Decreased glucocerebrosidase activity and substrate accumulation of glycosphingolipids in a novel GBA1 D409V knock-in mouse model

Abstract

Multiple mutations have been described in the human GBA1 gene, which encodes the lysosomal enzyme beta-glucocerebrosidase (GCase) that degrades glucosylceramide and is pivotal in glycosphingolipid substrate metabolism. Depletion of GCase, typically by homozygous mutations in GBA1, is linked to the lysosomal storage disorder Gaucher's disease (GD) and distinct or heterozygous mutations in GBA1 are associated with increased Parkinson's disease (PD) risk. While numerous genes have been linked to heritable PD, GBA1 mutations in aggregate are the single greatest risk factor for development of idiopathic PD. The importance of GCase in PD necessitates preclinical models in which to study GCase-related mechanisms and novel therapeutic approaches, as well as to elucidate the molecular mechanisms leading to enhanced PD risk in GBA1 mutation carriers. The aim of this study was to develop and characterize a novel GBA1 mouse model and to facilitate wide accessibility of the model with phenotypic data. Herein we describe the results of molecular, biochemical, histological, and behavioral phenotyping analyses in a GBA1 D409V knock-in (KI) mouse. This mouse model exhibited significantly decreased GCase activity in liver and brain, with substantial increases in glycosphingolipid substrates in the liver. While no changes in the number of dopamine neurons in the substantia nigra were noted, subtle changes in striatal neurotransmitters were observed in GBA1 D409V KI mice. Alpha-synuclein pathology and inflammation were not observed in the nigrostriatal system of this model. In summary, the GBA1 D409V KI mouse model provides an ideal model for studies aimed at pharmacodynamic assessments of potential therapies aiming to restore GCase.

Fig 1. Schematic depiction of GBA1 targeting strategy.

The GBA1 D409V KI mutation was introduced to the murine Gba1 gene through the constitutive knockin (KI) of a Gba1 D427V point mutation as the D427V mutation corresponds to the D409V mutation in the mature GCase protein (Xu et al, 2003). The targeting vector was designed to introduce the D427V point mutation and an additional silent mutation to generate a Psil restriction site for analytical purposes. Positive selection markers were flanked by FRT (Neo^R) and F3 (Puro^R) sites respectively. Following homologous recombination in embryonic stem (ES) cells and selection by double positive resistance to neomycin and puromycin, resistant clones were validated for correct integration. Validated ES cells were injected into blastocysts to generate chimers. Chimeric male animals were crossed to FLP recombinase-expressing females to remove the selection cassettes. The resulting animals express the GBA1 D427V point mutation corresponding to the D409V point mutation in mature protein. An additional feature of this model is the insertion of LoxP sites flanking exons 6–8. After Cre recombination, the murine GBA1 gene will be constitutively knocked out as a result of loss of gene function through deletion of part of the GCase domain and introduction of a frameshift from exon 5 to all downstream exons with a premature stop codon in exon 9.

Fig 2. Introduction of the D409V point mutation to the GBA1 gene does not affect GBA1 mRNA transcription or GCase protein translation.

(A) GBA1 mRNA levels assessed by qPCR in frontal cortex tissue from 4 month old C57Bl/6 wild type (WT) mice, GBA1 D409V KI heterozygous (HET) mice, and GBA1 D409V KI homozygous (HOM) mice. GBA1 mRNA levels are normalized to GAPDH as the housekeeping gene, with graphs depicting fold changes relative to WT expression levels. GBA1 mRNA levels are consistent between groups with no significant differences observed (n = 8/group; p > 0.05). WT levels measure 1.016±0.068, HET levels measure 1.185±0.140, HOM levels measure 0.9066±0.037. (B-C) GCase protein levels assessed by Western blot in forebrain tissue (caudal to frontal cortex) from the same mice used for mRNA analysis. (B) GCase protein levels are normalized to β-actin as the housekeeping protein, with graphs depicting fold changes relative to WT GCase protein expression levels. GCase protein is unchanged in the HET and HOM GBA1 D409V KI mice as compared to WT (n = 7-8/group; p > 0.05). WT levels measure 1.000±0.052, HET levels measure 1.091±0.120, HOM levels measure 1.114±0.057. (C) Representative image of Western blot results for alternating samples of WT, HET, and HOM GBA1 D409V KI mice. Abbreviations: SEM, standard error of the mean; WT, wild type; GBA1, gene encoding human glucocerebrosidase; GCase, glucocerebrosidase; KI, knockin; HET, heterozygous; HOM, homozygous; mo, month.

