WO2004013175A1

WO2004013175A1 - SYNTHETIC GENE CODING FOR HUMAN GRANULOCYTE-COLONY STIMULATING FACTOR FOR THE EXPRESSION IN E. coli

Info

Publication number: WO2004013175A1
Application number: PCT/EP2003/008308
Authority: WO
Inventors: Simona JEVSÈVAR; Viktor Menart
Original assignee: Lek Pharmaceuticals D.D.
Priority date: 2002-07-31
Filing date: 2003-07-28
Publication date: 2004-02-12
Also published as: US7655437B2; AR040521A1; EP1527095A1; AU2003253344A1; JP4445466B2; US20060228781A1; SI21272A; JP2006512089A

Abstract

The invention relates to the synthetic gene coding for hG-CSF which enables expression in E. coli with an improved expression level of the recombinant hG-CSF regarding the total cellular proteins after expression

Description

Title of the invention

Synthetic gene coding for human granulocyte-colony stimulating factor for the expression in E. coli

Field of the invention

The present invention relates to synthetic gene coding for human granulocyte- colony stimulating factor (hG-CSF) which enables expression in E. coli with an improved expression level, enabling an expression level being equal to or higher than 52% of the recombinant hG-CSF to the total proteins after expression. hG-CSF belongs to a family of stimulating factors which regulate the differentiation and proliferation of hematopoetic mammalian cells. They have a major role in the neutrophil formation and are therefore suitable for use in medicine in the field of hematology and oncology.

Two forms of hG-CSF are currently available for clinical use on the market lenograstim which is glycosylated and is obtained by the expression in mammalian cell line and filgrastim which is non-glycosylated and is obtained by the expression in the bacterium Escherichia coli (E. coli).

Background of the invention

The impact of several successive rare codons such as arginine codons (AGG/AGA; CGA), leucine codon (CTA), isoleucine codon (ATA) and proline codon (CCC), on the level of translation and consecutively on the decrease of the amount and quality of the expressed protein in E. coli are described in Kane JF, Current Opinion in Biotechnology, 6:494-500 (1995). There is a similar impact of individual rare codons if they occur in different parts of the gene.

The GC rich regions also have impact on the translational efficiency in E. coli if a stable double stranded RNA is formed in the mRNA secondary structure. This impact is the highest when the GC rich regions of mRNA are found either in the RBS, or in the direct proximity of the RBS or als^'o in the direct proximity of the start codon (Makrides SC, Microbiological Reviews, 60:512-538 (1996); Baneyx F, Current Opinion in Biotechnology, 10:411-421 (1999)). There are known several methods for the prediction of the secondary structure and calculating minimal free energy of individual RNA molecule which is supposed to be the basic rule for the most stable / most probable structure (SantaLucia J Jr and Turner DH, Biopolymers, 44:309-319 (1997)). The reliable algorithms for the prediction of the correct secondary structure are not known with the exception of some cases. There has been no evidence for the quantitative correlation with the expression level (Smit MH and van Duin JJ. Mol. Biol., 244, 144-150 (1994)). It is still impossible to predict the tertiary structures of RNA (Tinoco I and Bustamante C, J. Mol. Biol, 293:271-281 (1999)).

The increase of the expression level after the optimization of DNA sequence in the TIR region, in the RBS region and in the region between the start codon and the RBS region is described in McCarthy JEG and Brimacombe R, Trends Genet 10:402- 407 (1994). In this case the expression level increased due to more efficient translation initiation and its smooth continuation in the mRNA coding region.

The production of adequate amounts of hG-CSF for performing the in vitro biological studies by expression in E. coli is described in Souza LM et al, Science 232:61-65 (1986) and in Zsebo KM et al, Immunobiology 172:175-184 (1986). The hG-CSF expression level was lower than 1 %.

The patent US4810643 discloses the use of synthetic gene coding for hG-CSF which was first of all constructed on the basis of replacement of E. coli rare codons with the E. coli preference codons. The combination with thermoinducible pfiage lambda promoter led to the expression level of 3 to 5% of hG-CSF regarding the total cellular proteins. This level is not sufficient for the economical large-scale production of hG-CSF.

8-10% accumulation of hG-CSF to total cellular proteins was reached by changing the first four codons in the 5' end region of hG-CSF as is described in Wingfield P et al, Biochem. J, 256:213-218 (1988).

The expression of hG-CSF^"in E. coli with the yield up to 17% of hG-CSF to total cellular bacterial proteins is described in Devlin PE et al, Gene 65:13-22 (1988). Such yield was reached with partial optimization of DNA sequence in the 5' end of the G-CSF coding region (codons coding for the first four amino acids) whereby the GC region was replaced with AT region and a relatively strong lambda phage promoter was used. This expression level is not very high what leads to lower production yields and is less economical in the large-scale production.

The use of synthetic gene and the expression level of about 30% are described in Kang SH et al, Biotechnology letters, 17(7):687-692 (1995). This level was attained by the introduction of E. coli preference codons, by the modifications in the TIR region and with the additional modifications of codon sets whereby the 3' end of the gene was not essentially changed. Thus, for attaining the stated expression level the changes of the gene in the TIR region were needed and the expression level did not exceed 30%.

The patent US5840543 describes the synthetic gene coding for hG-CSF which was constructed by the introduction of AT rich regions at the 5' end of the gene and with the replacement of E. coli rare codons with E. coli preference codons. Under the control of the Trp promoter the expression with the yield of 11 % hG-CSF to total cellular proteins was reached. On the other hand, the addition of leucine and threonine or their combination into the fermentation medium (where the bacteria were cultivated) led to the accumulation of up to 35% of hG-CSF regarding total cellular proteins. Such expression level was therefore reached by the addition of amino acids into the fermentation medium what is an additional cost in the process for production of hG-CSF and is not economical for the industrial production. Only optimization of the gene coding for hG-CSF did not enable a higher expression level of hG-CSF.

