WO2009142767A1 - Lentivector-based protein production - Google Patents

Abstract

This invention relates to making proteins in cultures of mammalian cells through the use of lentiviral vectors encoding the protein or proteins of interest. The invention provides a method for making a mammalian cell that produces a protein by transducing a mammalian cell with lentiviral transduction vectors at least 2 times in 24 hours, wherein each lentiviral transduction vector comprises an expressible polynucleotide coding for the protein. In one embodiment, the transduced cell is an apoptosis-resistant cell. The invention further provides mammalian cells and cell lines produced by this method as well as cultures of such cells. The invention also provides a method for making a protein by culturing these cells and cell cultures under conditions effective to produce the protein. The invention further provides an assay for measuring the concentration of a protein in a sample having a high concentration of the protein.

Description

LENTIVECTOR-BASED PROTEIN PRODUCTION

This application claims priority to U.S. Provisional Patent Application No. 61/128,486, filed May 22, 2008, which is incorporated herein by reference in its entirety.

FIELD OF THE INVENTION

This invention relates generally to protein production and more specifically to making proteins in cultures of mammalian cells through the use of lentiviral vectors encoding the protein or proteins of interest.

BACKGROUND

Recent progress in biomedical sciences has resulted in a substantial number of proteins that find clinical use. These proteins may be hormones, interleukins, cytokines, enzymes, antibodies or even novel type of vaccines.^1"5 One of the main limiting factors to more ubiquitous use of the proteins is the complexity of manufacturing proteins and their availability in large enough quantities. The production issues also interfere with the rapid development of new products as well as the advancement of biosimilars. Although many proteins have been expressed in bacterial, fungal ⁶'⁷, insect ⁸'⁹, and plant ^1O'^U cells, several factors limit the use of these alternative manufacturing methods, mainly the proper folding, secretion, and secondary modification of the proteins of interest. Inappropriate protein modifications can occur include misfolding and aggregation, oxidation of methionine, deamidation of asparagine and glutamine, variable glycosylation, and proteolysis^12"14. Interestingly, even viral proteins intended for vaccines benefit from mammalian expression system, since the proper glycosylation is often essential for appropriate immunogenicity⁶'⁷'¹⁵'¹⁶ and is considered important factor when developing biosimilars¹⁷'¹⁸. Among these post-translational modifications, glycosylation is considered to be one of the most important factors^19"21. Developing an expression system for efficient production of mammalian proteins is a long, time consuming job that may take a year or more to complete. Many expression systems produce yields^22"24 in the range of 50-250 mg/L protein in 4-5 day long stirred batch cultures or 500-1000 mg/L in 7-10 day cultures as compared to the amounts produced by other, non-mammalian systems.

Lentivector-based transduction of mammalian cells results in the protein encoding genes being permanently inserted randomly into the chromosomes of the targeted cells, which allows a subsequent selection of a stable cell line that over-expresses one or more proteins of interest. If the mammalian cell line and the encoded protein(s) are suitable and the expression levels are sufficiently high, the system can be used for efficient and economically viable protein manufacturing. Also, lentiviral vectors have been shown to have a lower genotoxic potential than other retroviral vectors, such as those in the Oncornaviridae family (eg Moloney Murine Leukemia viral vectors).^25"30 Therefore, cell lines created from the use of lentiviral vectors and not Oncoretroviral vectors are safer for use in the production of biologic products.

Furthermore, oncoretroviral vector transduction is not as efficient as lentiviral vector transduction because the cells have to be a mitotic state for efficient oncoretroviral transduction to occur. Therefore, higher multiplicity of infection (MOI) does not translate into increased copy numbers, and it may lead to unnecessary cytotoxic effects.

The present invention overcomes these deficiencies.

SUMMARY OF THE INVENTION

This invention relates to making proteins in cultures of mammalian cells through the use of lentiviral vectors encoding the protein or proteins of interest. In one embodiment, the invention provides a method for making mammalian cells that produce high levels of properly glycosylated proteins, provides high density cultures of such cells, and provides an assay for measuring the concentration of the proteins in cultures having a high concentration of the proteins.

The invention provides a method for making a mammalian cell that produces a protein. A mammalian cell is transduced with lentiviral transduction vectors at least 2 times in 24 hours, wherein each lentiviral transduction vector comprises an expressible polynucleotide coding for the protein. In one embodiment, the cell that is transduced is an apoptosis-resistant cell. Such cells are made by exposing mammalian cells to an apoptosis- inducing chemical and selecting at least one cell that shows the characteristic of apoptosis- resistance, then transducing the selected cell with the lentiviral transduction vector. The invention further provides mammalian cells and cell lines produced by the method of the invention as well as cultures of such cells. In one embodiment, the cells are optimized for high density cell cultures. En one embodiment, the cell line produces the protein without significant deglycosylation of the protein. In one embodiment, the cell line comprises at least three copies of the expressible polynucleotide integrated into the genome of each cell. The copies of the polynucleotide are inserted without using a gene amplification technique. In another embodiment, the cell line comprises apoptosis-resistant cells.

The invention also provides a method for making a protein by culturing these cells and cell cultures under conditions effective to produce the protein. The protein may be isolated and purified from the cells or from the supernatant if it is excreted from the cells. The invention also provides a method for making high levels of multiple proteins from a single cell, either by transduction of the cell with multiple vectors, each coding for a different protein or with a single vector that where the expressible polynucleotide insert is polycistronic and can produce multiple proteins. The invention further provides an assay for measuring the concentration of a protein in a sample having a high concentration of the protein. The assay comprises the steps of: 1) contacting molecules of the protein, which are bound to the surface of a carrier, with labeled ligands to create complexes comprising the labeled ligands bound to the proteins; 2) measuring the signal produced by the labeled ligands bound to the proteins; 3) contacting molecules of the protein in the sample to be assayed with the labeled ligands to form complexes comprising the labeled ligands bound to the proteins, which complexes are not bound to the surface of the carrier; 4) measuring the signal produced by the labeled ligands bound to proteins; 5) calculating the difference between the two signals; and 6) correlating the difference with the concentration of the proteins in the sample.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1. Titration of GFP-encoding Lentivector on HEK293 cells. Increasing the MOI may increase the percentage of cells that carry at least one transgene, however the efficiency of transduction decreases, toxic effects of transductions become visible (by increased cell death rates) and the cost of transduction rises rapidly. Figure 2. Repeated hourly transductions were inefficient in increasing the copy transduction, indicating that the cellular mechanisms responsible for lentivector insertion can be saturated or disturbed by the addition of "fresh" vector.

Figure 3. The single high dose of lentivector can be delivered in repeated transductions of lower doses of the vector. Note that the transduction efficiency is linearly additive at low MOI, indicating an efficient use of lentivector particles. The repeated transductions were done at intervals of 2 hours.

Figure 4. Transducing cells with 2 different gene carrying Lentivectors does not interfere with insertion efficiency if the vectors follow each other by more than 1 hour. This finding allows to repeat a transduction with the same or with a different lentivector, delivering a larger dose while remaining efficient by avoiding the saturation seen in the case of single - large dose delivery. In this experiment, the repeated transductions were done at 2 hour intervals. Figure 5. Typical embodiment of the Rapid Transduction Protocol for Lentivector-based protein production. Upon selecting the target protein, the lentivector is designed, assembled, produced, purified if necessary and the quality of the manufactured vector is verified and the titer is determined if possible. If not used within a short time (several weeks, or as the stability of the vector dictates), aliquots of the vector may be prepared and stored at -70 C. Note, that a rather low initial cell number is possible, due to reduced total toxicity associated with the high doses of Lentivectors. This improves the efficiency by allowing us the reduce the amount of the vector needed to achieve high copy numbers. The number of target cells for transductions is typically comparable to the number of single cell clones the developer needs to test (typically lxl0⁴-2xl0⁵). In addition, the repeated transductions shorten the development time, since 10-12 transductions per day are feasible and very high copy numbers are attainable at low MOI in days. We found that the cells tolerate the rapid transductions so well that it is possible to follow the transductions with single cell cloning within 3 days, while the remaining cells can be cultured further in bulk for fast protein production, that is possible because the efficient transduction procedure the entire population of cells carries many stabile copies of the gene of interest. Bulk transduced cell lines are sufficient for most research level protein production project, while the industrial strength cell lines can be selected by single cell cloning. The initial evaluation of the single cell cones and the bulk cultures can happen 14-30 days after the transductions, depending on the sensitivity of the assay at hand, upon which the transductions can be repeated on selected clones or on the samples from the bulk culture or proceed with establishing the master cell bank and larger scale protein production.

Figure 5A. In situ hybridization of a fluorescent labeled gag probe to the chromatin of multiple-transduced DG44 cells having over 40 copies of EPO gene inserts. Figure 6. Lentivector-transduced DG44 cells over express highly glycosylated form of recombinant human erythropoeitin. All four single cell clones from 16 harvests from 4 different clones. When compared in Western blot, the bands corresponding to the erythropoietin migrated to similar distances from the top of the gel indicating high degree of similarity between the different samples including similar molecular weights, matching the molecular weight of the highly purified recombinant human EPO purchased from Sigma. Figure 7.

We found consistent over expression, high degree of glycosylation and high purity of erythropoietin, comparable to purified recombinant human erythropoietin when 2 different harvest from 4 different clones were compared by BioAnalyser. Figure 8. The distribution of EPO producing clones according to EPO productivity. Note that the frequency of high producer clones is very high, in the order of 5% of the population tested. This allows the developer to reduce the number of clones that needs to be tested before finding one that meets the criteria for efficient, high level protein production. Figure 9. Staurosporine-resistance can be established by culturing cell lines in the presence of increasing concentration of staurosporine. Maintaining the DG44 cell line, that has an estimated LD₅₀ of 1-2 nM for staurosporine, in increasingly higher concentration of staurosporine resulted in establishing the DG44str cell line, that is at least 1000 fold more resistant, and has an LD₅₀ of 1-5000 nM. Figure 10. Growth of DG44 and the str lines in ProCho medium in tissue culture flask, without medium change or feeding. The cell were maintained under conditions which can be considered adverse. The DG44 str cell line showed faster growth rates, reached about 30% higher cell density and remained alive longer than the parental cell line. Figure 11. Factor VIII Productivity of different DG44 str clones. Figure 12. Comparison of FVIII productivities by methods. Only a few milliunits of Human factor VIII are produced by retro virus-transduced cell lines. Expressing the rhpFVIII with transient, plasmid-based systems increased the productivity by approximately 50-100 fold, but the Lentivector-based systems yielded an additional 10 fold increase and using the cell line in high density culture mode resulted an additional 5 -to 10 fold increase even under non- optimized conditions.

Figure 13. Comparing assay methods for determining protein in supernatants. Figure 14. Competitive bead assay to rank single cell clones according to IgG production. Microspheres 5 uM in diameter were coated with murine polyclonal, affinity purified murine IgG using standard amino coupling method. The calibration curve was prepared using 1-64 ug/mL murine IgG and 1 : 1000 diluted polyclonal FITC labeled rabbit anti mouse IgG. The linear range of sensitivity can be expanded by increasing the amount of beads and the excess of FITC-labeled anti mouse IgG antibody and the assay total volume". Figure 15. Schematic for using tagged version of the target protein to develop immunoassay for detecting, measuring, and purifying the protein of interest.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to lentiviral transduction vectors used to introduce expressible polynucleotides of interest into host cells. This creates host cells that express a protein of interest but do not produce infectious lentiviral particles. As used herein, a "lentiviral transduction vector" is an enveloped virion particle that contains an expressible polynucleotide sequence, and which is capable of penetrating a target host cell, thereby carrying the expressible sequence into the cell. The terms "lentiviral vector," "lentivector," "vector", and their plurals will also be used herein to refer to lentiviral transduction vector(s). As used herein, a "host cell" refers to any mammalian cell, whether located in vitro or in vivo.

