Back

Acid sphingomyelinase inhibition restores RPE homeostasis and photoreceptor function in preclinical Stargardt macular degeneration models

Germer, C. J.; Williams, S.; Fernandes, V.; Espinosa Lima, R.; La Cunza, N. R.; Tan, L. X.; Panda, S. R.; Iorio, E.; Elahi, F.; Lakkaraju, A.

2026-02-22 neuroscience
10.64898/2026.02.20.706984 bioRxiv
Show abstract

Stargardt disease, which destroys central high-resolution vision in over 2 million people globally, lacks effective therapies. The primary site of damage in Stargardt disease is the retinal pigment epithelium (RPE), which safeguards photoreceptor health and function. Progressive loss of RPE integrity precedes visual deficits, yet insight into mechanisms driving RPE dysfunction and how this influences disease pathogenesis remains elusive. Here, we addressed this in cell-based and pigmented Abca4-/- Stargardt mice models using super-resolution imaging, bioinformatics, and biochemical approaches. We show that ceramide accumulation induced by bisretinoid-mediated overactivation of acid sphingomyelinase (ASM) in Abca4-/- RPE selectively disrupts Rab GTPases and ESCRT machinery involved in apical membrane trafficking and small extracellular vesicle (EV) biogenesis. Consequently, connexin 43 (Cx43) is misrouted from cell-cell junctions into EVs that are released apically by the RPE. This compromises RPE integrity and promotes subretinal immune cell recruitment, leading to photoreceptor dysfunction. Pharmacological ASM inhibition normalizes EV biogenesis and restores Cx43 localization. Decreasing RPE ceramide safeguards RPE structural integrity, limits subretinal microglia, and improves visual function in Abca4-/- mice. This study underscores the importance of the RPE as a communication hub in the retina and identifies ASM as a potential therapeutic target to prevent progressive vision loss.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.