Discovery of 6‐Fluoro-5-{4-[(5-fluoro-2-methyl-3-oxo-3,4- dihydroquinoxalin-6 yl)methyl]piperazin-1-yl}‐N‐methylpyridine-2- carboxamide (AZD9574): A CNS-Penetrant, PARP1-Selective Inhibitor
J. Med. Chem., 2024, 67, 21717–21728
DNA damage, such as double-strand breaks (DSB), single strand nicks (SSB), gaps and more occur naturally within cell nuclei and can also be induced by environmental factors such as radiation. Key defences against these cytotoxic or oncogenic genetic abnormalities that can otherwise build up are found in the numerous and complex pathways for DNA Damage Repair (DDR) which are part of our cellular machinery. One part of this involves the poly-(ADP-ribose)polymerase (PARP) family of enzymes, a function of which is to detect DNA damage such as SSBs and initiate the appropriate DDR response. The Breast Cancer (BRCA-2) gene is involved in the DDR response, and is known as a ”tumour-suppressor” gene. Mutations to BRCA-2 significantly increase the risk of developing breast cancer, however, PARP-1 inhibitors can be particularly effective against BRCA-2 mutant tumours based on data from numerous PARP inhibitors approved for use in the clinic.
Researchers at AstraZeneca (AZ) identified a gap in the portfolio of approved PARP inhibitors: that of PARP-1-selective, CNS penetrant inhibitors. The PARP selectivity was required in order to limit side effects and the CNS penetration was required to be able to treat brain metastases. As a stringent filter for compounds not expected to show CNS exposure, the team adopted a permeability assay with MDCK cells engineered to express both the MDR-1 and BCRP transporters, which play a significant role in keeping xenobiotics out of the CNS. The compounds representing the state of the art in terms of clinical PARP-1 inhibition displayed high efflux ratios (ERs) and negligible CNS exposure, apart from veliparib.
Having access to crystal structures of relevant inhibitors bound to PARP-1 proved to be useful in design decisions. Scaffold hopping and thorough R group exploration enabled the discovery of AZD9574, which showed: decent CNS exposure, no hERG liability and a remarkable >10,000-fold selectivity margin over PARP-2.
The final result of selectivity vs PARP-2 was the result of exploration of the 8-position of the quinoxazinone core. This was likely met with apprehension at first, due to the tentative R group contacting the PARP-1 surface in structural biology data. Sometimes the compounds just need to be made, “fortune favours the bold“ and other clichés come to mind!