Cell-controlled perfusion in continuous culture

Cells self-regulate perfusion rates in continuous culture by monitoring pH, using glucose and L-lactate, to optimize productivity and reduce medium consumption, addressing limitations in traditional perfusion methods.

JP7579289B2Active Publication Date: 2024-11-07BOEHRINGER INGELHEIM INT GMBH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2022001006
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2015-10-27
Filing Date
2022-01-06
Publication Date
2024-11-07
Estimated Expiration
2036-05-27

AI Technical Summary

Technical Problem

Existing continuous perfusion cell culture methods face limitations such as high medium consumption, long time to reach peak cell density, and complexity from using cell retention devices, particularly due to lactate accumulation and osmotic strength issues.

Method used

A method where cells self-regulate perfusion rates through pH control, using glucose, L-lactate, and sodium bicarbonate in the perfusion medium, activating pumps based on pH changes to maintain optimal conditions, reducing the need for manual intervention and medium volume.

Benefits of technology

This approach achieves twice the volumetric productivity of traditional methods with modest medium volumes, minimizing osmotic strength fluctuations, and reducing the complexity and cost of cell retention systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007579289000001
    Figure 0007579289000001
  • Figure 0007579289000002
    Figure 0007579289000002
  • Figure 0007579289000003
    Figure 0007579289000003
Patent Text Reader

Abstract

It overcomes current limitations in continuous perfusion cell culture and provides an alternative method to optimize cell growth and viability. [Solution] A continuous perfusion culture process includes: (a) monitoring the pH in a cell culture medium with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) inactivating the delivery of medium and the removal of permeate when the pH falls below a predetermined value.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] Related Applications This application claims the benefit of priority to U.S. Provisional Application Nos. 62 / 168,297, filed May 29, 2015; 62 / 199,388, filed July 31, 2015; and 62 / 246,774, filed October 27, 2015, the entire contents of each of which are incorporated by reference herein as if fully set forth herein.

[0002] [Technical field] The present invention relates to a method for protein production in cultured animal cells, preferably mammalian cells, using continuous perfusion cell culture in a bioreactor. [Background technology]

[0003] Microorganisms or single cells of multicellular species are grown in a variety of ways, including batch, fed-batch, continuous culture, continuous culture with cell retention (perfusion), or combinations of these. Batch culture requires that an inoculum of microorganisms or cells, e.g., yeast, bacteria, or fungi, is grown in a closed culture system with a limited or constant amount of growth medium. In batch culture, there are several growth phases that the microorganism goes through before cell death due to consumption of nutrients in the medium and production of toxic metabolic products. Fed-batch culture is similar to the batch culture method, except that in batch culture, nutrient medium is fed into the fermenter or bioreactor without removing the growth medium or growth products. As expected, the volume of culture fluid in the bioreactor increases over time, but typically higher cell densities are achieved with this culture method compared to batch culture. Continuous culture is another mode of operation in which the bioreactor volume is kept constant, fresh medium is added, and culture fluid is continuously removed. In such cultures, additional cell divisions are likely to occur, especially if there is continuous removal of cells from the bioreactor. In continuous cultures, a cell retention device can also be employed that retains all or nearly all the cells within the bioreactor. When cell retention is employed, cell densities can be much higher in the bioreactor with higher medium inflow and outflow rates to the bioreactor (higher perfusion rates) without the risk of cell washout. It is also possible to combine modes of bioreactor operation.

[0004] Due to limitations in fed-batch volume addition and issues with amino acid counterions, miscellaneous osmolytes, and cell growth inhibitor accumulation, fed-batch mode of bioreactor operation for cell production of protein therapeutics, such as Chinese Hamster Ovary (CHO) cell production, is inherently limited in terms of achievable productivity and cell density. Lactic acid is a growth inhibitor secreted during cell growth and is particularly problematic. It is produced by the breakdown of glucose in cells and causes a drop in bulk pH as its concentration increases. Although lactate can be neutralized by adding base titrants during cell culture, the addition of base results in a large change in the osmotic strength of the growth medium due to ion accumulation. The increase in both osmotic strength due to lactate accumulation and the lactate ion itself can ultimately slow cell growth and cause loss of cell productivity. Other efforts to reduce lactate formation include limiting glucose concentration, substituting alternative carbon sugars (C6 sugars) such as galactose, fructose, or mannose for glucose, and / or manipulating specific enzymes or substrate membrane transporters during cell culture. More recently, high-end pH-controlled delivery of glucose has been shown to limit lactate secretion during fed-batch culture (Luan et al., US Patent 7,429,491 B2 and Gagnon et al., Biotechnol. Bioeng., 2011; 108: 1328-1337).

[0005] Continuous or perfusion culture with cell retention can overcome some of these limitations, but suffers from the disadvantages of large volume medium consumption, long times to reach peak cell density, and the complexity of using cell retention devices. Therefore, there remains a need for alternative culture methods that overcome the limitations associated with fed-batch and / or traditional continuous perfusion culture. Summary of the Invention

[0006] It is an object of the present invention to overcome the current limitations in continuous perfusion cell culture and provide an alternative method to optimize cell growth and viability.

[0007] A method of continuous culture is described that overcomes many of the limitations outlined above, utilizing a unique technique that allows cells to control their own perfusion rate (i.e., self-regulating cells) in a continuous feedback manner. The term "High-End pH Control of Perfusion, or HIPCOP" is used to refer to the process by which cells control their own perfusion rate. For moderately specific productive cell lines, volumetric productivities of about 1 gram / L / day or more can be achieved with very modest medium volumes, relatively simple bioreactor operation, and batch lengths that fit in standard fed-batch windows. The volumetric productivity achieved is more than double the volumetric productivity that might be achieved with the same cell line in an optimized fed-batch culture.

[0008] In its broadest aspect, the present invention relates to a continuous perfusion culture process comprising: monitoring pH in a cell culture medium in a perfusion bioreactor with a pH controller or sensor; activating a medium perfusion pump that delivers fresh medium and a permeate perfusion pump that removes an approximately equal volume of permeate when the pH increases above a set point or a predetermined value; and deactivating the medium perfusion pump and the permeate perfusion pump when the pH drops below a set point or a predetermined value. In other words, in one embodiment, the present invention relates to a continuous perfusion culture process comprising: (a) monitoring pH in a cell culture medium with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) deactivating medium delivery and permeate removal when the pH drops below a predetermined value. In short, pH changes in the cell culture medium are the result of cell metabolism, which then triggers the perfusion process. Alternatively, the method of the present invention may also be described as a cell-controlled perfusion process without the need to measure glucose concentration. The pH triggers for turning on and off the perfusion pump, the fresh medium and the permeate pump are set to a predefined value. This predefined value is about pH 7 (e.g., 6.8-7.4). The medium contains glucose, L-lactate and a specific ratio of amino acids to glucose. According to the invention, activation and deactivation of the medium perfusion pump and the permeate perfusion pump may occur simultaneously or independently. Also, the permeate may be cell-free or cell-containing.

[0009] The described method also includes adding L-lactate to the fresh perfusion medium used in the continuous culture process. In one embodiment, the amount of L-lactate present in the perfusion medium is from about 0.1 g / L to about 7.0 g / L. More preferably, the amount of L-lactate present in the perfusion medium is from about 1 to about 4 g / L, from about 1 to about 3 g / L, or from about 1 to about 2.5 g / L. In a preferred embodiment, the L-lactate is sodium L-lactate or potassium L-lactate.

[0010] In another embodiment, rather than including L-lactate in the perfusion medium, additional sodium bicarbonate is delivered to the fresh medium or to the perfusion bioreactor when the pH of the perfusion medium exceeds a predetermined value (i.e., pH 7, or a pH between 6.8 and 7.4). Typical cell culture media contain about 1.0-2.5 g / L of sodium bicarbonate. In this embodiment, however, when the pH of the cell culture medium exceeds pH 6.8, or pH 7, or pH 7.4, or any pH in the range of 6.8-7.4, additional 1-3 grams of sodium bicarbonate is added to the perfusion medium for every liter of perfusion medium added to the perfusion cell culture bioreactor. Once the pH drops below the predetermined value, the delivery of sodium bicarbonate is stopped or interrupted. Any other physiologically acceptable base (e.g., sodium carbonate, potassium carbonate, HEPES buffer, etc.) known to those skilled in the art may be used in place of sodium bicarbonate, so long as such base is added to the perfusion bioreactor slowly and at a rate that will have an upward impact on the pH in a manner similar to that which would occur due to consumption of lactate from the perfusion medium as lactic acid. For example, sodium carbonate can be added slowly to the perfusion bioreactor in place of sodium bicarbonate, which would be added to the perfusion medium when the pH exceeds a predetermined value. In one exemplary embodiment, sodium carbonate is added to the perfusion bioreactor such that 1 molar carbonate enters the perfusion bioreactor at a rate of 8.7 mL per liter of perfusion medium utilized.

[0011] The fresh medium in this method requires at least glucose and amino acids. In one embodiment, the concentration of glucose ranges from about 0.5 to about 40 g / L. In another embodiment, the fresh medium contains L-lactate. The concentration of L-lactate ranges from about 0.1 to about 7.0 g / L. The ratio of the moles of glucose to amino acids is between about 0.25 and 1.0. In another embodiment, the fresh medium further contains sodium bicarbonate in an amount of about 2 to 5.5 g / L in addition to or instead of L-lactate. In another embodiment, the fresh perfusion medium contains both L-lactate (in an amount of about 0.1 to 7 g / L) and sodium bicarbonate (in an amount of about 2 to about 5.5 g / L) throughout the entire continuous perfusion period according to the present invention, so that the cells can control their perfusion rate.

