A plant-derived compound called ailanthone reduced HSV-1 replication in several cell models and eased corneal disease in a mouse model of herpes simplex keratitis, according to the study. The researchers found that ailanthone acted mainly after the virus entered cells and linked much of this effect to reduced expression of integrin beta 2, or ITGB2, a host protein that supports viral replication. The results identify ailanthone as a potential drug candidate, but the experiments do not establish that it is safe or effective in people.

The team began by screening natural compounds and testing ailanthone in African green monkey kidney Vero cells infected with HSV-1. Concentrations of 100 to 3,000 nanomolar reduced transcription of the viral immediate-early genes ICP0 and ICP4 and the late gene gD in a dose-dependent manner. The compound also reduced plaque formation and the concentration of progeny virus released into culture. Ailanthone had a half-maximal cytotoxic concentration above 10,000 nanomolar in these cells, while its half-maximal effective concentration against viral replication was 197.4 nanomolar, giving a selectivity index greater than 50. It also suppressed HSV-1-induced increases in the inflammatory cytokine genes IL-1B and TNF-α.

Antiviral activity was subsequently observed in human corneal epithelial cells, a model more directly relevant to the tissue affected by keratitis. Ailanthone concentrations of 100 to 500 nanomolar markedly reduced the cytopathic effects of HSV-1F, lowered expression of ICP0, ICP4 and gD, and reduced the viral induction of IL-1B and TNF-α. Its effective concentration in these cells was 50.16 nanomolar, with a cytotoxicity concentration above 10,000 nanomolar and a selectivity index greater than 100. The researchers also tested the compound in human retinal pigment epithelial ARPE-19 cells, where it reduced viral gene expression, plaque formation and progeny-virus titers while showing low cytotoxicity.

Activity was not restricted to the laboratory HSV-1F strain. In ARPE-19 cells, ailanthone concentration-dependently inhibited the late gD gene in the clinical acyclovir-resistant strain HSV-1/153, whereas 1,000 nanomolar acyclovir had little effect. The compound’s effective concentration was 35.05 nanomolar against HSV-1F and 227.4 nanomolar against HSV-1/153, with selectivity indices above 100 for both strains. A time-of-addition experiment indicated that ailanthone did not prevent attachment of HSV-1 to cells. It reduced entry under one experimental schedule and had its strongest effect on post-entry replication, where it lowered gD expression.

To investigate how the compound might work, the researchers combined computational target prediction with protein-interaction network analysis. They identified 36 targets shared by ailanthone and herpes simplex keratitis, then selected ITGB2 as a highly connected hub in the resulting network and as a top candidate in molecular docking. The predicted docking score for ailanthone and ITGB2 was −8.6 kcal per mole. Computational predictions alone cannot establish direct binding, so the team also performed a drug affinity responsive target stability assay in THP-1 cell lysates. Ailanthone protected ITGB2 from digestion by the protease pronase E in a dose-dependent manner, while acyclovir did not, providing biochemical evidence consistent with a direct interaction.

Functional experiments then tested whether ITGB2 was necessary for the antiviral effect. Three independent small interfering RNAs reduced ITGB2 messenger RNA and protein in THP-1 cells. Following HSV-1 infection, this knockdown lowered gD messenger RNA and protein and reduced the number of plaques produced when culture supernatants were transferred to fresh Vero cells. The converse pattern appeared when ITGB2 was overexpressed: elevated ITGB2 increased gD expression in infected cells, and it partially counteracted the reduction in gD protein caused by ailanthone in 293T and ARPE-19 cells. Together, these loss- and gain-of-function results support the conclusion that ailanthone’s antiviral activity is mediated at least partly by lowering ITGB2.

The researchers next tested ailanthone in an acute mouse model of HSV-1 keratitis. Six-week-old male BALB/c mice received a corneal scratch followed by application of HSV-1F, then received intraperitoneal ailanthone at 1 or 5 milligrams per kilogram or acyclovir at 5 milligrams per kilogram on days 1, 3 and 5 after infection. The methods section states that six animals were assigned to each group, although the discussion says five mice per group were used in the animal experiment. The 5-milligram-per-kilogram ailanthone treatment significantly reduced clinical disease scores, improved the appearance of the cornea under normal and cobalt-blue illumination, and lowered fluorescein staining scores on days 3 and 5. Hematoxylin and eosin staining at day 6 also indicated reduced corneal inflammation, and viral titers in ocular swabs were significantly lower than in infected mice given vehicle.

