Connection perfusion of continuous flow stirred tank reactor cell culture systems

Abstract

Methods of protein production in a linked cultivation and production bioreactor system are provided. Such methods include a cultivation bioreactor (N-1 bioreactor) linked to a production bioreactor (N bioreactor). More specifically, the method includes (a) cultivating cells having a gene encoding a protein of interest in a continuous perfusion cultivation bioreactor (N-1 bioreactor); inoculating a continuous stirred tank reactor (CSTR) production bioreactor (N bioreactor) with the cells obtained from step (a); and cultivating the cells in the CSTR production bioreactor under conditions that allow production of the protein of interest.

Description

[0001] This application is a divisional application of the invention patent application filed on January 25, 2017, with application number 201780019302.5 and invention title "Connection and perfusion of a cell culture system in a continuous flow stirred tank reactor". Technical Field

[0002] The subject matter relates to a method for protein production in cultured animal cells, preferably mammalian cells, using a connected bioreactor system comprising a culture bioreactor (N-1 bioreactor) connected to a production bioreactor (N bioreactor). More specifically, the subject matter relates to a continuous perfusion culture bioreactor connected to a chemostat or a continuous flow stirred tank reactor (CSTR) production bioreactor. Background Technology

[0003] Conventional perfusion cell culture systems have drawbacks such as large volume culture medium consumption, long time to reach peak cell density, and complexity of cell retention devices when used on a large scale.

[0004] For example, in the hybrid cell culture system disclosed in WO 2008091113, multiple CSTR bioreactors are connected in series, each using a cell retention device (e.g., a packed bed). This reference suggests that in an ideal system, the last bioreactor in the series would use complete cell retention, but earlier bioreactors would allow some cells to be discharged from the bioreactor. In the hybrid cell culture system disclosed in WO 2015003012, a single 'care' or N-1 bioreactor is periodically inoculated into a multiple N-stage production bioreactor with a cell retention device. In the hybrid cell culture system disclosed in WO 2015095809, an N-1 perfusion bioreactor is used to prepare inoculum for use in a production bioreactor operated in fed-batch or perfusion mode. However, all the data and examples in this reference suggest that N-1 is a perfusion bioreactor operated for a short period, producing a single inoculum for inoculating a production bioreactor, which is operated as a fed-batch bioreactor. Furthermore, the authors of this application have not indicated that the production bioreactor can be operated as a chemostat (or CSTR) without cell retention, which produces a continuous harvest for downstream purification operations.

[0005] Cell retention systems are difficult to operate and design for large-scale (>1,000 L) use. For example, almost any cell retention device using a membrane (like many currently in large-scale use) eventually becomes clogged with cell debris. This clogging is more likely to occur with low-cell-viability cultures because the cells are more fragile, and the resulting microparticles are often similar in size to the membrane pores (0.2–5 micrometers). Furthermore, when membranes are clogged with cell debris, they also begin to effectively act as ultrafiltration devices, retaining high-molecular-weight protein products within the bioreactor in an unpredictable or reproducible manner. This is a disadvantage because in continuous perfusion bioreactor systems, it is advantageous to continuously remove the target product from the bioreactor in a cell-free harvest and deliver it to downstream operations in a consistent manner.

[0006] Therefore, conventional continuous perfusion cell culture systems typically have a working volume of less than 2,000 L, and if operated under conditions with the highest productivity (e.g., high viable cell density and high perfusion rate), the membrane-based cell retention device needs frequent replacement due to clogging and product retention (ultrafiltration). Furthermore, cell retention devices capable of handling very large volumes of cell-free culture harvests can be very complex (e.g., many moving parts) and expensive, prone to damaging cells through excessive shear forces and failure. Effective cleaning and sterilization can also be challenging on a large scale, as cell retention devices are usually located outside the bioreactor. These factors can lead to high operating costs, significant productivity losses, and inefficiencies in the production of the desired protein. Therefore, alternative cell culture methods or systems that overcome the limitations associated with current conventional perfusion culture systems remain needed. Summary of the Invention

[0007] For example, the subject matter technology is illustrated by the various aspects and implementation schemes listed below.

[0008] In one aspect, the subject matter relates to a method for producing a target protein, comprising: (a) culturing cells containing a gene encoding the target protein in a culture bioreactor (N-1 bioreactor); (b) inoculating a production bioreactor (N bioreactor) with the cells obtained from step (a); and (c) culturing the cells in the production bioreactor under conditions allowing the production of the target protein. In one or more embodiments directly or indirectly relating to this aspect, the method further comprises step (d) harvesting the target protein from the production bioreactor; the culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor; the production bioreactor does not have a cell retention device; the volume ratio of the culture bioreactor to the production bioreactor is from about 1:1 to about 1:20; the volume ratio of the culture bioreactor to the production bioreactor is from about 1:1 to about 1:5; the volume ratio of the culture bioreactor to the production bioreactor is from about 1:5; the inoculation in step (b) is by transferring cells from the culture bioreactor to the production bioreactor; the cell transfer is by cell efflux in a continuous or semi-continuous mode. (bleed); cell transfer is in a semi-continuous mode, including cell transfer at intervals of 2 minutes to 24 hours or any interval in between; step (a) optionally alternates between a first culture bioreactor and a second culture bioreactor to allow for renewal and continuous production of cultured cells for step (b); the second culture bioreactor is a continuous perfusion culture bioreactor; periods of continuous operation of the production bioreactor for more than 3 weeks; periods of continuous operation of the production bioreactor for more than 4 weeks; periods of continuous operation of the production bioreactor for more than 5 weeks; periods of continuous operation of the production bioreactor for more than 6 weeks; harvest step (d) is continuous; cells are CHO cells, HEK-293 cells. VERO cells, NSO cells, PER.C6 cells, Sp2 / 0 cells, BHK cells, MDCK cells, MDBK cells, or COS cells; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day and a period of at least 14 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day and a period of at least 20 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day and a period of at least 30 days; the production bioreactor has a product residence time of about 1 to about 10 days; the production bioreactor has a dilution rate of about 1 to about 0.1 volume / day; the production bioreactor is fed with a dilution solution; the dilution solution is water or brine.

[0009] In another aspect, the subject matter technology relates to a culturing method for connecting cells, comprising: (a) culturing cells containing a gene encoding a target protein in a culture bioreactor (N-1 bioreactor); (b) inoculating a production bioreactor (N bioreactor) with the cells obtained from step (a); and (c) culturing the cells in the production bioreactor under conditions that allow the production of the target protein. In one or more embodiments directly or indirectly relating to this aspect, the connected culture method further includes step (d) harvesting the target protein from the production bioreactor; the culture bioreactor is a continuous perfusion culture bioreactor, and the continuous production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor; the production bioreactor does not have a cell retention device; the volume ratio of the culture bioreactor to the production bioreactor is about 1:1 to about 1:20; the volume ratio of the culture bioreactor to the production bioreactor is about 1:1 to about 1:5; the volume ratio of the culture bioreactor to the production bioreactor is about 1:5; the inoculation in step (b) is by transferring cells from the culture bioreactor to the production bioreactor; the cell transfer is by cell effluent in a continuous or semi-continuous mode; the cell transfer is in a semi-continuous mode, including cell transfer once every 2 minutes to 24 hours or at any interval therebetween; step (a) is optionally alternated between a first culture bioreactor and a second culture bioreactor to allow for renewal and continuous production for the culture in step (b). Cell culture; the second culture bioreactor is a continuous perfusion culture bioreactor; the production bioreactor operates continuously for a period of more than 3 weeks; the production bioreactor operates continuously for a period of more than 4 weeks; the production bioreactor operates continuously for a period of more than 5 weeks; the production bioreactor operates continuously for a period of more than 6 weeks; the harvest step (d) is continuous; the cells are CHO cells, HEK-293 cells, VERO cells, NSO cells, PER.C6 cells, Sp2 / 0 cells, BHK cells, MDCK cells, MDBK cells, or COS cells; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day for a period of at least 14 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day for a period of at least 20 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day for a period of at least 30 days; the production bioreactor has a product residence time of about 1 to about 10 days; the production bioreactor has a dilution rate of about 1 to about 0.1 volume / day; the production bioreactor is fed with a dilution solution; the dilution solution is water or brine.

[0010] In another aspect, the subject matter technology relates to a connected culture system comprising: a culture bioreactor (N-1 bioreactor) for culturing cells containing a gene encoding a target protein; and a production bioreactor (N bioreactor) connected to the culture bioreactor and receiving cells from the culture bioreactor as inoculum, wherein the production bioreactor operates under conditions that allow the production of the target protein. In one or more embodiments directly or indirectly relating to this aspect, the system is connected to a downstream purification system for harvesting the target protein from a production bioreactor; the culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor; the production bioreactor does not have a cell retention device; the volume ratio of the culture bioreactor to the production bioreactor is about 1:1 to about 1:20; the volume ratio of the culture bioreactor to the production bioreactor is about 1:1 to about 1:5; the volume ratio of the culture bioreactor to the production bioreactor is about 1:5; inoculation is performed by transferring cells from the culture bioreactor to the production bioreactor; cell transfer is performed by cell effluent in a continuous or semi-continuous mode; cell transfer in a semi-continuous mode includes cell transfer at intervals of 2 minutes to 24 hours or any interval therebetween; the system includes a second culture bioreactor that operates in parallel with the first culture bioreactor and alternately produces inoculum for transfer to the production bioreactor; The second type of culture bioreactor is a continuous perfusion culture bioreactor; the production bioreactor operates continuously for a period of more than 3 weeks; the production bioreactor operates continuously for a period of more than 4 weeks; the production bioreactor operates continuously for a period of more than 5 weeks; the production bioreactor operates continuously for a period of more than 6 weeks; the harvest is continuous; the cells are CHO cells, HEK-293 cells, VERO cells, NSO cells, PER.C6 cells, Sp2 / 0 cells, BHK cells, MDCK cells, MDBK cells, or COS cells; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day for a period of at least 14 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day for a period of at least 20 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day for a period of at least 30 days; the production bioreactor has a product residence time of about 1 to about 10 days; the production bioreactor has a dilution rate of about 1 to about 0.1 volume / day; the production bioreactor is fed with a dilution solution; the dilution solution is water or brine.

