WO2023198450A1 - Methods for high recovery of a cell culture clarification product - Google Patents

Abstract

The present invention relates to cell culture clarification. In particular the present invention relates to methods for high recovery of a biomolecule of interest, such as an antibody, from a cell culture clarification filter. More in particular, to methods for high recovery of a biomolecule of interest from a cell culture clarification filter using Arginine-HCl buffer. The present invention also discloses methods of cell culture clarification including a filtration step wherein the recovery of the biomolecule of interest from the filter is maximized by a recovery flush step performed with Arginine-HCl buffer.

Description

Methods for high recovery of a cell culture clarification product

TECHNICAL FIELD

The present invention relates to cell culture clarification. In particular the present invention relates to methods for high recovery of a biomolecule of interest, such as an antibody, from a cell culture clarification filter. More in particular, to methods for high recovery of a biomolecule of interest from a cell culture clarification filter using Arginine-HCI buffer. The present invention also discloses methods of cell culture clarification including a filtration step wherein the recovery of the biomolecule of interest from the filter is maximized by a recovery flush step performed with Arginine-HCI buffer.

BACKGROUND

The manufacturing process of a biopharmaceutical molecule, such as an antibody, is complex and it comprises different steps, each requiring extensive optimizations. The process begins with the selection of a cell line to use for expressing and secreting the biomolecule of interest. Once selected, this cell line is amplified and cultivated at larger scales in bioreactors where the biomolecule of interest is produced under controlled conditions. These steps are referred to as the upstream process (USP) part of the whole biomolecule manufacturing process.

In order to remove residual biomass and impurities from the culture medium containing the biomolecule of interest, such as an antibody, a clarification process is necessary. Cell culture clarification requires multiple steps directed to the removal of different types of impurities spanning a wide range of size, including residual cells, cell debris and eventually DNA and host cell proteins (HCPs). Importantly, the clarification process must meet the downstream process requirements in terms of product stability and purity and at the same time it needs to be robust enough to permit a high recovery of the obtained biomolecule of interest.

Cell culture clarification often starts with a primary clarification step which allows the removal of large particles such as whole cells and cell debris. The primary clarification is then followed by a secondary clarification step, which allows the removal of smaller particles such as HCPs and DNA. Primary and secondary clarification may require one or more clarification operational unit such as filters, centrifuges, acoustic separator, and their combinations. For instance, a clarification process may comprise one or more steps performed by a depth filter, namely by a filter whose porosity is such that it retains particles of a cell culture throughout the medium, rather than just on the surface. Depth filters can be used for primary or secondary clarification only, or for both primary and secondary clarification, alone or in combination with other clarification units. When using depth filters in the clarification process, a challenge that arises is the product recovery post-clarification. In fact, because of the binding between the filter media and the product, loss of the biomolecule of interest can be observed. Accordingly, given the increasing complexity of biological molecules and the development of various membranes for depth filtration, such as synthetic membranes, obtaining high product yield post-clarification still remains a challenge. The present invention allows to solve this problem so as that the recovery of the biomolecule of interest is maximized.

SUMMARY

In the manufacturing process of a biomolecule of interest, such as a therapeutic antibody, the cell culture clarification is the step which allows the removal of cell culture material such as cells, cell debris and impurities from the cell culture medium to obtain a clarified cell culture where mainly the biomolecule of interest (herein also called "product") is present in the cell culture medium. The present invention relates to cell culture clarification. In particular, the present invention relates to methods for high recovery of a biomolecule of interest, such as an antibody, from a cell culture clarification filter. More in particular, to methods for high recovery of a biomolecule of interest from a cell culture clarification filter using Arginine-HCI buffer. The present invention also discloses methods of cell culture clarification including a filtration step wherein the recovery of the biomolecule of interest from the filter is maximized by a recovery flush step performed with Arginine-HCI buffer.

The present invention relates to a method for recovering a biomolecule of interest from a cell culture clarification synthetic depth filter comprising the step of flushing said depth filter with Arginine-HCI buffer.

In particular wherein the biomolecule of interest is an antibody.

More in particular, wherein the biomolecule of interest is a non-naturally occurring antibody.

Even more in particular, wherein the biomolecule is a multispecific and/or multivalent antibody, such as a bispecific antibody.

In one embodiment, Arginine-HCI buffer is used at a concentration comprised between about 600 and about 1000 mM at a pH comprised between about 5 and about 7.

In a more specific embodiment, Arginine-HCI buffer is used at 800 mM at pH=6. The present invention also relates to a process for clarifying a cell culture including a biomolecule of interest comprising at least a filtration step performed by a synthetic depth filter, the process characterized in that it further comprises recovering said biomolecule of interest from said depth filter according to the method of the present invention for recovering a biomolecule of interest from a cell culture clarification synthetic depth filter comprising the step of flushing said depth filter with the Arginine-HCI buffer disclosed herein.

In particular, the process disclosed herein comprises the consequent steps of: i. flushing said depth filter with water for injection; ii. flushing said depth filter with 1 X phosphate buffer saline 140 mM; ill. connecting the cell culture to the filter; iv. flushing said depth filter with the cell culture fluid; v. discarding the dead volume; vi. collecting the clarified cell culture fluid; vii. recovering said biomolecule of interest from said depth filter by flushing said depth filter with Arginine-HCI buffer according to the method of the present invention.

More in particular, the process disclosed herein further comprises a step of subjecting said recovered biomolecule of interest to one or more steps of chromatography purification.

The present invention also discloses a process of production of a drug substance comprising the steps of: i. seeding cells expressing a biomolecule of interest in a cell culture medium; ii. culturing said cells for a period comprised between 10 and 18 days, preferably for 14 days; ill. subjecting the obtained cell culture to the clarification process according to the present invention; iv. add excipients to the biomolecule of interest purified.

Unless otherwise defined, all terms used in this disclosure, including technical and scientific terms, have the meaning as commonly understood by a person skilled in the art to which this disclosure belongs. By means of further guidance, terms definition are included to better appreciate the teaching of the present disclosure.

As used herein, the following terms have the following meanings: "a", "an", and "the" as used herein refers to both singular and plural unless the context clearly dictates otherwise. The term "biomolecule of interest" as used herein refers to a product of interest, which is desired to be purified or separated from one or more undesirable entities, e.g., one or more impurities, which may be present in a sample containing the product of interest. In a preferred embodiment of the present invention, the biomolecule of interest is a polypeptide. In a more preferred embodiment, the biomolecule of interest is a protein. In an even more preferred embodiment, the biomolecule of interest is an antibody or an antibody fragment thereof. In a particularly preferred embodiment of the present invention, the biomolecule of interest is an isolated polypeptide, for instance an antibody such as a monoclonal antibody or a monoclonal antibody fragment thereof. In certain embodiments the antibody is a non-naturally occurring or recombinant or engineered molecule, for instance a multispecific antibody. A "recombinant" molecule is one that has been prepared, expressed, created, or isolated by recombinant means.

An "isolated polypeptide" is one that: (1) is free of at least some other polypeptides with which it would normally be found, (2) is essentially free of other polypeptides from the same source, e.g., from the same species, (3) is expressed by a cell from a different species, (4) has been separated from at least about 50 percent of polynucleotides, lipids, carbohydrates, or other materials with which it is associated in nature, (5) is not associated (by covalent or noncovalent interaction) with portions of a polypeptide with which the "isolated polypeptide" is associated in nature, (6) is operably associated (by covalent or noncovalent interaction) with a polypeptide with which it is not associated in nature, or (7) does not occur in nature. Such an isolated polypeptide can be encoded by genomic DNA, cDNA, mRNA or other RNA, of synthetic origin, or any combination thereof. Preferably, the isolated polypeptide is substantially free from polypeptides or other contaminants that are found in its natural environment that would interfere with its use (therapeutic, diagnostic, prophylactic, research or otherwise).

