Hydrophobic Interaction Chromatography for Viral Elimination

Abstract

This application provides a method for characterizing and / or determining the viral removal capacity of hydrophobic interaction chromatography (HIC), including an experimental design for multivariate analysis of viral removal by HIC. The method provides an understanding of the mechanism of viral removal using HIC by executing a D-Optimal experimental design that includes evaluations of multiple factors, such as pH, buffer concentration, column loading concentration, column flow rate, and hydrophobic strength of the HIC column.

Description

Hydrophobic Interaction Chromatography for Viral Elimination Field of invention The present invention relates in general to methods for characterizing the viral removal capacity of hydrophobic interaction chromatography, including experimental designs for multivariate analysis. Background of the invention Biological products can be contaminated with unwanted viruses, posing a risk of viral disease transmission. Global health authorities require viral clearance assessments for the manufacture of biological or biotechnological products, as viral load can multiply during mammalian cell culture growth. Effective viral clearance studies are an essential part of process validation, which is critical for ensuring drug safety. Viral contamination can affect raw materials, cell culture processes, the bioreactor, and downstream purification processes. Viral validation studies are designed to provide evidence that selected operating conditions will effectively inactivate and / or eliminate viruses. The experimental design of viral elimination studies includes characterizing the manufacturing process to ensure its ability to eliminate viruses and to improve understanding of the processing conditions. When evaluating the removal of viral contaminants, selecting the worst-case conditions for assessment is warranted. Virus inactivation or removal processes include pH treatment, heat treatment, filtration, or chromatography. For the purification of biological products, chromatographic steps with the potential to provide viral reduction for viral removal can be used, such as protein A chromatography, anion-exchange chromatography, or hydrophobic interaction chromatography (HIC). When a chromatographic step is used to capture a monoclonal antibody, the virus could interact with the antibody and / or the chromatography resin. For antibody purification using HIC, there is limited knowledge regarding viral removal related to the continuous flow mode of HIC, for example, the selective binding of unwanted components as the antibody appears in the continuous flow. It will be appreciated that there is a need for methods to effectively characterize the viral clearance capacity of manufacturing processes to ensure drug safety, including the creation of a retrospective viral clearance database to explain the mechanisms and justify the selection of the worst conditions, such as improving the understanding of viral clearance capacity by HIC. bQ t onn / zznz / E / YiAi Brief description of the invention This disclosure provides methods for determining the impact of factors on the development of viral clearance capacity by HIC, including experimental design for multivariate analysis of viral clearance by HIC. This disclosure further provides an understanding of the viral clearance mechanism and the worst-case processing conditions for viral clearance to improve drug safety. In addition, this disclosure provides methods for constructing a retrospective database of viral clearance by HIC to explain the mechanisms and justify the selection of worst-case conditions. To enable the use of HIC for viral clearance, this disclosure provides the characterization of HIC related to the clearance of a model retrovirus to gain an understanding of the process. This disclosure further provides a method for purifying an antibody from a sample comprising one or more impurities including viral particles, the method comprising the steps of: (a) providing the sample comprising the antibody produced in a host cell, (b) adjusting a pH of the sample to a range of approximately 4.2 to approximately 8.0, (c) loading the sample onto a hydrophobic interaction chromatography (HIC) column wherein a concentration of the sample is from approximately 40 g / L to approximately 200 g / L, and (d) collecting the HIC-treated sample. In some example embodiments, a citrate buffer solution is used to adjust the pH of the method sample, where the concentration of the citrate buffer solution is approximately 10 mM to approximately 200 mM. In some aspects, the HIC column resin of the method is a phenyl or phenylcapto resin. In some aspects, the hydrophobic strength of the HIC column of the method ranges from weak to strong hydrophobic, where weak hydrophobic strength is achieved by using a phenyl resin or its equivalent, and strong hydrophobic strength is achieved by using a phenylcapto resin or its equivalent. In some respects, the antibody used in the method is a monoclonal antibody or a bispecific antibody, where the antibody has an IgG1 isotype or an IgG4 isotype. In some respects, the flow rate through the HIC column of the method has a linear velocity of approximately 100 cm / h or approximately 300 cm / h. In some respects, the method in this application also includes measuring the presence of viral genomic copies and / or viral particles. This disclosure, at least in part, provides a method for purifying an antibody from a sample comprising one or more impurities including viral particles. The method comprises the steps of: (a) providing the sample comprising the antibody produced in a host cell, (b) adjusting the pH of the sample to a range of approximately 4.2 to approximately 8.0, (c) loading the sample onto a hydrophobic interaction chromatography (HIC) column, wherein the concentration of the sample is approximately 40 g / L to approximately 200 g / L, (d) collecting the HIC-treated sample, and (e) measuring the presence of viral genomic copies and / or infectious viral particles in the HIC-treated sample from step (d). In some example embodiments, the method of this application further comprises the optimization of the removal of viral genomic copies and / or viral particles by performing a D-Optimal experimental design. In some aspects, the D-Optimal experimental design of this application evaluates the following factors: (a) the pH of the sample, from approximately 4.2 to approximately 8.(b) the column loading, wherein the sample concentration is from approximately 40 g / L to approximately 200 g / L, (c) a linear velocity of a flow rate through the HIC column from approximately 100 cm / h to approximately 300 cm / h, and (d) a hydrophobic strength of the HIC column from weak to strong hydrophobic; wherein the weak hydrophobic strength is achieved by the use of a phenyl resin or an equivalent thereof, and the strong hydrophobic strength is achieved by the use of a phenyl capto resin or an equivalent thereof. In some respects, the D-Optimal experimental design of the present application further evaluates an antibody isotype, wherein the antibody is either a monoclonal antibody or a bispecific antibody. This disclosure, at least in part, provides a method for purifying an antibody from a sample comprising one or more impurities including viral particles. The method comprises the steps of: (a) providing the sample comprising the antibody produced in a host cell, (b) adding citrate buffer solution to the sample, (c) adjusting the pH of the sample to a range of approximately 4.2 to approximately 8.0, (d) loading the sample onto a hydrophobic interaction chromatography (HIC) column, wherein the concentration of the sample is approximately 40 g / L to approximately 200 g / L, (e) collecting the HIC-treated sample, and (f) measuring the presence of viral genomic copies and / or viral particles in the HIC-treated sample from step (e). In some example modalities, the method of the present application further comprises optimizing the removal of viral genomic copies and / or viral particles by performing a D-Optimal experimental design, wherein the D-Optimal experimental design evaluates the following factors: (a) the pH of the sample from approximately 4.2 to approximately 8.