Oxysterol-binding protein ORP6 regulates lipid metabolism and brain Aβ production

Abstract

The mammalian brain is the most cholesterol-rich organ of the body, relying on in situ de novo cholesterol synthesis. Maintaining cholesterol homeostasis is crucial for normal brain function. Oxysterol-binding protein (OSBP)-related proteins (ORPs) are highly conserved cytosolic proteins that coordinate lipid homeostasis by regulating cell signaling, interorganelle membrane contact sites, and non-vesicular transport of cholesterol. Here, we show that ORP6 is highly enriched in the mammalian brain, particularly within neurons and astrocytes, with widespread expression across distinct brain regions, including the hippocampus, which is essential for learning and memory. Whole-body ablation of ORP6 (Osbpl6^-/-) in mice resulted in dysregulation of systemic and brain lipid homeostasis, with elevated levels of brain desmosterol and amyloid-beta oligomers (AβOs). Mechanistically, ORP6 knockdown in astrocytes altered the expression of cholesterol metabolism genes, promoting the accumulation of esterified cholesterol in lipid droplets, reducing cholesterol efflux and plasma membrane cholesterol content, and increasing amyloid-beta precursor protein (APP) processing. Our findings underscore the role of ORP6 in systemic and brain lipid homeostasis, highlighting its importance in maintaining overall brain health.

Keywords: amyloid beta; astrocyte; cholesterol efflux; high-density lipoprotein; lipid droplet; lipid metabolism; oxysterol-binding protein-like 6.

Fig. 1

Loss of ORP6 in mice alters plasma lipid and lipoprotein profiles. A: Immunoblot of ORP6 and total protein in organs from WT and Osbpl6^−/− mice. B: Viability of WT, Osbpl6^+/− and Osbpl6^−/− mice shown as a percentage from heterozygous crosses. C: Volcano plot of plasma lipid altered in Osbpl6^−/− mice versus WT mice, with log2 fold change (FC) against -log10 P-value. Significance was determined by t test (P < 0.05), with FC thresholds set at <0.8 and >1.5 (n = 6 per genotype). Colored points represent significantly dysregulated lipid species by class. D: Heatmap of significantly different lipid entities annotated by MS/MS normalized to the area under the curve between WT and Osbpl6^−/− mice (n = 3 per genotype per sex). E: Triglyceride, (F) Total cholesterol, and (G) HDL cholesterol levels in WT and Osbpl6^−/− mice, stratified by sex (n = 19–20 WT, n = 8 Osbpl6^−/− per sex). ∗P < 0.05, ∗∗P < 0.005 by one-way ANOVA with Tukey’s multiple comparisons test (E–G).

Fig. 2

Altered CNS lipid species and biological processes in ORP6^−/− mice. A: Volcano plot showing altered lipid species in the brains of Osbpl6^−/− mice versus WT mice, with log2 fold change (FC) plotted against -log10 P-value. Significance was assessed using a t test (P < 0.05), with FC thresholds set at <0.8 and >1.5 (n = 6 per genotype/sex). Colored points indicate significantly dysregulated lipid species by class. B: Heatmap of significantly altered brain lipid species annotated by MS/MS normalized to the area under the curve between WT and Osbpl6^−/− mice (n = 3 per genotype/sex). C: Normalized MS signal intensity (Log 2) of desmosterol (mean ± s.e.m, n = 6 per genotype). ∗P < 0.05. D: Volcano plot of altered protein expression in male Osbpl6^−/− versus WT brains from unbiased proteomics analysis, showing log2 fold change against -log10 adjusted P-value (n = 5 per genotype/sex). E: Treemap of altered biological processes, cellular components, and molecular functions in WT and Osbpl6^−/− brains from unbiased proteomics analysis (n = 5 per genotype/sex). F: Whole brain amyloid-beta oligomer (AβOs) concentrations (pg/mg protein) in age-matched WT and Osbpl6^−/− mice at 16 weeks (n = 5 per genotype/sex, mean ± s.e.m). ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.0005 by Student’s t test (C, F).

Fig. 3

ORP6 ablation in mice leads to brain hypotrophy and impaired neuromuscular function. A: Differential expression of ORP6 between AD and control cases across brain regions, shown as Log2 fold-change ± s.e.m. using the AGORA web application. B: Differential expression of Osbpl6 in the prefrontal cortex of AD versus healthy patients, expressed as Log2 fold change across brain cell types from snRNA-Seq data (Lau et al. (17)). C: qRT-PCR quantification of ORP6 mRNA in hippocampi of WT and APPswe mice (n = 4 per genotype, mean ± s.e.m). D: Immunoblotting of ORP6 and Plin2 in hippocampal tissue from WT and APPswe mice. E-F: Densitometry of ORP6 (E) and Plin2 (F) protein expression normalized to total protein (n = 4 mice/genotype, mean ± s.e.m). Grip strength of forelimbs (G) or all limbs (H) of WT and Osbpl6^−/− mice at 9 weeks based on the average of 3 independent trials normalized to body weight (mean ± s.e.m, n = 18–19 for WT and n = 8 for Osbpl6^−/−). I: Prepulse inhibition startle intensity at 110 dB of WT and Osbpl6^−/− mice at 10 weeks (mean ± s.e.m, n = 18–19 for WT and n = 8 for Osbpl6^−/−). J–L: Magnetic resonance imaging of WT as compared to Osbpl6^−/− mouse brains (n = 10 WT females, n = 10 Osbpl6^−/− females, n = 10 WT males, n = 10 Osbpl6^−/− males). Differences in total brain volume between genotypes are observed in (J), and a significant neuroanatomical effect of Osbpl6 deletion relative to WT is visualized using t-statistics on a structural (K) and voxel-wise (L) level. Regions larger or smaller in Osbpl6^−/− mutants relative to WT are given red-yellow and blue-turquoise colours, respectively, if effects are significant at an FDR of 5%. #P < 0.1, ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.0005 by Student’s t test (C, E, F) or by one-way ANOVA with Tukey’s multiple comparisons test (G–I).