[0012] Thus, in one aspect, the invention relates to a perfusion culture process comprising: (a) monitoring pH in a cell culture solution in a perfusion bioreactor with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) inactivating the delivery of medium and the removal of permeate when the pH falls below a predetermined value.In one or more embodiments related directly or indirectly to this aspect, the fresh medium comprises L-lactate; the L-lactate is present in the fresh medium in an amount of about 0.1 g / L to 7.0 g / L; the L-lactate is present in the fresh medium in an amount of about 1 to 4 g / L; the L-lactate is present in the fresh medium in an amount of about 1 to 3 g / L; the L-lactate is present in the fresh medium in an amount of about 1 to 2.5 g / L; instead of or in addition to L-lactate, additional sodium bicarbonate is added to the perfusion medium or perfusion bioreactor in an amount of about 1 to about 2.5 g / L such that the total sodium bicarbonate in the medium is about 2 to about 5.5 g / L, and when the pH exceeds a predetermined value, the additional sodium bicarbonate is delivered to the perfusion bioreactor; 1 molar at a rate of 8.7 mL per liter of perfusion medium utilized. Additional sodium bicarbonate is added to the perfusion bioreactor so that carbonate enters the perfusion bioreactor; instead of or in addition to L-lactate, any other physiologically acceptable base, such as sodium carbonate, potassium carbonate, etc., is added to the perfusion bioreactor in an amount that will have an upward impact on the pH in a manner similar to that which would occur with the addition of L-lactate; the fresh medium contains (a) glucose, (b) L-lactate and / or additional sodium bicarbonate, and (c) amino acids; the fresh medium contains (a) about 0.5 to about 40 g / L glucose, (b) about 0.1 to about 7 g / L L-lactate and / or about 2 to about 5.5 g / L sodium bicarbonate, and (c) amino acids in a molar ratio of glucose to amino acid of about 0.25 to about 1.0; the fresh medium contains glucose at about 70 mM The fresh medium contains amino acids and glucose in an amount equivalent to about 5.3 grams per liter of medium; the fresh medium contains glucose at a ratio of amino acids (mM) to glucose (g / L) selected from the group consisting of about 60 to about 4.2, about 90 to about 8, about 100 to about 12, about 120 to about 13, about 240 to about 42, and about 380 to about 70 per liter of medium; the predetermined pH value is about pH 7 or about 6.8 to about 7.4; measurement of glucose concentration in the cell culture medium or addition of glucose to the cell culture medium by a glucose pump is not required.

[0013] In another aspect, the invention relates to a method for achieving rapid cell growth in a perfusion culture process, the method comprising: (a) monitoring pH in a cell culture solution in a perfusion bioreactor with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) inactivating the delivery of medium and the removal of permeate when the pH falls below the predetermined value.In one or more embodiments directly or indirectly related to this aspect, the fresh medium includes L-lactate; the amount of L-lactate is between about 0.1 g / L and 7.0 g / L; instead of or in addition to L-lactate, additional sodium bicarbonate is added to the perfusion bioreactor in an amount of between about 1 and about 2.5 g / L such that the total sodium bicarbonate in the medium is between about 2 and about 5.5 g / L, and when the pH exceeds a predetermined value, the additional sodium bicarbonate is delivered to the perfusion bioreactor; and 1 molar at a rate of 8.7 mL per liter of perfusion medium utilized. Additional sodium bicarbonate is added to the perfusion bioreactor so that carbonate enters the perfusion bioreactor; instead of or in addition to L-lactate, any other physiologically acceptable base, such as sodium carbonate, potassium carbonate, etc., is added to the perfusion bioreactor in an amount that will have an upward impact on the pH in a manner similar to that which would occur with the addition of L-lactate; the fresh medium contains (a) glucose, (b) L-lactate and / or additional sodium bicarbonate, and (c) amino acids; the fresh medium contains (a) about 0.5 to about 40 g / L glucose, (b) about 0.1 to about 7 g / L L-lactate and / or about 2 to about 5.5 g / L sodium bicarbonate, and (c) amino acids in a molar ratio of glucose to amino acid of about 0.25 to about 1.0; the fresh medium contains glucose at about 70 mM The fresh medium contains glucose at a ratio of amino acid (mM) to glucose (g / L) selected from the group consisting of about 60 to about 4.2, about 90 to about 8, about 100 to about 12, about 120 to about 13, about 240 to about 42, and about 380 to about 70 per liter of medium; the predetermined pH value is about pH 7 or about 6.8 to about 7.4; no measurement of glucose concentration in the cell culture medium or addition of glucose to the cell culture medium by a glucose pump is required; about 60 x 10. 6 A viable cell density of 100 / mL is achieved within 4 days.

[0014] Another aspect of the invention relates to a hybrid perfusion / fed-batch process. This process involves using the continuous perfusion process in combination with a fed-batch process. According to the invention, the process will be run in perfusion mode for 4-6 days and then switched to fed-batch mode for another 4-8 days. By using this hybrid mode, a volumetric productivity of 1.0 grams / L / day or more is achieved. In this embodiment, the continuous perfusion process includes the steps of (a) monitoring the pH in the cell culture solution with a pH sensor, (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value, and (c) inactivating the medium delivery and permeate removal when the pH falls below a predetermined value.

[0015] Another aspect of the invention relates to a continuous perfusion bioreactor consisting of two significantly distinct phases. The first phase involves an initial continuous phase as described above (cell density and perfusion rate ramp-up using HIPCOP) utilizing a relatively dilute medium, followed by a second phase of perfusion where the perfusion rate is significantly reduced by utilizing a highly concentrated perfusion medium. Using this two-phase perfusion system, volumetric productivities of about 1.5 grams / L / day or more can be achieved while using very modest volumes of perfusion medium. In the examples shown in this application, no significant cell bleed (removal of cells from the bioreactor) occurs.

[0016] Another aspect of the invention relates to the use of a diluent liquid that is added to the bioreactor during the later stages of perfusion culture when concentrated perfusion medium is being added to the bioreactor. In one embodiment, such a diluent liquid is a saline solution of appropriate concentration. Since long-distance transportation of liquid nutrient medium can pose many challenges (cost of transportation, maintenance of sterility, temperature control), there is great value in using highly concentrated perfusion medium in industrial settings. With such a medium, perfusion rates as low as 0.05-0.30 reactor volumes per day for concentrated perfusion medium are possible. In related embodiments, such bioreactor systems require flushing of product material from the bioreactor, especially if a continuous downstream process is directly connected to capture continuously delivered upstream harvest material (e.g., material passing through a hollow fiber cell retention system) and if the protein being produced is highly labile.

[0017] In another aspect, the invention relates to a hybrid culture process comprising a first continuous perfusion culture process followed by a second continuous perfusion culture process, the first continuous perfusion culture process comprising the steps of (a) monitoring the pH in the cell culture solution with a pH sensor, (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value, and (c) inactivating the medium delivery and permeate removal when the pH falls below a predetermined value, and the second continuous perfusion culture process comprising the steps of (a) adding concentrated medium, and (b) adding diluent. In a related embodiment, the hybrid culture process optionally comprises an intervening fed-batch step. In this embodiment, the first continuous perfusion culture process is performed first, followed by a fed-batch step, and then followed by the second continuous perfusion culture process. In one or more embodiments related directly or indirectly to this aspect, the first culture process is a continuous perfusion culture process; the continuous perfusion culture process is followed by a fed-batch process; the continuous perfusion culture process is followed by a second continuous perfusion culture process; the second continuous perfusion culture process includes the steps of (a) adding an enriched medium, and (b) adding a diluent; the enriched medium includes 600 millimolar amino acids, 90 grams / liter glucose, 0 g / L sodium L-lactate; the diluent is selected from the group consisting of saline and water; the saline or water normalizes the osmotic strength of the medium to about 0-250 and / or 250-350 mOsm / kg; the saline includes 2.0 g / L sodium bicarbonate, 2.4 g / L polyvinyl alcohol, 20 mM potassium chloride, and 80 mM Contains sodium chloride; the hybrid culture process produces volumetric productivity of greater than 1 gram cells / L / day; the hybrid culture process is carried out at culture volumes of about 50 and about 150 L; the culture volume is about 70 L.

[0018] In another aspect, the invention relates to a bioreactor for performing a continuous perfusion culture process, the process comprising the steps of: (a) monitoring the pH in a cell culture solution in the perfusion bioreactor with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) inactivating the medium delivery and permeate removal when the pH falls below a predetermined value.In one or more embodiments related directly or indirectly to this aspect, the fresh medium comprises (a) glucose, (b) L-lactate, and (c) an amino acid; the amount of L-lactate is between about 0.1 g / L and 7.0 g / L; instead of or in addition to L-lactate, additional sodium bicarbonate is added to the perfusion bioreactor in an amount of between about 1 and about 2.5 g / L such that the total sodium bicarbonate in the medium is between about 2 and about 5.5 g / L, and when the pH exceeds a predetermined value, the additional sodium bicarbonate is delivered to the perfusion bioreactor; and 1 molar at a rate of 8.7 mL per liter of perfusion medium utilized. Additional sodium bicarbonate is added to the perfusion bioreactor so that carbonate enters the perfusion bioreactor; instead of or in addition to L-lactate, any other physiologically acceptable base, such as sodium carbonate, potassium carbonate, etc., is added to the perfusion bioreactor in an amount that will have an upward impact on the pH in a manner similar to that which would occur with the addition of L-lactate; the fresh medium contains (a) glucose, (b) L-lactate and / or additional sodium bicarbonate, and (c) amino acids; the fresh medium contains (a) about 0.5 to about 40 g / L glucose, (b) about 0.1 to about 7 g / L L-lactate and / or about 2 to about 5.5 g / L sodium bicarbonate, and (c) amino acids in a molar ratio of glucose to amino acid of about 0.25 to about 1.0; the fresh medium contains glucose at about 70 mM The fresh medium contains amino acids and glucose in an amount equivalent to about 5.3 grams per liter of medium; the fresh medium contains glucose at a ratio of amino acids (mM) to glucose (g / L) selected from the group consisting of about 60 to about 4.2, about 90 to about 8, about 100 to about 12, about 120 to about 13, about 240 to about 42, and about 380 to about 70 per liter of medium; the predetermined pH value is about pH 7 or about 6.8 to about 7.4; measurement of glucose concentration in the cell culture medium or addition of glucose to the cell culture medium by a glucose pump is not required.

