Cell culture method and system

Abstract

Provided is a culture method for maintaining the productivity of a product by dynamically controlling cell discharge in order to appropriately control viable cell density. Provided is a method for culturing cells using a perfusion bioreactor, the method including: measuring the capacitance of living cells in a culture medium in a perfusion bioreactor; and discharging live cells in the culture medium in the perfusion bioreactor on the basis of a set value of the capacitance. The set value for the capacitance is changed at a predetermined time interval, and the change is performed on the basis of the numerical value of a viable cell density calculated on the basis of a parameter other than the capacitance.

Description

Cell culture method and system

[0001] The present invention relates to a method and system for cell culture using a perfusion bioreactor.

[0002] There is a need for more efficient production of biopharmaceuticals that use therapeutic recombinant proteins as active ingredients. Many studies have been conducted on improving cell culture methods for producing recombinant proteins or other biological products (Patent Documents 1 to 4, Non-Patent Documents 1 to 7).

[0003] Cell culture methods are generally classified into batch, fed-batch, and continuous culture. Perfusion culture is a type of continuous culture, and control of parameters such as medium supply, recovery, and cell discharge rate affects production efficiency and product quality. In perfusion culture, a widely used method for appropriately controlling viable cell density (VCD) is the use of a capacitance probe to control cell discharge, which exploits the relationship between capacitance and viable cells (Patent Document 1, Non-Patent Documents 1 and 5). This goal is achieved by setting an arbitrary capacitance set point and controlling the discharge of cell-containing culture medium according to the capacitance value measured by the capacitance probe. The capacitance set point is often determined based on viable cell density, which is easy to measure. However, it has been reported that, depending on the cell state, capacitance values ​​are more strongly correlated with cell mass or viable cell volume than with viable cell density (Non-Patent Document 8). Therefore, when determining the capacitance set point based on viable cell density, it becomes necessary to adjust the capacitance set point depending on the cell state. As a means to avoid this problem, a method has been developed in which an estimated value of viable cell density obtained by Raman spectroscopy is directly used to control cell discharge without using a capacitance probe, but this method has the problem of not being able to perform real-time control like when using a capacitance probe (Patent Documents 2 to 4, Non-Patent Document 9). Although there have been reports on replacing capacitance measurement with Raman spectroscopy and on automatic feed addition methods (Non-Patent Document 10), these do not solve the problems associated with using Raman spectroscopy.

[0004] Special Publication No. 2020-533983 International Publication No. 2020 / 238918 Special Publication No. 2020-536521 Special Publication No. 2020-536497 Special Publication No. 2022-537650

[0005] Jean-Marc Bielser et al., Biotechnology Advances, 36, 1328-1340, 2018Jonathan Coffman et al., Biotechnol. Bioeng., 118, 1735-1749, 2021Cary Matanguihan and Paul Wu, Curr Opin Biotechnol., 78, 102828, 2022Adam Bergin et al., Biotechnology Advances, 61, 108048, 2022Izabella Surowiec and Jochen Scholz, Current Opinion in Biotechnology, 83, 102979, 2023Wlaschin, KF et al., Cell Culture Engineering, 101:43-74 (2006).Jason E. Dowd et al., Cytotechnology 42: 35-45, 2003.S. Metze et al., Bioprocess and Biosystems Engineering 43, 193-205 , 2020G. Chen et al., Biochemical Engineering Journal, 172, 108063, 2021Carl Raferty et al., Bioprocess and Biosystems Engineering (2020) 43: 1415-1429. Kompala et al., Cell Culture Technology for Pharmaceutical and Cell-Based Therapies, (2006), Taylor & Francis Group, LLC, pages 387-416

[0006] In one aspect of the present invention, a culture method is provided for dynamically controlling cell discharge to maintain product productivity in order to appropriately control viable cell density (VCD) in a perfusion culture system.

[0007] The present inventors have found that it is possible to maintain and culture a desired viable cell density by successively calculating the capacitance value to be set based on an estimated viable cell density obtained by Raman spectroscopy, and have completed the present invention.

[0008] [1-1] A method for culturing cells using a perfusion bioreactor, comprising: continuously measuring the capacitance of a culture medium containing live cells in the perfusion bioreactor; and, when the measured capacitance exceeds a set capacitance value, discharging the culture medium containing live cells in the perfusion bioreactor so that the measured capacitance approaches the set capacitance value; the set capacitance value is changed at predetermined time intervals; the change is made based on a numerical value of live cell density calculated based on a parameter other than capacitance; and the set capacitance value is a capacitance corresponding to a target live cell density.

[0009] [1-2] The method according to [1-1], wherein the parameter other than capacitance is a measurement value obtained by a spectroscopic method.

[0010] [1-3] The method according to [1-2], wherein the capacitance is measured using a capacitance probe, and the parameters other than the capacitance are measured using a probe used in a spectroscopic technique.

[0011] [1-4] The method according to [1-3], wherein the spectroscopic technique includes Raman spectroscopy, near-infrared spectroscopy, and infrared spectroscopy.

[0012] [1-5] The method according to [1-3], in which the initial capacitance setting value is set based on an offline measurement value or a measurement value of a parameter other than capacitance.

[0013] [1-6] The method according to any one of [1-1] to [1-5], wherein the calculation of the viable cell density is carried out based on a parameter other than capacitance by creating a regression model whose coefficient of variation with respect to offline measured values ​​of viable cell density is within 20%.

[0014] [1-7] The method according to any one of [1-1] to [1-6], wherein the change is performed by obtaining a coefficient by dividing the value of the viable cell density calculated based on the parameter other than the capacitance by the measured value of the capacitance at the time the parameter other than the capacitance is measured, and by dividing the target viable cell density by the coefficient to determine the set value of the capacitance.

