FIELD OF THE INVENTION

The present invention refers to methods for production and / or purification of hormones, particularly the follicle stimulating hormone (FSH) , from human or animal origin, through a human cells platform, as well as products thus obtained and their uses.

PRIOR ART

Hormones are involved in various stages of the mammal reproduction, in particular, the gonadotropins follicle stimulating hormone (FSH) , luteinizing hormone (LH) and chorionic gonadotropin (CG) , as well as thyrotropin or thyroid stimulating hormone (TSH) .

Because of their direct or indirect role in reproduction, such hormones have been used in the treatment of infertility and reproductive disorders in mammals, such as human, bovine, sheep, pig, goat, buffalo, horse, monkey, either male or female.

In nature, FSH, LH and TSH are pituitary hormones, whereas the CG is produced during the formation of the placenta.

For pharmaceutical use, the hormones have been produced by recombinant technology or from mammalian urine, the critical factor being the high degree of purity.

In particular, FSH is used in female human patients in ovulation induction and controlled ovarian hyperstimulation for assisted reproduction (e.g. assisted reproductive technologies, intra uterine insemination, in vitro fertilization and intracytoplasmatic sperm injection) . It is also used in male human patients suffering from oligospermia in order to induce and maintain spermatogenesis. FSH treatment is based on successive injections of the hormone, and requires a high purity and specific activity product.

Human FSH (hFSH) is a glycoprotein secreted by the anterior pituitary gland. The molecule is active as a non-covalently linked heterodimer containing alpha and beta subunits. The hFSH alpha subunit is homologous to the alpha subunits of several other glycoprotein hormones (LH, TSH and hCG) and contains 92 amino acids. The beta subunit is specific and contains 111 aminoacids. There are several isoforms of hFSH molecule, which differ among themselves depending on the molecule glycosylation .

Non-human animal FSH is used to induce super- ovulation in female donors and to induce follicular growth during in vitro fertilization programs in anestrous females .

The FSH currently used in the animal reproduction programs is purified from porcine pituitary extracts. This product is prepared by a technique, which affords a cheaper, but less effective, hormonal preparation. Additionally, commercial preparations depend on the availability of a sufficient number of fresh pituitaries for the preparation of each batch, leading to great variability in results during superovulation treatments with different purified product batches.

The bovine hormone FSH (bFSH) is composed of two glycoprotein subunits encoded by genes CGA (a chain, common to all pituitary hormones: FSH, LH and TSH) and bFSH (β chain, specific for this hormone), wherein CGA ^' has 363 base pairs and bFSH has 389 base pairs.

The recombinant bFSH has also been expressed in various recombinant systems, such as: yeast, plants, insect cells, mouse epithelioid cells, hamster ovary cells, transgenic mice milk and transgenic rabbits. However, only the bFSH expressed in mammalian cells showed in vivo biological activity. This is explained by the fact that the bFSH hormone is very complex, composed of two polypeptide chains, each of which containing post-translational modifications, i.e. glycosylation .

Although the animal embryo transfer techniques are widespread, the variability in response to treatment- based super ovulatory gonadotropin persists as a major limitation. This variability results from the lack of reproducibility of available commercial preparations (FSH extracted from porcine pituitary glands) . These products usually have immunogenic undesirable contaminants and potential vectors for transmission of viral particles or prions. Furthermore, FSH commercial preparations always contain some contamination by LH (luteinizing hormone) , contributing to the inconsistency of the super ovulatory treatment. Finally, the species-specific hormone is more effective in inducing ovulation and prevents immune responses in animals. For these reasons, from the point of view of biotechnology, biosafety and economy, development of recombinant FSH in animal cells is of great interest.

In recent years, new methods for preparation of hormones, especially with high purity, have been sought by several research groups, particularly concerning the FSH hormone. From a technical perspective, there are several options in terms of methodologies for purification of FSH. Several of these methodologies avoid immunoaffinity chromatography due to high costs and generation of impurities, such as antibodies immobilized onto the resin used to separate the product of interest.

The international patent application WO 98/20039, filed by IBSA Institut Biochimique AS, describes a process for purification of human urinary FSH from urinary extracts (menopausal gonadotropin) using ion exchange chromatography on DEAE weakly basic anion exchange resins followed by resin affinity chromatography with an anthraquinone derivative as a ligand.

The international patent application WO 00/63248, filed by Institute Massone AS, describes a process for the purification of gonadotropins, including FSH from human urine, using ion exchange chromatography with a strong cationic resin of the sulphopropyl type, ion exchange with strong anionic resin and hydrophobic interaction chromatography.

U.S. 5,990,288, filed by Musick et al. discloses a method for purifying FSH from biological samples, such as human pituitary glands or human post-menopausal urine, using cation exchange chromatography on Fractogel EMD S03- 650M, followed by dye affinity in a Mimetic Orange 1 resin and hydrophobic interaction in Bakerbond Wide Pore HI- Propyl resin.

The international patent application WO 88/10270, filed by Instituto di Ricerca Cesare Serono SPA, discloses a method for purifying human FSH from urine by immunochromatography (using monoclonal antibodies bound to Sepharose 4B) followed by reverse phase HPLC.

The international patent application WO 2005/063811, filed by Ares Trading SA, describes the purification of recombinant human FSH comprising the steps i

of ion exchange chromatography, immobilized metal ion chromatography and hydrophobic interaction chromatography, in any order.

The international patent application WO 2006/051070, filed by Ares Trading SA, describes the purification of recombinant human FSH, comprising the steps of dye affinity chromatography, hydrophobic interaction and reverse phase chromatography, in any order. Thus, there remains a need for new methodologies for the production of both human and animals hormones, especially those able to provide high purity products.

BRIEF DRAWING DESCRIPTION

Figure 1 is a flowchart showing the method for producing the recombinant bovine FSH (rbFSH) , from construction of the recombinant vectors by genetic engineering up to the tests of biological activity in vitro and in vivo, according to example 2.

Figure 2 shows the fractionation of the fragments corresponding to the bFSH a and β chains and of the pCXN2 vector on 1.5% agarose gel stained with ethydium bromide. In this figure, one may observe the size of the fragments with respect to the molecular weight markers (M) , namely: β chain (398bp), a chain (363bp) and vector (~ 6.000bp).

Figure 3 shows the fractionation on 1.5% agarose gel stained with ethydium bromide of the plasmid preparations containing the fragments of interest (both subunits of bFSH) , digested with restriction enzyme, along with the same undigested constructions, where M represents the molecular weight markers.

Figure 4 illustrates the transfection of both bFSH chains in a mammalian cell for their expression.

Figure 5 shows the autoradiography of the Western blot corresponding to rbFSH samples.

Figure 6 shows the in vivo biological activity assay .

Figure 7 shows the general flowchart of the purification method for human and bovine recombinant hormones (steps according to the present invention) .

Figure 8 shows the chromatographic profile of the FSH purification procedure observed at the different chromatographic steps. Figure 9 shows a Western Blot, wherein: 1) rhFSH sample after anion exchange chromatography; 2) rhFSH sample after gel filtration.

Figure 10 shows the method for identification and quantification of FSH using HPLC.

Figure 11 shows representative chromatograms of the method for identification and quantification of FSH hormone in cell culture medium (DMEM) .

Figure 12 shows the identification and quantification of progesterone performed by HPLC.

Figure 13 shows FSH in vivo activity assay. A) human and B) bovine.

Figure 14 shows the biological activity of FSH in bovines. The graph correlates the number of ovarian follicles of 8 or more mm diameterover time after treatment of the cows with the commercial product (Folltropin) or the rbFSH.

Figure 15 shows SDS-PAGE analysis of rhFSH levels in culture supernatant-conditioned media. Conditioned culture medium obtained from 3D and 6E cell clones upon culturing for 24h were subjected to SDS-PAGE gel electrophoresis under reducing conditions. After the electrophoretic run, the proteins were stained by Comasssie brilliant blue.

