Diabetic retinopathy is a devastating complication of diabetes and a major cause of vision loss in working-age adults. Epidemiological projections cited in the study’s background literature estimate that hundreds of millions of people will be living with the condition by mid-century as global diabetes rates climb. At the tissue level, the disease damages the retina’s delicate vasculature, but growing evidence indicates that the retinal pigment epithelium, a monolayer of support cells that nourishes and maintains the light-sensitive photoreceptors, is an early and critical casualty. When these cells falter, the entire retina suffers. Current treatments, ranging from laser therapy to injections of anti-VEGF drugs, manage symptoms and slow progression but do not halt the underlying cellular degeneration, leaving an urgent need for therapies that act on root causes.

In recent years, one root cause has attracted intense attention: ferroptosis. This form of regulated cell death, formally described only in 2012, differs fundamentally from apoptosis. Rather than quietly dismantling itself, a ferroptotic cell succumbs to an iron-dependent assault on its own membranes. Free iron catalyzes the peroxidation of polyunsaturated lipids in the cell membrane, and once lipid peroxides accumulate beyond a survivable threshold, the membrane ruptures and the cell dies in an inflammatory fashion. Because the retina is rich in both iron and polyunsaturated fatty acids and operates under high oxygen tension, it is uniquely vulnerable to this mode of death. Prior studies, including work by some of the same Kunming group, have linked oxidative stress-induced ferroptosis in retinal pigment epithelial cells to diabetic retinal damage, making the ferroptosis pathway a compelling therapeutic target.

To model diabetic conditions in the laboratory, the team exposed ARPE-19 cells, a widely used human retinal pigment epithelial cell line, to an elevated glucose concentration of 33 millimoles per liter. The consequences were exactly what the ferroptosis hypothesis predicts. Cell viability, measured with a CCK-8 assay, dropped significantly compared with cells grown in normal glucose. Inside the stressed cells, the researchers detected rising levels of ferrous iron and lipid peroxidation, visualized with the fluorescent probe BODIPY-C11, while the cellular reservoir of the antioxidant molecule glutathione, or GSH, shrank. Western blot analysis revealed that two key guardians of the antiferroptotic defense, the proteins GPX4 and SLC7A11, were downregulated. GPX4 is the only mammalian enzyme capable of reducing lipid hydroperoxides directly within membranes, and SLC7A11 is the transporter subunit that imports cystine, the raw material cells need to synthesize glutathione. Suppress both, and the cell loses its main protection against membrane oxidation.

Crucially, the researchers confirmed that ferroptosis was genuinely the mode of death at work. When they treated the high-glucose cells with ferrostatin-1, a classical ferroptosis inhibitor, or with deferoxamine, an iron chelator that starves the Fenton chemistry driving lipid peroxidation, cell viability recovered and lipid peroxidation receded. That pharmacological validation mattered, because it established that rescuing cells from iron-dependent lipid damage was sufficient to counter the glucose insult. Only then did the team introduce acteoside to see whether the natural product could reproduce that protection on its own.

It could. Acteoside intervention markedly reversed the high glucose-induced promotion of ferroptosis, restoring viability and damping the biochemical hallmarks of lipid peroxidation. The compound’s influence extended to the master regulatory layer of cellular antioxidant defense: NRF2. Under normal conditions, NRF2 is held in the cytoplasm by its inhibitor Keap1, but oxidative stress or pharmacological activation liberates it, allowing the transcription factor to migrate into the nucleus and switch on hundreds of cytoprotective genes, including those governing glutathione synthesis and GPX4 expression. Using immunofluorescence microscopy, the team observed that acteoside significantly promoted NRF2 nuclear translocation in the stressed cells—direct visual evidence that the compound was turning the pathway on.

The most decisive experiment came next. When the researchers silenced NRF2, acteoside’s protective grip loosened: the compound’s ability to inhibit ferroptosis was significantly weakened in cells lacking functional NRF2. That loss-of-function result elevates the findings from correlation to mechanistic argument. Acteoside does not merely mop up free radicals as a generic antioxidant; it appears to work substantially by engaging the NRF2 signaling axis, which in turn sustains the GPX4 and SLC7A11-dependent firewall against lipid peroxidation. This mode of action distinguishes pathway activators from simple radical scavengers and aligns with a broader therapeutic strategy, since NRF2 agonism is being pursued for a range of diseases involving oxidative damage and inflammation.

Importantly, the story did not end in the culture dish. The team also established a diabetic retinopathy model in C57BL/6 mice using streptozotocin, a chemical that destroys insulin-producing beta cells and induces diabetes, and treated the animals with different doses of acteoside. The results mirrored the in vitro findings with a clear dose-dependent trend. Mice receiving acteoside showed significantly alleviated retinal tissue damage on hematoxylin and eosin staining, along with reduced levels of ferrous iron in the tissue and diminished abundance of 4-HNE and MDA, two well-established byproducts of lipid peroxidation that serve as molecular fingerprints of ferroptotic stress. Glutathione content rose, and Western blots of retinal tissue showed upregulated expression of NRF2, GPX4, and SLC7A11. In short, the same protective circuitry identified in single cells operated in intact, diabetic retinae.