[0015] [1-8] The method according to [1-3], wherein the probe used in the spectroscopic technique is configured to collect Raman spectra in the culture medium in the perfusion bioreactor.

[0016] [1-9] The method according to [1-8], wherein a value of viable cell density is calculated based on the Raman spectrum, and the value of viable cell density is used to change the set value of capacitance.

[0017] [1-10] The method according to [1-9], wherein the change is performed by dividing the value of the viable cell density calculated based on the Raman spectrum by the measured capacitance value at the time the Raman spectrum was collected to obtain a coefficient, and by dividing the target viable cell density by the coefficient to determine the set value of the capacitance.

[0018] [1-11] The method according to any one of [1-8] to [1-10], wherein only Raman spectra are used to calculate viable cell density.

[0019] [1-12] The method according to any one of [1-1] to [1-11], wherein the set value of the capacitance is used to determine the time when a target cell density is reached.

[0020] [1-13] The target cell density is 30 × 10 ６ cells / mL or more, 40×10 ６ cells / mL or more, 50×10 ６ cells / mL or more, 60×10 ６ cells / mL or more, or 70 x 10 ６ cells / mL or more, or 200 x 10 ６ cells / mL or less, 190×10 ６ cells / mL or less, 180×10 ６ cells / mL or less, 170×10６ cells / mL or less, or 160 x 10 ６ The method according to any one of [1-1] to [1-12], wherein the concentration is 0.01 to 0.01 cells / mL or less.

[0021] [1-14] The method according to any one of [1-1] to [1-13], wherein the cells to be cultured are animal cells, plant cells, bacterial cells, or fungal cells, preferably animal cells.

[0022] [1-15] The method according to [1-14], wherein the cells to be cultured are mammalian cells, preferably Chinese hamster ovary (CHO) cells.

[0023] [1-16] The method according to any one of [1-1] to [1-15], wherein the cultured cells produce a desired product such as a monoclonal antibody or a recombinant protein.

[0024] [1-17] The method according to [1-16], wherein the cultured cells produce a monoclonal antibody.

[0025] [1-18] The method according to any one of [1-1] to [1-17], further comprising, when calculating the value of the viable cell density based on a parameter other than the capacitance, associating the parameter other than the capacitance with the viable cell density by machine learning using the parameter other than the capacitance.

[0026] [2-1] A system for culturing cells by perfusion culture, comprising: a perfusion bioreactor including: an internal chamber containing culture medium and viable cells; a port for supplying nutrients from a supply reservoir; a port for automatic cell discharge from the internal chamber; and a port for continuously collecting a culture supernatant containing a target product from the internal chamber; and a measuring device including: one or more capacitance probes immersed in the culture medium in the perfusion bioreactor for measuring capacitance; one or more probes immersed in the culture medium in the perfusion bioreactor for measuring parameters other than capacitance; a calculation unit that receives the measured capacitance values ​​and the parameters other than capacitance and is provided with a target viable cell density, and calculates and changes the set value of capacitance; and a control unit that controls the automatic cell discharge from the internal chamber based on the set value of capacitance and the measured capacitance values ​​obtained by the capacitance probe.

[0027] [2-2] The system described in [2-1], wherein the parameter other than the capacitance is a measurement value obtained by a spectroscopic technique.

[0028] [2-3] The system according to [2-1] or [2-2], wherein one or more probes for measuring parameters other than capacitance are probes used in spectroscopic techniques.

[0029] [2-4] The system described in [2-3], wherein the spectroscopic techniques include Raman spectroscopy, near-infrared spectroscopy, and infrared spectroscopy.

[0030] [2-5] A system according to [2-3], in which measurements of parameters other than capacitance are performed at predetermined time intervals.

[0031] [2-6] A system according to any one of [2-1] to [2-5], in which capacitance measurement is performed continuously.

[0032] [2-7] A system according to any one of [2-1] to [2-6], which calculates the viable cell density based on parameters other than capacitance by creating a regression model whose coefficient of variation with the offline measured value of viable cell density is within 20%.

[0033] [2-8] A system described in any one of [2-1] to [2-7], wherein the setting value of the capacitance is changed by dividing the numerical value of the viable cell density by the measured value of the capacitance at the time when a parameter other than the capacitance is measured to obtain a coefficient, and by dividing the target viable cell density by the coefficient to determine the setting value of the capacitance.

[0034] [2-9] The system described in [2-3], wherein the probe used in the spectroscopic technique is configured to collect Raman spectra in the culture medium within the bioreactor.

[0035] [2-10] The system described in [2-9], which calculates a value of viable cell density based on the Raman spectrum and uses the value of viable cell density to change the capacitance setting value.

[0036] [2-11] The system described in [2-10], wherein the change is performed by dividing the numerical value of the viable cell density by the measured capacitance at the time the Raman spectrum is collected to obtain a coefficient, and dividing the target viable cell density by the coefficient to determine the set value of the capacitance.

[0037] [2-12] A system described in any of [2-1] to [2-11], wherein the measuring device further includes a laser emission module that periodically emits a laser with a constant wavelength of 795 nm excitation wavelength into the bioreactor at intervals of 1 minute to 24 hours.

[0038] [2-13] A system described in any of [2-9] to [2-12], further including a cover configured to cover the bioreactor while the laser emission module of the measurement device emits a laser into the perfusion bioreactor at predetermined time intervals and a probe used in a spectroscopic technique collects a Raman spectrum by detecting the intensity of scattered light within the perfusion bioreactor.

[0039] [2-14] The method according to any one of [2-9] to [2-13], wherein only Raman spectra are used to calculate viable cell density.

[0040] [2-15] A system described in any of [2-1] to [2-14], further including, when calculating the value of viable cell density based on parameters other than capacitance, associating parameters other than capacitance with viable cell density by machine learning using parameters other than capacitance.

