Abstract
Recent work indicates a role for RING finger protein 11 (RNF11) in Parkinson disease (PD) pathology, which involves the loss of dopaminergic neurons. However, the role of RNF11 in regulating dopamine neurotransmission has not been studied. In this work, we tested the effect of RNF11 RNAi knockdown or overexpression on stimulated dopamine release in the larval Drosophila central nervous system. Dopamine release was stimulated using optogenetics and monitored in real-time using fast-scan cyclic voltammetry at an electrode implanted in an isolated ventral nerve cord. RNF11 knockdown doubled dopamine release, but there was no decrease in dopamine from RNF11 overexpression. RNF11 knockdown did not significantly increase stimulated serotonin or octopamine release, indicating the effect is dopamine specific. Dopamine clearance was also changed, as RNF11 RNAi flies had a higher Vmax and RNF11 overexpressing flies had a lower Vmax than control flies. RNF11 RNAi flies had increased mRNA levels of dopamine transporter (DAT) in RNF11, confirming changes in DAT. In RNF11 RNAi flies, release was maintained better for stimulations repeated at short intervals, indicating increases in the recycled releasable pool of dopamine. Nisoxetine, a DAT inhibitor, and flupenthixol, a D2 antagonist, did not affect RNF11 RNAi or overexpressing flies differently than control. Thus, RNF11 knockdown causes early changes in dopamine neurotransmission, and this is the first work to demonstrate that RNF11 affects both dopamine release and uptake. RNF11 expression decreases in human dopaminergic neurons during PD, and that decrease may be protective by increasing dopamine neurotransmission in the surviving dopaminergic neurons.
Introduction
Recent work has indicated a role for RING finger protein 11 (RNF11) in Parkinson disease (PD) pathology (Anderson et al., 2007; Pranski et al., 2013b). RNF11 shows homology to E3 ubiquitin ligases and functions as a part of the A20 ubiquitin-editing complex and HECT-type E3 ubiquitin ligase (Kitching et al, 2003). The A20 complex inhibits NF-κB activation in human neurons (Pranski et al., 2012) and persistent NF-κB activation is a hallmark of neurodegenerative disease (Glass et al., 2010; Bellucci et al., 2020). RNF11 variants in the PARK10 locus may increase susceptibility for PD (Lesage and Brice, 2009). RNF11 colocalizes with α-synuclein and Lewy bodies in PD tissue (Anderson et al., 2007). Analysis of genes and extracted brain tissue from Parkinson’s patients revealed RNF11 gene expression and protein level is decreased in brain tissue affected by neurodegenerative diseases (Li et al., 2014, Pranski et al., 2013b). In rats with the 6-OHDA model of PD, RNF11 knockdown protected against dopaminergic neuronal cell death and RNF11 overexpression enhanced toxicity (Pranski et al., 2013a). The enhanced toxicity from RNF11 overexpression corresponded to downregulation of NF-κB transcribed antioxidants glutathione synthetase and superoxide dismutase 1, anti-apoptotic factor BCL2, neurotrophic factor brain-derived neurotrophic factor (BDNF), and tumor necrosis factor alpha (TNF-α). Therefore the decreased expression of RNF11 found in surviving neurons is thought to be a compensatory response to PD pathology (Pranski et al., 2013a). This decrease is not correlated with age, making RNF11 an unchanging candidate as a biomarker of neurodegeneration (Li et al., 2014). While RNF11 downregulation has been discovered in the context of neurodegeneration, no studies have addressed the effect of RNF11 on neurotransmission.
Drosophila melanogaster is a powerful model organism for the study of neurodegeneration and genetic tools exist to study the effects of RNF11 expression levels on neurotransmission (Hirth, 2010; Shin et al., 2018). A true homolog of RNF11, CG32850, has been identified in the Drosophila melanogaster genome. This Drosophila RNF11 has 87% positive sequence identity with the RING domain of the human protein (Kitching et al., 2003). An A20 ubiquitin-editing complex has not been described in fruit flies. However, homologs for many NF-κB pathway proteins have been found (Khush and Lemaitre, 2000), so it is likely that Drosophila RNF11 works in concert with these proteins to edit ubiquitin and inhibit NF-κB function. Drosophila is also a good model system for studying dopamine as the dynamics of dopamine release can be measured using fast-scan cyclic voltammetry (FSCV) at carbon-fiber electrodes (Vickrey et al., 2009; Xiao et al., 2014; Privman and Venton, 2015; Pyakurel et al., 2018; Shin and Venton, 2018). FSCV provides information about both release and uptake of dopamine (Vickrey et al., 2013). Thus, dopamine can be measured in genetically altered flies to give a better understanding of how RNF11 affects dopamine release.
