Subretinal fibrosis is the body’s wound-healing response gone awry in the confined space between the retinal pigment epithelium and the photoreceptors. When abnormal choroidal blood vessels invade the subretinal space in neovascular AMD, bleeding, inflammation and vascular endothelial growth factor–driven remodeling follow. The retinal pigment epithelial (RPE) cells that normally act as quiet, polarized guardians of the photoreceptor layer respond by abandoning their identity: they lose tight junctions, reorganize their cytoskeleton, and transform into migratory, contractile, matrix-secreting cells. This process, epithelial-mesenchymal transition, or EMT, is driven centrally by transforming growth factor-beta (TGF-β) signaling and produces the collagen-rich scar that physically distorts and eventually destroys the overlying retina.

The research team, led by corresponding authors Chaoyang Zhang, Xiufeng Zhong, Haibin Tian and Jingfa Zhang, with first authors Xi Wang and Xueying Wang, reasoned that a molecule already known to suppress fibrosis in other organs could be repurposed for the eye. Sophocarpine has previously shown anti-fibrotic activity in hepatic and pulmonary models, and its alkaloid chemistry suggested a small molecule capable of crossing biological membranes and reaching intracellular targets. What remained entirely unknown was whether it could do anything for ocular fibrosis—and whether it could do so without harming the delicate neural retina it would have to coexist with.

The safety question came first, and the answer was reassuring. The investigators injected sophocarpine intravitreally into C57BL/6J mice at a concentration of 200 micromolar and then interrogated retinal function with electroretinography, a technique that measures the summed electrical responses of photoreceptors and inner retinal neurons to light flashes. Histological examination of the treated retinas found no detectable acute toxicity. That combination—a preserved electroretinographic waveform and an intact retinal architecture—is the basic license a compound needs before it can be considered seriously for intravitreal delivery, and many candidates never clear that bar.

With safety established, the team moved to a disease model that has become the standard for studying the fibrotic sequelae of neovascular AMD: laser-induced choroidal neovascularization. Focused laser burns to the mouse Bruch’s membrane trigger the growth of abnormal choroidal vessels into the subretinal space, followed by fibrotic scarring that closely mirrors the human lesion. When sophocarpine was delivered into the eyes of these mice, the CNV lesions shrank measurably, collagen deposition within the lesions decreased, and—critically—the co-localization of alpha-smooth muscle actin with RPE65, a marker of RPE cells that had acquired myofibroblast characteristics, was reduced. In plain terms, the RPE cells at the lesion edge were less inclined to become scar-forming cells.

The cellular mechanism was then dissected in culture using two complementary human RPE systems: the widely used ARPE-19 cell line and induced pluripotent stem cell–derived RPE cells, which recapitulate native RPE biology more faithfully. When these cells were stimulated with TGF-β1, they predictably ramped up the mesenchymal program: fibronectin, alpha-smooth muscle actin and vimentin all rose, while the tight-junction protein ZO-1, a hallmark of the healthy epithelial state, was lost. Sophocarpine treatment reversed this cascade. It suppressed the induction of fibronectin, α-SMA and vimentin, preserved ZO-1 expression, and curtailed the functional consequences of EMT—cell migration measured in wound-healing and Transwell assays, and the contractile behavior that scars use to wrinkle and detach the retina, captured in gel contraction assays. The findings were confirmed at transcript, protein and cellular localization levels using quantitative PCR, Western blotting and immunofluorescence.

The mechanistic heart of the paper lies in a striking signaling selectivity. TGF-β signals through two major intracellular routes: the canonical Smad pathway, in which Smad2 and Smad3 transcription factors are phosphorylated and carry the signal to the nucleus, and the non-canonical MAPK pathways, including the MEK/ERK cascade and p38 MAPK. Sophocarpine preferentially reduced the phosphorylation of MEK, ERK and p38 while leaving Smad2/3 phosphorylation untouched. This dissociation is significant because it suggests the compound does not blunt TGF-β signaling wholesale—which could interfere with the many homeostatic roles TGF-β plays in the eye—but instead specifically interrupts the MAPK arm that drives EMT-associated proliferation, migration and matrix production.

To probe how sophocarpine might achieve that selectivity at the molecular level, the team turned to computational structural biology. Molecular docking predicted that the alkaloid fits into binding pockets on MEK1, MEK2 and ERK2, the kinase pair that constitutes the canonical MAPK relay, and molecular dynamics simulations suggested the docked complexes remain stable over simulated time. These are predictive, not definitive, results—binding in silico must ultimately be validated with direct biophysical measurements and target-engagement assays in cells—but they furnish a concrete, testable hypothesis: that sophocarpine acts as a direct inhibitor of upstream MAPK kinases, thereby explaining both the reduced ERK and p38 phosphorylation and the suppression of the EMT program downstream.

The translational implications are considerable. Anti-VEGF therapy, the current mainstay for neovascular AMD, suppresses vascular leakage and growth but does nothing to stop the fibrovascular scar once it begins to organize; a substantial fraction of treated patients develop subretinal fibrosis and irreversible vision loss despite flushes of injections. A molecule that targets the fibrotic arm of the disease—by preserving RPE identity and restraining the TGF-β–MAPK–EMT axis—would address a therapeutic void that no approved drug currently fills. It would also be an unusually accessible candidate: sophocarpine is a defined small molecule with a long history of human exposure in herbal preparations, which could streamline preclinical development compared with entirely novel chemical entities.

That said, the distance from a laser-burn mouse and a culture dish to a clinic-ready therapy remains long. The study used a single dose level for the intravitreal safety assessment, relied on an acute injury model rather than spontaneous, slowly evolving human fibrosis, and its mechanistic target hypothesis awaits confirmation with techniques such as surface plasmon resonance and kinase assays. The authors themselves position sophocarpine as a promising preclinical candidate that warrants further investigation. Even so, the convergence of findings—in vivo reduction of scarring, in vitro reversal of EMT, and a coherent signaling rationale—makes this one of the more complete early-stage stories in ocular fibrosis research, and a reminder that some of ophthalmology’s next tools may still come from plants that traditional medicine has used for centuries.

Subject of Research: Therapeutic inhibition of subretinal fibrosis and retinal pigment epithelial EMT by the plant-derived alkaloid sophocarpine

Article Title: Sophocarpine, a bioactive alkaloid derived from Sophora flavescens, alleviates subretinal fibrosis by suppressing epithelial-mesenchymal transition

Article References: Sophocarpine, a bioactive alkaloid derived from Sophora flavescens, alleviates subretinal fibrosis by suppressing epithelial-mesenchymal transition. (n.d.). https://doi.org/10.1186/s12967-026-08995-1

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08995-1

Keywords: sophocarpine, Sophora flavescens, subretinal fibrosis, neovascular age-related macular degeneration, retinal pigment epithelium, epithelial-mesenchymal transition, TGF-beta, MEK/ERK, p38 MAPK, choroidal neovascularization, anti-fibrotic therapy, Journal of Translational Medicine