[0011] In another aspect, the subject matter relates to the production of a target protein by a method comprising: (a) culturing cells containing a gene encoding the target protein in a culture bioreactor (N-1 bioreactor); (b) inoculating a production bioreactor (N bioreactor) with the cells obtained from step (a); and (c) culturing the cells in the production bioreactor under conditions allowing the production of the target protein. In one or more embodiments directly or indirectly relating to this aspect, the method further comprises step (d) harvesting the target protein from the production bioreactor; the culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor; the production bioreactor does not have a cell retention device; the volume ratio of the culture bioreactor to the production bioreactor is from about 1:1 to about 1:20; the volume ratio of the culture bioreactor to the production bioreactor is from about 1:1 to about 1:5; the volume ratio of the culture bioreactor to the production bioreactor is from about 1:5; the inoculation in step (b) is by transferring cells from the culture bioreactor to the production bioreactor; the cell transfer is by cell effluent in a continuous or semi-continuous mode; step (a) optionally Alternating between the first and second culture bioreactors allows for the renewal and continuous production of cultured cells for step (b); the second culture bioreactor is a continuous perfusion culture bioreactor; the production bioreactor operates continuously for more than 3 weeks; the production bioreactor operates continuously for more than 4 weeks; the production bioreactor operates continuously for more than 5 weeks; the production bioreactor operates continuously for more than 6 weeks; the harvest step (d) is continuous; the target protein is an antibody or fusion protein; the cells are CHO cells, HEK-293 cells, VERO cells, NSO cells, PER.C6 cells, Sp2 / 0 cells, BHK cells, MDCK cells, MDBK cells, or COS cells; the production bioreactor is fed with a dilution solution; the dilution solution is water or saline.

[0012] Further advantages of the subject matter will become apparent to those skilled in the art from the following accompanying drawings and detailed description. The drawings and description are intended to be illustrative rather than limiting. Attached Figure Description

[0013] These and other features, aspects, and advantages of the claimed method, apparatus, and / or system can be better understood when the following detailed description is read in conjunction with the accompanying drawings: Figure 1This is an exemplary diagram illustrating a bioreactor configuration according to an embodiment described in detail in the experimental section of this application. In this configuration, a continuous supply of cells is made to a second bioreactor, operating as a continuous flow stirred tank reactor (CSTR) or chemostat, using an N-1 bioreactor with cell retention perfusion. The CSTR also receives an additional concentrated nutrient feed stream. The CSTR does not have a cell retention system and can be harvested continuously. In experiments conducted on a laboratory scale, the cell retention device consisted of a surface area of ​​850 cm². 2 It consists of a 0.2-micron microfiltration hollow tube (General Electric CFP-2-E-4X2MA). Cells are circulated through the lumen side of the hollow fiber tube at a rate of ~120 mL / min using a Watson-Marlow peristaltic pump with a 6.4 mm ID and 12.7 mm OD channel.

[0014] Figure 2A This is a graph showing the viable cell density as measured by trypan blue exclusion as described in the experimental section of this application. Hollow squares represent cell density in the N-1 perfusion reactor. Solid squares represent cell density in the CSTR or production bioreactor. Dashed vertical lines represent step changes in the composition of slightly different perfusion media flowing into the N-1 perfusion reactor, as shown in Table 2.

[0015] Figure 2B This is a graph showing the percentage of viable cells as measured by trypan blue exclusion. Hollow squares represent cell viability in an N-1 perfusion reactor. Solid squares represent cell viability in a CSTR or production bioreactor.

[0016] Figure 3 The perfusion and cell effluent rates for the N-1 bioreactor and the dilution rate for the production (CSTR) bioreactor are shown. The perfusion rate (hollow squares) is listed in reactor volume (RV) / day and corresponds to the left-hand y-axis. The perfusion rate is the volume of culture medium flowing into the N-1 perfusion bioreactor per day. The cell effluent rate (solid squares) is the volume of cell-containing culture continuously removed from the N-1 perfusion bioreactor per day and corresponds to the right-hand y-axis. The dilution rate (solid triangles) is the fraction of the cell-containing volume (the only flow leaving the CSTR) continuously removed from the production (CSTR) bioreactor per day and corresponds to the right-hand y-axis.

[0017] Figure 4 This is a graph showing the residual glucose concentration in the N-1 (hollow cube) and production (CSTR, solid cube) bioreactors.

[0018] Figure 5This is a graph showing the lactate concentration in the N-1 (hollow cube) and production (CSTR, solid cube) bioreactors.

[0019] Figure 6 It is a graph showing the concentration of antibodies or products in the large volume of fluid leaving the N-1 bioreactor (solid cube), the permeate leaving the N-1 bioreactor system (hollow cube), and the large volume of fluid continuously leaving the production bioreactor (CSTR, solid triangle).

[0020] Figure 7 This is a graph showing the volumetric productivity of CSTR production bioreactors, expressed in grams of product produced per reactor volume per day (solid triangles).

[0021] Figure 8 This is a graph showing cell densities in two N-1 perfusion bioreactors over time. The target steady-state viable cell density for the first N-1 perfusion bioreactor is ~40 x 10⁻⁶ cells / year. 6 The target steady-state viable cell density for the second N-1 perfusion reactor was ~80 x 10⁻⁶ cells / mL. 6 Cells / mL. As described in more detail in Example 2, in this configuration, each of these N-1 perfusion bioreactors is independently fed into a separate CSTR production bioreactor.

[0022] Figure 9 This is a graph showing cell viability in two N-1 perfusion bioreactors over time.

[0023] Figure 10 This is a diagram showing the residual glucose over time for two continuous perfusion bioreactors (hollow symbols) and two production bioreactors (solid symbols) operating as CSTRs, each production bioreactor connected to one of the perfusion bioreactors.

[0024] Figure 11 This is a graph showing the lactate concentration over time for the N-1 continuous perfusion bioreactor (hollow symbol) and the production bioreactor (solid symbol) operating as a CSTR.

[0025] Figure 12 This is a graph showing the perfusion rate of the N-1 continuous perfusion bioreactor - perfusion rate controlled by cells using the HIPCOP technique as described in Example 2.

[0026] Figure 13 This is a graph showing the cell effluent rates of two N-1 continuous perfusion bioreactors.

[0027] Figure 14This is a graph showing the live cell density over time in a production bioreactor operating as a continuous flow stirred tank reactor (CSTR). The time intervals in which the dilution rate remains constant are indicated on the graph. The simulated volume ratio of the N-1 perfusion bioreactor to the CSTR is also shown on the graph.

[0028] Figure 15 It is a graph showing the dilution rate of each in a production bioreactor operating as a continuous flow stirred tank reactor (CSTR) over time.

[0029] Figure 16 This is a graph showing cell viability measured over time, such as by trypan blue dye exclusion, in a production bioreactor operating as a continuous flow stirred tank reactor (CSTR).

[0030] Figure 17 This is a graph showing the product concentration or titer over time for a production bioreactor operating as a continuous flow stirred tank reactor (CSTR). The time intervals in which the dilution rate remains constant are indicated on the graph.

[0031] Figure 18 This is a graph showing the iso-osmotic gravity of the N-1 continuous perfusion bioreactor (hollow symbol) and the production bioreactor operating as a CSTR (solid symbol) over time. The time points indicated on the graph are when the perfusion medium was changed to sodium lactate in the form of DL.

[0032] Figure 19 This is a graph showing the instantaneous volumetric productivity of the N-1 continuous perfusion bioreactor when considered as an independent operation. The calculations assume that products are recovered from both cell effluent and cell-free permeate leaving the cell retention system.

[0033] Figure 20 This is a graph showing the concentration of products (antibodies) in the cell-free permeate leaving the N-1 perfusion reactor and the cell retention device.

[0034] Figure 21 This is a graph showing the steady-state volumetric productivity of the N-1 continuous perfusion bioreactor connected to the CSTR system, plotted according to the CSTR production dilution rate. The volume ratio of the N-1 continuous perfusion bioreactor to the CSTR production is indicated for several conditions. All points not specifically indicated use a volume ratio of 1:5. Detailed Implementation

[0035] Numerous specific details are set forth in the following detailed description to provide a complete understanding of the subject matter. However, it will be apparent to those skilled in the art that the subject matter can be practiced without some of these specific details. In other instances, well-known structures and techniques have not been shown in detail so as not to obscure the subject matter.

[0036] To facilitate understanding of the subject matter, the following defines a number of terms and phrases: definition: Unless otherwise stated, as used herein, the grammatical articles “a / kind,” “one,” “a kind,” and “the / described” are intended to include “at least one / kind” or “one or more / a combination of.” Therefore, the articles are used herein to refer to one or more (i.e., at least one) grammatical objects. For example, “component” means one or more components, and thus, more than one component is considered and may be adopted or used in the implementation of the described embodiments.

[0037] The term "about" generally refers to a slight error in a measurement and is often stated as a series of values ​​that contain the true value within a specific confidence level (typically ±1 for a 68% CI). The term "about" can also be described as an integer and an integer ±20% of the true value.

[0038] As used interchangeably herein, the terms "production bioreactor," or "N-bioreactor," or "CSTR bioreactor," or "CSTR production bioreactor" refer to a bioreactor that does not utilize a cell retention system or apparatus (e.g., a continuous flow stirred tank reactor (CSTR)). A production bioreactor is connected downstream to one or more culture bioreactors and receives inoculum from the culture bioreactors. Production bioreactors are homogeneously mixed and have a fluid inflow equivalent to a fluid outflow, thus maintaining a near-constant volume. Such production bioreactors often (though not necessarily) achieve a 'chemically static' or 'steady-state' environment when operated for a sufficiently long period, meaning that cell density and other aspects of the culture (e.g., nutrient concentrations, etc.) will reach a stable (i.e., steady-state or quiescent) state, and are therefore often referred to as 'chemostats'. The fluids flowing into and / or out of the production bioreactor can be independently continuous or semi-continuous. Such production bioreactors operate continuously for periods equal to or greater than 3, 4, 5, or 6 weeks.

[0039] In one embodiment, the subject-specific CSTR production bioreactor can operate continuously indefinitely, or as long as the cultured cells remain genetically stable. In another embodiment, the subject-specific CSTR production bioreactor is connected to at least two culture bioreactors that alternately feed fresh (i.e., genetically stable) inoculum to the production bioreactor, and thus, the production bioreactor operates continuously for a long period of time, for example, equal to or greater than 1, 2, 3, 4, 5, 6 months, or indefinitely, as long as the cultured cells fed to it remain genetically stable. Without a cell retention device, the fluid effluent from the CSTR production bioreactor comprises a mixture of cells, cell debris, and products. In one embodiment, this fluid effluent is directed to a cell separation device (e.g., an in-line centrifuge) before being transferred to the next unit operation (e.g., a protein A column). Cells and cell debris separated from the remaining fluid effluent in this manner are discarded and not returned to the production bioreactor.

[0040] As used herein, the term "cell retention system" or "cell retention device" refers to a device that selectively retains live cells within a bioreactor such that the cell density in the fluid leaving the bioreactor is lower than the cell density in the fluid inside the bioreactor. In this sense, a cell retention system or device differs from the cell separation device described above.