The goal of this study was to test the effect of RNF11 knockdown or overexpression on the release and clearance of dopamine in Drosophila larvae. The GAL4/UAS system was used to either overexpress RNF11 or knock it down using RNAi (Brand and Perrimon, 1993). Dopamine release and uptake were measured with FSCV in the Drosophila ventral nerve cord (VNC) (Vickrey et al., 2009). Release was controlled optogenetically using the red light sensitive channelrhodopsin CsChrimson (Privman and Venton, 2015). RNF11 RNAi flies released double the concentration of dopamine, while RNF11 overexpressing flies did not release dopamine differently than control. This effect was specific to dopamine, as knocking down RNF11 in serotonergic or octopaminergic neurons did not alter release. Pulsed stimulations were modeled to obtain the Michaelis-Menten kinetic parameters (Xiao et al., 2014), revealing that Vmax, the maximum rate of dopamine uptake, increased in RNF11 RNAi flies and decreased in RNF11 overexpressing flies. These experiments are the first to demonstrate that RNF11 affects the ability of cells to release and uptake dopamine. Thus, the decrease in RNF11 during PD may lead to increased dopamine neurotransmission, which could be compensating for dopaminergic cell loss.
Experimental Procedures
Chemicals
Chemicals and drugs were purchased from Sigma-Aldrich (St. Louis, MO) and solutions were made with Milli-Q water (Millipore, Billerica, MA). Dissections, recordings, and calibrations were performed in a physiological buffer (131.3 mM NaCl, 3.0 mM KCl, 10 mM NaH2PO4 monohydrate, 1.2 mM MgCl2 hexahydrate, 2.0 mM Na2SO4 anhydrous, 1.2 CaCl2 dihydrate, 11.1 mM glucose, 5.3 mM trehalose, pH = 7.4). A 40 mM stock solution of nisoxetine was made in buffer and added to the sample for a final concentration of 20 μM. A 10 mM stock solution of flupenthixol was made in DMSO, diluted in buffer, and added to the sample for a final concentration of 5 μM flupenthixol. For nisoxetine and flupenthixol testing, a pre-drug baseline recording was taken and then the drug was bath applied for 15 minutes before post-drug recordings were taken. Flupenthixol is known to decrease the sensitivity of the electrode for dopamine, so the post-drug data were corrected with a calibration factor to account for the loss in signal (Vickrey and Venton, 2011).
Electrochemical Measurements
Cylindrical carbon-fiber microelectrodes were made using T-650 carbon fibers (a gift of Cytec Engineering Materials, West Patterson, NJ) as previously described (Jacobs et al., 2011). Electrodes were calibrated in vitro using a flow cell to determine the electroactive surface area. The electrodes were exposed to buffer containing 1 μM dopamine, serotonin, or octopamine to determine the response to each neurotransmitter. This calibration factor was used to calculate the amount of neurotransmitter measured in vivo as peak oxidation current increases linearly with analyte concentration within the physiological range. Fast-scan cyclic voltammetry was performed using a ChemClamp potentiostat with an n=0.01 headstage (Dagan, Minneapolis, MN), PCI 6711 and 6052 computer interface cards (National Instruments, Austin, TX), and a home built break-out box. TarHeel CV software (a gift of Mark Wightman, University of North Carolina) was used to collect and background subtract the data. Every 100 ms the electrode was scanned from −0.4 to 1.3 V and back at 400 V/s for dopamine detection (Venton and Cao, 2020). For serotonin and octopamine detection, modified waveforms were used to minimize electrode fouling. For the serotonin waveform, the potential was held at 0.2 V, and then scanned up to 1.0 V, down to −0.1 V, and then back to 0.2 V at a rate of 1000 V/s every 100 ms.(Jackson et al., 1995; Borue et al., 2009) For octopamine detection, the potential was held at −0.4 V and scanned to 1.4 V at 100 V/s every 100 ms (Pyakurel et al., 2016). The tissue content of dopamine was measured from 3 pooled larval Drosophila ventral nerve cords using capillary electrophoresis coupled with FSCV as described previously (Denno et al., 2015, 2016).
