Flies prefer to consume food containing phytocannabinoids
The non-psychoactive phytocannabinoids are naturally occurring compounds derived from Cannabis sativa, and exert their biological effects via interaction with cannabinoid receptors or other G-protein coupled receptors 28. To investigate any potential role of cannabinoids in food preference in adult flies, we took advantage of the two-choice CAFE assay 29, in which flies are presented with cannabinoid food and proper control food in four capillaries (Figure 1A). We first assessed whether flies display a preference for four phytocannabinoids, including CBD, CBDV, CBC and CBG. Flies were presented with various concentrations of CBD (0.01 mg/ml, 0.1 mg/ml or 1 mg/ml), CBDV, CBC or CBG (0.01 mg/ml or 0.1 mg/ml) in each CAFE assay. Statistical analysis shows that the flies did not exhibit obvious preference for food with the individual phytocannabinoids in the first 2 days (Figures 1B-E). Interestingly, a significant preference for these phytocannabinoids was observed in the following two days. Flies showed the preference for consuming food with 0.1 mg/ml or 1 mg/ml CBD on day 3 and day 4, however, appeared to prefer 0.01 mg/ml CBD only on day 4 (Figure 1B). These data indicate that flies possess an ability to detect CBD and develop a dose-dependent CBD preference over time. Likewise, food containing high concentration of CBDV (Figure 1C), CBC (Figure 1D) or CBG (Figure 1E), was preferentially consumed by the flies on day 3 and day 4. Flies also displayed food preference to CBG at the lower concentration on day 4 (Figure 1E), similar to CBD (Figure 1B). Thus, these data indicate that flies can develop a delayed preference for phytocannabinoids in a dose-dependent manner.
Flies prefer to consume food with endocannabinoids and synthetic cannabinoids
We next extended the characterization of food preference to endocannabinoids and synthetic cannabinoids. Flies were presented with either of endocannabinoids, namely anandamide (AEA) and 2-arachidonoylglycerol (2-AG), at 0.01 mg/ml and 0.1 mg/ml (Figure 2A and 2B). The preference indexes of food intake indicate a robust preference for food with a high concentration of AEA (Figure 2A) and 2-AG (Figure 2B), but not with the lower concentration. The strong preference for 2-AG was sustained throughout 4 days of feeding (Figure 2B), whereas the preference for AEA remained consistent, albeit relatively lower, in the first 3 days (Figure 2A). Previous studies have reported the identification of the endocannabinoid-like signal lipid 2-LG, an evolutionary alternative to 2-AG in Drosophila16,20. Arachidonic acid (AA) supplement can promote the synthesis of 2-AG in Drosophila16. We also observed a weak but significant preference for food containing 2-LG only on day 3, in contrast to more robust preference for food with AA throughout the entire course of feeding (Figure 2C).
With the detection of food preference for endocannabinoids in flies, we further assessed potential preference for synthetic cannabinoids. Flies exhibited a significant preference for food containing 0.5 mM CP55940 over 4 days (Figure 2D), but not for another CB1/CB2 receptor agonist WIN55212-2 at the same concentration (Figure 2D). Given the absence of CB1/2 receptors in the fly genome, the results suggest that these two synthetic cannabinoids might target distinct receptors in Drosophila. Furthermore, we also investigated potential effect on food preference for AM251 (0.5 mM) or AM630 (0.5 mM), two antagonists of CB1 and CB2 receptors, respectively. The consumption for AM251 or AM630-containing food remained similar to that for the control food throughout 4 days of feeding, despite a weak preference for AM630 on day 3 (Figure 2E). Fatty acid amide hydrolase (FAAH), which is well conserved between mammals and Drosophila30, catabolizes AEA into AA and ethanolamine 31. Treatment with PF3845, an inhibitor of mammalian FAAH, has also been shown to increase the endocannabinoids in fly hemolymph 30. Flies displayed a significant preference for food with 0.5 mM PF3845, and the preference index remained relatively high throughout 4 days of feeding (Figure 2F). Taken together, these findings indicate a general preference for both endocannabinoids (AEA and 2-AG) and synthetic cannabinoid CP55940 in flies.
