Canonical receptors for endocannabinoids and phytocannabinoids such as those present in mammals are absent in the Drosophila genome (McPartland et al., 2006a,b). However, the fruit fly possesses its own endogenous cannabinoids (Khaliullina et al., 2015). Therefore, we hypothesized that this organism may be responsive to phytocannabinoids.

We developed a system for administrating vaporized substances to intact Drosophila flies. This method has been successfully extended to other models related to diseases and the consumption of substances, such as tobacco (Balcazar et al., 2017). Plant material was provided by a local NGO (www.mamacultiva.org), and consists of a cannabis strain that is currently used in the treatment of patients with refractory epilepsy. First, we determined the proportion of THC, CBD and CBN by gas chromatography coupled to mass spectrometry (Table S1). The marijuana strain used in the present study is rich in THC (114: 1 THC/CBD). Herbal cannabis was placed into a commercial vaporizer (Ascent by DaVinci™) connected to a custom device as shown in Fig. 1A. Temperature was set to 188°C (370°F). At this temperature, most cannabinoids were present in the air (Table S2). By controlling a three-way stopcock manifold, vapor was suctioned with the aid of a syringe and transferred into a vial containing the flies to be treated. Temperature inside the vial receiving the air coming from the vaporizer was measured with a temperature probe in order to verify that the vial's temperature (room temperature) did not change during exposition (data not shown). Flies were exposed to the resulting air containing cannabinoid compounds, which were incorporated through and distributed by the tracheal ramifications of the flies, eventually reaching the hemolymph. Although it is a passive process, we defined it as inhalation.

Survival and mortality studies were carried out to test if chronic administration of cannabis is a feasible approach to analyze heart function. In this aspect, consumption of cannabinoids along the entire life did not affect lifespan of flies (Fig. 1C).

Cardiac function was assessed in semi-intact heart preparations as we have previously described (Santalla et al., 2014; Santalla et al., 2016). A reporter system that expresses the green fluorescent protein (GFP) in cardiomyocytes and pericardial cells was used to track the movement of the individual cells during each cardiac cycle (systole and diastole). Individuals resulting from the cross between Canton-S and the homozygous handC-GFP strain harbored one copy of the reporter system, enough to follow the movement of the heart wall. Mechanical parameters were defined as previously described (Balcazar et al., 2018). Fig. 2A shows representative mechanical recordings of cardiac performance.

With the aim to assess the cardiac performance, flies that emerged from the puparium were assigned to different groups: individuals exposed to marijuana for 6–8 days and individuals exposed for 11–13 days. Other groups of flies received only ambient air (control) for the same periods of time as the treated flies. Cannabis treatment applied for 6–8 or 11–13 days correspond to 13% or 30%, respectively, of the flies' average total lifespan (Fig. 1B). Therefore, we consider these periods of exposure to cannabis as chronic treatments.

Short exposure to cannabis (6–8 days) induced an increase in arrhythmicity index (Fig. 2B), which indicates that this phenomenon is an early and transitory effect because it was not observed in individuals exposed to the drug for 11–13 days. An early increment of arrhythmicity index may be expected because heart performance should adapt to a stimulation induced by the compounds. This arrhythmic pattern might be responsible for the early incremented mortality in the treated flies compared to non-exposed individuals (Fig. 1C). Flies under the short period of treatment (5 days), but analyzed 5 days after interruption of cannabis exposure, did not exhibit heart-rate variability (Fig. S1A). In all groups, inhaled marijuana did not modify the heart rate as a function of the exposure time (Fig. 2D and Fig. S1C). Studies on patients show that exposure to smoked marijuana induces abnormalities on cardiac rhythm (Akins and Awdeh, 1981; Fisher et al., 2005; Johnson and Domino, 1971; Kosior et al., 2001; Rezkalla and Kloner, 2018; Singh, 2000). Early arrhythmogenesis is observed, for example, only during the onset of reperfusion in mammal heart after a period of ischemia (Vittone et al., 2002). It is feasible that arrhythmias of the Drosophila heart in response to cannabis treatment represents one of the early consequences of adaptive mechanisms that the heart employs for reaching an equilibrium or steady state under cannabis consumption.

We then explored the effect of different inhalation times on cardiac contractility (inotropism) evaluated by assessing cardiac wall displacement. Fig. 2C shows that wall displacement increased significantly after prolonged inhalation of Cannabis sp. (11–13 days), but not in response to short-term treatment (6–8 days). Moreover, heart wall displacement was not incremented in flies exposed to ambient air and also not in a group of flies exposed for a short period (5 days) and then grown for an additional 5 days after treatment interruption (Fig. S1B). Taken together, these results suggest that the increase in contractility observed in the 11–13 days-treated flies is a late effect induced by prolonged exposure to marijuana.

