Ethanol stimulates the in vivo axonal movement of neuropeptide dense core vesicles in Drosophila motor neurons

Abstract

Proper neuronal function requires essential biological cargoes to be packaged within membranous vesicles and transported, intracellularly, through the extensive outgrowth of axonal and dendritic fibers. The precise spatiotemporal movement of these cargoes is vital for neuronal survival and, thus, is highly regulated. In this study we test how the axonal movement of a neuropeptide containing dense core vesicle (DCV) responds to alcohol stressors. We found that ethanol induces a strong anterograde bias in vesicle movement. Low doses of ethanol stimulate the anterograde movement of neuropeptide-DCV while high doses inhibit bidirectional movement. This process required the presence of functional kinesin-1 motors as reduction of kinesin prevented the ethanol-induced stimulation of the anterograde movement of neuropeptide-DCV. Furthermore, expression of inactive GSK-3β also prevented ethanol-induced stimulation of neuropeptide-DCV movement, similar to pharmacological inhibition of GSK-3β with lithium. Conversely, inhibition of PI3K/AKT signaling with wortmannin led to a partial prevention of ethanol-stimulated transport of neuropeptide-DCV. Taken together, we conclude that GSK-3β signaling mediates the stimulatory effects of ethanol. Therefore, our study provides new insight into the physiological response of the axonal movement of neuropeptide-DCV to exogenous stressors.

Introduction

Normal brain development is reliant upon proper neuronal growth and function. The developing neuron is susceptible to many environmental stressors. Prenatal exposure to alcohol is known to have teratogenic effects by producing excessive reactive oxygen species that result in neuron apoptosis by altering the structural and physiological integrity of the cell. Neuronal tissue loss in the developing central nervous system is a pathological hallmark of the fetal alcohol spectrum disorders. However, despite early recognition of the toxic effects of ethanol (Jones et al. 1973), there remains a lack of understanding regarding the biological mechanisms underlying ethanol-related pathology.

Exposure to ethanol (EtOH) has long been known to modulate vesicle transport in neurons. Monitoring retrograde dye injected into the sciatic nerve revealed reduced transport in EtOH-fed rats (McLane 1990, McLane 1987). Similar findings were observed in cholinergic neurons (Malatova & Cizkova 2002) and electron micrographs that showed accumulations of organelles contained mitochondria, large and small clear vesicles and dense vesicles with EtOH exposure (McLane et al. 1992). In contrast, increased retrograde transport was observed in the vagus nerve in response to EtOH (Laduron & De Witte 1987). However, the molecular mechanisms underlying EtOH-induced effects on axonal transport remain elusive.

Drosophila has been widely used as an established model for studying the molecular pathology of EtOH-related disorders (Kaun et al. 2012, Guarnieri & Heberlein 2003, Scholz et al. 2000). In addition, Drosophila has a natural response to EtOH (Guarnieri & Heberlein 2003). Genetic and biochemical evidence in Drosophila has identified novel molecules with mammalian orthologues implicated in EtOH-mediated responses (Corl et al. 2005, Maas et al. 2005, Moore et al. 1998, Rodan et al. 2002, Rothenfluh et al. 2006, Wen et al. 2005, Park et al. 2000, Urizar et al. 2007). Further, EtOH exposure in Drosophila and other model organisms induce cellular stress pathways, such as reactive oxygen species production and membrane permeability, with significant functional overlap with those induced by other environmental stressors, such as heat shock (Scholz et al. 2005, Piper 1995). Mutations in the Drosophila homologue of PARK2 resulted in greater sensitivity to reactive oxygen stress (Pesah et al. 2004). In insects and mammalian cells, pesticides induce cellular stress by reactive oxygen species production (Abdollahi et al. 2004). The activation of cellular stress pathways by environmental stressors has been linked to severe perturbations in vesicular transport in neurons (Fang et al. 2012). Ultraviolet stress has been shown to perturb the transport of amyloid precursor protein (APP) in mammalian neurons (Almenar-Queralt et al. 2014). Therefore, intracellular trafficking of various cargoes appears to play a pivotal role in mediating neuronal responses to exogenous stressors; however, the mechanism of how such stress directly affects transport is unknown.

Several studies have implicated the PI3K/AKT stress pathway and its downstream targets of EtOH-induced signaling (He et al. 2006, Huang & Kim 2012, Pascual & Guerri 2007, Zeng et al. 2012). Activating PI3K has been shown to rescue EtOH-induced apoptosis (de la Monte et al. 2000). In both the fly and mammalian systems, Shaggy (SGG), the fly homologue of Glycogen Synthase Kinase 3β (GSK-3β), a major substrate of AKT, has been implicated in EtOH-induced neurotoxicity by NMDA receptor activity and proapoptotic protein, Bax (French & Heberlein 2009, Liu et al. 2009, Neznanova et al. 2009). Furthermore, ionizing radiation enhances the activation of SGG via decreased AKT activity (Tan et al. 2006). Ionizing radiation is a known inducer of reactive oxygen species production in Drosophila neurons similar to EtOH (Kim & Johnson 2014). SGG has also been shown to have an inhibitory role (Mudher et al. 2004) in regulating the transport of various cargoes including APP (Weaver et al. 2013), atrial natriuretic factor (ANF), mitochondria, and synaptobrevin (Dolma et al. 2014).

In the present study, we test the hypothesis that GSK-3β plays a role in modulating the response of vesicle motility to EtOH, using in vivo imaging coupled with genetics and pharmacology. Human and preclinical studies have previously established a connection between alcohol consumption and ANF plasma concentrations (Kovacs 2003). Furthermore, several Drosophila motor neurons are peptidergic and control behavioral responses to environmental stressors (Gorczyca et al. 1993, Monastirioti et al. 1995, Nassel & Wegener 2011). In Drosophila, an endogenous neuropeptide, Neuropeptide F, was found to be required for the behavioral effects of EtOH (Wen et al. 2005). Therefore, peptidergic neurons likely play a vital role in mediating EtOH responses in flies. To study this system, we used an established neuropeptide marker, rat atrial natriuretic peptide tagged with green fluorescent protein (ANF-GFP), using previously established protocols (Reis et al. 2012, Weaver et al. 2013, Gunawardena et al. 2013) to monitor changes to motility dynamics. ANF-GFP is specifically targeted to dense core vesicles (DCVs) in both flies (Rao et al. 2001, Barkus et al. 2008, Hill et al. 2012, Wong et al. 2012, Husain & Ewer 2004, Bulgari et al. 2017, Neisch et al. 2017, Cavolo et al. 2015, Wong et al. 2015, Kuznetsov & Kuznetsov 2015, Grygoruk et al. 2014) and mice (Xia et al. 2009). In mammalian systems, ANF has been implicated in water and salt homeostasis as well as vasodilation and plasma volume regulation (Curry 2005). Previous work showed that exogenous ANF-GFP is proteolytically processed, sorted into secretory granules, localized at peptidergic terminals, and released upon stimulation (Rao et al. 2001, Husain & Ewer 2004, Heifetz & Wolfner 2004, Kula et al. 2006, Shakiryanova et al. 2006, Loveall & Deitcher 2010). Furthermore, ANF-GFP has been used to identify peptidergic neurons in the central nervous system in flies (Husain & Ewer 2004). ANF is enriched in dense core vesicles and is widely used to identify these vesicles in Drosophila, with detailed tracking of DCV motility in vivo (Iacobucci et al. 2014, Bulgari et al. 2017, Cavolo et al. 2015, Grygoruk et al. 2014, Hill et al. 2012, Husain & Ewer 2004, Kuznetsov & Kuznetsov 2015, Neisch et al. 2017, Rao et al. 2001, Wong et al. 2015, Wong et al. 2012). Therefore, ANF-GFP is a reliable and a widely accepted marker for dense core vesicle behavior in Drosophila. We found striking shifts in the axonal motility of ANF during prolonged EtOH exposure. Low doses of EtOH enhanced the axonal anterograde motility of DCVs, while high doses of EtOH inhibited bidirectional movement. This novel observation of EtOH-induced stimulation of axonal DCV movement requires functional kinesin-1 and is modulated by GSK-3β. Our observations also provide new insights into how the long distance axonal transport pathway responds to exogenous stressors under physiological conditions.