(a) a citrate buffer concentration of approximately 10 mM to approximately 200 mM, (c) the column loading, wherein the sample concentration is approximately 40 g / L to approximately 200 g / L, (d) a linear velocity of approximately 100 cm / h to approximately 300 cm / h through the HIC column, and (e) a hydrophobic strength of the HIC column from weak to strong, wherein the weak hydrophobic strength is achieved by using a phenyl resin or its equivalent, and the strong hydrophobic strength is achieved by using a phenylcapto resin or its equivalent. In some respects, the method antibody is a monoclonal antibody or a bispecific antibody. These and other aspects of the invention will be better appreciated and understood when considered in conjunction with the following description and accompanying drawings. The following description, while indicating several embodiments and numerous specific details thereof, is provided for illustrative purposes and not for limitation. Many substitutions, modifications, additions, or rearrangements are possible within the scope of the invention. bQ / onn / zznz / E / YiAi Brief description of the drawings Figure 1 shows an overview of an example process for quantifying viral clearance using virus assays according to an example modality. Figure 2 shows the calculation of the total logarithmic reduction factor (LRF) for quantifying viral clearance according to an example modality. Figure 3 shows the scatter plot of the D-optimal designs that were used to elucidate the main effects, interactions, and quadratics to evaluate wide ranges of development factors, including pH, citrate, column loading, linear speed, or HIC resin, according to an example modality. Figure 4 shows the quantification of viruses for backup virus control, HIC load retention control, and HIC product backup according to an example modality. Figure 5 shows a predicted graph that was generated based on the actual reported LRF and the predicted reported LRF to assess X-MuLV viral clearance according to an example modality. Figure 6 shows the one-way analysis of residual X-MuLV infectivity by program using a reported elimination model applied to a retrospective dataset to predict the behavior of specific monoclonal antibodies according to an example modality. Twenty-three monoclonal antibodies were evaluated using a one-way analysis of residual X-MuLV infectivity by program according to an example modality. Figure 7 shows the assessment of X-MuLV viral clearance by generating a predicted graph based on actual and predicted qPCR LRF according to an example modality. A qPCR LRF model was used to identify multiple significant developmental factors according to an example modality. Figure 8 shows the one-way analysis of residual qPCR by program using a qPCR model to predict the behavior of specific monoclonal antibodies according to an example modality. Twenty-one monoclonal antibodies were evaluated by one-way analysis of residual GenReg qPCR by program according to an example modality. Figure 9A shows a graph and prediction profile generated by applying DoE to optimize development factors for both performance and virus LRF, by applying the reported LRF models to HIC viral elimination processes according to an example modality. Figure 9B shows a prediction profiler applied to lgG4 by applying DoE to provide instructions for optimizing development factors for both performance and LRF of the virus. Figure 10 shows comparisons of the relative molecular hydrophobicity of various HIC resins using monoclonal antibodies (mAb) according to an example modality. bQ / onn / zznz / E / YiAi Figure 11 shows a one-way analysis of X-MuLV LRF infectivity with or without DoE data by applying DoE studies for viral clearance by HIC with various ranges of development factors according to an example modality. Figure 12 shows the analysis of data obtained from DoE studies using bivariate adjustment of X-MuLV LRF by qPCR X-MuLV LRF to support the relationship between LRF by infectivity and qPCR assays according to an example modality. Figure 13 shows the analysis of the data obtained from the DoE studies in relation to the interaction profile of the LRF HIC DoE model reported according to an example modality. Figure 14A shows an original model of a retrospective database for infectivity assays prior to the application of HIC viral clearance DoE data according to an example modality. Figure 14B shows an improved model for applying HIC viral clearance DoE data to a retrospective database for infectivity assays according to an example modality. Figure 15A shows an original model before applying HIC viral clearance DoE data to a retrospective database for a qPCR model according to an example modality. Figure 15B shows a modified model by applying HIC viral clearance DoE data to a retrospective database for a qPCR model according to an example modality. Figure 16 shows a predicted graph and scaled estimates by applying HIC viral clearance DoE data to a retrospective database for mAb11 IgG1 isotype according to an example modality. Figure 17 shows a predicted graph and scaled estimates by applying HIC viral clearance DoE data to a retrospective database for the mAb14 lgG4 isotype according to an example modality. Figure 18A shows a predicted graph by applying HIC viral clearance DoE data to a retrospective database for phenyl capto HS according to an example modality. Figure 18B shows parameter estimates and a predicted profiler by applying HIC viral clearance DoE data to a retrospective database for phenyl capto HS according to an example modality. Figure 19A shows a predicted graph by applying HIC viral clearance DoE data to a retrospective database for phenylsepharose 6 FF HS according to an example modality. fq / οηη / ζζηζ / Ε / γίΛΐ Figure 19B shows parameter estimates and a predicted profiler by applying HIC viral clearance DoE data to a retrospective database for phenylsepharose 6 FF HS according to an example modality. Figure 20 shows a forecast graph and prediction profile that optimizes performance by applying HIC's DoE performance model to monitor the trend between LRF and performance according to an example modality. Detailed description of the invention Given that viral contamination can multiply during mammalian cell culture growth, viral clearance assessment for the manufacture of biological or biotechnological products is essential and critical to ensuring drug safety. Health authorities have provided guidelines equivalent to "Using good science to manage patient risk" for evaluating whether a step clears the virus, by understanding how clearance occurs, when steps operate independently, whether their capacity is additive or non-additive, and what affects performance. Viral clearance assessment should include demonstrating the clearance of a specific model virus for retrovirus-like particles inherent in the genome of Chinese hamster ovary (CHO) cells (Anderson et al.).Endogenous origin of detective retrovirus-like particles from a recombinant Chinese hamster ovary cell line, Virology 181 (1): 305-311, 1991). Xenotropic murine leukemia virus (X-MuLV) can be used as a model virus in the evaluation of viral inactivation in pharmaceutical proteins derived from CHO cells. Murine leukemia virus (MuLV) is a retrovirus and has a positive-sense single-stranded RNA that replicates by reverse transcription. MuLV can induce leukemia in inoculated mice. Ensuring viral removal is crucial when designing a purification process. A typical workflow for studying viral removal in a manufacturing process involves adding virus to the sample load, running the process in a reduced-scale experiment to mimic a full-scale step, and documenting the removal capacity of the added virus. Regulatory guidelines recommend using virus validation data to design in-process limits for determining critical process parameters, such as performing end-of-process validations. Testing can be conducted under worst-case conditions to demonstrate the minimum removal a process step can provide (1998, Q5A Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin. TIC f. H. o. TR f. P. f. H. Use).The worst conditions can be determined by factors that influence the viral clearance mechanism depending on the process used. The worst conditions can be evaluated to demonstrate the minimum viral reduction of a specific process step (Aranha et al., Viral clearance strategies for biopharmaceutical safety, part II: a multifaceted approach to process validation, BioPharm 14 (5), 43-54, 90, 2001). Viral validation studies can be designed to document selected operating conditions with respect to product quality and process specificity to ensure viral safety. Virus inactivation or removal processes include pH treatment, thermal treatment, solvent / detergent treatment, filtration, or chromatography. Low-pH incubation can be used to inactivate enveloped viruses, such as through irreversible capsid denaturation (Brorson et al., Bracketed generic inactivation of rodent retroviruses by low pH treatment for monoclonal antibodies