Fig 2: 144DG11 interacts with LAMP1 and its associated interactome AHeteroassembly forms around 144DG11 and not around endogenous molecules as shown by the liquid crystals formed in experiments 1‐3 (see Materials & Methods).BSTRING network of targets at the interactome of 144DG11. Targets also show strong protein–protein interactions among themselves, as connectors show (see legend). SLC17A5, Solute Carrier Family 17 Member 5; LAMP1, Lysosomal Associated Membrane Protein 1; LAMTOR4, Late endosomal/lysosomal adaptor, MAPK and MTOR activator 4; GBA, Glucosylceramidase; SCARB2 (aka LIMP2), Scavenger Receptor Class B Member 2; DPP4, Dipeptidyl peptidase‐4; GAA, Lysosomal alpha‐glucosidase; HEXA, Beta‐Hexosaminidase Subunit Alpha.CCellular thermal shift assay (CETSA) of different targets of the 144DG11heteroassembly (B). Only LAMP1 was significantly protected by 144DG11 from heat‐mediated denaturation (see right shift in its Tm in the lower panel) suggesting its specific interaction with 144DG11. n = 3 biological replicates. Error bars represent S.D.D–FSurface plasmon resonance assays demonstrate dose and pH‐responsive interaction between LAMP1 and 144DG11. Full interaction (but with apparent KD of 6.3mM) was demonstrated only at the lysosomal pH 4.5‐5. (E) Upper panel, LAMP1 was deglycoslated as detailed in Materials & Methods. RNase B is a positive control for a glycosylated protein. Glycosylation status is shown after short (24 h) or long (72 h) dialysis. The results demonstrate a full deglycosylation of both LAMP1 luminal part and RNase B, with unique bands appearing after a long dialysis. Lower panel, surface plasmon resonance, performed as in (D), showing that the luminal domain of deglycosylated LAMP1 specifically binds 144DG11 with apparent KD of 52.5 nM. (F) Three binding modes of 144DG11 (gray) according to LAMP1 grids predicted by SiteMap (green), fPocket (cyan) and FtSite (magenta). Two out of three binding modes (SiteMap and fPocket) are identical. In FtSite part of the molecule went through a rotation relative to the other two. See Materials & Methodss5.

Fig 3: LAMP1‐KD and 144DG11 enhance autophagic flux Left panel, Autophagic flux, determined by the extent of lysosomal inhibitor‐dependent increase in the ratio of lipidated to non‐lipidated LC3 (LC3II/LC3I), is increased by 144DG11 (50µM, 24h). Right panel, 144DG11‐mediated increase in autophagic flux, demonstrated by enhanced degradation of the autophagy substrate p62. Shown are representative immunoblots (grouped from different gels) and densitometric quantifications ±s.d. (for LC3II/LC3I, n = 4 biological replicates, *P < 0.0023; for p62, n = 5 biological replicates, *P < 0.0099; two‐tailed t‐test). Left panel, TEM images of liver tissue from vehicle or 144DG11‐treated Gbeys/ys mice. High‐magnification (right, scale bars=200 nm) and low‐magnification (left, scale bars=1 µm) images show higher levels of glycogen/polyglucosan in lysosomes and cytosol, respectively. Right panel, quantification (± s.d) of lysosomal glycogen particles (n = 3 biological replicates, *P < 0.03, two‐tailed t‐test). G, Glycogen/polyglucosan; L, Lysosomes; M, Mitochondria.144DG11 reduces LC3 and p62 in mouse liver, but not muscle (n = 3 biological replicates, *P < 0.04 for LC3, **P < 0.0007 for p62; two‐tailed t‐test, SEM). Scale bars, 10 µm.LAMP1 knockdown increases autophagic flux, which is further facilitated by 144DG11 (see densitometric quantification ±SEM (n = 3 biological replicates, *P < 0.1; **P < 0.05; ***P < 0.01 (two‐tailed t‐tests)). Immunoblots show LAMP1, LC3II/I, and actin in LAMP1 knocked down and control APBD fibroblasts treated or not with 144DG11 and lysosomal inhibitors (LI).LAMP1‐KD and 144DG11 treatment cause lysosomal acidification. Upper panel, flow cytometry results showing that 144DG11 slightly increased acidification (yellow to blue median fluorescence ratio (Y/B)) in control, GFP‐transduced, APBD fibroblasts (Y/B(GFP/144DG11)>Y/B(GFP), P < 0.12), but significantly acidified LAMP1‐KD, GFP‐shLAMP1‐transduced, APBD fibroblasts (Y/B(LAMP1‐KD/144DG11)>(Y/B(LAMP1‐KD), P < 0.03). LAMP1‐KD itself led to the most significant acidification (Y/B (LAMP1‐KD)>Y/B(GFP), P < 0.007). n = 3, two‐tailed t‐tests. Middle panel Lysosensor staining of the corresponding cells. Yellow fluorescence intensity correlates with acidification. Lower panel, PAS (glycogen) staining of the corresponding cells. Scale bars, 50 µm.ATP production rates (Fig 4B) ± SEM in serum‐starved (−S) and non‐starved (+S) LAMP1‐KD and GFP (Control) cells, untreated (UT), or treated for 24 h (CHR, chronic) or on assay (acute) with 144DG11. See text (144DG11 enhances LAMP1 knockdown‐induced autolysosomal degradation and catabolism of glycogen, third paragraph) for results of statistical analysis (multiple t‐tests, one‐way ANOVA with Sidak’s post hoc corrections, n = 3 biological replicates).144DG11 reduces lysosomal (LAMP1 positive) area in liver, but not muscle, of Gbeys/ys mice (n = 3 biological replicates, *P < 0.01, two‐tailed t‐test, SEM).Scale bars, 5 µm for liver (upper left panel), and 10 µm for muscle (lower left panel). Right panel shows quantification of the left panel. Source data are available online for this figure.

