Principles Of Chromatography

The fundamental principle behind the development of purification schemes for biological medicines should be to develop a robust process suitable for use at the scale required for manufacturing. Processes in manufacture should work the same each time, within predefined limits, and produce material meeting predetermined specifications. Yields need to be predictable and as high as possible, so that the scale can be set long before implementing transfer into manufacture. The key to ensuring robustness is to investigate limits around optimal values for parameters like pH, salt concentration, column loadings and flow rates. This should be done in the early stages, preferably during method screening.

Defining limits for variables involves testing different conditions, for example the salt concentration required to elute product from a column without eluting contaminants, and checking that the same result is achieved within a reasonable range either side of the optimum. It is useful to determine the effects of using parameters outside the range, for example overloading a column may cause product breakthrough and result in a low yield and suboptimal performance, whereas under-loading a column could result in heavy contamination of the product with impurities and a failed batch. Flow rates used may affect the elution volume and this may be important if the subsequent step relies on the loading volume, such as size exclusion. Variables that have a significant impact on product quality should be deemed to be critical variables (Seely et al. 1999). Critical variables are also those that are required to be very tightly controlled, such as the concentration of salt used for wash and elution of an ion exchange column, or variables that are difficult to control at scale, such as the temperature of the facility where, for example, hydrophobic interaction chromatography - which is temperature dependent - is performed.

Separation of product from host-cell protein impurities, DNA, and endotoxin can be achieved by a number of commonly used techniques. With the development of a purification scheme for biological medicines it is usual to end up with a process consisting of steps that function orthogonally, i.e. each with a different mechanism of separation to the others. The functions of these steps should be identified during screening, and the performance of each monitored in relation to its function as the process is put together.

Selectivity is the most important factor when choosing a chromatography matrix for use in a purification scheme. Selectivity is the ability of a matrix to interact with the product molecule in a different way to the impurities. A matrix that binds product and not the impurities has a high selectivity. A different matrix that has the same ligand chemistry but a different support, or a matrix from a different supplier, may have a different selectivity and might bind impurities differently. The selectivity is dependent on the properties of both product and impurities and is something that can only be determined empirically. The conditions yielding optimal selectivity on any given matrix also have to be determined experimentally.

A simple purification scheme may take clarified harvest material and use ion exchange as a capture step. Since salt is used to elute from the ion exchange matrix, hydrophobic interaction might be a useful choice as a second step because it requires salt to be present during loading. A polishing step might be size exclusion, which could simultaneously remove impurities from the product, and change the buffer into that required for formulation.

Any process step will be dependent on the performance of the preceding steps. Optimization of the performance of the process should be done as a complete unit by testing the effects of the optimization of any one step on the performance of the process as a whole. Inevitably, this will lead to compromises in the optimization of individual process steps (Ngiam 2003). A simple example may be where the optimization of the elution conditions of one step may affect the binding of the product onto the subsequent column. It is more likely that problems will occur where the optimization of one step alters the impurity profile and the challenge for a step further down the process; this can mean that new impurities or higher levels of impurities remain at the end of the process and it is important to know how the change arose. This can be resolved by running a full process with the changes to one step in place, along with full analytical support to test the purified drug substance. Changes to the process should, therefore, be introduced stepwise.

Such an approach should be taken to the purification development of any biological medicine, including proteins, viruses and gene therapy vectors, whatever cell type and cell culture method is used in the upstream process.

Chromatography matrices can be made of a variety of different ligand chemistries, with beads made of a number of different materials. Each manufacturer is likely to provide similar chemistries attached to their own type of bead. Most columns are run as packed beds, and the majority of resins are intended for use in this way. Beads are manufactured from soft materials such as agarose, or from materials that produce a more rigid bead such as methacrylate polymer, or poly(styrene divinyl benzene). Most of the leading manufacturers will be able to provide a drug master file (DMF, see also Section 18.4) for chromatography resins designed for use in biological medicines production; it is important to check that this is the case before using a resin intended for clinical production.