The enveloped particle is preferably pseudotyped with an engineered or native viral envelope protein from another viral species, including non-lentiviruses, which alters the host range and infectivity of the native lentivirus. The envelope polypeptide is displayed on the viral surface and is involved in the recognition and infection of host cells by a virus particle. The host range and specificity can be changed by modifying or substituting the envelope polypeptide, e.g., with an envelope expressed by a different (heterologous) viral species or which has otherwise been modified. This is called pseudotyping. See, e.g., Yee et al., Proc. Natl. Acad. Sci. USA 91 : 9564-9568, 1994. Vesicular stomatitis virus (VSV) protein G (VSV G) has been used extensively because of its broad species and tissue tropism and its ability to confer physical stability and high infectivity to vector particles. See, e.g., Yee et al, Methods Cell Biol., (1994) 43:99-1 12. An envelope polypeptide can be utilized without limitation, including, e.g., HIV gpl20 (including native and modified forms), Moloney murine leukemia virus (MoMuLV or MMLV), Harvey murine sarcoma virus (HaMuSV or HSV), murine mammary tumor virus (MuMTV or MMTV), gibbon ape leukemia virus (GaLV or GALV), Rous sarcoma virus (RSV), hepatitis viruses, influenza viruses (VSV-G), Moloka, Rabies, filovirus (e.g., Ebola and Marburg, such as GP1/GP2 envelope, including NP.sub.--066246 and Q05320), amphotropic, alphavirus, etc. Other examples, include, e.g., envelope proteins from Togaviridae, Rhabdoviridae, Retroviridae, Poxyiridae, Paramyxoviridae, and other enveloped virus families. Other example envelopes are from viruses listed in the following database located on the worldwide web at ncbi.nlm.nih.gov/genomes/VIRUSES/viruses.html.

Furthermore, a viral envelope protein can be modified or engineered to contain polypeptide sequences that allow the transduction vector to target and infect host cells outside its normal range or more specifically limit transduction to a cell or tissue type. For example, the envelope protein can be joined in-frame with targeting sequences, such as receptor ligands, antibodies (using an antigen-binding portion of an antibody or a recombinant antibody-type molecule, such as a single chain antibody), and polypeptide moieties or modifications thereof (e.g., where a glycosylation site is present in the targeting sequence) that, when displayed on the transduction vector coat, facilitate directed delivery of the virion particle to a target cell of interest.

Furthermore, envelope proteins can further comprise sequences that modulate cell function. Modulating cell function with a transducing vector may increase or decrease transduction efficiency for certain cell types in a mixed population of cells. For example, stem cells could be transduced more specifically with envelope sequences containing ligands or binding partners that bind specifically to stem cells, rather than other cell types that are found in the blood or bone marrow. Such ligands are known in the art. Non-limiting examples are stem cell factor (SCF) and Flt-3 ligand. Other examples, include, e.g., antibodies (e.g., single-chain antibodies that are specific for a cell-type), and essentially any antigen (including receptors) that is specific for such tissues as lung, liver, pancreas, heart, endothelial, smooth, breast, prostate, epithelial, vascular cancer, etc.

Lentiviral transduction vectors comprising the polynucleotide of interest are constructed by techniques known to those skilled in the art. Such techniques are disclosed in U.S. Patent Application No. 11/884,639, published as US 2008/0254008 Al, and in U.S. Patent Nos. 5,994,136, 6,013,516, 6,165,782, 6,294,165 Bl, 6,428,953 Bl, 6,797,512 Bl, 6,863,884 B2, 6,924,144 B2, 7,083,981 B2, and 7,250,299 Bl, the disclosures of which are incorporated herein by reference in their entireties. Such vectors include those constructed from Human Immunodeficiency Virus (HIV), Simian Immunodeficiency Virus (SIV), Bovine Immunodeficiency Virus (BIV), feline immunodeficiency virus (FIV), Caprine arthritis-encephalitis virus (CAEV)and Equine Infectious Anemia Virus (EIAV). In one embodiment, the vectors are constructed from HIV, including HIV-I and HIV-2. Cells. Cell Lines, and Cell Cultures

In one embodiment, the present invention provides mammalian cells and cell lines that produce a protein, methods of making the cells and cell lines, and methods of using them to produce a desired protein. In another embodiment, the cells produce the protein without significant deglycosylation of the protein. As used herein, "without significant glycosylation" is defined as not deglycosylated more than 30% from normal levels of natural glycosylation. In one embodiment, the cells produce high levels of protein. "High levels of protein" is somewhat relative, depending on the cells and the protein. For a protein that is not toxic to the cell, the cells can produce at least 5 pg/cell/day to 10 pg/cell/day and generally from about 20pg/cell/day to about 80pg/cell/day. Many high producing cells produce over over 80 pg/cell/day of protein.

The cell contains at least three copies of an expressible polynucleotide integrated into its genome. In one embodiment, the cell contains at least five copies of the polynucleotide integrated into its genome; in another embodiment, it contains at least 10 copies of the polynucleotide. In a further embodiment, the cell contains from about 3 to about 50 copies of the expressible polynucleotide integrated into its genome. In still another embodiment, the cell contains from about 20 to about 40 copies of the integrated polynucleotide. As used herein, an "expressible polynucleotide" means a polynucleotide operably linked to a promoter so that it codes for and expresses a protein once integrated into the genome of the cell. It may contain other regulatory sequences to facilitate or enhance coding of the protein. As contained in a lentiviral transduction vector, the expressible polynucleotide is an RNA molecule, which is reverse transcribed in the cell into complementary DNA (cDNA) that is integrated into the cell's genome.

As used herein, the term "polynucleotide" means a RNA, DNA, or cDNA sequence; i.e., a molecule that is a polymer of ribonucleotides or deoxyribonucleotides. The expressible polynucleotides are inserted into the cells' genomes by lentiviral transduction vectors. Such vectors are well known by those skilled in the art as a means for inserting polynucleotides into cells. When lentiviral vectors are used in multiple rounds of transductions, they integrate at many sites in the genome. In this way the resulting cell has multiple copies of the gene, but they are at scattered sites of the chromosomes. With many other systems of such as other viruses or plasmids are used, they generally insert at a single location in the cell and then gene amplification techniques, such as selection for a drug resistance gene located in tandom with the gene of interest in the vector, are used to increase the copy number of the gene of interest. All of the copies of the amplified genes are located at a single site on the genome. This can result in certain disadvantages, such as the possibility of gene silencing by changes in the state of the chromatin in that region. We believe it can also lead to deglycosylation due to local overloading of the glycosylation pathway.

The polynucleotide and expressed protein may be heterologous, i.e., not naturally in or produced by the cell, or they may be non-heterologous, i.e., naturally in or produced by the cell. In the latter case, the additional polynucleotides integrated into the genome of the cell produce additional protein, beyond that produced by the cell prior to the integration of the additional polynucleotides. In one embodiment, the protein is secreted from the cell into the surrounding medium or supernantant. The protein is any protein capable of being made by a mammalian cell. These include, but are not limited to, mammalian, human, viral, and bacterial proteins.

In one embodiment, the proteins are human ones, whether naturally occurring or genetically engineered. These include but are not limited to hormones, interleukins, cytokines, enzymes, and antibodies. Specific examples include but are not limited to Factor VIII, Factor IX, anti-CD20, anthrax antibody, rituximab, erlotinib, efalizumab, cituximab, etanercept, darbepoietin alpha, erythropoietin (EPO), growth hormone, alpha and beta interferon, insulin, and granulocyte colony stimulatory factor.

In another embodiment, the proteins are viral or bacterial ones, whether naturally occurring or genetically engineered. These include influenza proteins, such as hemagglutinin and neuraminidase, and chimeric antigens, such as hemagluttinin-adjuvant and hemagluttinin-cancer antigen, or surface proteins from a variety of other infectious agents including,, but not limited to ebola virus, Marburg virus, dengue virus, HIV, CMV, Herpes, Hepatitis. Thus, the protein production system of the invention can be used for antiviral vaccine production. Vaccines against viruses are often manufactured using hosts that are efficient in producing large amounts of proteins, such as eggs (influenza vaccine) monkey kidney cells (polio) or even insect, bacteria, or plant cells. However, the glycosylation pattern may, and often does, differ significantly. The present invention provides cells and methods for manufacturing viral proteins used in vaccines that do not have these defects in glycosylation.

The mammalian cell may be any one that can be transduced by lentiviral transduction vectors. These include, but are not limited to, primary cells, stem cells, embryonic cells, and cells from immortalized and non-immortalized cell lines. Specific examples include 3T3 cells, CHO (immortalized Chinese hamster ovary) cells, CHO-derived DG44 (DHFR-) cells, Scr cells, C 127 cells, CA cells, HeLa cells, NSO cells, HEK-293 (Human embryonic kidney) cells, BHK (Baby hamster kidney) cells, PER-C6 cells, SP2/0 cells, U937 cells, THP-I cells, Vero cells, and YAC-I cells. In one embodiment, the cells are human cells, such as HEK- 293 or HeLa cells. In another embodiment, the cells are CHO or DG44 cells. In another embodiment, the cells are human umbilical mesenchymal stem cells.

In another embodiment, the tranduced cells are made from starting cells that are apoptosis-resistant and, therefore, are themselves apoptosis-resistant. An apoptosis-resistant cell line is produced by the following steps: 1) exposing mammalian cells to an apoptosis- inducing chemical; 2) selecting at least one cell that shows the characteristic of apoptosis- resistance; and 3 ) growing it to form a cell line. In one embodiment, the resulting cells are about 1 ,000 - 10,000 more resistant to the apoptosis-inducing chemical than the starting cells. In another embodiment, the doubling time of the culture of apoptosis-resistant cells is about 95% to about 50% of the doubling time of a culture of the starting cells (the original mammalian cells). Apoptosis-inducing chemicals are known to those skilled in the art. One such chemical is staurosporine.

In one embodiment, a clonally selected mammalian host cell comprising from about 20 to about 50 lentiviral transduction vectors integrated into the genome of the host cell, wherein the integrated vectors comprise at least one polynucleotide operably linked to a promoter, said polynucleotide encoding a secreted protein, wherein the host cell does not produce infectious lentiviral particles. As used herein, "clonally selected" refers to a cell line that is derived from a single cell. Making the Cells The protein-producing cells of the invention are made by transducing a mammalian cell with a lentiviral transduction vector at least two times in 24 hours. Each lenti viral transduction vector carries an expressible polynucleotide that codes for the protein that is desired to be produced. In one embodiment, the cell is transduced at regular time intervals over the 24 hour period, where each time interval is about one hour to about twelve hours after the prior transduction. In another embodiment, the cell is transduced approximately every 6 hours over the 24 hour period. In still another embodiment, cell is transduced approximately every 2 hours over the 24 hour period.

