Use of raman spectroscopy in downstream purification
In situ Raman spectroscopy with chemometric modeling addresses the challenge of real-time CQA monitoring in protein purification, improving processing efficiency and quality control in monoclonal antibody production by accurately measuring protein concentration and excipient levels.
Patent Information
- Application Number
- JP2024139486
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2018-08-27
- Filing Date
- 2024-08-21
- Publication Date
- 2024-11-26
AI Technical Summary
Current methods for monitoring critical quality attributes (CQAs) in downstream protein purification processes are limited by the lack of real-time analysis, leading to prolonged processing times and increased risk of batch failures due to unmet quality standards.
In situ Raman spectroscopy is employed to characterize and quantify protein purification intermediates during downstream processing, enabling real-time monitoring of protein concentration, excipient levels, and other CQAs such as high molecular weight species and drug-to-antibody ratio, using chemometric modeling and signal processing techniques to improve accuracy.
This approach allows for precise, real-time control of protein concentration and excipient levels, reducing processing time and minimizing batch failures by ensuring that critical quality attributes are consistently met, thereby enhancing the safety and efficacy of monoclonal antibody production.
Smart Images

Figure 2024164120000001_ABST
Abstract
Description
[Technical field]
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of and priority to U.S. Provisional Patent Application No. 62 / 723,188, filed August 27, 2018, which is incorporated by reference in its entirety.
[0002] TECHNICAL FIELD OF THEINVENTION The present invention relates generally to systems and methods for monitoring and controlling one or more Critical Quality Attributes (CQAs) or parameters in a downstream protein purification process. [Background technology]
[0003] 2. Background of the Invention The number of monoclonal antibodies (mAbs) approved for therapeutic use has increased significantly over the past decade. This is due in part to improvements in large-scale manufacturing processes that facilitate the production of large quantities of mAbs. Furthermore, initiatives such as the Food and Drug Administration's (FDA) Process Analytical Technology (PAT) Framework have led to innovative solutions for process development, process analysis, and process control to improve process understanding and control product quality. Efficient recovery and purification of mAbs from cell culture media is a critical part of the production process. The purification process must ensure that mAbs are produced that are safe for human use. An example of this is monitoring critical quality attributes (CQAs), which include protein attributes and impurities, such as host cell proteins, DNA, viruses, endotoxins, aggregates, concentrations, excipients, and other species that may affect patient safety, efficacy, or potency. Protein concentration is often also a CQA of purified materials, and the appropriate protein concentration of process intermediates can be a critical process parameter for the performance of unit operations. These CQAs need to be monitored throughout production as well as throughout the program lifecycle.
[0004] mAb products are tested at various stages of downstream processing to ensure that the final formulation of the mAb does not contain impurities above a defined level. Quality control in the manufacture of biological products such as mAbs is typically achieved by analyzing purification intermediates and formulated drug substance samples using offline methods for each production lot. Samples are removed from the processing equipment, such as a UF / DF skid, and subjected to offline testing to measure product CQAs such as protein concentration (g / L), buffer excipients, size variants, etc. Lack of real-time monitoring and analysis during production increases processing times and the risk of batch failure due to non-fulfillment of CQAs. Thus, there is a need for a rapid in-line method of real-time quality control monitoring of mAbs.
[0005] It is therefore an object of the present invention to provide a system and method for real-time monitoring of critical quality attributes during downstream refinery processes. Summary of the Invention
[0006] An in situ Raman spectroscopy method and system are provided for characterizing or quantifying protein purification intermediates during production or manufacturing. In one embodiment, in situ Raman spectroscopy is used to characterize or quantify the critical quality attributes of protein drugs during downstream processing (i.e., after the protein purification intermediate is harvested from the cell culture medium). For example, the disclosed in situ Raman spectroscopy method and system can be used to characterize and quantify the critical quality attributes of protein purification intermediates as they are purified before formulation into final pharmaceutical products that are sold or administered. Critical quality attributes include, but are not limited to, protein concentration, excipients, high molecular weight (HMW) species, antibody titer, and drug-to-antibody ratio.
[0007] In one embodiment, a method is provided for producing a concentrated protein purification intermediate by concentrating / diafiltering the protein purification intermediate while simultaneously measuring the concentration of the protein purification intermediate in real time using in situ Raman spectroscopy and adjusting the parameters of the concentration step in real time to obtain the predetermined concentration target and excipient levels required for formulation of the drug substance. The protein purification intermediate product may have a concentration of 5 mg / mL to 300 mg / mL, preferably 50 mg / mL to 300 mg / mL for the subsequent formulation step. In one embodiment, ultrafiltration is used during the primary or final concentration to concentrate the protein purification intermediate to the desired concentration target. During processing after the primary concentration, diafiltration is used as a means of buffer exchange to achieve the desired final formulation components. The protein purification intermediate may be harvested from a bioreactor, fed-batch culture, or continuous culture. In another embodiment, measuring the concentration of the protein purification intermediate may be performed continuously or intermittently in real time. Quantification of protein concentration may be performed hourly or daily at intervals of about 5 seconds to 10 minutes. The protein purification intermediate can be an antibody or antigen-binding fragment thereof, a fusion protein, or a recombinant protein. Spectral data are shown for the peaks at 977-1027 cm -1 , 1408~1485cm -1 , 1621~1711cm -1 , 2823~3046cm -1 , and combinations thereof.
[0008] In another embodiment, a method for generating protein purification intermediates is provided by independently performing Raman spectroscopy on multiple protein purification intermediates to generate a universal model that can quantify any one of the multiple protein purification intermediates. The concentration of the protein purification intermediate can be measured by using in situ Raman spectroscopy on the universal model during the concentration of the protein purification intermediate from the start to the end of the concentration of the protein purification intermediate. In another embodiment, a method for generating protein purification intermediates is provided for generating a protein-specific model that can quantify the protein concentration used to enable commercial production of the protein.
[0009] Models can be generated using partial least squares regression analysis of the raw spectral data and using orthogonal methods on the offline protein concentration data. Pre-processing techniques such as standard normal variant (SNV) and / or point smoothing techniques can be performed on the Raman spectroscopic data to reduce variability and prediction error in the model. -1 It may be a smoothed first derivative. Further model refinement may be performed to identify spectral regions that correlate with CQA predictions such as protein concentration. In one embodiment, the model has a margin of error of 5% or less, preferably 3% or less.
[0010] Yet another embodiment provides a method for monitoring and controlling excipient levels in harvested cell culture fluid and / or protein purification intermediates during downstream purification by measuring excipient concentrations in real time using in situ Raman spectroscopy while purifying the cell culture fluid or protein purification intermediate, and adjusting parameters of the purification step in real time to obtain or maintain a predetermined amount of excipient in the harvested cell culture fluid and / or protein purification intermediate. The excipient can be acetate, citrate, histidine, succinate, phosphate, hydroxymethylaminomethane (Tris), proline, arginine, sucrose, or combinations thereof. The excipient can be a surface excipient such as polysorbate 80, polysorbate 20, and poloxamer 188. [Brief description of the drawings]
[0011] [Figure 1] FIG. 1 is a flow chart illustrating an exemplary protein purification process. [Diagram 2] Figure 2 is a representative spectrograph showing the initial model development for in-line Raman spectroscopy. The X-axis represents Raman shift. The Y-axis represents intensity. The legend on the right represents protein concentration. Spectral regions used during the initial model development included, from left to right, 977-1027 cm-1 (ring structure), 1408-1485 cm-1 (arginine), 1621-1711 cm-1 (secondary structure), and 2823-3046 cm-1 (CH stretch). [Diagram 3]FIG. 3 is a bar graph showing the protein concentration (g / L) of mAb1 during a standard UF / DF unit operation. The open bars are concentrations measured using a UV-Vis-based offline method using a C-technologies SoloVPE system. An error bar of ±5% is the goal for the in-line Raman prediction. Bars with wide hatching represent the model referred to herein as the universal model without mAb1 included in the Raman prediction, and bars with narrow hatching represent the universal model with mAb1. The shaded bars correspond to predictions based on the initial universal model in the range 0-120 g / L. The hatched bars correspond to predictions from the initial universal model above 120 g / L. [Figure 4] Figure 4 is a bar graph showing the absolute Raman model error for various mAbs based on the initial universal model development. The hatched bars represent the initial universal model 0-120 g / L (primary concentrate and diafiltration) and the empty bars represent the initial universal model >120 g / L (final concentrate). The horizontal lines represent the goal of the Raman model to keep the error below 5%. [Diagram 5] Figure 5 is a bar graph showing the absolute Raman model error for various mAbs. Bars with broad hatching represent 0-120 g / L (primary concentration and diafiltration) and bars with narrow hatching represent >120 g / L (final concentration). The horizontal line represents the goal of the Raman model to keep the error below 5%. Two versions of the universal Raman model are shown. The hatched bars represent the initial universal model and the cross-hatched bars represent the updated universal model. [Figure 6] FIG. 6 is a schematic diagram of an UF / DF system, including the