Glucocerebrosidase Deficiency in Drosophila Results in α-Synuclein-Independent Protein Aggregation and Neurodegeneration

Abstract

Mutations in the glucosidase, beta, acid (GBA1) gene cause Gaucher's disease, and are the most common genetic risk factor for Parkinson's disease (PD) and dementia with Lewy bodies (DLB) excluding variants of low penetrance. Because α-synuclein-containing neuronal aggregates are a defining feature of PD and DLB, it is widely believed that mutations in GBA1 act by enhancing α-synuclein toxicity. To explore this hypothesis, we deleted the Drosophila GBA1 homolog, dGBA1b, and compared the phenotypes of dGBA1b mutants in the presence and absence of α-synuclein expression. Homozygous dGBA1b mutants exhibit shortened lifespan, locomotor and memory deficits, neurodegeneration, and dramatically increased accumulation of ubiquitinated protein aggregates that are normally degraded through an autophagic mechanism. Ectopic expression of human α-synuclein in dGBA1b mutants resulted in a mild enhancement of dopaminergic neuron loss and increased α-synuclein aggregation relative to controls. However, α-synuclein expression did not substantially enhance other dGBA1b mutant phenotypes. Our findings indicate that dGBA1b plays an important role in the metabolism of protein aggregates, but that the deleterious consequences of mutations in dGBA1b are largely independent of α-synuclein. Future work with dGBA1b mutants should reveal the mechanism by which mutations in dGBA1b lead to accumulation of protein aggregates, and the potential influence of this protein aggregation on neuronal integrity.

Fig 1. A Drosophila dGBA1b deletion results in glucocerebrosidase deficiency.

(A) Comparison of protein sequence of human GBA1 and Drosophila dGBA1a and dGBA1b. Gray indicates similar residues, whereas black shading indicates identical residues. (B) Genomic organization of the Drosophila GBA1 homologs, dGBA1a and dGBA1b, and the intervening CG31413 gene. Orange and blue boxes represent coding and non-coding sequences, respectively. Black arrows indicate direction of transcription. Red arrows designate the breakpoints of the GBA1^ΔTT deletion allele. (C) There was no significant difference in the percentage of GBA1^ΔTT homozygotes and WT controls (GBA1^+/+) that survived the embryo to 1st instar larval transition, the 3^rd instar larval to pupal stage transition, or the pupal to adult stage transition. (D) Relative glucocerebrosidase (GCase) enzyme activity from isolated heads of 14-day-old controls and GBA1^ΔTT homozygotes. (E) Relative GCase enzyme activity from bodies excluding heads of 14-day-old controls and GBA1^ΔTT homozygotes. Error bars represent standard error of the mean (s.e.m.), ns indicates p>0.05, **p<0.005 by Student t test in all results shown in this figure.

Fig 2. GBA1^ΔTT homozygotes exhibit shortened lifespan and behavioral phenotypes consistent with neuronal dysfunction.

(A) Kaplan-Meier survival curves of WT controls (GBA1^+/+), GBA1^ΔTT heterozygotes (GBA1^ΔTT/+), GBA1^ΔTT homozygotes (GBA1^ΔTT/ΔTT), WT controls ectopically expressing Drosophila WT dGBA1b using the Actin GAL4 driver (Actin-GAL4>UAS-GBA1b;GBA^+/+), GBA1^ΔTT homozygotes ectopically expressing Drosophila WT dGBA1b using the Actin GAL4 driver (Act-GAL4>UAS-GBA1b;GBA1^ΔTT/ΔTT), GBA1^ΔTT homozygotes ectopically expressing human WT GBA1 using the Actin GAL4 driver (Act-GAL4>UAS-hGBA1^WT;GBA1^ΔTT/ΔTT), and GBA1^ΔTT homozygotes ectopically expressing human GBA1 harboring the p.N370S mutation using the Actin GAL4 driver (Act-GAL4>UAS-hGBA1^N370S;GBA1^ΔTT/ΔTT). (B) Climbing index of 5-day-old flies of indicated genotypes as described in A. (C) Recovery time from mechanical stress (bang sensitivity) of flies of given genotypes as described in A at given adult ages. (D) Recovery time from heat stress of flies of given phenotypes as described in A at given adult ages. Error bars represent s.e.m., *p<0.05, **p<0.005 by Student t test in all results shown in this figure.