Transduction does not need to stop after 24 hours. It can continue for several days. The timing will depend, among other things, on the cell, vector, and polynucleotide and can be determined by a person skilled in the art, given the teachings contained herein. In one embodiment of the invention, the cell is transduced over a 24 hour to 96 hour period. In another embodiment, it is transduced over a 24 hour to 72 hour period. In another embodiment, it is transduced over a 24 hour to 48 hour period. In practice, a plurality of cells are tranduced by a plurality of vectors during these time periods. The cells can then be screened to select a cell that produces high levels of the protein and has other desired characteristics.

The cells are transduced at a low multiplicity of infection (MOI). As used herein, the terms "multiplicity of infection" or "MOI" refer to the ratio of the number of lentiviral transduction vectors to the number of cells being exposed to the vectors. Herein all MOI descriptions, unless otherwise stated, are functional titers (e.g. levels of green fluorescent protein (GFP) expression) rather than based on qPCR titers which are about 10-fold higher. The cells are transduced at a MOI of less than 100. In one embodiment, the MOI is about 10 to about 99. In another embodiment, the MOI is about 10 to about 50. In still another embodiment, the MOI is about 50 to about 99. In a further embodiment, the MOI is about 50.

In one embodiment, the invention provides a method for transducing a host cell comprising the steps of: 1) contacting a host cell with a plurality of lentiviral transduction vectors at a multiplicity of infection of less than or equal to 100 under conditions whereby at least two of the vectors integrate into the genome of the host cell, wherein the vectors comprise at least one polynucleotide operably linked to a promoter, said polynucleotide encoding a secreted protein, wherein the host cell does not produce infectious lentiviral particles; and 2) clonally selecting a transduced host cell. In another embodiment, the invention provides a method of transducing a host cell comprising the step of contacting a host cell comprising least one expressible polynucleotide already integrated into the genome of the host cell, wherein the host cell does not produce infectious lentiviral particles, with a plurality of lentiviral transduction vectors comprising the expressible polynucleotide, said polynucleotide encoding a secreted protein, at a multiplicity of infection of less than or equal to 100.

The vector may comprise two different expressible polynucleotides coding for two different proteins. Alternatively, two different transduction vectors may be used; that is, a first lentiviral transduction vector comprising a first expressible polynucleotide encoding a first protein and a second lentiviral transduction vector comprising a second expressible polynucleotide encoding a second protein can be used to transduce the cell. For example, the use of two different polynucleotides permits the construction of a cell that produces both the heavy and light chains of an immunoglobulin molecule. Moreover, the cell may be transduced with with a plurality of lentiviral transduction vectors, wherein each vector comprises a different expressible polynucleotide coding for a different protein. As mentioned above, generally, a plurality of cells are tranduced by a plurality of the vectors in all of these embodiments.

The polynucleotides delivered by the vectors are part of a construct that includes promoters operably linked to the polynucleotides and may include other genetic elements, including but not limited to enhancer sequences, polyadenylation sequences, termination sequences, insulators, post-translation regulatory elements (PRE), microRNA, RNAi, inducible elements (e.g. tetracycline-inducible promoters), IRES 2 A, Furin 2 A, Kozak sequences, or retrotransposon elements. The genetic elements may be tissue specific or cell specific. The term "tissue specific" as it applies to a regulatory element refers to a regulatory element that is capable of directing selective expression of a nucleotide sequence of interest to a specific type of tissue (e.g., liver) in the relative absence of expression of the same nucleotide sequence of interest in a different type of tissue (e.g., lung).

Transduction can be facilitated by various transduction enhancing agents known to those skilled in the art. These include Polybrene, PEI, and phospholipids like Lipofectin, calcium phosphate, nucleofectin and other polymers and lipids. Transduction can also be facilitiated by physical techniques such as electroporation or magnetofectin.

The lentiviral vectors integrate into the chromosomes of the transduced cell. In one embodiment of the invention, each of the transduced, protein-producing cells of the invention has about 3 to about 50 vectors integrated into its genome. In another embodiment, the cell has about 20 to about 40 vectors integrated into its genome.

The transduced cells of the invention are expanded into cell lines by Standard cloning techniques. The cell lines are then cultured under conditions effective to produce the desired protein. The cultures can be high-density, where the concentration of the cells is about 10⁶ cells per mL to about 10⁸ cells per mL. In one embodiment, the concentration of the cells is about 1 x 10⁷ cells per mL to about 6 x 10⁷ cells per mL. For a protein that is not toxic to the cell, the cell cultures produce at least 5 pg/cell/day to 10 pg/cell/day and generally from about 20pg/cell/day to about 80pg/cell/day. Many high producing cells produce over over 80 pg/cell/day of protein.

In one embodiment, the invention provides a method of producing a secretable protein comprising the steps of: 1) providing a clonally selected host cells wherein the genome of the host cells comprises at least about 50 integrated copies of at least one integrated lentiviral transduction vector comprising a polynucleotide operably linked to a promoter, wherein the polynucleotide encodes the secretable protein , and 2) culturing the host cells under conditions whereby the protein is produced at rate greater than about 50 picograms per cell per day.

In one embodiment, the culture is maintained at about 30° C to about 36° C. In another embodiment, it is maintained at about 32° C to about 34° C, and in still another embodiment, the culture is maintained at about 34° C. These temperatures increase protein production. Assay for High Concentration of Proteins

The invention also provides an assay for measuring the concentration of a protein in a sample having a high concentration of the protein. The assay comprises the steps of: 1) contacting molecules of the protein, which are bound to the surface of a carrier, with labeled ligands to create complexes comprising the labeled ligands bound to the proteins; 2) measuring the signal produced by the labeled ligands bound to the proteins; 3) contacting molecules of the protein in the sample to be assayed with the labeled ligands in a manner such that the proteins in the sample cannot bind to the carrier but can bind to the labeled ligand, forming complexes comprising the labeled ligands bound to the proteins, which complexes are not bound to the surface of the carrier; 4) measuring the signal produced by the labeled ligands bound to proteins; 5) calculating the difference between the two signals; and 6) correlating the difference with the concentration of the proteins in the sample. In one embodiment, the assay comprises the steps of: 1) binding molecules of the proteins to the surface of a carrier through tags incorporated into the proteins and tag-specific ligands bound to the surface of the carrier, 2) contacting the tagged proteins that are bound to the surface of the carrier with labeled ligands to create complexes comprising the labeled ligands bound to the tagged proteins; 3) measuring the signal produced by the labeled ligands bound to the tagged proteins; 4) contacting untagged molecules of the protein in the sample to be assayed with the labeled ligands to create complexes comprising the labeled ligands and the untagged proteins, which complexes are not bound to the surface of the carrier; 5) measuring the signal produced by the labeled ligands bound to the untagged proteins that are not bound to the surface of the carrier; 6) calculating the difference between the two signals; and 7) correlating the difference with the concentration of the proteins in the sample.

The carriers are ones well known in the art. They include, but are not limited to, microtiter wells, beads, microspheres, liposomes, filters and filter paper, sensor chip surfaces, glass slides, and metal-coated surfaces (e.g. gold coated). In one embodiment, the carriers are beads and microtiter wells. In another embodiment, microspheres are used with flow cytometry.

The tag is any small molecule that can be added to a protein. It allows the protein to be bound or detected. They are well known to persons skilled in the art. In the present invention, tags include, but are not limited to, green, red, and yellow florescent protein, flurochromes, radioisotopes, enzymes, short peptides, and His-tag. In one embodiment, it is Strep-tag® peptide, an 8 amino acid peptide that binds to the biotin pocket of streptavidin and is commercially available from IBA US, St. Louis, Missouri, USA. A nucleotide sequence coding for this peptide can be incorporated into the expressible polynucleotide that is transduced into the cell. Thereby, it becomes part of the protein of interest.

Tag-specific ligands are well known to persons skilled in the art. They include biotin- avidin, gold surface, and sulfhydrl(-SH) containing amino acids. In one embodiment, the ligand is Strep»Tactin ® protein, a streptavidin derivative that binds to Strep-tag® peptide. It is commercially available from IBA US, St. Louis, Missouri, USA. Labeled ligands are also well known to persons skilled in the art. The ligands can be any molecule that will bind to the protein of interest, such as an antibody. The labels include, but are not limited to, fluorochromes, enzymes, radioisotopes, chromophores, magnetic beads, and optically active compounds, such as sugars and certain nanoparticles. The determination of the difference and correlating can be achieved a) by measuring the accumulation of the labeled ligand on the prepared surface or (b) measuring the remaining unbound ligand in the supernatant. The differential signal is the value obtained in the absence and in the presence of the protein of interest. A calibration curve can be generated by using known amounts of protein of interest and calculating the absolute amounts of the protein of interest or use a standard or series of standards which allow normalizing the differential values and ranking of unknown samples by their content of protein of interest.

The assay can measure concentrations of the protein ranging from about 20 ug/mL to about 10 mg/mL. In one embodiment, the concentration of protein is from about 50 ug/mL to about 1 mg/mL.

Uses of the Invention

The protein production system of the invention can be used to produce proteins for therapeutic, diagnostic, and research purposes. Specific applications include the following: transgene expression; production of material for biochemical analyses; production of assay standards; structural studies, including protein crystallization, protein structure, and NMR; protein-protein interaction experiments; immunogen for antibody, T-cell/T-cell receptor development, and vaccines; physiology and pathology studies; diagnostic applications; therapeutic applications; vaccine development; and protein engineering and mutagenesis studies. The invention can also be applied to ex vivo cell therapy, when cells harvested from patients or cultured cell lines (such as stem cells, including but not limited to mesenchymal stem cells, mesenchymal-like cells, embryonic stem cells, inducible embryonic stem cells, and primary cells converted to stem cells, or somatic cells such as blood cells, including umbilical cord blood cells, brain cells, liver cells, ) are transduced. In such cases, the copy number and total transduction rate needs to be controlled, and the cell toxicity needs to be kept at very low level.

The invention can also be applied to in vivo gene therapy where the patient is injected with lentiviral vectors multiple times over a 24 hour period of time.

EXAMPLES

The following examples illustrate certain aspects and embodiments of the invention and should not be construed as limiting the scope thereof. Example 1 : Rapid Transduction Protocol

In the test system presented in Figure 1, a lentivector (Ltg-173 encoding Green Fluorescence Protein as test gene of interest) is used. A fixed number of HEK293 cells were transduced with varying amount of vector (varying the multiplicity of infection or MOI). If successfully transduced, cells expressed GFP (Green Fluorescence Protein). When measured in Flow Cytometric Assay, the GFP positivity and GFP fluorescence intensity increased in the transduced cell population as the MOI increased. However, above a certain MOI, in this case above a MOI of 50 - 100, we reach the area of diminishing returns. The intensity of GFP fluorescence is not increasing (not shown), indicating that there may be a saturation-effect, and worse that substantial cytotoxicity is present due to high load of toxic viral elements

(such as VSVG) when the MOI is increased over a critical value. In addition to diminishing effectiveness, the cost of the Lentivector particles used for transduction is not negligible, especially if larger numbers of cells are targeted or there is an in vivo administering of the vector. We tested the possibility of repeating the transductions hourly at a low MOI of 20.

