Column Chromatography

Chromatographic separations involve the partition of a solute between a mobile and a stationary phase. Usually, the stationary phase consists of a particulate matrix packed into a tubular column through which the mobile liquid phase passes. The feedstock is introduced as a pulse into the column and individual components are separated by their differential distribution between the phases. For example, in size exclusion chromatography the stationary phase consists of the liquid held inside porous particles. The pores have a distribution of sizes which restrict the access of some species to the stationary phase and so separation is achieved on the basis of the relative sizes of the molecules in the mixture. In adsorptive chromatography, the retention of the solutes in the column is dependent on the strength of their interactions with the solid phase. There are various types of adsorptive chromatography which differ according to the chemical basis of the interaction between the solute and the stationary phase. Separations can therefore be achieved by means of different protein characteristics (see Table 4). To elute adsorbed proteins from the column, the characteristics of the mobile liquid phase are changed such that the interactions between the proteins and the solid phase are disrupted. Very high resolution separations can be achieved using a gradual change by means of an elution gradient. High resolution, combined with the availability of separations based on different protein characteristics, makes column chromatography an extremely powerful technique. By using several methods in sequence it is possible to purify to homogeneity a single enzyme from the thousands present in a cell extract. For this reason, column chromatography represents the state-of-the-art of protein purification.

In a survey of research papers (Bonnerjea et al, 1986), it was shown that ion-exchange chromatography was the most commonly used method of protein purification. On a process scale also, for the isolation of proteins for the pharmaceutical and fine biochemical industries, ion-exchange is used very effectively. However, there are limits to the scale on which such chromatographic techniques can be operated. The flow of liquid through the column is dependent on a pressure drop between the outlet and the inlet. As excessive pressure can damage the separation matrix, this limits the length of column which may be used if a reasonable flow rate is to be attained. Consequently, most process scale columns achieve large bed volumes by using low bed height and large column diameter. The largest process scale chromatography columns available (Amicon) are 200 cm diameter and 80 cm in height and therefore have volumes of up to 2500 1. Using Whatman DE52 anion-exchange medium (Levison et al, 1990), such a column would be capable of binding in the region of 200 kg of protein.

Table 4

Methods of column chromatography in protein separation

Table 4

Methods of column chromatography in protein separation

Chromatography application

Basis of separation

Means of elution

Size exclusion




Surface charge

Increasing ionic strength


Surface charge

Change of pH toward pi

Hydrophobic interaction

Surface hydrophobicity

Decrease in salt concentration


Biological interactions

Increasing concentration of free ligand, or change in pH or salt concentration

For most current uses of enzymes by the food industry, a high degree of purity is not required. The costs of column chromatography are therefore not justified. For those emerging food enzyme technologies which demand a higher degree of enzyme purity, process cost remains a major limitation to the use of column chromatography. In addition to the cost of the chromatography unit process, there are also the costs of associated processes to be considered. For example, particulate materials such as whole cells and cell debris can block the chromatography column and must therefore be removed. It may also be necessary to concentrate the protein or to change the solvent composition to achieve the conditions required for the adsorption of the target compound. Chromatography can rarely be carried out without preliminary processing of the feedstock by, for example, ultrafiltration. High resolution separation techniques are most effective early in the purification stream (Bonneijea et al, 1986). However, the presence of high concentrations of contaminating proteins in the feedstock at this stage reduces considerably the capacity of the matrix to bind the target protein. Consequently, larger and more expensive plant is required if chromatography is used in this way. If the target protein is in a relatively high state of purity at the start of chromatography then smaller columns may be used. In this case several preliminary low-resolution separation steps such as ultrafiltration and precipitation are often required to remove the majority of contaminants.

As chromatography is such a widely used method, a great deal of research is carried out by manufacturers of chromatography equipment and materials in the development of new products. If advances in chromatography are to be of benefit to the production of bulk enzymes, the reduction of material and process costs is an essential goal. There are a number of areas of development which could contribute to this. In a study of a-galactosidase purification, Porter & Ladisch (1992) demonstrated that the cost of the stationary phase can be a major component of total process costs, particularly at larger scales of operation. Total costs can therefore be reduced by increasing the operational life of the chromatography matrix, increasing the binding capacity and reducing the materials cost. Process time is also an important factor as this affects through-put, labour costs and overheads. Higher flow rates may be achieved using chromatography media which tolerate higher pressures, or have better flow characteristics. A reduction in the total number of unit processes which are required to prepare the feedstock for the chromatography step would also significantly reduce costs. The achievement of all of these aims in a single product is perhaps an impossible goal as novel chromatography media with higher binding capacities and flow rates than current products are likely to be more expensive. However, several new developments in chromatography do achieve some of the aims outlined above.

