Parkin knockdown modulates dopamine release in the central complex, but not the mushroom body heel, of aging Drosophila

Abstract

Parkinson’s disease (PD) is characterized by progressive degeneration of dopaminergic neurons leading to reduced locomotion. Mutations of parkin gene in Drosophila produce the same phenotypes as vertebrate models, but the effect of parkin knockdown on dopamine release is not known. Here, we report age-dependent, spatial variation of dopamine release in the brain of parkin-RNAi adult Drosophila. Dopamine was repetitively stimulated by local application of acetylcholine and quantified by fast-scan cyclic voltammetry in the central complex or mushroom body heel. In the central complex, the main area controlling locomotor function, dopamine is release is maintained for repeated stimulations in aged control flies, but lower concentrations of dopamine are released in the central complex of aged parkin-RNAi flies. In the mushroom body heel, the dopamine release decrease in older parkin-RNAi flies is similar to controls. There is not significant dopaminergic neuronal loss even in older parkin knockdown flies, which indicates that the changes in stimulated dopamine release are due to alterations of neuronal function. In young parkin-RNAi flies, locomotion is inhibited by 30%, while in older parkin-RNAi flies it is inhibited by 85%. Overall, stimulated dopamine release is modulated by parkin in an age and brain region dependent manner. Correlating the functional state of the dopaminergic system with behavioral phenotypes provides unique insights into the PD mechanism. Drosophila can be used to study dopamine functionality in PD, elucidate how genetics influence dopamine, and test potential therapies to maintain dopamine release.

Graphical Abstract

Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disease.¹ The hallmark of PD in mammals is degeneration of dopamine neurons that reduce dopamine neurotransmission in the nigrostriatal pathway, triggering locomotor and non-locomotor symptoms.² Similar to other neurodegenerative diseases, aging is one of the main risk factors in the progression of PD and its symptoms.³ While most PD cases are idiopathic, 5-10% of them are linked to genetic causes,⁴ and mutations of specific genes are used to study PD in animal models.⁵ Drosophila melanogaster (fruit fly) is a powerful, genetically-accessible model to study PD.⁶ Fruit flies have a relatively short life cycle and perform complex locomotor and learning behaviors, while the dopaminergic system functions similarly to mammals.⁷ Many PD-linked human genes have homologs in the Drosophila genome or have been expressed in fruit flies, including parkin,⁸ pink1,⁹ LRRK,¹⁰ and α-synuclein.¹¹ Targeting these genes, Drosophila PD genetic models recapitulate some of the same physiological features observed in vertebrate models.^12,13 The parkin gene, part of the PARK2 locus, acts downstream from multiple other PD gene targets. Several studies have demonstrated that the parkin gene is active in the same pathway as the pink1 gene.^14-16 While parkin downregulation has been linked with development of early-onset PD and its typical phenotypes,¹⁷ overexpression of parkin in pink1 flies rescues neurodegeneration and locomotor defects. Additionally, expression of parkin in LRRK2 mutant flies counteracts the loss of dopaminergic neurons.¹⁸ These results support the involvement of parkin mutations in PD development, but the effect of parkin on local dopamine release dynamics has not been documented in specific brain regions.

Our lab developed electrochemistry-based approaches to measure neurotransmitters using fast-scan cyclic voltammetry (FSCV)^19,20 in the brain of larval and adult Drosophila.^21-29 Neurotransmitter release is stimulated using acetylcholine, chemogenetics, or optogenetics.³⁰ Moreover, the implanted carbon-fiber microelectrode (CFME) is targeted to specific regions of the fly brain to measure local release dynamics.^29,30 The modulatory effect of PD-linked genes such as parkin has not been studied in the dopaminergic system of adult Drosophila. In recent work, we reported that downregulation of PD candidate gene RNF11 is associated with higher optogenetically-stimulated dopamine release and faster clearance in fruit fly larvae, which supports a potential protective role for RNF11 in maintaining dopamine signaling in PD pathology.³¹ Thus, FSCV provides rapid information about the functionality of dopaminergic neurons and quantitative dopamine dynamics. There are many genetic models of PD that have been made in fruit flies, but they have not been tested to explore their dopamine release. Measurements of dopamine release will improve our understanding of dopamine neurotransmission alterations in PD pathology and complement behavioral studies, neuronal population counting, and dopamine tissue content measurements.^14,32

The goal of this study was to quantify the age-dependent variation of dopamine in parkin-RNAi Drosophila in two different brain regions. We hypothesized that dopamine release will decrease with age in parkin-RNAi flies in a heterogeneous manner across different regions of the brain. We measured dopamine release after repeated acetylcholine (ACh) stimulations in the central complex and the mushroom body heel. Lower concentrations of dopamine are released in the central complex of aged parkin-RNAi flies, the main area controlling locomotor function, but dopamine release is maintained in the mushroom body heel of aged parkin-RNAi flies. The lack of significant degeneration of dopaminergic neurons suggests the quantitative changes in dopamine release are due to alteration of neuronal function but not death. The locomotor ability is reduced by about 30% in young parkin-RNAi flies, before changes in dopamine release are apparent, but locomotion is severely limited (85% decrease) in older parkin flies, where dopamine is reduced. Therefore, the parkin gene modulates dopamine release in an age and brain region dependent manner. Studying the characteristics of local dopamine release can contribute to a better understanding of the relationship between dopamine neurotransmission and behavioral phenotypes during the progression of genetic PD.

