The subject of the invention is a method for producing insulin and derivatives thereof, as well as a hybrid peptide used in this method.

Diabetes (diabetes meliitus), a metabolic disease (a group of metabolic diseases) of multifactorial etiology, is one of the most dangerous diseases of society and civilization, also sometimes called the epidemic of the XXI century. Due to serious complications, diabetes can lead to death of the patient if left untreated. In the case when chronic hyperglycemia is caused by insufficient secretion of insulin by the body - type 1 diabetes - the therapy requires continuous use of insulin and/or its analogues. The number of diabetic patients is 5-6% of the population, and the annual increase in the number of diabetics is approx. 4-6% depending on the region of the world. The incidence of this disease continues to rise; it is considered a disease of civilization - the current lifestyle greatly contributes to the development of diabetes - for this reason, so far efforts to stop the development of this disease have failed. Also, despite medical advances, insulin and its fast-acting and long-acting analogues form the basis for treatment of type 1 diabetes (10% of the total number of cases of the disease). Insulin and various derivatives thereof used in large quantities in the treatment of diabetes are produced in a large industrial scale, and for this reason production methods are constantly improved in order to increase process yield and lower production costs. The economics of the production process is also important for the growing number of biosimilar insulin manufacturers (e.g.: S. Gough, Practical Diabetes 2013, 30 (4), 146-147a, G. Dranitsaris, E. Amir, K. Dorwart, Drugs 2011, 71 (12) 1527-1536], In the first years following the discovery of insulin, animal insulin secreted from the bovine or porcine pancreas was used as a human drug. Humanization of those insulins by replacement of the differing amino acids with the synthetic method was not applicable in industrial scale. It was not until the biosynthetic methods solved the problem of patient access to human insulin, and the first process was biotechnological generation of separate recombinant A and B chains of human insulin, and then assembling them into the same hormone as produced endogenously in the human pancreas (D. V. Goeddel et ai, Proc. Nat. Acad. Sci., 76: 106-110, 1979). Since the end of the last century, biosynthetic human insulin and analogues thereof, produced by the transformation of a single chain fusion protein, are the main source of this hormone as an exogenous drug. It is produced in recombinant bacterial systems, particularly in the cells of Escherichia coli (Polish patent No. PL 127 843), or yeast (L Thim et ai., Proc. Nat. Acad. Sci. USA, 83: 6766-6770, 1986). The method of producing these drugs is generally described in numerous patent specifications and scientific publications and is based on the overexpression of a gene encoding preproinsulin, i.e. a hybrid polypeptide consisting of a leader protein and proinsulin, i.e. the insulin B-chain (or derivative thereof), a linker peptide and the insulin A-chain (or derivative thereof). Once the fusion protein is isolated, the next manufacturing step is removal of the leader protein and the linker peptide, followed by isolation of the pure hormone. One example of such a process is a method of producing human insulin as described in Polish patent No. PL180818, and these methods are being continuously improved (e.g. Polish patents: PL167810, PL168114, PL177002, PL178466, PL180968, PL183284, PL191901, PL196626, PL198190, PL203195, PL203254, PL210437 and Polish patent applications P.309882, P.310007, P.320644, P.356005, P.374949, P.385586 and P.391975).

The biosynthetic hu man insulin and analogues thereof are the prima ry therapeutics in the treatment of type 1 diabetes and have significant value in the treatment of type 2 with different glucose control issues. Although the treatment with huma n insulin has brought satisfactory results in most cases for over 30 years, a number of analogues were created for improved metabolic control of diabetes d ue to their faster or prolonged action. For several years, the sales of insulin analogues have exceeded the half of the global insulin market, while patent protection of those drugs is beginning to expire. This situation allows entering the medica l markets by new manufacturers of such insulins as biosimilar drugs, in accordance with the regulations of the Europea n Union and the United States. For these reasons, efficient biosynthesis methods are becoming important, to provide high expression of the proteins of interest.

Polish patent No. PL180 818 describes an efficient method to produce recombinant human insulin using E. coli bacterial strain lacking the cytR repressor (described in Europea n patent application No. EP.0303972) transformed with a plasmid with deo operon containing the coding region for the hybrid protein (precursor) including a leader peptide, 62 amino acids long derived from the N-terminus of modified human superoxide dismutase (CuZnSOD) preceded by the N-terminal a mino acid Met and ending with the C-terminal Arg residue linking it to the insulin B-chain. The insu lin B-chain is linked to the A-chain with a short linker peptide composed of the known linkers Lys-Arg (European patent specification No. EP195 691) or Arg (European patent specification No. EP347 781).

The aim of the invention was to provide an efficient process for producing insu lin and analogues thereof that would exceed by several percent the yield of the process using the peptide C as a linker protein, as well as the hybrid proteins that are used in this method .

The subject of the invention is a polypeptide having the amino acid sequence of the formula:

X _n -B-Arg-Arg-A

where:

A is a polypeptide of the insulin A-chain or analogue thereof, prefera bly a sequence selected from SEQ. I D No.: 1-4.

B is a polypeptide of the insulin B-chain or analogue thereof, preferably a sequence selected from SEQ. I D No.: 5-7.

n is 0 or 1, X is a leader protein polypeptide, preferably a sequence selected from SOD of SEQ. No.: 8 or U BI of SEQ. No. : 9.

Preferably, the polypeptide of the invention has an amino acid sequence selected from : SEQ. I D No.: 10-27.

The polypeptide of the invention is intended for the production of human insulin by expressing the polypeptide in a strain of Escherichia coli lacking deo repressors and/or deo (cytR) gene containing DNA encoding said polypeptide, wherein said DNA is present in a vector under the control of deoP ₁ P ₂ promoter, while the hybrid polypeptide comprises a insulin B-chain covalentiy bound to the A-chain via Arg-Arg dipeptide.

