Advancing adeno-associated virus (AAV) gene therapy for Duchenne muscular dystrophy

Abstract

Duchenne muscular dystrophy (DMD) is a lethal degenerative muscle disease. DMD results from mutations preventing the production of the subsarcolemmal protein dystrophin, a protein critical for muscle stability. In the absence of dystrophin, muscle fibers are damaged and degenerate, leading to an inflamed and fibrotic muscle phenotype. DMD patients have a life expectancy of 20 to 40 years old and succumb to cardiac and respiratory failure. Adeno-associated virus (AAV) is a promising vector to deliver gene therapies (such as truncated dystrophin proteins, i.e., microdystrophin) to the muscle of DMD patients, due to its safety profile and broad tissue tropism. Currently, naturally occurring AAV variants including serotypes 8, 9, and rh74 are used in clinical trials, and an AAVrh74-based microdystrophin therapy (Elevidys) was approved in 2023 to treat DMD. Despite regulatory success, limited results are available from patients treated with Elevidys and in other clinical trials, and many questions remain about the biology of AAV vectors in muscle. Individual muscle cells, or myofibers, have different phenotypes (myofiber type), closely linked to their genetically regulated expression of different isoforms of the contractile protein myosin heavy chain. Based on previous literature, we hypothesized that AAV serotypes 8 and 9 may have a selective tropism for certain myofiber types, or restricted tropism in others. Fast-twitch myofiber types are susceptible to greater injury in DMD, and thus, benefit greater from microdystrophin gene transfer. To determine the myofiber type preference AAV serotypes 8 and 9 we performed histological analysis of reporter gene and microdystrophin expression in archival tissues from preclinical studies in dystrophic canines (cDMD model). We did not detect a myofiber type preference. However, the results from our histological analysis indicated that microdystrophin expression is maintained long-term (8-40 months following systemic delivery); and that micro-dystrophin expressing fibers were protected from atrophy. These results offer important insight for clinical development of microdystrophin gene therapies. Selecting the best AAV variants for muscle-directed gene therapy is a critical step in developing effective treatments. Currently, there is no way to evaluate systemic delivery of AAV vectors to human muscle. To overcome this barrier, we adapted a human skeletal muscle xenograft model, to use as a platform for AAV delivery. We characterized this model using 222 transplanted muscles from human donors, and developed rigorous methods to evaluate muscle transplant quality, and AAV transduction of the muscles. We showed for the first time, that human muscles transplanted in immune-deficient mice are efficiently transduced by AAV vectors at following intravenous injection of the host mice. We first used this platform to compare the three AAV serotypes in clinical trials (8, 9, and rh74). We found that all three have similar delivery efficiency (i.e., infection) in human muscles, but only AAV9 led to high-level mRNA production. These results not only present a novel finding on the superior efficiency of AAV9 but also highlight a new way to evaluate AAV vectors in a preclinical context. Other human transplantation models have been used to discover next-generation AAV serotypes for liver-directed gene therapy. We expect that the human muscle xenograft model has similar potential for discovery of serotypes with enhanced efficiency in human muscle.