WO2005014801A1

WO2005014801A1 - Improvements in cell culture productivity

Info

Publication number: WO2005014801A1
Application number: PCT/GB2004/003399
Authority: WO
Inventors: Andrew Devitt; Christopher Darrell Gregory
Original assignee: University Court Of The University Of Edinburgh
Priority date: 2003-08-04
Filing date: 2004-08-04
Publication date: 2005-02-17

Abstract

The present invention relates to methods and apparatus suitable for improving productivity of cell cultures. The invention also relates to novel reagents for use in such methods and/or apparatus.

Description

IMPROVEMENTS IN CELL CULTURE PRODUCTIVITY

Field of the Invention The present invention relates to methods and apparatus suitable for improving productivity of cell cultures. The invention also relates to novel reagents for use in such methods and/or apparatus.

Background to the Invention It has previously been suggested that it may be undesirable to maintain dead cells in a cell culture, as the dead cells, or materials which leach into the cell culture on lysis of the dead cells may be detrimental to the cell culture. Previously described methods and apparatus designed to remove dead cells from the culture have generally relied on the use of sedimentation and/or centrifugation in order to remove cells based on size and/or weight differences compared to live cells, see for example EP599651 and Prod. Biol. Anim. Cells. Cult., 1991, ESCAT 10 Meeting, p394-399, Bjorling & Maelstrom. Nevertheless, such techniques may be physically too severe, impractical for large scale-up and can damage otherwise viable healthy cells. Moreover, there is little or no teaching of the ability to be able to remove dying cells i.e. cells which are committed to cell death and whose productivity may be reduced. US6,432,653 describes a device and process for separating cells of different types based on the fact that the living cells can possess different cell surface markers. However, there is no suggestion or teaching of how one would separate dead and/or dying cells from a cell culture, or what benefits this may provide. Reagents are known, such as Annexin V which preferentially binds to cells undergoing apoptosis, see for example the apoptotic cell separation kit by Sigma, and EP0755516. Nevertheless there is no suggestion in the art that this molecule could easily be used to remove dead and/or dying cells from a cell culture with the view to increasing productivity of the cell culture, or how the skilled man could go about achieving this. Consequently it is amongst objects of the present invention to provide a method and apparatus designed to improve productivity of cell cultures by removing dead and/or dying cells and/or parts of dead and/or dying cells therefrom. In a first aspect the present invention provides a method of improving productivity of a cell culture by removing dying, dead cells and/or parts thereof from the cell culture, comprising the steps of: a) providing a cell culture; b) subjecting the cell culture to conditions conducive to allow production of a desired product by the cell culture; c) contacting the cell culture with a reagent which is capable of preferentially binding to dying, dead cells and/or parts thereof from the cell culture; and d) removing at least a portion of said dying , dead cells and/or parts thereof from the cell culture which are bound to the reagent. The method may effectively be a continuous one, with steps b) to d) being repeated as necessary and in any order, or continuously during the cell culturing and product production. That is, the cell culturing process and the steps of removing dead and/or dying cells may be carried out contemporaneously. The method may also be carried out at the end of the cell culture process. The method according to the present invention is intended for improving productivity of a cell culture. That is, the method is designed to improve an amount and/or quality of product produced by the cell culture in comparison to a cell culture not employing the present method. Such products include biopharmaceuticals such as recombinant proteins and antibodies, which are routinely used today. It is often not possible to produce such complex products via synthetic means and the only way to produce them is by culturing cells which are capable of expressing such products. As the cost of producing such products can be extremely high, any degree of improvement can be seen as beneficial/advantageous. However, an improvement in production of greater than 2%, 5%, 7.5% 10% or 20% is preferred. Moreover, the product could in fact be the cells themselves and in this manner removal of dead and/or dying cells may lead to higher titres and/or improved quality of viable cells being produced. Typically the cell culture comprises eukaryotic cells e.g. mammalian, such as human cells. It is understood that such cells can be fragile to excessive physical stress, such that the present invention is designed to impose reduced physical stresses to cells as may occur by previously known processes for removal of dead cells, which employ, for example, centrifugation steps. Generally the present invention does not include a centrifugation step. The skilled addressee will readily know what conditions are conducive to allow production of a desired product by said cell culture. Indeed these are likely to already be being used in an existing process for producing the product (i.e. a manufacturer of say a biopharmaceutical will already have a defined set of culture conditions which have been optimised for production of the desired product, but dead and/or dying cells will likely be present in the cell culture) . Nevertheless, it is well within the knowledge of the skilled addressee to carry out optimisation of the culture conditions to obtain the highest levels of product production. The cell culture is contacted