Fly Strains
Fly stocks were all prepared on a white- genetic background. The following fly stocks were used in this study: UAS-RNF11JF11021-RNAi, UAS-RNF11HMJ22085-RNAi, TH-GAL4, Tph-GAL4, tdc-GAL4, Act5C-GAL4 (Bloomington stock center), and UAS-CsChrimson (Klapoetke et al., 2014). The UAS-CG32850 (dRNF11) line was generated as follows: an EcoRI/Acc65I fragment of RE41137 (Drosophila Genomics Resources Center, Indiana University), containing a full length cDNA of Drosophila RNF11, was ligated into pUASTattB vector and transgenic lines were established by using PhiC31integrase-mediated site specific germ line transformation. Fragments were then cloned into the pENTR Gateway entry clone, followed by LR clonase recombination into the Drosophila expression vector pDEST-TWF. Constructs were injected into embryos and progeny screened for positive transformants by Rainbow Transgenic Flies, Inc.
Tissue Preparation
For genetic expression in dopamine neurons, flies with TH-GAL4, UAS-mCD8-GFP were crossed with UAS-CsChrimson (control); UAS-CsChrimson, UAS-RNF11-RNAi (RNF11 knockdown); or UAS-CsChrimson, UAS-RNF11 (RNF11 overexpression). The light senesitive CsChrimson channel was used for optogenetic control of the cells (Klapoetke et al., 2014). For the serotonin experiments, the flies expressed Tph-GAL4 instead and for octopamine experiments they expressed Tdc2-GAL4. Resulting heterozygous larvae were shielded from light and raised on standard cornmeal food mixed 250:1 with 100 mM all-trans-retinal. The VNC of a third instar wandering larva (either sex) was dissected in PBS buffer. Isolated VNCs were prepared and recorded from as previously described (Borue et al., 2009). The electrode was allowed to equilibrate in the tissue for 10 minutes prior to data collection. A baseline recording was taken for 10 seconds prior to stimulation to facilitate background subtraction.
Stimulated Neurotransmitter Release
Red-orange light from a 617 nm fiber-coupled high-power LED with a 200 μm core optical cable (ThorLabs, Newton, NJ) was used to stimulate the CsChrimson ion channel. The power output at the fiber end was 0.75 mW and the fiber was centered positioned 75-100 μm from the ventral nerve cord using a micromanipulator. The light was modulated with Transistor-Transistor Logic inputs to a T-cube LED controller (ThorLabs, Newton, NJ), which was connected to the breakout box. TTL input was driven by electrical pulses controlled by the TarHeel CV software, which was used to control the frequency (60 Hz), pulse width (4 ms), and number of pulses (120) in the stimulations.
Quantitative PCR
For each genotype, 20 adults expressing the Act5C-GAL4 ubiquitous driver were flash frozen and stored at −80 °C before RNA extraction and RT-qPCR analyses. RNA was purified using the RNeasy Mini protocol (Qiagen, Hilden, Germany). Isolated RNA was quantified using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE, USA). 100 microgram of RNA was reverse transcribed using the iScript cDNA Synthesis kit (Bio-Rad, Hercules, CA), according to the manufacturer’s instructions. The cDNA was stored at −80 °C until use.
Primers were designed using primer-BLAST (http://www.ncbi.nlm.nih.gov) (Table 1). RT-qPCR was performed on a CFX Connect Detection System (Bio-Rad, Hercules, CA) using Sso Advanced Universal SYBR Green Mix (Bio-Rad, Hercules, CA). Reactions were carried out in 20 μl total volume, with 200 ng cDNA and 100 nM of each primer. The following amplification program was used in all RT-qPCRs: 95 °C for 2 min, and 40 cycles of 15 s at 95 °C and 15 s at 60 °C. The specificity of each amplified reaction was verified by a dissociation curve analysis after 40 cycles by 2 % agarose gel electrophoresis. Each sample was analyzed in quadruplicate wells and NTC “no-template controls” (without cDNA in the PCR) were included.
