VIRUS, VLP and PHAGE PURIFICATION
Downstream processing of viral vectors and virus particles for gene therapy and viral vaccines applications is one of the larger challenges in biomolecule manufacturing.
Virus particles are complex biomolecules, which differentiate by size, positive and negative charges, distribution of charges and hydrophobic properties. These physicochemical properties of viruses are defining features that can be used for isolation and purification.
The first parameter that must be taken into consideration is scalability of technique. Chromatography is well established as an easily scaleable technology for protein manufacturing. There exist different modes of chromatography (IEX, SEC, HIC…), which can separate molecules by various properties. These can be combined if necessary to achieve high purity and maximized yield for virus and vaccine preparations.
The second important factor is the size of the virus. Viruses have diameter from around 20 nm up to 500 nm or so, and frequently generate aggregates which extend to larger sizes. It is well known that virus cannot penetrate the pores of porous particles used for traditional protein chromatography. This results in extremely reduced Dynamic Binding Capacity (DBC), because virus material binds only to binding sites residing on the surface of particles. Such material can easily block chromatographic columns with small pores sizes, especially in the presence of DNA.
To overcome these challenges methacrylate monolithic materials (CIM monolith) were developed with extra large pore sizes (around 1.5um dia) that uniquely enable purification of big macromolecules such as viruses. Despite having a lower total surface area of internal larger pores, they enable dramatically enhanced capacity for viruses and other such massive species in comparison to conventional chromatographic media, which were optimized for protein purification, simply due to their unhindered accessibility for virus sized molecules.
Upstream production currently does not deliver high mass titers, so the other biologically derived impurities exist in relatively high concentration and must be removed during downstream purification.
The unique structure of CIM monolith provides a three dimensional matrix of interconnecting large pore channels which allows convectional pumped laminar flow of mobile phase. This carries adsorbates including virus sized materials directly to all the internal binding sites without reliance upon diffusion, as is the case with other chromatographic particulate materials. Binding and elution mass transfer is no longer slowed by diffusional constraints, and high flow rates provide high capacity and high resolution, even with such large virus adsorbates. Laminar flow at high flow rates helps to minimize shear forces which frequently reduce yield with particulate chromatography methods.
Most virus purification processes include at least one step using ion exchange chromatographic column (IEX). Which IEX column will be used depends on isoelectric point of virus and its stability under different conditions. This should be explored before attempting any chromatography. Stability in different buffers system with different NaCl concentration at various pH values are first examined.
Majority of viruses have isoelectric point below pH 6 and are stable in the range between pH 6,5 to 8,5. This means that an anion exchange column must be used. This is often chosen as a high purification factor capture step. In practice, the high ligand density of CIM monolith material assists virus binding at multiple sites, and many impurity proteins can be arranged to flow through monolith by loading at higher ionic strength, this helps to retain high capacity for the virus loading. DNA impurities will bind more strongly that virus material and this elutes later or remains on the column until a cleaning step is performed prior to reuse of the column. In order to improve purity and yield of active eluted virus, different buffers and pH values of these buffer systems should be tested. If virus is stable also at pH below isoelectric point, a cation exchanger can often be used. When used in addition to an anion exchange step, this can achieve extra valuable removal of coeluted proteins from the anion exchange step.
Viruses are large molecules and charge may not be uniformly distributed through across entire viral capsid. The result of this is a bipolarity that can cause virus to bind at the same pH value to both anion and cation exchanger supports. Influenza is one such virus.
For this reason, and certainly during a comprehensive process optimization regime, all chemistries should be tested prior to method optimization on the selected chemistry.
After elution conditions using linear gradient is determined, the method should be converted to stepwise operation and capacity must be determined. This should ideally be performed at small scale with purified virus material, but with high value materials this is not always possible. The availability of small scale devices and predictability of scaleup performance are essential to contain costs. Even when Dynamic Binding Capacity (DBC) and elution conditions are defined with pure virus, both parameters have to be later checked with real load material, because impurities can greatly influence the purification process.
For predictable scalability and method robustness a step gradient elution profile should be determined. It has to be considered that material from upstream can vary regarding virus titer and impurity concentration. A robust method will deliberately load less material than maximal DBC on to the column. In addition, good analytical methods should be developed to monitor the whole manufacturing process. In this way problems of upstream load reproducibility can be minimized and the method will tolerate such variability of source material.