Adsorption to ion-exchange matrices is facilitated by the attachment of charged groups to a largely inert base material such as cross-linked cellulose or dextran. Conventionally, the ionic groups are located on short linear molecules and are thus held away from the matrix surface so that they are more accessible to the solute. Miiller (1990) suggested an alternative strategy to maximise protein binding. Rather than a rigid array of charged groups, he synthesised adsorbents with the charges grafted laterally on vinyl polymer chains. The high flexibility of the charged chains of these "tentacle" ion-exchangers increased the contact between adsorbent and protein ionic groups, while decreasing the distortion of the protein required for maximum ion pairing. The brush-like layers of charged polymer also reduced contact between the protein and the matrix base material, thus reducing non-specific adsorption. It was demonstrated that the tentacle ion-exchangers had protein binding capacities of 70-140 mg/ml, compared with 35-55 mg/ml for comparable conventional materials. In addition, the resolution of protein separations was much improved.

Chromatographic stationary phase particles are usually porous in order to provide the maximum surface area, and therefore binding capacity. The transport of solute to the outside of the particle is by convective flow, while transport into the interior is by a much slower process of diffusion. If the solvent flow rate through the column is too great and there is insufficient time for diffusion into the particle, this results in a loss of resolution and binding capacity. A novel chromatography particle structure has been developed by PerSeptive Biosystems with the aim of overcoming this problem (Afeyan et al, 1990). Larger pores allow penetration of solute into the interior of the particle by rapid convective flow and give access to smaller "diffusive pores" which provide a large surface area for adsorption (Figure 3). It is claimed that perfusion chromatography using this type of medium reduces the need for a compromise between flow rate, binding capacity and resolution. The manufacturer's literature states a flow rate of 1800 cm/h for a perfusion ion-exchange medium. This compares with approximately 20-200 cm/h for conventional chromatography materials. Lehman et al (1993) applied perfusion chromatography media to the large-scale purification of recombinant tick anticoagulant peptide from yeast culture. In a low pressure capture step, 24.3 g of product was recovered from 4001 of diafiltered fermentation broth. In a subsequent high resolution step, 16.7 g of product was purified in a series of rapid cycles using a 800 ml high pressure column. Compared to previous methods using conventional chromatography, the process time was reduced by half, recovery was increased from 32% to 47% and the need for low temperature facilities was eliminated.

Figure 3. The transport of solute into the porous structure of chromatography particles. Transport into the interior of conventional particles is entirely by diffusion. In perfusion particles, large pores allow initial rapid access of solute by convective flow followed by diffusion further into the structure.

Conventional chromatography systems operate under axial flow, in which the mobile phase moves along the axis of the column. It is possible that some of the limitations of chromatography can be overcome by using an alternative column geometry. In radial flow chromatography, the movement of mobile phase occurs across the radius of the column from the outer surface to an inner compartment (Figure 4). In any chromatography column, it is the area over which the liquid phase is applied which defines the flow rate at a given pressure. In axial flow this is a function of ni2. With radial flow, the liquid phase is applied over the whole outer surface area of the column, which is a function of 2nr x length. As a result, faster flow rates can be applied under radial flow than under axial flow for a given column volume. Smaller columns and shorter cycle times can therefore be used to achieve the required throughput. In addition, process scale-up can be carried out by increasing the column length, which is directly proportional to the bed volume and the flow rate.

Huang et al (1988) demonstrated the purification of porcine trypsin by radial flow affinity chromatography. A radial flow cartridge (800 ml) was constructed from spiral-wound sheets of modified cellulose material formed around a central core. Using p-aminobenzamidine as affinity ligand, crude material containing 30 g protein in 4 1 was processed at a rate of 17.7 1/h. This resulted in a purification factor of 6.3 and 57% recovery. They also demonstrated a linear scale-up of flow rate and binding capacity with cartridge size,

0 0

Post a comment