Too little synuclein harms neurons

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

Oleg Gorbatyuk, Nicholas Muzyczka, Fredric Manfredsson and colleagues, University of Florida (Mol Ther 18:1450–1457, 2010; NIH grants listed). Timothy Collier, Fredric Manfredsson and colleagues (Front Neurosci 10:12, 2016; Benskey et al., Front Mol Neurosci 11:36, 2018, funded by the NIH and the Michael J. Fox Foundation). Zharikov, Burton and colleagues, with Greenamyre's group, University of Pittsburgh (J Clin Invest 125:2721–2735, 2015; NIH and VA grants listed).

What happened

Extra copies of the synuclein gene cause Parkinson's (landmark 14), and mice with no synuclein gene live normally (landmark 201). That suggested lowering it should be safe.

• 2010. Gorbatyuk et al. delivered small interfering RNAs against synuclein by an AAV vector into the rat substantia nigra. Within 4 weeks, tyrosine hydroxylase-positive (dopamine) cells and striatal dopamine fell on the injected side, with an amphetamine-induced turning bias. The loss was real cell death, not just a lower dopamine-making signal. Three control RNAs did not do this. Loss tracked how well each RNA cut synuclein, and adding back rat synuclein partly rescued it. The authors concluded both too much and too little synuclein raise the risk.
• 2015. Zharikov et al. cut synuclein by about 35% in rats. There was no degeneration, and the knockdown protected nigral neurons against rotenone.
• 2016. Collier et al. is a perspective article. It argues for a loss-of-function component and includes a pilot study in which knocking down synuclein in monkey dopamine neurons reproduced Parkinson-like nigrostriatal loss. Its size is not given in the abstract.
• 2018. Benskey et al. found that in rat nigra, silencing synuclein raised MHC-1 (an immune-tag molecule) on neurons and recruited microglia and T-cells before cell death, followed by about 50% loss of nigral neurons. Silencing synuclein in cerebellar neurons did not do this.

Why it moved us toward a cure

It bounds the synuclein-lowering strategy. The safety case from knockout mice has limits, because acute loss in adult neurons differs from lifelong loss in embryos. It argues for care over the degree of lowering in human trials; see the lowering drugs (landmark 29) and gene editing (landmark 117).

Outcome

Open. The rat and monkey findings are not human data, and they depended on dose and vector. The first human data on the dosing question will come from the antisense and siRNA trials. All the toxicity reports come from one extended group of labs using AAV-delivered RNA; I did not find an independent replication. No correction or retraction was found for any of the four papers.

Sources

• Gorbatyuk et al., Mol Ther 2010, DOI 10.1038/mt.2010.115
• Zharikov et al., J Clin Invest 2015, DOI 10.1172/JCI64502
• Collier et al., Front Neurosci 2016, DOI 10.3389/fnins.2016.00012
• Benskey et al., Front Mol Neurosci 2018, DOI 10.3389/fnmol.2018.00036
• Not yet verified: the monkey pilot's size and result details (abstract gives none); whether other groups saw no toxicity with similar knockdown (not searched); PubPeer not checked.