In the present invention, the term "antibody" and the term "immunoglobulin" are used interchangeably. The term "antibody" as referred to herein, includes the full-length antibody and antibody fragments. Antibodies are glycoproteins produced by plasma cells that play a role in the immune response by recognizing and inactivating antigen molecules. In mammals, five classes of immunoglobulins are produced: IgM, IgD, IgG, IgA and IgE. In the native form, immunoglobulins exist as one or more copies of a Y-shaped unit composed of four polypeptide chains: two identical heavy (H) chains and two identical light (L) chains. Covalent disulfide bonds and non-covalent interactions allow inter-chain connections; particularly heavy chains are linked to each other, while each light chain pairs with a heavy chain. Both heavy chain and light chain comprise an N-terminal variable (V) region and a C-terminal constant (C) region. In the heavy chain, the variable region is composed of one variable domain (VH), and the constant region is composed of three or four constant domains (CHI, CH2, CH3 and CH4), depending on the antibody class; while the light chain comprises a variable domain (VL) and a single constant domain (CL). The variable regions contain three regions of hypervariability, termed complementarity determining regions (CDRs). These form the antigen binding site and confer specificity to the antibody. CDRs are situated between four more conserved regions, termed framework regions (FRs) that define the position of the CDRs. Antigen binding is facilitated by flexibility of the domains position; for instance, immunoglobulin containing three constant heavy domains present a spacer between CHI and CH2, called "hinge region" that allows movement for the interaction with the target. Starting from an antibody in its intact, native form, enzymatic digestion can lead to the generation of antibody fragments. For example, the incubation of an IgG with the endopeptidase papain, leads to the disruption of peptide bonds in the hinge region and to the consequent production of three fragments: two antibody binding (Fab) fragments, each capable of antigen binding, and a cristallizable fragment (Fc). Digestion by pepsin instead yields one large fragment, F(ab')2, composed by two Fab units linked by disulfide bonds, and many small fragments resulting from the degradation of the Fc region.

The term "antibody" as referred to herein includes whole antibodies and any antigen binding fragments or single chains thereof. Naturally occurring antibodies typically comprise a tetramer. Each such tetramer is typically composed of two identical pairs of polypeptide chains, each pair having one full-length "light" chain (typically having a molecular weight of about 25 kDa) and one full-length "heavy" chain (typically having a molecular weight of about 50-70 kDa). The terms "heavy chain" and "light chain" as used herein refer to any immunoglobulin polypeptide having sufficient variable domain sequence to confer specificity for a target antigen. The amino-terminal portion of each light and heavy chain typically includes a variable domain of about 100 to 110 or more amino acids that typically is responsible for antigen recognition. The carboxy-terminal portion of each chain typically defines a constant domain responsible for effector function. Thus, in a naturally occurring antibody, a full-length heavy chain immunoglobulin polypeptide includes a variable domain (VH) and three constant domains (CHI, CH2, and CH3), wherein the VH domain is at the amino-terminus of the polypeptide and the CH3domain is at the carboxyl-terminus, and a full-length light chain immunoglobulin polypeptide includes a variable domain (VL) and a constant domain (CL), wherein the VL domain is at the amino-terminus of the polypeptide and the CL domain is at the carboxyl-terminus.

Human light chains are typically classified as kappa and lambda light chains, and human heavy chains are typically classified as mu, delta, gamma, alpha, or epsilon, and define the antibody's isotype as IgM, IgD, IgG, IgA, and IgE, respectively. IgG has several subclasses, including, but not limited to, IgGl, lgG2, lgG3, and lgG4. IgM has subclasses including, but not limited to, IgMl and lgM2. IgA is similarly subdivided into subclasses including, but not limited to, IgAl and lgA2. Within full-length light and heavy chains, the variable and constant domains typically are joined by a "J" region of about 12 or more amino acids, with the heavy chain also including a "D" region of about 10 more amino acids. See, e.g., FUNDAMENTAL IMMUNOLOGY (Paul, W., ed., Raven Press, 2nd ed., 1989), which is incorporated by reference in its entirety for all purposes. The variable regions of each light/heavy chain pair typically form an antigen binding site. The variable domains of naturally occurring antibodies typically exhibit the same general structure of relatively conserved framework regions (FR) joined by three hypervariable regions, also called complementarity determining regions or CDRs. The CDRs from the two chains of each pair typically are aligned by the framework regions, which may enable binding to a specific epitope. From the amino-terminus to the carboxyl-terminus, both light and heavy chain variable domains typically comprise the domains FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4.

The term "Fc" as used herein refers to a molecule comprising the sequence of a non-antigen-binding fragment resulting from digestion of an antibody or produced by other means, whether in monomeric or multimeric form, and can contain the hinge region. The original immunoglobulin source of the native Fc is preferably of human origin and can be any of the immunoglobulins. Fc molecules are made up of monomeric polypeptides that can be linked into dimeric or multimeric forms by covalent {i.e., disulfide bonds) and non- covalent association. The number of intermolecular disulfide bonds between monomeric subunits of native Fc molecules ranges from 1 to 4 depending on class {e.g., IgG, IgA, and IgE) or subclass {e.g., IgGl, lgG2, lgG3, IgAl, lgGA2, and lgG4). One example of a Fc is a disulfide-bonded dimer resulting from papain digestion of an IgG. The term "native Fc" as used herein is generic to the monomeric, dimeric, and multimeric forms.

A F(ab) fragment typically includes one light chain and the VH and CHI domains of one heavy chain, wherein the VH-CH1 heavy chain portion of the F(ab) fragment cannot form a disulfide bond with another heavy chain polypeptide. As used herein, a F(ab) fragment can also include one light chain containing two variable domains separated by an amino acid linker and one heavy chain containing two variable domains separated by an amino acid linker and a CHI domain.

A F(ab') fragment typically includes one light chain and a portion of one heavy chain that contains more of the constant region (between the CHI and CH2 domains), such that an interchain disulfide bond can be formed between two heavy chains to form a F(ab')2molecule.

The term "antibody fragments" as used herein, includes one or more portion(s) of a full-length antibody. Non limiting examples of antibody fragments include: (i) the fragment crystallizable (Fc) composed by two constant heavy chain fragments which consist of CH2 and CH3 domains, in IgA, IgD and IgG, and of CH2, CH3 and CH4 domains, in IgE and IgM, and which are paired by disulfide bonds and non-covalent interactions; (ii) the fragment antigen binding (Fab), consisting of VL, CL and VH, CHI connected by disulfide bonds; (iii) Fab', consisting of VL, CL and VH, CHI connected by disulfide bonds, and of one or more cysteine residues from the hinge region; (iv) Fab'-SH, which is a Fab' fragment in which the cysteine residues contain a free sulfhydryl group; (v) F(ab')2 consisting of two Fab fragments connected at the hinge region by a disulfides bond; (vi) the variable fragments (Fv), consisting of VL and VH chains, paired together by non-covalent interactions; (vii) the single chain variable fragments (scFv), consisting of VL and VH chains paired together by a linker; (ix) the bispecific single chain Fv dimers, (x) the scFv-Fc fragment; (xi) a Fd fragment consisting of the VH and CHI domains; (xii) the single domain antibody, dAb, consisting of a VH domain or a VL domain; (xiii) diabodies, consisting of two scFv fragments in which VH and VL domains are connected by a short peptide that prevent their pairing in the same chain and allows the non-covalent dimerization of the two scFvs; (xiv) the trivalent 10 triabodies, where three scFv, with VH and VL domains connected by a short peptide, form a trimer. (xv) half-IgG, comprising a single heavy chain and a single variable chain.

Depending on their nature, antibodies and antibody fragments can be monomeric or multimeric, monovalent or multivalent, monospecific or multispecific.

The term "monospecific antibody" as used herein, refers to any antibody or fragment having one or more binding sites, all binding the same epitope.

The term "multispecific antibody" as used herein, refers to any antibody or fragment having more than one binding site that can bind different epitopes of the same antigen, or different antigens. A non-limiting example of multispecific antibodies are bispecific antibody, which have two binding sites that can bind two different epitopes of the same antigen, or two different antigens.

