Ion-Exchange Purification of Proteins Using Magnetic Nanoclusters Andre Ditsch, Jin Yin, ‡ Paul E. Laibinis, † Daniel I. C. Wang, and T. Alan Hatton* Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 Polymer-coated magnetic nanoclusters were used for recovery and purification of proteins from both model systems and cell-free Pichia pastoris fermentation broth. The nanoclusters exhibited extremely high capacity for proteins, up to 900 mg/mL adsorbent, and were recovered by high gradient magnetic separation (HGMS) at flow rates of up to 3,600 cm 3 /cm 2 h (flow rates up to 15,000 cm 3 /cm 2 h are possible). The nanoclusters were coated with a primary coating of poly- (acrylic acid-co-styrenesulfonic acid-co-vinylsulfonic acid), which allowed both electrostatic and hydrophobic interactions with the protein to be used to enhance specificity for targeted products. With this dual mode separation, nearly pure protein could be recovered from complex mixtures, such as fermentation broth, in a few quick steps. Introduction A major expense in the production of a recombinant protein is that of the separation and purification of the product from the rest of the milieu in which it is produced. Typically, multiple separation steps are required and these operations can account for as much as 60-90% of the total manufacturing costs. While early protein drugs commanded a high selling price, rendering the cost of production almost irrelevant to the bottom line, it is becoming increasingly more important to reduce these separation costs for the new, higher volume, lower margin products now coming on line. This calls not only for improving individual purification steps but also for combining these steps into integrated unit operations to enhance the overall efficiency of the process train. Adsorptive separation operations, which play an important role in protein purification processes, suffer from limitations that must be overcome if the production costs are to be lowered. In column chromatography, for instance, which uses packed beds of porous beads, high pressure drops limit the flow rates through the column, and hence the overall production rates are low. Typically, the beads used to pack the columns are fairly large (∼100 μm) so that access to the internal pore surface areas is limited by pore diffusion, which is often the rate-limiting step in adsorptive purifications (1, 2). The mass transfer rates can be improved by reducing the size of the particles, but the attendant increase in the pressure drop makes this an unattractive option. In addition, the packed bed will foul and plug when challenged by feed streams containing cells and other colloidal debris. Thus fermentation broths must be clarified either by filtration or centrifugation prior to being introduced to a packed column, adding cost and complexity to the protein purification process. Whole cell broth may be processed directly with expanded bed systems, where the beads are fluidized by the feed stream to provide dynamic interstitial spaces through which the cells can pass freely. The dynamic flow rate range for such columns is limited on one hand by the need to maintain a fluidized bed, while on the other hand the rate should not be so high that the particles are entrained and lost with the exiting flow. The fluid residence times are short, and the capacity of such systems is typically quite low, ∼10 mg protein/g support, even though the equilibrium capacities for the separation media can be 3- to 10- fold larger. Longer contact times can be attained by suspending the beads in the fermentation broth in a continuously stirred tank, in which the residence time can be controlled at will. The particles can be recovered using either filtration or centrifuga- tion, both of which may be compromised by the presence of the cells and cellular debris in the system. To overcome many of these problems, it has been suggested that nano- (3, 4), sub-micron-, and micron-sized (5-8) magnetic adsorbent particles be used for the capture of the species of interest, whether they be proteins or other low molecular weight compounds, and that the magnetic particles be recovered using high gradient magnetic separation (HGMS) technology devel- oped initially for the minerals industry. The dispersed nature of the suspended particles and the open configuration of HGMS columns allow for the direct processing of whole cell broth, while overall production rates can, in principle, be significantly greater than those attainable with packed or expanded beds. Thomas and co-workers have used sub-micron- and micron- sized magnetic beads (>300 nm) functionalized with appropriate ligands to develop a robust technology for affinity and ion- exchange separation of proteins (5-8), while in independent studies we have investigated the prospects for using magnetic nanoparticles themselves (∼10-50 nm) as adsorbents for the recovery either of organic compounds from aqueous streams (4) or of proteins from protein mixtures (3). In this paper we explore further the prospects for using magnetic nanoparticles and small nanoclusters of these particles in protein recovery operations. Colloidally dispersed magnetic nanoparticles have already shown considerable promise for a wide range of applications (9) and have been used as sealants, damping agents, drug delivery vehicles, MRI contrast agents, and separation aids. In many cases, these colloidal dispersions of magnetic particles, or magnetic fluids, consist of magnetite (Fe 3 O 4 ) nanoparticles, typically ∼10 nm in size, coated with surfactants (10, 11) or * To whom correspondence should be addressed. Tel: 617-253-4588. Fax: 617-253-8723. Email: tahatton@mit.edu. † Department of Chemical Engineering, Rice University, Houston, Texas, 77005. ‡ Present address: Shire Biologics Inc. Northborough, MA 01532. 1153 Biotechnol. Prog. 2006, 22, 1153-1162 10.1021/bp050290t CCC: $33.50 © 2006 American Chemical Society and American Institute of Chemical Engineers Published on Web 06/08/2006