[0019] In another aspect, the invention relates to a protein of interest produced by a method comprising: (a) culturing cells comprising a gene encoding the protein of interest in a perfusion cell culture bioreactor under conditions for producing the protein of interest, including: (i) monitoring the pH in the cell culture medium in the perfusion bioreactor with a pH sensor; (ii) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (iii) inactivating the delivery of medium and removing permeate when the pH falls below a predetermined value; and (b) recovering the protein of interest from the cell culture bioreactor.In one or more embodiments related directly or indirectly to this aspect, the fresh medium comprises (a) glucose, (b) L-lactate, and (c) an amino acid; the amount of L-lactate is between about 0.1 g / L and 7.0 g / L; instead of or in addition to L-lactate, additional sodium bicarbonate is added to the perfusion bioreactor in an amount of between about 1 and about 2.5 g / L such that the total sodium bicarbonate in the medium is between about 2 and about 5.5 g / L, and when the pH exceeds a predetermined value, the additional sodium bicarbonate is delivered to the perfusion bioreactor; and 1 molar at a rate of 8.7 mL per liter of perfusion medium utilized. Additional sodium bicarbonate is added to the perfusion bioreactor so that carbonate enters the perfusion bioreactor; instead of or in addition to L-lactate, any other physiologically acceptable base, such as sodium carbonate, potassium carbonate, etc., is added to the perfusion bioreactor in an amount that will have an upward impact on the pH in a manner similar to that which would occur with the addition of L-lactate; the fresh medium contains (a) glucose, (b) L-lactate and / or additional sodium bicarbonate, and (c) amino acids; the fresh medium contains (a) about 0.5 to about 40 g / L glucose, (b) about 0.1 to about 7 g / L L-lactate and / or about 2 to about 5.5 g / L sodium bicarbonate, and (c) amino acids in a molar ratio of glucose to amino acid of about 0.25 to about 1.0; the fresh medium contains glucose at about 70 mM The fresh medium contains amino acids and glucose in an amount equivalent to about 5.3 grams per liter of medium; the fresh medium contains glucose at a ratio of amino acids (mM) to glucose (g / L) selected from the group consisting of about 60 to about 4.2, about 90 to about 8, about 100 to about 12, about 120 to about 13, about 240 to about 42, and about 380 to about 70 per liter of medium; the predetermined pH value is about pH 7 or about 6.8 to about 7.4; measurement of glucose concentration in the cell culture medium or addition of glucose to the cell culture medium by a glucose pump is not required. [Brief description of the drawings]