The term "monoclonal antibody" (MAb), as used herein, refers to a population of antibody molecules that contain only one molecular species of antibody molecule consisting of a unique light chain gene product and a unique heavy chain gene product. In particular, the complementarity determining regions (CDRs) of the monoclonal antibody are identical in all the molecules of the population. MAbs contain an antigen binding site capable of immunoreacting with a particular epitope of the antigen characterized by a unique binding affinity for it.

The biomolecule of interest, such as an antibody, can be produced by introducing genetic material encoding said biomolecule of interest in host cells. The term "host cells" refers to all the cells in which the biomolecule of interest codified by the artificially introduced genetic material is expressed, including those cells in which the foreign nucleic acid is directly introduced and their progeny. In the host cells it can be introduced an expression vectors (constructs), such as plasmids and the like, encoding the biomolecule of interest e.g., via transformation, transfection, infection, or injection. Such expression vectors normally contain the necessary elements for the transcription and translation of the sequence encoding the biomolecule of interest. Methods which are well known to and practiced by those skilled in the art can be used to construct expression vectors containing sequences encoding the protein of interest, as well as the appropriate transcriptional and translational control elements. These methods include in vitro recombinant DNA techniques, synthetic techniques, and in vivo genetic recombination. Cell lines suitable as host cells include and are not limited to bacteria, mammalian, insect, plant and yeast cells. Cell lines often used for the expression and production of therapeutic antibodies include mammalian cells lines such as Chinese hamster ovary (CHO) cells, NSO mouse myeloma cells, human cervical carcinoma (HeLa) cells and human embryonic kidney (HEK) cells.

In a particular embodiment of the present invention the cultured cells are mammalian cells, more in particular, they are CHO cells.

The terms "cell culture" and "culture" as used herein are interchangeable and refer to the growth and/or propagation and/or maintenance of cells in controlled artificial conditions, and they indicate a cell culture which comprises a cell culture medium and cell culture material comprising cells, cell debris, for instance generated upon cell death, colloidal particles, such as DNA, RNA and host cell proteins (HCP), and (bio)molecules secreted by the cultured cells, such as the biomolecule of interest. The cells of a cell culture can be cultured in suspension or attached to a solid substrate, in containers comprising a cell culture medium. For example, a cell culture can be grown in tubes, spin tubes, flasks, bags, roller bottles, bioreactors. In certain embodiments of the present invention, at the end of the cell culture the obtained titer of the biomolecule of interest is below 10 g/L, in other embodiments the obtained titer of the biomolecule of interest is comprised between about 1 g/L and about 10 g/L. Non limiting examples of the obtained titer of the biomolecule of interest include: about 0.1 g/L, about 0.5 g/L, about 1 g/L, about 1.5 g/L, about 2 g/L, about 2.5 g/L, about 3 g/L, about 3.5 g/L, about 4 g/L, about 4.5 g/L, about 5 g/L, about 5.5 g/L, about 6 g/L, about 6.5 g/L, about 7 g/L, about 7.5 g/L, about 8 g/L, about 8.5 g/L, about 9 g/L, about 9.5 g/L, about 10 g/L.

The cell culture to be clarified is also referred herein as cell culture fluid (CCF).

When the production of the biomolecule of interest has a commercial purpose, often the host cells are cultured in bioreactors, under conditions that aid their growth and the expression of said biomolecule of interest. The term "bioreactor," as used herein, refers to any manufactured or engineered device or system that supports a biologically active environment. Optimal culturing conditions are obtained by the control and adjustment of several parameters including: the formulation of the cell culture medium, the bioreactor operating parameters, the nutrient supply modality and the culturing time period. The formulation of the culturing medium has to be optimized to favorite cell vitality and multiplication; examples of constituents of the cell culture medium include but are not limited to essential amino acids, salts, glucose, growth factors and antibiotics. Important bioreactor operating parameters include: initial cell seeding density, temperature, pH, agitation speed, oxygenation and carbon dioxide levels. Nutrients can be supplied in different ways: in the batch mode culture all the necessary nutrients are present in the initial base medium and are used till exhausted while wastes accumulate; in the fed-batch culture additional feed medium is supplied to prevent nutrient depletion and prolong the culture; differently, in the perfusion modality, cells in culture are continuously supplemented with fresh medium containing nutrients that flows in the bioreactor removing cell wastes. The culturing period is important as it needs to be long enough to let the cells produce a consistent amount of product but it cannot be too long to impair their viability. Commonly used bioreactors are typically cylindrical, ranging in size from liters to cubic meters, and are often made of stainless steel. In the embodiments described herein, a bioreactor is made of plastic or of stainless steel. It is contemplated that the total volume of a bioreactor may be any volume ranging from 10 mL to up to 20000 L, e.g., from 100 mL to up to 20000 Liters or more, depending on a particular process. Non limiting examples of bioreactor volumes include about 100 mL, about 200 mL, about 500 mL, about 800 mL, about 1 L, about 5 L, about 10 L, about 20 L, about 30 L, about 40 L, about 50 L, about 60 L, about 70 L, about 80 L, about 90 L, about 100 L, about 200 L, about 300 L, about 400 L, about 500 L, about 600 L, about 700 L, about 800 L, about 900 L, about 1000 L, about 2000 L, about 3000 L, about 4000 L, about 5000 L, about 6000 L, about 7000 L, about 8000 L, about 9000 L, about 10000 L, about 15000 L, about 20000 L.

Bioreactors useful for present inventions include but are not limited to small scale bioreactors, single use bioreactors (SUB), shake flask vessels, large scale bioreactors, batch bioreactors, fed-batch bioreactors. In a particular embodiment of this invention cells are cultured in 3 to 5 L, or 50 L SUBs in fed-batch mode for a number of days comprised between 10 and 20 days, preferably between 12 and 14 days, most preferably for at least 12 days, even more preferably for 12 or 13 or 14 days in a cell culture medium.

The terms "cell culture medium," and "culture medium" and "medium" are used interchangeably herein and they refer to a nutrient solution used for growing cells, such as animal cells, e.g., mammalian cells. Such a nutrient solution generally includes various factors necessary for cell attachment, growth, and maintenance of the cellular environment. For example, a typical nutrient solution may include a basal media formulation, various supplements depending on the cell type and, occasionally, antibiotics. During cell culture the cell culture medium may also contain cell culture material such as cell waste products, host cell proteins (HCP) and material from lysed cells. The composition of the culture medium may vary in time during the course of the culturing of cells.

The terms "clarify", "clarification", "clarification step", "clarification process" as used herein are interchangeable and generally they refer to one or more steps that aid the removal of a part of the cell culture material from the cell culture (e.g. from the cell culture fluid (CCF)), such as removal of cells, cell debris and colloidal particles, to obtained clarified cell culture, also called clarified cell culture fluid (CCCF) herein, comprising the biomolecule of interest, the biomolecule of interest is also generally referred inhere as the "product".

The efficiency of the clarification step is crucial to facilitate the subsequent downstream processing steps of purification of the biomolecule of interest. Characteristics of the cell culture that have an impact on the clarification step include the total cell concentration, the cell viability, the initial turbidity of the cell culture to clarify, the concentration of biomolecule produced by the cultured cells. The term "Total cell concentration" (TCC) refers to the number of cells in a given volume of culture. The terms "Viable cell concentration" (VCC) refers to the number of live cells in a given volume of culture, as determined by standard viability assays (such as trypan blue dye exclusion method). The percentage of living cells is called "viability". In general terms, a higher TCC implies higher biomass to be removed from the cell culture and therefore higher impact on the clarification. Similarly for the lower viability, given the higher presence of cell debris. The term "turbidity" refers to the cloudiness or haziness of a liquid caused by large numbers of individual particles. In particular, the turbidity indicates the amount of material and small particles inside a liquid capable of light diffusion. More in particular, the turbidity of a cell culture may be due to the presence of cells, cell debris, colloidal particles, such as DNA, RNA and host cell proteins (HCP), and of the biomolecule of interest.