The minimum information for publication of quantitative real-time PCR experiments (MIQE) guidelines (Bustin et al. 2009) were used to promote implementation of experimental consistency and to increase the integrity of our results. Serial 10-fold dilutions of each primer pair were used to generate a standard curve and amplification efficiencies (E) were determined based on the slope (M) of the log-linear portion of each standard curve (E = 10-1/M −1). Serial 2-fold dilutions were used to generate a standard curve, and amplification efficiencies (E) were determined based on the slope (M) of the log-linear portion of each standard curve (E = 10−1/M −1 * 100).
Statistics and Data Analysis
Data are presented as mean +/- standard error of the mean (SEM) displayed as error bars. GraphPad Prism 6 (GraphPad Software Inc, La Jolla, CA) was used to perform all statistics, including one-way and two-way ANOVA with Bonferonni post-hoc tests. Stimulated changes in the extracellular concentration of dopamine were modeled as a balance between discontinuous release and continuous uptake (Wu et al., 2001). A nonlinear regression with a simplex minimization algorithm was used to fit the curves (Wu et al., 2001). The modeling software was stopped when the iteration number continued increasing without a change in the three floated parameters: Km, Vmax, and dopamine release per light pulse ([DA]p). The goodness of the fit was described using the square regression coefficient (R2). The average R2 for this work was 0.96 with a standard deviation of 0.02.
Results
To study the effects of RNF11 on dopamine release, RNF11 RNAi or overexpression was driven by the tyrosine hydroxylase (TH) GAL4 driver, which is a dopamine synthesis enzyme. Dopamine release was stimulated using red light to activate CsChrimson, and extracellular dopamine concentrations were monitored in real-time using FSCV (Privman and Venton, 2015). The pulsed optogenetic stimulation mimics the phasic neurotransmission that is important for reward and motivation in mammals (Venton et al., 2003). Figure 1A shows that knocking down RNF11 (unless otherwise noted all experiments for RNF RNAi are the JF11021 allele) caused larger dopamine release than control or RNF11 overexpression flies. Figure 1B shows example cyclic voltammograms, which confirm dopamine was detected with its characteristic oxidation peak at 0.6 V and reduction peak at −0.2 V. For average peak concentrations, extracellular dopamine concentration varies with genotype (Fig. 1C, one-way ANOVA, p=0.0058) and RNF11 RNAi flies have significantly larger dopamine release than control, about double the normal concentration (Bonferroni post hoc tests, p<0.01). There was no significant difference between the peak release for control and RNF11 overexpressing flies. To confirm that the increased dopamine release was not specific to the particular RNF11 allele tested, the RNF11HMJ22085 strain was also tested. Similar to the JF11021 allele, the HMJ22085 RNAi line also released about twice the amount of dopamine upon activation (Fig. 2).
In the present study, RNF11 was either overexpressed or knocked down with the use of the GAL4/UAS system available in Drosophila. To check the overall efficacy of the UAS-RNF11-RNAi and the RNF11 overexpression constructs, the UAS flies were crossed to the universal actin (Act5C) driver, and RNF11 mRNA levels were shown to be reduced to 75 ± 4% of control in the RNAiJF11021 allele. The overexpression of RNF11 by the actin driver increased mRNA levels to 366 ± 8% of control (Fig. 2B). Given the changes in gene expression observed with Act5C-GAL4, it is reasonable that similar changes would likewise be observed with other GAL4 driver lines.
To check the specificity of this effect for dopaminergic neurons, RNF11 RNAi was expressed along with CsChrimson in serotonergic (Tph-GAL4) or octopaminergic (Tdc2-GAL4) neurons. Neurotransmitter release was monitored using specialized waveforms for those neurotransmitters during a 2 s pulsed red light stimulation. Knocking down RNF11 in serotonergic neurons did not change stimulated serotonin levels (Figure 3A, t-test, p=0.8054); similarly, knocking down RNF11 in octopaminergic neurons did not change peak octopamine concentrations (Figure 3B, t-test, p=0.9345). Thus, RNF11 RNAi selectively increases dopamine release.