Results and Discussion

Knockdown of parkin gene leads to age-dependent decrease of locomotion

The goal of this study was to assess the effects of parkin knockdown on behavioral changes, dopamine cell loss, and phasic dopamine release in a common genetic model of PD. Here, we employed the GAL4/UAS system to drive the knockdown of the parkin gene in dopamine-containing neurons of Drosophila using a parkin^HMS01800-RNAi fly stock generated by the Drosophila Transgenic RNAi Project.³³ This parkin-RNAi mutant was associated with reduced locomotion both in larvae when driven with a muscle-specific driver³⁴ and in adults when driven with the pan-neuronal driver elav-GAL4.³⁵ To verify the efficacy of the UAS-parkin-RNAi construct, the RNAi flies were crossed to the ubiquitous actin (Act5C) driver, and gene expression levels were monitored by quantitative real-time PCR (Figure S1). Parkin mRNA levels were reduced by 57 ± 2% of the control in the RNAi allele in 1 day old flies (unpaired t-test, p < 0.0001). At 45 days of age, parkin mRNA levels were reduced by 94 ± 0.3% in the RNAi allele (unpaired t-test, p < 0.002).

The main behavioral characteristic of PD is a reduction in motor activity which is expected to be age-dependent. Here, we employed a tyrosine hydroxylase driver (TH-GAL4) that drives the reduced expression of parkin in dopaminergic neurons that contain tyrosine hydroxylase, the enzyme responsible for dopamine synthesis.³⁶ Figure 1 shows a comparison between the activity of young (1 day post-eclosion (dpe)) and old (45 dpe) flies for both UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies and UAS-mCD8-GFP; TH-GAL4 control flies. A one-way ANOVA shows a significant main effect (F_(3,141)=94.06, p<0.0001). Post hoc analysis (Šídák test) shows no effect of aging for the control flies (p=0.9927), but significantly decreased activity in 45 dpe parkin-RNAi flies (p<0.0001). Knocking down parkin causes a decrease in activity for both 1 dpe (p<0.0001, 1 dpe parkin vs. control flies) and 45 dpe flies (p<0.0001, 45 dpe parkin vs. control flies). Figure S2 shows that the decreased activity is not innate to the UAS-parkin-RNAi stock flies, as UAS-parkin-RNAi flies not crossed with a driver show high activity independent of age, similar to that of the control flies (see figure caption for detailed comparisons). Thus, the parkin-RNAi flies show two characteristic hallmarks of PD, reduced locomotion in comparison with the control genotype flies and progressive age-dependent activity reduction.

Figure 1.

The effect on locomotion of parkin-RNAi expressed in dopaminergic neurons using the TH-GAL4 driver is dependent on age. Activity monitoring comparison for UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies, at young (park., 1) and old (park., 45) age, and UAS-mCD8-GFP; TH-GAL4 flies, at young (cont., 1) and old (cont., 45) age. Activity levels are proportional to the average daily counts per fly recorded by the Drosophila activity monitor system. The number of individual flies (males only) monitored for each group are: control, 1 (n=32, average age at start of 7 days of activity monitoring 1.3 ± 0.1 dpe), control, 45 (n=32, average age 41.8 ± 0.2 dpe), parkin, 1 (n=29, average age 1.6 ± 0.1 dpe), and parkin, 45 (n=52, average age 41.7 ± 0.1 dpe). Error bars are SEM. See text for statistics **** p<0.0001.

