Class Lecture 3_Bioprocess Engineering.pptx

Class Lecture 3_Bioprocess Engineering

1. Upstream Bioprocessing

2. Developing a robust, high-yielding upstream process is a complex and challenging task, and a good result is a combination of powerful technology, high-quality starting materials, and clever process design. A thorough understanding of the basic process concepts and the underlying biological phenomena is required to succeed, not just in the upstream production stage, but in the overall bioprocess. An intriguing and challenging characteristic of upstream processes is the fact that living organisms are used, and they do not always behave as predicted or wanted. During the past decades, tremendous progress has been made in upstream processing, and modern fed-batch processes are capable of delivering 10 g/L of monoclonal antibody or more. This development can be explained by an increased understanding of cell cultures, which has led to better cell culture media, more advanced feeding strategies, more robust cell lines, and bioreactor control tailored for specific processes. The augmented upstream knowledge can further be used in systems approaches that target the entire bioprocess.

3. KINETIC MODELS Process characterization and modeling are important tools to optimize and document processes and to develop control and automation strategies. Models also form a core part of the “quality by design” ( QbD ) framework. Recent developments that have led to an increased knowledge about, for example, the metabolic pathways in animal cells, in combination with more powerful computers, have facilitated the development of very sophisticated and advanced upstream models. However, the more complex the process is, the more difficult the modeling becomes, and the modeling of bioprocesses is still in its infancy compared with modeling in many other industries. One underlying reason is that despite continuous progress, we still don’t understand the full complexity of a living cell.

4. The concepts described below are included to provide an elementary understanding of modeling that can be applied to upstream bioprocessing. These models are non-segregated and unstructured. Non-segregated means that all cells are assumed to behave in the same way. This is not true for most bioprocesses, but this assumption simplifies the calculations considerably. Unstructured means that the reaction rates are described only by input and output that depends on the culture environment. In other words, the cell is treated like a black box where it is known what comes in and what comes out. What happens in-between is not considered. The construction of a process model can be divided in the following basic steps: Defining system variables Performing a kinetic analysis Setting up mass balances Validating the model

5. In the first step, variables are defined for the system. There are different variable categories: state variables that characterize the system such as cell concentration, substrate and product concentrations, and operational variables that represent specific conditions, for example, an initial concentration or a feed rate. Some variables for upstream bioprocessing are seen in Table 5.1. In the second step, a kinetic analysis is performed. When the kinetic analysis is done, the changes of state variables are investigated, and the key parameters that will influence those are identified. Typically, at this stage, the specific rates for cell growth, cell death, substrate consumption, and product formation, as well as yield coefficients, can be calculated. This analysis can be done in various cultivation systems, such as small-scale bioreactors or shake flasks. However, care must be taken so that the system is not artificially limited, for example, by low oxygen limitations or suboptimal pH. Very accurate kinetic analyses can be done during steady state conditions, for example, in chemostat cultivations . Commonly batch or fed-batch approaches are used instead due to their simpler setup.

6. In the third step, mass balances are established. In this step, the operational variables are considered and arranged depending on the type of process (batch, fed-batch, perfusion, chemostat, etc.) Parameter fitting is done to minimize the discrepancy between the values calculated through the model and the experimental values. This can be done with, for example, changing values of constants. The final step is to validate the model and compare the predicted values with the reality. This is done through several experiments under different conditions to understand the applicability and how well the model predicts the outcome of the process. The primary validation is done in small scale, and this is verified in larger scale.

7. A GENERAL MASS BALANCE Mass balances are the basis for developing a good understanding of cell growth, substrate utilization, and product formation. They are also very useful for determining optimal bioreactor control parameters. A mass balance can be done for any type of cultivation vessel, and the basis for the general dynamic mass balance is the following equation: Change = In - Out+ net reaction An overview of commonly used variables in upstream bioprocessing.

