Dopaminergic modulation by quercetin: in silico and in vivo evidence using Caenorhabditis elegans as a model

Abstract

Quercetin is a flavonol widely distributed in plants and has various described biological functions. Several studies have reported on the ability to restore neuronal function in a wide variety of disease models, including animal models of neurodegenerative disorders such as Parkinson’s disease. Quercetin per se can act as a neuroprotector/neuromodulator, especially in diseases related to impaired dopaminergic neurotransmission. However, little is known about how quercetin interacts with the dopaminergic machinery. Here we employed the nematode Caenorhabditis elegans to study this putative interaction. After observing behavioral modulation, mutant analysis and gene expression in C. elegans upon exposure with quercetin at a concentration that does not protect against MPTP, we constructed a homology-based dopamine transporter protein model to conduct a docking study. This led to suggestive evidence on how quercetin may act as a dopaminergic modulator by interacting with C. elegans’ dopamine transporter and alter the nematode’s exploratory behavior. Consistent with this model, quercetin controls C. elegans behavior in a way dependent on the presence of both the dopamine transporter (dat-1), which is up-regulated upon quercetin exposure, and the dopamine receptor 2 (dop-2), which appears to be mandatory for dat-1 up-regulation. Our data propose an interaction with the dopaminergic machinery that may help to establish the effects of quercetin as a neuromodulator.

Graphical Abstract

Introduction

The incidence of neurodegenerative diseases is increasing worldwide. For instance, nearly 50 million people (0.62%) have Alzheimer’s disease (AD) [1], while 10 million people (0.12%) are affected by Parkinson’s disease (PD) [2]. Disturbance of dopaminergic signaling is a common feature associated with both diseases [3, 4], and dopamine-dependent (DA) signaling dysfunction in the brain is inherent of PD etiology.

The current pharmacologic management for PD is based mainly on three approaches: (1) dopamine replacement by levodopa administration; (2) dopaminergic agonists such as ropinirol and pramipexole; and (3) dopamine metabolism inhibition by selegiline and entacapone [5]. However, there are side effects associated with their use such as motor fluctuations, dyskinesia [6], hallucinations and sleep dysfunctions [7, 8]. Thus, new therapeutic options are necessary to attenuate and ameliorate neuronal loss and signaling.

A promising strategy to develop novel and safe pharmacological therapies is to look at natural products and single biomolecules derived from them. Their therapeutic potential acting on pathologies associated with central nervous system (CNS) dysfunction has been reported in the literature [9, 10], and their advantages include safety and low incidence of adverse events in preclinical [11] and clinical trials [12]. Amongst the mostly studied bioactive metabolites are the flavonoids [13]. Studies demonstrate that flavonoids intake reduce the risk of dementia, AD [14, 15] and PD [16, 17].

The flavonol quercetin is one of the most promising neuroprotective molecules in preclinical studies [18–22]. Current evidence demonstrate quercetin may be safely administered to humans [23] and results in beneficial clinical outcomes such as reduction of blood pressure, inhibition of LDL oxidation and inflammatory markers [24], reduction of lung cancer and asthma incidence [25] and decreased oxidative stress and inflammation [26]. Recently, a clinical trial using quercetin + dasatinib has been designed to investigate the safety and feasibility of using this combination as a senolytic therapy to mitigate Alzheimer’s disease (AD) progression (ClinicalTrials.gov Identifier: NCT04063124), reinforcing the promising potential of this flavonoid in neurodegenerative diseases.

Despite the potential for quercetin to act as a neuroprotector, no specific mechanism has been established for quercetin’s action in the CNS. There is evidence that quercetin crosses the blood-brain barrier (BBB) [27, 28], however the hypothesis that quercetin targets the DAergic machinery has been merely speculative [29–31]. Considering the structural similarity between the flavonol group and DA, it is noteworthy that quercetin presents a double ring that could increase affinities/interactions with aminoacids/groups of hydrophobic regions in the DA binding sites. Therefore, in silico docking analysis could be a preliminary strategy to verify this possibility [32].

Animal models could allow easy and fast screening of putative quercetin molecular and biochemical targets, which display biological functions. In this context, the nematode Caenorhabditis elegans is an ideal choice, since it has a completely characterized nervous system; the hermaphrodite worm has 302 neurons and eight of them are DAergic [33, 34]. The DAergic machinery is highly conserved from worms to mammals, from DA biosynthesis, receptors, uptake to catabolism [35] (Figure 1). In mammals, the loss of DAergic function causes motor impairment symptoms such as decrease of fine motor movements, constant tremors and reduced mobility in a progressive manner [36]. Notably, these are all characteristics of PD. Some locomotor functions controlled by DA can be monitored in C. elegans. The head thrash movement’s behavior, although is not specific by dopamine, are stimulated by the DAergic system also and are promoted by sensorial neurons located at the nose tip which control food seeking behavior [34, 37]. The typical sinusoidal movement pattern (body bends) is also dependent on DA response, i.e., under the presence of food, neurons release DA which slows down worm movements, whereas starved worms move at a faster rate, searching for food [38].