The cell culture clarification can start with a primary clarification step. The terms "primary clarification" and "primary recovery" as used herein are interchangeable and refer to the removal of large particles such as whole cells and cell debris. The primary clarification can be followed by a secondary clarification step. The terms "secondary clarification" and "secondary recovery" as used in the present patent application are interchangeable and indicate the removal of smaller particles.

Primary and secondary clarification may require one or more clarification operational unit such as filters, centrifuges, acoustic separator etc., and their combinations.

In particular embodiment of the present invention, the primary clarification step is a depth filtration. As used herein, the term "depth filtration" refers to a technology that exploits filters with a certain porosity to retain particles of a medium throughout the medium, rather than just on the surface. Examples of depth filters used for instance in biopharmaceutical industry include single-use devices in the form of a lenticular disks or cartridges which contain the filter sheets at small scale. Disks or cartridges can be assembled into multilayer housings at larger scales. Depth filters are typically composed of natural fibers such as cellulose fibers and filter aids like diatomaceous earth or perlite bound together into a polymeric resin. In some cases, cellulose fibers can be replaced by fully synthetic polymeric fibers like polyacrylic or polystyrene. As defined herein synthetic depth filters are the ones made of synthetic fibers, for instance of polyacrylic or polystyrene. Synthetic filters may further comprise other synthetic or non-synthetic fiber (i.e. natural fibers) and a binder. For instance, a synthetic filter may comprise silica fibers and a binder resin. Filters fibers form a three-dimensional network with a certain porosity. The permeability and retention characteristics of the filters are directly correlated with the length and compaction of those fibers. The porosity of the filters matrix can vary along with the filter depth, allowing the coverage a wide range of particle exclusion. Non limiting examples of depth filters useful for the present invention comprise filters made of cellulose and/or resin, and/or synthetic media and/or polypropylene and/or filters aids. A depth filter can be suitable for both primary and/or secondary recovery, or for primary or secondary recovery only. Filters suitable for primary and/or secondary recovery are also known as "single filters". Single filters can be applied alone to carried out both the primary and the secondary recovery, or they can be used as filters for primary recovery and coupled with the subsequent use of a filter for secondary recovery; or they can be used as filters for secondary recovery for instance after a centrifugation step for primary recovery. In the present invention single filters are also referred as "SF", filters for primary clarification are referred as "PCF" and filters for secondary clarification are referred as "SCF".

The present invention relates to methods for high recovery of a biomolecule of interest, such as an antibody, from a cell culture clarification filter, such as a depth filter or a sterile filer for bioburden reduction. More in particular, the present invention relates to methods for high recovery of a biomolecule of interest from a cell culture clarification depth filter, specifically from a cell culture clarification synthetic depth filter, in particular using Arginine-HCI buffer. The present invention also discloses methods of cell culture clarification including a filtration step wherein the recovery of the biomolecule of interest from the filter, in particular from a depth filer, preferably from a synthetic depth filter, is maximized by a recovery flush step performed with Arginine-HCI buffer, wherein the term "flush step" or "flushing" is defined as making the filter in contact with a fluid, such as the water, water for injection, PBS, the cell culture fluid, a buffer, Arginine-HCI buffer, and allows the fluid to go through the filter. In certain embodiments Arginine-HCI buffer has a concentration comprised between about 100 mM and 1600 mM, e.g., between about 400 mM and about 1200 mM, preferably comprised between about 600 mM and 1000 mM, more preferably comprised between about 700 mM and 900 mM; particularly Arginine-HCI buffer has a concentration selected from the group comprising about 400 mM, about 200 mM, about 300 mM, about 400 mM, about 500 mM, about 600 mM, about 700 mM, about 800 mM, about 900 mM, about 1000 mM, about 1100 mM, about 1200 mM, about 1300 mM, about 1400 mM, about 1600 mM, about 1600 mM. Preferably the concentration of Arginine-HCI buffer is about 800 mM. In certain embodiments Arginine-HCI buffer has a pH comprised between about 5 and 7, in particular Arginine-HCI buffer has a pH selected from the group comprising about 5, about 6, about 7. Preferably the pH of Arginine-HCI buffer is about 6. The present invention also includes concentration and pH of Arginine-HCI buffer at any intermediate value of the above said ranges. In a particular embodiment Arginine-HCI buffer has a concentration of about 800 mM and a pH of about 6.

The present invention also relates to a process for clarifying a cell culture including a biomolecule of interest comprising at least a filtration step performed by synthetic a depth filter, characterized in that said process further comprises recovering said biomolecule of interest from the depth filter by a recovery flush step performed with Arginine-HCI buffer.

More in particular, the process for clarifying a cell culture of the present invention comprising the consequent steps of: i. flushing said depth filter with water for injection; ii. flushing said depth filter with 1 X phosphate buffer saline 140 mM; ill. connecting the cell culture to the filter; iv. flushing said depth filter with the cell culture fluid; v. discarding the dead volume; vi. collecting the clarified cell culture fluid; vii. recovering said biomolecule of interest from said depth filter by flushing said depth filter with Arginine-HCI buffer according to the method of the present invention.

More in particular the process disclosed herein further comprises a step of subjecting said recovered biomolecule of interest to one or more steps of chromatography purification.

The term "chromatography" refers to the operation of separating compounds of a mixture based on their capability to interact with a stationary phase of a chromatography column, from which they can be retained or eluted. The present invention also discloses a process of production of a drug substance comprising the steps of: i. seeding cells expressing a biomolecule of interest in a cell culture medium; ii. culturing said cells for a period comprised between 10 and 18 days, preferably for 14 days; ill. subjecting the obtained cell culture to the clarification process according to the present invention; iv. add excipients to the biomolecule of interest purified.

In one aspect of the present invention, the primary clarification is performed by a first depth filter selected from the group comprising depth filters for primary clarification and single filters. In another aspect of the present invention the first depth filter has an exclusion range equal to or greater than about 0.1 pm and equal to or less than about 50 pm, more particularly the exclusion range is comprised between about 0.25 pm and about 30 pm. In certain particular embodiments the exclusion range is comprised between about 1 pm and about 20 pm, or comprised between about 5 pm and about 30 pm, preferably comprised between about 6 pm and about 30 pm, or comprised between about 0.5 pm and about 10 pm, preferably comprised between about 0.55 pm and about 8 pm, or comprised between about 0.2 pm and about 2 pm, or comprised between about 1.5 pm and about 10 pm, or comprised between about 0.7 pm and about 5 pm, or comprised between about 0.25 pm and about 5 pm. In a specific aspect the first depth filter has an exclusion range selected from the group comprising at least about 0.1 pm, at least about 0.2 pm, at least about 0.5 pm, at least about 0.7 pm, at least about 1 pm, at least about 1.5 pm, at least about 2 pm, at least about 5 pm, at least about 7 pm, at least about 10 pm, at least about 20 pm, at least about 30 pm. The present invention also includes first depth filters with an exclusion range at any intermediate value of the above said ranges.

In another particular embodiment of the present invention, the secondary clarification step is carried out by a second depth filtration.

In one aspect of the present invention, the second clarification is performed by a second depth filter selected from the group comprising depth filters for secondary clarification, single filters and postflocculation filters. In another aspect of the present invention the second depth filter has an exclusion range comprised between about 0.01 pm and about 10 pm, more in particular comprised between about 0.05 pm and about 5 pm, even more in particular comprised between about 0.1 pm and about 4 pm, in a further particular embodiment the second depth filter exclusion range in comprised between about 0.2 pm and about 3.5 pm. In a specific aspect, the second depth filter has an exclusion range equal to or less than about 3.5 pm, or equal to or less than about 3 pm, or equal to or less than about 2 pm, or equal to or less than about 1 pm, or equal to or less than about 0.5 pm, or equal to or less than about 0.2 pm, or equal to or less than about 0.1 pm. The present invention also includes second depth filters with an exclusion range at any intermediate value of the above said values.