The results indicate that the "saturation" effect appeared at an even lower concentration, as soon as the third transduction, and the transduction efficiency was low, only 20-25 % of the cells harbored one or a small number of genes (Figure 2).

The efficiency of transduction increased significantly when we increased the time window between subsequent transductions to two or more hours. In that experiment (Figure 3), the single dose of 100 MOI was compared to several smaller doses delivered in 2 hour intervals: two times at 50 MOI, 4 times 25 MOI, etc. The results showed the same highly efficient transduction in all cultures. The titers were additive, which allows a seamless control of copy numbers in the cell population. The low MOI in the individual transductions even reduced any toxicity as washing steps can be introduced between transductions, and very high copy numbers can be built up with successive transductions. In general, cells can be transduced with a preferred MOI of less than or equal to 100 and greater than 10, with a more preferred MOI equal or greater than 25 and less than or equal to 75 and with a most preferred MOI of approximately 50 lentiviral particles by functional titer. The transductions are done multiple times over a 24 hour period with from twice during the 24 hours up to a maximum of 12 times in the 24 hour period.

We conducted another experiment to confirm that, if a second dose of Lentivector is added to the cell culture at least 2 hours after the first, the effectiveness of the second transduction is not compromised by the previous dose and that there is no or little interference between the two doses of Lentivectors. In this experiment (Figure 4), transduction with varying doses ( up to —100 MOI GFP titer) of a luciferase gene-carrying Lentivector (glue) was followed by administering 50 MOI of an Erythropoietin-encoding (EPO) Lentivector (that is otherwise identical to the first one) just 2 hours after the first administration. The copy numbers (represented by t-crit. values of respective qPCR assays) glue gene insertion titrated as expected, while the EPO insertion was not affected negatively by the presence of previous transduction step using luciferase encoding lentivector. This observation was true in the opposite direction as well; the completion of the insertion of the luciferase gene was not affected by the addition of EPO 2 hours later.

Subsequently, we clearly demonstrated that adding the Lentivectors in 2 hours intervals results in tightly controlled, reproducible delivery of the desired gene.

We concluded that administering Lentivectors repeatedly in rapid successions, preferably bi-hourly (but not shorter than 1 hour intervals) at relatively low MOI improves the efficiency of the vector delivery, increases the control over the copy numbers of transduced genes in the targeted cells, and substantially reduces the potential cytotoxic effects of the vector in vitro and in vivo.

A typical application of the procedure is presented in Figure 5. A target protein is selected, and the cDNA or the optimized DNA encoding the target protein is inserted into a plasmid that is used in conjunction with a packaging vector or cell line to make the

Lentivector. The biological activity of this vector is determined by titration to help in the selection of the appropriate MOI for transductions.

A typical transduction consists of the addition of consecutive doses of Lentivector (typically 10 to 100 MOI GFP titer equivalent ) to the cell culture. The vector is administered every 2 hours in the presence or absence of enhancing agents, such as Polybrene, depending on the sensitivity of the cell line. Cells may be washed between the repeated addition of the vector and should be washed at the end of the day to remove the excess lentiviral particles, transduction enhancers and cell debris. Washing is by pipetting off the solution from adherent cells and adding fresh medium. For suspension cells, we generally centrifuge the cells, pipette of the liquid and add fresh medium. Following one or more days of transductions, the cells can be bulk cultured, or single cell cloned by any means appropriate, including limited dilution or cell sorting, and the transductions can be repeated with a bulk cultured cells or selected single cell clones to further increase the copy numbers and over expression of the selected protein. Figure 5 shows a general scheme for such transductions with three days of multiple transductions (from 2 to 12 each day) followed by culturing the cells for three days in fresh medium and then either growing as a bulk culture or initiation of single cell cloning (or both in parallel). Rapid transductions, i.e. multiple transductions per day over several days can rapidly and more robustly lead to a higher copy number than multiple transductions over a longer period of time. Since the transduced cell will have limited time for cell division, one has a greater effective MOI using the same amount of vector on essentially the same number of cells. Example 2: Repeated Transduction of DG44 Cells with Human rEPO

We repeatedly transduced DG44 (CHO-cell line derived) cells with a human recombinant Erythropoietin (EPO) gene and generated single cell clones. The clone EPO- 118 was selected for in situ hybridization (FISH) assay.

Two cell lines were used for the in situ hybridization study. DG44 (a control, untransduced cell line) and DG44 EPO-118 (the same cell line multiply transduced with a lentiviral EPO vector. The cell lines were maintained and expanded in CD-CHO media. A single passage was performed on each line in order to seed the cells for Colcemid® treatment. Mitotic division was arrested using a Colcemid® treatment for 12 hours for each cell line. The cells were harvested, treated with hypotonic solution, and fixed in methanol/acetic acid. In each cell line we observed a high number of metaphase spreads suitable for FISH analysis.

Metaphase spreads were prepared for the DG44 and DG44 EPO-1 18 cell lines. Spread were aged and viewed by phase contrast microscopy prior to hybridization to ensure sufficient spreading of chromosomes. The provided probe to a portion of the lentiviral vector was labeled by nick translation using Spectrum Green-dUTP. Labeled probe was purified before hybridization to remove unincorporated nucleotides.

Slides and probes were denatured prior to hybridization. Hybridizations were performed in a standard hybridization buffer for 18 hours. Slides were washed in a low stringency washing buffer and mounted in mounting medium. Parental DG44 and the EPO cell line were treated equivalently in all experiments.

Slides were visualized at 6OX magnification and individual images were captured using filters appropriate for DAPI or Spectrum Green. There was no hybridization of the probe to the control DG44 but the DG44 EPO-118 line showed many fluorescent dots on various chromosomes indicating regions of insertion of the lentiviral vectors and hybridization with the probe (Figure 5A). We found that a single nucleus harbors a high number of insertions (over 40 noted on one image) and the copies are distributed on many chromosomes within the outline of the nucleus. It was also clearly noted that in several cases, where insertion occurred in one arm of a chromosome, it also was noted in the other corresponding arm — both arms of the chromosome were transduced. This is further indication of the high efficiency of the transduction process. Such high levels of transduction, where transduction occurs in both arms (or region) of the chromosome, has never been seen before. This should result in a more balanced expression of mRNA, and cells should be able to over express a properly glycosylated form the human recombinant Erythropoietin. The cells were repeatedly transduced 16 times and over expressed EPO, such that in high density cultures, over 10 g/L of EPO was produced. When the molecular weight of the EPO was assessed in Western blots, the migration of the EPO was similar to that of commercially available, purified, fully glycosylated EPO standard (Figures 6 and 7).

Example 2A: Genetic Modification of umbilical mesenchymal stem cells (UMSCs)

Lentiviral vectors transduce primary human cells with extraordinary efficiency. In studies with primary HUCPVC cells, we have shown that essentially 100% of cells can be genetically modified with an HIV-based lentiviral vector, as assayed using GFP as a reporter gene. A Lentiviral vector expressing the GFP gene (LTGl 73) was transduced onto

HUCPVC cells. While no cells were positive in mock-transduced cultures, essentially 100% of the cells were GFP positive in cells transduced with the Lentiviral vector expressing GFP. (HUCPVCs are a type of stem cell - a "mesenchymal stromal cell" (MSC) - also called mesenchymal stem cell — cells which have many varied and special properties.) MSCs give rise to specialized cells in the body that make the musculoskeletal system (bone, cartilage, fibrous tissue and muscle). Example 2B: Virus-Like Particle (VLPI Production for H5N1 Influenza

A tri-cistronic lentiviral vector, LTG985, was constructed using DNA synthesized at DNA2.0. DNA was codon-optimized for human expression. Genes from two influenza strains were synthesized: Vietnam 1203 and Indonesia 05. The 6.5 kb insert contained the Hemagglutinin (HA)5, Neuraminadase (NA)I, and Matrix 1 (Ml) genes, separated by self- cleaving 2 A sequences, and followed by IRES-puromycin (HA-2A-NA-2A-M1 -IRED- puromycin). Lentiviral vector with this sequence was generated with high titre (>lE8/ml), and used to transduce HEK293 human producer cells. Cells were transduced at an MOI of 50 at intervals of four hours, three times per day, for three days, with 4 ug/ml polybrene. Cells were allowed to recover, and then subjected to puromycin selection. Cells were passaged, frozen, and expanded for characterization. Characterization of the copy number indicated approximately 32 copies per cell after 12 transductions and up to 50 copies per cell after 15 transductions.

Cells were expanded in 0.5L NewBrunswick Fibrastage reactors, and harvested daily for 10 days. Harvests were concentrated by centrifugation and analyzed for HA and Ml protein by Western Blot. HA and Ml were detected in supernatant, and samples were also positive for HA by ELISA and Hemagglutination and NA, strongly arguing for intact particle expression. Example 3 : Increase in Frequency of High Producer Clones

Currently, to find a high producing cell line among the bulk cells transduced with plasmids or other viral vectors (other than retroviral vectors), one must single cell clone and testing thousands or tens of thousands of clones to identify those with high productivity. However, the rapid and predictable increase of copy number achieved through the use of Lentiviral vectors as well as the high cell transduction frequency reduces that task. Using the method we describe, it is sufficient to screen just a hundred clones or less to find high producers that occur at relatively high frequency. As illustrated in Figure 8, the rate of occurrence of high producer clones is almost 5%. Similarly, high producer cell lines were identified among as few as 120 single cell clones producing IgG. Limiting the screening to a few hundred clones allows us to use efficient methods that were previously only applicable with methods generally thought to be low throughput, such as HPLC, Surface Plasmon resonance, or Flow cytometry, cost-effectively compared to even the fully automatic ELISA techniques. That could save significant cost in clone selection and expands the choices for clone selection to areas previously not possible, such as inclusion of functional assays (enzyme activity), binding strength and specificity, etc. Example 3 A: Stability of high expressing clones

The high copy numbers generated by our transduction protocols did not diminish the stability of our over-expressing clones. We cultured a bulk culture of transduced HEK297 cells expressing green fluorescent protein (GFP) for 47 weeks without any selection or cloning and noted that there was little, if any dimunition of expression level or the number of expressing cells in the culture over this period of time. Similarly we have examined the stability of our high-expressing EPO clones over 5 Vz months with little or no dimunition of expression.

Example 3 B: Copy number of transduced cells and productivity

From the high EPO producing clones, we have characterized the number of inserted copies of the EPO gene per cell. The copy number in DG44 cells which underwent 6 transductions was 62, and increased to 94 by the 8^th transduction step, as determined by qPCR assay using a gag specific probe. The end result of high copy numbers achieved by repeated transductions is high productivity, which exceeded 80 pg/cell/day values. Combined with high cell density, very high volumetric productivity rates (5-20g/L) were achieved in perfusion cultures (CL-1000 or hollow fiber cartridges).

We selected 4 of the high producer clones for further testing in high density perfusion culture for volumetric productivity determination by inoculating CL-1000 tissue culture flasks with 5x10⁶/mL cells. The cells performed well, the viability was maintained high, 95- 96% and the cell concentrations reached 40-60xl0⁶/mL. The supernatants were harvested every 3 days while the cell concentration was reset to 2OxIO⁶AnL along with the cells when exceeded that in the previous period. The medium in the upper chamber was changed every 2 days. The supernatants were tested for erythropoietin in ELISA. Figure 6 shows that the productivity was consistently high in all 16 harvests tested, except for E- 104, which had a median erythropoietin production of 4200 mg/L, while the other 3 clones (E-108, E-I lO and E-118) the median productivity was significantly higher: 13880± 2893, 16520± 6800 and 13920± 9237 mg/L, respectively.