8. In and Out represent flow of a compound into and out from the bioreactor vessel, respectively. This can be exemplified with a bioreactor with a volume V and an arbitrary state variable y. Y could be, for example, product concentration or concentration of glucose. Then the mass balance becomes the following basic equation

9. δ is a separation factor that is used in the case of recirculation; if no recirculation occurs, then δ=1. It is also assumed that the vessel is completely mixed, and that there are no concentration gradients. This generic mass balance is then modified according to the specific process and the variable y that is being considered. The volumetric reaction rate r, is further written according to Eq. (5.2): where q is the specific rate by which a reaction is catalyzed, and X is the concentration of the biocatalyst, for example, cell concentration. This simple equation can be used to describe, for example, cell growth and death, substrate consumption, and production rates.

10. CELL GROWTH Cell reproduction is called proliferation, and this is done through a series of highly regulated events during a cell cycle. The cell cycle is regulated differently in different organisms. In bacteria, proliferation is tightly linked to cell size, because critical functions such as intracellular transport and nutrient uptake are constrained if a bacterial cell becomes too large. Bacterial DNA replication starts at the cell wall near the mid-point of the cell, and a new round of replication can be initiated before the first one is completed, a feature which enables very fast growth. For example, a fast-growing E. coli culture can have a doubling time of 15–20min. The eukaryotic cell cycle, by contrast, is longer and much more complex. The eukaryotic cell cycle can be divided into two main parts: interphase and mitosis. During interphase, the cell increases in size, doubles the genome, and prepares for division. Interphase can be further divided into the gap phases G1 and G2, and the DNA synthesis phase S. Mitosis is the part of the cell cycle when the newly replicated sister chromatids are separated into two nuclei. When mitosis and the subsequent cell division, cytokinesis, are complete, then the original cell has divided in two identical daughter cells.

11. Many cell cycle control mechanisms have been conserved throughout all eukaryotic cells, including plant cells. Multiple checkpoints are included where the eukaryotic cell will ensure that the conditions are favorable for division and that the resulting daughter cells are replicas of the original cell. If this is not the case, the cell cycle will be halted. If the conditions become favorable again, then cell cycle progression will continue, but otherwise, the cell can remain in a non-dividing quiescent stage, or alternatively go into programmed death, apoptosis. The chosen path depends on multiple factors, for example, if external growth factors are present, if nutrients are available, and if the DNA is damaged. The cell cycle is under strict regulation in non-transformed cells, but many cell lines used for biomanufacturing are transformed and carry mutations in genes that control cell cycle progression. This means that critical control proteins can be dysfunctional or transcribed constitutively, and a new cell cycle can be initiated in the absence of an external growth factor. Consequently, transformed cells are easier to handle, less dependent on external growth factors, and can often be grown in less complex culture medium, for example, medium without serum.

12. The time to complete one cell cycle varies between different organisms and between cell lines from the same organism. Typically, it is between 12 and 24h for animal cells in culture, but the variations are considerable. The cell cycle time will not just depend on the cell type, but also on the history of a specific cell line and the conditions by which a culture is maintained. When a culture is transferred to a new laboratory, there will be a selection for cells with a phenotype that is suited for the cultivation conditions in that specific laboratory. Any difference in media composition, incubator temperature, gassing scheme, maintenance routine, or operator handling will impact the culture. This can have implications for many cell characteristics including the growth pattern, and thorough documentation of the cell maintenance is required to minimize the risk for changes that can impact the process. A production cell line should have a short doubling time, as this will simplify scale-up and minimize the required batch time in the production bioreactor.

13. A short batch time will contribute to a large overall number of batches that can be produced annually, and a fast-growing cell line can also simplify the seed-train process. The doubling time can be decreased, for example, through optimization of the cell culture medium and repeated re-inoculation to select for a fast-growing clone. Optimization of the bioreactor parameters are also important to prolong the time during which the culture grows exponentially. From a process perspective, a culture can be divided into the following four stages: lag phase, exponential growth phase, stationary phase, and death phase (Fig. 5.2A). During the lag phase, the culture adjusts to the new environment before proliferation starts. A too-low start cell concentration can prolong the lag phase, especially with serum-free culture. It has been hypothesized that for some cell types, this is linked to an autocrine system and the absence of growth factors released by the cells themselves. Next, the culture enters the exponential growth phase. This phase can vary in length, depending on the cell and the specific growth conditions. After this comes a period with growth declination and a stationary phase where new cells are formed at an equal rate as cells die.