location of the in-line Raman probe placement. [Figure 7]Figure 7 is a bar graph showing protein concentration of mAb10 using the universal model or the mAb10-specific model for final concentrated pool (FCP) measurement. Protein concentrations are shown for inline real-time Raman predictions (bars with wide hatching), updated model (bars with narrow hatching), and SoloVPE (empty bars). The X-axis represents experimental group and the Y-axis represents protein concentration. [Figure 8] 8A-8B are bar graphs showing the Raman model error of bench-scale DoE modeling of protein concentration. In FIG. 8A, the Raman model error for real-time predictions at various stages of UF / DF processing (primary concentration, diafiltration, and final concentrate pool) is shown. In FIG. 8B, the Raman model error of the final DoE model at various stages of UF / DF processing (primary concentration, diafiltration, and final concentrate pool) is shown. [Figure 9A] Figure 9A is a bar graph showing the error rate of model prediction ability of model scale-up to pilot processing equipment for mAb11. Bench scale model data is compared to bench scale model date incorporating pilot scale data. The X-axis represents experimental group and the Y-axis represents the error rate (%) of model prediction ability. [Figure 9B] Figure 9B is a bar graph showing the model prediction performance for pilot-scale processing of various monoclonal antibodies. The X-axis represents the experimental group, and the Y-axis represents the error rate (%) of the model prediction performance. [Figure 10A] FIG. 10A is a schematic diagram of an exemplary automated batch UF / DF with Raman feedback. [Figure 10B] FIG. 10B is a schematic diagram of an exemplary automated single-pass / TFF with Raman feedback. [Figure 11] Figure 11 is a graph showing the concentration of mAb14 during UF / DF processing. The X-axis represents throughput (L / m2) and the Y-axis represents the concentration of mAb14 (g / L). [Figure 12A]Figure 12A is a bar graph showing the percentage of high molecular weight (HMW) species of mAb2 during various steps of processing (primary concentration, diafiltration, final concentration) as determined by either SE-UPLC or Raman modeling. The X-axis represents experimental groups and the Y-axis represents HMW percent (%) of mAb2. [Figure 12B] Figure 12B is a bar graph showing the percentage of high molecular weight (HMW) species predicted using various scan times (10 s, 20 s, 30 s) for mAb 15. The X-axis represents experimental group and the Y-axis represents HMW percent (%). [Figure 13] 13 is a dot plot showing actual titer (g / L) versus predicted Raman titer for monoclonal antibody samples derived from mAb 14. The X-axis represents predicted Raman titer and the Y-axis represents actual titer (g / L). [Figure 14A] 14A is a scatter plot showing Raman histidine predictions in various monoclonal antibodies, where the X-axis represents the histidine concentration predicted by Raman modeling and the Y-axis represents the actual histidine concentration measured by amino acid analysis. [Figure 14B] FIG. 14B is a dot plot showing the measured and Raman predicted histidine concentrations of monoclonal antibody samples. [Figure 14C] FIG. 14C is a dot plot showing the measured and Raman predicted arginine concentrations of monoclonal antibody samples. [Figure 15] FIG. 15A is a scatter plot showing the actual drug-to-antibody ratio (DAR) and Raman predicted DAR for monoclonal antibody samples derived from mAb 3. The X-axis represents the Raman predicted DAR. The Y-axis represents the actual DAR measured by UV spectroscopy. FIG. 15B is a scatter plot showing the actual drug-to-antibody ratio (DAR) and Raman predicted DAR for monoclonal antibody samples derived from mAb 1. The X-axis represents the Raman predicted DAR and the Y-axis represents the actual DAR. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0012] I. Definition It is to be understood that the disclosure is not limited to the compositions and methods described herein, and the experimental conditions described, as they may vary. Also, since the scope of the disclosure is limited only by the appended claims, it should be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting.
[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this disclosure belongs.In addition, any compositions, methods, and materials similar or equivalent to those described herein may be used to carry out or test the present invention.All publications mentioned are incorporated herein by reference in their entirety.
[0014] Unless otherwise indicated herein or clearly contradictory from the context, in the context of describing the invention as claimed, and particularly in the context of the claims, the use of "a," "an," "the," and similar referents should be construed to cover both the singular and the plural.
[0015] Unless otherwise stated herein, the recitation of ranges of values herein is merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, and each separate value is incorporated herein as if it were individually recited herein.
[0016] When the term "about" is used, it is intended that the value describes a value that is either above or below the stated value by about ±10%. In other embodiments, the value may be in a range of values that are either above or below the stated value by about ±5%. In other embodiments, the value may be in a range of values that are either above or below the stated value by about ±2%. In other embodiments, the value may be in a range of values that are either above or below the stated value by about ±1%. The preceding ranges are intended to be clear by context, and no further limitations are implied. Unless otherwise indicated herein or otherwise clearly contradicted by context, all methods described herein can be performed in any suitable order. Any and all examples provided herein, or the use of exemplary language (e.g., "such as") are intended solely to clearly illustrate the invention, and do not limit the scope of the invention unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the invention.
[0017] As used herein, "protein" refers to a molecule that comprises two or more amino acid residues linked together by peptide bonds. Proteins include polypeptides and peptides, and also include modifications such as glycosylation, lipid attachment, sulfation, gamma-carboxylation of glutamic acid residues, alkylation, hydroxylation, and ADP-ribosylation. Proteins can be scientific or commercial objects, including protein-based drugs, and proteins include enzymes, ligands, receptors, antibodies, and chimeric or fusion proteins, among others. Proteins are produced by various types of recombinant cells using well-known cell culture methods, and are generally introduced into cells by transfection of genetically engineered nucleotide vectors (e.g., sequences encoding chimeric proteins, or codon-optimized sequences, intronless sequences, etc.), where the vectors may be present as episomes or may be integrated into the genome of the cells.
[0018] An "antibody" refers to an immunoglobulin molecule that consists of four polypeptide chains, two heavy (H) chains and two light (L) chains, bound together via 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 are further subdivided into regions of hypervariability called complementarity determining regions (CDRs), which are interspersed with regions of high conservation called framework regions (FRs). Each VH and VL is 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 both glycosylated and non-glycosylated immunoglobulins of any isotype or subclass. The term "antibody" also includes antibody molecules prepared, expressed, produced or isolated by recombinant means, such as antibodies isolated from a host cell transfected for antibody expression. The term antibody also includes bispecific antibodies, which include heterotetrameric immunoglobulins capable of binding to multiple, distinct epitopes. Bispecific antibodies are generally described in U.S. Patent Application Publication No. 2010 / 0331527, which is incorporated herein by reference.
[0019] "Secondary structure" refers to the local folded structures formed within a polypeptide by interactions between the atoms of the backbone. The most common types of secondary structure are the α-helix and the β-pleated sheet. Both structures are held in place by hydrogen bonds formed between the carbonyl O of one amino acid and the amino H of another.
[0020] As used herein, "excipient" refers to a pharmacologically inactive substance used as a stabilizer for long-term storage of a formulated drug substance. Although additional excipients are typically added to the final concentrate pool to produce a formulated drug substance, excipient levels are monitored during UF / DF processing to ensure that they do not impact the desired formulation strategy. Excipients provide bulk to the pharmaceutical formulation, promote drug absorption or solubility, provide stability, and prevent denaturation. Common pharmaceutical excipients include, but are not limited to, amino acids, fillers, binders, disintegrants, coatings, adsorbents, buffers, chelating agents, lubricants, glidants, preservatives, antioxidants, flavors, sweeteners, colors, solvents and cosolvents, and viscosity agents. In one embodiment, the excipient is polyethylene glycol, including but not limited to PEG-3550.
[0021] "Polyethylene glycol" or "PEG" is a polyether polymer of ethylene oxide commonly used in food, pharmaceuticals, and cosmetics. It is a non-ionic polymer useful as a molecule to reduce the solubility of biomolecules. PEG is commercially available in a variety of molecular weights ranging from 300 g / mol to 10,000,000 g / mol. Exemplary types of PEG include, but are not limited to, PEG 20000, PEG 8000, and PEG 3350. PEG is available in a variety of configurations including linear, branched (3-10 chains attached to a central core), star (10-100 chains attached to a central core), and comb (multiple chains attached to the polymer backbone).
[0022] "Raman spectroscopy" is a spectroscopic technique used to measure the wavelength and intensity of inelastically scattered light from molecules. It is based on the principle that monochromatic radiation incident on a material is reflected, absorbed, or scattered in a specific manner that depends on the particular molecule or protein that receives the radiation. Most of the energy is scattered at the same wavelength, called elastic or Rayleigh scattering. A small amount (less than 0.001%) is scattered at a range of wavelengths, called inelastically or Raman scattering. Raman scattering is associated with rotational, vibrational, and electronic level transitions. The chemical and structural composition of a sample can be revealed by Raman spectroscopy.
[0023] "Ultrafiltration" is a membrane process frequently used for protein concentration in downstream processing of protein therapeutics; during purification of recombinant proteins, ultrafiltration is a size-based separation in which species larger than the membrane pores are retained and smaller species pass freely. During processing, the protein solution is pumped tangentially across the surface of a semi-permeable parallel flat-sheet membrane. The membrane is permeable to buffers and buffer salts, but generally impermeable to monoclonal antibodies. A driving force for osmosis is applied against the transmembrane pressure (TMP) induced by a flow restriction at the outlet of the membrane channel (TMP = (P feed +P retentate ) / 2-P permeate ).