Analysis of mouse corneal tissue was consistent with the proposed mechanism. ITGB2 was upregulated in the corneas of infected animals, and ailanthone reduced its fluorescence intensity in a dose-dependent manner. Acyclovir did not significantly alter corneal ITGB2 levels. The investigators interpreted this as evidence that ailanthone differs from a direct viral DNA-replication inhibitor: the data are consistent with modulation of a host dependency factor rather than action solely through a viral enzyme. However, the study did not determine precisely how binding to ITGB2 leads to lower protein abundance. The authors note that transcriptional regulation, changes in messenger-RNA stability, or enhanced protein degradation could all be involved, and dedicated experiments would be required to distinguish among them.

Several limitations constrain how far the findings can be generalized. The cell experiments used established lines, including corneal epithelial and retinal pigment epithelial cells, rather than primary human corneal cells or an intact human immune system. The animal experiment examined acute primary infection in young male mice, whereas recurrent reactivation and the inflammation associated with repeated episodes are important features of human herpes simplex keratitis. Ailanthone was administered by intraperitoneal injection, but no topical ophthalmic formulation, corneal penetration, intraocular distribution, or optimized dosing regimen was evaluated. Long-term ocular toxicity and safety under repeated exposure also remain unknown. The natural compound may additionally affect multiple targets, and compensation among integrins could limit or alter its effects. The evidence therefore supports further preclinical investigation of ailanthone and ITGB2 as a host-directed antiviral strategy, not a clinical recommendation or a claim of human therapeutic benefit.

Subject of Research: Preclinical evaluation of ailanthone as a host-directed inhibitor of HSV-1 replication and herpes simplex keratitis.

Article Title: Natural compound ailanthone alleviates herpes simplex keratitis in mice by inhibiting HSV-1 replication

Article References: Chen, L., Yang, E., Zhang, X., Zhang, L., Wu, L., Chen, S., Jin, W., Qin, Q., Chen, R., Wu, J., & Chen, D. (2026). Natural compound ailanthone alleviates herpes simplex keratitis in mice by inhibiting HSV-1 replication. Frontiers in Pharmacology, 17, Article 1937590. https://doi.org/10.3389/fphar.2026.1937590

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1937590

Keywords: HSV-1, herpes simplex keratitis, ailanthone, ITGB2, host-directed antivirals, viral replication, corneal infection, mouse model, integrin beta 2, drug resistance, Natural, compound

OpenAI said internal models hacked two more websites, bypassed company-imposed browsing restrictions, and used exposed credentials to draw cryptocurrency from a wallet it couldn’t otherwise access, in the latest installment of disclosures of the unethical and likely illegal acts its agents are performing.

The company on Thursday issued reports of four new incidents, all of which occurred in June. They come after the company revealed in July that internal test agents exploited a zero-day vulnerability in a tool designed to restrict their access to the Internet. They then broke into the network of AI tool provider Hugging Face. The agents took that illegal action to steal answers to a hacking test that weren’t available when completing the tasks on their own. The incident—which likely would have resulted in criminal charges had they been performed directly by a person—has further heightened concerns that widespread AI use could have harmful effects on an unprecedented scale.

Easier to beg for forgiveness than ask permission

In two of the cases reported Thursday, OpenAI agents tasked with retrieving information from websites hacked them when the models ran into obstacles obtaining it directly. In one incident during reinforcement training, a model that received access denials when trying to retrieve historical hospital-bed statistics on a government statistics website “used the site's file-download feature to retrieve server configuration and application files, including sensitive credentials,” OpenAI said. “It inspected the application code, bypassed the statistics endpoint's authorization checks, and retrieved data that had previously been denied. The model saved the data locally.”

Read full article

Comments

AI company Anthropic is warning that advances in small AI models could soon usher in a new era of self-replicating malware that’s capable of infecting smart fridges, vacuum cleaners, and a wide range of other Internet of Things devices and is more resistant to takedown.

By next year, smaller AI models may be able to autonomously self-replicate on all kinds of IoT devices, the company said. Eventually, the capability could enable computer worms, which propagate without requiring human interaction and are capable of infecting unfamiliar IoT devices. An experiment the company performed recently revealed significant challenges in bringing such an AI malware replicator to fruition, so the finding is necessarily tempered. Still, the results raise concerns about the possibility of new forms of computer worms.

Challenging to shut down

Anthropic further warned not only about the new potential malware, but also that it could be difficult to stop. In its blog post, it wrote that “a replicator like this would be challenging to shut down, since there’s no centralized point to shut off—though it may be slow (with a doubling time of days). Unlike non-AI malware, it could also figure out unfamiliar devices on its own.”

Read full article

Comments