and G166R with a factor of 1.64 and 1.73 respectively. The hydrophobicity of the isoleucine has a relevant impact on the activity and affinity of the wild type V8- GluC. The mutant G166I was chosen for further characterization as it showed better activity and affinity at the cuvette scale.

1.6 Mutation of the prosequence The screening system cultivation implementation could have been more optimized consequently but this discovery does not affect the screening system. If a mutant is more active, its prosequence is faster digested and the improvement of the mutant is immediatly observed by the screening. However, hypothesis was ventured that the V8-GluC could completely autocatalyze its prosequence without addition of aureolysin, if the last amino acid of the prosequence, an asparagine, was substituted by a glutamic acid.

A mutation N-IE was introduced by site directed mutagenesis on the wild type and the mutant G166I to confirm this hypothesis. Wild type and mutant G166I V8-GluCs were designated E-Wt-H and E-M-H respectively. The H designates the polyhistidine tag cloning, the E, the N-I E mutation in the prosequence and wild type and G166I mutant GluCs are abbreviated by Wt and M.

According to this mutagenesis strategy, more than 60 amino acid positions were screened after directed evolution of the wild type V8-GluC gene. One relevant mutant came out from the library screening at the position 166. Saturated mutagenesis demonstrated the importance of the hydrophobicity at this position on the protease activity and affinity towards glutamate substrate at pH 7.8. The mutant G166I was generated with an additional mutation N-IE in the prosequence in order to facilitate the prosequence auto- hydrolysis and the activation of the V8-GluC mutant.

Example 2

2.1 Expression and cultivation The wild type V8-GluC was initially cloned in B. subiilis so that the protease could be expressed extracellularly and the screening could be facilitated. After screening, the different wild type and mutant V8-GluCs were expressed in B. subiilis with or without the N-1E mutation to accelerate the prosequence autoproteolysis. The effect of the polyhisdine tag presence was also studied during the expression. Cell growth and enzymatic activity were measured during the cultivation.

B. subtilis cells were first cultivated at a lower temperature, 30°C, in order to facilitate a correct protease folding during the protease expression. As the cell growth was too slow at 30°C, the cultivation was always performed at 37°C afterwards for the expression of the V8-GluC. Acetoin concentration variations did not interfere with the Bacillus subtilis growth and the wild type V8-GluC expression. The same concentration, 0.5 % of acetoin (10 %), was therefore always used. The cell growth of all these V8-Gluc variants reached an OD600nm plateau phase of about 4 after 24 hours of cultivation. The polyhistidine tag had no negative effect on the culture growth and on the activity of the GluCs. Wild type and G166I mutant GluCs activities were detected after 48 hours of cultivation. Interestingly, the activity continued increasing even if the cell growth had already reached a plateau. All these V8-GluC variants were expressed with the prosequence to avoid their own audigestion. Observed activities of the cultivation cannot be considered as the final activities as the prosequence was not properly cleaved. However, the N-1 E mutation of the prosequence caused a strong impact on the activation of the different V8- GluCs and therefore on the prosequence cleavage. The measured activity was multiplied by a factor of two for the wild type and mutant V8-GluCs, which harboured the N-1 E mutation of the prosequence.

The protein digestion of each expressed V8-GluC protease was followed during expression. The molecular weights were investigated by ESI-MS at different time points of the cultivation. Expressed protein sequences and the prosequence cleavage were deduced by their analysis. The results showed that, for each one given V8-GluC variant, different molecular weights could be found at different cultivation times. These fluctuations of molecular weights were representative of different protease forms and emphasized the hypothesis of degradation of the protease and/or the auto-hydrolysis of its prosequence. The intact E-Wt-H GluC sequence could be observed at a 0.50 U/mL.OD activity. The intact E-M-H GluC sequence was detected at a 0.10 U/mL.OD activity. The E-Wt-H and E-M-H GluC cultivations were always stopped with these similar activities, at the early stage of the cultivation to get the intact native protease. A complete reproducibility of the protease expression was however difficult.

