Downstream process development of next generation challenging proteins
Posted on 12/02/2018
The Knowledge Transfer Network (KTN), in partnership with GlaxoSmithKline (GSK) and GE Healthcare Life Sciences (GE), organised an event on downstream processing, held at the GlaxoSmithKline Medicines Research Centre in Stevenage (UK) on the January 30 2018 and attended by over 100 delegates from industry, academia and research institutes.
Aligning with KTN’s activities in medicines bioprocessing and manufacturing, the programme offered presentations from SMEs, large companies and academics. Will Lewis (GSK) kindly chaired the morning session with Sarah Goulding, KTN, highlighting the Medicines Manufacturing Industry Partnership (MMIP) and some funding available soon from Innovate UK in support of collaborative research and development in this area including an Engagement Workshop on March 7 in York for detailed information.
RECAP – morning session key highlights:
Nigel Darby (GE) set out the drivers and future challenges for downstream processing of biopharmaceuticals. He concluded that the main driver for improving bioprocessing yields is the high proportion of fixed costs on total biomanufacturing cost of goods, which can be reduced by process intensification as well as getting things right first time. Some of the challenges he highlighted are the diversification of biopharmaceutical portfolios, the fragmentation of the manufacturing base and controlling the complex supply chain.
Richard Kucia-Tran and Will Lewis (GSK) presented GSK’s approach to process improvement through three case studies. The first showed an approach of marginal gains to reduce cost and increase productivity using two examples: repetitive batch cycling and a modified protein A elution step. The second showed the optimisation of an innovative nanofiber resin developed by Puridify which allows for much shorter elution times and associated advantages such as smaller column sizes and increased number of cycles per day. Finally, a new paradigm called ‘OneStream’ was presented. This showed that by integrating the upstream and downstream process optimisation experiments it may be possible to design processes that present optimal outcomes for the combined upstream and downstream process that cannot be achieved by combining processes that have been optimised separately. In addition, the process scientists are also working earlier with discovery scientists in order to assess developability and use this as a selection criterion.
Dan Bracewell (UCL) showed data on the application of X-ray and neutron beams to packed beds, individual beads and molecular interfaces. The X-ray techniques generate 2D images that, using powerful computational technology, can be brought together to provide the impressive 3D images and videos that Dan presented. Applications of these technologies range from mapping tortuosity in beds as well as beads to characterising biopharmaceutical molecules binding to resin surfaces. Dan finished his presentation with work on HCPs which showed that Heat Shock Proteins increase in concentration in the cell culture supernatant over time and co-purify over protein A purification.
RECAP – afternoon session:
The afternoon session was chaired by Rachel Horsley (GE) and started with a panel consisting of Alison Tang (MedImmune), Richard Kucia-Tran, Nigel Darby, John Liddell (National Biologics Manufacturing Centre, CPI) and Dan Bracewell who discussed the future of bioprocessing.
The wide-ranging discussion explored engineering host cells with favourable HCP profiles, to cell-free expression systems to membrane crystallisation technology and selective nano-template recognitions directly from the fermentation broth instead of traditional chromatography methods, which is envisaged to be part of a continuous manufacturing process. Continuous processing also was a topic for lively debate with the consensus reached that there can be cost savings relative to batch production but that the regulatory hurdle is currently too high for the level of productivity improvement offered. Finally, the challenge that data management and analysis posed for the industry was discussed and it was argued that this is an area that needs to be addressed. This may also link to the current under-utilisation of automation in downstream processing.
Alison Tang presented on the challenges for process development of DuetMab bispecific antibodies. She echoed the approach previously presented by Richard Kucia-Tran of working closely with cell line development, bioreactor, analytical and formulation colleagues to facilitate development of a good process, dealing with the main challenge of DuetMabs which is mis-pairing. Alison presented on her work with a liquid handling platform to find optimal conditions for the chromatography steps to address this challenge. In addition, she optimised viral filtration resulting in a GMP-ready process.
To cover the subject of new resins we were grateful to have two other presenters. First, Oliver Hardick (Puridify, recently acquired by GE) talked about his company’s innovative chromatography resin Fibroselect which, as described above, due to its fibrous structure allows for much faster cycling time as compared to spherical resins. He argued that this provides an advantage relative to other process development tools. In addition, they are looking to develop the resin for GMP manufacture. Andy Masters (GE) then presented on their new protein A resin called MabSelect PrismA which, over a period of 40 years, is the eleventh protein A resin that GE have developed to match the dynamics of manufacturing monoclonal antibodies at scale. The improvements of the new material over previous generations are the increased dynamic binding capacity (25-40% depending on residence time) as well as improved resistance to cleaning with hydroxide (0.5-1.0M NaOH).
Focusing on the challenges posed to downstream processing by antibody fragments as biotherapeutics was Gráinne Dunlevy who presented on the automation Crescendo Biologics employs to develop downstream processes for their Humabody® VH molecules, as well as clone picking and sample dilution. The implementation of plate and liquid handling robots increased sample throughput from cherry picking to purified product by 60× whilst increasing data quality as well. Jim Pullen representing FujiFilm Diosynth Biotechnologies (FDB) then spoke about their work on a Horizon 2020 (EU) project called Amecrys which intends to study the application of the aforementioned membrane crystalisation technology for purification. The company’s role in this project is to supply two model biotherapeutics for the project: an anti-CD20 monoclonal antibody and a VH domain antibody HEL4. For the former a platform, scaleable process is available so Jim’s presentation focussed on the optimisation of a downstream process for the latter, for which they only had an “academic” purification method. Using liquid handling and other HTP technology, he showed the development of scaleable processes on affinity and mixed mode resins.
Ruth Rowland-Jones (GSK) then spoke about analytical approaches used in upstream bioprocessing. She presented case studies on the assessment of spectroscopy techniques. Raman spectroscopy came out as the superior tool and Ruth discussed potential applications to these technologies to downstream processing.
The final presentation of the day was given by Jelena Rusic (Cell and Gene Therapy Catapult) who introduced the Catapult and also presented analyses on the ease of optimisation and scalability of steps traditionally used in viral vector purification with a view of identifying those that can be used for GMP manufacture.
A lively and successful event was closed by Will Lewis (GSK) who thanked the organisers, contributors, attendees and sponsors of the event. Many positive conclusions were taken away from this collaborative event focussing on the future of downstream process development of next generation challenging proteins.