Methods

Drosophila Genetics and feeding experiments

Fly stocks and crosses were maintained on cornmeal food at room temperature in a 12hr light/dark cycle. Males containing the motor neuron driver pin^88K/cyo; D42-GAL4/TM2 (RRID: BDSC_8816) were crossed with either virgin female UAS-ANF-GFP (RRID: BDSC_7001), UAS-SYT-EGFP (RRID: BDSC_6924), or UAS-htt15Q-mRFP (gift from Dr Littleton). To reduce functional kinesin motors, male UAS-ANF-GFP/7; +/CyO; +/D42-GAL4 were crossed with virgin female klc^8ex94/TM6B (RRID: BDSC_31997). To express sgg^A81T, male UAS-ANF-GFP/7; +/CyO; +/D42-GAL4 were crossed with virgin female UAS-sgg^A81T (RRID: BDSC_5359). For dual imaging, male UAS-ANF-GFP/7; +/CyO; +/D42-GAL4 were crossed with virgin female UAS-htt15Q-mRFP. Crosses were maintained at 29°C for protein overexpression.

For feeding experiments, flies were allowed to mate for 48 hrs. on cornmeal food made with either 10 mM LiCl or 100 nM wortmanin dissolved in DMSO for a final concentration of 0.02% v/v in fly food, as previously done. Control feedings were done in fly food containing 0.02% v/v DMSO. Adult flies were transferred to normal cornmeal food. Laid eggs were allowed to mature to third instar larvae in food containing compounds.

In Vivo Vesicle imaging

Larvae were dissected and imaged as detailed previously (Kuznicki & Gunawardena 2010). Vesicle imaging in larvae was done using a Nikon Eclipse TE-2000E inverted fluorescence microscope and Metamorph Imaging software. Time-lapse movies were collected using a CoolSnap HQ camera (Roper Scientific, Surrey, BC, Canada). Five larvae were dissected per genotype. 150-frame movies were recorded consecutively for 30 min from each larva at 100X/1.40 NA (90 μm field-of-view) oil objective with a 2x2 binning factor yielding a 0.126 micron/pixel spatial resolution and a 200 msec exposure for each frame. Movies were cropped and oriented for analysis in Metamorph (MDS Analytical Technologies, Sunnyvale, CA) and analyzed using a MATLAB 2010b-based (Mathworks) custom single particle tracking program (Gunawardena et al. 2013, Reis et al. 2012, Weaver et al. 2013). Segmental velocities were defined as the mean velocity of a trajectory segment uninterrupted by a pause, reversal, or movie termination event. Duration-weighted segmental velocity evaluates the average velocity behavior that vesicles exhibit per time spent moving.

TUNEL assay

Cell death was assessed in ventral ganglion of isolated larval brains of female pin^88k/cyo; D42-GAL4/TM2. Brains were incubated in either 0, 10, 50, 100, or 300 mM EtOH in dissection buffer (2X stock contains 128 mM NaCl, 4 mM MgCl₂, 2 mM KCl, 5 mM HEPES, and 36 mM sucrose, pH 7.2) for 20 min. Brains were subsequently fixed in 4% paraformaldehyde in PBS for 30 min at 25°C. After washing in PBS, cells were permeablized with 5% saponin for 30 min at 25°C. After washing in PBS, brains were mounted in Vectashield mounting medium (Vector Labs) for imaging. TUNEL assay was performed using an In Situ Cell Death Detection Kit (Roche Life Science) per manufacturer’s instructions. Positive controls were incubated with DNAase I. Negative controls were incubated in labeling solution only. The number of puncta were quantified in ImageJ (NIH) using the Threshold tool and Analyze Particles tool.

Larval immunohistochemistry

Third instar larvae were dissected, fixed, and immunostained as described (Fye et al 2010, Gunawardena & Goldstein 2001). Briefly, larvae were dissected in dissection buffer (2X stock contains 128 mM NaCl, 4 mM MgCl₂, 2 mM KCl, 5 mM HEPES, and 36 mM sucrose, pH 7.2). Following dissection, larvae were treated with 0, 10, or 50 mM EtOH at 25°C for 20 min. Larvae were fixed in 4% formaldehyde and incubated with primary antibodies against either rabbit monoclonal cysteine string protein (DCSP-3, 1:10, Developmental Studies Hybridoma Bank; RRID:AB_528184) overnight. Larvae were incubated with the neuronal membrane marker Texas Red-conjugated horse radish peroxidase (HRP) and secondary antibody (Alexa anti-rabbit 488, 1:100, Invitrogen) for 2 hrs at room temperature, mounted using Vectashield mounting medium (Vector Labs) and imaged using a Nikon TE-2000E inverted microscope at 60X.

Statistical analysis

To select the appropriate statistical test, transport parameter distributions were first checked for normality using the nortest package of R: Lilliefors test and Anderson-Darling test. Statistical significance of normal distributions was calculated by a two-sample two-tailed Student’s t-test while the non-normal segmental velocity distributions were compared using the nonparametric Wilcoxon-Mann-Whitney rank sum test. Percent of cargo population tended to follow normal distributions. Duration-weighted segmental velocity, flux, and run length often followed a mixture of normal distributions or a non-normal distribution. Hill function fitting was performed in OriginPro 8 (OriginLab).

Results

Ethanol exposure stimulates anterograde motility of neuropeptide-DCV within larval axons

Various alcohols are known to have deleterious effects on biological systems (Baker & Kramer 1999). Previous studies evaluated the effect of EtOH (in the diet or by incubation of nerve ligations) on axonal transport by monitoring radiolabeled markers in animals (Malatova & Cizkova 2002, McLane 1987, McLane 1990, McLane et al. 1992). These methods only evaluated the long-term chronic effects of EtOH and the temporal resolution of immediate responses to EtOH was not evaluated. Further, nerve ligations can themselves induce aberrant artifacts that affect normal axonal transport via injury responses. To investigate the immediate response of long distance axonal transport to EtOH at high resolution, we evaluated in vivo vesicle motility in a filleted, but living Drosophila third instar larvae (Fig 1a) (Dolma et al. 2014, Gunawardena et al. 2013, Reis et al. 2012, Weaver et al. 2013). We utilized the UAS-GAL4 expression system in Drosophila to express GFP-tagged mammalian atrial natriuretic factor (ANF-GFP) specifically in motor neurons using the D42-GAL4 driver. ANF-GFP which enriches in dense core vesicles is processed, localized, and released, as an endogenous neuropeptide when expressed in the nervous system of Drosophila larvae (Rao et al. 2001, Husain & Ewer 2004, Heifetz & Wolfner 2004, Kula et al. 2006, Shakiryanova et al. 2006, Loveall & Deitcher 2010), and exhibits no known deleterious effects on all processes studied.