[0024] As used herein, "primary concentration" refers to the first step in which transmembrane pressure drives water and salt across a permeable membrane, thereby reducing the volume of liquid and increasing protein concentration. The degree of concentration of the primary concentration may be optimized to achieve a good balance between throughput, protein stability, processing time, and buffer consumption.
[0025] As used herein, "diafiltration" refers to a technique that uses a semipermeable membrane to transfer a product of interest from one liquid medium to another. Buffer exchange and desalting are usually performed using diafiltration mode, where small impurities and buffer components are effectively washed out of the product by continuously adding buffers aimed at adjusting the protein to a stable pH and excipient concentration, thereby allowing the product to be highly concentrated. This can be performed in continuous or discontinuous mode, depending on the processing technique.
[0026] Diafiltration is often combined with ultrafiltration to remove impurities and salts while achieving the desired volume reduction. UF / DF is the final unit operation of downstream purification, which adjusts the mAb to achieve pH, excipient content, and protein concentration conducive to long-term storage and addition of stabilizing excipients to produce the formulated drug substance (FDS).
[0027] As used herein, "final concentration" refers to the final step in which transmembrane pressure drives water and salts across a permeable membrane, thereby reducing the volume of liquid and increasing the protein concentration to a desired target for storage and / or formulation. The resulting pool from the final concentration step is the final concentrated pool (FCP). This final concentration step can be performed in a continuous or discontinuous processing mode.
[0028] The terms "bioproduct" and "protein purification intermediate" may be used interchangeably and refer to any antibody, antibody fragment, modified antibody, protein, glycoprotein, or fusion protein, as well as the final drug substance purified from a bioreactor process.
[0029] The terms "control" and "controlling" refer to adjusting the amount or concentration level of a critical quality attribute in a harvested cell culture fluid to meet a predefined set point.
[0030] As used herein, the term "upstream processing" refers to the first step in which an antibody or therapeutic protein is produced in a bioreactor, usually by a bacterial or mammalian cell line. Upstream processing includes medium preparation, cell culture, cell separation and harvesting. When the cells reach a desired density, they are harvested and moved to the downstream section of the bioprocess. The term "downstream processing" refers to the isolation and purification carried out after the antibody or therapeutic protein is harvested from the bioreactor. This usually means the recovery of the product from an aqueous solution through several different modalities. The harvested product is processed to meet the requirements of purity and critical quality attributes during downstream processing.
[0031] As used herein, the term "protein purification intermediate" refers to the protein harvested from the bioreactor and any intermediate in downstream processing.
[0032] The term "concentrated protein purification intermediate" refers to a protein purification intermediate having a concentration of greater than 5 mg / mL, more preferably between 50 mg / mL and 300 mg / mL.
[0033] The terms "monitor" and "monitoring" refer to periodically checking the amount or concentration level of a critical quality attribute in a cell culture or in a harvested cell culture fluid.
[0034] The term "harvested cell culture fluid" refers to the liquid removed from a bioreactor containing cells engineered to secrete any protein of interest. The "harvested cell culture fluid" optimally contains any secreted protein of interest, such as a monoclonal antibody.
[0035] As used herein, "Critical Quality Attribute (CQA)" refers to a physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to ensure the desired product quality of a biological therapeutic product. These attributes may affect safety, efficacy, and / or potency. Critical quality attributes include, but are not limited to, protein concentration, high molecular weight species, buffer excipients, and pH.
[0036] As used herein, "formulated drug substance" refers to an active ingredient intended to provide pharmacological activity, but does not include the intermediates used in the synthesis of such ingredient.
[0037] As used herein, a "universal model" refers to a mathematical correlation of the spectral properties of various recombinant proteins that is used to predict important quality attributes.
[0038] As used herein, "mAb-specific model" refers to the mathematical correlation of the spectral characteristics of one particular protein that is used to predict important quality attributes.
[0039] As used herein, "titer" refers to the amount of antibody or protein molecules in solution.
[0040] II. Systems and methods for evaluating the properties of downstream protein purification products Systems and methods are provided for monitoring and controlling protein concentration during protein production. Accurate measurement of concentrated protein solutions is difficult due to the high viscosity of the solutions (>10 cP). Accurate quantification requires specialized offline equipment as well as typically diluting the solution. At high UF / DF concentrations, the final excipient level is a function of protein concentration due to the Gibbs-Donnan effect. Lack of real-time monitoring and analysis during production lengthens processing times and increases the likelihood of batch failure due to non-fulfillment of CQAs. The systems and methods disclosed herein can be used for in-line monitoring of protein concentration and other critical quality attributes.
[0041] In one embodiment, the Raman spectroscopy system is an in-line or in situ Raman spectroscopy system used during the production of a final concentrated pool with high concentration (150 g / L or more). Typically, the Raman spectroscopy system is used downstream of the production of the protein purification intermediate, for example, during the processing and subsequent purification of the protein purification intermediate after harvest from a bioreactor or fed-batch system. An exemplary protein purification process is as illustrated in FIG. 1. Typically, the protein purification intermediate from the cell culture is processed through various purification steps, such as harvest (100), affinity capture (110), viral inactivation (120), polishing chromatography (130 and 140), viral retaining filtration (150), and ultrafiltration / diafiltration (160) to generate a final concentrated pool, which is then formulated into a drug substance. In one embodiment, the protein concentration in the harvested cell culture fluid is monitored by in situ Raman spectroscopy.
[0042] A. Raman Spectroscopy In one embodiment, the protein concentration in the harvested cell culture fluid is monitored and controlled by Raman spectroscopy. Raman spectroscopy is a type of vibrational spectroscopy that provides information about molecular vibrations that can be used to identify samples and quantify key quality attributes. In situ Raman analysis is a method of analyzing samples in situ without extracting a portion of the sample for analysis by a Raman spectrometer. In situ Raman analysis is advantageous in that the Raman spectrometer is non-invasive and non-destructive, reducing the risk of contamination and loss of protein quality. In-line Raman analysis can be implemented to enable continuous processing while monitoring the protein concentration of the harvested cell culture fluid, protein purification intermediates, and / or final concentrated pool.
[0043] In situ Raman analysis can provide a real-time assessment of protein concentration in a protein purification intermediate. For example, the raw spectral data provided by in situ Raman spectroscopy can be used to obtain and monitor the current protein concentration of the protein purification intermediate. In this embodiment, to ensure that the raw spectral data is continuously updated, the spectral data from the Raman spectroscopy should be obtained about every 5 seconds to 10 hours. In another embodiment, the spectral data should be obtained about every 15 minutes to 1 hour. In another embodiment, the spectral data should be obtained about every 20 minutes to 30 minutes.
[0044] Monitoring of protein concentration in protein purification intermediates can be analyzed by any commercially available Raman spectrometer that allows in situ Raman analysis. The in situ Raman analyzer must have the capability of obtaining raw spectral data in protein purification intermediates. For example, the Raman analyzer must be equipped with a probe that can be inserted in-line into the fluid circuit. Suitable Raman analyzers include, but are not limited to, the RamanRXN2 and RamanRXN4 analyzers manufactured by Kaiser Optical Systems, Inc., Ann Arbor, Michigan.
[0045] The raw spectral data obtained by in situ Raman spectroscopy may be compared to an offline protein concentration measurement to correlate peaks in the spectral data to protein concentration. The offline protein concentration measurement may be used to determine which spectral regions indicate protein signal. The offline measurement data may be collected through any suitable analytical method. For example, for protein concentration, an offline measurement may be collected using SoloVPE from C-technologies. Additionally, any type of multivariate software package, including SIMCA 13 from MKS Data Analytic Solutions, Umeå, Sweden, may be used to correlate peaks in the raw spectral data to offline measurements of protein concentration. However, in some embodiments, the raw spectral data may need to be pre-processed with a spectral filter to remove fluctuating baselines. For example, the raw spectral data may be pre-processed with any type of point smoothing or normalization technique. Normalization may be required to compensate for probe, optics, laser power variations, and exposure times using a Raman analyzer. In one embodiment, the raw spectral data is collected using a 21 cm -1The pre-processing techniques may be combined for specific spectral regions to improve model predictions.
[0046] Chemometric modeling may be performed on the resulting spectral data. In this embodiment, one or more multivariate methods, including but not limited to partial least squares (PLS), principal component analysis (PCA), orthogonal partial least squares (OPLS), multivariate regression, canonical correlation, factor analysis, cluster analysis, graphical procedures, etc., may be used on the spectral data. In one embodiment, the resulting spectral data is used to create a PLS regression model. The PLS regression model may be created by projecting the predictor and observed variables into a new space. In this embodiment, measurements from the Raman analysis and offline measurements may also be used to create the PLS regression model. The PLS regression model provides predicted process values, such as predicted protein concentration. In one embodiment, the model provides predicted protein concentration values with an error of 5% or less compared to offline protein concentration values. In a preferred embodiment, the model provides predicted protein concentration values with an error of 3% or less compared to offline protein concentration values.
