Microglia can turn astrocytes toxic

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

Shane Liddelow, Ben Barres, Beth Stevens, Valina Dawson, Ted Dawson and colleagues (Nature, 2017); Seung Pil Yun, Han Seok Ko, Seulki Lee, the Dawsons, Liddelow and Barres, among 30+ authors (Nature Medicine, 2018).

What happened

Astrocytes are star-shaped support cells. Liddelow's team found that activated microglia release three signals (IL-1α, TNF and C1q) which together switch astrocytes into a harmful "A1" state. A1 astrocytes stopped supporting neurons and instead killed them, including about a quarter of dopamine neurons in dish experiments. Staining of human brain tissue found many A1-marked astrocytes in several neurodegenerative diseases, Parkinson's included. Nature 541:481-487, 18 Jan 2017. In 2018, Yun et al. tested NLY01, a brain-penetrant, long-acting GLP-1-receptor drug, in a synuclein-fibril mouse model and a transgenic synuclein mouse. It protected dopamine neurons, reduced behavioural deficits and extended survival, apparently by stopping microglia from creating A1 astrocytes. Nat Med 24:931-938.

Why it moved us toward a cure

It offered a concrete chain from immune activation to neuron death (landmark 52) and a drug-able point to break it, helping justify GLP-1-type drugs (landmark 20). The human test of that drug is in landmark 143.

Outcome

Open. The mouse result did not carry over: NLY01 missed its main goal in a 255-person Phase 2 in 2023 (landmark 143). Separately, a 2021 expert consensus (Escartin et al., Nat Neurosci) argued the A1/A2 two-box labelling is too simple. Reactive astrocytes matter, but the "A1" label is now used cautiously.

Sources

• Liddelow et al., Nature 2017;541:481-487, DOI 10.1038/nature21029 (Alzforum summary with expert caveats)
• Yun et al., Nat Med 2018;24:931-938, DOI 10.1038/s41591-018-0051-5
• Escartin et al., Nat Neurosci 2021;24:312-325
• Not yet verified: the number of Parkinson's samples stained in Liddelow 2017; whether those samples were postmortem.