Dopamine Release Impairments Accompany Locomotor and Cognitive Deficiencies in Rotenone-Treated Parkinson’s Disease Model Zebrafish

Abstract

In this work, we carried out neurochemical and behavioral analysis of zebrafish (Danio rerio) treated with rotenone, an agent used to chemically induce a syndrome resembling Parkinson’s disease (PD). Dopamine release, measured with fast-scan cyclic voltammetry (FSCV) at carbon-fiber electrodes in acutely harvested whole brains, was about 30 percent of that found in controls. Uptake, represented by the first order rate constant (k) and the half-life (t_1/2) determined by nonlinear regression modeling of the stimulated release plots, was also diminished. Behavioral analysis revealed that rotenone treatment increased the time required for zebrafish to reach a reward within a maze by more than 50 percent and caused fish to select the wrong pathway, suggesting that latent learning was impaired. Additionally, zebrafish treated with rotenone suffered from diminished locomotor activity, swimming shorter distances with lower mean velocity and acceleration. Thus, the neurochemical and behavioral approaches, as applied, were able to resolve rotenone-induced differences in key parameters. This approach may be effective for screening therapies in this and other models of neurodegeneration.

Graphical Abstract

INTRODUCTION

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that results in motor and cognitive dysfunction¹. Patients with PD lose dopaminergic neurons in the substantia nigra²; this loss decreases levels of dopamine in regions to where these neurons project, including the caudate and putamen³. The degeneration of dopaminergic neurons and other neuronal systems in PD may result in bradykinesia⁴, diminished short-term memory⁵, and motor learning deficits⁶ as well as other cognitive and outward physical symptoms.

The zebrafish (Danio rerio) is rapidly becoming a useful model organism in PD research. Zebrafish have the advantage of enabling higher throughput studies, compared to rodents, while still having a complex vertebrate central nervous system. This enhanced throughput makes them useful for screening potential drug therapies. Strategies for modeling PD in zebrafish have included genetic alterations that target proteins associated with PD, including pink1 ^7–16, parkin ¹⁷, and others ^18–22. Another convenient strategy, which has been used in rodents primarily, is to chemically induce parkinsonism by treatment with a toxin either in the food or by injection²³. Toxins previously used in zebrafish for this purpose include 6-hydroxydopamine (6-OHDA) ^24–28, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) ^{8, 10, 14, 29–34}, paraquat ^35–46, and rotenone ^47–55.

Fast-scan cyclic voltammetry (FSCV) at carbon-fiber microelectrodes is an electrochemical technique that allows measurement of electroactive neurotransmitters and other signaling molecules with good temporal resolution, limits of detection, and selectivity. This technique has found broad use in rodents, especially mice ^56–58 and rats ^59–62. Recently, our group ^63–65 and others ^{66, 67} have applied FSCV or constant potential amperometry to measure dopamine release and uptake in zebrafish brain tissue. We have also measured the effects of carboplatin, a toxic chemotherapy agent, on dopamine release and uptake⁶³. These studies collectively indicate that zebrafish are a useful model of neurological disease.

Numerous behavioral paradigms have been employed to assess various aspects of locomotor and cognitive function of zebrafish (reviewed by Benvenutti et al. 2021⁶⁸). Latent learning involves the passive, unmotivated retention of information. Once this information is internalized, it may be used to complete a motivated task⁶⁹. A latent learning paradigm, first published by Gómez-Laplaza and Gerlai in 2010⁷⁰, employs a maze that consists of two pathways, left or right, that zebrafish passively learn in groups prior to testing. Each zebrafish then undergoes testing with both arms open as an assessment of how well they learned the correct path. This paradigm has found use in assessing the effects of alcohol exposure⁷¹, selenium exposure⁷², and modulation of dopamine receptor function on latent learning ability⁷³.

In this work, we have dosed zebrafish with rotenone, a mitochondrial complex I inhibitor ⁷⁴, to induce a syndrome that resembles PD. We then behaviorally analyzed treated and control fish with the latent learning paradigm to assess cognitive and locomotor impairments. We also measured dopamine release and uptake in treated and control zebrafish with FSCV. Our results reveal that our approach can resolve neurochemical and behavioral changes induced by rotenone treatment and could be useful for studying other disease models as well as screening potential therapies with enhanced throughput compared to rodents.

