Bioseparation Engineering

Chapter 1

Adsorption, Chromatography and Membrane Separations

Recent Advances in Membrane Technology that Could Improve Resource Recovery and Recycle: Fluid Mechanics, Surface Science and Bioaffinity

Georges Belfort    Howard P. Isermann Department of Chemical Engineering Rensselaer Polytechnic Institute, Troy, NY 12180 (USA)

THE GLOBAL ENVIRONMENT

With the realization that enormous investments will be needed to balance economic activity with environmental protection (called sustainable development), new clean and cleaning technologies will be needed to address the global conditions of excessive pollution, increasing population and increasing industrialization¹. Of these technologies, synthetic membrane technology is expected to be a major player. See Fig. 1. The reasons for this are that pressure-driven membrane processes are very attractive because they do not involve a phase change (i.e. do not consume large amounts of energy), are often linearly scalable, do not need additives, are relatively fast (rate governed rather than equilibrium processes), operate in a continuous mode, are easily combined with other processes, and are completely contained. However, several limitations, still need to be addressed. Foremost among these are concentration polarization (CP) and fouling phenomena which can substantially reduce performance through osmotic effects and solute adsorption and deposition on the membrane surface. These limitations can readily result in additional energy requirements and larger capital and maintenance costs, thus reducing the attractiveness of pressure-driven membrane technology. Various approaches have been used to address these limitations including improved methods of operation through the use of positive displacement pumps for controlling permeation rate and minimizing transmembrane pressure drop, operating at or below a prescribed protein wall concentration, modifying the chemical properties of the membrane surface so as to minimize solute-membrane interactions, and improved fluid mechanics and module design for reducing solute concentration and deposition on the membrane.

Fig. 1 Sustainable development how can synthetic membrane technology contribute? Ref: B. Zechendorf, TIBTECH 17,219 (1999)

SYNTHETIC MEMBRANE TECHNOLOGY

The success of synthetic membrane technology has depended on a collaboration between polymer and surface scientists, who have developed suitable commercial membranes, and chemical engineers with an expertise in mass transfer and fluid mechanics, who have designed modules for optimizing filtration performance. Recent developments in these two fields will be emphasized in this presentation with a particular focus on biotechnology and the need to recover valuable proteins from solution. We argue that the need to understand the behavior of fluid flow with imposed centrifugal vortices can assist in designing optimal flow paths with minimal fouling and reduced concentration polarization²,³. Similarly, the connection between a fundamental understanding of intermolecular forces between a model protein, hen egg lysozyme (Lz), and polymeric membranes is crucial for the development of new and improved membrane materials for this application⁴,⁵.

THREE FUNDAMENTAL EXAMPLES

An example of the first module design without moving parts especially designed for suspensions commonly found in the biotechnology industry is our new Da Vinci module. By flowing sufficiently fast along a helical twisted membrane tube, counter rotating Dean vortices can be used to clean the membrane surface and reduce particulate build-up and fouling. See Fig. 2.

Fig. 2 Schematic of (a) the current commercial liner hollow fiber and (b) a new helical hollow fiber in which centrifugal vortices are produces that clean the membrane surface.

Koehler et al.⁴,⁵ have explained the well-known phenomenon of increased protein fouling on hydrophobic (poly(sulfone), PES) as compared to hydrophilic (hydroxyethyl methacrylate-PES, HEMA/PES) surfaces by using a correlation between adhesion forces and filtration fluxes. See Fig. 3. They show that protein-protein and protein-polymer interactions are about equally important for the PES-Lz system, while only protein- polymer interactions are important for the HEMA/PES-Lz system. How these two surfaces effect the stability of Lz and the fouling of membranes is discussed in detail.

Fig. 3 Protein-membrane force-distance curves for (a) a HEMA-modified poly(ethersulfone) membrane and (b) an unmodified poly(ethersuifone) membrane. The feed contained hen egg white lysozyme at 50 mg/1 and pH 6.6

Synthetic membranes or porous chromatographic beads are attractive binding media for affinity separations of fusion proteins because they overcome diffusion limitations with convective flow. In our final example, we illustrate the development and application of a new linker with controllable cleavage activity between the binding domain and the desired protein⁶. See Fig. 4. Both batch and column examples of the resulting one-step purification using temperature and pH excursions to induce cleavage are presented. Excellent purity and yield are obtained in all cases.

CONCLUSIONS

Cost estimates for achieving sustainable development up to the year 2,000 are about twice the current world pharmaceutical market of US$308 billion!⁷,⁸. Whether the advanced societies will be prepared to spend such a large amount without a crisis or environmental disaster, is open to question. Clearly, attractive technologies that utilize less energy and produce less waste such as biotechnology and synthetic membrane processes are prime candidates for such an effort.

ACKNOWLEDGEMENTS

The author thanks his past and current graduate students, post-docs and research collaborators. Technical support was obtained from Millipore Corp., Bedford, MA., while funding was supplied by Bob Peterson, Dow Chemical Co. and FilmTec Corp., NWRI, NSF (CTS-9400610), DOE (DE-FG02-90ER1414)Millipore Corp., and the NATO Scientific Committee.

Fig. 5 Development of a new controllable linker for fusion protein

REFERENCES

1. Zechendorf B. Trends in Biotechn. 1999;17:219–225.

2. Gehlert G, Luque S, Belfort G. Biotechnology Progress. 1998;14:931–942.

3. Luque S, Mallubhotla H, Gehlert G, Kuriyel R, Dzengeleski S, Pearl S, Belfon G. Biotechnology Bioengineering. 1999 in press.

