177P Effect of gene therapy on the epigenetic and transcriptional stability of SMA

• Published in: Neuromuscular disorders : NMD Vol. 43; p. 104441
• Main Authors: Smeriglio, P., Grandi, F., Arnould, A., Pezet, S., Marais, T., Mazzucchi, S., Astord, S., Cohen-Tannoudji, M.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2024
• ISSN: 0960-8966
• DOI: 10.1016/j.nmd.2024.07.621

Adeno-associated virus (AAV) based gene therapies are poised to have a large impact on rare diseases. ZOLGENSMA treatment has been life altering for spinal muscular atrophy (SMA) Type I infants treated perinatally, with incredible improvements in life expectancy and gain of developmental milestones. While this therapy is clearly life changing, it also highlights the importance of understanding how AAV-mediated gene therapies are regulated after injection, especially in the case of childhood diseases that require stable and reliable gene expression over a lifetime. The epigenetic landscape, which consists of chemical modifications to DNA and histones, ensures the correct temporal regulation of gene expression and changes to the epigenetic landscape have been found to contribute to many diseases. However, the effect(s) of gene therapies like ZOLGENSMA on the long-term stability of the epigenome is unknown. The goal of this work is to understand how the epigenome reacts to ZOLGENSMA administration and to use this information to create better gene therapies. We had previously shown that ICV injection of AAV9-PGK-SMN1 resulted in efficient transduction of the brain, spinal cord, muscle and other peripheral tissues, with a viral genome copy number/ nucleus of ∼40 in the spinal cord and muscle, allowing us to study the effect AAV therapy in these tissues. To study the epigenetic landscape, we began by profiling 5-hydroxymethylcytosine (5hmC) location and chromatin accessibility of treated and untreated tibialis anterior muscle. 5hmC is a chemical modification to the cytosine base, whose accumulation at enhancers and gene bodies allows for gene expression. Thus, by profiling the 5hmC landscape in combination with RNA-sequencing, we can observe genes which are currently active (5hmC + RNA expression) and those that will be activated (5hmC only). As expected, we observed vast transcriptional and 5hmC differences between SMA and WT muscle. In line with our physiological data, gene therapy restored the transcriptome of the SMA mice at postnatal day 14 and most of the SMA-disease associated 5hmC changes were restored by the AAV therapy at an early timepoint, in line with the restoration of the transcriptome. However, the 5hmC profile of the injected mice also contained new changes not observed in the WT or SMA mice, suggesting that these 5hmC-changes may result from the AAV injection. Indeed, we observed similar changes in WT AAV9-GFP injected mice. These 5hmC alterations affected genes involved in innate immune activation, apoptosis, and metabolism. Ongoing work is tracing these effects long-term (4.5 and 6.5 months after mice treatment), when variable responses to gene therapy in our animal cohort are evident. Preliminary results suggest that while gene therapy ensures short-term transcriptional restoration, many aspects of the epigenetic landscape are not fully restored and they are predictive of the variable long-term response. Ongoing work is testing the effect of the route of injection, the capsid type and the promoter used for expression on these parameters, as well as determining the molecular mechanisms that explain the heterogeneity in treatment response.