METHODS

Animals and treatment

All animal procedures were approved by the University of Kansas Institutional Animal Care and Use Committee. Wild-type zebrafish were bred and housed in the University of Kansas Synthetic Chemical biology core. Adult zebrafish were divided into three groups (n=12 per group). One group of fish was treated for 4 weeks with rotenone (Millipore Sigma, Burlington, MA) at a final concentration of 3 μg/L in their water. Treatment duration and rotenone concentration were based on previous studies ⁵⁴. This concentration was achieved by addition of 30 μL of a 1000 mg/L stock solution in dimethyl sulfoxide (DMSO) to 1 L of fish system water. The second group of fish was treated for 4 weeks after addition of 30 μL DMSO vehicle without rotenone to 1 L system water. The final group, naïve fish, received no chemical treatment. The water in the tanks was changed daily and fish were continuously exposed to rotenone and vehicle throughout the 4-week dosing period.

Chemicals

Dopamine hydrochloride was purchased from Sigma-Aldrich (St. Louis, MO). Sodium chloride (NaCl), potassium chloride (KCl), and potassium phosphate (monobasic) was purchased from Fisher Scientific (Waltham, MA). Sodium bicarbonate (NaHCO₃) and rotenone was purchased from Sigma-Aldrich (St. Louis, MO). Magnesium sulfate (anhydrous) was purchased from Amresco (Solon OH). Calcium chloride dihydrate and glucose (anhydrous) were purchased from VWR (Rachor, PA). HEPES (N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid) was purchased from IBI (Peosta, IA).

Electrochemical measurements

After the probe trial, dopamine release measurements were made in isolated fish brains ex vivo with fast-scan cyclic voltammetry. Zebrafish were euthanized by rapid chilling method followed by decapitation. Brains were harvested by dissecting the head under a stereoscope (Leica Microsystem, Bannockburn, IL). The harvested brain was transferred to a superfusion chamber in which they were kept viable by a continuous flow of artificial cerebrospinal fluid (aCSF) that had been oxygenated with a mixture of 95% oxygen and 5% carbon dioxide and was maintained at a temperature of 28 °C. The aCSF was modified for zebrafish, and consisted of 131 mM NaCl, 2 mM KCl, 1.25 mM KH₂PO₄, 20 mM NaHCO₃, 2 mM MgSO₄, 10 mM glucose, 2.5 mM CaCl₂·H₂O, and 10 mM HEPES dissolved in purified (18.2 MΩ) water. The pH was adjusted to 7.4.

A nylon mesh harp was used to immobilize the brain in the superfusion chamber, and the brains were allowed to equilibrate for 30 minutes before inserting electrodes and starting to collect dopamine release data.

We used previously published procedures to make cylindrical carbon-fiber microelectrodes. Briefly, single carbon fibers (7 μm diameter, Goodfellow Cambridge LTD, Huntingdon, UK) were aspirated into glass capillary tubes (1.2 mm D.D. and 0.68 mm LD, 4 in long; A-M Systems Inc, Carlsberg, WA) and pulled on a PE-22 heated coil puller (Narishige Int. USA, East Meadow NY). The exposed carbon fiber, which served as the active sensing region, was cut to a length of 100 μm from the pulled glass tip. Electrodes were sealed by dipping into epoxy resin (EPON resin 815C and EPIKURE 3234 curing agent, Miller-Stephenson, Danbury, CT) for 45 s, gently rinsing with acetone, and curing at 100 °C for 1 h. Electrodes were calibrated by flow injection analysis against standard dopamine solutions. Just prior to use, each electrode was soaked in isopropanol for 10 min. The electrode was then backfilled with 0.5 M potassium acetate for electrical connection and the exposed electroactive surface was electrochemically pretreated by applying the waveform −0.4 to +1.3 V to −0.4 V at an application frequency of 60 Hz for 15 min followed by 10 Hz for 10 min. The Ag/AgCl reference electrode was created by chloriding a silver wire. The electrochemical equipment used consisted of a Dagan Chem-Clamp potentiostat (Dagan, Minneapolis, MN), modified to allow decreased gain settings, a personal computer with Tar Heel CV software (provided by R.M. Wightman, University of North Carolina, Chapel Hill, NC), a breakout box, and two National Instruments computer interface cards (PCI 6052 and PCI 6711, National Instruments, Austin, TX). The waveform was applied every 100 ms to the electrode at a scan rate of 400 V/s. To evoke dopamine release in the whole brain, multiple biphasic stimulus pulses (25 pulses, 60 Hz, 2 ms, 350 μA) were applied locally to the dorsal nucleus of the ventral telencephalon (Vd)⁷⁵ with two tungsten electrodes positioned 200 μm apart. Brains were given a 10 min recovery time between electrical stimulations.

