Mapping DNA in 3D yields insights into Alzheimer's disease

Single-cell approaches show that alterations in the 3D architecture of the genome are tied to changes in gene activity in Alzheimer's disease.

Much research into Alzheimer's disease has focused on characterising the genetic factors underlying the condition, investigating the abnormal accumulation of amyloid-beta and tau proteins, and identifying potential environmental and lifestyle risk factors. A new study published in Science now considers the role of the 3D structure of the genome – how chromatin is organised spatially inside a cell.

'We know the classic hallmarks of Alzheimer's disease – accumulation of amyloid-beta plaques and tau tangles – but our results establish higher-order chromatin alterations as a component of the molecular pathology associated with the disease,' said Dr Hansruedi Mathys from the University of Pittsburgh, Pennsylvania, a co-senior author of the paper.

The intricate, dynamic way the genome folds and organises itself within the tight space of the nucleus has emerged as a key regulatory layer for gene activity. Chromatin interactions can bring distant regions of the genome into contact, helping control how and when genes are turned on. At a larger scale, A/B compartmentalisation segregates transcriptionally active and inactive sequences.

Disruptions in the 3D genome architecture have been linked to various neurological disorders, but how this gene regulatory layer is affected in Alzheimer's disease remains poorly understood. To address this question, the team combined several approaches using postmortem prefrontal cortex tissue from healthy individuals and people with Alzheimer's disease.

The team used a multi-omics tool, GAGE-seq, to assess 3D genome folding and gene expression in individual cells, alongside spatial transcriptomics to map gene activity in intact tissues. An AI model, Hicformer, was then used to investigate how genome folding might contribute to gene expression changes associated with Alzheimer's disease.

The experiments revealed that brain cells from people with Alzheimer's disease featured 'increased compartment mingling', whereby the sharp boundaries between A and B compartments became blurred. Cells in which mingling was more severe had lower gene activity overall. Short-range chromatin interactions declined in cells from Alzheimer's disease patients while some midrange interactions were strengthened.

The AI model showed that 3D genome architecture was important for predicting gene activity changes associated with Alzheimer's disease. Disrupted genome architecture was linked to transcriptional changes in gene networks involved in neuronal and synaptic function, metabolism and stress responses, and ageing-related processes in microglia, a type of brain immune cell.

These results do not demonstrate causal mechanisms, as they establish correlative associations based on static snapshots. Further work is needed to determine whether genome changes contribute to Alzheimer's disease pathology, or result from it.

'Alzheimer's disease cannot be understood one layer at a time,' said co-senior author Professor Jian Ma from Carnegie Mellon University, Pennsylvania. '... By integrating genome folding, cell state and tissue context, we can move beyond cataloguing disease-associated changes toward understanding how they fit together and which mechanisms to test next.'