[0020] These and other features, aspects, and advantages of the claimed methods, apparatus, and systems will become better understood from the following detailed description when taken in conjunction with the accompanying drawings. [Figure 1] 1 is a schematic diagram showing a perfusion device. [Figure 2A] FIG. 1 is a pictorial representation of a hypothetical sequence of events occurring in the bulk cell culture fluid during the growth phase of a bioreactor utilizing the HIPDOG control scheme to limit lactate accumulation in fed-batch culture. [Figure 2B] FIG. 1 is a pictorial representation of a hypothetical sequence of events occurring in the bulk cell culture fluid during the growth phase of a bioreactor utilizing the HIPCOP control scheme for cell-controlled perfusion rate ramp-up in continuous perfusion culture. [Figure 3A] Figure 3 shows results of CHO cell line A growth in a 2L bioreactor (as described in Example 1). Open symbols represent results when using HIPCOP technology. Filled symbols represent non-glucose limited conditions. Perfusion is "self-initiated" when the pH of the culture begins to rise as lactate is first removed from the bulk culture causing an increase in pH. Figure 3A shows viable cell density (E5 cells / mL) over time (days). Figure 3B shows viability (%) over time (days). Figure 3C shows glucose (g / L) over time (days). Figure 3D shows L-lactate concentration (g / L) over time (days). Figure 3E shows osmolality (mOsm / kg) over time (days). [Figure 3B]Figure 3 shows results of CHO cell line A growth in a 2L bioreactor (as described in Example 1). Open symbols represent results when using HIPCOP technology. Filled symbols represent non-glucose limited conditions. Perfusion is "self-initiated" when the pH of the culture begins to rise as lactate is first removed from the bulk culture causing an increase in pH. Figure 3A shows viable cell density (E5 cells / mL) over time (days). Figure 3B shows viability (%) over time (days). Figure 3C shows glucose (g / L) over time (days). Figure 3D shows L-lactate concentration (g / L) over time (days). Figure 3E shows osmolality (mOsm / kg) over time (days). [Figure 3C] Figure 3 shows results of CHO cell line A growth in a 2L bioreactor (as described in Example 1). Open symbols represent results when using HIPCOP technology. Filled symbols represent non-glucose limited conditions. Perfusion is "self-initiated" when the pH of the culture begins to rise as lactate is first removed from the bulk culture causing an increase in pH. Figure 3A shows viable cell density (E5 cells / mL) over time (days). Figure 3B shows viability (%) over time (days). Figure 3C shows glucose (g / L) over time (days). Figure 3D shows L-lactate concentration (g / L) over time (days). Figure 3E shows osmolality (mOsm / kg) over time (days). [Figure 3D] Figure 3 shows results of CHO cell line A growth in a 2L bioreactor (as described in Example 1). Open symbols represent results when using HIPCOP technology. Filled symbols represent non-glucose limited conditions. Perfusion is "self-initiated" when the pH of the culture begins to rise as lactate is first removed from the bulk culture causing an increase in pH. Figure 3A shows viable cell density (E5 cells / mL) over time (days). Figure 3B shows viability (%) over time (days). Figure 3C shows glucose (g / L) over time (days). Figure 3D shows L-lactate concentration (g / L) over time (days). Figure 3E shows osmolality (mOsm / kg) over time (days). [Figure 3E] Figure 3 shows results of CHO cell line A growth in a 2L bioreactor (as described in Example 1). Open symbols represent results when using HIPCOP technology. Filled symbols represent non-glucose limited conditions. Perfusion is "self-initiated" when the pH of the culture begins to rise as lactate is first removed from the bulk culture causing an increase in pH. Figure 3A shows viable cell density (E5 cells / mL) over time (days). Figure 3B shows viability (%) over time (days). Figure 3C shows glucose (g / L) over time (days). Figure 3D shows L-lactate concentration (g / L) over time (days). Figure 3E shows osmolality (mOsm / kg) over time (days). [Figure 4A] Figure 4 shows the results of CHO cell line B growth in a 2L bioreactor using the HIPCOP method of controlling perfusion rate (as described in Example 1). Figure 4A shows either viable cell density (E5 cells / mL) or viability (%) over time (days). Figure 4B shows metabolite concentration (g / L) over time (days). Figure 4C shows osmolality (mOsm / kg) over time (days). [Figure 4B] Figure 4 shows the results of CHO cell line B growth in a 2L bioreactor using the HIPCOP method of controlling perfusion rate (as described in Example 1). Figure 4A shows either viable cell density (E5 cells / mL) or viability (%) over time (days). Figure 4B shows metabolite concentration (g / L) over time (days). Figure 4C shows osmolality (mOsm / kg) over time (days). [Figure 4C] Figure 4 shows the results of CHO cell line B growth in a 2L bioreactor using the HIPCOP method of controlling perfusion rate (as described in Example 1). Figure 4A shows either viable cell density (E5 cells / mL) or viability (%) over time (days). Figure 4B shows metabolite concentration (g / L) over time (days). Figure 4C shows osmolality (mOsm / kg) over time (days). [Figure 5A]Figure 5 shows the results of CHO cell line A growth in a 150 L bioreactor (70 L working volume) using the HIPCOP method (as described in Example 2) to control the perfusion rate. Figure 5A shows either viable cell density (E5 or 105 cells / mL) or viability (%) over time (days) for CHO cell line A in a 70 liter working volume pilot scale bioreactor using the HIPCOP method to control the perfusion rate. Figure 5B shows metabolite concentration (g / L) over time (days). Figure 4C shows osmolality (mOsm / kg) over time (days). [Figure 5B] Figure 5 shows the results of CHO cell line A growth in a 150 L bioreactor (70 L working volume) using the HIPCOP method (as described in Example 2) to control the perfusion rate. Figure 5A shows either viable cell density (E5 or 105 cells / mL) or viability (%) over time (days) for CHO cell line A in a 70 liter working volume pilot scale bioreactor using the HIPCOP method to control the perfusion rate. Figure 5B shows metabolite concentration (g / L) over time (days). Figure 4C shows osmolality (mOsm / kg) over time (days). [Figure 5C] Figure 5 shows the results of CHO cell line A growth in a 150 L bioreactor (70 L working volume) using the HIPCOP method (as described in Example 2) to control the perfusion rate. Figure 5A shows either viable cell density (E5 or 105 cells / mL) or viability (%) over time (days) for CHO cell line A in a 70 liter working volume pilot scale bioreactor using the HIPCOP method to control the perfusion rate. Figure 5B shows metabolite concentration (g / L) over time (days). Figure 4C shows osmolality (mOsm / kg) over time (days). [Figure 6A]Figure 6 shows the results of four different glutamine-sythetase expressing CHO cell lines (cell lines A, B, C, and D) producing IgG antibodies grown in 2L bioreactors using HIPCOP perfusion technology until day 4 and then converted to fed-batch mode for the remainder of the culture for 12 days. Figure 6A shows the viable cell density (E6 or 106 cells / mL). Figure 6B shows the percentage of viable cells measured by trypan blue dye exclusion. Figure 6C shows the L-lactate concentration over time. Figure 6D shows the osmotic strength (mOsm / kg) over time. Figure 6E shows the product titer (IgG antibody) accumulation over time. [Figure 6B] Figure 6 shows the results of four different glutamine-sythetase expressing CHO cell lines (cell lines A, B, C, and D) producing IgG antibodies grown in 2L bioreactors using HIPCOP perfusion technology until day 4 and then converted to fed-batch mode for the remainder of the culture for 12 days. Figure 6A shows the viable cell density (E6 or 106 cells / mL). Figure 6B shows the percentage of viable cells measured by trypan blue dye exclusion. Figure 6C shows the L-lactate concentration over time. Figure 6D shows the osmotic strength (mOsm / kg) over time. Figure 6E shows the product titer (IgG antibody) accumulation over time. [Figure 6C] Figure 6 shows the results of four different glutamine-sythetase expressing CHO cell lines (cell lines A, B, C, and D) producing IgG antibodies grown in 2L bioreactors using HIPCOP perfusion technology until day 4 and then converted to fed-batch mode for the remainder of the culture for 12 days. Figure 6A shows the viable cell density (E6 or 106 cells / mL). Figure 6B shows the percentage of viable cells measured by trypan blue dye exclusion. Figure 6C shows the L-lactate concentration over time. Figure 6D shows the osmotic strength (mOsm / kg) over time. Figure 6E shows the product titer (IgG antibody) accumulation over time. [Figure 6D]Figure 6 shows the results of four different glutamine-sythetase expressing CHO cell lines (cell lines A, B, C, and D) producing IgG antibodies grown in 2L bioreactors using HIPCOP perfusion technology until day 4 and then converted to fed-batch mode for the remainder of the culture for 12 days. Figure 6A shows the viable cell density (E6 or 106 cells / mL). Figure 6B shows the percentage of viable cells measured by trypan blue dye exclusion. Figure 6C shows the L-lactate concentration over time. Figure 6D shows the osmotic strength (mOsm / kg) over time. Figure 6E shows the product titer (IgG antibody) accumulation over time. [Figure 6E] Figure 6 shows the results of four different glutamine-sythetase expressing CHO cell lines (cell lines A, B, C, and D) producing IgG antibodies grown in 2L bioreactors using HIPCOP perfusion technology until day 4 and then converted to fed-batch mode for the remainder of the culture for 12 days. Figure 6A shows the viable cell density (E6 or 106 cells / mL). Figure 6B shows the percentage of viable cells measured by trypan blue dye exclusion. Figure 6C shows the L-lactate concentration over time. Figure 6D shows the osmotic strength (mOsm / kg) over time. Figure 6E shows the product titer (IgG antibody) accumulation over time. [Figure 7] FIG. 1 shows day 12 product titers (grams of IgG antibody / liter) of a hybrid perfusion / fed-batch process using HIPCOP during the cell growth phase compared to optimized titers of a 12-day fed-batch process. [Figure 8] FIG. 8 is a table showing the appropriate volumes of perfusion medium (expressed in reactor volumes of perfusion medium) to be used during the HIPCOP phase of perfusion to achieve the high cell densities and productivities shown in FIGS. 6A-E and 7. The table also shows the calculated highest volumetric productivity that can be obtained from FIG. 7 at any time point in the culture by dividing by the number of days in culture. [Figure 9A]Figure 9F shows cell density, cell viability, L-lactate concentration, osmotic strength, product titer, and rate of perfusion as a function of time for a 2 L continuous perfusion culture of IgG antibody producing CHO cell line A using HIPCOP initially during the cell growth phase (days 1-6) with manually adjusted perfusion rates from day 6 to the end of the culture. In Figure 9F, the total perfusion rate is indicated by filled diamonds, concentrated perfusion medium by filled squares, and saline diluent by open squares. [Figure 9B] Figure 9F shows cell density, cell viability, L-lactate concentration, osmotic strength, product titer, and rate of perfusion as a function of time for a 2 L continuous perfusion culture of IgG antibody producing CHO cell line A using HIPCOP initially during the cell growth phase (days 1-6) with manually adjusted perfusion rates from day 6 to the end of the culture. In Figure 9F, the total perfusion rate is indicated by filled diamonds, concentrated perfusion medium by filled squares, and saline diluent by open squares. [Figure 9C] Figure 9F shows cell density, cell viability, L-lactate concentration, osmotic strength, product titer, and rate of perfusion as a function of time for a 2 L continuous perfusion culture of IgG antibody producing CHO cell line A using HIPCOP initially during the cell growth phase (days 1-6) with manually adjusted perfusion rates from day 6 to the end of the culture. In Figure 9F, the total perfusion rate is indicated by filled diamonds, concentrated perfusion medium by filled squares, and saline diluent by open squares. [Figure 9D] Figure 9F shows cell density, cell viability, L-lactate concentration, osmotic strength, product titer, and rate of perfusion as a function of time for a 2 L continuous perfusion culture of IgG antibody producing CHO cell line A using HIPCOP initially during the cell growth phase (days 1-6) with manually adjusted perfusion rates from day 6 to the end of the culture. In Figure 9F, the total perfusion rate is indicated by filled diamonds, concentrated perfusion medium by filled squares, and saline diluent by open squares. [Figure 9E]Figure 9F shows cell density, cell viability, L-lactate concentration, osmotic strength, product titer, and rate of perfusion as a function of time for a 2 L continuous perfusion culture of IgG antibody producing CHO cell line A using HIPCOP initially during the cell growth phase (days 1-6) with manually adjusted perfusion rates from day 6 to the end of the culture. In Figure 9F, the total perfusion rate is indicated by filled diamonds, concentrated perfusion medium by filled squares, and saline diluent by open squares. [Figure 9F] Figure 9F shows cell density, cell viability, L-lactate concentration, osmotic strength, product titer, and rate of perfusion as a function of time for a 2 L continuous perfusion culture of IgG antibody producing CHO cell line A using HIPCOP initially during the cell growth phase (days 1-6) with manually adjusted perfusion rates from day 6 to the end of the culture. In Figure 9F, the total perfusion rate is indicated by filled diamonds, concentrated perfusion medium by filled squares, and saline diluent by open squares. [Figure 10] This table lists the overall volumetric productivity calculated using the total process length (17 days) as a divisor, assuming all of the product of interest is captured from the final bioreactor harvest and permeate collected from days 6 through 17. Also listed are the total volumes of initial perfusion medium (days 1-6), concentrated nutrient perfusion medium (days 6-17), and saline diluent (days 6-17) used in the experiments in Figures 9A-F. [Figure 11]Figure 1 is a bar graph showing the increase in total process volumetric productivity that can be achieved using HIPCOP technology in conjunction with either the hybrid perfusion fed-batch operating mode (second bar from the left, 12-day process) or the continuous perfusion operating mode (high intensity-low volume perfusion, 17-day process, third bar from the left) compared to the optimized fed-batch process (first bar from the left, 12-day process). All data in this figure was collected using cell line A as described in the previous figures and examples. The calculation of total process volumetric productivity assumes that the product exiting the bioreactor system in the permeate and that remaining in the bioreactor are available. The material in the bioreactor will likely be captured by a final harvest step at the end of the process. The calculation does not include the product lost to the permeate stream during the growth phase (when HIPCOP technology is in operation, days 0-4 for the hybrid perfusion fed-batch process, and days 0-6 for the continuous perfusion process) while utilizing a relatively dilute perfusion medium. Additionally, the calculation utilizes the total batch length (from seeding to the end of harvest) as the divisor. [Figure 12A] Figure 12 shows results for CHO cell line E derived from DG-44 producing a recombinant protein of approximately 130 kDa. Growth occurs in a 3 L bioreactor using the HIPCOP method (as described in Example 1) to control the perfusion rate. Figure 12A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 12B shows metabolite concentration (g / L) over time (days). Figure 12C shows osmolality (mOsm / kg) over time (days). [Figure 12B] Figure 12 shows results for CHO cell line E derived from DG-44 producing a recombinant protein of approximately 130 kDa. Growth occurs in a 3 L bioreactor using the HIPCOP method (as described in Example 1) to control the perfusion rate. Figure 12A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 12B shows metabolite concentration (g / L) over time (days). Figure 12C shows osmolality (mOsm / kg) over time (days). [Figure 12C]Figure 12 shows results for CHO cell line E derived from DG-44 producing a recombinant protein of approximately 130 kDa. Growth occurs in a 3 L bioreactor using the HIPCOP method (as described in Example 1) to control the perfusion rate. Figure 12A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 12B shows metabolite concentration (g / L) over time (days). Figure 12C shows osmolality (mOsm / kg) over time (days). [Figure 13A] Figure 13 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Growth occurs in a 3L bioreactor using the HIPCOP method (as described in Example 1) to control the perfusion rate. Figure 13A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 13B shows metabolite concentration (g / L) over time (days). Figure 13C shows osmolality (mOsm / kg) over time (days). [Figure 13B] Figure 13 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Growth occurs in a 3L bioreactor using the HIPCOP method (as described in Example 1) to control the perfusion rate. Figure 13A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 13B shows metabolite concentration (g / L) over time (days). Figure 13C shows osmolality (mOsm / kg) over time (days). [Figure 13C] Figure 13 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Growth occurs in a 3L bioreactor using the HIPCOP method (as described in Example 1) to control the perfusion rate. Figure 13A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 13B shows metabolite concentration (g / L) over time (days). Figure 13C shows osmolality (mOsm / kg) over time (days). [Figure 14A]Figure 14 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Cultures were maintained in 3L bioreactors using the HIPCOP method of perfusion rate control (as described in Example 1). Figure 14A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 14B shows metabolite concentration (g / L) over time (days). Figure 14C shows average daily perfusion rate (VVD) over time (days). Figure 14D shows cell specific perfusion rate (pL / cell / day) over time (days). Figure 14E shows osmolality (mOsm / kg) over time (days). [Figure 14B] Figure 14 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Cultures were maintained in 3L bioreactors using the HIPCOP method of perfusion rate control (as described in Example 1). Figure 14A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 14B shows metabolite concentration (g / L) over time (days). Figure 14C shows average daily perfusion rate (VVD) over time (days). Figure 14D shows cell specific perfusion rate (pL / cell / day) over time (days). Figure 14E shows osmolality (mOsm / kg) over time (days). [Figure 14C] Figure 14 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Cultures were maintained in 3L bioreactors using the HIPCOP method of perfusion rate control (as described in Example 1). Figure 14A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 14B shows metabolite concentration (g / L) over time (days). Figure 14C shows average daily perfusion rate (VVD) over time (days). Figure 14D shows cell specific perfusion rate (pL / cell / day) over time (days). Figure 14E shows osmolality (mOsm / kg) over time (days). [Figure 14D]Figure 14 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Cultures were maintained in 3L bioreactors using the HIPCOP method of perfusion rate control (as described in Example 1). Figure 14A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 14B shows metabolite concentration (g / L) over time (days). Figure 14C shows average daily perfusion rate (VVD) over time (days). Figure 14D shows cell specific perfusion rate (pL / cell / day) over time (days). Figure 14E shows osmolality (mOsm / kg) over time (days). [Figure 14E] Figure 14 shows results for GS-CHO cell line B producing recombinant immunoglobulin G. Cultures were maintained in 3L bioreactors using the HIPCOP method of perfusion rate control (as described in Example 1). Figure 14A shows either viable cell density (E6 cells / mL) or viability (%) over time (days). Figure 14B shows metabolite concentration (g / L) over time (days). Figure 14C shows average daily perfusion rate (VVD) over time (days). Figure 14D shows cell specific perfusion rate (pL / cell / day) over time (days). Figure 14E shows osmolality (mOsm / kg) over time (days).

[0021] [Detailed Description of the Invention] The present technology relates to a method for continuous perfusion culture by having cells self-regulate the rate of addition of perfusion medium to a bioreactor via pH change, where the perfusion medium contains glucose, L-lactate (and / or sodium bicarbonate) and specific ratios of amino acids to glucose. Advantages of the technology include optimizing process conditions, such as using smaller volumes of liquid medium, to increase protein production.

[0022] The present invention is based in part on the surprising discovery that L-lactate, a potentially growth-inhibiting compound, can be added to the perfusion medium to the advantage of cells and control their perfusion rate over the entire period of continuous perfusion.

[0023] definition The term "about" generally refers to a small degree of error in measurement and is often presented as a range of values ​​that includes the true value within a certain confidence level (usually ±1σ for a 68% confidence interval (CI)). The term "about" may also be stated as an integer and a value of ±20% of that integer.

