The nearest neighbor nuclei method to objectify analysis of pertussis toxin-induced clustering

The in vivo histamine sensitization test (HIST) has historically been performed to guarantee the safety of acellular per­tussis vaccine batches. Non-compliance of batches is primarily associated with the presence of low levels of pertussis toxin (PTx). Because of ethical, standardization and scientific reasons, a variety of alternative in vitro approaches have been studied to replace the lethal HIST. A broadly applied and partially accepted method is the CHO cell clustering test, which is based on the clustered growth pattern of CHO cells when exposed to minute amounts of PTx. One of the major hurdles for global application of the CHO clustering test is the manual assessment of the clusters, which is associated with suboptimal reproducibility of test outcomes and is time-consuming. Here, various parameters of CHO cell nuclei were evaluated in search for a reliable, objective read-out parameter. We demonstrate that the distance between each nucleus and its nearest neighbor (3N method) is the most suitable parameter to assess clustered cell growth. This method detects 2.8 mIU PTx/mL and thereby complies with the requirement set for the sensitivity of the CHO clustering test based on visual reading. In commercial acellular pertussis vaccines spiked with PTx, the method detects 45 mIU/mL PTx, which is substantially lower than the 181-725 mIU/mL PTx detected by visual interpretation. The 3N method thus allows objective and sensitive assessment of CHO clustering and thereby encourages broad and global implementation of the in vitro test as an alternative to the HIST.