Thus, a further subject of the invention is a method for prod ucing huma n insulin comprising microbia l expression of a recombinant protein containing insulin characterized in that it comprises:

(a) obtaining a hybrid polypeptide comprising proinsulin by expression of the hybrid polypeptide in a cell of an Escherichia coli strain lacking deo repressor and/or deo (cytR) gene containing DNA encoding the hybrid polypeptide, wherein the DNA is present in a vector under the control of promoter, while the hybrid polypeptide com prises the insulin B-chain covalentiy bound to

the A-chain via Arg-Arg dipeptide,

(b) recovering of the hybrid polypeptide,

(c) folding and formation of disulfide bonds within the hybrid polypeptide obtained in {b), avoiding prior subjection of the hybrid polypeptide to sulfitolysis,

(d) subjecting the hybrid polypeptide obtained in (c) folded, with disulfide bonds formed, to enzymatic digestion in order to produce insulin,

(e) purifying the insulin so produced.

The hu man insulin A-chain and the human insulin B-chain according to the invention is intended to mean both a polypeptide having the naturally occurring amino acid sequence and a seq uence of its known analogue.

Prefera bly, the hybrid peptide is a peptide of the invention as defined above.

Preferably, at the step (a) the expression is performed in E. coli strain of the genotype F ara Δ(pro lac) rpsL thi cytR recA λ~ devoid of active cytR repressor protein.

Preferably, at the step (a) a plasmid pIBA of the sequence SEQ I D NO: 28 is used as the vector.

Preferably, step (a) comprises fermentation in the presence of glucose, glycerol or galactose.

Preferably, step (b) comprises:

(i) destruction of the cell wall of the bacterial cell or its fragments to form a lysate, preferably by the action of lysozyme and Triton X100 followed by sonication or disintegration;

(ii) isolation of intracellular inclusion body precipitate from the lysate by centrifugation; and

(iii) dissolving the inclusion body precipitate. Prefera bly, step (c) comprises incubating the hybrid polypeptide, preferably in the presence of ascorbic acid, in particular at a concentration of about 2 moles per mole of SH groups present in the mixture, at a temperature of 4 to 37°C for about 1 to 30 hours, preferably about 5 hours, at a pH of 8.5 to 12, preferably at a pH of 11.0 to 11.25.

Preferably, the step (d) further com prises:

(i) adjusting the pH to about 8.8-9.0; and

(ii) digestion of the hybrid polypeptide with trypsin or optionally with deubiquitinating protease, e.g. UBPIAC or U BP1AC2, a nd subsequently optionally with carboxypeptidase B at a temperature of 16- 37°C for 30 minutes to 16 hours.

Preferably, step (e) fu rther com prises purifying the solution by low-pressure liquid chromatography on Sepharose Q prior to trypsin digestion.

Preferably, step (e) further comprises purifying the solution by low-pressure chromatography on DEAE-Sepharose prior to carboxypeptidase B digestion.

Preferably, step (e) fu rther comprises purifying the resulting insulin by high-performance liquid chromatography, preferably to a purity of at least 98%, followed by crystallization, filtration a nd drying.

Preferably, the insu lin purified at the step (e) is a protein selected from the group comprising: human insulin, human insulin Lys ^B28 Pro ⁸²⁵ (lispro insulin), insulin Gly ^A22 Arg ^B31 (GR insulin), human insulin Ser ^A22 Arg ^B31 (SR insulin), insulin Ser ^A22 Lys ^B3 Arg ^K1 (SK3 R insulin), proinsuiin B-ArgArg- A:dezAsn ^A21 Gly ^A21 .

Detailed description of the invention

As part of the search for high yield expression systems of insu!ins, studies have been planned on the effects of the type of linker peptide linking the A and B chains (proinsuiin) on the yield of insulin and its analogues in the gene system described in the Polish patent No. PL 180 818.

The recombinant preproinsulin according to the invention is understood as a connection of proinsuiin and an additional leader polypeptide, for example ubiquitin or superoxide dtsmutase

(SOD), or fragments thereof. In the light of recognized terminology, the recombinant proinsuiin is understood as a polypeptide chain, wherein the A and B chains of human insulin or an analogue thereof are connected to any (poly)peptide (linker), that allows them to fold by creating two disulphide bonds between the chains A and B and the third one - within the chain A.

The recombina nt insulin according to the invention is meant hu man insulin or its recombinant analog of hypoglycaemic activity.

