Research Project

How do chromosome ends decide a cell's fate?

Both ends of every human chromosome carry a telomere, a protective cap. When a cap opens, the cell may stop dividing, die, or, rarely, become unstable enough to start down the road to cancer. We study how telomeres switch between protected and unprotected states, and what happens to a cell when chromosome ends become fused. Our aim is to understand the earliest steps of ageing and cancer, one chromosome at a time.

Telomere switches — how cells deliberately open their telomeres during mitosis
One fusion, many fates — following a single fused chromosome in living cells
Broken chromosomes and immunity — when chromosome fragments alert the cell's defence system

Background

Telomeres: caps that open and close

Telomeres are long stretches of a repeated DNA sequence (TTAGGG) coated by a six-protein complex called shelterin. The very end of the DNA tucks back into itself to form a loop, the t-loop (Figure 1), which hides the chromosome end from the cell's DNA-damage alarm (Griffith et al. 1999; Doksani et al. 2013).

Figure 1. Structure of a human telomere. Telomeres consist of TTAGGG repeats bound by the shelterin complex. The single-stranded 3′ overhang at the very end folds back and invades the double-stranded repeats, forming the t-loop. This loop hides the chromosome end from the DNA-damage response.

Telomeres shorten every time a human cell divides (Harley et al. 1990). When some become too short, the cell stops dividing for good, a state called replicative senescence; supplying the telomere-extending enzyme telomerase prevents this (Bodnar et al. 1998). Cells that ignore this stop signal keep dividing until they reach crisis: chromosome ends fuse, most cells die, and rare survivors switch on telomerase and become immortal (Counter et al. 1992). Cells in crisis die by at least two routes: death after a prolonged mitotic arrest caused by telomere fusions (Hayashi et a. 2015), and autophagy-dependent death triggered through the cell's innate immune sensors (Nassour et al. 2019; Nassour et al. 2023).

We describe telomeres as existing in three states, shown in Figure 2 (Cesare et al. 2013; Hayashi & Cesare 2026):

Closed/Capped — looped and fully protected.
Intermediate — opened into a linear end that raises a damage signal, but still protected from fusion. This state can stop the cell cycle without harming the genome.
Open/Uncapped — unprotected ends that can fuse with other chromosomes.

The transitions between these states are where our questions begin.

Figure 2. Telomeres in cellular ageing and tumorigenesis. Each circle represents one telomere; a human cell has 92, and their lengths vary. Green, closed state: looped and protected. Orange, intermediate state: shortened telomeres that signal DNA damage but are still protected from fusion by residual repeats and TRF2; persistent signalling causes replicative senescence. Red, uncapped state (labelled "open" in the figure): telomeres that have lost repeats and TRF2 and can fuse with other DNA ends. Cells with defects in the DNA-damage response bypass senescence and enter crisis, where chromosome fusions cause chromosome instability and most cells die. Rare cells that switch on telomerase escape crisis and become immortal.

Pillar 1 — A telomere switch inside mitosis

Cells usually pass through mitosis quickly. When they are held there for hours, for example by anticancer drugs such as taxol, something unexpected happens: the cell actively opens a subset of its own telomeres. We discovered this mitotic arrest-dependent (MAD) telomere deprotection and have spent more than a decade working out how it is switched on.

What we found

Prolonged mitosis triggers a damage signal specifically at telomeres. It depends on the mitotic kinase Aurora B and the damage sensor ATM, and normal cells then halt in the next cell cycle (Hayashi et al. 2012).
The opened telomeres sit in the intermediate state: they signal damage but do not fuse (Cesare et al. 2013).
The switch contributes to cell killing by antimitotic drugs, during telomere crisis (Figure 3) and under lethal DNA replication stress (Hayashi et al. 2015; Masamsetti et al. 2019).
The full circuit: Aurora B marks the shelterin proteins TRF1 and TRF2, and the BLM–TOP3A–RMI enzyme complex then untangles the t-loop without shortening the telomere (Romero-Zamora et al. 2025).
A brake: a region of the WRN protein, the protein lost in the premature-ageing disease Werner syndrome, restrains the switch without using its enzyme activity (Romero-Zamora & Hayashi 2023).

