How to Scale-Up Animal Cell Culture? | Zoology

It is the surface area for cell growth. A unit volume of medium is only capable of giving a finite yield of cells. Factors which affect the yield are: pH, oxygen limitation, accumulation of toxic products (e.g. NH₄), nutrient limitation (e.g. glutamine), spatial restrictions, and mechanical/shear stress.

As soon as one of these factors comes into effect, the culture is finished and the remaining resources of the system are wasted. The aim is, therefore, to delay the onset of any one factor until the accumulated effect causes cessation of growth, at which point the system has been maximally utilised.

Simple ways of achieving this are: a better buffering system (e.g. Hepes instead of bicarbonate), continuous gassing, generous headspace volume, enriched rather than basal media, with nutrient-sparing supplements such as non-essential amino acids or lactalbumin hydrolysate, perfusion loops through ultrafiltration membranes or dialysis tubing for detoxification (3) and oxygenation, and attention to culture and process design.

Growth Kinetics:

The standard format of a culture cycle beginning with a lag phase, proceeding through the logarithmic phase to a stationary phase and finally to the decline and death of cells, is well documented. Although cell growth usually implies increase in cell numbers, increase in cell mass can occur without any replication. The difference in mean cell mass between cell populations is considerable, as would be expected, but so is the variation within the same population.

Growth (increase in cell numbers or mass) can be defined in the following terms:

a. Specific growth rate

b. Doubling time, td (i.e., the time for a population to double in number/mass).

c. Degree of multiplication, n or number of doublings (i.e. the number of times the inoculum has replicated)-

n = 3.32 log(x/Xo) 6

Medium and Nutrients:

A given concentration of nutrients can only support a certain number of cells. Alternative nutrients can often be found by a cell when one becomes exhausted, but this is bad practice because the growth rate is always reduced (e.g. while alternative enzymes are being induced).

If a minimal medium, such as Eagle’s basal medium (BME) or minimum essential medium (MEM) is used with scrum as the only supplement, then this problem is going to be met sooner than in cultures using complete media (e.g. 199), or in media supplemented with lactalbumin hydrolysate, peptone, or BS A (which provides many of the fatty acids).

Nutrients likely to be exhausted first are glutamine, partly because it spontaneously cyclizes to pyrrolidone carboxylic acid and is enzymically converted (by serum and cellular enzymes) to glutamic acid, leucine, and isoleucine. Human diploid cells are almost unique in utilising cystine heavily.

A point to remember is that nutrients become growth-limiting before they become exhausted. As the concentration of amino acids falls, the cell finds it increasingly difficult to maintain sufficient intracellular pool levels. This is exaggerated in monolayer cultures because as cells become more tightly packed together, the surface area available for nutrient uptake becomes smaller.

Glucose is often another limiting factor as it is destructively utilised by cells and, rather than adding high concentrations at the beginning, it is more beneficial to supplement after two to three days. In order to maintain a culture, some additional feeding often has to be carried out either by complete, or partial, media changes or by perfusion.

Many cell types are either totally dependent upon certain additives or can only perform optimally when they are present. For many purposes it is highly desirable, or even essential, to reduce the serum level to 1% or below. In order to achieve this without a significant reduction in cell yield, various growth factors and hormones are added to the basal medium. The most common additives are insulin (5 mg/litre), transferrin (5-35 mg/litre), ethanolamine (20 (AM), and selenium (5 mg/litre).

Cell aggregation is often a problem in suspension cultures. Media lacking calcium and magnesium ions have been designed specifically for suspension cells because of the role of these ions in attachment. This problem has also been overcome by including very low levels of trypsin in the medium (2 mg/ml).

i. pH:

Ideally pH should be near 7.4 at the initiation of a culture and not fall below a value of 7.0 during the culture, although many hybridoma lines appear to prefer a pH of 7.0 or lower. A pH below 6.8 is usually inhibitory to cell growth. Factors affecting the pH stability of the medium are buffer capacity and type, headspace, and glucose concentration.

The normal buffer system in tissue culture media is the carbon dioxide bicarbonate system analogous to that in blood. This is a weak buffer system, in that it has a pKa well below the physiological optimum. It also requires the addition of carbon dioxide to the headspace above the medium to prevent the loss of carbon dioxide and an increase in hydroxyl ions. The buffering capacity of the medium is increased by the phosphates present in the balanced salt solution (BSS).

Medium intended to equilibrate with 5% carbon dioxide usually contain Earle’s BSS (25 mM NaHCO₃) but an alternative is Hanks’ BSS (4 mM NaHCO₃) for equilibration with air. Improved buffering and pH stability in media is possible by using a zwitterionic buffer, such as Hepes (10-20 mM), either in addition to, or instead of, bicarbonate (include bicarbonate at 0.5 mM).

