METHOD FOR OBTAINING HIGH-YIELD, STABLE EXPRESSION CELL CLONES AND ANTIBODY MOLECULES OBTAINED THEREBY

Abstract

Provided is a method for obtaining high-yield, stable expression cell clones from myeloma cell lines in a protein-free culture medium. The method is used for industrial production of a recombinant antibody, and includes three stage: (1) adapting to a protein-free culture medium, statically culturing cells at a low density, and gradually reducing a fat-rich supplement to a chemical culture medium; (2) adapting to a protein-free culture medium; culturing cells at a high density, and using a perfusion fermentation system in a laboratory scale; and (3) screening high-yield, stable expression cell clones from the cells after fermentation ends. The cell clone may be used to produce a humanized anti-NeuGcGM3 14F7 recombinant antibody.

Claims

1. A method for obtaining stable high producer cell clones from myeloma cell lines in protein-free medium producing recombinant antibodies for industrial purposes that comprise three stages: I. Adaptation to protein-free medium by a stepwise reduction of a lipid-enriched supplement to chemically defined medium in low density stationary cell culture. II. Adaptation to protein-free medium in high density cell culture using perfusion fermentation system at lab scale. III. Selection of stable high producer cell clones from cells at the end of fermentation.

2. The method of claim 1 wherein the first stage comprises the following steps: I. Adaptation to grow in PFHMII cell culture medium+3.5 g/L Cell Boost. Serial passage is performed until Xv reach a constant value. II. Adaptation to grow in PFHMII cell culture medium+1 g/L Cell Boost. Serial passage is performed until Xv reach a constant value. III. Adaptation to grow in PFHMII cell culture medium. Cells are allowed to grow in protein-free medium without any supplement for 60 days.

3. The method of claim 1 wherein the second stage comprises the following steps: I. Adaptation to grow in PFHMII cell culture medium at high cell density (5-1010.sup.6 cells/ml) using perfusion fermentation system in 5 L bioreactor for more than 21 days. II. Cells at the end of fermentation are frozen and storage in liquid nitrogen (end production cell bank, EPCB).

4. The method of claim 1 wherein the third stage comprises the following steps: I. Defrosted cells from EPCB are subjected to cell cloning by the limiting dilution method. II. Antibody-secreting clones are detected by a highly sensitive and specific ELISA method using an anti-idiotypic antibody as capture antigen. III. Quantification of the antibody-producer cell sub-population in selected clones by FACS using intracellular immunoglobulin staining. IV. Kinetic studies to select clones with higher specific growth rate () and specific production rate (QP).

5. The method of claim 1 wherein the myeloma cell line is the NS0 cell line.

6. A myeloma cell line according to claim 5 wherein the NS0 cell line contains a sequence encoding for the humanized recombinant antibody anti-NeuGcGM3 14F7h.

7. The humanized recombinant antibody anti-NeuGcGM3 14F7h secreted by the cell lines obtained by the method of claim 6.

8. The humanized recombinant antibody anti-NeuGcGM3 14F7h secreted by the cell line according to claim 7 wherein said cell line is the clone B58-31E7.

9. The humanized recombinant antibody anti-NeuGcGM3 14F7h according to claim 8 wherein said humanized recombinant antibody anti-NeuGcGM3 14F7h has the following molecular phenotype: TABLE-US-00007 Identity attributes Molecular features Data Primary structure Molecular Mass Heavy Chain: 50770 Da Light Chain: 23665 Da Peptide mapping Profile Conform to Example High order Secondary structure CD Profile Conform to Example structure Tertiary structure Max emission 333 nm Absorbance Ratio.sub.330 nm/350 nm 1.23 Glycosylation HPLC Profiling G0F % = 33.1 11.7 G1F % = 42.0 7.5 G2F % = 10.9 5.1 Fuc % = 98.4 0.6 Sial % = 7.2 4.2 Heterogeneity Charges Main Peak: 96.0 1.8% Aggregation Main Peak: 99.8 0.3% Function Antigen recognition EC.sub.50 = 0.245 0.180 ug/mL

10. The humanized recombinant antibody anti-NeuGcGM3 14F7h according to claim 9 having in vitro and in vivo anti-tumor activity.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0020] FIG. 1: Cell adaptation to PFHM-II medium at different concentrations of Cell Boost 5.