14. A typical growth curve with the five growth phases indicated: (A) lag phase, (B) exponential growth phase, (C) stationary phase, and (D) death phase. The exponential growth phase can be very short for an animal cell culture, whereas a microbial culture can grow exponentially for a prolonged period unless it is limited by, for example, availability of nutrients or oxygen. (B) A typical growth curve and specific rate of proliferation from an animal cell culture illustrated by an Sf9 S. frugiperda insect cell culture.

15. The last phase is the death phase. In a biomanufacturing process the goal is to keep the culture in a state where productivity is maximized and viability is high. If the product formation is growth-associated, the growth phase should be maximized, whereas in the case of non-growth associated product formation, considerable productivity gains can be achieved from prolonging the stationary phase. However, implications for the downstream process should always be considered when the upstream process is designed. A very late harvest can lead to challenges downstream and/or a compromised product quality. To ensure genetic and phenotypic stability and high viability, animal cell lines should be kept in the exponential growth phase during maintenance. This means that the cells should be subcultured on a regular basis before they enter the stationary growth phase. The split ratio during subculturing typically varies from 1:2 for primary cultures to 1:10 or even higher for very fast growing, continuous cell lines. The passage number is the number of times the cells have been split from when the cell line was first established. This is important information for biomanufacturing processes, because with each passage comes a risk for genetic drift.

16. where X is the concentration of biomass or cells. Common practice is that X, in microbial fermentation, represents the biomass concentration (g/L), and in animal cell cultivation, the viable cell concentration (number of cells/mL). As can be seen, the growth rate is very low during the lag phase, after which it peaks and then it declines again during the stationary phase. The passage number should be recorded for each experiment, and processes must be validated with cells at a defined passage number. A useful tool to monitor the process and determine the culture status is the specific growth rate μ . In microbial fermentation, the specific growth rate is calculated from dry-weight data. However, accurate dry weight is challenging to retrieve from animal cell cultures due to low cell densities, and here the cell number is used instead. μ is calculated from a curve fitted to the experimental data according to Eq. (5.3)

17. CELL DEATH Cell death decreases overall product yield, and this can be a problem in biopharmaceutical processes. Animal cells die in two different ways, through necrosis or apoptosis. Necrosis is a sudden death caused by very unfavorable conditions that physically damage the cell such as extreme pH and high shear. A necrotic cell is unable to maintain the osmotic pressure across the cell membrane, which leads to cell swelling followed by lost membrane integrity. In the final necrotic stage, the intracellular components leak out into the surrounding environment. Apoptosis is a programmed, genetically controlled, suicide-like death. Nutrient starvation, oxygen limitation, build-up of toxic metabolites, loss of attachment (for adherent cell lines), and growth factor deficiency are examples of events that can induce apoptosis.

18. Apoptotic cells are commonly seen in the final stages of a batch culture. Apoptosis is managed by a class of enzymes, caspases, which crosslink proteins in the cytoplasm, cleave the DNA in smaller pieces, and assemble the interior of the cell into smaller, membrane bound vesicles, apoptotic bodies. In vivo, these vesicles are easily digested by phagocytes and removed with no inflammatory response. However, in a culture, no phagocytes exist, and the apoptotic bodies will remain in the culture medium until they are removed in the downstream purification process. With time, the vesicles might end up in a secondary apoptotic stage where membrane integrity is lost, and the contents are released. A variety of methods can be used to detect and monitor apoptosis, both on a culture level and in individual cells.