FIGURE 1. Dopamine metabolism.

TYR : Tyrosine. L-DOPA: Levodopa. AAAD: Aromatic L-Aminoacid Decarboxylase. VMAT: Vesicular Monoamine Transporter. DAT: Dopamine Transporter. MAO: Monoamine Oxidase. DOP-1,2,3: Dopamine receptor. MP: Methylphenidate.

Several studies have shown that quercetin treatment protects C. elegans from amyloid beta aggregation-induced dysfunction [39]; mitochondrial defects [40] and neuronal expression of human exon-1 huntingtin (128Q) [41]. In addition, quercetin has been described as a potential neuroprotective agent in models of PD, such as exposure to 6-hydroxydopamine (6-OHDA) and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [19, 42–44]. Biochemical, and most importantly, behavioral studies, confirmed quercetin’s efficacy as a neuromodulator molecule against MPTP and 6-OHDA, toxicants that are intrinsically linked to the DAergic machinery, thus pointing out to a role for quercetin as a DA signaling modulator.

In this work, we sought to investigate whether quercetin modulates the DAergic machinery of C. elegans. Based on the initial observations, at non-lethal acute exposure to quercetin at a concentration that does not protect against MPTP, increased worm head movements (Head Thrash, HT), caused swimming induced paralysis and reduced locomotor activity decline (slowing basal) in the absence of food. Therefore, we hypothesized that quercetin acts as a DAergic signaling modulator in C. elegans, by a mechanism that is dependent on the dopamine transporter (DAT-1) and on the D2-like dopamine receptor 2 (DOP-2). Furthermore, in silico evidence (DAT-1/quercetin docking) suggest that quercetin may have higher affinity to the C. elegans DAT-1 homology-based model protein when compared to DA, exerting a competitive agonist-like effect and maintaining higher DA levels in the synaptic cleft.

2. Materials and Methods

2.1. Chemicals

Quercetin (Q), dopamine (DA), methylphenidate hydrochloride (MP), 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), and were obtained from Sigma-Merck. TaqMan probes were acquired from Thermo-Fisher.

2.2. C. elegans strains and handling of the worms

C. elegans Bristol N2 (wild type), MT7988 [bas-1 (ad446) III]; CB1112 [cat-2 (e1112) II]; RM2702 [dat-1 (ok157)]; LX703 [dop-3 (vs106) X]; LX702 [dop-2 (vs105)]; LX645 [dop-1 (vs100)] and BY200 (dat-1p∷GFP) worms were handled and maintained at 20 °C on NGM (nematode growth medium)/Escherichia coli OP50 Petri dishes. All strains were provided by the Caenorhabditis Genetics Center (University of Minnesota, Twin Cities, MN, USA). Synchronous L1 populations were obtained by isolating embryos from gravid hermaphrodites using bleaching solution (1% NaOCl; 0.25 M NaOH), according to standard procedures previously described [45]. All experiments were conducted at 20±1°C in a humidified-controlled environment. Double mutants were generated by crossing dat1(ok157)III males with dop-2 (vs105) hermaphrodites and mutant lines were confirmed using worm PCR. No ethical approval was needed for the study with invertebrate model.

2.3. Protocol of exposure

1,500 synchronized L1 or L4 worms were exposed to quercetin (200–2000μM); dopamine (500–10,000 μM), methylphenidate (10,000 μM) or MPTP (10μM) for one hour in M9 (0.3% KH₂PO₄, 0.6% Na₂HPO₄, 0.5% NaCl) buffer under constant agitation (Figure 2). Concentration of MPTP was established based on a previous survival assay (Figure S1) in a liquid medium exposure method, because in the literature, higher concentrations (400 μM- 1mM) were used, however, using exposures in plates [46, 47]. The administered chemicals were removed by subsequent washing with 85 mM NaCl solution. Worms exposed to quercetin and MPTP were placed on E. coli OP50-seeded NGM plates and the number of surviving worms on each plate was scored 24h later. Behavior and gene expression were measured 2h or 6h after exposure to quercetin, using L1 or L4 worms, detailed in experimental design (Figure 2).

FIGURE 2. Experimental design.

In order from left to right: Quercetin (Q), Dopamine (D), MPTP, and Methylphenidate (MP). Worms have been exposed to the drugs in liquid media inside microtubes at either L1 or L4 for 1h. Survival has been scored 24h after exposure. Head thrashing, ROS, and mRNA expression have been conducted immediately after L1 exposure. Food sensing behavior, dopaminergic degeneration, and mRNA expression have been conducted on L4 animals 2h or 6h after exposure to the drugs. Swimming paralysis (alongside mRNA levels, L4 worms) has been conducted after 50min of exposure to quercetin (30min in OP50, 20min swimming in liquid media).