The highest accumulation of hG-CSF regarding total cellular proteins found in the prior art is described in v Jeong et al, Protein Expression and Purification 23, :311-318 (2001 ) and is 48%. Such accumulation was obtained by the changes in the N-terminal end and by the induction with 1 m IPTG.

In general, there are no reports on possible predictions of the expression level of native human genes in prokaryotic organisms, e.g. bacterium E. coli. The described expression levels are relatively low or difficult to detect even when the expression plasmids with strong promoters, e.g. from lambda or T7 phage are used. From the prior art literature it can be gathered that many parameters (rare codons or their clustering; GC base pairs rich regions, unfavorable mRNA secondary structures, unstable mRNA) have an impact on the accumulation of a human protein in E. coli. Until now there has been no entirely developed rule known on how to combine the codons in order to obtain the secondary or tertiary mRNA structures which are optimal for expression. Although there exist some mathematical and structural models for predicting and thermodynamical stability of secondary structures, but they are too unreliable to predict the secondary structures. On the other hand, there are no such models for predicting the tertiary structures. These currently accessible models therefore do not enable the prediction of the impact of the codons on the expression level.

There are no reports in either the patent or the scientific literature on the more efficient way for solving the problem of low expression level of the native gene coding for hG-CSF in E. coli.

Summary of the invention

It is thus an object of the present invention to provide a DNA sequence coding for hG-CSF or biologically active G-CSF, which DNA sequence enables an improved expression level (accumulation) in E. coli, and to provide a process for the construction of such a DNA sequence.

The object is solved by a DNA sequence according to claim 1 , and by a process for the construction of such a DNA sequence according to claim 15. The present invention also provides an expression plasmid according to claim 6 or 7, an expression system according to claim 11 or 12, a process for the expression of hG- CSF according to claim 20 and a process for the manufacture of a pharmaceutical composition according to claim 24. Preferred embodiments are defined in sub-claims.

The significant feature of the present invention is that the use of synthetic gene coding for hG-CSF enables to attain an expression level (accumulation) in E. coli being equal to or higher than 52% of recombinant hG-CSF regarding the total proteins in E. coli. Preferably, an expression plasmid containing a strong T7 promoter is used for the expression. The synthetic gene coding for hG-CSF is constructed by using a complex combination of two methods which enable the construction of optimized synthetic gene (coding for hG-CSF) for its expression in E. coli. The first nethod includes the replacement of some rare E. coli codons which are unfavorable for expression in E. coli by E. coli preference codons for which are more favorable for the expression in E. coli. The second method includes the replacement of some GC rich regions by AT rich regions. Some parts of the synthetic gene of the present invention are constructed by using one of the two methods, for some parts the combination of the two methods is used, whereas some parts of the gene are not changed, i n the construction procedure of the synthetic gene coding for hG-CSF, which is also the subject of the present invention, the non coding (5'-untranslated) regions are preferably not changed. Advantageously, this means that there are no modifications in either the translation initiation region (TIR) or in the ribosome binding site (RBS), or in the region between the start codon and RBS.

Brief description of the drawings:

Figure 1 schematically shows an optimized construction of a synthetic gene coding for hG-CSF according to a preferred embodiment of the present invention.

Figure 2 shows the DNA sequence of the native gene sequence coding for hG-CSF (Fig. 2A) (GenBank: NM_000759) and the DNA sequence of the optimized (Foptδ) gene coding for hG-CSF (Fig. 2B). The bases which differ from native gene are bolded.

Figure 3 shows an SDS-PAGE analysis of samples of proteins obtained from the expression of native hG-CSF DNA sequence (lanes 1 to 4) and of optimized (Foptδ) gene coding for hG-CSF (lanes 6 and 7) in induced and noninduced cultures of E. coli, as evaluated by dye staining (Fig. 3A) and by Western blot using antibody specific for hG-CSF protein (Fig. 3B).

Figure 4 shows an SDS-PAGE analysis of samples of proteins obtained from the expression of optimized (Foptδ) gene coding for hG-CSF in induced culture of E. coli, as evaluated by dye staining.

Figure 5 shows an SDS-PAGE analysis of samples of proteins obtained from the expression of optimized (Fopt5) gene coding for hG-CSF in induced culture of E. coli according to an alternative embodiment, as evaluated by dye staining

Description of the invention and the preferred embodiments thereof It has been found that the problem with the low expression level of the gene coding for hG-CSF in E. coli can be solved by the optimization of the gene sequence coding for hG-CSF. The native gene coding for hG-CSF is changed, leading to the construction of a particular synthetic gene coding for hG-CSF. The particular synthetic gene is defined by the DNA sequence of SEQ ID NO: 1 or by a nucleotide sequence comprising suitable modifications of SEQ ID NO: 1 or of the native hG- CSF gene sequence.

In comparison with the data described in the art, surprisingly high expression level can be obtained according to the present invention.

The term 'hG-CSF', as used herein, refers to human granulocyte-colony stimulating factor, comprising the recombinant hG-CSF obtained by the expression in E. coli.

The synthetic gene encoding hG-CSF of the present invention was obtained by introducing changes in the nucleotide sequence of the gene encoding the native hG-CSF. Thus the amino acid sequence was not changed and remained identical to the native hG-CSF.

The present invention further comprises a process for the expression of the synthetic gene in E. coli and concerns the level of expression of the synthetic gene.

The term 'expression level', as used herein, refers to the proportion of hG-CSF obtained after the heterologous expression of the gene encoding hG-CSF regarding the total cellular proteins after expression. The expression level may be quantified from the quantification of appropriately separated proteins after expression, e.g. quantifying the staining of protein bands separated by SDS-PAGE.

The term 'heterologous expression', as used herein, refers to the expression of the genes which are foreign to the organism in which the expression occurs.

The term 'homologous expression', as used herein, refers to the expression of the genes which are proper to the organism in which the expression occurs.