According to the research results, the particularly efficient method to produce insulin and analogues thereof according to the invention, comprising the microbial overexpression of a gene encoding the recombinant hybrid polypeptide, comprising proinsuiin, in cells of Escherichia coli bacterial strains lacking the cytR repressor transformed with an appropriate vector with relevant gene inserted under the control of a deoPlP2 promoter, consists in that the proinsulin produced comprises the insulin B- chain covalently bound to the A-chain via Arg-Arg dipeptide as a linker peptide. In order to perform the studies, piBA plasmid has been prepared on the basis of patent specification No. PL 180 818, containing a region encoding the deoPlP2 promoter and a gene fragment of h uman superoxide dismutase (SOD, 62 aa) or ubiquitin (UBI) as described in U.S. patents 8,158,382 and 8,956,848, as well as by A. Wojtowicz er a/., M icrobial Cell Fact. 2005, 4: 17 and Microb Cell Fact. 2014; 13 (1): 113. Further, a fragment encoding the relevant proinsulin is in the same reading frame - optionally modified B-chain, the linker peptide and optionally modified A-chain. The regulatory elements of the plasmid are derived from pBR322 vector (ATCC 31344). The deoPlP2 promoter is derived from the bacterial deo operon composed of four closely related genes that regulate nucleotide and deoxynucleotide catabolism in Escherichia coli bacteria; its sequence was amplified from chromosomal DNA of E. coli K12 strain based on the nucleotide sequence from gene database [GenBank: AP009048]. Based on the strain of E. coli CSH50 {ATCC: 39111; Cheng et al., Gene, 14 Suppl. 1-2: 121-130, 1981), the E. coli I BA strain was derived with the F ara Δ(pro lac) rpsL thi cytR recA A ^" genotype, devoid of active cytR repressor protein by introducing a mutation into the gene using the method described in European patent application No. EP0303972. After transformation of E. coli I BA host strain with the pI BA plasmid, a system was obtained as described in Polish patent PL 180 818 to express the encoded fusion protein in a pIBAINS vector under the control of the deoPlP2 promoter, which is negatively controlled by a chromosomal repressor protein, the product of cytR gene (K. Hammer-Jesperson et al., M olec. Genet. 1975; 137: 327-335). The construction of plasmids used elements-DNA fragments encoding various leader proteins and various insulin derivatives, wherein the (pro)insulin A-chain was linked to the insulin B-chain via various linker peptides - examples of which are given below.

The production yield of insulin and its analogues in this expression system and with different linker peptides was com pared with expression of these hormones with other strains of E. coli and a vector which is derived from the pIGAL plasmid (Gene Bank AY 424310), investigating in particular the relationship between the yield and the type of peptide linker.

The linker peptide sequences used were:

1. Lys-Arg

2. Arg

3. Arg-Arg

4. C-peptide

According to the invention the thoroughly investigated insulin a nd its derivatives-analogues were: recombinant human insulin (Polish patent specification No. PL128 599) Lys Pro h uman insulin (lispro insulin, Polish patent specification No. PL180 968) Gly ^A22 Arg ⁸³¹ human insu lin; (Polish patent specification No. PL219335 or US 08,618,048, GR insulin), Ser ^A22 Arg ^B31 human insulin (description of Polish patent application P.219335, SR insulin),

Ser ^A22 Lys ^B3 Arg ^B31 human insulin (description of Polish patent application P.399287, SK3R insulin) and preproinsulin (precursor) G: leader protein - B-chain - linker - A-chain: dezAsnA21GlyA21.

The £ coli IBA was used as the E. coli host strain to study the yield; the essence of its properties is the inactive gene of cytR repressive protein, which is disclosed a nd described in the literature (e.g. Miller J H., Experiments in molecular genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 1972) or in U.S. patent description 4,480,038 or Polish patent description PL 180 818 for E. coli strains deposited under accession numbers ATCC 69361 and 69363. Other strains widely used in research - E. coli DH5, E. coli DHSa!pha. and E. coli HB101 - were used for comparative purposes.

The pIBA vector - deoPlP2 promoter, tetracycline resistance, terA Trp tra nscription terminator - was used as the plasmid for protein expression in the yield studies, but for comparative purposes also other vectors were used : pIGALl derivative (Gene Bank AY 424310) - pms promoter, ampicillin resistance, ST1 transcription terminator; pIGALl derivative - pms promoter, tetracycline resistance, terA Trp transcription terminator and pIGALl derivative - deo P1P2 promoter, tetracycline resistance, ST1 transcription terminator. A modified fragment of SOD, as in Polish patent specification No. PL 180 818 and a modified ubiquitin (UBI), as described in U.S. patent specification 8,158,382 and 8,956,848, were primarily used as the leader protein in the yield studies. In the present description, the terms SOD and UBI represent compounds described respectively in these patent specifications. The process for producing each of the insulin which is the subject of the present application is carried out with a classical method of genetic engineering and biotechnology. It consists in that a suitable strain is constructed, which in a process of biosynthesis (fermentation) produces the desired hybrid polypeptide (relevant preproinsulin), which is then transformed into the desired product, purified and isolated.

The system is universal to produce recombinant human insulin and its studied - slightly different - analogues. For this pu rpose modifications of the recombinant hu man preproinsulin gene were constructed using genetic techniques, such as site-specific mutagenesis reaction . Point mutagenesis reaction was carried out using a kit from Stratagene (Cat. No. 200518-5); plasmid DNA of pIBA a nd pIGAL plasmids was used as the template. Also any other DNA comprising the relevant sequence coding for recombined human proinsulin or preproinsulin can be used as the template.