Behavioral testing

We used a behavior test apparatus similar to that described by Gómez-Laplaza & Gerlai⁷⁰. This latent learning maze contained water at a depth of 10 cm and was made of transparent Plexiglas to facilitate visibility of landmarks and conspecifics to zebrafish. It had two tunnels (left and right) connecting the start chamber and the reward chamber. During the training period, the fish received a 30 second acclimation period in the start chamber. At the end of the acclimation period fish were allowed to leave the start chamber and explore the maze without any stimulations or interferences. After leaving the start box, fish would enter a straight path led to an intersection that offered four choices: 1) swim back to the start chamber, 2) go straight through the middle tunnel, 3) turn left and swim through the left tunnel to the reward chamber, and 4) turn right and swim through the right tunnel to the reward chamber. During the training period, each fish group (rotenone-treated, vehicle-treated, and chemically naïve) were divided into two subgroups. One subgroup was trained while the right tunnel was open, but the middle and left tunnel were closed, and the other group was trained while the left tunnel was open, and the middle and right tunnel were closed. All the fish were trained for 15 days, 30 minutes per day while there were no conspecifics in the reward chamber. After this training period (Day 16), one 10-minute probe trial was conducted for each fish while 5-stimulus fish were in the reward chamber. During the probe trial, both left and right tunnels were open while the center tunnel was blocked. While obtaining behavioral measurements, fish movement was recorded using a Canon VIXIA HF R800 camcorder. EthoVision XT video tracking software (Noldus Info Tech., Wageningen, The Netherlands) was used to analyze the recordings.

Statistical analysis

Statistical analyses were performed using Graph Pad Prism 6 (Graph Pad Software, Inc, La Jolla, CA). All data are reported as mean plus or minus the standard error of the mean. For all analyses, the N value is equal to the number of zebrafish brains used. One way ANOVA with Dunnett’s multiple comparisons test (unless otherwise specified) was used to compare the group differences between control vs treated fish.

Modeling to determine dopamine reuptake rate constants was performed using GraphPad software. Briefly, the files were loaded into Tarheel software, and the peak current (i_max) was determined. The portion of the plot in which current decreases from 90% to 40% of i_max was modeled to determine the kinetic parameters. The x and y values of the 90% value and the y value for the 40% value were used as constraints in the nonlinear regression-plateau followed by one phase decay model. Data files were excluded from modeling if they did not have a peak current over 0.6 nA and if there were less than 3 points between the 90% and 40% values. Once the modeling was completed, the 1^st order rate constant (k), and the half-life (t_1/2), were obtained.

RESULTS AND DISCUSSION

Decreased Dopamine Release and Uptake in Rotenone-Treated Zebrafish.

Loss of dopaminergic neurons and associated dopamine depletion plays a crucial role in Parkinson’s disease⁷⁶. We used FSCV to measure dopamine release that was electrically stimulated in brains from rotenone-treated, DMSO vehicle-treated, and naïve zebrafish. Representative raw data reveal that dopamine release in rotenone-treated fish was greatly diminished compared to naïve and vehicle-treated control fish (Fig. 1). The color plots did not reveal the presence of other electroactive species released. Analysis of multiple fish indicates that the decrease in dopamine release found in rotenone-treated fish compared to controls is significant (Fig. 2A) (p < 0.0001, one-way ANOVA, Dunnett’s multiple comparisons test, n = 9). Furthermore, release did not differ between male and female fish (Fig. 2B). This result agrees with previous findings in PD model rodents generated either genetically⁷⁷ or chemically^78–81 in which dopamine release was impaired. However, this study is the first, to our knowledge, to examine subsecond dopamine release events in a zebrafish model of PD.

Figure 1. Sample raw FSCV data.

(A) Plots of stimulated release with cyclic voltammograms (inset) to confirm the presence of dopamine. Stimulation indicated by vertical dotted lines. (B) Color plots that indicate current in false color (legend on right). Applied potential (in V) is shown on the ordinate and time is shown on the abscissa. Cyclic voltammograms are sampled at the vertical dashed lines and current-time plots are sampled on the horizontal dashed lines.

Figure 2. Treatment with rotenone decreases dopamine release and uptake in zebrafish.