Figure 1. Experimental setup and Ethanol exposure stimulates anterograde motility of neuropeptide-DCV within larval axons.

(A) Schematic diagram detailing the experimental strategy. Time values on right indicate approximate duration of each step. First, larvae are dissected in saline buffer. This buffer is then replaced with fresh buffer containing EtOH. The start of imaging the region of interest (ROI) is referred to as t₀. Larvae are imaged under constant exposure of buffer or EtOH. (B) Left, schematic diagram of the larval nervous system. Box shows the 90 micron, ROI for live-imaging of vesicle transport. Middle, representative image of ANF-GFP vesicles within larval segmental nerves. Ante, anterograde; Retro, retrograde, identified with respect to the localization of the soma and synapse. Bar = 10 μm. Top, right, a representative kymograph showing 30 sec of ANF-GFP vesicle movement. Bottom, right, individual particle trajectories are traced on the kymograph (colored lines) and analyzed for vesicle motility dynamics. (C) A representative kymographs of ANF-GFP motility after 25mM alcohol exposure at (t₀) and for an additional 10, 20, and 30 min after exposure in EtOH compared to buffer treated control (0mM EtOH). (D) Quantification of anterograde duration-weighted segmental velocities (Ante Velocity) averaged over N = 5 larvae treated with either 0 mM EtOH (gray) or 25 mM EtOH (red). Data shown as mean ± sem. ‘Early’ time window (green shaded area) and ‘Late’ time window (blue shaded area) are defined by pooling four, 30 sec movies of continuous imaging at t₀ + 5 min and t₀ + 25 min, respectively. Note the specific increase in segmental velocities within the ‘Early’ time window. (E) Mean ± sem of pooled data from the Early and Late time windows for treated (red) and untreated (gray) larvae.*p<0.05, Mann-Whitney U-test. Values compared at each time point to control values at corresponding time points.

Drosophila larvae expressing ANF-GFP are carefully dissected on agarose to isolate an intact neuromuscular system, (Fig 1a, b ventral ganglion with radiating segmental nerves innervating cuticle muscles). Once isolated, the preparation is exposed to a freshly prepared EtOH-containing solution before mounting the larvae on a glass coverslip (Fig 1a) and immediately imaged. Movies showing the movement of GFP-positive puncta along individual tracks within the segmental nerves are obtained (Fig 1b, middle). Vesicles can be definitive ly classified as moving anterogradely or retrogradely based on their trajectory relative to the ganglion (where the cell bodies are located) and neuromuscular junction (nerve terminals). In addition, since we start imaging as soon as the EtOH buffer is applied, the time between EtOH exposure and imaging (approx. 5 min) is substantially reduced compared to previous methods (McLane 1987, McLane 1990, McLane et al. 1992). Note that we define time zero (t₀) as the time we start imaging once we have found the region of interest (ROI).

Long-term imaging (30 min) of vesicle motility was performed immediately after finding the ROI in the continued presence of EtOH-containing buffer. Vesicle movement was imaged ~100 microns from the brain region to systematically rule out possible effects of heterogeneity in vesicle motility along the length of the nerve (Fig 1b, boxed region). Larvae were continuously imaged at a spatial resolution of 0.126 micron/pixel. Individual movies contained 30 seconds of live imaging obtained at a 200 msec temporal resolution. Movies were sequentially captured for a total time of 30 minutes for each condition. For each 5 minute time point (0, 5, 10, 15, 20, 25, 30 min), four sequential movies were analyzed for vesicle transport parameters using a custom particle tracker program as previously detailed (Gunawardena et al. 2013, Reis et al. 2012) (Fig 1b, right). Kymographs show both the temporal and spatial changes that EtOH exposure has on ANF-GFP motility compared to control conditions (Fig 1c). We measured duration-weighted segmental velocities as a reporter of vesicle motility. Each vesicle trajectory is divided into segments of continuous unidirectional motion flanked by either a pause in movement or reversal in direction. The duration-weighted segmental velocity of a vesicle represents the sum of its individual segmental velocities weighted by their respective durations and divided by the sum of the vesicle’s segmental durations. Thus, this metric reports the average velocity behavior that vesicles exhibit per time spent moving during the time imaged (Gunawardena et al. 2013, Reis et al. 2012).

In control, untreated larvae, no significant changes in bidirectional ANF motility were qualitatively observed (Fig 1c). Note that ANF motility remains unperturbed and does not significantly vary for the 30 min duration of the assay (Fig 1d) consistent with previous ob servations that the larvae remains viable throughout the assay (Kuznicki & Gunawardena 2010). There was also no difference in stalled vesicles at the ‘early’ (18.89%) and ‘late’ (19.95%) time windows after EtOH exposure (see below). In contrast, an initial stimulatory effect on the anterograde duration weighted segmental velocity was seen with exposure to 25 mM EtOH starting at t₀, which was disrupted 25 min causing a decrease in velocities by 30 min. This concentration of EtOH is physiologically relevant to what a Drosophila neuron may be exposed to because similar concentrations accumulated in EtOH-fed larvae (Robinson et al. 2012) and adult flies exposed to EtOH vapor for 30 min (Scholz et al. 2000, Moore et al. 1998, Devineni & Heberlein 2012). To differentially analyze how neuropeptide-DCV motility was influenced by early responses to EtOH, from delayed or late responses to EtOH after prolonged exposure, we defined two categorical time windows. An ‘early’ window, which represented 5 min after EtOH exposure, and a ‘late’ window, which represented 25 min after EtOH exposure (Fig 1d, shaded areas). For each window, the data for each movie within the defined time frame for each larvae was pooled. Strikingly we observe an early stimulatory effect in neuropeptide-DCV velocity at the early time window and an inhibitory effect at the late time window (Fig 1e). At the early time, ANF anterograde velocities were significantly faster compared to control. In contrast, ANF anterograde velocities were significantly slower at the later time window. While no significant stimulatory or inhibitory effects were seen on ANF retrograde velocities with EtOH exposure as compared to control at the early time point, significant inhibitory effects were seen on retrograde velocities during later time points (Fig S1). Thus, while prolonged EtOH exposure inhibits both anterograde and retrograde ANF velocities, short exposure appears to selectively stimulate anterograde ANF velocities.