[0041] As used interchangeably herein, the terms "culture bioreactor," or "N-1 bioreactor," or "N-1 perfusion bioreactor" refer to a perfusion culture or perfusion bioreactor (e.g., a continuous perfusion culture bioreactor) used to culture cells that will be used to inoculate a production bioreactor. Such culture bioreactors have cell retention systems. Many different forms of cell retention systems exist for industrial use. Some of these cell retention systems remove 100% of the live cells from the liquid leaving the bioreactor system; however, many others can remove only a variable portion of the cells from the liquid leaving the bioreactor system. The transfer of liquid into and out of the culture bioreactor can be continuous or semi-continuous. In particular, the cell effluent or transfer from the culture bioreactor to the production bioreactor can be continuous or semi-continuous.

[0042] In the context of liquid transfer into and / or out of a bioreactor, as used herein, the term “semi-continuous” means “periodic” or refers to a process in which liquid (e.g., culture medium alone and / or containing cells, cell effluent) is added to and / or removed from the bioreactor at long intervals. For example, once every 1, 2, 5, 10, 15, 30, 45, or 60 minutes, or once per hour, or once every 2-3 hours, or once every minute to 24 hours, for periods ranging from a few seconds (e.g., 1 second, 2 seconds, 5 seconds, 10 seconds, 20 seconds, or 60 seconds) to several minutes (e.g., 2 minutes, 5 minutes, 10 minutes, 25 minutes, 50 minutes, 120 minutes, or 240 minutes). In this context, the term “continuous” refers to constant or non-periodic liquid transfer. In one embodiment, the cell effluent / transfer rate from the culture bioreactor to the production bioreactor in a continuous or semi-continuous mode is less than the growth rate of the cells in the culture bioreactor. In another embodiment, the cell effluent / transfer rate from the culture bioreactor to the production bioreactor is from about 0.1 reactor volume / day (RV / day) to about 1.3 RV / day. In another embodiment, in a continuous or semi-continuous mode, the cell effluent / transfer rate from the culture bioreactor to the production bioreactor is less than 1.3 RV / day, or less than 1.0 RV / day, or less than 0.8 RV / day, or less than 0.6 RV / day, or less than 0.4 RV / day, or less than 0.2 RV / day to about 0.1 RV / day, or varies within the range of about 0.1 to about 1.3 RV / day.

[0043] Using cell retention systems for perfusion culture of mammalian cells offers significant productivity advantages over batch, fed-batch, or chemostat / CSTR cultures. Because cells are retained in the bioreactor system, they can be perfused with large volumes of medium and achieve much higher cell densities without the elution issues that occur in CSTRs at high dilution rates. Furthermore, when a continuous perfusion bioreactor is connected to a continuous downstream purification train, the size of the purification train can be drastically reduced, which simplifies the overall operation by eliminating the retention step (and the required tanks) and reducing sampling and analysis of pools in the process. However, perfusion culture of mammalian cells also has several disadvantages, particularly when implemented at large scale (>1,000 L).

[0044] Typically, most perfusion cultures are operated for extended periods (>3 weeks). This requires cell lines to be exceptionally genetically stable, enabling cells to continue producing the target protein, and the rate of epigenetic change to be sufficiently slow. These long-duration perfusion cultures can be defined as “sustainable” perfusion cultures because the culture can be considered operational as long as the genetic profile of the culture does not drift far enough to negatively impact the productivity or product quality profile (specifications) of the target protein. In such sustainable perfusion cultures, it is necessary to maintain high cell viability, which typically requires cells to maintain a non-zero growth rate to continue dividing and compensate for cells that die from apoptosis, shearing, oxidation, or other stresses. In most cases, a limited cell effluent rate (removal of intact culture medium containing cells) is required to maintain some cell growth and continuously remove some inactive biomass from the culture. This 'cell effluent rate' is typically referred to as a fraction of the culture volume removed per day or based on a countdown of days. Cell efflux rates are typically in the range of 0.05-0.5 / day. Higher rates generally produce conditions for higher cell viability, higher cell growth rates, and more sustainable perfusion systems, but at the same time promote lower cell density and generally lower volumetric productivity in perfused cultures.

[0045] Additionally, depending on the cell line, CHO (Chinese hamster ovary) cell cultures typically exhibit higher specific productivity (protein quality produced per cell per cycle) when cell growth rates are low. This is likely because cells dedicate more of their resources to continuous cell division, as opposed to producing the target protein when growth rates are high. Fed-batch culture is one of the most common operating modes for large-scale production of CHO cell cultures and typically involves a short period of cell growth followed by a few days of quiescent period in which most of the product is produced. As a result, there are advantages to culture systems where the periods of culture are separated and the conditions for each period are optimized; one period has a high cell growth rate to quickly achieve high cell density, and a separate period has a near-zero cell growth rate but a higher level of specific cell productivity.

[0046] Continuous perfusion systems may be difficult to implement on a large scale (>1,000 L) because cell retention systems do not necessarily perform well at scale. Furthermore, almost any cell retention device using a membrane (like many currently used on a large scale) eventually becomes clogged with cell debris. This clogging is more likely to occur with low-viability cultures because the cells are more fragile, and the resulting microparticles are often similar in size to the membrane pores (0.2 micrometers). Therefore, minimizing the size of the container in which perfusion takes place and maintaining high cell viability to minimize problems related to membrane fouling is advantageous.

[0047] As mentioned above, cellular microparticles can slowly clog membrane-based cell retention systems, particularly those utilizing microfiltration membranes for long-term perfusion cultures. Cell lysis and the interaction of host cell proteins with the target product, forming substances with higher molecular weights, are also possible. These substances can also contribute to a phenomenon known as gel layer formation or product sieving. Under these conditions, cell retention devices utilizing microfiltration membranes can begin to function as ultrafiltration membranes, selectively retaining high molecular weight proteins within the bioreactor system. Because immunoglobulin molecules (one of the most common proteins produced in cell culture) have relatively high molecular weights (~150,000 Da and above), product sieving has become a major problem for large-scale perfusion systems. Typically, continuous perfusion systems are linked to continuous downstream purification systems, so the target protein must flow out of the bioreactor, and ideally, this protein should be of consistent quantity and quality. Minimizing the retention of the target protein is particularly important for minimizing the size of the downstream system and is critical when the target protein is unstable and may degrade due to overexposure to the conditions within the bioreactor.

[0048] For the many reasons mentioned above, in the following manner and as Figure 1 The two bioreactors connected in continuous operation, as shown, offer significant advantages. The first bioreactor, operating as a continuous perfusion culture operation utilizing a high perfusion rate, allows for high cell densities. A high cell effluent rate will also be used in this reactor, which will maintain a high cell growth rate and thus also high cell viability. The second bioreactor, the production bioreactor, can advantageously be 4-5 times the volume of the first bioreactor (a typical volume ratio used in large-scale facilities for N-1 and production bioreactors) and, according to the preferred mode of the subject matter art, operate as a chemostat (or continuous stirred tank reactor [CSTR]) reactor without a cell retention system. The terms CSTR or chemostat are used interchangeably to refer to the production bioreactor throughout this document, as in most cases, the CSTR will rapidly reach steady-state or 'chemostatic' conditions, where most or all cell culture parameters will reach nearly constant values. In one embodiment, the volume ratio of the N-1 continuous perfusion culture bioreactor to the CSTR production bioreactor is from about 1:1 to about 1:20. In another embodiment, the volume ratio of the N-1 continuous perfusion culture bioreactor to the CSTR production bioreactor is about 1:1 to about 1:5. In another embodiment, the volume ratio of the N-1 continuous perfusion culture bioreactor to the CSTR production bioreactor is about 1:10. In yet another embodiment, the volume ratio of the N-1 continuous perfusion culture bioreactor to the CSTR production bioreactor is about 1:4.

[0049] In the first aspect, the subject matter technology relates to a method for producing a target protein, comprising: (a) culturing cells including a gene encoding the target protein in a culture bioreactor (N-1 bioreactor); (b) inoculating a production bioreactor (N bioreactor) with the cells obtained from step (a); and (c) culturing the cells in the production bioreactor under conditions that allow the production of the target protein.

[0050] In the second aspect, the subject matter technology relates to a method for culturing cells, comprising: (a) culturing cells containing a gene encoding a target protein in a culture bioreactor (N-1 bioreactor); (b) inoculating a production bioreactor (N bioreactor) with the cells obtained from step (a); and (c) culturing the cells in the production bioreactor under conditions that allow the production of the target protein.

[0051] In a third aspect, the subject matter technology relates to a connected culture system comprising: a culture bioreactor (N-1 bioreactor) for culturing cells including a gene encoding a target protein; and a production bioreactor (N bioreactor) connected to the culture bioreactor and receiving cells from the culture bioreactor as inoculum, wherein the production bioreactor operates under conditions that allow the production of the target protein.

[0052] In a fourth aspect, the subject matter technology relates to the production of a target protein by means of: (a) culturing cells containing a gene encoding the target protein in a culture bioreactor (N-1 bioreactor); (b) inoculating a production bioreactor (N bioreactor) with the cells obtained from step (a); and (c) culturing the cells in the production bioreactor under conditions that allow the production of the target protein.

[0053] In one or more embodiments that directly or indirectly relate to any of the foregoing aspects, the method, process, or system further includes step (d) harvesting the target protein from a production bioreactor; the culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor; the production bioreactor does not have a cell retention device; the volume ratio of the culture bioreactor to the production bioreactor is about 1:1 to about 1:20; the volume ratio of the culture bioreactor to the production bioreactor is about 1:1 to about 1:5; the volume ratio of the culture bioreactor to the production bioreactor is about 1:4; the inoculation in step (b) is by transferring cells from the culture bioreactor to the production bioreactor; the cell transfer is by cell effluent in a continuous or semi-continuous mode; the cell transfer is in a semi-continuous mode, including cell transfer at intervals of every 2 minutes to 24 hours or any interval therebetween; step (a) is optionally alternated between a first culture bioreactor and a second culture bioreactor to allow for renewal and continuous production of cultured cells for step (b); the second culture bioreactor is a continuous perfusion culture bioreactor. The production bioreactor is operated continuously for more than 3 weeks; the production bioreactor is operated continuously for more than 4 weeks; the production bioreactor is operated continuously for more than 5 weeks; the production bioreactor is operated continuously for more than 6 weeks; the harvest step (d) is continuous; the cells are, for example, CHO cells, HEK-293 cells, VERO cells, NSO cells, PER.C6 cells, Sp2 / 0 cells, BHK cells, MDCK cells, MDBK cells, or COS cells, or any cells genetically derived therefrom, such as those with specific metabolic conditions and / or selection systems such as glutamine. The selected derivatives, the genetically derived cells and / or metabolic conditions and / or selection systems are known in principle to those skilled in the art; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day and a period of at least 14 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day and a period of at least 20 days; the production bioreactor has a volumetric productivity of at least 0.6 g / L / day and a period of at least 30 days; the production bioreactor has a product residence time of about 1 to about 10 days; the production bioreactor has a dilution rate of about 1 to about 0.1 volume / day.

[0054] Non-limiting examples of mammalian cells that can be used in this invention are summarized in Table 1.

