AI-driven drug discovery identifies potential RIPK3 inhibitors for glaucoma treatment.

Glaucoma is a progressive condition of the eye caused by the accumulation of fluid, leading to increased eye pressure. It is projected that by 2040, 111.8 million individuals worldwide will be affected by glaucoma, which can result in blindness if not treated. While treatments are available to control eye pressure, there is still no cure for glaucoma (1^✔ ✔Trusted Source
Artificial intelligence-enabled discovery of a RIPK3 inhibitor with neuroprotective effects in an acute glaucoma mouse model

Go to source

).

RIPK3: A Potential Target for Glaucoma Therapy

Retinal ganglion cells (RGCs) are crucial for transmitting visual signals from the eyes to the brain, and their degeneration leads to optic nerve damage, which is a hallmark of glaucoma. In recent years, scientists have focused on developing neuroprotective drugs that can rescue RGCs and restore the optic nerve pathways. Necroptosis, a pathway responsible for programmed cell death, plays a significant role in the loss of RGCs. Unlike other forms of cell death, necroptosis shares features of both apoptosis (natural cell breakdown) and necrosis (injury-related cell damage). Receptor-interacting protein kinase 3 (RIPK3), a key signaling molecule, is known to play a critical role in necroptosis, making it a promising target for therapeutic intervention.

AI-Driven Drug Screening Approach

To identify potential RIPK3 inhibitors, a team of researchers from different research facilities and medical centers in China came together to conduct an artificial intelligence (AI)-driven drug screening technique and identify the potential target compounds of RIPK3. The study, led by Dr. Yuanxu Gao from Macau University of Science and Technology and Professor Zhang Kang from Guangzhou National Laboratory, was published in the Chinese Medical Journal. “AI provides reliable tools and methods for drug discovery, such as virtual screening, quantitative structure-activity relationship modeling, and de novo drug design,” says Dr. Gao, speaking about the motivation behind the study.

In Silico Analysis: ADMET Predictions

The team utilized advanced AI models, including a large language model and graph neural network models, to identify RIPK3-targeting molecules. Even though the command, “small-molecule compounds targeting RIPK3,” was the same throughout numerous engagements with ChatGPT 3.5, the team got a randomized AI-generated list. The team also used multiple AI models, including DynamicBind, to predict the drug-RIPK3 affinity and interaction pattern. Complex-conformation prediction using molecular dynamics simulation along with in silico analysis of absorption, distribution, metabolism, excretion, and toxicity (ADMET) of the molecules was also performed. To validate these findings, the team conducted biological experiments, including cell viability assays, immunofluorescence, histological analysis, and protein quantification.

Using AI, the team identified HG9-91-01, dabrafenib, AZD5423, GSK840, and HS-1371 as the most potent small-molecule compounds that can effectively target RIPK3. According to affinity-score prediction models, HG9-91-01 was considered to be the most promising candidate for targeting RIPK3. The suitability of HG9-91-01 was confirmed by molecular simulations, demonstrating that it formed a more stable complex with RIPK3 compared to other options. The compound also exhibited good results in safety and drug-related tests as per the ADMET predictions.

In Vitro Model of Optic Nerve Damage

The compound’s efficacy was validated through laboratory experiments. In an in vitro model mimicking optic nerve damage, RGCs exposed to oxygen-glucose deprivation (OGD) showed higher survival rates when treated with HG9-91-01 compared to other candidates. The compound also reduced the presence of gasdermin D (GSDMD)-positive cells, a marker of pyroptosis, a type of inflammatory cell death. Highlighting an interesting aspect of this study, Prof. Kang says, “Although numerous studies have focused on anti-apoptotic, anti-necroptotic, and anti-pyroptotic drugs for treating acute ocular hypertension (AOH), strategies targeting PANoptosis, including cell-cell communication and cascade reactions of cell death, are rarely mentioned. In this study, we investigated potential drug treatments targeting RIPK3 to prevent RGC death and explored their role in preventing PANoptosis.”

Inhibition of Cell Death Pathways: Apoptosis, Pyroptosis, and Necroptosis

In vivo experiments involving mouse models also showed positive results. Retinal thinning is often associated with glaucoma and increased ocular pressure. However, treatment with HG9-91-01 molecule prevented the loss of retinal thickness. It also reduced the activation of signaling molecules associated with apoptosis, pyroptosis, and necroptosis, suggesting its potential in preventing PANoptosis.

AI Technologies Aiding Drug Discovery

The findings of this study provide interesting insights into the combination of AI algorithms and traditional drug screening methods, that might lead to a logical and data-driven drug development model. Biological research investigating drug-target interaction necessitates a substantial amount of time and resources. AI-based prediction models simplify and expedite this process. “AI technologies are useful for handling computationally intensive tasks and making rational decisions based on complex multimodal knowledge. However, possible concerns such as data privacy, transparency, and bias should be addressed with caution,” concludes Dr. Gao.

In order to confirm that HG9-91-01 can protect the retinal structure in patients with AOH, the team plans on conducting further confirmatory retinal assessments. With AI-driven breakthroughs on the horizon, the future of glaucoma treatment looks brighter than ever!

Reference:

1. Artificial intelligence-enabled discovery of a RIPK3 inhibitor with neuroprotective effects in an acute glaucoma mouse model – (https://journals.lww.com/cmj/fulltext/9900/artificial_intelligence_enabled_discovery_of_a.1374.aspx)

Source-Eurekalert