Rutin Disrupts hnRNP L Condensates to Rewire Fibrotic Metabolism in Pulmonary Fibrosis

Pulmonary fibrosis is one of medicine’s most stubborn progressive diseases, a condition in which the delicate air sacs of the lung are slowly smothered by scar tissue, leaving patients with a median survival of only three to five years after diagnosis. At the heart of this process lie fibroblasts, the ordinary structural cells of the lung, which transform into myofibroblasts—scar-forming cells that churn out extracellular matrix, stiffen lung architecture, and progressively rob the organ of its ability to exchange oxygen. Only two antifibrotic drugs, nintedanib and pirfenidone, are currently approved, and while both can modestly slow the decline in forced vital capacity, neither halts nor reverses the disease, and both carry prominent gastrointestinal side effects. Against this grim backdrop, a team of researchers at Shandong Medical and Pharmaceutical University has now reported that a humble plant flavonoid—rutin, the glycoside abundant in buckwheat, apples, and licorice—may interfere with a previously unrecognized molecular circuit driving fibrogenesis, one that connects RNA splicing, cellular metabolism, and epigenetic regulation in a single self-reinforcing loop.

The new study, published in Frontiers in Pharmacology, set out to move beyond rutin’s well-documented antioxidant and anti-inflammatory reputation and identify a concrete molecular target. Using TGF-β1-stimulated MRC-5 human fetal lung fibroblasts and bleomycin-induced pulmonary fibrosis in C57BL/6 mice, the investigators first confirmed that rutin inhibits fibroblast-to-myofibroblast differentiation, suppresses myofibroblast proliferation and migration, and reduces extracellular matrix deposition. Notably, the half-maximal inhibitory concentration of rutin for TGF-β1-activated cells was roughly 60 micrograms per milliliter at 24 hours, dramatically lower than the approximately 250 micrograms per milliliter measured in quiescent cells, indicating a selective sensitivity of the pathological phenotype rather than indiscriminate cytotoxicity. In mice, oral rutin at 10 and 20 milligrams per kilogram improved forced vital capacity, dynamic lung compliance, and total lung capacity, while microCT imaging revealed reduced honeycomb lung, preserved alveolar structure, and thinned alveolar walls. Histological staining with hematoxylin and eosin and Masson’s trichrome confirmed diminished collagen deposition, and Western blot analysis showed downregulation of fibroblast activation protein 1, vimentin, collagen I, and α-SMA—the canonical molecular signatures of myofibroblast activation.

The mechanistic hunt then zeroed in on heterogeneous nuclear ribonucleoprotein L, or hnRNP L, an RNA-binding splicing factor previously implicated in pulmonary fibrosis through its control of circRNA back-splicing. Molecular docking against the AlphaFold-predicted structure suggested potential interactions with residues including ASN-395, ARG-398, ASN-401, GLN-566, LYS-475, ASP-397, ASN-482, and PHE-561, alongside hydrophobic contacts with PHE-477 and PHE-400. Surface plasmon resonance on a CM5 chip confirmed that rutin binds recombinant hnRNP L with an equilibrium dissociation constant of approximately 19.48 micromolar—a moderate but measurable affinity. Critically, the cellular thermal shift assay showed that in rutin-treated cells, hnRNP L exhibited higher thermal stability across a 45 to 70 degree Celsius range, with a significant rightward shift in the denaturation curve, demonstrating that the compound engages its target inside living cells and stabilizes the protein against heat-induced unfolding. Quantitative PCR and Western blotting further revealed that rutin suppressed hnRNP L expression at both the mRNA and protein levels, and rescue experiments showed that overexpressing hnRNP L reversed rutin’s inhibitory effects on α-SMA expression, cell migration, and fibrotic protein markers.