2.4. Abnormal Neuronal Morphology

To evaluate dopaminergic neurons, we used transcriptional GFP reporter driven by dat-1 (dopamine transporter 1) promoter (dat-1p∷GFP), resulting in the expression of GFP in dopaminergic neurons. We analyzed nematodes at the L4 stage exposed as described in 2.3 and illustrated Figure 2, and a scoring the lesioned neurons was performed as described in Soares et al. [48]. At least 15 animals per group were scored and 3 independent experiments were conducted for each condition.

2.5. Head Thrashes (HT)

We considered head thrashes as the high angled movement described by Shimomura et al. [49] and Hills et al. [50], with minor modifications. One thrash was counted each time the tip of the head of the animals was able to achieve its maximal or minimal angle in relation to the posterior part of the pharynx. To score L1 worms head thrashes, we allowed them to recover for 2 h on NGM/OP50 plates after quercetin, dopamine or MPTP exposure. After that, they were transferred individually to a NGM/E.coli-free plate. This food-free environment mimics the approach by Hills et al. [50]. After a 1-min recovery period, the head thrashes were counted for 1 min. At least 20 animals per genotype and treatment conditions were scored. At least 5 independent experiments were conducted for each condition. For the experiments, nematodes were randomly chosen, and as selection criteria, animals that responded to the touch of a pencil were selected.

2.6. Reactive species levels

We have measured the amount of oxidant chemical entities inside the worms through the H₂DCF-DA fluorophore, which upon cleavage of the acetate groups by intracellular esterase and oxidation by reactive species is converted to the fluorescent 2’,7’-dichlorofluorescein (DCF). The L1 treated worms were transferred to a microplate after the treatments and H₂DCF-DA (50 μM) was added in the buffered medium. The fluorescence levels were measured for 30 min each 10 min in a plate reader CHAMELEON^™V Hidex Model 425–106. The mean values were expressed as percentage of the control group.

2.7. Food sensing behavior

Slowing basal was performed as described in literature, with some modifications [51]. 6–10 L4 stage worms that’s had just been exposed to quercetin for 1h per group were washed out of bacterial food source with M9 buffer for 5min. A recover period in the presence of food was allowed for 2h or 6h, as indicated in the respective figures. Those animals were then placed either in plates with a standard bacterial lawn (E. coli OP50, seeded on the previous day and stored at 4°) or plates without food. After 2 min, for acclimation, the number of body bends was counted for 20 seconds. One body bend was counted each time the middle part of the body reached its maximum or minimal angle in relation to the posterior pharynx. Mutants used in this assay were: N2; bas-1 (ad446) III; cat-2 (e1112) II; dat-1 (ok157); dop-2 (vs105) and double mutant dat-1 (ok157)/dop-2 (vs105). At least 3 independent experiments were conducted for each condition. For the experiments, nematodes were randomly chosen, and as selection criteria, animals that responded to the touch of a pencil were selected.

2.8. Swimming- induced paralysis (SWIP)

L4 animals have been exposed to quercetin/methylphenidate/dopamine (400μM or 2mM) for one hour. Afterwards, quercetin was removed with M9 buffer, the animals were left to rest in an OP50 seeded NGM plate for 30min. The animals were removed from the plates by applying a small drop of M9 buffer over the plate with a micropipette. Groups of 4 animals each were separated in 200μL microtubes with 100μL of basal media, at the point in which a chronometer was set to synchronize the time elapsed in between each group, so that the scoring of paralysis could take place at 20min timepoint for all groups. A worm has been considered paralyzed when at the bottom of the tube and no visible body movement was shown for more than 5s. 3 tubes with 4 worms each have been scored for each genotype and quercetin concentration, resulting in one biological replicate. This procedure has been repeated 4x in different days.

2.9. Gene expression

Relative levels of gene expression following quercetin exposure were measured using TaqMan gene expression assay probes (Life Technologies ^®). For this purpose, we used Trizol (Life Technologies ^®) in order to extract total RNA following compounds exposure in L1 or L4 worms, extraction was conducted immediately in L1 worms exposed and after 2 hours in L4 exposed worms. 1μg of total RNA was used to synthesize cDNA using High Capacity cDNA Reverse Transcription Kit (Life Technologies ^®). Quantitative reverse-transcription PCR (BioRad ^® CFX96) was conducted in triplicates using afd-1 (β-actin homolog) gene as housekeeping. We determined the fold difference using the comparative 2^ΔΔCt method. The following probes were used: dat-1 (Ce02450900_g1), dop-1 (Ce02494345_m1), dop-2 (Ce02479824_m1), dop-3 (Ce02496463_m1), and afd-1 (Ce02414573_m1).