In order to determine the extent to which dopamine release and uptake are affected by RNF11 knockdown, the current versus time plots (examples shown in Figure 1A) were modeled using Michaelis-Menten kinetics (Figure 4) (Wu et al., 2001). The parameters determined were [DA]p, the amount of dopamine released with each pulse of stimulating light, Vmax, the maximum rate of clearance by the dopamine transporter (DAT), and Km, the affinity of DAT for dopamine. Dopamine released per pulse of light is affected by genotype (Fig. 4A, one-way ANOVA, p=0.0006) and RNF11 knockdown significantly increases dopamine release per pulse compared to control (Bonferroni post hoc test, p<0.001). RNF11 overexpression does not affect [DA]p. Vmax is also significantly affected by genotype (Fig. 4B, one-way ANOVA, p<0.0001). RNF11 knockdown significantly increases Vmax compared to control (Bonferroni post-hoc test, p<0.05) and RNF11 overexpression significantly decreases Vmax compared to control (Bonferroni post-hoc test, p<0.05). Km is not affected by genotype (Fig. 4C, one-way ANOVA, p=0.3056). Fig. 4D shows the DAT mRNA levels in adult flies where RNF11 is either knocked down or overexpressed by the ubiquitous Act5C-GAL4 driver. There is a main effect of genotype (one-way ANOVA, p=0.0022, n=4) and RNF11 knockdown produces double the DAT mRNA (p<0.01), while RNF11 overexpressing flies show similar DAT mRNA levels to controls.
The amount of dopamine release may be due to increases in the total pool of dopamine. Tissue content was measured using capillary electrophoresis coupled to FSCV (Figure 5) (Fang et al., 2011). The dopamine content per larval central nervous system (CNS) was significantly affected by the genotype of the flies (one-way ANOVA, p=0.0003). Post hoc tests showed that the RNF11 RNAi tissues had significantly more dopamine than the RNF11 overexpressing tissue (p=0.0002). However, neither the RNF11 RNAi (p=0.1152) nor the overexpressing (p=0.0510) flies differed significantly from control. This indicates that tissue content did not increase as much as release.
To test whether DAT was contributing to release through reverse transport, DAT was competitively inhibited by bath application of 20 μM nisoxetine for 15 minutes (Pörzgen et al., 2001). Figure 5A shows example data of the pre-drug baseline stimulation in an RNF11 RNAi VNC (orange) and the slowed dopamine uptake after nisoxetine (purple), as evidenced by the increased time to clear dopamine to half the maximum concentration (t50). Figure 6B summarizes the data and a two-way ANOVA showed a significant main effect of genotype (p=0.0002) and nisoxetine (p=0.0012), but no interaction. Post-tests show a higher t50 for RNF11 overexpressing flies than both RNF11 RNAi and control. The percent increase in t50 from baseline upon application of nisoxetine was calculated for each genotype, and one-way ANOVA revealed no significant difference (p=0.3558). This indicates that nisoxetine is affecting dopamine uptake equally among the genotypes, despite the slower uptake in RNF11 overexpressing flies. These results also indicate that the larger dopamine release in RNF11 RNAi flies cannot be accounted for by reverse transport through DAT.
Extracellular dopamine levels can be controlled through negative feedback of the D2 dopamine autoreceptor. If RNF11 knockdown decreases D2 receptor levels in the presynaptic cell membrane, the cell would release more dopamine due to a lack of negative inhibition. D2 was inhibited by bath application of 5 μM flupenthixol for 15 minutes (Vickrey and Venton, 2011). Figure 6A shows example traces of the pre-drug baseline in an RNF11 RNAi VNC (light green) and the post-drug increased dopamine release (dark green). Figure 6B shows the averaged data. Stimulated dopamine release doubled in all genotypes after flupenthixol bath application. A two-way ANOVA showed a significant main effect of genotype (p=0.0105) and flupenthixol (p=0.0007), but no interaction (p=0.2963). The percent increase in dopamine concentration from baseline upon application of flupenthixol was calculated for each genotype, and one-way ANOVA revealed no significant difference (Fig. 6C, p=0.0659, n=4-7). This indicates that D2 autoreceptors function similarly in all genotypes and RNF11 knockdown is not increasing dopamine by modulating D2 autoreceptor function.