The term 'preference codons', as used herein, refers to the codons used by an individual organism (e.g. E. coli) for the production of most mRNA molecules. The organism uses these codons for expressing genes with high homologous expression. The term 'rare codons' as used herein, refers to the codons used by an individual organism (e.g. E. coli) only for expressing genes with low expression level. These codons are rarely used in the organism (low homologous expression).

The term 'GC rich regions', as used herein, refers to the regions in the gene where the bases guanine (G) and cytosine (C) prevail.

The term 'AT rich regions', as used herein, refers to the regions in the gene, where the bases adenine (A) and thymine (T) prevail.

The term 'synthetic gene', as used herein, refers to the gene prepared from short double stranded DNA fragments which are composed of synthetic complementary oligonucleotides. This synthetic gene differs from the native gene (e.g., cDNA) only in the nucleotide sequence whereby the amino acid sequence remains unchanged. The synthetic gene is obtained by the techniques of the recombinant DNA technology.

The term 'native gene', as used herein, refers to the DNA sequence of a gene which is identical to the native DNA sequence.

The term 'segment', as used herein, refers to the parts of the genes which are bounded by single restriction sites on both ends. These sites serve as subcloning sites for the synthetically constructed parts of the gene. In the following the restrictions sites are numbered according to the nucleotide position in the 5'-3' direction from the start codon.

The term 'segment I', as used herein, refers to the 5' end of the gene encoding hG-CSF between the nucleotide positions 3 and 194 (notably the restriction sites Ndel (3) and Sacl (194)), i.e. 191 bp long sequence. Segment I may be de novo synthesized.

The term 'segment II', as used herein, refers to the part of the gene for hG- CSF between the nucleotide positions 194 and 309 (notably the restriction sites Sacl (194) and Apal (309)), i.e. 115 bp long central part of the gene. Segment II may be de novo synthesized. ^~ "

The term 'segment III', as used herein, refers to the part of the gene for hG- CSF between the nucleotide positions 309 and 467 (notably the restriction sites Apal (309) and Nhel (467)), i.e. 158 bp long part of the gene where the native DNA sequence for hG-CSF is preserved with the exception of codons for Arg148 and Gly1 50.

The term 'segment IV, as used herein, refers to the 3' terminal end of the gene encoding hG-CSF between the nucleotide positions 467 and 536 (notably the restriction sites Nhel (467) and BamHI (536)), i.e. 69 bp long terminal part of the gene. Segment IV may be de novo synthesized.

The synthetic gene encoding hG-CSF of the present invention is constructed by the combination of the following methods:

• replacement of E. coli rare codons with E. coli preference codons: in the segment II (between restriction sites Sacl (194) and Apal (309)) and in the segment IV (between restriction sites Nhel (467) and BamHI (536))

• replacement of GC rich regions with AT rich regions, whereby the rarest E. coli codons are replaced, but mostly not with the E. coli preference codons: in the segment I (between restriction sites Ndel (3) and Sacl (194)).

• completely unchanged native sequence of 46 codons (between CCC for Pro102 and CGC for Arg147) in the segment III.

• replacement of two E. coli rare codons (CGG->CGT (Arg148) and GGA→GGT (Gly150)) at the terminal end of the segment 111.

Optimization of the gene coding for hG-CSF of the present invention does not include changes in the TIR, RBS and in the regions between the start codons and RBS.

The synthetic gene of the present invention encoding hG-CSF enables expression of the constructed synthetic gene encoding hG-CSF with the expression level in E. coli equal to or higher than 52%. Furthermore, the expression level of about 55% or even about 60% can also be obtained. High expression level of the synthetic gene coding for hG-CSF of the present invention enables high yields of hG- CSF production, faster and simpler purification and isolation of heterologous hG- CSF, easier in-process control, and the whole production process is - more economical. Therefore, the efficient production of hG-CSF in industrial scale is enabled. The produced hG-CSF is suitable for clinical use in medicine.

The construction of the synthetic gene of the present invention begins with the initial preparation of the hG-CSF native gene and of the plasmids. Gene coding for native hG-CSF can be of human origin, but the same principle can be used for every gene which is homologous in the regions which comprise single restriction sites which are used for subcloning of de novo synthesized gene segments. The plasmid for mutagenesis was chosen according to its ability to be capable of enabling the successive introduction of point mutations. Selection or enrichment of the piasmids containing desired mutation was obtained by using an additional selection primer that changed unique restriction site EcoRI into EcoRV or vice-versa (Transformer™ Site- Directed Mutagenesis Kit (Clontech)). The gene and the plasmid are constructed in such a way that the introduction of point mutation by cassette mutagenesis is possible.

After the initial preparation of native gene coding for hG-CSF and of piasmids the optimization of the native gene coding for hG-CSF is performed. This means that the synthetic gene coding for hG-CSF is constructed. The optimization begins with the division of the native gene coding for hG-CSF into four (I, II, III in IV) segments, which are or will be separated with single restriction sites after the oligonucleotide mutagenesis and in the individual segments the changes are introduced. In some individual segments the changes in the gene sequence are introduced whereas in certain segments the gene is not changed (Figure 1 ). The obtained optimized synthetic gene coding for hG-CSF therefore consists of partially preserved native sequence (segment III) and of 5' and 3" coding regions which are synthesized de novo (segments I, II and IV).