and recombinant proteins, Biotechnol Bioeng 82(3): 321-329, 2003). Filtration is a size-based removal method that can be used to remove both enveloped and non-enveloped viruses (Lute et al.).Phage passage after extended processing in small-virus-retentive filters, Biotechnol Appl Biochem 47(Pt3): 141-151, 2007). Chromatography steps can be used to purify biological products with the potential to provide viral reduction for viral clearance, such as protein A (Bach et al., Clearance of the rodent retrovirus, XMuLV, by protein A chromatography, Biotechnol Bioeng 112(4): 743-750, 2015) or anion exchange chromatography (Strauss et al., Anion exchange chromatography provides a robust, predictable process to ensure viral safety of biotechnology products, Biotechnol Bioeng 102(1): 168-175, 2009a). Some chromatographic steps can contribute to viral removal, such as the use of anion-exchange chromatography or hydrophobic interaction chromatography (HIC) for logarithmic reductions on the order of 4 to 5 log (Brown et al., A step-wise approach to define binding mechanisms of surrogate viral particles to multi-modal anion exchange resin in a single solute system. Biotechnol. Bioeng., 114(7), pp. 1487-1494, 2017). Multimodal anion-exchange resins typically exhibit high and robust viral removal over a wide pH and conductivity range. Cation exchange and protein A affinity also contribute to viral reduction on the order of 2 to 3 log (Ruppach, Log Reduction Factors in Viral Clearance Studies, BioProcess. J., 12(4), 24-30 https: / / www.bioprocessingjournal.com / , online publication date January 7, 2014).Several critical variables in chromatographic processes can affect viral removal, including sample loading concentrations (such as antibody loading), contaminant concentrations, buffer solutions, pH, flow rates, wash volumes, and temperatures, depending on the resin and binding mode. Changing these conditions can provide indications of the viral reduction capacity of the process. The requirements for viral clearance assessment in the manufacture of biological or biotechnological products imposed by global health authorities have led to increased demand for characterizing the viral clearance capabilities of manufacturing processes. This disclosure provides an experimental design for multivariate analysis of viral clearance by HIC to meet this demand, which can contribute to understanding the viral clearance mechanism and improve drug safety. This disclosure provides methods for characterizing the viral clearance capabilities of HIC to evaluate and validate HIC by identifying the impacts of developmental factors, including experimental design for multivariate analysis. Experimental design, for example, design of experiments (DoE), for multivariate analysis includes fundamental characterizations of the HIC process by identifying significant developmental factors to improve the understanding of processing conditions for maximizing viral clearance. This disclosure provides the ability to build a retrospective data base of viral clearance by HIC to explain the mechanism and justify the selection of the worst-case conditions. Design of Experiments (DoE) is a methodology that allows for the systematic variation of multiple development factors within the context of an experimental design. The results of DoE can be used to create mathematical models of the process under examination. The true optimum of the process can be identified by applying these mathematical models. Applications of DoE results include eliminating inconsequential development factors, identifying critical development factors for further study, and predicting the performance of the process under examination. DoE is performed in a systematic, logical flow that includes stating objectives, selecting variable factors and models, creating experimental designs to support the models, collecting data based on the designs, running the analysis, verifying the models with control points, and reporting the results. The typical workflow for studying viral removal in a chromatography step includes adding virus to the sample load, running the chromatography step on a reduced-scale column, and documenting the removal capacity of the added virus. Viral removal studies are often performed using a reduced-scale model by employing a scaled-down chromatography column to mimic a full-scale step with the same bed height and flow rate. The Design of Experiments (DoE) can facilitate the determination of worst-case scenarios to identify factors influencing the viral removal mechanism based on the process used. These worst-case scenarios can then be evaluated to demonstrate the minimal viral reduction achieved by a specific process step. For antibody purification, there is limited knowledge of viral removal related to the negative mode (continuous flow mode) of hydrochloric acid (HCA), for example, the selective binding of unwanted components as the antibody appears in the continuous flow. One limitation of HCA is the need for high salt concentrations for protein binding, which can lead to protein aggregation. To enable the use of HCA for viral removal, this disclosure provides a characterization of HCA related to the removal of a model retrovirus to gain an understanding of the process. The impact of developmental factors on viral removal capacity by HCA can be determined to improve the understanding of the worst-case processing conditions for viral removal, thus enabling a path toward maximizing removal. Viral elimination studies are measurements of the ability of manufacturing process steps to inactivate or eliminate viruses. Model viruses can be added to a specific process, and experiments can then be performed to demonstrate the inactivation or elimination of the added viruses during subsequent processing steps. The viral loads of intermediate and relevant process samples can be determined for estimating reduction factors. The virus quantification methods described in this application include virus-specific cell-based infectivity assays and quantitative polymerase chain reaction (qPCR). Viruses comprise DNA or RNA encapsulated by a protein coat, with or without membranes. Viral reduction refers to the difference between the total amount of virus in the input sample and the output sample after performing a specific process step, such as the bQ / oni / zzi / E / glA chromatography process. Viral reduction capacity can be defined as the logarithmic reduction value (LRV) or the logarithmic reduction factor (LRF) of a process step. The reduction factor is calculated based on the total virus load before and after the removal step. Viral validation studies can be conducted to document the removal of known product-associated viruses and to estimate the process's effectiveness in removing potential adventitious viral contaminants by characterizing the process's ability to remove nonspecific model viruses. The evaluation and validation of a viral reduction process includes a critical process analysis to determine potential pathogenic sources of viral contamination or to characterize the process in order to identify manufacturing steps with the potential to achieve viral elimination. Each process step under examination can be evaluated by its viral elimination mechanism, such as inactivation, removal, or a combination thereof. It is advisable to select an effective and robust step that can eliminate viral contamination regardless of varying process parameters. (Aranha et al.) This disclosure also provides methods for determining the impact of growth factors on viral clearance by HIC by identifying multiple significant growth factors, including buffer pH, sodium citrate concentration in the buffer, sample loading, linear flow rate, hydrophobic strength of the HIC resin, and monoclonal antibody isotype. This disclosure also provides an understanding of the worst-case processing conditions for viral clearance. Growth factors in the overall worst-case scenario for HIC clearance include high pH, ​​low-to-medium citrate buffer concentration, high column loading, fast linear flow rate, an IgG4 monoclonal antibody isotype, and a phenylsepharose 6 FF HS (weak HIC resin).The DoE results and resulting models can be used to confirm, reject, or alter the existing understanding of the HIC mechanism for viral clearance. The example modalities disclosed herein satisfy the aforementioned demands by providing methods and systems for characterizing viral clearance capacity by HIC, including experimental design for multivariate analysis in order to identify the impact of developmental factors. In some example modalities, methods are provided for purifying an antibody from a sample