The number of dopaminergic neurons is not significantly reduced in parkin-RNAi flies

PD progression is generally characterized by an age-dependent progressive dopaminergic neuronal loss,³⁷ which could lead to a reduction in dopamine release. To identify dopaminergic neurons, we employed both an antibody marker for tyrosine hydroxylase (anti-TH) and GFP, which shows TH-containing neurons expressed by the TH-GAL4 driver. Dopaminergic neurons were counted in both hemispheres of the Drosophila brain from six major clusters: paired posterior lateral 1 and 2 (PPL1 and PPL2), paired posterior medial 1 and 2 (PPM1/2), paired posterior medial 3 (PPM3), paired anterior lateral (PAL), and paired anterior medial (PAM). The PAM and PAL clusters are identified in the anterior side of the brain (Figure 2A-B), while the PPM1/2, PPM3, PPL1, and PPL2 clusters are found in the posterior side of the brain (Figure 2C-D). Representative confocal z-stack projection images of the anterior (Figure 2B) and posterior (Figure 2D) side of a brain from a 1 dpe UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi fly show the overlapped expression of GFP and anti-TH staining marking all six neuron clusters within the brain. Figure S3 shows images of parkin-RNAi flies at 1 dpe (Figure S3A) and 45 dpe (Figure S3B) while Figure S4 shows control flies at 1 dpe (Figure S4A) and 45 dpe (Figure S4B). The overall neuronal counts averaged from multiple samples are given in Table S1 for GFP and Table S2 for anti-TH staining.

Figure 2.

Dopaminergic cells in parkin-RNAi flies. The dopaminergic system in Drosophila is comprised of distinct dopamine neuron clusters identified by GFP and anti-TH labeling in UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi. Six different dopamine neuron clusters are found in the anterior (A, B) and posterior (C, D) side of the Drosophila brain. Representative confocal z-stack projection images show the GFP (green) and anti-TH (red) labeling with overlap looking yellow across a whole brain from a 1 dpe parkin-RNAi fly, in the anterior (B) and posterior (D) side. The PAM and PAL clusters are on the anterior side, while the PPM1/2, PPM3, PPL1, and PPL2 clusters are on the posterior side.

Our focus was to analyze the variation in the number of dopaminergic neurons per cluster, especially for clusters projecting to the central complex and the mushroom body heel, where dopamine was measured. The PPM3 cluster projects to the central complex^38,39 and was identified in all samples by both anti-TH and GFP staining (Figure 3A). Based on the anti-TH staining (Figure 4D), there is no main effect of the fly group on neuron count (One-way ANOVA, F_(3,72)=2.633, p=0.0564). For GFP expression (Figure S5A), there is a main effect of fly group on number of neurons (one-way ANOVA, F_(3,76)=10.14, p<0.0001), with significant differences observed between 1 dpe and 45 dpe flies, for both parkin-RNAi flies (p<0.0001) and control flies (p=0.0367).

Figure 3.

Effects of aging or knocking down the parkin gene on dopaminergic neurons in the clusters projecting towards the central complex or the mushroom body. (A-C) Confocal z-stack projection images show the (A) PPM3, (B) PPL1, and (C) PAM clusters identified in UAS-mCD8-GFP; TH-GAL4 control flies, 1 dpe (male, left hemisphere) and 45 dpe (male, right hemisphere), and UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies, 1 dpe (female, left hemisphere) and 45 dpe (male, right hemisphere). All images show the overlap (yellow) of anti-TH (red) and GFP (green) staining and they are close ups taken from the whole brain images in Figures S2 and S3. Scale bar is 10 μm. (D) Comparison of mean number of TH-positive neurons in the PPM3 cluster. One-way ANOVA shows no significant main effect (F_(3,72)=2.633, p=0.0564). (E) Comparison of mean number of TH-positive neurons in the PPL1 cluster. (F) Comparison of mean number of TH-positive neurons in the PAM cluster. Error bars are SEM for “n” individual brain hemispheres where the cluster was identified (see Tables S1 and S2 for detailed numbers, text for statistics). (****) p<0.0001.

Figure 4.

(A) Stimulation sequence timeline. After the fly brain is dissected, the brain rests for ~ 20 min while the electrode and capillary are implanted. ACh (0.2 pmol) is applied for 6 consecutive stimulations, with 1 min recording time for each stimulation and 10 min rest between stimulations. (B) Microscope image of a brain from a UAS-mCD8-GFP; TH-GAL4 fly with the microelectrode and injection capillary implanted in the central complex (CC). GFP expression (green) driven by TH-GAL4 helps identify the region. (C) Microscope image of a brain from a UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi fly with the microelectrode and injection capillary implanted in the mushroom body (MB) heel.

The mushroom body heel receives projections from dopaminergic neurons in the PPL1 and PAM clusters.⁴⁰ The PPL1 cluster was identified through both anti-TH and GFP staining (Figure 3B). With anti-TH staining (Figure 3E), there is a main effect of group (One-way ANOVA, F_(3,74)=19.34, p<0.0001). Post hoc analysis shows significant differences between young and old control flies (p<0.0001), and older parkin-RNAi flies and control flies (p<0.0001). GFP expression (Figure S5B) marks largely the same number of neurons as anti-TH. The PAM cluster is identified by anti-TH staining (Figure 3C, F) and there is no main effect of genotype and age (F_(3,64)=6.885, p=0.7075). In comparison, the TH driver was relatively ineffective in targeting the PAM dopaminergic neurons by GFP (Figure S5C), especially in older flies.