[0021] FIG. 2: Fermentation run in 5 L bioreactor using the cells adapted to PFHM-II medium without cell boost 5.

[0022] FIG. 3: Absorbance values from the recognition ELISA. Black line is the median of the results. Purified 1 g/mL of T1h and hR3 were used as negative controls. Selected clones 35D8, 35D6, 31E7, 31C7, 23E9, 23C2, 73F4, 73E5, 72G5, 72F6, 72D3 (gray color), discarded clones 35F7, 33E8, 23E4, 22D8, 21E6, 2105, 73C4, 72D9, 71F2, 62E7, 61E3 (black color).

[0023] FIG. 4: Selected cells were analyzed by flow cytometry and the percentage of high producer cell sub-population was determined. NS0 myeloma cell line was used as negative control and recombinant NS0 myeloma cell line expressing hR3 antibody was used as positive control. Selected clones 31E7, 23E9, 33E8, 35D6, 23C2, 72F6, 73F4, 72G5 (gray color), discarded clones 72D3, 73E5 (black color).

[0024] FIG. 5: MFI value of high producer cell sub-population was determined. The myeloma cell line expressing hR3 and T1h antibodies were used as positive controls (light gray color). Selected clones 31E7, 35D6, 72G5 (gray color); discarded clones 23E9, 33E8, 72D3, 73E5, 23C2, 72F6, 73F4 (black color).

[0025] FIG. 6: Growth curves of the evaluated clones, cell viability and integral of viable cells from kinetic studies in roller bottles and 5 L bioreactor are plotted (y axis) versus time (x axis).

[0026] FIG. 7: Comparative analysis of maximum specific growth rate and maximum IgG concentrations between selected clones growing in roller bottles and 5 L fermenter.

[0027] FIG. 8: Recognition of NeuGcGM3 ganglioside by 14F7h antibody produced by different selected clones. Supernatants were taken from 5 L bioreactor. The anti-EGFR antibody was used as negative control, 14F7h antibody produced by parental cell was used as positive control.

[0028] FIG. 9: IgG intracellular concentration measured by flow cytometry at different steps of the selection process.

[0029] FIG. 10: Comparative kinetic studies between parental cells and clone 31E7 performed in 5 L bioreactor. Xv: viable cell concentration; SXv: integral of viable cells; IgG: maximum antibody concentration.

[0030] FIG. 11: Kinetic studies to evaluate the stability of clone 31E7 were performed at 30, 60 and 90 days after cell thawing. Cell concentration (Xv), cell viability (%), integral of viable cells (SXv), IgG concentration, maximum growth rate (p) and specific production rate (QP) were measured.

[0031] FIG. 12: IgG intracellular content measured by flow cytometry in samples from the stability study of clone 31E7.

[0032] FIG. 13: Deconvoluted mass spectrum of 14F7h Light Chain in its native (upper panel) and deglycosylated (lower panel; using PNGase F) states for the samples being reduced/alkylated. Conventional LC-MS conditions were used with a C8 column for sample separation/desalting and Acetonitrile/Formic acid buffer system for running. Inserted figure corresponds to magnification of main peak region.

[0033] FIG. 14: Deconvoluted mass spectrum of 14F7h Heavy Chain in its native (upper panel) and deglycosylated (lower panel; using PNGase F) states for the samples being reduced/alkylated. Conventional LC-MS conditions were used with a C8 column for sample separation/desalting and Acetonitrile/Formic acid buffer system for running. Inserted figure corresponds to magnification of main peak region.

[0034] FIG. 15: Deconvoluted mass spectrum of 14F7h Whole Molecule in its native (upper panel) and deglycosylated (lower panel; using PNGase F) states. Conventional LC-MS conditions were used with a C8 column for sample separation/desalting and Acetonitrile/Formic acid buffer system for running. Inserted figure corresponds to magnification of main peak region.

[0035] FIG. 16: Peptide mapping profile of 14F7h obtained after trypsin digestion and reverse phase HPLC separation using a C4 column and a conventional Acetonitrile/TFA buffer system.

