Stress-Triggered SMN Droplets Act as a Rescue System for Misplaced Splicing Machinery

Every cell in the body depends on a precise division of labor between its nucleus and cytoplasm, and few molecular workflows illustrate this better than the assembly of the spliceosome, the machine that removes intervening sequences from messenger RNA. A key component of this machinery, the survival motor neuron protein SMN, has long been known for its role in the nuclear Cajal bodies, where it shepherds the final maturation of small nuclear ribonucleoproteins, or UsnRNPs. Now a team at the Institute of Genetics of the Rheinische Friedrich-Wilhelms Universität in Bonn, led by Oliver J. Gruss with co-first authors Yannick L. Riedel and Jessica Dressler, reports that SMN does something remarkable in the cytoplasm as well. When cells experience molecular crowding stress, cytoplasmic SMN rapidly condenses into droplets and filamentous structures that capture UsnRNPs which have strayed out of the nucleus, holding them until the cell can dispose of them safely.

The study, published open access in Cellular and Molecular Life Sciences, describes these cytoplasmic SMN condensates as S-bodies. Under stress conditions such as osmotic challenge, oxidative stress induced by arsenite, or reducing stress induced by dithiothreitol, human lung fibroblasts responded within minutes by assembling these structures along microtubules. Live-cell imaging of cells stably expressing a fluorescently tagged SMN showed that the process is strikingly fast and reversible: when mechanical stress was applied, condensates appeared within seconds, and when osmotic stress was lifted, the assemblies reorganized over the course of recovery. The researchers used indirect immunofluorescence, fluorescence in situ hybridization, and super-resolution STORM microscopy to map the molecular composition and dynamics of these bodies in detail.

What makes S-bodies particularly intriguing is their selectivity. The team showed that they exclude G3BP1, a defining marker of classical stress granules, demonstrating that S-bodies are not simply another flavor of generic stress condensate. Instead, they specifically sequester UsnRNPs that have been mislocalized to the cytoplasm by stress. The assembly of S-bodies also proved independent of ongoing UsnRNP biogenesis: blocking translation with cycloheximide or transcription with NVP2 did not prevent their formation, indicating that the condensation machinery responds to the physical state of the cell rather than to the flux of newly made spliceosomal components.

The recovery phase revealed a second, equally elegant act. As stressed cells returned to normal conditions, the S-bodies did not simply dissolve. Instead, they underwent a microtubule-dependent reorganization into split condensates containing both SMN and CLNS1A, a protein known from the PRMT5 complex that acts earlier in the UsnRNP assembly pathway. The researchers named these dual-compartment structures Janus bodies, after the two-faced Roman god, because they bridge two distinct stages of the assembly line. Within these Janus bodies, SMN colocalized with its interaction partners Gemin2 and DDX20, with methylated Sm proteins, and with the hypermethylated cap structures of UsnRNAs, suggesting that the condensates serve as processing hubs where captured UsnRNPs are evaluated and cleared.

To understand the biophysical basis of this behavior, the authors draw on the concept of liquid-liquid phase separation, or LLPS, the process by which mixtures of proteins and RNAs demix into dense condensates within the cell. Nuclear SMN is already known to condense in Cajal bodies, where it promotes late maturation steps of the spliceosomal machinery. The new work extends this principle to the cytoplasm and assigns it a protective function: rather than participating in assembly, cytoplasmic SMN condensation acts as a buffer, physically sequestering misplaced UsnRNPs so that they cannot interfere with cytoplasmic processes or form aberrant aggregates. The biphasic nature of the response, first sequestration in S-bodies and then resolution through Janus bodies, provides a complete lifecycle for the stress response.

The most clinically resonant finding concerns spinal muscular atrophy, or SMA, a devastating neuromuscular disease caused by loss or deficiency of functional SMN. SMA arises because humans carry a duplicated SMN gene, SMN2, which produces predominantly a truncated protein, SMND7, that lacks exon 7 and is unstable. In cellular models of the disease, including mouse embryonic fibroblasts carrying a humanized SMN2 allele, the researchers found that S-bodies failed to assemble and that cytoplasmic UsnRNPs were not cleared after stress. This failure identifies an impaired tolerance to molecular crowding as a previously unrecognized hallmark of SMA, extending the disease picture beyond the well-established defects in UsnRNP biogenesis and motor neuron survival.

Importantly, the team tested whether pharmacological intervention could restore this protective function. Risdiplam, a small molecule approved for the treatment of SMA that splices SMN2 transcripts productively, was applied to the humanized mouse embryonic fibroblasts. Western blot analysis confirmed that a 24-hour treatment restored SMN expression levels in these cells, raising the possibility that reconstituting sufficient SMN protein also reconstitutes the cytoplasmic condensation response. While the study focuses on cellular models rather than patient neurons, the result suggests that some benefits of SMN-restoring therapies may include the reestablishment of this stress-buffering quality control system, a hypothesis that will require testing in disease-relevant contexts.

The work also reframes how scientists think about the relationship between condensates and the cytoskeleton. The dependence of both S-body maintenance and the transition to Janus bodies on microtubules indicates that phase separation and cytoskeletal transport are not parallel, independent systems but are mechanistically coupled. Microtubules appear to provide the tracks along which the condensates organize, mature, and redistribute during recovery. This coupling may ensure that captured UsnRNPs are routed to specific cellular locations for processing rather than being scattered randomly through the cytoplasm, adding a layer of spatial control to what might otherwise be a simple sequestration mechanism.

Broader implications reach into several active areas of cell biology. The finding that distinct stressors, including osmotic, oxidative, reducing, and mechanical stress, all trigger S-body formation suggests a general response to conditions that perturb molecular crowding. Because UsnRNPs contain uridine-rich small nuclear RNAs, their cytoplasmic accumulation could in principle seed unwanted RNA interactions, and the S-body system may represent an evolved safeguard against such molecular misbehavior. The study also connects to the growing literature on membraneless organelles, stress granules, processing bodies, and U bodies, positioning S-bodies as a distinct member of this family with a dedicated cargo and a defined lifecycle.

For the SMA research community, the identification of condensation defects as a disease phenotype opens new avenues for biomarker development and therapeutic evaluation. If the ability to form S-bodies and clear cytoplasmic UsnRNPs correlates with clinical outcomes, assays based on this cellular response could complement existing measures of SMN protein levels. More fundamentally, the work exemplifies how a protein famous for one role, nuclear assembly of the splicing machinery, can moonlight in an entirely different compartment under stress, and how the failure of that second role may contribute to disease. As the field of biomolecular condensates matures, studies like this one demonstrate that these droplets are not curiosities of physical chemistry but essential, regulated, and clinically relevant machines of cellular housekeeping.

Subject of Research: Stress-induced cytoplasmic SMN condensates that sequester and clear mislocalized UsnRNPs and their failure in spinal muscular atrophy

Article Title: Biphasic SMN condensates safeguard stress-induced cytoplasmic mislocalization of splicing RNPs

Article References: Riedel, Y. L., Dressler, J., Schilling, M. T., & Gruss, O. J. (2026). Biphasic SMN condensates safeguard stress-induced cytoplasmic mislocalization of splicing RNPs. Cellular and Molecular Life Sciences, 83(1), Article 350. https://doi.org/10.1007/s00018-026-06459-9

Image Credits: AI Generated

DOI: 10.1007/s00018-026-06459-9

Keywords: SMN, spinal muscular atrophy, UsnRNP, liquid-liquid phase separation, stress granules, microtubules, Cajal bodies, CLNS1A, spliceosome, Risdiplam, molecular crowding, condensates