Separation of intrinsically magnetic cells using magnetic filters

Cell separation is a common industrial, clinical, and laboratory technique that is typically accomplished via centrifugation. However other technologies for cell separation are employed when the samples need to be separated based on biological/physical properties apart from cell density. Magnetism is a technique that can be employed to separate biological entities. Cellular magnetism can be intrinsic (i.e., paramagnetic), such is the case for deoxygenated red blood cells (RBCs) and certain cancer cells that exhibit abnormal iron metabolism (e.g., glioblastoma), or this magnetism can be artificially induced based on magnetic particle labeling. In this work, we investigate magnetic filtration using commercial, high gradient magnetic separation (HGMS) packed columns for the separation of intrinsically magnetic RBCs. Specifically, flow rate, cell concentration, and external field strength were investigated using Miltenyi LS columns as magnetic filters. A generalized model for magnetic filtration was applied and used to investigate potential scale up of our RBC magnetic filtration process. COMSOL Multiphysics numerical simulations were performed to quantify the effect of the external field strength on the local magnetic energy gradient generated inside the ferromagnetic packed column. It was found that varying the externally supplied magnetic field from 0.1 to 1 T produces a non-linear increase in the magnetic energy gradient local to the ferromagnetic packing structure in the separator. Finally, the magnetic separator was used to demonstrate the capability of a binary separation of an artificially produced mixture of methemoglobin (metHb) containing RBCs (paramagnetic) from non-oxidized ferrous oxyhemoglobin (oxyHb) containing RBCs (diamagnetic). It was shown that 99 % of the magnetically isolated cells had a magnetic signature indicative of metHb-RBCs. In conclusion, we demonstrated the application of a magnetic filtration model for RBC separation and its utility to study the process variable effects on the separation's performance.