with the reagent designed to remove dead and/or dying cells from said cell culture. Contacting under low shear forces may be achieved by for example, passing the cell culture over a substrate to which is adhered the reagent. The substrate can be any suitable material which is generally applicable for use with the reagent to be adhered and includes for example glass, natural or synthetic polymers, metal and the like. The substrate may be in the form of a flat planar surface, or could be for example the inner surface of a tube or the available surface of a foam material. Alternatively a cartridge, monolith or the like comprising a plurality of surfaces with adhered reagent may be used and the cell culture passed through the cartridge or monolith such that the culture can contact the reagent and dead and/or dying cells be removed from the culture. Desirably, the substrate, cartridge and/or monolith can be replaced as required. In one embodiment the reagent may be coated and adhered on the surface of, for example, small beads, e.g. microspheres using known techniques. Suitable methods of adhering the reagent to the substrate include natural (passive) binding, as many substrates such as plastics have a high binding capacity in their native state and molecules such as proteins adhere without needing to modify the surface of the substrate. Alternatively the substrate can comprise or be modified to include reactive groups on the substrate surface, e.g. carboxyl groups which can react with the free amines present in proteins and other amine-containing-biomolecules. Covalent coupling to the carboxyl groups can be achieved using for example carbodiimide reagents such as EDAC (l-ethyl-3-[3- dimethylaminopropyl]carbodiimide hydrochloride) . It is also possible to coat the substrate with antibodies or other molecules that can specifically bind to the reagent to be adhered, or avidin that will bind to reagents comprising biotin attached thereto. As such the reagent to bind dead and/or dying cells may be bound directly to the substrate or indirectly through intermediate molecule (s). In one embodiment of the invention, soluble CD14 fused to the Fc portion of, for example, human IgG (sCD14Fc) is used and this can be directly bound to the substrate, or the substrate can be coated with an anti- human-Fc, which will adhere to the sCD14Fc. Preferably, the substrate is also "blocked" as known in the art, with, for example, a non-specific protein such as bovine serum albumin, before passing cells over the substrate. It is also possible to render the beads magnetic such that the beads, once the dead and/or dying cells have been adhered thereto, can be separated from the body of the cell culture using a magnetic field or magnets. In all embodiments, desirably the substrate can be reused and/or replenished. For example dead and/or dying cells may be washed off the substrate such that the substrate can be re-used or the substrate with adhered reagent could be replaced. In one embodiment of the invention using antibody covalently bound to the substrate, dead and/or dying cells and/or parts of dead and/or dying cells may be removed from the antibody using standard techniques known to individuals skilled in the art, for example washing the antibody-coated substrate with a solution of low pH (e.g. lOOmM glycine pH 2.8). According to the present invention dead/dying cells are cells which are apoptotic (cells undergoing apoptosis) and/or necrotic (cytolysed or leaky cells) and include cell fragments such as apoptotic bodies or debris as a result of cell lysis and/or leakage. The skilled addressee can easily discern dead and/or dying cells from healthy and/or viable ones using suitable techniques including microscopy and detection of phosphatidylserine on the surface of the cell membrane. Necrotic cells can be identified by their inability to exclude dyes such as trypan blue or propidium iodide (PI) using well established methods. Apoptosis can be defined by staining with annexin V and exclusion of PI. Other tests can be used to define apoptosis e.g. nuclear morphology, loss of mitochondrial membrane potential, activation of one or more caspases. Not all apoptotic cells will show all of these parameters at all times as they are cell dependent and dependent upon the stage of the apoptosis programme. All of these are well established methods with details available from the literature and a range of companies (e.g. Molecular Probes, Alexis Corp., Sigma, Beckman-Coulter, Pro ega) . The amount of dead and/or dying cells in a culture is likely to vary depending on the type of culture, cells used, length of culture, type of product etc. However, any improvement in the amount of product produced as a result of removing dead and/or dying cells from the culture is encompassed by the invention. It may not be possible to remove all dead and/or dying cells or parts of dead and dying cells from culture, but desirably greater than 10%, for example 25%, 50%, 75%, 90% or 95% of whole dead and/or dying cells are removed. The removal may be accompanied by removal of debris. By "removal" or "removing" is meant that the dead and/or dying cells are separated from the bulk or body of the cell culture. The dead and/or dying cells may be removed from the cell culture or otherwise isolated from the viable cells in the cell culture, so as to not interfere with the viable cells. The reagent which is capable of preferentially binding to dying and/or dead cells includes macrophages and acrophage-like cell lines e.g. THP-1, HL60, U937, Mono-mac 6; and J774; and cells expressing molecules capable of binding dead and dying cells e.g. Cos cells expressing recombinant membrane-associated CD14 (see Devitt et al (1998) . Human CD14 mediates recognition and phagocytosis of apoptotic cells. Nature: 392 :pp505-509) . and proteins such as Annexin V (see Sigma catalogue and