The eukaryotic yeast Pichia pastoris is an increasingly applied platform for
heterologous protein production for therapeutic and diagnostic purposes. Recent
advances in high production titre have increased pressure on downstream
processes to reduce the overall cost of production. The work in this thesis
investigates a means of reducing the cost of P. pastoris production by combining
three steps (centrifugation or filtration, concentration and affinity purification) in one
step and comparing the primary capture of recombinant antibodies between radial
and expanded bed adsorption chromatography systems, using Immobilized Metal
Affinity Chromatography (IMAC) as a primary capture step.
A hexa-histidine (His6) tagged single-chain antibody shMFELL2cys was used for
the study. The protein is secreted from P. pastoris strain X-33 and has an affinity
towards the carcinoembryonic antigen (CEA), an onco-foetal tumour antigen
overexpressed in cancer cells used for therapeutic as well as diagnostic purposes.
A high cell density culture of P. pastoris, which had been induced to express
shMFELL2cys was directly applied to (i) Cellthru™ resin in a radial bed and (ii)
STREAMLINE™ chelating resin in expanded bed adsorption chromatography.
Copper-mediated IMAC capture of the tagged protein occurred while cells and cell
debris were allowed to flow through the matrix bed.
To conduct this study, first, the binding conditions were optimised at a small scale
using 1 mL IDA chelating Cellthru™ and STREAMLINE™ chelating resins used in
RBA and EBA chromatography columns, respectively. Second, a comparison of
the static and dynamic binding capacity of the RBA and EBA resins was performed
at a small scale using the purified His6 tagged scFv antibody. Third, an investigation
of the effect of the residence time on the recovery of the His6 tagged scFv antibody
using the RBA column was performed using a fermentation culture fluid.
The His6 tagged scFv antibody shMFELL2cys was successfully expressed in the
high cell density P. pastoris fermentation at a concentration of 680 to 600 mg/L,
and captured using the RBA and EBA chromatography processes, respectively.
The calculated step recovery for RBA and EBA chromatography was 30% and 47%, respectively. A total amount of 801 mg and 611 mg of the scFv antibody
shMFELL2cys was purified from the EBA and RBA chromatography processes,
respectively.
The scFv antibody shMFELL2cys purified using RBA and EBA chromatography
has specificity for carcinoembryonic antigens (CEA). A single monomer peak of the
purified scFv antibody was seen on the analytical SEC column and a scFv antibody
band at ~ 27 kDa was present on the SDS-PAGE for the scFv which was purified
using RBA and EBA chromatography. The HCP concentration in the final purified
scFv antibody using EBA chromatography was 161.8 ng/mL, equivalent to 39
ng/mg of scFv antibody. The HCP concentration in the final purified scFv antibody
using RBA chromatography was 83 ng/mL, equivalent to 22.3 ng/mg of scFv
antibody. The high HCP concentration in the EBA eluted protein in comparison to
the RBA eluted protein was also calculated.
Cost of goods analysis of the RBA and EBA chromatography processes were
performed from 8L lab scale to 2000L large production scale using Excel-based
BioSolve software®. At 2000L production scale, factors affecting the CoG/g and
CoG/batch of the His6 tagged scFv antibody were analysed using DoE software for
the RBA and EBA chromatography processes. BioSolve software® cost
comparisons between the capture steps revealed a significant decrease in the total
CoG/g of the product manufactured at 2000L scale. At best-case scenario of 6 g/L
fermentation yield, 40 mg/mL resin binding capacity, 60 resin CIP cycles and 60%
step recovery, calculated CoG/g were ÂŁ54 and ÂŁ197, respectively for the RBA and
EBA processes. In the worst-case scenario of 1 g/L fermentation yield, 20 mg/mL
resin binding capacity, 20 resin CIP cycles and 20% step recovery, calculated
CoG/g were ÂŁ556 and ÂŁ1400, respectively for the RBA and EBA processes.
Primary chromatography process RBA out performed the EBA chromatography
process, at both best-case and worst-case scenario levels.
Therefore, preference is given to introducing RBA chromatography as a primary
capture step to purify the His6 tagged scFv antibody from a high cell density
P.pastoris fermentation run, following the evaluation of quality, quantity and cost of
goods to manufacture the antibody.