Scaling Up of Steric Exclusion Membrane Chromatography for Lentiviral Vector Purification

Abstract

Lentiviral vectors (LVs) are widely used in clinical trials of gene and cell therapy. Low LV stability incentivizes constant development and the improvement of gentle process steps. Steric exclusion chromatography (SXC) has gained interest in the field of virus purification but scaling up has not yet been addressed. In this study, the scaling up of lentiviral vector purification by SXC with membrane modules was approached. Visualization of the LVs captured on the membrane during SXC showed predominant usage of the upper membrane layer. Furthermore, testing of different housing geometries showed a strong influence on the uniform usage of the membrane. The main use of the first membrane layer places a completely new requirement on the scaling of the process and the membrane modules. When transferring the SXC process to smaller or larger membrane modules, it became apparent that scaling of the flow rate is a critical factor that must be related to the membrane area of the first layer. Performing SXC at different scales demonstrated that a certain critical minimum surface area-dependent flow rate is necessary to achieve reproducible LV recoveries. With the presented scaling approach, we were able to purify 980 mL LVs with a recovery of 68%.

Impact of the Membrane Structure of the Stationary Phase on Steric Exclusion Chromatography (SXC) of Lentiviral Vectors

“…As previously described [18], it can be concluded that 10-20 membrane layers, as often described in the literature [14,[21][22][23], are not required for efficient purification and that the separation takes place in the first membrane layer. The influence of the membrane height on the purification performance is further evaluated in Section 3.3.…”

“…Third-generation lentiviral vectors were produced by transient transfection of suspension HEK293T/17 SF cells (ACS-4500, ATCC) with four plasmids in a UniVessel ® 10 L bioreactor operated by a BIOSTAT ® B-DCU (Sartorius, Gottingen, Germany). Detailed information on LV production and nucleic acid digestion can be found in [18]. The cell culture broth was clarified using Sartoclear Dynamics ® Lab V50 (0.45 µm polyethersulfone membrane version) with 5 g•L −1 diatomaceous earth (Sartorius).…”

“…In a recent publication, the capture of the LVs on the membrane was investigated, and it was found that the LVs were mainly captured on the first membrane layer [18]. We were further interested in discovering if the location of the captured LVs was influenced by the membrane structure and pore size, in order to gain further insights into the separation mechanism.…”

“…A volume of 50 mL was loaded, corresponding to a 25 mL LV solution mixed in a 1:1 ratio with a PEG buffer. The loading volume was defined based on previous experiments on membrane capacity [18]. The SXC runs were stopped before the elution step, and the five membrane layers were removed from the MA15 chromatography device and compared against a pre-wetted control membrane without LVs.…”

“…Understanding the capture location aids in understanding the fact that five membrane layers are sufficient; thus, material can be saved. A more detailed discussion of how this affects the scaling of the process can be found in [18].…”