EP0755516) ; antiphosphatidylserine antibodies e.g 1H6 (Upstate UK Ltd.) and soluble CD14 (sCD14) , further described in detail hereinafter, or a combination thereof. Advantageously sCD14 and antibodies do not require Ca ⁺ for binding to apoptotic cells in contrast to Annexin V. Additionally antibodies or other reagent which preferentially react to intracellular proteins or proteins expressed on the surface of dead and/or dying cells, but not to the same extent on the surface of living and/or viable cells, may also be utilised. Such reagents would include antibodies to cytoskeletal components and other reagents such as the fungal toxin phalloidin. Reagents capable of binding to intracellular components may preferentially bind dead and dying cells as those cells lose their membrane integrity. According to the present invention soluble CD14 is CD14 in any format that is not membrane associated and will include CD14 bound to surfaces (e.g. beads) and CD14 conjugated to reagents (e.g. fluorochromes) . The present invention also relates to the use of reagents which are capable of preferentially binding dead, dying cells and/or fragments thereof for improving cell culture productivity. The present invention also relates to new uses of sCD14 that have not hitherto been suggested. As SCD14 has been shown by the present inventors to bind apoptotic cells, sCD14 may be used in in vitro or in vivo experiments to identify apoptotic or dead cells. In vivo, it can be of importance to identify apoptotic cells. For example, tumours have areas of apoptosis within the tumour and certain tumours have many apoptotic cells. sCD14 could be labelled in some manner e.g. fluorescently (directly or indirectly conjugated) or isotopically e.g. Technetiu in order to allow imaging of the site of apoptosis. This may be of use in studying the efficacy of anti-tumour therapy (e.g. responding tumours may show an increase in apoptosis following therapy. Also, sCD14 can be administered as a so-called "magic-bullet" when conjugated to a toxic/therapeutic moiety or prodrug that could be used to deliver the moiety or prodrug to site of dying or dead cells (e.g. tumours) . It may also be used to detect tumours or other sites of apoptosis in health screens. In vitro, it can be of importance to identify apoptotic cells. For example, researchers may wish to identify and/or quantify apoptotic cells within a culture. sCD14 could be labelled in some manner e.g. fluorescently (directly or indirectly conjugated) and by using appropriate detection technologies e.g. flow cytometry or fluorescence microscopy A range of molecules are known in the literature to bind to apoptotic cells. An embodiment of the invention is to use any of these alone or in combination to improve productivity in culture through removal of dead and/or dying cells. These reagents may be used as soluble factors or be expressed on cells. These reagents include: Integrins (e.g. α_vβ₃; α_vβ₅), CD36, SR-A, MER, CD14, ABCA1, PSR, CR3 , CR4 , CD91/calreticulin, CD31, FcγR, SHPS-1, SRB1, asialoglycoprotein receptor, thrombospondin-1, Gas-6, protein S, MFG-E8 , Del-1, β2GPI, C-reactive protein and serum amyloid protein. In a further aspect the present invention provides an apparatus for improved culturing of cells, the apparatus comprising a culture vessel for culturing cells, and a cell separation system for separating and enabling removal of dead and/or dying cells from a cell culture, the cell separation system comprising a solid substrate or substrates to which is/are adhered a reagent or reagents which is/are capable of preferentially binding to dead and/or dying cells of said cell culture, using, for example, a panning process (as defined below) . The cell culture vessel and separation system conveniently in its own vessel are typically connected to one another, such that a cell culture medium comprising cells may be passed from the cell culture vessel to the separation system and thereafter back to the culture vessel. Alternatively, the cell culture vessel and cell separation system may be separate and unconnected. Moreover, the cell separation system may simply be adapted to be added or incorporated within the cell culture vessel itself and removed, for example, for replenishing, as required. The reagent or reagents capable of preferentially binding dead and/or dying cells is/are typically proteins, such as Annexin V, sCD14 or others mentioned hereinabove, macrophage, macrophage-like cell or other phagocytic cell (e.g. fibroblasts) , antibodies specifically reactive with a cell surface receptor or intracellular moieties which are present on/in dead and/or dying cells and/or combinations thereof. According to the present invention "panning" refers to the process whereby a culture is brought into contact with the reagent (s) and substrate to facilitate removal of dead and/or dying cells and/or parts of dead and/or dying cells. The present invention will now be further described by way of example and with reference to the Figures that show: Figure 1 shows a schematic representation of a culture apparatus for use in a method according to one embodiment of the present invention; Figure 2 shows sCD14 binding to apoptotic human B cells. The Burkitt lymphoma cell line Mutu I was induced to undergo apoptosis by UN irradiation (100 mJ/cm²) , cultured for 16-18 h and labelled with sCD14-Fc (A) or sCD14-His (B) . Nominally, 50-100μl of 20 μg/ml SCD14 was used to label 200,000 cells. sCD14-Fc was visualised using goat anti-human γ-chain-PE conjugate; sCD14-His was visualised using mouse anti-CD14 mAb 63D3, goat anti- mouse-IgG-biotin, followed by streptavidin-PE. "Control' indicates non-irradiated cultures. Analyses were undertaken using the EPICS-XL flow cytometer (Beckman Coulter) ; Figure 3 shows sCD14 binding to apoptotic human T cells. The human T cell line Jurkat was induced to