Sexually dimorphic mechanisms of VGLUT-mediated protection from dopaminergic neurodegeneration

Abstract

Parkinson's disease (PD) targets some dopamine (DA) neurons more than others. Sex differences offer insights, with females more protected from DA neurodegeneration. The mammalian vesicular glutamate transporter VGLUT2 and Drosophila ortholog dVGLUT have been implicated as modulators of DA neuron resilience. However, the mechanisms by which VGLUT2/dVGLUT protects DA neurons remain unknown. We discovered DA neuron dVGLUT knockdown increased mitochondrial reactive oxygen species in a sexually dimorphic manner in response to depolarization or paraquat-induced stress, males being especially affected. DA neuron dVGLUT also reduced ATP biosynthetic burden during depolarization. RNA sequencing of VGLUT⁺ DA neurons in mice and flies identified candidate genes that we functionally screened to further dissect VGLUT-mediated DA neuron resilience across PD models. We discovered transcription factors modulating dVGLUT-dependent DA neuroprotection and identified dj-1β as a regulator of sex-specific DA neuron dVGLUT expression. Overall, VGLUT protects DA neurons from PD-associated degeneration by maintaining mitochondrial health.

Keywords: Drosophila; Parkinson’s disease; RNAseq; VGLUT2; dVGLUT; dopamine; glutamate; neurodegeneration; paraquat.

Figure 1.. dVGLUT-expressing DA neurons localize to distinct cell clusters.

(A) Illustrations of DA neuron clusters in the anterior and posterior regions of the adult Drosophila brain including: subesophageal ganglion (SOG), tritocerebrum (T1), protocerebral posterior medial 1 (PPM1), protocerebral posterior lateral 1 (PPL1), and protocerebral posterior lateral 2 (PPL2) clusters. (B) Representative multiplex RNAscope 4x images (left) and magnified 20x views of insets (3 right panels) of wild-type w¹¹¹⁸ fly brains (14d post-eclosion), displaying dVGLUT (green) and TH (magenta) mRNA co-expression in several DA neuron clusters. TH⁺/dVGLUT⁺ cells were found in PPL2 subclusters comprised of both PPL2ab and PPL2c. 4× scale bar=100μm, 20× scale bar=25μm.

Figure 2.. Dynamic regulation of dVGLUT expression mediates sex-specific DA neuron resilience.

(A) Schematic of the intersectional luciferase reporter to estimate dVGLUT expression in DA neurons (dVGLUT>GAL4/LexAop>B3R;TH>LexA/UAS>B3RT-STOP-B3RT-Luciferase). Panel i: The TH promoter drives LexA to express B3 recombinase (B3R) in TH⁺ DA neurons. Panel ii: B3R excises a transcriptional stop cassette within UAS>Luciferase. Panel iii: This excision permits successful dVGLUT>GAL4-driven transactivation of UAS>Luciferase specifically in TH⁺/dVGLUT⁺ cells. (B) Among adult flies (14d post-eclosion), control (Con) females exhibited higher baseline DA neuron dVGLUT expression compared to males (p=0.048, Bonferroni multiple comparisons test). 5d exposure to 10mM paraquat significantly raised DA neuron dVGLUT expression compared to 3d exposure (p=0.047, Bonferroni multiple comparisons test). Luminescent DA neuron dVGLUT reporter data were normalized to % untreated male and female controls. N=4–8 homogenates of 5 brains per group. (C) Among older flies (60d post-eclosion), paraquat exposure (10mM, 12h) significantly raised DA neuron dVGLUT expression compared to the vehicle-treated Con group (p=0.039, Bonferroni multiple comparisons test). N=4–8 homogenates of 5 brains per group. Related analyses are shown in Figure S1. (D) Representative images showing dVGLUT (green) and TH (magenta) mRNA expression in TH-driven Control (TH>GAL4/UAS>LexA-RNAi) and dVGLUT RNAi (TH>GAL4/UAS>dVGLUT-RNAi) flies at 14d post-eclosion. Scale bar=10μm. Related analyses are shown in Table S1. (E) At day 2 of 10mM paraquat exposure, a genotype × paraquat interaction was observed in male flies (4d post-eclosion) (left panel; F_1,202=10.3, p=0.0016, two-way ANOVA). Male TH-driven dVGLUT RNAi flies exhibited decreased locomotion in response to paraquat (p<0.001), but male w¹¹¹⁸ wild-type control flies did not (p>0.05, Bonferroni multiple comparisons test). Female w¹¹¹⁸ wild-type and dVGLUT RNAi flies were not significantly affected by paraquat (right panel; p>0.05, Bonferroni multiple comparisons test). N=50–67 flies per group. (F) Representative projection images of GFP-labeled DA neurons in TH-driven Control (TH>GAL4,UAS>GFP/UAS>LexA-RNAi) and dVGLUT RNAi (TH>GAL4,UAS>GFP/UAS>dVGLUT-RNAi) brains (14d post-eclosion). Images are from protocerebral anterior lateral (PAL, purple), protocerebral anterior medial (PAM, green) and subesophageal ganglion (SOG, yellow) clusters. Scale bar=100μm. (G) There was no effect of paraquat exposure (10mM, 5d) on DA neuron number (p>0.05, three-way ANOVA). There was a significant effect of dVGLUT RNAi on SOG DA neuron number (F_1,46=4.95, p=0.031, three-way ANOVA). N=4–8 per group. *p<0.05, ***p<0.001; all data are plotted as Mean±SEM.