The study adds to a rapidly growing dossier on acteoside, a molecule also known as verbascoside that has drawn attention for antimicrobial, anti-inflammatory, and immunomodulatory properties. The Kunming group itself had previously shown that acteoside protects retinal ganglion cells in experimental glaucoma through the PI3K/AKT pathway and that it relieves diabetic retinopathy by regulating Müller cell reactivity via TXNIP and Kir4.1 channels. Other laboratories have reported that acteoside inhibits ferroptosis in contexts as diverse as hepatic ischemia-reperfusion injury, doxorubicin-induced cardiotoxicity, and neurodegeneration in a Parkinson’s disease model, in the latter case through Nrf2-mediated mitophagy. The new work extends this pattern to the retinal pigment epithelium in diabetes, and it lands alongside a parallel literature in which other natural NRF2 activators, such as corilagin, resveratrol, and ghrelin, have been shown to suppress ferroptosis in diabetic retinal cells.

For the millions of people at risk of diabetic vision loss, the research offers a genuinely encouraging proof of concept, though the usual caveats apply. The findings come from one human cell line and a chemically induced mouse model, and translating a plant-derived compound into a clinically validated therapy will require pharmacokinetic studies, dosing optimization, safety evaluation, and eventually controlled human trials—steps that often stall at the gap between mouse and patient. Still, the study’s central contribution is clear and technically meaningful: it identifies ferroptosis as a driver of high glucose-induced retinal pigment epithelial injury and demonstrates, with both gain- and loss-of-function evidence, that acteoside counteracts that cell death by activating NRF2 signaling and its downstream glutathione-GPX4 antioxidant machinery. If subsequent work confirms the effect in clinical settings, a humble compound long buried in herbal leaves could become a foundation for protecting the sight of a diabetic generation.

Subject of Research: Acteoside-mediated NRF2 activation protecting retinal pigment epithelial cells from high glucose-induced ferroptosis in diabetic retinopathy

Article Title: Acteoside alleviates high glucose-induced ROS accumulation and ferroptosis in ARPE-19 cells by activating the NRF2 signaling pathway

Article References: Acteoside alleviates high glucose-induced ROS accumulation and ferroptosis in ARPE-19 cells by activating the NRF2 signaling pathway. (n.d.). https://doi.org/10.1007/s11033-026-12608-x

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12608-x

Keywords: diabetic retinopathy, acteoside, ferroptosis, NRF2, retinal pigment epithelium, lipid peroxidation, GPX4, SLC7A11, oxidative stress, ARPE-19 cells, glutathione, natural products

The retraction notice, published in the same journal, explains that after publication, concerns were raised regarding image irregularities. When the authors were contacted to clarify the issues, they did not respond. In the world of research integrity, that silence is often decisive. Journals typically give authors an opportunity to explain anomalies in their figures, and when those explanations fail to materialize, editors are left with little choice but to retract. The notice is unambiguous: the editors no longer trust the findings, and the scientific record has been corrected accordingly.

The specific concerns are technical but instructive. According to the retraction notice, a panel in Figure 3B of the retracted paper, labeled as an IL-1 beta condition at a concentration of 10 nanograms per milliliter, appears to overlap with a panel in Figure 6C of a separate 2019 article on nasopharyngeal carcinoma published in Cancer Cell International. Strikingly, the same panel also appears to overlap with a panel in Figure 2D of a second, unrelated paper on pancreatic cancer that has itself since been retracted. In plain terms, identical or nearly identical images seem to have been used to represent entirely different experiments in entirely different disease contexts, spanning osteoarthritis, nasopharyngeal cancer, and pancreatic cancer.

This kind of image duplication is one of the most common triggers for retractions in the modern era of research integrity screening. Western blots, microscopy fields, and other gel-based or imaging outputs are supposed to be unique records of individual experiments. When the same image shows up in multiple papers, or in multiple figures within a single paper, it suggests either careless figure assembly or deliberate fabrication. Software tools and vigilant readers now routinely flag such overlaps, and journals have invested heavily in automated screening pipelines that compare figures across vast swaths of the published literature. What once might have slipped through unnoticed for decades can now be detected within months or years of publication.

The retraction also contains a second, more troubling element: a disputed authorship claim. One of the listed authors, Zhiguang Qiao, stated that he did not contribute to the study and did not consent to being listed as an author of the article. The remaining authors, according to the notice, have not responded to correspondence regarding the retraction. Authorship disputes of this kind raise questions that go beyond the integrity of the data itself. Authorship carries professional credit, responsibility, and accountability, and being listed on a paper without consent is a serious matter. Conversely, when a paper is retracted, every listed author shares in the reputational damage, which is precisely why consent and genuine contribution matter so much.