[0047] After chemometric modeling, signal processing techniques may be applied to the protein concentration predictions. In one embodiment, the signal processing techniques suppress the variability and prediction error of the model. In this aspect, one or more pre-processing techniques may be applied to the protein concentration predictions. Any pre-processing techniques known to those of skill in the art may be utilized. For example, the noise reduction techniques may include data smoothing and / or signal removal. Smoothing is achieved by a series of smoothing algorithms and filters, while signal removal uses signal characteristics to identify data that should not be included in the analyzed spectral data. In one embodiment, the protein concentration predictions are reduced in noise by a noise reduction filter. The noise reduction filter results in a final protein concentration prediction. The noise reduction approach in this aspect combines the raw measurements with a model-based estimate of what the measurements will result in according to the model. In one embodiment, the noise reduction technique combines the currently predicted protein concentration value with its uncertainty. The uncertainty can be determined by the reproducibility of the predicted protein concentration value and the current protein concentration value. When the next protein concentration prediction is observed, the estimate of the predicted protein concentration value is updated using a weighted average where more weight is given to estimates with higher certainty. Using an iterative approach, the final protein concentration value can be updated based on previous measurements and the current measurement. In this embodiment, the algorithm should be recursive and can be run in real time to utilize the current protein concentration prediction, previous values, and experimentally measured constants. The denoising technique improves the robustness of the measurements received from the Raman analysis and PLS prediction by reducing the noise on which the automatic feedback controller operates.
[0048] B.How to use The disclosed method can be used to monitor and control protein concentration in harvested cell culture and / or protein purification intermediate fluids during downstream protein purification processes. Common downstream purification processes include, but are not limited to, centrifugation, direct depth filtration, Protein A affinity purification, viral inactivation steps, ion exchange chromatography, hydrophobic interaction chromatography, size exclusion chromatography, ultrafiltration / diafiltration, viral retention filtration, and combinations thereof. These unit operations are used in defined sequential combinations to separate any protein of interest and ensure that impurities and / or critical quality attributes are monitored prior to manufacturing a formulated drug substance. In one embodiment, the disclosed method includes a Raman probe in-line with the fluid circuit. In another embodiment, the disclosed method can be used to generate a concentrated protein purification intermediate or a final concentrated pool.
[0049] 1. Antibody Titer and Protein Concentration Both antibody titer and protein concentration are important factors in the purification of biological products. Antibody titer measured after the first harvest of cell culture fluid is important to measure column loading and ensure a robust purification process to remove impurities. Monitoring protein concentration throughout the purification steps is important to ensure both the proper concentration of the final product and the proper performance of the purification unit operations being performed. Inadequate protein concentration can lead to the production of ineffective drug product or formulated drug substance.
[0050] In one embodiment, the harvested cell culture fluid is subjected to Raman spectroscopy immediately after harvest but before additional purification is initiated. Raman spectroscopy data can be used to quantify antibody titers in the harvested cell culture fluid after harvest. The disclosed method can be used to measure protein concentration at multiple steps during the protein purification process, for example, during affinity capture, polishing chromatography, virus-retentive filtration, or UF / DF. In-line Raman probes allow for detection of Raman scattering of harvested cell culture fluid and / or protein purification intermediates in the fluid circuit without removing the sample from the system, providing analytical characterization that is typically determined offline.
[0051] In one embodiment, if the protein concentration during UF / DF falls outside a predetermined concentration range, the system is notified and the protein purification intermediate is modified accordingly. For example, if the protein concentration of the protein purification intermediate is lower than a predetermined protein concentration, UF / DF can be performed to further concentrate the protein purification intermediate.
[0052] In one embodiment, the enrichment step is carried out by Protein A affinity chromatography.
[0053] 2. Drug-to-Antibody Ratio In another embodiment, the method of the present disclosure can be used to monitor and control the drug-to-antibody ratio (DAR). DAR is a quality attribute monitored during the development of antibody-drug conjugates (ADCs), antibody-radionuclide conjugates (ARCs), and general protein conjugates (potent steroids, non-cytotoxic payloads, etc.) to ensure consistent product quality and facilitate subsequent labeling with payloads. DAR is the average number of drugs or other therapeutic molecules conjugated to an antibody and is an important quality attribute in the manufacture of therapeutic conjugates. DAR value affects the efficacy of the conjugated drug, as low drug loading reduces drug efficacy and high loading can negatively affect pharmacokinetics and safety.
[0054] In one embodiment, the antibody-radionuclide conjugate is subjected to Raman spectroscopy immediately after conjugation but before any additional purification takes place. Raman spectroscopy data can be used after conjugation to determine the DAR.
[0055] In one embodiment, if the DAR during processing falls outside a predetermined concentration range, the system is notified and the ADC intermediate is modified accordingly.For example, if the DAR of the ADC intermediate falls below a predetermined DAR, the components of the conjugation reaction may be modified.For example, reactant concentrations may be optimized, linker types may be changed, temperature may be changed, or other manufacturing variables may be optimized.
[0056] 3. Buffer Excipients The disclosed methods can be used to monitor and control the levels of buffer excipients in harvested cell culture fluid and / or protein purification intermediates during downstream purification. Buffer excipients commonly used in monoclonal antibody production include, but are not limited to, acetate, citrate, histidine, succinate, phosphate, and hydroxymethylaminomethane (Tris), as well as proline and arginine. Surfactant excipients include, but are not limited to, polysorbate 80 (Tween 80), polysorbate 20 (Tween 20), and poloxamer 188. Polyol / disaccharide / polysaccharide excipients include, but are not limited to, mannitol, sorbitol, sucrose, and dextran 40. Antioxidant excipients include, but are not limited to, ascorbic acid, methionine, and ethylenediaminetetraacetic acid (EDTA). Two commonly used amino acid excipients are histidine and arginine. In a preferred embodiment, the excipients being monitored and controlled are histidine and arginine.
[0057] The final concentrate pool excipient concentration differs from the diafiltration buffer composition due to a combination of excluded volume and Donnan effects. The Donnan effect is a phenomenon that occurs due to the retention of net positively charged proteins by the membrane during UF / DF, necessitating charge neutrality in both the retentate and the permeate. To balance the positively charged proteins, negatively charged buffer components are enriched in the retentate compared to the diafiltration buffer, while positively charged buffer components are flushed out. This effect can lead to the pH and buffer excipient concentrations of the FCP being significantly different from the composition of the diafiltration buffer (Stoner, et al., J Pharm Sci, 93:2332-2342 (2004)).
[0058] Volume exclusion describes the behavior of highly concentrated samples where the protein occupies a significant portion of the solution volume. As buffer is excluded from the volume occupied by the protein, the solute concentration of the buffer decreases as the protein concentration, expressed as moles (or mass) of solute per solution volume, increases. As buffer is excluded from the volume occupied by the protein, the solute concentration of the buffer decreases as the protein concentration, expressed as moles (or mass) of solute per solution volume, increases.
[0059] Based on both these principles and on buffer excipient levels being critical quality attributes, the in-line Raman probe minimizes offline analytical characterization and provides additional understanding of the process to ensure sufficient excipient levels prior to formulation.
[0060] 4. High molecular weight impurities In monoclonal antibody production, low levels of product-related impurities are often present even after extensive purification steps. High molecular weight (HMW) species (e.g., antibody dimer species) are product-related impurities that contribute to the size heterogeneity of mAb products. The formation of HMW species in therapeutic mAb drugs as a result of protein aggregation can compromise both efficacy and safety. HMW species are considered a CQA to be routinely monitored during drug development and as part of purified protein drug release testing during manufacturing.
[0061] In one embodiment, the method of the present disclosure can be used to identify protein pharmaceuticals that contain HMW species. HMW species can be detected by Raman spectroscopy at various steps during the purification process, such as, but not limited to, affinity capture, virus inactivation, polishing chromatography, virus retention filtration, ultrafiltration / diafiltration, or a combination thereof.
[0062] In one embodiment, the disclosed method detects HMW species in harvested cell culture fluid and the fluid is further processed to remove HMW species. Methods of removing HMW species for cell culture fluid include additional polishing steps, including but not limited to cation exchange chromatography and anion exchange chromatography.
[0063] C.UF / DF system FIG. 6 illustrates an UF / DF processing system including, for example, various locations of an in-line Raman probe (circled numbers 1-4). Protein purification intermediates are pumped by diafiltration pump 200 into retentate vessel 205. This diafiltration pump can be a peristaltic pump, rotating lobe, pressure transfer, or diaphragm pump. Fluid from retentate vessel 205 flows to feed pump 210, which can be either a rotating lobe, peristaltic pump, or diaphragm pump, passes through feed pressure valve 215, and enters tangential flow filtration module (TFFM) 220. In TFFM 220, the protein purification intermediates are subjected to ultrafiltration through a membrane. The bioproduct of interest is retained in the fluid (retentate), while low molecular weight solutes, including water and buffer excipients, pass through the membrane in the permeate (filtrate), which exits the system by passing through permeate pressure valve 225 and into waste tank 230. The retentate exits the TFFM 220 and flows through the retentate pressure valve 235, the transmembrane pressure control valve 240, and the retentate return channel 245 back into the retentate vessel 205. This process can be repeated as necessary to concentrate the bioproduct, remove impurities, and ensure that CQAs are within acceptable limits. During diafiltration, the same flow path is followed as above, with permeable solutes being exchanged as new buffer is pumped into the product stream. If new buffer is added at the same rate that permeate is removed from the system, the sum of the retentate tank and skid hold-up volumes defines the system capacity. One turnover volume (TOV) is defined as the amount of diafiltration buffer added to the UF / DF process that is equal to the system capacity. Typically, an exchange of eight times the system volume (8 TOV) ensures greater than 99.9% buffer exchange (Schwarts, L., Scientific and Technical Report, PN 33289).