According to the expression study of the V8-GluC variants, the presence of the polyhistidine tag did not interfere with the cell growth and protease expression. The N-1 E mutation was observed to have a strong impact on the prosequence. The wild type and G166I mutant GluCs were consequently expressed extracellularly from B. subtilis with the N-1E mutation and the polyhisdine tag. The E-Wt-H and E-M-H GluCs were collected at the early stages of the cultivation, as they were present under different digested forms after several days of expression. 2.2 Purification by Nickel-NTA affinity chromatography

After extracellular expression from B. sublilis, the E-Wt-H and E-M-H GluCs were purified for further analytical investigations. The E-Wt-H and E-M-H GluCs were purified on a Ni-NTA column as the polyhistidine tag was not completely digested. Purification of 10 mL culture sample was first pretested on Ni-NTA spin columns. As the E-Wt-H and E- M-H GluCs were found to bind onto these columns, the purification was scaled up to purify the cultivation rest of the E-Wt-H and E-M-H GluCs. A purity effect was observed with a factor of 4 for the E-Wt-H GluC and a factor of 3 for the E-M-H GluC. However, these yields and specific activities are not representative as the GluCs were not activated. The molecular weight measurement by ESI-MS confirmed that the prosequence was still present in the expressed proteases.

According to the molecular weight assessment by ESI-MS, in the first purification trial of the E-Wt-H GluC, 90 % of the purified E-Wt-H GluC were activated with the right sequence, [V40-D305] and a molecular weight of 29850 Da. 5 % had still a prosequence piece, [G26-H310], with a molecular weight of 30979 Da and the last 5 % of the purified E-Wt-H GluC were an autodigestion of the E-Wt-H GluC, [N44-H309] with a molecular weight of 28880 Da. In the second purification trial of the E-Wt-H GluC, 55 % of the purified E-Wt-H GluC was activated with the right sequence: [V40-D305] and a molecular weight of 29880 Da, the rest still contained the prosequence with missing amino acids at different positions. 32 % of the purified GluC had [H37-H31 1 ] or [A36-H312] as sequence for a molecular weight of 29 890 Da, 6 % had a [ 39-N303] or [V40P304] for sequence and a molecular weight of 28780 Da and the 5 % rest was a non determined mixture of E- Wt-H GluCs with some parts of the prosequence. The E-Wt-H purification did not show the same reproducibility because of the difficulty to stop the cultivation at the same activity point. Furthermore, the E-Wt-H was quite completely activated after the first purification and this could not be reproduced. The first purification trial of the E-Wt-H was an exception whereas the second one was more representative of the E-Wt-H GluC purification. After the E-M-H GluC purification, 100 % of the mutant GluC were found not to be activated and still contained the prosequence with [R35-H31 1 ] as sequence and a molecular weight of 30240 Da.

Some GluCs were lost during the flow through and wash steps. Some carboxyl- terminal digested side of the E-Wt-H or E-M-H GluCs were found in the flow through and the wash. The whole amount of E-WT-H and E-M-H GluCs, which were expressed, could not bind onto the Ni-NTA because of the carboxyl-terminal side hydrolysis. During this purification step, both E-Wt-H and E-M-H GluCs were purified by Ni-

NTA chromatography but they were obviously degraded during the purification. The activities and the purification balance were not representative as the E-Wt-H and E-M-H GluCs were not activated. The E-Wt-H and E-M-H GluC samples were dialyzed against a 50 mM Tris-HCl buffer pH 8 for long-term storage before further analysis such as the protease activation by prosequence cleavage.

