Synuclein fibrils kill via PARP1

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Tae-In Kam, Ted Dawson, Valina Dawson (Johns Hopkins) and colleagues. Science 362(6414):eaat8407, 2 Nov 2018. Funding: US NIH/NINDS grants, the JPB Foundation and the Adrienne Helis Malvin Medical Research Foundation. Five authors, including both Dawsons, have filed a US patent on detecting PAR in spinal fluid of Parkinson's patients (competing-interest statement).

What happened

Misfolded synuclein spreads between neurons (landmark 51) and can be made in the lab as "pre-formed fibrils", used in standard models (landmark 43). How the fibrils then kill a neuron was unclear.

PARP1 is a repair enzyme that attaches chains of poly(ADP-ribose), or PAR, to proteins. Overactive PARP1 triggers parthanatos, a cell death in which PAR chains damage mitochondria. The team found that fibrils activated PARP1 in cultured mouse neurons, and that PAR then sped up the formation of toxic synuclein and made a fibril strain about 25 times more toxic in culture (a supplementary figure). In mice given one striatal injection of 5 µg fibrils, about half the nigral dopamine neurons were lost at 6 months. That loss, and the dopamine and movement deficits, did not occur in PARP1-knockout mice or in wild-type mice fed a diet containing the PARP inhibitor veliparib (ABT-888, 125 mg/kg); stereology used 5 to 7 mice per group. PAR was also higher in spinal fluid from Parkinson's patients than controls in two cohorts (31 controls and 80 patients; 33 and 21) and in nigral tissue.

Why it moved us toward a cure

It connected the seeding models to a specific, druggable cell-death step. PARP inhibitors are already approved for cancer, so the idea could in principle be tested quickly. The paper proposes inhibiting PARP1 as a possible disease-modifying strategy.

Outcome

Open. It is mouse and cell data, and PAR in spinal fluid was based on two small cohorts. Follow-up work has used PAR plus synuclein to build a faster monkey model (Liu et al., Cell Rep Methods 2024), but no independent test of PARP inhibitors for slowing the disease was found. A ClinicalTrials.gov search on 20 Oct 2026 found no Parkinson's trial of a PARP inhibitor. Long-term PARP inhibition has side effects in cancer patients, which are not assessed here. No correction or retraction was found; four commentaries accompanied the paper.

Sources

• Kam et al., "Poly(ADP-ribose) drives pathologic α-synuclein neurodegeneration in Parkinson's disease," Science 2018;362:eaat8407, DOI 10.1126/science.aat8407, PMID 30385548, PMC6431793 (full text read)
• Liu et al., Cell Rep Methods 2024;4:100876, DOI 10.1016/j.crmeth.2024.100876 (title only)
• Not yet verified: whether anyone has replicated the PARP1 result independently (one search; the 2024 monkey-model paper was seen by title only); the sample sizes for nigral PAR.