Fully T2T pedigree assemblies reveal genetic stability and epigenetic plasticity of human centromeres across inheritance and cell-fate transitions

By leveraging fully phased telomere-to-telomere pedigree assemblies and matched long-read epigenomes across cell-fate transitions, this study reveals that while human centromeric dip regions maintain positional stability across generations and differentiation, their epigenetic architecture exhibits remarkable plasticity, characterized by dynamic methylation changes during reprogramming and differentiation, insulation from X-chromosome status, and a distinct pattern of de novo mutations that are enriched in centromeric regions but depleted within functional cores.

The Gist

Imagine your DNA as a massive, 30,000-page instruction manual for building a human. For decades, scientists could read most of the book, but there was a specific, crucial chapter in the middle of every page that was completely unreadable. This chapter was written in a language of repeating gibberish (like "AAAAA... AAAAA..."), making it impossible to assemble the pages in the right order.

This unreadable chapter is the centromere. It's the "waist" of your chromosome, the spot where the cell's machinery grabs on to pull the chromosomes apart when a cell divides. If this waist breaks or gets confused, the cell dies or becomes cancerous.

This new study is like finally getting a high-resolution, color-coded map of that unreadable chapter. The researchers didn't just look at one person; they looked at a three-generation family (grandparents, parents, and children) and tracked how this "waist" behaves when cells change their jobs (like turning a blood cell into a stem cell, and then into a brain cell).

Here is what they discovered, explained through simple analogies:

1. The "Anchor" vs. The "Paint Job"

The most surprising finding is a split personality of the centromere.

• The Position is an Anchor: The exact location of the centromere is incredibly stubborn. Whether you pass it down from a grandparent to a grandchild, or turn a blood cell into a brain cell, the centromere stays in the exact same spot on the chromosome. It's like a house address that never changes, no matter how many times you repaint the house or who lives there.
• The "Paint Job" is Fluid: While the address stays the same, the chemical coating (epigenetics) on that spot changes wildly. Think of the centromere as a building. The foundation (the DNA sequence) is solid, but the interior decoration (methylation, proteins) gets completely renovated every time the cell changes its identity.

2. The "Reset Button" and the "Recovery"

The researchers watched what happened when they took adult blood cells and hit the "Reset Button" to turn them into Stem Cells (iPSCs), and then let them grow up to become Brain Cells (Neural Progenitor Cells).

• The Reset (Stem Cells): When the cells became stem cells, the "special zone" in the middle of the centromere (called the Centromeric Dip Region or CDR) got messy. It lost its special chemical markers. Imagine a highly organized command center suddenly getting covered in fog and losing its security guards. The researchers call this "attenuation." The cell forgets exactly where the "grab point" is, even though the location hasn't moved.
• The Recovery (Brain Cells): When those stem cells started turning into brain cells, the command center got cleaned up! The fog lifted, the security guards returned, and the chemical markers were restored. However, it wasn't a perfect 100% return to the original state; it was a "partial restoration," like renovating a house to be functional again, but with a slightly different style than the original.

3. The "X-Chromosome" Mystery

In women, one of the two X chromosomes is usually "shut down" (inactivated) to prevent double-dosing of genes. This shut-down state affects the whole chromosome, like putting a "Do Not Disturb" sign on a whole floor of an office building.

The study asked: Does this "Do Not Disturb" sign affect the centromere?
The answer: No.
Even though the rest of the X chromosome changes its chemical state dramatically, the centromere's "waist" remains completely unaffected. It's like the "Do Not Disturb" sign is on the walls and windows, but the front door (the centromere) ignores it and keeps working normally. This shows the centromere is incredibly insulated and protected from the rest of the chromosome's drama.

4. The "Danger Zone" for Mutations

Finally, the researchers looked for "typos" (mutations) that happen when cells are reprogrammed.