Figure 16 shows Western Blot analysis of rhFSH levels in culture supernatant-conditioned media. Conditioned culture medium obtained from the parental cell line and cell clones upon culturing for 24h were prepared, resolved by SDS-PAGE, electrophoretically transferred to nitrocellulose membrane and immunostained using an antibody to human FSH β subunit.

Figure 17 shows the cell growth and recombinant protein (rhFSH) production under high and low serum concentration conditions. Cell growth kinetics was quantified in terms of population doubling time. rhFSH production was determined throughout the growth curve by quantification of hFSH present in the conditioned medium using the chemioluminescence assay.

Figure 18 shows the production of rhFSH under both serum-containing and serum-free cell culture media. rhFSH production was determined in cell-conditioned media upon culturing for 24, 48 and 72h and quantified using the chemioluminescence assay.

Figure 19 shows the increment in progesterone production by rFSHR-17 cells upon stimulation with rhFSH compared to nonstimulated cultures. rFSHR-17 cells were stimulated with the culture supernatants obtained from the recombinant clones upon culturing for 24h. Progesterone released to the culture medium was measured by the chemiluminescence method.

Figure 20 shows the dose-effect relationship of recombinant human follicle stimulating hormone (rhFSH) upon rat ovary weight.

Figure 21 shows progesterone production, in ng per mL, by the GFSHR17 cell line in response to rbFSH stimulus, measured by chemiluminescence. 293T: conditioned medium from non-transfected 293T cells; 3A: negative control (empty pCXN2 vector without insert); 6E through 6K: 293T/rbFSH cell clones; cell pop (mix) : conditioned medium of the transfected 293T culture before cell clone isolation. Experiments performed in triplicates.

Figure 22 shows Western Blot analysis of rbFSH present in culture supernatant-conditioned media and in a partially purified preparation. 48h conditioned medium from 293T/rbFSH 6E, partially purified conditioned medium from the same clone, commercial purified porcine FSH ( Folltropin-V®) and conditioned culture medium obtained from the 293T/negative control (3A) were prepared, resolved under reducing 12% SDS-PAGE, electrophoretically transferred to nitrocellulose membrane, incubated with polyclonal antibody to bFSH and chemiluminescently detected using radiographic film. 1: Conditioned medium from the negative control 293T/3A cells transfected with the empty pCXN2 vector without insert; 2-4: Folltropin preparations (2, 5 and lOug, respectively); 5-6: concentrated 48h conditioned medium from 293T/rbFSH clone 6E containing 2 and lOug of rbFSH, respectively; 7-8: partially purified rbFSH preparations containing 5 and 50ug of rbFSH, respectively. The top black arrow indicates the holoprotein, while the middle black arrow indicates the single beta FSH chain and the bottom black arrow indicates the single alpha FSH chain.

Figure 23 shows the effect of recombinant bovine follicle stimulating hormone (rbFSH) in rat ovary weight augmentation .

DESCRIPTION OF THE INVENTION

Recombinant proteins generally exhibit maximum biosafety, better performance in superovulation and no interference with other hormones and/or contaminants due to the purity of the preparation, being, therefore, of great interest for assisted reproduction.

The recombinant technology proved to be feasible to generate proteins with different complexity degrees in a reproducible manner. However, the structure of gonadotropins is difficult to reproduce because, besides the presence of the already mentioned post-translational modifications, the gonadotropins are composed of two chains, whose coding sequences are independently transcribed and translated, and their products are subsequently combined.

The advantage of the recombinant protein expression system in mammalian cells is the possibility to add the complex glycosidic to the newly synthesized chains, which is absolutely essential for the correct tertiary structure of each subunit, being a prerequisite for folding with the counter-subunit . This method facilitates the formation of disulfide bridges, which is a critical factor in the folding and maturation of functional subunits of the hormone. Furthermore, glycosylation directly affects the half-life of the molecule in the circulation and recognition of the molecule by the receptor, as well as hormone exposure to the in vivo regulatory mechanisms, interference in its solubility, and protection against proteases and thermal inactivation . Therefore, sugar addition causes great impact on the biological properties of glycoprotein hormones.

However, even in mammalian cells expression systems, there is great variability in the pattern of sugars added to the expressed recombinant protein, since the structure and composition of carbohydrates attached to the glycoprotein are determined by the glycosylation pattern (types and levels of enzymes responsible for addition of glycans: the glycosyltransferases ) available in each specific type of cell, which may generate specific glycoforms according to their origin. Thus, different cell lines containing different kinds of glycosyltransferases will catalyze the same protein glycosylation in a cell- specific manner.

Nowadays, one of the most effective ways of producing pharmacologically active glycoproteins uses expression systems based on Chinese . hamster ovary (CHO) cells, which, despite containing a variety of oligosaccharide-processing enzymes, are not able to add the penultimate N-acetyl galactosamines and terminal sulphates, being deficient to perform type 2,β fucolizations and signalizations . These sugars are essential to ensure the prolonged half-life of eCG, as described above.

Therefore, taking into consideration that sugars attached to both FSH chains are essential for signal transduction, metabolism and half-life of the protein, folding and stability of the heterodimer, and that the glycosylation profile of a human cell line, the phylogenetically closest to the species of interest (whether human or animal) would be less immunogenic and better recognizable by the hormone receptors, expression of the hormones in 293T cells was developed as a biotechnologically effective expression system.

Furthermore, the HEK cells family, which includes 293T, as used in the method according to the present invention, exhibits great transfection capability, allowing a greater influx of copies of the vector constructs of interest in cells.

The system of expression in cell lines of human embryonic kidney (HEK) cells or derived cells (293, 293T) is capable of producing glycoproteins with N-glycans complexes, which are, in their great majority, biantennaries , relatively homogeneous and partially marked, with approximately 60% of the molecules carrying terminal sialic acid residues.

According to the present invention, hormones include, without limitation, follicle stimulating hormone (FSH), luteinizing hormone (LH) and chorionic gonadotropin (CG), thyroid stimulating hormone (TSH) . Preferably the hormone is FSH.

The production strategy using HEK-derived cells, such as 293T, may be used for expression at high expression levels of any other homo- or multimeric proteins having complex post-translational modifications.

Furthermore, according to the present invention, the mammalian species include human, bovine, sheep, pig, goat, buffalo, horse, monkey, either male or female. The preferred species are human and bovine.

Thus, in a first embodiment, the present invention relates to a method for production and / or purification of hormones comprising the following steps:

1) Extraction of RNA from a tissue sample;

2) Synthesis of cDNA by reverse transcription from RNA obtained in step 1;

3) Amplification of cDNA fragments corresponding to the full a and β chains;

4) Cloning of each of the amplified fragments (a and β chains) in an expression vector for mammalian cells;

5) Co-transfection of both a and β chain- containing vectors (i.e. recombinant plasmid constructs) obtained in step 4, along with an antibiotic resistance gene-containing vector in a mammalian cell line of human embryonic kidney (HEK) cells or HEK-derived cells (293, 293T) ;

6) Selection of recombinant cell clones through their resistance .to the antibiotic (preferably hygromycin) ;

7) Optional detection of the recombinant protein expression by Western blotting and testing the in vitro and in vivo biological activity;

8) Optional purification of the obtained hormone. The vector used in step 4 is a plasmid for expression of the protein chains in mammalian cells, preferably pCXN2. However, other plasmids may be used, provided they comprise the same elements of the pCXN2, i.e. one or more strong promoters, multiple cloning sites, antibiotic resistance sequence for bacterial cloning and recombination sites.

The way to insert plasmids in the cell ( transfection) may be carried out by various systems, such as lipofection, cationic molecules, calcium mediated transfection, polyethylenimine electroporation, since the transfection rate is maintained at the same level.

The proportion of antibiotic resistance plasmids versus the ones used in co-transfection strategy varies from 1:1 to 1:5,000.

The concentration of antibiotic in culture medium (such as hygromycin) used for selection of stable cell clones varies from lOug/mL to 2.000ug/mL.

The cell line used for expression of the product may be replaced by any other derived from HEK 293 cells, since such a cell adds similar post-translational modifications and has the same behavior when cultured in vitro (similar replication capacity, protein expression and transfection abilities, etc.).

