A Nature study published on November 11, 2026, reports direct measurements of the protein barrier inside living-cell nuclear pores. The researchers found that the barrier remains highly mobile rather than hardening like some isolated laboratory condensates. Two cellular factors, the transport receptor importin-beta and a sugar modification called O-GlcNAc, help maintain that dynamic state. The findings offer a more realistic explanation for how a tiny gateway can be selective enough to protect the genome while remaining fast enough to support the cell’s constant exchange of molecules.

The cell nucleus has a guarded molecular gateway

Human cells keep their DNA inside a nucleus, a compartment surrounded by a double membrane. The nucleus is protected, but it cannot operate as a sealed box. Messenger RNA must leave to help direct protein production, while proteins that regulate genes, repair DNA or organise chromosomes must enter. The traffic passes through nuclear pore complexes, large molecular assemblies embedded in the nuclear envelope. Each pore is a selective gateway: it allows many small molecules to diffuse while controlling the passage of larger cargoes.

The central channel of a nuclear pore contains flexible proteins known as FG nucleoporins, named for repeated phenylalanine-glycine motifs in their sequences. These disordered protein chains form a permeability barrier that helps distinguish cargoes that should pass from molecules that should be excluded. Scientists have studied the pore’s architecture for decades, but the physical state of the barrier inside a living cell has been debated. Some experiments on isolated proteins suggested liquid-like condensates, while others pointed toward gels, polymer brushes or more solid assemblies.

Why experiments in living cells were needed

Many earlier models relied on purified proteins outside their natural cellular environment. In a test tube, isolated FG nucleoporins can form dense condensates that initially behave like liquids and then gradually become more rigid. That behaviour is scientifically useful, but it does not prove that the same transition happens inside a living nuclear pore. A cell contains transport receptors, chemical modifications, membranes, molecular crowding and continuous traffic. Those components can change how proteins interact and how long they remain mobile.

The new study, led by researchers from Johannes Gutenberg University Mainz, the Max Planck Institute for Biophysics and Wuhan University, addressed the gap by measuring local protein motion inside living mammalian cells. The team used fluorescent dyes attached to proteins in the nuclear pore and measured their rapid movement using time-resolved fluorescence methods. The researchers focused on the intrinsic dynamics of the FG barrier rather than relying only on the shape of a purified protein assembly. The work was published in Nature on November 11 and is available as an open-access research article.

The barrier stays highly mobile inside living cells

The central finding is that FG proteins in living nuclear pores remained highly mobile throughout a two-hour observation period. The fluorescent probes moved rapidly, showing that the protein chains were not locked into a rigid gel during the measurements. The researchers compared this behaviour with an in-vitro model made from isolated proteins. In the test tube, the protein material became progressively more rigid after roughly two hours, while the barrier in living cells remained dynamic.

The difference suggests that the native pore is not simply a miniature version of a test-tube condensate. Its surrounding cellular environment actively maintains a material state that supports rapid and reversible molecular interactions. This matters because nuclear transport depends on many transient contacts. A cargo bound to a transport receptor must repeatedly interact with FG nucleoporins as it crosses the channel. Those interactions must be strong enough to provide selectivity but short-lived enough to permit efficient passage.

Importin-beta helps keep the protein network dynamic

One of the factors identified by the researchers is importin-beta, a major nuclear transport receptor. Importin-beta binds to FG motifs and helps carry selected cargoes through the pore. In the study’s experiments, adding importin-beta to purified FG condensates made them more swollen and more liquid-like. The receptor reduced close contacts between protein segments that are prone to aggregation, helping to delay the transition from a fluid condensate to a more solid material.

The team also used large-scale molecular simulations to explore how importin-beta behaves inside the pore. The model suggested that importin-beta becomes enriched toward the periphery of the channel, while extended FG domains occupy more of the central region. The researchers supported this arrangement with high-resolution imaging of importin-beta in mammalian cells. The model does not mean that every transport receptor follows one identical route, but it offers a physical explanation for how a dynamic protein network can create a selective pathway through a crowded nanoscale channel.

