LGALS3BP, also known as galectin-3-binding protein or 90K, is a heavily glycosylated protein that circulates in the blood and coats the surface of vesicles released by many tumor types. Previous work from the same Italian research group, based at the University G. d’Annunzio of Chieti-Pescara, had shown that LGALS3BP is abundantly enriched on neuroblastoma-derived extracellular vesicles and that elevated levels correlate with increased metastatic potential in preclinical models. The team had also demonstrated that an antibody-drug conjugate targeting secreted LGALS3BP can boost lymphocyte infiltration into tumors and synergize with anti-PD-1 immunotherapy in mouse models of the disease.

In the new study, the researchers asked a more fundamental question: does the presence of LGALS3BP change what these vesicles actually carry? To find out, they exploited an unusual experimental asset. The patient-derived neuroblastoma cell line hNB, isolated from a non-MYCN-amplified tumor that had metastasized to the neck of a three-year-old patient, naturally lacks LGALS3BP expression entirely, at both the messenger RNA and protein levels. Using lentiviral gene delivery, the team created a matched counterpart that stably re-expresses the protein at levels comparable to SK-N-AS, a neuroblastoma line that naturally produces high amounts of LGALS3BP. The result was an isogenic pair of cell lines differing essentially only in LGALS3BP status, an ideal platform for isolating the protein’s effects.

The first set of experiments focused on the tumor microenvironment, the cellular ecosystem that surrounds a tumor and often determines whether it can grow and spread. High-risk neuroblastoma is notorious for its immunologically cold, poorly infiltrated microenvironment, which limits the effectiveness of current immunotherapies. As a representative stromal component, the researchers used bone marrow-derived mesenchymal stem cells, or BMMSCs, key players in the bone marrow niche where metastatic neuroblastoma often takes hold. When these stromal cells were exposed to conditioned medium from LGALS3BP-expressing tumor cells, a multiplex Luminex assay measuring thirty soluble mediators revealed a broad shift in their secretory profile, with increased release of inflammatory factors such as IL-6, IL-8, and TNF-alpha, myeloid-related cytokines including M-CSF and IL-10, and the angiogenic driver VEGF-A.

Validation by ELISA confirmed a statistically significant increase in VEGF and a parallel rise in IL-6, both well-established pro-tumorigenic factors. Importantly, these molecules were virtually undetectable in the tumor-conditioned media themselves, confirming that the stromal cells, not the cancer cells, were their source. The team also detected the mature, hyperglycosylated form of LGALS3BP, roughly 100 kilodaltons in size, inside the recipient BMMSCs after exposure to LGALS3BP-containing medium, and this signal dropped markedly when the anti-LGALS3BP therapeutic antibody 1959 was added. Purified vesicles from LGALS3BP-positive cells produced the same result, suggesting that tumor-derived extracellular LGALS3BP physically associates with stromal cells.

The heart of the study, however, lies in the molecular analysis of the vesicles themselves. Extracellular vesicles were isolated by differential ultracentrifugation from the culture supernatants of both cell lines and profiled for their microRNA content using TaqMan array cards covering 94 vesicular miRNAs. Thirty-three showed differential abundance between LGALS3BP-positive and control vesicles, with a clear predominance of upregulation in the LGALS3BP-containing vesicles. In the donor cells, by contrast, the pattern was reversed: among 118 profiled miRNAs, 29 were differentially expressed, mostly downregulated in LGALS3BP-expressing cells. Seven miRNAs stood out as simultaneously enriched in vesicles and depleted in the cells that released them, a signature consistent with preferential sorting into the vesicular compartment.

These seven candidates, including miR-26a-5p, miR-27b-3p, miR-29a-3p, miR-103a-3p, miR-152-3p, miR-340-5p, and miR-374b-5p, were validated by highly sensitive stem-loop RT-PCR. When BMMSCs were then incubated with fluorescently labeled vesicles, confocal microscopy confirmed that the vesicles interacted with the recipient cells, and four of the seven miRNAs, specifically miR-340-5p, miR-27b-3p, miR-103a-3p, and miR-152-3p, were detected at increased levels inside the stromal cells after exposure to LGALS3BP-positive vesicles. Notably, the remaining three did not rise in recipients despite their enrichment in the vesicles, a discrepancy the authors interpret as evidence that vesicular enrichment does not automatically translate into increased abundance in recipient cells, and that additional factors govern the fate of vesicle-associated miRNAs after uptake.

