Hugo Bellen’s lab watches fruit flies die in slow motion. Not all of them, just the ones carrying two particular genetic flaws. These flies develop a tremor, lose their balance, struggle to move. Their neurons fire less reliably. Their vision fades. Watch them closely enough across their brief lifespans and you’re witnessing a molecular murder mystery: how does Parkinson’s disease choose its victims?
The question has nagged at researchers for years. We’ve known that people carrying one mutant copy of the GBA1 gene face a five-fold increased risk of developing Parkinson’s. Yet most never develop the disease. Something else has to tip the scales. Bellen, a neurogeneticist at Baylor College of Medicine, suspected that “something else” might be hiding in the genome’s recycling system—the lysosomes that break down and recycle cellular waste.
His team screened dozens of lysosomal genes in flies, looking for ones that interact lethally with Gba1b, the fly version of GBA1. They found their answer in anne, the fly equivalent of the human gene ATP13A2. Flies missing one copy of either gene muddle through life relatively unscathed. But flies missing one copy of both develop a devastating neurodegeneration. “A second factor must be in place for the condition to arise,” Bellen says.
The destruction unfolds across 30 days of fly time with an almost choreographed precision. First author Mingxue Gu tracked the damage day by day. Nothing obvious happens initially. Then, around day 15, glial cells—the brain’s support crew—begin to swell and detach from the neurons they’re meant to protect. By day 30, neurons are failing outright, their electrical signals stuttering into silence. The flies can barely walk. Their vision dims. They’re living out a compressed version of human Parkinson’s.
What surprised Gu most was where the damage started. “The earliest signs didn’t appear in neurons but in glial cells,” she says. That runs counter to how we typically think about neurodegenerative disease. But it makes sense once you understand what anne and Gba1b actually do, and where they work.
Gba1b operates primarily in glia, where it helps break down a fatty molecule called glucosylceramide, or GlcCer. Anne works mainly in neurons, where its job is maintaining the proper acidity inside lysosomes—the cellular compartments where GlcCer gets degraded. When both systems are partially compromised, neurons lose their ability to keep their lysosomes acidic. Without that acidity, GlcCer production goes haywire. The neurons start churning out excess GlcCer, which they then export to neighboring glial cells.
Now you have a logistics nightmare. Glial cells already running on reduced Gba1b capacity are suddenly flooded with more GlcCer than they can process. The molecule accumulates in glial lysosomes like garbage piling up at an understaffed recycling center. The glia swell, their internal architecture warps, and they begin pulling away from the neurons they’re meant to nourish and protect. Without glial support, neurons eventually collapse, particularly those controlling movement and vision.
The researchers demonstrated this neuron-to-glia transfer isn’t speculation. They used lipidomics to track GlcCer levels in different cell types and found exactly the predicted pattern: elevated in both neurons and glia in double-mutant flies, with the worst accumulation in glia. They also showed the damage is activity-dependent. When they reduced neuronal firing, the degeneration slowed. More neural activity means more GlcCer production, which overwhelms the glia faster.
Bellen’s team then tested whether they could intervene. Three compounds reduced the damage: ML-SA1, which improves lysosomal trafficking; myriocin, which blocks GlcCer synthesis; and DFMO, which inhibits polyamine production (polyamines being molecules that seem to exacerbate lysosomal dysfunction). Each treatment restored some degree of normal function to the dying flies.
These aren’t ready-made therapies for human Parkinson’s, but they point to specific biological pathways that might be druggable. More immediately, they validate the two-hit model. If Parkinson’s penetrance requires both GBA1 mutation and some second genetic modifier, we should find people carrying both.
Bellen’s collaborators surveyed genetic data from multiple Parkinson’s cohorts. They found what they were looking for: multiple patients carrying pathogenic variants in both GBA1 and ATP13A2. The cases are rare, which makes sense—you need to win an unlucky genetic lottery twice. But they exist. And their existence suggests the fly model isn’t just mechanistically interesting; it may explain real human disease.
The digenic mechanism also offers a partial answer to a long-standing puzzle. If GBA1 mutations are common and confer substantial risk, why don’t more carriers develop Parkinson’s? Perhaps because most don’t carry the second hit. Their glial cells can still clear GlcCer adequately. Their neurons maintain lysosomal pH well enough. The system limps along. It’s only when both systems are compromised that the recycling infrastructure collapses and neurodegeneration cascades.
This raises uncomfortable questions about genetic risk assessment. Should we screen Parkinson’s patients with GBA1 mutations for variants in ATP13A2 or other lysosomal genes? Would that information be actionable? Perhaps not yet, given we lack targeted therapies. But it might be soon, especially if researchers can develop drugs targeting lysosomal acidification or sphingolipid metabolism.
It also suggests Parkinson’s genetics may be more complex than current models acknowledge. Most genetic studies focus on single-gene effects. But if disease penetrance depends on combinations of variants, we’ve been looking at the problem through too narrow a lens. The genome isn’t a collection of independent risk factors but an interacting network where vulnerabilities compound in unpredictable ways.
Bellen remains cautious about overselling the findings. “These treatments don’t point to an immediate cure,” he says. But they reveal what he calls “promising biological pathways” that could be explored. For families carrying GBA1 mutations and watching for tremors, watching for stiffness, watching for the first stumble, that’s not nothing. It’s a map of where the disease might be vulnerable, drawn from watching flies die slowly under the microscope, one mutation at a time.
Study link: https://link.springer.com/article/10.1186/s13024-025-00923-z
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.