The turnover of dopamine was assessed by performing multiple stimulations in quick succession. Stimulations were performed every 30 s, which excludes the contribution of dopamine synthesis to the releasable pool since synthesis occurs on the order of 5-10 minutes (Xiao and Venton, 2015). Figure 7 shows that stimulated release decreases with multiple stimulations, but that RNF11 RNAi flies did not exhibit as large a decrease in evoked dopamine as did control or RNF11 overexpressing flies. There was a significant main effect of both stimulation number (two-way ANOVA, p<0.0001) and genotype (p<0.0001), but no interaction (p=0.1468). Post hoc tests showed that RNF11 RNAi flies have a higher plateau than both control and RNF11 overexpressing flies (p<0.0001 for both). The data were fit to a single-phase exponential decay, which were significantly different (comparison of fits, p<0.0001). Control flies plateaued at 51% of the initial stimulated release, RNF11 overexpressing plateaued at 59%, and RNF11 RNAi flies plateaued at 75%. Thus, significantly more dopamine is being recycled for reuse in RNF11 RNAi flies than in control or RNF11 overexpressing flies.
Discussion
RNF11 downregulation has been observed in the late pathogenesis of PD; however early effects of RNF11 depletion have not been studied. In addition, no study had examined the effect of RNF11 knockdown on dopamine release and clearance. Here, we describe for the first time that knocking down RNF11 increases dopamine neurotransmission in Drosophila larvae. This effect was specific for dopamine, as knocking down RNF11 in serotonin or octopamine neurons did not affect stimulated release. RNF11 RNAi larvae had faster rates of dopamine uptake and more dopamine available for rapidly repeated stimulations than control or RNF11 overexpressing flies, indicating that RNF11 knockdown promotes recycling of dopamine and maintenance of the releasable pool. RNF11 downregulation in neurodegenerative tissue has been postulated as a compensatory response to protect dopaminergic neurons from the insult of PD. Our work indicates that dopaminergic cells with downregulated RNF11 have more phasic dopamine release. PD patients do not exhibit motor symptoms until a majority of the dopaminergic neurons are dead. Therefore, downregulation of RNF11 may be a mechanism by which the remaining cells compensate for that loss to maintain the long, symptom-free progression that characterizes this movement disorder.
RNF11 knockdown specifically increases dopamine release
RNF11 RNAi flies had about a quarter of the normal levels of RNF11 expression and significantly higher stimulated dopamine release than control flies. This doubling in release is about equivalent to blocking the D2 autoreceptors with flupenthixol, which shows that release is greatly affected by RNF11 knockdown. However, the increase in release is not due to modulation of D2 autoreceptors, since application of flupenthixol doubled the release in RNF11 RNAi flies as well as controls. One explanation for greater release could be greater synthesis of dopamine. However, RNF11 RNAi flies did not have significantly higher dopamine tissue content than control flies. Dopamine tissue content in RNF11 RNAi was significantly higher than RNF11 overexpressing flies, indicating that synthesis might be somewhat affected by changing RNF11 levels, but the effect is certainly not a doubling of tissue content. RNF11 knockdown flies also have greater stimulated release when stimuli were closely applied, showing that they have more dopamine turnover to maintain a larger releasable pool. Thus, the main finding of this work is that RNF11 RNAi flies are able to increase dopamine release and maintain the releasable pool, and this effect is not primarily mediated by D2 autoreceptors or synthesis.