Additionally, we identified and quantified other dopaminergic neuronal clusters which do not project towards either the central complex or the mushroom body heel. Figure S6 shows the number of dopaminergic neurons for the PAL, PPM1/2, and PPL2 clusters and there are generally no significant differences between the different fly groups based on either anti-TH or GFP staining.

Stimulated dopamine release in the central complex is reduced in parkin-RNAi flies in an age-dependent manner

The central objective of this research is to assess the dopamine dynamics in discrete regions in the brain of parkin-RNAi flies. We employed acetylcholine (ACh) to stimulate endogenous dopamine release in a dissected Drosophila brain, and detected dopamine at an implanted microelectrode using FSCV. Six consecutive injections of a 0.2 pmol bolus of ACh were applied every 10 min, with 1 min of data recording period for each injection (Figure 4A). In previous studies, this rest time between repeated injections allows the releasable dopamine pool to recover between stimulations.²⁹ Measurements in the central complex (CC) were performed in flies crossed with a GFP responder (UAS-mCD8-GFP) that expresses GFP in dopaminergic neurons across the brain, allowing for easy visualization of the CC during electrode implantation (Figure 4B).

The release of dopamine was assessed in the CC of young and old UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies and in UAS-mCD8-GFP; TH-GAL4 flies as genetic controls. Figure 5A, B (parkin) and Figure S7A, B (control) show a set of representative color plots for the six consecutive stimulations. The presence of two redox peaks, an oxidation peak on the forward scan and a reduction peak on the back scan, indicates the selective detection of dopamine. Example concentration versus time traces taken at peak oxidation current are shown in Figure 5C-F. In young flies, control (C) and parkin-RNAi (E) have similar trends; dopamine release is constant over all six stimulations. However, older control (D) and parkin-FRNAi (F) flies exhibit decreasing dopamine release with each stimulation. The half-decay time (t_1/2) of each trace does not vary significantly between fly groups (Figure S8A, one-way ANOVA, F_(3,36)=1.442, p=0.2468).

Figure 5.

Dopamine release in the central complex (CC) of parkin-RNAi flies. (A-B) Representative 15 s long color plots of measurements performed in the CC of UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies at (A) 1 dpe and (B) 45 dpe. ACh (0.2 pmol) was applied after 5 s with 10 min between stimulations. (C-F) Dopamine concentration versus time traces for 6 stimulations performed in the CC of control (UAS-mCD8-GFP; TH-GAL4) flies at (C) 1 dpe and (D) 45 dpe. Dopamine release for 6 stimulations in parkin-RNAi flies at (E) 1 dpe and (F) 45 dpe. (G) Averaged dopamine concentration across all six stimulations. (H) Normalized dopamine concentration (to the first stimulation value). Error bars are SEM (shown in one direction only for clarity) for n=10 individual fly brains (5F/5M). See text for statistics.

Figure 5G shows the concentration of CC dopamine release as a function of stimulation number for control and parkin-RNAi flies at young and old age. Three-way ANOVA shows a significant effect of genotype (F_(1,216)=4.629, p=0.0326), age (F_(1,216)=7.824, p=0.0056), and stimulation number (F_(5,216)=11.61, p<0.0001) on dopamine release in the CC. In young parkin-RNAi flies, dopamine release is relatively constant over all six stimulations, similar to the age-matched control flies. However, the 45 dpe parkin-RNAi flies show decreasing dopamine at late stimulations, with concentration values smaller than the 1 dpe parkin-RNAi flies or the age-matched older control. Dopamine release is affected by aging across multiple stimulations (age x stimulation number two-way interaction, F_(5,216)=2.884, p=0.0153), and by genotype (age x genotype two-way interaction, F_(1,216)=15.08, p=0.0001). However, independent of age, the effect of parkin is not statistically significant (genotype x stimulation number, two-way interaction, F_(5,216)=0.413, p=0.8394). To simplify statistics, we also looked at the difference between the last stimulation and the first stimulation (Fig. S9B) and found that the delta was significantly different between control young and old flies, but not for parkin-RNAi young and old flies.

Figure 5H compares normalized concentration data (concentration values normalized to the first stimulation value for each individual fly dataset) to better observe the differences between trends. Three-way ANOVA indicates a significant effect on normalized dopamine concentrations of age (F_(1,216)=53.44, p<0.0001) and stimulation number (F_(5,216)=19.5, p<0.0001), but not genotype (F_(1,216)=1.838, p=0.1766). At young age, both parkin-RNAi and control flies exhibit a similar pattern of dopamine release, decreasing by about 30%. In comparison, 45 dpe aged flies show a steeper decrease for both parkin-RNAi and control flies. There is not a significant effect of genotype and stimulation number (genotype x stimulation number two-way interaction, F_(5,216)=0.2355, p=0.9466), but there are significant interactions of age with genotype (genotype x age two-way interaction, F_(1,216)=7.952, p=0.0053) and age with stimulation number (age x stimulation number two-way interaction, F_(5,216)=4.096, p=0.0014).