undergo apoptosis by UN irradiation (A) or staurosporine treatment (B) . Cells were labelled and analysed after 16-18 h. Soluble ICAM-3 (sICAM-3) is used as a negative control protein; in (A), "Secondary" indicates labelling with goat anti-human γ-chain-PE conjugate alone and "Control* indicates cells not induced to undergo apoptosis; Figure 4 shows sCD14 binding to apoptotic and necrotic human B cells. A: Cells were treated as in Figure 1 and labelled and analysed up to 18 h after irradiation. Histograms were created separately for apoptotic or viable cells according to light scatter properties. BU12 is a mouse CD19 mAb and is visualised using biotinylated goat anti-mouse immunoglobulin, followed by streptavidin-PE (a positive control for staining) . LPS-binding protein (LBP) was included with some samples at 1.2 μg/ml and is a protein that promotes binding of CD14 to LPS (one of its ligands) . B: Cells were induced into necrosis by incubating for 30 minutes at 56°C. Viable and dead cells were again discriminated by light scatter; Figure 5 shows BU65 hybridoma cells stained with AxV-biotin before and after removal of dead and/or dying cells with Annexin V-biotin and streptavidin-magnetic beads ; Figure 6 shows improved productivity of Ab by hybridoma cell culture (9el0) depleted of dead and/or dying cells. Hybridoma culture of 9el0 was panned on macrophages at a ratio of 20 non-viable cells per monocyte-derived macrophage (MDM) as per example 4. Productivity was assessed over time following depletion of non-viable cells; Figure 7a shows anti-phosphatidylserine (anti-PS) Ab (1H6) staining of human B cells. The human B cell line was stained using indirect immunofluorescence (standard methodology as used by people skilled in the art) and analysed using flow cytometry. Strong preferential staining of non-viable cells was noted. Using cytometry analysis techniques (as used by people skilled in the art) the FS/SS plots of events staining strongly with Ab 1H6 and weak/non-staining events were generated. Clearly the strongly stained events included dead and dying cells and particles smaller than whole cells (i.e. cell debris) ; Figure 7b shows agglutination of 1H6 coated beads by supernatants from cultures where cells had been removed by centrifugation- In this case sub-cellular debris caused beads to agglutinate; Figure 8 shows anti-phosphatidylserine (anti-PS) Ab (1H6) staining of hybridoma cells and debris of non- viable cells. A hybridoma culture was stained using indirect immunofluorescence (standard methodology as used by people skilled in the art) and analysed using flow cytometry. Strong preferential staining of non-viable cells was noted. Using cytometry analysis techniques (as used by people skilled in the art) the FS/SS plots of events staining strongly with Ab 1H6 and weak/non- staining events were generated. Clearly the strongly stained events included dead and dying cells and particles smaller than whole cells (i.e. cell debris). Figure 9 shows the effect of panning a culture of hybridoma cells on anti-PS Ab (1H6) or isotype control Ab. Ab was coated to polyether foam and cells sucked into the foam. After incubation unbound cells were cultured for 7 days and productivity assessed. Culture exposed to the non-viable and debris-binding Ab 1H6 showed improved productivity. Figure 10 shows staining of dead and dying cells with reagents reactive with intracellular molecules. A culture of human B lymphoma cells (induced to apoptosis with ionomycin) was stained with a monoclonal antibody to actin (AC15, Sigma Ltd) using indirect immunofluorescence (standard methodology as used by people skilled in the art) and analysed using flow cytometry (figure 10a) . Further a hybridoma culture (9el0 cells) was stained with a fungal toxin (fluorescein-labelled) that is known to bind to actin (Phalloidin-FITC, Sigma Ltd) using direct immunofluorescence (standard methodology as used by people skilled in the art) and analysed using flow cytometry (figure 10b) . Using cytometry analysis techniques (as commonly used by people skilled in the art) the FS/SS plots of events staining strongly with Ab AC15 or phalloidin-FITC were generated. Clearly the strongly stained events included dead and dying cells and exemplifies the use of reagents reactive to intracellular antigens to detect and bind dead and dying cells. Other such targets would include other cytoskeletal components. Figure 1 shows a schematic representation of a culture apparatus in accordance with an embodiment of the present invention. The apparatus 2, has a culture vessel 4 for culturing cells 6, agitation of the culture medium 6 can be carried out by an impellor 8. The culture medium 6 in which the cells are grown flows, together with the cells, along pipe 10 into a separation vessel 12. Within the separation vessel 12 is found a solid substrate 13 which has coated thereon a reagent or reagents (e.g. acrophages) designed to preferentially bind dead and/or dying cells The cells in the culture solution come into contact with the reagent and dead and/or dying cells bind to the reagent whilst healthy viable cells substantially do not. The culture medium 6 is then returned to the culture vessel 4 via pipe 14 and culturing continued. The process of removing dead and/or dying cells from the culture medium 6 can be a continuous process during the culturing of the cells. That is, the culture medium can continuously be cycled through the separation vessel and returned to the culture vessel. In this manner dead and/or dying cells can continuously be removed from the cell culture. The skilled man can easily appreciate what other features may be present in the apparatus, such as heating systems, O₂/CO₂ control mechanisms, product extraction devices and the like. The skilled man will also appreciate that such a device need not be used in a continuous fashion as depicted.