19. The goal in biomanufacturing is to inhibit or slow the onset of cell death and keep the viability of the culture high, typically above 90%–95%, over an extended period to maximize productivity and facilitate downstream purification. In addition, a culture with high viability does not generate large amounts of cell debris, which simplifies the purification process. Consequently, the most efficient downstream purification process starts with an upstream culture with high viability. Two main strategies have been used to prevent cell death: medium optimization (nutrient supplementation) and genetic engineering. Medium optimization has been performed through supplementation of, for example, glucose, glutamine, insulin, IGF-1, transferrin, and amino acids.

20. Genetic engineering has included overexpression of the transcription factor E2F, overexpression of the anti-apoptotic Bcl-2 protein, caspase inhibition, overexpression of human telomerase reverse transcriptase (hTERT), and pro-apoptotic factor knockouts. Overexpression of E2F, for example, makes the cell line less dependent on external growth factors. However, the apoptotic machinery is closely linked to the overall mechanisms for proliferation, and changing a key factor might have unexpected implications, for example, for cell line stability. Another strategy to consider for minimizing apoptosis is to optimize process monitoring and control. Fast-responding and accurate sensors used together with well-tuned control loops will minimize fluctuations in the culture and consequently decrease the risk for apoptosis. A good understanding of the optimal temperature, pH, and DO levels is also required.

21. METABOLIC CONCEPTS Efficient design and optimization of bioprocesses require a good understanding of cell metabolism. This is beneficial, for example, for cell culture media optimization, engineering of high-productive cell lines, and development of robust bioreactor control strategies. There is growing evidence that the metabolism and the signaling pathways that control proliferation are connected more closely than what was previously thought. For example, the metabolic demands of an actively proliferating cell will differ from those in a cell in a non-dividing, quiescent state. The metabolism is highly regulated in undifferentiated cells, but genetic alterations and transformations that are used, for example, when developing cell lines for biomanufacturing, can disturb this natural pattern.

22. The consequence can be an increased uptake of nutrients, for example, glucose and glutamine, that exceed the cell’s bioenergetic demand. This leads to unwanted byproduct formation and variability in manufacturing processes. The complete cellular metabolism is immensely complex with thousands of reactions, interrelations, and dependencies, and much is yet to be discovered despite considerable progress during the past decades. Metabolic profiling and Metabolic Flux Analysis (MFA) are examples of approaches that are increasingly used to increase the understanding of metabolic reactions. The goal is to find opportunities to intervene in the metabolism, steering it towards robust growth and high productivity.

23. GLUCOSE, GLUTAMINE, LACTATE AND AMMONIA For most biomanufacturing cell lines, only glucose and glutamine are catabolized in larger quantity, and lactate and ammonia are the two major byproducts. Glucose serves as the main carbon source, and glutamine is the main source of nitrogen. Glycolysis is a sequence of reactions by which glucose is converted to the key intermediate pyruvate, which is further metabolized depending on the environment and specific cellular needs. The glycolysis pathway is conserved in all kinds of organisms and cells, but the fate of pyruvate is variable. Two pyruvate reactions are of primary importance for cells from multicellular organisms: reduction to lactate and oxidation in the tricarboxylic acid (TCA) cycle, also known as the citric acid cycle or the Krebs cycle. In yeast and some other microorganisms, reduction to ethanol is a third, important reaction. Lactate is not inhibitory in concentrations normally seen in culture, typically below 35–40mM, but excessive lactate production is an indicator for a poor overall process and low product yields. The main problem associated with lactate is that it acidifies the culture medium, and base addition is required to maintain the pH set-point, causing an unwanted osmolality increase.

