Despite extensive investigation into the anti-cancer activity of the natural polyphenol curcumin, its therapeutic application is restricted by its inherent physicochemical properties. Synthetic curcumin analogues, however, offer a promising strategy to improve the drug-like potential of curcumin. In this study, we evaluated three curcumin-based benzothiazepane analogues for their ability to selectively target colon cancer cells. Their cytotoxicity was assessed on intestinal cancerous HCT-116 and non-cancerous IPEC-J2 cells using cell viability assays and microscopic imaging. Two analogues, AT007 and AT096, demonstrated enhanced anti-cancer selectivity compared to curcumin. Interestingly, this effect correlated with the aggregation of these compounds in cell medium, which was influenced by compound concentration and medium composition (particularly the presence of albumin). Confocal microscopy confirmed the presence of particles up to 12 µm inside both cell lines, yet downstream metabolic and transcriptomic responses revealed distinct coping mechanisms that may underlie the higher survival of IPEC-J2 cells. Rather than direct molecular interactions typical of soluble compounds, the observed selectivity appears to result from indirect, particle-driven physical effects, potentially involving (intracellular) membrane disruption. Our findings suggest that aggregation behaviour can be a key determinant in improving the potency and selectivity of bioactive compounds, opening new opportunities for the design and screening of more selective anti-cancer therapeutics.
Curcumin, a natural polyphenol with anti‐cancer potential, faces limitations due to metabolic instability, and non‐specific toxicity. This study focused on four curcumin analogs (FM038, FM044, FM045, FM050) featuring oxobutenylidene‐substituted benzothiazepane cores bearing pyridinyl or furan substituents. We systematically evaluated their selective toxic effect on human colon (Caco‐2) and liver (HepG2) cells. This study reveals that structural modifications critically govern the selective cytotoxicity of curcumin analogs towards proliferating Caco‐2 and HepG2 cells. Treatment with analogs featuring pyridin‐2‐yl substituents (FM045/FM050) resulted in a slightly reduced potency towards Caco‐2 cells compared to curcumin, whereas FM044 (furan‐2‐yl) was non‐toxic and FM038 (pyridin‐3‐yl, methylated core) maintained similar activity. In contrast, direct exposure to proliferating HepG2 cells resulted in less cytotoxicity compared to the Caco‐2 cell line, which points to a cell‐specific mode‐of‐action in growing cells. When combined in a co‐culture, metabolites generated and/or transported by differentiated Caco‐2 cells significantly decreased HepG2 viability at an apical concentration of 15 μM. Efflux analysis of respiration and acidification showed that FM050 induced the strongest mitochondrial dysfunction in Caco‐2 cells, directly driving cytotoxicity, whereas HepG2 toxicity correlated with intrinsic detoxification capacity rather than mitochondrial disruption. A metabolic shift towards glycolysis was observed in both cell lines, indicating energy pathway modulation to a more dysfunctional phenotype. These findings highlight the dual role of structural optimization (e.g., methylation, heterocyclic substituents) for (cancer) cell‐specific bioactivity and colonic metabolism and may guide future development of curcumin‐based chemopreventive agents after in vivo validation and metabolite characterization.