Example 1: Preferential binding of SCD14 to apoptotic and necrotic cells; The extracellular domains of CD14 (whole molecule lacking the GPI-anchor site and stop codon) were PCR amplified with restriction site (Hind III/Bam HI) . This product was cloned (directionally) into the polylinker of a pCDM8-based vector named plgl (this is published in the reference below) . A splice donor site was incorporated into the 3 ' end of the PCR product also so that the resulting pre-mRNA is spliced to yield a mature mRNA with the CD14 extracellular domain directly abutting the hinge region of the IgGl Fc. The vertebrate splice donor concensus is a 5 * : C/AAGGTAAGT 3 ' ) . On translation this then produces a fused chimaeric protein with CD14 N- terminal to the Fc region. This method and the plas id are published: D.L. Simmons. Cloning cell surface molecules by transient expression in mammalian cells. Chapter 5 in: Cellular Interactions in development: A practical approach. Edited by D.A. Hartley. A deposit of this construct has been made in accordance with the Budapest Treaty on the 1st August, 2003 and has the accession No. NCIMG 41188. From the DNA the protein is produced as follows: It was produced by transfecting the DNA into 293 cells using standard calcium phosphate-mediated transfection methods - see for example Sambrook et al: Molecular Cloning: A. Laboratory Manual (2000) . Cold Spring Harbor Laboratory Press) and after culture for 3-7 days the supernatent was harvested . The produced recombinant protein was purified by binding to protein G in a column (Pharmacia Biotech) . After washing the column to remove impurities the desired protein was eluted at low pH using lOOmM glycine pH2.7 and neutralised by adding 50μl 1M Tris at pH 9.0 per ml of eluate. Protein was assessed for purity by western blot and ELISA and concentration of the protein was estimated by using a colorimetric technique (Bradford assay supplied by Bio-Rad) . The present inventors have found that soluble forms of the CD14 molecule can be shown to bind preferentially to apoptotic and necrotic cells, as compared to viable cells. Results obtained are based on flow-cytometric assays of soluble recombinant human CD14 binding to human cells in suspension. Examples are shown in Figures 2-4. Two forms of soluble CD14 (sCD14) have been produced and analysed, one fused to the Fc portion of human IgG (sCD14-Fc) , the other tagged with multiple histidine residues (CD14-His) . The protocol adopted for assessing binding of recombinant sCD14 to cells is immunofluorescence staining and flow cytometry (see Becton-Dickinson Manufacturers Protocols for details) . In summary, cells are incubated with the recombinant protein, washed, and the bound recombinant protein is detected using an antibody (anti-Ig or anti-His) that is either directly conjugated to a fluorochrome or that is detected in turn by a tertiary reagent that is fluorochrome conjugated. The labelled cells are run on a flow cytometer and histograms are generated that plot the relative number of cells (fluorescent events, on the y- axis) versus the fluorescence intensity (normally on a log scale, x-axis) . By comparing the histograms of cells subjected to different treatments and labelled with different recombinant proteins, it is possible to ascertain the capacity of sCD14 preparations to bind to viable, apoptotic and necrotic cells. In studies to date the inventors have used sICAM-3 as a control recombinant protein for comparison with sCD14. The results shown in Figures 2-4 indicate that sCD14 binds with highest intensity to apoptotic or necrotic cells as compared with viable cells. In Figure 2, a human B cell line induced to undergo apoptosis by UN irradiation are compared with untreated (Control) cells. Figure 2A shows preferential binding of sCD14-Fc, and 2B that of sCD14-His, to cells induced to undergo apoptosis. Differences in staining intensities are due to differences in concentrations of recombinant proteins and/or differences in visualisation protocols. In Figure 3, a human T cell line that has been subjected to similar treatments and analyses are shown. In Figure 3A, untreated (Control) T cells are compared to cells induced to undergo apoptosis by UN irradiation. Preferential binding of sCD14-Fc is observed in the case of the UV-treated cells. In this experiment comparisons are made not only with untreated cells, but also with sICAM-3-Fc labelling and secondary visualisation reagents alone - all of which show histograms of relatively lower fluorescence intensity. Similar results are obtained when T cells are induced to undergo apoptosis with staurosporine (Figure 3B) . Figure 4A shows a time course of induction of apoptosis in a human B cell line with labelling carried out at various times up to 18 hours. In these experiments, cells were stained with recombinant protein and the apoptotic and viable cells identified on the basis of light scatter properties (forward versus side scatter) using flow cytometry (see Dive, C, Gregory, CD., Phipps, D.J., Evans, D.L., Milner, A.E., and Wyllie, A.H. (1992) . Analysis and discrimination of necrosis and apoptosis (programmed cell death) by multiparameter flow cytometry, Biochimica Et Biophysica Acta 1133 , 275-285, copy attached) were individually analysed for their ability to bind sCD14. Note that the apoptotic cells at time zero are present as a result of a low level of spontaneous, constitutive apoptosis that is a feature of these B cells. Maximal labelling with sCD14 is observed at 12-15 hours post apoptosis-induction. LBP (which enhances binding of LPS to CD14) is utilised but does not appear to enhance sCD14 binding to apoptotic cells - if anything it causes a minor reduction in binding. BU12 is a monoclonal antibody that binds to CD19 on the B cells and is included as a positive labelling control. Soluble ICAM-3 and the secondary visualisation reagent are included as negative controls. In Figure 4B, cells that were deliberately induced into primary necrosis by heating were similarly labelled and analysed. In this case there are virtually no viable cells available for analysis. Necrotic cells were preferentially labelled by sCD14 as compared with sICAM-3 or secondary reagent alone. It appears that necrotic cells are not as strongly labelled with sCD14 as the apoptotic cells, particularly those analysed at the 12-15 hour time points (see Figure 4A) .