Figure 3.. DA neuron dVGLUT modulates sex differences in mitochondrial oxidative stress.

(A) Representative projection images of MitoTimer-labeled DAergic projections to the fan-shaped body (FSB) of male and female dVGLUT RNAi flies (TH>GAL4,UAS>MitoTimer/UAS>dVGLUT-RNAi) versus Control (TH>GAL4,UAS>MitoTimer/UAS>LexA RNAi) flies (14d post-eclosion). Green and red fluorescent channels were overlaid to generate final images. All groups were exposed to either paraquat (10mM, 5d) or vehicle control (0mM, 5d). Scale bar=25μm. (B) Paraquat exposure (10mM, 5d) significantly increased the MitoTimer red:green ratio in DAergic projections to the FSB of male dVGLUT RNAi flies compared to vehicle (paraquat × sex × RNAi interaction: F_1,59=6.05, p=0.017, three-way ANOVA; male RNAi vehicle versus 10mM: p=0.0049, Bonferroni multiple comparisons test). N=7–10 per group. (C) Representative projection images of MitoTimer-labeled DAergic projections to the FSB of adult male dVGLUT RNAi flies. Flies were exposed to 5d of either 10 mM paraquat, 10 mM paraquat + antioxidant 3mM TEMPOL, or vehicle control. Scale bar=25μm. (D) Co-treatment with 3mM TEMPOL blocked paraquat-induced increases in the MitoTimer red:green ratio in male RNAi flies (p=0.038, unpaired t-test). N=7–8 per group. *p<0.05, **p<0.01; all data are plotted as Mean±SEM. See also Figures S2, S3 and S4.

Figure 4.. dVGLUT modulates activity-driven intracellular ATP and mitochondrial ROS generation in DA neurons.