[0024] The term "about pH 7" refers to pH 7 ± 1 pH unit. In one embodiment, about pH 7 refers to pH 7 ± 0.2 pH units. In another embodiment, about pH 7 refers to a pH between 6.8 and 7.4. In another embodiment, about pH 7 refers to pH 7.10 ± 0.025 pH units. For example, the pH set point (or predetermined value) and dead band during perfusion may be set to 7.10 ± 0.025. At this value, a perfusion pump is triggered at the high end of the range, for example, a pH of 7.125. If the pH rises above 7.125, the perfusion pump is turned on, and if the pH drops below 7.125, the pump is turned off. A separate pump may be activated to add an alkaline solution when the pH drops to the low end of the range, below 7.075.

[0025] Lactic acid, or 2-hydroxypropanoic acid (CH3CHOHCOOH), is an organic acid produced and consumed by certain cells during cultivation. Lactic acid is chiral and has two optical isomers. One is known as L-(+)-lactic acid (chiral, (S)-lactic acid) and the other is its enantiomer, D-(-)-lactic acid (chiral, (R)-lactic acid). A mixture of equal parts of both is called DL-lactic acid.

[0026] L-lactate refers to esters or salts of lactic acid. Lactate is a by-product of culture and is produced as glucose breaks down during cellular respiration. Esters of lactic acid may include, but are not limited to, methyl L-lactate, ethyl L-lactate, butyl L-lactate, ethylhexyl L-lactate, lauryl L-lactate, myristyl L-lactate, or cetyl L-lactate. Salts of L-lactate may refer to alkali metal L-lactates such as potassium L-lactate, sodium L-lactate, lithium L-lactate, or ammonium L-lactate, as well as alkaline earth metal L-lactates such as calcium L-lactate, magnesium L-lactate, strontium L-lactate, or barium L-lactate. In addition, L-lactates of other divalent, trivalent, and tetravalent metals may include zinc L-lactate, aluminum L-lactate, iron L-lactate, chromium L-lactate, or titanium L-lactate. According to the present invention, any ester or salt of lactic acid may be used.

[0027] The term "volumetric productivity" refers to the amount of material produced per hour of run per volume. For mammalian cell culture, this value can be expressed in grams / L / day.

[0028] The term "permeate" refers to the liquid (containing spent medium and expressed protein) that exits a bioreactor after passing through one or more filters, membranes, or other cell retention devices. Depending on the type of filter / membrane or other cell retention device it passes through, the permeate may be cell-free or may contain residual amounts of cells.

[0029] Continuous perfusion During continuous perfusion culture of mammalian cells, the cell mass remains contained within the bioreactor by a cell retention device, while the medium is perfused by the culture fluid (Figure 1). In suspension culture systems, the cell retention device is usually some type of filter, but many other methods can be employed, including some devices that may not be completely "cell-free" (sonic separation, tilted plate sedimentation, external centrifuges, internal filters such as rotating or vibrating, external hydrocyclones, etc.). As the cell mass continues to grow and increase in number and weight, the rate of perfusion is increased to remove metabolic by-products and provide necessary nutrients. The perfusion rate is usually increased in steps, often calculated, to maintain a specific ratio of perfusion medium per time to cell number (CSPR, i.e., cell-specific perfusion rate (often in nanoliters / cell / day, typically in the range of 0.05-0.5 nL / cell / day). See Ozturk, Cytotechnology, 1996; 2: 3-16 and Konstantinov et al., Adv. Biochem. Eng / Biotechnol., 2006; 101: 75-98. In some instances, the perfusion rate is set to control the concentration of glucose or L-lactate (Konstantinov et al., Biotechnol. Prog, 1996; Jan-Feb; 12(1): 100-9 and Ozturk et al., Biotechnol. Bioeng., 1997; Feb 20; 53(4): 372-8) or is based on oxygen uptake rate measurements. See Feng et al., J. Biotechnol., 2006; April 20; 122(4): 422-430.

[0030] Compared to fed-batch cultures producing similar amounts of product protein, continuous perfusion cultures typically utilize much larger volumes of cell culture medium. Because nutrient feeds can be highly concentrated and added in low volumes as in fed-batch cultures, larger volumes of medium are not required to supply nutrients. The larger volumes of medium used in perfusion cultures are typically employed to wash away metabolic by-products of the cells. Cells in perfusion cultures typically experience higher levels of shear and other environmental insults, necessitating at least some growth of cells to replace cells that die. In addition, continuous cultures are expected to operate for much longer periods of time, weeks to months, compared to fed-batch cultures, which may last at most 10-18 days. As a result, cells in continuous cultures should be maintained in conditions that allow at least some cell division, conditions that require inhibitory metabolic by-products to be kept below certain levels.

[0031] The main inhibitory compound produced by mammalian cells is lactate. This is especially true when cells are growing rapidly. Lactate inhibits pH, making it necessary to add a base titrant to keep the pH in the appropriate range for cell growth. Lactate is neutralized to L-lactate, and the ion sodium (or potassium) enters the culture in large quantities as a typical counterion for high pH titrants. While the L-lactate ion is itself inhibitory to growth at high enough levels, the mere presence of L-lactate and additional sodium ions can also significantly increase the osmotic strength of the medium to outside the normal physiological range of 280-320 mOsm (see Gagnon et al., Biotechnol. Bioeng., 2011; 108: 1328-1337). At sufficiently high osmotic strength, cell growth will slow and ultimately the productivity of the culture will decrease.

[0032] Conventional continuous perfusion cultures often ramp up the perfusion rate to flush out accumulated L-lactate and keep cells growing at the beginning of the culture, consuming large volumes of medium that might not have been needed if L-lactate had not been produced.

[0033] When mammalian cells are exposed to freely available glucose in culture (concentrations perhaps above 2 mM), they typically produce high levels of lactate and continue to maintain high glucose consumption rates. However, when glucose levels are low (below 2 mM), the mammalian cells stop producing lactate and instead transport lactate from the bulk medium back across their membranes for consumption. The ultimate uptake of lactate from the bulk fluid causes a rapid increase in the pH of the culture. If the increase in pH triggers the slow addition of glucose to the culture, for example from a pump delivering a glucose-containing nutrient solution, then the high-end pH can be controlled near a steady-state value as glucose (and potentially other nutrients) is delivered to the culture, resulting in the accumulation of lactate and suppressing its deleterious effects on the culture. This process is the basis of the previously described high-end pH delivery of glucose (HIPDOG) shown in Figure 2A (Gagnon et al., Biotechnol. Bioeng., 2011; 108: 1328-1337, disclosing experiments in fed-batch cultures). High-end pH controlled delivery of glucose effectively suppresses L-lactate accumulation, for example in CHO fed-batch cultures. The HIPDOG process can also be shown as a cell-self-determination of the rate of their glucose consumption, as cells indicate a need for additional glucose by uptake of lactate and triggering an increase in pH.

[0034] The HIPDOG fed-batch process can be extended to continuous perfusion culture of mammalian or CHO cells as follows: When the glucose concentration becomes too low in continuous culture, the cells take up lactate, triggering an increase in pH. In fed-batch culture, a pump is activated to deliver a highly concentrated glucose solution; in other words, the glucose concentration in the fed-batch is usually between about 50-500 g / L, whereas in continuous perfusion culture, the pump can be activated to deliver a perfusion medium containing a lower concentration of glucose, e.g., between about 4-40 g / L of glucose. In continuous perfusion culture, the volume of the bioreactor is kept constant. Therefore, every addition of perfusion medium is simultaneously accompanied by the removal of an equal volume of medium from the bioreactor. This can be accomplished, for example, by using a laboratory balance that signals the pump to remove the permeate whenever the weight of the bioreactor exceeds the tare weight. Once glucose levels in the culture are no longer limiting, the cells again secrete lactate, which suppresses the bulk culture pH and inactivates the fed-batch and permeate pumps that deliver and remove medium to the culture, respectively. This cycle repeats over and over as the cells grow and metabolize. As the cell density of the culture increases, the cells are able to trigger the perfusion pumps to turn on more frequently, which allows their perfusion rate to ramp up without any manual intervention.

[0035] During the process, it was observed that the level of L-lactate in the bulk culture fluid dropped over time, and as a result, it was sometimes necessary to supplement the perfusion medium with sodium L-lactate, for example, at about 1-7 grams / L, to prevent the control procedure from failing.

[0036] Additional considerations regarding the selection of nutrient and glucose levels in the perfusion medium were necessary. For example, during the early phase of the culture it was advantageous to rapidly increase cell mass and therefore maintain a high growth rate. To keep the growth rate high, a relatively high rate of perfusion was required to remove growth inhibitors other than L-lactate.

[0037] According to the present invention, high-end pH control of perfusion or HIPCOP allows the cells to determine their rate of perfusion, which ultimately depends on their rate of glucose consumption. At any one time, the volume of perfusion medium delivered to the culture was dependent on the glucose concentration in the perfusion medium. A higher concentration of glucose in the perfusion medium corresponds to a lower volume of perfusion medium delivered. For this reason, there is value in a low concentration of glucose in the perfusion medium when a high perfusion rate is preferred, and a high concentration of glucose in the perfusion medium when a low perfusion rate is preferred. Since many operating parameters (oxygen transfer, carbon dioxide removal, etc.) limit the final peak viable cell density that can be practically maintained in a large-scale perfusion bioreactor, it may be worthwhile to limit cell growth after the initial proliferation phase. If it is desirable to limit additional cell divisions, it may be effective to reduce the perfusion rate and utilize a perfusion medium with a high concentration of glucose that would accumulate high levels of inhibitory molecules in the culture.

[0038] According to the present invention, the glucose concentration used in the basal medium in which the cells are initially seeded is important. If the initial glucose level is too high, it is possible that early in the culture the cells may produce too much lactate, which will slow growth before perfusion can begin. For CHO cell cultures, initial glucose concentrations in the range of, for example, about 2-4 grams / L may be used with the present technology.

[0039] Another consideration for the present technology is the importance of the ratio of glucose to other nutrients (mainly amino acids) in the perfusion medium. To sustain cell growth and / or recombinant protein production rates, this ratio must be balanced to ensure that the amount of amino acids delivered to the culture is not too high or too low. Slow cell biomass production may require a different (possibly lower) ratio of glucose to amino acids. Alternatively, concentrated, slow feeds of pure amino acids or pure glucose could correct inaccurate estimates of the appropriate ratio of glucose to amino acids in the perfusion medium. If pure glucose addition is required, this addition may require linking to a pH controller / sensor as well as a perfusion medium pump.