4. Koehler JA, Ubricht M, Belfort G. Langmuir. 1997;13:4162.

5. Koehler JA, Ubricht M, Belfort G. Langmuir. 1999 in review.

6. Wood D, Wei W, Derbyshire V, Belfort G, Belfort M. Nature Biotechnology. 1999 in press.

7. MacNeil J. In: Scientific Amer. 1989:105–113.

8. Walker S. Plenary lecture at the-Recovery of Biological Products IX. Canada: Whistler; May 23, 1999.

Stabilization of target protein during bioseparation

X.-L. Fenga; Y.-T. Jinb; Z.-G. Sua    a National Laboratory of Biochemical Engineering, Institute of Chemical Metallurgy, Chinese Academy of Science, Beijing 100080, The People's Republic of China

b Laboratory of Biochemical Engineering, Dalian University of Technology, Dalian 116012, The People's Republic of China

Denaturation of target protein by various separation and purification steps contributes significant part to the total product loss in bioseparation. This report classifies the denaturation into four types including thermal denaturation, shear denaturation, solution denaturation and adsorption denaturation. For stabilization of target protein, three strategies are proposed including careful selection of unit operation to avoid detrimental action, process optimization to reduce the number of steps and the total processing time, and utilization of protective reagents such as PEG during bioseparation. It is important to understand the structure and property of the product to design the best bioseparation route.

1 INTRODUCTION

Low recovery is a major problem in production of pharmaceutical proteins. The loss of target protein can be classified into two aspects. The first one is physical loss in the flow stream, such as the leakage through an ultrafiltration membrane during concentration operation, the carry-away during a washing step in chromatography after loading, or even the residual left in the dead volume of a process device and the pipelines. This part of loss should not contribute to more than 15%, and is often controllable by proper process design and operation. The second loss is the denaturation of the target protein by various separation or purification steps. This part is significant, much more than 15%, and is difficult to control.

Any separation step in a bioprocess relies on its physical, chemical or biological action to distinct one or a group of proteins from the other. The product, or the target protein, has a limited stability undergoing the treatment. Even there is no change in the molecular weight or in the one dimensional structure, a minor alteration of the molecular conformation would result in loss of its biological activity. While molecular biologists are trying to construct artificial proteins that are more stable and functional, biochemical engineers are working hard in designing optimal separation routes to maintain the three dimensional integrity of the products and to achieve the desired purification during bioseparation [1].

2 AVOIDANCE OF DETRIMENTAL ACTION

In order to decrease the denaturation loss, care has to be exercised in choosing suitable separation methods to avoid detrimental actions, such as increasing temperature, excessive stirring, marked changes in pH, adding organic solvents and exposure to ultraviolet light. Table 1 lists the frequently used unit operations, its separation principles and possible damage to proteins. In general, protein denaturation in bioseparation can be classified into four categories, i.e. thermal denaturation, shear denaturation, solution denaturation and adsorption denaturation. Other denaturations such as those induced by high pressure and ultraviolet light are not common, and will not be discussed here.

Table 1

Denaturation of proteins in separation and purification

Thermal denaturation is caused by temperature increase, resulting in disorder of the three dimensional structure by breakage of the forces stabilizing the spatial conformation, such as hydrogen bonds, electrostatic and hydrophobic interactions. In mechanical cell disintegration such as homogenization and bead milling, part of the mechanical energy transferred to heat energy, increasing the temperature of the homogenate. For example, one passage through a homogenizer at 600 bars can increase the homogenate temperature by 2-5 °C depending on cell concentration and viscosity of the homogenate. Cooling is necessary for multiple passage of homogenization.

Shear denaturation is associated with high liquid flow rates. The mechanism is still unclear. Many observations have proved that protein may lost its activity in a high liquid shear field. For shear sensitive proteins, cross-flow microfiltration and ultrafiltration may cause denaturation due to high shear used for minimization of concentration polarization. Pumping is a process associated with liquid shear. Peristaltic pumps are normally regarded as mild operators and preferred choice for less contamination. However, studies have demonstrated that peristaltic pumps could denature proteins by generation of protein aggregates. The solution of serum albumin, in which aggregates had been removed, when being pumped again with a peristaltic pump, produced aggregates again. The pumping period and concentration of the protein determine the magnitude of aggregate formation [2].

For solution denaturation, several mechanisms may be involved, including protease hydrolysis, chemical hydrolysis, interaction with salts, surfactants, organic solvents etc.[3]. In fact these actions in solution may be going on all the time during bioseparation with varied degrees for different proteins, even the solution is in cold storage. When a separation requires addition of certain substances to the protein solution and process it under certain condition, denaturation by the substances present in the solution may occur. For example, chemical disruption of the cells requires addition of organic solvents, surfactants or chaotropic agents such as guanidine hydrochloride. These reagents break down cell membranes to release the intracellular protein. However, the released product is also under the attack of the reagents. Aqueous two-phase extraction in general is good for maintaining the activity of the protein, but the high concentration of salts and type of salts may affect the protein activity in salt- polymer system. Solution denaturation depends on the concentration of the solutes that denature the product. In freeze drying, much of the protein activity may be lost during freezing stage because water forms ice and solute concentrations are