For each insulin plasmid constructs - vectors - were prepared that encoded the studied (pre)proinsulin with a linker (poly) peptide, and elements-DNA fragments encoding various proinsu lins were used for the plasmid construction - consisting of various linker peptides and of various insulin derivatives. These vectors were used to transform competent cells of a suitable Escherichia coli strain, for exam ple IBA, DH5a, DH5 or HBlOl, and it is possible to use cells of other E. coli strains or cells of other microorganisms, or other known cell lines suita ble for recombinant protein expression. The plasmid containing the specific gene modification of recombinant human preproinsulin was isolated and sequenced to verify the correct transformation a nd nucleotide sequence. Single clones of the transformed strains were typically cultured in appropriate medium supplemented with selection antibiotic to prepare the bacterial material for the resea rch ban k, and the samples of bacterial cu ltures and 40% glycerol in the ratio 1:1 were deposited at -70°C. Ail research bank strains were subjected to extensive microbiological, biochemical and physicochemical testing, which confirmed their stability and compatibility of their properties with the requirements applicable to industrial banks. The course of subsequent biosynthesis was based on Polish patent specification No. PL 180 818. Following biosynthesis, the variants of the recombinant preproinsulin produced in £ coli stra ins were isolated after cell disruption as inclusion bodies, which were separated. Hybrid polypeptide with insulin or analogue thereof (corresponding preproinsulin) was isolated from the inclusion bodies, and then subjected to folding (renaturation). The resulting solution of folded hybrid protein was subjected to the controlled action of trypsin, similarly as in the case of a number of methods known and described previously (e.g. by Kem mler et al., J. Biol. Chem., Vol. 246, pp. 6786-6791 (1971} or patent specifications US 6,686,177 or US 6,100,376), or deubiquitinating protease, as described in patent specifications US 8,158,382 and US 8,956,848, as well as by A. Wojtowicz et al., M icrobial Cell Fact. 2005, 4: 17 and M icrob Cell Fact. 2014; 13 (1): 113, or both of these enzymes. In some cases (hu man insulin, lispro insulin) additional amino acids of the linker peptide were removed by carboxypeptidase B action; in other cases, this step was omitted. The resulting insulin and its analogues were subjected to a process of purification using known methods, mainly low-pressure chromatography, ultrafiltration, and HPLC. From the sufficiently pu rified - meeting pharmacopoeia l requirements - solution of insulin or its analogue, the product was crystallized and dried.

In order to compare the expression yield for each host and each variant of the system, the biosynthesis, isolation and purification of the product was developed and optimized for each strain. The biosynthesis of Escherichia coli strains lacking the cytR repressor transformed with a suitable vector with relevant gene inserted under the control of appropriate deoPlPl promoter was carried out according to the description of Polish patent specification No. PL 180 818, in fermenters of 7.5 dm ³ in volume, using approx. 3-5% of the inoculation material a nd minimum, industrial medium for inoculum and production culture and controlling the induction of a desired product with glucose, according to example 2 of Polish patent specification No. PL 180 818. The biosynthesis of other E. coli strains was performed in accordance with the general methods of their culture. I noculum culture was ca rried out for approx. 10 hours, until the end of the logarithmic phase of bacterial growth, and production culture for a pprox. 16 hours. The subsequent process of product separation, tra nsformation, and purification was also conducted as described in Polish patent No. PL 180 818 - optimized example 2 a nd 3. I n particular, on completion of the production biosynthesis the resulting broth was cooled to approx. 10°C, £ coli cells were centrifuged, subjected to action of lysozyme a nd Triton X100, and then sonicated with ultrasound or disintegrated with pressure at 5-10°C. The inclusion bodies were centrifuged again, washed, and then subjected to renaturation, as genera lly described in exam ple 3A of the Polish patent PL 180 818. Except for the proteins obtained according to Example 6, the further optimized process consisted in subjecting the solution to low-pressure liquid chromatography on DEAE Sepharose following citraconylation and digestion with trypsin or digestion with deubiquitinating protease a nd trypsin. After decitraconylation and precipitation and dissolution of the protein, the solution was subjected to another low-pressure liquid chromatography using-Sepharose Q resin, followed by optional digestion of the linker peptide with carboxypeptidase B. After the final high-performance chromatography on Kromasil C8-18, the product purified at least to 98% purity was subjected to crysta llization, filtration and drying. Detailed description of the entire process, com pletely com parable between specific insu lins, is given in the exam ples. The process was performed in a pilot scale, using the methods and devices used in industry. After process optimization, three more batches of each of insulin were carried out to study the yield of the process, and the presented results are the average of each of them.

To compare the expression yield, at a fixed production process, analytical tests of five operations were selected from the unified process control:

(1) optical density (OD ₆₀₀ ) of completed production culture;

(2) weight of the culture (isolated bacteria), expressed in g per 1 dm ³ of production culture {wet and/or dry);

(3) weight of the isolated inclusion bodies, expressed in g per 1 dm ³ of production culture (wet and/or dry or as dry weight of the slurry in a buffer);

(4) total protein (after precursor folding) as determined by a spectrophotometric method a nd expressed in AU/dm ³ or H PLC and expressed as % folded precursor in the total protein;

(5) weight of the purified final product expressed in mg per 1 dm ³ of production culture.

In the case of lispro insulin the analytical results of most steps are given - total protein determined by spectrophotometry and expressed in AU/dm ³ and the Bradford method and expressed in grams per 1 dm ³ of production culture.

A general block diagram of the production of insulin and its derivatives is shown in Scheme 1.

The invention discloses that the final yield of biosynthetic human insulin and its analogues, using the optimized technological process of biosynthesis, secretion, transformation, a nd purification, depending on the degree of overexpression of the gene encoding preproinsulin in the modified Escherichia coli cells (with inactive cytR repressors that negatively regulate deo operon} transformed with a specific plasmid (with deoPlP2 promoter) surprisingly depends primarily on the peptide linker used, which links the A and B chains of insulin.

According to the invention it was surprisingly found that the highest yield in this system is obtained when the peptide lin ker is Arg-Arg dipeptide.

Through the invention, specific conditions for preparation of human insulin and its analogues by genetic engineering are given, using the microbial overexpression of the relevant preproinsu lin in the cells of Escherichia coli strains devoid of deo repressors and/or deo gene, containing plasmid DNA coding for an appropriate hybrid polypeptide in a vector under the control of deoPlP2 promoter at the optimum performance - significantly increased when compared to the published data.

In the following em bodiments of the invention the structu re, yield and purity of all products were routinely determined by known methods of HPLC and UV spectrophotometry; to fully cha racterize the produced insulins, n umerous modern methods and analytical techniques were used, including MS, CIEF, LC-MS/MS, GC, peptide mapping, NMR, crystallography and powder X-ray diffraction, CD, FT1R, ELISA or PCR and sequencing of amino acids and nucleotides.

