AI-POWERED DISCOVERY OF RIPK-INHIBITOR

 

Artificial intelligence (AI) technologies have found multiple roles and advantages in drug discovery.

Those technologies have significantly advanced the discovery of newer drugs through target identification and high-throughput screening.

AI can utilize experimentally validated conformational and physicochemical features of protein-ligand compounds to create statistical models. These models enable predictions in three directions—binding site, binding affinity, and binding pose—to provide predictions that are more applicable to real-world scenarios.

AI technologies are also useful for handling computationally intensive tasks and making rational decisions based on complex multimodal knowledge.

Tu et al from China have reported the Artificial Intelligence-enabled discovery of a RIPK3 inhibitor with neuroprotective effects in an acute glaucoma mouse model.

An acute ocular hypertension model was used to simulate pathological ocular hypertension in vivo.

The study involved a series of AI methods, including large language models (LLM) and graph neural network models, to identify the target compounds of RIPK3. Subsequently, these target candidates were validated using molecular simulations (molecular docking, absorption, distribution, metabolism, excretion, and toxicity [ADMET] prediction, and molecular dynamics simulations) and biological experiments (Western blotting and fluorescence staining) in vitro and in vivo.

In this study, the authors used LLM as an AI-driven drug screening tool and ChatGPT to conduct targeted drug queries for RIPK3.

The authors sorted the candidate drugs based on the prediction of their binding affinity to RIPK3 using graph neural network (GNN) models and the predicted results were validated using biological experiments.

The study involved AI models (GraphDTA, MGraphDTA, and WGN-NDTA), and reconstructed DynamicBind, another model architecture that uses deep equivariant geometric diffusion networks to predict affinity.

The authors also performed molecular docking and molecular dynamic simulations to validate the predictions of the GNN models.

In conclusion, these studies identified a compound called HG9-91-01 using AI methods. The compound exerts neuroprotective effects in acute glaucoma.

Retinal ganglion cells (RGCs) had a high survival rate and reduced loss of retinal layers on exposure to HG9-91-01.

The neuroprotective effects of HG9-91-01 were attributed to the inhibition of PANoptosis (apoptosis, pyroptosis, and necroptosis).

Also, HG9-91-01 can regulate key proteins related to PANoptosis, indicating that this compound exerts neuroprotective effects in the retina by inhibiting the expression of proteins related to apoptosis, pyroptosis, and necroptosis.

ROLE OF RIPK3 IN NECROPTOSIS:

The loss of RGCs during glaucoma progression involves various types of cell death, and necroptosis, a process similar to apoptosis, may play a significant role in RGC death. Necroptosis is initiated by receptor-interacting protein kinase (RIPK) 1, RIPK3, and mixed-lineage kinase domain-like (MLKL). RIPK3 is an important signaling molecule located downstream of RIPK1, and phosphorylation of RIPK3 or RIPK1 activates MLKL, thus initiating necroptosis. However, the relationship between RIPK1 and RIPK3 is non-linear. As RIPK3, but not RIPK1, is essential for necroptosis, “necroptosis” is more accurately described as “RIPK3-dependent cell death”.

RIPK3 is highly expressed in the ganglion cell layer (GCL) of injured retinas in vivo. Compounds that target RIPK3 and regulate the necroptotic cascade can be used to treat various retinal diseases. Therefore, RIPK3 inhibitors are promising candidates for the treatment of neurodegenerative ocular diseases.

REFERENCE:

Tu X, Zou Z, Li J, Zeng S, Luo Z, Li G, Gao Y, Zhang K. Artificial intelligence-enabled discovery of a RIPK3 inhibitor with neuroprotective effects in an acute glaucoma mouse model. Chin Med J (Engl). 2025 Jan 20;138(2):172-184.