Fig 3. GBA1^ΔTT homozygotes display a memory deficit and neurodegeneration, but do not have dopaminergic neuron loss.

(A) Latency time to initiate courtship in untrained 14-day-old males of indicated genotype. (B) Latency time to initiate courtship in 14-day-old males of indicated genotypes at 1 hour, 6 hours and 24 hours following training using a conditioned mating assay. Note the longer latency times in trained males (B) relative to untrained males (A). (C) Representative paraffin-embedded H&E-stained brain sections from 30-day-old controls and GBA1^ΔTT homozygotes. Yellow arrows indicate vacuoles. (D) Representative image of a projected Z-series of a control adult Drosophila brain stained with anti-Tyrosine Hydroxylase to label dopaminergic (DA) neurons. DA neurons within the PPL1 cluster are indicated by the circled regions. Scale bar, 200 μm. (E) Relative number of DA neurons within the PPL1 cluster of 30-day-old GBA1^ΔTT homozygotes (GBA1^ΔTT/ΔTT) N = 19, normalized to age-matched WT controls (GBA1^+/+) N = 21. There was no significant difference between the number of DA neurons within the PPL1 cluster per genotype by Student t test. Error bars represent s.e.m., ns indicates p>0.05, *p<0.05, **p<0.005 by Student t test in all results shown in this figure.

Fig 4. GBA1^ΔTT/ΔTT exhibit increased accumulation of soluble and insoluble ubiquitinated proteins.

(A) Western blot using antiserum to ubiquitin (Ub) of Triton-soluble (S) and Triton–insoluble (I) protein fractions from whole 30-day-old WT controls (GBA1^+/+) and GBA1^ΔTT homozygotes (GBA1^ΔTT/ΔTT), using anti-β-Actin (βAct) as a loading control. The intense ubiquitin-positive band immediately below the 72kD marker and the upper βAct band likely represents arthirin, a 55kD mono-ubiquitinated form of β-Actin that is present exclusively in the indirect flight muscle [68]. (B) Western blot probed for Ub of Triton–soluble (S) and Triton–insoluble (I) protein fractions from heads of flies of 30-day-old GBA1^+/+, GBA1^ΔTT/ΔTT, WT controls ectopically expressing Drosophila WT dGBA1b using the Actin GAL4 driver (Actin-GAL4>UAS-GBA1b;GBA^+/+), and GBA1^ΔTT homozygotes ectopically expressing Drosophila WT dGBA1b using the Actin GAL4 driver (Act-GAL4>UAS-GBAb;GBA1^ΔTT/ΔTT), using βAct as a loading control. (C) Densitometric quantification of Ub signal in the Triton-insoluble fraction from the heads of flies of the indicated genotypes. Levels of Ub signal per genotype were normalized to respective βAct loading controls, and these ratios were in turn normalized to the insoluble Ub level of GBA1^+/+. (D) Representative immunofluorescent staining of thoracic muscle from 30-day-old GBA1^+/+ and GBA1^ΔTT/ΔTT flies with anti-Ub and anti-F-Actin. Scale bars, 20 μm. (E) Quantification of Ub-positive objects within 10-μm-thick Z-stacks of thoracic muscle of 30-day-old GBA1^+/+ and GBA1^ΔTT/ΔTT flies. At least 9 separate Z-stacks were analyzed per genotype. Error bars represent standard deviation. (F) Representative anti-Ub immunofluorescent staining of 30-day-old whole brains from GBA1^+/+ and GBA1^ΔTT/ΔTT flies. Scale bar, 200 μm. Arrowhead indicates punctate staining pattern, arrow indicates filamentous staining pattern. (G) Western blot using antiserum to Ref(2)P of whole 10-day-old flies of the indicated genotypes, including Atg7 null flies (Atg7^d77/d77). βAct was used as a loading control. (H) Densitometric quantification of Ref(2)P signal from 10-day-old whole flies of the indicated genotypes. Levels of Ref(2)P signal per genotype were normalized to respective βAct loading controls, and these ratios were in turn normalized to the Ref(2)p level of GBA1^+/+. (I) Cathepsin D activity in whole 7-day-old GBA1^ΔTT homozygotes relative to age-matched controls. (J) Hexosaminidase activity in bodies excluding heads, and isolated heads of 7-day-old GBA1^ΔTT homozygotes relative to age-matched controls. Error bars represent s.e.m. unless indicated, ns indicates p>0.05, *p<0.05, **p<0.005 by Student t test in all results shown in this figure.