(A) Representative images of TH-driven iATPSnFr fluorescence (F) in DAergic projections to the SOG of adult (14d post-eclosion) LexA RNAi (Control, full genotype: UAS>iATPSnFr/w¹¹¹⁸;TH>GAL4/UAS>LexA-RNAi) and dVGLUT RNAi (RNAi, full genotype: UAS>iATPSnFr/w¹¹¹⁸;TH>GAL4/UAS>dVGLUT-RNAi) flies before and after 40mM KCl stimulation. White circles highlight regions that feature differences in intracellular DA neuron ATP levels during KCl-induced depolarization. Scale bar=25μm. (B) Quantification showed increased peak iATPSnFr fluorescence (Peak ΔF/F_i) in SOG DAergic projections of RNAi flies (main effect of RNAi: F_1,20=5.93, p=0.024, two-way ANOVA). N=4–7 per group, *p<0.05. Additional analyses are shown in Figure S5. (C) Regional differences in DA neuron baseline MitoTimer red:green ratio and iATPSnFr peak ΔF/F after KCl-induced depolarization were strongly correlated (r²=0.998, p=0.034). N=4 genotype/sex groups compared for correlation analysis. (D) Representative images of TH-driven MitoTimer labeling in DAergic projections to the SOG, EB, and FSB of Control and RNAi flies before and after 40mM KCl treatment. Scale bars=25μm. (E) DA neuron dVGLUT RNAi knockdown increased mitochondrial ROS acutely in response to 40mM KCl stimulation compared to Control flies (KCl × RNAi interaction, SOG: F_5,45=6.38, p<0.001; EB: F_5,43=6.63, p<0.001; FSB: F_5,43=6.08, p<0.001, two-way ANOVA). **p<0.01, ***p<0.001 compared to baseline (pre-KCl), ^#p<0.05, ^##p<0.01, ^###p<0.001 compared to Control flies (Bonferroni multiple comparisons test). N=8–9 per group. (F) Schematic showing that in DA neurons, dVGLUT decreases ATP levels during depolarization and decreases mitochondrial ROS both during depolarization and in response to paraquat. dVGLUT’s modulation of intracellular ATP and mitochondrial ROS accumulation may be due to its impacts on: 1) cytoplasmic glutamate availability, enabling efficient glutathione production to prevent toxic ROS accumulation, and 2) reducing levels of cytoplasmic DA available for oxidation that contribute to ROS generation. All data are plotted as Mean±SEM.

Figure 5.. RNAseq reveals differentially expressed genes in VGLUT-expressing DA neurons conserved between flies and mice.

(A) Annotations of adult Drosophila brain 10X single-cell RNAseq (scRNAseq) data via Uniform Manifold Approximation and Projection (UMAP) plots identified discrete clusters of dVMAT⁺ DA neurons. Expression of TH, DAT and dVGLUT across all 2,642 dVMAT-expressing cells is represented, with highest expression shown in purple and lowest/no expression shown in gray. Specific subpopulations of dVMAT⁺/TH⁺/DAT⁺/dVGLUT⁺ cells are highlighted by red boxes. (B) Volcano plot showing differential expression of genes in dVGLUT⁺ DA neurons compared to dVGLUT⁻ DA neurons. (C) Gene ontology (GO) analysis of Drosophila scRNAseq data identifying top pathways based on differential gene expression in dVGLUT⁺ DA neurons. (D) Volcano plot showing differential expression of genes in mouse VGLUT2⁺ DA neurons compared to all DAT⁺ DA neurons. Additional analyses relating to VGLUT2⁺ DA neurons are shown in Figures S6 and S7 and Videos S1 and S2. (E) Schematic summarizing the mitochondrial findings from the comparative fly and mouse RNAseq studies. These data revealed overall downregulation of genes implicated in mitochondrial processes in both Drosophila dVGLUT⁺ and mouse VGLUT2⁺ DA neurons. This included downregulation of mitochondrial ribosomal genes in mouse and overall downregulation of TCA cycle genes in flies. Among the mitochondrial respiratory complexes in VGLUT2⁺ DA, Complex I gene expression was downregulated in both mice and flies, while Complex IV and ATP synthase gene expression were specifically downregulated in mice.

Figure 6.. Forward genetic screens identify modulators of VGLUT-associated DA neuron resilience.

(A) Flow chart depicting the process by which candidate genes were selected for forward genetic screening. (B) Primary paraquat screen: 187 candidate genes differentially expressed in TH⁺/VGLUT⁺ DA neurons were knocked down via RNAi. Adult (2d post-eclosion) male and female TH-driven RNAi flies were exposed to either paraquat (10mM, 4d) or vehicle control. Paraquat-induced alterations in locomotion over days 2–4 of exposure are plotted as % locomotion relative to each candidate’s vehicle-treated control. For reference, TH-driven dVGLUT RNAi flies (orange dashed lines) were tested alongside candidates. Average paraquat locomotion of 5 Bloomington Drosophila Stock Center (BDSC)/Harvard Transgenic RNAi Project (TRiP) control lines tested are represented by green dashed lines. Two screen hits that are orthologs of human genes with known roles in PD are highlighted (i.e., dj-1β, Tom20). N=3–37 flies per group; mean N=18 flies per group. (C) GO pathway analysis of paraquat screen candidate hits. (D) Secondary αSyn^A53T screen: The top 32 candidate genes from the primary screen were knocked down via TH-driven RNAi alongside TH-driven αSyn^A53T expression. αSyn^A53T-induced changes in locomotion (at 4–6d post-eclosion) are plotted as % UAS Control. N=12–39 flies per group; mean N=21 flies per group. (E) GO pathway analysis of candidate hits from the αSyn^A53T secondary screen. Each candidate’s results in the screens along with control line results are shown in Table S5. Further analyses of candidate genes are included in Figure S8. All data are plotted as Mean±SEM.