2.10. Molecular modeling

Molecular modeling studies were performed with quercetin, methylphenidate and dopamine aiming to support the understanding of their activity relative to the C. elegans dopaminergic machinery. Methylphenidate was chosen as a positive control (inhibitor) to perform a comparison to quercetin prediction as an antagonist. The geometry of quercetin, methylphenidate and dopamine was fully optimized using Merck Molecular Force Field (MMFF94) and Density Functional Theory (DFT) B3LYP/6-311G* basis in gas phase methodologies using Spartan’08 for Windows (Wavefunction Inc., Irvine, USA) software, at least 15 scoring poses were analyzed from each one of 24 runs (total of 360 clusters), and we selected the best score pose to be anchored in the pocket site (Table S4). This local minimal energy structure was submitted to systematic conformational analysis using DFT with torsion angle increment, which was set at 30° with a range from 0 to 360°. The lowest energy conformer was saved as a .pdb file in order to perform docking studies.

The 3D structure of dopamine transporter 1, DAT-1 (PDB ID: 4XP6, Organism Drosophila melanogaster and Expression system Homo sapiens), which represents (+)-methamphetamine ligand complexed at its active site, was downloaded from Protein Data Bank (PDB). The 3D structure of C. elegans DAT-1 (Figure S2A) was obtained from the Robetta server in a homology-based approach. To construct the homology DAT-1 model, we choose Drosophila melanogaster based in some studies that demonstrated a great similarity between fly’s and worm’s development, cells, tissues, transcriptional similarities, proteomes and orthologous proteins in relation to humans [52–54].

2.11. C. elegans DAT-1 protein structure prediction

The prediction of DAT-1 C. elegans secondary protein structure was performed in five independent servers, Robetta [55], SWISS-MODEL [56], Phyre2 [57], RaptorX [58] and I-TASSER [59] according to each configurations. The models obtained from these five different servers were validated as previously described, using the Ramachandran diagram (Table S2) and Verify3D program (Table S3). The best model, Robetta, showed 98.4% of residues in favorable regions in the Ramachandran diagram and greater than 86.34% of coherence in the Verify3D, indicative of excellent quality. The final model (Figure S2A) was deposited in the Protein Model Database (PMDB), access code PM0081131. Figures were generated using PyMOL (https://www.pymol.org/).

2.12. Quercetin and dopamine docking studies with C. elegans DAT-1 homology model

Homology model protein were prepared by removing water molecules and adding polar hydrogens using Autodock Tools 1.5.6 [60] and the docking studies were performed using two softwares: iGemdock [61] and Docking Server (Bikadi, Hazai, 2009). The docking was run at drug screening Docking Accuracy Setting with GA parameters set for population size, generation and number of solutions as 200, 70 and 3, respectively, Gemdock score function of hydrophobic and electrostatic (1:1 preference), and binding site radius of 3.0 to 8.0 Å using by bound ligand option. iGemdock software was used to infer if the pharmacological interactions and results obtained allow for combining the pharmacological interactions and energy-based scoring function of iGemdock.

2.13. Multiple alignment generation

Protein sequences were uploaded from the Uniprot website (http://http://www.uniprot.org). Multiple alignments were generated using Clustal [62] and the alignment editing with Jalview [63]. The alignment was generated (Figure S2B) to compare the binding sites for quercetin between the two models and reinforce the predictive value of our homology-based C. elegans DAT-1.

2.14. Statistical Analysis

Statistical analyses were performed using the GraphPad Prism 6.0 software. All data are presented as mean ± SE. Assessment of the normality of the data was conducted using Shapiro–Wilk test. For survival and gene expression assay we employed an ordinary One-Way ANOVA, followed by Holm-Sidak multiple comparisons test. For behavioral assays we employed a One-way ANOVA, also followed by a Holm-Sidak multiple comparisons test, unless different genotypes and conditions were involved at the same time. In that case a Two-way ANOVA, followed by a Sidak post hoc. A p value smaller than 0.05 is indicated by one *, p<0.01 by **, p<0.001 by ***, and p<0.0001 is indicated by ****.

3. Results and Discussion

Numerous studies have corroborated quercetin’s antioxidant properties; however, only recently it has been demonstrated that flavonoids, quercetin included, may interact with multiple molecular targets to elicit a variety of cellular effects [64, 65]. Notably, quercetin is known by its neuroprotective effects [66, 67]. Then, this was confirmed against the DAergic neurotoxin MPTP and to eliminate a possible quercetin effect on MPTP-induced mortality, a survival assay was concomitantly conducted, showing that the treatment with quercetin after pre-exposure to MPTP did not reduce lethality induced by the neurotoxin (Figure 3A). Of note, we did not observe mortality (Figure 3A) or delayed development after quercetin exposure (data not shown). However, quercetin ameliorated the reduction in head thrashes induced by MPTP (Figure 3B). MPTP damages DAergic neurons by causing mitochondrial dysfunction and ROS generation, thus leading to neuronal death, as already demonstrated in previous studies with both worms and mice [47, 68].

FIGURE 3. Quercetin prevents behavioral changes induced by MPTP.