2.3 Prosequence cleavage and glutamyl endopeptidase activation

The purified and dialyzed E-Wt-H and E-M-H GluCs were expressed as inactive with their prosequence. The aim was to avoid its toxic effect towards B. subtilis during its expression and to avoid the auto-hydrolysis of the protease itself. As the E-Wt-H and E-M- H GluCs had the N-IE mutation to facilitate the prosequence hydrolysis by autodigestion, it was also tried to activate it by incubation with other GluCs such as the Roche and NEB commercial lyophilisate GluCs. They were added at different concentrations of 0.5 U, 1 U and 3 U after incubation at 25°C for several days. Any activation could be observed after two days of incubation with these commercial V8-GluC lyophilisates. As some of the expressed GluCs were found under different protein forms during the ESI-MS analysis, it was assumed that the expressed GluCs were partially activated during the cultivation at 37°C. The protease activation could not be observed without the N-IE mutation during the cultivation. At the opposite, the E-Wt-H and E-M-H GluCs continued to auto-hydrolyze themselves on a long-term storage. The purified and dialyzed E-Wt-H and E-M-H GluC samples were consequently incubated for several days at different temperatures: 25°C, 30°C and 37°C to try to activate the prosequence auto- hydrolysis. Samples were incubated at 25°C in case of too rapid activation at the other temperatures. Results are shown in figure 46. The E-Wt-H GluC had a specific activity of about 45 U/mg after incubation for 2 days at 30°C or 37°C in the 1.5 mL pre-test sample. This specific activity was reproduced at 37°C at a larger scale with the rest of E-Wt-H GluC volume after the purification step. This purified and dialyzed E-Wt-H GluC sample contained 80 % of the following sequence, [V40-D305], with a molecular weight of 28880 Da and 20 % of the sample were digested under the [V40-D294] form with a molecular weight of 27730 Da after incubation at 37°C for 2 days. The major form of the E-Wt-H GluC was already autodigested at the carboxyl-terminal side, where the polyhisidine tag and the two last alanines were missing.

After activation in 1.5 mL pre-test sample, the E-M-H GluC reached a specific activity of 160 U/mg after incubation for 4 days at 37°C. The total volume of the purified and dialyzed E-M-H GluC was incubated according to this pre-test and after activation, the E-M-H GluC showed a specific activity of 190 U/mg after incubation for 2 days and 3.5 hours at 37°C. The specific activity of the E-M-H GluC changed from 160 to 190 U/mg. This fluctuation could be explained by the larger volume and therefore the amount of GluC, which was incubated. The greater amount of GluC was incubated, the more the GluC was activated. The E-M-H GluC contained 57 % of the [V40-D305] sequence with a molecular weight of 28940 Da, 27 % of the [V40-H313] sequence with a molecular weight of 29936 Da and 16 % of the [V40-H309] as sequence with a molecular weight of 29310 Da. The two last alanines of the carboxyl-terminal side were also digested as in the E-Wt-H GluC. The E-Wt-H and E-M-H GluCs are observed to be pure as much as the standards on the SDS-gel after purification and activation.

The E-Wt-H and E-M-H GluCs should have a molecular weight of 29846 and 29902 Da respectively with the intact sequence, [V40-A307], and the polyhisdine tag. This molecular weight should decrease to 29023 and 29079 Da for the complete sequence without polyhisidine tag. However, the majority of the E-Wt-H and E-M-H GluCs were activated under the [V40-D305] form with a molecular weight of 28880 Da and 28940 Da respectively. The SDS gel confirmed the presence of the autodigested E-Wt-H and E-M-H GluCs in the flow through of the Ni-NTA purification. The E-Wt-H and E-M-H GluCs were completely purified. The presence of autodigested forms were also observed in the eluted sample. The E-Wt-H GluC is strongly autodigested on this SDS gel. It could come from the continuous E-Wt-H GluC autocatalysis. This autodigestion was also observed for the E-M-H GluC. The first six amino acids from the amino-terminal side were sequenced and the amino-terminal sequencing confirmed that the prosequence was correctly cleaved. The first six amino acids of the active GluC sequence, VILPNN, were distinctly sequenced without additional amino acids of the prosequence.