In silico analysis of public neuroblastoma datasets added clinical context. miR-27b and miR-340 were significantly upregulated in tumor samples compared with normal controls, and predicted targets of the enriched miRNAs include PPARG, a master regulator of adipogenic differentiation and mesenchymal plasticity, and DNMT1, a key epigenetic enzyme. Because PPAR-gamma is a validated target of miR-27b, the authors speculate that increased delivery of this miRNA to bone marrow stromal cells could alter adipogenic programs and push the stroma toward a more tumor-supportive, activated state. The picture is nuanced, however, since miR-27b, miR-340, and miR-103a have each been reported to act as either tumor suppressors or promoters depending on context, targeting molecules such as PD-L1 and PTEN.

Proteomic analysis by nanoLC-MS/MS of the same vesicle preparations revealed a parallel story at the protein level. LGALS3BP-positive vesicles were selectively enriched in RNA-binding proteins, notably members of the serine/arginine-rich splicing factor family, and the splicing factor SRSF1 was validated by western blotting as more abundant in LGALS3BP-positive vesicles than in control vesicles, with no corresponding difference in whole-cell lysates. This finding is particularly intriguing because SRSF1 has recently been implicated in the selective sorting of miRNAs into tumor-derived extracellular vesicles, and its vesicular form appeared to migrate at a different apparent molecular weight than the intracellular protein, hinting at distinct post-translational modifications or processing. Patient dataset analyses reinforced the clinical relevance: SRSF1 expression was significantly higher in high-risk than in non-high-risk neuroblastoma in the SEQC cohort, and high SRSF1 levels were associated with poorer overall survival in the Kocak cohort.

The authors are careful to frame these findings as associative rather than causal. The miRNA profiling was exploratory, without correction for multiple testing, and some results derive from a single experiment. Whether LGALS3BP physically interacts with RNA-binding proteins such as SRSF1, alters endosomal trafficking, or shapes vesicle biogenesis pathways remains unknown, and functional studies and in vivo validation will be needed to determine whether the enriched cargo genuinely drives stromal reprogramming, metastatic niche formation, or therapeutic resistance. Nevertheless, the study broadens the portfolio of LGALS3BP, already under development as a biomarker and antibody-drug conjugate target, by suggesting that the protein may act not only as a passenger on tumor vesicles but as a marker of a fundamentally altered cargo program, one that could reshape how the youngest cancer patients’ tumors talk to their surroundings.

Subject of Research: The association of LGALS3BP with altered microRNA and protein cargo in neuroblastoma-derived extracellular vesicles and stromal cell modulation

Article Title: LGALS3BP is associated with selective cargo enrichment in neuroblastoma extracellular vesicles

Article References: Cela, I., Konstantinidou, F., Amodei, L., Mercatelli, A., Piermatteo, F., Santoro, S., Pece, A., Staropoli, M. C., Lamolinara, A., Capone, E., Gatta, V., & Sala, G. (2026). LGALS3BP is associated with selective cargo enrichment in neuroblastoma extracellular vesicles. Frontiers in Oncology, 16, Article 1966526. https://doi.org/10.3389/fonc.2026.1966526

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1966526

Keywords: neuroblastoma, LGALS3BP, extracellular vesicles, tumor microenvironment, microRNA, SRSF1, mesenchymal stem cells, proteomics, pediatric cancer, biomarker, antibody-drug conjugate, tumor-stroma communication

The scientific foundation for this work was laid two decades ago. In 2006, Shinya Yamanaka demonstrated that over-expressing just four transcription factors, now universally known as the Yamanaka factors, could reprogram mature cells all the way back to a pluripotent, embryonic-like ground state. The discovery earned a Nobel Prize and reshaped developmental biology. What followed was an even more tantalizing observation for aging researchers: when the Yamanaka factors are expressed only in short, carefully timed bursts, they can partially reprogram old cells into a more youthful condition, reversing multiple hallmarks of aging while the cells retain their specialized functions. The implication was profound. Aging, it seemed, does not have to be a one-way street, and even old cells retain a latent capacity for rejuvenation that modern biology might learn to unlock.