Figure 3. One route to cell death during telomere crisis. Telomere fusions cause a prolonged mitotic arrest that depends on the spindle assembly checkpoint, the cell's quality control for chromosome separation. The arrest triggers mitotic telomere deprotection, which leads to death in mitosis or in the following cell cycle. A second route, autophagy-dependent cell death triggered through innate immune signalling, has also been reported.

What we are asking now

Why did cells evolve a way to open their own telomeres? We have proposed two hypotheses: a "mitotic timer" that limits how long a cell can stay in mitosis, and a role in separating sister telomeres that are entangled (Hayashi & Cesare 2026).
Could tuning this switch make anticancer drugs work better?

Pillar 2 — Following one fused chromosome

Chromosome fusions are a hallmark of telomere crisis, but in most cells they occur at random places and times, which makes their consequences hard to trace. We built a tool that creates one defined fusion and makes the cell that carries it glow; the Fusion Visualization system (FuVis).

What we found

FuVis (Fusion Visualization system; Figure 4): a reporter placed near the tip of the X chromosome in human colon cancer cells (HCT116). When the genome-editing enzyme Cas9 cuts it, a yellow fluorescent protein appears only if the two sister copies of the chromosome end are joined together (Kagaya et al. 2020).
A single sister chromatid fusion was enough to generate X-chromosome fragments lacking a centromere and micronuclei (small extra nuclei holding stray chromosome pieces). It raised the chance of a micronucleus about 10-fold, and micronuclei lengthened the following cell cycle about 2.3-fold (Kagaya et al. 2020).
FuVis2 adds a second colour as a built-in control in the same cells, so that the effects of a fusion can be separated from those of the DNA cut alone (Sato & Hayashi 2024).

Figure 4. The FuVis system. A reporter cassette is inserted near the end of the short arm of the X chromosome (Xp subtelomere) in HCT116 cells. The gene for a yellow fluorescent protein (mCitrine) is split in two: the first part is driven by a promoter and spliced to a neomycin-resistance gene, while the second part sits upstream in the reverse orientation. Cells carrying the cassette are drug-resistant but not fluorescent. When Cas9 cuts the cassette, most breaks are repaired with small mutations. In rare cases the two sister chromatids are joined, which brings the two halves together so that the full fluorescent protein is made by splicing. Only cells with this sister chromatid fusion glow.

What we are asking now

How does one chance fusion lead to such diverse outcomes, from repair to chromosome shattering to cell death?
We are extending FuVis to watch these events in more settings.

Pillar 3 — When broken chromosomes alert the immune system

Cells carry an alarm for foreign DNA, the cGAS/STING pathway, which normally detects viruses. Recently, cGAS has been shown to also gather on micronuclei –– abnormal forms of self-DNA in the cytosol associated with tumorigenesis –– and micronuclei have been proposed as a major source of the cell's own DNA that triggers inflammation. Using FuVis2, we tested this idea by following micronuclei born from a single chromatid fusion in living cells.

What we found (Sato & Hayashi 2024)

cGAS reached micronuclei mainly during mitosis, when the nuclear envelope breaks down, and rarely during interphase.
This capture did not switch on STING or its downstream immune genes.
After low-dose radiation, STING activation did not depend on micronuclei. It required DNA released from mitochondria through the channel protein VDAC1.
Conclusion: in these cells, micronuclei are not a potent trigger of the cGAS/STING alarm. (Shown in one human cancer cell line, HCT116.)

What we are asking now

If not micronuclei, which forms of chromosome-derived DNA do alert the cell's immune system, and what does that mean for cancer and ageing?

Our roots — Makoto's PhD work mapped DNA replication origins across the fission yeast genome and showed that the heterochromatin protein Swi6/HP1 activates replication origins (Hayashi et al. 2007; Hayashi et al. 2009).

We have more ongoing projects. If you are interested in our latest research topics, please contact the lab.

RESEARCH INTERESTS

Telomere

Tumorigenesis

Chromosome

Cell Cycle

Genome Editing

The old site (mthayashilab.com) will close in fiscal year 2026.
Please visit our new site at https://mthayashilab.med.kyoto-u.ac.jp.

テロメア・老化・がん化研究