[0041] [3-1] A method for controlling the viable cell density in a medium during cell culture using a perfusion bioreactor, comprising: i) continuously measuring the capacitance of a medium containing viable cells in the perfusion bioreactor to confirm a correlation between the measured capacitance value and the viable cell density; ii) calculating a capacitance corresponding to a predetermined viable cell density and setting the calculated capacitance as a set capacitance value; and iii) discharging the viable cell-containing medium in the perfusion bioreactor based on a comparison between the set capacitance value and the measured capacitance value, wherein the set capacitance value is changed at predetermined time intervals, and the change is made so as to correspond to a viable cell density value calculated based on a parameter other than capacitance.

[0042] [3-2] The method according to [3-1], wherein the parameter other than capacitance is a measurement value obtained by a spectroscopic method.

[0043] [3-3] The method according to [3-2], in which the capacitance is measured using a capacitance probe and the parameters other than the capacitance are measured using a probe used in a spectroscopic technique.

[0044] [3-4] The method according to [3-3], wherein the spectroscopic technique includes Raman spectroscopy, near-infrared spectroscopy, and infrared spectroscopy.

[0045] [3-5] The method described in [3-3], in which the initial capacitance setting value is set based on an offline measurement value or a measurement value of a parameter other than capacitance.

[0046] [3-6] The method according to any one of [3-1] to [3-5], wherein the calculation of the viable cell density is carried out based on a parameter other than capacitance by creating a regression model whose coefficient of variation with respect to offline measured values ​​of viable cell density is within 20%.

[0047] [3-7] The method according to any one of [3-1] to [3-6], wherein the change is performed by dividing the numerical value of the viable cell density by the measured value of the capacitance at the time when a parameter other than the capacitance is measured to obtain a coefficient, and by dividing the target viable cell density by the coefficient to determine the set value of the capacitance.

[0048] [3-8] The method described in [3-3], wherein the probe used in the spectroscopic technique is configured to collect Raman spectra in the culture medium within the perfusion bioreactor.

[0049] [3-9] The method according to [3-8], wherein a value of viable cell density is calculated based on the Raman spectrum, and the value of viable cell density is used to change the set value of capacitance.

[0050] [3-10] The method according to [3-9], wherein the change is performed by dividing the numerical value of the viable cell density by the measured capacitance at the time the Raman spectrum is collected to obtain a coefficient, and by dividing the target viable cell density by the coefficient to determine the set value of the capacitance.

[0051] [3-11] The method according to [3-8] or [3-9], in which only Raman spectra are used to calculate viable cell density.

[0052] [3-12] The method according to any one of [3-1] to [3-11], wherein the set value of the capacitance is used to determine the time when a target cell density is reached.

[0053] [3-13] The target cell density is 30 × 10 ６ cells / mL or more, 40×10 ６ cells / mL or more, 50×10 ６ cells / mL or more, 60×10 ６ cells / mL or more, or 70 x 10 ６ cells / mL or more, or 200 x 10 ６ cells / mL or less, 190×10 ６ cells / mL or less, 180×10 ６ cells / mL or less, 170×10 ６ cells / mL or less, or 160 x 10 ６The method according to any one of [3-1] to [3-12], wherein the total amount of the antibody is 0.01 to 0.01 cells / mL or less.

[0054] [3-14] The method according to any one of [3-1] to [3-13], wherein the cells to be cultured are animal cells, plant cells, bacterial cells, or fungal cells, preferably animal cells.

[0055] [3-15] The method according to [3-14], wherein the cells to be cultured are mammalian cells, preferably Chinese hamster ovary (CHO) cells.

[0056] [3-16] The method according to any one of [3-1] to [3-15], wherein the cultured cells produce a desired product such as a monoclonal antibody or a recombinant protein.

[0057] [3-17] The method according to [3-16], wherein the cultured cells produce a monoclonal antibody.

[0058] [3-18] The method according to any one of [3-1] to [3-17], further comprising, when calculating the value of the viable cell density based on a parameter other than the capacitance, associating the parameter other than the capacitance with the viable cell density by machine learning using the parameter other than the capacitance.

[0059] [4-1] A method for culturing cells using a perfusion bioreactor, comprising: i) continuously measuring the capacitance of a medium containing live cells in the perfusion bioreactor to confirm a correlation between the measured capacitance and the live cell density; ii) calculating a capacitance corresponding to a predetermined live cell density and setting the calculated capacitance as a set capacitance value; and iii) controlling the discharge of the medium containing live cells from the perfusion bioreactor based on a comparison between the set capacitance value and the measured capacitance value, wherein the set capacitance value is changed at predetermined time intervals so as to correspond to a value of the live cell density calculated based on a parameter other than the capacitance.

[0060] [4-2] The method according to [4-1], wherein the parameter other than capacitance is a measurement value obtained by Raman spectroscopy.

[0061] [4-3] The method according to [4-2], in which the capacitance is measured using a capacitance probe and the parameters other than the capacitance are measured using a Raman probe.

[0062] [4-4] The method according to any one of [4-1] to [4-3], wherein the predetermined time is a time in the range of 1 minute to 24 hours.

[0063] [4-5] The method according to any one of [4-1] to [4-4], wherein the calculation of the viable cell density is carried out based on a parameter other than capacitance by creating a regression model whose coefficient of variation with respect to offline measured values ​​of viable cell density is within 20%.

[0064] [4-6] The method according to any one of [4-1] to [4-5], wherein the capacitance measurement value is used to measure the time point at which a target cell density is reached.