L1 Worms exposed for 1 hour (Basal Media) to 400μM of quercetin followed by 10μM of MPTP also for 1 hour. (A) survival rate (B) HT score; (C) Oxidative species measured by H₂DCF-DA; (D) percentage of worms with abnormal morphology in dopamine neurons and; (E) L4 worms previously exposed to 400μM of quercetin followed by 10μM of MPTP also for 1 hour are shown in representative images of dopaminergic neurons (strain BY200); red arrows indicate blebbing in the axons/dendrites of the nematodes. In (A), (B), (C) and (D) * and ** indicates p<0.05 and p<0.01, respectively, when comparing the indicated groups.

Quercetin treatment did not rescue the oxidative stress induced by MPTP, as measured by the amount of oxidant chemical species (shown in Figure 3C). This suggests that the mechanism by which quercetin improves C. elegans behavior after MPTP exposure, under our experimental conditions, does not rely on its classic antioxidant effect. To corroborate this hypothesis, when dopaminergic neurons were analyzed, it was observed that quercetin (400μM and 1mM) did not attenuate MPTP-mediated dopaminergic degeneration in the nematodes (Figures 3D and 3E and Figure S4). Accelerated aging induced by MPTP has been attributed to, among other factors, a disrupted DAergic system, therefore these observations might point towards the same direction as previous findings about the potential role of flavonols to modulate the DAergic machinery [69, 70].

Based on these results, we hypothesized that quercetin (400μM or 1mM) treatment did not rescue neuronal damage from MPTP toxicity itself, but did improve the behavioral functions of DA by promoting its signaling in neurons that remained functional. On other hand, only the higher concentrations (2mM and 3 mM) showed a neuroprotective potential against MPTP exposure (Figure S4), and studies reported that its antioxidant activities are concentration-dependent [66, 71, 72]. Furthermore, in other models (in vitro and in vivo), the neuroprotective effects mediated by quercetin are observed at higher concentrations, long-term exposures or both [67, 73, 74]. In addition, it was demonstrated that protection against toxins exposures like LPS [67, 75], rotenone [76] and MPTP [42, 43] is only observed after a long-term intake of quercetin.

Therefore, perhaps quercetin acts in two pathways: at lower concentrations (400 μM and 1 mM) acts as a neuromodulator, and at higher concentrations (2 mM and 3 mM) as a classic antioxidant. Currently, the strategies to treat patients with PD are based on the attenuation of the locomotor symptoms and not by rescuing neurons, but instead by improving the signaling in a neuromodulation strategy [77, 78]. Initially, we tested the hypothesis that quercetin could act as a DA mimetic. Indeed, we observed that in the absence of food, quercetin increased the number of head thrashes in C. elegans (Figure 4A) in a similar manner as exogenous DA (Figure 4B).

FIGURE 4. Quercetin modulates head thrash movements.

Head thrash score for L1 worms exposed for 1 hour (Basal Media) to different concentrations of A) Quercetin (Q) and B) Dopamine hydrochloride (D); C) Head thrashes score in L1 mutant worms exposed for 1 hour (Basal Media) to 400μM of quercetin. In (A) and (B) * and ** indicate p<0.05 and p<0.01, respectively, compared to the control group. In (C) *, **, ***, and **** indicate p<0.05, p<0.01, p<0.001, and p<0.0001 differences when compared to its own quercetin-free group or difference in between the indicated lines; # Indicates difference when compared to WT quercetin-free group.

In C. elegans, DA controls numerous functions, including locomotion [34, 79]. The DAergic system is comprised of eight neurons in the hermaphrodite (4 CEP; 2 ADE and 2 PDE). DA is synthetized from tyrosine by a reaction catalyzed by tyrosine hydroxylase (encoded by C. elegans cat-2), followed by a second step of catalysis by aromatic amino acid decarboxylase (encoded by bas-1). The release of DA results in its binding to specific DAergic receptors: DOP-1, DOP-2, DOP-3 and DOP-4. While dop-1 encodes a D1-like receptor (G-protein-coupled DA receptor type 1), dop-2 and dop-3 encode their antagonistic counterpart D2-like receptors (dop-4 is much less characterized) [80, 81]. These concepts might be extended to DOP-2 regarding its D2-like classification with respect to pharmacological responsiveness to DA. Additionally, free extra synaptic DA undergoes uptake by DAT-1 or is degraded by monoamine oxidases (MAO) or by catechol-O-methyl transferase (COMT) (Figure 1).