Yet the therapeutic promise of the Yamanaka factors has always been shadowed by a serious risk. Because these factors can drive cells all the way back to a pluripotent state, their over-expression raises the danger of cancer formation, and controlling exactly how far the reprogramming proceeds remains a formidable technical challenge. Alex Plesa, a scientist at the Wyss Institute and first co-author of the new study, framed the dilemma succinctly: in reprogramming, there is a fundamental tradeoff between rejuvenation and cell identity, and achieving one without losing the other requires walking a fine line. That tradeoff has motivated a worldwide search for alternative factors that might rejuvenate aging tissues in safer and more controllable ways, and it is precisely this search that the Harvard-led team set out to systematize.

Their first challenge was measurement. How do you score rejuvenation reliably enough to screen hundreds of candidate genes? The most widely used tool, the epigenetic aging clock, reads chemical modifications to the genome that accumulate with age. But as Plesa explained, that approach does not scale well to screening because of high costs and the difficulty of interpreting its widespread changes. Instead, the team reasoned that any genuine rejuvenation process must ultimately manifest itself in altered gene expression programs that restore more youthful cellular functions. Gene expression, they concluded, was the right readout. Working with Antonio del Sol, an Ikerbasque Research Professor at CIC bioGUNE-BRTA in Spain, and postdoctoral fellow Sascha Jung, the study’s co-first author, they built a machine learning algorithm trained on transcriptomic profiles spanning multiple aging processes in human skin fibroblast cells from donors across a wide age range.

The resulting tool, which the team calls a transcriptomic aging model, functions something like a set of odometers for different parts of a car. It tracks changes in the expression of genes involved in distinct biological processes and integrates them into a single accurate overview of a cell’s age and health. Trained on gene expression data from 464 fibroblast samples, the model could pinpoint the transcriptomic age of skin cells from individual donors. Crucially, it also proved sensitive to real biological change: when the researchers treated fibroblast cell lines derived from donors aged 55, 65, and 79 with experimental lifespan-altering interventions, including the Yamanaka factors, the predictor accurately registered the resulting shifts in transcriptomic age. The team now had a scorable, quantitative yardstick for rejuvenation.

With that yardstick in hand, the researchers assembled a focused library of genes predicted to influence aging processes and over-expressed each one in aged fibroblast cell lines, hunting for candidates that could shift the cells’ aging score toward a more youthful value. One gene rose decisively to the top: SRSF1, which encodes an RNA processing factor. Over-expression of SRSF1 reduced features corresponding to several recognized hallmarks of aging, including cellular senescence, the state in which damaged cells stop dividing but refuse to die; proteostatic stress, a disruption of the normal balance and turnover of proteins; and inflammaging, the chronic low-grade inflammation that accompanies aging. The rejuvenation effects also registered on the conventional epigenetic aging clock and extended to the restoration of DNA repair programs, suggesting a broad and genuine reversal of age-associated molecular damage.

Just as important as what changed was what did not. The rejuvenated fibroblasts maintained their identity throughout the intervention. They kept their characteristic spindle-shaped morphology, continued expressing the genes that encode typical fibroblast functions, and did not activate the pluripotency programs that Yamanaka factors switch on. In other words, SRSF1 appears to rejuvenate cells by working through machinery that all cells already use, rather than overriding their identity, which is exactly the property that makes it a safer candidate for therapy than the pluripotency factors. As Plesa put it, the team worked on the premise that over-expressing a single gene with validated rejuvenation outcomes could eventually be advanced into an organ- or tissue-specific gene therapy if combined with the right delivery system.

The decisive test came in living animals. The researchers packaged the SRSF1 gene into a liver-tropic adeno-associated virus, or AAV, a widely used gene delivery vehicle, and treated aged mice. The gene primarily accumulated in the animals’ livers, and the results were remarkable on two fronts. First, the livers themselves showed signs of rejuvenation. Second, and more strikingly, the benefits were systemic: SRSF1-treated animals grew frail more slowly than controls and lived longer, with median lifespan extended by 14 percent. That a single RNA processing factor, delivered mainly to one organ, could slow the aging of the whole organism positions SRSF1 as what the researchers call a geroprotective target, and it hints that rejuvenation programs initiated in one tissue may be communicated and amplified throughout the body.

Exactly how SRSF1 accomplishes its multifaceted effects remains an open and actively pursued question. The protein’s best-known role is that of an RNA splicing factor, one that helps include alternative exon sequences in a subset of mature messenger RNA molecules. These alternatively spliced mRNAs are then translated by ribosomes into proteins with altered functions, effectively allowing one gene to produce a repertoire of protein variants. The team observed a strong trend of alternatively spliced mRNAs following SRSF1 over-expression, but as Plesa acknowledged, they are only just beginning to understand how these splicing changes bring about the multi-faceted rejuvenation program in tissues. Unraveling that mechanism could reveal additional targets and open entirely new avenues for intervention.