As a conclusion, the E-Wt-H and E-M-H GluCs were activated by auto-hydrolysis of the prosequence. Even if they were present under degraded forms, only some histidines of the polyhistine tag and the two last carboxyl-terminal alanines were mainly missing. The specific activities of the E-Wt-H and E-M-H GluCs were assessed at 45 U/mg and 190 U/mg respectively in Tris buffer, pH 8 with the Z-Phe-Leu-Glu-pNA substrate.

Once the E-Wt-H and E-M-H GluCs were expressed, purified and activated, further biochemical and analytical characterizations were investigated. The Roche Lyo GluC, which comes from the Drapeau GluC, was taken as a positive control. The NEB GluC, which was discovered by Carmona, was used as a reference for the recombinant wild type GluC, E-Wt-H.

2.4 Storage stability at -20°C

The purified and activated E-Wt-H and E-M-H GluCs were stored before further investigations. Their stability was controlled for storage at -20°C. Storage stability at - 20°C was tested by freezing the E-Wt-H and E-M-H GluCs at -20°C overnight and defreezing them the day after. This experiment was repeated 5 times over one week, as samples were once frozen over the week-end. Their activity was measured after defreezing. 5 mM of Glu-Glu dipeptide were also added before freezing to check a potential stabilization effect. The E-Wt-H and E-M-H GluCs were stable through the freezing and defreezing over 1 week. The addition of the Glu-Glu dipeptide did not bring any effect. The liquid E-Wt-H and E-M-H GluCs could be stored at -20°C.

2.5 Specific activity in different buffers

After activation, the final specific activity of the E-Wt-H and E-M-H GluCs were calculated and compared to the other glutamyl endopeptidases. Specific activities of the E- Wt-H and E-M-H GluCs were assessed in different buffers and with different substrates. Tris buffer (Tris 0.1 M pH 7.8) and potassium phosphate buffer (KH2P04 0.1 M pH 7.8) were first tried out with the Roche substrate, Z-Pheu-Leu-Glu-pNA. Activities were also tested in potassium phosphate buffer (KH2P04 0.1 M pH 5) and ammonium acetate buffer (NH4-Ac. 50 mM pH 4.3) with the HbA0(l -6) pNA substrate. Outcomes are summed up in figure 48. The specific activities of the E-M-H GluC were compared to the Roche Lyo V8- GluC, the GluC of Drapeau. The activities of the E-Wt-H and NEB GluCs were also calculated as Carmona GluC reference.

The E-M-H GluC is 4 times more active in Tris buffer pH 7.8 with Z-Phe-Leu- Glu-p A as substrate than the Roche, NEB, E-Wt-H wild type GluCs. The E-M-H GluC is 4 times more active in ammonium acetate buffer at pH 4.3 with HbA0(l-6)-pNA as substrate than the Roche, NEB and E-Wt-H wild type GluCs. The E-M-H GluC is also around 2 times more active than the Roche, NEB and E-Wt-H wild type GluCs in potassium phosphate buffer at pH 5 but there is no activity difference in KH2P04 buffer at pH 7.8.

The E-M-H GluC has a higher specific activity than the Drapeau and Carmona GluCs whatever the buffer, pH or glutamate substrate is tested. This improvement is due to only one amino exchange of a glycine to a isoleucine at the position Gl 661. 2.6 Km-value

One other screening characteristic after the activity was the V8-GluC affinity towards the Z-Phe-Leu-Glu-pNA substrate. Once the E-Wt-H and E-M-H GluCs were purified and activated, their affinity could be defined by their Km-Value. The Km-values of the different GluCs, Roche, NEB, E-Wt-H and E-M-H, were assessed by the Lineweaver Burk plotting. The average was calculated after repeating the Km-value experiment 3 times as it is shown in Figure 2.