Differential sensitives of ‘early’ and ‘late’ ANF motility responses to EtOH

To characterize the differential effects of EtOH on ANF velocities, we determined the sensitivity of several motility parameters at multiple EtOH doses (Fig S2-S6). We monitored the fold change of axonal ANF velocities relative to untreated controls at each time window by normalizing the response in the treatment condition to that in the untreated condition. This data was fit with the Hill function (Goutelle et al. 2008) to extrapolate the half-maximal response dose (EC₅₀) to assess the sensitivity of ANF motility to EtOH. A smaller EC₅₀ value indicates higher sensitivity to EtOH, while a EC₅₀ larger indicates lower sensitivity. We observed that at the early time point, the increase in anterograde velocity was dose-dependent and highly sensitive to EtOH (Fig 2a, left; EC₅₀ = 4.46 mM). In contrast, the decrease in anterograde velocities at the ‘late’ time window was also dose-dependent, but by an order of magnitude less (Fig 2a, right; EC₅₀ = 24.63 mM). To identify whether the increased anterograde velocity could be attributed to an increase in the processivity of kinesin, the anterograde motor, we monitored the run length, which is defined as a segment of continuous motion for a given vesicle flanked by either pauses in motion or reversals in each direction. We found that the anterograde run length also increased at the ‘early’ time window with high sensitivity (Fig 2b, left; EC₅₀ = 2.48 mM). Similarly, run lengths decreased at the ‘late’ time window with lower sensitivity (Fig 2b, right; EC₅₀ = 26.33 mM).

Figure 2. Differential Early and Late sensitivities of anterograde ANF transport to EtOH.

(A) Dose-response of duration-weighted segmental velocity in the ‘early’ time window, left, and the ‘late’ time window, right. All values are normalized to value of the indicated transport parameter recorded in untreated larvae to determine fold change. (B) Dose-response of particle run length in the ‘early’ time window, left, and the ‘late’ time window, right. (C) Dose-response of average particle flux in the ‘early’ time window, left, and the ‘late’ time window, right. (D) Dose-response of percentage of anterograde moving vesicles in the ‘early’ time window, left, and the ‘late’ time window, right. All data were fit with the Hill function(Goutelle et al. 2008) to extrapolate the half-maximal response dose (EC₅₀) to assess the sensitivity of ANF motility to EtOH. A smaller EC₅₀ value indicates higher sensitivity to EtOH, while a EC₅₀ larger indicates lower sensitivity. Note the Early stimulatory effects for all parameters for anterograde ANF motility with EtOH and Late inhibitory effects.

In addition to alterations in ANF-GFP motility, we observed similar changes in vesicle flux. This parameter assesses the number of vesicles crossing a defined boundary along the nerve over a given time. Early responses exhibited greater anterograde flux (Fig 2c, left; EC₅₀ = 2.02 mM) whereas, later responses to prolonged exposure exhibited less antero-grade vesicle flux (Fig 2c, right; EC₅₀ = 16.39 mM). To test whether these changes in anterograde flux could be correlated with changes in the number of anterogradely moving vesicles, we quantified the fractional percentage of ANF-GFP particles moving anterogradely out of the total population of vesicles analyzed. An increase in the fraction of anterogradely moving ANF-GFP vesicles at the ‘early’ time window (Fig 2d, left; EC₅₀ = 1.51 mM) and a de crease in the fraction of anterogradely moving vesicles at the ‘late’ time window (Fig 2d, right; EC₅₀ = 19.21 mM) was seen. However, retrograde vesicle velocities did not exhibit the same biphasic responses to EtOH. Strikingly, retrograde velocities were not stimulated at any EtOH dose tested. However, at higher doses, retrograde velocities decreased (Fig S1). Thus, the effects of EtOH do not globally affect ANF motilities but displays differential sensitivities with early stimulatory effects on anterogradely moving vesicles, and later inhibitory effects on both anterograde and retrograde velocities.

To determine whether the observed effects of EtOH were specific to axonal transport and not due to effects caused by cell death, we assessed cell death in response to EtOH using the TUNEL assay. At all tested concentrations, no observable change in cell death was detected relative to untreated controls (Fig S7a). Since ANF-GFP axonal accumulations or blockages were not detected from live-imaging at any EtOH concentration (Fig 1b, Fig S2-S6), we tested whether the reduction in ANF motility at later time points was the result of blockages formed by other vesicle types physically impeding axonal transport (Gunawardena & Goldstein 2001). Immunostaining of larval segmental nerves with the synaptic vesicle marker, cysteine string protein (CSP), revealed no accumulation of vesicles at low or high doses of EtOH (Fig S7b). Analysis of microtubule (MT) fidelity by expressing Tau-GFP in motor neurons in both treated and untreated larvae, did not show any defects. MTs were continuous and no puncta were observed indicating that there was no loss of structural integrity of MTs at any of the concentrations of EtOH (Fig S8).

To determine whether EtOH affected the number of vesicles entering the axon, we quantified the total number of vesicles within the larval axon in control, low dose (5 mM), and high dose (50 mM) EtOH. In the control condition of 0 mM EtOH, a total of 2061 vesicles were tracked with an average of 412.2 vesicles per larvae over 30 min of recording. We observed no significant difference in the total number of vesicles in larvae treated with 5 mM EtOH. A total of 2102 vesicles were tracked with an average of 420.3 per larvae over 30 min of recording. However, a slightly lesser number of vesicles were observed in 50 mM EtOH; a total of 1734 vesicles were tracked with an average of 289.4 per larvae over 30 min of recording. Since the slight decrease in vesicle number at the higher EtOH dose correlates with a greater number of stalled vesicles, we conclude that vesicle entry into axons is not affected by EtOH within the timeframe of our experiment.

Kinesin-1 mediates the stimulatory effect of EtOH on anterograde neuropeptide-DCV motility

To determine whether the stimulatory increase in anterograde ANF motility at the early time points after EtOH application was due to the anterograde motor kinesin-1, we expressed ANF-GFP in motor neurons in the context of 50% genetic reduction of kinesin-1 using the loss of function mutation of the kinesin light chain (KLC) subunit (klc^8ex94 +/−). We hypothesized that if motor activity was required for the effects of EtOH on ANF motility, then reducing the available kinesin-1 pool should dampen the stimulatory anterograde effects of EtOH. In untreated control larvae, decreasing KLC (ANF-GFP;klc^8ex94 −/+) resulted in decreased ANF-GFP anterograde velocities for the duration of image recording (Fig 3a, top left, Table 1, 2). These observations are consistent with previous observations that kinesin-1 is involved in the transport of ANF-containing dense-core vesicles (Djagaeva et al. 2012). When larvae expressing ANF-GFP in the context of 50% reduction of kinesin (ANF-GFP;klc^8ex94 −/+) were treated with 5 mM EtOH no increase in ANF velocities was detected (Fig 3a, bottom left, and right), in contrast to the stimulation of anterograde velocities seen in larvae expressing ANF-GFP with 5 mM EtOH (Fig 2).Therefore, the observed effects of EtOH on anterograde ANF-GFP motility at early times is likely the result of changes in the activity of kinesin-1 motors.

Figure 3. Stimulatory effect of EtOH requires a functional pool of kinesin-1.