What elevates this work above a routine drug-target pairing is its embrace of liquid–liquid phase separation, the biophysical process by which proteins containing intrinsically disordered regions spontaneously condense into dynamic droplet-like biomolecular condensates. Sequence analysis using the PONDR database identified two intrinsically disordered segments in hnRNP L, spanning amino acids 1–101 and 336–383, marking it as a phase-separation-prone protein. The researchers purified an EGFP-tagged hnRNP L fusion protein and observed that, in the presence of 10 percent PEG6000 as a molecular crowding agent, the protein formed spherical droplets under laser scanning confocal microscopy. Fluorescence recovery after photobleaching showed that when a single condensate was bleached with a high-intensity laser, fluorescence recovered within roughly four minutes, confirming the liquid-like diffusivity and internal mobility characteristic of genuine phase-separated condensates rather than static aggregates. When TGF-β1-stimulated cells were transfected with EGFP-hnRNP L, phase-separated puncta formed in the nucleus; rutin treatment prevented this assembly, and photobleached condensates in the rutin group showed no fluorescence recovery, indicating that the drug had dismantled the condensate network entirely.

The disordered regions proved functionally indispensable. When the researchers constructed a truncation mutant lacking amino acids 1–111 and 274–384, the protein failed to form droplets both in vitro and in cells, and overexpression of this mutant had no significant impact on fibrotic protein expression or on α-SMA mRNA levels in IMR-90 cells. To test whether phase separation was required for hnRNP L’s pro-fibrotic activity in living animals, the team packaged full-length hnRNP L and the truncation mutant into adeno-associated viral vectors (AAV2/5) and delivered them via aerosol directly into the lungs of bleomycin-treated mice. Overexpression of full-length hnRNP L deteriorated lung function, reversed rutin’s beneficial effects on forced vital capacity, dynamic lung compliance, and total lung capacity, worsened honeycomb-like fibrotic structures on microCT, and elevated collagen, vimentin, α-SMA, and FAP1 on Western blot. In striking contrast, overexpression of the phase-separation-deficient mutant did not impair lung function, did not reverse rutin’s therapeutic efficacy, and did not counteract the drug’s inhibitory effects on fibrotic markers. These in vivo rescue data, obtained from a single randomized and blinded cohort of six mice per group, support the model that the anti-fibrotic action of rutin is tied to its interference with hnRNP L condensate assembly.

Having established that hnRNP L undergoes phase separation and that this capacity is required for its pro-fibrotic function, the investigators next asked which downstream splicing events the condensates control. Transcriptome sequencing of TGF-β1-stimulated MRC-5 cells after hnRNP L knockdown, analyzed through Gene Ontology and KEGG enrichment, revealed that glycolysis was among the most significantly altered biological processes. A heatmap of 21 differentially expressed glycolysis-related genes placed pyruvate kinase M, or PKM, among the top-ranking hits. Using the SUPPA2 pipeline on Oxford Nanopore long-read data, the team identified 17,958 significant differential splicing events between control and hnRNP L-depleted groups, with 25 of these events located in genes involved in glycolysis and gluconeogenesis—suggesting that hnRNP L phase separation may orchestrate a broader metabolic reprogramming beyond PKM alone. PKM can be alternatively spliced to produce two isoforms, PKM1 and PKM2, and the PKM2 isoform is a rate-limiting glycolytic enzyme previously shown to promote myofibroblast differentiation and pulmonary fibrosis progression. Quantitative PCR showed that rutin reduced PKM2 expression and increased PKM1 expression at the mRNA level, and Western blotting confirmed the downregulation of PKM2 protein. An RNA immunoprecipitation experiment demonstrated that PKM pre-mRNA physically binds hnRNP L, and that this binding decreases under rutin treatment, consistent with the drug disrupting the splicing factor’s engagement with its target transcript.