To better understand the DAergic modulatory effects of quercetin, we exposed various DAergic system related mutants to quercetin, including cat-2 (tyrosine hydroxylase), dat-1 (dopamine transporter), dop-1 (which is a D1-like receptor), dop-2, and dop-3 (which are D2-like receptors) mutants. We observed that the increase in head thrashes caused by quercetin was dependent on dat-1 and dop-2, since these mutants did not respond to the flavonoid as WT animals do (Figure 4C), thus corroborating our hypothesis that quercetin could modulated the DAergic system. Furthermore, we observed that in quercetin-free conditions, cat-2 mutants (which have drastically diminished DA levels) had a diminished number of head thrashes, while dat-1 mutants (which have increased dopamine levels) showed the opposite behavioral pattern (Figure 4C). To further access at which level quercetin could regulate the DAergic system in C. elegans, we measured the mRNA of key genes involved in the synthesis and signaling of dopamine, such as DOP-1, DOP-2, DOP-3 and DOP-4 (dopamine receptor), tyrosine 3-monooxygenase (TH) and DAT-1 (dopamine transporter) in L1 and L4 worms after exposure. Our data shows (Figure 5) that in both stages dat-1 was upregulated, and to our surprise, when measuring its expression in different mutant backgrounds (data not shown), dop-2 deletion prevented dat-1 overexpression, pointing towards a possible expression regulation of dat-1 by dop-2. Since dop-2 is an auto-receptor, which regulates itself and dopamine levels upon dopamine exposure [82], it would not be surprising if that could be achieved by regulating dat-1 expression.

FIGURE 5. mRNA levels of key dopamine metabolism genes after quercetin exposure.

A) mRNA levels for different genes involved in the dopaminergic machinery in L1 C. elegans worms, and B) L4 worms. mRNA levels are expressed as fold changes relative to the quercetin-free group, which is given a value equal to 1 and is represented by the dotted line. In letters (A) and (B) n.s. indicates non-significant difference and **** indicates p<0.0001.

To better clarify this role of quercetin in DAergic function, we tested a food behavior that is more specific to assess the DAergic signaling, the basal slowing. This behavior is mediated mainly by the DAergic system through signaling the presence or absence of food in the environment and modulating the locomotor activity [51]. In the first moment, 2 hours after exposure, quercetin did change food perception (Figure 6A), slowing down worms at a higher degree than the bacterial lawn; However, 6 hours after exposure, worms decreased the number of body bends executed not only in the presence of food, but also in the absence of food (Figure 6B). Since further slowing down was induced by increasing concentrations of quercetin, even in the absence of food, a behavior that was not expected since we expect quercetin to increase head thrashes off-food (Figure 4), we decided to verify if that was indeed dopamine related. Figure 6C displays dat-1 mutants, which is known for accumulating dopamine in the synaptic cleft [83], after 6h and 2h of quercetin exposure.

FIGURE 6. Food sensing behavior is modulated by quercetin.

Basal slowing score in L4 worms exposed to 400 μM, 1mM or 2mM of quercetin, 10mM of dopamine, or 10mM of methylphenidate. A) Basal slowing of WT worms 2h after exposure; B) Basal slowing of WT worms 6h after exposure; C) Basal slowing of dat-1 (ok157) after 2h and 6h from exposure to quercetin 2mM; and D) Basal slowing of WT worms exposed to 10mM of methylphenidate or dopamine. Symbols (−) and (+) indicate on food and off food conditions, respectively. A two-way ANOVA with Tukey post hoc has been used to analyze the effect of both treatment and food across the different time points and genotypes. Symbols (−) and (+) indicate on food and off food, respectively. In (A) *, ***, and **** indicate p<0.05, p<0.001, p<0.0001 and n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines; In (B) *, **, ***, and **** indicate p<0.05, p<0.01, p<0.001, and p<0.0001. n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines; In (C) * and ** indicate p<0.05 and p<0.01, and n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines; In letter (D) ***, and **** indicate p<0.001, and p<0.0001, and n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines.

Similarly to our previous observation regarding a food-dependent movement at the 6h timepoint in comparison to 2h, off-food movement is also decreased at 6h (Figure 6C). Additionally, quercetin does not further enhance slowing off-food in these animals, suggesting that quercetin and dopamine act in similar non-additive manner. Finally, we tested if exogenous dopamine and methylphenidate, a dopamine reuptake inhibitor [84] would exert the same effect when comparing times. Indeed, Figure 6D shows that dopamine and methylphenidate further decrease movement when comparing to bacteria alone, which has been previously shown, but also induce slowing off food after six hours, and slight bending increase off-food at 2h. Our experimental setup returns the animals to OP50 seeded plates after quercetin/dopamine/methylphenidate exposure for 2 or 6h. Animals starved for 30min further slow their movement when in contact with a bacterial lawn [85]. We believe further DAergic activation caused by quercetin and the other drugs might overstimulate C. elegans after 6h, compromising their ability to move at the same pace as just after exposure off-food.

However, it was still unclear whether this was due to a DA mimetic or modulatory effect. To better understand this mechanism, we used cat-2 and bas-1 mutants (knockouts of genes needed for DA biosynthesis) and evaluated their movement after 2h, the timepoint at which we observed consistent movement slowing only on-food. Figure 7A and 7B shows that quercetin did not further decrease body bends on food, neither increased them off-food. Therefore, quercetin cannot have the same effect in the absence of endogenous dopamine, ruling out the possibility of quercetin directly mimicking dopamine, but more likely being a DA levels modulator (Figure 7A and 7B). As the head thrash assay had indicated that quercetin might act in a dat-1 and dop-2 dependent manner, here we also observed no increased movement off-food and a loss of slowing in dat-1 mutants exposed to quercetin (Figure 7C).