comprising one or more impurities, including viral particles. The removal of viral particles and / or viral genomic copies is assessed using HIC for viral clearance. The article "a" should be understood as "at least one"; and the expressions "around" and "approximately" should be understood to allow for standard variation, as people of middle-level trade would understand it; and where ranges are provided, endpoints are included. bQ / οηη / ζζηζ / Ε / γίΛΐ As used herein, the expressions include, including and that includes are not intended to be limiting and mean to include, includes and that includes, respectively. In some example embodiments, this disclosure provides a method for purifying an antibody from a sample comprising one or more impurities including viral particles, the method comprising the steps of: (a) providing the sample comprising the antibody produced in a host cell, (b) adjusting a pH of the sample to a range of approximately 4.2 to approximately 8.0 and (c) loading the sample onto a hydrophobic interaction chromatography (HIC) column, wherein a concentration of the sample is from approximately 40 g / L to approximately 200 g / L, and (d) collecting the HIC-treated sample from step (c). As used herein, the term antibody refers to immunoglobulin molecules consisting of four polypeptide chains: two heavy chains (H) and two light chains (L) interconnected by disulfide bonds. Each heavy chain has a heavy chain variable region (HCVR or Vh) and a heavy chain constant region. The heavy chain constant region contains three domains: Ch1, Ch2, and Ch3. Each light chain has a light chain variable region and a light chain constant region. The light chain constant region consists of one domain (Cl). The Vh and Vl regions can be further subdivided into hypervariability regions, called complementarity-determining regions (CDRs), interspersed with more conserved regions called framework regions (FRs).Each Vh and Vl can be composed of three CDRs and four FRs, arranged from the amino terminus to the carboxy terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The term antibody includes reference to glycosylated and non-glycosylated immunoglobulins of any isotype or subclass. The term antibody includes, among others, those prepared, expressed, created, or isolated by recombinant means, such as antibodies or bispecific antibodies isolated from a host cell transfected to express the antibody. An IgG comprises a subset of antibodies. As used herein, the term impurity may include any undesirable protein present in the protein biopharmaceutical product. Impurities may include process-related and product-related impurities. Furthermore, impurities may be of known, partially characterized, or unidentified structure. Process-related impurities may originate from the manufacturing process and may include three main categories: those derived from the cell substrate, those derived from cell culture, and those derived from downstream steps. Cell substrate-derived impurities include, but are not limited to, proteins derived from the host organism and nucleic acid (genomic, vector, or total host cell DNA). Cell culture-derived impurities include, but are not limited to, inducers, antibiotics, serum, and other medium components.Impurities derived from subsequent steps include, but are not limited to, enzymes, chemical and biochemical processing reagents (e.g., cyanogen bromide, guanidine, oxidizing and reducing agents), inorganic salts (e.g., heavy metals, arsenic, non-metal ions), solvents, vehicles, ligands (e.g., monoclonal antibodies), and other leachates. Product-related impurities (e.g., precursors, certain degradation products) may be molecular variants that arise during manufacturing and / or storage and do not have properties comparable to those of the desired product with respect to activity, efficacy, and safety. Such variants may require considerable isolation and characterization effort to identify the type of modifications. Product-related impurities may include truncated forms, modified forms, and aggregates.Truncated forms are formed by hydrolytic enzymes or chemicals that catalyze the cleavage of peptide bonds. Modified forms include, but are not limited to, deamidated, isomerized, mismatched SS bond, oxidized, or conjugated forms with altered forms (e.g., glycosylation, phosphorylation). Modified forms may also include any form of post-translational modification. Aggregates include dimers and higher multiples of the desired product. (Q6B Specifications: Test Procedures and Acceptance Criteria for Biotechnological / Biological Products, ICH August 1999, U.S. Department of Health and Human Services). In some example embodiments, the method of the present application further comprises optimizing the removal of viral genomic copies and / or viral particles by running a D-Optimal experimental design, wherein the D-Optimal experimental design of the present application evaluates the following factors: (a) the pH of the sample from approximately 4.2 to approximately 8.0, (b) the column loading, wherein a sample concentration is from approximately 40 g / L to approximately 200 g / L, (c) a linear velocity of a flow rate through the HIC column from approximately 100 cm / h to approximately 300 cm / h, (d) a hydrophobic strength of the HIC column from a weak hydrophobic strength to a strong hydrophobic strength, and (e) an antibody isotype. As used herein, the term isotype refers to different isotypes of immunoglobulins. Immunoglobulins are heterodimeric proteins composed of two heavy chains and two light chains. Immunoglobulins have variable domains that bind to antigens and constant domains that specify effector functions. The Fe portion of the heavy chains defines the antibody class, of which there are five in mammals: IgG, IgA, IgM, IgD, and IgE. The classes differ in their biological properties, also known as effector functions, and in their functional localization to ensure an appropriate immune response to a given antigen. There are five main classes of heavy-chain constant domains. Each class defines the isotypes of IgM, IgG, IgA, IgD, and IgE. IgG can be further classified into four subclasses, for example, IgG1, IgG2, IgG3, and IgG4. IgA can be classified into IgA1 and IgA2.When the antibody is a human antibody, an isotype of the human antibody can be IgG1, IgG2, IgG3, IgG4, IgG1, IgG2, IgM, or IgE. When the antibody is a monkey antibody, an isotype of the monkey antibody can be IgG1, IgG2, IgG3, IgG4, IgM, or IgA. Example modalities The methods disclosed herein provide a means of purifying an antibody from a sample comprising one or more impurities, including viral particles. The method includes optimizing the removal of viral genomic copies and / or viral particles by executing a D-Optimal experimental design. In some example embodiments, this disclosure provides a method for purifying an antibody from a sample comprising one or more impurities including viral particles. The method comprises the steps of: (a) providing the sample comprising the antibody produced in a host cell, (b) adjusting the pH of the sample to a range of approximately 4.2 to approximately 8.0, (c) loading the sample onto an HIC column, wherein the concentration of the sample is from approximately 40 g / L to approximately 200 g / L, and (d) collecting the HIC-treated sample from step (c). In some respects, a resin of the HIC column of the method is a phenylsepharose 6 FF HS resin or a phenylcapto HS resin. In some respects, the hydrophobic strength of the HIC column of the method ranges from weak to strong, where weak hydrophobic strength is achieved by using a phenylsepharose 6 FF HS resin or its equivalent, and strong hydrophobic strength is achieved by using a phenylcapto HS resin or its equivalent. In some respects, the HIC resin comprises a hydrophobic group that is phenyl, phenylcapto, octyl, butyl, hexyl, or propyl. It is understood that the method or system is not limited to any of the hydrophobic interaction chromatography techniques mentioned above, nor to the processing conditions thereof. The sequential order of the steps of the method, as provided herein by numbers and / or letters, is not intended to limit the method or any of its embodiments to the particular order indicated. Throughout this specification, various publications are cited, including patents, patent applications, published patent applications, accession numbers, technical articles, and academic articles. Each of these cited references is incorporated herein by reference, in its entirety and for all purposes. Unless otherwise described, all technical and scientific terms used herein have the same meaning as they are commonly