[0065] [4-7] The target cell density is 30 × 10 ６ cells / mL or more, 40×10 ６ cells / mL or more, 50×10 ６ cells / mL or more, 60×10 ６ cells / mL or more, or 70 x 10 ６ cells / mL or more, or 200 x 10 ６ cells / mL or less, 190×10 ６ cells / mL or less, 180×10 ６ cells / mL or less, 170×10 ６ cells / mL or less, or 160 x 10 ６ The method according to any one of [4-1] to [4-6], wherein the total amount of the antibody is 0.01% or less than 0.01%.

[0066] [4-8] The method according to any one of [4-1] to [4-7], wherein the cells to be cultured are animal cells, plant cells, bacterial cells, or fungal cells, preferably animal cells.

[0067] [4-9] The method according to [4-8], wherein the cells to be cultured are mammalian cells, preferably Chinese hamster ovary (CHO) cells.

[0068] [4-10] The method according to any one of [4-1] to [4-9], wherein the cultured cells produce a desired product such as a monoclonal antibody or a recombinant protein.

[0069] [4-11] A method according to any one of [4-1] to [4-10], further comprising, when calculating the value of the viable cell density based on a parameter other than the capacitance, associating the parameter other than the capacitance with the viable cell density by machine learning using the parameter other than the capacitance.

[0070] [5-1] A method for producing a protein, comprising culturing cells in a medium in a perfusion bioreactor by the method according to any one of [1-1] to [1-17] and [4-1] to [4-11], wherein the cells are cells that secrete a protein of interest into the medium, and obtaining the protein of interest from the medium recovered from the perfusion bioreactor.

[0071] [5-2] The method according to [5-1], wherein the target protein is an antibody.

[0072] FIG. 1 is a schematic diagram of a perfusion bioreactor according to one embodiment of the present invention. FIG. 2 is a graph based on data values ​​before and after processing of historical Raman spectral data used in a machine learning model to make it usable data. FIG. 3 is a graph of capacitance values ​​and viable cell density under conditions in which the capacitance setpoint was corrected based on offline measurements. FIG. 4 is a graph of capacitance values ​​and viable cell density under conditions in which the capacitance setpoint was corrected based on offline measurements and under conditions in which the capacitance setpoint was kept constant. FIG. 5 is a schematic diagram of a perfusion bioreactor according to one embodiment of the present invention. FIG. 6 is a graph of capacitance values ​​and viable cell density under conditions in which the capacitance setpoint was corrected based on offline measurements and under conditions in which automatic capacitance setpoint correction was performed using Raman analysis. FIG. 7 is a graph of inline capacitance measurements. FIG. 8 is a graph of viable cell density calculated based on inline Raman spectral measurements. FIG. 9 is a graph of capacitance values ​​and viable cell density under conditions in which the capacitance setpoint was corrected based on offline measurements and under conditions in which automatic capacitance setpoint correction was performed using Raman analysis. Figure 10 is a graph of in-line capacitance measurements, and Figure 11 is a graph of calculated viable cell density based on in-line Raman spectroscopy measurements.

[0073] Hereinafter, embodiments of the present invention will be described in detail.

[0074] In one aspect of the present invention, a method for culturing cells using a perfusion bioreactor is characterized in that, when discharging live cells, a set value calculated based on capacitance measurement is changed based on a parameter other than capacitance.

[0075] As used herein, "perfusion" refers to the addition and simultaneous removal of medium from a reactor. As used herein, "perfusion culture" refers to continuous cell culture performed by perfusing medium while retaining cells in the reactor. In one aspect of the present invention, perfusion constantly supplies new nutrients to the reactor and constantly removes components secreted by the cells. In one aspect of the present invention, perfusion culture is performed at a higher cell density than batch or fed-batch culture in conventional bioreactors. Thus, perfusion culture can achieve high productivity per volume. Components secreted by the cells include the protein product of interest and cell-derived waste products. In one aspect of the present invention, the protein product of interest, also referred to as product, is continuously recovered while retaining cells in the reactor by, for example, filtration, alternating tangential flow (ATF), cell sedimentation, ultrasonic separation, hydrocyclone, or any other method known to those skilled in the art or described in non-patent document 11. In one embodiment, perfusion culture is performed under uniform culture conditions, for example, by culturing mammalian cells in suspension culture. In another embodiment, perfusion culture is performed by culturing mammalian cells under heterogeneous culture conditions, for example, by attaching the cells to the surface of a culture vessel. In one aspect of the present invention, cells are also removed when the medium is withdrawn. In one embodiment, the harvest rate and cell withdrawal (liquid removal, cell removal) are performed so that the working volume in the bioreactor is maintained constant, taking into account the perfusion rate.

[0076] As used herein, a "perfusion bioreactor" is characterized by the ability to continuously remove medium from a cell culture and replace it with fresh medium. Furthermore, a perfusion bioreactor can provide the cells in a cell culture with the nutrients necessary to achieve a high cell concentration by constantly adding fresh medium while removing waste products. In one aspect of the present invention, perfusion can be continuous, stepwise, intermittent, or a combination of any or all of these.

[0077] As used herein, "viable cell density (VCD)" refers to the number of viable cells in a given volume of culture medium as determined by a standard viability assay (e.g., trypan blue exclusion). Viable cell density determination can be performed using online or offline sampling. In one embodiment of the present invention, viable cell density is calculated by Raman spectroscopy. Machine learning may be used for Raman spectroscopy calculations. Viable cell density can also be measured offline, for example, using a cell density measuring device. Viable cell density control can be achieved by pumping out the culture volume when the viable cell density exceeds a predetermined level. In one embodiment of the present invention, the predetermined level can be a capacitance setpoint. Viable cell density control can also be achieved by automatic discharge of viable cells.

[0078] Generally, viable cell density is one of the key measurements used to regulate perfusion cultures. Depending on the method of viable cell density measurement, viable cell density can be adjusted daily, hourly, or in real time. Several in-line probes have been developed for estimating viable cell density and are known to those skilled in the art. Examples include capacitance probes, such as the Incyte probe (Hamilton) or the Futura probe (Aber Instruments).