(A) Pooled data representing the dopamine release from rotenone-treated (3 μg/L), vehicle-treated and chemically naïve zebrafish brains. (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, ****p < 0.0001) (B) Pooled data representing dopamine release in male vs female fish in each treatment group. No significant difference was found between male and female. (n=4, two-way ANOVA). (C) Sample model of a dopamine reuptake curve. The current vs time graph is shown with black while the model is the blue line. This model uses the time point at approximately 90% of maximum current and ends at the time point at approximately 40% of maximum current. The model gave k, t_1/2, and R² values. (D) Pooled data representing the first order rate constant (k) of dopamine reuptake (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, *p < 0.05). (E) Pooled data representing the half-life values obtained by the reuptake modeling. (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, *p < 0.05 and *p<0.01)

Membrane-bound dopamine transporter (DAT) protein molecules regulate extracellular dopamine levels by actively taking up dopamine into the cell. DAT expression and dopamine uptake are impaired in rodents treated with rotenone and other PD models^82–84. Therefore, we used nonlinear regression in GraphPad prism to model the decrease in dopamine signal and determine first order rate constants for uptake (k) and the half-lives of decay (t_1/2)(Fig. 2C). Pooled results from the modeling indicate a diminished value of k in rotenone-treated fish compared to the vehicle and naïve zebrafish (Fig. 2D). Furthermore, rotenone-treated fish had a significantly greater half-life (t_1/2) compared to the vehicle and naïve zebrafish (Fig. 2E). These results collectively demonstrate that we can quantitatively measure neurochemical deficits with FSCV in this research model. We have also shown that this PD zebrafish model undergoes similar neurochemical impairments compared to other PD model organisms^85,86,87,88 These findings underscore the relevance of using zebrafish to evaluate potential therapies with higher throughput compared to rodents.

Motor Impairment in Rotenone-Treated Zebrafish

PD is primarily a movement disorder, with bradykinesia being a defining clinical symptom⁸⁹. To assess how rotenone influences motor function, we analyzed video files of the experimental fish performing a task in the latent learning paradigm. The representative heat maps in Fig. 3 depict the movement of a single zebrafish for 10 minutes during the probe trial. All three heat maps were created using zebrafish trained with the right tunnel of the maze open. Repeated analysis of multiple fish revealed that rotenone-treated fish tended to move slower, have decreased acceleration, and swim less distance than vehicle-treated and naïve control fish (p < 0.05, one-way ANOVA, Dunnett’s multiple comparisons test, N = 9) (Fig. 4). Previous studies have also analyzed the behavioral effects of rotenone. Open tank studies have revealed motor deficits in various models of chemically induced PD, including rotenone-treated fish^{26, 48, 54}. Our results are consistent with these findings, demonstrating that we can use this latent learning paradigm to quantitatively measure locomotor function in PD model zebrafish.

Figure 3. Representative heat maps of zebrafish movement in the maze during the probe trial.

A heatmap that shows where fish swam during the 10 min trial is overlaid on the maze. Locations where fish spent longer periods of time are denoted by longer wavelengths (e.g., red), while locations where fish spent shorter periods of time are denoted by shorter wavelengths (e.g., blue). During training fish had access to the right tunnel. (A) chemically naïve fish, (B) vehicle-treated fish, and (C) rotenone-treated fish.

Figure 4. Pooled data representing the locomotion of the fish during the probe trial.

(A) Total distance moved in the maze, (B) Mean swimming velocity (cm/s), and (C) Acceleration (cm/s²) (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, *p < 0.05).

Cognitive Impairment in Rotenone-Treated Zebrafish

The heat maps in Fig. 3 suggested that zebrafish, regardless of the treatment condition, spent a long period of time with their conspecifics. Moreover, these heatmaps indicated the side preference (left vs right tunnels) of the maze. Chemically naïve and vehicle-treated fish appeared to prefer the tunnel that remained unobstructed while they were trained; conversely, rotenone-treated fish spent time over the full area of the maze. We directly observed that the fish actively explored the maze without any obvious signs of anxiety, including stopping in place, moving erratically, or jumping out of the water.

In-depth analysis of the video data obtained from multiple fish revealed significant differences in performance. We counted fish that chose each tunnel during the evaluation trial, when left and right maze arms were unobstructed, and compared their choices against their training conditions. The initial path taken by the fish after leaving the start chamber was used as the measure to evaluate correct and incorrect actions. Most chemically naïve and vehicle-treated fish chose the correct tunnel while most rotenone-treated fish chose the incorrect tunnel (Fig. 5). This result indicates that rotenone treatment has affected the memory acquisition and consolidation in zebrafish.

Figure 5. Rotenone-treated zebrafish made more incorrect decisions than control zebrafish.

Naïve, fish untreated with vehicle or rotenone, Veh, fish treated with vehicle, Rot, fish treated with rotenone. N = 9 fish in each group.