The high specific productivity rates were preserved for at least 5 ¹A months as the experiments comparing early and late passages of the same 4 clones showed no statistically significant differences in average duplication times or specific productivity rates. The stability of the cell lines was also assessed by comparing the clones from 2^nd passage and after maintaining the corresponding cell line for over 5 Vz months with nearly bi-weekly serial passages at 1 : 10 dilutions in completed ProCho5 medium, without any selection or re- cloning. One million cells from each of the four selected clones was plated in 1 mL triplicates using cells from the early passage as well as from the passages 5 Vz later in 12 well tissue culture plates and kept at 37, 34 or 31⁰C for four days (36 cultures in total). The cells were counted at the end of the fourth day and samples were taken from supernatants to measure the erythropoietin content, apparent molecular weight and purity using BioAnalyser. The overall viability of the cells after four days of culture was excellent: >90% (96.3+- 2.79). The average purity of erythropoietin regarding other proteins in the supernatants was, as previously noted, extraordinary, 98.1% +-0.798 and did not change with the number of passages. To see possible trends, we calculated the mean doubling times of the cells for the four cell lines. The increase in number of passages tended to decrease the variability of doubling times by different clones and we found a convergence from 56.2± 14.3 to 51.6±5.3 hours, but the change in averages is not significant in paired t-test; regardless to number of passages, the doubling times were much longer than the parental DG44 cell line (23 hours), a value less than half of the average doubling times of the derived erythropoietin- secreting clones, indicating that producing erythropoietin at these levels is overwhelming the cells. The average specific productivity data support this hypothesis: the mean productivity for early passages was 156±63 pg/cell per day and for the cell lines maintained in culture for 5 ¹A months was 180±60 pg/cell per day, a difference statistically not significant.

We picked randomly 2 harvests for each clone and tested the purity and apparent molecular weight of the produced erythropoietin using BioAnalizer 2000 and P230 protein capillary electrophoresis chips. The purity of the supernatant was surprisingly high, the peaks corresponding to erythropoietin dominated the detectable protein population and the purity consistently exceeded 95% (average 98% ± 1.72). The chip also allowed direct quantitation using low and high molecular weight internal standards. The analysis of samples confirmed the high productivity rates being similar to concentrations determined by ELISA; mapping the median erythropoietin concentration into 8-11,000 mg/L range. The apparent molecular weight determined by BioAnalyser capillary electrophoresis was somewhat higher than expected, 37.5kDa, with a very low standard deviation (0.372). The higher than expected molecular weight for a fully glycosylated erythropoietin (37.5 kDa) is most probably caused by the interaction of heavily glycosylated (and charged) erythropoietin with the gel as well as with the gel in the capillary system indicating consistently high glycosylation of the samples. Similar increase in apparent molecular weight can be seen in Western blots (Figure 6, panel B.) in which the erythropoietin samples were detected at the molecular weight corresponding the purified, fully glycosylated human recombinant erythropoietin standard purchased from SIGMA. The very low standard deviation for the apparent molecular weight values over the set of erythropoietin samples produced by different clones in different cultures at different times indicates substantial stability and reproducibility of production of recombinant erythropoietin using the Lenti vector-based system.

To verify the identity and glycosylation we purified the supernatant from E-118 clone supernatant using size exclusion chromatography as first step. The high erythropoietin concentration and excellent initial purity (>95%) allowed us to use SEC early in the process and combine purification with buffer change in single step, "desalting mode" in which the column was partially overloaded by using as much as 0.8 mL of sample in a 8 mL column (approximately 1/10 of column volume) resulting in elimination of the low molecular weight components of ProCho5 medium eluted at and after 11 minutes (Figure 7, panel A). The increased load resulted in substantial peak widening without deteriorating the erythropoietin- peak recovery, which took 1.5 minutes (6-7.5 minutes), corresponding to 7.5 mL fractions. Western blot analysis confirmed that erythropoietin was mainly eluted in that 2^nd peak at 7.1 minute. The buffer used for replacing ProCho5 medium was 5OmM phosphate buffer containing 1OmM TRIS, pH 6.8, recommended as formulation buffer as erythropoietin was found to be stabile in it, substantially free or aggregation [refj. The same buffer was also appropriate to be used as loading and buffer DEAE cellulose ion exchange chromatography. The column was loaded with 35 mL of erythropoietin - fraction collected from 5 SEC runs, and eluted with linear salt gradient. After washing the column with 20OmL loading buffer, erythropoietin was eluted at 10-15% of salt gradient in the first and largest peak (>95% of the total eluate peak areas) with 2 additional minor peaks at 25 and 28% of salt but all 3 of the eluted peaks contained erythropoietin as shown by Western blot analysis indicating that only a minor fraction of erythropoietin was isomorphic, separable by ion exchange chromatography. To assess the purity of the main peak, the sample was analysed using Protein 230 chip on Bio Analyser. The data showed that the purity of the main erythropoietin fraction (Pl) was at least 99.3.

Example 4: Quasi-Simultaneous Transductions with More Than One Gene

Our method allows a quasi-simultaneous transduction of the same cell suspension and same set of cells with two or more genes by alternating the vector added to the culture in intervals more than one hour, typically 2 hours. An example of use of this methodology is manufacturing a protein that consists of two or more chains, such as the light and heavy chains of immunoglobulin genes. By modifying the amount of heavy and light chain encoding plasmids, a precise and optimal ratio of heavy and light chains can be achieved. Graphing IgG Production vs. heavy chain copy number indicates that there is a strong correlation between the number of IgG heavy chains inserted into the genome and the amount of IgG produced by the cells. However the correlation only explains approximately 60% of change, the rest depends on the cell line and the number of light chain genes expressed alongside the heavy chains. Indicating that most of the clones can be further improved by adjusting the heavy -light chain ratio and choosing a better cell line for IgG expression. Graphing IgG Production vs. light chain copy number graph indicates that increasing the number of light chains can only increase IgG moderately, and has value only in conjunction with sufficient number of heavy chains which is the main rate-limiting factor in IgG expression.

A strong correlation (R²=0.663) exists between the heavy chain copy numbers and overall productivity in single cell clones. The correlation is far weaker when the light chains are concerned (R²=0.354) but still perceptible. Increasing the L/H chain ratio further improved the productivity in antibody production, However, it was largely masked by the strong influence of heavy chain copy numbers. By analyzing the resulting clones, the clones that have a bottle neck in protein production — folding, assembly and secretion - can be identified through accumulation intracellularly, while those that efficiently secrete the heavy and light chains as antibodies with the appropriate structure can be identified and improved further. Clone 4 produced significant amount of H and L chains, and assembled successfully into IHlL heteromers. However the heteromers do not dimerized (2H2L) and there is very little secreted IgG. The clone 3 lysate indicates that the cell does not produce sufficient amount of H and L chains, and although these assemble into the heteromer, the assembly of dimers is also inefficient, but most of what is made is secreted efficiently into the medium. The clone 2 produces both the H and L chains but the heteromer is not accumulated in the cells, all that is formed, is converted into dimers and immediately secreted, we see accumulation in the supernatant. Clone 1 in contrast accumulates the H and L chains in form of dimers and secrets the IgG very efficiently into the supernatant. Increasing the copy numbers in the case of clone 4 and 3 apparently could not help increasing the productivity since the cell line itself has bottlenecks in the assembly of the complex IgG molecule or the secretion of the dimers, however the clone 1 and clone 2 which efficiently assemble and secrete, could be improved by increasing the copy numbers IgG H and /or L chains or changing their ratios. When we examined the copy number of the heavy and light chains in the transduced cells, we determined using PCR probes specific for the gene that the individual clones contain the copy numbers ranging from 19 to 91.

Example 5: DG44 str. a Cell line Optimized for Lentivector-Based Protein Production Cells respond to a great number of stressors by undergoing programmed cell death.

This process is rather complex, but the diverse apoptosis-inducing signals converge into a relatively well defined but redundant apoptotic pathway. Many pathways and signals lead to apoptosis, but there is only one mechanism that actually causes the death of a cell. After a cell receives stimulus, it undergoes organized degradation of cellular organelles by activated proteolytic caspases.

Apoptotic cell death is a very important factor that puts limits on achievable cell densities, limits stress endurance of cells in terms of low pH starvation, and if apoptosis is induced, limits the ultimate level of production of the desired over-expressed protein(s). Developing a cell line resistant to apoptotic signals would be an improvement over standard cell lines as it would improve the ability of the cell to grow and produce proteins under stressful condition ubiquitous in high density cell cultures.

Apoptosis can be induced by an alkaloid, staurosporine, isolated from Streptomyces. The mechanism of staurosporine-induced apoptosis is complex. Staurosporine may have multiple points of attack, all resulting in apoptosis. Perhaps one could even activate the intrinsic apoptotic pathway, the one associated with mitochondria. If so, staurosporine could be used to select a cell line that will be defective in this apoptotic pathway or have more resistant mitochondria, which could have several positive consequences for protein production and survival of the cell line under conditions when the mitochondria would normally trigger apoptotic cell death. Since staurosporine seems to activate multiple pathways that result in apoptotic cell death, the number of variants in the cell population will be rather heterogeneous. These changes could be very valuable for special purposes, including protein production. Staurosporine treatment allowed us to generate a great variety of cells that can serve as basis for further selection of clones of interest. In order to develop a staurosporine-resistant cell line, serial dilution of staurosporine was made in 6 well plates, and each well was seeded with 10⁶/mL DG44 cells. Initially, the LD₅₀ for staurosporine on DG44 cells was 0.1-5 nM, with significant toxicity even below a single nM. The staurosporine-induced apoptosis was compete in 3 days; all or most of the cells died. The cells were left in staurosporine-containing medium until cell division was seen in one or more wells, than collected from the well that had the highest staurosporine concentration. This step was repeated for 4 months. The resulting cell line showed an approximately 1,000-10,000 fold increase in resistance to staurosporine as shown in Figure 9. The question remains, whether these cells have the ability to grow properly and have better features than the original DG44 cell line. The growth curves of DG44 and DG44 str have been compared in shaker flask cultures (Figure 10). The tissue culture medium was ProCho5, the seeding density 100,000 cells per mL. Under these conditions, the DG44 cell culture reached its peak in 120 hours and had a very short plateau of high viability (48 hours). The str line reached its peak earlier, indicating a significantly shorter doubling time, and the plateau of high viability lasted for several days (typically 4-6).