given by persons skilled in the art to whom the invention pertains. The disclosure will be more fully understood by reference to the following Examples, which are provided to describe the disclosure in greater detail. They are intended to illustrate and should not be construed as limiting the scope of the disclosure. Examples Methods for experimental design 1.1 Selection of development factors Various development factors (parameters) were selected for the design of experiments (DoE) of hydrochloric acid (HCA) to investigate their impact on HCA's viral clearance capacity and improve the understanding of processing conditions to maximize viral clearance. Development factors with theoretical impacts on HCA viral clearance were selected. Factors routinely studied in phase designs relevant to low-level risk assessment (LLRA) were also selected. Factors such as monoclonal antibody isotypes and the hydrophobic strength of the HCA resin were evaluated to enhance the understanding of the viral clearance mechanism for communication with regulatory agencies.The selected development factors and their retrovirus safety risk classification bQ / οηη / ζζηζ / E / γίΛΐ are listed in Table 1, including the concentration of the sample loaded onto the column (g / L), buffer pH, linear flow rate (cm / h), host cell protein loading (HCP in ppm), high molecular weight (HMW) dimer loading (%), higher-order high molecular weight (HMW) loading (%), sodium citrate concentration (mM), HIC resin cycle number, protein loading concentration, and operating temperature (Celsius). Other development factors not selected in the studies were evaluated by residual analysis of pre-existing experimental data. The results indicate that operating temperature factors are expected to have limited impacts on phase design. Controlling operating temperature factors in the virus laboratory proved difficult (Lu et al., Recent Advancement in Application of Hydrophobic Interaction Chromatography for Aggregate Removal in Industrial Purification Process, Current Pharmaceutical Biotechnology, 2009, 10, 427-433). Recycled (reused) HIC resin showed no impact in the viral removal studies. Measuring and controlling impurity factors in the load (external study) proved challenging. Table 1. Selected development factors bQ j onn / zznz / E / YiAi Factors (X) Retrovirus Safety Risk Classification Load (g / L) 9 pH 6 Linear Speed ​​(cm / h) 6 Load HCP (ppm) 6 Load HMW, Dimer (%) 3 Load HMW, Higher Order (%) 3 Sodium Citrate Concentration (mM) 3 Cycle Number 3 Load Concentration 3 Operating Temperature (Celsius) 3 2.1 Selection of development factors for experimental designs for viral elimination by HIC Various ranges of development factors were selected for the design of experiments (DoE) for viral clearance (VC) from HIC. To maximize the signal-to-noise ratio, broad ranges of development factors were selected, as shown in Table 2, including pH ranges, citrate concentration, column loading, linear speed, hydrophobic strength (HIC resin), and monoclonal antibody isotypes. The rationale for selecting the pH range of 4.2–8.0 is that lower pH has been shown to improve protein clearance from the host cell for some programs. In particular, pH 4.2 was selected as the lower pH limit that was expected not to impact product quality. The rationale for selecting the citrate range of 10–200 mM is to select a broad range of citrate concentrations, as the strength of the cosmotrope can modulate virus adsorption to the column.The rationale for selecting a column loading range of 40–200 g / L is that a higher column loading may represent the worst-case scenario for viral removal due to competitive binding. The rationale for selecting a linear velocity of 100–300 cm / h is that variations in contact time can limit virus adsorption. In particular, a shorter contact time is expected to decrease diffusion, which could limit virus adsorption. IgG1 and IgG4 isotypes were selected to include the loading attribute to meet potential regulatory requirements. bQ / onn / zznz / E / YiAi Table 2. Range selections of developmental factors for DoE for viral clearance by HIC Factors DoE Ranges for VC Phase Design References Ranges in Preexisting Experimental Data pH 4.2-8.0 The common multivariate DoE range for HIC is 4.5-6.0. 4.4-8.0 Citrate 10-200 mM The common multivariate DoE range for HIC is 10-50 mM. 20-150 Column Loading 40-200 g / L The common multivariate DoE range for HIC is 50-150 g / L 80-160 Linear Velocity 100-300 cm / h Multivariate DoE range for HIC 150-200 Hydrophobic Strength (HIC resin) Weak (phenylsepharose 6 FF HS); strong (phenyl capto HS) 10 resins evaluated by high-throughput screening Only phenyl capto mAb isotype lgG1, lgG4 IgG 1 / lgG4 Mainly lgG4 The development factor ranges were further verified to ensure that wide factor ranges did not result in impossible factor combinations, as operations with such wide factor ranges could risk failure in HIC runs, possibly due to irreversible bonding or high column pressure. Pre-study experiments were conducted to confirm the worst-case design space for HIC performance, such as yield percentage and cleaning strategy. These pre-study experiments included screening tests at low pH 4.2. Low loading at a concentration of 40 g / L, high citrate concentration at 200 mM for each monoclonal antibody (mAb) on each resin; estimation of the area under the curve (AUC) analysis to verify the column cleaning effects using 6 N guanidine HCl; and evaluation of the impacts of the freeze / thaw process on the loading material. Pre-study tests did not show significant failure modes, as shown in Table 3. bQ j οηη / ζζηζ / Ε / γίΛΐ Table 3. Preliminary study experiments mAb Resin Yield (%) Post-guanidine HMW ABC % removal factor (% of total) mAb 14 Phenyl capto HS 57.0 19 0.01 mAb 14 Phenylsepharose 6 FF HS 87.0 5.6 0.05 mAb11 Phenyl capto HS 28.0 16 0.09 mAb11 Phenylsepharose 6 FF HS 72.0 3.3 0.01 Because some uncontrolled variables may be present, certain conditions were monitored, such as the number of column tests, processing temperature (such as ambient temperature), virus batch consistency, and the AKTA system (a preparative chromatography system for methods and processes development), capturing uncontrolled variables in residuals or RMSE (root square of the variance of the residuals). Different variable loading concentrations were used for different loading pH or different citrate concentrations (mM) to allow for the column loading shown in Table 2. For the validation of the viral elimination assays, preliminary tests were performed only under the worst-case conditions, such as low pH or high citrate concentrations. In addition, a preliminary test dilution for the worst-case conditions was applied to all runs. 3.1 General description of the process Virus quantification was performed using two viral assays: infectivity and qPCR (Xu et al., An overview of quantitative PCR assays for biologicals: quality and safety evaluation, Dev Biol (Basel) 113: 89-98, 2003). The infectivity assay is a cell-based assay relevant to specific viruses, performed by measuring infectious virus particles. A lack of infectivity indicates viral inactivation or clearance. Infectivity assays are relevant to patient safety because they represent viruses that could potentially infect a patient. Quantification using qPCR only detected the presence of viral genomic copies. A lack of genomic copies indicates viral clearance. To maximize the number of tests, sampling plans were limited.The use of standard volume assays may limit the sensitivity of the assay; however, LRF ranges of 0-4 were achievable. The overall process for this example is illustrated in Figure 1. The starting material, e.g., the HIC load adjusted with freeze / thaw at -80 °C, was filtered using a 0.2 µm filter. Subsequently, the model virus, e.g., X-MuLV at 1.0% v / v, was added to the material followed by filtration using a 0.2 µm filter. The enriched and filtered material was then fed into an HIC column to obtain a pool of HIC products, which were subsequently subjected to a standard volume assay. A portion of the added and filtered material was retained as a retention control. The total LRF can be calculated by combining the results of pH inactivation, HIC resin inactivation, and physical HIC removal. The LRF contributed by HIC resin inactivation and physical HIC removal can be quantified using infectivity assays. The LRF contributed by physical HIC removal can also be quantified using qPCR. Regarding the LRF contributed by pH inactivation, the loss of infectivity can be observed in the retention control due to low pH retention, as shown in Figure 2. The HIC method involves