[0079] In one aspect of the present invention, a capacitance probe uses capacitance to measure the density of live cells in a culture. The capacitance probe can measure the charge from live cells in the culture. Specifically, as used herein, a capacitance probe refers to a probe that uses capacitance to measure the density of live cells in a culture. Live cells act as capacitors in a residual electric field. The capacitance probe can measure and report the charge from these cells.

[0080] In one embodiment of the present invention, capacitance is constantly measured. In this specification, the value obtained by the measurement is referred to as a "measured capacitance value." In one embodiment of the present invention, the measured capacitance value is obtained in real time.

[0081] In one embodiment of the present invention, a "capacitance set point" can be determined and used to control the automatic ejection of viable cells. The capacitance set point is determined from capacitance measurements and a viable cell density obtained offline or based on parameters other than capacitance. In one embodiment of the present invention, the capacitance set point corresponds to a target viable cell density.

[0082] In one aspect of the present invention, the set value of the capacitance may be determined from the cell mass or cell volume obtained from parameters other than the capacitance, using the cell mass or cell volume as an indicator in addition to the viable cell density.

[0083] In one embodiment of the present invention, parameters other than capacitance include values ​​of viable cell density calculated based on spectroscopic techniques, such as Raman spectroscopy, and these parameters can be used to determine the capacitance setting and remove excess cells from the bioreactor, thereby controlling the viable cell density to a desired target cell density.

[0084] In one embodiment of the present invention, a dip tube and a peristaltic pump with a defined flow rate can be used to remove viable cells from a bioreactor. The tube should be the correct size; too narrow a tube will tend to cause cell clumping and clogging, while too large a tube may cause cells to settle. In one embodiment of the present invention, viable cell removal is achieved by automatically discharging viable cells in the medium within the perfusion bioreactor when the measured capacitance exceeds a set capacitance value. Automatic discharging is also commonly referred to as bleeding.

[0085] In one aspect of the invention, viable cells are removed at a rate that achieves or maintains a target cell density.

[0086] In one aspect of the present invention, the target cell density is 30 x 10 ６ cells / mL or more, 40×10 ６ cells / mL or more, 50×10 ６ cells / mL or more, 60×10 ６ cells / mL or more, or 70 x 10 ６ cells / mL or more, or 200 x 10６ cells / mL or less, 190×10 ６ cells / mL or less, 180×10 ６ cells / mL or less, 170×10 ６ cells / mL or less, or 160 x 10 ６ cells / mL or less.

[0087] In one aspect of the invention, physical parameters such as pH, dissolved oxygen, and temperature in the perfusion bioreactor are monitored online and controlled in real time or at regular intervals.

[0088] In one embodiment of the invention, the dissolved oxygen is between 10% and 100%, between 20% and 90%, or between 30% and 80%.

[0089] In one aspect of the invention, the agitation speed in the perfusion bioreactor is 50 to 600 rpm, 100 to 500 rpm, or 200 to 400 rpm.

[0090] In one aspect of the present invention, the capacitance probe can be used to continuously measure the capacitance of living cells. As used herein, "continuously" means to measure at all times.

[0091] In one embodiment of the present invention, the parameters other than the capacitance include spectra obtained by Raman spectroscopy, near-infrared spectroscopy, and infrared spectroscopy.

[0092] In one aspect of the present invention, Raman spectroscopy is a form of vibrational spectroscopy that provides information about molecular vibrations, which can be used by inserting a Raman probe in-line for sample identification and quantification. In some embodiments of the present invention, biochemical indicators are monitored using in-line Raman spectroscopy. In-line Raman spectroscopy is a method of analyzing samples in situ without the need to extract a portion of the sample for analysis by a Raman spectrometer. In-line Raman spectroscopy is advantageous in that the Raman spectrometer is non-invasive, thereby reducing the risk of contamination, and is non-destructive, without affecting the viability or protein quality of the cell culture. In-line Raman spectroscopy can provide real-time assessment of one or more biochemical indicators in a cell culture.

[0093] In one aspect of the present invention, the Raman probe can measure Raman spectra at regular intervals over a predetermined period of time. Based on the measured Raman spectra, an estimated value of viable cell density is calculated over a predetermined period of time based on a pre-prepared machine learning model. In one aspect of the present invention, the estimated value is calculated by creating a regression model with a coefficient of variation of within 15%, 20%, 30%, or preferably within 20% from the offline measured value of viable cell density. In one aspect of the present invention, the set capacitance is changed over a predetermined period of time by dividing the estimated value of viable cell density calculated based on the Raman spectroscopy measurement by the capacitance to calculate the viable cell density per capacitance. Any coefficient or function may be used in the calculation formula. In near-infrared spectroscopy and infrared spectroscopy, spectra are measured in the same manner as in Raman spectroscopy, and an estimated value of viable cell density is calculated over a predetermined period of time based on the measured spectrum and a pre-prepared machine learning model.

[0094] In one embodiment of the present invention, the perfusion rate and automatic cell discharge are adjusted in consideration of the supply of nutrients required by living cells in the reactor and the production efficiency of the target protein recovered from the discharged medium. In one embodiment of the present invention, the perfusion rate and automatic cell discharge are adjusted in consideration of, for example, the pH of the medium, pCO ２ , or pO２ The concentration of the medium is controlled based on the measured values ​​of the concentration of the target protein, or the measured values ​​of the concentrations of Na+ ions, K+ ions, glucose, glutamine, glutamic acid, lactate, or ammonium ions contained in the medium.

[0095] In one aspect of the present invention, pH, pCO ２ , or pO ２ The concentrations of Na+ ions, K+ ions, glucose, glutamine, glutamic acid, lactate, or ammonium ions contained in the medium; or the concentration of the target protein are measured by a general method in the present invention.