The metabolic consequences of this splicing switch proved equally consequential. PKM2 drives the final step of glycolysis and enhances lactate production, and lactate serves as the substrate for histone lactylation—a recently discovered epigenetic modification in which lactyl groups are covalently attached to histone lysine residues, directly influencing gene transcription. The researchers screened three lactylated histones—H3K9la, H4K12la, and H4K16la—and found that rutin significantly reduced H3K9 lactylation. Inhibition of PKM2 with the selective inhibitor PKM2-IN-1 reduced H3K9la, while overexpression of PKM2 reversed this effect, and exogenous lactic acid supplementation attenuated rutin’s inhibitory effects on both H3K9la and α-SMA. Co-immunoprecipitation revealed that P300, the acetyltransferase enzyme responsible for lactylation, increased H3K9la in the TGF-β1 group, whereas rutin inhibited this modification. Chromatin immunoprecipitation followed by PCR demonstrated that TGF-β1 enhanced the occupancy of H3K9la at the PKM promoter, while rutin decreased this interaction—and critically, overexpression of full-length hnRNP L reversed rutin’s effect, whereas the phase-separation-deficient truncation mutant did not. Supplementation with exogenous lactic acid largely restored PKM pre-mRNA and PKM2 transcript abundance in rutin-treated cells, supporting a model in which lactate functionally maintains PKM transcription through H3K9la, forming a self-limiting regulatory circuit in which reduced H3K9la feeds back to suppress PKM2 transcription.

The authors are careful to frame their findings as a mechanistic framework rather than a definitive causal chain, acknowledging that the binding affinity of rutin for hnRNP L is moderate and that the compound’s well-documented polyphenol properties may contribute to its overall anti-fibrotic effects through additional targets. They also note that the intrinsically disordered region truncation mutant lost both phase separation ability and pro-fibrotic activity simultaneously, leaving open the possibility that the truncation impairs other functional domains independent of condensate assembly. Direct visualization of endogenous hnRNP L phase separation in fibrotic lung tissue has not yet been achieved, and the mechanistic studies were conducted in fetal lung fibroblast lines that may not fully recapitulate the behavior of primary adult fibroblasts from patients with idiopathic pulmonary fibrosis. The models also do not incorporate the mechanical microenvironment—increased matrix stiffness is itself a principal regulator of PKM2 in lung fibroblasts—and future work using stiffness-tunable substrates or decellularized fibrotic lung matrices will be needed to determine whether the hnRNP L–PKM2–lactate–H3K9la axis is also modulated by mechanical cues. Lactate may derive from pathways other than PKM2-dependent glycolysis, and H3K9la may be regulated by histone modifiers beyond P300, so locus-specific manipulation of H3K9la at the PKM promoter will be necessary to establish a causal link between this epigenetic mark and PKM transcription.

Perhaps the most significant translational caveat concerns rutin’s pharmacokinetics. The compound suffers from poor oral bioavailability, stemming from low aqueous solubility, limited intestinal absorption, and rapid metabolism—drawbacks that have long hindered its clinical translation despite promising in vitro bioactivity. The researchers therefore positioned rutin as a tool compound to dissect a previously unrecognized molecular mechanism, while recognizing that future chemical optimization, nanocarrier formulation, or structural derivatization will be required to translate this mechanism into therapeutic benefit. No pharmacokinetic profiling of rutin or its metabolites was performed in the present study, and future LC-MS/MS-based analyses of plasma, lung tissue, and urine will be necessary to correlate systemic exposure with target engagement and antifibrotic efficacy. Nevertheless, the study represents a conceptual advance on multiple fronts: it identifies a natural product that pharmacologically targets phase separation, links an RNA splicing factor to metabolic and epigenetic regulation through a single druggable axis, and proposes a self-limiting feedback circuit in which reduced histone lactylation suppresses the very transcript that generates the lactate substrate. Whether this framework survives validation in primary human fibroblasts, independent animal cohorts, and more physiological systems remains to be seen—but it offers a compelling new lens through which to view the molecular logic of fibrotic disease, and a template for how natural products might be repurposed as precision modulators of condensate biology.

Subject of Research: Pulmonary Fibrosis

Article Title: Rutin interferes with hnRNP L phase separation and modulates the PKM2–lactate–H3K9la axis in pulmonary fibrosis

Article References: Luan, J., Wang, Y., Jiang, M., Zhang, H., Lv, C., Zhang, J., & Song, X. (2026). Rutin interferes with hnRNP L phase separation and modulates the PKM2–lactate–H3K9la axis in pulmonary fibrosis. Frontiers in Pharmacology, 17, Article 1787570. https://doi.org/10.3389/fphar.2026.1787570

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1787570

Keywords: Rutin, interferes, hnRNP, phase, separation, modulates, PKM2, lactate, H3K9la, axis, pulmonary, fibrosis