[0064] Additionally, it is necessary to mix the protein solution in the retentate vessel 205 during the UF / DF process. The difference in density between the diafiltration buffer, the retentate return, and the bulk retentate during diafiltration requires that the agitation in the tank be sufficient to ensure adequate buffer exchange, yet gentle enough to avoid shear, as these have been observed to result in protein aggregation and the generation of subvisible particles (SVPs) in certain products. Additionally, it is important to ensure adequate mixing of the retentate return in the concentration step to prevent polarization of the protein concentration in the retentate tank, which would result in a higher protein concentration being fed to the UF / DF membrane.
[0065] In one embodiment, the Raman probe is placed at position 1 in the retentate vessel 205 downstream of the diafiltration pump 200. Alternatively, the Raman probe can be placed in-line at position 2 downstream of the retentate vessel 205 and before the feed pump 210. In another embodiment, the Raman probe is placed in-line at position 3 between the feed pressure pump 215 and the tangential flow filtration module 220. In yet another embodiment, the Raman probe is placed in-line in the retentate return 245. In complex systems with engineering and processing constraints, the location of the Raman probe is critical to ensure accurate in-line measurements.
[0066] D. Cell culture The harvested cell culture fluid can be taken from a bioreactor containing cells engineered to produce monoclonal antibodies. The term "cells" encompasses any cell suitable for expressing a recombinant nucleic acid sequence. Cells include prokaryotic and eukaryotic cells, such as bacterial cells, mammalian cells, human cells, non-human animal cells, avian cells, insect cells, yeast cells, or cell fusions, such as hybridomas and quadromas. In certain embodiments, the cells are human, monkey, ape, hamster, rat, or mouse cells. In other embodiments, the cell is selected from the following cells: Chinese Hamster Ovary (CHO) (e.g., CHO K1, DXB-11 CHO, Veggie-CHO), COS (e.g., COS-7), retinal cells, Vero, CV1, kidney (e.g., HEK293, 293 EBNA, MSR 293, MDCK, HaK, BHK21), HeLa, HepG2, WI38, MRC 5, Colo25, HB 8065, HL-60, lymphocytes such as Jurkat (T lymphocytes) or Daudi (B lymphocytes), A431 (epidermal), U937, 3T3, L cells, C127 cells, SP2 / 0, NS-0, MMT cells, stem cells, tumor cells, and cell lines derived from the aforementioned cells. In some embodiments, one or more viral genes are contained within the cell, e.g., a retinal cell (e.g., a PER.C6® cell) expressing the viral genes. In some embodiments, the cell is a CHO cell. In other embodiments, the cell is a CHO K1 cell.
[0067] In protein production, "fed-batch cell culture" or "fed-batch culture" refers to batch culture in which cells and medium are first fed into a culture vessel. Additional culture nutrients are slowly fed to the culture in discrete increments during the culture, with or without periodic harvesting of cells and / or product before the end of the culture. Fed-batch culture includes "semi-continuous fed-batch culture" in which the entire culture (which may include cells and medium) is periodically removed and replaced with fresh medium. Fed-batch culture is distinguished from simple "batch culture", in which all components for cell culture (including animal cells and all culture nutrients) are fed into the culture vessel at the beginning of the culture process in batch culture. Fed-batch culture can differ from "perfusion culture" insofar as the supernatant is not removed from the culture vessel during the standard fed-batch process. In perfusion culture, cells are confined in the culture, for example by filtration, and culture medium is continuously or intermittently introduced and removed from the culture vessel. In contrast, during fed-batch cell culture, it is contemplated that samples may be removed for testing purposes. The fed-batch process is continued until it is determined that maximum working volume and / or protein production has been reached, after which the protein is harvested.
[0068] The phrase "continuous cell culture" refers to a technique used in applications where cells are continuously grown, usually in a particular growth phase. It may be necessary to maintain a cell culture at a particular growth stage, for example when a constant supply of cells is required or when it is required to produce any protein of particular interest. Therefore, to maintain cells at that particular stage, conditions need to be continuously monitored and adjusted accordingly.
[0069] The terms "cell culture medium" and "culture medium" refer to a nutrient solution used to grow mammalian cells, and typically provide nutrients necessary to promote cell growth, such as carbohydrate energy sources, essential amino acids (e.g., phenylalanine, valine, threonine, tryptophan, methionine, leucine, isoleucine, lysine, and histidine) and non-essential amino acids (e.g., alanine, asparagine, aspartic acid, cysteine, glutamic acid, glutamine, glycine, proline, serine, and tyrosine), trace elements, energy sources, lipids, vitamins, and the like. Cell culture media may contain extracts, such as serum or peptones (hydrolysates), to provide raw materials to support cell growth. Instead of including animal-derived extracts in the medium, yeast-derived or soybean extracts may be included. Chemically defined media refers to cell culture media in which all chemical components are known (i.e., have a known chemical structure). Chemically defined media do not contain any animal-derived components, such as serum or animal-derived peptones. In one embodiment, the medium is a chemically defined medium.
[0070] The solution may also contain components, including hormones and growth factors, that enhance growth and / or survival rates above minimum rates. The solution may be formulated to provide an optimal pH and salt concentration for survival and proliferation of the particular cells being cultured.
[0071] E. Any Protein of Interest Any protein of interest suitable for expression in prokaryotic or eukaryotic cells can be monitored using the method of the present disclosure.For example, any protein of interest includes, but is not limited to, an antibody or its antigen-binding fragment, a chimeric antibody or its antigen-binding fragment, an ScFv or its fragment, an Fc fusion protein or its fragment, a growth factor or its fragment, a cytokine or its fragment, or an extracellular domain or its fragment of a cell surface receptor.Any protein of interest can be a simple polypeptide consisting of a single subunit, or a complex multi-subunit protein that contains two or more subunits.Any protein of interest can be a biopharmaceutical, a food additive or preservative, or any protein product that is subject to purification and quality standards.
[0072] In some embodiments, the antibody is selected from the group consisting of anti-programmed cell death 1 antibodies (e.g., anti-PD1 antibodies, source: U.S. Pat. No. 9,987,500), anti-programmed cell death ligand-1 (e.g., anti-PD-L1 antibodies, source: U.S. Pat. No. 9,938,345), anti-Dll4 antibodies, anti-angiopoietin-2 antibodies (e.g., anti-ANG2 antibodies, source: U.S. Pat. No. 9,402,898), anti-angiopoietin-like 3 antibodies (e.g., anti-AngPtl3 antibodies, source: U.S. Pat. No. 9,018,356), anti-platelet derived growth factor receptor antibodies (e.g., anti-PDGFR antibodies, source: U.S. Pat. No. 9,265 ,827), anti-Erb3 antibodies, anti-prolactin receptor antibodies (e.g., anti-PRLR antibodies, source: U.S. Patent No. 9,302,015), anti-complement 5 antibodies (e.g., anti-C5 antibodies, source: U.S. Patent No. 9,795,121), anti-TNF antibodies, anti-epidermal growth factor receptor antibodies (e.g., anti-EGFR antibodies, source: U.S. Patent No. 9,132,192 or anti-EGFRvIII antibodies, source: U.S. Patent No. 9,475,875), anti-proprotein convertase subtilisin kexin-9 antibodies (e.g., anti-PCSK9 antibodies, source: U.S. Patent No. 8,062,640 or U.S. Patent No. 9,540,440 9), anti-growth differentiation factor 8 antibodies (e.g., anti-GDF8 antibodies, also known as anti-myostatin antibodies, source: U.S. Patent No. 8,871,209 or U.S. Patent No. 9,260,515), anti-glucagon receptor (e.g., anti-GCGR antibodies, source: U.S. Patent No. 9,587,029 or U.S. Patent No. 9,657,099), anti-VEGF antibodies, anti-IL1R antibodies, interleukin 4 receptor antibodies (e.g., anti-IL4R antibodies, source: U.S. Patent Application Publication No. 2014 / 0271681(A1) or U.S. Patent No. 8,735,095 or U.S. Patent No. 8,945,559), anti-interleukin 6 receptor antibodies ( Examples include anti-IL6R antibodies, see U.S. Pat. Nos. 7,582,298, 8,043,617, or 9,173,880; anti-IL1 antibodies, anti-IL2 antibodies, anti-IL3 antibodies, anti-IL4 antibodies, anti-IL5 antibodies, anti-IL6 antibodies, anti-IL7 antibodies, anti-interleukin 33 (e.g., anti-IL33 antibodies, see U.S. Pat. Nos. 9,453,072 or 9,637,535); anti-respiratory rash antibodies (e.g., anti-RSV antibodies, see U.S. Pat. App. Pub. No. 9,447,173); anti-cluster of differentiation 3 (e.g., anti-CD3 antibodies, see U.S. Pat. Nos. 9,447,173 and 9,447,173, and U.S. Patent Application No. 62 / 222,605), anti-cluster of differentiation 20 (e.g., anti-CD20 antibodies, see U.S. Patent No. 9,657,102 and U.S. Patent Application Publication No. 20150266966(A1), U.S. Patent No. 7,879,984), anti-CD19 antibodies, anti-CD28 antibodies, anti-cluster of differentiation 48 (e.g., anti-CD48 antibodies, see U.S. Patent No. 9,228,014), anti-Fel d1 antibodies (e.g., those described in U.S. Pat. No. 9,079,948), anti-Middle East Respiratory Syndrome virus (e.g., anti-MERS antibodies, source: U.S. Pat. No. 9,718,872), anti-Ebola virus antibodies (e.g., U.S. Pat. No. 9,771,414), anti-Zika virus antibodies, anti-lymphocyte activation gene 3 antibodies (e.g., anti-LAG3 antibodies or anti-CD223 antibodies), anti-nerve growth factor antibodies (e.g., anti-NGF antibodies, source: U.S. Pat. App. Pub. No. 2016 / 0017029 and U.S. Pat. Nos. 8,309,088 and 9,353,176), and anti-activin A antibodies.