In order to better explain the principle of the invention, the present description is supplemented with a detailed discussion of the exemplary embodiments, including the drawings, showing the general structure (Fig. 1) and the sequence (Fig. 2) of the pIBA plasmid, both ready for insertion of a hybrid protein gene and as an exemplary embodiment in the form of plasmid containing a gene encoding the insulin hybrid protein (Fig. 3 and Fig. 4).

No detailed descriptions of conventiona l bioengineering methods used in the construction of specific genes, plasmids, vectors with inserted genes encoding studied preproinsuiins or methods of transformation of bacterial strains with these vectors, were given in the examples, as well as no methods for testing the stability of these strains. Such methods are well known to those skilled in the art and used in biotechnological laboratories, as well as described in numerous publications, of which the key ones are provided herein. The final products produced with method according to the invention are known insulins, most of the specific bioengineering methods for the construction of these recombinant bacterial strains are disclosed in the Polish patent PL1S0968 and Polish patent applications P.385586 and P.3992S7.

No detailed descriptions of standard a nalytical methods used for determination of content and concentrations of proteins and insulins at the specific manufacturing steps are given, since these methods are also well known to relevant skilled in the art and routinely used for protein determination. The following examples, not limiting the scope of the invention, describe in detail the research methodology that allows in a fully controlled way to compare the obtained representative research results on the relationship between the process yield and the structure of the linker peptide.

Examples

I. General procedure of the process

Step 1. Research Bank - primary culture

Primary cultures of E. coli strain I BA, DH5 DHSalpha a nd H8101 containing the pI BA vector and pIGALl derivative as protein expression plasmids (Gene Bank AY 424310) were performed according to the requirements ( ICH Topic Q 5 D: Quality of Biotech nologica! Products: Derivation and Characterization of Cell su bstrates Used for Prod uction of Biotechnological/Biological Products and ICH Topic Q 5 B: Qua lity of BiotechnologicaI Products: Analysis of the Expression Construct in Cell Lines Used for Production of r-DNA Derived Protein Products) in the basal medium :

The basal medium:

- casein hydrolyzate 20 g

- yeast extract 10 g

- NaCl 5.0 g

- water to a volume of 1 dm ³

- a ntibiotic (tetracycline or ampici!lin) to a final concentration of 12.5 μg/cm ³ for TET or 100 μg/cm ³ for Amp

The flask with the basal medium and antibiotic was inoculated with a single bacterial colony obtained after transformation a nd shaken (220 rpm) at 37°C to an optical density OD ₆₀₀ of approx. 1.

The resu lting culture was mixed with the freezing medium (1:1) and frozen at -70°C in 1 ml aliquots.

For the resulting bacterial culture the following assays were performed:

• measurement of the optical density OD of the culture at 600 n m;

• checking the homogeneity of bacterial culture by preparation of live specimen.

Bacterial culture purity, morphological characteristics - appearance, shape of the cells - were evaluated. Observation of the specimen using Olympus CX 40 microscope;

• calculating the most probable number of colony forming units on the T5A medium and T5A with a specific antibiotic - cfu/mi;

• checking microbiological purity of the culture using Sabouraud media (to identify fungal contamination) and TSA (to identify contamination by other bacteria).

Step 2. Inoculum

The inoculum was prepared in flasks of 250-300 ml in volume filled with 50 ml of inoculum medium. To each flask the following sterile solutions were added:

• antibiotic (tetracycline or am picillin, 12.5 mg/ml) 50 μΙ

• glucose 50% 500 μΙ

Additionally added to the cultures of f. coli 1BA were:

• proline (100 mg/ml) 300 μΙ

• thiamine (50 mg/ml) 2-5 μΙ.

After inoculation of the flasks with primary culture (working bank), the inoculum was incubated at 30-37°C for 16-18 hou rs to OD ₆₀₀ a pprox. 0.7-1.0 on a shaker at 180-200 rpm. The inoculum medium (components for 1 dm ³ ):

Step 3. Production culture (NewBrunswIck BioFlo 310 fermenter, 7.5 dm ³ {starting volume 4 dm ³ )

Initial cultu re parameters:

• Temperature: 37°C

• pH: 7.1 ± 0.01

• DO: 30-35%

• rotations: 250 rpm

• air flow: 5 dm ³ /min.

To the sterile fermenter with 4 dm ³ of the production medium and antifoaming agent Antifoam 204 (0.5-1 ml, Sigma) sterile solutions were added :

• 6 g of MgS0 ₄ x 7H ₂ O in 80 cm ³ of water

• 0.2 g CaCl ₂ x 2H ₂ O in 60 cm ³ of water

• glucose 50%, 4 cm ³

Additionally added to the E. coli \ BA cu ltures:

• proline (100 mg/ml} 24 cm ³

• thiamine (50 mg/ml) 200-500 μΙ.

The prod uction medium :

Medium in the fermenter is inoculated with inoculum and cultured in appropriate conditions. 40% glucose solution, as a source of carbon and energy for biomass prod uction, is continuously fed into the fermenter (supplemented with proline 10-30 mg/ml in the case of E. coli I BA) so that its concentration is 70-180 mg/dm ³ .

pH (7.1 +/- 0.1) is controlled by cascade using 16% ammonia. The oxygen level in the medium is maintained at a level not lower than 30-35%. This parameter is controlled depending on the reading from DO electrode by the increase in impeller speed from 250 to 1000 rpm and in air flow from 5 to 10 dm ³ /min. After reaching the maximum speed of the impeller and the maximum air flow (approx. 7-10 hours) the addition of glucose is stopped; OD ₆₀₀ value (in the case of E. coli I BA/plBA) is approx. 25-35.