[0096] Generally, "cell," "cell line," and "cell culture" are used interchangeably, and all such designations herein include cellular progeny. For example, cells "derived from" CHO cells are cellular progeny of Chinese hamster ovary cells, which may have been removed from the original primary cellular parent by any number of generations and may include transformed progeny. Transformants and transformed cells include the primary subject cell and cultures derived therefrom, regardless of the number of times of introduction. It is also understood that all progeny may not be precisely identical in DNA content, due to deliberate or inadvertent mutations. Mutant progeny that have the same function or biological activity as screened for in the originally transformed cell are included.

[0097] In one embodiment of the present invention, the cultured cells can be animal cells, plant cells, bacterial cells, or fungal cells, preferably animal cells, such as mammalian cells, for example, Chinese hamster ovary (CHO) cells. Also, in one embodiment of the present invention, the cultured cells can be cells that produce a desired product, such as a monoclonal antibody or a recombinant protein.

[0098] In one aspect of the present invention, the culture method is not particularly limited and can be used to culture various cells (e.g., bacterial cells, fungal cells, insect cells, plant cells, animal cells, etc.). For example, it is possible to culture COS cells or CHO cells into which a gene encoding a desired protein has been inserted by genetic engineering, or fusion cells such as mouse-human, mouse-mouse, and mouse-rat hybridomas that produce antibodies. The method in one aspect of the present invention can also be used to culture animal cells to obtain native proteins produced by the animal cells, and can be used to culture BHK cells, HeLa cells, etc. in addition to the above-mentioned cells.

[0099] Among the cells, particularly preferred animal cells are CHO cells transfected with a gene encoding a desired protein. The desired protein is not particularly limited and may be any protein, such as an antibody (natural antibody, minibody, chimeric antibody, human antibody, etc.) or a physiologically active protein (granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-SCF), erythropoietin, interferon, interleukins such as IL-1 and IL-6, t-PA, urokinase, serum albumin, blood coagulation factors, etc.), with antibodies being particularly preferred.

[0100] Vectors can be introduced into host cells by, for example, the calcium phosphate method, the DEAE-dextran method, a method using cationic liposome DOTAP (Boehringer Mannheim), electroporation, lipofection, or the like.

[0101] Antibodies produced by the production method in one embodiment of the present invention include not only monoclonal antibodies derived from animals such as humans, mice, rats, hamsters, rabbits, and monkeys, but also artificially modified recombinant antibodies such as chimeric antibodies, humanized antibodies, and bispecific antibodies. Furthermore, the immunoglobulin of the antibody is not particularly limited and may be any class, such as IgG (e.g., IgG1, IgG2, IgG3, and IgG4), IgA, IgD, IgE, or IgM, although IgG and IgM are preferred for pharmaceutical use. Furthermore, antibodies in one embodiment of the present invention include not only whole antibodies, but also antibody fragments such as Fv, Fab, and F(ab)2, and minibodies such as monovalent or multivalent single-chain Fvs (e.g., scFv and sc(Fv)2) in which antibody variable regions are linked via a linker such as a peptide linker.

[0102] Protein production in animal cells can be achieved by simply culturing the cells, or by special procedures. The procedures and conditions may be appropriately determined depending on the animal cells being cultured. For example, CHO cells transformed with a vector containing a gene encoding a mouse-human chimeric antibody through genetic engineering can produce the antibody in the medium under the conditions described below. The desired protein can be obtained by isolating and purifying the antibody according to standard methods (see, for example, "Introduction to Antibody Engineering," Chijin Shokan, pp. 102-104; "Affinity Chromatography Principles & Methods," Amersham Pharmacia Biotech, pp. 56-60).

[0103] In addition, various culture apparatuses for animal cell culture may be used, such as a fermenter-type tank culture apparatus, an airlift-type culture apparatus, a culture flask-type culture apparatus, a spinner flask-type culture apparatus, a microcarrier-type culture apparatus, a fluidized bed-type culture apparatus, a hollow fiber-type culture apparatus, a roller bottle-type culture apparatus, and a packed bed-type culture apparatus.

[0104] Proteins secreted into the medium from cultured animal cells can be recovered from the culture medium by conventional methods. Alternatively, proteins can be recovered from host cell lysates by conventional methods. Specifically, the desired protein can be recovered by removing cells and cell debris from the cell culture medium or cell lysate by centrifugation or other methods, followed by application of common protein isolation and purification techniques. Examples of techniques that can be used include salting out (e.g., ammonium sulfate fractionation), alcohol precipitation (e.g., ethanol precipitation), PEG, electrophoresis, ion exchange chromatography, ultracentrifugation, gel filtration, hydrophobic chromatography, and affinity chromatography. When the desired protein is an antibody, protein A chromatography is preferably used, but is not limited to this. Furthermore, various affinity-based separation or fractionation methods can be used to separate antibodies into immunoglobulin classes or to separate them based on their antigen binding ability.

[0105] In one aspect of the present invention, recombinant antibodies (natural antibodies, antibody fragments, minibodies, chimeric antibodies, humanized antibodies, bispecific antibodies, etc.), recombinant proteins (granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), erythropoietin, interferons, interleukins such as IL-1 and IL-6, t-PA, urokinase, serum albumin, blood coagulation factors, etc.), and the like can be produced in high yields while maintaining high homogeneity.

