Development of a laminar shear apparatus for characterising the shear-mitigating effect(s) of bovine serum albumin in mammalian cell suspension culture

peer-reviewedThe stirred tank bioreactor is one of the most widely-utilised technologies in the biopharmaceutical industry for large-scale mammalian cell culture resulting in the production of therapeutic monoclonal antibodies. In these reactors, cells are exposed to fluid shear stress from various sources within the culturing environment which, in the absence of intervention, can result in cell lysis. The shear protectant Pluronic F-68 (PF-68) is commonly added to large-scale mammalian cultures in order to mitigate this shear-induced cell damage. However, as the industry moves towards ever higher cell densities and product titres there have been reports that the protective effects of PF-68 become limited at these scales. Bovine serum albumin (BSA) is an animal-derived protein which has also been suggested to play a role in mitigating shear-induced lysis in mammalian cell culture but the specific effects and the mechanism(s) by which this has been proposed to occur have not been well-characterised. Further research into the function and method of action of BSA in this capacity may highlight its potential role as an alternative to PF-68. Previously published literature on the action of shear protectants has proposed a framework for categorising their method of action as either (i) physical, (ii) slow-acting physiological or (iii) fast-acting physiological in nature. To investigate the effect of BSA on mammalian cells in a shearing environment, a laminar flow shear apparatus was developed which can subject cells in suspension culture to well-defined, fluid shear stress utilising a combination of concentric cylinder and cone-and-plate geometries. In order to maintain aseptic test conditions, all cell-contacting surfaces of the apparatus were manufactured using corrosion-resistant materials that were easy to sterilise between tests. The apparatus was also designed to operate in the absence of a liquid/air interface to enable the user to distinguish between any observed physical or physiological protective effects of BSA. Functionality of the test setup was validated through a series of experiments tracking cell circulation, culture viability and shear-induced cell lysis. A novel aspect of the device is that it produces a more representative model of the long-term effects of shear on cells in stirred tank bioreactor culture than previously described Couette flow devices by enabling the shearing of cells in continuous culture for up to 24hrs. It was shown that BSA provides concentration-dependent protection from shear stress over the range studied (0-4g/L) in both long-term, shear-in-culture experiments and short-term shear experiments. This protection was observed to take full effect v immediately after the BSA was added to the culture and fully lose effect immediately after being removed from the culture, indicating that the action of BSA is transient in nature and functions in part through a fast-acting, physiological mechanism. The dynamic viscosity of the culture medium was not significantly affected when BSA concentrations over the range 0-4g/L were supplemented. However, the surface tension of the medium was reduced by ~6% over this range, suggesting that BSA also provides a physical protective effect in cultures that utilise direct sparging or have a liquid/air interface. The protective effect of BSA was still seen in the absence of cholesterol supplementation and when thermal denaturation of the BSA at 100°C was carried out prior to testing. Additionally, it was shown that BSA associates directly with NS0 cells and that the damage-mitigating effect of the protein is not masked when the plasma membrane of the NS0 cells is saturated with high concentrations of PF-68. It is thus proposed that the fast-acting, physiological protective effect is manifested through the formation of a BSA coating around the external surface of the cell via adsorption of BSA molecules to the glycocalyx layer