Cognitive impairment is common among PD patients and about 78% of PD patients develop dementia⁹⁰. Previous attempts at characterizing learning and memory of PD patients have demonstrated deficits in executive control, visuospatial functioning, and memory retrieval⁹¹. As a crucial part of the limbic system, hippocampus is known to be associated with the spatial and other forms of learning as well as memory⁹². Even though zebrafish do not have a specific brain region resembling the mammalian hippocampus, they can acquire and remember the spatial information^93–95. The previous approaches on latent learning have confirmed that this paradigm can assess the memory acquisition of the zebrafish^{70, 71, 73}.

In a previous study, experimental zebrafish displayed appropriate side bias during the probe trial that corresponded to the condition under which they were trained (right or left arm open)⁷⁰. Studies with rodents have demonstrated that latent learning experiments are useful for evaluating memory and thus can be used in drug screening tests for dementia^{96, 97, 98, 99}. Furthermore, recently, a novel latent learning-based approach has been proposed as a diagnosis tool for dementia patients¹⁰⁰. Thus, the findings of this experiment agree with previous latent learning experiments carried out using other organisms and may be beneficial in developing zebrafish as a useful experimental model for enhanced throughput drug screening for PD.

Fig. 6A demonstrates the latency of the fish to swim out of the start box during the probe trial. This behavioral measure indicates and quantifies the level of motivation fish have to move throughout the maze versus freezing in place or lacking the ability to move⁷⁰. Most fish, irrespective of the treatment condition, left the start box and explored the maze. Also, latency to leave the start box did not differ between rotenone-treated, vehicle-treated, and naïve zebrafish (Fig. 6C). We also determined how much time fish spent in the start chamber. The experimental data demonstrated no significant difference between groups. Thus, we conclude that rotenone-treated fish do not suffer from increased anxiety compared to controls.

Figure 6. Pooled data representing the fish behavior during the probe trial.

(A) Latency to leave the start chamber (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, ns). (B) Latency to first enter the reward chamber. (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, p < 0.05). (c) Cumulative duration in the start chamber (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, ns). (D) cumulative duration in the reward chamber. (n=9, one-way ANOVA, Dunnett’s multiple comparisons test, p < 0.05 and p<0.001)

The latency to enter the reward chamber, which contains conspecifics, may reflect motivation of the fish to access and stay with the stimulus fish⁷⁰. This parameter may also reflect the familiarity of the experimental fish with the spatial layout of the maze. Fig. 6B shows that fish treated with rotenone took a significantly longer time to reach the reward chamber compared to controls (p < 0.05, one-way ANOVA, Dunnett’s multiple comparisons test, n = 9). Latencies did not significantly differ between the vehicle-treated and naïve fish. The increased latency occurring in rotenone-treated fish may indicate the presence of impairments in latent learning and spatial memory. It is possible that diminished capacity to release dopamine contributes to these impairments. Rotenone-treated fish also spent less time in the reward chamber (Fig. 6D), possibly reflecting a decreased tendency to shoal¹⁰¹.

Few studies have examined latent learning in zebrafish; none of these, to our knowledge, involved PD models. The latent learning paradigm employed in this study was published by Gómex-Laplaza and Gerlai (2010)⁷⁰ and has been used by the same group to examine the effects of alcohol administration cognitive performance⁷¹. Other recent studies have used this paradigm to show that dopamine receptors probably influence the processing of information in latent learning⁷³ and that degradation of the dopamine system is associated with cognitive impairment⁷². Our study demonstrates not only that rotenone, a chemical commonly used to induce parkinsonism, affects the locomotor behavior of zebrafish, but also that it impairs latent learning, assayed using the same paradigm. Taken with the neurochemical results, our work suggests that rotenone induced impairments in dopamine function negatively affect the ability of zebrafish to learn and locomote. This work is especially relevant given that a large percentage of PD patients suffer from cognitive impairments^{102, 103}.

CONCLUSION

In this study, we dosed zebrafish with rotenone and measured its effects on dopamine release and uptake, locomotor activity, and latent learning in zebrafish. This work is, to our knowledge, the first to carry out these neurochemical and behavioral analyses in PD model zebrafish. Rotenone treatment decreased dopamine release and uptake and impaired locomotion and learning. These results demonstrate that we can employ these techniques to resolve neurochemical and behavioral changes in PD model zebrafish and use this approach to evaluate potential therapies. Moving forward, it will be important to evaluate this approach in other zebrafish PD models, such as MPTP- and 6-OHDA-treated fish. We also expect that this approach will be useful in the analysis of zebrafish that model other motor and cognitive disorders and in the evaluation of therapeutic interventions with increased throughput over more complex models, such as rodents.