Cultured Macrophage Models for the Investigation of Lysosomal Glucocerebrosidase and Gaucher Disease

Abstract

Macrophages are specialised cells that degrade a range of substrates during their lifetime. In inherited lysosomal storage disorders, particularly the sphingolipidoses, macrophages transform into storage cells and contribute to pathology. An appropriate cultured macrophage model is desired for fundamental research and the assessment of considered therapeutic interventions. We compared commonly used macrophage cell lines, RAW264.7, J774A.1, and THP-1 cells, with human monocyte-derived macrophages (HMDMs) isolated from peripheral blood. Specific lysosomal glucosidases were analysed by enzymatic activity measurements and visualised with fluorescent activity-based probes. Special attention was given to lysosomal glucocerebrosidase (GBA1), the enzyme deficient in Gaucher disease in which lipid-laden macrophages are a hallmark. In macrophage cell lines and HMDMs, various (glyco)sphingolipids relevant to GBA1 activity were determined. Finally, the feasibility of inactivation of GBA1 with a cell-permeable suicide inhibitor was established, as well as the monitoring of uptake of therapeutic recombinant human GBA1. Major differences among various cell lines were noted in terms of morphology, lysosomal enzyme expression, and glycosphingolipid content. HMDMs appear to be the most suitable model for investigations into GBA1 and Gaucher disease. Moreover, they serve as a valuable model for mannose-receptor mediated uptake of therapeutic human GBA1, effectively mimicking enzyme replacement therapy for Gaucher disease.

Keywords: Gaucher disease; glucocerebrosidase; lysosome; macrophage; monocyte.

Figure 1

Enzyme activities of a selection of lysosomal enzymes in macrophage models. The activity in lysates of human monocyte-derived macrophages (HMDM), THP-1 macrophages, J774A.1 and RAW264.7 cells was determined by measuring the release of fluorogenic 4-MU from substrates specific for lysosomal/non-lysosomal β-glucosidase (GBA1/GBA2), α-glucosidase (GAA), β-galactosidase (GALC), α-galactosidase (GLA), β-hexosaminidases (HEXA&B) and β-glucuronidase (GUSB). Values are expressed as mean ± SD, at least n = 3 with technical triplicates. ns = non significant, * p < 0.033, ** p < 0.002 and *** p < 0.001.

Figure 2

In gel fluorescent detection of a selection of exo-glycosidases labelled with an activity-based probe (ABP) in lysates of human monocyte-derived macrophages (HMDM), THP-1 macrophages, J774A.1 and RAW264.7 cells. (A) Labelling with specific ABP 5 for lysosomal β-glucosidase (GBA1), ABP 6 non-lysosomal β-glucosidase (GBA2) and ABP 8 for α-galactosidase A (GLA) and α-N-acetylgalactosaminidase (NAGA) (B) Labelling with specific ABP 9 for α-glucosidase (GAA), ABP 7 for β-galactosidases (GALC and GLB1) and ABP 10 for β-glucuronidase (GUSB). In the labelling of GAA, GALC/GLB1 and GUSB, before ABP treatment, lysates were pre-incubated with GBA1 inhibitor 1. Each ABP was incubated for 30–60 min at specific pH for which labelling of the individual glycosidase is optimal at 37 °C. Gels are representative examples of n = 3 experiments.

Figure 3

Sphingolipids and lyso-Sphingolipids levels (pmol/mg protein) in macrophage models. Human monocyte-derived macrophages (HMDM), THP-1 macrophages, J774A.1 and RAW264.7 cells (n = 3) were harvested, and lipids were extracted and quantified by LC-MS/MS. (A) The levels of (glyco)-sphingolipids (B) The levels of (glyco)lyso-sphingolipids. Values are expressed as mean ± SD with n = 3 and technical duplicates. Statistics were performed if samples were above the limit of detection with significance: ns = non significant, * p < 0.033, ** p < 0.002 and *** p < 0.001.

Figure 4

Confocal microscopy of glucocerebrosidase (GBA1) by activity-based probe (ABP) labelling and LAMP1 antibody staining. (A) Human monocyte-derived macrophages and (B) RAW264.7 cells. Both samples were incubated in situ with 5 nM ABP 3 (Red). Afterwards, samples were fixed, stained with α-LAMP1-antibody (green) and nuclei was stained with 0.5 µM SYTOX Green for (A) or 10 µg/mL DAPI for (B) (blue). Scale bar: 20 µm.

Figure 5

Effect of glucocerebrosidase inhibition by small molecule inhibitor in human monocyte-derived macrophages (HMDMs). HMDMs were cultured in the presence of 100 nM compound 1, a cell permeable inhibitor of glucocerebrosidase (GBA1), for 2 or 7 days. Fresh medium with inhibitor was added every 2 days and cells were harvest at 21 days post isolation and lysate was used for the analysis of (A) unbound GBA1 detection by activity-based probe ABP 5 (B) residual relative GBA1 enzyme activity (C) chitotriosidase (Chito) enzyme activity in culture medium and cell lysate and (D) glycosphingolipids (pmol/mg protein) quantification of hexosylceramide (HexCer), the deacylated for hexosylsphingosine (HexSph) and hexosylcholesterol (HexChol). GBA1 irreversible inhibition with compound 1 for 7 days (E) did not show any morphological changes to HMDMs. SDS-Page gel is a representative example of n = 3 experiments. Values are expressed as mean ± SD using n = 3 for (B,D), and n = 1 for (C) all with technical triplicates and ns = non significant, * p < 0.033, ** p < 0.002 and *** p < 0.001. Scale bar: 25 µm for (E).

Figure 6

Uptake of therapeutic recombinant human glucocerebrosidase by human monocyte-derived macrophages. Endogenous glucocerebrosidase (GBA1) in human monocyte-derived macrophages (HMDMs) was blocked with ABP 3 and cells were incubated with recombinant human glucocerebrosidase (rhGBA1) pre-labelled with ABP 4 for 6 h. Lysate of HMDMs analysed on SDS-PAGE gel (A) showing endogenous GBA1 (red) and internalised rhGBA1 (green) (B) western blot of the same gel stained for mannose-receptor (MR) and β-Actin loading control. (C) Relative intensities of internalised rhGBA1 and MR normalised towards β-actin of gel and blot seen in (A,B). (D) Confocal microscopy visualization of endogenous GBA1 (red) and internalised rhGBA1 (green) in HMDMs; after fixation, the nuclei were stained with 10 µg/mL DAPI (blue). Scale bar: 20 µm. Gels and Western Blots are representative example of n = 3 experiments.