Example 2: Presence of apoptotic cells inhibits productivity in hybridoma cultures In this experiment, apoptotic hybridoma cells were added to viable cultures and the effect on antibody (IgGl) production was assessed at 3 and 5 days. The results show that antibody productivity is reduced by the presence of apoptotic cells in a dose dependent fashion (table 1) .

Table 1: Effect of Adding Spontaneously Occurring Apoptotic Cells to Hybridomas on Their Antibody Production

Notes

• Hybridomas were from an exponentially growing culture containing 3.9% apoptotic cells. Control cells were from the same culture.

• "Neglected" hybridoma cells were not sub-cultured for 5 days previously and contained 47% apoptotic cells

• Cultures were seeded at 2x10 /ml and subsequently neglected (apoptotic) cells or control cells were added

• Final volume was the same for all cultures (10ml)

• Culture medium was not replenished during the experiment

• Apoptosis was measured by DAPI staining and fluorescence microscopy. Example 3: Use of Annexin V to remove apoptotic cells from hybridoma cultures The inventors have shown that Annexin V-coated beads can be used to remove apoptotic cells from hybridoma cultures (figure 5) . Details how dead/dying cells were removed from cell culture using Annexin V coated magnetic beads are as follows: Hybridoma cells (BU65) were washed into binding buffer and stained with Annexin V-biotin as per manufacturer's instructions (Bender Medsystems) . Annexin V stained cells were removed by incubation with streptavidin-coated magnetic beads (Dynal Ltd) prior to exposure to a magnet. The level of apoptotic cells was assessed before and after the removal procedure using annexin V-FITC and flow cytometry as per manufacturer's instructions (Bender Medsystems) . In a second experiment, Table 2, the percentage of apoptotic cells was reduced from 5% to less than 0.2% or from 29% to 2% or less as follows: 4C/6 mouse hybridoma cells were exposed to UN irradiation (lOOmJ/cm ) 20 hours prior to study. A mixture of these cells (30%) and untreated cells (70%) or untreated cells alone, each at 5x10 /ml, were stained with Annexin V-biotin (Bender MedSystems Ltd) lμl per 10 cells in high calcium buffer for 15 min. The cells were washed 3 times with high calcium buffer and resuspended in the same. Various volumes of streptavidin-coated magnetic Dynabeads (6.7xl0⁸/ml in high calcium buffer) were added to 250μl aliquots of cells in the wells of a 48 well tissue culture plate which was rocked on a GyroRocker STR-9 moving platform (Bibby Sterilin Ltd) sufficiently to keep the cells in suspension at room temperature for 30 min. Reactants were transferred to 5ml of high calcium buffer in a 15ml polypropylene round bottom tube and AxV-binding cells were removed magnetically. The proportion of apoptotic cells was determined by staining with DAPI. Table 2. Removal of apoptotic hybridomas by magnetic depletion of AxV binding cells

% APOPTOTIC CELLS

Hybridoma Start Post depletion of AxV-bindinq cells Streptavidin-coated 40 20 10 5 Dynabeads (μl)

Control 5 <0.2 <0.2 <0.2 <0.2

UV-treated* 29 <0.3 <0.3 1.7 2.0

*comprised 30% UV-treated cells and 70% untreated cells

Additionally, the present inventors wished to investigate alternative procedures that did not require a centrifugation step, since the present inventors have found that centrifugation itself has a detrimental effect on hybridoma cultures, the present inventors therefore sought a gentler method that would be more broadly applicable to a final product.

Example 4: Use of Macrophage monolaγers to remove apoptotic cells from hybridoma cultures In the present experiments, a human monocytoid cell line, THP-1 was used which, after differentiation to macrophage-like cells was used as a monolayer to capture apoptotic cells by "panning" .

Method THP-1 cells (10⁶/ml) were stimulated to differentiate by culture with phorbol myristate acetate (250nM) (PMA also known as TPA) and di-hydroxy-vitamin D3 (lOOnM) for 3 days. Both reagents are widely available (e.g. from Sigma) The differentiated cells form a confluent monolayer after this treatment and previously have been shown to be capable of the clearance of apoptotic cells. Immediately before panning: - • Differentiated THP-1 cells were washed x2 in warm (37°C) culture medium (CM) to remove non-adherent cells. Residual CM was tipped off. • Hybridoma cells (containing various proportions of non- viable and apoptotic cells) were resuspended in pre- warmed CM from an exponentially growing culture.

Panning • Hybridoma cells were added to culture flasks of differentiated THP-1 cells. • Flasks were incubated at 37°C/5%C0₂ for various intervals with gentle rocking every 10 min. • Supernatants were tipped off and recovered hybridoma cells were counted, the percentage viability and percentage apoptosis were determined.

Results Experiment 1: 8.5.03

1x10⁷ hybridomas were panned on 2x10⁷ THP-1 cells. Ratio of apoptotic hybridomas:THP-1 cells = 1:60 The percentage of apoptotic cells was reduced from 3.4% to less than 1% with over 90% of cells recovered (Table 3).

Table 3

The percentage of apoptotic cells was reduced. The efficiency of panning was increased by increasing the http: apoptotic cell ratio and the panning time (table 4) .

Table 4

Example 5: Increased IgG Production by Hybridomas after Removal of Non-Viable/Apoptotic Cells by "Panning" on Human Macrophages

Methods Preparation of Human Monocyte-derived Macrophages • Human peripheral blood mononuclear cells (MNC) from a healthy adult volunteer were fractionated by density gradient centrifugation on Percoll (Amersham-Pharmacia) . • Monocytes were enriched by adherence: MNC (10ml, 4x10⁶ per ml) were incubated in Iscove's Modified Dulbecco's Culture Medium (IMDM) containing antibiotics and glutamine in 25cm tissue culture flasks at 37°C in a humidified atmosphere containing 5% C0₂. Supernatant culture medium containing non- adherent cells was discarded after one hour and adherent cells were rinsed with RPMI 1640 culture medium (RPMI) . • Monocyte-derived macrophages (MDM) were obtained by culture of adherent MNC (approximately 10 monocytes per flask) for 7 days in IMDM containing 10% autologous serum at 37°C in a humidified atmosphere containing 5% CO . Culture medium was replenished after 3 days. • Immediately before panning, flasks containing MDM were rinsed 3 times with prewarmed (37°C) RPMI containing 10% (v/v) foetal bovine serum, antibiotics and glutamine (RPMI-FBS) .

Hybridoma Cells • 4C/6 (IgGl-producing hybridoma) or 9el0 (IgGl- producing hybridoma) hybridoma cells were cultured in RPMI-FBS. Exponentially growing stocks (viability > 93% as defined by Trypan Blue exclusion) were maintained by subculturing every 2-3 days. • Cells used for panning were not subcultured for 5-7 days in order to increase the proportion of dead and apoptotic cells. Immediately before panning, these "neglected" hybridomas were resuspended in prewarmed RPMI-FBS containing 10% (v/v) culture medium harvested from exponentially growing cells.

