TL;DR
This paper introduces a novel "drug attack on network robustness" model to understand how traditional Chinese medicine (TCM) like HXMM tablet works against retinal degenerative diseases. By simulating drug effects on disease networks, they found HXMM targets oxidative stress and inflammation, offering a new strategy for precise TCM application.
What Problem Does This Paper Address?
Retinal degenerative diseases, such as age-related macular degeneration (AMD) and diabetic retinopathy (DR), lead to irreversible vision loss and pose a significant health burden. While anti-VEGF therapies have shown some success, they don't work for all patients, highlighting the need for new treatment strategies. Traditional Chinese Medicine (TCM) is known for its multi-target approach to complex diseases, but its mechanisms are often poorly understood. This paper aims to elucidate the specific mechanisms of HXMM, a TCM tablet used for retinal diseases, by developing a computational model to predict drug effects on disease networks and then validating these predictions experimentally.
Methods
The core of this research is the development and application of a "drug attack on network robustness" model. This approach treats disease as a network of interacting genes and pathways. The model simulates a "drug attack" by virtually removing drug targets from this network and then quantifies how this disruption affects the network's stability and robustness. Key network topology metrics like average degree (AD), average shortest path length (ASPL), degree centrality (DC), and closeness centrality (CC) are used to measure this robustness.
The process involves:
- Disease Network Construction: Identifying genes associated with AMD and DR from databases and building protein-protein interaction (PPI) networks. These networks are then functionally enriched using Gene Ontology (GO) terms to identify key pathological processes like oxidative stress, inflammation, and angiogenesis.
- Drug Attack Simulation: Predicting potential drug targets of HXMM using bioinformatics tools (BATMAN-TCM) and then simulating the "attack" on the constructed disease networks by removing these targets.
- Robustness Evaluation: Calculating the change in network robustness before and after the simulated drug attack using the defined metrics. A normalized robustness index (RI) is generated to compare the drug's effect against random network perturbations.
- Experimental Validation: Using an in vitro model of retinal injury (ARPE-19 cells treated with H₂O₂) to test HXMM's protective effects. This includes measuring cell viability, oxidative stress markers (LDH, GSH), inflammatory cytokines (IL-6, TNF-α), and angiogenesis markers (VEGFA, VEGFB).
Datasets
- Disease Gene Data: DisGeNET database was used to identify genes associated with AMD and DR.
- Protein-Protein Interaction (PPI) Data: STRING database was used to construct PPI networks.
- Gene Ontology (GO) Enrichment: Metascape software was used for functional enrichment analysis.
- Drug Target Prediction: BATMAN-TCM was used to predict targets for HXMM, and the LINCS program was used for FDA-approved drug targets.
- Experimental Data: ARPE-19 cells were used for in vitro experiments.
Key Insights and Ideas
- Network Robustness as a Drug Efficacy Predictor: The study demonstrates that quantifying how a drug disrupts a disease network's robustness can predict its therapeutic efficacy, moving beyond single-target analysis.
- HXMM's Multi-Target Action: Unlike anti-VEGF drugs that primarily target angiogenesis, HXMM appears to exert its effects by reducing network robustness in the oxidative stress and inflammatory subnetworks.
- TCM's Unique Mechanism: The findings suggest that TCM like HXMM may offer a broader therapeutic approach by addressing multiple interconnected pathological processes, rather than a single pathway.
- Computational-Experimental Synergy: The study successfully integrates computational predictions with experimental validation, showcasing a powerful strategy for understanding complex drug mechanisms and enabling precision medicine.
Key Results
- The "drug attack on network robustness" model successfully predicted the therapeutic characteristics of drugs.
- HXMM showed a significant reduction in network robustness for the oxidative stress and inflammatory subnetworks in both AMD and DR models.
- In vitro experiments confirmed HXMM's protective effects on ARPE-19 cells against H₂O₂-induced injury.
- HXMM elevated GSH and reduced LDH levels, indicating antioxidant activity.
- HXMM suppressed IL-6 and TNF-α expression, demonstrating anti-inflammatory activity.
- HXMM also showed some anti-angiogenesis activity by modulating VEGFA and VEGFB.
- Compared to anti-VEGF agents, HXMM exhibited stronger antioxidant activity and comparable anti-inflammatory and anti-VEGF effects.
Limitations and Future Work
- In Vitro Model Limitations: The study primarily used an in vitro model (ARPE-19 cells) and H₂O₂-induced injury. While informative, this model may not fully replicate the complex in vivo environment of retinal degenerative diseases.
- Need for Further Validation: The authors acknowledge the need for more strong evidence and further exploration to fully validate their findings and the potential of HXMM.
- Animal Studies: While an animal study was mentioned for preparing the HXMM intestinal absorption liquid, a comprehensive animal study to validate the drug's efficacy in vivo was not detailed as part of the core findings.
- Compound-Target-Pathology Network: The paper suggests further investigation into the "compound-target-pathological process" network, particularly exploring the relative amounts of compounds and their target interactions to discover more valuable disease targets and active compounds.
- Broader Animal Validation: The authors encourage further animal experiments to validate the findings.
Summary updated