24. The lactate metabolism is cell line and clone specific, and the amount produced is highly dependent upon the nutrient situation and the culture stage. Most cell types produce only a small amount of lactate in an aerobic environment. However, fast-growing cells like cancer cells and many cell lines used for bioproduction can convert a substantial quantity of glucose to lactate also under fully aerobic conditions. This phenomenon is called aerobic glycolysis or the “Warburg effect”. It has been hypothesized that glutaminolysis , the conversion of glutamine to lactate, also contributes to high lactate production at high proliferation rates. Glutaminolysis occurs in all cells, but it is upregulated in immortalized cells and many types of cancers. It has been speculated that the altered glucose and glutamine metabolism in fast-growing cells is caused by an increased need for building blocks required to sustain a high proliferation rate.

25. In a bioreactor, a net production of lactate is commonly observed during the exponential growth phase. When the growth declines in the stationary phase, some (but not all) cultures switch from lactate production to lactate consumption. This switch contributes to an overall low lactate concentration in the final process stage, and this phenomenon has been linked to high productivity and process robustness. The lactate-consuming phenotype has been associated with a high mitochondrial oxidative metabolism , but the complete mechanisms behind the switch remain to be understood. Selecting a clone with a low net production of lactate is a good starting point for keeping lactate under control in an animal cell biomanufacturing process. Several other complementary strategies have been used to decrease lactate production (Table 5.2).

26. The classic technique is to limit the amount of glucose in the culture medium. However, a too-low glucose concentration will impair cell growth, and the target level should include a safety margin. In practice, the glucose concentration is often kept at 1–6g/L. Another option is to use alternate carbon sources instead of glucose, but this can lead to slower growth and impaired productivity. Enzyme manipulation has been investigated, but in some cases, the effects were only temporal. Some of the recently developed, promising techniques for lactate control are based on adaptive feeding. In these processes, the feeding rate is adjusted based on real-time measurement of a specific bioreactor parameter. Regardless of the strategy chosen, it is critical that it is fully scalable, and this can be investigated in, for example, scale-down models.

27. Ammonia is a second key metabolite for animal cells, and it has a more direct toxic effect compared with lactate. Ammonia can be formed in two ways: as a byproduct in glutamine metabolism or from decomposition of glutamine in the culture medium. The spontaneous decomposition is temperature dependent, and it increases with an increased temperature. Ammonia concentrations as low as 2 mM can be toxic for sensitive cell lines. Reported effects from high ammonia concentrations include cessation of cell growth, impaired productivity of recombinant protein, inhibition of virus propagation in cells, and changes in the glycosylation pattern. For example, a decrease in sialylation of all glycans, and a decrease in the proportion of the O-linked glycan was observed in one study. It has been hypothesized that the observed glycan heterogeneity at high ammonia concentrations is caused by an increased pH in the Golgi apparatus and a subsequent inhibition of the activity of glycosylation enzymes.

28. Maintaining low levels of glutamine (and/or glucose) in culture is one of the most commonly used strategies to minimize ammonium production, as it is relatively easy to implement. Substitution of glutamine with other substrates that generate less ammonia, such as pyruvate, glutamate and α-ketoglutarate, has also been proposed. In addition, efforts should be taken to minimize the spontaneous decomposition of glutamine in the culture medium. The half-life of glutamine at 37°C and pH 7.2 is only 7 days, and this should be considered when scheduling the medium preparation in a manufacturing environment.

29. Some cell lines, for example, CHO, insect cells, and NS0, produce glutamine synthetase (GS), an enzyme that enables synthesis of glutamine from glutamate and ammonium. If the endogenous GS activity is sufficiently high, then cells will be able to proliferate in glutamine-free medium. This feature is used in the GS expression system and can contribute to minimizing ammonia formation in the cultures. The specific substrate consumption can be calculated according to Eq. (5.5), and the specific by-product formation can be calculated according to Eq. (5.6).

30. Another useful concept related to metabolites is the yield factor Y i|j . Yield factors can be used to understand the relation between two arbitrary compounds i and j to measure units consumed and units formed. For example, a yield coefficient can be used to measure substrate conversion (such as glucose) into cells according to Eq. (5.7): where Y X / S is the coefficient for conversion of substrate S to cells, Δ X is the change in biomass concentration, and Δ S is the change in substrate concentration. The assumption is that S is the sole limiting substrate for the culture. Similarly, the yield factor can be used to calculate conversion of substrate to product. Yield coefficients are not constant, but they depend on environmental factors.