Panning • Supernatant culture medium from flasks containing MDMs was discarded and hybridoma cells (2ml per flask) were added at a concentration that resulted in a ratio of 20 non-viable hybridomas per MDM. • Flasks were incubated at 37°C in a humidified atmosphere containing 5% C0₂ for 2 hours and were rocked gently every 30 minutes.

• Hybridoma cells that were not attached to or ingested by MDMs were recovered by decanting culture medium then rinsing the cell monolayer with warm hybridoma CM (10ml) .

• Cells recovered from panning and unpanned controls were resuspended at 2x10⁵/ml in fresh RPMI-FBS for reculture (5ml in 25cm² flasks) .

• Viability of recultured cells and IgGl production was measured at intervals up to 11 days. IgGl was measured by single radial immunodiffusion using a sheep anti-mouse IgG antiserum and purified mouse IgGl as calibrator.

Results

Panning resulted in >90% depletion of non-viable hybridoma cells with <1% loss of viable cells (Table 5).

Table 5 - Efficiency of Panning on Human Macrophages for Removal of Non-viable Hybridoma Cells.

Hybridoma Cells Non-viable(%) Yield of Viable Cells (%)

Unpanned control 43.1

Post-panning 1.9 >99

Hybridoma cells depleted of non-viable cells produced 31% more IgG than unpanned controls after 4 days culture; 43% more after 11 days (Table 6) And in a second experiment, 100% more after 14-23 days (figure 6) . Consistent with this, twice as many viable cells (3.5xl0⁵/ml) remained in cultures of panned cells compared with controls (1.7x10⁵/ml) after 11 days. Table 6 - Hybridoma IgG Production after Removal of

Non-viable/Apoptotic Cells by Panning on Human Macrophages . IgGl concentration in culture supernatant (μμ n )

Hybridoma Cells Day 4 Day 11

Unpanned control 10.3 16.5

Post-panning 13.5 23.6

Example 6; Anti-phosphatidγlserine (anti-PS) antibody binds preferentially to dead and dying cells and parts of dead and dying cells ("debris').

The inventors have shown that an antibody directed to phosphatidylserine binds preferentially to dead and dying cells. Using standard indirect immunofluorescence techniques (as known to people skilled in the art) a population of a human B cell line was stained with anti-PS Ab. By presenting light scatter data from strongly staining cells it is clear that the Ab binds preferentially to dead and dying cell zone cells (figure 7a) . Staining of sub-cellular sized particles (debris) is evident (figure 7a) . Further evidence that this 1H6 (anti-PS) antibody binds sub- cellular debris was noted (figure 7b) : cultures were subjected to different centrifugation protocols which removed a proportion or all cells leaving sub-cellular debris within the culture supernatants. Incubation of these supernatants (but not culture medium alone) with lH6-coated beads (anti-PS antibody) resulted in bead agglutination in the absence of cells. Using similar analytical techniques the inventors show that the anti-PS antibody stains non-viable cells from a hybridoma culture with clear staining of sub- cellular particles (figure 8) . Example 7: Increased IgG production by hybridomas after panning on anti-PS antibody

The inventors show that panning a culture of hybridoma cells on anti-PS antibody 1H6 improves productivity as follows:

Methods and Materials.

Anti-Phosphatidylserine (Anti-PS) mAb: Clone 1H6, mouse IgG2a from Upstate UK Ltd.

Preparation of Foam: Anti-PS mAb was passively coated onto pre-wetted polyether form cores (0.5cm dia x 1.5cm) by overnight incubation of the core in a solution containing 50μg/ml antibody in 0.05M carbonate/bicarbonate buffer, pH9.6 at fridge temperature .

Panning of hybridoma cells on foam coated with 1H6: Panning was carried out by passing hybridoma cells through the antibody coated foam as follows:

• Cells were not subcultured for 5-7 days in order to increase the proportion of dead and apoptotic cells. Immediately before adding to foam cores, these ""neglected" hybridomas were resuspended in fresh RPMI-FCS. • Antibody-coated foam cores were washed 3x in RPMI- FCS then evacuated by compression using forceps and transferred to wells of a 24 well polystyrene tissue culture plate each containing hybridoma cells (0.5ml, 2xl0⁶/ml) . Cells were drawn into foam by negative pressure. • Loaded foam cores were placed on their sides in wells and plates were incubated at fridge temperature for 45 mins., rotating cores through 180° after 20 mins to give further opportunity for exposure of cells to solid phase. • Cores were transferred to a 6ml capacity vial containing 2ml cold RPMI-FCS, capped and inverted for 30s. • Cores were removed with due care to avoid compression and cells counted and viability estimated.

Assessment of antibody productivity: cells released from the panning procedure above were cultured for 7 days post-panning and the produced antibody assessed by single radial im uno-diffusion assay:

Results

The inventors found that panning by passage of cells through foams coated with anti-PS produced 50% more IgG after 7 days reculture (Fig.9) suggesting that depletion of dead and dying cells and cell debris are important in improving productivity.