using a three-column volume for equilibration, a six-column volume step for washing, and using 6 M guanidine HCl to strip the column after each cycle. Table 4 shows examples of the DoE method for viral removal by HIC. fq / onn / zznz / E / YiAi Table 4. DoE methods for viral elimination by HIC Step Name Solution Column Volume Linear Velocity (cm / h) Flow Direction Pre-stripping Purified Water 2 200 Downflow Equilibrium N / A 3 200 Downflow Loading N / A / N / A / A Downflow Washing N / A 6 N / A Downflow Stripping 1 Guanidine HCl 6 N 4 200 Upflow Stripping 2 Purified Water 2 200 Upflow Stripping 3 NaOH 1 N 2 200 Downflow Column Storage NaOH 0.1 N 2 200 Downflow 4.1 D-optimal design for evaluating factor ranges D-optimal designs were used to elucidate main effects, interactions, and quadratic effects, as shown in Figure 3. Twenty-eight D-optimal DoE tests were used to evaluate broad ranges of development factors, including: pH at 4.2, 6.1, or 8.0; citrate buffer concentration at 10 mM, 105 mM, or 200 mM; column loading concentration at 40 g / L, 120 g / L, or 200 g / L; linear velocity at 100 cm / h or 300 cm / h; HIC resin with phenylsepharose 6 FF (weak) or phenylcapto HS (strong); or IgG1 or IgG4 monoclonal antibody. D-optimal designs were generated using computer algorithms to correlate the estimated effects. Optimizations of the D-optimal designs were generated based on the chosen optimization criteria and the fitted models. The optimization of a given D-optimal design depended on the model. The computer algorithm selected the optimal set of design tests from a candidate set of possible treatment design tests according to the total number of treatment tests required for a specific experiment and model. The candidate set was a collection of treatment combinations from which the D-optimal algorithm selected the treatment combinations to include in the design. The candidate set of treatment tests comprised possible combinations of various factor levels for inclusion in the experiment. Three responses were obtained, including X-MuLV LRF by infectivity (physical elimination and inactivation), X-MuLV LRF by qPCR (physical elimination only), and passage yield (%). All interactions and quadratics were included in the designs. Example 1. Low pH for viral inactivation to characterize multivariate analysis of DoE Twenty-eight DoE D-optimum assays were used to evaluate broad ranges of growth factors, including pH at 4.2, 6.1, or 8.0. qPCR data served as representative measures for residual virus under low pH conditions, as previous studies have shown that quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR) can be used to replace the infectivity assay when the virus clearance mechanism is physical removal. A previous study showed that the infectivity and qRT-PCR assays were closely correlated (r = 0.85, P < 0.05, n = 22) (Anwaruzzaman et al., Evaluation of infectivity and reverse transcriptase real-time polymerase chain reaction assays for detection of xenotropic murine leukemia virus used in virus clearance validation, Biologicals 43: 256-265, 2015). Eleven of the twenty-eight tests performed at a low pH of 4.2 showed complete inactivation of viral load infectivity, independent of other load properties. Virus quantification was performed for the stock virus control, the HIC load retention control, and the HIC product group, as shown in Figure 4. The mean difference between infectivity and qPCR within the retrospective dataset is 0.5 LRF, which is within the accepted assay variability. Based on the experimental results, two distinct models for viral elimination of HIC-related DoE were generated: a reported LRF model and a qPCR LRF model. The reported LRF model was generated from a combination of LRFs determined by infectivity assays and qPCR. This model included qPCR data for low-pH tests, as the infectivity assay based on observed chemical inactivation unrelated to HIC is not appropriate. The reported LRF model used infectivity LRFs for pH 6.1 and pH 8.0 tests. The qPCR LRF model was generated solely from LRFs obtained using qPCR data that measure the presence of viral genomic copies (physical elimination). Both models demonstrate elimination orthogonal to other unit operations for viral inactivation, such as dedicated low-pH retention at approximately pH 3.60.The LRF qPCR model is less variable compared to the reported LRF model due to the unique elimination mechanism for qPCR. Example 2. Identification of significant development factors using the LRF model reported bQ / onn / zznz / E / YiAi Viral clearance of X-MuLV was assessed by generating a predicted graph based on the actual reported LRF and the predicted reported LRF, as shown in Figure 5. The reported LRF model was used to identify multiple significant development factors, as shown in Figure 5. All variance inflation factors (VIFs) were approximately 1 without removing any outliers. When qPCR data were omitted, this resulted in a poor model fit due to the low signal-to-noise ratio. While some scaled estimates were statistically significant, they may not be considered significant in practice when applied to each monoclonal antibody program within the context of standard process parameter tolerances. Furthermore, the effect of the monoclonal antibody's specific isotype cannot be observed. Example 3. Predictability through the application of the reported elimination model for monoclonal antibodies A reported elimination model was used to predict the behavior of specific monoclonal antibodies. When the reported elimination model was applied to a retrospective dataset, the results showed adequate prediction for the behavior of specific monoclonal antibodies, as shown in Figure 6. Twenty-three monoclonal antibodies were evaluated using one-way analysis of residual X-MuLV infectivity (i.e., actual - predicted) per program. Nineteen of the twenty-three monoclonal antibodies evaluated showed a 95% confidence interval (CI) for mean residual LRF that was within 0.5 LRF of zero (where actual = predicted). The LRF of mAb4 was outside 0.5 LRF. mAb4 had the lowest pl in the dataset at 6.2 with the highest non-DoE citrate buffer concentration of 150 mM. Example 4. Identification of significant developmental factors using an LRF qPCR model Viral clearance of X-MuLV was assessed by generating a predicted graph based on actual and predicted LRF qPCR, as shown in Figure 7. The LRF qPCR model was used to identify multiple significant development factors, as shown in Figure 7. The results show differences compared to the reported LRF model. All variance inflation factors (VIFs) were approximately 1, with the removal of two outliers. The results show a higher R-squared with a lower RMSE compared to the reported LRF model due to the measurement of a single viral clearance mechanism by qPCR. The results obtained from the LRF qPCR model indicate a greater impact of pH compared to the reported LRF model. The results suggest that the citrate development factor may not be significant for viral clearance. Example 5. Predictability through the application of the qPCR model for monoclonal antibodies A qPCR model was used to predict the behavior of specific monoclonal antibodies. When the qPCR model was applied to the retrospective dataset, the results showed adequate prediction for the behavior of specific monoclonal antibodies, as shown in Figure 8. Twenty-one monoclonal antibodies were evaluated using one-way residual LRF (i.e., actual - predicted) analysis per program. Twenty of the twenty-one monoclonal antibodies evaluated showed a 95% confidence interval (CI) for mean residual LRF that was within 0.5 of the predicted LRF without the presence of mAb4 qPCR data. The results indicate less inter-test variation compared to reported LRF (mainly infectivity). The DoE dataset was compared to the retrospective database to better understand viral clearance by HIC, as shown in Table 5. The retrospective dataset has a higher number of specific monoclonal antibodies with limited process variance. The DoE dataset has a lower number of specific monoclonal antibodies with wider ranges of process variance. When the DoE models were applied only to viral clearance studies with fewer than 1 LRF, the results led to a greater understanding of viral clearance by HIC, as shown in Table 6. mAb indicates monoclonal antibody, and mAb12 indicates monoclonal antibody 12 with HIC process in Table 6. bQ t onn / zznz / E / YiAi Table 5. Understanding the viral clearance process by HIC Retrospective dataset DoE dataset (+) Number of mAbs (large) (-) Number of mAbs (small) (-) Limited process variance (+) Wide process variance Table 6. Application of DoE models to the study of