Cannabinoid preference is independent of sensory inputs in flies
To examine whether the preferential response to cannabinoids is due to sensory stimuli in flies, we examined possible involvement of gustatory and olfactory functions in food preference using taste-deficiency mutants (poxn70-23 and poxn ΔM22-B5) (Figure 3A and 3B), and olfaction-deficiency mutants (orco1 and orco2) (Figure 3C and 3D), respectively. In comparison to the control w1118 flies, both poxn and orco homozygotes displayed similar preference for 1 mg/ml CBD and 0.5 mM CP55940 over 4 consecutive days of feeding, suggesting that the preference for cannabinoids in flies is modulated through a mechanism independent of gustatory and olfactory sensory functions.
An inhibitory effect of phytocannabinoids on food intake
The lack of gustatory and olfactory sensory inputs in influencing cannabinoid preference suggests a potential pharmacological effect of these compounds on metabolism in Drosophila. Phytocannabinoids have been shown to affect food intake, and can be developed as potential therapeutic agents for obesity treatment 12. We next investigated if phytocannabinoids regulate food intake in flies. The consumption of normal food by flies was quantified for two consecutive days following two days of cannabinoid training (Figure 4A). Pre-treatment of the flies with higher concentrations of CBD at 1 mg/ml and 2 mg/ml, but not with the low concentration at 0.1 mg/ml, significantly decreased the total food intake on both days (Figure 4B). The movement ability of flies was assessed to examine whether the effect observed in food intake was due to a decline in locomotion. In the negative geotaxis assay, the treatment with high concentration of CBD did not impair the climbing ability of adult flies (Figure S1A), suggesting that decreased food intake by CBD treatment is not attributed to altered locomotion. However, food intake was significantly decreased only on day 2 following the treatment with 0.1 mg/ml CBC and CBDV (Figure 4C). No or negligible effect was detectable upon CBG treatment in flies (Figure 4C). Subtle effects of CBC, CBDV and CBG could be due to their lower feeding concentration. We were unable to further increase the feeding concentration, given the low concentration of commercially available stock solutions. During the first two days of training, the amounts of food containing phytocannabinoids or the respective control solutions consumed by the flies remained similar (Figure S2). These control experiments suggest that the decrease in food intake is not caused by altered food consumption during the initial cannabinoid training. Thus, our results indicate that phytocannabinoids elicit a functional effect on the regulation of food intake in flies.
Endocannabinoids inhibit food intake
Growing studies have reported that endocannabinoid signaling functions to enhance food intake through the activation of CB1 receptor in mammals 25. However, the sequence homology analysis indicated that canonical CB1 receptor appears to be absent in Drosophila15,32. Cannabinoids might function in an alternative way in flies, for example, via non-canonical cannabinoid receptors. To explore whether the endocannabinoids regulate food consumption, flies were pre-treated with AEA and 2-AG at the concentration of 0.01, 0.1 and 0.5 mg/ml in the first two days prior to the measurement of food intake in the next two days (Figure 4A). The pre-treatment with higher concentrations of AEA exhibited profound effects on food intake whereby consumption of normal food was significantly attenuated on both days (Figure 5A). Consistently, a significant reduction in food intake was observed with 2-AG treatment at the higher concentrations (Figure 5B). However, this effect was only observed on the first day following the pre-treatment. Strikingly, the amounts of food containing the higher concentrations of AEA or 2-AG consumed by flies were significantly lower as compared to those of the normal food (Figure S3A and S3B). As a control, the climbing ability was not affected by either AEA or 2-AG treatment (Figure S1B and S1C). Next, we attempted to determine the possible effect of 2-LG (0.01 and 0.1 mg/ml) on food intake. Similar to 2-AG treatment, flies consumed significantly less food with 2-LG at the higher concentration of 0.1 mg/ml as compared to the normal food (Figure S3C) during the initial training. The pre-treatment with higher concentration of 2-LG also significantly decreased food consumption in the next two days (Figure 5C).
