Ultra scale-down of elution chromatography.
Doctoral thesis, UCL (University College London).
It is highly advantageous to be able to develop bioprocesses early on in the product design lifecycle, where strategic options and alternatives can be considered cheaply and effectively. Resources are however often in very short supply at this point in the process, typically with very limited quantities of feedstock available and with minimal access to capital equipment. As such there is significant advantage in being able to develop chromatographic processes at very small scales that will only require small quantities of feedstock and can utilise common laboratory equipment. If the height of the packed bed is not maintained during scale up then it becomes difficult to predict chromatographic behaviour and efficiency. This seriously limits how small a chromatography system can be and still be representative of how a process scale chromatographic separation will behave. Ultra scale-down (USD) methodologies for chromatography take a different approach. A small scale device, which may or may not have geometric similarity to the process scale equipment, generates data which when combined with a specific methodology and mathematical model allows the prediction of the large scale equivalent. The work in this thesis sought to develop a USD methodology and model to accurately predict large scale chromatographic behaviour using very small scale devices. The development of a model separation was required to act as a source of realistic experimental data for the development of a USD methodology. This model separation required a suitable feedstock, a suitable USD device and a suitable sorbent on which to perform the separation. It was found that the most suitable feedstock of those tested was a FAb fragment containing periplasmic lysate produced in E. coli, due to the ease of pre-chromatographic processing and industrial relevance. Several designs of very small column were investigated and it was found that PRC pre-packed columns (Pall Life Sciences) had superior separation characteristics and were therefore selected as the USD device of choice. This was combined to produce a viable separation process using MEP HyperCel presented in the PRC pre-packed column format with FAb fragment containing periplasmic lysate as a feedstock. A linear pH gradient produced a clearly resolved two peak system with excellent FAb fragment purity that was deemed very suitable as a reference separation for the USD methodology development. The premise of the USD methodology was the deconvolution of each relevant peak within an elution profile into its four curve coefficients, namely height, width at half height, skew and peak location. These four curve coefficients for each peak in the small scale chromatogram could then be individually modelled using a transformation function into the large scale equivalents and then recreate a large scale chromatogram prediction from these values. This was also used to predict how the chromatograms will change with respect to altering the packed bed height and linear velocity of the loading and elution steps. The methodology that was developed was shown to be effective and was typically accurate to under 5% difference normalised root mean square when the predicted large scale chromatograms and real large scale data was compared. The methodology was further validated by testing with a range of different chromatographic systems and processes. These included changing the feedstock by reducing the FAb fragment titre by 50%, changing the chromatographic ligand to PPA and also a multi-variate change that altered the ligand, the sorbent backbone and the feedstock all at once. The transformation functions were found not to be generic and required system specific alterations. However, the methodology itself was shown to be very effective across a wide range of chromatographic conditions and systems and as such would be a good basis for ultra-scale down development in elution chromatography.
|Title:||Ultra scale-down of elution chromatography|
|Open access status:||An open access version is available from UCL Discovery|
|UCL classification:||UCL > School of BEAMS > Faculty of Engineering Science > Biochemical Engineering|
Archive Staff Only