Interrupting this cycle would slow, or even stop, the spread
For the more than one million Americans living with Parkinson's disease, medicine has long offered only accommodation — ways to soften the tremors and rigidity while the disease itself marches on unchecked. Now, researchers at the University of Pennsylvania have identified a protein called GPNMB that appears to act as an accelerant in the brain's own unraveling, and have shown that blocking it with engineered antibodies can interrupt the spread of the misfolded proteins at Parkinson's core. Validated against tissue from nearly 1,700 human brains, this discovery offers something the field has not had before: a credible biological target and a means to reach it.
- Parkinson's disease currently has no treatment that slows its progression — only symptom management — leaving over a million Americans on an inevitable decline with no way to interrupt it.
- A newly identified protein, GPNMB, creates a self-reinforcing loop in which brain immune cells release it in response to damage, and it in turn accelerates the spread of toxic alpha-synuclein clumps to healthy neurons.
- UPenn scientists engineered monoclonal antibodies that block GPNMB, successfully halting cell-to-cell alpha-synuclein transmission in laboratory neuron cultures.
- Analysis of 1,675 donated human brains confirmed the mechanism is clinically relevant — individuals with gene variants driving higher GPNMB production showed significantly more widespread Parkinson's pathology.
- The discovery is specific to Parkinson's rather than neurodegeneration broadly, sharpening its therapeutic promise, though clinical trials are still needed before any patient benefit can be realized.
Parkinson's disease moves through the brain like a slow chain reaction, neurons falling one after another as clumps of misfolded alpha-synuclein spread from cell to cell. A person may begin with only a tremor, but over years the pathology claims walking, swallowing, speech. More than a million Americans live with this progression, and roughly 90,000 more are diagnosed each year. Despite decades of effort, medicine can ease the symptoms — levodopa for rigidity, deep-brain stimulation for tremors — but nothing stops the disease itself.
Researchers at the University of Pennsylvania believe they may have found a way to break the chain. Publishing in Neuron, lead researcher Alice Chen-Plotkin and her team identified a protein called GPNMB that functions as an accelerant. When neurons are damaged, the brain's immune cells — microglia — produce more of it. Enzymes release it into the space between cells, where it facilitates the transfer of alpha-synuclein from one neuron to the next. Damage produces GPNMB; GPNMB spreads damage; more damage produces more GPNMB. The cycle is self-reinforcing.
The team developed monoclonal antibodies designed to block GPNMB, and in cultured neuron experiments, those antibodies successfully prevented alpha-synuclein from spreading between cells. To test whether the mechanism holds in human disease, they examined brain tissue from 1,675 individuals in the Penn Brain Bank, looking for genetic variants associated with higher GPNMB production. People carrying those variants showed more extensive alpha-synuclein pathology — and the correlation was specific to Parkinson's, not other neurodegenerative conditions like Alzheimer's.
The road to clinical use remains long. Safety must be established, dosing refined, and trials designed and conducted. But for a disease that has never had a treatment capable of slowing its course, the identification of a specific, human-validated target — and a tool already capable of hitting it — marks a genuine turning point. The question that remains is whether what works in the laboratory will hold in living patients, and whether it will arrive in time to matter.
Parkinson's disease spreads through the brain like a chain reaction, one neuron infecting the next with clumps of a misfolded protein called alpha-synuclein. A person diagnosed early might have only a tremor or stiffness, but as the pathology moves through different brain regions over years, the disease steals walking, swallowing, speech—the basic machinery of living. More than a million Americans have Parkinson's, with roughly 90,000 new diagnoses each year. Yet despite decades of research, doctors have nothing that slows this progression. They can manage the symptoms—levodopa for rigidity, deep-brain stimulation for tremors—but they cannot stop the disease itself.
Researchers at the University of Pennsylvania may have found a way to interrupt that chain. In a study published in Neuron, they identified a protein called GPNMB that acts as a kind of accelerant in Parkinson's pathology. When neurons are damaged or dying, the brain's resident immune cells, called microglia, produce more of this protein. Enzymes then release it from the cell surface, allowing it to move freely between neurons and facilitate the spread of alpha-synuclein from one cell to the next. It is, in essence, a self-reinforcing cycle: damage triggers GPNMB release, which spreads more damage, which triggers more GPNMB.
Lead researcher Alice Chen-Plotkin and her team developed monoclonal antibodies—engineered immune proteins—that block GPNMB. In laboratory experiments using cultured neurons, these antibodies prevented alpha-synuclein from spreading cell to cell. The finding is significant because it suggests a way to interrupt the cycle entirely. "If we can block GPNMB, we might slow or even stop the spread of alpha-synuclein through the brain," Chen-Plotkin said, "and the neurodegeneration that follows."
But laboratory results are one thing. To test whether this mechanism actually matters in human disease, the team analyzed brain tissue from 1,675 individuals in the Penn Brain Bank. They looked for genetic variants that increase GPNMB production. What they found was striking: people with those variants showed more extensive alpha-synuclein pathology throughout their brains. The correlation was specific to Parkinson's—elevated GPNMB was not associated with other neurodegenerative diseases like Alzheimer's. This human evidence validates the therapeutic target and suggests that blocking GPNMB could have real clinical relevance.
The path from bench to bedside remains long. Chen-Plotkin acknowledged that much work lies ahead before these antibodies can be tested in patients. Clinical trials would need to be designed and conducted, safety profiles established, dosing optimized. But for a disease that currently offers no way to slow its relentless progression, the discovery of a potential mechanism and a way to interrupt it represents a genuine shift. For the first time, researchers have a credible target and a tool to hit it. The question now is whether that tool will work in living brains, in real patients, in time to make a difference.
Notable Quotes
Many patients with Parkinson's disease are diagnosed in the early stages, when symptoms are relatively mild, but there is currently no treatment that slows the progression.— Alice Chen-Plotkin, MD, lead author
These results are promising for laboratory models and human brain tissue analysis, but we still have a lot of work to do before we can translate this therapy into humans.— Alice Chen-Plotkin, MD
The Hearth Conversation Another angle on the story
So GPNMB is produced by immune cells in the brain. Why would the immune system be making a protein that spreads disease?
It's not intentional harm. Microglia are trying to clean up—they're responding to damaged neurons. But in doing that, they produce GPNMB, which actually accelerates the very pathology they're trying to contain. It's a well-meaning response that backfires.
And the antibodies just block it. They prevent GPNMB from doing its job.
Yes. In the lab, when you block GPNMB, alpha-synuclein can't spread between cells as easily. The chain reaction slows or stops.
Why hasn't this been found before? GPNMB must have been studied.
It has, but mostly in cancer research. Chen-Plotkin's team made the connection to Parkinson's relatively recently, in 2022. Once they identified it as central to the spread mechanism, blocking it became an obvious strategy.
The brain tissue analysis—1,675 brains—that's the part that makes this real, isn't it?
Exactly. Lab work is promising, but it doesn't prove the mechanism matters in actual human disease. Finding that genetic variants linked to higher GPNMB production correlate with worse pathology in real brains—that's the evidence that this isn't just a laboratory curiosity.
What happens next?
Clinical trials. They need to test these antibodies in actual patients with early Parkinson's and see if blocking GPNMB slows progression. That's years away, but it's the logical next step.