Figure 7.. Top candidate transcription factors impact DA neuron numbers in a region- and sex-specific manner.

(A) UMAP plots showing expression of selected top candidates from primary and secondary genetic screens in dVMAT-expressing cells. Highest expression levels are shown in purple and lowest/no expression is shown in gray. dVMAT⁺/TH⁺/DAT⁺/dVGLUT⁺ cells are highlighted by red boxes. (B) Representative images of GFP-labeled DA neurons demonstrating the impact of TH-driven RNAi knockdown of top gene candidates and LexA RNAi control (Con) on DA neuron number (14d post-eclosion) in the PAL, PPL2, PPM1/2 and PPM3 cell clusters. PAL, PPL2 scale bars=100μm, PPM1/2, PPM3 scale bars=250μm. (C) TH-driven RNAi of candidate genes mirr, trh, oc, and Dll differentially altered GFP-labeled DA neuron numbers in region-specific and sex-specific manners compared to controls. N=5–21 fly brains per group. *p<0.05, **p<0.01, ***p<0.001, two-way ANOVA with Bonferroni multiple comparisons test. All data are plotted as Mean±SEM.

Figure 8.. dj-1β modulates sex differences in DA neuron dVGLUT expression.

(A) Representative images of DA neuron dVGLUT mRNA expression (in yellow); TH⁺ DA neurons are GFP-labeled (in green) in adult LexA RNAi control (Con) flies. DA neuron dVGLUT expression is shown in the PPL1 and SOG (both T1 and lateral SOG dVGLUT-expressing DA neurons), in contrast to the PAL and PPM3 DA neuron clusters. Scale bars=100μm. (B) Quantification of DA neuron dVGLUT expression showed higher percentages of dVGLUT⁺ DA neurons in the PPL1 and SOG compared to other DA neuron clusters (F_5,168=3.59, p=0.0041, one-way ANOVA). N=25–31 fly brains per group. (C) Representative images of DA neuron dVGLUT mRNA expression in response to TH-driven dj-1β RNAi (dj-1β); images focus on the PPL2ab cluster of GFP-labeled DA neurons from dj-1β RNAi flies exposed to 10mM paraquat (10d) or vehicle control. (D) Quantification in the PPL2 DA neuron cluster showed a paraquat × sex × genotype interaction (F_1,63=4.29, p=0.043, three-way ANOVA) where TH-driven dj-1β knockdown in males boosted the percentage of DA neurons expressing dVGLUT compared to male Con flies (p=0.018, Bonferroni multiple comparisons test). N=5–22 fly brains per group. (E) Schematic summarizing mechanisms of dVGLUT-mediated DA neuron resilience. dVGLUT expression protects DA neurons by decreasing ATP production and mitochondrial ROS levels, which is reflected by lower mitochondrial gene expression in dVGLUT⁺ DA neurons. Mitochondrial oxidative stress vulnerability may also be lower in dVGLUT-expressing DA neurons due to increased glutathione and decreased cytoplasmic DA oxidation. Additionally, vulnerability may be impacted by differential expression of transcription factors that impact DA neuron development and/or differentiation. Finally, during periods of cell stress, DA neuron dVGLUT expression is raised as part of a neuroprotective response to accumulating mitochondrial ROS. This increase in dVGLUT expression is triggered by the concomitant loss of dj-1β during periods of cell stress/insult, particularly in males. *p<0.05, **p<0.01, ^#p<0.05 compared to Con Male receiving Control food. All data are plotted as Mean±SEM.