3D print your own chromatography column: material development and characterisation of ion exchange monoliths for protein purification

Abstract

Chromatography is one of the most important tools for the purification of therapeutic proteins and in analytical separations. Commonly employed chromatographic stationary phases are characterised by a randomly organised bed of porous spherical beads. Computer simulations have demonstrated that ordered structures would greatly improve purification performance. With the inception of 3D printing, fabrication of such ordered structures is now possible, but their application for protein separations is limited due to the lack of suitable materials compatible with both chromatography and 3D printing. This PhD thesis aims to overcome this limitation by developing novel materials for the 3D printing of functional ion exchange monoliths in one simple fabrication step.

To achieve this, Digital Light Processing (DLP) printing, based on photopolymerisation, was employed as 3D printing technique. Through co-polymerisation of bifunctional monomers, i.e. bearing both the desired chromatographic ligand and a polymerisable group, monoliths with ion exchange properties were fabricated. DLP printing allows the creation of features in the range of 100 – 500 µm. In order to generate pores at nanometre scale, pore forming agents, such as organic solvents, were introduced into the printing formulation. These agents cannot be polymerised, hence can be washed out after the printing process, leaving a porous network. The 3D printed monoliths were optimised in terms of ligand density, protein binding and recovery, chemical and mechanical stability, printability and porous morphology.

In an initial proof-of-concept study, acrylate quaternary ammonium (QA) functionalised monoliths were 3D printed, enabling a maximum bovine serum albumin (BSA) static binding capacity of 74 ± 6 mg/mL and the possibility to print features as small as 200 µm. Low protein recovery (<10 %), high compressibility and extensive shrinkage at high ionic strength rendered its application unsuitable for dynamic separation experiments. Subsequently, the initial formulation was optimised by introducing a multifunctional crosslinker and replacing the pore forming agents. This resulted in improved chemical and mechanical stability, the creation of pores in the order of 271 ± 120 nm, a 40 % higher BSA binding capacity and protein recoveries up to 50 % in static experiments. In dynamic experiments, monolithic columns with Schoen gyroid bed geometry (50 % porosity, 500 µm walls) demonstrated reproducible anion exchange behaviour for model proteins and successfully captured Cphycocyanin from crude cell harvests. The flexibility of the formulation was demonstrated by alternating the bifunctional monomer to introduce carboxylic groups. The resulting weak cation exchanger showed binding capacities of up to 110 mg/mL lysozyme.

To improve the monolith’s stability in alkaline solutions, a third material based on methacrylate was developed. The methacrylate monoliths had pores one order of magnitude larger than the previous acrylate monoliths (1.0 ± 0.5 µm), at the expense of available surface area and binding capacity (34 ± 2 mg/mL BSA, static conditions). However, minimal swelling, protein recovery up to 80 % in dynamic conditions and sodium hydroxide compatibility were achieved. Material optimisation allowed to 3D print Schoen gyroid monoliths with 250 µm wall and channel width. Their dynamic binding capacity was highly depending on the employed flow rate due to mass transfer limitations. Total binding capacities of up to 6.6 mg/mL BSA were achieved and model proteins were successfully separated.

Overall, this work demonstrates, for the first time, the direct 3D printing of ion exchange monoliths for chromatography in one step. Their suitability as chromatography media was proven by separating model proteins as well as capturing proteins from complex samples. The material provided properties in line with commercial materials, however, their printable feature size requires further reduction to compete with commercial columns. Nevertheless, the presented approach proves that co-polymerisation of bifunctional monomers is a simple but versatile approach, and it is expected that in the future chromatography users will be able to 3D print their own customised chromatography columns in their lab, on demand, cheaply, consistently, and rapidly.