[0036] FIG. 17: Circular Dichroism spectral analysis of 14F7 in the Far UV (205-260 nm) region obtained at 25 C. using a 2 mm path length cuvette. The spectra was obtained with a sample concentration of 0.6 mg/mL.

[0037] FIG. 18: Fluorescence emission of Tryptophan, using a Varioskan Flash equipment, read between 310 and 390 nm and obtained after exciting 14F7h molecule at 280 nm. A 96 well plate format was used, with 200 uL per well and a sample concentration of 0.2 mg/mL.

[0038] FIG. 19: Glycosylation profiling of 2-AB labeled glycans isolated from 14F7h and separated using normal phase HPLC with fluorescent detection (Ex: 330 nm/Em: 420 nm). Squared broken lines show a magnification of the minor peaks.

[0039] FIG. 20: Glycosylation parameters from different samples (N=3) of 14F7h used to study ranges of variation for glycosylation profiling.

[0040] FIG. 21: Weak cation exchange profile of 14F7h using a Propac-WCX10 column, obtained at 280 nm. 30 ug of sample were injected.

[0041] FIG. 22: SEC-HPLC profile of 14F7 using a TSK-G3000sxl column, obtained at 280 nm.

[0042] FIG. 23: Flow cytometer dose-response curve (% positive vs log of 14F7h concentration (ug/mL)) using L1210 target cells for different samples of antibody tested on different days.

[0043] FIG. 24: Cytotoxic effect induced by 14F7h mAb in X63 mouse myeloma cells. (A) Binding properties of 14F7h mAb. X63 cells were stained with 10 g/mL of the antibody followed by a FITC-conjugated rabbit anti-human IgG antibody. Herceptin mAb (anti-human Her-2) was used as negative control. (B) X63 cells were treated with 100 g/mL of 14F7h mAb. Cell viability after 6 h incubation at 37 C. was evaluated by propidium iodide (PI) uptake and flow cytometry analysis. Cytotoxicity is expressed as percentage of PI-stained cells. Herceptin mAb was used as negative control.

[0044] FIG. 25: In vivo anti-tumor effect of 14F7h mAb on a mouse myeloma model. (A) Schedule of mAb administration. X63 mouse myeloma cells (0.210.sup.6) were inoculated subcutaneously at day 0 to BALB/c mice and the antibodies were administered intravenously (300 g) at days 2 to 5. (B) Kaplan-Meier curves of tumor-free survival up to day 106. Humanized T1h mAb (anti-human CD6) was used as negative control. Statistical analysis was performed with the log-rank test.

EXAMPLES

[0045] The following Examples are intended to illustrate the invention but not to limit its scope in any way. Detailed descriptions of state of the art methods are not provided.

Example 1: Adaptation to Protein-Free Medium by a Stepwise Reduction of a Lipid-Enriched Supplement to Chemically Defined Medium in Low Density Stationary Cell Culture

[0046] The cell line 14F7htb58 was adapted to growth in PFHMII cell culture medium without supplement cell boost 5. The process of reduction of supplement cell boost 5 was performed in 75 cm.sup.2 T flasks, in stirred shaker (80 rpm) at 36.5 C. and 5% CO.sub.2. Every 48-72 hours the concentration of viable cells and the percentage of cell viability were determined. Cell concentration was adjusted to 0.4-0.510.sup.6 cell/ml in every culture passage.

[0047] Frozen 14F7htb58 cells were thawed in PFHMII supplemented with 3.5 g/L cell boost 5. After thawing cell viability was 75%, and then it increased up to over 90% (FIG. 1). During the adaptation process, serial culture passages were performed to maintain a maximal concentration of viable cells in the range of 0.9-1.810.sup.6 cell/ml, with percentages of cell viability higher than 90%. After eleven culture passages the concentration of cell boost 5 was reduced to zero, with 95% of cell viability and 1.2-1.810.sup.6 cell/ml. The adaptation process to PFHMII cell culture medium without lipids and cholesterol took 60 days approximately.