Dopamine release dynamics in parkin-RNAi flies in the mushroom body heel

Knocking down parkin may have different effects on dopamine in other brain regions. To test this hypothesis, we assessed dopamine release patterns in a second region, the mushroom body (MB) heel of the same parkin-RNAi flies used for CC experiments (Figure 4C). We employed DAT-GAL4, UAS-mCD8-GFP flies as control to facilitate localizing the MB heel (Figure S10A).⁴¹

Representative color plots recorded in the MB heel of both young and old parkin-RNAi (Figure 6A, B) and control flies (Figure S7C, D) confirm dopamine is detected. Figures 6C-F show the concentration versus time plots. In young flies, dopamine release is constant over 6 stimulations for control flies in the MB (Figure 6C), but for parkin-RNAi flies (Figure 6E) it starts at higher concentrations before decreasing more quickly to reach similar baseline values for the later stimulations. Dopamine release decreases with increasing stimulation number in both control aged flies (Figure 6D) and parkin-RNAi aged flies (Figure 6F). The half-decay time of dopamine release is not significantly different between the groups (Figure S8B). Figure 6G shows the comparison of average dopamine concentrations in the MB heel as a function of stimulation number for each genotype and age tested. Three-way ANOVA shows a significant change in dopamine release in the MB heel based on genotype (F_(1,216)=4.927, p=0.0275), age (F_(1,216)=47.16, p<0.0001), and stimulation number (F_(5,216)=6.586, p<0.0001). For parkin-RNAi, the young flies have high dopamine release that slowly decreases with stimulation number, while old flies exhibit a decreasing trend with lower concentration values overall. In young control flies, dopamine release in MB heel is constant over all stimulations, but in older control flies, dopamine concentrations are lower and decreasing with stimulation number. There are no statistically significant interactions between genotype, age, and stimulation number (age x genotype, F_(1,216)=0.4386, p=0.5085; age x stimulation number, F_(5,216)=0.2932, p=0.9164; genotype x stimulation number, F_(5,216)=0.3967, p=0.8508). Examining at the delta between the first and last stimulation (Fig. S9D), there are no significant differences among young and old control or parkin flies. Figure 6H shows normalized data to compare the trends. There is a difference between young control flies which have constant release and young parkin-RNAi flies which have release decreasing 40%. Both control and parkin-RNAi aged flies exhibit about a 50% decrease in release for the 6^th stimulation. There is a main effect of genotype (F_(1,216)=4.927, p=0.0275), age (F_(1,216)=47.16, p<0.0001), and stimulation number (F_(5,216)=6.586, p<0.0001). The effect of age varies significantly with genotype (genotype x age two-way interaction, F_(1,216)=5.35, p=0.0217), but there are not significant interactions of stimulation number with genotype (F_(5,216)=0.6065, p=0.6950) or age (F_(5,216)=1.375, p=0.2348). Thus, there are fewer effects of parkin knockdown on dopamine release in the MB heel.

Figure 6.

Dopamine release in the mushroom body (MB) heel of parkin-RNAi flies. (A-B) Representative 15 s long color plots of measurements performed in the MB heel of (A) 1 dpe and (B) 45 dpe UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies. ACh (0.2 pmol) was applied after 5 s. (C-F) Dopamine concentration versus time traces in the MB heel of DAT-GAL4, UAS-mCD8-GFP flies, (C) 1 dpe and (D) 45 dpe, and UAS-mCD8-GFP; TH-GAL4/UAS-parkin-RNAi flies, (E) 1 dpe and (F) 45 dpe. (G) Averaged dopamine concentration across all six stimulations. (H) Normalized dopamine concentration (to the first stimulation value). See text for statistics. Error bars are SEM (shown in one direction only) for n=10 individual fly brains (5F/5M).

Dopamine release under repetitive stimulations is altered by aging

In this study, we examined the dynamic release profile of dopamine under repeated stimulation, which is a model of stress that allow us to understand dynamics of release. Repetitive stimulation assesses the ability of the dopaminergic system to respond to continued stress. Aging induces a progressive impairment of behaviors and functions of the nervous system of Drosophila^42,43 and the total dopamine pool decreases with age.⁴⁴ Our results demonstrate aging leads to a change of the dopamine release patterns during repeated neuronal activity. These are the first results to show deficits in dopamine signaling during aging in Drosophila. Young flies maintain a constant release of dopamine when repeatedly stimulated, but older flies release a higher concentration at first that drops dramatically with repeated stimulation. Both the CC and MB heel have similar trends in older flies, with decreasing dopamine concentration with repeated stimulations, suggesting dopamine neurotransmission function is altered across the brain with aging. The number of dopamine neurons was largely maintained with aging, except in the PPL1. Thus, declines in dopamine from aging were not due mainly to neuronal loss, but from greater depletion of dopamine pools. Future studies could examine dopamine tissue content or basal levels, to give insight into the extent that these contribute to changes in dopamine dynamics. However, the measurements of functional dopamine in response to repeated stimulations suggest older flies would have more decreased responses to repeated salient stimuli in their environment.