The bulk cell line was also tested for response to the staurosporine treatment by using the Guava apoptosis kit that measures the activity of multiple caspases. The assays employ a FLICA (Fluorescent Labeled Inhibitor of Caspases) reagent that specifically identifies active Caspase 3 and 7 molecules or Caspase 8 molecules. Each FLICA is comprised of three subunits: a caspase-specific amino acid inhibitor containing a recognition sequence for Caspase 3/7 (DEVD) or Caspase 8 (LETD); a fluoromethylketone moiety (FMK) that forms a covalent linkage with the active enzyme; and a carboxyfluorescein (FAM) reporter. The nuclear DNA stain, propidium iodide (PI), is also included in the assay to simultaneously evaluate membrane integrity and cell viability. The population of CD44 str cells was clearly highly heterogeneous: live cells both with and without activated caspase-3 were present, indicating that in the cell line many different clones with different handicaps in apoptotic pathway were present. This observation made it important to clone the best suited sub-lines for protein production from the bulk cell culture. The growth rates and the maximum cell concentration of 42 resistant single cell clones have been determined. The growth rates were substantially higher than that of in the original DG44 cell line. The doubling times were significantly lower than the 23-24 hours in the DG44 parental cell line. The average in the clones was 13.5 hours, with some clones close to 10-11 hours. While proliferating faster, the size of the str cells is not reduced as can be estimated by the Forward side scatter parameters measured in flow cytometry, and one expects that the biomass produced by the str clone sis larger than by the parental DG44 cell line in a give time. The ability to be transduced by lentivector is similar in both cell lines and the same multiplicity of infection (MOI) results in equal or only slightly lower transduction rate than what observed in the case of the DG44 cell.

This is a significant change, since the shortened doubling time is not a result of generating smaller cells. On the contrary, the flow cytometric measurements indicated that the cell diameters of the str resistant clones are not smaller, but are of the same size or even slightly larger than the parental cell line. It indicates that these resistant cells produce greater biomass in a given time than the original DG44 line; i.e., have a significantly increased protein production rate. This indicates that the metabolic rate or utilization of the resources by the cells may have changed as well; consequently these cells seem to be good candidates for protein manufacturing.

Another result from the lower doubling time is a reduction in the time needed for single cell cloning and bulking up single cell clones. The change is significant, especially in the case of single cell cloning steps, which need to be followed by expanding the cell population as the growth rate increases from 3% to 5.5%. This affects the number of cells minimally necessary for transductions as fewer cells are sufficient to achieve the same final cell number or the time to test the clones is shorter, which speeds up the clone development significantly. It takes 15-18 days to have 100,000 cells usually needed for the parental cell line and only 9-10 days for the str cell line. The same advantage is realized in every stage of biomass production in scaled up tissue cultures. To use the new cell line for protein production, we also tested the ability of the resistant clones to be transduced by Lentivectors compared to clones of the original DG44 line. The experiments showed that the resistant clones are similar in their ability to be transduced as the original DG44 line.

An additional advantage of the DG44 str clones is seen during transduction. When DG44 and many other cell lines are transduced at high MOI in the presence of polybrene, a significant fraction of the cells (ranging from 5 to 50 %) are killed and the population of viable cells frequently have a long lag period before they resume growth. In contrast, transduction of DG44str resulted in significantly greater fraction of viable cells (approximately95%), and these cells show little or no lag period as they resume growth essentially immediately. The question that remained is whether the genes integrated into the str clones are expressed as efficiently as in the parental cell line.

Example 6: Single cell clones of DG44 str cell lines selected for protein production.

To answer this question, we transduced the cells with LTGl 73, a GFP encoding vector and compared the fluorescence intensities in the cell, assuming that the fluorescence intensity measured in flow cytometry is proportional with the amount of GFP produced. The total amount of product is estimated by multiplying the fluorescence intensity with the number of cells in the culture. The resulting indices show that most of the str clones are equal or even superior in productivity to the parental DG44 cell line. The ranking allows us to identify the most promising clones among the str clones.

To estimate the protein productivity, we considered the fact that these cells divide very rapidly, and there is no time for the GFP to accumulate in the cell. It would be more justified to compare the production rate per biomass than to compare simple fluorescence intensities. The measure of productivity in this calculation is the number of cells in the culture multiplied by the fluorescence intensity: indeed half of the DG44 str clones made significantly more GFP per unit volume of tissue culture than the parental DG44 cell line .

Example 7: Increasing the Productivity of DG44 str Cell Line by Reducing the Proliferation Rate

We believed that if we could reduce the proliferation rate, these cells would be able to accumulate more GFP and the production rate per cell could significantly increase. One method to achieve such a goal is to culture the cells at lower temperature. This increases the doubling time and slows the cell proliferation while the GFP, which is driven by a strong promoter which is not affected by the lower temperatur, would not be affected to the same extent. The experiment comparing the GFP expression in selected DG44 str clones indicated that this method indeed results in rapid accumulation of GFP in the cytoplasm: the fluorescence intensity increased by more than 6,000 fold after just 4 days in culture at 34° C as compared to 37°C. Similarly, when recombinant human-porcine Factor VIII was produced by the same cells, reducing the temperature resulted in a 30-50% increase in the amount of secreted protein. (Figure 11) We have also evaluated whether other steps necessary for effective protein production are as efficient in the DG44 str clones as in the parental cell line. Namely, folding and secretion of the target protein product. To evaluate the ability to make and secrete proteins, we chose 6 DG44str clones for further evaluation: 8, 14, 15, 28, 30 and 42. We transduced these clones 9 times under very demanding conditions: 5000 MOI of Factor Vlll-encoding lentivector (a toxic level when research-grade, unpurified Lentivector is used) in the presence of 8 mM polybrene (known to be toxic to DG44 cells). All colonies survived the transduction, except the parental DG44 cell line in which the viability was reduced to a few surviving cells ( -1%). The remaining cell lines proliferated for 5-6 days. However, at that point the cells in 3 out of 6 clones (8, 28 and 30) vacuolated, shriveled and died around days 8-10, even those in single cell cloning plates (96 well). This, we assumed, was the result of accumulation of intracellular rhpFVIII, while the rest of the clones propagated and expanded as expected. A very few cells from the parental line have also survived, and we were able to rescue a dozen clones in the following 2 months. A set of single cell clones have been tested for their ability to secrete functionally active recombinant human-porcine factor VIII.

Recombinant Factor VIII is considered a very difficult protein to produce for several reasons. First, it is a large protein, containing over 1400 amino acids; it is difficult to fold correctly; and cell lines secreting it have been found to produce very low yields. In addition, it seems to stress the cells strongly, resulting in loss in cell viability if any attempt is made to over express it. When compared to reported Factor VIII productivity (Herlitska S. E, Schocklat U., Falkner F., Dorner F. High Expression of a B-domain deleted rhpFVIII gene in human hepatic cell line. Biotechnology 61, 165-173 (1998).), the levels we achieved are significantly higher. In addition, the differences in productivities are minimal between the different clones, indicating that the cells obtained after transductions by the Lentivectors are rather uniform. There was no difference in productivity between the unselected bulk culture and the selected single cell clones. When comparing the productivities, we find that only a few mU of Human factor VIII are produced by retrovirus-transduced cell lines. Expressing the rhpFVIII with transient, plasmid-based systems increased the productivity by approximately 50-100 fold, but the Lentivector-based systems yielded an additional 10-fold increase. Also, a stable cell line has been isolated that can be used continuously and in optimized high density cultures. Using one of our FVIII expressing STR cell lines (DG44str42.7) in high density culture resulted an additional 5-to 10 fold increase even under non-optimized conditions, totaling over 50-100,000 mU of rhpFVIII activity per mL of culture supernatant. (Figure 12). Example 8: Efficient Competitive Immunoassay to Assess Protein Production The protein production platform we have developed typically results in variation in protein production from clone to clone. The frequency of high producing clones isolated from the bulk culture, however, following the repeated transduction with Lentivectors is rather high. This reduces the number of clones that need to be tested compared to other transduction methods from thousands or tens of thousands to hundreds or even less than a hundred clones.

On the other hand, it poses new challenges, namely the protein concentration in the supernatant of high-producer cells is high. Some traditional assays (ELISA, bead-assay that are optimized for assaying minuscule amounts of proteins) may produce false data or require serial dilutions of the supernatant, which increases the workload and the time needed for the screening. An example of such a problem is the ELISA assays for the lentiviral protein p24, which measure in the range of tens-hundreds of pg/mL of p24, and when Lentivectors or lentivirus-based vaccines are manufactured and purified, the p24 concentrations may reach hundreds of ng/mL range or higher. In the case of measuring Erythropoietin production that reached 5-20 mg/mL, one would have to dilute our sample 10,000,000 fold to be in the linear range of the ELISA kit currently available on the market. The HPLC-based methods are often not appropriate because the supernatant carries a large number of proteins that may interfere with the measurement, and the assay validation could be impossible or at least very difficult. In addition, all of the methods mentioned above require a positive control and a purified sample of the target protein. Often, it is not available and/or there is no established method for purification of the protein to begin with.

Using a competitive bead assay technology can overcome most of these issues. See Figure 13, panel C. The idea is to generate two version of a protein. One is tagged and the other is not. Tagging is used to efficiently coat the surface of wells, beads, microspheres, or liposomes (carriers, if beads or microspheres should be of 0.1-100 uM in diameter) with the protein of interest. We used an appropriately diluted ( but sufficient to saturate the binding sites on the carrier) directly labeled ligand, which recognizes and binds with high affinity to the protein of interest as well as to the tagged protein (i.e. to some shared sequence in the proteins). The signal is generated by the directly labeled ligand on the carrier surface after binding to the tagged version of the protein is the reference signal. Then, in a separate procedure, a sample of untagged protein of interest is added to the beads prior to adding the directly labeled ligand. This will cause interference and reduces the binding of the ligand to the proteins on the bead by the nature of competition for binding sites, and thereby reduces the signal generated by the labeled antibody on the surface of the bead. The difference between the reference signal and the second signal correlates with the amount of untagged protein of interest in the sample.

In order to change the sensitivity of the system and expand the linear range in order to quantitate large concentrations of proteins of interest, two factors can be varied: the amount of the carriers per sample and the excess of the ligand.

The timing of the development of reagents for immunoassays is such that it does not cause significant delays in the single cell cloning of the untagged version of the protein. We used this method to measure murine IgG antibody production by lentivector- transduced mammalian cells. In this case, the tag is the constant domains of the antibody

(murine gamma chain, constant region), and the protein of interest is an immunoglobulin with different variable region having a unique F(ab) or F(ab)2 specific for each product. Microspheres (PolyScience, 5uM in diameter) were coated with a murine IgG by standard (NHS) amino coupling method. A serial dilution of FITC -labeled anti-murine IgG antibody was added to aliquots of the microspheres ( 100 uL of 10⁶/mL of microspheres per well in a 96 well plate) and incubated without washing until the maximal fluorescence intensity was detected using flow cytometry, ensuring that the FITC-labeled anti murine IgG is in slight excess. Known amounts of murine igG is used as a calibration standard for the batch of beads in the range of 1-100 ug/mL. Tissue culture supernatants were directly added to wells, 20 uL to each well.

The octameric peptide StrepTag: Trp-Ser-His-Pro-Gln-Phe-Glu-Lys (US Patent No. 6,103,493) in conjunction with streptavidin or its engineered version, StrepTactin, is used widely for chromatographic purposes and for detecting StrepTag-tagged proteins. A new use of the system relies on the fact that the composition and the encoding nucleotide sequence of the protein of interest (P) are already known: a. When cloning a protein of interest (P) into a Lentivector, it is possible to generate an additional variant that carries as direct label a StrepTag sequence (P-S fusion protein). See Figure 15. a. Transduce two cell cultures with the tagged and untagged genes inserted into different lenti viral vectors. b. The cells transduced with the untagged protein undergo standard protein- expressing cell line development. c. The cells with the tagged version of the protein are rapidly bulked up (because the Lentivectors are very efficient and result in stable cell lines) and the supernatant can be used as source of P-S. Beads coated with StrepTactin or commercially available affinity columns can be used to collect the secreted P- S from the supernatant (or cell lysate) and purify the protein using a single step affinity procedure. d. The P-S protein can be used to immunize animals to get appropriate antiserum (3-4 weeks are sufficient for this step). e. The antiserum can be affinity purified if necessary, and the anti-tag antibodies can be removed in a second step. f. The purified antibodies can be directly labeled with enzymes (such as horse radish peroxidase, HRPO) or fluorochromes (such as FITC or A488) and used with Western blots and other immunoassays. g. StrepTactin- coated beads, which capture and are saturated with P-S, can be used as reagents in competitive bead assays.