In certain embodiments of the present invention, when the cell culture has a turbidity comprised between 1000 NTU and 5000 NTU less than about 3000 NTU, e.g. selected from the group comprising about 1000 NTU, about 1500 NTU, about 2000 NTU, about 2500 NTU, about 3000 NTU, about 3500 NTU, about 4000 NTU, about 4500 NTU and about 5000 NTU. The present invention also comprises turbidity at any intermediate values of the ones said above.

In certain embodiments of the present invention, when a combination of a depth filter for primary clarification and a depth filter for secondary clarification is used, the surface ratio between the first and the second depth filter is selected from the group comprising 1:1, 2:1, 1:2.

In another embodiment of the present invention, when the turbidity of the cell culture is less than 3000 NTU, the primary clarification step and the secondary clarification step are performed by a single filter with an exclusion range comprised between about 1 pm and about 20 pm.

In one embodiment of the present invention the second clarification is followed by a bioburden reduction.

The term "bioburden reduction" as used herein refers to the reduction of the number of microorganisms in the fluid obtained after the primary and secondary clarification. Normally bioburden reduction is considered the final step of the clarification process and comprises one or more steps of sterile purification. In one embodiment of this invention the bioburden reduction is performed by at least one sterile filter having exclusion range equal to or less than about 0.5 pm. In a more specific embodiment, the bioburden reduction is performed by at least one sterile filter having exclusion range from about 0.2 pm to about 0.45 pm.

In one aspect of the present invention, the secondary depth filtration may be followed by a further filtration performed by a membrane absorber with exclusion range equal to less than 0.2 pm.

At the end of the clarification process, the clarified cell culture may be further subjected to one or more steps of purification to isolate and recover the biomolecule of interest. Standard purification methods include chromatographic techniques, including ion exchange, hydrophobic interaction, affinity, sizing or gel filtration, and reversed-phase, carried out at atmospheric pressure or at high pressure using systems such as FPLC and HPLC. Purification methods also include electrophoretic, immunological, precipitation, dialysis, and chromatofocusing techniques. Ultrafiltration and diafiltration techniques, in conjunction with protein concentration, are also useful. Figure 1: Molecule A (left panel) and B (right panel) cell culture VCC and viability.

Figure 2: Molecule A (left panel) and B (right panel) cell culture titer concentration over the fed-batch concentration.

Figure 3: Molecule A (experiments 1 to 8, 12 and 13) and B (experiments 9 to 11) cell culture PCV (Packed cell volume) and turbidity.

Figure 4: Molecule A CCCFs comparison at small and bench scale (Each error bar is constructed using 1 standard deviation from the mean. N = 9 for bench scale and N = 4 for small scale).

Figure 5: Summary of reference process CCCF results with molecule A and molecule B CCF (Each error bar is constructed using 1 standard deviation from the mean. N = 13 for molecule A and N = 7 for molecule B).

Figure 6: High flow impact on product yields (For experiment 8, no reference process run was performed with the same initial material. For comparison, the reference process product yield average is displayed. Each error bar is constructed using 1 standard deviation from the mean. N = 13 for reference experiment 8).

Figure 7: High flow impact on product yield summary (Each error bar is constructed using 1 standard deviation from the mean. N = 16 for reference condition and N = 4 for high flow condition).

Figure 8: Screening of equilibration and recovery buffers runs overview.

Figure 9: Screening of equilibration buffers yields results PBS equilibration 140 mM = reference condition.

Figure 10: Screening of different salt concentration of recovery PBS yields results.

Figure 11: Calcium chloride 250 mM recovery flush yield result.

Figure 12: Product yields obtained post screening of arginine-HCI recovery flush.

Figure 13: Evaluation of the dead volume recovery flush number (Each error bar is constructed using 1 standard deviation from the mean. N = 2 for each bar).

Figure 14: Screening on molecule B CCFs with arginine-HCI flush.

Figure 15: Impact of arginine-HCI flush on molecule A and molecule B (Each error bar is constructed using 1 standard deviation from the mean. N = 7 for molecule A bars and N = 2 for molecule B bars).

Figure 16: Correlation between product recovery increase with arginine-HCI (%) and product titer (%). Example 1: Material and methods

Cell culture process

Molecule A (bispecific antibody) and B (IgG-like antibody) were produced by two different CHO cell lines derived from the same mother cell line.

Table 1. Overview of the cell culture fluid (CCF) used for the optimization of a reference clarification process.

1) Expansion and fed-batch

CHO cells were thawed and expanded in Power-CHO-2-CD medium (serum free and chemically defined) from Lonza/Sartorius. The duration of a passage was either 2 or 3 days with the aim of reaching 3.0 x 10^s cells/mL at the end of the passage. Cells were diluted to respectively 0.4 x 10^sor 0.9 x 10^s cells/mL for a 2-day or a 3-day passage. Cells were scaled-up in shake flasks before the fed-batch inoculation at 0.8 x 10^scells/mL.

CHO cells were grown in fed-batch mode either in Thomson 5 L or in single-use BioBlu 3c bioreactors with a final working volume of 2 L or 3 L, respectively. Power-CHO-2-CD medium from Lonza/Sartorius supplemented with 1 g/L of an in-house formulated Kolliphor (shear stress protector) solution was used as a basal medium. The cell culture process was a 14 days fed-batch culture. The feeding started at day 2 consisting in a daily addition of a total of 30%f_eed voiume/working volume of Cell Boost 7a (feed "7a") and 3%feed voiume/working volume of Cell Boost 7b (feed "7b"). Both feeds 7a and 7b are manufactured by Cytiva (previously General Electric) and are chemically defined, animal-derived component-free supplements. Feed 7a supplements cell culture with amino acids, vitamins, salts, trace elements and glucose while feed 7b is a concentrated supplement of amino acids with an alkaline pH. From day 5, glucose threshold and target were respectively set at 5 g/L and 6 g/L meaning that when the glucose concentration went below 5 g/L, in-house formulated glucose was added to reach 6 g/L.

CCFs were harvested when the viability dropped below 75 % or at day 14 even if the viability was still above 75 %.

An overview of the CCFs used for all the experiments is given in Table 1.

Low titers and viability were obtained for the first experiments with molecule A. Overall, molecule A and molecule B CCFs represent two extremes of CCFs for clarification experiments in terms of PCV, viability, turbidity and titer.

2) Cell culture monitoring

The cell culture monitoring was performed off-line using specific devices.

Viability, Viable Cells Concentration (VCC) and average cell diameter were daily determined using Vi- Cell™XR (Beckman Coulter Ref 38353, USA) cell counter. A sample of 500 pL (diluted if needed) from the cell culture was pumped into the analyser and then automatically mixed with trypan blue to evaluate the amount of dead and viable cells.

Metabolite concentrations (glucose, lactate, glutamine, glutamate, ammoniac, lactate dehydrogenase and IgG) were monitored thanks to the CEDEX Bio HT device from Roche which is a high-throughput automated metabolite analyser based on enzymatic assays coupled with spectrophotometry. A 500 pL centrifuged sample was analysed each day of the cell culture process. pH of the cell culture was daily measured off-line for bioreactor culture or at day 0, day 7 and harvest day for Thomson culture using ABL80 Flex analyser from Radiometer. pCO2 and pO2 were also measured using the ABL80.