The changes in the individual segments: Segment I: Replacement of E. coli rare codons with E. coli preference codons and replacement of GC rich regions with AT rich regions

Italic: GC/AT rich replacement; Italic and underlined: rare/preference codon replacements and GC/AT rich replacement; underlined: rare/preference codon replacements; Gly101 (GGT→GGG) introduction of Apal (309) restriction site. Thr2 (ACC→ACA), Pro3 (CCC→CCA), Gly5 (GGC→GGT) Pro6 (CCT→CCA), Ala7 (GCC→GCT), Ser8 (AGC→TCT), Ser9 (TCC→TCT), Pro11 (CCC→CCG), Gln12 (CAG→CAA), Phe14 (TTC→TTT), Leu16 (CTC→TTG), Lvs17 (AAG →AAA). Cys18 (TGC→TGT), Glu20 (GAG→GAA). Val22 (GTG →GTT), Arq23 (AGG→CGT), Lys24 (AAG→AAA) He25 (ATC→ATT), Gln26 (CAG→CAA), Gly27 (GGC→GGT), Gly29 (GGC→GGT), Ala31 (GCG→GCT), Leu32 (CTC→TTA), Gln33 (CAG→CAA), Glu34 CGAG→GAA), Lys35 (AAG→AAA), Ala38 (GCC→GCA),^' Thr39 (ACC→ACT), Tyr40 (TAC→TAT), Lvs41 (AAG→AAA). Cys43 (TGC→TGT), His44 (CAC→CAT), Pro45 (CCC→CCA), Glu46 (GAG→GAA). Glu47 (GAG→GAA). Val49 (GTG→GTT), Leu51 (CTC→TTA), Glv52 (GGA→GGT , His53 (CAC→CAT), Gly56 (GGC→GGT), He57 (ATC→ATT), Pro58 (CCβ→CCG), Pro61 (CCC→CCT)

Segment II: Replacement of E. coli rare codons with E. coli preference codons. Cys65 (TGC→TGT), Pro66 (CCC→CCG), Ala69 (GCC→GCG), Leu76 (TTG→CTG), Leu79 (CTC→CTG), Gly82 (GGC→GGT), Leu83 (CTT→CTG), Phe84 (TTC→TTT), Leu85 (CTC→CTG), Tyr86 (TAC→TAT), Glv88 (GGG→GGTl Leu89 (CTC→CTG), Ala92 (GCC→GCG), Glv95 (GGG→GGC), Ile96 (ATA→ATT), Pro98 (CCC→CCG), Glu99 (GAG→GAA), Leu 100 fTTG→CTG), Gly101 (GGT→GGG)

Segment III: Replacement of two E. coli rare codons situated just before the restriction site Nhel

Arq 148 (CGG →CGT), Glv150 (GGA→GGT)

Segment IV: Replacement of a long cluster of E. coli rare codons at the terminal end of the gene with E. coli preference codons.

Gln159 (CAG→CAA), Ser160 (AGC→TCT), Phe161 (TTC→TTT), Glu163 (GAG→GAA), Val164 (GTG→GTT), Ser165 (TCG→AGC), Tyr166 (TAC→TAT), Arg167 (CGC→CGT), Leu 169 (CTA→CTG), Arg170 (CGC→CGT), His171 (CAC→CAT), Leu172 (CTT→CTG), Ala173 (GCG→GCT), Pro175 (CCC→CCG)

After the construction of the synthetic gene coding for hG-CSF the optimized synthetic gene is subcloned in the final plasmid vector suitable for the expression in E.coli. Preferably, the plasmid vector is selected from the group of pET vectors (available from Novagen). These vectors contain a strong T7 promoter. More preferably the plasmid vector pET3a comprising an ampicilline resistance gene, and particularly the plasmid vector pET9a comprising a kanamycin resistance- gene is used. The expression plasmid which is thereby constructed is then transformed into an appropriate E. coli production strain. Preferably, the E. coli production strain is selected from the group of strains which carry on the chromosome or expression plasmid gene for T7 RNA polymerase. Most preferably, E. coli BL21 (DE3) is used.

The procedure is continued with the preparation of inoculum and with the fermentation process in a suitable culture medium. Preferably, IPTG is used for induction, suitable at a concentration in the range of about 0.1 mM to about 1 mM. Preferably at a concentration of about 0.3 to 0.6 mM. The fermentation can be performed at about 37°C, but is preferably performed below 30°C, more preferably at about 20 to 30°C, particularly at about 25°C. Performing the fermentation process at such a lower temperature than conventionally used can advantageously assist in the accumulation of precursor molecules of biologically active G-CSF in inclusion bodies.

The fermentation process may be performed in the presence or in the absence of the antibiotic that corresponds to resistance gene which is inserted into the plasmid vector, e.g. with ampicilline or kanamycin at an appropriate concentration or in the absence thereof. It has been found that the fermentation and thus the accumulation of hG-CSF was highly effective also without a selection pressure.

The accumulated heterologous hG-CSF is found in the inclusion bodies and is suitable for the renaturation process and use in the isolation procedures.

Suitable techniques for the isolation and/or purification of the hG-CSF or biologically active G-CSF protein are known to the person skilled in the art and can be used, e.g., classical or expanded-bed chromatography using any of well known principles, e.g., ion-exchange, hydro phobic-interaction, affinity or size-exclusion, as well as continuous and batch-mode extractions using appropriate matrices or solutions. The preferred technique is immobilised metal affinity chromatography (IMAC), as it enables a highly efficient preparation of pure and biologically active protein in high yield and under native conditions.

The isolated and/or purified hG-CSF or biologically active G-CSF obtained according to the present invention can be used in a process for the manufacture of a pharmaceutical composition containing it as an effective ingredient. The pharmaceutical composition comprises an amount of hG-CSF or biologically active G-CSF that is therapeutically effective to treat a desired disease in a patient. Suitable pharmaceutically acceptable carrier or auxiliary substances include suitable diluents, adjuvants and/or carriers useful in G-CSF therapy.

Biologically active G-CSF which was obtained by using the process of the present invention can be used for preparation of medicaments, which are indicated for the indications selected from the group, which comprises: neutropenia and neutropenia-related clinical sequelae, reduction of hospitalisation for febrile neutropenia after chemotherapy, mobilisation of hematopoietic progenitor cells, as alternative to donor leukocyte infusion, chronic neutropenia, neutropenic and non- neutropenic infections, transplant recipients, chronic inflammatory conditions, sepsis and septic shock, reduction of rist, morbidity, mortality, number of days of hospitalisation in neutropenic and non-neutropenic infections, prevention of infection and infection-related complications in neutropenic and non-neutropenic patients, prevention of nosocomial infection and to reduce the mortality rate and the frequency rate of nosocomial infections, enteral administration in neonates, enhancing the immune system in neonates, improving the clinical outcome in intensive care unit patients and critically ill patients, wound/skin ulcers/bums healing and treatment, intensification of chemotherapy and/or radiotherapy, pancytopenia, increase of anti- inflammatory citokines, shortening of intervals of high-dose chemotherapy by the prophylactic employment of filgrastim, potentiation of the anti-tumour effects of photodynamic therapy, prevention and treatment of illness caused by different cerebral disfunctions, treatment of thrombotic illness and their complications and post irradiation recovery of erythropoiesis.