Alternative buffer systems are provided by using specialist media such as Leibovitz’s L- 15. This medium utilises the buffering capacity of free amino acids, substitutes galactose and pyruvate for glucose, and omits sodium bicarbonate. It is suitable for open cultures.

The headspace volume in a closed culture is important because, in the initial stages of the culture, 5% carbon dioxide is needed to maintain a stable pH in the medium but, as the cells grow and generate carbon dioxide, it builds up in the headspace and this prevents it diffusing out of the medium.

The result is an increase in weakly dissociated bicarbonate producing an excess of hydrogen ions in the medium and a fall in pH. Thus, a large headspace is required in closed cultures, typically tenfold greater than the medium volume (this volume is also needed to supply adequate oxygen). This generous headspace is not possible as cultures are scaled-up, and an open system with a continuous flow of air, supplied through one filter and extracted through another, is required.

The metabolism of glucose by cells results in the accumulation of pyruvic and lactic acids. Glucose is metabolised at a far greater rate than it is needed. Thus, glucose should ideally never be included in media at concentrations above 2 g/litre, and it is better to supplement during the culture than to increase the initial concentration.

An alternative is to substitute glucose by galactose or fructose as this significantly reduces the formation of lactic acid, but it usually results in a slower growth rate. These precautions delay the onset of a non- physiological pH and are sufficient for small cultures. As scale-up increases, headspace volume and culture surface area in relation to the medium volume decrease.

Also, many systems are developed in order to increase the surface area for cell attachment and cell density per unit volume. Thus pH problems occur far earlier in the culture cycle because carbon dioxide cannot escape as readily, and more cells means higher production of lactic acid and carbon dioxide. The answer is to carry out frequent medium changes or use perfusion, or have a pH control system.

The basis of a pH control system is an autoclavable pH probe (available from Pye Ingold, Russell). This feeds a signal to the pH controller which is converted to give a digital or analogue display of the pH. This is a pH monitor system. Control of pH requires the defining of high and low pH values beyond which the pH should not go.

These two set points on the pH scale turn on a relay to activate a pump, or solenoid valve, which allows additions of acid or alkali to be made to the culture to bring it back to within the allowable units. It is rare to have the pH rise above the set point once the medium has equilibrated with the headspace, so the addition of acid can be disregarded. If an alkali is to be added, then sodium bicarbonate (5.5%) is recommended.

Sodium hydroxide (0.2 M) can be used only with a fast stirring rate, which dilutes the alkali before localised concentrations can damage the cells, or if a perrusion loop is installed. Normally, liquid delivery pumps are supplied as part of the pH controller. Gas supply is controlled by a solenoid valve and <95% air directly.

Above the set point carbon dioxide will be mixed with the air, but below this point only air will be delivered to the culture. This in itself is a controlling factor in that it helps remove carbon dioxide as well as meeting the oxygen requirements of the cultures. pH regulation is very readily adapted to computer control systems.

ii. Oxygen:

The scale-up of animal cell cultures is very dependent upon the ability to supply sufficient oxygen without causing cell damage. Oxygen is only sparingly soluble in culture media (7.6 n-g/mi) and a survey of reported oxygen utilisation rates by ceils reveals a mean value of 6 |ig/ 10⁶ cells/h.

A typical culture of 2 x 10⁶ cells/ml would, therefore, deplete the oxygen content of the medium (7.6 |o,g/ml) in under 1 h.

It is necessary to supply oxygen to the medium throughout the life of the culture and the ability to do this adequately depends upon the oxygen transfer rate (OTR) of the system-

OTR = Kla (C* – C) 7

Where OTR is the amount of oxygen transferred per unit volume in unit time, K1 is the oxygen transfer coefficient, and a is the area of the interface across which oxygen transfer occurs (as this can only be measured in stationary and surface-aerated cultures, the value Kla is used; this is the mass transfer coefficient (vol/h). C* is the concentration of dissolved oxygen when medium is saturated, and C is the actual concentration of oxygen at any given time.

The Kla (OTR/C when C = 0) is in units of h¹ hours¹ and is thus a measure of the time taken to oxygenate a given culture vessel completely under a particular set of conditions. A culture can be aerated by one, or a combination, of the following methods- surface aeration, sparging, membrane diffusion, medium perfusion, increasing the partial pressure of oxygen, and increasing the atmospheric pressure.

Surface Aeration (Static Cultures):

In a closed system, such as a sealed flask, the important factors are the amount of oxygen in the system and the availability of this oxygen to the cells growing under 3-6 mm of medium. Normally a headspace/ medium volume ratio of 10:1 is used in order to provide sufficient oxygen.

Oxygen Concentrations on the Gas and Liquid Phases of a Roux Bottle Culture:

Oxygen in 900 ml air,

900 x 0.21 x 32/22400 = 0.27 g

Oxygen in 100 ml medium,

100 x 7.6 x 10⁴ = 0.0076 g

Notes-

0.21 = proportion of oxygen in air;

32 = molecular weight of oxygen;

22400 = gram molecular volume;

x 10^{^1} = solubility of oxygen in water at 37°C when equilibrated with air.

