The contacts that matter for gene expression find each other by other means.
Every time a cell divides, it faces a remarkable logistical problem: the entire architecture of its genome — the precise spatial relationships between genes and the distant switches that control them — must be dismantled during mitosis and then rebuilt from scratch. How that rebuilding happens, and what machinery drives it, has been one of the central puzzles of modern cell biology. A new study published in Nature offers a surprising answer to part of that puzzle: the contacts that matter most for gene regulation can reassemble on their own, without the molecular motor scientists had assumed was doing the heavy lifting.
The motor in question is a process called loop extrusion, carried out by a protein complex called cohesin. The way it works is elegant: cohesin latches onto DNA and reels it in, forming expanding loops until it encounters a roadblock — typically a protein called CTCF sitting at a convergently oriented binding site. The result is a stable structural loop, and the genome is organized into a series of these loops and the domains they define. For years, the assumption has been that this same extrusion process is responsible not just for structural organization but for bringing enhancers — the regulatory switches — into physical contact with the promoters they activate.
To test that assumption directly, the researchers focused on a protein called NIPBL, which acts as the essential activator of cohesin's loop-extruding activity. Without NIPBL, cohesin can still bind to DNA but loses its ability to extrude loops efficiently. The team used a chemical degron system to rapidly degrade NIPBL at a precise moment: the transition from mitosis back into the G1 phase of the cell cycle, exactly the window when chromatin architecture is being rebuilt after division. This allowed them to ask, cleanly, whether loop extrusion is required for the de novo establishment of regulatory contacts.
The answer split neatly in two. For structural loops — the large-scale organizational features anchored at CTCF sites — NIPBL depletion caused clear impairment, and the damage scaled with loop length. Longer loops, which require more sustained extrusion activity to form, were more severely disrupted. Computational modeling of cohesin dynamics supported this finding, suggesting NIPBL contributes both to loading cohesin onto chromatin and to driving the extrusion process itself.
But for the loops connecting cis-regulatory elements — the enhancer-to-promoter contacts that directly govern which genes get turned on — the picture was strikingly different. The vast majority of these regulatory contacts formed normally even when NIPBL was absent and loop extrusion was impaired. Gene activation after mitosis was only mildly affected. A subset of contacts between weaker regulatory elements did show delayed formation kinetics under NIPBL depletion, suggesting that extrusion can assist in some cases, particularly where the regulatory signals are faint. But the overall conclusion was clear: the genome does not need loop extrusion to reestablish the contacts that drive gene expression.
This distinction — structural loops dependent on extrusion, regulatory loops largely not — has real conceptual weight. It suggests that the genome maintains two partially separable organizational logics. One is mechanical and length-dependent, built by cohesin motors grinding along DNA. The other is driven by something else: the intrinsic affinity of regulatory elements for one another, perhaps mediated by transcription factors, co-activators, or the physical properties of active chromatin itself. The study does not fully resolve what that alternative mechanism is, but it establishes clearly that it exists and that it is robust enough to operate independently of the extrusion machinery.
The findings carry implications beyond basic cell biology. Mutations in NIPBL and other cohesin pathway components cause a group of developmental disorders collectively called cohesinopathies, the most well-known being Cornelia de Lange syndrome. These conditions have long been interpreted partly through the lens of disrupted loop extrusion and the gene-regulatory failures that might follow. If enhancer-promoter contacts are largely extrusion-independent, the disease mechanisms may be more nuanced — and more targeted — than previously appreciated. The structural disorganization caused by cohesin pathway mutations may matter differently than the regulatory disorganization, and the two may need to be disentangled in future research.
What the study leaves open is the identity of the extrusion-independent mechanism and whether it operates the same way in all cell types and developmental contexts. The experiments were conducted during a specific and somewhat unusual cellular moment — the immediate aftermath of mitosis — and it remains to be seen how broadly the findings generalize. Still, the core result stands: when the cell rebuilds its regulatory genome after division, it does not wait for the motors to run. The contacts that matter for gene expression find each other by other means.
Notable Quotes
Postmitotic establishment of regulatory contacts and gene transcription can occur independently of chromatin loop extrusion.— Study authors, as summarized in the paper's conclusions
The Hearth Conversation Another angle on the story
So the headline finding is that enhancers and promoters connect without loop extrusion — but wasn't that already suspected?
Suspected in some quarters, yes. There was evidence that cohesin loss didn't always destroy regulatory contacts. But this study is the first to watch it happen in real time, right at the moment the genome is being rebuilt after cell division.
Why does that timing matter so much?
Because at mitosis, everything is wiped clean. The loops are gone, the contacts are gone. So when you watch the cell re-enter G1, you're seeing de novo establishment — not maintenance of something that was already there.
And NIPBL depletion at that exact moment is the key experimental move.
Right. NIPBL is what activates cohesin's motor function. Take it away and you've essentially stalled the extrusion machinery before it can do anything. If contacts still form, they're forming by some other route.
The structural loops were impaired though — and in a length-dependent way. What does that tell us?
It tells us that long-range structural organization genuinely needs the motor. The longer the loop, the more extrusion cycles required to build it, so the more sensitive it is to losing NIPBL. That part of the story fits the standard model perfectly.
So the genome has two organizational systems running in parallel?
That's one way to read it. One system is mechanical — cohesin motors building large-scale structure. The other is something more like chemical affinity — regulatory elements finding each other because of what they are, not because a motor pushed them together.
What might that second system actually be?
The study doesn't pin it down, but the candidates include transcription factors that bind both enhancers and promoters simultaneously, phase-separation-like clustering of active chromatin, or proteins like LDB1 that are known to bridge regulatory elements independently of cohesin.
Does this change how we should think about cohesinopathy diseases?
It should at least complicate the picture. If gene regulation is largely extrusion-independent, then the developmental defects in conditions like Cornelia de Lange syndrome may not be primarily about broken enhancer-promoter contacts. The structural disorganization might matter in different ways — maybe through insulation failures, or effects on replication, or something else entirely.
What's the most important thing this study doesn't yet answer?
What the extrusion-independent mechanism actually is, and whether it works the same way in every cell type. The experiments were done in one specific cellular context. The next question is whether this is a universal feature of genome regulation or something particular to how certain cells handle the mitotic transition.