Researchers at UCL developed a hydrogel-based axon model to improve early testing for multiple sclerosis remyelination therapies. Published in Nature Methods, the platform uses tunable polyacrylamide hydrogel micropillars to mimic the geometry and approximate softness (~5 kPa) of native axons, aiming to address failures that arise when drug candidates are screened on overly rigid in vitro substrates. The team seeded human and rodent oligodendrocytes and reported myelin formation around the tuned pillars. When pillars were adjusted to realistic softness, drug performance dropped—supporting the premise that rigid lab models can produce misleading “hits.” For translational oncology-adjacent neurology programs, the immediate impact is a potential upgrade to preclinical screening fidelity, enabling better selection of remyelination candidates before expensive clinical evaluation.