The cell may be grown in batches in monolayers or in suspension in different containers, such as culture bottles of various sizes , plates, T flasks, spinners, culture bags, bioreactors, etc.

The culture medium may be any one which complies with the cell metabolic needs, for instance, adult or fetal serum from several species.

The temperature of cultivation varies from 30 to

38 °C.

The rotation, when the cultivation is carried out in spinners, bags and bioreactors, varies from 10 to 200rpm.

The relative proportion of C0 ₂ used for the cell culture ranges from 1 to 10%.

The cultivation pH varies from 5.0 to 7.5.

The recombinant hormone obtained according to the present invention has high purity, for instance, rbFSH is free from LH, when compared with the commercial preparations for bovine species that includes about 15% LH .

The high purity is guaranteed by a complementary purification method, as necessary. Therefore, in another embodiment, the present invention refers to a method for purification of recombinant hormones comprising subjecting the culture supernatant containing recombinant hormone to the following steps:

a) Dye affinity chromatography (Blue Sepharose) : the hormone does not bind to the matrix, which allows cleaning the original sample by removing contaminants that have affinity for the resin.

b) Ion exchange chromatography (DEAE Sepharose): the hormone binds to the matrix, therefore, its elution and isolation from other contaminants is possible.

c) Size exclusion chromatography (Sephacryl S-

100): separation of aggregates and remaining contaminants.

Steps a) and b) may be carried out in any sequence without compromising the result since the buffer exchange is appropriate for the subsequent chromatography.

The recombinant hormone containing supernatant may be previously prepared before the purification step. In this case, the starting material obtained from the method for production described above is optionally clarified and/or concentrated (by ultrafiltration) and / or subjected to buffer exchange by diafiltration or chromatography before the first chromatographic step.

The buffer exchange and / or concentration are carried out by tangential filtration in hollow-fiber membrane using 20 mM sodium phosphate buffer pH 7.0. Alternatively, the buffer exchange is performed by gel filtration using G25 Sepharose resin (for instance, as specified below) and the same buffer as mentioned above.

Table I - Gel Filtration

The affinity chromatography of step a) is carried out in a resin having a dye as the immobilized ligand - Cibacron Blue F3G-A. The dye affinity chromatography is performed using as the binding buffer 20 mM sodium phosphate pH 7.0. The sample elution is then performed by applying a gradient between 20 mM sodium phosphate buffer pH 7.0 and 20mM sodium phosphate buffer pH 7.0 + 2 M NaCl, optionally in the presence of anti-oxidant . The fraction which does not interact with the resin is separated and used in the following step, for instance, as in Table II below .

Table II - Affinity Chromatography

The buffer exchange and concentration may also be performed between steps a) and b) .

After completion of the dye affinity chromatography, fractions of interest are subjected to tangential filtration in hollow-fiber membrane (5kDa) for buffer exchange and / or concentration in ammonium acetate buffer 0.16 M pH 7.0.

Ion exchange chromatography is carried out using DEAE Sepharose FF, for instance, as shown in Table III. Binding to the resin occurs in ammonium acetate buffer 0.16 M pH 7.0 and elution occurs by applying a gradient of ammonium acetate buffer and 0.16 M ammonium acetate buffer 0.16 M + 2.0 NaCl pH 7.0.

Table III - Ion Exchange Chromatography

After the ion-exchange chromatography fractions of interest are eluted and subjected to . buffer exchange (PBS pH 7.4) and separated by size exclusion using the Sephacryl S -100 resin (GE Healthcare) , for instance, as shown in Table IV below.

Table IV - Exclusion Chromatography

Fractionation

1 x 10 ³ - 1 x 10 ⁵

range (M _r )

Spherical allyl dextran and N, N' -

Matrix methylenebisacrylamide

Stable in buffers commonly used:

Chemical 1 M acetic acid, 8 M urea, 6 M guanidine- stability HC1, 1% SDS, 2 M NaCl, 24% ethanol, 30% propanol, 30% acetonitrile, 0.5 M NaOH (Cleaning)

3- 11 (long term)

Stability at pH

2-13 (short term)

Storage 20% ethanol

Storage

4°C a 30°C

temperature 1

Antimicrobial

20% ethanol

agent

The concentration of the product after the last stage of purification may be performed by 5 or lOkDa hollow fiber membrane ultrafiltration, preferably 5kDa membrane.

Alternatively, depending on the volume (< or= to lOOmL) Centricon may be used to concentrate the final product.

The purified recombinant hormone may be frozen or lyophilized after concentration for storage. Alternatively, it may be formulated to provide "ready for use" preparations. For instance, such preparations include m- cresol solution or benzyl alcohol.

In other embodiments, the present invention also refers to purified or non-purified recombinant hormones, their use in infertility treatment or as a fertility enhancer for mammalian species, diagnostic _. ki. s , pharmaceutical or veterinary compositions or cell culture supplements .

The following examples serve to illustrate aspects of the present invention without having, however, any limiting character beyond the content of the claims presented further on.

EXAMPLES

Example 1 - Production of Recombinant Human FSH (rhFSH)

RNA Isolation

Total RNA was extracted from tissue frozen in liquid nitrogen using the isothioguanidine/CsCl cushion method described by Chirgwin et al., 1979. RNA integrity was electrophoretically verified by ethydium bromide staining and through the Abs ₂ 6o/Abs ₂₈₀ nm absorption ratio which was >1.95.

a- and β-hFSH cDNA cloning

Approximately Ιμς of total RNA was reverse- transcribed using the Superscript ^® II RT (Invitrogen) enzyme and random hexamers in a 20μΙ reaction, according to the manufacturer's instructions. One μΐ. of the single- stranded cDNA reaction mixture was used as the template for subsequent PCR reactions.

Efficient cDNA synthesis was confirmed by amplification of the GAPDH housekeeping transcript using primers and conditions described in the PCR-Select ^® cDNA Subtraction Kit from BD Clontech.

PCR amplification of the hFSH a and β subunits cDNAs was achieved using 25μ1, reactions containing 1* PCR Buffer Promega ^® , 1.5mM gCl ₂ , 0.2 mM dNTPs, 0.4μΜ oligos and 0.5U of Taq DNA polymerase.

For amplification of a- and β-hFSH chains, the same PCR cycling was used, as follows: reactions were started with a denaturation step of 94°C for 2min, followed by additional 35 cycles of 94°C for lmin, 60°C for 30sec and 72°C for 30sec, plus a final extension step of 72°C for 5min, and then cooling to 4°C. Primers FSHl (5'- AGTATCCGCCCTGAACACAT-3 ' ) and FSH2 ( 5 ' -CTGTAGGATAAGGAGGAAGG- 3') were used for amplification of the cDNAs spanning the coding region for the common FSH a-chain, and primers FSH4 (5' -ACCGTTTTCAAGTGACCAGG-3 ' ) and FSH5 (5'-

GGCCTGAAATGTCCACTGAT-3 ' ) were used for amplification of the specific FSH β-chain.

The PCR products for these two cDNAs (567 and 432 bp, respectively) , were gel-purified and blunt-end-cloned into the Smal site of pUC18, using the SureClone ^® PCR Cloning Kit (GE Healthcare), according to the manufacturer's protocol. Recombinant clones (pUC18-ahFSH and pUC18- hFSH) were screened by colony PCR and checked by restriction digestion and automated DNA sequencing. Recombinant clones, in the correct orientation, were screened and confirmed using the aforementioned basic molecular cloning techniques.

The mutagenic primers used for a-hFSH cDNA re- amplification were AF ( 5 ' -CCCAGAGAAATTACCGCCAT-3 ' ) and AR (5' -TGTCGACTCATCAAGACAGCA-3 ' ) , whereas for β-hFSH cDNA, the oligos used were BF ( 5 ' -GTTTTCAAGTGACCGCCATG-3 ' ) and BR (5' -GGCCTGAAATGTCGACTGAT-3 ' ) . Kozak consensus sequences are highlighted in bold, and Sail restriction sites are underlined. These PCR products were blunt-end-cloned into the pUC18 vector using the SureClone ^® PCR Cloning Kit (GE Healthcare) , and the resulting recombinant clones were screened by colony PCR followed by confirmation by restriction digestion and automatic DNA sequencing, to yield the pUC18-a (Ko) hFSH and pUC18-p (Ko) hFSH vectors. The expression-ready cDNA inserts for a and β hFSH were released from the pUC18-a ( Ko ) hFSH and pUC18-p (Ko) hFSH vectors by digestion with Sail, followed by gel purification and then subcloned into the Xhol site of the pCXN2 mammalian expression vector. The pCXN2 expression vector was derived from the pCXN vector (Niwa et al . , 1991) containing the neo gene, which confers resistance to Geneticin (G418), and a strong promoter gene from cytomegalovirus .