Example 8: Binding of dead and dying cells with reagents reactive with intracellular components

The inventors have shown that an antibody directed to actin binds preferentially to dead and dying cells. Further they have shown that a non-antibody reagent reactive with actin similarly binds preferentially to dead and dying cells. Using standard indirect immunofluorescence techniques (as known to people skilled in the art) a population of a human B cell line was stained with anti-actin Ab (AC15, Sigma Ltd) . By presenting light scatter data from strongly staining cells it is clear that the Ab binds preferentially to dead and dying cell zone cells (figure 10a) . Further evidence that reagents reactive with intracellular components can be used to bind dead and dying cells is provided through the use of phalloidin, a fungal toxin well known to react to actin. Using standard direct immunofluorescence techniques (as known to people skilled in the art) a population of antibody-producing hybridoma cells was stained with phalloidin-FITC (Sigma Ltd) . By presenting light scatter data from strongly staining cells it is clear that the phalloidin binds preferentially to dead and dying cell zone cells (figure 10b) . This clearly exempli ies that reagents reactive to intracellular components such as cytoskeletal components may be used to bind and discriminate dead and dying cells.

Claims

CLAIMS 1. A method of improving productivity of a cell culture by removing dying, dead cells and/or parts thereof from the cell culture, comprising the steps of: a) providing a cell culture; b) subjecting the cell culture to conditions conducive to allow production of a desired product by the cell culture; c) contacting cells from the cell culture with a reagent which is capable of preferentially binding to dying, dead cells and/or parts thereof from the cell culture; and d) removing at least a portion of said dying, dead cells and/or parts thereof from the cell culture which are bound to the reagent.

2. The method according to claim 1 wherein steps b) to d) are repeated as necessary.

3. The method according to claim 1 wherein essentially steps b) to d) are carried out in a continuous manner during the cell culturing process.

4. The method according to any preceding claim wherein improved productivity results in increased production of a biopharmaceutical, recombinant protein, antibody or cells.

5. The method according to any preceding claim wherein an improvement in production of greater than 2% is observed.

6. The method according to any preceding claim wherein the cell culture comprises eukaryotic cells.

7. The method according to any preceding claim wherein cells from the cell culture are passed over or otherwise allowed to contact a substrate to which is adhered the reagent (directly or indirectly) .

8. The method according to claim 7 wherein the substrate is glass, a natural or synthetic polymer material or metal.

9. The method according to either of claims 7 or 8 wherein the substrate is in the form of a flat planar surface; the inner surface of a tube; available surface of an open-celled or partially opened-cell foam, or the surface of a bead.

10. The method according to either of claims 7 or 8 wherein the substrate is in the form of a cartridge or monolith comprising a plurality of surfaces to which the reagent is adhered.

11. The method according to any preceding claim wherein greater than 10% of intact dead and/or dying cells are removed from the cell culture.

12. The method according to any preceding claim wherein the reagent which is capable of preferentially binding to dying and/or dead cells, is a macrophage, macrophage-like (e.g. THP-1) cell or cells or other phagocytic cell (e.g. fibroblast) , protein, or cell expressing a protein capable of binding dead and/or dying cells (e.g. integrins (e.g. α_vβ₃; _vβ₅) , CD36, SR-A, MER, CD14, ABCA1, PSR, CR3 , CR4 , CD91/calreticulin, CD31, FcγR, SHPS-1, SRB1, asialoglycoprotein receptor, thrombospondin-1, Gas-6, protein S, MFG-E8, Del-1, β2GPI, C-reactive protein or serum amyloid protein, annexin V, soluble CD14, phalloidin) , proteins or antibodies which preferentially bind with intracellular molecules or molecules expressed on the surface of dead and/or dying cells, or combinations thereof.

13. An apparatus for improved culturing of cells, the apparatus comprising a culture vessel for culturing cells, and a cell separation system for separating and enabling removal of dead, dying cells and/or parts thereof from a cell culture, the cell separation system comprising a solid substrate or substrates to which is/are adhered (directly or indirectly) a reagent or reagents which is/are capable of preferentially binding to dead, dying cells and/or parts thereof of said cell culture.

14. The apparatus according to claim 13 wherein the cell culture vessel and cell separation system are connected to one another, such that cell culture medium comprising cells may be passed from the cell culture vessel to the separation system and thereafter back to the culture vessel.

15. The apparatus according to claim 13 wherein the cell culture vessel contains the cell separation system, such that cell culture medium comprising cells contacts the separation system within the culture vessel.

16. The apparatus according to either of claims 13- 15 wherein the reagent or reagents capable of preferentially binding dead and/or dying cells, is a macrophage,macrophage-like (e.g. THP-1) cell or cells or other phagocytic cell (e.g. fibroblasts) , protein, or cell expressing a protein capable of binding dead and/or dying cells (e.g. integrins (e.g. α_vβ₃,* a_vβs), CD36, SR-A, MER, CD14 , ABCA1, PSR, CR3 , CR4 , CD91/calreticulin, CD31, FcγR, SHPS-1, SRB1, asialoglycoprotein receptor, thrombospondin-1, Gas-6, protein S, MFG-E8, Del-1, β2GPI, C-reactive protein or serum amyloid protein, annexin V, soluble CD14, phalloidin) , proteins or antibodies which preferentially bind with intracellular molecules or molecules expressed on the surface of dead and/or dying cells, or combinations thereof.

17. Use of soluble CD14 to bind apoptotic cells or dead cells wherein soluble CD14 refers to any non- membrane-associated CD14

18. Use according to claim 17 carried out in vitro or in vivo .

19. Use according to either of claims 17 or 18 further comprising the step of identifying said apoptotic or dead cells.

20. Use according to claim 19 wherein the SCD14 is labelled.

21. Use of sCD14 to target a toxic/therapeutic moiety or prodrug to a site of dying or dead cells.