Testing the Broth

Cell viability in a bioreactor can be measured in real time in a bulk vessel. Measurements can be done on-line via a bypass loop, inline (inside the vessel), and offline, wherein the measurement device is disconnected physically from the vessel, experts from Sartorius including Stuart Tindal, PhD, product manager, PAT and automation; Jochen Scholz, PhD, senior scientist, R&D instrumentation and control; Mario Becker, director, PAT and automation; and Marek Hoehse, PhD, senior scientist, R&D spectroscopy and MVDA, recently told BioPharm International. These individuals, along with John Carvell, PhD, sales and marketing director of Aber Instruments Ltd., and John Gleeson, director at process sensor sales at Hamilton Company, spoke with the publication to explain how to best measure cell viability in a bioreactor and ensure that process parameters stay stable.

BioPharm: What are the most popular inline methods of analyzing cell quality in bioreactors?

Sartorius: Cell quality is a very loosely worded term; there is no one inline probe that can directly measure viability or cell health. However, combining information from multiple techniques does enable the measurement of viable biomass (alive) and total biomass (alive and dead), which can give a viability ratio. For example, capacitance and transmission-based turbidity measurement could be used. In the future, capacity scanning/impedance spectroscopy is an upcoming method to get information about cell parameters and some groups investigate the possibility of obtaining “cell health” information from this method.

Gleeson: We take cell quality to mean the cell viability. In this case, the most popular way to determine cell quality inline is to use capacitance-based measurements. For these measurements, an electric field is applied to polarize cells within the medium. Because only cells with an intact membrane will be polarized, the measured permittivity will only reflect cells that are viable.

Carvell: I am not aware of any specific instrument that measures cell quality in bioreactors. In yeast fermentations, there has been a requirement for a reliable offline measurement to measure the yeast vitality, but academics and end-users cannot agree on a method.

BioPharm: What are the most popular inline methods of measuring cell concentration/density/biomass?

Carvell: The most popular inline methods methods for measuring cell concentration/density/biomass are radiofrequency impedance and optical probes. Inline cell density probes based on optical transmission (e.g., Optek, Exner), or reflectance (e.g., Mettler Toledo), are most commonly used in microbial cultures. The measurements are only linear with cell concentrations at a high viability and deviate as the viability drops towards the end of fermentation. The measurements are also very sensitive to gas aeration and the media; with a significant and varying suspended solids content, it becomes difficult to distinguish the true biomass from the debris. To overcome such limitations, radiofrequency impedance (RFI) spectroscopy (e.g., Aber Instruments) is now used in both cell culture and fermentation. RFI spectroscopy measures the live cell biovolume and the gas bubbles and debris do not contribute to the signal. RFI has a further advantage in that the probe can be now supplied in a single-use form as well as a robust probe for multiple steam sterilizations.

Sartorius: The answer here can be split into market sectors: bacterial, microbial (yeast), and mammalian cell culture. Bacterial cases normally have very high viabilities and tend to use optical or near-infrared (NIR) transmission or reflectance-based probes such as turbidity or optical density. Microbial cell culture would be more split, [and we would] see a mix of capacitance and turbidity methods. Lastly, cell-culture applications are commonly monitored with capacitance-based technology and supplemented with offline optical cell counting.

Gleeson: Total cell concentration/density/biomass are most popularly measured using optical density sensors. Optical density sensors measure the aforementioned parameters based on the amount of light absorbed by the medium. As previously discussed, viable cell density is best measured through capacitance-based technology.

BioPharm: What new technology has recently been released to aid in the inline measurements of cells within a bioreactor?

Gleeson: There are two main new technologies for inline cell measurement. The first is capacitance-based measurements, and the second is the technology for an online measurement using a microscope and flow cytometry to provide a 3D image. Capacitance-based measurements provide simple, real-time quantitative analysis without the need for grab-sampling. This eliminates risk of contamination and the trouble and time inherent in offline analysis. The microscope and flow cytometry combination quickly provides visual information about the cells while lessening burden and contamination.

Sartorius: As the biopharmaceutical industry shifts more towards single-use bioreactors, the recent release of an integrated single-use capacitance measurement now allows the same measurement capabilities within gamma-irradiated bags. The use of this type of bag removes the requirements for autoclaved reuseable probes, which can be a source of contamination risk.

Spectroscopic techniques such as NIR and Raman spectroscopy are now being used to measure directly in the bioreactor cell parameters such as viable cell density, viability, viable cell diameter, and total viable cell volume.

