Mycoplasma Clearance and Risk Analysis in a Model Bioprocess: Robustness with Respect to Mycoplasma Species.

Capture bioprocessing unit operations were shown previously to clear or kill several log10 of a model mycoplasma Acholeplasma laidlawii in lab-scale spike/removal studies. Here, we confirm this observation with two additional mollicute species relevant to biotechnology products for human use: Mycoplasma orale and Mycoplasma arginini. Clearance of M. orale and M. arginini from Protein A column purification was similar to that seen with A. laidlawii, though some between cycle carryover was evident, especially for M. orale. However, on-resin growth studies revealed that residual mycoplasma in a column slowly die off over time rather than expanding further for all three species. Solvent/detergent exposure completely inactivated M. arginini though detectable levels of M. orale remained. A small-scale model of a commercial low-pH hold step did inactivate live M. orale, but this inactivation required a lower pH set point and occurred with slower kinetics than previously seen with A. laidlawii. Additionally, UVc irradiation was shown to be effective for A. laidlawii and M. orale inactivation while virus-retentive filters for upstream and downstream processes, as expected, cleared A. laidlawii. These data argue that M. orale and M. arginini overall would be largely cleared by early bioprocessing steps as shown previously for A. laidlawii, and barrier technologies can effectively reduce risk from media components. For some unit operations, the other two species may be hardier, and require more stringent processing or equipment cleaning conditions to assure effective mycoplasma reduction. By exploring how some of the failure modes in commercial antibody manufacturing processes can still eliminate mycoplasma burden, we demonstrate that required best practices assure biotechnology products will be safe for patients.