VC with less than 1 LRF mAb pH Citrate (mM) Column Loading (g / L) Measured LRF (Log 10) DoE Model Predicted LRF (Log 10) mAb12 8.0 (high) 30 (low medium) 260 (highest) 0.48 (infectivity) 0.98 (qPCR) 1.15 (reported LRF) 0.95 (qPCR) Example 6. Application of DoE to target the improvement of viral clearance by HIC When applied, DoE can provide guidance for optimizing development factors to further improve viral clearance processes by HIC. For example, when various LRF models were applied to viral clearance processes by HIC, improved viral clearance was achieved. For instance, a yield of over 90% of the monoclonal antibody was achieved when more than two LRFs were provided. The results indicate that development factor conditions that resulted in high-yield negative-mode HIC can also achieve high viral clearance. Most conditions that can promote monoclonal antibody adsorption can also promote X-MuLV adsorption.As shown in Figure 9A, three observed data points have 2–3 LRFs with over 90% monoclonal antibody yield, where all pH values ​​were less than or equal to 6, the citrate concentration was low at 10 mM, and the column loading was high at ≥ 120 g / L. Figure 9B shows the prediction profiler applied to IgG4 using DoE to provide instructions for optimizing development factors for both yield and virus LRF. The high yield of approximately 95% can still be maintained with a complementary 2–3 LRF removal. A lower mean removal was expected compared to IgG1, which was consistent with the DoE findings. Example 7. Relative hydrophobicity of HIC resins The relative molecular hydrophobicity of several HIC (GE Healthcare Life Sciences) resins was compared using monoclonal antibodies (mAbs), as shown in Figure 10. The relative molecular hydrophobicity was comparable between phenyl capto HiSub and phenylsepharose 6 Fast Flow HiSub (PS6FFHS). Some reversals were observed for phenylsepharose HP (low sub) and butyl capto (aliphatic). mAb2 was mostly retained under all conditions. mAb9 was not the most strongly retained resin under any condition. Example 8. DoE study for viral clearance by HIC DoE studies for viral clearance by HIC with varying ranges of growth factors were applied to a retrospective database for comparison. DoE studies for viral clearance by HIC with varying ranges of growth factors were applied using a one-way X-MuLV LRF infectivity analysis with or without DoE data, as shown in Figure 11. The DoE data were comparable to a larger HIC viral clearance dataset, where the observed LRF range for DoE was 0.52–3.95, similar to the 0.48–4.01 LRF range in the database. The DoE data had a slightly lower mean LRF with a higher standard deviation (SD), which is statistically equivalent to the larger viral clearance dataset, as shown in Figure 11. Data obtained from DoE studies support the relationship between infectivity LRF and qPCR assays. Infectivity and qPCR assays show significant comparability (R-squared of 0.73) for the retrospective dataset, as shown in Figure 12 using a bivariate fit of infectivity X-MuLV LRF by qPCR X-MuLV LRF. The observed assay relationship can be modeled as Formula (I): Formula (I) Infectivity X-MuLV LRF = -0.516447 + (1.6481014 x qPCR X-MuLV LRF) The interaction profiler of the reported LRF HIC DoE model is shown in Figure 13. When the HIC viral clearance DoE data were applied to the retrospective database for infectivity assays, this contributed to a significant improvement in the model, increasing the R-squared from 0.37 to 0.48 with a similar RMSE, as shown in Figures 14A and 14B. Figure 14A shows the original model. Figure 14B shows the model improved by applying DoE tests. Some significant terms were removed from the original model. Two mechanisms were modeled, e.g., inactivation and clearance. When the viral clearance DoE data by HIC were applied to the retrospective database for a qPCR model, this contributed to a minimal decrease in R-squared from 0.78 to 0.76 with an increased RMSE from 0.36 to 0.42, as shown in Figures 15A and 15B. Figure 15A shows the original model. Figure 15B shows the model modified by applying DoE tests. The DoE data support the retrospective qPCR model, which is strongly influenced by three factors. The load pH development factor was found to be the strongest factor. The design space of the modified model was filled with product-specific data that confirmed the LRF range of the original model. It was found to be easier to model a mechanism, such as clearance. Viral clearance DoE data by HIC were applied to a retrospective database for the monoclonal antibody isotype. The reported LRF of HIC DoE by monoclonal antibody isotype for mAb11 IgG1 is shown in Figure 16. Figure 16 shows a predicted graph and scaled estimates obtained by applying viral clearance DoE data by HIC to a retrospective database for the mAb11 IgG1 isotype according to an example modality. The modified model has a significant model fit with a higher RMSE compared to the total DoE model. Only the load pH and citrate development factors were found to be significant. The mean response of 2.6 LRF in the modified model is higher than the total DoE dataset's 2.3 LRF. Viral clearance DoE data by HIC were applied to a retrospective database for the monoclonal antibody isotype. The reported LRF of HIC DoE by monoclonal antibody isotype for mAb14 lgG4 is shown in Figure 17. Figure 17 shows a predicted plot and scaled estimates obtained by applying viral clearance DoE data by HIC to a retrospective database for the mAb14 lgG4 isotype according to an example modality. The modified lgG4 model has greater sensitivity compared to the lgG1 model with a lower mean LRF (2.1 vs 2.6 Iog10). The charging pH and the quadratic citrate equation were found to have larger estimates. The different structures between lgG4 and lgG1 may allow for virus binding due to weaker partitioning.Relevant contributing mechanisms may include specific properties of monoclonal antibodies, such as hydrophobicity, aggregate scoring, or molecular charge. Viral clearance DOE data by HIC were applied to a retrospective database for different resin types. The reported LRF of HIC DOE by resin for phenylcapto is shown in Figures 18A and 18B. Figure 18A shows a predicted graph obtained by applying HIC viral clearance DOE data to a retrospective database for phenylcapto according to an example modality. Figure 18B shows parameter estimates and a predicted profiler obtained by applying HIC viral clearance DOE data to a retrospective database for phenylcapto according to an example modality. The results indicate a strong model fit, with an R-squared value greater than 0.9. The importance of the monoclonal antibody isotype is the same as the results shown previously, with IgG4 being the worst case. Viral elimination DoE data by HIC were applied to a retrospective database for different resin types. The reported LRF of HIC DoE by resin for phenylsepharose 6 FF HS is shown in Figures 19A and 19B. Figure 19A shows a predicted graph obtained by applying HIC viral elimination DoE data to a retrospective database for phenylsepharose 6 FF according to mi / onn / zznz / E / YiAi with an example modality. Figure 19B shows parameter estimates and a predicted profiler obtained by applying HIC viral elimination DoE data to a retrospective database for phenylsepharose 6 FF according to an example modality. The results indicate a strong model fit, with R-squared greater than 0.95 and RMSE lower than 0.19. The mean LRF is approximately 0.5 log10, which is less than the phenyl capto model (2.1 vs 2.6 log 10).The factors of pH charge and antibody isotype development have the largest effect size. Example 9. Application of HIC's DoE performance model The HIC Design of Experiments (DoE) performance model was applied to monitor the trend between LRF and yield, as shown in Figure 20. Figure 20 displays a predicted graph and profile obtained by applying the HIC DoE performance model to monitor the trend between LRF and yield according to an example modality. The application of the HIC DoE performance model indicated an inverse trend between LRF and yield. Large ranges of development factors were relative to the typical HIC DoE design phase. The broad yield response of 34–101% was relative to the retrospective dataset of 80–100%. Maximizing HIC X-MuLV LRF to 4.3 Iog10 can result in a 28% pass rate. Conditions that promoted extensive monoclonal antibody adsorption may also promote extensive X-MuLV adsorption.The results indicate that the virus binds to the resin under conditions that also promote the adsorption of monoclonal antibody against HIC, since the hydrophobic sites of the virus may be potentially exposed during the HIC process.