[0106] Cell Culture System In one embodiment of the present invention, a system for culturing cells by perfusion culture may include a perfusion bioreactor, a measurement device, a cover for the perfusion bioreactor, and a supply reservoir. The perfusion bioreactor may include an internal chamber, a port for supplying nutrients from the supply reservoir, a port for automatic cell discharge from the internal chamber, a port for collecting culture supernatant from the internal chamber, and a rotatable shaft connected to a stirrer. The port for automatic cell discharge from the internal chamber is used to automatically discharge the medium containing viable cells in the bioreactor. Each port is connected to a pump. For example, the port for automatic cell discharge from the internal chamber is connected to a cell discharge pump. Each pump is connected to a measurement device. The perfusion bioreactor may be connected to a cell retention device. The cell retention device is equipped with a filter. The filter may be a hollow fiber filter with a particle size of 0.05 μm to 5 μm, 0.1 μm to 2 μm, or 0.2 μm to 0.5 μm, preferably 0.2 μm. The measurement device may include a capacitance probe, a probe for measuring parameters other than capacitance, such as a Raman probe, a calculation unit, and a control unit. The measurement device receives culture device data and operates the cell culture system. The culture device data is real-time information on the operating conditions of the culture device, such as temperature, agitation rate, DO, pH, capacitance value, and balance weight. The capacitance probe and the probe measuring parameters other than capacitance are connected to the calculation unit and transmit measurement results from the probes. The calculation unit receives measurement results from the capacitance probe and the probe measuring parameters other than capacitance and calculates the viable cell density. The measurement results from the capacitance probe are received continuously. The measurement results from the probe measuring parameters other than capacitance are received at predetermined time intervals. The calculation unit obtains a coefficient by dividing the viable cell density value by the measured capacitance at the time the parameter other than capacitance is measured, and determines the set value of the capacitance by dividing the target viable cell density by the coefficient. When the measured capacitance value exceeds the set value of the capacitance, the control unit discharges the culture medium containing the viable cells from the perfusion bioreactor so that the measured capacitance approaches the set value. The measurement device further includes a laser emission module.The laser emitting module emits a laser at a certain wavelength within the bioreactor.

[0107] In one aspect of the present invention, the laser emission module can emit a laser having an excitation wavelength of 532, 785, or 1000 nm. The laser emission module can also emit a laser periodically. The laser emission can be performed every predetermined time, for example, every 10 minutes, 15 minutes, 20 minutes, 30 minutes, 45 minutes, 1 hour, 1.5 hours, 2 hours, 2.5 hours, 3 hours, 4 hours, 5 hours, 6 hours, 12 hours, or 24 hours.

[0108] Example 1 (Cell Culture) A CHO cell line was cultured in a 2 L bioreactor (BCP, Biot) using a dedicated medium. The temperature during culture was set and controlled at 37°C, agitation at 270 RPM, pH at approximately 7.00, and dissolved oxygen (DO) at 50%. To perform perfusion culture, an ATF2 cell retention device (Repligen) equipped with a 0.2 μm hollow fiber filter was connected to the bioreactor. The medium supply and discharge rates were set to 2.5 times the volume of the culture medium per 24 hours. Using the ATF2 cell retention device, cells were retained in the bioreactor without passing through the hollow fiber filter, while proteins and nutrients passed through the filter.

[0109] During cell culture, the capacitance in the culture medium was measured sequentially using a capacitance probe (Futura, manufactured by ABER). To control the viable cell density during culture, an arbitrary capacitance value was set, and cells were sequentially discharged when the measured capacitance value exceeded the set value. The capacitance value was set based on the viable cell density measured using an offline cell measurement device during culture.

[0110] During cell culture, Raman spectral data were also collected sequentially using a Raman probe (Endress+Hauser Optical Analysis). The viable cell density was estimated from the Raman spectral data using a machine learning model constructed using the spectral data and offline viable cell density measurement data.

[0111] (Cell culture system) Culture apparatus data and weight data of the bioreactor and the waste bottle containing the discharged cells were sent to a computer (measuring device) running control software via RS232 communication, Ethernet (registered trademark), or other connection. The culture apparatus data is real-time information on the operating conditions of the culture apparatus, such as temperature, stirring rate, DO, pH, capacitance value, and balance weight. The supply of medium during perfusion culture and the recovery of medium containing the target substance were controlled by a pump based on each weight data via the control software (Figure 1).

[0112] The capacitance data acquired through the capacitance probe was converted into digital values ​​from the MODBUS analog output module and sent to a computer (measuring device) running control software. The cell efflux pump was configured to operate when the capacitance value exceeded a preset value set in the control software until the value fell below the preset value (Figure 1). The capacitance setting is usually performed using the viable cell density measured using an offline cell measuring device.

[0113] The spectral data acquired via the Raman probe was processed by a Raman measurement system RNX2 (manufactured by Endress+Hauser Optical Analysis), and an estimate of the viable cell density was calculated based on a machine learning model previously stored in the system. The calculated viable cell density was sent via an Ethernet connection or other means to a computer (measurement system) running the control software for the culture system (Figure 1).

[0114] (Construction of a machine learning model for calculating viable cell density from Raman spectral data) Construction of a machine learning model requires spectral data previously measured using Raman spectroscopy and viable cell density data measured offline at a time close to the time the spectral data was acquired. The spectral data underwent preprocessing, including differentiation, smoothing, and SNV processing, to make it usable (Figure 2). Approximately 100 points of data covering the viable cell density data range were prepared, and machine learning was performed. The machine learning method used was PLS regression analysis using SIMCA software (registered trademark). Details of the machine learning method are described in Non-Patent Document 9.

[0115] (Cell Culture Results) An example of the cell culture results is shown in Figure 3. After the start of cell culture, the number of cells in the perfusion bioreactor measured by offline viable cell density measurement reached 120 x 10 by the 10th day of culture. ６ The cell density reached approximately 130 cells / mL. After the desired cell number was reached, the capacitance was set to 130 pF / cm, and the cell discharge operation was initiated (Figure 3). After the cell discharge operation began, the viable cell density was measured periodically, and the capacitance value was manually reset (Figure 3). When the capacitance setting was maintained at a constant value without being changed, the VCD of CHO cells decreased (Figure 4).

