This study accelerates aptamer drug discovery through AI, offering a novel anti-fibrotic therapeutic strategy for DMD. By targeting the CT-domain of CTGF, the design circumvents the compensatory upregulation of TGF-β1 commonly triggered by conventional antibody therapies, providing direct guidance for drug development in related fields.
Literature Overview
This article, 'AI-powered therapeutic aptamer drug discovery: Targeting the CT-domain of CTGF for duchenne muscular dystrophy,' published in Acta Pharmaceutica Sinica. B, systematically explores the use of the generative AI model AptGEN to rapidly screen aptamer drug candidates targeting the CT-domain of CTGF, particularly Apc003OA, and validates their anti-fibrotic effects in DMD models. The study overcomes the traditional limitations of SELEX—lengthy cycles and low efficiency—enabling full pipeline acceleration from target validation to candidate drug identification.Background Knowledge
Duchenne muscular dystrophy (DMD) is an X-linked lethal genetic disorder caused by mutations in the dystrophin gene, leading to progressive muscle degeneration and fibrosis, ultimately resulting in respiratory and cardiac failure. Although restoring dystrophin expression remains the fundamental therapeutic goal, clinical progress has been limited. Fibrosis, a core pathological process in DMD, severely impairs muscle regeneration and accelerates functional decline, making it a critical intervention point. Among various pro-fibrotic factors, CTGF (CCN2) is highly expressed in the muscle tissues of DMD patients and correlates strongly with the extent of fibrosis, making it a promising therapeutic target. However, the antibody drug FG-3019 (Pamrevlumab), which targets the VWC domain of CTGF, failed to meet primary endpoints in Phase III clinical trials, suggesting that the VWC domain may not be the optimal target. This study reveals that the CT domain of CTGF exhibits stronger pro-fibrotic activity and does not induce compensatory TGF-β1 upregulation, whereas targeting the VWC domain paradoxically activates TGF-β1, undermining therapeutic efficacy. Therefore, targeting the CT domain of CTGF, combined with the advantages of aptamer-based therapeutics, represents a key strategy to overcome current bottlenecks in anti-fibrotic DMD treatment.
Research Methods and Experiments
The study employed mdx mice as an animal model of DMD and used CRISPR/Cas9 to generate CTGF-knockout Rat 2 fibroblasts to validate the functional impact of domain-specific mutations. Binding affinity between aptamers and the CT-domain of CTGF was assessed using bio-layer interferometry (BLI) and ELONA. The AI model AptGEN, based on a hybrid architecture of variational autoencoder (VAE) and generative adversarial network (GAN), was trained on NGS data from early SELEX rounds to predict high-affinity aptamer sequences. Structural predictions using RosettaFold2NA, followed by truncation and mutation analyses, led to the optimized sequence Apc003. Chemical modifications with 2'-O-methyl groups at C/G sites enhanced nuclease resistance, and fatty acid (OA) conjugation was used to extend in vivo half-life. Pharmacokinetics were evaluated in SD rats, while tissue distribution and therapeutic efficacy were tested in mdx mice using Western blot, Masson’s staining, and muscle strength assessments to evaluate fibrosis markers and functional improvements.Key Conclusions and Perspectives
Research Significance and Prospects
This study establishes a new paradigm for AI-driven aptamer discovery in drug development, particularly suitable for targeting flexible protein domains that are traditionally difficult to drug with small molecules or antibodies. The advancement of Apc003OA to receive FDA Orphan Drug and Rare Pediatric Disease designations highlights its clinical translatability. Future work may extend this strategy to other fibrotic diseases, such as pulmonary or hepatic fibrosis, broadening the therapeutic indications for CTGF-targeted treatments.
Conclusion
This study integrates AI modeling with experimental validation to establish the CT-domain of CTGF as a superior target for anti-fibrotic therapy in DMD and develops Apc003OA, a highly effective and safe aptamer drug. The drug not only overcomes the compensatory activation of TGF-β1 associated with antibody therapies but also achieves superior muscle tissue penetration and retention due to its small size. The entire process—from target identification to candidate drug advancement—was completed in just 10 months, demonstrating the immense value of AI in accelerating drug development for rare diseases. The clinical potential of Apc003OA offers new therapeutic hope for DMD patients and provides a replicable precision intervention strategy for other genetic muscle disorders characterized by fibrosis. This achievement marks an efficient translation from laboratory-based mechanistic exploration to clinical candidate, positioning itself as a potential cornerstone in future comprehensive DMD management.