To test the specificity of release, RNF11 RNAi knockdown was also performed in serotonergic or octopaminergic neurons. Stimulated release of serotonin or octopamine was not affected by RNF11 RNAi knockdown. This demonstrates that the effect of RNF11 knockdown is not universal to all neurotransmitters, but is more specific for dopamine. Dopamine, serotonin, and octopamine are all packaged into vesicles by the same mechanism: vesicular monoamine transporters (VMAT). Humans encode two VMAT genes, VMAT1 and VMAT2, while Drosophila has only one, dVMAT (Martin and Krantz, 2014). Since RNF11 RNAi affects only dopamine release, it is unlikely that the mechanism is through dVMAT, since that would also affect other amine neurotransmitters such as serotonin and octopamine.
While RNF11 knockdown flies had dramatically larger stimulated dopamine release, overexpressing RNF11 did not significantly change the level of stimulated release. However, the dopamine tissue content of dopamine in RNF11 overexpressing flies was significantly lower than in RNF11 knockdown flies and close to significantly lower (p=0.0510) than control flies. These data indicate that there is some compensation happening in the dopamine neurons that allows the RNF11 overexpressing flies to maintain normal stimulated release, even when the tissue content is somewhat lower. D2 autoreceptors do not appear to be responsible for this effect, as the magnitude of change with flupenthixol was not different than control or RNF11 RNAi flies. Interestingly, the turnover of dopamine is similar to control, meaning the tissue can support the same level of release with RNF11 overexpression even though tissue levels are somewhat lower.
In humans, RNF11 functions as part of the A20 ubiquitin editing complex to dampen the signaling cascade that leads to NF-κB activation (Pranski et al., 2013b). Overexpressing RNF11 leads to a decrease in NF-κB function and knocking down RNF11 increases NF-κB function (Shembade et al., 2009). In this work we see effects on release, tissue content, and dopamine recycling in dopaminergic neurons with decreased RNF11, which may be due to increased NF-κB function. NF-κB has multiple targets and future studies would be needed to discern whether the increase in release works through NF-κB and how its downstream targets work together to affect release. However, we see a limited effect on release and tissue content with RNF11 overexpression. This may be due to NF-κB already functioning at a low baseline in healthy dopaminergic cells, and further downregulation having limited downstream effects.
RNF11 RNAi flies have increased dopamine uptake
RNF11 RNAi and overexpressing flies also exhibit clear changes in dopamine uptake. RNF11 RNAi flies had significantly faster uptake than control flies and RNF11 overexpressing flies had a significantly slower uptake than control. The affinity of the transporter for dopamine (Km) did not change. The DAT mRNA levels were significantly higher for RNF11 RNAi flies, but were not significantly lower in RNF11 overexpression flies. While the qPCR results give a general picture of the amount of DAT expressed, the Vmax corresponds to the number of functioning transporters expressed on the surface. Descrepancies between these two experiments could be due to different rates of internalization, and may explain why measured uptake is slower in RNF11 overexpression flies while the DAT mRNA level is the same. RNF11 may affect the trafficking of DAT and surface expression by modulating the endocytic pathway (Santonico et al., 2010; Gabriel et al., 2013). RNF11 localizes to complexes that mediate endocytosis and endosome trafficking (Pisitkun et al., 2004; Gonzales et al., 2009; Hong et al., 2009; Kostaras et al., 2014), and RNF11 regulates the switch between recycling and degradation of endosomes (Santonico et al., 2014). Therefore, RNF11 knockdown could lead to a higher proportion of DAT being recycled to the membrane rather than degraded in the endosome pathway. However, there are also NF-κB downstream pathways that could regulate uptake. Overexpressing RNF11 leads to less activated NF-κB and thus less transcription of brain derived neurotropin factor (BDNF). In mice with 50% of normal BDNF levels, the Vmax for dopamine uptake was reduced (Bosse et al., 2013), which is consistent with the decrease in Vmax we observed in RNF11 overexpressing Drosophila. However, BDNF+/- mice also showed a decrease in dopamine release, which was not found in the RNF11 overexpression Drosophila model. It is important to note that RNF11 has not yet shown to directly assemble ubiquitin chains on any substrate but itself, and in fact might act as an adapter for other E3 ligases (Santonico et al., 2015; Budhidarmo et al., 2018). Thus, the RNF11 effects on release and uptake cannot be explained by just one of the known downstream targets of NF-κB and certainly involves a more complicated protein machinery.
