FOUNDIN-PD dopamine-neuron atlas

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

Elisangela Bressan (first author), Cornelis Blauwendraat, Andrew Singleton, Mark Cookson, Peter Heutink and others, on behalf of the FOUNDIN-PD consortium. Cell Genomics 3(3):100261, 2023; preprint on bioRxiv, June 2021. The paper states that it is supported by the Michael J. Fox Foundation (as part of its PD Pathogenesis consortium). The cell lines came from the Golub Capital iPSC substudy of PPMI (landmark 24), itself funded by MJFF and corporate sponsors. The paper does not name ASAP.

What happened

GWAS gave 90 risk signals (landmark 19), but most lie outside genes, so the gene each one acts on is unclear. The team differentiated 95 PPMI stem cell lines (landmark 47) into dopamine neurons with one protocol on a robotic system. Donors included people with LRRK2 G2019S (25) and R1441G (1), GBA1 N370S (20) and SNCA A53T (4), people with idiopathic Parkinson's, healthy controls, and people with high or low polygenic risk. Donor whole genomes were available.

Layers: bulk RNA-seq at day 0, 25 and 65 of differentiation; single-cell RNA-seq on 83 lines at day 65 (416,216 cells); chromatin and methylation data; and time-lapse imaging of neuron survival. Efficiency varied a lot between lines. Of the 90 GWAS signals, multiple variants showed expression-linked (eQTL) effects in at least one cell type, for example a locus whose lead variant rs11950533 was linked to CAMLG expression. In the imaging, neurons from GBA1 lines had a higher risk of dying than healthy-control neurons (the paper's Figure 5; sample sizes are in the paper).

Why it moved us toward a cure

It is the first large public dataset that ties Parkinson's risk variants to gene activity in the cell type that dies, from people already followed in PPMI. That can help rank which of the GWAS genes to target. Others have since reused it: for example a 2024 study of cholesterol-degrading enzymes used FOUNDIN-PD to test whether risk variants alter regulation, and a 2026 reanalysis of its single-cell data found shared changes in neuron-development and synapse pathways (both secondary reuse, not clinical).

Outcome

Advanced as a resource. It does not show causation or a drug target by itself, the neurons are immature and cultured, and line-to-line variation is large. It is not a standalone landmark for any therapy.

Sources

• Bressan et al., Cell Genom 2023;3(3):100261, DOI 10.1016/j.xgen.2023.100261, PMID 36950378 (full text via Europe PMC, PMC10025424)
• bioRxiv preprint, 2021, DOI 10.1101/2021.06.03.446785
• Petkova-Kirova et al., Front Pharmacol 2024, DOI 10.3389/fphar.2024.1477009 · Macchi et al., Mov Disord 2026, DOI 10.1002/mds.70505
• Not yet verified: senior and corresponding authors; total cost and ASAP involvement (not in the paper); how many of the 90 GWAS signals had a significant eQTL (the paper's Table S7 was not read); the size of the GBA1 survival effect; the Frontiers paper describes 89 lines against the 95 in the Cell Genomics paper (I did not resolve the difference).