This procedure can be used for new proteins of interest without developing new purification techniques. The time needed for the development of reagents will be very short, in fact so short that it can be done in parallel with the single cell cloning of the untagged protein. Comparison with Standard Assays for Protein Concentration

Currently, the high throughput assays used generally for determining protein production by mammalian cell lines rely on various variants of automated forms of ELISA assays. The usual configuration uses a capturing antibody-coated well that binds the ligand from the culture supernatant. The captured ligand is bound to a second, enzyme-labeled antibody, specific for other than the capturing epitope, which is then used to estimate the amount of ligand captured. The assay however, suffers from a number of problems:

1) It is labor intensive and often uses costly robotics to assay thousands or tens of thousands of samples in search of high producer clones.

2) The assay includes at least 6-8 washing steps, which increases the potential for errors and generates waste.

3) A number of parallel samples (usually 3-5 wells per sample) are required, due to limited accuracy at given concentrations.

4) It requires two independent set of antibodies. 5) It requires a separate calibration curve for each plate.

6) The ELISA assays are highly sensitive, and the range of linearity is limited to very low concentrations (nano grams per well, ug/mL), much lower than what is expected from a highly efficient protein producing cell line accumulating several milligrams of the target protein per mL. (Figure 13, panel A.) This, in turn, requires that a serial dilution of the sample be done, which increases the number of wells further by 3-5 fold. These assays require purified protein for calibration curves and purified antigen for antibody production; i.e., access to the protein that is often not available commercially before developing the cell lines that will be used for production. The inadequacies of the ELISA type assays can be partially overcome by the bead assays and flow cytometry (Figure 13, panel B). Although the number of samples that can be processed by flow cytometry may be lower than the number of ELISA plates because the read-out is serial, it is partially compensated for by two factors: 1) the accuracy of the assay is so high that the number of parallel samples is low; often a single value gives acceptable results and 2) the washing steps can be eliminated. In a typical commercially available bead assay, the capturing antibody is conjugated to the bead surface and, after capturing the analyte, a second, fluorochome-labeled antibody.

However, the bead assay has drawbacks for high concentrations of protein. The inadequacy of the bead assay for high concentrations of the analyte is manifested in variants of Hook-effect, which results in strong ambiguity when the resulting data are evaluated. One cannot distinguish between data obtained from the high producer clones and the low producer clones (Figure 19, panel B). The Hook effect occurs as follows. As one increases the concentration of analyte in the sample, the intensity increases. However, beyond a certain point, the excess analyte results in a competition within the reaction and actually quenches the intensity. As a result, there are two places on the curve that can result in the same intensity, but they will differ by a significant amount in terms of amount of analyte. In contrast, our approach, as shown in Figure 13 C and in Figure 14, shows that, as the concentration of analyte increases, the top producing samples will ultimately result in a plateau of low intensity samples. These are selected for further analysis after appropriate dilution to identify the top producing samples.

In contrast, our assay is optimized for detecting and measuring high concentrations of proteins. It requires no purification of the analyte/ligand protein other than a single step affinity purification. All of the high producing clones can be identified. Then one can use the limited set of top clones with appropriate dilutions of the supematants to obtain accurate readings on protein production in the linear range of the assay. Therefore, this method allows a straightforward, labor and materials minimizing approach to identification of top clones.

REFERENCES

1. Werner,R.G., Noe,W., Kopp,K., & Schluter,M. Appropriate mammalian expression systems for biopharmaceuticals. Arzneimittelforschung. 48, 870-880 (1998).

2. Birch,J.R. & Racher,A.J. Antibody production. Adv. Drug Deliv. Rev. 58, 671-685 (2006).

3. Hesse,F. & Wagner,R. Developments and improvements in the manufacturing of human therapeutics with mammalian cell cultures. Trends Biotechnol. 18, 173-180 (2000).

4. Molowa,D.T. & Mazanet,R. The state of biopharmaceutical manufacturing. Biotechnol. Annu. Rev. 9, 285-302 (2003). 5. Reiter,M. & Bluml,G. Large-scale mammalian cell culture. Curr. Opin.

Biotechnol. 5, 175-179 (1994).

6. Chemier,J.A., Fowler,Z.L., Koffas,M.A., & Leonard,E. Trends in microbial synthesis of natural products and biofuels. Adv. Enzymol. Relat Areas MoI. Biol. 76, 151-

217 (2009). 7. Graumann,K. & Premstaller,A. Manufacturing of recombinant therapeutic proteins in microbial systems. Biotechnol. J. 1, 164-186 (2006).

8. Negrete,A. & Kotin,R.M. Strategies for manufacturing recombinant adeno- associated virus vectors for gene therapy applications exploiting baculovirus technology.

Brief. Funct. Genomic. Proteomic. 7, 303-311 (2008). 9. Cox,M.M. & Hollister,J.R. FluBlok, a next generation influenza vaccine manufactured in insect cells. Biologicals(2009).

10. Plasson,C. et al. Production of recombinant proteins in suspension-cultured plant cells. Methods MoI. Biol. 483, 145-161 (2009).

1 1. Mett,V., Farrance,C.E., Green,B.J., & Yusibov,V. Plants as biofactories. Biologicals 36, 354-358 (2008).

12. Dunker,A.K. et al. The unfoldomics decade: an update on intrinsically disordered proteins. BMC. Genomics 9 Suppl 2, Sl (2008). 13. Wandinger,S.K., Richter,K., & Buchner,J. The Hsp90 chaperone machinery. J. Biol. Chem. 283, 18473-18477 (2008).

14. Jenkins,N., Murphy,L., & Tyther,R. Post-translational modifications of recombinant proteins: significance for biopharmaceuticals. MoI. Biotechnol. 39, 1 13-118 (2008).

15. Erbacher,A., Gieseke,F., Handgretinger,R., & Muller,I. Dendritic cells: Functional aspects of glycosylation and lectins. Hum. Immunol. (2009).

16. Reading,P.C, Tate,M.D., Pickett,D.L., & Brooks,A.G. Glycosylation as a target for recognition of influenza viruses by the innate immune system. Adv. Exp. Med. Biol. 598, 279-292 (2007).

17. Kresse,G.B. Biosimilars - Science, status, and strategic perspective. Eur. J. Pharm. Biopharm.(2009).

18. Schellekens,H. Recombinant human erythropoietins, biosimilars and immunogenicity. J. Nephrol. 21, 497-502 (2008). 19. SoIa₅RJ. & Griebenow,K. Effects of glycosylation on the stability of protein pharmaceuticals. J. Pharm. Sci. 98, 1223-1245 (2009).

20. Werner,R.G., Kopp,K., & Schlueter,M. Glycosylation of therapeutic proteins in different production systems. Acta Paediatr. Suppl 96, 17-22 (2007).

21. Palomares,L.A., Estrada-Mondaca,S., & Ramirez,O.T. Production of recombinant proteins: challenges and solutions. Methods MoI. Biol. 267, 15-52 (2004).

22. Derouazi,M. et al. Genetic characterization of CHO production host DG44 and derivative recombinant cell lines. Biochem. Biophys. Res. Commun. 340, 1069-1077 (2006).

23. Schmidt,F.R. Recombinant expression systems in the pharmaceutical industry. Appl. Microbiol. Biotechnol. 65, 363-372 (2004). 24. Grillberger,L., Kreil,T.R., Nasr,S., & Reiter,M. Emerging trends in plasma- free manufacturing of recombinant protein therapeutics expressed in mammalian cells. Biotechnol. J. 4, 186-201 (2009).

25. Modlich et al. Cell-culture assays reveal the importance of retroviral vector design for insertional genotoxicity. Blood. 108, 2545-2553 (2009). 26. Montini et al. Hematopoietic stem cell gene transfer in a tumor-pronemouse model uncovers low genotoxicity of lentiviral vector integration. Nature Biotechnology. 24, 687-696 (2006). 27. Montini et al.The genotoxic potential of retroviral vectors is strongly modulated by vector design and integration site selection in a mouse model of HSC gene therapy J. Clinical Investigation, 119, 964-975 (2009).

28. Zychlinski et al. Physiological Promoters Reduce the Genotoxic Risk of Integrating Gene Vectors. Amer. Soc. Gene Therapy, 16, 718-725 (2008).

29. Levine et al. Gene transfer in humans using a conditionally replicating lentiviral vector. PNAS, 103, 17372-17377 (2006).

30. Wang et al. Analysis of Lentiviral Vector Integration in HIV+ Study Subjects Receiving Autologous Infusions of Gene Modified CD4+ T Cells. Amer. Soc. Gene Therapy, 17, 844-850 (2009).

All publications, including issued patents and published applications, and all database entries identified by url address or accession number are incorporated herein by reference in their entirety.

Although this invention has been described in relation to certain embodiments thereof, and many details have been set forth for purposes of illustration, it will be apparent to those skilled in the art that the invention is susceptible to additional embodiments and that certain of the details described herein may be varied considerably without departing from the basic principles of the invention.

Claims

WHAT IS CLAIMED IS:

1. A method for making a mammalian cell that produces a protein comprising the step of: transducing a mammalian cell with lentiviral transduction vectors at least 2 times in

24 hours, wherein each lentiviral transduction vector comprises an expressible polynucleotide coding for the protein.

2. The method of claim 1 wherein high levels of the protein are produced.

3. The method of claims 1 or 2 wherein the protein is produced without significant deglycosylation of the protein.

4. The method of claims 1-3 wherein the lentiviral transduction vector comprises two or more different expressible polynucleotides coding for two or more different proteins.

5. The method of claim 1-3 wherein the cell is transduced with a second lentiviral transduction vector that comprises a second expressible polynucleotide coding for a second protein.

6. The method of claims 4 or 5 wherein one polynucleotide codes for the heavy chain of an immunoglobulin molecule and the other polynucleotide codes for the light chain of an immunoglobulin molecule.

7. The method of claims 1-3 wherein the cell is transduced with a plurality of lentiviral transduction vectors wherein each vector comprises a different expressible polynucleotide coding for a different protein.

8. The method of claims 1-3 wherein the lentiviral transduction vector comprises an expressible heterologous polynucleotide coding for a heterologous protein.

9. The method of any one of claims 4-7 wherein at least one of the polynucleotides comprises a heterologous polynucleotide coding for a heterologous protein.

10. The method of any one of claims 1-9 wherein the protein is secreted from the transduced cell.

11. The method of any one of claims 1-10 wherein the number of transduction vectors integrated into the genome of each transduced cell is about 3 to about 50.

12. The method of claim 11 wherein the number of transduction vectors integrated into the genome of each transduced cell is about 20 to about 40.