Example 2: Adaptation to Protein-Free Medium in High Density Cell Culture Using Perfusion Fermentation System at Lab Scale

[0048] 14F7htb58 cells adapted to growth in protein-free PFHMII cell culture medium were inoculated in a 5 L fermenter at a concentration of 0.4-0.510.sup.6 cell/ml. Fermenter operation parameters were: pH 6.9-7.0; 105 RPM; 40% dissolved oxygen; 36.5-37.0 C.; work volume 3.5 L. Cell viability and cell concentration were monitored daily by Trypan Blue exclusion method (Sigma) using a Neubauer chamber. Fermenter operation in perfusion mode was performed after cell concentration reached the value of 2.510.sup.6 cell/ml. A hollow fiber cartridge was employed for the perfusion mode and a dilution rate of 0.3-0.7 VVD was used.

[0049] During the fermentation process cell viability kept over 90% up to 384 hours, and then it was reduced to 80%. Antibody concentration was in the range of 30-40 mg/L. The maximal cell concentration reached was 910.sup.6 cell/ml when a dilution rate of 0.7 VVD was used (FIG. 2). Cells were taken at the end of the fermentation process and seeded in roller bottles at a concentration of 0.810.sup.6 cell/ml, after 72 hours in culture, cells were frozen at 1210.sup.6 cell/vial (End Production Cell Bank, EPC). After thawing cell viability was 90% and after 96 hours in culture cell viability was 97%.

Example 3: Selection of Stable High Producer Cell Clones from Cells at the End of Fermentation

[0050] Thawed cells from EPC were cloned to a single cell per well in 96 well plates using the limiting dilution method and DMEM-F12 1:1 cell culture medium supplemented with 5-10% of fetal bovine serum. Culture plates were incubated at 36.5 C. in 5% CO.sub.2 atmosphere. Cloning efficiency was below 2%. Twenty days after cell cloning culture supernatants were taken to assess IgG concentration by sandwich-type ELISA. The anti-idiotypic antibody 4G9 (3 ug/ml) was used as capture antigen and an anti-human heavy chain goat antibody coupled to alkaline phosphatase as a probe. All samples were diluted 1/500, and the median of supernatant absorbance was determined. Clones with absorbance values higher than the median value were selected (23C2, 23E9, 31C7, 31E7, 35D6, 35D8, 72D3, 72F6, 72G5, 73E5, 73F4) (FIG. 3). Selected clones were expanded to 24 well plates in DMEM-F12 1:1 cell culture medium supplemented with 5-10% of fetal bovine serum. Further cell expansion in T culture flasks were performed using PFHMII cell culture medium, in stirred shaker (80 rpm) at 36.5 C. and 5% CO.sub.2.

[0051] Intracellular IgG content was determined by flow cytometry (FACS) in selected clones. An anti-human IgG antibody coupled to FITC (fluorescein iso-thiocyanate) (Sigma) diluted 1:200 was used as probe. To determine labeling percentage 410.sup.5 cell/sample were analyzed. Clones having more than 95% of high producer cell sub-population were selected (FIG. 4). Another selection criterion was the normalized Fluorescence Median Intensity (FMI) and clones 35D6, 72G5 and 31E7 were selected (FIG. 5).

[0052] Such clones were expanded in roller bottles and kinetic studies were performed also in 5 L bioreactor (36.5-37 C., 100-105 RPM, pH<7 and dissolved oxygen higher than 40%). Clone 31E7 showed the highest cell concentration in roller bottles and 5 L bioreactor, with values of 510.sup.6 cell/ml and 410.sup.6 cell/ml respectively. Integral of viable cells provided similar results, values of 4.510.sup.8 cell/ml*h in roller bottles and 2.510.sup.8 cell/ml*h in 5 L bioreactor, these values were 1.3-1.4 higher than the one obtained for parental cells (FIG. 6). Cell viability was higher than 85% in the growth exponential phase for all evaluated clones (FIG. 6). Clone 31E7 showed the highest growth specific rate (>0.025 h.sup.1) both in roller bottles and 5 L bioreactor. Antibody concentration between 50-70 mg/L was found for all clones in 5 L bioreactor however clone 31E7 showed the highest antibody concentration in roller bottles (FIGS. 7A and 7B).