Incorporation of the a- and β-hFSH cDNAs into a human host cell line

Equimolar amounts of the a and β FSH expression vector (pCXN2-ahFSH and pCXN2-phFSH) were co-transfected with the pX343 vector, derived from pY3 vector containing the bacterial hygromicin B resistance gene, using a higher proportion of the former relative to the latter, into a well characterized mammalian cell line, the human embryonic kidney cells transformed by the wild type T antigen of SV40 (293T cell line, ATCC, CRL1591) (DuBridge et al., 1987; Pear et al., 1993), employing the Lipofectamine Plus Reagent (Invitrogen, Carlsbad, CA) .

Bacterial endotoxins were assayed by the gel-clot LAL assay (GeneScript USA, Piscataway, New Jersey) ) performed by mixing equal volumes of a test sample and lysate with a known sensitivity in a test tube.

Generation of r-hFSH overproducing 293Tj human cell lines

Six different experimental conditions were used, as shown in Table below.

Table V - Experimental Conditions

Following transfection with the pCXN2-ochFSH, pCXN2- hFSH and pX343 expression vectors, colonies which were resistant to hygromicin (lOC^g/mL) were isolated. Genetically stable cell clones were selected for higher levels of FSH production and secretion to the cell culture medium, determined by chemiluminescence immunoassay (CRIESP, 2001) . For the purpose of bioproduction, stable cell lines that expressed the FSH a and β dimer, in relatively abundant amounts, were selected. This was followed by a very detailed evaluation of one particular cell line which displayed promising characteristics (high and stable levels of FSH expression) . Culture supernatants collected at different periods of time were collected, stored and cryopreserved until testing for biological potency .

Establishment of the Master and Working Cell Bank

Following a series of cell expansion and evaluation steps, one particular 293T cell transfectant (6E cell line) was chosen on the basis of its higher levels of rhFSH production. This cell line was used to establish the Master Cell Bank (MCB) , consisting of individual vials containing samples of the same culture of identical cell preparations of the 6E cell clone, which were cryopreserved in liquid nitrogen.

A Working Cell Bank (WBC) was then established by expansion of the cells recovered from a single vial of the MCB. The cells were successively expanded and aliquots were placed into vials and cryopreserved. Cells from one or more vials are cultured for each rhFSH production cycle.

Both the MCB and WCB have been tested for sterility and absence of Mycoplasma and bacterial endotoxins contamination, according to ECC and FDA guidelines .

Analysis of rhFSH preparations by SDS-PAGE and Western Blot

The r-hFSH preparations were reduced with β- mercaptoethanol, denatured by boiling in the presence of SDS and applied onto a 12% polyacrylamide gel. Following fractionation, the proteins were visualized by Coomassie brilliant blue staining and transferred to nitrocellulose membranes using standard techniques. The membrane was blocked in 5% BSA for 2h and then incubated overnight in a 1:6250 dilution of an anti-human FSH β subunit monoclonal antibody. The membrane was subsequently washed three times and then incubated in 1:1,000 dilution of a peroxidase- conjugated polyclonal antibody to the mouse Ig (Vector Laboratories) for Ih. The bands were visualized upon incubation with a chemiluminescent detection reagent (GE Biosciences) and exposed to X-Ray film. The a and β rhFSH subunits co-migrated under reducing and denaturing conditions, appearing as a single broad band on both Comassie brilliant blue and immunostained gels. The apparent molecular weight of the dissociated subunits, estimated using appropriate molecular weight markers, was 16kDa.

Growth characteristics of the rhFSH overexpressing 6E cell line

Cell growth kinetics was assessed by automated count (Celm, Sao Paulo, SP, BR) , and quantified in terms of population doubling time. To this end, cells from conditions 3 (3D cell line) and 6 (6E cell line) were cultured in 6 wells plates (5 x 10 ⁴ cells per well) , in growth medium (DMEM plus 5% FCS or DMEM plus 0.5% FCS), for up to 9 days. rhFSH production was determined throughout the growth curve by quantification of hFSH in the conditioned medium using the chemioluminescence assay.

In vitro bioassay for FSH

The biological activity of the rhFSH was measured using the rat preovulatory follicle cell line (rFSHR-17), overexpressing the rat FSH receptor, which responds to human FSH stimulus by progesterone production. Cell culture supernatants from clones obtained under conditions 3 (3D cell line) and 6 (6E cell line) were chosen for in vitro characterization. rFSH-17 cells were cultured in DMEM/F12 medium containing 5% fetal calf serum (FCS) to 80% of confluence and then stimulated with the culture supernatants obtained from the recombinant clones upon culturing for 24h at 37°C. Progesterone released to the culture medium was measured by the chemiluminescence method .

In vivo bioassay for FSH

The potency of the rhFSH preparations was assessed by determining the rat ovarian weight gain in response to FSH administration, according to the conventional Steelman-Pohley assay. To this end, 45 female Sprague-Dowley rats, aged 22 days were housed (three to five rats per cage) and given standard ad libitum food and water administration. Animals were randomized and divided into 8 experimental groups. Each animal received one daily rhFSH and hCG injection over a period of 3 days. On day 4, the animals were sacrificed and the ovaries were removed and dissected free from the surrounding tissue, and weighed. The reference FSH standard was the Folltropin a (Gonal-F®, Serono International S.A. Geneva, Switzerland).

Results

Construction of a a- and β-hFSH mammalian expression vector

The RNA source for cloning of the cDNAs encoding the human FSH a and β polypeptide chains was isolated from a pituitary tumor surgical specimen, which overexpressed FSH.

PCR amplification of the hFSH a and β subunits cDNAs was achieved as described in Material and Methods. The PCR products for these two cDNAs (567 and 432 bp, respectively) were cloned into the pUC18 resulting in recombinant clones (pUC18-cchFSH and pUC18-phFSH) .

To generate the mammalian expression vectors for hFSH, the inserts obtained from the pUC18-ahFSH and pUC18- phFSH vectors were re-amplified using mutagenic primers to incorporate both the Kozak consensus sequence (ACCGCC) juxtaposed to the initial ATG codon as well as a Sail restriction site at the 3' -end of each cDNA. These PCR products were cloned into the pUC18 vector to yield the pUC18-a(Ko) hFSH and ρϋΌ18-β (Ko) hFSH vectors. Finally, expression-ready cDNA inserts for a and β hFSH were released from the pUC18 vectors and then subcloned into the pCXN2 mammalian expression vector, which was chosen for high expression of the subunits mRNAs .

Expression of human rFSH in 293 cells Since FSH requires glycosylation for full biological activity, rhFSH was produced by genetically engineered mammalian cells, into which the genes coding for the a and β FSH subunits were inserted.

The 293T human embryonic kidney cell line, was selected as recipient cell, since these cells are easily transfected with foreign DNA, and are capable of synthesizing and secreting glycoproteins. Furthermore, they may be culturds in a large scale.

Six different conditions were used for generation of rhFSH overproducing cells, as described on Table V. Following transfection, genetically stable cell clones were isolated for each conditions (Table VI) and selected according to the levels of production and secretion of rhFSH to the cell culture medium. As expected, the cell clones derived from control conditions 2 to 5 displayed low rhFSH production. However, condition 6, containing the gene sequence coding for both a- and β-FSH cDNAs, presented high values of hFSH production (Table VIII) .

Table VII - Selection of rhFSH overproducing cells . Number of rhFSH cell transfectants obtained from each experimental condition.

