Synthesis and phospholipidosis effect of a series of cationic amphiphilic compounds: a case study to evaluate in silico and in vitro assays

In recent years, a large number of in silico and in vitro assays have been developed for safety assessment in early drug discovery. These methods are usually validated using datasets of known drugs with large chemical diver- sity, while application to homologous series has been rarely explored. Here we report a case study about phospholipi- dosis (PLD) risk evaluation for a dataset of nine com- pounds, designed and synthesized to modulate the physico- chemical properties typical of cationic amphiphilic com- pounds (CADs), representing the main class of PLD indu- cers. Our aim was to investigate the effect of structure modification on PLD induction according to a number of standard in silico and in vitro methods. As a result, we found that different in silico methods lead to conflicting results when applied to our series of weak PLD inducers, thus the apparently easy-to-use definition of CADs requires special attention. Moreover, when weak inducers are tested in vitro, the revealed PLD effect may vary based on the purity grade of the tested compound and the features of the selected assay. Finally, we have shown that slight mod- ifications on a chemical scaffold can have an impact on the PLD effect. This study also exemplifies that current in silico methods possibly overestimate the PLD induction effect of cationic amphiphilic compounds compared to the in vitro, with the risk of discarding promising compounds based on incorrect safety liabilities.