[0040] Thus, in one aspect, the invention relates to a perfusion culture process comprising: (a) monitoring pH in a cell culture solution in a perfusion bioreactor with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) inactivating the delivery of medium and the removal of permeate when the pH falls below the predetermined value.In one or more embodiments directly or indirectly related to this aspect, the fresh medium comprises L-lactate; the L-lactate is present in the fresh medium in an amount of about 0.1 g / L to 7.0 g / L; the L-lactate is present in the fresh medium in an amount of about 1 to 4 g / L; the L-lactate is present in the fresh medium in an amount of about 1 to 3 g / L; the L-lactate is present in the fresh medium in an amount of about 1 to 2.5 g / L; instead of or in addition to L-lactate, additional sodium bicarbonate is added to the perfusion bioreactor in an amount of about 1 to about 2.5 g / L such that the total sodium bicarbonate in the medium is about 2 to about 5.5 g / L, and when the pH exceeds a predetermined value, the additional sodium bicarbonate is delivered to the perfusion bioreactor; 1 molar at a rate of 8.7 mL per liter of perfusion medium utilized. Additional sodium bicarbonate is added to the perfusion bioreactor so that carbonate enters the perfusion bioreactor; instead of or in addition to L-lactate, any other physiologically acceptable base, such as sodium carbonate, potassium carbonate, etc., is added to the perfusion bioreactor in an amount that will have an upward impact on the pH in a manner similar to that which would occur with the addition of L-lactate; the fresh medium contains (a) glucose, (b) L-lactate and / or additional sodium bicarbonate, and (c) amino acids; the fresh medium contains (a) about 0.5 to about 40 g / L glucose, (b) about 0.1 to about 7 g / L L-lactate and / or about 2 to about 5.5 g / L sodium bicarbonate, and (c) amino acids in a molar ratio of glucose to amino acid of about 0.25 to about 1.0; the fresh medium contains glucose at about 70 mM The fresh medium contains amino acids and glucose in an amount equivalent to about 5.3 grams per liter of medium; the fresh medium contains glucose at a ratio of amino acids (mM) to glucose (g / L) selected from the group consisting of about 60 to about 4.2, about 90 to about 8, about 100 to about 12, about 120 to about 13, about 240 to about 42, and about 380 to about 70 per liter of medium; the predetermined pH value is about pH 7 or about 6.8 to about 7.4; measurement of glucose concentration in the cell culture medium or addition of glucose to the cell culture medium by a glucose pump is not required.

[0041] In one embodiment, a ratio of 70 mM amino acids to 5.3 grams of glucose per liter, or 60:4.2 respectively, is used in perfusion medium during the growth phase. For slower growth conditions, ratios of 90:8, 100:12, 120:13, 240:42 and 380:70 (mM amino acids:grams / L glucose) are used.

[0042] The "cell-controlled" perfusion system of the present invention has important advantages including reducing the volume of perfusion medium required, lowering the load on the cell retention system (e.g., smaller filter area, smaller device size, more efficient separation at low flow rates, etc.), minimizing or eliminating accumulation of sodium and lactate ions in the culture, and minimizing osmotic strength increase. The improved culture conditions also allow for faster cell growth rates that are closer to the maximum growth rate. Low osmotic strength and low levels of lactate are particularly beneficial if the culture is to be transitioned at the end of a fed-batch culture where there is no easy way to reduce accumulated ions. The cell-controlled aspect of the present invention means that the cells will "self-initiate" perfusion and even self-control the perfusion rate ramp-up. Such "cell-controlled" perfusion may allow for near-instantaneous minor corrections of the desired perfusion rate due to minor perturbations in the culture's environment. In many conventional perfusion bioreactors, the perfusion rate is increased based on the cell density reached or on the inhibitory metabolite reaching a given concentration. The need for sampling the bioreactor may be reduced in HIPCOP controlled perfusion systems, since the perfusion rate is controlled continuously by the pH increase signaled almost instantly by the cell metabolism. This opportunity to self-correct is particularly useful if HIPCOP control is to be used for long duration continuous perfusion cultures. In addition, such a 'cell controlled perfusion rate' process requires less time for process development, since the optimal perfusion rate for glucose delivery is determined by the cells themselves.

[0043] Alternatively or additionally, the HIPCOP or "cell controlled perfusion rate" process of the present invention has little or no reliance on glucose measurements, reducing the dependency of the culture process on such values ​​and thus greatly simplifying the process. Furthermore, the HIPCOP process of the present invention does not include a pump to deliver a highly concentrated glucose solution to the bioreactor. In addition, the HIPCOP process uses a smaller volume of perfusion medium than typical conventional perfusion processes (generally less than two reactor volumes total during the cell growth phase), which makes the process even more advantageous than conventional processes.

[0044] Thus, in one embodiment, the present invention relates to a HIPCOP or cell-controlled perfusion culture process, which includes the steps of: (a) monitoring the pH in the cell culture medium with a pH sensor; (b) delivering fresh medium and removing permeate when the pH exceeds a predetermined value; and (c) deactivating the medium delivery and permeate removal when the pH falls below a predetermined value. This process is different from, for example, monitoring the pH in the cell culture medium with a pH sensor; delivering fresh medium and removing permeate when the pH falls below a predetermined value; and deactivating the medium delivery and permeate removal when the pH exceeds a predetermined value. The latter process is based on the idea that if a lot of lactic acid is being produced and the pH drops, the perfusion pump will be turned on. If the incoming medium pH is high, adding medium will actually push the pH of the medium higher, but it will also add excess glucose, triggering more lactate production, continuing the "vicious cycle" until it finally ends up flushing out the lactate with a large amount of perfusion medium. The HIPCOP process of the present invention does not have these limitations.

[0045] The technology of the invention can also be used in N-1 (or seed) bioreactors, where perfusion is used to increase cell density and provide optimal health high cell density inoculation for the production bioreactor. The technology of the invention can also be used in production reactors, either as part of a continuous perfusion culture, or during a short time span of perfusion (in the production reactor) before switching to a conventional fed-batch mode of operation.

[0046] It is advantageous to grow cells quickly to very high densities using short periods of perfusion (continuous culture with cell retention) and then complete the production culture as a simple fed-batch bioreactor. In such a perfusion run of the culture, the perfusion rate would likely be "ramp-up" continuously as the cell density increases, keeping the cells growing at a near-exponential rate. Although adding complexity, it is also worthwhile to perform short periods of perfusion with dilute medium (which would simultaneously add nutrients and remove waste products) followed by very slow perfusion with highly concentrated medium (which no longer "effectively" flushes out inhibitors, but still adds enough nutrients to support production of the product of interest). Such a culture would allow very fast growth to high densities and high productivity, generally minimizing the length of the culture to a similar length as a more conventional fed-batch culture.

[0047] Thus, for a hybrid perfusion / fed-batch process, the process may be carried out as follows (numbers used in this example are taken from process data shown for cell line A as shown in Figures 6-8): For bioreactor seed generation, approximately 5 x 10 6Batches with high seeding densities of cells / mL are cultivated in an initial medium of 4 g / L glucose. After 1 day and 3 subsequent days, the batch is switched to a continuous perfusion process where the cells "self-control" the perfusion ramp. Approximately 2.5 g / L sodium L-lactate is present in the perfusion medium. Perfusion is performed at a working volume of approximately 60 L, which requires approximately 160 L of total perfusion medium. After 3 days, the process is switched to a standard fed-batch process and the feed continues until the bioreactor reaches 100 L on day 12 of the culture. By using this hybrid perfusion / fed-batch process, volumetric productivities of more than 1 gram cells / L / day have been achieved. In addition, if a typical conventional perfusion ramp were used, perhaps 2-3 times more perfusion medium would be required, and additional filter area for the cell retention system may also be required to achieve similar results.

[0048] In one embodiment of a hybrid perfusion / fed-batch process, a perfusion process is performed first for cell culture, and a fed-batch process is performed second. In another related embodiment, the perfusion process starts first and continues for 1-12 days, followed by a fed-batch process of the perfusion process that continues for another 1-12 days. In an exemplary embodiment, the perfusion process continues for 3-5 days (typically starting automatically within 24 hours of seeding), followed by a fed-batch process of the perfusion process that continues for another 9-11 days. The advantages of this hybrid system are very high volumetric productivity, fitting standard fed-batch windows, and fitting into existing facilities (e.g., single harvest). HIPCOP in this system allows the cells to control their own perfusion rate, adjusting to minor process deviations, so that lactate, ammonia, and osmolality are very low just before the start of the fed-batch.

[0049] Another embodiment of the invention relates to a continuous perfusion bioreactor consisting of two significantly different phases: the first phase, as previously described, involves an initial continuous phase (the rise in cell density and perfusion rate using HIPCOP) in which a relatively dilute medium is utilized, followed by a second phase of perfusion in which the perfusion rate is significantly reduced through the use of a highly concentrated perfusion medium.

[0050] Another embodiment of the invention relates to the use of a diluent liquid that is added to the bioreactor during the later stages of perfusion culture when concentrated perfusion medium is being added to the bioreactor. In one embodiment, such a diluent liquid is a saline solution of suitable concentration (e.g., 2.0 g / L sodium bicarbonate, 2.4 g / L polyvinyl alcohol, 20 mM potassium chloride, and 80 mM sodium chloride). Since long-distance transportation of liquid nutrient medium can pose many challenges (cost of transportation, maintaining sterility, temperature control), there is great value in using highly concentrated perfusion medium in industrial settings. With such a medium, perfusion rates as low as 0.05-0.30 reactor volumes per day for concentrated perfusion medium are possible. In related embodiments, such bioreactor systems require flushing of product materials from the bioreactor, especially if a continuous downstream process is directly connected to capture the upstream harvest material that is continuously delivered, and if the protein being produced is highly unstable. In addition, it may be advantageous to control the daily mass of product that the downstream process falls within a small range to prevent the downstream from becoming too large. This can also be facilitated by manipulating the flow rate of the diluent, e.g., saline. It is also advantageous to maintain the osmotic strength of the bioreactor environment close to the physiological range of 250-350 mOsm. Cells often produce exogenous metabolites that increase the osmotic strength of the culture, which can have a negative impact on culture health and cell productivity. In addition, since the perfusion medium at the end of the continuous culture is highly concentrated and there can be variability in the cellular uptake rate of amino acids, some amino acids may accumulate in the bioreactor, which can also have a negative impact on the culture environment due to toxicity or simply due to their contribution to the increase in osmotic strength. Both of these objectives (flushing product or accumulating unconsumed nutrients out of the bioreactor and maintaining adequate culture osmotic strength) may be facilitated by feeding a saline solution to the bioreactor at an optimal concentration (e.g., about 250-350 mOsm / kg or about 0-250 mOsm / kg).Such diluents could be added continuously to the bioreactor and the saline content could be continuously adjusted by addition of sterile water such that a near optimal culture environment could be maintained at any time point in the culture (again, potentially minimizing the need to transport / ship large volumes of liquid). Feedback control using in-line or off-line analysis of osmotic strength and product concentration might also facilitate the addition of appropriate amounts and concentrations of salts and aqueous solutions. Additionally, the saline or diluent solution could be near saturated with sodium chloride or, more preferably, contain a mixture of sodium chloride and potassium chloride near saturation so that additional potassium can be provided to the culture as a nutrient or a more physiologically relevant ratio of potassium to sodium ions can be maintained in the culture. EXAMPLES