It can be also used for treatment of all other illnesses, which are indicative for G-CSF.

The pharmaceutical composition containing the pure and biologically active G- CSF obtained by the process of the invention can thus be administered, in a manner known to those skilled in the art, to patients in a therapeutically amount which is effective to treat the above mentioned diseases.

The present invention will be explained in more detail by the examples below and by reference to the accompanying drawings, which examples and drawings are however merely illustrative and shall not considered as limiting the present invention. Examples:

Example 1 : Construction of the optimal gene: Foptδ Example 1 a: The initial gene and plasmid preparations

The gene coding for hG-CSF was amplified from BBG13 (R&D) with the PCR method, which was also used to introduce by using the start oligonucleotides the restriction sites Ndel and BamHI at the start and terminal end of the gene.^" The gene was then incorporated in the plasmid pCytexΔH.H (see the description below) between the restriction sites Ndel and BamHI. All other optimization steps for the expression of the gene in E. coli were also performed in this plasmid.

During the initial gene preparation the EcoRV restriction site was annihilated (oligo M20z108) by point mutation. This was performed with the aim to ensure the possibility of introduction of (individual) mutations by using the oligonucleotide- directed mutagenesis in the plasmid pCytexΔH.H with the kit Transformer™ Site- Directed Mutagenesis Kit (Clontech). The selection of mutants in the plasmid pCytexΔH,H-G-CSF via the restriction sites EcoRI/EcoRV was therefore possible.

The starting plasmid pCYTEXPI (Medac, Hamburg) was reconstructed in a way to enable the constitutive expression. This was performed by the excision of the part of the gene coding for cl857 repressor between both restriction sites Hindlll. The obtained plasmid was named pCytexΔH.H.

The oligonucleotide for the annihilation of EcoRV site from the gene coding for hG- CSF: 20z108 5' -CCT GGA AGG AAT ATC CCC CG-3'

Example 1 b: Codon optimization (Figure 1 )

In the first optimization step the synthetic gene between the restriction sites Ndel and Sacl was constructed by ligation of five cassettes (A, B, C, D, E) which were composed of complementary oligonucleotides. This synthetic part of the gene represents the segment I. With the segment I the part of the native gene for hG-CSF between the restriction sites Ndel and Sacl was replaced. This was performed by the excision of the first part of the gene between the restriction sites Ndel and Sacl and its replacement with the synthetically prepared cassette. The process was performed in two steps. In the first step, the cassette A was ligated to the Ndel site and the cassette E was ligated to the Sacl site. After 16 hours at 16°C the ligation mixture was precipitated with ethanol to remove the excess of (not bound) oligonucleotides. In the second steps the central part of the whole cassette (cassettes B, C and D) from the three previously ligated complementary oligonucleotides was added and the ligatjon was performed for 16 hours at 16°C.

In the second optimization step the two for E. coli most critical codons located in the segment III, namely, CGG→CGT (Arg148) and GGA→-GGT (Giy150), were replaced, by using the oligonucleotide-directed mutagenesis (TransformerTM Site-Directed Mutagenesis Kit (Clontech)).

In the third optimization step the segment IV was constructed in a similar way as the segment I with the exception of intermediate ethanol precipitation. The segment IV represents the last part of the gene between the restrictions sites Nhel and BamHI and is composed of two pairs of complementary oligonucleotides (cassettes F and

G).

In the fourth step of optimization the rare codon coding for Ile96 was replaced (ATA→ATT) (segment II) by using the oligonucleotide-directed mutagenesis (TransformerTM Site-Directed Mutagenesis Kit (Clontech)) and the restriction site for Apal (309) (GGT→GGG (Gly101)) was introduced at the 3' end of the segment II. Apal restriction site was then used in the fifth optimization step with the aim to replace the native gene between Sacl and Apal with the synthetic DNA (segment II). This synthetic DNA is composed of three pairs of complementary oligonucleotides (cassette H, I and J). This was performed similarly as in the first step with the later addition of the cassette I.

1^st optimization step: complementary pairs of oligonucleotides (Ndel - Sacl; segment I in Figure 1): Cassette A: composed of complementary oligonucleotides zgl osl in sp1os2: zglosl 5' TAT GAC ACC ACT GGG TCC AGC TTC TTC TCT GCC GCA AAG 3' sp1os2 5' GCA GAG AAG AAG CTG GAC CCA GTG GTG TCA 3'

Cassette B: composed of complementary oligonucleotides zg2os3 in sp2os4: zg2os3 5' CTT TCT GTT GAA ATG TTT AGA ACA AGTTCG TAA AAT TCA AG 3' sp2os4 5' GAA CTT GTT CTA AAC ATT TCA ACA GAA AGC TTT GCG 3'

Cassette C: composed of complementary oligonucleotides zg3os5 in sp3os6: zg3os5 5' GTG ATG GTG CAG CTT TAC AAG AM AAC TGT GTG 3' sp3os6 5' GTT TTT CTT GTA AAG CTG CAC CAT CAC CTT GAA TTT TAC 3'

Cassette D: composed of complementary oligonucleotides zg4os7 in sp4os8: zg4os75^* CAA CTT ATA AAC TGT GTC ATC CAG AAG AAC TGG TTC TGT TAG

3' sp4os8 δ' CAG TTC TTC TGG ATG ACA CAG TTT ATA AGT TGC ACA CA 3'