Parkin knockdown decreases dopamine in the central complex of aged flies, but not in the mushroom body heel

Dopamine neurotransmission in the central complex of Drosophila modulates locomotor functions, including flight, walking, and visual orientation,^45-47 and mutations in dopaminergic neurons projecting to this brain region lead to decreased locomotor activity.⁴⁸ Therefore, we hypothesized that there would be changes in dopamine dynamics in the central complex during PD. Stimulated dopamine concentrations significantly decreased in the central complex of older parkin-RNAi flies, corresponding to the age-dependent neurodegeneration characteristic of PD. While aging independently exerts an inhibitory effect on stimulated dopamine release, this effect is compounded when parkin is downregulated. Parkin is responsible for ubiquitination of proteins that regulate synaptic vesicle release or recycling, as well as post-synaptic membrane receptors.⁴⁹ The pre-synaptic function of parkin may be conserved in Drosophila in a similar manner as in vertebrates, with its loss of function leading to a decrease in dopamine release with age. Additionally, parkin is involved in mitochondrial quality control.⁵⁰ In the absence of parkin, dysfunctional mitochondrial activity can indirectly limit the extent to which dopamine is released through energy-consuming exocytosis events. The effect of parkin being limited to aged flies is particularly striking and it may be related to the parkin expression being knocked down, but not completely knocked down, in dopaminergic neurons, which conserves its function in young adult flies. Dopamine compensation mechanisms in early parkin-induced PD development could be responsible for limiting the measurable effects on dopamine release in young flies.⁵¹

While there is reduced dopamine in the central complex of aged parkin-RNAi flies, dopamine release in the mushroom body heel is not decreased in aged parkin-RNAi in comparison with control flies. Additionally, we knocked down parkin using a DAT-GAL4 driver that more specifically targets the mushroom body cells and there was also no significant effect on dopamine release (Figure S10). Thus, parkin has a heterogeneous effect on modulating dopamine across different brain regions. While the mushroom body is mainly known for its involvement in learning and memory processes,^52,53 there are dopaminergic projections from the PAM cluster innervating the general area of the heel and they are involved in motor-related functions.^54-56 However, the low number of GFP-positive PAM neurons in the TH-GAL4 flies suggests the driver does not target the PAM neurons effectively, as observed in previous studies.^36,57 Therefore, few effects of parkin knockdown on dopamine dynamics are observed in the mushroom body heel.

Dopamine dynamics are maintained after locomotion deficits are apparent

Motor symptoms associated with PD progression develop after dopamine neurotransmission declines substantially due to dopaminergic neuronal loss.^2,58 Here, parkin downregulation leads to a small decrease in locomotion in young flies (30% decrease), while the dopamine release pattern in the central complex is conserved. Thus, locomotor deficits are apparent before a decrease in dopamine release in the central complex. Our studies only measure phasic dopamine, and there could be changes in tonic basal levels. In addition, other neurotransmission pathways may be altered, such as octopamine and tyramine, which are also linked to locomotion.⁵⁹ Other brain regions also control locomotion and changes may be more evident there in young flies. On the other hand, locomotor behavior is altered to a greater extent in older parkin-RNAi flies (85% decrease) concomitantly with the overall decrease of released dopamine concentrations in the central complex. In comparison, wild-type flies maintain their locomotor behavior during aging. Future studies are needed to precisely correlate changes in dopamine with specific locomotor functions in order to understand how flies compensate differently for aging and parkin-mediated alterations of dopamine release.

Dopamine release is reduced despite no significant dopaminergic cell loss

The decrease in dopamine neurotransmission in PD models is usually attributed to a loss of dopaminergic neurons,² although there is conflicting evidence regarding the extent of neuronal loss in Drosophila PD models. Some studies show parkin mutations induce a significant age-dependent loss of dopaminergic neurons in specific clusters.^8,60-62 Other studies have reported no dopaminergic neuronal loss,^63,64 such as the comprehensive study by Navarro et al. for parkin and pink PD models.⁶⁵ In our parkin model, there is not a consistent loss of dopaminergic neurons across the brain before the 45 days old age point. Specifically, there is no decrease in TH-positive neurons in the PPM3 cluster, which is projecting towards the central complex site where reduced dopamine release is observed at advanced age. The relationship between neuronal loss, locomotor activity, and dopamine neurotransmission in this parkin knockdown model is different from the typical PD progression mechanism, when locomotion and dopamine neurotransmission alterations appear after a significant loss of dopaminergic neurons.⁶⁶ Therefore, we suggest that knocking down parkin specifically in TH-positive neurons decreases dopamine release by affecting the functionality of dopaminergic neurons, and not through neuronal death.

