These switches, found in noncoding regions of the genome, regulate how genes behave in astrocytes, cells that play a key role in brain health and inflammation.
The findings, led by scientists at UNSW Sydney, introduce a resource called AstroREG, designed to identify how changes in DNA regulation, rather than in the genes themselves, may contribute to Alzheimer’s. The study brings fresh focus to the 98% of human DNA that doesn’t make proteins but still influences which genes turn on or off.
Gene control is often overlooked in Alzheimer’s research, which has traditionally focused on mutations in protein-coding genes. But most genetic risk factors lie in noncoding DNA, making this work a step toward understanding how small regulatory changes can have wide effects on gene activity and disease progression.
Astrocytes Under The Microscope
The study centers on astrocytes, cells that help regulate inflammation, clean up excess neurotransmitters, and support neurons. Changes in astrocyte gene expression can influence nearby brain cells, especially since Alzheimer’s often begins with inflammation even before neurons show signs of damage.
Using the AstroREG platform, the researchers screened almost 1,000 candidate enhancer regions, stretches of DNA that don’t code for proteins but help control nearby genes. They used a method called CRISPR interference (CRISPRi), which silences DNA without cutting it, to test whether these enhancers were active inside cultured human astrocytes.
According to Earth.com, only a subset of these candidate regions turned out to be functional enhancers, capable of regulating gene expression. These switches were identified by their ability to change RNA levels in the same DNA neighborhood when silenced, helping scientists pinpoint exactly which genes were affected.
Tracing The Effect Of Each Enhancer
To measure how gene activity responded to enhancer silencing, the team used single-cell RNA sequencing. This allowed them to track how each individual astrocyte changed its gene expression after CRISPRi treatment, rather than relying on averages across cell populations. This level of detail revealed specific gene targets and showed how even small regulatory changes can shift gene behavior.
Many of the target genes uncovered in the experiment were already known to be dysregulated in Alzheimer’s brain tissue, either expressed too much or too little. The presence of these genes in the enhancer dataset strengthens the connection between astrocyte regulation and Alzheimer’s pathology.
Interestingly, some enhancers skipped over nearby genes entirely to control more distant ones. The team cautions against assuming that the nearest gene to a regulatory region is always the one being affected. This insight helps refine how researchers interpret genetic association studies, where noncoding variants are often linked to disease but hard to trace to specific genes.
From Experimental Data To Machine Learning
Beyond identifying active enhancers, the researchers also used their results to train computational models. A random forest model, a type of algorithm based on multiple decision trees, was tested using the data from AstroREG to improve predictions about enhancer activity in the brain.
According to Professor Irina Voineagu, who led the study, the resource now serves as valuable training material for computational biologists. But even the best models still need lab confirmation, as enhancer activity can vary based on cell type, developmental stage, or environmental stress.
The study also acknowledges its limitations. The astrocytes used were derived from cultured human cells with fetal characteristics, which might not fully mimic aging brain environments. CRISPRi can dampen enhancer activity but may not completely block it, meaning some functional regions could have gone undetected.
Still, as reported by the same source, the results lay important groundwork for future studies of Alzheimer’s genetics.