31. Oxygen Adequate oxygen supply is essential for a successful upstream process. At too-low oxygen levels proliferation is compromised (for aerobic organisms), and at too-high levels, the risk for oxidative damage increases. Several studies have shown a direct link between the oxygen level in the medium and the metabolism, and a tight control is necessary for an optimized product yield. Oxygen supply can be a limiting factor in microbial processes, but it has historically not been considered a problem in animal cell culture. However, recent developments, for example, in the field of process intensification, are changing this picture. Modern animal cell fed-batch processes can reach 30 million cells/mL or more, and perfusion processes can be operated at 50–60 million cells/mL and higher over extended periods of time. The highest reported cell concentrations are above 200 million cells/mL, and this is at the limit of what conventional bioreactors for animal cells can handle. Hence, the oxygen requirements in each process must be well understood, especially during scale-up and large-scale manufacturing, to avoid limitations and ensure a successful manufacturing operation. In other words, what is essential is that the oxygen uptake rate (OUR) of the culture does not exceed the oxygen transfer rate (OTR) that the bioreactor system is capable of.

32. The concentration of dissolved oxygen is difficult to measure in culture, and it is instead indicated as percentage dissolved oxygen, DO, according to Eq. (5.8). C L is the actual oxygen concentration in the medium and C L * is the oxygen concentration when the medium is saturated with air. The difference between C L * and C L is the driving force for oxygen transfer in the culture and the value of C L * is given by Henry’s law. In distilled water C L * is approximately 7 mg/L at 30°C. In a typical cell culture medium, oxygen solubility is between 5% and 25% lower than in water because of solute effects from various medium components like electrolytes and organic compounds.

33. The oxygen uptake rate (OUR) is calculated according to Eq. (5.9). OUR is the volumetric oxygen uptake rate, q O2 is the specific oxygen uptake rate, and X is the viable cell/biomass concentration in the culture. The specific oxygen uptake rate q O2 is influenced by, for example, the cell type, the metabolic status of the culture and the concentration of metabolites such as glucose and glutamine in the culture medium. The specific oxygen uptake in animal cell cultures is between 9.4 x 10 -15 and 6.2 × 10 -13 mol/cell/h with a reported average of commonly used mammalian cell lines in the range of 2 × 10 -13 mol/cell/h . However, the variability is high. For example, the specific oxygen uptake rate of CHO cells varied between 1.8 × 10 -13 and 3.2 × 10 -13 mol/cell/h in the same culture depending on culture phase.

34. A typical animal cell culture with an average cell concentration can deplete the available oxygen within one hour unless new oxygen is added. Microbial cultures have much higher oxygen demand and can deplete the available oxygen within just seconds. Headspace aeration, sparging, membrane aeration, and oxygen enrichment of the inlet gas are methods by which oxygen can be added. Headspace aeration can suffice in small scale, but in large-scale production, sparging is the preferred method, potentially in combination with oxygen enrichment. If a low DO is required and the culture’s own oxygen consumption is not sufficient to decrease the DO to the set-point, nitrogen can be added to the inlet gas to strip oxygen from the culture medium. This can sometimes be required in the initial culture stage or during cultivation of primary cells.

35. The oxygen transfer rate OTR that the bioreactor is capable of is given by a mass balance in the liquid phase according to Eq. (5.10): k L a is the volumetric oxygen mass transfer coefficient of the bioreactor, C L * is the dissolved oxygen concentration in saturated liquid, and C L is the dissolved oxygen concentration. k L a is bioreactor-specific characteristic that influences the oxygen transfer capability of the system. Information about k L a is often provided by the bioreactor vendor, but the true k L a value is process specific and influenced by many factors, for example, temperature, culture medium, and specific bioreactor configuration. In addition, k L a is not a constant, but it can change during a process with variations in, for example, cell concentration, the amount of cell debris in the bulk liquid, nutrient supplementation, and antifoam addition.