[0073] In some embodiments, the bispecific antibody is selected from the group consisting of: anti-CD3x anti-CD20 bispecific antibodies (sources: U.S. Pat. No. 9,657,102 and U.S. Patent Application Publication No. 20150266966(A1)), anti-CD3x anti-mucin 16 bispecific antibodies (e.g., anti-CD3x anti-Muc16 bispecific antibodies), and anti-CD3x anti-prostate specific membrane antigen bispecific antibodies (e.g., anti-CD3x anti-PSMA bispecific antibodies). In some embodiments, any protein of interest is selected from the group consisting of abciximab, adalimumab, adalimumab-ato, ado-trastuzumab, alemtuzumab, alirocumab, atezolizumab, avelumab, basiliximab, belimumab, benralizumab, bevacizumab, bezlotoxumab, blinatumomab, brentuximab vedotin, brodatumab, canakinumab, capromab pendetide, certolizumab pegol, cemiplimab, cetuximab, denosumab, dinutuximab, dupilumab, durvalumab, eculizumab, elotuzumab, emicizumab-kxwh, entansinellirocumab, evinacumab, evolocumab, fasinumab, golimumab, guseni, Selected from the group consisting of lucumab, ibritumomab tiuxetan, idarucizumab, infliximab infliximab-abda, infliximab-dive, ipilimumab, ixekizumab, mepolizumab, necitumumab, nesbacumab, nivolumab, obiltoxaximab, obinutuzumab, ocrelizumab, ofatumab, olaratumab, omalizumab, panitumumab, pembrolizumab, pertuzumab, ramuciltunab, ranibizumab, raxibacumab, reslizumab, rinukumab, rituximab, sarilumab, secukinumab, siltuximab, tocilizumab, tocilizumab, trastuzumab, trevoglumab, ustekinumab, and vedolizumab. EXAMPLES
[0074] Example 1: Universal In-Line Protein Concentration Model for UF / DF Applications material and method Spectral data included in the model data collection were from Raman Rxn2 and Rxn4 analyzers utilizing MR-Probe-785 (Kaiser Optical Systems, Inc. Ann Arbor, MI) and Raman RxnProbehead-758 (Kaiser Optical Systems, Inc. Ann Arbor, MI). Additionally, several different optics were used throughout the development based on availability. The operating parameters of the Raman analyzer were set to a scan time of 10 seconds for 6 accumulations repeated 5 times. Peaks in the spectral data were correlated to offline protein concentration measurements using SIMCA 13 (MKS Data Analytic Solutions, Umea, Sweden). In-line measurements were made at various points in the UF / DF unit operation, including primary concentration, diafiltration, and final concentration. Offline protein concentrations were measured using a SoloVPE (C Technologies, Inc.). SoloVPE measurements were made in triplicate.
[0075] The spectral regions used to develop the chemometrics model are shown in Figure 2. These regions include: Region 1: 977–1027 cm -1 (Ring structure), Region 2: 1408-1485 cm -1 (arginine), region 3: 1621-1711 cm -1 (secondary structure), and region 4: 2823–3046 cm -1 (CH stretching) were included. Spectral filtering performed on the raw spectral data was as follows: 21 cm -1 Point smoothed first derivative.
[0076] result With the goal of clarifying the feasibility of a universal in-line protein concentration model for ultrafiltration / diafiltration (UF / DF) applications, mAb1 was analyzed using Raman spectroscopy. Protein concentration measurements were taken before diafiltration (primary concentrate), during diafiltration (diafiltration), and after diafiltration (final concentrate). Concentrations calculated from the model were compared to protein concentrations measured by SoloVPE (Figure 3). When training set data was included (mAb1 spectra were incorporated within the PLS model), the model error for 0-120 g / L (primary concentrate and diafiltration) was 3.1%, and for >120 g / L (final concentrate) was 1.8%.
[0077] Processing errors can be detected by Raman spectroscopy. Air entrainment in the system during final recirculation through the UF / DF system was detected by Raman spectral data as shown in Figure 3. As shown by the rectangular bar, the predicted concentration of mAb1 by SoloVPE was >200 g / L, whereas the Raman prediction was approximately 65 g / L.
[0078] The absolute Raman model errors for 10 representative mAbs are shown in Figure 4. As revealed from this data, model development was successful for the depicted mAbs, including different mAb isotypes (IgG1 and IgG4) and bispecific molecules. 14 out of 17 model predictions met an error of 5% or less. However, specific models (0-120 g / L and >120 g / L) were created for each probe used during development as a probe to demonstrate that variability inflated the error predicted by the PLS model.
[0079] To optimize the model, various probes and lasers were tested in-line with multiple mAbs. Model refinement was performed where the updated universal model was focused on only one spectral region, i.e. 2823-3046 cm. -1(CH stretching), spectral filtering of standard normal variant (SNV) as baseline correction, correction for laser power fluctuations and probe fluctuations. A comparison of the two developed model components is summarized in Table 1. A partial least squares (PLS) regression model was developed using three runs of the corresponding offline SoloVPE measurements. Details of the partial least squares regression model are shown in Table 2. The updated dataset predicted with the optimized laser / probe universal model shows that 15 out of 17 model predictions meet the error of 5% or less compared to 14 out of 17 previous ones (Figure 5).
[0080] (Table 1) Comparison of Universal Model Components TIFF2024164120000002.tif69166
[0081] (Table 2) Protein concentration partial least squares regression universal model (v.2) details TIFF2024164120000003.tif40166R 2 X - the proportion of variation explained by the model, goal: R 2 >0.9 Q 2 - Proportion of variation predicted by the model during cross-validation, goal: Q 2 >0.8 RMSECV: Root mean square error of cross-validation
[0082] Example 2: Scale-up performance of protein concentration model material and method The optimized universal model (v.2) (see Table 1) was tested with mAb10 in a scaled-up 1 / 2 inch single-use tangential flow filtration system experiment from Pall Corporation. The mAb10 loading material was the formulated drug substance rather than the normal processing loading material. The FDS material was diluted into a representative UF / DF loading source containing protein concentration and buffer excipients. However, a mAb10-specific model was created because there are additional excipients present in the loading source that were not tested during the development of the universal model. See Example 1 for information on Raman data collection method and scan length. For the mAb-specific model >120 g / L, the same spectral regions and pre-processing techniques were used as for the universal model (v.2). Meanwhile, for the mAb-specific model 0-120 g / L, the four spectral regions were used as shown in Figure 2 with standard normal variant pre-processing. The difference in the 0-120 g / L model may be due to the additional excipients in the loading source.
[0083] result As shown in Table 3, the model goal of keeping the error below 5% for the 2 / 4 updated models was achieved when the training set was included in the model prediction. A second run was performed based on notable differences during the preliminary experiments at IOPS, such as loading source, laser, and scale (bench scale vs. scale-up). Prior to the second run, modifications were made to the mAb 10 specific model for 0-120 g / L. All four regions were included using SNV pre-processing, but three regions, namely 977-1027 cm -1 , 1408~1485cm -1 , and 1621-1711 cm -1 is 21cm -1A point smoothed first derivative was additionally used. A summary of the compiled experimental results is summarized in Table 3. During the second experiment, the model goal of keeping the error below 5% for the 3 / 4 updated model was achieved when the training set was included in the model prediction, as shown in Table 3. An additional observation made during data analysis was that the main source of error increase was the loading source. Removal of the loading sample reduces the error of the universal model from 8.6% to 5.7%. In Figure 7, the results of the in-line prediction (real-time, bars with wide hatching) and the updated model (bars with narrow hatching) are illustrated for the final concentrated pool, and the protein concentration is compared to the offline measurement of SoloVPE (empty bars). For the final concentrated pool, the error of the in-line and updated models of the universal model was 2.7%, while the in-line and updated models of the mAb 10 model were 12.0% and 4.2%, respectively. The observed increase in error in the mAb 10 model is likely due to the fact that the universal model has a larger data set, whereas the data is limited to the range of approximately 250 g / L. An additional factor contributing to the increased error is the inability of the model to extrapolate outside the characterized range (i.e., above 250 g / L in the mAb 10 model).