When cells were grown in bioreactor, osmolality was closely monitored using Micro-Osmometer OsmoPro Multi-Sample from Advanced instruments. A 20 pL centrifuged sample was analysed each day of the cell culture process. Clarification process

1 ) Process set-up and flow

A reference clarification process is composed of 5 steps: a water for injection (WFI) flush, a 1 X phosphate buffer saline (PBS) 140 mM flush, a dead volume (DV) discard step, the filtration of the CCF and a 1 X PBS 140 mM recovery flush. The aim of the WFI flush is to remove all leachables that could have remained in the filters from their manufacturing. The PBS flush aimed to equilibrate the filter media to a suitable pH. The dead volume discard corresponds to the discard of the filters void volume at the beginning of the filtration step to limit the dilution of the product. Finally, the recovery flush goal is to recover a maximum of product that is still remaining in the filters. At least one DV of PBS needs to be flushed to recover the harvest volume remaining in the filters and tubing. Another DV of PBS is then flushed to potentially unbind the remaining product bind in the filter media. The reference process steps, and flow details for bench scale clarification are shown in Table 2.

Table 2. Reference process steps and flow details at bench scale. Bench scale clarification was performed with a primary and a secondary depth filters made of high capacity synthetic media, respectively 270 cm² and 140 cm² each. The primary clarification depth filter had an exclusion range of 0.55-8 pm; the secondary clarification depth had an exclusion range of «0.1 pm.

The clarification runs were conducted based on the reference process limits, meaning a maximum of pressure of respectively 3 bars and 2.4 bars at small and bench scale (filter limit - supplier recommendation) and a turbidity of 20 NTU post clarification for both scales.

2) Online measurements

For each scale, a balance was placed under the collected clarified harvest to determine the exact amount of material that went through the filtration system to ensure the accuracy of the yield calculation.

The classical filtration train used is composed of either a tri-headed FilterTech pump at small scale or peristaltic MasterFlex pumps at bench scale connected to Scilog pressure sensors. The tri-headed pump allows to run three independent CCF clarifications in parallel at the same flow with three different conditions.

The clarification process was thus monitored with online pressure and weight measurements over the time.

3 ) Off tin e m easurem en ts

The packed cell volume (PCV) as well as the turbidity of the initial cell broth to be clarified were determined for each experiment.

The PCV corresponds to the proportion of solids (cells, debris) present in the CCF. It is measured by transferring 5 mL of CCF in a 15 mL falcon which is then centrifuged for 5 min at 3000 g. The supernatant is then removed, and the falcon is weighted again to determine the mass of the cell pellet. The PCV is then the calculated with the following equation:

Cell pellet weiqht

PCV = - x 100

Initial weight

The turbidity is measured thanks to a turbidimeter from Hach which measured the diffused light at 90° at a wavelength of 860 nm in NTU.

The turbidity indicates the amount of material and small particles inside a liquid capable of light diffusion and is directly correlated with the haziness of a solution. Samples are also taken during the filtration to measure the titer by CEDEX, as well as the glucose and the LDH concentrations. The LDH rate is directly correlated to the cell lysis rate as LDH is an intracellular enzyme which will be detected in the supernatant only when cell lysis increase. Glucose yield was always compared to product yield to ensure no loss of product by dilution effect. Glucose yield was also compared to product yield to detect if any binding of the product was happening. Indeed, glucose is not expected to bind the filters. Therefore, almost 100 % glucose is expected to be recovered.

Example 2: CCF characterization

1) VCC and viability

Cell culture and viability profiles (VCC) of Molecules A and B used for clarification experiments are depicted in Figure 1.

Overall, molecule B CCFs reached higher viabilities on the harvest day than molecule A CCFs. Indeed, on average, 54 % viability was obtained with molecule A cell culture while 71 % viability was reached with molecule B cell culture.

The noteworthy viability drop on day 10 in molecule A cell culture was unexpected. Two of the hypotheses that could have explained this viability drop before day 14 were the high-volume cell culture and the Thomson vessel 5 L shape.

Indeed, the cell culture process was performed with 2 L initial cell culture reaching 3 L final cell culture volume. Supplier recommendation is to run fed-batches with 2 L final cell culture volume to maximize the gas exchange. Therefore, with 2 L initial culture and 3 L final culture, the gas exchange may have been limiting. This would explain the viability drop.

Further, the Thomson 5L shape could have contributed to minimizing the gas exchange. Due to the specific shape of the Thomson 5L, the gas exchange surface gets smaller with increasing height of the cell culture volume.

Corning 5 L shake flasks were proposed as a solution to assess if the hypothesis of Thomson shape could have explained the issue.

No significant difference in terms of viability was observed when the cell cultures were conducted in Corning (Figure 1, Experiment 3) instead of Thomson (Figure 1, Experiment 2) excluding the Thomson shape hypothesis.

In a following experiment, the final cell culture volume was decreased from 3 L to 2 L to assess the impact of the final cell culture on the viability. When the final cell culture volume decreased to 2 L (Figure 1, Experiment 5, 6 and 7), the cells stayed alive slightly longer than the ones with a final cell culture volume at 3 L (Figure 1, Experiment 1 to 4). Therefore, the high cell culture volume is the most likely cause explaining an early viability drop probably due to limited gas exchange.

Regarding the VCC profile, the cell density peak is quite similar for both cell lines with around 15 - 20 x 10^scells/mL at day 8.

2) Titer

Figure 2 presents the molecules A and B concentration over the fed-batch duration.

The fed-batches producing molecule B were also more successful in terms of production compared to cell culture producing molecule A. Indeed, on average fed-batches producing molecule B reached 9 times more titer than fed-batches producing molecule A (0.3 g/L of molecule A and 2.8 g/L of molecule B).

Unexpected low titers of molecule A were reached in the first cell culture process (Figure 2, Experiment 1 to 4). Nearly, 2 times more molecule A was obtained when the final volume cell culture were decreased to 2 L (Figure 2, Experiment 5 to 7) and 3 times more molecule A were obtained when the cell culture was performed in bioreactor (Figure 2, Experiment 8 and 12).

3) T urbidity an d PCV

The initial turbidity and PCV (Packed cell volume, i.e., proportion of solids in the solution - cells and debris) of molecules A and B CCF are depicted in Figure 3.

In terms of turbidity as well as in PCV, molecule B CCFs were again more challenging than molecule A CCFs. In average, 2 times more solids are present in molecule B CCFs and 1.7 more turbidity is reached in molecule B CCFs.

To sum up, both molecule B and molecule A CCFs were challenging for clarification. Indeed, the high turbidity of molecule B CCFs challenged the filters with more impurities and HCPs while the low viability of molecule A CCFs challenged the filters with more cell debris.

4) Offline measurements

The packed cell volume as well as the turbidity of the initial cell broth to be clarified were determined for each experiment.

The PCV (see equation in Experiment 1, paragraph "3) Offline measurements") corresponds to the proportion of solids (cells, debris) present in the CCF. It is measured by transferring 5 mL of CCF in a 15 mL falcon which is then centrifuged for 5 min at 3000 g. The supernatant is then removed, and the falcon is weighted again to determine the mass of the cell pellet. The turbidity is measured by a turbidimeter from Hach which measured the diffused light at 90° at a wavelength of 860 nm in NTU.

Example 3: Clarified cell culture fluids (CCCFs) characterization

A screening on molecule A CCFs was performed at small scale and confirmation runs were conducted at bench scale. Figure 4 illustrates the reference clarification process CCCFs results for molecule A CCF clarification at small and bench scale. On average, small and bench scale clarification lead to similar product yield around 76 %. However, a higher standard deviation is observed at small scale than at bench scale. This variability at small scale can be explained by the low number of small-scale clarifications. Indeed, only 4 small-scale clarifications were performed compared to 9 bench-scale clarifications. Therefore, small-scale clarifications can be considered representative to bench-scale clarifications and CCCFs reference results from both scales can be pooled.

A summary of the clarification reference process results obtained with both molecules CCFs is depicted in Figure 5. Overall, molecule B CCF clarification led to 10 % more product recovery than molecule A CCF clarifications with the same clarification process. Indeed, on average, 76.1 % product is recovered when molecule A CCFs are clarified while 85.7 % product is recovered in molecule B CCCFs. These CCFs reference process results are used as reference average product yields to compare the optimized parameters. Two main hypotheses can explain this difference in product recovery.

