GC-t8-linked pyrrolobenzodiazepine (PBD)-biaryl conjugates with femptomolar in vitro cytotoxicity and in vivo antitumour activity in mouse models of pancreatic and breast cancer.

As a class, minor-groove non-covalent DNA-binding small molecules generally have A/T rather than G/C selectivity, thought to be due to the narrower minor groove in A/T regions of DNA, and the presence of exocyclic guanine C2-amino groups which project into the minor groove and prevent multiple close van der Waals contacts in GC-rich regions. We have developed a set of biaryl building blocks based on phenyl-substituted heterocycles with significant GC-selectivity and sufficient length to span two DNA base pairs. These have been conjugated to DNA minor-groove covalent-binding pyrrolobenzodiazepine (PBD) molecules via a four-carbon linker to produce C8-linked PBD-MPB hybrid molecules. In particular, the 4-(1-methyl-1H-pyrrol-3-yl)benzenamine (MPB) biaryl motif either alone or conjugated to a PBD molecule has a strong preference for GC-rich sequences as demonstrated by the results of FID, HPLC-MS, FRET-based and DNA footprinting assays. Molecular modeling studies support these observations, suggesting that the high GC-affinity may be due to a combination of overall shape and the formation of key hydrogen bonds. Some PBD-MPB conjugates have sub-picomolar IC50 values in MCF7, A431, A2780, A549, MIA PaCa2 and MDA-MB-231 human tumour cell lines in vitro, while being up to six orders of magnitude less cytotoxic in the non-tumour cell line WI38, suggesting that key DNA sequences may be relevant targets in these ultra-sensitive cancer cell lines. One conjugate, which has femptomolar activity in the breast cancer cell line MDA-MB-231 (IC50 = 0.065 picomolar), has significant dose-dependent antitumour activity in MDA-MB-231 (breast) and MIA PaCa2 (pancreatic) human tumour xenografts in nude mice. It is well tolerated at concentrations up to 350 μg/kg, with no signs of toxicity at this dose level. Preliminary results based on cell culture and Western blotting experiments, and on histology studies on xenograft biopsies, have led to a suggested mechanism of action involving selective transcription factor inhibition, which is supported by molecular modeling studies.