[0053] Samples were taken from the kinetic studies performed in 5 L bioreactor to evaluate the biological activity of the secreted antibody. A sandwich-type ELISA was performed. Polysorp plates were coated with NeuGcGM3 ganglioside solution in methanol (10 g/ml). As secondary antibody an anti-human heavy chain goat antibody coupled to alkaline phosphatase was used. All samples were adjusted to 1 g/ml antibody concentration and diluted 1/10 in the ELISA test. All tested samples showed recognition of the ganglioside antigen, while clones 31E7 and 35D6 showed values between 70-80% with respect to the positive control (FIG. 8).

[0054] The intracellular IgG content evolved along the process of phenotypic adaptation and cloning. Parental cells 14F7htB58 growing in PFHMII plus Cell Boost 5 showed a bimodal distribution because of the existence of a non-producer cell sub-population. After adaptation to grow at high densities in PFHMII medium 14F7htB58 cells were enriched in the producer cell sub-population (unimodal distribution), however for clone 31E7 a narrower single peak was obtained suggesting a more homogeneous cell sub-population (FIG. 9). In fact, clone 31E7 has higher maximal cell concentration, integral of viable cells and antibody production rate than parental cells (FIG. 10)

Example 4: Long-Term Stability Studies of High Producer Cell Clones

[0055] The stability of antibody production by clone 31E7 was evaluated during 90 days in cell culture. Kinetic studies in roller bottles were performed at 30, 60 and 90 days after cell thawing. Samples were taken to carry out flow cytometry studies. No differences were found between cells having different times in cell culture. Maximal cell concentration varied from 3.5-4.510.sup.6 cell/ml, cell viability was higher than 90% and the antibody concentration ranged from 50-80 mg/L (FIG. 10). The growth specific rate () kept higher than 0.025 h.sup.1 while the production specific rate (QP) was higher than 0.18 pg/cell*h (FIG. 11). After 90 days in cell culture clone 31E7 showed a narrow single peak at same FMI representative of a homogeneous high producer cell population (FIG. 12).

Example 5: Identity Attributes of the Secreted Immunoglobulin by Selected Stable High Producer Cell Clone 31E7

[0056] Several identity attributes are defined to cover the basic molecular properties that later will allow to assess quality of the molecule and to monitor product consistency as well as cell line stability. Primary structure is studied by determination of the mass of the whole molecule, and its individual chains, by LC-ESI-MS analysis of the native and disulfide bridges reduced/alkylated sample (glycosylated and deglycosylated). See Table 1 for a summary of the results and FIGS. 13, 14, 15. Additionally, the peptide mapping of the molecule is used to monitor its first order structure and further exclude any possible post-translational modification truncation or sequence changes (FIG. 16).

TABLE-US-00002 TABLE 1 Masses of the complete molecule and its chains with and without glycosylation. Mass (Da) No PNGase F Species PNGase F Observed Theoretical % Difference Whole molecule 149863.41 146983.31 147009.30.sup.a 0.018 Heavy Chain 52217.63 50770.69.sup.b 50784.12.sup.c 0.026 Light Chain 23665.93 23664.91.sup.b 23664.25.sup.d 0.0028 .sup.aConsidering lack of K in the C-terminal region of both heavy chains .sup.bReduced with Dithiothreitol and alkylated with Iodoacetic acid .sup.cConsidering lack of K in the C-terminal region and Iodoacetic acid modification .sup.dConsidering Iodoacetic acid modification

[0057] Higher order structures are tested at two levels by using Far UV CD spectra for the analysis of secondary structures (yielding the profile shown in FIG. 17) and Intrinsic Fluorescence properties that allows to detect differences in conformation and protein stability (Weichel et al, 2008, BioProcess International, June). FIG. 18 shows the results obtained for this latter test. In this case, emission maximum was obtained at 333 nm and the value of Absorbance Ratio.sub.330 nm/350 nm was 1.23.

[0058] Glycosylation, being the major post-translational modification occurring in human IgG1 antibody molecules, is studied by normal phase HPLC profiling. FIG. 19 shows the results obtained for three different samples, and their glycosylation parameters are shown in Table 2 (Montesino et al, 2012, Biologicals 40:288-298). The behavior and dispersion of these glycosylation parameters are shown in FIG. 20.