(A) Top, left, A representative movie from a larval segmental nerve expressing ANF-GFP in the context of heterozygous klc imaged over 30 sec of continuous recording within the ‘early’ time window. Ante, anterograde; Retro, retrograde are notes. Note that the kymograph shows stalled vesicles with only a few moving vesicles. Bottom, left, A representative movie from a larval segmental nerve expressing ANF-GFP in the context of heterozygous klc treated with 5 mM EtOH. Note the axonal blockages. The kymograph shows stalled vesicles. Right, top, anterograde duration-weighted segmental velocities quantified over 30 min. Each data point represents the mean ± sem of pooled data from N = 5 larvae. Early and Late time points are noted. Right, bottom, quantification of anterograde duration-weighted segmental velocities pooled from the ‘early’ and ‘late’ time windows. Note that heterozygous reduction of KLC significantly decreases anterograde velocities. However, 5mM EtOH treatment does further decrease anterograde velocities significantly at the early or late time points. (B) Left, top, A representative image of a larval segmental nerve expressing SYT-EGFP imaged over 30 sec of continuous recording within the ‘early’ time window. The representative kymograph shows robust bi-directional motility. Bottom, left, A representative larval segmental nerve expressing SYT-EGFP treated with 5 mM EtOH. The representative kymograph is also shown. Right, quantification of anterograde duration-weighted segmental velocity pooled from the ‘early’ and ‘late’ time windows. Note that there are no significant changes to anterograde SYT vesicle velocities with 5mM EtOH. (C) Left, top, A representative image of a larval segmental nerve expressing htt15Q-mRFP imaged over 30 sec of continuous recording within the ‘early’ time window. The kymographs show robust bi-directional motility of HTT. Bottom, left, A representative image of a larval segmental nerve expressing htt15Q-mRFP treated with 5 mM EtOH. Note that the representative kymograph still shows bi-directional motility. Right, quantification of anterograde duration-weighted segmental velocity pooled from the ‘early’ and ‘late’ time windows. Note that there are no significant changes to anterograde HTT vesicle velocities with 5mM EtOH. Bar = 10 μm. *p<0.05, using Mann-Whitney U-test.

Table 1.

Summary of Anterograde Motility Parameters

Condition	Duration-weighted Segmental Velocity (μm/s)		Percent Vesicle Population (%)		Run Length (μm)		Flux (event/s)
	Early	Late	Early	Late	Early	Late	Early	Late
ANF-GFP (0 mM EtOH)	0.849 ± 0.092	0.898 ± 0.097	13.425 ± 4.125	16.844 ± 4.340	5.614 ± 1.635	5.311 ± 1.852	0.759 ± 0.245	0.877 ± 0.327
ANF-GFP (1 mM EtOH)	0.806 ± 0.078	1.225 ± 0.125	21.899 ± 1.578	49.026 ± 3.471	5.935 ± 1.337	12.115 ± 1.829	1.301 ± 0.229	2.276 ± 0.241
ANF-GFP (5 mM EtOH)	1.145 ± 0.126	1.194 ± 0.115	31.642 ± 4.498	37.662 ± 4.054	10.644 ± 0.859	11.784 ± 1.744	1.965 ± 0.175	2.216 ± 0.234
ANF-GFP (10 mM EtOH)	1.322 ± 0.098	1.388 ± 0.110	32.829 ± 4.550	37.494 ± 4.803	11.720 ± 1.383	11.052 ± 1.435	2.200 ± 0.185	2.153 ± 0.182
ANF-GFP (25 mM EtOH)	1.348 ± 0.110	0.766 ± 0.077	35.831 ± 2.769	21.716 ± 4.595	11.200 ± 0.571	8.769 ± 1.262	2.231 ± 0.192	0.236 ± 0.056
ANF-GFP (50 mM EtOH)	1.312 ± 0.139	0.442 ± 0.061	34.912 ± 2.384	4.220 ± 1.356	10.851 ± 0.450	4.629 ± 0.735	2.089 ± 0.232	0.093 ± 0.073
ANF-GFP; klc8ex94 (0 mM EtOH)	0.530 ± 0.106	0.567 ± 0.074	-	-	-	-	-	-
ANF-GFP; klc8ex94 (5 mM EtOH)	0.609 ± 0.090	0.663 ± 0.099	-	-	-	-	-	-
ANF-GFP; sggA81T (0 mM EtOH)	1.555 ± 0.135	1.199 ± 0.136	-	-	-	-	-	-
ANF-GFP; sggA81T (5 mM EtOH)	1.651 ± 0.099	1.459 ± 0.137	-	-	-	-	-	-
ANF-GFP (Li-fed, 0 mM EtOH)	1.342 ± 0.123	1.079 ± 0.122	-	-	-	-	-	-
ANF-GFP (Li-fed, 5 mM EtOH)	1.301 ± 0.158	1.240 ± 0.133	-	-	-	-	-	-
ANF-GFP (wort-fed, 0 mM EtOH)	0.418 ± 0.091	0.266 ± 0.064	-	-	-	-	-	-
ANF-GFP (wort-fed, 5 mM EtOH)	0.734 ± 0.083	0.705 ± 0.075	-	-	-	-	-	-
SYN-eGFP (0 mM EtOH)	0.740 ± 0.079	0.769 ± 0.071	-	-	-	-	-	-
SYN-eGFP (5 mM EtOH)	0.750 ± 0.065	0.523 ± 0.038	-	-	-	-	-	-
Htt15Q- mRFP (0 mM EtOH)	0.804 ± 0.082	0.759 ± 0.070	-	-	-	-	-	-
Htt15Q- mRFP (5 mM EtOH)	0.835 ± 0.080	0.855 ± 0.091	-	-	-	-	-	-
ANF-GFP (25 mM MeOH)	1.164 ± 0.110	0.732 ± 0.063	-	-	-	-	-	-
ANF-GFP (25 mM PrOH)	0.740 ± 0.059	0.797 ± 0.082	-	-	-	-	-	-

Data represent mean ± sem of N = 5 larvae.

Table 2.

Summary of Retrograde Motility Parameters

Condition	Duration-weighted Segmental Velocity* (μm/s)		Percent Vesicle Population* (%)		Run Length* (μm)		Flux* (event/s)
	Early	Late	Early	Late	Early	Late	Early	Late
ANF-GFP (0 mM EtOH)	0.809 ± 0.053	0.759 ± 0.058	26.978 ± 5.120	25.498 ± 5.775	5.808 ± 1.287	4.296 ± 1.037	1.209 ± 0.239	0.968 ± 0.215
ANF-GFP (1 mM EtOH)	0.708 ± 0.039	0.698 ± 0.087	21.928 ± 4.259	11.222 ± 5.032	6.513 ± 1.076	5.895 ± 0.807	0.793 ± 0.115	0.967 ± 0.224
ANF-GFP (5 mM EtOH)	0.848 ± 0.153	0.898 ± 0.170	26.161 ± 5.099	25.154 ± 5.820	6.731 ± 1.229	5.057 ± 1.074	0.822 ± 0.267	0.913 ± 0.311
ANF-GFP (10 mM EtOH)	0.828 ± 0.159	0.898 ± 0.170	30.021 ± 6.688	20.075 ± 4.557	5.410 ± 0.853	5.507 ± 0.865	1.020 ± 0.288	0.780 ± 0.254
ANF-GFP (25 mM EtOH)	0.813 ± 0.157	0.502 ± 0.006	27.200 ± 5.731	3.649 ± 1.410	7.333 ± 1.404	5.830 ± 1.101	1.074 ± 0.248	1.084 ± 0.285
ANF-GFP (50 mM EtOH)	0.902 ± 0.163	0.542 ± 0.010	18.016 ± 4.292	11.177 ± 1.239	6.597 ± 0.865	1.851 ± 0.173	0.790 ± 0.211	0.690 ± 0.195
ANF-GFP; klc8ex94(0 mM EtOH)	0.414 ± 0.088	0.586 ± 0.113	-	-	-	-	-	-
ANF-GFP; klc8ex94(5 mM EtOH)	0.554 ± 0.095	0.503 ± 0.103	-	-	-	-	-	-
ANF-GFP; sggA81T(0 mM EtOH)	1.555 ± 0.135	1.199 ± 0.136	-	-	-	-	-	-
ANF-GFP; sggA81T(5 mM EtOH)	1.651 ± 0.099	1.459 ± 0.137	-	-	-	-	-	-
ANF-GFP (Li-fed, 0 mM EtOH)	0.910 ± 0.213	0.872 ± 0.187	-	-	-	-	-	-
ANF-GFP (Li-fed, 5 mM EtOH)	0.806 ± 0.154	0.882 ± 0.147	-	-	-	-	-	-
ANF-GFP (wort-fed, 0 mM EtOH)	0.455 ± 0.064	0.425 ± 0.040	-	-	-	-	-	-
ANF-GFP (wort-fed, 5 mM EtOH)	0.551 ± 0.084	0.420 ± 0.085	-	-	-	-	-	-
SYN-eGFP (0 mM EtOH)	0.701 ± 0.063	0.723 ± 0.088	-	-	-	-	-	-
SYN-eGFP (5 mM EtOH)	0.698 ± 0.102	0.417 ± 0.042	-	-	-	-	-	-
Htt15Q- mRFP (0 mM EtOH)	0.751 ± 0.097	0.710 ± 0.067	-	-	-	-	-	-
Htt15Q- mRFP (5 mM EtOH)	0.768 ± 0.045	0.777 ± 0.054	-	-	-	-	-	-
ANF-GFP (25 mM MeOH)	0.727 ± 0.059	0.642 ± 0.112	-	-	-	-	-	-
ANF-GFP (25 mM PrOH)	0.703 ± 0.020	0.687 ± 0.077	-	-	-	-	-	-