For George Church, the Wyss Institute core faculty member and Professor of Genetics at Harvard Medical School who led the study, the work represents a step toward a long-standing ambition: devising clinically useful de-aging strategies by identifying genes that, when engineered for under- or over-expression in aging target tissues, extend human health-span. In his view, aging research is transiting an inflection point where exponential progress in biotechnology is beginning to impact the core aging processes that would otherwise kill 90 percent of us. Significant hurdles remain before SRSF1 could reach the clinic, particularly around delivery and dosing, and further research into its systemic effects is clearly needed. But the study demonstrates something once thought impossible: that a precisely targeted gene, measured against a rigorous transcriptomic yardstick, can turn back key aspects of the aging clock in cells and in a living mammal, without sacrificing what those cells are meant to be.

Subject of Research: Gene expression reprogramming to reverse cellular and organismal aging using the RNA splicing factor SRSF1

Article Title: Reversing age by reprogramming gene expression networks

Article References: Reversing age by reprogramming gene expression networks. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: aging, rejuvenation, SRSF1, gene expression, transcriptomic aging clock, Yamanaka factors, cellular senescence, RNA splicing, gene therapy, lifespan extension, machine learning, geroprotection

The study began with a computational foundation. The researchers built an interpretable transcriptomic age predictor specifically calibrated in human fibroblasts, the connective tissue cells that have long served as a workhorse of aging research. Drawing on transcriptomes from 133 donor lines spanning ages 1 to 96 years, the model distilled aging down to 292 genes whose activity tracks the passage of biological time. These genes cluster into recognizable biological processes, including WNT signaling, translation initiation, actin organization, histone methylation, and endoplasmic reticulum stress. Importantly, the model captured distinct kinetic patterns of aging: some processes, such as ER quality control and translation initiation, decline steadily and continuously with age, while others, like histone methylation and actin polymerization, respond late and exponentially, and still others, such as sodium transport, follow a biphasic trajectory. This architecture gave the team a quantitative yardstick for biological age that could be applied to any fibroblast sample.

With the age predictor in hand, the researchers turned to an unbiased search for factors capable of turning back the transcriptomic clock. They assembled a library of 95 plasmids for overexpression screening: 89 candidate genes prioritized through a network-based approach that combined protein-protein interaction data from the STRING database with the strength of age-associated expression changes, plus five known aging modulators including ATG5, CXCR2, HES1, KAT7, and the OSKM reprogramming cocktail, along with a blue fluorescent protein control. Each candidate was introduced into aged fibroblasts, and the resulting transcriptomes were scored by the age predictor. The screen’s standout hit was SRSF1, a well-known splicing factor that binds RNA and controls how messenger RNA transcripts are cut and pasted together before they are translated into protein. When SRSF1 was induced in aged fibroblast lines, the cells’ predicted transcriptomic age dropped by a median of 16.2 years after just three days of expression, a striking and technically reproducible rejuvenation effect.

What makes this result particularly notable is what did not happen. Partial reprogramming with OSKM factors is famously accompanied by dramatic morphological changes: fibroblasts lose their characteristic spindle shape, shrink in size, and drift toward pluripotency programs, raising the specter of uncontrolled cell growth. In the new study, OSKM expression produced exactly these changes by day three of induction, and even short one-to-three-day inductions left lasting positive age shifts of more than 37 to 54 years in some donor lines after the factors were withdrawn. SRSF1-expressing cells, by contrast, retained their spindle-shaped fibroblast morphology and cell body area throughout induction, showed no sustained age shifts after doxycycline withdrawal, and did not activate pluripotency programs. The rejuvenation was also confirmed with an orthogonal measurement: the Skin and Blood DNA methylation clock, an entirely independent biomarker of aging based on chemical tags on the genome rather than gene expression, also registered reduced age in SRSF1-treated cells.

