1. Bioseparations

    Bioseparations are the key processes in the production of biologics such as therapeutic proteins, antibodies, vaccines and DNA. Various chromatographic methods have been playing the central role in the bioseparations. There are various chromatographic methods based on the different interactions between ligands and proteins, such as size-exclusion chromatography (SEC), affinity chromatography, ion exchange chromatography (IEC), hydrophobic interaction chromatography (HIC), and mixed-mode chromatography (MMC) (also called hydrophobic charge induction chromatography. HCIC). Except SEC, all other chromatographic methods can be regarded as adsorptive chromatography. Packed-bed chromatography are mostly employed. Besides, there are other operation modes for adsorptive chromatography to meet the needs of high separation efficiency in some specific cases. These include expanded-bed adsorption (EBA) or fluidized-bed adsorption (FBA), and magnetically stabilized fluidized bed adsorption. Our lab has long worked on the various aspects of chromatography as described below in detail. Our recent work has focused more on affinity ligand design and polymeric ligand-based chromatography.

    Ion exchange chromatography (IEC) has been playing the most important role in the downstream processing of therapeutic proteins. During the past decade, polymer-functionalized IEC adsorbents have been developed and widely studied. Compared with traditional IEC resins whose ion-exchange groups are located on matrix surfaces, polymer-grafted resins generally show significant increases in dynamic binding capacities of proteins, which reflects significant improvements in both equilibrium capacity and mass transport kinetics. Yet, the mechanism for the fast uptake behavior was not well understood. To address this problem and to develop more high-performance ion exchangers for protein IEC, we have fabricated different charged polymer-based ion-exchangers and studied their protein adsorption equilibria and uptake kinetics.
    We have emphasized on the use of poly(ethylenimine) (PEI) of 60 kDa as an anion-exchange ligand of IEC by modification onto Sepharose FF gel. By studying the effect of its modification degree (ionic capacity, IC), we found that at the narrow IC range from 600 to 740 mmol/L, over six-fold increase in uptake rate occurred. The results indicate that there was a critical IC (cIC, 600 mmol/L) or critical PEI chain density at which happened a "Chain Delivery" effect that facilitated the transport of bound protein molecules by the interactions of neighboring chains mediated by the bound protein molecules when the adjacent PEI chains became close enough. The chain delivery effect was demonstrated by using different proteins and by a detailed study on the effects of ionic strength. Because the highly charged PEI chains might be unfavorable for its working as an ion exchanger, we then reduced the charge density of PEI chains by neutralizing the amine groups on PEI-modified Sepharose FF at IC=740 mmol/L. It was interesting to note that when the IC value was decreased to 440 mmol/L, a three-fold increase in the uptake rate as compared to FF-PEI-740 was observed. It is convincing that the charge density reduction significantly lowered the number of binding sites of protein molecules, leading to an enhanced chain delivery effect. Finally, it was found that the surface ligands assisted the transport of bound proteins on polymer chains in mixed-ligand resins. That is, surface ligands worked as "transfer stations" between two neighboring chains, resulting in enhanced transport of bound proteins on chains. The research thus disclosed the unique role of surface ligands in facilitating protein uptake kinetics onto polymer-grafted ion-exchangers.

    Affinity chromatography (AC) is an adsorptive chromatography based on the specific interaction between a protein and its ligand immobilized on a solid matrix. Due to the high selectivity and purification power, AC has become one of the best ways to the purification of biomolecules. However, the difficulty in the use of AC is the lack of specific ligands for the target molecules. So, our research focus on AC has recently been shifted toward selecting and designing ligands of high affinity and specificity. Flexible docking simulation has been used to screen peptide libraries and affinity peptide ligands have been screened for several proteins. At present, the screening strategy has been extended to a comprehensive approach consisting of amino acid location, molecular docking, molecular surface analysis and MD simulations. The approach is expected to help fast and reliable design of new and robust affinity ligands for therapeutically important proteins such as monoclonal antibody.

    Research in this area is mainly concerned with the acquisition of fundamental knowledge on the equilibrium theory and process kinetics of protein adsorption in chromatography. The systems involved in the studies include dye-ligand affinity adsorption, hydrophobic adsorption and ion-exchange or electrostatic adsorption. Predictive models for the salt effect on dye-ligand and hydrophobic adsorption equilibria have been established and pH and ionic strength dependence of model parameters in the steric mass-action model is studied based on chromatography and batch adsorption experiments. In addition, we have performed extensive studies on diffusion models describing the adsorption process of proteins by finite batch adsorption and chromatography incorporating visualization of intraparticle concentration profiles by confocal laser scanning microscopy. These are expected to contribute to the development of chromatography theory of proteins.