[0055] Table 1: Suitable Exemplary Mammalian Production Cell Lines

[0056] The production cells are preferably cultured under conditions that allow for cell proliferation. Furthermore, the production cells are preferably cultured under conditions favorable to the expression of the desired gene and / or target protein. The target protein is then isolated from the cells and / or cell culture supernatant. Preferably, the target protein is recovered from the culture medium as a secreted polypeptide, or, if expressed without a secretion signal, it can be recovered from the host cell lysis products.

[0057] The culture from bioreactor N-1 will be fed into the production bioreactor at a continuous and constant flow rate, equal to the cell effluent rate of bioreactor N-1. Additionally, the production bioreactor will have a continuous feed of nutrient medium, as the cells in this bioreactor will continue to metabolize, produce products, and may undergo some limited cell division. The working volumes of both bioreactors are likely to remain constant; therefore, the effective dilution rate of the production bioreactor will be determined by the continuous cell effluent flow from N-1, the continuous feed of nutrient medium, and any pH-controlled titrants added directly to the production bioreactor.

[0058] Such bioreactor systems can alleviate many of the problems of current continuous perfusion bioreactors. Perfusion bioreactors and their associated cell retention systems are likely to be significantly simpler to operate, as they will operate at approximately one-fifth the scale of production bioreactors. If gel layer / product sieving occurs in the cell retention device, this will only increase the overall productivity of the system, because the target protein now generated in the N-1 bioreactor will flow into the production bioreactor and eventually enter downstream purification processes, rather than being discharged with cell-free permeate, such as... Figure 1 As shown in the image.

[0059] In cases where the cell lines are not genetically stable enough for long-term operation, a second N-1 bioreactor will be used. This second bioreactor will be periodically started with an inoculum of fresh cells expanded from frozen vials, and once the second bioreactor has reached the appropriate cell density, it can replace the first N-1 bioreactor, which will then be removed for cleaning and resterilization. The cell density in the N-1 culture bioreactor can vary depending on the size of the production bioreactor to which it is connected. In some embodiments, the cell density in the culture bioreactor is approximately 10 x 10⁻⁶. 6 / liter or higher, or approximately 20 x 10 6 / liter or higher, or approximately 40 x 10 6 / liter or higher, or in 10 x 10 6 and 200 x 10 6 Between / liter, or at 40 x 10 6 Up to 120 x 10 6A specific density between 1 / L and 2 / L, or within any of these ranges. Because cell division in the production bioreactor may be low, this would allow for a semi-continuous renewal of protein-producing cells from a genetically younger population. The production bioreactor could then be able to operate continuously with minimal interruption and at high cell density and high productivity, producing feed streams with consistent quality parameters for downstream purification operations for months at a time. Thus, in one embodiment, the subject-specific connected cell culture system comprises an N-1 continuous perfusion culture bioreactor (N-1 bioreactor) in which cells encoding genes for the target protein are cultured before being transferred to a CSTR production bioreactor without cell retention devices. In another embodiment, the subject-specific connected cell culture system comprises two N-1 culture bioreactors that operate alternately to produce cultured cells for transfer to the CSTR production bioreactor. In another embodiment, the N-1 continuous perfusion culture bioreactor of the subject technology is operated in the following mode: high-end pH control of glucose (HIPDOG) as described in Biotechnol Bioeng. 2011 Jun; 108(6):1328-37, or high-end pH control of the perfusion rate (HIPCOP) as described in co-pending application WO 2016 / 196261 entitled “Cell-Controlled Perfusion in Continuous Culture” which describes this method.

[0060] Cell lines with high specific productivity at low growth rates, or those requiring different cell culture conditions to be highly productive or to produce products with specific quality characteristics, should particularly benefit from such bioreactor configurations, as the growth rate in the production bioreactor will likely be very low and more similar to the conditions typically achieved in the later stages of a fed-batch bioreactor. This may also mean that the product quality of the target protein produced in such a connected, continuous N-1 to CSTR production bioreactor may be more similar to the product quality of proteins produced in fed-batch production. Therefore, in one embodiment, the CSTR production bioreactor is operated under conditions that promote the highest cell productivity, which may in some cases also be conditions that slow cell growth. To achieve high productivity conditions, known chemicals that improve per-cell productivity but slow or stop growth can be added to the CSTR production bioreactor. Alternatively, the CSTR bioreactor can be operated, for example, at low or high pH or low temperature (with the addition of Cu, low Ca, galactose, etc.), which is beneficial for obtaining cells that produce proteins with specific quality properties, but those same conditions do not allow for high-rate cell growth.

[0061] As mentioned above, cell specific productivity can, in some cases, be inversely proportional to the cell growth rate in the production (CSTR) bioreactor. Operating a CSTR at a high dilution rate should result in a higher growth rate because inhibitory metabolites will be flushed out of the bioreactor more efficiently. The CSTR dilution rate will depend on the cell effluent rate from the N-1 continuous perfusion bioreactor, the feed rate directly into the CSTR, and any titrants used for pH control. Manipulating the concentration of the feed medium to the CSTR will allow for a balance between optimal growth, nutrient availability, and the flushing out of inhibitory metabolites from the system. Many of these are related multiple factors that may contribute to the maximum productivity of the combined N-1 continuous perfusion and CSTR system, including cell density and cell growth rate in both bioreactors, the perfusion rate in the N-1, and the CSTR dilution rate. It is advantageous to manipulate the CSTR dilution rate using a single concentrated feed medium, diluted with water or water mixed with saturated saline solution as needed. Controlling the dilution rate with a variable concentration feed medium also allows for control of the residence time of recombinant proteins produced from the system. High dilution rates will result in low product residence time and may be advantageous for highly unstable proteins.

[0062] Many industrial applications are currently exploring the use of cell retention / perfusion bioreactors as production systems for mammalian cell cultures. Such systems increase volumetric productivity and enable continuous downstream purification runs. Because, like continuous perfusion bioreactors used as production bioreactors, the N-1 perfusion bioreactor described in this work, connected to a chemostat or CSTR production bioreactor, continues to generate a product stream constant for downstream operations, it continues to offer the advantages of smaller and continuous downstream operations.

[0063] Additionally, the subject matter technology relates to target proteins produced by the methods described above. Such target proteins include, but are not limited to, antibodies or fusion proteins, such as Fc-fusion proteins. Others may include, for example, enzymes, cytokines, lymphokines, adhesion molecules, receptors and their derivatives or fragments, and any other peptides and scaffolds that can act as agonists or antagonists and / or have therapeutic or diagnostic uses. Other target recombinant proteins include, but are not limited to, insulin, insulin-like growth factor, hGH, tPA, and cytokines such as interleukins (ILs).

[0064] Preferred recombinant secreted therapeutic proteins are antibodies or fragments or derivatives thereof. Therefore, the present invention can be advantageously used to produce antibodies, such as monoclonal antibodies, multispecific antibodies, or fragments thereof, preferably monoclonal antibodies, bispecific antibodies, or fragments thereof. Antibody fragments include, for example, “Fab fragments” (fragment antigen binding = Fab). Fab fragments consist of variable regions of two chains held together by adjacent constant regions. These can be formed from conventional antibodies by proteolytic digestion, for example, with papain, but similar Fab fragments can also be produced through genetic engineering. Other antibody fragments include F(ab')2 fragments, which can be prepared by proteolytic cleavage with pepsin.

[0065] Using genetic engineering methods, it is possible to produce shortened antibody fragments consisting only of variable regions of the heavy chain (VH) and light chain (VL). These are called Fv fragments (fragment variable = fragment with variable portion). Because these Fv fragments lack the covalent bond between the two chains via cysteine ​​residues of the constant chain, Fv fragments are often stable. It is advantageous to link the variable regions of the heavy and light chains by, for example, short peptide fragments having 10 to 30 amino acids, preferably 15 amino acids. In this way, a single peptide chain consisting of VH and VL linked by a peptide linker is obtained. Such antibody proteins are called single-chain-Fv (scFv). Examples of scFv-antibody proteins are known to those skilled in the art. The preferred secreted recombinant therapeutic antibody according to the invention is a bispecific antibody. Bispecific antibodies typically combine antigen-binding specificity for target cells (e.g., malignant B cells) and effector cells (e.g., T cells, NK cells, or macrophages) in a single molecule. Exemplary bispecific antibodies are, but are not limited to, biantibodies, BiTE (bispecific T-cell conjugate) forms, and DART (dual affinity retargeting) forms. Microantibodies are also contemplated in the context of this invention. By microantibodies, those skilled in the art mean divalent homodimer scFv derivatives.

[0066] Recombinant secreted therapeutic proteins, especially antibodies, antibody fragments, or Fc-fusion proteins, are preferably recovered / separated from the culture medium as secreted peptides. It is necessary to purify the recombinant secreted therapeutic proteins from other recombinant proteins and host cell proteins to obtain a substantially homogeneous formulation of the recombinant secreted therapeutic protein. As a first step, cell and / or microparticle cell debris is removed from the culture medium or lysis products. Furthermore, the recombinant secreted therapeutic proteins are purified from contaminant-soluble proteins, peptides, and nucleic acids, for example, by fractionation on immunoaffinity or ion exchange columns, ethanol precipitation, reversed-phase HPLC, Sephadex chromatography, and chromatography on silica or cation exchange resins such as DEAE. Methods for purifying heterologous proteins expressed by host cells are known in the art.

[0067] Example Example 1: A continuous N-1 perfusion bioreactor (1-2 liter scale) using feed CSTR production bioreactors to produce humanized IgG using CHO cells. Figure 1 A simplified diagram of the two-stage connected bioreactor system used in the experiment is shown. For experimental purposes, the N-1 perfusion bioreactor, including the perfusion loop (cell retention system), has a working volume of 1.25 L, and the production bioreactor has a working volume of 1.0 L. In a full-scale industrial setting, the N-1 bioreactor would be considered to be approximately one-fifth the volume of the production (CSTR or chemostat) bioreactor. For this reason, the experimental bioreactor was operated to simulate this volume ratio. This means that the majority of the cell effluent from the N-1 perfusion bioreactor (~81% of the total cell effluent) is discharged. Only an appropriate volume (~19% of the total) of cell effluent is pumped into the production bioreactor. The only exception to this rule occurs during the first 3 days of operation of the production (CSTR) bioreactor, in which all cell effluent from the N-1 is added to the production bioreactor. This allows the production bioreactor to reach high density slightly faster than by other methods.

[0068] Figure 2A Initially, only the cell density of the N-1 perfusion bioreactor was shown. This density rapidly increased to approximately 40 x 10⁻⁶ cells / year. 6 The target density was set at 100 live cells / mL, after which cell efflux was initiated and manually adjusted during culture to obtain approximately 40 x 10⁻⁶ cells / mL in the N-1 bioreactor. 6 A constant live cell density of [number] live cells / mL. For example... Figure 3 As shown, for most experiments, the cell efflux rate of the N-1 perfusion bioreactor is approximately 0.4 / day, or 0.4 reactor volume is removed per day.