First, the initial difference between molecule A and molecule B CCFs could lead to significant differences in the composition of the CCCFs. Indeed, higher product yields obtained with molecule B CCCFs compared to molecule A CCCFs could be linked to saturation of binding sites on the membranes. If their quantity is constant, it should be expected that with high titer concentration CCFs, like molecule B CCFs, the saturation of the filters could be negligible compared to molecule A where low titer concentration CCFs are clarified.

In addition, molecule B CCFs contained much more impurities than molecule A CCFs.

Therefore, it can be expected that when a higher amount of impurities is present in the CCF, there could be a competition for potential binding sites on the membranes, leading to less binding of the product and a better recovery.

Further, the structure difference between molecule (bispecific antibody) and B (IgG-like antibody) could be responsible for the differences observed in clarification performances. Example 4: Post-clarification yield titers was enhanced by arginine-HCI recovery flush

In order to enhance the yield titers post-clarification and therefore to minimize the product loss, the impact of different process parameters was assessed, including among others, the flow rate, the use of different recovery and equilibration buffers, including PBS at various concentrations and CaCI2, and the use of arginine-HCI recovery flush.

1) High flow impact

The first parameter to assess was the clarification flow rate. A higher flow rate was expected to decrease the time of interaction between the product and filter potentially limiting the unwanted binding phenomenon and therefore less product loss.

A two times higher flow rate was applied from the DV discard to the recovery flush step (see Table 2).

A summary of the product yields obtained between the reference process and higher flow conditions is given in Figure 6.

On a first run (Figure 6, experiment 2), a two times higher flow rate was applied on molecule A CCF at bench scale. The product yield between the reference process and the high flow condition was similar.

A second experiment (Figure 6, experiment 4) was set-up to confirm the previous trends. However, this experiment was done with a high flow and another recovery flush buffer (i.e., Arginine-HCI). Again, in this experiment, the results between standard flow and high flow conditions were similar.

An additional run (Figure 6, experiment 8) was performed on molecule A CCF with a higher flow and a higher initial titer. As no reference conditions was run in parallel with the same initial material, the reference process average yield is used for comparison in Figure 6.

The product yield obtained through this high-flow clarification (69.4 %) was lower than the reference process average product yield (76.1 %).

The glucose yield was used to assess whether there was an error in dilution (as glucose is not supposed to bind to the membrane and should be fully recovered). Therefore, glucose yields compared to the product yields obtained with all these reference experiments were plotted to determine if a correlation was observed.

No correlation was observed between the glucose and the product yield for the clarification of molecule A material, meaning that there was no unexpected loss of culture liquid.

On a last high flow experiment (Figure 6, experiment 11), molecule B CCF was clarified to determine if any difference would be observed between molecule A CCFs and molecule B CCFs clarifications at higher flow. Product yield obtained to assess the influence of a 2 times higher flow on the product recovery is given in Figure 6 (experiment 11). No significant difference was observed between the high flow and the reference condition at the end of the clarification (respectively 91.6 and 91.1 %) meaning that the high flow did not impact the product recovery on molecule B.

To summarize, the experiments above lead to the conclusion that a high flow is not impacting the product recovery. Therefore, a 2 times higher flow showed no positive impact on the product recovery on molecule A and molecule B.

Following up this experiment, product quality on a high flow experiment was analysed to ensure that the high flow does not impact the molecules. No significant difference was observed in terms of product quality between the reference process and the high flow conditions (data not shown).

Figure 7 presents the high flow impact on the product yield. No significant difference was observed between the reference process and the high flow conditions meaning that the high flow did not impact the product recovery in terms of product yield. The high flow did also not impact the DSP process and the product quality (data not shown).

2) Screening of equilibration and recovery buffers

The aim of this experiment was to screen different equilibration and recovery buffers to either prevent the binding of the product on the filters or unbind the maximum of product potentially bound to the membrane filters.

First, equilibration buffers were tested to ensure that the filters were correctly equilibrated and that all the charges of the membranes were saturated limiting the binding of our product in the membranes.

Secondly, recovery buffers were tested to unbind the maximum of product remaining in the filters.

Different levels of PBS salt concentration were tested as equilibration and recovery buffers as well as calcium chloride as a recovery buffer.

With higher PBS salt concentration or calcium chloride, higher product recovery was expected either because less charged site where the product can bind are available in the membrane or because interactions between the filter and the product were weakened.

An overview of the small-scale clarification runs is given Figure 8.

Screening of equilibration buffers

Three small scale clarifications were performed on the same CCF (see experiment 5, Table 1) to evaluate the impact of different salt concentration of PBS at equilibration step. The aim of these experiments was to assess if the equilibration of the membrane with higher salt concentration PBS enhances the product recovery as the membrane filters would be more saturated and less charged sites would be available for the antibody. Therefore, higher product yields were expected when the filters were equilibrated with higher salt concentration PBS.

Figure 9 depicts the yields obtained with the different salt concentration of PBS equilibration.

No difference was observed in terms of product yield between each condition, as the product yield when 350 mM PBS (79.8 %) and 700 mM PBS (83 %) were flushed remained close to the 82.1 % yield obtained with the reference experiment.

An increase of nearly 40 % in the LDH yield is observed between the reference process and the PBS 700 mM conditions. This showed that the higher concentration is also impacting the cell lysis in the filters.

Therefore, the salt concentration of the equilibration buffer does not have an impact on the product recovery but negatively impacted the state of the cells in the clarification filters (more lysis).

Screening of recovery buffers

Three small scale clarifications were performed on the same CCF (see experiment 6, Table 1) to evaluate the impact of different salt concentration PBS at recovery step.

The control condition (reference process - PBS recovery 140 mM) was taken from experiment 5 (screening of equilibration buffers) runs as experiments 5 and 6 CCFs were considered similar.

The aim of these experiments was to assess if the recovery flush with higher salt concentration PBS enhances the product recovery as the higher ionic strength should easily weaken the charged interactions between the filters and the product. Therefore, higher product yields were expected when the filters were flushed with higher salt concentration PBS.

Figure 10 illustrates the yields obtained with the different salt concentration of PBS recovery.

No significant difference was observed between the reference condition at 82 % product yield and the three experiments where the PBS salt concentration of the recovery flush step increased. Product yields of respectively 82.0 %, 74.3 % and 85.7 % were obtained from PBS 350 mM to PBS 1400 mM.

Therefore, surprisingly, it seems that the salt concentration of the PBS recovery flush buffer (similar than an elution buffer) did not have a positive impact on the product recovery.

Calcium chloride (CaCL) was also tested as a recovery buffer, as described in Figure 8. CaCL is a chaotropic salt well known to decrease the hydrophobic effect in solution thus weakening the interaction and causing dissociation. Therefore, the aim of this experiment was to assess if flushing with CaCI₂ enhances the product recovery.

Yields obtained when calcium chloride at 250 mM was flushed are presented in Figure 11.

Regarding the impact on the calcium chloride flush, a slight increase in terms of product yield was observed when the recovery flush was performed with CaCI₂ instead of PBS 140 mM (respectively 92.6% and 88.7%). However, this difference was not significant.

For the control condition, 10 % more than the average reference process product yield (76.1 %) was obtained (PBS recovery 140 mM - experience 7).

One of the hypotheses that could explain this 7 % difference in the final product yield for the same experiment could be the initial CCFs which were different. The CCF used for the experiment 7 had 1000 NTU more (2520 NTU) than the one used for experiment 5 (1532 NTU) and 3 % more PCV meaning that much more impurities (HCP and other molecules) were in the CCF of this experiment 7. It should be expected that when higher impurities are present in the CCF, they also could come in contact with the filter membranes and a competition phenomenon between the antibody and the impurities could appear leading to less binding of our product and a better recovery.

A noteworthy precipitation phenomenon had been observed when the recovery flush was conducted with CaCI₂.