After a decline in glucose concentration, its addition is maintained so that the concentration is approx. 35-50 mg/dm ³ and the pH was kept at 7.1 +/- 0.1. This phase of the biosynthesis is carried out until the cells reach the stationary growth phase (increase in optical density of the culture is no longer observed) and lasts approx. 7-8 hours. At the end of the recombinant protein production phase OD ₆₀₀ is 50-60 in the case of E. coli IBA/plBA strains. The culture is then cooled to a temperature below 20°C, preferably approx. 10°C.

Step 4. Centrifugation

The biomass is centrifuged using a fixed centrifuge at 6,000-8,000 rpm for 15 minutes and at a tem perature of 4°C. The fermented broth may be centrifuged using a flow centrifuge.

Step 5. Isolation of inclusion bodies

The biomass is resuspended in 1 dm ³ of 50 m M Tris buffer, 0.5 M NaCl and 5 mM beta- mercaptoethanol, pH 7.5, and lysozyme is added to a concentration of 0.43 mg/cm ³ and 0,2 M EDTA to a concentration of 0.5%, then incubated for 30 minutes at room tem perature and subjected to disintegration by sonication or pressure.

Step 6 - option 1: Sonication.

The material is cooled to approx. 10°C, divided into 4 parts, Triton is added to 1% and each portion is sonicated for 30 min utes with the amplitude of 33%. The sonicated sample is centrifuged at 8,000 rpm for 15 minutes at room temperature. The peliet is then resuspended in 1 dm ³ of 50 mM Tris buffer, 0.5 M NaCl and 5 m M beta-merca ptoethanol, pH 7.5 and sonicated again for 10 minutes. The inclusion bodies are centrifuged at 8,000 rpm for 15 minutes at room temperature.

Step 6 - option 2: Pressure disintegration.

The material is cooled to approx. 10°C, subjected to pressure disintegration (Panda+lOOO, GEA Niro Soavi) at a pressure of 800 bar at a rate that ensures total destruction of the cells (microscopic inspection), and cooling the suspension so that the temperature does not exceed approx. 10°C. Step 7. Dissolution of inclusion bodies

The inclusion bodies were dissolved in adequate amount of 12 mM bicarbonate buffer and 0.2 mM EDTA to obta in absorbance values below 9 at a wavelength of 280 nm. The solution is adjusted to a pH va lue of 11.5 ± 0.5 with 2 M NaOH and stirred for 30 minutes at room temperature. Then the pH is adjusted to 10.8 ± 0.4 with 2 M HC1, The solution is clarified by centrifugation at 12,500 rpm for 15 minutes at 4°C.

Step S. Renaturation

After dissolving the inclusion bodies, the protein is renatured by aeration - vigorous stirring for approx. 16-18 h at room temperature. The pH is then adjusted to pH 9.0 with 2 M HCl.

Step 9. Citraconylation

Citraconic anhydride is added portionwise to the protein solution in an amount caicuiated according to the formula : anhydride volume (cm ³ ) = 0.15 x solution volume [dm ³ ] x A ₂₈₀ and 2 M NaOH to maintain the pH in the range of 8.7-9.3. After adding the entire amount of the anhydride, the solution is stirred for 1 hour, then solution of 2 M ethanola mine in the volume of three times anhydride volume is added, followed by stirring for 30 minutes.

Step 10. Trypsination

After citraconylation the pH of the solution is adjusted to a value of 8.8 with 2 M HC! and diluted with water so that the conductivity reaches less than 5 mS. Then 1% trypsin solution is added in an amount calculated according to the formula:

volume of trypsin = A ₂₈₀ x volume of a solution of [dm ³ ] / 95)

and stirred for 16-18 hours at room temperature. The reaction is inhibited with aprotinin solution with a concentration of 1 mg/mi, added in an amount 21-times less than the volume of trypsin.

Step 11. Cutting off the leader protein - ubiquitin.

20 mM phosphoric acid is added to the protein solution at a temperature of 10°C to obtain a pH value of 7.5, and then at 37°C 19.2 cm ³ U BPIΔC protease solution is added at a concentration of 1 mg/cm ³ and stirred for 1 hour. After cooling to 4°C the solution is clarified by centrifugation at 12,000 rpm for 15 minutes.

Step 12. The low-pressure chromatography on a bed of DEAE Sepharose

The solution at pH 8.6 ± 0.2 is loaded onto a column packed with DEAE resin (200 cm ³ ) a nd eq uilibrated with 0.5 M Tris, pH 8.6 ± 0.2 followed by 20 mM Tris, pH 8.6 ± 0.2. After loading, the column is washed with buffers at pH 8.6 + 0.2: 20 mM Tris with NaCl until conductivity is 5 + 1 mS, and then 20 mM Tris with NaCl until conductivity is 2 ± 1 mS and with 20 ± 5% isopropanoi. The insulin protein is eluted with a buffer at pH 8.6 ± 0.2, consisting of 20 m M Tris with NaCi until conductivity is 5 ± 1 mS and supplemented with 25 + 5% isopropanoi. Step 13. Decitraconylation

The main fraction e!uted from the colum n is diluted 2 times, cooled to 4°C and acidified with 0.1 M HCI to a pH of 2.9 ± 0.2, then stirred 10 hours.

Step 14. Precipitation of insulin with zinc chloride.

A volume of 1 M zinc chloride, which corresponds to 1/50 of the volume of the main fraction, is added to the diluted main fraction, pH adjusted to 5.5 ± 0.5 and stirred for 1 hour. The suspension is then centrifuged at 9,000 rpm for 15 minutes at 4°C. The resulting pellet is suspended in water. Step 15. Low-pressure chromatography on Sepharose Q resin.

