A type I combi-targeting approach for the design of molecules with enhanced potency against BRCA1/2 mutant- and O6-methylguanine-DNA methyltransferase (mgmt)- expressing tumour cells.

BACKGROUND: Mutations of the DNA repair proteins BRCA1/2 are synthetically lethal with the DNA repair enzyme poly(ADP-ribose) polymerase (PARP), which when inhibited, leads to cell death due to the absence of compensatory DNA repair mechanism. The potency of PARP inhibitors has now been clinically proven. However, disappointingly, acquired resistance mediated by the reactivation of wild type BRCA1/2 has been reported. In order to improve their efficacy, trials are ongoing to explore their combinations with temozolomide (TMZ). Here, in order to enhance potency in BRCA1/2-mutant cells, we report on the design of single molecules termed "combi-molecules" capable of not only inhibiting PARP but also damaging DNA like TMZ, which is known to induce a large number of DNA adducts. The majority of these lesions are processed through PARP-dependent base-excision repair machinery. Paradoxically, the least abundant lesion, the O6-methylguanine adduct is the most cytotoxic. Its repair by the O6-methylguanine DNA methyl transferase (MGMT) confers robust resistance to TMZ. Thus, we surmise that a combi-molecule designed to generate the same DNA adducts as TMZ, with an additional ability to block PARP, could induce BRCA1/2 mutant selective potency and a growth inhibitory profile independent of MGMT status. METHODS: The hydrolysis of EG22 and its stabilized form ZSM02 was analyzed by HPLC and fluorescence spectroscopy. Growth inhibitory potency was determined by SRB assay. PARP inhibition was determined by an enzyme assay and DNA damage by the comet assay. Subcellular distribution was visualized by confocal microscopy. RESULTS: Studies on EG22 showed that: (a) it inflicted anomalously higher levels of DNA damage than TMZ (b) it induced PARP inhibitory potency in the same range as ANI, a known PARP inhibitor (IC50 = 0.10 µM) (c) it showed strong potency in both BRCA1/2 wild type and mutated cells with 6-fold selectivity for the mutants and it was 65-303-fold more potent than TMZ and 4-63-fold than ANI alone and 3-47-fold than their corresponding equimolar combinations and (d) its potency was independent of MGMT expression. CONCLUSION: The results in toto suggest that a combi-molecular approach directed at blocking PARP and damaging DNA can lead to single molecules with selective and enhanced potency against BRCA1/2 mutant and with activity independent of MGMT, the major predictive biomarker for resistance to TMZ.

15N-, 13C- and ¹H-NMR Spectroscopy Characterization and Growth Inhibitory Potency of a Combi-Molecule Synthesized by Acetylation of an Unstable Monoalkyltriazene.

6-(3-Methyltriaz-1-en-1-yl)-1H-benzo[de]isoquinoline-1,3(2H)-dione referred to as EG22 (8a), is an open-chain 3-alkyl-1,2,3-triazene termed "combi-molecule" designed to inhibit poly(adenosine diphosphate ribose) polymerase (PARP) and damage DNA. To delay its hydrolysis, acetylation of N3 was required. Being a monoalkyl-1,2,3-triazene, EG22 could assume two tautomers in solution or lose nitrogen during the reaction, thereby leading to several acetylated compounds. Instead, one compound was observed and to unequivocally assign its structure, we introduced isotopically labeled reagents in its preparation, with the purpose of incorporating 15N at N2 and 13C in the 3-methyl group. The results showed that the 1,2,3-triazene moiety remained intact, as confirmed by 15N-NMR, coupling patterns between the 15N-labeled N2 and the 13C-labeled methyl group. Furthermore, we undertook heteronuclear single quantum coherence (HSQC) and heteronuclear multiple bond correlation (HMBC) experiments that permitted the detection and assignment of all four nitrogens in 6-(3-acetyl-3-methyltriaz-1-en-1-yl)-1H-benzo[de]isoquinoline-1,3(2H)-dione, referred to as ZSM02 (9a), whose structure was further confirmed by X-ray crystallography. The structure showed a remarkable coplanarity between the N-acetyltriazene and the naphtalimide moiety. Thus, we unequivocally assigned 9a as the product of the reaction and compared its growth inhibitory activity with that of its precursor, EG22. ZSM02 exhibited identical growth inhibitory profile as EG22, suggesting that it may be a prodrug of EG22.

Design, synthesis and biological activity of novel molecules designed to target PARP and DNA.

In order to enhance the cytotoxic potential of poly(ADP-ribose) polymerase (PARP) inhibitors in BRCA1 or 2 deficient tumours, we designed a series of molecules containing a 1,2,3-triazene moiety tethered to a PARP targeting scaffold. A cell-based selectivity assay involving a BRCA2-deficient Chinese hamster cell line and its corresponding BRCA2 wild type transfectant, was used to predict the PARP targeting potential of the latter agents. The results showed that adding a DNA damaging function to the PARP inhibitors decreased but did not abrogate the selective targeting of the BRCA2-deficient cells. The DNA damaging moiety augmented the potency in BRCA2 deficient cells by 2-20 fold. The most selective dual PARP-DNA targeting agent 14b was found to possess dual DNA and PARP targeting properties.

Positional isomerization of a non-cleavable combi-molecule dramatically altered tumor cell response profile.

To potentiate the quinazoline-based inhibitor of the epidermal growth factor receptor (EGFR), a chloroethyl alkylating moiety was appended to its 6-position. This led to molecules with extremely strong EGFR inhibitory potency and anomalously strong DNA-damaging potential. To assess the role of the chloroethyl group on potency, we designed a molecule in which it is shifted to the 7-position where it would be less reactive and away from the cys773 of the EGFR ATP site. The results showed that (i) ZR2009 was 10-fold less potent than its positional isomer ZR2003 in EGFR tyrosine kinase inhibition, (ii) it consistently exhibited significantly weaker antiproliferative potency than ZR2003, (iii) in reversibility assays, while ZR2003 induced sustained inhibition of EGFR phosphorylation, ZR2009 inhibitory activity was partially reversed, and (iv) likewise, ZR2009 significantly lost its activity in short exposure growth inhibitory assays and induced lower levels of DNA damage than ZR2003. Molecular modeling suggested that while the chloroethylamino group in ZR2003 was at 3.5 Å away from Cys773, that of ZR2009 was at 6.3 Å. The results in toto suggest that, while the chloroethyl is a strong alkylating group, its appendage to the 6-position is optimal for DNA damage, sustained EGFR, and growth inhibition.