ATP10B loss-of-function raises risk

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

Shaun Martin (first author), Peter Vangheluwe (KU Leuven) and Christine Van Broeckhoven (VIB and University of Antwerp, last author), with Veerle Baekelandt, Peter De Deyn, Patrick Cras and the Belgian BELNEU consortium (Acta Neuropathologica 139:1001-1024, online 14 March 2020). I did not read the funding statement or confirm affiliations.

What happened

The team sequenced the exomes (protein-coding DNA) of 52 early-onset patients and found three with compound heterozygous ATP10B variants (a different damaging change on each copy of the gene). Screening a Belgian cohort added four more: 6 of 617 Parkinson's patients (0.97%) and 1 of 226 with dementia with Lewy bodies (0.44%).

ATP10B codes for a flippase, an enzyme in the lysosome membrane that moves fats across it. It moves glucosylceramide (GluCer) and phosphatidylcholine. The patient variants were inactive. Cells lacking the enzyme were not protected against rotenone and manganese, and cortical neurons lost lysosome function and died.

Replication was mixed:

• The International Parkinson's Disease Genomics Consortium (Real et al., Acta Neuropathol 2020;140:401) checked whole-genome data from 1,647 patients and 1,050 controls, and imputed genome-wide data from 14,671 cases and 17,667 controls. They found no enrichment of putative compound heterozygotes (1.9% of patients versus 2.0% of controls) and no association of rare or common variants, and noted that two variants called harmful in the cell work were homozygous in a healthy 52-year-old control. They concluded that evidence for ATP10B as a disease gene is limited. The Martin group replied (Smolders and Van Broeckhoven, 2020).
• A French group (Tesson et al., 2020) wrote on segregation of ATP10B variants in recessive families, and a Chinese mainland group (Zhao et al., 2021) reported a large cohort; each drew a reply from Martin's group. I read titles only.
• A Japanese screen of 245 young-onset and 33 recessive patients (2021) found only single-copy variants with little sign of harm.
• A southern Spanish study (1,162 patients, 2024) saw no link of rare variants to risk.
• In 2025, rats with Atp10b knocked down in the nigra developed parkinsonian movement deficits and lost dopamine neurons over a year.

Why it moved us toward a cure

If real, it adds a second gene to the GluCer story. GBA (landmark 16) breaks GluCer down; ATP10B moves it out of the lysosome. Lysosomal storage genes as a group were already enriched in patients (landmark 233).

Outcome

Open. The genetic evidence is thin: the largest checks (IPDGC, Spain, Japan) did not confirm it, while the biology is supported by cell and rat work. No drug targets it. The original paper has no correction or retraction; six comment-and-reply items are attached to it (Europe PMC).

Sources

• Martin et al., Acta Neuropathol 2020;139:1001–1024, DOI 10.1007/s00401-020-02145-7
• Ishiguro et al., Parkinsonism Relat Disord 2021, DOI 10.1016/j.parkreldis.2021.05.020 · Díaz-Belloso et al., 2024, DOI 10.1016/j.parkreldis.2024.106989 · Sanchiz-Calvo et al., Acta Neuropathol 2025, DOI 10.1007/s00401-025-02908-0
• Correspondence: Real et al. 2020, DOI 10.1007/s00401-020-02172-4, PMC7540943 · Tesson et al. 2020, DOI 10.1007/s00401-020-02219-6 · Zhao et al. 2021, DOI 10.1007/s00401-021-02280-9
• Not yet verified: the findings of the Tesson and Zhao letters and the three replies (no abstracts or open text found); funding; whether any large analysis has confirmed ATP10B at genome-wide level (the IPDGC check did not).