Fig 2: Transcriptome profiling of the VT retina in Slc6a4-/- and WT mice after injury.(A) Schema of RNA preparations and sequencing for transcriptome profiling of the VT retina in Slc6a4-/- and WT mice one day after ONC. (B) Enriched cellular processes and pathways in the VT Slc6a4-/- retina compared to WT retina one day after ONC are shown. Bars indicate the degree of -log10 P value. (C-D) Heatmap (C) and volcano plots in red (D) showing the most significant 29 genes (FDR value < 0.05 and change in expression of more than 1.5-fold) in the VT retina of Slc6a4-/- compared to WT mice one day after ONC (n = 3 samples/condition, each sample includes three VT retinal tissues of three Slc6a4-/- or WT mice). In the heatmap (C), color values indicate log2-transformed expression values shown in the color key (bottom) relative to the average expression in WT mice. The columns show individual samples from Slc6a4-/- and WT mice in triplicate. (E) GPNMB is expressed in RBPMS+ RGCs of the VT and DN retina. (F) GPNMB+ cells are reduced in the VT retina of WT mice two days after ONC compared to the intact retina, while GPNMB+ cells in the DN retina of WT mice or the VT retina of Slc6a4-/- or Slc6a4::Cre;Itgb3flox/flox mice are still clearly detected after ONC. (G) Quantitative analysis of GPNMB+ cell detection rate two days after ONC compared to the intact counterpart (n = 4 mice/condition, one-way ANOVA). Data presented as mean ± SD. Scale bars represent 20 µm (E, F).

Fig 3: HEPES drives lysosomal biogenesis in cultured RAW264.7 macrophages. (A) Flow cytometric analysis (FL1) of LTG-stained RAW cells grown in either DMEM (31966), DMEM (32430; containing HEPES), RPMI (61870), or RPMI (22409; containing HEPES). (B) Time-course analysis of LTG staining in cells grown in DMEM supplemented with HEPES (25 mM) for 6–72 h. RPMI-grown cells served as a positive control. (C) Fluorescence microscopy analysis of LTG-stained RAW cells cultured in DMEM or DMEM+H for 24 h. (D-E) RAW cells were adapted to grow in DMEM (32430; containing HEPES) for 7 d, after which culture media were replaced by HEPES-free DMEM (31966) for 6–72 h. A time course for (D) LTG staining and (E) Immunoblot analysis of GPNMB and CTSD protein levels. (F) Transmission electron microscopy (TEM) analysis of RAW cells grown in either DMEM or DMEM+H for 24 h. Scale bar: 1 µm. (G) Phase-contrast and immunofluorescence microscopy analysis of LAMP1-stained cells. (H) Analysis of GBA1 enzymatic activity using a 4-MU-based assay in RAW cells grown in DMEM or DMEM+H for 24 h. (I) Schematic illustration of the DQ-BSA reagent used for quantifying lysosomal proteolytic activity. DQ-BSA added to culture media is rapidly endocytosed, but only emits a fluorescent signal after cleavage by proteases inside lysosomes. (J) Flow cytometric analysis of DQ-BSA cleavage (FL1) in RAW cells grown in DMEM or DMEM+H for 24 h. (K) Western blot analysis and (L) quantification of LC3-II protein levels in RAW cells grown in DMEM or DMEM+H for 24 h, and where indicated treated with bafilomycin A1 (BAF A1; 100 nM) for the last 2 h. Values are expressed as mean ± SEM, n = 3-4 in A-L. **P<0.01.

Fig 4: HEPES affects host-immune gene programs in RAW macrophages. (A-B) Rank-rank hypergeometric overlap (RRHO) analysis comparing the gene ranking (log FC) affected by HEPES (relative to DMEM) to an (A) M1 polarization state induced by LPS (100 ng/ml), or an (B) M2-specific state induced by IL4 (50 ng/ml). Pixel values in the RRHO map represent the log10-transformed hypergeometric overlap of subsections of 2 ranked gene lists (step size 100 genes). Red values indicate a higher than expected number of overlapping genes in the subsections, and blue values signify a lower than expected overlapping gene number. Below the heatmaps, the metric values (log FC) used for the differential expression levels are plotted in a bar graph along x- and y-axes. A scatter plot (A-B) of the datasets is shown for comparing the RRHO map to a standard metric of correlation (Pearson). The RRHO result and Pearson correlation coefficient reflect a similar relationship. (C) Cytokine array blots on culture media derived from DMEM or DMEM+H-grown cells for 24 h. Secreted cytokines and chemokines in culture media were detected using a Mouse Cytokine Array kit and quantified with the Odyssey V3.0 software (fold-increase relative to DMEM ctrl). Secreted GPNMB levels were measured as a positive control. (D-E) RT-PCR analysis of the specified (D) M1- and M2-specific markers and (E) MiT/TFE target genes in RAW cells grown in DMEM or DMEM+H for 24 h and pulsed with either vehicle Ctrl (-), LPS (100 ng/ml) or IL4 (50 ng/ml) for another 24 h. Gene expression was normalized to Rplp0. Values are expressed as mean ± SEM, n = 3-4 in A-E. *P < 0.05, **P < 0.01. N.S., nonsignificant.

Fig 5: GPNMB/IBA-1-positive microglia cells cluster around individual plaque cores in 5XFAD brains. Triple immunofluorescence staining using antibodies against GPNMB (a), Aß (b) and IBA1 (c) demonstrated the spatial co-localization of GPNMB-positive microglia cells around amyloid plaque cores in 12-month-old 5XFAD brains (d). Even though APP23 mice showed numerous activated microglia cells (g) clustered around amyloid plaques (f), no GPNMB signal could be detected (e,h). (i) RT-PCR analyses revealed significantly increased mRNA levels of CST7, TREM2, APOE, CLEC7A and CCL2 in 5XFAD brains when compared to WT and APP23 mice. However, levels of AIF1 and TMEM119 were comparable in all groups tested. All data are given as mean ± SD. ***P < 0.001; **P < 0.01. Scale bar: A-H = 33 µm