A sharp change at the start of S phase
Genome compartments in mouse embryonic stem cells changed most sharply as cells moved from late G1 into early S phase, when DNA synthesis begins. The compartment-strength signal then stayed broadly stable through S phase before declining in late S and G2. At the transition, A-B contacts were depleted while A-A contacts were enriched, indicating that the large-scale interaction pattern was being reorganized in a narrow window of interphase.
Researchers tracked that timing with a drug-free Fucci workflow and Hi-C, a method that maps how often different DNA regions contact one another. Reciprocal Cdt1 and Geminin fluorescence, combined with DNA content, sorted highly pure mouse embryonic stem-cell populations without chemical synchronization. Samples covered early, mid and late G1; early, mid and late S; and G2. Each experiment harvested 50 million to 100 million cells.
The analysis called A/B compartments, the two broad interaction patterns, at 1-megabase resolution using the first principal component of the Hi-C data. Its compartment-strength score compared A-A and B-B contact frequencies with A-B contacts, so a stronger score represented a sharper separation between same-type and mixed interactions. That measure provided the basis for comparing the seven phase groups across interphase.
The shift followed an S-like state
An arrest comparison sharpened the interpretation. INK-128-treated cells resembled late G1 in their Hi-C trajectory. Cells in the thymidine-treated G1/S-block condition had replication-program markers and followed an S-like trajectory even though DNA-synthesis elongation was blocked. In this comparison, the compartment pattern aligned more closely with an S-like state than with continued DNA-synthesis elongation.
A related test measured insulation at RAD21/CTCF sites, a readout of local boundary behavior. During normal cycling, compartment strength rose as insulation weakened. In G1/S-arrested cells, the compartment shift was still observed while insulation increased slightly. The two measurements therefore did not move in lockstep, although this comparison does not establish that cohesin has no role in compartment organization.
A compartments pulled together
Changes within the A side of the map added another layer. The strongest A category, rank 1, accumulated from early G1 and peaked in mid S, then decreased in late S and G2. Medium-ranked A categories declined between early G1 and mid S. The result was a stronger representation of the top A state during part of S phase, consistent with consolidation rather than a simple uniform increase across all A categories.
Long-range A-A interactions were enriched in early S and reduced in G2. The late-G1-to-early-S change was largest for inter-A contacts, with p=1.74 × 10⁻¹⁰ and a median change of 0.64. DNA-FISH imaging, which measures distances between selected DNA probes, also showed larger inter-probe distances from G1 to early or mid S, with no significant later change. The authors noted substantial cell-to-cell heterogeneity and said additional contributors to the Hi-C pattern could not be excluded.
The pattern also appeared in a developmental reanalysis. Across four selected mouse developmental cell types with sufficient Hi-C coverage, pooled data recapitulated increased compartment strength and A-A interactions, along with enrichment of rank-1 strong-A bins in mid S. Because the analysis was limited to selected cell types, it extends the comparison without showing that every developmental context follows the same course.
A population-level model
To translate contact patterns into a possible three-dimensional picture, the researchers generated LorDG models. The models matched the experimental Hi-C patterns with an average absolute Spearman correlation of 0.77. They indicated that large A domains extended without a change in the distances between their boundaries, while B domains became more compact, with shorter boundary-to-center and boundary-to-boundary distances.
Those results underpin a four-stage picture of interphase organization: chromosome unfolding in G1, compartment maturation at G1/S, stabilization during S, and chromosome refolding in G2. The sequence is a proposed model of how the measurements fit together, not a claim that every nucleus passes through identical structural steps.
The central caution is that the 3D structures were derived from population-averaged Hi-C. They are heuristic frameworks rather than definitive representations of individual nuclei. The broader findings therefore describe a population-level transition, not the exact behavior of every single cell.
Paper data and sources
Original title: Cell-cycle-resolved Hi-C reveals unexpected plasticity of A/B compartments across interphase
Authors: Choubani L, Miura H, Ichinose T et al.
Journal/Repository: eLife
Status: Peer-reviewed
First online: 2026-10-07
DOI: Not available
Original paper · Full text