Roche, NEB and the E-Wt-H GluCs have a similar Km-value of 0.50 mM towards the Z-Phe-Leu-Glu-pNA substrate in Tris buffer at at pH 7.8. The E-M-H GluC Km-value is 0.17. The E-M-H GluC has 3 times more affinity towards the Z-Phe-Leu-Glu-pNA substrate in Tris buffer at pH 7.8. than the E-Wt-H GluC, NEB and Roche lyophilisate GluCs.

The Km-values of the E-M-H mutant GluC could not be determined with the hemoglobin hexapeptide substrate. However, the Roche GluC was not at the maximal speed, Vmax, when the hemoglobin hexapeptide was applied with the same amount as the protocol with the Z-Phe-Leu-Glu-pNA substrate. The hemoglobin hexapeptide was not in this case at the saturation concentration. The Roche V8-GluC is therefore less affine towards the hemoglobin hexapeptide than the Z-Phe-Leu-Glu-pNA substrate. The Roche V8-GluC Km-value should be higher towards the hemoglobin hexapeptide substrate than the Z-Phe-Leu-Glu-pNA one.

By the G166I mutation, the E-M-H GluC mutant presents more affinity towards the glutamate substrate in Tris buffer at pH 7.8. Its affinity could not be investigated with the hemoglobin hexapeptide substrate or at other pHs.

2.7 Temperature stability As mutagenesis could influence other enzymatic parameters of the E-Wt-H wild type GluC, the thermostability of the E-M-H GluC, was investigated. The stabilities of the E-Wt-H and E-M-H GluCs were tested stepwise from 25 to 55°C by GluC incubation for 30 minutes. 5 mM of Glu-Glu dipeptide were also added to control a potential stabilization effect. Results are plotted in Figure 3.

The E-M-H GluC is more stable than the E-Wt-H GluC over 50°C but any activity can be detected over 55°C for both GluCs. The stabilization effect of the Glu-Glu dipeptide cannot be observed neither on the E-Wt-H GluC nor on the E-M-H GluC. Both E- Wt-H and E-M-H GluCs are not stable anymore after 50°C and 55°C respectively.

2.8 Melting temperature assessment by DSC

The thermostability could be also checked by measuring the melting temperature by DSC (Differential Scanning Calorimetry). Differential Scanning Calorimetry (DSC) measures the amount of required energy to raise the temperature of a protein in solution. The protein unfolding due to heat denaturation, is correlated to this energy variation by referring to a standard. The unfolding transition is recognized as a sharp endothermic peak centered at the transition midpoint, Tm, also called melting temperature. This measured Tm indicates the protein thermostability. After the thermostability study of the E-Wt-H and E-M-H GluCs by incubation and measurement of the activity, the melting temperature of the E-Wt-H, E-M-H, Roche and NEB GluCs were also measured by DSC.

The melting temperatures of Roche GluC was assessed at 51.4°C, the NEB GluC at 52.33°C, the E-Wt-H at 52.4°C and the E-M-H GluC at 55.15°C. The Roche GluC is more sensitive to the temperature than the other GluCs due to the eleven amino acid position difference. The NEB GluC has only 0.1 °C difference with the E-Wt-H, which can be due to the two missing alanines at the carboxyl-terminal side of the E-Wt-H. The E-M- H mutant is more thermostable than the other GluCs. The melting temperature assessment confirmed the higher thermostability of the E-M-H GluC, which was observed during the temperature stability experiment. The mutant E-M-H GluC is slightly less susceptible to unfolding and denaturation.

