due to inadequate yield, purity or poor overall economics. In particular, it has been reported that many proteins only have biological activity in their natural un-denatured state (i.e. native conformational state). The chemicals or heat introduced in the traditional processes can however easily denature the protein components.
| |
Figure
1: Selectivity for separation of 1.0 g/L α-LA and 2.5 g/L β-LG as a function of solution pH through a 30 kD CRC membrane. |
|
|
In recent years, membrane filtration has emerged as a potentially attractive alternative method for protein fractionation because membrane separation is a non-chemical process that allows the production of protein at high purity (more than 90%) and with virtually no denatured proteins. Two important applications of membranes in bioprocess are ultrafiltration (UF) and cross-flow filtration. UF is currently used for concentrating enzyme and protein, removing virus, dewatering whey, clarifying juices and so on. Cross-flow filtration with ultrafiltration membranes achieves highly selective separation between proteins and thus provides an alternative to conventional chromatographic process or affinity binding for protein purification at a lower cost.
In this study, experiments were conducted firstly in a total recirculation mode to study the effect of various physicochemical conditions on the separation of protein. Both the filtrate and retentate were recycled back to the feed tank to maintain constant protein concentrations. The filtration velocity was controlled and the filtrate and feed samples were collected for analyzing the protein concentrations.
Actual protein separations were then performed in a two-stage closed-loop cascade diafiltration mode to effectively remove the more permeable protein(s). The filtrate from the first-stage was directed to the feed tank of the second-stage separation system. The system was installed with 10 kD composite regenerated cellulose (CRC) membranes. The filtrate from the second-stage was returned to the first-stage feed tank. During the filtration, the filtrate flow rates were controlled at the same rate for the two stages, and thus the liquid levels in the two feed tanks were maintained constant.
Cross-flow ultrafiltration process was used to separate α-Lactalbumin and β-lactoglobulin that have a similar molecular size. Total recirculation ultrafiltration was first used to determine the effect of several physicochemical conditions including membrane pore size and material, solution pH and ionic strength, bovine serum albumin content and transmembrane pressure. Selectivity of up to 24 could be achieved with a 30 kD CRC membrane at pH 7.5 and ionic strength 50 mM (see Figure 1). A two-stage closed-loop cascade membrane process was then used for the separation of α-Lactalbumin and β-lactoglobulin. An 80% β-lactoglobulin yield was achieved as the first-stage retentate product after 12 diavolumes, with α-Lactalbumin less than 1% of its initial concentration; a 16 diavolume resulted in more than 99% α-Lactalbumin yield as the second-stage retentate product, but about 25% of initial β-LG was also accumulated
in this solution simultaneously (see Figure 2). The results demonstrate the prospects of using membrane separation technology to achieve highly selective separation and purification of biopharmaceutical products.
