CAMBRIDGE, Massachusetts; ROTTERDAM, Netherlands; and SHANGHAI, China, May 18, 2026
Harbour BioMed announced promising preclinical data for LET003, its first AI-enabled monoclonal antibody drug candidate developed using the company’s proprietary Hu-mAtrIx™ artificial intelligence platform. The next-generation therapy targets ACVR2A and ACVR2B receptors, key regulators of fat and muscle metabolism, and demonstrated significant potential in enhancing fat reduction while preserving lean muscle mass when combined with semaglutide, a leading GLP-1 obesity therapy.
AI-Designed Antibody Shows Potential Best-in-Class Obesity Profile
According to Harbour BioMed, LET003 displayed superior pharmacokinetic properties compared with several competing molecules in both mouse and cynomolgus monkey studies. Researchers observed significantly slower blood clearance rates following subcutaneous administration, suggesting the therapy may achieve effective results using lower doses or less frequent dosing schedules compared with rival obesity treatments.
In preclinical obesity studies using wild-type mice, combining LET003 with semaglutide produced a 76% reduction in fat mass versus vehicle controls and significantly outperformed semaglutide alone. Importantly, the combination therapy also preserved lean muscle mass more effectively than semaglutide monotherapy, addressing one of the major limitations associated with current GLP-1 weight loss drugs.
Additional experiments using high-fat diet obesity models further confirmed that the combination reduced body fat ratios while significantly increasing lean mass ratios compared with semaglutide alone. Researchers stated the findings support LET003’s potential to improve overall body composition management, a growing focus in next-generation obesity drug development.
LET003 Demonstrates Strong Lean Muscle Preservation Effects
Advertisement
Harbour BioMed also compared LET003 directly with bimagrumab, an earlier antibody targeting the same metabolic pathway. Results showed LET003 achieved comparable lean mass-promoting effects at a much lower dose level, highlighting the drug’s strong pharmacological efficiency.
In normal-diet animal models, LET003 significantly increased lean muscle mass and overall body weight compared with both vehicle controls and competitor molecules. The company said the data suggest the antibody may offer meaningful advantages for patients requiring weight reduction without excessive muscle loss, a critical issue in obesity treatment and metabolic disease management.
The ACVR2A/2B signaling pathway has become an increasingly important target in obesity and metabolic research because of its role in balancing fat accumulation and muscle maintenance. Analysts believe therapies capable of reducing fat while preserving or increasing lean mass could become major competitors in the rapidly expanding global obesity drug market currently dominated by GLP-1 therapies.
Hu-mAtrIx AI Platform Accelerates Drug Discovery Strategy
Harbour BioMed emphasized that LET003 is the first ACVR2A/2B dual-target antibody generated through its Hu-mAtrIx™ AI-driven antibody engineering platform. The system combines proprietary Harbour Mice® data, large language model-based sequence generation, and AI prediction tools designed to optimize antibody development and improve drug candidate performance
Dr. Jingsong Wang, Founder, Chairman, and CEO of Harbour BioMed, stated that the company is encouraged by LET003’s favorable preclinical profile and plans to advance the candidate rapidly into clinical development. The company believes the data further validate the growing role of artificial intelligence in biologic drug discovery and metabolic disease research.
The announcement highlights the increasing integration of AI technologies within the global biopharmaceutical sector, where companies are using machine learning and computational biology to accelerate the development of next-generation therapeutics for obesity, oncology, and immune-related diseases.
Source: Harbour BioMed press release
