Breakthrough AI finds a cause and treatment candidate for Alzheimer’s

The causes of Alzheimer’s disease have long puzzled researchers, especially when most older adults show signs of the illness even without carrying the well-known mutations linked to it. In fact, nearly everyone over 65 develops some level of Alzheimer’s pathology.

But many lack the mutations in genes like APP, PSEN, or MAPT, which are typically found in early-onset or inherited forms of the disease. Even fewer carry risk-related versions of the genes APOE4 or TREM2. So why does this disease appear so often, even in people without the usual risk factors?

A new study led by scientists at the University of California San Diego offers a surprising answer. It reveals that a gene previously thought to be just a marker of Alzheimer’s may actually cause it—by disrupting how brain cells manage gene expression.

This discovery, published in the journal Cell, opens the door to an entirely new way to detect and possibly treat the disease before symptoms begin.

The gene in question is PHGDH, which produces an enzyme involved in creating serine, an essential amino acid. This enzyme helps glial cells—especially astrocytes—support neurons. For years, researchers have known that PHGDH levels rise in Alzheimer’s patients. Its protein and RNA appear in larger amounts in the hippocampus, prefrontal cortex, and other parts of the brain affected by the disease.

High PHGDH levels were linked with more severe cognitive symptoms and advanced disease stages, known as Braak stages. This correlation was confirmed across patient groups and brain samples. Even in the bloodstream, PHGDH’s extracellular RNA could be detected, showing up before symptoms in some patients. That made it a promising early detection biomarker.

But this team, led by Professor Sheng Zhong at UC San Diego’s Jacobs School of Engineering, wanted to know more. Could PHGDH actually cause Alzheimer's disease?

To find out, the team ran experiments on mice and on lab-grown human brain organoids. These models did not contain the mutations commonly tied to inherited Alzheimer’s disease. When the scientists increased PHGDH activity in these models, the disease worsened. Reducing PHGDH slowed it down. This clear connection suggested that the gene was not just a marker—it played a direct role.

But the gene’s known function didn’t explain how. Its job in serine production is vital for brain development. Mice without PHGDH have severe brain defects, and in humans, mutations can be fatal during development. That meant removing its enzyme activity couldn’t be the answer.

The puzzle led researchers to think PHGDH might have another, previously unknown function.

At first, the scientists had no clue what that function might be. Their tests targeting the enzyme’s role in metabolism didn’t explain the disease link. But a separate study in Zhong’s lab had shown that Alzheimer’s brains have trouble controlling which genes are switched on and off. This led the team to ask: could PHGDH be involved in that process?

They turned to artificial intelligence to examine the protein’s 3D shape. This analysis revealed something stunning—a small part of the PHGDH protein closely resembled the structure of known transcription factors, which help regulate gene activity. That shape allowed the protein to latch onto DNA and change which genes are expressed in astrocytes.

This moonlighting function—acting like a transcription factor—was completely unknown before. And it helps explain how PHGDH could lead to Alzheimer’s without affecting its metabolic job.

The researchers showed that this hidden function causes an imbalance in gene regulation, which disrupts cellular cleanup processes like autophagy. This imbalance increases the buildup of harmful plaques made of beta-amyloid, a hallmark of Alzheimer’s.

“It really demanded modern AI to formulate the three-dimensional structure very precisely to make this discovery,” said Zhong.

Now that the team knew PHGDH had this secret function, they looked for ways to block it—without harming its main job in serine production. They focused on a small molecule called NCT-503. This compound doesn’t affect serine creation much, but it can cross the blood-brain barrier, a key feature for any drug targeting the brain.

Using AI tools, they saw that NCT-503 could fit into a binding pocket near the protein’s DNA-binding region. Lab tests confirmed it blocked the gene-regulating function of PHGDH. When the researchers gave NCT-503 to two different mouse models of Alzheimer’s, the results were striking. Mice treated with the compound showed fewer amyloid plaques and improved behavior in tests for memory and anxiety.

These findings matter because most Alzheimer’s treatments target symptoms or work too late in the disease. Many focus on clearing amyloid plaques after they’ve formed. But by then, brain damage may already be too severe.

Zhong’s team took a different approach. By stopping the process that leads to plaque buildup in the first place, they showed that early intervention could work better.

“Now there is a therapeutic candidate with demonstrated efficacy that has the potential of being further developed into clinical tests,” said Zhong. He believes this could lead to a new class of treatments—possibly even pills—that would target the disease early and more effectively.

While this discovery is promising, there are still challenges ahead. Mouse models aren’t perfect copies of human brains, and most available models are based on inherited forms of Alzheimer’s. The spontaneous, late-onset form—by far the most common—is harder to mimic.

Still, Zhong’s team is optimistic. The next step is to refine the drug candidate and begin studies needed for FDA approval. These are called IND-enabling studies, which determine whether a compound is safe enough for testing in humans.

As scientists dig deeper into the biology of Alzheimer’s, this study suggests we may need to rethink the disease. Rather than just a result of toxic protein buildup, it might also be a breakdown in how brain cells manage their genes—triggered by a gene hiding in plain sight.

The discovery of PHGDH’s moonlighting role doesn’t just explain how Alzheimer’s starts in many people. It also points to a powerful way to catch the disease early and perhaps stop it before it robs more people of their memories.

Alzheimer’s disease often begins quietly, long before symptoms appear. Scientists believe that a mix of factors leads to its development. Genetics, lifestyle choices, and environmental exposures all play important roles. Though no single cause explains every case, certain patterns have become clear.

One major risk comes from genetics. A small number of people inherit specific gene mutations that almost guarantee Alzheimer’s. More often, it’s genes like APOE4 that raise the chances but don’t seal the fate. Having a close family member with Alzheimer’s also makes it more likely, but it’s not a sure thing.

Age remains the biggest risk factor by far. Most people who get Alzheimer’s are over 65. As the brain grows older, it becomes less able to clear waste proteins. These sticky proteins, like amyloid-beta and tau, clump together and damage brain cells. Over time, the damage spreads, making memory and thinking worse.

Lifestyle choices matter, too. Poor heart health, lack of exercise, smoking, and unhealthy diets can raise risk. Conditions like high blood pressure, diabetes, and obesity seem to speed up brain aging. Meanwhile, staying mentally and socially active may help protect the brain for longer.

Head injuries also raise the risk. People who have suffered major blows to the head, especially repeated ones, may develop Alzheimer’s earlier. Scientists think that trauma may trigger the buildup of harmful proteins.

Finally, the environment plays a role we are still trying to understand. Exposure to air pollution, toxins, or even chronic stress might add to brain decline. Although these factors don’t guarantee disease, they seem to push vulnerable brains closer to the tipping point.

In the end, Alzheimer’s disease comes from a storm of influences, not just one spark. While age and genes set the stage, lifestyle and environment often write the final act. Early action on health habits can make a real difference, even if it cannot promise total protection.

Note: The article above provided above by The Brighter Side of News.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.