Data represent mean ± sem of N = 5 larvae.

In addition to kinesin-1 (Djagaeva et al. 2012), ANF-dense core vesicles are also transported by kinesin-3 (unc-104) (Barkus et al. 2008, Zahn et al. 2004). To test whether the observed early effect of EtOH was specific to kinesin-1, we measured the effect of EtOH on eGFP-tagged synaptotagmin (SYT-EGFP), a synaptic vesicle (SV) protein transported by kinesin-3 (Pack-Chung et al. 2007). Previous work showed that synaptic vesicle trafficking and recycling was impaired in response to high EtOH concentrations (Gioia et al. 2017, McLane et al. 1992) and synpatotagmin expression was increased in response to high EtOH (Varodayan et al. 2011). However, whether EtOH exposure has an early effect on synaptic vesicles was unknown. We found that EtOH incubation failed to elicit the same early stimulatory effect on anterograde SV velocities (Fig 3b) as observed for DCVs. However, a modest but significant inhibitory effect on anterograde SV velocities was observed at the later time point. EtOH incubation also failed to elicit an early stimulatory effect on the anterograde velocities of huntingtin (HTT)-containing vesicles (Fig 3c). Thus, the stimulatory effect of EtOH on anterograde ANF-GFP velocities at the early time point appears to be specific to neuropeptide-DCV and is likely due to altered kinesin-1 activity.

GSK-3β modulates EtOH mediated neuropeptide-DCV motility changes

EtOH has been shown to regulate AKT activity both by direct (Huang & Kim 2012) and indirect means (Zeng et al. 2012, Bjork et al. 2010). One protein that is thought to regulate anterograde motility of vesicles through phosphorylation of AKT is huntingtin (Colin et al. 2008). This study also demonstrated that huntingtin can co-sediment with kinesin-1. Biochemical evidence has shown that huntingtin-associated protein-1 can directly interact with kinesin light chain (McGuire et al. 2006) and a genetic interaction between huntingtin and kinesin-1 was also seen (Gunawardena et al. 2003), indicating that huntingtin likely uses kinesin-1 for anterograde motility. To test whether huntingtin plays a role in the motility of neuropeptide-DCV we co-expressed a nonpathogenic mRFP-tagged huntingtin, HTT15Q-mRFP and ANF-GFP in the same larvae and imaged their comotility using simultaneous dual-view live-imaging (White et al. 2015). We found that HTT and ANF do not move together on the same vesicle in vivo. HTT and ANF were also not present on stationary vesicles (Fig S9). Therefore, huntingtin does not appear to play a role in mediating EtOH’s stimulatory effect on neuropeptide-DCV motility at early time points.

There is mounting evidence that glycogen synthase kinase 3β (GSK-3β) plays an important regulatory role during axonal transport (Dolma et al. 2014, Weaver et al. 2013). We previously showed that constitutively active GSK-3β resulted in increased binding of kinesin-1 to membranes containing synaptic proteins, while reduction of GSK-3β resulted in decreased binding of kinesin-1 to these membranes (Dolma et al. 2014). Increased GSK-3β-mediated binding of kinesin to synaptic membranes correlated with axonal blockages, while decreased GSK-3β-mediated binding of kinesin stimulated anterograde motility. Therefore, GSK-3β could play a key role in the initial stimulatory increases we observe in neuropeptide-DCV anterograde motility in response to EtOH. To test this proposal we co-expressed ANF-GFP and a dominant negative GSK-3β (sgg^A81T) in motor neurons using the D42-GAL4 driver. Sgg^A81T has been previously proposed to act as a null kinase similar to a loss of function mutation of GSK-3β (Bourouis 2002). Strikingly we found that anterograde ANF velocities were increased at both the early and late time points in larvae co-expressing ANF-GFP and sgg^A81T compared to control larvae expressing only ANF-GFP (Fig 4a). Similar to what was seen previously seen with APP (Weaver et al 2013); we note that both anterograde and retrograde bidirectional velocities of ANF vesicles were increased, however we only focused on anterograde velocities in the context of EtOH. EtOH treatment of larvae co-expressing ANF-GFP with the dominant negative GSK-3β did not further stimulate early anterograde ANF-GFP vesicle velocities (Fig 4a, top right). No significant changes were observed at both the ‘early’ or the ‘late’ time points compared to larvae not exposed to EtOH (Fig 4a, bottom right). Thus, a pool of active GSK-3β is likely required to regulate and stimulate ANF-GFP axonal motility at the early time point.

Figure 4. GSK-3β mediates the early stimulatory effect of EtOH.

