LAUSR

Gene therapies to target disorders of the central nervous system (CNS) are an evolving field. Viral vectors, such as recombinant adeno-associated viruses (AAVs), are increasingly used because of their low pathogenicity and long-term transgene expression. Most AAV serotypes have restrictive packaging capacity and broad tissue infectivity, which limits their ability to carry large transgenes and accurately target specific tissues and organs. Moreover, they have a propensity for non-CNS tissues, particularly the liver. Reduced transduction to the CNS posed by the blood–brain barrier has traditionally required either intracranial or intrathecal administration, but such approaches are invasive and cannot treat pathologies that require broad coverage (eg, to the whole cortex or striatum). However, an AAV9 serotype with improved CNS targeting was used in Zolgensma (onasemnogene abeparvovec), the first approved virus-mediated treatment for any neurological disorder, to replace missing or defective survival motor neuron-1 (SMN1) in spinal muscular atrophy. Further engineering of this variant resulted in AAV-PHP. eB, a capsid subtype with better CNS transduction efficiency, but with the persistent limitation of off-target binding. To develop AAV serotypes with more precise CNS targeting, Goertsen and colleagues employed MultiplexedCre-recombination-based AAV targeted evolution (M-CREATE), a newly-described method of capsid engineering. This system implements multiple positive and negative selection criteria to detect novel capsid variants that target the CNS while avoiding peripheral binding. Sequential engineering of AAVPHP.eB between the surface-exposed AA452 and AA460 of VP3 identified the AAV.CAP-B10 variant, which can robustly transduce the cortex, subcortical region, cerebellum, spinal cord, and dorsal root ganglia in both C57BL/6J mice and adult marmosets following intravenous administration. Notably, AAV. CAP-B10 demonstrated reduced liver targeting compared to other AAV variants. This study highlights how iterative modifications of the viral capsid can increase cellular and regional specificity of transduction towards tissues of interest, while decreasing off-target transduction. It further demonstrates that using M-CREATE as a screening platform prior to validation in non-human primates (NHPs) can increase the efficiency of engineering viral vectors. Developing AAVs with specific CNS tropism could facilitate development of gene therapies for movement disorders. However, conceptual challenges like off-target effects from global CNS transduction need to be addressed before strategies such as gene silencing in Huntington’s disease and disease modification in non-monogenetic causes of Parkinson’s disease can be safely employed. Nevertheless, this study demonstrates notable progress towards safe, efficient, and effective delivery of gene therapies to the CNS of NHPs, and is ultimately promising for their benefit in humans.