At the functional level, SRSF1 induction produced a multidimensional rejuvenation across molecular hallmarks of aging. In aged fibroblasts, the factor reduced senescence, the state of permanent growth arrest that accumulates in aging tissues and secretes inflammatory signals; it eased proteostatic stress, the burden of misfolded and damaged proteins; and it dampened inflammatory signaling pathways. When delivered as synthetic mRNA rather than DNA, SRSF1 significantly reduced senescence-associated beta-galactosidase activity, a classic marker of cellular aging, in both young and old donor cell cohorts. The factor also improved stress resilience: aged fibroblasts carry roughly 64 percent higher levels of reactive oxygen species after a hydrogen peroxide challenge than young cells, and SRSF1 treatment trended toward reducing this oxidative burden. Collagen synthesis, which declines modestly with age, increased by about 14 percent in aged cells treated with SRSF1, and the treated cells showed a tendency toward faster closure of scratch wounds in vitro, a measure of regenerative capacity.

The critical question was whether these cellular effects would translate into benefits in living animals. To test this, the team used inducible viral delivery, packaging the Srsf1 gene into adeno-associated virus vectors under the control of a doxycycline-activated Tet-On system, allowing them to switch expression on and off in aged mice. The results were remarkable: inducible SRSF1 expression reduced the progression of frailty, a composite measure of age-related decline in strength, endurance, and health, by 44 percent, and was associated with a 14 percent extension of median lifespan, with 9 to 10 mice per group. Sustained frailty onset, defined as a frailty index exceeding 0.18, was delayed by approximately 123 days in the SRSF1-treated animals. In skin wound-healing experiments, wounds in aged mice closed significantly faster with SRSF1 expression, reaching a partial restoration of the regenerative capacity seen in young animals, although the effect depended on the delivery regimen and was not observed with all expression strategies.

Molecular analysis of the treated mice revealed that SRSF1’s benefits extended across tissues. In the liver, SRSF1 expression reduced epigenetic age acceleration as measured by an intervention-optimized DNA methylation clock, shaving off about half a year of apparent age. Gene set enrichment analysis across tissues showed that SRSF1 modulated pathways involved in bioenergetics, immune signaling, and cellular homeostasis, and many of these processes were classified as age-opposing when compared against the Tabula Muris Senis reference atlas of mouse aging. Strikingly, a cross-species comparison revealed shared signatures between SRSF1-treated human fibroblasts and mouse liver tissue, including reduced stress response, reduced inflammation, and altered cell adhesion and cell cycle programs, suggesting that the factor acts through conserved longevity pathways rather than species-specific mechanisms.

The evolutionary conservation of this pathway received direct experimental support from studies in the roundworm Caenorhabditis elegans. The worm homolog of SRSF1 is a gene called rsp-3, and when the researchers knocked it down with RNA interference in long-lived raga-1 mutant worms, a strain in which the nutrient-sensing mTOR pathway is dampened, the worms’ characteristic lifespan extension was suppressed. The result, confirmed in independent repeats with 120 to 159 worms per group, places SRSF1 squarely within one of the most intensively studied longevity networks in biology, the same nutrient-sensing axis targeted by rapamycin and dietary restriction. It also connects the new findings to an older body of evidence showing that splicing factor expression changes with age in both mice and humans, and that variations in splicing factor transcripts correlate with lifespan.

SRSF1 is not an obvious anti-aging target. The factor has a well-documented dual personality: it is essential for normal cell function, governing progenitor-specific alternative splicing in epithelial tissues and acting as a master switch of gene expression in the immune system, but it is also a known proto-oncogene whose overexpression is associated with cancer cell biology. The authors and several coauthors are listed as inventors on a patent application covering the therapeutic use of SRSF1 and related factors, and two of them are cofounders of an early-stage company developing therapeutics for age-related diseases, disclosures that underscore the translational ambition behind the work. The tension between geroprotection and oncogenesis will need careful resolution, and the mouse data, while encouraging, come from relatively small groups and a defined delivery system.

Nevertheless, the study represents a conceptual advance in how rejuvenation targets are found. Rather than starting from the Yamanaka factors and trying to make them safer, the team started from a quantitative definition of the aged state and screened for any gene that could reverse it while preserving cell identity. The answer, an RNA-binding splicing factor rather than a transcription factor, suggests that the aging transcriptome may be held in an aged configuration partly at the level of RNA processing, and that correcting the splicing machinery can ripple outward to restore youthful patterns of gene expression, protein homeostasis, and tissue function. If the findings hold up in larger animal studies and can be delivered safely, SRSF1 and the screening platform that uncovered it could open a new chapter in the search for interventions that extend not just lifespan, but the healthy years of human life.