A recent study has reported that AA facilitates 2-AG synthesis by the fly ortholog of diacylglycerol lipase dDAGL 16. Since both 2-AG and 2-LG account for the reduction of food intake, we postulated that treatment with AA in flies may also have an inhibitory role in food intake. In agreement with our hypothesis, a similar inhibitory effect on food consumption was observed upon AA treatment. Normal food consumed by flies treated with higher concentration of AA at 0.1 mg/ml was significantly reduced as compared to the control group (Figure 5D). Similarly, food consumption was remarkably lower during the first 2 days of 0.1 mg/ml AA pre-treatment (Figure S3D), and the climbing ability was not affected by AA treatment (Figure S1D). Together, these results suggest that endocannabinoids, like phytocannabinoids, elicit an inhibitory effect on the regulation of food consumption.
AM251 attenuated the inhibitory effect of AEA on food intake
It has been documented that the activity and levels of AEA are negatively regulated via at least two models. First, AEA is degraded by FAAH to AA and ethanolamine 31. Second, fatty acid-binding protein (FABP) facilitates the transportation of AEA to FAAH, which leads to subsequent degradation and inactivation of AEA 33. Overexpression of FABP significantly promotes AEA uptake and hydrolysis in neuroblastoma cells 34. Given that AEA treatment led to a reduction in food intake, we next examined whether changes in endogenous AEA levels through FAAH inhibition or FABP overexpression can affect food intake. Importantly, treatment with the FAAH inhibitor PF3845, which inhibit endocannabinoid degradation in flies 30, led to a significant reduction in food intake on day 1 (Figure 6A). Likewise, food consumption was significantly reduced during PF3845 pre-treatment (Figure S3E). Conversely, overexpression of fly dFABP significantly enhanced the food intake on two consecutive days (Figure 6B). These results are in line with an inhibitory effect of AEA on food intake.
Next, we sought to determine whether the CB1 receptor antagonist AM251 (0.5 mM) can counteract AEA (0.1 mg/ml) or 2-AG (0.1 mg/ml) effect in terms of food intake. Flies pre-fed with AM251 did not alter the levels of food intake in the following two days (Figure 6C). Flies also consumed similar amount of food containing AM251 during the pre-treatment (Figure S3F). These results indicate that the treatment with this CB1 receptor antagonist did not affect food consumption in flies. Treatment with 0.1 mg/ml AEA or 2-AG led to significant reductions in normal food consumption (Figures 6D and 6E). Interestingly, co-treatment of AM251 with AEA (Figure 6D) or 2-AG (Figure 6E) significantly attenuated their inhibitory effects on food intake. Due to the lack of canonical CB1/2 receptors in Drosophila genome, AEA and 2-AG regulate food intake, likely via another unknown receptor that can be blocked by AM251.
AEA treatment promotes starvation resistance and inhibits lipid metabolism
Given that phyto- and endo- cannabinoids can modulate food intake in flies, we next tested whether they also affect starvation resistance and lipid metabolism. Starvation resistance was quantified by measuring the survival percentage of the flies following cannabinoid treatment. After treatment with 0.1 mg/ml AEA, 0.1 mg/ml AA or 0.5 mM CP55940 for 2 days, flies exhibited significant increases in the survival rate (Figures 7A, 7C and 7E), suggesting an enhanced resistance to starvation. Unexpectedly, neither 2-AG nor CBD pre-treatment showed any effect on starvation resistance (Figures 7B and 7D). The observation of enhanced starvation resistance further prompted us to investigate whether body fat deposition was affected by AEA. Following 18 h starvation, the levels of triglyceride (TAG) were measured in endocannabinoid-treated flies. TAG levels in AEA- or AA- treated flies were significantly elevated when compared to the control flies (Figure 7F). In line with no effect on starvation resistance, we did not observe a significant alteration in TAG levels upon 2-AG treatment (Figure 7F). Thus, AEA and AA can enhance the resistance to starvation likely through the inhibition of lipid metabolism.