[0084] Table 3. Average model error for protein concentration prediction during scale-up TIFF2024164120000004.tif47166 1 : Removal of the loading sample reduces the error: Universal (0-120g / L): 8.6%-5.7% and mAb11 (0-120g / L): 3.5%-2.7%
[0085] Example 3: Scale-up of protein concentration model to pilot processing equipment material and method During process development of a commercially viable process, UF / DF is characterized as a unit operation that follows the principle of quality by design to understand the critical process parameters and critical quality attributes. Raman and model development were included during the development of mAb 11 to improve process understanding and propose a streamlined approach to model development. See Example 1 for information on Raman data collection methods, other than adjusting the scan time length from 10 s to 5 s. The developed mAb11 model used SNV pre-processing for all four spectral regions, but only three regions, namely 977-1027 cm. -1 , 1408~1485cm -1 , and 1621-1711 cm -1 is 21cm -1 A point smoothed first derivative was additionally used.
[0086] result : Bench-scale models were developed using four DoE experiments and four spectral regions. The Raman model error of the real-time predictions from the four DoE experiments is shown in Figure 8A. The Raman model error of 15 additional experiments using the bench-scale model is shown in Figure 8B. mAb-specific protein concentration models were developed from 0-120 g / L and above 120 g / L with an error of 5% or less.
[0087] Using the bench-scale model (n=15), the prediction error for mAb11 from pilot-scale runs (n=3) is between 0.6% and 10.6% (Figure 9A). When the pilot-scale data was incorporated within the bench-scale model, the pilot-scale prediction error decreased to 0.6-2.2% (n=18). The increase in bench-scale model error may be a result of the effect of temperature on Raman spectral shifts due to heat dissipation associated with scale-up equipment. Temperature will be a factor to be considered during future Raman development and model validation is solicited.
[0088] In seven pilot-scale experiments using three different monoclonal antibodies (mAb J, mAb K, and mAb L), the prediction error of the final model ranged from 0.6 to 2.2%, well within the goal of 5% (Figure 9B).
[0089] Example 4: Use of Raman models for real-time concentration measurements to enable processing decisions material and method : For Raman automation, the same protocol was used for Raman spectral collection and modeling. See Example 3 for details. An automated control strategy was developed to use Raman spectral data to achieve the final protein concentration target. A predictive model was created to filter the data, which was used to provide input to the instrumentation on the UF / DF to terminate the unit operation when the protein concentration target was achieved. SoloVPE measurements were performed in triplicate.
[0090] result : Exemplary screen setups for automated monitoring of batch UF / DF and single pass TFF protein concentration are shown in Figures 10A and 10B. As shown in Figure 11, predictive modeling can be used to monitor the concentration of mAb14 in real time throughout the various processing steps, or the concentration unit operation can be triggered to stop when the desired concentration target is achieved. The final concentrate pool had a Raman predicted value of 260 g / L compared to an offline SoloVPE measurement of 262 g / L. This resulted in an error of 0.8%, achieving an error goal of 5% or less. Raman is a suitable application to use in making automated processing decisions that confirm desired protein concentration targets are being met.
[0091] Example 5: Proof of concept for modeling high molecular weight (HMW) species in UF / DF material and method : Spectral data included in the data collection of HMW species were obtained from a Raman Rxn2 analyzer (Kaiser Optical Systems, Inc. Ann Arbor, MI) and a Raman Rxn Probehead-758 (Kaiser Optical Systems, Inc. Ann Arbor, MI). Additionally, several different optics were used throughout the development based on availability. The operating parameters of the Raman analyzer were set to a scan time of 72 seconds for one accumulation, repeated 25 times. In-line measurements were taken at various points in the UF / DF unit operation, including primary concentration, diafiltration, and final concentration. The spectral range was measured from 110 to 3415 cm. -1 The raw spectral data were preprocessed using SNV and -1 Further filtering was performed using a smoothed first derivative. Offline measurements of HMW species were performed using size-exclusion ultra-performance liquid chromatography.
[0092] result : Another attribute considered as a preliminary critical quality characteristic in protein purification is HMW species. Current technology does not allow real-time monitoring of HMW species during processing. As illustrated in FIG. 12A, the Raman modeling method of the present disclosure can be used to monitor HMW species during protein purification. The prediction of HMW species by Raman modeling was compared to measurements collected using SE-UPLC throughout the entire purification (primary concentration, diafiltration, and final concentration). Raman modeling effectively predicted the proportion of high molecular weight species in real time during protein processing. The model was generated with an average error of 3.4%.
[0093] Example 6: Proof of concept for modeling high molecular weight (HMW) species in polishing chromatography material and method : Spectral data included in the data collection of HMW species were from a Raman Rxn2 analyzer (Kaiser Optical Systems, Inc. Ann Arbor, MI) utilizing an MR-Probe-785. The operating parameters of the Raman analyzer were set to a scan time of 10, 30, or 60 seconds for one accumulation with five replicate measurements. Offline measurements were performed using anion exchange chromatography (AEX) pools with total HMW ranging from 6.2% to 76.2%. The spectral ranges used for the modeling and pre-processing techniques employed are listed in Table 4. Offline HMW species measurements were measured using size-exclusion ultra-performance liquid chromatography.
[0094] result : The disclosed Raman modeling method can be used to monitor HMW species during polishing chromatography protein purification. The predictions of HMW species from Raman modeling were compared to measurements collected using SE-UPLC from the generated AEX pool. As summarized in Table 4, the RMSEP of the methods evaluated ranged from 3.2 to 7.6%. In Figure 12B, the summary data set of 6.2% to 19.7% HMW was used to measure the RMSEP from 350 to 3100 cm -1 The spectral regions and the models generated by the SNV preprocessing technique were evaluated. By reducing the HMW range, the RMSEP was reduced to 1.2%, 1.4%, and 2.1% for 10, 30, and 60 seconds, respectively. Based on these results, the HMW content can be determined by Raman of the AEX pool.
[0095] Table 4. HMW model prediction error (RMSEP) for HMW contents from 6.2% to 76.2% TIFF2024164120000005.tif69147
[0096] Example 7: Proof of concept for potency modeling material and method : The training set model was 35 Protein A flow-through samples spiked with FCP (265 g / L) to achieve titers ranging from 0.36 to 9.8 g / L. The model was evaluated against diluted depth filtrate samples with titers ranging from 1.3 to 8.8 g / L. Spectral data included in the titer model data collection was from a Raman Rxn2 (Kaiser Optical Systems, Inc. Ann Arbor, MI) utilizing an MR-Probe-785. Spectral data was generated using an off-line immersion probe with operating parameters set to 20 seconds scan time for 1 accumulation, repeated five times. Spectral ranges were 977–1027, 1408–1485, 1621–1711, and 2823–3046 cm -1 The raw spectral data were preprocessed using SNV and -1 Further filtering was performed using a smoothed first derivative. The model characteristics are shown in Table 5.
[0097] Table 5. Characteristics of antibody titer model TIFF2024164120000006.tif52128
[0098] result : Antibody titer is a process attribute in protein purification that is needed to inform subsequent downstream purification unit operations, including affinity column loading, production consistency, and intermediate volume constraints in the process. Inaccurate column loading can affect subsequent preliminary critical quality attributes, and thus monitoring techniques such as Raman spectroscopy are needed. For monoclonal antibodies, the actual antibody titer and Raman predicted antibody titer are shown in Figure 13. In this experiment, the model error is 26%, which is above the desired goal of 5% or less. By developing a model using longer scan lengths and diluted and undiluted depth filtrates, the model error is reduced.
[0099] Example 8: Raman model for measuring buffer excipients that meets current orthogonal assay error of approximately 10% material and method : Data was collected from previous concentration model development runs of various antibodies. See Example 1 for information on Raman data collection methods. Model components for detecting histidine and arginine in samples are as shown in Table 6. Spectral regions were based on known histidine / arginine peaks (Zhu, et al., Spectrochim Acta A Mol Biomol Spectrosc, 78(3):1187-1195(2011)). Following initial model development of histidine and arginine, further model characterization was performed using mAb 14. Using a non-contact optical probe, the operating parameters of the Raman analyzer were set to a scan time of 20 seconds with 5 accumulations. The spectral range for histidine was 1200-1480 cm -1 , and the spectral range for arginine is 860–1470 cm -1 The raw spectral data for both buffer excipients were preprocessed using SNV and 21 cm -1 Further filtering was performed using a smoothed first derivative. Offline histidine and arginine species measurements were determined using an ultra-performance liquid chromatography (UPLC) amino acid quantification-based method.