[0069] The N-1 perfusion bioreactor uses a pH-based feedback mechanism that allows the culture to determine its own perfusion rate. The perfusion medium contains a small amount of sodium L-lactic acid (listed in Table 2).

[0070] Table 2 lists the details of the composition of the perfusion medium used in the N-1 bioreactor.

[0071] Table 2. Composition of the perfusion medium in the N-1 biosensor.

[0072] The culture signals a lack of available glucose by initiating lactate uptake from the bulk culture. Lactate removal from the bulk culture by the cells causes an increase in the pH of the bioreactor, which in turn activates a pump that delivers glucose-containing perfusion medium to the N-1 perfusion bioreactor, thereby increasing the perfusion rate of the culture (a horizontal controller removes cell-free medium [osmotic fluid] from the culture via the shell side of a hollow fiber cell retention device to maintain a constant bioreactor volume). As the cells in the culture begin to uptake the excess glucose being delivered, a portion of it is converted into lactate by the cells. When this lactate is secreted from the cells, it subsequently suppresses the pH of the bulk culture, which in turn prompts the pH controller to deactivate the perfusion medium addition pump (and in turn, the osmotic pump). This control scheme is repeated as a cycle every few minutes during the experimental process. For a description of this method, see co-pending application WO 2016 / 196261 entitled “Cell-Controlled Perfusion in Continuous Culture”. In summary, in the method described in co-pending application WO 2016 / 196261, pH triggers for turning the perfusion pump, fresh culture medium, and osmotic pump on and off are set to predetermined values. These predetermined values ​​are approximately pH 7 (e.g., between 6.8 and 7.4). The culture medium contains glucose, L-lactate, and a specific ratio of amino acids to glucose. The method further includes adding L-lactate to the fresh perfusion medium used in a continuous culture process. The L-lactate present in the perfusion medium is in an amount from about 0.1 g / L to 7.0 g / L. Alternatively or additionally, the L-lactate present in the perfusion medium is in an amount from about 1 to 4 g / L, about 1 to 3 g / L, or about 1 to 2.5 g / L. In a preferred embodiment, the L-lactate is sodium L-lactate or potassium L-lactate. In another embodiment, the lactate used in the perfusion medium may be a D / L mixture of sodium lactate or potassium lactate. In cases where a D / L mixture of lactate is used, sufficient mixture is added to the perfusion medium such that the amount of L-lactate is the same as when L-lactate is used alone. In one embodiment, additional sodium bicarbonate is added to the perfusion medium in place of L-lactate or in addition to L-lactate. Typical cell culture media contain about 1.0 to 2.5 g / L of sodium bicarbonate. In one embodiment, the perfusion medium for the N-1 perfusion bioreactor used in the subject technology has an additional 1 to 3 g / L of sodium bicarbonate to achieve a total sodium bicarbonate concentration of about 2 to 5.5 g / L. The perfusion medium in this method requires at least glucose, L-lactate (or alternatively or additionally, sodium bicarbonate), and amino acids. The concentration of glucose is about 0.5 to about 40 g / L. The concentration of L-lactate is about 0.1 to about 7.0 g / L. The concentration of sodium bicarbonate is about 1 to 5.5 g / L.The molar ratio of glucose to amino acids is between approximately 0.25 and 1.0. In HIPCOP mode, as cell density increases (…). Figure 2A The cells are effectively 'requiring' additional perfusion medium on an increasingly frequent basis, leading to an increase in the perfusion rate. Figure 3 ).

[0073] A similar control scheme for adding glucose to fed-batch cultures has been described in the literature (Biotechnol Bioeng. 2011 Jun; 108(6):1328-37). This technique of high-end pH control at the perfusion rate (HIPCOP) was used for N-1 perfusion cultures throughout the entire duration of the experiment. Similar glucose restriction techniques are also used in CSTRs or production bioreactors to control lactate formation when necessary or useful. In CSTRs, a pump is triggered to add a concentrated nutrient feed containing glucose, or a feed containing pure glucose dissolved in water, when the pH rises due to lactate uptake by the cells. Table 3 lists some of the feed medium or pure glucose solution used as part of this high-end pH control (HIPDOG) strategy for glucose.

[0074] Table 3 details the composition of the concentrated feed medium delivered on demand to the CSTR (chemostat) via the high-end pH control system (HIPDOG) when the system was started on days 12–19 and again on days 29–33.

[0075] Table 3. Composition of concentrated feed culture medium.

[0076] The HIPDOG strategy was used to control lactate formation in the CSTR only from days 12–19 and 29–33. From days 19–35, a separate fixed-rate feed of concentrated nutrient solution was administered to the CSTR. The addition rates of these feeds were determined by occasionally analyzing offline samples for residual amino acid levels in the CSTR and by changing the feed rate to the CSTR if excess (above 30 mmol) or depleted (below 30 mmol) total amino acids were detected. Using the feed protocol described in this paragraph, we believe that cells in the CSTR are unlikely to be excessively restricted in terms of glucose availability, nor are they restricted in terms of availability of any particular amino acid. When the combined N-1 and CSTR reactors reached steady-state or near-steady-state conditions at approximately days 25–28 and 32–35, the feed rate was adjusted accordingly. Figure 3We can observe that the cell effluent from the N-1 bioreactor is approximately 0.43 / day, and the dilution rate of the CSTR production reactor is approximately 0.13-0.15 / day. In some embodiments, the dilution rate of the production bioreactor is approximately 0.05-0.4, 0.1-0.3, or 0.15-0.2 / day. Due to the simulated 1 / 5 volume ratio using these two connected bioreactors, it can be calculated that the cell effluent from the N-1 perfusion reactor contributes approximately 53-62% of the liquid volume entering the CSTR, with the remainder consisting of concentrated culture medium or glucose feed.

[0077] The pipeline from the N-1 perfusion bioreactor to the production bioreactor (CSTR, or chemostat) is first connected and opened at approximately day 12, allowing cells to begin flowing into the CSTR. In an industrial setting, there is no powerful actuator to delay the start-up of the CSTR. It is hypothesized that it could begin at the very first point in time when cell effluent is removed from the N-1 perfusion bioreactor. In the current experiment, the CSTR was started after the N-1 was considered to have reached near-steady-state conditions.

[0078] From day 12 to day 15 (the first 3 days of CSTR operation), all cell effluent from N-1 is pumped into the production (CSTR) bioreactor. This is likely the normal start-up method on an industrial scale, where the N-1 bioreactor is approximately 1 / 5 the volume of the CSTR. However, in small-scale experimental systems, the bioreactor volumes are nearly identical, and this means that a much larger number of cells are added to the CSTR system during the first 3 days of its operation than would be possible on a large scale. This is done on a small scale to accelerate the attainment of high densities in the CSTR, but should have a negligible effect on the eventual steady state achieved in the combined bioreactor system. On day 15, the flow from the N-1 perfusion to the production CSTR reactor is adjusted to precisely simulate a 1:5 volume ratio.

[0079] Around day 23, it first reached relative to most parameters: cell density ( Figure 2A ), cell vitality ( Figure 2B ), bioreactor volumetric flow ( Figure 3 ), concentration of metabolites ( Figure 4 and 5 ), and product concentration and volumetric productivity ( Figure 6 and 7 The steady state of the CSTR was reached. At this point, the volumetric productivity of the CSTR was approximately 0.9 g / L reactor volume / day. For the next 12 days, this volumetric productivity was maintained or slightly exceeded. Figure 7 Although some product screening or selective retention of antibody products occurs across hollow fiber cell retention devices in the N-1 bioreactor ( Figure 6The solid and hollow cubes), but considering the daily liquid volume entering the CSTR from the N-1 bioreactor (~65% of the total volume entering the CSTR) and its product concentration (0.4-0.6 g / L), and a simple mass balance calculation of the daily volume and the product concentration leaving the CSTR on day 23 (6.4-7.4 g / L), the calculated volumetric productivity for the CSTR is primarily (~93-98%) due to product generation in the CSTR. Table 4 lists the details of the composition of the concentrated feed medium, which was manually delivered to the production (CSTR) reactor at various fixed rates from day 19-35.

[0080] Table 4. Composition of concentrated feed medium in the production reactor.

[0081] It is valuable to compare the steady-state volumetric productivity (~0.9 g / L / day) of the combined bioreactor system on day 23 with that of the N-1 perfusion bioreactor, which operates independently as a production bioreactor. Using the product concentration and volume exiting the N-1 perfusion bioreactor, and assuming that downstream operations can capture material from two streams (cell-free permeate and cell-containing effluent), the volumetric productivity of the N-1 perfusion bioreactor on day 23 was approximately 0.46 g / L / day, roughly half that of the combined bioreactor system (0.9 g / L / day). Furthermore, the culture medium consumption rate of the N-1 perfusion bioreactor operating independently on day 23 was approximately 1.95 reactor volumes / day. The culture medium consumption rate of the combined N-1 / CSTR system (including both N-1 perfusion medium and concentrated nutrient feed to the CSTR system) was only 0.44 reactor volumes / day (based on CSTR volume), less than a quarter of the culture medium consumption rate of the N-1 perfusion bioreactor operating independently as a production bioreactor.

[0082] It should be noted that in this particular experiment, high-end pH control of the perfusion rate (HIPCOP) was used in the N-1 perfusion bioreactor, and these control modes are considered optional when a high-end pH control strategy of glucose (HIPDOG) is used intermittently in the CSTR production bioreactor. Alternative or additional control mechanisms can be applied by considering the metabolic conditions required for each cell type cultured using the methods and systems described herein. For example, various CHO cell lines and other mammalian cell expression systems have unique cellular metabolisms, which can be utilized using optional glucose restriction techniques for the connected bioreactors described herein.

[0083] Example 2: Production of IgG (Immunoglobulin G) In the preparation of IgG protein, the β-glutamine synthase expression system CHO (Chinese hamster ovary) cell line was used. The experimental design of the two-stage connected bioreactor system used in this embodiment is the same as that used in Example 1. For experimental purposes, the N-1 perfusion bioreactor, including the perfusion loop (cell retention system), has a working volume of 1.36 liters, and the production bioreactor (CSTR) has a working volume of 1.1 liters. In an industrial setting, the N-1 bioreactor is considered to be approximately one-fifth the volume of the production (CSTR or chemostat) bioreactor. For this reason (in a similar manner to Example 1), the experimental bioreactors are operated to simulate this volume ratio at the start of the experiment. Later in the experiment (day 32), the experimental bioreactor configuration is changed in such a way that they simulate a 1:10 volume ratio, and again on day 44 to simulate a 1:20 volume ratio. In each of these cases, primarily the cell effluent from the N-1 continuous perfusion bioreactor is discharged rather than entering the production bioreactor, solely for the purpose of simulating the theoretical bioreactor volume difference. As the simulated volume ratio increased from 1:5 to 1:10 and then to 1:20, most of the cell volume from the N-1 perfusion bioreactor was discharged.