Experimental r-hFSH cell

conditions transfectants

2 12 3 16

4 12

5 17

6 25

Table VIII - Recombinant protein (rhFSH) production in cell clones supernatant-conditioned media, according to CRIESP protocols .

Control and rhFSH producing cell clones (shown in bold in Table VIII) were selected for protein production characterization and for Master (MCB) and Working (WCB) Cell Bank establishment to allow continuous and reproducible production of FSH . The cell productivity, cellular morphology and cell growth kinetics of the MCB and WCB were found to be similar (data not shown) .

Characterization of the rhFSH protein The electrophoretic mobility pattern in SDS-PAGE under reducing conditions of the 6E cell-conditioned media showed a band which probably includes the dissociated a- and β- subunits with an apparent molecular weight of 16kDa, consistent with the relative molecular size predicted from laser densitometric mass spectrometry (Figure 15) . As expected, this band exactly matched the single well-defined band of the standard reference preparation (Puregon®) .

Western immunobloting analysis, using anti-human FSH β subunit monoclonal antibody and performed under reducing conditions, showed a band migrating with an apparent molecular weight of 16kDa for the 6E cell- conditioned media preparation (Figure 16) . As expected, this band also matched the single well-defined band of the standard reference preparation (Puregon®) . No cross- reacting material was detected in cells transfected with the empty vector or in parental cells.

rhFSH producing cell clone characterization is shown in Figures 17 and 18.

In vitro activity of human rFSH

The in vitro activity of rhFSH was evaluated using a rat granulose cell bioassay. This assay measures bioactive FSH by assessing the induced production of progesterone by rat granulose cells (FSHR-17 cell line) . Preparations from the 6E cell clone conditioned medium increased progesterone production by FSHR-17 cells in a dose-dependent manner (Figure 19).

In vivo activity of human rFSH

The in vivo activity of rhFSH was determined by the Steelman Pohley assay. The behavior of the biological activity of the FSH preparations was verified by checking the dose effect relationships. The dose regimens administered are summarized in Figure 20. FSH preparations were diluted in phosphate-buffered saline to final concentrations of 1, 2, and 3 IU FSH/ml . Samples were each administered to groups of rats given s.c. injections of 0.5ml /rat daily, for. 3 consecutive days, yielding final cumulative doses of 1.5, 3, an 6 IU FSH/rat. A total dose of 60IU human chorionic gonadotropin (HCG) was also administered to each animal. Animals were sacrificed 72h after the first administration, the ovaries were then removed, dissected free of surrounding tissue and weighed. The cumulated mean results of two independent experiments are reported.

These results demonstrate that rhFSH administration increased the ovarian weight in a dose- dependent manner.

Example 2 - Production of Recombinant Bovine FSH (rbFSH)

Figure 1 shows a flowchart of the complete method .

Preparation of total RNA

A pituitary gland was obtained from a mature Bos taurus taurus female. The gland was removed, placed into the RNAlater ^® solution (Ambion, Applied Biosystems, Carlsbad, CA) , subsequently frozen in liquid nitrogen and manually masserated using a crucible and pistil, under liquid nitrogen. Total pituitary RNA was extracted from the crude pituitary material by the cesium chloride method. The quality of the RNA preparation was evaluated by Absorbance (A) at 260nm and ₂ 80nm (Hitachi High-Technologies UV/visible spectrophotometer, U2000), and only the RNA preparations displaying A ₂ 60nmA¾80nm ratio above 1.8 were used for RT- PCR.

Reverse-transcription-polymerase chain reaction (RT-PCR)

A sample (500ng) of the total pituitary RNA was reverse-transcribed using the Superscript™ II Reverse Transcriptase ( Invitrogen©, Carlsbad, CA) enzyme and 125ng of oligo(dT)18 primer (Fermentas Life Science, Burlington, Ontario) in a 20uL reaction, according to the manufacturer's protocol.

Primers Design

For amplification of the sequences coding for the or and β chains of bovine FSH, two pairs of mutagenic primers were designed, namely:

bFSHa-Forward

( 5 ^' -GCGCGTCGACgccaccATGGATTACTACAGAAAATATGCAGCTGTCA-3 ^' ) and bFSHa-Reverse

(5 ^' -GCGCGTCGAGTTAGGATTTGTGATAATAACAAGTGCTGCAGTGGCACT-3 ^' ) , for bFSH a chain; and

bFSH -Forward

(5 ^' -GCGCGTCGACgccaccATGAAGTCTGTCCAGTTCTGTTTCCTTTTCTGT-3 ^' ) and

bFSH -Reverse

( 5 ^' GCGCGTCGAGTTATTCTTTGATTTCCCTGAAGGAGCAGTAGCT-3 ' ) for the bFSH β chain.

Both Forward primers contain: 1) a restriction site spacer (italic); 2) a Sail restriction site (underlined); 3) the Kozak consensus enhancer sequence (bold); and 4) the 5 ^' -end sequence encoding for bFSHa or bFSHp. Both the reverse primers contain: 1) a restriction site spacer (italic); 2) a Sail restriction site (underlined); and 3) the reverse-complement of the 3 ^' -end sequence encoding for bFSHa or bFSHp. The oligonucleotide primers were designed basing on the NCBI posted sequences regions corresponding to the bovine FSH a and β chains.

Amplification of cDNAs corresponding to bFSHa and bFSHp

sequences

Amplification of the cDNA sequences encoding the a and β chains of bFSH, was carried out using the Triple Master Taq (Eppendorf ^® , Westbury, NY), according to the manufacturer's manual. The following conditions were used for the polymerase chain reactions: 93 °C for 2min; 35 cycles: 93 °C for 20sec and 68 °C for 3min; final extension at 72 °C for 5min. The amplified products were subjected to agarose gel electrophoresis and the corresponding bands were purified using the QIAquick ^® Gel Extraction Kit (QIAGEN) and restriction enzyme digested with Sail (Fermentas Life Sciences - Thermo Fisher Scientific Inc.) at 37 °C for 2h and 30min and inactivated at 65 °C for 20min before cloning into the mammalian expression vector.

Cloning of the cc and β bFSH cDNA sequences into the

mammalian expression vector

In order to clone the sequences corresponding to both subunits of the bovine FSH, we have used the pCXN2 mammalian expression vector, derived from the pCXN plasmid. This vector is composed of a strong Cytomegalovirus and beta-actin derived promoter, which allows efficient selection - conferred by the neomycin phosphotransferase-II encoding gene - of high vector copy number cell transfectants, which are consequently able of express high levels of foreign proteins. The nucleotide sequences of the cloned rbFSH a and β constructs were determined using the fluorescent dye termination reaction (BigDye ^® Terminator v3.1 Cycle Sequencing Kit - Applied Biosystems™) and analyzed using an automated DNA sequencer (ABI Prism 3700 DNA Analyzer, Applied Biosystems/Hitashi ) . The chromatograms were assembled and transformed into contigs using the Phred/Phrap software, by comparison with the NCBI posted sequence for all isoforms. The restriction sites, as well as the Kozak fragment, were carefully checked and all sequences proved to be correct. Plasmid mini-scale (GFX miniprep Kit, GE Healthcare, Pittsburgh, PA) preparations were obtained according to the manufacture's protocol and purified plasmids were subsequently used for cellular transfection .

Figure 3 shows the fractionation on 1.5% agarose gel stained with ethydium bromide plasmid of the preparations containing the fragments of interest (both subunits bFSH) , digested with restriction enzyme, and the same undigested constructions, where M represents the molecular weight marker.

Cell line and culture conditions

293T/17 cells, the well characterized human embryonic kidney cells transformed by the SV40 wild type T antigen were purchased from the American Type Culture Collection (ATCC™ - ATCC® Number: CRL-11268™) and cultured in Dulbecco's Modified Eagle's (DME) medium supplemented with 10 % bovine calf serum (HyClone, Thermo Scientific) under a 2.5% C0 ₂ atmosphere, at 37 °C, in adherent culture. For the serum free media culture, cells were maintained in HyQ ^® SFM4HEK293™ medium (HyClone, Thermo Scientific) .