(A) Top, left, A representative image from a larval segmental nerve co-expressing ANF-GFP with the dominant negative sgg^A81T imaged over 30 sec of continuous recording within the ‘early’ time window. Note that the kymograph shows robust bi-directional motility. Bottom, left, A representative image from a larval segmental nerve expressing co-expressing ANF-GFP with dominant negative sgg^A81T treated with 5 mM EtOH imaged over 30 sec of continuous recording within the ‘early’ time window. Note that the kymograph still shows robust bi-directional motility. Right, top, anterograde duration-weighted segmental velocity quantified over 30 min. Each data point represents the mean ± sem of pooled data from N = 5 larvae. Early and Late time points are noted. Right, bottom, quantification of anterograde duration-weighted segmental velocities pooled from the ‘early’ and ‘late’ time windows for untreated larvae expressing ANF-GFP with endogenous levels of sgg (Cntrl, black), untreated larvae co-expressing ANF-GFP with sgg^A81T (gray), and 5mM EtOH treated larvae co-expressing ANF-GFP with sgg^A81T (red). Note that expression of sgg^A81T stimulates early and late anterograde vesicle velocities. However, 5mM EtOH treatment does not further stimulate anterograde velocities. (B) Top, A representative kymograph from a larvae fed with LiCl expressing ANF-GFP. Middle, A representative kymograph from a larvae fed with LiCl expressing ANF-GFP and treated with 5mM EtOH. Bottom, quantification of anterograde duration-weighted segmental velocity pooled from the ‘early’ and ‘late’ time windows for untreated larvae reared on normal food (Cntrl, black), untreated larvae fed with LiCl (gray), and LiCl fed larvae treated with EtOH (red). (C) Top, A representative kymograph from a larvae fed with ANF-GFP movement from larvae fed with wortmannin. Middle, the representative kymograph showed stalled ANF-GFP vesicles and only few vesicles in motion. Bottom, quantification of the anterograde duration-weighted segmental velocities pooled from the ‘early’ and ‘late’ time windows for larvae expressing ANF—GFP reared on normal food (Cntrl, black), larvae expressing ANF-GFP fed with wortmannin (gray), and EtOH-treated larvae expressing ANF-GFP fed with wortmannin (red). Bar = 10 μm. *p<0.05, using the Mann-Whitney U-test.

Since GSK-3β plays vital roles in neuronal development, the effects we observe in living larvae could be linked to general effects due to loss of GSK-3β. To rule out this possibility we pharmacologically inhibited GSK-3β using the inhibitor lithium. We reared ANF-GFP expressing larvae on food containing 10 mM lithium chloride (LiCl) until larvae reached wandering third-instar stage. Consistent with previous in vivo reports (Dolma et al. 2014, Weaver et al. 2013), we found that LiCl-fed larvae resulted in faster ANF-GFP particle velocities (Fig 4b), similar to larvae co-expressing ANF-GFP and a dominant negative GSK-3β (Fig 4a). When we applied EtOH to LiCl-fed ANF-GFP expressing larvae, no additional stimulation of anterograde vesicle velocities was observed. Since EtOH treatment failed to elicit additional enhancement of vesicle motility in LiCl-fed larvae similar to larvae expressing inactive GSK-3β, we propose that the population of active GSK-3β was insufficient for EtOH to mediate any further enhancement of ANF-GFP motility. Additionally, the lack of an EtOH effect with dominant negative GSK-3β is likely not due to early developmental requirements of GSK-3β.

To further investigate the mechanism by which GSK-3β may function, we pharmacologically elevated active levels of GSK-3β using wortmannin, which increases active GSK-3β via the inhibition of PI3K, a negative regulator of GSK-3β. Previous work has shown that wortmannin results in increased levels of active GSK-3β in both Drosophila (Parisi et al. 2011) and mammals (Zhu et al. 2007, Liu et al. 2003, Fang et al. 2000). Since genetic overexpression of GSK-3β resulted in vesicle blockages and impaired motility within larval nerves (Dolma et al. 2014), we chose to pharmacologically manipulate GSK-3β using wortmannin which constitutes a less chronic alteration of endogenous GSK-3β levels. We found that larvae reared on 100 nM wortmannin led to an observable decrease in ANF-GFP vesicle velocities (Fig 4c), as expected (Parisi et al. 2011, Zhu et al. 2007). Interestingly, when these wortmanninfed larvae were treated with EtOH, a modest increase in anterograde vesicle velocities was seen (Fig 4c). However, although the velocities of moving ANF-GFP vesicles were increased in these wortmanninfed larvae upon EtOH treatment, the total fraction of moving vesicles remained decreased (Table 1, 2). Because wortmannin acts by inhibiting PI3K, and given substantial evidence suggesting that GSK-3β is the primary mediator regulating vesicle motility (Dolma et al. 2014, Weaver et al. 2013) and that EtOH modulates levels of active GSK-3β and AKT (Fang et al. 2000, Liu et al. 2003, Parisi et al. 2011), we predict that wortmannin likely increases the levels of active GSK-3β by inhibiting PI3K/AKT. Therefore, the rescue of wortmannin mediated effects by EtOH is consistent with EtOH acting downstream of PI3K, perhaps at the level of AKT or GSK-3β.

EtOH’s early stimulatory effect is conserved for other alcohols native to the environment of Drosophila

In addition to EtOH, methanol (MeOH) is among the most common alcohols found in fermenting fruits, which is the native environment of the developing Drosophila larvae. Drosophila has evolved distinct metabolic pathways for MeOH consumption (Wang et al. 2013). However, propanol (PrOH) does not constitute a large fraction of the alcohols in this environment. To test whether the effect we observed for EtOH can be generalized to other dietary alcohols, we compared the effects of MeOH and PrOH exposure on ANF-dense core motility. Exposure to 25 mM MeOH lead to a significant enhancement in vesicle velocity within the ‘early’ time, with no change observed at the ‘late’ time point (Fig 5a). In contrast, 25 mM PrOH did not modulate ANF-GFP motility at either time points (Fig 5b). Thus, the observed response of ANF DCV transport likely represents an endogenous response to dietary alcohols Drosophila larvae would encounter in its natural habitat.

Figure 5. Effect of ethanol is conserved for other Drosophila dietary alcohols.

(A) Left, a representative kymograph from a 30 sec continuous movie from the ‘early’ time window (top) and the ‘late’ (bottom) time window. Ante, anterograde; Retro, retrograde. Right, quantification of anterograde duration-weighted segmental velocities pooled from the ‘early’ and ‘late’ time windows for untreated (gray) and 25 mM MeOH-treated larvae (red). Note that a significant increase in anterograde vesicle velocities is seen at the early time with 25mM MeOH treatment. (B) Left, A representative kymograph from a 30 sec continuous movie from the ‘early’ time window (top) and the ‘late’ (bottom) time window. Right, quantification of anterograde duration-weighted segmental velocities pooled from the ‘early’ and ‘late’ time windows for untreated (gray) and 25 mM PrOH-treated larvae (red). Note that no significant changes are seen for either the early or late time points. Bar = 10 μm. *p<0.05, using the Mann-Whitney U-test.

Discussion

In this study, we provide novel evidence that EtOH has multiple functional effects on neuropeptide-DCV motility in vivo. Initially, low doses of EtOH stimulate the anterograde movement of DCV, in part, through modulating the activity of GSK-3β on kinesin-1 function. In contrast, high doses of EtOH have a global inhibitory effect on both anterograde and retrograde DCV motility.