Bead sizes also vary, and in general a small bead size of, say, 20 |im, although giving higher resolution at small scale, is impractical for large-scale use, mainly because of the resultant high backpressures requiring expensive pumps that are impractical at industrial scales. Large beads of 150 |im or more can give poor dynamic binding capacity, due to the increased volume between functional groups on the large beads, and may require low flow rates to achieve high loading levels. Low flow rates can impact on the time a step will take, and hence adversely affect production efficiency. Low loading levels may make the step less economic, requiring more matrix for a given load. Larger beads may be of particular use in virus- or gene-therapy applications, as the large beads can allow the large virus particles to flow through the column without being filtered out by the matrix. Virus particles will not diffuse into the beads and will bind to the bead surface, and so binding capacities are likely to be low, though 109 particles per ml of packed bed can be achieved. Bead sizes of around 50 |im are most likely to suit production of most protein products; pressures of around 0.5 bar are normal, high flow rates can be achieved and protein loading levels of around 50 mg/ml packed bed are achievable, for example, with ion exchange resins.

Column packing efficiency is an important factor to take into account during process development for biological medicines. The normal measurements are number of theoretical plates (N) and asymmetry (As), the calculations of which are shown in Figure 18.2. The height equivalent to a theoretical plate (HETP) is often referred to where HETP = L/N, and L is the length of the column.

The number of theoretical plates and asymmetry are used to determine how much resolution a column of a certain bed height will give, and how much an elution peak will spread during elution, normally by peak tailing, respectively. It is important that a process in development does not rely on column packing efficiencies that cannot be replicated in manufacturing at large scale. For example, a size exclusion column with N of around 2000 and As of around 2 should be achievable at large scale. Although values of over 10 000 for N and under 1.5 for As may be possible at large scale, they are more readily achieved in the laboratory. A separation where the resolution is critical and is dependent on the packing may be unachievable upon scale-up, and may not be transferable, entailing expensive and time-consuming rework at a critical stage in the project.

A technique that has been increasingly applied to the capture of biological products is expanded bed adsorption (EBA) (Thommes 1996). In such a system, the resin is not loaded as a packed bed. The upward flow of crude unclarified harvest material causes the beads to disperse up the length of a column about a metre high. The flow rate is balanced against the rate of the settling of the dense beads. The material is loaded onto the column and as the crude material passes through the column the beads remain suspended in the fluid. Once loaded, the matrix can be packed into a conventional bed, then washed and eluted as in conventional chromatography. The advantage of this technique is that unclarified material can be applied to the column, which may have advantages in some applications. For example, EBA can eliminate the need for clarification by centrifugation or large-scale filtration where unclarified material would otherwise block a conventional column.

Detector output (Absorbance or conductivity)

Detector output (Absorbance or conductivity)

Number of theoretical plates:

The elution volume (Ve) can be measured on the chromatogram from the point of sample application to the apex of the peak. The width at half the peak height is measured (W,/2) on the chromatogram.

Number of theoretical plates:

The elution volume (Ve) can be measured on the chromatogram from the point of sample application to the apex of the peak. The width at half the peak height is measured (W,/2) on the chromatogram.

Number of theoretical plates (N) = 5.54 x ( Ve / WK )


A vertical is drawn from the apex of the peak. At 10% of the peak height the distances from the vertical to the peak are measured; A at the front of the peak, B after the apex. Asymmetry (As) = B / A

Figure 18.2 The calculation of Number of Theoretical Plates and Asymmetry of a column.

Was this article helpful?

0 0
Healthy Chemistry For Optimal Health

Healthy Chemistry For Optimal Health

Thousands Have Used Chemicals To Improve Their Medical Condition. This Book Is one Of The Most Valuable Resources In The World When It Comes To Chemicals. Not All Chemicals Are Harmful For Your Body – Find Out Those That Helps To Maintain Your Health.

Get My Free Ebook

Post a comment