Fig 4: Integrated analysis of single-cell RNA sequencing and spatial transcriptomics reveals elevated LAMP1 expression in MDSCs and CAFs with enrichment in tumor regions compared to adjacent normal tissue. (A) K-means clustering analysis of single-cell RNA sequencing data highlights distinct cellular populations within the tumor microenvironment. Normalized log2 expression levels of CD68 (B), LAMP1 (C), CD74 (D), and FXYD3 (E) across all identified cell populations. (F) Quantitative analysis shows high expression of LAMP1 in cancer cells as well as macrophages and CAFs, which are prominent components of the tumor microenvironment. (G) Subtype clustering of macrophages identifies functional subsets, allowing for finer delineation of LAMP1 expression within specific macrophage populations. (H) Normalized log2 expression of LAMP1 within macrophage subtypes highlights differential expression patterns. (I) LAMP1 expression was significantly elevated in the MDSC subtype of macrophages. (J) Combined analysis of tumor marker expression within the spatial transcriptomics dataset reveals high expression of FXYD3, human epidermal growth factor receptor 2 (HER2), estrogen receptor (ER), MKI67, topoisomerase II alpha (TOP2A), and epithelial cell adhesion molecule (EPCAM). (K) Annotation of cancer cell regions in breast cancer tissue highlights tumor-specific regions. Spatial expression maps show CD74 (L), CD68 (M), and LAMP1 (N) in breast cancer tissue, with notable localization in cancer cells and tumor-associated immune cells. (O) Spatial co-expression analysis of CD74, CD68, and LAMP1 with a magnified view and (P) comparative analysis of LAMP1 expression between cancerous regions and adjacent normal tissue confirms a significant elevation of LAMP1 in tumor regions. ***: p < 0.001.

Fig 5: Assessment of LAMP1 expression in human pan-cancer samples and tumor-bearing murine model. (A) Representative immunofluorescence staining of LAMP1 (red channel) in major carcinomas and their respective normal tissue. Nuclei are stained with DAPI (blue channel). (B) The overall LAMP1 fluorescent signal intensity in tumor cores (red) versus normal cores (green). (C) Organ-based comparison of fluorescent signal intensity demonstrated significantly higher LAMP1 fluorescence in prostate, pancreas, colon, breast, and uterine endometrium carcinomas (red) compared to normal tissue (green). *, p < 0.05; **, p < 0.001; ****, p < 0.0001.