Cassette E: composed of complementary oligonucleotides zgδos9 in sp5os10: zg5os9 5' GTC ATT CTC TGG GTA TTC CGT GGG CTC CTC TGA GCT 3' sp5os10 δ' CAG AGG AGC CCA CGG AAT ACC CAG AGA ATG ACC TAA CAG AAC 3'

2^nd optimization step: oligonucleotides for the replacement of the most critical codons by using the oligonucleotide-directed mutagenesis replacement CGG→CGT (Arg 148) and GGA→GGT (Gly 160) m38os16

5' CTC TGC TTT CCA GCG CCG TGC AGG TGG GGT CCT GGT TG 3'

3^rd optimization step: complementary pairs of nucleotides (Nhel - BamHI; segment IV on Figure 1):

Cassette F: composed of complementary nucleotides zg6os11 in sp6os12: zg6os1 5' CTA GCC ATC TGC AAT CTT TTC TGG AAG TTA G 3' sp6os12 5' ACG ATA GCT AAC TTC CAG AAA AGA TTG CAG ATG G 3'

Cassette G: composed of complementary oligonucleotides zg7os13 in sp7os14: zg7os13 5' CTA TCG TGT TCT GCG TCA TCT GGC TCA GCC GTG ATA AG 3' sp7os14 5' GAT CCT TAT CAC GGC TGA GCC AGA TGA CGC AGA AC 3' 4^th optimization step: oligonucleotides for the introduction of Apal (309) (GGT→GGG (Gly101)), and the replacement of the rare codon ATA→ATT (Ile96) by using the ol igonucleotide-directed mutagenesis insertion of Apal (309) (GGT→GGG (Gly101 )), and replacement ATA→ATT (lie 96):

ApaIos15

5^* GCC CTG GAG GGG ATT TCC CCC GAG TTG GGG CCC ACC TTG GAC AC 3'

5. optimization step: complementary pairs of oligonucleotides (Sacl - Apal; segment II in Figure 1 ):

Cassette H: composed of complementary oligonucleotides zg8os18 in sp8os19: zg8os18 δ' CCT GTC CGA GCC AGG CGC TGC AGC TGG CAG GCT GCC TGA G 3" sp8os19 δ' CCT GCC AGC TGC AGC GCC TGG CTC GGA CAG GAG CT 3'

Cassette I: composed of complementary oligonucleotides zg9os20 in sp9os21 : zg9os20 δ' CCA ACT GCA TAG CGG TCT GTT TCT GTA TCA GGG TCT GCT G

3' sp9os21 δ' CTG ATA CAG AAA CAG ACC GCT ATG CAG TTG GCT CAG GCA G

3'

Cassette J: composed of complementary oligonucleotides zg10os22 in sp10os23: z:g10os22 δ' CAG GCG CTG GAA GGC ATT TCC CCG GAA CTG GGG CC 3' sp10os23 δ' CCA GTT CCG GGG AAA TGC CTT CCA GCG CCT GCA GCA GAC C 3'

Example 2: Expression of the synthetic gene coding for hG-CSF in E. coli

The optimized gene Foptδ was excised from the plasmid pCyΔH,H with the restriction enzymes Ndel and BamHI and the gene was then subcloned in the final expression plasmid pET3a (Novagen, Madison USA), which contains an ampicilline esistance gene, which was then transformed into the production strain E. coli BL21

DE3).

The cultures were prepared on a shaker at 160 rpm for 24 hours at 2δ°C or 16 hours at 42°C:

- in LBG10/amp100 medium (10 g/l tryptone, δg/l yeast extract, 10 g/TNaCI, 10 g/l glucose, 100 mg/l ampicillin). The induction was performed with the addition of IPTG to the final concentration of 0.4 mM.

The cultures were prepared on a shaker for 24 hours at 160 rpm at 2δ°C:

- in GYSP/amp100 medium (20 g/l phytone, δg/l yeast extract, 10 g/l NaCI, 10 g/I glucose, metals in traces, 100 mg/l ampicillin). The induction was performed with the addition of IPTG into the medium to the final concentration of 0.4 mM.

- in LYSP/amp100 medium (20 g/l phytone, δg/l yeast extract, 10 g/l NaCI, 6 g/I glycerol, 4 g/l lactose, metals in traces, 100 mg/l ampicillin). The induction was performed with the addition of lactose into the medium.

The inoculum was prepared in LBG/amp100 medium (10 g/l tryptone, δ g/l yeast extract, 10 g/l NaCI, 2.δ g/l glucose) and 100 mg/l ampicillin at 2δ°C, 160 rpm overnight.

For analysis 8 ml of the culture was centrifuged at δOOO rpm. The pellets were then resuspended in 10 mM TrisHCI/pH=8.0 in a proportion of 0.66 ml buffer added to calculated 1 unit OD₆oonm- The loaded amounts were thereby equalized. Namely, the final ODeoonm of the cultures in the stated examples were not equal. The samples were mixed in the proportion of 3:1 with 4x SDS - sample buffer with DTT (pH=8.7) and heated 10 minutes at 9δ°C, centrifuged and loaded onto he gel.

Samples of various expression examples, using the optimized gene construction and the conventional hG-CSF cDNA, were compared by SDS-PAGE evaluations. The SDS-PAGE conditions were as follows, giving results are shown by Figs. 3 and 4.