A sugar modification also changes how proteins interact

The second factor is O-linked beta-N-acetylglucosamine, usually shortened to O-GlcNAc. It is a post-translational modification in which a sugar group is attached to certain amino acids in a protein. FG nucleoporins are among the proteins that can carry this modification. The researchers found that reducing O-GlcNAc levels slowed the local motion of FG domains and also slowed nuclear import in their experimental system. This linked the chemical modification to both the material behaviour of the barrier and its transport function.

In purified protein experiments, glycosylation delayed the ageing of FG condensates and helped them remain mobile for longer. When the researchers combined glycosylated FG domains with importin-beta, the resulting material more closely reproduced the motion measured in living cells. The two factors appear to act together: the sugar groups reduce excessive protein-protein sticking, while importin-beta changes the pattern of contacts between FG chains. The findings show that the barrier’s physical properties are regulated by both molecular binding and chemical modification.

A route along the pore’s edge helps explain transport

The simulations also addressed a long-standing question about where transport occurs inside the nuclear pore. In many tubes and channels, flow is fastest through the centre, but earlier experiments suggested that transport receptors often move along the edge of the nuclear pore. Researchers had considered whether a biological plug in the centre might force cargoes toward the periphery. The new model suggests that a central plug is not required to explain the observed trajectories.

Instead, the arrangement of FG domains and importin-beta can produce a route that is enriched near the pore’s rim. The simulation used the degree of molecular connectivity measured in the live-cell experiments and predicted how the transport receptor would distribute itself within the channel. The results do not rule out every local structure or every transient obstruction, but they offer a physical explanation for why the main transport path can be closer to the edge. This helps reconcile earlier observations with the dynamic material properties measured in living cells.

Why the discovery matters beyond cell biology

Nuclear pores are involved in gene regulation, cell division, stress responses and the movement of viral and cellular molecules. Their function depends on maintaining a carefully balanced barrier. Changes in nucleoporin expression, modification or organisation have been linked to cancer, viral infection and neurodegenerative disease. The new study does not identify a treatment for any of those conditions, but it gives researchers a clearer physical framework for understanding how changes to pore proteins might affect molecular traffic.

The findings may also be relevant to other biological structures built from intrinsically disordered proteins. Such proteins do not fold into one rigid three-dimensional shape and can form dynamic assemblies that respond to their environment. Similar principles may apply to biomolecular condensates involved in gene expression, RNA processing or cellular stress. By combining live-cell fluorescence measurements with computational models, researchers can study how local chemical conditions influence protein motion in places where traditional structural methods struggle to capture rapid changes.

What the experiments do not yet establish

The work is a major advance in measuring the native barrier, but it has limits. The fluorescence measurements focused on selected regions of FG nucleoporins and on particular cell models. Nuclear pores contain many different nucleoporins, and their composition can vary between cell types and physiological states. The simulations simplify some molecular details so that the system can be studied at large scale. Their predictions are supported by imaging, but they do not replace direct measurements of every protein and every cargo in an intact living organism.

The study also does not show that altering O-GlcNAc or importin-beta would be a safe way to treat disease. These factors participate in many cellular processes, and changing them could have effects beyond nuclear transport. A molecular mechanism is not automatically a therapeutic target. The next steps include testing additional nucleoporins, examining how the barrier changes under stress or during cell division and studying whether disease-associated mutations alter the same dynamics in a measurable way.

The next questions for nuclear transport research

Future research can use the study’s measurement approach to examine how the barrier responds to different cargo sizes, transport receptors and chemical signals. Researchers may ask whether importin-beta’s peripheral enrichment changes when a cell increases nuclear import, whether different nucleoporins have distinct mobility profiles and how the pore adapts when the cell is under metabolic or inflammatory stress. These questions could help explain why the same nuclear pore architecture can support a wide range of transport tasks.

The broader lesson is that a living cell’s molecular machines are not static structures. Their function can depend on a carefully maintained material state that is continually adjusted by binding partners and chemical modifications. The nuclear pore is both a physical barrier and an active transport system. By showing how cellular context keeps its protein network fluid, the Nature study provides a more realistic picture of one of the cell’s most important gateways and a framework for investigating other dynamic protein assemblies.