Chromosomal instability (aneuploidy) drives cancer progression and genetic disease, yet the molecular brakes that normally prevent centromere dysfunction remain poorly understood. A new study from an international collaboration reveals that DNA methylation acts as a critical boundary to maintain centromere architecture, with implications for understanding how epigenetic dysregulation contributes to cancer.
Salinas-Luypaert et al. (2025) in Nature Genetics demonstrate that centromeric DNA methylation patterns are not merely decorative but functionally essential for controlling where the centromeric histone CENP-A localizes on chromosomes. When this methylation is disrupted, centromeres destabilize, leading to chromosome missegregation and cell death.
This finding bridges two areas that often seem disconnected in the literature: epigenetics and centromere biology. For researchers working on cancer genomics, chromosomal instability mechanisms, or epigenetic regulation, this work clarifies how these domains intersect at the centromere.
The Finding in Plain Terms
At the heart of every cell lies a problem: chromosomes must be copied and separated accurately during cell division. The centromere is the physical knot that holds sister chromatids together and recruits the molecular machinery needed for this division. The centromere’s location is marked by a specialized histone protein called CENP-A, which replaces normal histone H3 in centromeric chromatin.
For decades, scientists knew that centromeres are heavily methylated. These centromeres sit within massive stretches of repetitive DNA called satellite repeats, and this DNA is densely methylated. Within this hypermethylated landscape, there is a dip in methylation directly at the CENP-A-containing region. But no one had experimentally tested whether this methylation pattern actually mattered.
The new study shows that it does. Using targeted epigenetic editing tools including dCas9-DNMT and dCas9-TET fusion proteins, Salinas-Luypaert et al. selectively removed methylation from centromeric regions in human cells. The result: CENP-A spread outward from its normal boundaries, centromeric proteins accumulated abnormally, and chromosomes began to mis-segregate during mitosis.
Why It Matters
The clinical relevance is immediate. Chromosomal instability is a hallmark of cancer cells. If centromeres destabilize, you get aneuploidy, and aneuploidy drives oncogenic transformation. By identifying DNA methylation as a physical boundary that constrains CENP-A, this work opens a new lens on cancer progression: epigenetic disruption at centromeres could be a driver of aneuploidy independent of the chromosome translocations or copy number changes typically cited in cancer genomics.
Additionally, the study reveals a temporal dimension: the speed at which methylation is lost determines the cellular outcome. Rapid demethylation triggers immediate centromere dysfunction and cell death, while gradual loss allows cells to adapt. This distinction has practical implications for therapies targeting DNA methyltransferases (DNMTs). If DNMT inhibitors cause rapid centromeric demethylation, they may trigger aneuploidy and cell death in cancer cells but also create genomic instability that could be dangerous in normal tissues.
The work also clarifies a fundamental principle of centromere biology: centromere position is not fixed by CENP-A alone. Instead, CENP-A localization is maintained by the epigenetic landscape surrounding it. This suggests that centromere instability in cancer may not always require mutations in centromeric proteins themselves; epigenetic changes may be sufficient.
How They Did It
The researchers used a combination of approaches to test whether centromeric methylation causally influences CENP-A positioning. First, they employed dCas9 (catalytically inactive Cas9) fused to DNMT3A to add methylation or to TET1 (a demethylase) to remove methylation at specific centromeric loci in cultured HeLa and HCT116 cells. They combined this targeted epigenetic editing with:
- Immunofluorescence and chromatin immunoprecipitation (ChIP) to measure CENP-A and CENP-B localization
- Next-generation sequencing (NGS) to assess methylation patterns genome-wide and at centromeres specifically
- Live-cell microscopy and FISH (fluorescence in situ hybridization) to track chromosome segregation errors
- Flow cytometry to measure aneuploidy
This multi-pronged approach ensured they could detect the molecular and cellular consequences of centromeric methylation loss. The sample sizes were reasonable (multiple cell lines, multiple replicates), and the use of targeted epigenetic editors allowed them to avoid off-target effects of global DNMT inhibitors.
Limitations and Caveats
Several important caveats apply. First, this work was conducted in cultured human cells, not in tissues or organisms. While centromere biology is likely conserved across human tissues, the specific epigenetic thresholds that trigger aneuploidy in vivo could differ from those in rapidly dividing cancer cell lines.
Second, the study focuses on HeLa and HCT116 cell lines, both of which are cancer-derived and already genetically unstable. It remains unclear whether centromeric methylation operates identically in normal diploid cells.
Third, the relationship between rapid and gradual demethylation and cellular adaptation is intriguing but mechanistically underexplored. The authors identify that gradual loss allows adaptation, but the molecular basis of this adaptation remains unknown.
Finally, while this work is foundational, it does not directly show that centromeric methylation loss is a driver of cancer in vivo. This is an early-stage mechanistic finding, not evidence that disrupting centromeric methylation is a viable cancer therapy strategy.
What This Means in Practice
For researchers working on chromosomal instability in cancer, this work suggests a new angle for investigation: epigenetic dysregulation at centromeres. If you are studying aneuploidy drivers, asking whether centromeric methylation is altered in your tumor samples could yield insights.
For those developing DNMT inhibitors or other epigenetic therapies, this work underscores that blocking methylation globally carries risks of centromere destabilization and unintended aneuploidy. Tissue-specific or locus-specific approaches might be safer.
For cell biologists studying centromere organization, this is a reminder that chromatin context shapes protein localization in ways that go beyond the canonical histone codes. The centromere’s position is not hardwired but maintained by a dynamic equilibrium of histone variants, DNA methylation, and binding proteins.
For computational and bioinformatics researchers, the underlying methods (dCas9 epigenetic editing combined with sequencing and imaging) represent a general strategy for testing whether regulatory elements (methylation, histone modifications, transcription factor binding) actually cause phenotypic changes, as opposed to merely correlating with them. This causal inference approach is increasingly important as genomics becomes more predictive.
Source and Further Reading
For deeper context, see the accompanying editorial in the same issue: “Dipping into the function of DNA methylation at centromeres” in Nature Genetics.
For related work on centromere biology, the 2024 Nature paper on centromere structure and evolution provides excellent background: “The variation and evolution of complete human centromeres”.
The Bottom Line
DNA methylation is not just a regulatory tag that modulates gene expression; at the centromere, it is a structural necessity that prevents chromosomal chaos. This work reveals how epigenetic changes can destabilize the genome without requiring mutations in centromeric proteins. For researchers studying cancer evolution or epigenetic inheritance, this adds an important mechanism to your mental model of how genomes stay stable.
If you’re interested in how epigenetic changes drive disease, this paper exemplifies the power of combining targeted molecular tools (dCas9 editors) with quantitative cellular and genomic readouts to move from correlation to causation.