Physiological and behavioral implications of EtOH

Our study has several potential physiological implications. Adult Drosophila flies have been extensively used in the study of EtOH-related behaviors (Guarnieri & Heberlein 2003, Kaun et al. 2012). However, the pathways that are affected to mediate these natural responses to EtOH are not completely known. Since the axonal transport pathway is essential for viability, perhaps this pathway is sensitive to EtOH. The use of larvae has proven to be particularly amenable to the study of axonal transport in vivo (Dolma et al. 2014, Gunawardena et al. 2013, Reis et al. 2012, Weaver et al. 2013). However, unlike adult Drosophila flies, which undergo sedative effects of EtOH (Wolf et al. 2002), larval locomotion was unaffected by EtOH (Robinson et al. 2012), but low doses of EtOH impaired olfactory associative learning (Robinson et al. 2012). Since the natural habitat of larvae consists of fermented fruits, environmental EtOH concentration gradients likely play significant roles in larval foraging behaviors (McKenzie 1979). Indeed, our observation that low doses of EtOH stimulated the anterograde motility of neuropeptide-DCVs (Fig 2; Fig S2-6) is strikingly consistent with previous observations that enhanced neuropeptide signaling prompted feeding behaviors (Wu et al. 2005). Therefore, such a physiological response to low EtOH concentrations could serve to stimulate larval crawling toward food sources in response to environmental EtOH concentration gradients. As EtOH concentrations increase perhaps in over-fermented potentially deleterious food sources, decreased neuropeptide motility and signaling could inhibit larval locomotion and larval foraging. Consistent with this proposal, we found that high concentrations of EtOH inhibit neuropeptide-DCV motility (Fig 2). Thus, the physiological effects we observe for EtOH could likely be part of the intricate behavioral dynamics Drosophila larvae use to navigate environmental gradients of many sensory cues (odors, tastes).

Although in Drosophila ANF is processed, localized, and released, as an endogenous neuropeptide and localizes to neuropeptide-containing DCVs (Barkus et al. 2008, Rao et al. 2001, Xia et al. 2009), endogenous Drosophila neuropeptides, such as neuropeptide F, was shown to be critical in regulating foraging behaviors in Drosophila (Wu et al. 2005). Since neuropeptides exclusively regulate neural circuits involved in metabolism, stress, and energy homeostasis, modulation of stress by peptidergic neurons during periods of starvation can stimulate food-searching behaviors (Nassel & Wegener 2011). Indeed, type II and III synaptic boutons within the Drosophila larval NMJ are peptidergic (Gorczyca et al. 1993, Monastirioti et al. 1995), suggesting that neuropeptides are critical in locomotion. Therefore, our observation of enhanced neuropeptide-DCV trafficking in motor neurons in response to dietary alcohols may also provide evidence for a link between stress and foraging.

Drosophila also responds to neural stress and injury perhaps via the regulation of neuropeptides. Live-imaging of immobilized larvae revealed that nerve crush increased the accumulation of anterograde neuropeptide-DCV density in aCC and RP2 motor neurons proximal to the site of injury (Barkus et al. 2008, Ghannad-Rezaie et al. 2012). In addition, nerve crush led to the accumulation of prominent anterograde synaptic cargos proximal to the site of injury: vesicular glutamate transporter, a protein critical for glutamate recycling and NMJ development, and APP, a kinesin-1-transported synaptic vesicle marker. Such anterograde accumulations caused increased expression of JNK phosphatase, puckered, and phosphorylated JNK in the cell bodies of injured motor neurons. Functional dynein was required for injury-induced signaling to the cell body (Xiong et al. 2010, Cavalli et al. 2005). The role of axonal vesicle motility in mediating cellular stress responses is underscored by observations that UV-induced neural stress disrupts APP vesicle motility within axons (Almenar-Queralt et al. 2014). In addition, selective perturbation of axonal transport by deleting KLC1 gradually led to an increase in cellular stress as determined by JNK activation (Falzone et al. 2009). Since EtOH exposure has been shown to increase ROS production (Dey & Cederbaum 2006, Hoek & Pastorino 2002) via JNK activation (Kapfhamer et al. 2012, Lam et al. 2010), perhaps neural responses to exogenous stressors is also inextricably linked to the axonal transport pathway.

Mechanisms of EtOH effects on neuropeptide-DCV transport within axons

We found that inhibition of GSK-3β stimulates neuropeptide-DCV motility within axons (Fig 4a, b) consistent with previous studies (Dolma et al. 2014, Weaver et al. 2013). However, a low dose of EtOH did not further enhance vesicle motility when GSK-3β was inactive. Thus, the stimulatory effect of EtOH on neuropeptide-DCV movement is likely due to reducing pools of active GSK-3β. In contrast, inhibition of PI3K/AKT, which causes increases in active GSK-3β, resulted in reduced ANF vesicle velocities (Fig 4c). Strikingly, a low dose of EtOH was able to partially rescue wortmannin-induced decreases in neuropeptide-DCV motility, suggesting that EtOH has a negative effect on GSK-3β-mediated roles in vesicle transport. This is consistent with mammalian studies showing that EtOH can both directly and indirectly activate AKT. At least two possible models could likely explain our experimental results. While these models may not be mutually exclusive, in the first model, EtOH acts upstream of GSK-3β to modulate transport by PI3K/AKT activation which leads to the accumulation of inactive GSK-3β. Alternatively, EtOH directly inhibits GSK-3β. Therefore, while lithium likely inhibits a saturating fraction of endogenous GSK-3β prior to EtOH exposure, no further inactivation by EtOH is possible under these conditions. In this model, the partial rescue of wortmannin-induced decreases in neuropeptide-DCV motility is likely the result of EtOH-induced inactivation of a fraction of the active pool of GSK-3β.

Furthermore, these GSK-3β- mediated effects appear to regulate the anterograde motor kinesin-1. Indeed, previous work has shown that ANF-GFP likely uses both kinesin-1 (Djagaeva et al. 2012) and kinesin-3 (Barkus et al. 2008). We found that kinesin-1 reduction impaired the ability of EtOH to modulate ANF-GFP motility, suggesting that at least one of these molecular motors is influenced by GSK-3β (Fig 3a). Intriguingly, both Drosophila kinesin-1 subunits, kinesin heavy chain (KHC) and KLC have putative GSK-3β-phosphorylation sites, and could likely be the mechanism of how active GSK-3β regulates anterograde motility. Therefore, further studies are needed to fully understand the exact molecular mechanism of GSK-3β-mediated EtOH effects on DCV motility within axons.

In conclusion, we provide the first evidence that low EtOH concentrations can stimulate the transport of neuropeptide-containing DCVs in Drosophila larvae. Signaling via GSK-3β appears to mediate the effects of EtOH on neuropeptide-DCV movement. In mammalian models, GSK-3β activity is required for the neurotoxic effects of EtOH and is ameliorated by feeding LiCl (Liu et al. 2009). Similarly, activation of PI3K by insulin removes the AKT-dependent inhibition of GSK-3β (de la Monte et al. 2000) consistent with the role of PI3K/AKT signaling in alcohol-induced fatty liver disease (Zeng et al. 2012). Thus, PI3K/AKT upregulation may serve to counteract cell stress signaling. Similarly, in humans, the hypothalamic-pituitary-adrenal axis is linked to addictive behavior. This pathway is tightly regulated by ANF and altered serum levels have been associated with craving, withdrawal intensity, and relapse (Kiefer & Wiedemann 2004, Kovacs 2003). Our novel observation of enhanced anterograde neuropeptide-DCV motility in response to EtOH could also provide a mechanism for altered ANF serum levels. Further work is needed to validate this phenomenon in mammalian systems and to understand its role in alcohol-related disorders.

Abbreviations

DCV

Dense core vesicles

ANF-GFP

Green fluorescent protein-tagged atrial natriuretic factor

SYT-eGFP

Enhanced green fluorescent protein-tagged synaptotagmin

Htt

Huntingtin

GSK-3

Glycogen synthase kinase 3

EtOH

Ethanol

MeOH

Methanol

PrOH

Propanol

APP

Amyloid precursor protein