According to SciTechDaily, scientists at UNSW Sydney have identified elusive DNA switches in brain support cells that influence genes tied to Alzheimer’s disease. In a study published in Nature Neuroscience, the team, led by Dr. Nicole Green and Professor Irina Voineagu, used a CRISPRi technique to test close to 1000 candidate regulatory regions called enhancers in human astrocytes grown in a lab. They combined this with single-cell RNA sequencing to track gene activity, finding that about 150 of those enhancers were functional. Strikingly, a large fraction of those functional switches controlled genes already implicated in Alzheimer’s, dramatically narrowing where to look in the vast non-coding “junk” DNA. The dataset is already being used by Google’s DeepMind to benchmark its AlphaGenome AI model.
The Wiring Diagram Behind the Genes
Here’s the thing about our DNA: the genes themselves are just the components. The enhancers are the wiring that decides when and where those components get power. And for decades, that wiring—making up about 98% of our genome—was mostly a mystery, often dismissed as “junk.” This study is a big step in mapping that circuitry specifically in astrocytes, the brain’s crucial support cells. By using CRISPRi to temporarily silence a potential enhancer and watching what happened to gene expression, they could directly prove which switches controlled which genes. It’s painstaking lab work, but it gives researchers a verified catalogue. Now they don’t have to guess; they have a list of real, functional switches to investigate further.
Why This Matters Beyond Alzheimer’s
Professor Voineagu makes a great point. When we do big genetic studies looking for links to diseases, we often find the problematic changes aren’t in the genes themselves, but in these in-between regulatory regions. But until now, it’s been incredibly hard to prove which of those millions of potential switches actually *does* anything in a specific human cell type. This work provides a blueprint. And it’s not just about understanding disease. This kind of precise mapping is the foundational step towards any future gene therapy that wants to be truly precise. Think about it: if you can target an enhancer that only works in astrocytes, you could theoretically tweak gene activity there without touching neurons or any other cell. That’s the dream of precision medicine.
From Lab Benches to AI Models
Maybe the coolest part? This hard-won experimental data is now fuel for AI. Testing a thousand enhancers manually is a huge task. But that dataset becomes the perfect training ground and benchmark for computational tools like DeepMind’s AlphaGenome. The goal is to teach AI to predict which DNA sequences are real enhancers, saving years of lab work in the future. It’s a perfect synergy: messy, complex biological experimentation providing the ground truth for sleek algorithms. One accelerates the other. So while therapies are still a long way off, this research is speeding up the basic science phase dramatically. We’re finally starting to read the instruction manual hidden in our own “junk” DNA.