There are many possible mechanisms for how parkin knockdown affects dopamine release. For example, reduced synthesis of dopamine in neurons or differential vesicular packing could lead to lower dopamine neurotransmission. Aging does not lead to dopaminergic neuronal loss, but neurons are smaller in size and exhibit reduced TH expression, indicating a reduced dopamine pool for neurotransmission.⁶⁷ Targeted genetic manipulation of dopamine synthesis pathways and vesicular packaging can lead to the development of PD phenotypes such as reduced locomotor activity,^68-70 and dopamine transporter changes may alter dopamine neurotransmission in PD models.⁷¹ However, here we saw no significant variation of dopamine uptake in any of the fly genotypes and ages studied, similar with a previous report in a pink1 Drosophila model.³² These results show that dopaminergic neuronal loss is not an essential physiological feature in order to trigger the reduced locomotion and altered dopamine signaling characteristic to PD.

Conclusions

This study examined the effect of a loss-of-function parkin mutation on local dopamine neurotransmission in the brain of Drosophila. Repetitive dopamine release is inhibited in the central complex of aging parkin-RNAi flies, but not the mushroom body heel, suggesting an age-dependent and heterogeneous effect of parkin on local dopamine neurotransmission. The variation in dopamine release dynamics is attributed to changes in the function of dopaminergic neurons, as no significant dopaminergic neuronal loss has been observed due to aging or parkin knockdown. Parkin-RNAi flies have severe locomotor deficits at older age, and a decrease in repeated dopamine release. However, younger parkin-RNAi flies exhibit a 30% reduction in locomotor activity without changes in central complex dopamine, which indicates other brain regions and neurotransmission pathways that control motor functions are altered. Measurements of local dopamine dynamics in Drosophila offers unique information about the modulatory effect of PD-linked genes on the functional state of the dopaminergic system. Future research can measure dopamine release in Drosophila PD models to understand the mechanism of PD and study potential therapeutic approaches to slow dopamine neurotransmission deficits during disease progression.

Methods

Drosophila stocks and aging protocol

The study design is shown in Figure 7. The TH-GAL4 (3^rd chromosome) fly line was donated by the Jay Hirsh lab at University of Virginia. Fly stocks DAT-GAL4 (3^rd chromosome, stock #48359), UAS-mCD8-GFP (2^nd chromosome, stock #5137), UAS-mCD8-GFP (3^rd chromosome, stock #5130), UAS- parkin^HMS01800 RNAi (3^rd chromosome, stock #38333), and Act5C-GAL4 (2^nd chromosome, stock #4414) were obtained from the Bloomington Drosophila Stock Center (Indiana University, Bloomington). The stable fly lines UAS-mCD8-GFP; TH-GAL4 and DAT-GAL4, UAS-mCD8-GFP (3^rd chromosome) were obtained by crossings between stocks and double balancer fly lines and used as control genotype. The UAS-parkin-RNAi fly line was crossed with the two control lines to obtain parkin knockdown flies. For aging, flies from each line were collected daily soon after eclosion and segregated based on sex in tubes with fresh fly food (LabExpress). The food tubes were replaced every 7 days. Aging flies were kept at 25 °C in a 12h:12h light-dark cycle incubator.

Figure 7.

Diagram showing the study design. Control and parkin-RNAi flies were assessed at two age points, 1 dpe (day post-eclosion) and 45 dpe, for stimulated dopamine (DA) release measured using FSCV, locomotor ability measured by an activity monitoring system, and dopaminergic neuronal count using tyrosine hydroxylase (TH) immunohistochemistry.

Activity monitoring

Locomotion was assessed using a Drosophila Activity Monitor (DAM) system (TriKinetics) and data was analyzed using the ShinyR-DAM open-source cloud-based application developed by the Jay Hirsh lab.⁷² The DAM systems records the number of times (counts) a fly breaks an infrared beam while walking in a glass tube. The number of counts is directly proportional with the locomotor activity of the flies. Male flies of different genotypes and ages were loaded in activity monitor tubes, with food fly at one end and a cotton plug at the other end. Locomotor activity was monitored for 7 days using 12h:12h light:dark cycle. Average number of counts per fly per day were extracted from the ShinyR-DAM application. In total, 145 flies were used for 4 experimental groups (n=29-52 for each group).

