Identification and characterization of two novel, neuroprotective PARP-1 inhibitors

Amos Fatokun1, Valina Dawson2, Ted Dawson3. 1Johns Hopkins University School of Medicine, Neuroregeneration and Stem Cell Programs, Institute for Cell Engineering and the Department of Neurology, MD 21205, Baltimore, Maryland, United States, 2Johns Hopkins University School of Medicine, Neuroregeneration and Stem Cell Programs, Institute for Cell Engineering and the Departments of Neurology, Neuroscience and Physiology, MD 21205, Baltimore, Maryland, United States, 3Johns Hopkins University School of Medicine, Neuroregeneration and Stem Cell Programs, Institute for Cell Engineering and the Departments of Neurology and Neuroscience, MD 21205, Baltimore, Maryland, United States.

Excessive activation of the nuclear enzyme poly (ADP-ribose) polymerase-1 (PARP-1) leads to the formation and accumulation of a prodigious amount of poly (ADP-ribose) polymer (PAR). As a consequence, PAR signals to the mitochondrially-localized apoptosis-inducing factor (AIF), causing its translocation from the mitochondria to the nucleus, a process that engenders large-scale DNA fragmentation and chromatin condensation, and, eventually, cell death. This type of cell death, now referred to as “parthanatos,” is caspase-independent and has been implicated in a wide range of pathological conditions, including neurological and neurodegenerative diseases. The pathway therefore represents a very attractive target for neurotherapeutics development. We have thus employed high-throughput screening (HTS) to identify novel, small-molecule compounds that can potently block parthanatos. The primary screening assay was developed and optimized in 96-well plates and involved 25-min exposure of HeLa cell cultures to N-methyl-N’-nitro-N-nitrosoguanidine (MNNG) (50 µM), a DNA alkylating agent that induces parthanatos. The PARP-1 inhibitor DPQ (30 µM) and the pan-caspase inhibitor Z-VAD-fmk (100 µM) served as positive and negative controls, respectively. Extensive quantification of Z’ factor values, coefficients of variation and signal-to-background ratios confirmed attainment of sufficient optimization and assay robustness. The fluorescent Alamar Blue (AB) dye (Invitrogen) was used to quantify viability as the endpoint. Small molecules from three carefully selected chemical libraries (BIOMOL, MSSP (Spectrum 2K), Hopkins Drugs Library) were screened at 10 µM (added to the recovery medium only), and those that significantly protected against the reduction in viability induced by MNNG were taken as positive hits, while those that elicited significant toxicity in the primary screen were re-screened at lower concentrations. In order to weed out false positives, initial hits were taken through rounds of secondary screening by employing another viability reagent (CellTiterGlo, Promega) that required a different readout (luminescence) from what was required by AB. In total, about 5,000 small molecules were screened. Following primary and secondary screens, only 4’-methoxyflavone (4-MF) was confirmed as a true hit. 4-MF was found to elicit concentration-dependent protection (at a low micromolar range) against the toxic effect of MNNG and the formation/accumulation of PAR. It also protected significantly (12.5 µM – 50 µM) against the death of cortical neuronal cultures induced by a brief (5 min) exposure to excitotoxic N-methyl-D-aspartate (NMDA, 500 µM), which is an established in vitro model of parthanatos. A small library of structurally-related small molecules was then customized and screened, from which was identified a second PARP-1 inhibitor, 3’,4’-dimethoxyflavone (DMF), which also protected concentration-dependently against the toxic effect of MNNG and the formation/accumulation of PAR, as well as against NMDA-induced cell death of cortical neurons. These compounds are now to be examined for neuroprotection in animal models of neurodegeneration. Taken together, the data suggest that our model HTS assay has specificity for PARP-1-dependent cell-death (parthanatos) pathway, and natural compounds, especially flavones, are a potential source of novel PARP-1 inhibitors for probes or therapeutics development. In addition, methoxylation on the 4′ position of the flavone ring may be obligatory for the PARP-inhibiting activity of flavones.