Fig 3. GCase activity is significantly decreased in the brain and liver of the homozygous GBA1 D409V KI mouse model at 4, 8, and 12 months of age.

Two independent groups evaluated GCase activity in whole brain homogenate (A,C) and liver homogenate (B,D) using different methods—The CBE/4-MU method (A,B) and the MDW941 method (C,D). In whole brain homogenate (A,C) and liver homogenate (B,D), GCase activity is significantly decreased in the GBA1 D409V KI homozygous (HOM) mice as compared to C57Bl/6 wild type (WT) mice (n = 7/group). GCase activity measured by 4-MU in brain (A) measured as follows: WT 4mo 27.63±0.68, HOM 4mo 2.53±0.07, WT 8mo 27.65±0.69, HOM 8mo 3.01±0.20, WT 12 mo 27.49±0.46, HOM 12mo 3.15±0.26. GCase activity measured by 4-MU in liver (B) measured as follows: WT 4mo 93.69±7.16, HOM 4mo 3.90±0.21, WT 8mo 112.40±6.84, HOM 8mo 1.50±0.19, WT 12 mo 96.79±7.44, HOM 12mo 3.50±0.58. GCase activity measured by MDW941 in brain (C) measured as follows: WT 4mo 15.54±0.13, HOM 4mo 1.33±0.04, WT 8mo 15.99±0.23, HOM 8mo 1.36±0.02, WT 12 mo 18.61±0.24, HOM 12mo 2.37±0.02. GCase activity measured by MDW941 in liver (D) measured as follows: WT 4mo 59.48±3.14, HOM 4mo 2.15±0.10, WT 8mo 69.27±2.36, HOM 8mo 2.03±0.13, WT 12 mo 79.14±1.48, HOM 12mo 1.79±0.29. Significant differences are indicated as follows: ****p < 0.0001. Abbreviations: SEM, standard error of the mean; RFU, relative fluorescence units; GBA1, gene encoding human glucocerebrosidase; GCase, glucocerebrosidase; WT, wild type; HOM, homozygous; KI, knockin; mo, month.

Fig 4. GCase substrate levels are altered in the brain and liver of homozygous GBA1 D409V KI mice.

(A-B) Glucosylceramide (GlcCer) and (C-D) glucosylsphingosine (GlcSph) levels measured by LC-MS/MS in C57Bl/6 wild type (WT) and GBA1 D409V KI homozygous (HOM) mice at 4, 8, and 12 months of age (n = 7/group). (A) GlcCer levels in the brain are increased in GBA1 D409V KI mice at 12, but not 4 and 8, months of age. (B) GlcCer levels in the liver are increased in GBA1 D409V KI mice at all ages as compared to WT mice. (C) GlcSph levels in the brain are significantly increased in GBA1 D409V KI mice as compared to WT mice at all ages. (D) GlcSph levels in the liver are significantly increased in GBA1 D409V KI mice at all ages as compared to WT mice. Brain GlcCer (A) measured as follows: WT 4mo 11.20±0.72, HOM 4mo 13.20±1.06, WT 8mo 20.30±0.34, HOM 8mo 21.60±0.95, WT 12 mo 15.17±0.84, HOM 12mo 18.64±1.09. Liver GlcCer (B) measured as follows: WT 4mo 73.56±9.39, HOM 4mo 183.60±12.09, WT 8mo 84.40±5.22, HOM 8mo 228.10±25.32, WT 12 mo 70.30±5.72, HOM 12mo 255.40±35.80. Brain GlcSph(C) measured as follows: WT 4mo 78.98±1.31, HOM 4mo 1032.00±52.91, WT 8mo 66.64±1.38, HOM 8mo 710.30±29.01, WT 12 mo 57.24±2.73, HOM 12mo 770.70±44.97. Liver GlcSph (D) measured as follows: WT 4mo 23.41±2.07, HOM 4mo 3446.00±232.40, WT 8mo 24.47±0.72, HOM 8mo 2071.00±122.30, WT 12 mo 14.73±0.78, HOM 12mo 2291.00±206.50. Significant differences are indicated as follows: *p < 0.05, ***p < 0.001; ****p < 0.0001. Abbreviations: GlcCer, glucosylceramide; GlcSph, glucosylsphingosine; SEM, standard error of the mean; GBA1, gene encoding human glucocerebrosidase; GCase, glucocerebrosidase; WT, wild type; HOM, homozygous; KI, knockin; mo, month; ns, non-significant (p > 0.05).