TABLE-US-00003 TABLE 2 Glycosylation parameters from different samples (N = 3) of 14F7h used to study ranges of variation for glycosylation profiling. Samples G0F % G1F % G2F % G0F/G1F G1F ratio Fuc % Sial % 1 31.31 41.83 10.78 0.75 3.28 98.15 8.81 2 29.52 45.15 13.04 0.65 3.18 98.50 7.41 3 38.47 39.05 8.99 0.99 2.67 98.53 5.50 Average 33.1 11.7 42.0 7.5 10.9 5.1 0.8 0.3 3.0 0.9 98.4 0.6 7.2 4.2 G0F %, G1F %, G2F %: % of agalactose, monogalactose and digalactosylated fucosylated glycans, Fuc %: % of Fucosylated glycans, Sial %: % of Sialylated glycans

[0059] The heterogeneity of the molecule is defined by two orthogonal methods. In the first place Weak Cation Exchange (WCX) is used to detect the different charged species, allowing detecting product truncations, deamidations, some glycosylation variants, etc. For antibodies, results of this method mainly monitor C-terminal lysine truncation, a common modification found in hlgG1 molecules (Dionex Application Note 127, http://www.dionex-france.com/library/literature/application_notes_updates/AN127_LPN1047.pdf). The profiles obtained for different samples are shown in FIG. 21 and Table 3 shows the result of their integration. Additionally, Size Exclusion Chromatography (SEC) is used to monitor the aggregation state of the molecule. FIG. 22 shows the profiles obtained from different samples and Table 4 shows the result of their analyses.

TABLE-US-00004 TABLE 3 WCX-HPLC integration results (main peak percentage) for different 14F7h samples. Samples Main Peak % 1 96.2 2 95.0 3 95.7 4 96.5 5 96.6 Average 96.0 1.8

TABLE-US-00005 TABLE 4 SEC-HPLC integration results (main peak percentage) for different 14F7h samples. SE-HPLC Samples Main Peak % 1 99.9 2 99.9 3 99.9 4 99.8 5 99.7 Average 99.8 0.3

[0060] Finally, the function of the molecule is primarily assessed by studying its ability to recognize its antigen on target cells measured by Flow Cytometry. FIG. 23 shows the dose-response curve results obtained using different samples, and Table 5 shows the EC.sub.50 calculated from them.

TABLE-US-00006 TABLE 5 EC.sub.50 results for different tests of flow cytometer dose-response curve (% positive vs log of 14F7h concentration (ug/mL)) using L1210 target cells for different samples of antibody tested on different days. Test Sample EC.sub.50 (g/mL) 1 1 0.1633 2 0.2121 2 1 0.2524 3 2 0.2409 4 2 0.3554 Average 0.245 0.180

Example 6: Evaluation of the In Vitro and In Vivo Anti-Tumor Effect

[0061] Antigen expression by X63 mouse myeloma cells was measured by flow cytometry. 100% of X63 mouse myeloma cells were stained by 14F7h mAb (FIG. 24A). X63 cells were incubated with 10 g/mL of the antibody followed by a FITC-conjugated rabbit anti-human IgG antibody. Herceptin mAb (anti-human Her-2) was used as negative control.

[0062] In vitro cytotoxic effect induced by 14F7h mAb in X63 mouse myeloma cells was assessed. X63 cells were treated with 100 g/mL of 14F7h mAb. Cell viability after 6 h incubation at 37 C. was evaluated by propidium iodide (PI) uptake and flow cytometry analysis. Cytotoxicity is expressed as percentage of PI-stained cells. More than 50% of tumor cells were dead after treatment (FIG. 24B). Herceptin mAb was used as negative control.

[0063] To assess the in vivo anti-tumor effect of 14F7h mAb on X63 mouse myeloma model, 0.210.sup.6 X63 tumor cells were inoculated subcutaneously at day 0 to BALB/c mice and the antibodies were administered intravenously (300 g) at days 2 to 5 (FIG. 25A). Kaplan-Meier curves of tumor-free survival up to day 106 are shown in FIG. 25B. Statistically significant increase in tumor-free survival was obtained 60% of mice treated with 14F7h mAb were still alive after the follow up time. Humanized T1h mAb (anti-human CD6) was used as negative control. Statistical analysis was performed with the log-rank test.