Calcium signaling is essential for the cardiac contraction (systole) and relaxation (diastole) that occur when electrical activity is followed by a mechanical action in the cardiomyocytes. The link between the electrical and mechanical activity is known as excitation-contraction coupling (ECC), which depends on the intracellular Ca²⁺ concentration (Ca²⁺ transient). When an action potential depolarizes the cell membrane, the intracellular Ca²⁺ concentration increases due to a massive Ca²⁺ release from the sarcoplasmic reticulum (SR) into the cytosol, where Ca²⁺ binds to the myofilaments, thus inducing its contraction. For relaxation to occur, Ca²⁺ is quickly removed from the cytosol mainly by the SR Ca²⁺ATPase pump (SERCA) and the Na⁺/ Ca²⁺ exchanger (NCX). Slow removal systems like mitochondrial Ca²⁺ transporter and the sarcolemmal Ca²⁺ATPase pump (PMCA) contribute to a lesser extent to the reduction of Ca²⁺_i (Desai-Shah et al., 2010). These mechanisms govern cardiac contractility in Drosophila and mammals (Bers, 2001). Therefore, we studied whether mishandling of Ca²⁺ is involved in the cardiac phenotypes induced by the exposure to marijuana.

Ten-day-old flies bearing one copy of the genetically-encoded Ca²⁺ reporter (GCaMP3) in the heart, were vaporized with cannabis or ambient air (control). Cytosolic Ca²⁺ transients were assessed and SR Ca²⁺ content was estimated by application of a 10 mM caffeine pulse applied onto the semi-intact heart preparation (Fig. 3A). This pharmacological intervention, widely used in mammalian studies, opens the SR Ca²⁺ channels (ryanodine receptors, RyR) allowing SR- Ca²⁺ release. The amplitude of caffeine-induced Ca²⁺ transient estimates the SR- Ca²⁺ content. Comparison between pre-caffeine and caffeine-induced Ca²⁺ transients' decay, allows calculating the relative importance of each rapid removal system during relaxation. The results obtained from the analyses of the flies that consumed cannabis were compared with those obtained from flies which received air only.

We observed an increment in cytosolic Ca²⁺ transients in marijuana-treated flies (Fig. 3B), which was associated to incremented contractility in response to prolonged exposure to marijuana (Fig. 2C). Unexpectedly, the observed increment in the Ca²⁺ transient cannot be attributed to an increase in either SR Ca²⁺ load or fractional Ca²⁺ release because these two parameters were not significantly modified by marijuana inhalation (Fig. 3C,D). Moreover, estimated activity of the main proteins involved in Ca²⁺ handling remain unchanged (Fig. 3E,F). The reason for these somewhat unexpected results is not apparent to us. A possible explanation relates to myofilaments and Ca²⁺ interaction. Myofilament calcium sensitivity results from the balance between Ca²⁺ association and dissociation rates (k_on and k_off) with the components of troponin/tropomyosin complex. This complex modulates actin/myosin interaction and therefore the contractile force produced by the cardiac muscle (Chung et al., 2016). In this context, possible reduced association rates and/or incremented dissociation rates could be contributing to an increase in cytosolic Ca²⁺ transient. Marijuana inhalation might produce a simultaneous decrease in myofilament Ca²⁺ sensitivity, which would increase the intracellular Ca²⁺ transient – and consequently an increment in the activity of cross-bridges – which would be responsible for the observed increase in cardiac shortening. Another possibility to be considered is that marijuana produces a prolongation of the action potential that allows a higher Ca²⁺ influx that can be extruded from the cell through the NCX. Further experiments are needed to explore these possibilities. Altogether, effects of chronic cannabis exposure on heart performance can be categorized as follows: 1, early changes like heart-rate variability; and 2, alterations caused by prolonged exposure to cannabis like the increment in Drosophila heart contractility (Fig. S2).

Finally, one additional set of experiments was carried out to test whether cannabis treatment induced relevant alterations of behavior, e.g. movement behavior, that could affect the heart performance of Drosophila. Three images were captured from video recordings comparing flies treated with air or marijuana. To assess individual activities, we counted the number of individuals in the top and the bottom half of vials utilized for treatments. We observed that flies exposed to cannabis crowded more frequently close to the top of the vials than flies that received air (Fig. S3). Of note, experiments using CP-55,940 – a synthetic cannabinoid – showed restored negative geotaxis activity (i.e. climbing capability) of flies that were exposed to paraquat (PQ) toxicity via a receptor-independent mechanism. Antioxidant activity of CP-55,940 and preservation of the dopaminergic pathway could be the cause of these effects (Jimenez-Del-Rio et al., 2008). In our conditions, we observed that flies exposed to marijuana crowed at the top of the vial. Although the treated group exhibited a significant difference compared to air-subjected individuals, we cannot deduce whether this behavior is due to cannabinoid-induced negative geotaxis or associated to alterations in flight muscle activity or other metabolic adjustments from the preference triggered by the odor accumulated in the top of the vial. This aspect should be studied more deeply.

Our findings demonstrate that cannabis affects cardiac function in D. melanogaster. However, we currently cannot discriminate between the beneficial or detrimental effects of cannabinoids. In view of our observations, short-term exposure appears to be deleterious, because arrhythmic patterns of heart activity manifested concomitant with an incremented mortality of flies. On the other hand, prolonged exposure to marijuana incremented contractility. An adaptive long-term mechanism might re-establish proper heart performance and this change could be interpreted as being beneficial. This appraisal is in concordance with the similarity in survival between non-treated and treated groups observed at this period of treatment (Fig. 1C).

Herein, we provide the first evidence of cardiac responsiveness to vaporized cannabinoids in Drosophila. These findings indicate that the fruit fly as a useful, low-cost and high-performance model for future tests of cardiac phenotypes induced by phytocannabinoids.