Bioseparations using surfactant-aided size-exclusion chromatography

Abstract

In most bioprocesses one or more chromatographic steps are used in the purification of the product. Size-exclusion chromatography (SEC) is often one of these chromatographic steps. It is based on the difference in size and shape of the components to be separated and is used for the separation of molecules with a near identical chemical composition such as dimer or oligomers from monomeric products. In SEC, the selectivity is only depends on the characteristics of the gel material: the volume of the gel fibers and the diameter ratio of the pores versus the components to be separated. These parameters cannot be changed in-situ and each specific separation therefore requires a specific gel. Beside this low flexibility, SEC is characterized by a low efficiency due to the limited selectivity of the gel material. High resolution is possible but will result in high eluent and resin consumption, diluted products and high process times, which all have a negatively effect on the costs of the production process. This indicates that there is a need for improvements or alternative concepts for this polishing step. This thesis describes an alternative method which is based on the integration of SEC and a selective mobile phase containing non-ionic micelles. Chapter 2 gives a detailed description of this new alternative method: surfactant-aided size-exclusion chromatography (SASEC). The way in which biomolecules and bioparticles partition towards a mobile phase containing inert micelles of nonionic surfactants, depends on the same type of parameters as in size-exclusion chromatography: the volume fraction of micelles and the diameter ratio of solute and micelles. The larger component will be excluded to a higher extent from the micellar mobile phase than the smaller component, which will elute first. In theory, the gel matrix should act as a practically non-selective storage phase for the components to be separated but selectively exclude micelles. Small species elute first, thereby reversing the "normal" chromatographic behavior. In chapter 2 a proof a principle is given with the model proteins Bovine Serum Albumin (BSA) and Myoglobine (Myo). Pulse experiments have been performed using the nonionic surfactants C12E23 or C16E23 in the mobile phase. The elution time of both proteins was increased at increasing micelle concentration. For BSA, the larger protein, this effect was larger than for Myo. This indicates that the selectivity can be changed in-situ by changing the micelle concentration. This distribution behavior is described by a model based on the excluded volume interactions between the proteins and the gel fibers, the proteins and the micelles and the micelles and the gel fibers. This model adequately predicts the distribution behavior of both proteins in SASEC. The ability to change the selectivity in-situ makes SASEC also suitable for gradient Simulated Moving Bed SMB chromatography (SMB). High affinity in the top sections and low affinity in the bottom sections can be established by using a high concentration of micelles in the top sections and a low concentration in the bottom sections. The distribution behavior of the micelles is depending on the micelle concentration and is implemented in the flow selection procedure for positioning the micelle gradient in the SMB. This procedure is described in chapter 3 and verified with several gradient positioning experiments. In chapter 4 viral clearance using SASEC is studied by the separation of a bacteriophage f29 and bovine serum albumin (BSA). Although this bacteriophage can be seen as a model for a relative small virus (r=20 nm) it was too large to enter the pores of the gel matrix and adding micelles had no influence on the elution time of f29. The added micelles only caused an increase the elution time of BSA. As a result, the resolution of the separation was increased at increasing surfactant concentrations. Experiments further showed that SASEC results in higher productivity and lower solvent consumptions than compared with conventional SEC. This is not only shown in fixed bed chromatography but also in micellar gradient SMB. Gradient SASEC-SMB also shows an extra advantage of a non-diluted product. Appendix A contains a patent that was filed on part of this work. In chapter 5 the separation of BSA and IgG was studied. Two resins have been tested (Sephacryl S200 and Sephacryl S300) as well as two different surfactants (C12E23 and Tween 20). Pulse experiments using SASEC showed that both BSA and IgG are more distributed towards the solid phase than compared to using SEC. This effect is larger on IgG, the largest component, than on BSA. As a consequence the mixture forms an azeotrope at a specific micelle concentration. Above this concentration the selectivity is reversed. Experiments show that at 7.5 % (w/w) C12E23, BSA elutes before IgG. When the concentration is further increased the selectivity increases to values higher than obtained with conventional SEC. The productivity at 10 % (w/w) C12E23 is increased with a factor 3 while the eluent consumption is decreased with a factor 1.2. Mathemetical simulations of the separation of BSA and IgG show a large increase in productivity (factor 40) and a decrease in eluent consumption (factor 15) when SASEC-SMB is applied compared to conventional SEC-SMB. The obtained product IgG is also concentrated by a factor 2 with SASEC-SMB. Size-exclusion effects do not only appear between micelles and proteins or proteins and gel fibers but also between proteins and proteins. This leads to non-linear distribution behavior of these proteins. This is studied in chapter 6 by break-through experiments with highly concentrated BSA. The results show an increase in elution time with increasing concentration of BSA. This effect is however less then predicted by only excluded volume-effects. Apparently other ineractions also play a role. Nevertheless, these concentration effects are still large enough to play an important in the design of separation processes using SEC in fixed bed or SMB chromatography systems. The final chapter discusses the choice of surfactant-gel systems to be used in SASEC. The selectivity is depending on two radius ratio's; the radius ratio of the solute versus the gel fiber and the radius ratio of the micelle versus the gel fiber. The effect of these ratio's is first evaluated on a theoretical bases. This evaluation shows that the ratio of the solute versus the gel fiber should be as large as possible while the ratio of the micelle versus the gel fiber should be as small as possible. The second part of this chapter discusses the characteristics of some of the available micelles and gel materials to see the real limits of these ratio's. The ratio's needed to improve the selecitivity are well inside the limits of the real available ratios. It is therefore possible to find a surfactant-gel system for a specific separation. Chapter 7 finishes with some remarks concerning viscosity, acceptance of micelles in bioprocesses, and possible removal of the surfactants afterwards.