The original study, published on 7 October 2019, had explored a question of real clinical interest. Osteoarthritis is the most common form of arthritis, affecting hundreds of millions of people worldwide, and its hallmark is the slow degeneration of joint cartilage accompanied by chronic inflammation. Current treatments largely manage symptoms with painkillers and anti-inflammatory drugs, but nothing on the market halts or reverses the underlying disease. That gap has fueled intense interest in natural compounds with anti-inflammatory properties, and acteoside has been one of the candidates. Found in plants such as those used in traditional medicine systems, acteoside had been reported in various laboratory studies to modulate inflammatory signaling, and the JAK/STAT pathway, a chain of molecular interactions that transmits signals from the cell surface to the nucleus, is a well-established regulator of immune and inflammatory responses. Drugs targeting this pathway are already approved for conditions including rheumatoid arthritis, so the hypothesis that a plant compound might act similarly in osteoarthritis was scientifically plausible.

Plausible, however, is not the same as demonstrated. The retraction does not merely erase a single data point; it removes the evidentiary basis for the paper’s central claim that acteoside inhibits inflammatory responses via JAK/STAT signaling in osteoarthritic rats. Readers who encounter the paper in databases or citation lists will now find it flagged as retracted, and responsible researchers will exclude it from systematic reviews and meta-analyses. This matters because complementary and alternative medicine research is frequently cited in public discussions of herbal remedies, and retracted findings can linger in the popular imagination long after the formal correction. A retraction notice is a small document, but its function is to stop the spread of unreliable conclusions before they harden into accepted wisdom.

The case also illustrates the interconnected nature of research integrity problems. The two articles whose figures allegedly overlap with the retracted osteoarthritis paper are themselves notable: one, on circular RNA in nasopharyngeal carcinoma, remains in the literature as far as the notice indicates, while the other, a study of solasodine against pancreatic cancer, is explicitly labeled as retracted. When the same images surface across papers from what appear to be different research groups or different disease areas, integrity investigators often look for common threads, such as shared figure-preparation services, paper mills, or overlapping authorship networks. The retraction notice itself does not draw such conclusions, and it would be wrong to speculate beyond what the editors have stated. But the pattern of a single panel appearing in three different disease contexts is exactly the kind of red flag that has, in other cases, led to much larger sweeps of retractions.

For the journal, BMC Complementary Medicine and Therapies, the retraction is part of a broader reckoning. Journals in the complementary medicine space have faced particular scrutiny in recent years, in part because the field attracts a high volume of submissions and in part because several journals in adjacent areas experienced mass retractions linked to paper mills. Editors have responded with stricter checks at submission, including image forensics, statistical scrutiny, and requirements for raw data. The fact that this retraction was issued years after the original publication, and that the notice was published in 2026 with a new article number and DOI, reflects the often slow but methodical pace of post-publication correction. Investigations take time, correspondence with authors can drag on for months, and legal considerations sometimes complicate matters further.

What should readers take away from this episode? First, that the self-correcting machinery of science, however slow, does function. A paper published in 2019 has been formally retracted, its problems documented, and its authors given the chance to respond. Second, that images are evidence. The figures in a scientific paper are not illustrations; they are the data themselves, and their integrity is inseparable from the integrity of the conclusions. Third, that authorship is a responsibility, not just a credit line, and disputes about who did or did not contribute are treated with seriousness by editors. And finally, that single studies, especially early animal studies of natural compounds, should always be viewed as provisional. The hypothesis that acteoside might calm inflammation in arthritic joints may yet be tested rigorously by other groups with clean data. Until then, the scientific record now accurately reflects what is and is not known, and that accuracy is the foundation on which all future research depends.

Subject of Research: Retraction of a study on acteoside and the JAK/STAT inflammatory pathway in osteoarthritis due to image irregularities

Article Title: Retraction Note: Acteoside inhibits inflammatory response via JAK/STAT signaling pathway in osteoarthritic rats

Article References: Qiao, Z., Tang, J., Wu, W., Tang, J., & Liu, M. (2026). Retraction Note: Acteoside inhibits inflammatory response via JAK/STAT signaling pathway in osteoarthritic rats. BMC Complementary Medicine and Therapies, 26(1), Article 285. https://doi.org/10.1186/s12906-026-05627-5

Image Credits: AI Generated

DOI: 10.1186/s12906-026-05627-5

Keywords: retraction, acteoside, osteoarthritis, JAK/STAT pathway, research integrity, image duplication, BMC Complementary Medicine and Therapies, authorship dispute, inflammation, natural compounds, scientific publishing, figure manipulation