2.9 pH optimum

The pH is an important enzymatic factor, which is strongly correlated to the enzyme activity. Very acidic or basic pHs can even affect the enzyme up to a complete loss of activity. The pH optimum is defined as the pH value, for which the enzyme activity is the most efficient. As the mutagenesis of the wild type V8-GluC at the position 166 improves its activity, the pH optimum may be also influenced. The pH optimum was determined in ammonium acetate (50 mM) between pH 4 and 6, KH2P04 (0.1 M) between pH 6 and 7.5 and Tris-HCl buffer (0.1 M) between pH 7.5 and 9 with the hemoglobin hexapeptide substrate. This substrate is the only one, which can be diluted in all buffers and which gives a direct correlation to the HbAO functional test by ES1-MS measurement. The pH optima of the E-Wt-H and E-M-H Glues are respectively at 7.8 and 7 with the hemoglobin hexapeptide substrate. The E-M-H GluC remains at its highest activity on a wide pH range from pH 6.5 to 8 whereas the E-Wt-H GluC is mainly at its highest activity only at one pH peak at 7.8. An activity loss of 20 % can be already observed at pH 7 for the E-Wt-H wild type GluC compared to the E-M-H mutant GluC. Only one amino exchange in the E-Wt-H GluC sequence, a glycine into isoleucine, can create a shift of the pH optimum from pH 7.8 for the E-Wt-H GluC to pH 7 for the E-M-H GluC. The E-M-H GluC pH optimum is extended to a wider pH range. Only one pH optimum peak for both E-Wt-H and E-M-H Glues is observed whereas two optimum pH peaks are usually present at pH 4 and 7.8 with the whole hemoglobin as substrate. The pH optimum of the E-M-H GluC could not be determined with the whole hemoglobin substrate to compare it with the Roche GluC. The pH optimum of the V8-GluC depends on the substrate. 2.10 Isoelectric point

Enzymes are characterized by their isoelectric point (pi). The isoelectric point, also abbreviated to IEP, is defined as the pH value at which the molecule carries no electrical charge or the negative and positive charges are equal. As the G166I mutation created the amino exchange of a glycine into an isoleucine, the isoelectric point of the E- M-H GluC may have been affected. The isoelectric points of the different possible GluC constructs were calculated theoretically by the bioinformatics program of Roche and are summarized in Table 1.

V8-GIUC Pro- N-1 E Wild type G166I 2 last His Pi sequenc mutatio mutan Alanines Tag

P-Wt X X 4.5753

PE-Wt X X X 4.5282

Table 1. Isoelectric points of different GluCs by bioinformatics. Legend: P: prosequence, E: N-JE mutation in the prosequence, Wt: Wild type, M: G166I mutant, H: polyhisidine tag, D: 2 last missing alanine at the carboxyl-terminal side. The G 166I and N- 1 E mutations do not interfere on the isoelectric point unlike the prosequence. The eleven different positions in the Roche GluC sequence give a more acidic isoelectric point of 4.1477 than the the NEB, E-Wt-H and E-M-H GluCs with a pi of 4.5955.

The wild type and mutant G166I with or without the N-1 E mutation and/or polyhisdine tag have the same pi. The whole prosequence causes a strong impact of 0.5 pH units on the isoelectric point. The N-1E mutation brings a slight effect on the GluCs of 0.1 pH unit. The polyhisidine tag adds 0.3 pH units more in the GluCs sequence in general. The NEB GluC has a similar isoelectric point as the recombinant E-Wt-H and E-M-H GluCs after suppression of the prosequence even if there are two missing alanines in the recombinants, as they do not play any role on the isoelectric point calculation. The eleven different positions in the Roche GluC sequence give a more acidic isoelectric point of 4.1477 than the the NEB, E-Wt-H and E-M-H GluCs with a pi of 4.5955.