• The Good News: The most important part of the centromere (the actual "grab point" where the cell machinery attaches) is very well protected. It rarely gets typos.
• The Bad News: The areas surrounding that grab point are like a construction site. They are messy, repetitive, and prone to errors. When cells are reprogrammed, these surrounding areas accumulate mutations at a rate three times higher than the rest of the genome.

Why Does This Matter?

This study is a big deal for two main reasons:

1. Stem Cell Safety: We use stem cells to try to cure diseases. This study warns us that while the "address" of the chromosome is safe, the "neighborhood" around the centromere is a hotspot for errors during the reprogramming process. We need to be careful to check these areas to ensure the cells are safe for patients.
2. Understanding Life: It solves a 50-year-old mystery. We now know that nature keeps the centromere's location rock-solid (so chromosomes don't fall apart) but allows its chemical state to be flexible (so cells can change from blood to brain).

In a nutshell: The centromere is the most stable address in the genome, but the house at that address gets completely remodeled every time a cell changes its job. And while the main room is safe, the backyard is a bit of a construction zone where mistakes happen easily.

Technical Summary

1. Problem Statement

Centromeres are essential for chromosome segregation but have historically been poorly understood due to their highly repetitive $\alpha$-satellite DNA architecture, which resists short-read sequencing and assembly. While recent Telomere-to-Telomere (T2T) assemblies have provided a reference framework, critical gaps remain regarding:

• Inheritance: How centromeric genetic and epigenetic features are transmitted across generations.
• Cell-Fate Transitions: How centromeric chromatin (specifically DNA methylation and nucleosome organization) responds to massive epigenetic resetting during cellular reprogramming (to iPSCs) and differentiation (to Neural Progenitor Cells, NPCs).
• Genetic Stability: The mutational landscape of centromeric satellites during reprogramming, a region largely uncharacterized due to technical limitations.
• Insulation: Whether centromeres are insulated from chromosome-wide epigenetic programs, such as X-chromosome inactivation (XCI).

2. Methodology

The study employed an integrated multi-omics approach across a three-generation African American pedigree (Grandparents, Daughter, Granddaughter):

• Genomic Resource: Utilized fully phased, high-quality T2T diploid assemblies for all four individuals (derived from Lymphoblastoid Cell Lines). These assemblies were validated for accuracy (flagging <0.15% of bases as potentially erroneous).
• Cell Models: Reprogrammed Peripheral Blood Mononuclear Cells (PBMCs) from all four individuals into Induced Pluripotent Stem Cells (iPSCs) and subsequently differentiated them into Neural Progenitor Cells (NPCs).
• Long-Read Epigenomics: Generated PacBio Fiber-seq data (and Oxford Nanopore for specific clones) for all cell types. Fiber-seq allows simultaneous detection of:
  • CpG methylation (5mC).
  • Nucleosome footprints (via m6A labeling of accessible DNA).
  • Protein-bound regions (FIREs).
• Allele-Resolved Analysis: Reads were phased against the T2T assemblies to track maternal and paternal haplotypes independently.
• Computational Metrics:
  • MARS (Methylation Area Score): A novel metric quantifying the integrated hypomethylation (depth $\times$ breadth) of Centromeric Dip Regions (CDRs).
  • FAS (FIRE Area Score): Quantified protein occupancy within CDRs.
  • Variant Calling: Identified de novo mutations by comparing iPSC/NPC assemblies against parental PBMC assemblies.

3. Key Contributions

• First Allele-Resolved Centromere Maps: Generated single-basepair resolution maps of centromere genetics and epigenetics across inheritance and cell states.
• Definition of CDR Dynamics: Established that while the positional location of functional centromere cores (CDRs) is stable, their epigenetic architecture is highly plastic.
• Discovery of Reprogramming-Induced Remodeling: Revealed that reprogramming collapses CDR hypomethylation and alters nucleosome organization, which is partially restored upon differentiation.
• Mutational Landscape: Mapped the de novo mutation burden in centromeric satellites, revealing they are hotspots for mutation but that functional cores are protected.
• Insulation Mechanism: Demonstrated that centromeres are epigenetically insulated from chromosome-wide X-inactivation states.