Fig. 4

ORP6 is enriched in hippocampal astrocytes. A: Osbpl6 expression in mouse brain, heart, liver, spleen and white adipose tissue (WAT), expressed as fold change relative to liver, extracted from the mouse gene expression database (GXD) (mean ± s.e.m.). B: Osbpl6 gene expression (Log₂ fold-change) in specific brain regions from the Allen Brian Atlas. C, D: Mouse Osbpl6 and human OSBPL6 mRNA enrichment in astrocytes and neurons (brainrnaseq.com). E: qRT-PCR of Osbpl6 mRNA in murine CNS cell lines, normalized to BV2 cells (mean ± s.e.m.). F: Immunofluorescence of GFAP (red) and ORP6 (green) colocalization (yellow) in mouse brain. DAPI is shown in blue. G: Immunofluorescence of NeuN (purple), GFAP (red), DAPI (blue), and ORP6 (green) with ORP6 fluorescence and Pearson’s correlation for overlap with GFAP⁺ or NeuN⁺ areas (n = 6–8 mice/sex/genotype). Scale bar = 1 mm (F, G). ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.0005, ∗∗∗∗P < 0.00005 by one-way ANOVA with Tukey’s multiple comparisons test (A, C, G).

Fig. 5

ORP6 regulates cholesterol homeostasis in astrocytes. A: Immunoblot and qRT-PCR for ORP6 in C8-D1A astrocytic cells treated with control (ctrl) or Osbpl6 siRNA. B: qRT-PCR of lipid metabolism genes in astrocytes treated as in (A), showing fold-change relative to control (mean ± s.e.m.). C: immunofluorescence staining for BODIPY (green) and DAPI (blue) in C8-D1A cells treated with ctrl siRNA or ORP6 siRNA. D, E: qRT-PCR of Plin2 and Soat1 mRNA in ORP6 versus control siRNA-treated C8-D1A cells (n = 5). F: immunoblot of ORP6, PLIN2, SOAT1 in ORP6 versus control siRNA-treated C8-D1A cells. G: Cholesterol ester quantification by thin layer chromatography in cells with control or ORP6 siRNA loaded with ³H-cholesterol for 24 h and effluxed to BSA (n = 3). H, I: ³H-cholesterol efflux to HDL (H) or apoA-1 (I) in control or ORP6 siRNA-treated cells over 4h. A–I: Data are mean ± s.e.m., ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.0005 by Student’s t test (A, D, E, G–I) or by one-way ANOVA with Tukey’s multiple comparisons test (B). Scale bar = 25 μm.

Fig. 6

ORP6 knockdown enhances amyloid-beta production by lowering plasma membrane cholesterol and increasing APP processing. A: Relative desmosterol content in C8-D1A cells transfected with ctrl or ORP6 siRNA and loaded with ³H-mevalonate for 24 h (n = 3, mean ± s.e.m). B: Schematic overview of the Bloch Pathway for cholesterol biosynthesis in astrocytes, highlighting desmosterol as a precursor to cholesterol, catalyzed by DHCR24. C: qRT-PCR of Dhcr24 mRNA in ORP6 versus ctrl siRNA-treated C8-D1A cells (n = 3 with, mean ± s.e.m.). D: Immunoblotting of DHCR24 in C8-D1A cells transfected with ctrl or ORP6 siRNA. E: Quantification of de novo synthesized plasma membrane (PM) ³H-cholesterol content extracted from cells radiolabelled as in (A). Cholesterol was extracted from the PM by incubation of cells with 10 mM mβ-CD at 4˚C for 15 min (n = 3, mean ± s.e.m.). F: Immunoblotting of ORP6, full-length amyloid precursor protein (APP) and GAPDH in protein lysates of C8-D1A cells transfected with ctrl or ORP6 siRNA. G: Extracellular amyloid-beta oligomer (AβOs) concentrations (pg/mg protein) in the supernatants of cells treated as in (F) (n = 3, mean ± s.e.m.). H: Model: ORP6 knockdown decreases DHCR24 and increases ACAT1, leading to esterified cholesterol accumulation, reduced cholesterol efflux, lower plasma membrane cholesterol, and increased APP processing and neurotoxic AβO production. ∗P < 0.05, ∗∗P < 0.005, ∗∗∗P < 0.0005 by Student’s t test (A, C, E, G, H).