The mutation G166I, unlike the prosequence autodigestion, does not influence the isoelectric point of the wild type V8-GluC. By contrast, the Roche GluC pi differs from the NEB, E-Wt-H and E-M-H GluCs because of its 1 1 amino acid exchange. This pi differences must be considered for experiments, which are based on this analytical feature, such as ion exchange chromatography for example. 2.11 Inhibitor

Enzyme activities can be affected by inhibitors. As the activity was improved in the E-M-H mutant GluC, its inhibition was also studied. The Pefabloc ^® Sc. (AEBSF: 4-(2- Aminoethyl)-benzenesulfonyl fluoreide hydrochloride) belongs to the family of sulfonyl fluorides, which irreversibly block serine proteases. 4 mM of the Pefabloc were incubated for 2 hours at 37°C with the E-Wt-H and E-M-H GluCs. The E-M-H GluC was two times more sensitive to the Pefabloc inhibitor than the E-Wt-H GluC. The E-M-H GluC lost half of its activity because of the inhibitor. The amino acid exchange of glycine into isoleucine at the position 166 of the wild type V8-GluC improved its activity, affinity and thermostability but on the other hand, the G166I mutation made the protease more sensitive to serine protease inhibitor.

2.12 Hemoglobin functional test

The functional test of the IFCC is the measurement of the glycated hemoglobin (HbAlc) by ESI-MS. The glycated hemoglobin is digested by the Roche GluC in specific experiment conditions in order to assess by ESI-MS the ratio between glycated and non- glycated hemoglobin hexapeptides. The hemoglobin was not glycated (HbAO) for the ESI- MS analysis.

2.12.1 Wild type V8-GluC The specificity pattern of the hemoglobin cleavage by the Roche V8-GluC was defined in order to compare it to the E-M-H GluC mutant. They were assessed after ESI- MS analysis. The alpha-chain of hemoglobin was digested into three residues by the Roche V8-GluC. The beta-chain was cleaved into ten residues The hexapeptide of the beta-chain was eluted with a molecular weight of 694.4 Da at a retention time of 1 1.6 minutes. The same pattern specificity was obtained with the E-M-H GluC. The G166I mutation of the E- M-H GluC does not affect the cleavage specificity of the hemoglobin in the functional test conditions. The HbAO test is usually performed under strict conditions after 18 hours incubation at 37°C (1 : 100, enzyme: substrate) in 50 mM ammonium acetate pH 4.3. As the substrate specificity is buffer dependent and pH and buffers have a strong impact on the activity of the glutamyl endopeptidase, the hemoglobin was incubated with the Roche GluC in different buffers, Tris buffer at pH 7.8, ammonium acetate at pH 4.3 and pH 7. The formation of the hexapeptide was compared to the initial amount of total hemoglobin by ESI-MS analysis.

According to the results obtained in the present invention, the Roche GluC is around two times more active and specific towards the hexapeptide in ammonium acetate buffer at pH 4.3. The hexapeptide is already completely cleaved after 5 hours of incubation whereas it did not start before 7 hours in Tris buffer at pH 8.5. The cleavage is however better in ammonium acetate buffer than Tris buffer. The conditions of the ESI-MS standard IFCC test could not be changed. Directed evolution emerged as the solution to generate a faster V8-GluC variant by keeping its other enzymatic parameters.

2.12.2 Mutant V8-GluC

First, the pH stability of the E-M-H GluC was compared to the Roche GluC in the IFCC conditions. HbAO functional test was then performed to test the specificity of the E-M-H GluC mutant compared to the Roche GluC. The E-M-H GluC mutant was incubated in the exact conditions of the IFCC protocols, for 18 hours at 37°C of the glutamyl endopeptidase with the haemoglobin in a ratio 1 : 100. The specificity towards the non-glycated hemoglobin was checked by ESI-MS measurement. The parameters of the IFCC protocols were also tested. Different buffers were also tried during the HbAO functional test. According to the previous specific activity study of the mutant, the ammonium acetate buffer was replaced by a potassium phosphate buffer at pH 5.5 or 7 and a Tris buffer at pH 7, all buffers with the same concentrations of 50 mM. The Roche and the E-M-H GluCs did not show any relevant activity loss after incubation at 37 °C for 18 hours in 50 mM ammonium acetate pH 4.3. They were both stable under these conditions. The E-M-H GluC seems to have consequently the same pH stability in ammonium acetate buffer pH 4.3 as the Roche GluC.