Co-transfection of-293T cells with the pCXN2-rbFSHa and pCX 2-rbFSH constructs along with the pX343 selection vector for generation of cell clones

In order to generate stable cell clones expressing both the a and β chains of bFSH, 293T cells were co-transfected with _.. each of the _ pCXN2-rb.FSH. constructs _and- the pX343 Hygromycin B resistance vector (figure 4), using a 40:1 DNA mass ratio (8 iq pCXN2-rbFSHa plus 8 iq pCXN2- rbFSHP to 200 ng pX343) . The experimental control was performed under the same conditions, but using the empty pCXN2 vector without any insert. The transfection procedure was carried out using Lipofectamine™ 2000 Reagent (Invitrogen ^® ), according to the manufacturer's protocol. 24h after the transfection procedure, the cell population of each condition (control and experimental) were plated in five different dilutions (1:5, 1:10, 1:25, 1:50 and 1:100) and submitted to selection by adding lOOug/mL of the Hygromicin B selective antibiotic (Invitrogen, Carlsbad, CA) into the culture medium. This selective medium was renewed every 48h until the emergence of recombinant colonies, which occurred in approximately two weeks. Recombinant cell colonies of each condition were isolated from the remaining population by using cloning rings and transferred to 48-well plates for cell clones establishment .

Protein Quantification by RIA

After Hygromicin selection, the cell clones were cultured in adherent culture and samples of the conditioned media were collected to evaluate the levels of rbFSH expression. To this end, 10 ⁶ cells of each of the Hygro- resistant cellular clones, as well as the negative control, were plated and samples of the conditioned supernatants were collected after culturing for 48h. These supernatant samples were subsequently submitted to bFSH quantification by radioimmunoassay.

In Vitro Biological Activity

The biological activity of the rbFSH was assayed using the rat pre-ovulatory follicular cell line (GFSHR- 17), which overexpresses the rat FSH receptor, responding to FSH stimulus by increasing the levels of progesterone production, in a dose-dependent manner. Cell culture supernatants obtained from chosen cell clones from both control and experimental conditions were used for the in vitro characterization assays. GFSH-17 cells were cultured to 80% confluence in DMEM/F12 medium containing 5% FCS, plated onto a 24-weel plate and subjected to in vitro stimuli. The stimulatory treatments were performed in serum-free media supplemented with 10% of the conditioned supernatants obtained from 10 ^s cells of each recombinant or control clone, upon 24h of culturing under the described conditions. Progesterone released to the culture medium was assessed by chemilumine.scence . Protein Analysis by Western Blot

The rbFSH supernatants of selected clones were collected and concentrated using the Centricon ^® filter concentrators (10 kDa cutoff) (Millipore, Billerica, MA) . Concentrated samples were diluted to lx SDS-PAGE reducing sample buffer (60 mM Tris-Cl pH 6.8; 2.0% SDS; 10% glycerol; 0.025% bromophenol blue; 700 mM β- mercaptoethanol) and boiled for 5min at 95 °C. Samples were applied to 12% polyacrylamide mini-gels at 100V, 100 and 350mA for approximately 2h and then transferred to a Hybond-ECL nitrocellulose membrane (GE Healthcare, Little Chalfont, Buckinghamshire, UK) , blocked in the Starting Block™ solution (Pierce) plus 0.05%(v/v) Tween 20 for 12h at 4°C and then incubated for two hours in 1:1,000 dilution of a polyclonal anti-FSH β antibody (Santa Cruz Biotechnology, Santa Cruz, CA) . The membrane was subsequently washed three times in PBSA plus 0.01% (v/v) Tween 20 and then incubated in a 1:2000 dilution of a peroxidase-conjugated polyclonal antibody to goat IgG (Vector Laboratories) for lh. The bands were visualized upon incubation with the ECL chemiluminescent detection reagent (GE Healthcare, Little Chalfont, Buckinghamshire, UK) and exposed to autoradiography film (Kodak, Rochester, NY) .

Figure 5 shows the autoradiograph of the Western blot corresponding to rbFSH samples, wherein:

- Folltropin: FSH commercial preparation obtained from pig pituitary extracts;

rbFSH: recombinant bovine FSH produced according to the present invention;

- rbFSH: the same product after a preliminary purification stage.

In this figure, the numbers below the keys show the amount (micrograms - ug) applied to each channel of the gel. The whitish bands refer to the high consumption of the substrate by the antibody conjugated peroxidase enzyme, demonstrating the high amount of protein in the gel. On this autoradiograph, one may verify the presence of bFSH dimer (~ 34kDa) , oligomers (~ 68kDa) and the beta subunit (~ 22kDa) .

In Vivo Biological Activity

The rbFSH physiological activity present in samples of cell clones culture supernatants was assessed by determining the rat ovarian weight gain of sexually immature rats in response to three consecutive exogenous FSH doses, according to the Pharmacopeia FSH assay. To this end, Sprague-Dawley female rats, aged 22 days, obtained from the University of Sao Paulo Chemistry Institute Animal facility, were housed (three to five rats per cage) and submitted to ad libitum standard food and water administration. Animals were randomized and distributed into the different groups. Each animal received one daily rhFSH plus hCG injection over a period of three days. On day 4, the animals were sacrificed, their ovaries were removed and dissected free from the surrounding tissue and subsequently weighed. The reference FSH preparation used as an experimental control was the commercial porcine FSH preparation, Folltropin ^® -V (Bioniche Animal Health, Belleville, Ontario, Canada) . This experiment was performed in biological triplicates and two independent experimental replicates .

Figure 6 shows the in vivo biological activity assay results. The first two pairs of ovaries refer to negative controls of the experiment: the animals received saline (PBSA) or hCG (human chorionic gonadotropin, 10UI - hCG) used to synergize the FSH effect. The numbers (0.05 mg, 0.1 mg, 0.2 mg) below the dividing line reflects the dosage applied in each experimental condition. The treatments were independently carried out: Folltropin (commercial product) , the rbFSH over producer recombinant cell clone grown in culture medium without fetal bovine serum (FBS 1), the same culture medium after a previous purification step 1, the rbFSH over producer recombinant cell clone grown in culture medium without FBS 2, the same culture medium after a previous purification 2. In the rbFSH medium 1, the size of the ovaries decreases despite increasing doses of rbFSH applied in treatment, which results from a negative feedback due to excessively applied high dosage.

Adaptation of the cell clones to suspension culture and stability of overexpression .

The highest overexpressing cell clone was adapted to suspension culture in non-adherent T-flasks by replacing the medium for HyQ®SF 4HEK293TM either directly (100% fresh suspension culture medium) or sequentially (50% fresh suspension culture medium, 50% conditioned adherent culture medium) each 3-4 days. Viability was checked by treating cell samples with 50% v/v Trypan Blue (Gibco, Life Technologies, Carlsbad, CA) , with cells being considered totally adapted to suspension culture as soon as they reached 95% of viability in 100% of HyQ®SFM HEK293TM . After adaptation, conditioned medium was collected and submitted to the previously described biological assays.

Results

Construction of the pCX 2-bFSHa and pCX 2-bFSH mammalian expression vectors

The cDNA obtained from a bovine pituitary gland was used for amplification of the cDNA sequences encoding the bovine FSH a and β chains. The one-round amplification and mutation of both chains was performed as described in Experimental Procedures, and the PCR products were fractionated by agarose gel electrophoresis, in which the 389bp and 416bp amplicons, respectively, bearing the Sail restriction site, the digestion spacers and the Kozak fragment were clearly identified. These PCR products were gel purified and subsequently digested with Sail (Fermentas, Burlington, Ontario, Canada) before cloning into the Xhol site of the pCXN2 vector, as previously described. These steps generated both pCXN2-abFSH and pCXN2-pbFSH recombinant constructs, which were checked for integrity by DNA sequencing.

Expression of rbFSH in human 293T cells The recombinant plasmids bearing both bFSH chains (pCXN2-abFSH and pCXN2-pbFSH) were co-transfected along with the Hygromicin-resistance pX343 plasmid in a 40:1 ratio - into previously plated 293T cells. 24h after transfection, the experimental and control cell populations were submitted to selection in the presence of Hygromicin B. until the emergence of cellular colonies, which were isolated as described in Material and Methods.