Tyrosine hydroxylase (TH) immunohistochemistry

The immunohistochemistry procedure was adapted from Barone et. al.⁷³ and it is presented in detail in the Supporting Information. Brain samples were incubated with anti-tyrosine hydroxylase (anti-TH) polyclonal primary antibody (from rabbit, Chemicon^®, Sigma-Aldrich) and goat anti-rabbit IgG (H+L) highly cross-adsorbed secondary antibody with Alexa Fluor 594 (ThermoFisher Scientific). Samples were visualized using a Nikon A1Rsi upright confocal microscope (Nikon Instruments Inc.) with GaAsP detectors and a 40X NIR Apo water immersion objective. Red channel (561.6 nm) was used for the anti-TH staining and the green channel (487.8 nm) was used for the GFP staining. Z-stack images were recorded simultaneously for both channels with a 1 nm incremental z step. Images were processed using the Fiji software to create z-stack projection images.⁷⁴ Individual neurons were counted using the cell counter plugin. In total, 42 flies were used for 4 experimental groups (n=7-12 for each group).

Electrochemical measurements

CFMEs with a ~60 μm protruding tip were prepared as described before.²⁹ Preparation is detailed in the Supporting Information. A Waveneuro FSCV potentiostat (5 MΩ headstage, Pine Research Instrument) and a PCIe-6363 multifunctional I/O device (National Instruments) were used to apply the waveform and to collect data. HDCV software (provided by R. M. Wightman, University of North Carolina) was used to collect data and for background subtraction. The dopamine waveform (−0.4 V to 1.3 V and back to −0.4 V versus chloridized Ag wire reference electrode) was applied to the CFME every 100 ms at a scan rate of 400 V/s for conditioning and measurements. Electrodes were pre-calibrated and post-calibrated with 1 μM dopamine in a flow cell injection system. For picospritzing, an empty glass capillary was pulled, trimmed, and filled with acetylcholine (ACh) for stimulation. A Picospritzer III (Parker Hannfin) was used to pressure eject ACh into the brain tissue. Capillaries were calibrated by picospritzing a droplet of ACh in oil and the diameter of the pressure-ejected droplet was measured using DS-Qi2 monochrome CMOS camera and NIS-Elements BR imaging software (Nikon Instruments Inc.). For all experiments, a volume corresponding to 0.2 pmol ACh was ejected using a constant applied pressure of 20 psi.

Brain tissue preparation and stimulation experiments

Dissected brains from adult Drosophila of appropriate age were prepared for measurements as previously described.²⁹ Briefly, an adult fly was anesthetized by rapid chilling on ice as a quick and effective approach to minimize animal suffering before dissection. The fly was transferred to a Petri dish with cold dissection buffer (see supporting information) under a stereo microscope (AmScope), the head was removed, and the brain tissue was isolated by removing all external tissue with sharp tweezers. The harvested brain was transferred to another Petri dish containing dissecting buffer and fixed at the bottom of the dish, with the anterior side up. The stimulation experiments were performed under a SMZ800N stereoscope (Nikon Instruments Inc.). The CFME was placed using a micromanipulator (Narishige International) either in the central complex or the mushroom body heel, using the GFP marker expression to guide the implantation. The ACh-filled pipette was positioned using another micromanipulator near the CFME tip. The brain tissue was allowed to equilibrate for 10 min before CFME implantation, and another 10 min after implantation and before the first stimulation. The stimulation sequence consisted of six injections of 0.2 pmol ACh applied every 10 min. For each injection, electrochemical data were recorded for 60 s, with the ACh application performed after 5 s. In total, 100 flies were used for 10 experimental groups (n=10, 5 females / 5 males, for each group).

Statistical analysis

All statistical analysis tests were performed using GraphPad Prism 9 (Version 9.3.0, GraphPad Software Inc.). Data are presented as mean ± standard error of the mean (SEM). Activity monitoring data and dopaminergic neuron numbers were analyzed using one-way ANOVA with Šídák post-hoc test. Dopamine release data were analyzed using three-way ANOVA.

Abbreviations:

ACh

acetylcholine

ANOVA

analysis of variance

CFME

carbon fiber microelectrode

DA

dopamine

DAM

Drosophila activity monitor

DAT

dopamine transporter

dpe

days post-eclosion

F

female

FSCV

fast-scan cyclic voltammetry

GFP

green fluorescent protein

M

male

PAL

paired anterior lateral

PAM

paired anterior medial

PBS

phosphate buffer solution

PD

Parkinson’s disease

PPL1

paired posterior lateral 1

PPL2

paired posterior lateral 2

PPM1/2

paired posterior medial 1 and 2

PPM3

paired posterior medial 3

TH

tyrosine hydroxylase