The insu lin suspension is supplemented with water to a volume calculated according to the formula: total volume = total amount of protein obtained after DEAE [AU A ₂₈₀ ] / 7, Tris pH 8.6 is added to a concentration of 30 m M and 0.2 M EDTA solution is added in an amount calculated according to the formula: volume [cm ³ ] = total amount of protein obtained after DEAE [AU A ₂₈₀ ] / 190, and then the pH is adjusted to 8.6. The solution of dissolved zinc salt of insulin at pH 8,6 is loaded onto a column packed with the Q resin (SO cm ³ ) and equilibrated with 0.5 M Tris, pH 8,6 and then 20 mM Tris, pH 8.6. After loading, the column is washed with 20 mM Tris buffer with 20 ± 5% isopropanol, pH 8.6, and then the insulin protein eluted with a buffer at pH 8.6 consisting of 20 mM TRIS with NaCI until conductivity is 3 ± 2 mS and with 25 ± 5% isopropanoi.

Step 16. Reaction with carboxypeptidase B.

The main fraction eluted from the column is diluted two times and a solution of carboxypeptidase B at concentration of 2.5 mg/cm ³ is added in an amount calculated according to the formula: volu me [μm ³ ] = total amount of protein obtained from Q. [AU A ₂₈₀ ] / 15. pH of the solution is adjusted to 8.8 a nd the solution is stirred for 16-18 hours at room temperature. The pH is then adjusted to 3 and the solution is purified by high-pressure chromatography.

Step 17. High-pressure chromatography RP-HPLC.

The separation is carried out on a Kromasil C8 or C18 column using an acetonitriie gradient:

Phase A Phase B

95% 5%

75% 25%

Mobile phase A: 2,000 cm ³ of a 0.2 M sodium sulfate solution with 440 cm ³ of acetonitriie, pH 2.3 and a temperature of 25-30°C;

Mobile phase B: 1,250 cm ³ of a 0.2 M sodium sulfate solution with 1,250 cm ³ of acetonitriie, pH 2.3 and a temperatu re of 25-30°C;

The solution after reaction with carboxypeptidase B or a main fraction after purification on the Q Sepharose resin, following a double dilution and adjustment of the pH to 3, is purified in portions by loading onto the column a single volume of solution having a total protein content of 200 AU, and then the main fraction is collected, so that product purity (HPLC) is at least 98%.

Step 18. Precipitation of insulin with zinc chloride.

The combined main fractions from the high-pressure chromatography are diluted 2 times and 1 M ZnCI ₂ is added in a proportion of 3 cm ³ of zinc chloride per 150 cm ³ of the sample, and then the pH is adjusted to 6 + 1, followed by stirring for 1 hour. The suspension is then centrifuged at 9,000 rpm for 15 minutes at 4°C and the pellet is resuspended in 10-15 cm ³ of water.

Step 19. Chromatography on a bed Sephadex G-25 and freeze-drying.

The slurry is adjusted to pH 3, filtered through a 0.22 μm filter and loaded onto a column filled with Sephadex G 25 resin (120 cm ³ ). Before loading the column is equilibrated with 5 m M ammonium acetate, pH 4, and the protein is eluted with 5 m M ammonium acetate, pH 4. The resulting material is subjected to freeze-drying.

II. Specific examples

Example 1. LysB28ProB29 human insulin (lispro insulin)

According to the general procedure of the process given above (section I ), the production of biosynthetic lispro insulin was performed following steps 1-6, 6b, 7-10 and 12-19 a nd for the expression system 1.2. and 1.5. additionally step 11 (Table 1. Lispro insulin). The process was studied in 20 variants of 8 protein expression systems. The data obtained shows that the strains of Escherichia coli DH5a, DH5 and H B101, which perform well in laboratory conditions, with piasmids with deoPlP2 and pms promoters are not suitable for production culture in the minimal medium (system 1.3., 1.4., 1.5,, 1.6. and 1.7.). These strains are characterized by low growth on minimal technological medium, demonstrating a much lower optical density value in the production fermenters. A similarly low feasibility for production is shown by the IBA strain system with disrupted cytR repressor and pIGAL piasmid (system 1.8.), characterized by relatively better growth, but low content of inclusion bodies (microscopic observations).

The best expression yield is shown by the I BA/pl BA strain system, described in Polish patent specification No. PL 180 818, which was tested in 4 linker peptide variants and 2 leader protein variants (system 1.1. and 1.2.). All have good, comparable growth parameters in the minimal medium, but differ in weight of inclusion bodies and relative yield of the fusion peptide renaturation, and - consequently - the final product yield. The highest yield of lispro insulin was obtained using the E. coli IBA/pl BA expression system producing A-chain linked to the B-chain of lispro insulin with ArgArg dipeptide (Table 1, Example No. 1.1.1. and 1.2.1.}, exceeding the performance of other systems by at least 13-15%.

Example 2 GlyA22ArgB31 insulin (insulin GR).

According to the general procedure of the process given above (section I), the production of biosynthetic GR insulin was performed following steps 1-6, 6b, 7-10, 12-15 and 17-19 and for the expression system 2.2. additionally step 11 (Table 2, GR insulin}. The process was studied in 12 variants of 5 protein expression systems (not using the variant with LysArg linker, as is does not lead to G R insulin). The data obtained confirms the findings of Example 1 related to the superiority of the expression system described in the Polish patent specification No. PL ISO 818 over the other systems studies and shows similar differences in the yield of the process depending on type of the linker peptide that links the A and B chains of this insulin analogue. In particular, the highest yield of GR insulin was obtained using the E. coli I BA/pl BA expression system, producing A-chain linked to the B- chain of G R insulin with ArgArg dipeptide (Table 2, Example No. 2.1.1. and 2.2.1.) and exceeding the performance of other systems by at least 16-22%. The presented results show technological advantage of ArgArg linker regardless of the type of the leader protein.