FIGURE 7. dop-2 and dat-1 are required for quercetin behavioral effects.

Basal slowing score after 2h for L4 mutant worms exposed to 2mM of quercetin for 1 hour (liquid medium) (A) cat-2 (e1112); (B) bas-1 (ad446); (C) dat-1 (ok157); (D) dop-2 (vs105); (E) dat-1/dop-2 and (F) Swimming-induced paralysis was evaluated in L4 worms after 50min (30min in OP50, 20min swimming in liquid media) of regular 1h exposure to quercetin 400uM or 2mM, in WT, dop-2 (vs105) and dat-1 (ok157). Symbols (−) and (+) indicate on food and off food, respectively. A two-way ANOVA with Tukey post hoc has been used to analyze the effect of both treatment and food across the different time points and genotypes. In (A) and (B) n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines; In (C) and (D) **** indicate p<0.0001 and n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines; In (E) and (F) ** and **** indicate p<0.01 and p<0.0001, and n.s. indicates no statistical difference when compared to its adjacent group or difference in between the indicated lines Symbols indicate significant differences between (−) and (+) groups.

Even though there is loss of the normal further slowing on-food effect of quercetin, the loss of normal basal slowing points towards the possibility of another player interacting with quercetin, eliciting the opposite effect of dopamine, or altering food perception upon dat-1 deletion and quercetin exposure. We also and no further slowing or increased movement in dop-2 mutants (Figure 7D). This again points towards a role of these two genes in the effect of quercetin over C. elegans movement. However, a double mutant dat-1/dop-2 was generated (Figure 7E) and: 1) quercetin could not induce off-food movement, but 2) further slowdown was observed. On one hand this indicates that off-food movement induced by quercetin it is dependent on both dat-1 and dop-2. On the other hand, slowing on food induced by quercetin might be preferentially, but not exclusively regulated by dat-1 and dop-2. DOP-3, a D2-like receptor, such as DOP-2, which is required for the effect of different drugs over the DAergic system in C. elegans [86–88], shows a slight trend to be upregulated in L1 worms after quercetin exposure, and could be a potential additional target of quercetin.

The D2 receptors, which include C. elegans DOP-2, are common targets of anti-psychotic drugs [89, 90]. It is well known that DA levels are controlled by the combination of DAT-1 transporter and DOP-2 auto-receptor [91, 92]. In mammals, DAT-1 expression can be regulated by D2 receptors [93], since upon DA binding there is an increased expression of DAT-1 in the cell surface to reuptake the excess of DA in the synaptic cleft. Notably, drugs that are well-known for requiring DA influx/efflux through its transporters (amphetamine, 6-OHDA, reserpine) were already described as requiring C. elegans DAT-1 as well [94–96]. We have further tested the importance of these proteins by testing a highly specific DAergic behavior, the swimming-induced paralysis (SWIP) (Figure 7F), which is dependent on the release of dopamine and its binding to D2-like receptor DOP-3 [82]. Indeed we observed that dat-1 mutants have a increased SWIP per se, and dop-2 mutants are unable to display SWIP, in the same way its related DOP-3 mutant has been shown to [82]. Increasing doses of quercetin promoted SWIP, something that could not be replicated in dat-1 and dop-2 genetic backgrounds, further reinforcing their importance in the phenotypes elicited by quercetin.

Notably, quercetin and dopamine display similar chemical groups in their structure, as shown in Figure 8A, especially a conserved catechol group (red circle) that can participate in H-bond interactions. With that in mind, we tested whether quercetin might directly interact with DAT-1. We conducted docking simulations in a C. elegans DAT-1 homology model protein constructed from a previously described DAT protein derived from Drosophila melanogaster [97]. To obtain the protein model, we used Robetta server and validated the model with a Ramachandran diagram. Using Verify3D, we have found the sequence alignment, which allowed us to verify that the same putative interactions between the established Drosophila melanogaster DAT are similar to those in the predicted worm protein. In addition, we conducted comparative simulation between DA, quercetin and methylphenidate affinities to DAT-1. Tables S2 and S3 show that the C. elegans model has 98.4% of residues in favorable regions in the Ramachandran diagram and more than 86.34% of coherence in Verify3D, which is in agreement with our hypothesis that quercetin interaction with C. elegans DAT-1 is indeed occurring at its active site and not at any other region.

FIGURE 8. Structural analogy representation and docking analysis demonstrating that quercetin promotes DAT-1 inhibition and blocks dopamine transport.

Comparison between quercetin and dopamine structures (A). Circles indicate conserved activity-structure relationship and the catechol group (in red) present in the two compounds. Interaction of dopamine (C); quercetin (D); methylphenidate (E) and a merge of ligands (yellow: quercetin, blue: methylphenidate and red: dopamine (F); Free energy calculated to Dopamine, Quercetin and Methylphenidate (G); binding in model at the active site of C. elegans DAT-1 homology model (B). The analyzes were obtained using two softwares to represent binding site (images) and the best cluster was used; more details of clusters obtained in analysis and ΔG_bind for dopamine, quercetin and methylphenidate with the DAT-1 homology model are demonstrated in Supplementary Figure 3, Table S1 and Table S4.