All recombinant cellular clone supernatants were submitted to RIA in order to assess the rbFSH expression levels obtained from each clone. Values ranging from 1.2 to 20ug of rbFSH were obtained per mL of cellular conditioned medium, and seven cellular clones expressing higher levels of rbFSH were selected for further characterizations, as described below.

In Vitro Biological Activity

The in vitro activity of rbFSH was evaluated as described in Material and Methods. Conditioned media obtained from seven selected rbFSH cell clones, as well from the original rbFSH transfected population, were able to increase progesterone production by GFSHR-17 cells in a dose-dependent manner (Figure 21), with the highest in vitro biological activity being achieved by clone 293T/rbFSH 6E. This clone was selected to undergo molecular characterization by Western blot.

rbFSH overexpression revealed by Western Blotting

In order to determine the pattern of rbFSH expression by clone 293T 6E, Western immunoblotting analysis was carried out. Supernatants from 293T 6E cells cultured for 48h were concentrated and compared to commercial bFSH ( Folltropin ^® -V) and partially purified rbFSH (data not shown) , in a 12% SDS-PAGE under reducing conditions. After fractionation, proteins were transferred to nitrocellulose membranes, which were incubated with the polyclonal anti-FSH β antibody. Figure 22 shows the pattern obtained for each preparation upon autoradiography. It is possible to observe a band of an apparent molecular mass of 19kDa in the 6E cell-conditioned medium preparation (Figure 22, wells 5 and 6) , consistent with the single-cc chain subunit of bFSH, which is also present in the commercial FSH (wells 2-4) and in the partially purified rbFSH (wells 7-8), but is absent from the negative control (well 1). The autoradiography also shows other bands of 37 and 60kDa, corresponding to the single β chain of bFSH and to the αβ dimer, respectively, in both the rbFSH and in the partially purified rbFSH conditions (wells 5-6 and 7-8), which are absent from the negative control condition (well 2) . This last 60kDa band appears fainter but is clearly present in the commercial FSH band, but the 37kDa band does not appear under this condition. The black arrows, shown in figure 22, point to the FSH subunits - the one at the top indicates the holloprotein, while the one in the middle indicates the single β FSH subunit and the arrow at the bottom indicates the single a FSH subunit.

Adaptation of clone 293T/rbFSH 6E to Suspension culture

Switching from adherent to suspension culture is a critical step to scale-up the production of recombinant proteins in mammalian cells, as well as for improving the purification procedure. Therefore, we looked into adapting the 293T/rbFSH 6E cell clone to such condition, both directly and sequentially in order to remove SFB from the culture. Our results revealed that sequential adaptation results in higher cell viability in comparison to direct adaptation, therefore, the suspension-adapted 293T/rbFSH 6E clone was selected to undergo the activity assay.

In Vivo Biological Activity

In order to assess the expression levels of the rbFSH constructs, we used an in vivo biological activity assay based on the Pharmacopeia Steelman & Pohley assay, as described in Material and Methods. The response of immature female rats to the stimuli of different FSH preparations was evaluated by the ovary weight augmentation upon three consecutive daily doses of hormonal preparations, in a dose-dependent fashion. Conditioned medium from the recombinant cell clone 293T/rbFSH 6E (1) maintained in DMEM 10% FBS and collected after 48h in DMEM lacking FBS and (2) after adaptation to serum-free medium. In this experiment, HPLC-purified rbFSH was also utilized. The experimental groups, as well as the response obtained after each treatment, are summarized in Figure 23.

FSH preparations were diluted in phosphate- buffered saline and different doses (0.05; 0.1 and 0.2mg) were injected. Samples were administered, subcutaneously, to groups of five rats in injections of 0.2mL/rat daily, for three consecutive days. A total dose of 10IU human chorionic gonadotropin (hCG) was also administered to each animal, except for the negative control, at day 1, according to the Pharmacopeia. Animals were sacrificed 72h after the first administration; their ovaries were removed, dissected free from surrounding tissue and weighed. The cumulated mean results of three independent experiments are shown. 1: PBSA (phosphate-buffered saline); 2:10 IU of hCG (human chorionic gonadotropin); 3-5: 10 IU of hCG plus Folltropin (0.05, 0.1 or 0.2mg, consecutively); 6-8: 10 IU of hCG plus 48h DMEM serum-free conditioned medium of 2931/rbFSH 6E containing 0.05, 0.1 or 0.2mg, consecutively, of rbFSH; 9-11: 10 IU of hCG plus 0.05, 0.1 or 0.2mg of partially purified rbFSH obtained from previous DMEM serum free conditioned medium; 12-14: 10 IU of hCG plus 48h conditioned medium of 293T/rbFSH 6E adapted to serum-free medium containing 0.05, 0.1 or 0.2mg of rbFSH; 15-17: 10 IU of hCG plus 0.05, 0.1 or 0.2mg of partially purified rbFSH obtained from cells adapted to serum-free medium.

These results demonstrate that the Folligon commercial product (3-5) displayed a discrete response by the animals, which was very similar to that of the 10IU hCG administration (2) . It is also possible to notice that rbFSH produced by cells cultured regularly in DMEM 10% FBS before changing to DMEM without FBS (6-8) had the greatest biological response among all groups. It is important to note that the 0.05mg group (6) provoked a more dramatic augmentation in ovaries weight in comparison to higher doses (0.1 and 0.2mg) of the same preparation (7 and 8), in a proportionally-inverse response manner. The second best response group was achieved by the rbFSH produced under the same conditions as groups 6-8, but being HPLC purified (9- 11) . However, this group displayed a proportional dose- response ovaries weight augmentation, according to the applied rbFSH dose (0.05, 0.1 or 0.2mg). The rbFSH produced by 293T/rbFSH 6E cells adapted to SFM medium (12-17) displayed lower biological response, in comparison to the non-adapted cells (6-11) . The levels obtained for the non- purified medium (12-14) and the purified one (15-17) were similar, in all doses, with the purified rbFSH produced under this SFM condition displaying a better dose-response profile .

In this study, it has been demonstrated high levels of expression of an in vitro and in vivo biologically active rbFSH produced in the modified 293T human embryonic kidney cell line.

The coding sequences for a and β chains of bovine FSH were successfully amplified and cloned into the expression vector pCXN2. Adopting a 40:1 ratio in co- transfection of the rbFSH constructs relative to the pX343 hygromycin-resistance selection vector proved to be very efficient to isolate cell colonies derived from single cells, resulting in the establishment of several overexpressing cell clones. The pCXN2 expression vector utilized for the expression of the rbFSH is a derivative from the pCXN plasmid. The elements present in the pCXN2 vector allow high copy number and high yield of cell transfectants expressing high levels of the foreign proteins. After transfection and hygromicin selection of low-density plated cell populations, we isolated 12 rbFSH cellular clones. Production of rbFSH by these cell clones, evaluated by RIA, revealed that the expression levels varied from 1.2 to 20ug of rbFSH per mL of culture medium conditioned by these cell clones.

Some of the cell clones expressing the highest amounts of rbFSH (from 5ug to 20ug of rbFSH/mL) were selected for protein characterization. The in vitro activity of these overexpressing clones was evaluated by assessing the levels of progesterone produced by the FSH- responsive GFSHRlVcell lineage upon rbFSH stimulus. All cell clones were able to induce progesterone production in this responsive lineage, showing homogenous recombinant protein activity. Conditioned media from selected rbFSH clones were used as stimulus, with two cell clones, namely 293T/rbFSH 6E and 6K, displaying the highest in vitro activity. Clone 6E, the most in vitro biologically active clone, was selected to undergo further assays.