As IgG was still detected in large quantity by the CEDEX in the last sample from CaCI₂ condition, IgG did not precipitate. Given the large amount of the precipitate, calcium which is well known to precipitate in presence of phosphate could be part of the precipitate as phosphates are found in a large quantity in feed A, Power CHO and PBS. Feed B which is also well known because of its ability to precipitate could have led to similar situation. However, low amount of feed B are added during the culture process. Therefore, feed B might not be the cause of this precipitation.

No difference was observed when flushing with calcium chloride instead of PBS meaning that CaCI2 flush does not have an impact on the product recovery.

In summary, at the equilibration stage as well as at the recovery flush step, the salt concentration of the PBS used did not have an impact on the product recovery. Flushing with calcium chloride instead of PBS did not have a significant impact on the product recovery neither.

3) Arginine-HCI recovery assessment

The aim of the recovery buffer is to remove the volume of CCF remaining in the filters at the end (filling the void volume of the filters), to ensure a maximum product recovery. PBS is commonly used to flush 1 after clarification to enhance the product recovery. However, PBS flush was not efficient for recovery of molecule A mainly and molecule B in a less critical manner, as shown above. Other recovery flush, such as WFI were tested, however the use of WFI to make hydrophobic interactions weaker, did not help to recover more product. Surprisingly we found herein that the use of Arginine-HCI buffer as recovery buffer assures maximum product recovery.

Arginine-HCI buffer 800 mM at pH=6 was used for the following experiments.

On a first run of experiments (Figure 12, experiment 1 and 2, trials 1), the arginine-HCI recovery flush was tested on molecule A CCFs by addition of a final flush step with 2 dead volumes (DVs) of arginine- HCI flush after the current clarification process. This meant that a total of 4 DVs were flushed for these experiments instead of 2 DVs usually flushed in the reference clarification process.

Figure 12 shows the yields obtained for each experiment for the reference process (after 2 DV of PBS) and after the 2 added DV of arginine-HCI flush.

For each experiment, Arginine-HCI flush allowed to recover between 15 to 20 % more product leading to product yields of around 90 %. Arginine-HCI surprisingly enhanced the yield titer, unbinding the product remaining in the filters. Additionally, a second experiment, where PBS flush was directly replaced by arginine-HCI flush, was performed. For this purpose, the PBS flush is directly replaced by an arginine-HCI flush for the trials 2 (both experiment 1 and 2).

Results from an additional experiment to assess the impact of the number of DV recovery flushes and the arginine-HCI flushes are presented in Figure 13.

Two main conclusions can be drawn for these runs.

The first one is that after 2 DVs of arginine-HCI flush, on average 15 % (85.4 - 69.9) more product is recovered than when 2 DVs of PBS are flushed showing that arginine-HCI was responsible for unbinding more product.

In addition, the 2 added DVs of PBS flush after the 2 DVs of arginine-HCI flush did not recover more product (82.7% vs. 85.4% after 2 DVs of Arginine). This observation means that all the product remaining in the filters was already collected in the filtrate thanks to arginine-HCI action.

In fact, according to the product yield, less product was recovered after 4 DV of flush than after 2 DV of flush. This result was unexcepted. The difference might come from the analytical device (CEDEX) used for product quantification and the important standard deviation observed for the samples after 2 DV of arginine-HCI flush. Therefore, the experimental data has confirmed that no more product was recovered when increasing the number of flushes from 2 DV to 4 DV. Another observation was that 4 DVs of PBS flush allowed to recover the same amount of product that was recovered with only 2 DVs of arginine-HCI flush. This means that PBS is certainly less efficient in weakening the interaction between the filters and the product than arginine-HCI.

In summary, arginine-HCI flush showed great promise by increasing the product yield by 15 % on average only after a 2 DV flush while a 4 DV flush was needed to reach the same product yield with PBS.

Following these two experiments, a third experiment (Figure 12, experiment 8) on molecule A CCF was performed with an arginine-HCI flush. The product yield obtained after 2 DVs of PBS flush (69.4 %) and after 2 DVs of arginine-HCI (79.9 %) confirmed again the positive impact of arginine-HCI flush on the product recovery.

Finally, arginine-HCI recovery flush was tested on molecule B CCFs to assess if the previous results mentioned earlier could be confirmed.

Product yields obtained with arginine-HCI flush on molecule B CCFs are given in Figure 14.

For experiment 1, 5 % more product was recovered with the arginine-HCI flush. The lower percentage of recovery obtained for molecule B compared to molecule A can be explained by the fact that saturation becomes negligible at higher titer, as well as by the difference between molecule A and B CCFs and between molecule A and B structure.

For experiment 2, less than 4 % more product was recovered following the arginine-HCI flush. However, this experiment was combined with the high-flow condition. One hypothesis in addition to the previous ones above that could explain this low recovery would be that the higher flow rate did not allow time for the arginine-HCI to weaken the interactions leading to less product recovery than it has been previously observed.

The product quality on the conditions with an arginine-HCI flush was analysed to ensure that arginine- HCI buffer was not having an undesired impact. No significant differences were observed in terms of product quality between the reference process and the arginine-HCI conditions (data not shown).

A summary of the impact of arginine-HCI on molecule A and molecule B CCFs is depicted in Figure 15.

In conclusion Arginine-HCI flush leaded to the highest product recovery post-clarification.

Correlations between the percentage of product recovered with arginine-HCI and the initial CCF parameters have also been plotted. The aim was to explain why the recovery worked better for molecule A than for molecule B. Overall, no clear correlation has been found between the initial CCF state and the percentage of product recovered with Arginine-HCI flush. However, a trend was observed for the initial CCF titer (Figure 16, black dots are molecule A, grey dots are molecule B). Indeed, the lower the initial CCF titer is, the higher the product recovery thanks to arginine-HCI is (Rsquare = 86 %).

Molecule B cell culture process was optimized when compared to historical data and the product titer was increased by a factor of 10. When flushing with arginine-HCI using the high titer broth, an increase of 4% in recovery was observed (Figure 14). In comparison, when arginine-HCI was tested on the low molecule B titer broth before process optimization (Figure 16), recovery increased by 14%.

In conclusion Arginine-HCI flush leaded to the highest product recovery post-clarification, especially in the case of low initial titers projects.

Claims

1. A method for recovering a biomolecule of interest from a cell culture clarification synthetic depth filter comprising the step of flushing said depth filter with Arginine-HCI buffer.

2. The method of claim 1 wherein the biomolecule of interest is an antibody.

3. The method of claims 2, wherein the biomolecule of interest is a non-naturally occurring antibody.

4. The method of any one of the preceding claims, wherein Arginine-HCI buffer is used at a concentration comprised between about 600 and about 100 mM at a pH comprised between about 5 and about 7.

5. The method of claim 1 wherein Arginine-HCI buffer is used at 800 mM at pH=6.

6. A process for clarifying a cell culture including a biomolecule of interest comprising at least a filtration step performed by a synthetic depth filter, characterized in that said process further comprises recovering said biomolecule of interest from said depth filter according to the method of claims 1 to 5.

7. The process of claim 6 comprising the consequent steps of: i. flushing said depth filter with water for injection; ii. flushing said depth filter with 1 X phosphate buffer saline 140 mM; ill. connecting the cell culture to the filter; iv. flushing said depth filter with the cell culture fluid; v. discarding the dead volume; vi. collecting the clarified cell culture fluid; vii. recovering said biomolecule of interest from said depth filter by flushing said depth filter with Arginine-HCI buffer according to the method of claims 1 to 5.

8. The process of claims 6 or 7 further comprising a step of subjecting said recovered biomolecule of interest to one or more steps of chromatography purification.

9. A process of production of a drug substance comprising the steps of: i. seeding cells expressing a biomolecule of interest in a cell culture medium; ii. culturing said cells for a period comprised between 10 and 18 days, preferably for 14 days; ill. subjecting the obtained cell culture to the clarification process according to claims 6 to 7; iv. add excipients to the biomolecule of interest purified according to claim 8.