Nonionic surfactants are commonly used to stabilize proteins during upstream and downstream processing and drug formulation. Surfactants stabilize the proteins through two major mechanisms: (i) their preferential location at nearby interfaces, in this way precluding protein adsorption; and/or (ii) their association with protein into "complexes" that prevent proteins from interacting with surfaces as well as each other. In general, both mechanisms must be at play for effective protein stabilization against aggregation and activity loss, but selection of surfactants for protein stabilization currently is not made with benefit of any quantitative, predictive information to ensure that this requirement is met.
In certain circumstances the kinetics of surface tension depression (by surfactant) in protein-surfactant mixtures has been observed to be greater than that recorded for surfactant alone at the same concentration. We compared surface tension depression by poloxamer 188 (Pluronic® F68), polysorbate 80 (PS 80), and polysorbate 20 (PS 20) in the presence and absence of lysozyme and recombinant protein, at different surfactant concentrations and temperatures. The kinetic results were interpreted with reference to a mechanism for surfactant adsorption governed by the formation of a rate-limiting structural intermediate (i.e., an "activated complex") comprised of surfactant aggregates and protein. The presence of lysozyme was seen to increase the rate of surfactant adsorption in relation to surfactant acting alone at the same concentrations for the polysorbates while less of an effect was seen for Pluronic® F68. However, the addition of salt was observed to accelerate the surface tension depression of Pluronic® F68 in the presence of lysozyme. The addition of a more hydrophobic, surface active protein (Amgen recombinant protein) in place of lysozyme resulted in greater enhancement of surfactant adsorption than that recorded in the presence of lysozyme. A simple thermodynamic analysis indicated the presence of protein caused a reduction in ∆G for the surfactant adsorption process, with this reduction deriving entirely from a reduction in ∆H. We suggest that protein accelerates the adsorption of these surfactants by disrupting their self associations, increasing the concentration of surfactant monomers near the interface.
Based on these air-water tensiometry results, it is fair to expect that accelerated surfactant adsorption in the presence of protein (observed with PS 20 and PS 80) will occur with surfactants that stabilize protein mainly by their own adsorption at interfaces, and that the absence of accelerated surfactant adsorption (observed with F68) will be observed with surfactants that form stable surfactant-protein associations. Optical waveguide lightmode spectroscopy was used to test this expectation. Adsorption kinetics were recorded for surfactants (PS 20, PS 80, or F68) and protein (lysozyme or Amgen recombinant protein) at a hydrophilic solid (SiO₂-TiO₂) surface. Experiments were performed in sequential and competitive adsorption modes, enabling the adsorption kinetic patterns to be interpreted in a fashion revealing the dominant mode of surfactant-mediated stabilization of protein in each case. Kinetic results confirmed predictions based on our earlier quantitative analysis of protein effects on surface tension depression by surfactants. In particular, PS 20 and PS 80 are able to inhibit protein adsorption only by their preferential location at the interface, and not by formation of less surface active, protein-surfactant complexes. On the other hand, F68 is able to inhibit protein adsorption by formation of protein-surfactant complexes, and not by its preferential location at the interface.