Thus a 1 litre flask (e.g. a Roux bottle) with 900 ml of air and 100 ml of medium will initially contain 0.27 g oxygen. This amount will support iO⁸ cells for 450 h and is thus clearly adequate. The second factor is whether this oxygen can be made, available to the cells. The transfer rate of oxygen from the gas phase into a liquid phase has been calculated at about 17 |jig/cm²/h. Again, this is well in excess of that required by cells in a 1 litre flask.

Saturated with dissolved oxygen and the concentration at the cell sheet is almost zero then, applying Fick’s law of diffusion, the rate at which oxygen can diffuse to the cells is about 1.5 jig/ cm²/h. At this rate there is only sufficient oxygen to support about 50 x 10⁶ cells in a 1 litre flask, a cell density which in practice many tissue culturists take as the norm.

These calculations show the importance of maintaining a large headspace volume; otherwise, oxygen limitation could become one of the growth-limiting factors in static closed cultures. A raller explanation of these calculations, including Pick’s law of diffusion, can be found in a review by Spier and Griffiths.

iii. Sparging:

This is the bubbling of gas through the culture, and is a very efficient means of effecting oxygen transfer (as proven in bacterial fermentation). However, it may be damaging to animal cells due to the effect of the high surface energy of the bubble on the cell membrane.

This damaging effect can be minimised by using large air bubbles (which have lower surface energies than small bubbles), by using a very low gassing rate (e.g. 5 ml/1 min), and by adding Pluronic F-68. A specialised form of sparging, the airlift fermenter has also been used in large unit process monolayer cultures (e.g., multiple plate propagators).

When sparging is used, efficiency of oxygenation is increased by using a culture vessel with a large height/diameter ratio. This creates a higher pressure at the base of the reactor, which increases oxygen solubility.

iv. Membrane Diffusion:

Silicone tubing is very permeable to gases, and if long lengths of thin- walled tubing can be arranged in the culture vessel then sufficient diffusion of oxygen into the culture can be obtained. However, a lot of tubing is required (e.g., 30 m of 2.5 cm tubing for a 1000 litre culture).

This method is expensive and inconvenient to use, and has the inherent problem that scale-up of the tubing required is mainly two-dimensional while that of the culture is three-dimensional. However, several commercial systems are available (e.g., Braun).

v. Medium Perfusion

A closed-loop perfusion system contiguously (or on demand) takes medium from the culture, passes it through an oxygenation chamber, and returns it to the culture. This method has many advantages if the medium can be conveniently separated from the cells for perfusion through the loop.

The medium in the chamber can be vigorously sparged to ensure oxygen saturation and other additions, such as sodium hydroxide for pH control, which would damage the cells if put directly into the culture, can be made. This method is used in glass bead systems and has proved particularly effective in micro-carrier systems, where specially modified spin filters can be used.

vi. Environmental Supply:

The dissolved oxygen concentration can be increased by increasing the head-space pO₂ (from atmospheric 21 % to any value, using oxygen and nitrogen mixtures) and by raising the pressure of the culture by 100 kPa (about 1 atm) (which increases the solubility of oxygen and its diffusion rate).

These methods should be employed only when the culture is well advanced, otherwise oxygen toxicity could occur. Finally, the geometry of the stirrer blade also affects the oxygen transfer rate.

vii. Scale-up:

Oxygen limitation is usually the first factor to be overcome in culture scale-up. This becomes a problem in conventional stirred cultures at volumes above 10 litres. However, with the current use of high density cultures maintained by perfusion, oxygen limitation can occur in a 2 litre culture.

viii. Redox Potential:

The oxidation-reduction potential (ORP), or redox potential, is a measure of the charge of the medium and is affected by the proportion of oxidative and reducing chemicals, the oxygen concentration, and the pH. When fresh medium is prepared and placed in the culture vessel it takes time for the redox potential to equilibrate, a phenomenon known as poising. The optimum level for many cell lines is +75 mV, which corresponds to a dissolved p0₂ of 8-10%.

Some investigators find it beneficial to control the oxygen supply to the culture by means of redox, rather than an oxygen electrode. Alternatively, if the redox potential is monitored by means of a redox electrode and pH meter (with mV display), then an indication of how cell growth is progressing can be obtained. This is because the redox value falls during logarithmic growth and reaches a minimum value approximately 24 h before the onset of the stationary phase.

This provides a useful guide to cell growth in cultures where cell sampling is not possible. It is also useful to be able to predict the end of the logarithmic growth phase so that medium changes, addition of virus, or product promoters can be given at the optimum time. The effect of redox potential on cell cultures has been reviewed by Griffiths.