[0084] Table 5 lists the various process control parameters and set points (pH, dissolved oxygen, temperature, etc.) for the bioreactors used in these experiments.

[0085] Table 5. Operating parameters and control setpoints used in the N-1 continuous perfusion and CSTR production bioreactors.

[0086] As mentioned in Table 5, a dual-jet strategy was used, in which most of the oxygen for the culture was delivered using a 15-micron sintered steel jet injector containing pure oxygen, and most of the carbon dioxide removal was achieved by jetting atmosphere through a perforated (7 x 1 mm) jet injector that generated large bubbles. The air jetting rate through the perforated jet injector varied between 2.5 and 4 times the volume of oxygen used in the 15-micron jet injector. This strategy was sufficient to maintain dissolved oxygen at 40% of the air saturation setpoint and to maintain dissolved carbon dioxide levels between 5% and 13% for the first 13 days of culture, and between 4% and 8% from day 13 onwards (when most of the key data were collected).

[0087] In this embodiment, two N-1 perfusion bioreactors are operated, and each of these reactors independently supplies a continuous source of cells to a production bioreactor operating as a continuous flow stirred tank reactor (CSTR). The target steady-state viable cell density is ~40 x 10⁻⁶ cells / year for the first N-1 perfusion reactor.6 Cells / mL, and ~80 x 10 for the second N-1 perfusion reactor. 6 Cells / mL. The N-1 perfusion bioreactor was started 8 days before the first cells were transferred to the production bioreactor. The production bioreactor (CSTR) was started at full volume, meaning that the bioreactor was completely filled with culture medium when the first cell effluent from the N-1 perfusion bioreactor began to flow into the production bioreactor. The production bioreactor was started at full volume, partly because the experimental protocol for experiments on connected bioreactor systems was designed to explore several different effective dilution rates in the production bioreactor (CSTR) and to attempt to collect steady-state data for each of these conditions. Such experiments (dilution rate studies) are generally facilitated by first exploring high dilution rates before moving to lower dilution rates, which typically require a longer time to reach steady-state conditions. Additionally, low dilution rates can result in lower viability cultures and, when moving later to higher dilution rates with more favorable culture conditions, can lead to long lag times regarding cell response. For this reason, high dilution rates in the production bioreactor were investigated first.

[0088] Starting the production bioreactor at full volume (before adding cells from the N-1 perfusion bioreactor) is likely not the optimal approach for initiating such cultures. From a culture medium and facility utilization perspective, starting the bioreactor at a partial volume may be beneficial, ensuring that when the first material is removed from the production bioreactor (when the bioreactor is full), the cell density, target product concentration, and potential protein product quality are already close to their final steady-state values. This optimal starting volume (which depends on the growth and product production kinetics of any particular cell line) can be determined through routine experiments and computer modeling simulations. The optimal starting volume for the production bioreactor also depends on the optimal (maximum productivity and actual operating conditions) steady-state conditions of the N-1 perfusion and CSTR combined system.

[0089] Two N-1 continuous perfusion bioreactors diverted cells maintained in Erlenmeyer shake flasks at ~1.2 x 10⁻⁶ ppm. 6 1 live cells / mL seeding. The bioreactor was initially operated in batch mode (no perfusion). As cell density increased ( Figure 8 The glucose concentration decreased from approximately 4 g / L to below 0.4 g / L on day 3. Figure 10 Meanwhile, lactate initially accumulates to slightly above 2 g / L ( ). Figure 11Once glucose levels drop to a sufficiently low concentration, cells begin to take up lactate from the culture (day 3), raising the pH of the large volume of fluid and triggering the addition of perfusion medium (and simultaneous removal of cell-free permeate via a cell retention system to maintain a constant working volume in the bioreactor). For a description of this method, see co-pending application 62 / 246,774 entitled “Cell-Controlled Perfusion in Continuous Culture.” This method is referred to as High-End pH Control of Perfusion (HIPCOP). Over the following days, cells continue to control and increase their own perfusion rate (…). Figure 12 After approximately day 8, the perfusion rate stabilized for both bioreactors and remained relatively stable over the duration of the experiment. Throughout the experiment, the perfusion rate was controlled using the HIPCOP technique. From day 10 to day 51, the perfusion rate was controlled at a target depth of 40 x 10 mm. 6 The average perfusion rate of the N-1 perfusion bioreactor can be calculated as 1.76 reactor volumes / day (this is defined as the total volume of culture medium pumped into the bioreactor). From day 10 to day 40, regarding the target 80 x 10 6 The average perfusion rate of the N-1 perfusion bioreactors (cells / ml) can be calculated as 3.60 reactor volumes / day. Beyond day 5, the residual glucose levels in both N-1 bioreactors were very close to zero for the entire experiment, as is typical for bioreactors operated using the HIPCOP technique. Figure 10 ).

[0090] On day 6, the cell-containing culture was initially removed directly from the N-1 perfusion reactor (cell effluent). Most long-duration or sustainable perfusion bioreactor operations require a certain amount of cell effluent to help maintain cell viability in the bioreactor and prevent issues with overall levels of inert biomass. Cell viability data for the N-1 perfusion bioreactor are shown in [data missing]. Figure 9 The viability of both cultures was very high, exceeding 90% for the entire experimental length. Viability was slightly lower in the higher-density cultures. This is likely due to slightly higher shear forces, possibly experienced in the higher-density cultures due to the almost twice-higher gas injection levels necessary to maintain dissolved oxygen and carbon dioxide within appropriate ranges.

[0091] From day 6 to day 8, cell effluent was directed to waste because it was believed that, for experimental purposes, it might be easier to start the production reactor (CSTR) once the cell effluent rate and cell density in the N-1 perfusion culture had stabilized. The cell effluent rate was adjusted daily in an attempt to maintain the cell density in the N-1 perfusion reactor. Figure 8(As close as possible to its 40 and 80 x 10) 6 Target density of cells / mL. Cell efflux rate at... Figure 13 The text states that after day 8, manual adjustments were made, and from day 10 to day 51, adjustments were made to the target size of 40 x 10. 6 The average cell effluent rate of the N-1 perfusion bioreactor can be calculated as 0.72 reactor volumes / day. From day 10 to day 40, regarding the target 80 x 10 6 The average cell effluent rate of the N-1 perfusion bioreactor can be calculated as 0.74 reactor volumes / day.

[0092] On day 8, the cell effluent from the N-1 perfusion reactor was diverted to two independently operating production bioreactors (CSTRs). As mentioned earlier in the text, the correct volume of cell effluent was introduced into the production bioreactors to simulate a 1:5 volume ratio between the N-1 and production bioreactors. Given that the trajectory of cell density and cell effluent rate, as well as the optimal operating conditions for connecting the bioreactors, were known in industrial applications prior to the start of the experiment, immediately diverting the cell effluent (from day one of cell effluent) to the production (CSTR) reactors was likely the most efficient approach.

[0093] Once the cell effluent from the N-1 perfusion reactor is introduced into the production bioreactors (CSTRs), the cell density in those reactors increases very rapidly to over 20 x 10⁻⁶. 6 live cells / mL ( Figure 14 As cell density increases, the flow rate of the concentrated feed added directly to the production bioreactor is increased as necessary to maintain sufficient nutrients for sustained cell growth and protein production. Nutrients are monitored by amino acid analysis (HPLC), and other metabolites (lactate, glucose, ammonia, osmotic pressure, etc.) are monitored using a NovaFlex analyzer (Nova Biomedical, Waltham, MA). The feed rate is adjusted so that the total residual amino acids (excluding alanine, as alanine is often present at high levels in the culture as a metabolic byproduct) are maintained above 30 mmol, and any specific amino acid is not less than 0.2 mmol. The feed consists of concentrated solutions of glucose (50 g / L), amino acids (600 mmol), shear protectant (in 5.12 g / L polyvinyl alcohol), vitamins, and trace elements (some of which are detailed in Table 6).

[0094] Table 6. Details of Culture Medium and Feed Composition

[0095] Dipeptide glycine-tyrosine is also used in the concentrated feed to reduce the complexity of tyrosine solubility. For every 100 mL of concentrated feed medium added to the production bioreactor, 2.5 mL of a 400 mmol / L acidic cystine stock solution is also added.

[0096] Because the purpose of the experiment was to explore different dilution rates in the production bioreactor, it was recognized that a brine diluent solution needed to be fed into the production bioreactor, especially when high dilution rates were being explored. This brine diluent consists of water with 20 mmol of potassium chloride and a balanced sodium chloride solution with a final osmotic strength of 250 mOsm / kg. Therefore, in one embodiment, the production bioreactor is fed with a brine solution to help adjust the dilution rate and also control the osmotic strength of the CSTR production bioreactor. In another embodiment, the CSTR production bioreactor is fed with a brine solution in an amount sufficient to bring the culture medium in the production bioreactor to an optimal final osmotic strength for increased cell productivity. In yet another embodiment, the CSTR production bioreactor is fed with a brine solution in an amount sufficient to bring the culture medium in the production bioreactor to a final osmotic strength of about 100 to about 500 mOsm / kg, or about 150 to about 400 mOsm / kg, or about 150 to about 350 mOsm / kg, or any specific value between these values. In another embodiment, the brine solution added to the production bioreactor is sufficient to produce a final osmotic strength of approximately 250 mOsm / kg.

[0097] Initially, lactate levels in the production bioreactor increased very rapidly (days 8-10). Figure 11 For this reason, once the glucose is depleted in the production bioreactor (day 10), Figure 10 The feed of glucose to the production bioreactors is limited. Glucose levels are controlled by a continuous feed of concentrated nutrients and glucose, which is at a level predicted to be below the amount of amino acids contained in the equilibrium feed (details in Table 6 and as previously described). An additional feed of concentrated 500 g / L glucose is fed directly and slowly to both production bioreactors by means of a pump activated when the pH reaches a high-end setpoint of 7.125. This method of controlling lactate, high-end pH controlled delivery of glucose (HIPDOG), is described in detail in Gagnon et al. 2011, Biotechnol Bioeng 108:1328-37. The HIPDOG control uses a 40 x 10 g glucose sample from the target. 6 The cell effluent from the N-1 perfusion bioreactor was used in the production bioreactor until day 26, and was obtained from a target of 80 x 10 6The cell effluent from the N-1 perfusion bioreactor was used in the production bioreactor until day 22. After these time points, on days 22 and 26, residual glucose levels in the production bioreactor were maintained within a considered unrestricted range, generally above 0.3 g / L, but more typically in the range of 1–3 g / L for the remainder of the experiment via continuous feeding of a concentrated glucose solution (500 g / L). Figure 10 ).