[0051] Example 1 Application of HIPCOP technology at 1-2 liter scale

[0052] Numerous tests have been performed using HIPCOP (High-End pH Control of Perfusion). This process was favorable during the perfusion ramp-up period as cell density increased. L-lactate was kept low and osmotic strength was maintained in the optimal range. Figure 3A-E compares two continuous perfusion cultures, one using HIPCOP and one maintained under non-limiting glucose conditions, i.e., additional glucose was added to the culture to maintain approximately 0.5-3.5 g / L glucose. Both cultures had identical set points and dead bands for pH, dissolved oxygen, and temperature control. Perfusion started automatically on days 2-3 when glucose in the HIPCOP bioreactor dropped to low levels, lactate was taken up by the cells, and the bulk pH rose, triggering the start of the perfusion medium addition and permeate removal pumps. The rates of perfusion medium addition to the two cultures were identical and ramped up at the same rate controlled by otherwise identical HIPCOP cultures. Despite the fact that the HIPCOP bioreactor maintained a near-zero glucose concentration while perfusion was occurring, the rates of cell growth were nearly identical in the two cultures. Cell viability was also similar, although slightly lower in the HIPCOP condition (Figure 3B).

[0053] The glucose-non-limited conditions produced excess lactate and a base titrant was automatically added to maintain pH, resulting in a slightly higher overall volume of fluid (9% more fluid by volume) passing through the cell mass compared to the HIPCOP conditions. Although cell densities and growth rates were very similar, the glucose-non-limited cultures produced much more L-lactate and reached much higher osmotic strengths as the culture progressed (see Figures 3D and 3E).

[0054] Figures 4A-C demonstrate the application of the HIPCOP technique to another CHO cell line (cell line B). Again, cells divided at a growth rate close to the maximum growth rate for the CHO cell line. Again, L-lactate and osmotic strength of the culture medium were constrained, providing a culture environment nearly optimal for continued operation.

[0055] A specific example of a fresh perfusion medium that may be used in the present invention contains 8 g / L glucose, 2.5 g / L sodium L-lactate (i.e., 22 mM L-lactate ions) and 90 mM amino acids.

[0056] Example 2 Application of HIPCOP technology at 70 litre scale

[0057] The following data shows that the HIPCOP technique can be readily implemented at the 70 liter scale with similar results as at the 1-2 liter scale.

[0058] The techniques used to develop the present invention at 1-2 liter scale were implemented in a 150 L stainless steel bioreactor system fitted with a scaled-up hollow fiber filtration module and recirculation loop. The surface area of ​​the hollow fiber module was 2.55 square meters and the pore size was 0.2 microns. Liquid was recirculated through the external perfusion loop at 8-9 liters per minute and the working volume of the bioreactor was 70 liters. At pilot scale, the perfusion medium composition was 90 millimolar amino acids, 8 g / L glucose, and 2.5 g / L sodium L-lactate (i.e., 22 mM L-lactate ion) with a final osmolality of 345 mOsm.

[0059] Figure 5A-C shows the application of the HIPCOP technique with cell line A at 70 liter scale. Growth rate and cell viability were slightly lower at large scale compared to previous data from 1-2 L bioreactor. Additional optimization of the equipment used in the perfusion loop of the large scale bioreactor (minimizing cell shear damage) could be possible. The seeding density in the pilot scale experiment was significantly higher than previous data for cell line A from 1-2 L bioreactor. The final cell density reached at pilot scale was somewhat higher than that obtained at small scale. The L-lactate and resulting osmolality of the culture at pilot scale were controlled as in the small scale experiment. As expected as a result of the method of the present invention, the glucose level dropped quickly and remained very low throughout the HIPCOP perfusion. There was no need to manually adjust the perfusion rate during the pilot scale experiment. Only one maximum pump speed was set and the pumps (both fed-batch and permeate pumps) were turned on and off by a high-end pH controller / sensor system to virtually ramp up the culture perfusion rate.

[0060] Example 3 Application of HIPCOP technology to additional cell lines in hybrid perfusion / fed-batch processes

[0061] Further data for four different glutamine synthetase CHO cell lines producing monoclonal antibodies are shown in Figures 6, 7, and 8. All data were obtained with an almost identical process, i.e., a 4-day continuous perfusion period (perfusion ramp-up using HIPCOP), after which the culture process was switched to a fed-batch process for the remainder of the culture for 12 days. Thus, the cells were cultured under a hybrid (i.e., HIPCOP perfusion / fed-batch) process. The perfusion medium composition was the same as that used in Example 2, 90 millimolar amino acids, 8 g / L glucose, and 2.5 g / L sodium L-lactate (i.e., 22 mM L-lactate ion), with a final osmolality of 345 mOsm. As shown in Figure 6, in this hybrid process, the cell lines grew at 67-85 x 10 6A maximum cell density of 100 cells / mL was reached, with percent viability high at 87% to 97% throughout the length of the process. In some pilot-scale (approximately 100 L) hybrid processes, the cultured cells were 95 × 10 6 Densities of over 10000 cells / mL were reached (data not shown). The results are summarized in Figure 6E, where titers between 6 and approximately 12 grams / liter were reached for the four different cell lines. Figure 7 compares the titers achieved for the four CHO cell lines in the optimized fed-batch process with those achieved when applying a hybrid perfusion fed-batch process utilizing HIPCOP. Figure 8 shows a very reasonable volume of medium used during the HIPCOP perfusion stage of the hybrid perfusion fed-batch process. The reactor volume is calculated based on the final volume of the bioreactor that would be required to accommodate the fed-batch that occurs during the fed-batch portion of the bioreactor operation (the fed-batch volume is also shown in Figure 8). These results show that applying HIPCOP technology to an optimized late-stage fed-batch process can more than double productivity with minimal development effort.

[0062] Example 4 Application of HIPCOP technology to continuous perfusion processes

[0063] 9A tracks the viable cell density of a 2 L bioreactor using CHO cell line A in a continuous perfusion process where perfusion was controlled using HIPCOP technology for the first 6 days, then reverted to continuous perfusion at a manually controlled but relatively low perfusion rate over 6 days. The perfusion medium used during the first 6 days was of the same composition as used in Example 3 (90 millimolar amino acids, 8 g / L glucose, and 2.5 g / L sodium L-lactate (i.e., 22 mM L-lactate ion), final osmolality of 345 mOsm). From days 6 to 17, cultures were perfused with a highly concentrated perfusion nutrient solution of 600 millimolar amino acids (90 grams / liter glucose, 0 g / L sodium L-lactate, final osmolality of approximately 1300 mOsm / kg) and saline diluent (2.0 g / L sodium bicarbonate, 2.4 g / L polyvinyl alcohol [shear protectant], 20 mM potassium chloride, and 80 mM sodium chloride, pH of 7.10, final osmolality of 250 mOsm / kg) in varying proportions as shown in FIG. 9F. Subsequent experiments (data not shown) confirmed that similar results could be obtained by removing the sodium bicarbonate and polyvinyl alcohol from the saline diluent. In this case, the level of sodium chloride was increased relative to the diluent (20 mM potassium chloride and 105 mM sodium chloride) to achieve a similar final osmolality of 250 mOsm / kg. To minimize the volumes of liquid that may need to be shipped, on a large scale it may be optimal to use nearly saturated solutions of potassium chloride and sodium chloride in a molar ratio of approximately 1:5, respectively, diluted with additional water as needed.

[0064] Viable cell densities in this system can reach extremely high levels, up to 100 × 10 for several days. 6The HIPCOP perfusion was performed as shown in FIG. 9C, and L-lactate levels were well controlled and remained below 2 grams per liter during the growth phase of the culture. Osmolarity was also controlled as shown in FIG. 9D. Osmolarity control was facilitated by the addition of saline diluent solution at varying rates, beginning on day 6 and ramping up the perfusion rate of concentrated nutrient perfusion medium toward the end of the experiment (FIG. 9F). The rate of perfusion, achieved by adding saline diluent and concentrated perfusion nutrient solution, was manipulated with the goal of simultaneously adding enough nutrients to sustain cell biomass to produce the protein of interest, facilitating removal of the protein of interest from the bioreactor through a cell-retaining hollow fiber filtration device and subsequent downstream processing (at a controlled rate within a reasonably narrow band of total mass / day), and also maintaining the osmolarity strength of the culture near an optimal window of 250-350 mOsm.

[0065] FIG. 9E shows the titer achieved in the bioreactor and in the permeate that passes through the hollow fiber filtration device and leaves the bioreactor system. The concentration of antibody in the bioreactor is higher than that left in the permeate, likely due to protein accumulation on the surfaces of the microfiltration hollow fiber module. Such selective enrichment of the protein of interest inside the bioreactor may be undesirable if continuous downstream processes are used. There may be minor modifications to the cell retention system that can be used to eliminate or minimize this problem (increasing the microfiltration pore size, increasing the filter area, changing the composition of the microfiltration material, more aggressive tangential flow, backwashing the filter), or cell retention methods that are less prone to product retention or less permeable (spin filters, sonic cell settling devices, inclined surface gravity cell settling devices) may be employed.

[0066] Figure 10 lists some important parameters for the continuous perfusion process (high intensity-low volume perfusion). The table shows the overall volumetric productivity of the process (grams / liter of reactor volume / day) and the total volume (in reactor volumes) of the different perfusion media utilized during the entire experiment. As evidenced by the data in the table, the volumetric productivity is very high and the overall volume of perfusion medium is very moderate due to the continuous perfusion process using HIPCOP during the growth phase (days 0-6) and then using saline diluent to control the osmotic strength when highly concentrated perfusion medium is utilized on days 6-17. Figure 11 compares the volumetric productivity of both the hybrid perfusion fed-batch (second bar from the left) and continuous perfusion processes (third bar from the left) (both processes using HIPCOP during the cell growth phase) with the optimized fed-batch process (first bar on the left) that had been developed for the model cell line A. As can be seen, the continuous perfusion process using HIPCOP (high intensity, low volume perfusion) during the growth phase increases productivity per volume by almost 4-fold compared to the optimized fed-batch process.