36. k L a is not influenced by the amount of oxygen in the inlet gas (whereas C L * is). k L a values of bioreactors designed for animal cell cultivation are typically in the range of 1–25h −1 as compared to 100–400h −1 for microbial fermentors . In a bioreactor, the accumulation of oxygen in the liquid phase is given by Eq. (5.11): where dC L /dt is the accumulation of oxygen in the liquid phase, OTR is the oxygen transfer rate of the reactor, and OUR is the oxygen uptake rate of the culture. The goal is to operate under process conditions where dC L /dt is a positive value and OTR>OUR. This requires a good understanding of OUR and selecting an appropriate bioreactor and suitable process conditions to not exceed the oxygen transfer limits of the system.

37. CARBON DIOXIDE Carbon dioxide (CO 2 ) plays several roles in animal cell culture. First, CO 2 is produced by cells through oxidative phosphorylation in mitochondria. Second, many culture media for animal cells are buffered with carbonate, and addition of CO 2 is required, at least initially, to ensure that the correct pH is reached at equilibrium. Third, CO 2 is required as an intermediate in the fatty acid synthesis, and an abundance in the initial culture stage can lead to a prolonged lag phase. Atmospheric CO 2 dissolves in the culture medium according to the reaction:

38. An increased atmospheric CO 2 will shift the reaction to the right, and the net result will be an acidification of the medium. Culture media that rely on a carbonate buffer system will have a recommended bicarbonate concentration and CO 2 tension for achieving the correct pH and osmolality. In processes where culture medium that is formulated with noncarbonate buffer systems is used, for example, in the case of insect cell processes, no atmospheric CO 2 is needed. CO 2 addition can be done through headspace aeration and/or directly into the bulk liquid through sparging. When the cells proliferate, they will produce CO 2 through respiration. The respiration can enable the amount of CO 2 in the inlet gas to be decreased and sometimes removed entirely.

39. Rather, CO 2 accumulation and a too high pCO 2 can be a problem in later culture stages, especially for high-density processes and in larger-scale bioreactors. Physiological CO 2 levels are in the range of 4–7 kPa, but high-density cultures can have pCO 2 in the order of 20–30 kPa. High CO 2 levels can have an adverse effect on growth and production, and it can also impact protein glycosylation. In the case of an excessive CO 2 accumulation, sparging is necessary to strip out the CO 2 for keeping pCO 2 low and pH on target, without having to add base that would lead to an unwanted increase in osmolality. CO 2 removal has shown to be a strong function of sparge rate and bubble size, where large bubbles have been found to be most effective.

40. PRODUCT FORMATION The productivity of a bioprocess is a function of a many different cellular processes including gene transcription, mRNA stability, translation, post-translational processing, ER-associated degradation, and intracellular transport. As already mentioned, some products are formed mainly when cells grow, whereas other products are formed also during other phases. In practice, many processes follow an intermediate pattern. The specific productivity q P is a very useful tool to understand product formation kinetics, to compare production clones and processes, and to determine the optimal harvest point. The sought-after characteristic of a production cell line is to have a high and sustained specific productivity over a long period. In 1986, the typical antibody production process had a specific mAb productivity of less than 10pg/cell/day. Today, advanced processes were able to reach productivities of 100pg/cell/day or more, with extreme cases of 200pg/cell, day.

41. The specific production rate qp is calculated according to Eq. (5.7): where P is the product amount and X is the biomass/cell concentration. When comparing specific productivities between different cell lines and processes, it should be noted that different approaches can be used for the underlying calculations. For example, specific production rates can have been calculated for (a) the entire culture from inoculation to harvest, (b) the exponential growth phase, and (c) just the time point when it peaks. When comparing data between experiments and cell lines, it is essential to understand how the calculation was done in each specific case to ensure that a relevant comparison is made.