[0100] result : To determine whether the disclosed Raman modeling method and system can be used to measure buffer excipients in processed antibody samples, data collected from a previous concentration model development run was analyzed for histidine and arginine. Predictions using Raman modeling were compared to values calculated based on the UPLC-based amino acid method, as shown in FIG. 14A. Preliminary histidine / arginine Raman modeling predictions and average model errors are shown in Table 6. A dot plot for the model of histidine predicted vs. histidine actual for mAb 14 is shown in FIG. 14B. For buffer excipients, the average model error is 8.2%, meeting the goal of 10% or less. The goal of 10% or less is based on the variability of the current UPLC orthogonal assay. A dot plot for the model of arginine predicted vs. arginine actual for mAb 14 is shown in FIG. 14C. The average model error is 2.9%, meeting the goal of 10% or less.
[0101] This data demonstrates that Raman modeling can be used to predict the levels of buffer excipients from UF / DF and FCP materials during manufacturing. Successful quantification of these excipients ensures that the UF / DF provides a final concentrated pool that allows for subsequent formulation.
[0102] Table 6: Universal Concentration Model (Example 1) Processing, model components and data collected for histidine / arginine TIFF2024164120000007.tif54149Histidine range: 0-25mM; Arginine range: 0-81mM SNV - Standard normal variate: centered and normalized mean R 2 - The proportion of variation in the training set explained by the model, R 2 > 0.9 Q 2 - Proportion of the variation in the training set predicted by the model during cross-validation, Q2 > 0.8 (RMSEP) Root Mean Square Error Prediction
[0103] Example 9: Raman model for drug-to-antibody ratio measurement material and method : DAR is a quality attribute monitored during the development of antibody drug conjugates (ADCs), antibody radionuclide conjugates (ARCs), and general protein conjugates (potent steroids, non-cytotoxic payloads, etc.) to ensure consistent product quality and facilitate subsequent labeling with payloads. Raman was evaluated as a technique to monitor DAR levels that can be used as a control strategy for reactions. Two different mAbs in development (mAb1 and mAb3) were evaluated for the feasibility of DAR determination using Raman. The operating parameters of the Raman analyzer were set to a scan time of 10 s with 10 accumulations using a non-contact optical probe. The mAbs were pretreated using SNV and were then tuned to a 21 cm -1 The raw spectral data was further filtered using a smoothed second derivative of -1 A spectral range of 100 nm was used. The model components for determining the DAR in the samples are shown in Table 7. Offline DAR measurements were measured using a UV spectroscopy based method.
[0104] Table 7. Model components for drug-to-antibody ratio measurements TIFF2024164120000008.tif42143
[0105] result : As illustrated in Figures 15A-B, Raman modeling can be used to measure the drug-to-antibody ratio for iPET drug conjugates of mAb 1 and mAb 3. For both models, the root mean square error of cross-validation was 0.6 DAR. The current orthogonal UV-based assay has an associated variability of 0.3 DAR (1 standard deviation). Initial Raman predictions were within 2 standard deviations, suggesting that with further model refinement, the DAR can be successfully predicted by Raman.
[0106] While the foregoing specification has described the invention in relation to certain specific embodiments thereof, and numerous details have been set forth for purposes of illustration, it will be apparent to those skilled in the art that the invention is susceptible to additional embodiments, and that the specific details described herein may be varied considerably without departing from the underlying principles of the invention.
[0107] All references cited herein are incorporated by reference in their entirety. The present invention may be embodied in other specific forms without departing from its spirit or essential attributes. Therefore, reference should be made to the appended claims, rather than the foregoing specification, as indicating the scope of the invention.
Claims
1. 1. A method for producing a concentrated protein purification intermediate, comprising: concentrating the protein purification intermediate and simultaneously measuring the concentration of said protein purification intermediate in real time using in situ Raman spectroscopy; adjusting parameters of the enrichment step in real time to obtain the enriched protein purification intermediate and / or the final enriched pool; The method comprising:
2. 2. The method of claim 1, wherein the concentrated protein product has a concentration that is between 5 mg / mL and 300 mg / mL.
3. 3. The method of claim 1, wherein the concentration of the protein purification intermediate is at least 50 mg / mL.
4. 3. The method of claim 1, wherein the concentration of the protein purification intermediate is at least 150 mg / mL.
5. 5. The method of any one of claims 1 to 4, wherein the protein purification intermediate is concentrated using ultrafiltration, buffer exchange, or both.
6. The method of any one of claims 1 to 5, wherein the protein purification intermediate is harvested from a bioreactor, a fed-batch culture, or a continuous culture.
7. The method of any one of claims 1 to 6, wherein measuring the concentration of the protein purification intermediate is performed continuously or intermittently in real time.
8. The method of any one of claims 1 to 6, wherein the quantification of protein concentration is performed at intervals of from 30 seconds to 10 minutes, hourly or daily.
9. The method of any one of claims 1 to 8, wherein the protein purification intermediate is an antibody or antigen-binding fragment thereof, a fusion protein, or a recombinant protein.
10. Spectral data is 977-1027 cm -1 , 1408-1485cm -1 , 1621-1711cm -1 , 2823-3046cm -1 10. The method of claim 1, wherein the signal is collected at one or more wavenumber ranges selected from the group consisting of:
11. 1. A method for producing a protein purification intermediate, comprising: performing Raman spectroscopy independently on a plurality of protein purification intermediates to generate a universal model capable of quantifying any one of the plurality of protein purification intermediates; measuring the concentration of a protein purification intermediate using in situ Raman spectroscopy with the universal model during concentration of the protein purification intermediate; generating the concentrated protein purification intermediate when the concentrated protein purification intermediate reaches a predetermined concentration or final concentrated pool target; The method comprising:
12. The method of claim 11 , wherein the model is generated using partial least squares regression analysis of the raw spectral data and offline protein concentration data.
13. The method of claim 12 , further comprising performing a normalization technique or point smoothing on the Raman spectroscopy data.
14. The method of claim 13 , wherein the normalization technique comprises a standard normal variate.
15. The point smoothing is 21 cm -1 The method of claim 13 including smoothing first derivatives.
16. 16. The method of any one of claims 11 to 15, wherein the model provides predicted values of protein concentration with an error of 5% or less compared to offline protein concentration values.
17. 16. The method of any one of claims 11 to 15, wherein the model provides predicted protein concentrations with an error of no more than 3% compared to offline protein concentration values.
18. 12. The method of claim 11, wherein the concentrated protein purification intermediate has a concentration that is between 5 mg / mL and 300 mg / mL.
19. 12. The method of claim 11, wherein the concentration of the concentrated protein purification intermediate is at least 100 mg / mL.
20. 12. The method of claim 11, wherein the concentration of the concentrated protein purification intermediate is at least 150 mg / mL.
21. 12. The method of claim 11, wherein the concentrated protein purification intermediate is concentrated using ultrafiltration, diafiltration, or both.
22. 12. The method of claim 11, wherein the concentrated protein purification intermediate is harvested from a bioreactor, a fed-batch culture, or a continuous culture.
23. The method of any one of claims 1 to 22, wherein measuring the concentration of the protein purification intermediate is performed continuously or intermittently in real time.
24. The method of any one of claims 1 to 23, wherein the quantification of protein concentration is performed at intervals of from 30 seconds to 10 minutes, hourly or daily.
25. The method of any one of claims 11 to 24, wherein the protein purification intermediate is an antibody or antigen-binding fragment thereof, a fusion protein, or a recombinant protein.
26. A protein purification intermediate produced according to any one of claims 1 to 25.
27. 1. A method for monitoring and controlling a critical quality attribute of a protein purification intermediate during downstream protein purification processing, comprising: Quantifying one or more critical quality attributes of said protein purification intermediate using in situ Raman spectroscopy; adjusting the one or more critical quality attributes of the protein purification intermediate to conform to a predetermined critical quality attribute level; The method comprising:
28. 28. The method of claim 27, wherein the critical quality characteristic is selected from the group consisting of antibody titer, protein concentration, high molecular weight species, drug to antibody ratio, and buffer excipients.
29. 29. The method of any one of claims 27 or 28, wherein the downstream protein purification processing is ultrafiltration / diafiltration.
30. 1. A method for monitoring and controlling excipient levels in cell culture fluid and / or protein purification intermediates harvested during downstream purification, comprising: purifying the cell culture fluid or protein purification intermediate while measuring the concentration of the excipient in real time using in situ Raman spectroscopy; adjusting parameters of said purification step in real time to obtain or maintain a predetermined amount of said excipient in said harvested cell culture fluid and / or protein purification intermediate; The method comprising:
31. 31. The method of claim 30, wherein the excipient comprises a buffer excipient.
32. 32. The method of any one of claims 30 or 31, wherein the excipient is selected from the group consisting of acetate, citrate, histidine, succinate, phosphate, hydroxymethylaminomethane (Tris), proline, arginine, sucrose, or a combination thereof.
33. 31. The method of claim 30, wherein the excipient comprises a surfactant excipient.
34. 34. The method of any one of claims 30 or 33, wherein the surfactant excipient is selected from the group consisting of polysorbate 80, polysorbate 20, and poloxamer 188.
35. The method of claim 11 , wherein the excipient comprises polyethylene glycol.
Citation Information
Patent Citations
Antibody purification using pseudo-mobile bed chromatography
JP2013537235A
Raman spectroscopy for bioprocess operations
JP2013541711A
Analysis method and analysis device
JP2016156807A
