A new glucocerebrosidase-deficient neuronal cell model provides a tool to probe pathophysiology and therapeutics for Gaucher disease

Abstract

Glucocerebrosidase is a lysosomal hydrolase involved in the breakdown of glucosylceramide. Gaucher disease, a recessive lysosomal storage disorder, is caused by mutations in the gene GBA1 Dysfunctional glucocerebrosidase leads to accumulation of glucosylceramide and glycosylsphingosine in various cell types and organs. Mutations in GBA1 are also a common genetic risk factor for Parkinson disease and related synucleinopathies. In recent years, research on the pathophysiology of Gaucher disease, the molecular link between Gaucher and Parkinson disease, and novel therapeutics, have accelerated the need for relevant cell models with GBA1 mutations. Although induced pluripotent stem cells, primary rodent neurons, and transfected neuroblastoma cell lines have been used to study the effect of glucocerebrosidase deficiency on neuronal function, these models have limitations because of challenges in culturing and propagating the cells, low yield, and the introduction of exogenous mutant GBA1 To address some of these difficulties, we established a high yield, easy-to-culture mouse neuronal cell model with nearly complete glucocerebrosidase deficiency representative of Gaucher disease. We successfully immortalized cortical neurons from embryonic null allele gba(-/-) mice and the control littermate (gba(+/+)) by infecting differentiated primary cortical neurons in culture with an EF1α-SV40T lentivirus. Immortalized gba(-/-) neurons lack glucocerebrosidase protein and enzyme activity, and exhibit a dramatic increase in glucosylceramide and glucosylsphingosine accumulation, enlarged lysosomes, and an impaired ATP-dependent calcium-influx response; these phenotypical characteristics were absent in gba(+/+) neurons. This null allele gba(-/-) mouse neuronal model provides a much-needed tool to study the pathophysiology of Gaucher disease and to evaluate new therapies.

Keywords: Gaucher disease; Glucocerebrosidase; Glucosylceramide; Glucosylsphingosine; Neuron.

Fig. 1.

Expression analysis and in vivo tumor formation of SV40-T immortalized mouse cortical cells. (A) EGFP expression in primary C57BL/6 mouse cortical cell cultures five days after infection with EF1α-eGFP lentivirus at MOI 40. (B) Western blot analysis of protein lysates of gba^+/+ (lane 1) and gba^−/− (lane 2) immortalized cortical cells with antibodies against SV40-T and β-actin (protein loading control). SV40-T protein expression was detected in both gba^+/+ and gba^−/− immortalized cells. (C) In vivo intraperitoneal tumor formation after injection of 10⁶ cells of gba^+/+ and gba^−/− immortalized cells in five Swiss nu/nu mice per genotype. Four weeks after injection, mice injected with gba^+/+ cells developed visible solid tumors (left panel) whereas mice injected with gba^−/− cells developed solid bloody tumors (right panel). Immortalized gba^+/+ and gba^−/− mouse cortical cells were stained for DAPI (D,E), anti-SV40-T (F,G), and the image merged with bright-field for visualization of neuron morphology (H,I). DAPI and SV40-T co-localize in the nucleus (H,I). Scale bar: 20 μm.

Fig. 2.

Establishment of CD24-positive gba^+/+ and gba^−/− neuron cultures. Expression of (A,E) MAP-2 and (B,F) GFAP in SV40-T immortalized gba^+/+ and gba^−/− cells before FACS. Both MAP-2 and GFAP are expressed in gba^+/+ cultures. GFAP is only sporadically expressed in gba^−/− cultures. (C,G) FACS of SV40-T immortalized gba^+/+ and gba^−/− cells with antibodies against CD24 and CD29 surface markers. (D,H) Expression of MAP-2 and GFAP in SV40-T immortalized gba^+/+ and gba^−/− neuronal cultures after FACS. No GFAP-positive cells are detected. Scale bar: 20 μm.

Fig. 3.

Karyotyping, GCase enzyme activity, protein expression and substrate levels in immortalized gba^+/+ and gba^−/− neurons. (A,B) Karyotyping of immortalized CD24-positive gba neurons. Two examples each of heterogeneous karyotypes observed for immortalized gba^+/+ and gba^−/− mouse neurons. (C) GCase enzyme activity in SV40-T immortalized gba^+/+ and gba^−/− mouse neurons. Relative percentage GCase enzyme activity of gba^−/− (3.40±0.36%) immortalized neurons is significantly lower compared with gba^+/+ immortalized neurons (93.44±12.26%) (P=0.0017). Each enzyme assay was performed three independent times (n=3) in triplicates. The data are represented as relative mean values±s.e.m. (D) 15 µg of neuronal protein lysate was incubated with 100 nM of the GCase-specific MDW933 fluorescent inhibody. GCase protein expression is absent in gba^−/− immortalized neurons (lane 2). Recombinant imiglucerase was loaded as a positive control (lane 3). (E) Gba^−/− immortalized neurons show significantly increased GlcCer storage compared with gba^+/+ immortalized neurons (1500±118.9 vs 165.1±6.14 GlcCer/mg protein; P=0.0005). (F) Gba^−/− immortalized neurons show significantly increased GlcSph storage compared with gba^+/+ immortalized neurons (485.7±21.24 vs 0.071±0.02 GlcSph/mg protein; P<0.0001). Each substrate storage assay was done four independent times (n=4). The data are represented as mean values±s.e.m.

Fig. 4.

Quantitative analysis of lysosomes and ATP-dependent Ca²⁺ influx in immortalized gba ^+/+ and gba^−/− neurons. (A) FACS analysis of LysoTracker stained gba^−/− neurons and gba^+/+ (fold-change 17.7±0.83 vs 7.13±0.23). Data from three independent experiments (n=3) showed significantly increased LysoTracker staining in gba^−/− neurons (P=0.0003). (B) Total LysoTracker^® pixels per cell was significantly higher in gba^−/− cells compared with gba^+/+ cells (80.22±1.79 vs 22.12±1.85; P=0.0005). (C) The number of ROIs acquired per cell was significantly higher in gba^−/− cells compared with gba^+/+ cells (6.79±0.11 vs 2.36±0.13; P=0.0005). (D) After treatment of cells with different concentrations of ATP (3 µM, 1 µM, 0.3 µM, 0.1 µM), Ca²⁺ influx in gba^+/+ (black dots) was significantly higher (P=0.0036) compared with gba^−/− (white dots). Experimental data was normalized to numbers of cells with the CellTag 700 assay. Each experiment included quadruplicate samples and was repeated two times (n=2). The data are represented as mean values±s.e.m.