Figure 3 A: SDS-PAGE (4 % stacking, 15 % separating; stained with Coomassie brilliant blue) of the samples of the proteins from the induced and noninduced cultures of production strains E. coli BL21 (DE3) with the expression plasmid pET3a at 25° C and 42° C. The cultures were cultivated in the LBG10/amp100 medium. Legend:

Load 1 : BL21 (DE3) pET3a-hG-CSF non-induced at 2δ°C (10 μl) (no traces of hG- CSF)

Load 2: BL21(DE3) pET3a-hG-CSF induced with IPTG at 2δ°C (10 μl) (slight trace hG-CSF)

Load 3: BL21 (DE3) pET3a-hG-CSF non-induced at 42°C (10 μl) (no traces hG-CSF) Load 4: BL21 (DE3) pET3a-hG-CSF induced with IPTG at 42°C (10 μl) (under 1 % hG-CSF)

Load 5: standard filgrastim 0.3 μg for Coomassie brilliant blue Load 6: BL21 (DE3) pET3a-Fopt5 non-induced at 2δ°C (δ μl) (6 % hG-CSF) Load 7: BL21 (DE3) pET3a-Foptδ induced with IPTG at 2δ°C (δ μl) (over 60% hG- CSF) Figure 3 B: Detection, with antibodies (Western blot); primary rabbit antibodies; secondary goat anti-rabbit IgG antibodies conjugated with horseradish peroxidase, substrate β-naphthol. The samples for the detection with antibodies were loaded in the same amount and in the same sequence as at SDS-PAGE (Figure 3a) with the exception of the standard which load was 0.08 μg.

Figure 4: SDS-PAGE (4 % stacking, 15 % separating; stained with Coomassie brilliant blue) samples of proteins from induced culture of the production strain E. coli BL21 (DE3) with the expression plasmid pET3a at 2δ° C. The cultures were cultivated in GYSP/amp100 and LYSP/amp100 medium.

Legend:

Load 1 : LMW (BioRad)

Load 2: BL21 (DE3) pET3a/P-Fopt5, the culture cultivated in LYSP/amp100; (60% hG-CSF)

Load 3: BL21 (DE3) pET3a/P-Fopt5, the culture cultivated in LYSP/amp100; (over

64% hG-CSF)

Load 4: rhG-CSF (0.6 μg) Load 5: rhG-CSF (1.5 μg)

Load 6: BL21 (DE3) pET3a/P-Foptδ, the culture cultivated in GYSP/amp100 (4 μl); (65% hG-CSF)

Load 7: BL21 (DE3) pET3a/P-Foptδ, the culture cultivated in GYSP/amp100 (δμl); (62% hG-CSF)

The content (%) of accumulated hG-CSF found in the form of inclusion bodies for the native and optimized gene are described in Table 1.

Table 1 Comparison of the accumulation levels of hG-CSF for the native and the optimized gene (Foptδ)

The indicated values for hG-CSF contents are obtained by the densitometric analysis of SDS-PAGE gels stained with Coomassie brilliant blue in the case of Foptδ (Figure 3A and Figure 4) and by using the detection with antibodies (in the case of unoptimized gene (Figure 3B). In the case of Foptδ the relative amount of hG-CSF for the estimation of expression level was determined with^" the profile analysis (program Molecular analyst; BioRad) of the gels by using the apparatus Imaging densitometer Model GS670 (BioRad).

The results show a drastically improved expression level when the optimized synthetic gene Foptδ was used.

Example 3: Expression of the synthetic gene coding for hG-CSF in E. coli

(kanamycin resistance)

The optimized gene Foptδ was excised from the plasmid pET3a/P-Fopt5 bearing the ampicilline resistance with the restriction enzymes Ndel and BamHI and the gene was then subcloned in the final expression plasmid pET9a bearing the kanamycin resistance (Novagen, Madison USA) which was then transformed in the production strain E. coli BL21 (DE3).

The cultures were prepared on a shaker at 160 rpm for 24-30 h at 25°C.

- in GYSP/kan30 medium (20 g/l phytone, 5g/l yeast extract, 10 g/l NaCI, 10 g/l glucose, metals in traces, 30 mg/l kanamycin). The induction was performed with the addition of IPTG into the medium to the final concentration of 0.4 mM.

- in GYSP/kan15 medium (20 g/l phytone, δg/l yeast extract, 10 g/l NaCI, 10 g/I glucose, metals in traces, 15 mg/l kanamycin). The induction was performed with the addition of IPTG into the medium to the final concentration of 0.4 mM.

- in GYSP medium without the addition of an antibiotic (20 g/l phytone, δg/l yeast extract, 10 g/l NaCI, 10 g/I glucose, metals in traces). The induction was performed with the addition of IPTG into the medium to the final concentration of 0.4 mM.

The inoculum was prepared in LBPG/kan30 medium (10 g/l phytone, δ g/l yeast extract, 10 g/l NaCI, 2.5 g/I glucose) and 30 mg/l kanamycin at 2δ°C, at 160 rpm overnight. For SDS-PAGE analysis (the estimation of the content of hG-CSF; expression level) 8 ml of the culture was centrifuged at 5000 rpm. The pellets were then resuspended in 10 mM TrisHCI/pH=8.0 in a proportion of 0.66 ml buffer added to calculated 1 unit

ODeoonm--

The samples were mixed in the proportion of 3:1 with 4x SDS - sample buffer with DTT (pH=8.7) and heated 10 minutes at 95°C, centrifuged and the clear supernatant was loaded on the gel. The content (%) of the accumulated hG-CSF, found in the form of inclusion bodies for the optimized gene are described in Table 2.

Table 2 Accumulation level of hG-CSF for the optimized gene (Foptδ) in pET9a vector bearing the kanamycin resistance

Figure 5 shows the SDS-PAGE (4 % stacking, 1 δ % separating; stained with Coomassie brilliant blue) of the samples of the proteins from the induced culture of production strain E. coli BL21 (DE3) with the expression plasmid pET9a-Fopt5 at 25° C. The cultures were cultivated at two different kanamycin concentrations and without kanamycin, specifically in GYSP/kan30, GYSP/kan1δ and GYSP medium.