Western blot analysis (Figure 22) revealed that the a and β chains display similar molecular mass as the highly purified commercial FSH ( Folltropin) , being also possible to identify not only the isolated a and β bands (MM of 19 and 37 kDa, respectively), but, also, the bFSH dimer (MM around 60kDa) , which remained partially folded even under reducing conditions and upon denaturation. Since bFSH is N-linked glycosylated, the band pattern observed is in accordance with the expected size of fully-glycosylated bFSH chains and dimer 2, which may be demonstrated upon treatment with PNGase F. The lack of the β FSH chain signal in Folltropin (lines 2-4) may be explained by the low primary antibody recognition of this polypeptide chain, since Folltropin is purified from porcine tissue and, even though it shows similar molecular weight and activity, when compared to rbFSH, it has different aminoacid residues composition .

Another critical assessment is the in vivo biological activity of rbFSH. Using the Pharmacopeia Steelman & Pohley FSH in vivo bioactivity assay, the results were similar among animals from the same group, especially among the purified conditions (3-5; 9-11 and 15- 17). For this assay, we used conditioned medium from the 293T/rbFSH 6E recombinant cell clone under two different conditions, namely: (1) cells cultured in DMEM 10% FBS and conditioned medium collected after 48h in the presence of DMEM lacking FBS and (2) after cells adaptation to serum- free medium. Both conditions were able to induce in vivo biological response, but conditioned medium from cells cultured in the presence of FBS elicited a significantly higher response in comparison to those adapted to the serum-free medium. Considering the response of groups 6-8, described above, the higher response was obtained with the lower dosage (0.05mg - group 6), when compared to the lower levels of ovary weight augmentation found for higher doses (0.1 and 0.2mg - groups 7 and 8) and the proportionally- inverse response which occurred, we assume that this effect it probably due to a negative feedback mechanism. Since it was the most active preparation, the maximum dose limit was rapidly reached, and, even though a known amount of rbFSH was applied, the response observed was higher than expected, probably due to a higher specific activity of this preparation. It is important to compare this result with the response displayed by the second more active group (9-11), which displayed a proportional dose-response ovary weight augmentation, according to the applied rbFSH dose (0.05, 0.1 or 0.2mg) .

A protocol for purifying rbFSH from conditioned media has been developed and, despite a loss in biological activity for groups 9-11 (DMEM) , the results showed that the biological activity was maintained for groups 14-17 (serum-free medium) . In addition, for both conditions, the biological activity of purified rbFSH was higher than that of the commercial ^' FSH (Folltropin - groups 3-5), demonstrating that our expression and purification protocol generates a fully biological active and stable preparation. Biological activities of rhFSH or rbFSH are also shown in figure 13 and figure 14.

Example 3 - Purification Assays

Recombinant hormone samples obtained according to examples 1 or 2 were purified according to the flowchart shown in Figure 7. As may be observed, the method comprises a step of buffer exchange (diafiltration or gel filtration / desalting) and concentration which aims to prepare the sample for the first chromatographic step. After the first chromatographic step, a new buffer exchange and concentration is performed before performing the next step. After the second chromatography step the material is gel filtrated and the product of interest is lyophilized or frozen stored.

The chromatographic profile of the purification method in the different chromatographic steps is shown in Figure 8. The arrows indicate the fractions collected and used in subsequent chromatographic steps:

A - Step 1, collecting the fraction 3 obtained during buffer exchange by gel filtration / desalination,

B - Step 2, flow through collection obtained during dye affinity chromatography,

C - Step 3, collecting the peak eluted during ion exchange chromatography. This fraction is subjected to gel filtration for final purification. Concentration and buffer exchange between steps between B and C are not shown. The identification and characterization of the biomolecule (rbFSH or rhFSH) was performed by Western blotting and electrophoresis (SDS-PAGE) , respectively. The quantification of biomolecules is performed by radioimmunoassay or assay for colorimetric method.

Figure 9 shows Western blotting, wherein: 1) rhFSH sample after anion exchange chromatography; 2) rhFSH sample after gel filtration.

Figure 10 shows the method for identification and quantification of FSH on HPLC. SDS-PAGE gel stained with Colloidal Coomassie Blue. 1) material passed through the column without binding to it (Flow Through); 2) representative fraction of the column washing material (Wash-out); 3) Standard Molecular Weight; 4) Fraction 1 eluted from the column; 5) Fraction 2 eluted from the column; 6) Fraction 3 eluted from the column, 7) Fraction 4 eluted from the column, 8) Puregon 200 ng;. 9) culture supernatant before purification.

Alternatively to the colorimetric or radioimmunoassay methods and to the protein quantitation by gel densitometry, we also developed a method for identification and quantification of FSH in HPLC, as shown in Figure 11, wherein A = chromatogram obtained for Folltropin (lOug/ml) in methanol, indicating the retention time of the solution used as a standard; B = representative chromatogram analysis of culture medium (in different situations); C = chromatogram obtained for Folltropin (lOug / ml) in DMEM, indicating the maintaining of the solution retention time (3.4 minutes) used as standard.

The method may be used for process control because it is able to assess the presence of the biomolecule in the culture medium (culture supernatant) as well as to characterize the purified product.

Regarding the rhFSH, it is possible to identify the molecule by mass spectrometry. Two peptides of the alpha chain of hFSH were identified in the 15 kDa band with the rate of error of 5% by culture supernatant producing cells in the absence of fetal serum. The beta chain hFSH features 2 glycosylation sites predicted by UNIPROT. Furthermore, it is possible to perform estimates based on densitometry of bands separated in the gel using IMAGEJ software analysis and also by means of approximations obtained by counting peptides identified (spectral counts approach) . From these analyzes it is possible to estimate that FSH is 0.1 to 2% of the total mass present in samples from cell culture supernatant, representing concentrations between 0.2 and 4 ug/mL.

ELISA was performed as a quantification method using a commercial kit. The cell culture supernatant (serum free media) showed a productivity of 0.14 IU hFSH / cell (equivalent to 100,000 IU/L) .

By using qPCR it is also possible to determine the number of mRNA copies, which has a great relevance in the design of production methods. It is possible to detect the expression of mRNA encoding the human FSH a and β chains by qRT-PCR to determine and quantify the number of mRNA copies of a and β hormone chains transcribed in the hFSH producing cells, compared to the GAPDH endogenous control gene. Syber Green fluorophore is used to establish a Melting curve to verify the specificity of amplification. The relative expression of the FSH alpha chain was 50 times greater than the host cell, which was not transfected with the FSH gene and the relative expression of FSH beta chain was 400 times greater than the one used in the qRT-PCR assay.

The biological activity of rhFSH is analyzed based on progesterone production by cells GFSHR-17. These cells overexpress the FSH receptor and respond to the stimulation of FSH secreting progesterone to the culture medium. Thus, supernatants of the production method or already purified FSH are evaluated by applying to GFSHR-17 cells. After incubation with the culture conditioned medium supernatants, the 17-GFSHR cells which were stimulated with different concentrations of rhFSH were collected and the amount of progesterone produced by these cells was determined by electrochemiluminescence.

The activity of purified human recombinant FSH was higher than 9UI/ug (for bovine recombinant IU activity unit is not used and the market product is presented in milligrams ) .

Alternatively to the FSH activity detection (measured by progesterone production) by electrochemiluminescence, we also developed a methodology for progesterone detection by HPLC, as illustrated in Figure 12. In this figure, A-D: representative chromatograms of the selectivity and specificity assay of the method; A = progesterone in methanol, B = culture media evaluated as biological matrices, C = culture media with FSH ( Folltropin) ; D = culture media with progesterone (lOug / ml) and FSH; E: Calibration / Linearity of the method; F: Initial evaluation of the accuracy (Kruskal- allis p> 0.05); G: Specificity and Selectivity, A-D = culture media evaluated as biological matrices; MeOH = methanol as matrix for analysis of progesterone.

Briefly, this methodology was developed for directly analysing FSH and progesterone in cell culture medium samples. Liquid chromatography with UV detection (254 nm) and a C18 column were used for samples analysis. The runs were performed with a methanol, ethanol and water mobile phase. The flow was lmL/min, 40°C incubation temperature and the injection volume of 20 L sample.

It should be understood that the embodiments described above are only illustrative and that modification throughout may occur to the one skilled in the art. Therefore, the invention should not be considered limited to the embodiments described herein.