[0098] Figure 14 The diagram shows the live cell density in the production bioreactor. Cell density initially increased rapidly but then began to stabilize between days 10 and 14. The volume and concentration of cells entering the production bioreactor from the N-1 perfusion reactor remained constant from approximately day 10 until approximately day 32. At this point, the volume of cells entering the production bioreactor was halved to simulate a larger bioreactor volume ratio difference of 1:10 (N-1 to production bioreactor volume). This ratio was changed again at approximately day 43 to simulate a 1:20 ratio. The timing of these changes is indicated in text on the graph of cell density in the production (CSTR) bioreactor. Figure 7 ).

[0099] From day 13 to day 18, the nutrient feed volume to both production bioreactors was kept approximately constant. The dilution rate of the production bioreactors (measured in reactor volumes / day or reciprocal time) was... Figure 15 as well as Figure 14 and 17 The figure above (as shown in the text) represents the total volume of cell effluent, concentrated nutrient feed and brine diluent, concentrated glucose feed, and any small amounts of defoamer and alkaline titrant added to maintain the pH at a low level, divided by the reactor volume. For both production bioreactors, the dilution rate remained relatively constant at an average of 0.64 reactor volumes / day from days 13–18. Figure 15 During this period, when the dilution rates of the two production bioreactors remained approximately constant, the cell density ( Figure 14 ), cell vitality ( Figure 16 ), titer or product concentration ( Figure 17 The levels of 1, 2, and most other metabolites reach almost constant values, indicating that steady-state conditions have been reached.

[0100] After day 18, the dilution rate of the production bioreactor over several days was compared with that of the target 40 x 10 6 The culture of cell effluent from the N-1 perfusion reactor was slowly reduced to 0.30 reactor volumes / day, and for those receiving 80 x 10 live cells from the target... 6The cell effluent culture from the N-1 perfusion reactor was slowly reduced to 0.33 reactor volumes / day. This reduction in dilution rate was achieved by gradually decreasing the volume ratio of brine diluent to concentrated feed added directly to the production bioreactor. As before, nutrient levels of the culture were monitored, and the concentrated nutrient feed rate was adjusted as needed to provide sufficient nutrients for varying cell densities.

[0101] By day 21, both production bioreactors had reached the next target minimum dilution rate (0.30 and 0.33 reactor volumes / day). The production bioreactors were maintained at the target dilution rate for another 5 days until most culture parameters had reached nearly constant values. Since the dilution rate being tested at this point was relatively low, and for the remainder of the experiment, waiting until the cultures had reached values ​​that were completely constant for every parameter was impractical.

[0102] From day 26 to day 28, the dilution rate of the production bioreactor was further reduced by decreasing the volume ratio of brine diluent to concentrated feed added directly to the production bioreactor. To maintain a constant, decreasing dilution rate from day 28 to day 31 (0.21 and 0.25 reactor volumes / day as shown in Figure 15), the use of brine diluent had to be completely eliminated. From day 28 onwards, the feed to the production bioreactor consisted only of cell effluent from the associated N-1 perfusion reactor, concentrated nutrient feed, concentrated glucose feed, and trace amounts of antifoaming agent. Because the cells had transitioned to a metabolic state in which no net lactate was produced, negligible amounts of alkali titration were required for the remainder of the experiment. Again, most metabolic parameters stabilized to near-constant values ​​by day 31.

[0103] On day 31, the volume of cell effluent from the N-1 perfusion bioreactor to be added to its associated production bioreactor (CSTR) was reduced by half to simulate a production bioreactor with a volume ten times that of the N-1 reactor. To reiterate, the volume of cell effluent removed from the N-1 perfusion bioreactor was not changed, but a larger portion of this cell effluent was now sent to waste instead of being added to the associated production bioreactor. This change in the simulated volume ratio was necessary to attempt to study dilution rates lower than those of the production bioreactors already tested without limiting nutrient availability to the culture. Following the change in simulated volume, cells were fed from a target 80 x 10 6 The production of cell effluent from an N-1 perfusion bioreactor resulted in a reduced live cell density to just over 50 x 10⁻⁶ cells / mL. 6Cells / mL. When this occurs, the need for concentrated nutrient feed is slightly reduced until the effective dilution rate of the culture reaches approximately 0.15 reactor volumes / day by day 34. This dilution rate remains constant until day 41, at which point the experiment is performed for receiving 80 x 10⁸ cells / mL from the target. 6 The production of cell effluent from the perfusion bioreactor at a rate of 1 cell / mL N⁻¹ is completed. Similarly, the bioreactor receives cell effluent from the target at a rate of 40 x 10⁻⁶ cells / mL. 6 The effective dilution rate of the cell effluent from the N-1 perfusion bioreactor decreased to approximately 0.14 reactor volumes / day by day 34 and remained there until day 43, when most parameters reached steady-state values.

[0104] On day 43, the simulated volume ratio of the N-1 perfusion bioreactor to the production bioreactor was changed again, this time to 1:20. This allows for the intake of 40 x 10⁻⁶ cells from the target. 6 A bioreactor for producing cell effluent from an N-1 perfusion bioreactor was used to explore a dilution rate of 0.09 reactor volumes / day from day 46 to day 50. Figure 15 Also on day 43, the amount added to the target was slightly changed to 40 x 10. 6 The perfusion medium was prepared in an N-1 perfusion bioreactor using cells / mL. The L-lactic acid sodium level in the perfusion medium was 2.1 g / L prior to day 43. At day 43, a switch was made to use a perfusion medium containing DL-lactic acid sodium, a more commercially available and economical alternative to pure L-lactic acid sodium chemicals. It was assumed that the obtained concentrated DL-lactic acid sodium solution contained equal proportions of the D and L-isomers of lactate, and that the cultured cells might only take up the L-isomer form. DL-lactic acid sodium was added to a level of 4.2 g / L, so that the molar concentration of L-lactic acid sodium in the perfusion medium remained at approximately 18.8 mmol. The perfusion medium was prepared both before and after the change to DL-lactic acid sodium with the same osmotic strength. This was achieved by reducing the amount of sodium chloride added to the perfusion medium to compensate for the additional osmotic contribution of the added D-lactic acid sodium. Osmotic strength was measured using a freezing point osmoremeter as a final step before aseptic filtration during the medium preparation process. Sodium chloride is typically added to the culture medium formulation to adjust the isotonic molar concentration of the solution to near the desired value, which in this case is approximately 319 mOsm / kg. Despite attempts to match the osmotic strength of the perfusion medium, a slight increase in the steady-state osmotic strength of the N-1 bioreactor was observed shortly after conversion to a perfusion medium containing DL-lactate, around day 43. Figure 18 ).

[0105] In separate flask tests of pure form sodium D-lactate, it was determined that within the normal operating range (0-5 g / L), the D-lactate form could not be recorded using standard analytical methods with analytical equipment. For this reason, after changing to a perfusion medium containing DL-lactate (day 43), as shown in the target 40 x 10 6 It is not surprising that there was no significant change in lactate levels measured in a per-cell / mL N-1 perfusion bioreactor. Figure 11 ).

[0106] Figure 19 The figure shows the volumetric productivity of the N-1 perfusion bioreactor operating independently as a perfusion bioreactor. This figure assumes the perfusion bioreactor operates as a production bioreactor and that protein products are recovered from all streams leaving the bioreactor (cell effluent and cell-free permeate leaving the cell retention system). The volumetric productivity also accounts for minute variations in product concentration within the bioreactor, occurring primarily when small amounts of selective product concentration occur within the bioreactor. This is mainly only observed with a target size of 80 x 10... 6 It is evident in N-1 perfusion bioreactors with a cell / mL density ( Figure 20 (Days 15-25).

[0107] Even though the operating conditions of the perfusion bioreactors did not change significantly after approximately day 10, and for both perfusion bioreactors, cell density, cell viability, and other culture parameters remained almost unchanged from day 10 to the end of the experiment, the volumetric productivity remained consistent for the target 40 x 10⁴ cells / year. 6 Cells / mL N-1 perfusion bioreactor (from day 12 to day 48) decreased slowly by approximately 29%, and for the target 80 x 10 6 The number of cells / mL in the N-1 perfusion bioreactor (from day 12 to day 40) decreased slowly by approximately 36%. This loss in productivity is likely a result of the moderate genetic instability of the producer CHO cell line B.

[0108] Figure 21The steady-state volumetric productivity of the combined N-1 perfusion connected to the production bioreactor (CSTR) system is shown. Volumetric productivity is plotted against the dilution rate used in the production bioreactor. The steady-state volumetric productivity of the connected N-1 perfusion and CSTR production bioreactor is calculated as follows: Material balance is performed to determine the production rate of the product protein in the production bioreactor. Material balance takes into account the product concentration and liquid volume entering the production bioreactor from the N-1 bioreactor, the rate of increase in product concentration in the production bioreactor, and the product concentration and its removal rate (dilution rate) in the production bioreactor at any time point. This calculation yields the total mass of product produced in the CSTR, which is then plotted against time. Linear regression is then used to more accurately determine the mass production / time / reactor volume values, considering only time points when the dilution rate remains constant and cell density has reached a relatively constant value. This estimate is then added to the production rate of the N-1 bioreactor (again assuming the correct simulated volume ratio of N-1 to the production bioreactor, and considering only the fluid volume entering the production bioreactor) to estimate the final volumetric productivity of the connected N-1 perfusion bioreactor and the production CSTR system for any given dilution rate.

[0109] Multiplying the product concentration in the CSTR production bioreactor at the final time point for any given steady state by the dilution rate (in reciprocal time) will also yield an estimate of the overall volumetric productivity of the connected bioreactor system. However, unless the product concentration remains constant at the time points used, this calculation will give a slight underestimation of the actual steady-state volumetric productivity of the system.

[0110] like Figure 21 As can be seen, for many steady-state conditions, the combined volumetric productivity of the N-1 perfusion and production bioreactors is significantly higher than that of the N-1 perfusion bioreactors operating independently. An exception to this generalization might be... Figure 19 The earliest time point on the target, 80 x 10 6 The perfusion bioreactor achieved a volumetric productivity of 1 cell / mL N⁻¹. This perfusion bioreactor operated at a perfusion rate of 3.6 reactor volumes / day and a very high cell density. Practical operation of such bioreactors at very large scales can present significant challenges due to the volume of culture medium required for preparation and storage, and the difficulty in designing cell retention systems that can function at very large volumes. Furthermore, the very high viable cell density (80 x 10⁻¹⁰) becomes problematic on a large scale because delivering sufficient oxygen and removing sufficient carbon dioxide becomes even more problematic. 6 (cells / mL) could cause a great deal of engineering complexity.

[0111] The N-1 perfusion system, connected to the CSTR production bioreactor, mitigates many of these problems by drastically reducing the size of the perfusion bioreactor and the total volume of culture medium required. The connected bioreactor system maintains approximately the same volumetric productivity as a continuous perfusion bioreactor, which operates at high density and the high perfusion rate required to maintain high cell density. Figure 21 The optimal dilution rates for operating the CSTR production bioreactor occurred over a fairly wide range, from approximately 0.1 to approximately 0.35 reactor volumes per day. This is because... Figure 19 The slow loss of cell line productivity shown, and the fact that high dilution rate conditions were investigated first during the experiment, may be relevant to low dilution rates such as... Figure 21 The steady-state volumetric productivity presented may slightly underestimate the actual achievable productivity.