[0116] (Automatic Resetting of Capacitance Values ​​Based on Raman Spectroscopy) To maintain a constant VCD without offline viable cell density measurement, we automatically changed the capacitance setting using an estimated viable cell density based on Raman spectroscopy (Figure 5). Specifically, the viable cell density value based on Raman spectroscopy was divided by the capacitance value at the same time as the calculation to determine the viable cell density per capacitance value, thereby calculating the capacitance value required for the desired viable cell density. The Raman spectroscopy measurement interval was set to every hour.

[0117] Figure 6 shows the data obtained when the capacitance setting value was automatically changed using Raman spectroscopy. The capacitance value was controlled to a constant value until the 22nd day of culture, and after confirming a decrease in the viable cell density, automatic change of the capacitance value based on the Raman spectrum was initiated. After starting the automatic change of the capacitance value, an increase in the capacitance setting value and viable cell density was confirmed, and control was achieved as expected. For example, at 80 x 10 ６ If the capacitance value at cells / mL is 120, then 80 / 120 = 0.67. The target value is 120 x 10 ６ In the case of cells / mL, the result is 120 / 0.67 = 179, so the capacitance setting is 179. This calculation is performed each time spectral data is acquired. The viable cell density value based on Raman spectra will fluctuate depending on the machine learning model used, but it is sufficient if the viable cell density can be estimated to be within the range of fluctuation observed during offline viable cell density measurements. Figure 7 shows the data from inline capacitance measurements. After automatic capacitance value change control was initiated, we confirmed that correction of the capacitance setting value had begun. Figure 8 shows the inline viable cell density based on Raman spectra and the viable cell density measured offline (■).

[0118] Figure 9 shows data from an automatic change of the capacitance setting value based on Raman spectroscopy on a different experimental day. Automatic change of the capacitance value based on Raman spectroscopy was also performed after the 19th day of culture. After starting the automatic change of the capacitance value, an increase in the capacitance setting value and viable cell density was confirmed, demonstrating control as expected. Figure 10 shows data from inline capacitance measurements. Figure 11 shows the inline viable cell density based on Raman spectra and the viable cell density measured offline (■).

Claims

1. A method for culturing cells using a perfusion bioreactor, comprising: continuously measuring the capacitance of a culture medium containing live cells in the perfusion bioreactor; and, when the measured capacitance exceeds a set capacitance value, discharging the culture medium containing live cells in the perfusion bioreactor so that the measured capacitance approaches the set capacitance value; the set capacitance value is changed at predetermined time intervals; the change is made based on a numerical value of a live cell density calculated based on a parameter other than capacitance; and the set capacitance value is a capacitance corresponding to a target live cell density.

2. The method of claim 1, wherein the parameter other than capacitance is a spectroscopic measurement.

3. The method of claim 2, wherein the capacitance measurement is performed with a capacitance probe and the measurement of the parameter other than capacitance is performed with a probe used in spectroscopic techniques.

4. The method of claim 3, wherein said spectroscopic techniques include Raman spectroscopy, near-infrared spectroscopy, and infrared spectroscopy.

5. The method according to any one of claims 1 to 4, wherein the calculation of the viable cell density value based on a parameter other than capacitance is carried out by creating a regression model with a coefficient of variation of 20% or less with respect to offline measured values ​​of viable cell density.

6. A method according to any one of claims 1 to 5, wherein the change is carried out by obtaining a coefficient by dividing the value of the viable cell density calculated based on the parameter other than the capacitance by the measured value of the capacitance at the time the parameter other than the capacitance is measured, and by dividing the target viable cell density by the coefficient to determine the set value of the capacitance.

7. A system for culturing cells by perfusion culture, comprising: a perfusion bioreactor including: an internal chamber containing culture medium and viable cells; a port for supplying nutrients from a supply reservoir; a port for automatic cell discharge from the internal chamber; and a port for continuously collecting a culture supernatant containing a target product from the internal chamber; and a measuring device including: one or more capacitance probes immersed in the culture medium in the perfusion bioreactor for measuring capacitance; one or more probes immersed in the culture medium in the perfusion bioreactor for measuring parameters other than capacitance; a calculation unit that receives the measured capacitance and the parameters other than capacitance and is provided with a target viable cell density, and calculates and changes the set value of capacitance; and a control unit that controls the automatic cell discharge from the internal chamber based on the set value of capacitance and the measured capacitance obtained by the capacitance probe.

8. The system of claim 7, wherein the parameter other than capacitance is a spectroscopic measurement.

9. A system according to claim 7 or claim 8, wherein the one or more probes for measuring parameters other than capacitance are probes used in spectroscopic techniques.

10. The method of claim 9, wherein the spectroscopic techniques include Raman spectroscopy, near-infrared spectroscopy, and infrared spectroscopy.

11. A system according to any one of claims 7 to 10, wherein the one or more probes for measuring parameters other than capacitance within a perfusion bioreactor are Raman probes and configured to collect Raman spectra within the bioreactor.

12. The system of claim 11, wherein a viable cell density value is calculated based on the Raman spectrum, and the viable cell density value is used to change the capacitance setting.

13. The system of any one of claims 7 to 12, wherein the modification is performed by dividing the viable cell density value by the measured capacitance value at the time the Raman spectrum was collected to obtain a coefficient, and dividing the target viable cell density by the coefficient to determine the set capacitance value.

14. The method according to any one of claims 1 to 6, wherein the cells to be cultured are animal cells, plant cells, bacterial cells or fungal cells, preferably animal cells.

15. The method of claim 14, wherein the cells cultured are mammalian cells, preferably Chinese hamster ovary (CHO) cells.

Images