    The microscopic behavior in various chromatographic processes is crucial for the separation process and performance. Research in this area is mainly concerned with the microscopic description of adsorption mechanism and its implications to protein chromatography by molecular simulations. The systems involved in the studies include modeling of adsorbents, adsorption and desorption processes, protein conformational transition within adsorbent pores. At present, a coarse-grained adsorbent pore model of HCIC has been established. Protein adsorption, desorption and conformational transition in the HCIC pore and its implications to the separation performance has been extensively by molecular dynamics (MD) simulations. Moreover, researches into the molecular and microscopic performance of affinity adsorption and displacement chromatography are ongoing. The efforts would not only offer better understanding of the processes, but also be beneficial for the rational design of relevant separation materials and the process optimization towards high-performance protein chromatography.

    Ion-exchange chromatography (IEC) and hydrophobic interaction chromatography (IEC) for protein purification operation have some limitations, such as loading at low ionic strength for IEC and at high ionic strength for HIC, and limited loading capacity of HIC due to low ligand density. Mixed-mode chromatography (MMC) uses charged or chargeable chemical ligands that bind proteins by different mechanisms, such as electrostatic and hydrophobic interactions, hydrogen bonding, and pi-pi stacking. This kind of chromatography is more flexible in protein purification operations.
    One of the MMC is sometimes known as hydrophobic charge induction chromatography (HCIC). HCIC is based on the application of an ionizable and dual-mode ligand containing a heterocyclic structure and ionizable group. Generally, phenolic, pyridyl and imidazolyl derivatives are considered to be the most suitable candidates as ligands for HCIC. Dual-mode ligands developed for HCIC are very similar to those used in hydrophobic interaction chromatography (HIC) except that a high ligand density is employed. At such a high ligand density, protein can bind to chromatographic beads over a wide range of salt concentrations by hydrophobic interaction, exhibiting typical salt-independent characteristics, whereas hydrophobic binding of protein to HIC beads (having a low density of hydrophobic ligand) is promoted with lyotropic salts. Therefore, clarified feedstock can be applied to chromatographic step without the adjustment of pH and ionic strength. Elution of bound protein is easily accomplished by electrostatic repulsion between protein and ionized ligand after buffer pH is modulated to mild acidic conditions (3.5-5.5), in which the ligand takes on a net positive charge because of the dissociation of the ionizable group. HCIC is currently considered as a potential alternative to Protein A chromatography in the purification of antibodies. Up to now, only a couple of commercial HCIC beads (e.g., MEP Hypercel) are available. Moreover, the development of HCIC ligands specifically designed for high value-added recombinant and natural proteins and the exploitation of corresponding chromatographic processes suitable for industrialized production are other important topics in this field. A deep insight into the mechanism of protein binding is crucial to these researches mentioned herein. This subgroup is currently engaged in the following topics: (1) the development of novel HCIC ligands for antibodies, recombinant proteins and natural proteins from tissue fluids (e.g. blood); (2) a thorough understanding of thermodynamic and kinetic behaviors in protein binding on HCIC beads immobilized with the new ligands. Recently, several exclusive ligands have been discovered, which show attractive properties for the selective binding of antibodies and other proteins in a salt-independent manner. In thermodynamic studies, isothermal titration calorimetric measurement and adsorption isotherms are combined to find the binding mechanisms. An example is lysozyme binding to histamine-Sepharose; the result indicates that there is typical enthalpy-entropy compensation in protein adsorption on the bead and the adsorption is initiated by the dehydration of protein and ligands.

    Adsorption chromatography is one of the main techniques for bioseparations, and the design of high-performance solid matrices (adsorbents) is a main task in the development of efficient chromatographic technology. One research in this field has concentrated upon the design and fabrication of novel solid matrices with high column efficiency, high throughput, and low backpressure in biomacromolecular chromatography. To this end, we have developed novel methods for preparation of bidisperse porous materials. The methods for creating gigapores involve the incorporation of micron-sized solid granules and (W/O)/W emulsification. The gigapores created as such provide channels for intraparticle convective flow of mobile phase, so the interior mass transport is greatly intensified. Moreover, micropores (diffusive pores) connecting the gigapores provide a large surface area for high-capacity protein adsorption. Thus, the solid phase can be made with high column efficiency and dynamic capacity at mobile phase flow rate up to 30-50 cm/min. This implies that the adsorbents are promising for the high-speed purification of proteins and plasmid DNA for gene therapy.