Claims

CLAIMS 1. A method for purifying an antibody from a sample comprising one or more impurities including viral particles, the method comprising the steps of: providing the sample comprising the antibody produced in a host cell, adjusting a pH of the sample to a range of approximately 4.2 to approximately 8.0, loading the sample onto a hydrophobic interaction chromatography (HIC) column wherein a concentration of the sample is from approximately 40 g / L to approximately 200 g / L, and collecting the HIC-treated sample.

2. The method according to claim 1, wherein a citrate buffer solution is used to adjust the pH of the sample, wherein the concentration of the citrate buffer solution is from approximately 10 mM to approximately 200 mM.

3. The method according to claim 1, wherein one resin of the HIC column is a phenyl resin. bQ t onn / zznz / E / YiAi 4. The method according to claim 1, wherein one of the HIC column resins is a phenylcapto resin.

5. The method according to claim 1, wherein the hydrophobic strength of the HIC column is within a range from a weak hydrophobic strength to a strong hydrophobic strength.

6. The method according to claim 5, wherein the weak hydrophobic strength is achieved by using a phenyl resin or an equivalent thereof.

7. The method according to claim 5, wherein the strong hydrophobic strength is achieved by using a phenylcapto resin or an equivalent thereof.

8. The method according to claim 1, wherein the antibody is a monoclonal antibody or a bispecific antibody.

9. The method according to claim 1, wherein the antibody has an IgG1 isotype or an lgG4 isotype.

10. The method according to claim 1, wherein a flow rate through the HIC column has a linear velocity of approximately 100 cm / h or approximately 300 cm / h.

11. The method according to claim which further comprises measuring the presence of viral genomic copies.

12. The method according to claim which further comprises measuring the presence of viral particles.

13. The method according to claim which further comprises measuring the presence of both viral genomic copies and viral particles.

14. A method for purifying an antibody from a sample comprising one or more impurities including viral particles, the method comprising the steps of: providing the sample comprising the antibody produced in a host cell, adjusting the pH of the sample to a range of approximately 4.2 to approximately 8.0, loading the sample onto a hydrophobic interaction chromatography (HIC) column wherein the concentration of the sample is approximately 40 g / L to approximately 200 g / L, collecting the HIC-treated sample, and measuring the presence of viral genomic copies and / or viral particles in the collected HIC-treated sample.

15. The method according to claim 14 further comprising optimizing the elimination of viral genomic copies and / or viral particles by performing an optimal experimental design.

16. The method according to claim 14, wherein the D-Optimal experimental design evaluates the following factors: the pH of the sample from approximately 4.2 to approximately 8.0, the column loading, wherein the sample concentration is from approximately 40 g / L to approximately 200 g / L, a linear velocity of a flow rate through the HIC column from approximately 100 cm / h to approximately 300 cm / h, and a hydrophobic strength of the HIC column from a weak hydrophobic strength to a strong hydrophobic strength.

17. The method according to claim 16, wherein the weak hydrophobic strength is achieved by using a phenyl resin or an equivalent thereof, wherein the strong hydrophobic strength is achieved by using a phenyl capto resin or an equivalent thereof.

18. The method according to claim 16, wherein the D-Optimal experimental design further evaluates an antibody isotype.

19. The method according to claim 14, wherein the antibody is a monoclonal antibody or a bispecific antibody.

20. A method for purifying an antibody from a sample comprising one or more impurities including viral particles, the method comprising the steps of: providing the sample comprising the antibody produced in a host cell, adding a citrate buffer solution to the sample, adjusting the pH of the sample to a range of approximately 4.2 to approximately 8.0, loading the sample onto a hydrophobic interaction chromatography (HIC) column wherein the concentration of the sample is approximately 40 g / L to approximately 200 g / L, collecting the HIC-treated sample, and measuring the presence of viral genomic copies and / or viral particles in the collected HIC-treated sample.

21. The method according to claim 20 further comprising optimizing the elimination of viral genomic copies and / or viral particles by performing an optimal experimental design.

22. The method according to claim 21, wherein the D-Optimal experimental design evaluates the following factors: the pH of the sample from approximately 4.2 to approximately 8.0, a citrate buffer concentration from approximately 10 mM to approximately 200 mM, the column loading, wherein the sample concentration is from approximately 40 g / L to approximately 200 g / L, a linear velocity of a flow rate through the HIC column from approximately 100 cm / h to approximately 300 cm / h, and a hydrophobic strength of the HIC column from a weak hydrophobic strength to a strong hydrophobic strength.

23. The method according to claim 22, wherein the weak hydrophobic strength is achieved by using a phenyl resin or an equivalent thereof, wherein the strong hydrophobic strength is achieved by using a phenyl capto resin or an equivalent thereof.

24. The method according to claim 20, wherein the antibody is a monoclonal antibody or a bispecific antibody.