In a previous hit-to-lead research program targeting anticancer agents, two promising lead compounds, 1a and 1b, were found. However, the poor solubility of 1a and 1b made difficult further in vivo studies. To solve this problem, a lead optimization was conducted through introducing N-methyl-piperazine groups at the 2-position and 6-position. To our delight, the optimized analogue 1d showed comparable antiproliferative activity in vitro with better solubility, compared with 1a. Based on this result, the replacement of the benzothiazole scaffold with benzimidazole and benzoxazole moieties afforded 1f and 1g, whose activities were fundamentally retained. In the preliminary in vitro biological evaluation, the immunofluorescence staining of HCT116 cells indicated that 1d, 1f and 1g led to cytosolic vacuolization which was not induced by 1a at low micromolecular concentrations. These results suggest that these optimized compounds might potentially constitute a novel class of anticancer agents, which merit further studies.