The turnover of dopamine in RNF11 RNAi flies is also faster, indicating that the mechanisms for dopamine repackaging and release are upregulated when RNF11 is knocked down. The increased uptake would lead to more dopamine being available for recycling, thus dopamine release is maintained for closely repeated stimulations. Previous studies have shown that the releasable pool of dopamine is dependent on transporters in this time frame,(Xiao and Venton, 2015) and the finding of faster uptake in RNF11 RNAi flies is consistent with this being the main mechanism of releasable pool regeneration. This maintenance of the pool is not explained by the reverse transport of dopamine through DAT or the downregulation of D2 dopamine autoreceptors. There is also a decrease in the uptake of dopamine in RNF11 overexpressing flies, allowing dopamine to be active in the extracellular space for longer. Release and turnover levels are maintained in RNF11 overexpressing flies, despite the lower tissue content and slower uptake. This normal maintenance of the releasable pool points to other mechanisms that are helping to maintain release in RNF11 overexpressing flies and future studies could probe the possible mechanism of the regulation of dopamine turnover. The increased tissue content, increased release, and faster turnover of dopamine seen in RNF11 RNAi flies could also occur due to a decreased metabolic breakdown of dopamine in the cytoplasm after uptake. The dopamine metabolic pathways are poorly characterized in Drosophila; however they would also likely metabolize serotonin and octopamine (Suh and Jackson, 2008; Freeman et al., 2013; Ueno and Kume, 2014; Yamamoto et al., 2014). Therefore, metabolism is unlikely to be the cause of the dopamine release and turnover effects upon RNF11 knockdown.
Implications of RNF11 effects on dopamine neurotransmission in PD
RNF11 affects many pathways important to proper neuronal functioning that could make it important in the cellular response to PD. While dopamine neurons die in early PD, there is a prolonged compensation period, during which remaining neurons continue to support dopaminergic neurotransmission (Hawkes et al., 2010). RNF11 is downregulated in neurodegenerative diseases, including PD (Pranski et al., 2012, 2013b). Here we present the first evidence that RNF11 knockdown increases dopamine release and uptake, which may represent a mechanism by which dopaminergic cells maintain functional levels of dopamine despite the loss of a substantial percentage of the local neuronal population. Future studies could examine the importance of RNF11 in PD by simultaneously manipulating RNF11 and known PD genes in Drosophila, such as PINK1 or parkin (Park et al., 2006) to see the phenotypic effects of RNF11 on disease progression as well at the effects on dopamine neurotransmission. In addition, studies to change the levels of RNF11 knockdown could examine the extent to which further decreases in RNF11 increase dopamine release and whether RNF11 could be a target for future PD therapies.
Conclusions
RNF11 has been implicated in the pathophysiology of human PD, but its effect on dopamine neurotransmission was unknown. Here, in Drosophila, we show that knocking down RNF11 doubled dopamine release, while RNF11 overexpression did not change release. This effect was dopamine specific, as RNF11 knockdown does not have a similar effect on the serotonergic or octopaminergic neurotransmitter pathways. RNF11 knockdown increased, and RNF11 overexpression decreased the rate of dopamine uptake, showing that RNF11 affects both dopamine release and uptake. Future studies could identify the extent to which the effect of RNF11 is mediated by NF-κB or endosome recycling. These results are important for understanding PD, as the upregulation of dopamine signaling may help maintain dopamine function in PD patients that have reduced RNF11 expression. RNF11 may also be a future biomarker or drug target for detecting or treating PD.
Acknowledgments
The authors would like to thank the Gaultier, Hirsh, and Jayaraman labs for reagents and fly stocks and Karol Cichewicz for his help with fly selection. Work was funded by NIH R01MH085159 and by Biogen-Idec.
Abbreviations
- BDNF
- Brain derived neurotrophic factor
- [DA]p
- Dopamine release per light pulse
- DAT
- Dopamine transporter
- PD
- Parkinson disease
- RNF11
- Ring finger protein 11
- TH
- Tyrosine hydoxylase
- VMAT
- Vesicular monoamine transporter
- VNC
- Ventral nerve cord