by Michael Mietzner, IPT
You might have been faced with a mammalian batch not having the "expected density". Immediately it could be construed to be "poor cell growth". This guide will discuss how to determine "poor cell growth", whether density is in the "expected" range, along with what phenotypic traits can be analyzed and how to uncover process issues using data profiles.
An actual hemacytometer count tells you very little. Growth rate is a more useful tool in determining expected culture growth. Be mindful that logarithmic growth allows a small density error on inoculation day to provide a noticeable density difference on harvest day. At the same time, growth rate should be constant in log growth phase (with the exception of biological variation or noise) regardless of small differences in inoculation density.
Once it is determined that the cell density is outside of averaged historical ranges, the investigation can continue.
Calculations and Record Review
A very simple step that is often overlooked is re-calculating the media formulations and cell density. A simple mathematical mistake can cause havoc.
When a cell is genetically manipulated for product production it causes stress on the cell. Temperature, dissolved oxygen, and pH swings including offset from optimal conditions will also produce stress. Cellular stress will cause cell swelling and lysis under severe conditions. Stress may make the cell glycolytic, where lactic acid production reduces product production capacity (this can be observed by looking at a lactate data profile).
Carbon Dioxide Levels
Carbon Dioxide (CO2) is a cellular bi-product and is toxic to cells at certain concentrations. Cultures sparged with oxygen can approach toxic levels by harvest day, whereas sparging with air gives a 5-fold volumetric increase in order to reach the same O2 levels, which facilitates CO2 stripping.
It is assumed that the reactor operates as a continuously stirred tank reactor, or CSTR, but if solutions (titrant, nutrients) are added to the top of the liquid, rather than within the mixing zone, there exists a possible gradient of pH. The area near the pH addition experiences elevated levels momentarily and could stress or kill cells that are in the localized region.
There exists many ways of identifying the physical state of a cell using equipment. Two methods for tracking the physical condition of the cells are discussed as follows.
Particle Size Counters
Particle size counters can produce a particulate size profile and average particle diameter for a given culture. The size limit is set so that cells are counted above the limit, and cell debris is counted below the limit. The unit is cleaned and a sample of solution (used for dilution) is analyzed, then a sample of culture is analyzed and profiled. The solution profile is subtracted from the culture profile to eliminate solution debris. This yields a debris count and a cell count. Some units include measurement of the mean particle diameter, which, for particles above the size limit, equates to the average cell diameter. The cell diameter can swell in size from normal culturing stress (agitation, byproduct build-up). In more stressful conditions such as perfusion or absence of glucose, the size increase is more noticeable. In summary, the particle size counter provides a method for checking hemacytometer cell counts, provides measurements of debris in the culture, and determines the average cell diameter. This data allows for cell line profiling and insight into process troubleshooting.
Optical Density Probe
Optical density probes are very useful data collection and troubleshooting tools. The profiles captured with this equipment equate directly with cell density. This becomes very useful for monitoring that the batch growth trends as expected and any growth deviations would be identified long before the next off-line cell count.
Upon identifying a problem, determine if the cells are stressed and use data profiles to search for unexpected results. Also, calculate the rate of usage/production (Q) for air, titrant, glucose, lactate, and product and compare to historical rates. The calculated Q should trend closely between batches. Determine interrelations between phenotypic traits by running a comprehensive statistical analysis on the data. Use these interrelations to track batch performance and profile expected results.
Finally, vessel characteristics can be evaluated. Calculations and experimentation should be performed to verify that operational abilities are within required ranges and that design criteria was properly implemented into the system.
Page last updated: 5 March 2009