Legend:

Lane^" 1 : LMW (BioRad)

Lane 2: BL21(DE3) pET9a-Foptδ in GYSP/kan30 medium induced with IPTG at 2δ°C

(δ μi) (above 52% hG-CSF)

Lane 3: LMW (BioRad)

Lane 4: BL21(DE3) pET9a-Foptδ in GYSP/kan15 medium induced with IPTG at 25°C

(δ μl) (above 64% hG-CSF)

Lane δ: BL21 (DE3) pET9a-Foptδ in GYSP medium induced with IPTG at 2δ°C (δ μl)

(above 63% hG-CSF)

Lane 6: hG-CSF standard

Lane 7: LMW (BioRad)

The above cited amounts of the hG-CSF content are obtained with the densitometric analysis of the SDS-PAGE gels stained with Coomassie brilliant blue. The relative amount of hG-CSF for the estimation of expression level was determined with the profile analysis (program Molecular analyst; BioRad) of the gels by using the apparatus Imaging densitometer Model GS670 (BioRad).

The results show that the accumulation of hG-CSF is of the same order (more than 53%) also in the culture without kanamycin, i.e. without the selection pressure. This indicates that the strain is particularly suitable for use on the industrial scale.

Claims

Patent claims

1. A DNA sequence coding for hG-CSF characterized in that the sequence comprises the nucleotide sequence of SEQ ID NO: 1.

2. A DNA sequence characterized in that the sequence comprises a nucleotide sequence selected from the group-comprising: ^{" ~} - -

(a) a partial sequence of SEQ ID NO: 1 ;

(b) nucleic acids which hybridize with the sequence of SEQ ID NO: 1 under stringent conditions; and

(c) at least one and preferably a combination of the following modifications with respect to the native hG-CSF sequence: in a "segment I" (located at the 5' terminal end between the nucleotide positions 3 and 194): a plurality of replacements which include replacements of E. coli rare codons by E. coli preference codons and replacements of GC rich regions by AT rich regions, in a "segment II" (located between the nucleotide positions 194 and 309): a plurality of replacements of E. coli rare codons by E. coli preference codons, in a "segment III" (located between the nucleotide positions 309 and 467): no change or essentially no change, in a "segment IV" (located at the 3' terminal end between the nucleotide positions 467 and 536): a plurality of replacements of E. coli rare codons by

E. coli preference codons.

3. The DNA sequence according to claim 2, which encodes for a biologically active G-CSF.

4. The DNA sequence according to any one of claims 1 to 3, wherein the nucleotide sequence is capable of providing an expression level of at least 50%, preferably at least 52% in an expression system.

5. The DNA sequence according to claim 1 or 2, further comprising the 5'- untranslated region of the hG-CSF gene which are not changed relative to the native hG-CSF gene.

6. An expression plasmid, characterized in that the plasmid comprises the DNA sequence according to claim 1 or δ and a plasmid vector.

7. An expression plasmid, characterized in that the plasmid comprises a DNA sequence according to any one of claims 2 to δ and a plasmid vector.

8. The expression plasmid according to claim 6 or 7, characterized in that the plasmid vector comprises a T7 promoter sequence.

9. The expression plasmid according to claim 6 or 7,~ characterized- in that the plasmid vector is selected from the group of pET vectors.

10. The expression plasmid according to any one of claims 6 to 9, characterized in that the plasmid vector comprises a resistance gene, preferably an ampicilline or a kanamycine resistance gene.

11. An expression system for the expression of DNA sequence according to claim 1, characterized in that the system comprises the expression plasmid according to any one of claims 6 and 8 to 10 and a production strain E. coli.

12. An expression system for the expression of the DNA sequence according to claim 2, characterized in that the system comprises the expression plasmid according to any one of claims 7 and 8 to 10 and a production strain E. coli.

13. The expression system according to claim 11 or 12, characterized in that the production strain is E. coli BL21 (DE3).

14. The expression system according to any one of claims 11 to 13, characterized in that it is used without an antibiotic.

16. A process for construction of DNA sequence according to claim 1 or claim 2, characterized in that the process comprises (i) applying methods in order to provide a DNA sequence which is changed relative to the native seque^'nce coding for hG-CSF by:

-replacement of some E. coli rare codons with E. coli preference codons, and/or

-replacement of some GC rich regions with AT rich regions; and

(ii) maintaining a completely unchanged part in a substantial portion of the native sequence coding for hG-CSF.

16. A process for construction of DNA sequence according to claim 15, wherein the

DNA sequence further comprises δ'-untranslated region of the hG-CSF gene, characterized in that the process does not involve changes in the δ'-untranslated region in one or more of the following partial regions: translation initiation region, ribosome binding site and the region between the start codon and the ribosome binding site.

1 7. The process for construction of DNA sequence according to claim 16 or 16, wherein a completely unchanged sequence according to (ii) is maintained in segment 111 in a sequence of at least 99 nucleotides in length. -

18. The process for construction of DNA sequence according to any one of claims 15 to 17, further comprising inserting said constructed DNA sequence into a plasmid vector which comprises a T7 promoter sequence.

1 9. The process for construction of DNA sequence according to any one of claims 1δ to 18, which constructed DNA sequence is capable of providing an expression level of at least 50%, preferably at least 52% in a suitable expression system.

20. A process for the expression of hG-CSF, comprising expressing the DNA sequence according to any one of claims 1 to 5, or the expression plasmid according to any one of claims 6 to 10 in E. coli.

21. The process for the expression of hG-CSF according to claim 20, wherein IPTG is used for induction at a concentration in the range of at least 0.1 mM to less than 1 mM, preferably at a concentration of about 0.3 to 0.6 mM.

22. The process according to claim 20 or 21 , which comprises a fermentation step that is performed at a temperature of about 20°C to 30°C, preferably at around 25°C.

23. The process according to claim 20 or 21 , wherein the expression level is at least

50%, preferably at least 52%.

24. A process for the manufacture of a pharmaceutical composition containing, as an effective ingredient, hG-CSF or biologically active G-CSF, comprising the steps of:

(a) carrying out a process according to any one of claims 20 to 23,

(b) isolating and/or purifying the hG-CSF or biologically active G-CSF obtained by step (a), and

(c) mixing the isolated and/or purified hG-CSF or biologically active G-CSF with a pharmaceutically acceptable carrier or auxiliary substances.