The cleavage specificity of the E-M-H GluC after the 18-hour incubation of the IFCC test remains the same as the Roche GluC, as described in the chapter 5.12.1. However, the 18-hour kinetic showed that the velocity of the hemoglobin hexapeptide cleavage was 3 times greater with the Roche GluC than the E-M-H GluC. The Roche GluC cleaves still faster the hemoglobin hexapeptide than the E-M-H GluC. The buffer exchange did not improve this result. The Roche GluC activity was already tested in Tris acetate buffers with 100 mM and 50 mM concentrations in previous experiments and any activity improvement could have been also observed. The E-M-H GluC mutant is as stable as the Roche GluC in the IFCC test. Its hemoglobin cleavage specificity is similar to the Roche GluC one. The cleavage speed of the hexapeptide is however still lower even if different buffer and pH conditions were tested. In summary, once the E-M-H mutant GluC was targeted by screening, it was expressed extracellularly from B. subtilis, purified and activated without major degradation. Further characterizations were then investigated. Its activity is 4 times better than the other GluCs in Tris buffer at pH 7.8 with Z-Phe-Leu-Glu-pNA substrate and ammonium acetate at pH 4.3 with the hemoglobin hexapeptide. Its affinity is improved with a factor 3 towards the Z-Phe-Leu-Glu-pNA substrate. Its pH optimum is extended up to pH 6. It is also more thermostable and its isoelectric point is not affected by the mutation. Its stability and specificity in the hemoglobin test does not differ compared to the Roche GluC. The E-M-H GluC mutant is however two times more sensitive to serine protease inhibitor and any velocity improvement could be observed in the hemoglobin hexapeptide cleavage.

References

Bunn, H.F. 1981. Modification of hemoglobin and other proteins by nonenzymatic glycosylation. Prog Clin Biol Res. 51 :223-39.

Brockmeier, U., Caspers, M., Freudl, R., et al. 2006. Systematic screening of all signal peptides from Bacillus subtilis: a powerful strategy in optimizing heterologous protein secretion in Gram- positive bacteria. J Mol Biol. 362:393-402.

Carmona, C, and Gray, G.L. 1987. Nucleotide sequence of the serine protease gene of

Staphylococcus aureus, strain V8. Nucleic Acids Res. 15:6757. DCCT. 1993. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N Engl J Med. 329:977-86.

Drapeau, G.R. 1978. Unusual COOH-terminal structure of staphylococcal protease. J Biol Chem. 253:5899-901 .

Gavin, J.R., Alberti, .G.M.M., Davidson, M.B., and al. 2003. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 26 Suppl 1 :S5-20.

Lineweaver, H and Burk, D. 1934. The Determination of Enzyme Dissociation Constants. Journal of the American Chemical Society.56: 658-666 Nakamoto, M. 2004. H-B-A- l -C what it is and why it matters. Diabetes Self Manag. 21 :84, 86-9.

Nathan, D.M., Turgeon, H., and Regan, S. 2007. Relationship between glycated haemoglobin levels and mean glucose levels over time. Diabetologia. 50:2239-44.

Prasad, L., Leduc, Y., Hayakawa, K., et al. 2004. The structure of a universally employed enzyme: V8 protease from Staphylococcus aureus. Acta Crystallogr D Biol Crystallogr. 60:256-9. Pearson and Lipman (1988), Proc. Natl. Acad. Sci.USA. 85: 2444.

Rawlings, N.D., F.R. Morton, C.Y. Kok, et al. 2008. MEROPS: the peptidase database. Nucleic Acids Res. 36:D320-5.

Silbersack, J., Jurgen, B., Hecker, M., et al. 2006. An acetoin-regulated expression system of Bacillus subtilis. Appl Microbiol Biotechnol. 73:895-903.