Andrew Zydney | Research
Membrane filtration is a potentially attractive process for separating and purifying complex protein mixtures in the preparation of important biotechnological products. Recent work in our group has demonstrated that it is possible to obtain very high degrees of protein separation by combining the inherent size-based selectivity of these membrane devices with specific electrical interactions between the proteins and membrane pores.
For example, it is possible to operate the membrane device under conditions where one protein is charged while the others are neutral, resulting in very high throughput of the neutral protein with the charged species being "electrostatically" rejected from the membrane pores. This makes it possible, at least in principle, to use membrane systems for much finer separations than were previously believed possible. Alternatively, one can use charged membranes to obtain much higher filtration rates with equivalent product retention due to the strong electrostatic exclusion.
Our current research is directed in two areas.
- First, we are performing fundamental experimental and theoretical studies to determine the optimal surface charge characteristics for different protein separation processes. This includes studies of protein retention, fluid flow, and membrane fouling, all of which can have a dramatic effect on the overall performance of these membrane systems.
- Second, we are examining the use of membrane systems for the purification of whey proteins. Whey is a complex mixture of many different proteins, and there is considerable interest in purifying the individual whey proteins to take advantage of their unique physical, chemical, and biological properties. This is also an attractive model system for examining the use of membrane systems for purifying complex protein mixtures.
Membrane devices are particularly attractive for artificial organ applications. For example, the artificial kidney (hemodialysis) uses semipermeable membranes as a barrier between the blood and dialysate solution, thereby retaining all cells and plasma proteins while allowing relatively free transmission of metabolic wastes. The bioartificial pancreas uses a semipermeable membrane to immunoisolate transplanted tissue, with the membrane providing both a support for tissue growth and a barrier that shields the foreign cells from immune rejection. In both of these applications, the transport properties of the membranes, and thus the performance of the overall device, can be dramatically by interactions with blood cells and proteins.
Our current research is focused on understanding the effects of blood-membrane interactions in hemodialysis, including the clinical implications of these interactions on dialyzer reprocessing. Although reprocessing is an accepted practice, there are still major concerns about the clinical impact of this approach.
- First, current cleaning solutions are largely inadequate at unblocking hollow fiber membranes that become clotted during dialysis. These blocked fibers are largely unavailable for solute transport, thus reducing the overall clearance of these solutes during dialysis.
- Second, existing reprocessing techniques are generally unable to completely remove the plasma proteins from the membrane surface and pore structure. These adsorbed proteins provide an additional resistance to solute transport, further reducing the rate of solute clearance.
- Finally, existing cleaning solutions contain strong oxidizing agents like bleach, which are known to degrade the polymers used in many high-flux hemodialysis membranes. We are examining a new approach to dialyzer reprocessing using a two-phase air-liquid mixture in the mist regime with very high local shear stresses to remove surface deposits and unblock fibers. This approach has enormous potential for use in hemodialysis applications if the solution chemistry and flow conditions can be chosen to provide the desired cleaning without damaging the membranes.
Our group is also involved in a number of fundamental studies of membrane processes, with particular emphasis on understanding membrane transport and fouling. Recent work in our group has demonstrated that the underlying pore morphology can have a large effect on the flux decline that occurs during membrane filtration. We have developed a new model for protein fouling of larger pore size microfiltration membranes that explicitly accounts for the effects of the internal membrane pore connectivity, something that has been almost entirely neglected in prior studies of membrane fouling. This approach is now being extended to the analysis of protein fouling during virus filtration, an important new application of membrane systems in bioprocessing.
We have also developed a novel approach for studying the transport of charged species in membranes using "charge ladders". Protein charge ladders are developed by chemical modification of a base protein to block one or more of the charged amino acids, thereby generating a set of protein derivatives differing only in electrical charge. These charge ladders are an ideal tool for probing electrostatic interactions in membrane systems. Recent work in our group has demonstrated the importance of hydrogen ion partitioning in determining the electrostatic exclusion of charged proteins from charged membrane pores, a phenomenon that has not previously been recognized in membrane systems.
Page reviewed on 01/17/12, Top of page