A new molecular map of infected human cells shows how influenza A virus takes apart and repurposes paraspeckles — oddly named nuclear structures whose everyday job is holding certain genes in check — to help itself replicate. The study, published July 20, 2026 in Nature Microbiology, is one of the first to trace virus-host protein contacts inside living, infected cells rather than in test-tube extracts, and it turns up a genuinely new piece of how flu takes over a cell.
Primary source: Kotova, I., Mühlberg, L., Gilep, K. et al. “Mapping in-cell protein contact sites reveals hijacking of paraspeckles during influenza A virus infection.” Nature Microbiology (2026).
What are paraspeckles, and why does “hijacking” one matter?
Paraspeckles are membrane-less nuclear bodies built around a long non-coding RNA scaffold (NEAT1) and a set of RNA-binding proteins. They are not a fixed, permanent structure — cells assemble and disassemble them in response to stress, and their main known role is sequestering specific proteins and RNAs to fine-tune gene expression. Because they are dynamic and stress-responsive, paraspeckles sit in a category of cell biology — biomolecular condensates — that has drawn intense research interest over the past decade, but their relationship to viral infection has been comparatively under-studied.
That is what makes this finding notable. Rather than simply blocking or evading a host defense, influenza A virus (IAV) appears to actively re-engineer an existing, non-viral cellular compartment and repurpose the components it releases when the structure comes apart. Finding a specific, mechanistic route by which a virus dismantles a condensate to its own advantage — rather than as a side effect of infection stress — is a new addition to the known playbook of how influenza remodels its host cell.
How the researchers mapped it: contact sites inside living cells
The team, based at EMBL Hamburg and EMBL Heidelberg with collaborators including the Leibniz-Forschungsinstitut für Molekulare Pharmakologie, Charité – Universitätsmedizin Berlin, the Leibniz Institute of Virology, and the University of Veterinary Medicine Hannover, used in-cell cross-linking mass spectrometry to capture protein-protein contacts as they actually exist inside intact, infected human cells — combined with structural modeling and follow-up functional assays to test what the captured contacts meant.
Cross-linking mass spectrometry works by chemically “freezing” proteins that are physically touching at the moment of treatment, then identifying which pairs were linked. Doing this in-cell, rather than on purified proteins or lysed extracts, is the methodological advance that matters here: it captures contacts in their native spatial and temporal context, including transient interactions that would be lost or artefactually rearranged once a cell is broken open. That is precisely the kind of interaction older interactome methods tend to miss, and it is what let the team follow a chain of events happening inside the nucleus during infection rather than inferring it indirectly.
What the map revealed
Two separate storylines emerged from the same in-cell contact map. In the cytoplasm and secretory pathway, the researchers identified host factors involved in producing distinct glycoforms of the viral hemagglutinin protein as it moves through the endoplasmic reticulum and Golgi — relevant to how the virus’s surface protein is built and finished before new virus particles are released.
In the nucleus, the map showed progressive paraspeckle disassembly across multiple infected cell lines, and traced it to several converging viral mechanisms working together rather than one single cause:
- The viral nucleoprotein and the non-structural protein NS1 were found engaging directly with paraspeckle proteins;
- the viral PA-X endonuclease was linked to degradation of the long non-coding RNA scaffold that holds paraspeckles together; and
- inhibition of host RNA polymerase II by the virus further drove paraspeckle disruption.
The net effect, per the paper, is that dismantling paraspeckles releases host factors that were otherwise sequestered inside them — and those released factors go on to enhance influenza A replication. In other words, the virus does not merely disable an antiviral defense; it converts a normal gene-regulatory structure into a source of components useful to itself.
Why the method itself is part of the story
For a research-administration and infrastructure audience, the notable part of this paper is not only the biology but the demonstration that in-cell, native-context interactome mapping can resolve mechanistic detail that lysate-based proteomics and yeast two-hybrid screens have historically struggled to reach. Structural cross-linking mass spectrometry paired with computational modeling is increasingly used to bridge the gap between “these two proteins interact” and “here is the physical architecture and functional consequence of that interaction, inside a living cell, during an active infection.” Reproducibility of that kind of large, structurally-annotated interaction dataset depends on the same computational-methods discipline — documented pipelines, versioned software, and shared reference data — that underlies CASRAI’s own FAIR4RS and reproducible-computational-research work.
The open-data angle
Structural interactome studies of this kind typically fall under journal data-availability requirements that call for depositing the underlying mass-spectrometry and structural datasets in a public, community repository so other groups can re-analyze the raw contact data rather than take the paper’s conclusions on faith. We were not able to confirm the exact repository name or accession number for this specific dataset from the sources available at the time of writing, so we are not stating one here — but the general practice of pairing a data availability statement with actual repository deposition, rather than a promise to share “on request,” is exactly the kind of infrastructure norm that turns a single high-profile finding into something the wider field can verify and build on. Readers who want the primary confirmation should check the paper’s own Data Availability section directly.
What’s next
The paper’s authors frame this as a first systematic in-cell map of influenza-host contact sites rather than a complete picture — the same approach is a natural fit for other RNA viruses that interact with nuclear condensates, and for testing whether blocking the specific paraspeckle-disassembly steps identified here (rather than the process more broadly) could blunt viral replication without disrupting the paraspeckle’s normal regulatory role in uninfected cells.
Frequently asked questions
What is a paraspeckle?
A paraspeckle is a membrane-less nuclear body built around the long non-coding RNA NEAT1 and a set of RNA-binding proteins. Cells assemble and disassemble them dynamically, mainly to sequester specific proteins and RNAs and adjust gene expression, including under cellular stress.
What did the study actually find?
Using in-cell cross-linking mass spectrometry, structural modeling, and functional assays, researchers found that influenza A virus infection drives progressive disassembly of paraspeckles through several converging viral mechanisms, releasing host factors that go on to enhance viral replication.
How is this different from earlier virus-host interaction studies?
Most prior interactome studies rely on purified proteins or cell lysates, which can lose or distort transient, spatially-organized contacts. Mapping contacts inside intact, infected cells preserves that native context, which is what allowed the team to trace a multi-step mechanism inside the nucleus.
Where was this research done?
The work was led by researchers at EMBL Hamburg and EMBL Heidelberg, with collaborators at the Leibniz-Forschungsinstitut für Molekulare Pharmakologie, Charité – Universitätsmedizin Berlin, the Leibniz Institute of Virology, and the University of Veterinary Medicine Hannover, published in Nature Microbiology in July 2026.