The docking approach aimed to identify the conformational state of minimal energy of quercetin in comparison to DA when bound to DAT-1 active site. Results are shown in Table S1 (all amino acids involved in the interactions) and the active site interacts between dopamine, quercetin and methylphenidate in the C. elegans DAT-1 homology model are shown in Figure 8B–E. In addition, an overlap of these three ligands is shown in Figure 8F, reinforcing the shared binding site. Data obtained from our docking tests indicate that quercetin interacted at the active site of the homology model DAT-1 and shows that the most important amino acids for DAT-1 are PHE313; SER314; PHE319; VAL133 and ASP69, and putative interacts are conserved between dopamine and quercetin. In order to confirm the interactions visualized in our protein model, we performed docking analysis with other software and the results were similar (Figure S3A, dopamine and S3B, quercetin).

The computational analysis predicted that quercetin would have a higher binding affinity with DAT-1 than DA itself, as demonstrated by the free energy calculations and interactions between ligands with aminoacids of active site in DAT-1 (Table S1 and Table S4). The Free energy estimate for quercetin was – 8.996 and for dopamine is – 6.829 (Figure 8G), characterizing a potential antagonist effect. Based on the literature, to measure protein/ligand binding affinities, the Free energy calculations (Gibbs energy of binding, ΔG) provide the most accurate estimate to predict binding affinity. Considering that a favorable reaction has negative values, the greater the negative free energy, the tighter the binding and the affinity [98–100]. Importantly, DAT blockage has been considered an innovative approach in pharmacological strategies aiming to treat DAergic dysfunction [101].

Finally, it is important to reiterate that since our model protein is based on a DAT crystallographic model that is bound to a methamphetamine molecule, the interactions observed herein may resemble those of that psychoactive molecule. Methamphetamine is known for competing with DA for its transporters, and also for causing the so called “reverse DA transporter” effect [102]. If DA is preserved in the synaptic cleft by the effect of quercetin competitively preventing its binding to the DAT-1 transporter, it would be reasonable to suggest that this transporter undergoes overexpression as a negative feedback modulation to counteract the high DA levels. If that is the case, DA could be competing with quercetin for the same transporter, therefore increasing free DA at the synaptic cleft and increasing DA-related movements. This is in accordance with our in-silico findings, where quercetin demonstrated a higher affinity by DAT-1 in relation to DA and a DAergic-like behavior in vivo. However, more studies remain to be conducted in order to shed light over the molecular mechanisms regarding DAT-1 and quercetin interactions, as well as DOP-2 function in this context.

4. Conclusion

Herein, we described how quercetin may target C. elegans nervous system upon a single acute exposure, more specifically modulating the DAergic machinery as well as muscular cells that respond to quercetin to control body movements known as head thrashes (HT), basal slowing and swimming paralysis. We posit that quercetin may modulate movement in C. elegans by interacting with the DAergic system, more specifically, but not exclusively, with the DAT-1 transporter and DOP-2 auto-receptor. Such a conclusion can be suggested because there was a modification in C. elegans behavior, which was dependent on the DA transporter (DAT-1) and D2-like DA receptor (DOP-2). This was accompanied by up-regulation of dat-1 mRNA in a possible negative feedback loop dependent on DOP-2. This possibility was further reinforced by the fact that quercetin interacted with a C. elegans homology-based model of DAT-1, which we built from a D. melanogaster crystallographic model, suggesting that quercetin may interact and be recognized by that protein with higher affinity than DA. Finally, our data provide new evidence that quercetin may be a potential dopaminergic selective neuroactive molecule. Overall, our study contributes to elucidating how quercetin may act to counteract the symptoms of dopaminergic-associated diseases such as PD.

Highlights.

Quercetin increases head thrashes, swimming paralysis, and further slowing down in bacteria;

Dopamine transporter (DAT-1) and dopamine receptor (DOP-2) are required for these phenotypes;

dat-1 mRNA overexpression is dop-2 dependent;

Quercetin interacts with C. elegans DAT-1 homology based protein model in silico;

Quercetin might act by competing with dopamine for dopamine transporter;

Abbreviations:

6-OHDA

6-hydroxydopamine

DA

Dopamine

CAT-2

C. elegans tyrosine hydroxylase

DAergic

Dopaminergic system

DAT-1

Dopamine transporter 1

DOP-1

Dopamine receptor 1

DOP-2

Dopamine receptor 2

DOP-3

Dopamine receptor 3

HAT

High-angled turn

HT

Head thrash

OP50

Escherichia coli uracyl auxotrophic

PD

Parkinson’s disease

TH

Mammal tyrosine hydroxylase

WT

Wild type