    Preparative electrochromatography has shown its huge potential in protein separation. However, in the commercial application, some overwhelming challenges have to be faced, including the generation of Joule heat and electrolysis gas. Our Goal in this research area is focused on the development of high efficient, stable electro-chromatography system in virtue of the redesign of column architecture and deep understanding in electo-kinetic mechanism in electrochromatography. Recently, an electrochromatography with an oscillatory transverse electrical field was proposed. Due to the electric field is perpendicular to column axis, only a low voltage was needed to keep the suitable field strength in the column. In order to meet the requirement of introduction of electric field, the novel three-compartment chromatography column was designed, in which the column compartment and electrode compartments was separated by ceramic plate filled with polyacrylamide gel. Being packed with Sephadex G-75, the novel-designed column with a height of 12 cm showed a complete resolution of the mixture of bovine serum albumin and myoglobin.

    Expanded bed is a low back–mixing liquid fluidized bed achieved by the purpose-design of column configuration and solid matrix with a defined size and/or density distribution. Expanded bed technique offers potential advantages of both packed and fluidized beds. The upward flow through the bed of adsorbent provides higher void fraction within the bed, which makes it possible for particulate materials to pass through whilst the target bioproduct is adsorbed to the solid phase. The purpose-design of the column and adsorbents could ensure the expanded bed in an identical way to a packed bed chromatography to purify desired products from an unclarified feedstock. We have investigated the size and density distributions of two commercial media, and developed a mathematical model to express the size and concentration distributions of solid phase in expanded bed system. A theoretical model for the separation process has also been developed and evaluated based on the above studies. Various proteins including recombinant human interferon are purified by the expanded bed adsorption technology. For effective application of the expanded bed technology, our research also concentrates on the development of composite small-sized dense pellicular solid phase for expanded bed adsorption of proteins.
    Similar to expanded bed, magnetically stabilized fluidized bed (MSFB) is a low back-mixing liquid fluidized bed of magnetically susceptible supports by applying a weak, external magnetic field oriented axially or transversely relative to the flow. As an attractive alternative to conventional column chromatography, the liquid MSFB exhibits a combination of the properties of both packed and fluidized bed. These include the efficient fluid-solid mass transfer properties, elimination of solid mixing, low-pressure drop, resistance to clogging, and continuous counter-current operation. The lab has fabricated a novel magnetic agarose support for application in MSFB, which shows even somewhat higher dynamic capacity in MSFB than that in the packed bed. Current work is concerned with the elucidation of particle size and concentration distributions in MSFB and their dependence on various operating conditions.

    Amphiphilic molecules dissolved in organic solvents can self-organize to form thermodynamically stable reverse micelles containing nanometer-sized water droplets. The water pools can accommodate proteins, enzymes and nucleic acids, so reverse micelles have potential applications in bioseparation, biocatalysis, protein refolding and protein structure studies. This group has devoted to design new reverse micellar systems with for biotechnology application. The reverse micellar systems we have developed include new nonionic surfactants-based reverse micelles, dye-ligand affinity reverse micelles and metal-chelate affinity reverse micelles. The new reverse micelles show advanced performance in selective protein extraction. The current efforts are made to apply the metal-chelate reverse micelles to the purification and refolding of his-tagged proteins overexpressed in Escherichia coli.

    Surfactant molecules dissolved in water can form self-assembly micelles. Under appropriate conditions, the homogeneous aqueous micellar solution spontaneously separates into two macroscopic phases, both containing micelles, but one having a higher and the other having a lower surfactant concentration. Biomolecules will partition unevenly in the system attributed to the difference in physical-chemical properties between the micelle-rich and the micelle-poor phases. In addition, the high water content provides a gentle environment for biomaterials. So two-phase aqueous micellar systems have potential applications in the separation, purification and concentration of proteins. This group has been devoted to design new two-phase aqueous micellar systems for biotechnology applications. The current efforts are made to design metal-chelating two-phase aqueous micellar systems for the purification of his-tagged proteins overexpressed in Escherichia coli.

    The interest in plasmid DNA (pDNA) has increased due to the rapid evolution of gene therapy and DNA vaccines. In gene therapy applications, non-viral vectors would be preferred in clinical applications to minimize the risk of viral infection. This increases the demand for large-scale processes to manufacture pDNA of a high level of purity for use as therapeutic agent. Liquid chromatography is one of the most effective methods for the analytical and preparative separations of bio-molecules. Currently, pDNA and DNA vaccines have been purified by different types of chromatography such as gel filtration, ion-exchange, reversed-phase, hydrophobic interaction, hydroxyapatite, and affinity chromatography as well as other adsorbents. We have developed biporous adsorbents for high-speed plasmid purification by hydrophobic interaction chromatography and thiophilic interaction chromatography in a flow-through mode. Now, we are involved in the development of novel affinity ligands for the purification of supercoiled plasmid DNA from its isomers (open circular and linear plasmids).