The downstream affinity capture step has been recognized as a major bottleneck to productivity increases in biopharmaceutical manufacturing. Current commercial protein A affinity resins suffer from the following limitations:
# They offer only a limited capacity for diffusive mass transfer because they occupy a large amount of intrapore space in porous media and thereby hinder diffusion of mAb adsorbate molecules to their binding site within the porous matrix. Low molecular mass ligands are much better suited to meet operational requirements in downstream processing.
# They are typically eluted by denaturing buffers at extreme pH. This is cost efficient but disadvantageous, because this denaturing step is a potential trigger for product aggregation, in particular for some therapeutic antibodies which are sensitive to pH extremes. Conventional low pH elution and its negative effect on product integrity also hampers the development of true continuous approaches to downstream biomanufacturing that typically involve longer resin paths and column residence times resulting in extended contact time between product and eluent.
# Protein A affinity ligands tend to leach off from the column matrix and these “leachables” are a risk to patient safety.
Therefore, the industry needs new types of affinity ligands that are small and nonproteinaceous, that allow product capture and elution at physiological pH. and have a low safety risk profile. Carbohydrate structures of glycoproteins are an ideal candidate structure for ligand development since they are small, non-toxic, low immunogenic and still reach a level of complexity that may potentially enable highly specific and multivalent binding to other carbohydrate structures offering multiple sites for binding. Such ligands offer the possibility for mild competitive elution of bound glycoproteins at near neutral pH and offer potential for advanced specificity and selectivity towards product as well as towards product- and process-related impurities. Last but not least, they are likely to account only for low or even non-immunogenic column leachables. Such carbohydrate ligands may also hold potential for simultaneous product capture and depletion of undesirable Man5-glycosylated product variants. Such a feature would enable great cost-saving potential for industrial downstream processing of therapeutic glycoproteins. In summary, multivalent carbohydrate-based ligands are likely to have distinct advantages over current state-of-the art. Therefore, the aim of this project is to validate the potential of the proposed oligosaccharide structures as ligand scaffolds for multivalent transition metal or boron mediated binding of protein-linked N-glycans and to demonstrate that multivalent carbohydrate-based ligands as proposed in this project allow binding and elution of glycoproteins with a much higher selectivity as currently achievable with current monovalent ligands such as phenylboronic acid. The low molecular weight carbohydrate ligands proposed in this project are expected to enable the synthesis of chromatography supports with higher ligand density and increased dynamic binding capacity per column bed volume useful in continuous manufacturing approaches. Specific objectives of this project are the following:

# Proof that boronic acid conjugated and metal coordinated oligosaccharide structures can bind to protein-linked glycans in a multivalent fashion. Identify optimal binders and binding conditions. Investigate whether an additive and synergistic increase in ligand affinity can be observed.
# Evaluate the selectivity of ligand binding.
# Optimize ligand elution under physiological conditions.
# Verify the low immunogenic potential of column leachables.
Relevance: Downstream capture is the most expensive step in the biologics manufacturing process and the most dominant contributor to overall manufacturing Cost of Goods. Expenses for consumables including affinity resins account for a larger part of the operational expenditure for downstream processing compared to salary and wages for deployed personnel. Approximately 80% of the total process costs for a typical mAb production process are allocated to the steps following fermentation, with up to 60% of the downstream costs coming from chromatography [5]. The most significant contributor to the total cost is the Protein A resin. The cost of Protein A resin is nearly 50% higher than traditional chromatographic media with nonproteinaceous ligands [5]. Resin cost for a large column (>1-m diameter) can exceed $1 million [5]. The global Preparative and Process Chromatography Market is expected to reach $9 billion USD by 2019 from $6.4 billion in 2014, growing at a CAGR of 7.1% from 2014 to 2019 [48]. Depending on the pace of adoption of the proposed types of novel affinity resins by industry, a sizeable fraction of this global market for process chromatography consumables may be captured. While it is less likely for the technology to be employed in the manufacturing of current legacy products, it holds much more potential to be employed in bioprocessing of new therapeutic proteins currently in preclinical development. Fortunately, the market demand for process development for therapeutic proteins is highest in the preclinical phase of drug development. According to Research and Markets: Recombinant Therapeutic Proteins Market & Pipeline Analysis, 2014, 128 recombinant therapeutic proteins were in the preclinical phase of development and 340 drugs had already entered the clinical phase [48]. Only 83 marketed recombinant therapeutic protein drugs were reported for the same timeframe [48].

• Prof. Dr. rer. nat. Hans Henning von Horsten (Projektleitung)

• Dr. Karin Habermann (Projektmitarbeiter_in)
• Dr. Grit Sandig (Projektmitarbeiter_in)

Deutsche Forschungsgemeinschaft (DFG)

Sachbeihilfe