Cracking Parkinson’s Code with Google AlphaFold

If proteins are life’s building blocks, AlphaFold is the architect who finally cracked the blueprint.

In 2020, DeepMind’s AlphaFold stunned the scientific world by solving a 50-year mystery in biology: how to predict the 3D structure of a protein from its amino acid sequence.

Since then, the open-source AlphaFold Protein Structure Database has become a global resource, offering more than 200 million predicted protein structures. With over 2.5 million users across 190 countries—and over one million of those based in Asia Pacific—the tool has gone from research novelty to scientific necessity. So much so that the groundbreaking impact of AlphaFold was recognised with a Nobel Prize in Chemistry in October 2024.

In Singapore, that leap is already translating into life-altering research, particularly in the area of Parkinson’s Disease.

Inside the Parkinson’s Puzzle

At the Agency for Science, Technology and Research (A*STAR), scientists are turning to AlphaFold not just to accelerate discovery, but to reveal what’s long been invisible: the molecular story behind Parkinson’s Disease.

Parkinson’s Disease is a progressive neurodegenerative disorder. Over time, it affects motor function, balance, and coordination – often leading to tremors, muscle rigidity, and eventually, loss of independence.

Despite decades of research, its underlying cause has remained frustratingly complex. But new insights are emerging through proteins like STIP1 (Stress-Induced Phosphoprotein 1), a lesser-known player with a potentially outsized role.

STIP1 and AlphaFold

STIP1 is what’s known as a co-chaperone protein. In healthy cells, it acts like a molecular mediator – helping other proteins fold into their correct shape by working alongside heat shock proteins like HSP70 and HSP90. These chaperones ensure that proteins don’t misfold, because when they do, they can clump together and form toxic aggregates, a hallmark of Parkinson’s and other neurodegenerative diseases.

Using AlphaFold, researchers at A*STAR were able to model the precise 3D structure of STIP1 and its various domains. This is work that would have taken months or years using conventional methods. Significant research dollars and years in research time have been saved.

This structural model gave scientists a peek into how STIP1 behaves in both healthy individuals and those with Parkinson’s. What they found was both intriguing and concerning: patients with Parkinson’s had elevated levels of autoantibodies targeting STIP1.

These faulty antibodies appeared to bind specifically to domains on the STIP1 protein responsible for its interaction with HSP70 and HSP90.

In other words, the body’s own immune system was interfering with the very machinery needed to prevent protein misfolding.

How Does AlphaFold Help?

With the help of AlphaFold’s predictions, researchers could visualise how this interference may be disrupting the normal “hinge” motion of STIP1, thereby limiting its chaperoning ability. Without this support, misfolded proteins like the infamous alpha-synuclein (whose abnormal aggregation and accumulation forms Lewy bodies) found in Parkinson’s are more likely to aggregate and damage brain cells.

It’s both a sobering and a hopeful realisation. Because now, scientists not only understand the “what” but also the “how.” And that shift from symptom to mechanism marks a turning point in the way we think about diagnosis and treatment.

Beyond the Brain

AlphaFold’s applications in Singapore go far beyond neurology.

At Nanyang Technological University, it’s being used to study the structures of viral and insecticidal proteins. The information provides insights to their function and interactions, guiding experimental work aimed at developing new therapies and pest control strategies, which is important for combating viral diseases and protecting crops that are vital for food security.

At A*STAR, AlphaFold is also driving evolutionary research into ancient microorganisms like the Asgard archaea, aiding in the foundational work that could power future biotechnologies.

This is the promise of AlphaFold: an open-access AI tool that gives scientists the clarity to see what was once unseen and to question what was once assumed.

For Singapore’s 21,000 AlphaFold users, it’s not just about using AI to solve problems faster. It’s about using AI to ask better questions, and find answers that might change the course of disease itself.