LASER couples damage sensing to ESCRT assembly for lysosome repair

Article LASER couples damage sensing to ESCRT assembly for lysosome repair https://doi.org/10.1038/s41586-026-10604-6 Received: 26 August 2025 Accepted: 28 April 2026 Published online: xx xx xxxx Open access Check for updates Claire S. Goul1,6, Aakriti Jain1,5,6 ✉, Samira Yitiz2,3, Zahra E. Soltani4, Serim Yang1, Simon Rapp1, Martina Spacci1, Scot Federman2,3, James Sacco2,3, Huinan Li2,3, Lauren D. Enriquez2,3, Nalan Liv4, Laralynne Przybyla2,3 & Roberto Zoncu1 ✉ Lysosomal membrane integrity is essential for cell survival, but how damage sensing is spatiotemporally coupled to repair remains poorly understood. Recruitment and assembly of endosomal sorting complex required for transport (ESCRT) I–III rapidly counteracts membrane damage, but it is unclear how ESCRT-I recognizes defective lysosomal membranes. Here, leveraging genome-wide CRISPRi screens in a damagesensitized genetic background, we identified LC3/GABARAP-assisted stimulator for ESCRT recruitment (LASER), a multicomponent protein assembly that forms rapidly upon calcium release from damaged lysosomes and couples sensing of lysosomal membrane damage to ESCRT-dependent repair. At the core of LASER is TFG, an endoplasmic reticulum exit-site-resident protein that translocates to damaged lysosomes by binding to ATG8 family proteins (LC3 and GABARAP) conjugated to lysosomal phospholipids. ATG8-bound TFG forms oligomeric assemblies that directly recruit the essential ESCRT-I subunit TSG101 via conserved motif recognition enhanced by avidity-driven interactions. TFG binding to TSG101 stimulates sequential ESCRT-I– II–III polymerization and promotes membrane repair. TFG mutations that drive hereditary spastic paraplegia disrupt its oligomerization and impair lysosomal ESCRT recruitment and membrane resealing, implicating defective repair as a driver of TFG-associated neurodegeneration. Thus, LASER promotes ESCRT polymerization at damaged lysosomes and couples damage sensing to membrane repair. Lysosomal membrane integrity is essential for cellular homeostasis and survival, yet lysosomes are continuously exposed to stresses that threaten membrane stability, including lipid oxidation, lipid asymmetry and the accumulation of aggregation-prone luminal cargo1. In response to membrane disruption, cells activate a multi-tiered repair system that monitors and restores lysosomal integrity, preventing leakage of luminal contents and the resulting cell death1. Lysosomal membrane damage occurs along a continuum, ranging from subtle loss of ion gradients and lipid packing defects to overt membrane rupture, raising the possibility that distinct damage-sensing and repair mechanisms operate at early versus advanced stages of disruption2–6. The membrane remodelling ESCRT-I, -II and -III complexes provide one of the earliest-acting repair mechanisms, undergoing sequential assembly and oligomerization on damaged lysosomes that help to prevent, stabilize or reseal membrane tears7–10. Accordingly, depletion of the essential ESCRT-I subunit tumor susceptibility gene 101 (TSG101) compromises ESCRT polymerization and leads to accumulation of severely damaged lysosomes7,8. However, despite its obligate role in membrane repair, how sensing of lysosomal damage is coupled to ESCRT recruitment remains unclear. In contrast to intraluminal vesicle formation, in which ESCRT-0 recognizes ubiquitylated endocytosed proteins to initiate sequential ESCRT-I–II–III polymerization11, ESCRT-0 is dispensable for lysosomal membrane repair, and ubiquitylation is instead detected during bulk removal of irreversibly damaged lysosomes (lysophagy)2,7,12,13. The calcium sensor apoptosis linked gene-2 (ALG2) was proposed to recruit ESCRT to damaged lysosomes in a calcium-dependent manner14,15. However, in vitro and cell-based evidence suggests that the requirement for calcium-ALG2 in ESCRT polymerization at damage sites may be insult-specific and not universal14–16. Recently, recruitment of ESCRT subunits and of the accessory ESCRT assembly factor ALG2-interacting protein X (ALIX) to damaged lysosomes was shown to require conjugation of ATG8 proteins to single membranes (CASM). During CASM, ATG8 proteins are conjugated to lysosomal membrane phospholipids by the E3-like ATG5–ATG12 complex, which is recruited to damaged lysosomes either by ATG16L1, bound to stalled vacuolar ATPases, or by tectonin β-propeller repeat-containing 1 (TECPR1), bound in turn to inner leaflet sphingomyelin that becomes exposed to the cytosol15,17–21. Despite evidence for both ATG16L1- and TECPR1-dependent CASM contributing to ESCRT-I recruitment, the specific molecular links between lysosome-bound ATG8 and ESCRT-I subunits have not been identified. Another unresolved question is the relationship between ESCRT assembly and the extensive endoplasmic reticulum (ER)–lysosome 1 Department of Molecular and Cell Biology, University of California at Berkeley, Berkeley, CA, USA. 2Laboratory for Genomics Research, San Francisco, CA, USA. 3Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, CA, USA. 4Center for Molecular Medicine, University Medical Center Utrecht, Institute of Biomembranes, Utrecht University, Utrecht, The Netherlands. 5Present address: Department of Physiology, University of Texas Southwestern Medical Center, Dallas, TX, USA. 6These authors contributed equally: Claire S. Goul, Aakriti Jain. ✉e-mail: ; Nature | www.nature.com | 1 Article membrane contact sites that are established upon lysosomal damage22–24. The proximity created by these contacts enable the ER-to-lysosome transfer of phospholipids by oxysterol binding protein-related proteins (ORPs) and bridge-like lipid transport protein (BLTP) family proteins, which together help restore the optimal topology, surface area and composition of the lysosomal limiting membrane20,22–24. However, whether ER–lysosome contacts have additional repair-promoting roles, specifically whether and how they contribute to the regulation of ESCRT assembly on damaged lysosomes, is currently unknown. Neuronal lysosomes operate close to a damage threshold and are therefore uniquely reliant on efficient membrane repair3,25. Excess or aberrant proteolytic cargo, compounded by genetic or environmental perturbations in cholesterol and sphingolipid metabolism, accelerates membrane injury26–28. Increasing evidence suggests that even healthy neurons undergo constitutive, albeit transient, lysosomal permeabilization3,5,29. Mutations in several components of the lysosomal repair machinery are increasingly linked to neurodegenerative conditions. Most notably, the ESCRT-III subunit charged multivesicular body protein 2B (CHMP2B) is mutated in amyotrophic lateral sclerosis and frontotemporal degeneration11,30,31, whereas mutations in the ESCRT-I subunit ubiquitin associated protein 1 (UBAP1) underlie hereditary spastic paraplegia (HSP)32. However, given the pleiotropic roles of ESCRT in membrane remodelling (...truncated)