4. Key Results

A. Genetic Stability and Inheritance

• Positional Stability: The physical location of Centromeric Dip Regions (CDRs) is highly stable across three generations and cell-fate transitions (positional shifts <50 kb).
• Recombination: Meiotic recombination events are strongly depleted within centromeric cores, consistent with classical patterns.
• Haplotype Tracking: Successfully traced specific centromeric haplotypes through meiosis, confirming the fidelity of inheritance for centromeric sequences.

B. Epigenetic Plasticity (Methylation & Chromatin)

• CDR Collapse in iPSCs:
  • Reprogramming to iPSCs causes a ~49% reduction in CDR hypomethylation (MARS score), indicating a loss of the "dip" architecture.
  • CDR width narrows significantly in iPSCs compared to PBMCs.
  • Protein occupancy (FAS) decreases, and nucleosome organization shifts toward canonical sizes, losing specialized structures.
• Partial Restoration in NPCs:
  • Differentiation into NPCs partially restores CDR hypomethylation (~91% recovery of PBMC levels).
  • Restoration correlates with the demethylation of active $\alpha$-satellite arrays outside the CDR.
  • Nucleosome Remodeling: NPCs show a re-emergence of short nucleosome footprints (75–125 bp, likely CENP-A containing) and large protected regions (200–300 bp, likely large protein complexes), distinct from the canonical nucleosomes seen in iPSCs.
• MARS Metric: The MARS score is stable across generations but varies by cell type, capturing the "tunable" nature of centromeric epigenetic capacity.

C. Insulation from X-Chromosome Inactivation

• Chromosome Arms vs. Centromeres: While X-chromosome inactivation (Xi) and erosion (Xe) cause massive, global hypomethylation differences across chromosome arms, these effects do not extend to the centromere.
• Stability: Centromeric methylation patterns and CDR positions remain stable regardless of whether the X chromosome is active (Xa), inactive (Xi), or eroded (Xe). This suggests centromeres are governed by distinct, local regulatory mechanisms.

D. De Novo Mutations and Genetic Stability

• Mutation Hotspots: Centromeric satellites (cenSat) accumulate the highest burden of de novo mutations during reprogramming, with an enrichment 1.88-fold higher than the genome average and 3.31-fold higher than coding exons.
• Domain-Specific Protection:
  • Depleted: Functional active HORs (kinetochore-forming) and centromere transition (ct) regions show strong depletion of mutations, suggesting selective pressure to preserve kinetochore integrity.
  • Enriched: Inactive $\alpha$-satellites and pericentromeric repeats accumulate significantly more variants.
• Mutational Signatures: Centromeric regions exhibit distinct mutational spectra (enriched for T>A and T>G substitutions) and are dominated by signatures associated with replication/repair (SBS3) and CpG deamination (SBS1), unlike the clock-like SBS5 signature dominant in chromosome arms.

5. Significance

• Conceptual Model: The study proposes a model where centromeres possess a stable positional scaffold (anchored by CENP-A inheritance) but a tunable epigenetic state that responds to developmental cues.
• Safety in Regenerative Medicine: The finding that centromeric satellites are mutation hotspots during iPSC derivation highlights a critical, previously overlooked risk for genomic instability in stem cell therapies and disease modeling.
• Mechanistic Insight: It reveals that centromere identity is not just a static sequence feature but a dynamic epigenetic state that undergoes specific remodeling (collapse and recovery) during cell-fate transitions, independent of global chromosomal epigenetic programs.
• Technical Advancement: Demonstrates the power of combining T2T assemblies with long-read epigenomics (Fiber-seq) to resolve complex, repetitive genomic regions that were previously inaccessible.

In summary, this paper redefines human centromeres as genomic domains characterized by the coexistence of architectural stability (positional and genetic) and epigenetic flexibility, providing a foundational framework for understanding centromere biology in development, inheritance, and disease.