[0067] In high-intensity, low-volume perfusion processes, a diluent solution is added to the bioreactor while the concentrated perfusion medium is being added to the bioreactor during the later stages of perfusion culture. For example, a diluent solution (saline solution) of appropriate concentration (e.g., 2.0 g / L sodium bicarbonate, 2.4 g / L polyvinyl alcohol, 20 mM potassium chloride, and 80 mM sodium chloride) is added to the bioreactor. Using a diluent solution (e.g., saline) in combination with the concentrated perfusion medium allows perfusion rates as low as 0.05 to 0.30 reactor volumes per day for the concentrated perfusion medium. The addition of diluent may facilitate flushing of product material out of the bioreactor, especially if a continuous downstream process is directly connected to capture the upstream harvest material that is continuously delivered, and if the protein being produced is highly labile. In addition, it may be advantageous for the downstream process to control the daily mass of product to fall within a small range to prevent the downstream from becoming too large. This can also be facilitated by manipulating the flow rate of the diluent, e.g., saline. The addition of diluent can also be advantageously facilitated to control or maintain the osmotic strength of the bioreactor environment above the physiological range of 250-350 mOsm.

[0068] Example 5 Use of a DG44-derived CHO cell line expressing a recombinant protein of approximately 130 kDa and the HIPCOP technique with high sodium bicarbonate in the perfusion medium

[0069] The use of additional sodium bicarbonate in the perfusion medium in place of sodium L-lactate or sodium D / L-lactate was tested with an additional cell line, cell line "E". The perfusion medium contained 90 millimolar amino acids, 8 grams / L glucose, 10 millimolar glutamine, 3.87 grams / L sodium bicarbonate (approximately 1.87 grams / L additional than would be expected in the perfusion medium if 2.5 grams / L of sodium L-lactate were used), with a final osmotic strength of 330 mOsm / kg. The initial basal medium in this experiment consisted of 120 millimolar amino acids, 6 grams / L glucose, 10 millimolar glutamine, and 10 mg / L recombinant insulin. Cells were cultured at approximately 10 x 10 6 Cells were seeded at 2000 viable cells / ml. Cell growth was rapid and perfusion began with the cells approximately 23 hours after seeding. Perfusion continued for approximately 29 hours, and a total of 0.85 reactor volumes of perfusion medium was used for the culture over this time. The perfusion rate for the final 4 hours of perfusion was approximately 1.4 reactor volumes per day.

[0070] As in the previous examples, the HIPCOP method of cell-controlled perfusion worked well with this cell line. Cells were cultured at 47×10 6 The cells grew to approximately 1.3 g / ml of viable cells (Figure 12A). Figure 12B shows that using the techniques of the present invention, glucose was quickly consumed and then maintained at a low concentration. In addition, Figure 12B shows that lactate was controlled at a low level, ending up at approximately 1.3 g / L. Figure 12C shows that the osmotic strength of the culture medium was also maintained very close to the physiologically ideal range of approximately 300 mOsm / kg throughout the time that perfusion was occurring.

[0071] Example 6 Use of a glutamine synthetase CHO cell line expressing recombinant immunoglobulin G and the HIPCOP technique with high sodium bicarbonate in the perfusion medium

[0072] The use of additional sodium bicarbonate in the perfusion medium in place of sodium L-lactate or sodium D / L-lactate was tested with an additional cell line, cell line "B". The perfusion medium contained 90 millimolar amino acids, 10 grams / L glucose, 3.87 grams / L sodium bicarbonate (approximately 1.87 grams / L additional than would be expected in the perfusion medium if sodium L-lactate was at 2.5 grams / L), with a final osmotic strength of 366 mOsm / kg. The initial basal medium in this experiment consisted of 120 millimolar amino acids, 4 grams / L glucose. Cells were cultured at approximately 1.2 x 10 6 Cells were seeded at 1000x viable cells / ml. Perfusion was initiated with the cells approximately 2.3 days after seeding. Perfusion continued for approximately 3.7 days, and a total of 2.09 reactor volumes of perfusion medium were used for the culture over this time. The perfusion rate during the final 4 hours of perfusion was approximately 1.23 reactor volumes per day.

[0073] As in the previous examples, the HIPCOP method of cell-controlled perfusion worked well with this cell line. Cells reached a density of 57×10 within 5.8 days after seeding. 6 The cells grew to approximately 1.47 g / L of viable cells / ml (Figure 13A). Figure 13B shows that using the techniques of the present invention, glucose was quickly consumed and then maintained at a low concentration. In addition, Figure 13B shows that lactate was controlled at a low level, ending up at approximately 1.47 g / L. Figure 13C shows that the osmotic strength of the culture medium was also maintained very close to the physiologically ideal range of approximately 300 mOsm / kg throughout the time that perfusion was occurring.

[0074] Example 7 Use of a glutamine synthetase CHO cell line expressing recombinant immunoglobulin G and HIPCOP technology with sodium carbonate fed-batch

[0075] This is an example of the use of HIPCOP to control the perfusion rate of a "sustainable" continuous perfusion bioreactor. This example starts with a perfusion reactor operating at near steady-state conditions using HIPCOP and lets the cells control their own perfusion rate with sodium L-lactate in the perfusion medium. The composition of the perfusion medium is then changed to remove sodium L-lactate.

[0076] The use of alternative carbonate fed-batch methods in place of sodium L-lactate or sodium D / L-lactate in the perfusion medium was tested with cell line "B". Data for two steady states are shown. Perfusion medium for steady state 1 contained 110 millimolar amino acids, 10 grams / L glucose, 2.6 g / L sodium lactate, 2.0 grams / L sodium bicarbonate with a final osmotic strength of 405 mOsm / kg. Perfusion medium for steady state 2 contained 110 millimolar amino acids, 12.1 grams / L glucose, 2.0 grams / L sodium bicarbonate with a final osmotic strength of 403 mOsm / kg. Note that the total carbon source for glucose and lactate was kept approximately the same between the perfusion medium used during both steady states. Perfusion was maintained by the cells and cell bleed was adjusted once a day to 40×10 6A target viable cell density of viable cells / ml was maintained. The continuous steady-state perfusion rate that the cells are "determining" is approximately 1.0 reactor volume per day. No carbonate was added during steady-state 1. During steady-state 1, the cells apparently consume a significant fraction of the lactate that enters the perfusion medium, since the residual level of lactate in the bioreactor is lower than in the perfusion medium. They presumably consume the lactate as lactate. The cellular consumption of lactate maintains a continuous upward influence on the pH of the medium and allows the HIPCOP technology to function properly. Once steady-state 2 is initiated, this continuous upward influence on the pH of the medium is instead provided by the semi-continuous addition of a 1 molar carbonate solution (in this case the carbonate is a mixture of sodium and potassium carbonate in a molar ratio of 0.94 sodium:0.06 potassium) at a rate of approximately 8.7 mL per liter of bioreactor volume. Therefore, the addition of 1 molar carbonate will come in at a rate of 8.7 mL per liter of perfusion medium utilized.

[0077] As in the previous examples, the HIPCOP method of cell-controlled perfusion worked well with this cell line in both steady-state periods. The average perfusion rate and cell-specific perfusion rate during each steady-state period were 1.0 reactor volume per day and 25 picoliters / cell / day, respectively (Figures 14C and 14D). Figure 14B shows that lactate was maintained in the range of 1.5-1.8 g / L during both steady states. Figure 14E shows that the osmotic strength of the culture was maintained in the physiologically ideal range of approximately 300 mOsm / kg. There was a slight change in the steady-state osmotic strength of the culture from steady-state 1 to steady-state 2, which could be attributed to the additional sodium in the 1 molar carbonate solution entering the bioreactor, which was not fully accounted for when the perfusion medium composition was designed and prepared (a small fraction of sodium chloride normally used in the preparation of the perfusion medium should be excluded).

[0078] In this example, a mixture of continuously added sodium and potassium carbonate was used to provide the upward pressure that consumption of lactate from the perfusion medium would provide to the pH, if sodium L-lactate was present in the perfusion medium, but presumably any suitable, non-toxic, basic substance added continuously or semi-continuously to the culture could have the same effect. Examples of such bases include sodium or potassium hydroxide, among many others.

[0079] Although the present invention has been described in detail with reference to specific embodiments thereof, it will be apparent to those skilled in the art that various changes and modifications can be made to the claimed invention without departing from its spirit and scope. Thus, for example, those skilled in the art will recognize, or be able to ascertain, using no more than routine experimentation, many equivalents to the specific substances and procedures described herein. Such equivalents are considered to be within the scope of the present invention and encompassed by the following claims.

[0080] [Industrial application field] The devices and methods disclosed herein are useful for perfusion biomanufacturing and therefore for improving industrial processes for producing recombinant, therapeutic proteins.

Claims

1. A bioreactor for carrying out a continuous perfusion culture process, the continuous perfusion culture process comprising: (a) monitoring the pH in the cell culture medium with a pH sensor; (b) delivering fresh medium and removing the permeate when the pH exceeds a predetermined value; and (c) inactivating medium delivery and permeate removal when the pH falls below a predetermined value Including, The fresh medium contains L-lactate, 1. The bioreactor comprising: (i) a perfusion bioreactor; (ii) a cell retention device; (iii) pH controller (iv) a medium perfusion pump that supplies fresh medium; (v) a permeate perfusion pump to remove the permeate (vi) a reservoir containing fresh medium Including, (vii) if the pH increases above a set or predetermined value, the pH controller activates a medium perfusion pump to supply fresh medium and a permeate perfusion pump to remove permeate; and when the pH falls below a set value or a predetermined value, the pH controller deactivates the medium perfusion pump and the permeate perfusion pump; the cell is a mammalian cell; The bioreactor.

2. Fresh medium is (a) glucose, (b) 0.1 to 7.0 g / L L-lactate, and (c) amino acid 2. The bioreactor of claim 1 comprising:

3. The bioreactor according to claim 2, wherein glucose is supplied at 0.5 to 40 g / L.

4. 3. The bioreactor of claim 2, wherein the amino acids and glucose are provided in an amount equivalent to 70 mM amino acids and 5.3 grams of glucose per liter of medium, or in a ratio selected from the group consisting of 60:4.2, 90:8, 100:12, 120:13, 240:42 and 380:70 mM amino acids:grams / L glucose per liter of medium.

5. 60×10 6 2. The bioreactor of claim 1, wherein a viable cell density of 100 / mL is achieved within 4 days.

6. 2. The bioreactor of claim 1, wherein no measurement of the glucose concentration in the cell culture medium or no addition of glucose to the cell culture medium by a glucose pump is required.

7. 2. The bioreactor of claim 1, wherein the predetermined pH value is between 6.8 and 7.4.