Subject of Research: Identification of the splicing factor SRSF1 as a conserved regulator of aging through transcriptomic reprogramming screens

Article Title: Transcriptomic reprogramming screen identifies SRSF1 as a regulator of aging

Article References: Plesa, A. M., Jung, S., Wang, H. H., Omar, F., Shadpour, M., Choy Buentello, D., Perez-Matos, M. C., Horwitz, N., Cai, G., Ngian, Z. K., Magalhaes, C., Wagers, A. J., Mair, W. B., del Sol, A., & Church, G. M. (2026). Transcriptomic reprogramming screen identifies SRSF1 as a regulator of aging. Nature Aging, 6(10), 2057-2073. https://doi.org/10.1038/s43587-026-01234-8

Image Credits: AI Generated

DOI: 10.1038/s43587-026-01234-8

Keywords: SRSF1, aging, transcriptomic clock, partial reprogramming, splicing factor, senescence, geroprotection, lifespan extension, C. elegans, frailty, gene expression, epigenetic clock

The nucleolus is organized into three subcompartments: the fibrillar centers, the dense fibrillar component and the granular component, each hosting a distinct stage of ribosomal RNA production and processing. Recent work has shown that these subcompartments are maintained as coexisting liquid phases, and that a pH gradient across them helps set up the biochemical conditions each phase requires. Yet while pH is recognized as a key variable controlling condensate dynamics and function, how cells actually control pH inside these compartments under physiological and pathological conditions has remained poorly understood. The Xiamen team set out to close that gap, focusing on a structure that appears when nucleolar transcription is disrupted.

When rRNA synthesis is interrogated by stress, whether from DNA damage or from drugs that block transcription, the nucleolus undergoes a characteristic reorganization. Some of its components segregate into cap-like structures at the nucleolar periphery, structures that were first described more than sixty years ago in cells treated with the chemotherapeutic agent actinomycin D. Among these are the so-called dark nucleolar caps, or DNCs, marked by proteins such as SFPQ and DDX18. The researchers discovered that under specific stressed conditions, SRSF1, a canonical splicing factor that normally resides in nuclear speckles, is recruited into these DNCs. Using super-resolution imaging and knock-in fluorescent tags in human cells, they tracked this redistribution in detail, finding that it occurs not only in one cell line but across multiple human cell types, including 293T, HCT 116 and Huh7 cells, after ultraviolet treatment followed by a recovery period.

Recruitment, however, is not random. Through immunoprecipitation combined with mass spectrometry, the team identified a molecular interaction between SRSF1 and DDX18, a DEAD-box RNA helicase that localizes to the granular component and the DNCs. When DDX18 was depleted with short hairpin RNA, SRSF1’s ability to accumulate in the dark nucleolar caps was impaired, indicating that this interaction is partially required for SRSF1’s stress-induced localization. The finding adds a new dimension to the biology of SR proteins, a family of splicing factors already known to redistribute to nuclear stress bodies and segregated nucleolar components in response to DNA damage, and it suggests that the nucleolar caps act as staging grounds where splicing machinery performs an unexpected, non-splicing job.

That job, the study shows, is pH management. Using live-cell imaging with the ratiometric pH-sensitive dye BCECF-AM and a genetically encoded pHluorin biosensor, the researchers measured the pH inside individual nucleolar subcompartments with remarkable precision. They found that the pH microenvironments of the granular component, the dense fibrillar component and the fibrillar centers remain largely unchanged upon ultraviolet exposure, but the dark nucleolar caps behave differently. When SRSF1 was knocked down, the pH homeostasis of the DNCs was disrupted, and this disturbance cascaded into a failure of the whole organelle to recover. Cells lacking SRSF1 could not properly restore the nucleolar multiphase architecture after stress, and functional readouts of nucleolar health deteriorated accordingly.

The functional consequences were striking. In control cells subjected to ultraviolet treatment or actinomycin D and then allowed to recover, the granular component protein NPM1 repartitioned back into the nucleolus, nascent RNA synthesis resumed as measured by 5-ethynyl uridine incorporation, and precursors of ribosomal RNA detected by probes against the 5-prime external transcribed spacer returned to normal levels. In SRSF1-depleted cells, all of these recovery markers were blunted. The team also observed elevated levels of gamma-H2AX, a marker of DNA damage, in the stressed cells lacking SRSF1, consistent with the idea that impaired nucleolar recovery leaves the genome more vulnerable. Rescue experiments in which shRNA-resistant SRSF1 was reintroduced restored DNC pH and nucleolar function, confirming the specificity of the effect.