Carvell: NIR spectroscopy has been widely used in cell culture and fermentations. A version from Sartorius can be used to measure the cell count, viability, and glucose concentration. ASL Analytical has also developed a NIR method for the same parameters in cell culture, but in this case, they use a closed-loop recirculation with media and cells returned back to the bioreactor. Ovizio has also developed a platform based on an external recirculation loop with a microscope capturing the holograms of animal cells. This allows a label-free estimation of cell density, cell morphology, and viability. Ovizio, however, has a limited concentration range and cannot deal with the high cell densities seen in some perfusion cultures.

A true online probe using Raman spectroscopy is now available from Kaiser Optical. Cell concentrations and viability can be derived from the measurements after multiple calibration sets and chemometrics are applied. The derived cell concentrations and viability predicted by the online Raman probe is not based on a Raman effect but is more likely to be based on auto-fluorescence from products made by the cells.

BioPharm: What are some barriers to properly measuring osmolality and pH of a broth inline? What are typically the main causes of undesirable pH drift in bioreactors?

Carvell: There are not too many barriers to measuring pH; it’s been a standard measurement inside of bioreactors for a long time. In terms of pH drift, the main cause is typically a rush to calibrate. Most instruments will track the pH for 10 seconds, sometimes 20 seconds, during calibration, then pronounce the pH as “stable” and calculate an offset and span. It’s best to place the electrode in the first buffer and let it acclimate for a few minutes, then start the calibration process, let it set in the second buffer a few minutes and let it acclimate, then finish the calibration. It will improve the readings by several hundredths or even a tenth of a pH unit.

Gleeson: There are no significant barriers for inline pH measurement. Drift may be caused by changes in temperature or issues with the probe such as glass degradation, reference poisoning, poor maintenance, clogged junctions, and older or marginal probes.

Sartorius: Osmolality can be calculated by stoichiometric understanding of your process broth. Thus, by keeping track of your base media and tallying all additional liquid volume concentrations, an osmolality value can be produced. Performing those calculations manually is prone to error, thus, the use of supervisory control and a data acquisition system are considered best practice.

pH drift is generally due to the pH probe measurement being off-set over time, and a running recalibration would adjust for that off-set. However, it is challenging to remove a pH probe from a running process and recalibrate it without creating a sterility concern.

BioPharm: What are some recent advances in single-use sensors for disposable bioreactors? Are there many or few options of types of disposable probes?

Sartorius: These days, the consequences of gamma irradiation on sensor patches are better understood by the industry and the principle requirements to translate a multi-use sensor probe technology into a single use format are clearer. Recently, we have seen new options (viable biomass) and more improvements to single-use sensors for use in upstream operations. Single-use technologies now come integrated (both in the bag and the parameter signal in the local controller), qualified, and ready to use.

Gleeson: Recent advances in single-use include new glass and reference formulations that can withstand gamma sterilization. Currently, there are few options for pH sensors because of the instability of some of the optical probes and more standard glass probes that are being used.

Carvell: The most notable addition to single-use sensors for disposable bioreactors is the biomass probe. Sartorius Stedim Biotech now produce single-use biomass sensor discs under license from Aber Instrument. The radiofrequency impedance electronics for the single-use biomass sensor have been branded BioPAT ViaMass.

BioPharm: Are there any sensors currently on the market that measure more than one parameter at a time? What are some of the advantages of a multi-sensor system?

Gleeson: Oxidation reduction potential/pH/temperature combination sensors are currently on the market. Hamilton Company, Mettler-Toledo, and Van London are some of the manufacturers of these types of sensors.

Carvell: The main advantage of multi-sensor systems is that only one port is required and on an already-crowded bioreactor; this can save space. The main challenges remain the calibration of such sensors and then assessing if the calibration is robust when there is a significant change in the process conditions, change of media, cell line, etc. The sensors are also expensive and Raman runs have to be conducted in a darkened fermenter. There is also evidence that fluorescence from products from the cells interfere with the Raman measurement. 

Sartorius: Impedance-based technology will give both the conductivity (mS/cm) and the capacitance (pF/cm). Spectroscopy can be used to create a quantitative model of more than one parameter, covering a broad range of chemical (e.g., nutrients, product titer, etc.) and physical parameters (e.g., cell count, viability).

Insert port space, data external inputs, and space on a bioreactor are always limited; having a probe that can acquire more than just one parameter measure gives better utilization. However, one critical point is to have all the parameter data in one place so you can do multivariate data analysis and trigger system actions, such as alarms, guided sampling, and process control. Inline monitoring is generally not the user goal; the goal is typically inline control.