Example 3. Human insulin.

According to the general procedure of the process given above (section I), the prod uction of biosynthetic human insulin was performed in 8 va riants of 2 protein expression systems. Steps 1-6, 6b, 7-10 and 12-19 and for the expression system 1.2. and 1.5. additionally step 11 were used in the studies. The data obtained shows (Table 3. H uman insulin), that the IBA/plBA strains system is characterized by the best expression yield, showing good, com parable growth parameters in the industrial minimal medium and differing in the weight of inclusion bodies produced and the relative yield of renaturation. Again the highest yield of human insulin was obtained using the E. coli IBA/plBA expression system producing A-chain linked to the B-chain of human insulin with ArgArg dipeptide (Table 3 Example No. 3.1.1. and 3.2.1.). The ana lytical studies performed at different steps of the process indicate, that the higher yield in these expression system va riants - by about 14-16% - is reached after the step of inclusion body preparation and renaturation, and it is kept along all other steps of the process.

Example 4. SerA22ArgB31 human insulin (SR insulin).

Studies of the yield of the SR insulin production process were performed according to the general procedure given above (section I), following steps 1-6, 6b, 7-10, 12-15 and 17-19 and for the expression system 4.2. additionally step 11 (Table 2. GR insulin). The process was studied in 12 variants of 5 protein expression systems (not using the variant with LysArg linker, as is does not lead to SR insulin). The data obtained supports the previous results, constantly showing a higher yield of the E. coli IBA/pl BA expression system, producing A-chain linked to the B-chain of SR insulin with ArgArg dipeptide (Table 4 Exa mple No. 4.1.1. and 4.2.1.) in comparison with other linker peptides of this system - by at least 15%.

Example 5. SerA22LizB3ArgB31 insulin (SK3R insulin)

According to the general procedure of the process given above, the production of biosynthetic SK3R insulin were performed in 6 variants of 2 protein expression systems. Steps 1-6, 6b, 7-10, 12-15 and 17-19 and for the expression system 5.2. additionally step 11 were used in the studies. The data obtained shows (Table 5. SK3R insulin) that the IBA/plBA strains system is characterized by the best expression yield, showing good, comparable growth parameters in the industrial minimal medium and differing in the weight of inclusion bodies produced and the relative yield of renaturation . Again the highest yield of human insulin was obtained using the E. coli IBA/plBA expression system producing A-chain linked to the B-chain of human insulin with ArgArg dipeptide (Table 5, Example No. 5.1.1. and 5.2.1.). The extensive analytical studies performed at different steps of the process indicate, that the higher yield in these expression system variants is kept along all other steps of the process after the step of inclusion body preparation and renaturation, the yield being higher than that obtained with other expression system variants by at least 14-16%.

Example 6. Preproinsulin (precursor) G: leader protein - B-chain - linker - A-chain:

dezAsnA21GlyA21.

The versatility of the ArgArg linker advantage over others for preproinsulin expression was also demonstrated by studying the yield of a protein consisting of a leader protein preceding the B-chain, which was linked with the studied linkers to a modified A-chain of insulin (A: dezAsnA21G lyA21). The studies were performed using critical steps 1-8 of the general procedure of the process in 8 variants of 2 expression systems. The research results, presented in Table 6. Preproinsulin (precursor) G: leader protein - B-chain - linker - A-chain: dezAsnA21G!yA21 show that in the case of the ArgArg linker the yield of the process is 10% higher than in the other cases.

Based on the final product yield, confirmed by the protein content before purification, it was unexpected ly found that the best resu lts in the production of recombinant human insulin a nd analogues thereof are obtained in an expression system consisting of an expression plasmid with deoPlP2 promoter in bacterial strains of Escherichia coli lacking the cytR repressor gene (Polish patent No. PL 180 818) if in the plasmid encoding the fusion protein the A-chain of insulin is linked to the B-chain with ArgArg dipeptide. Moreover, the difference in performance between the ArgArg dipeptide variant of the system and LyzArg is 13-22%, and is comparable to the improvement in system performance described in the Polish patent No. PL 180 818 for LizArg dipeptide and Arg in comparison with the C-peptide, which is 23% and 17%, respectively.

Genetic construction of plBA/INSLys(31B)Arg(32B} pIBA/INS plasmid

The pIBA/I NS plasmid of 4354 base pairs in size is composed of the following regulatory sequences and genes:

- from 7 bp to 942 bp a regulatory region with deoPlP2 promoters is located,

- from 946 bp to 1137 bp there is a sequence encoding a fragment of human SOD gene,

- from 1138 bp to 1299 bp a sequence encoding the A a nd B chains, as well as the C chain (in the form of Lys-Arg linker) of recom binant human insulin INS is contained, the nucleotide sequence of insulin has been changed in accordance with the frequency of codon usage in E. coli.

- from 1312 bp to 1338 bp a region coding for Ter trpA transcription terminator is located,

- from 1508 bp to 2698 bp there is a tetracycline resistance gene Tet.

The structure of the pIBA plasmid is schematically shown in Figure 1, and its nucleotide sequence and amino acid sequence in Figure 2. The structure of the pI BA/INS plasmid comprising an exemplary gene encoding recombinant human insulin protein, wherein the A-chain is linked to the B-chain with LysArg dipeptide is schematically shown in Figure 3, and its nucleotide sequence and amino acid sequence in Figure 4.