Fig 5. Brain and liver GCase activity are significantly decreased in heterozygous GBA1 D409V KI mice, with varying differences in glucosphingolipid lipids in brain and liver.

(A-B) GCase, (C-D) gluocosylceramide (GlcCer), and (E-F) glucosylsphingosine (GlcSph) levels measured in whole brain homogenate and liver homogenate of C57Bl/6 wild type (WT) and GBA1 D409V KI heterozygous (HET) mice at 5 months of age (n = 10/age). (A-B) GCase levels as assessed by the CBE/4-MU method are significantly decreased in the het GBA1 D409V KI mice at 5 months of age in both brain (A) and liver (B). (C-D) GlcCer levels in the brain (C) and liver (D) are unchanged in het GBA1 D409V KI mice as compared to WT mice at 5 months of age. (E-F) GlcSph levels are significantly increased in het GBA1 D409V KI mice as compared to WT mice in brain (E) but not liver (F). GCase activity measured by 4-MU in brain (A) measured as follows: WT 30.82±1.11, HET 17.65±0.32. GCase activity measured by 4-MU in liver (B) measured as follows: WT 136.20±8.82, HET 61.20±2.69. Brain GlcCer (C) measured as follows: WT 16.39±1.18, HET 13.31±1.59. Liver GlcCer (D) measured as follows: WT 65.80±3.90, HET 71.88±7.37. Brain GlcSph (E) measured as follows: WT 37.96±0.68, HET 44.13±0.80. Liver GlcSph (F) measured as follows: WT 12.64±0.47, HET 13.65±0.59. Significant differences are indicated as follows: ****p < 0.0001. Abbreviations: GlcCer, glucosylceramide; GlcSph, glucosylsphingosine; GBA1, gene encoding human glucocerebrosidase; GCase, glucocerebrosidase; WT, wild type; HET, heterozygous; KI, knockin; SEM, standard error of the mean; mo, month; ns, non-significant (p > 0.05).

Fig 6. Dopamine turnover, but not total dopamine levels or nigral dopamine neuron numbers, are altered in aged homozygous GBA1 D409V KI mice as compared to controls.

(A) Stereological estimates of dopaminergic neurons in the substantia nigra pars compacta (SNpc) as denoted by tyrosine hydroxylase immunoreactivity (TH+) reveal no differences between the GBA1 D409V KI homozygous (HOM) mice and C57Bl/6 wild type (WT) mice at any age (n = 9/group). (B) Dopamine levels as assessed by UHPLC/MS/MS in striatal tissue are unaltered in HOM GBA1 D409V KI mice at all ages (n = 6/group). (C) Dopamine turnover is significantly increased in 12 month old HOM GBA1 D409V KI mice as compared to WT controls (n = 6/group). TH neurons (A) measured as follows: WT 4mo 7402±306.7, HOM 4mo 7263±181.9, WT 8mo 7102±503.5, HOM 8mo 7319±412.9, WT 12 mo 8169±429.1, HOM 12mo 7566±294.9. Dopamine levels (B) measured as follows: WT 4mo 9.13±0.62, HOM 4mo 10.05±0.58, WT 8mo 9.84±1.31, HOM 8mo 9.89±1.33, WT 12 mo 9.50±1.70, HOM 12mo 10.18±0.45. Dopamine turnover (C) measured as follows: WT 4mo 0.43±0.03, HOM 4mo 0.52±0.04, WT 8mo 0.38±0.05, HOM 8mo 0.46±0.05, WT 12 mo 0.29±0.03, HOM 12mo 0.55±0.06. Significant differences are indicated as follows: ***p < 0.001. Abbreviations: TH+, tyrosine hydroxylase immunoreactive; SNpc, substantia nigra pars compacta; DOPAC, 3,4-dihydroxyphenylacetic acid; HVA, homovanillic acid; DA, dopamine; GBA1, gene encoding human glucocerebrosidase; WT, wild type; HOM, homozygous; KI, knockin; SEM, standard error of the mean; mo, month; ns, non-significant (p > 0.05).