The mechanistic heart of the paper lies in SRSF1’s arginine/serine-rich, or RS, domain. Because of its high positive charge, this domain is accountable for alkalizing the DNC microenvironment. When the researchers mutated the charged residues of the RS domain to uncharged alanines, the mutant SRSF1 failed to modulate the pH of the dark nucleolar caps even though it could still localize there. Importantly, the pH-modulating activity of SRSF1 was shown to operate independently of its canonical role in splicing regulation, separating this newly discovered function from the protein’s day job in RNA processing. The work resonates with a growing body of evidence that biomolecular condensates can sustain electrochemical gradients and modulate reactions at their interfaces, and with theoretical work suggesting that charge neutralization allows condensates to maintain pH gradients at equilibrium.

Perhaps the most translationally exciting result is the demonstration that the pH-restoring function can be mimicked by a synthetic molecule. The team designed arginine-rich dipeptide derivatives based on the SRSF1 RS domain, varying their length and net charge, and expressed them in stressed cells. These synthetic dipeptides localized to the nucleolus, enriched in the DNC region as measured against the DDX18 fluorescence mask, and safeguarded the nucleolus from pH and functional disturbance. In cells treated with actinomycin D and then allowed to recover, the arginine-arginine dipeptides preserved nucleolar size, maintained 5-prime ETS and rRNA levels, and protected the multiphase organization of the organelle. In other words, a minimal positively charged peptide was sufficient to stand in for the splicing factor’s pH-buffering role, offering a proof of principle that nucleolar pH can be pharmacologically tuned.

The implications extend well beyond basic cell biology. The nucleolus is increasingly viewed as a therapeutic target in human disease, and its disorganization is a hallmark of stress responses in cancer chemotherapy, since many chemotherapeutic drugs inhibit ribosome biogenesis at various levels. Nucleolar dysfunction has also been implicated in neurodegenerative conditions: dipeptide repeat proteins produced by the C9orf72 repeat expansion, the most common genetic cause of amyotrophic lateral sclerosis and frontotemporal dementia, are known to disturb biomolecular phase separation and disrupt nucleolar function. A pathway that actively restores nucleolar pH homeostasis after stress, and that can be enhanced with simple arginine-rich peptides, suggests new angles for intervening in diseases where condensate chemistry goes wrong. It also reframes SRSF1, long studied as a splicing factor and oncoprotein, as a multifunctional regulator whose charged domains carry out electrochemical work inside membrane-less organelles.

Technically, the study stands out for the rigor of its measurements. The team combined structured illumination microscopy with knock-in fluorescently tagged endogenous proteins, ratiometric live-cell pH imaging validated against standard buffers, fluorescence recovery after photobleaching to assess condensate dynamics, quantitative RNA FISH, immunostaining and transcriptome-wide analyses. Mass spectrometry datasets were deposited to the ProteomeXchange Consortium and RNA-seq data to the Gene Expression Omnibus, with source data provided for every figure. By connecting a single charged protein domain to the electrochemical state of a nucleolar subcompartment, and that state to the recovery of an entire organelle, the work delivers a mechanistic model in which stress-recruited SRSF1 alkalizes the dark nucleolar caps, stabilizes the pH gradient of the multiphase nucleolus and enables the cell’s ribosome factory to resume production. It is a vivid demonstration that the chemistry of condensates, not merely their composition, governs how cellular compartments endure and recover from adversity.

Subject of Research: pH regulation of nucleolar condensates by the splicing factor SRSF1 during stress recovery

Article Title: SRSF1 modulates the dark nucleolar cap pH to restore nucleolar integrity and function

Article References: Ma, S., Guo, J., Zhan, X., Wu, F., Yang, S., Huang, C., & Wang, B. (2026). SRSF1 modulates the dark nucleolar cap pH to restore nucleolar integrity and function. Nature Chemical Biology. https://doi.org/10.1038/s41589-026-02316-9

Image Credits: AI Generated

DOI: 10.1038/s41589-026-02316-9

Keywords: nucleolus, SRSF1, pH gradient, biomolecular condensates, dark nucleolar caps, DDX18, RS domain, ribosome biogenesis, cellular stress, phase separation, arginine-rich dipeptides, chemical biology