13. The method of claims 11 or 12 wherein both arms of the same chromosome are transduced.

14. The method of any one of claims 1-13 wherein the cell is transduced at a MOI of less than or equal to 100.

15. The method of any one of claims 1-13 wherein the MOI is about 10 to about 99.

16. The method of any one of claims 1-13 wherein the MOI is about 10 to about 50.

17. The method of any one of claims 1-13 wherein the MOI is about 50 to about 99.

18. The method of any one of claims 1-13 wherein the MOI is about 50.

19. The method of any one of claims 1-18 wherein the cell is transduced at regular time intervals over the 24 hour period and wherein each time interval is about one hour to about twelve hours after the prior transduction.

20. The method of any one of claims 1-18 wherein the cell is transduced approximately every 6 hours over the 24 hour period.

21. The method of any one of claims 1-18 wherein the cell is transduced approximately every 2 hours over the 24 hour period.

22. The method of any one of claims 1-18 wherein the cell is transduced over a 24 hour to 96 hour period.

23. The method of any one of claims 1-18 wherein the cell is transduced over a 24 hour to 72 hour period.

24. The method of any one of claims 1-18 wherein the cell is transduced over a 24 hour to 48 hour period.

25. The method of any one of claims 1-24 wherein the cell is transduced in the presence of a transduction enhancing agent.

26. The method of claim 25 wherein the transduction enhancing agent is selected from the group consisting of Polybrene, PEI, and Lipofectin.

27. The method of any one of claims 1-26 wherein the protein is selected from the group consisting of: mammalian, human, viral, and bacterial proteins.

28. The method of claim 27 wherein the protein is a human protein.

29. The method of claim 27 wherein the heterologous protein is a viral protein.

30. The method of any one of claims 1-29 wherein the lentiviral transduction vectors are selected from the group consisting of HIV, SIV, EIAV, FIV, CAEV, and BIV vectors.

31. The method of any one of claims 1 -29 wherein the lentiviral transduction vectors are HIV vectors.

32. The method of any one of claims 1-30 wherein the mammalian cell that is transduced is an apoptosis-resistant cell produced by the process of: exposing mammalian cells to an apoptosis-inducing chemical; and selecting at least one cell that shows the characteristic of apoptosis-resistance.

33. The method of claim 32 wherein the apoptosis-inducing chemical is staurosporine.

34. The method of any one of claims 1-32 wherein the cell is a primary cell.

35. The method of any one of claims 1 -32 wherein the cell is selected from a cell line.

36. The method of any one of claims 1-32 wherein the cell is a human cell.

37. The method of any one of claims 1-32 wherein the cell is selected from the group consisting of stem cells, embryonic cells, 3T3 cells, CHO (immortalized Chinese hamster ovary) cells, CHO-derived DG44 (DHFR-) cells, umbilical mesenchymal stem cells (uMSCs), Scr cells, C 127 cells, CA cells, HeLa cells, NSO cells, HEK-293 (Human embryonic kidney) cells, BHK (Baby hamster kidney) cells, PER-C6 cells, SP2/0 cells, U937 cells, THP-I cells, Vero cells, and YAC-I cells.

38. The method of any one of claims 1-33 wherein the cell is selected from the group consisting of CHO, DG44, and HEK-293 cells.

39. The method of any one of claims 1-38 wherein multiple cells are transduced by the vectors and the transduced cells are screened to select a cell that produces high levels of the protein.

40. A method for transducing a host cell comprising the steps of: contacting a host cell with a plurality of lentiviral transduction vectors at a multiplicity of infection of less than or equal to 100 under conditions whereby at least two of the vectors integrate into the genome of the host cell, wherein the vectors comprise at least one polynucleotide operably linked to a promoter, said polynucleotide encoding a secreted protein, wherein the host cell does not produce infectious lentiviral particles; and clonally selecting a transduced host cell.

41. A method of transducing a host cell comprising the step of: contacting a host cell comprising least one expressible polynucleotide integrated into the genome of the host cell, wherein the host cell does not produce infectious lentiviral particles, with a plurality of lentiviral transduction vectors comprising the expressible polynucleotide, said polynucleotide encoding a secreted protein, at a multiplicity of infection of less than or equal to 100.

42. A mammalian cell produced by the method of any one of claims 1-41.

43. A cell line produced by cloning the mammalian cell of claim 42.

44. A cell culture produced by culturing the mammalian cell line of claim 43.

45. The cell culture of claim 44 wherein the concentration of the cells is about 10⁶ cells per mL to about 10 cells per mL.

46. The cell culture of claim 44 wherein the concentration of the cells is about 1 x 10⁷ cells per mL to about 6 x 10⁷ cells per mL.

47. The cell culture of claims 45 or 46 wherein the protein is not toxic to the cells and the cells produce at least about 5 pg/cell/day.

48. The cell culture of claim 45 or 46 wherein the protein is not toxic to the cells and the cells produce at least about 10 pg/cell/day.

49. The cell culture of claim 45 or 46 wherein the protein is not toxic to the cells and the cells produce from about 20pg/cell/day to about 80pg/cell/day.

50. The cell culture of claim 45 or 46 wherein the protein is not toxic to the cells and the cells produce over 80 pg/cell/day.

51. The cell culture of any one of claims 44-50 wherein the culture is maintained at about 30° C to about 36° C.

52. The cell culture of any one of claims 44-50 wherein the culture is maintained at about 32° C to about 34° C.

53. The cell culture of any one of claims 44-50 wherein the culture is maintained at about 34° C.

54. A mammalian cell line comprising at least 3 integrated copies per cell of an expressible polynucleotide encoding a protein, said copies of the polynucleotide having been inserted into the genome without using a gene amplification technique.

55. The cell line of claim 54 wherein the cells produce the protein without significant deglycosylation of the protein.

56. The cell line of claims 54 or 55 comprising at least 5 integrated copies per cell of the polynucleotide.

57. The cell line of claims 54 or 55 comprising at least 10 integrated copies per cell of the polynucleotide.

58. The cell line of any one of claims 54 or 55 wherein the cells comprise from about 3 to about 50 integrated copies of the polynucleotide.

59. The cell line of claim 58 wherein the cells comprise from about 20 to about 40 integrated copies of the polynucleotide.

60. The cell line of any one of claims 54-59 wherein the polynucleotide comprises a heterologous polynucleotide coding for a heterologous protein.

61. The cell line of any one of claims 54-60 wherein the protein is secreted from the cells.

62. The cell line of any one of claims 54-61 wherein the protein is selected from the group consisting of: mammalian, human, viral, and bacterial proteins.

63. The cell line of claim 62 wherein the protein is a human protein.

64. The cell line of claim 62 wherein the protein is a viral protein.

65. The cell line of any one of claims 54-64 wherein the cells are resistant to apoptosis.

66. A clonally selected mammalian host cell comprising from about 20 to about 50 lentiviral transduction vectors integrated into the genome of the host cell, wherein the integrated vectors comprise at least one polynucleotide operably linked to a promoter, said polynucleotide encoding a secreted protein, wherein the host cell does not produce infectious lentiviral particles.

67. A mammalian cell useful for high density cell cultures, said cell produced by the process comprising the steps of: exposing mammalian cells to an apoptosis-inducing chemical; selecting at least one cell that shows the characteristic of apoptosis-resistance.

68. The cell of claim 67 wherein the apoptosis-inducing chemical is staurosporine.

69. An apoptosis-resistant mammalian cell line produced by cloning the cell of claims 67 or 68 to form a cell line.

70. A cell culture produced by culturing the mammalian cell line of claim 69.

71. The cell culture of claim 70 wherein the cells are about 1 ,000 - 10,000 more resistant to the apoptosis-inducing chemical than the starting culture of the mammalian cells prior to being exposed to the apoptosis-inducing chemical.

72. The cell culture of claims 70 or 71 wherein the doubling time of the culture of apoptosis-resistant cells is about 95% to about 50% of the doubling time of a culture of the original mammalian cells.

73. A method for making a protein comprising the step of: culturing the cell culture of any one of claims 44-53 under conditions effective to produce the protein.

74. The method of claim 70 wherein high levels of the protein are produced.

75. The method of claim 73 or 74 wherein the protein is produced without significant deglycosylation of the protein.

76. The method of claim 73 further comprising the step of isolating the protein from the culture.

77. The method of claim 76 further comprising the step of purifying the isolated protein.

78. A method of producing a secretable protein comprising the steps of: providing a clonally selected host cells wherein the genome of the host cells comprises at least about 50 integrated copies of at least one integrated lentiviral transduction vector comprising a polynucleotide operably linked to a promoter, wherein the polynucleotide encodes the secretable protein , and culturing the host cells under conditions whereby the protein is produced at rate greater than about 50 picograms per cell per day.

79. An assay for measuring the concentration of a protein in a sample having a high concentration of the protein, comprising the steps of: contacting molecules of the protein, which are bound to the surface of a carrier, with labeled ligands to create complexes comprising the labeled ligands bound to the proteins wherein all of the molecules of bound protein are complexed with the labeled ligands; measuring the signal produced by the labeled ligands bound to the proteins thereby creating a reference signal; contacting molecules of the protein in the sample to be assayed with the labeled ligands to form complexes comprising the labeled ligands bound to the proteins, which complexes are not bound to the surface of the carrier; measuring the signal produced by the labeled ligands bound to the proteins; calculating the difference between the two signals; and correlating the difference with the concentration of the proteins in the sample.

80. The assay of claim 79 wherein the molecules of the proteins are bound to the surface of the carrier through tags incorporated into the proteins and tag-specific ligands bound to the surface of the carrier, and comprising the steps of: contacting the tagged proteins that are bound to the surface of the carrier with labeled ligands to create complexes comprising the labeled ligands bound to the tagged proteins wherein all of the molecules of bound protein are complexed with the labeled ligands; measuring the signal produced by the labeled ligands bound to the tagged proteins thereby creating a reference signal; contacting untagged molecules of the protein in the sample to be assayed with the labeled ligands to create complexes comprising the labeled ligands and the untagged proteins, which complexes are not bound to the surface of the carrier; measuring the signal produced by the labeled ligands bound to the untagged proteins that are not bound to the surface of the carrier; calculating the difference between the two signals; and correlating the difference with the concentration of the proteins in the sample.

81. The assay of claims 79 or 80 wherein the concentration of protein is from about 20 ug/mL to about 10 mg/mL.

82. The assay of claims 79 or 80 wherein the concentration of protein is from about 50 ug/mL to about 1 mg/mL.

83. The assay of any one of claims 79-82 wherein the proteins are tagged with a molecule selected from the group consisting of green, red, and yellow florescent protein, flurochromes, radioisotopes, enzymes, short peptides, and His-tag.

84. The assay of any one of claims 79-83 wherein the tag-specific ligands are selected from the group consisting of biotin-avidin, gold surface sulfhydrl(-SH) containing amino acids.

85. The assay of any one of claims 79-84 wherein the labels for ligands are selected from the group consisting of antibodies, fluorochromes, enzymes, radioisotopes, chromophores, and optically active compounds sugars and nanoparticles.

86. The assay of any one of claims 79-85 wherein the carrier is selected from the group consisting of microtiter wells, beads, microspheres, liposomes, filter surfaces, glass surfaces such as microchips, biochips or slides and slide cover slips, gold coated surfaces.