Fig 5. Glucocerebrosidase deficiency mildly enhances α-synuclein toxicity in neurons and increases the abundance of α-synuclein aggregates.

(A) Relative number of DA neurons within the PPL1 cluster of 30-day-old flies, normalized to WT controls. WT controls (GBA1^+/+) N = 12, GBA1^ΔTT homozygotes (GBA1^ΔTT/ΔTT) N = 22, WT controls expressing α-synuclein in DA neurons (TH-G4>UAS-syn^WT;GBA1^+/+) N = 16, and GBA1^ΔTT homozygotes expressing α-synuclein in DA neurons (TH-G4>UAS-syn^WT;GBA1^ΔTT/ΔTT) N = 28. (B) Western blot analysis using antiserum to α-synuclein (α-syn) of Triton-insoluble protein fractions from heads of 10-day-old WT controls (GBA1^+/+), GBA1^ΔTT homozygotes (GBA1^ΔTT/ΔTT), WT controls ubiquitously expressing α-synuclein with the p.A53T mutation (Act-G4>UAS-syn^A53T;GBA1^+/+), and GBA1^ΔTT homozygotes ubiquitously expressing α-synuclein with the p.A53T mutation (Act-G4>UAS-syn^A53T;GBA1^ΔTT/ΔTT). βAct was used as a loading control. (C) Densitometric quantification of α-syn signal in the Triton-insoluble fraction from 10-day-old heads of flies of indicated genotypes as described in B. The level of HMW α-syn signal in the Triton-insoluble fraction in GBA1^ΔTT/ΔTT was normalized to its βAct loading control, and this ratio was in turn normalized to the insoluble α-syn/βAct ratio of GBA1^+/+, N = 3. (D) Western blot analysis using an antiserum to Ubiquitin (anti-Ub) of Triton–soluble (S) and–insoluble (I) protein extracts from heads of controls ectopically expressing WT α-synuclein using the TH-GAL4 driver (TH-G4>UAS-syn^WT;GBA1^+/+) and GBA1^ΔTT homozygotes ectopically expressing WT α-synuclein using the TH-GAL4 driver (TH-G4>UAS-syn^WT;GBA1^ΔTT/ΔTT). βAct was used as a loading control. (E) Densitometric quantification of anti-Ub levels in Triton-insoluble protein fractions from heads of flies of indicated genotypes as described in D. Levels of Ub signal per genotype were normalized to respective βAct loading controls, and the ratio of Ub/βAct in Th-G4>UAS-syn^WT;GBA1^ΔTT/ΔTT was normalized to the ratio of α-syn/βAct in TH-G4>UAS-syn^WT;GBA1^+/+. Error bars represent s.e.m. unless indicated, *p<0.05, **p<0.005 by Student t test in all results shown in this figure.

Fig 6. α-synuclein expression does not enhance dGBA1b deficient fly phenotypes.

(A) Kaplan-Meier survival curves of lifespans of WT controls, GBA1^ΔTT homozygotes, WT controls expressing α-synuclein in dopaminergic neurons (TH-G4>UAS-syn^WT;GBA1^+/+), and GBA1^ΔTT homozygotes expressing α-synucleinin dopaminergic neurons (TH-G4>UAS-syn^WT;GBA1^ΔTT/ΔTT). (B) Climbing index of 5-day-old flies of given genotypes. (C) Recovery time from temporary paralysis due to mechanical stress (bang sensitivity) of flies of given phenotypes at given adult age. (D) Recovery time from heat stress of flies of given genotypes at given adult ages. Error bars represent s.e.m., *p<0.05, **p<0.005 by Student t test in all results shown in this figure.