[0112] Table 7 lists the following: Figure 21 As mentioned above, there are some additional details regarding the maximum volumetric productivity conditions of the two N-1 infusions connected to the CSTR system.

[0113] Table 7. Various process parameters related to the two highest steady-state volumetric productivity of the connected N-1 perfusion and CSTR production bioreactor systems. Comparison of the perfusion bioreactor operating independently as a production bioreactor and as part of a connected bioreactor system.

[0114] When compared to stand-alone perfusion bioreactors, the efficiency of culture medium or feed utilization is significantly higher in connected bioreactor systems. Even when compared to mammalian cell culture bioreactors operating in fed-batch mode, the conditions for achieving the highest efficiency of culture medium utilization are quite high, with 3.83 grams of product protein produced per liter of culture medium or feed used in the connected system as a whole.

[0115] While in the current experiments, the cell effluent rate was manually adjusted daily to ensure that both N-1 perfusion cultures maintained cell densities close to their target values, in an industrial setting using well-understood methods, cell density could be monitored with any number of sterilizable probes that utilize capacitance measurements (or other techniques) to estimate the mass of viable cells in the operating bioreactor. Computer control algorithms could then be programmed to continuously vary the amount of cell effluent removed from the N-1 perfusion bioreactor. It is hypothesized that in an N-1 perfusion connected to a CSTR production bioreactor system, all cell effluent would be immediately transferred to the production bioreactor.

[0116] Because the effective dilution rate of the CSTR production bioreactor depends largely on the volume of liquid entering the vessel from the continuous N-1 perfusion bioreactor, in one embodiment, the cell effluent from the N-1 perfusion bioreactor is further concentrated (by removing additional liquid from the cell effluent) before being transferred into the CSTR production bioreactor. This concentration of the cell effluent is particularly valuable when the N-1 perfusion bioreactor is operating at low cell densities, and / or when it is experimentally determined that a low dilution rate in the CSTR production bioreactor results in higher volumetric productivity for the system as a whole. Further concentration of the cell effluent, if desired, is carried out by any practical cell retention method (continuous microfiltration, acoustic sedimentation, hydrocyclones, etc.).

[0117] Because the cell effluent from the N-1 perfusion reactor carries a significant portion of the total fluid volume into the production reactor (depending on conditions), in one embodiment, a very high nutrient content medium is used to perfuse the N-1 bioreactor to minimize the addition of feed medium to the production bioreactor. Since the cells in the N-1 perfusion reactor cannot consume all the nutrients from the perfusion medium, a large amount of nutrients will subsequently flow into the production bioreactor. However, this capacity is limited because cell growth rates can be quite limited in the presence of very high levels of certain amino acids. Furthermore, because most of the perfusion medium entering the N-1 bioreactor is lost to waste in the cell-free permeate, the additional cost of discarding large volumes of high nutrient content perfusion medium must be considered.

[0118] In a manner similar to the rationale for potentially concentrating cell effluent prior to transfer, highly concentrating the nutrient feed directly to the CSTR production bioreactor is important for the efficient operation of the connected system and for achieving high volumetric productivity. Very diluted feed media can dilute the biomass in the CSTR production bioreactor by increasing the effective dilution rate of the bioreactor. For mammalian cells, which typically have a maximum doubling time close to or greater than one day, any effective dilution rate close to or greater than 1.0 reactor volumes / day (or 1 / day) can result in very low cell densities in the CSTR production bioreactor, and thus low productivity in the connected bioreactor system.

[0119] In the embodiments described above, where the dilution rate is higher than 0.20 reactor volumes / day (1 / day), brine diluents as described in Table 3 are used to dilute the concentrated feed to the CSTR production bioreactor. Figure 21A reasonable volumetric productivity can be achieved using a connected bioreactor system at dilution rates even up to 0.64 reactor volumes / day. High dilution rates dilute the quality of live cells in the CSTR and tend to drive down productivity, but they also simultaneously flush out inhibitory compounds of cell production from the bioreactor system, thus increasing cell growth rates. These competing factors partially balance each other, allowing cell densities to remain reasonably high even at high dilution rates, and therefore, reasonably high volumetric productivity can still be achieved. Thus, in one embodiment, the dilution rate of the production bioreactor is from about 0.1 volumes / day to about 1.25 volumes / day. In another embodiment, the dilution rate of the production bioreactor of the subject technology is from about 0.2 volumes / day to about 1.0 volumes / day; or from about 0.2 to about 0.75 volumes / day; or about 0.2 volumes / day or about 0.65 volumes / day. In another embodiment, the dilution rate of the production bioreactor of the subject technology is about 0.1 volume / day; or about 0.2 volume / day; or about 0.25 volume / day; or about 0.3 volume / day; or about 0.35 volume / day; or about 0.4 volume / day; or about 0.45 volume / day; or about 0.5 volume / day; or about 0.6 volume / day; or about 0.7 volume / day; or about 0.8 volume / day; or about 1.0 volume / day; or about 1.25 volume / day.

[0120] At a reactor volume of 0.64 volumes / day, the average residence time of the product generated in the bioreactor is approximately 1.56 days. This residence time is significantly shorter than the average residence time of the product prepared in a fed-batch bioreactor. For this reason, in the case of unstable proteins, N-1 perfusion connected to the CSTR production bioreactor system may have significant advantages over the fed-batch mode of bioreactor operation. Furthermore, by using a volume-increasing brine diluent added to the production bioreactor, the dilution rate of the connected bioreactor system can be manipulated to reduce the residence time of the product protein without unduly reducing the overall volumetric productivity of the system. The same result can, of course, be achieved by simply manipulating the concentration of the feed medium flowing directly into the production bioreactor. Therefore, in one embodiment, the product residence time in the production bioreactor of the subject technology is from approximately 1 day to approximately 10 days. In another embodiment, the product residence time in the production bioreactor of the subject technology is about 1 day, or about 2.0 days; or about 2.5 days; or about 3.0 days; or about 3.5 days; or about 4 days; or about 5 days; or about 6 days; or about 7 days; or about 8 days; or about 9 days; or about 10 days. In another embodiment, the cell residence time in the production bioreactor of the subject technology is less than about 10 days.

[0121] As mentioned earlier in this paper, a strategy of limiting glucose availability in the CSTR production bioreactor (HIPDOG technique) was used at the start of the experiment. This limited the accumulation of lactate in the bioreactor. Because low dilution rates have proven to be the most productive for CSTRs, it may be valuable to immediately or very rapidly switch to low dilution rate conditions in industrial applications. It is presumed that when the first cells flow from the N-1 perfusion bioreactor into the production bioreactor CSTR, the CSTR bioreactor will be at least partially filled with fresh basal medium. In a highly proliferative state with freely available glucose, these cells may begin to produce large amounts of lactate, and with pH control, this also means high osmotic strength when the lactate is neutralized. Any lactate (and its associated high osmotic strength) that accumulates in the CSTR in the early stages of culture when the bioreactor is full, or that is subsequently produced immediately at very low dilution rates, will be diluted out of the culture system very slowly. For this reason, utilizing glucose-limiting techniques such as high-end pH-controlled glucose delivery (HIPDOG) is of great value in the initial start-up phase of the CSTR production bioreactor. Without methods to control initial lactate production, high levels of lactate can accumulate in the production bioreactor, resulting in little or no growth after only a few days of operation. When CHO cells in the culture are exposed to high lactate concentrations, positive feedback conditions sometimes occur, where lactate production from glucose is significantly upregulated. Such conditions can significantly delay or even prevent the achievement of steady-state conditions with high volumetric productivity at low dilution rates. Although the HIPDOG glucose restriction strategy works well in current experiments, switching to a lower pH control setpoint or a temperature below 36.5°C can be used, or alternatively, to minimize lactate formation in the CSTR production bioreactor for a period of time when it is first started. Both lower pH and lower temperature can slow growth and reduce cell productivity, so the use of these setpoint control changes must be balanced against the need to control lactate formation. For this purpose, the optimal pH setpoint is likely below 7.0, possibly between 6.7 and 7.0. The optimal temperature is likely between 30 and 35°C. It will be apparent to those skilled in the art that various modifications and variations can be made to the described embodiments without departing from the spirit and scope of the claimed method. Therefore, it is contemplated that the methods claimed herein cover modifications and variations to the embodiments described herein, provided they fall within the scope of the appended claims and their equivalents.

[0122] Industrial applicability The apparatus and methods disclosed herein can be used for link-perfusion to CSTR cell culture biomanufacturing and thus for improving industrial methods for manufacturing recombinant and / or therapeutic proteins.

Claims

1. A method for producing a target protein, comprising: (a) In a culture bioreactor, cells containing genes encoding a target protein are cultured; (b) Using the cells obtained from step (a) to inoculate and produce a bioreactor; and (c) The cells are cultured in the production bioreactor under conditions that allow the production of the target protein; The culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor. The method further includes step (d) harvesting the target protein from the production bioreactor; and The production bioreactor described therein does not have a cell retention device.

2. The method according to claim 1, wherein the volume ratio of the culture bioreactor to the production bioreactor is 1:1 to 1:

20.

3. The method according to claim 1, wherein the volume ratio of the culture bioreactor to the production bioreactor is 1:1 to 1:

5.

4. The method according to claim 1, wherein the volume ratio of the culture bioreactor to the production bioreactor is 1:

5.

5. A method for ligating cultures, comprising: (a) In a culture bioreactor, cells containing genes encoding a target protein are cultured; (b) Using the cells obtained from step (a) to inoculate and produce a bioreactor; and (c) Cells are cultured in the production bioreactor under conditions that allow for the production of the target protein. The culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor. The method further includes step (d) harvesting the target protein from the production bioreactor; and The production bioreactor described therein does not have a cell retention device.

6. A ligation culture system comprising: A culture bioreactor used to culture cells containing genes encoding target proteins; and A production bioreactor, which is connected to the culture bioreactor and receives cells from the culture bioreactor as inoculum, wherein the production bioreactor operates under conditions that allow the production of the target protein; The culture bioreactor is a continuous perfusion culture bioreactor, and the production bioreactor is a continuous stirred tank reactor (CSTR) production bioreactor; the production bioreactor does not have a cell retention device.

7. The method according to any one of claims 1-5 or the culture system according to claim 6, wherein the cells are mammalian cells.

8. The method according to any one of claims 1-5 or the culture system according to claim 6, wherein the cells are CHO cells, HEK-293 cells, VERO cells, NSO cells, PER.C6 cells, Sp2 / O cells, BHK cells, MDCK cells, MDBK cells or COS cells.

Images