**3. Discussion**

#### *3.1. Effects of Phytocannabinoids in Wild-Type Zebrafish*

The use of zebrafish to test the toxicity of phytocannabinoids dates back to a 1975 study in which THC was dissolved in aquarium water (acute exposure), and its median lethal dose (LD50) calculated in zebrafish embryos was found to range between 2 and 5 mg/L [104]. Interest in studying cannabis/cannabinoids in the zebrafish model, however, has grown only in the past 10–15 years. The harmful effects of cannabinoid administration during zebrafish embryonic development have been well studied: embryos treated with THC and/or CBD exhibited shorter body lengths and mild deformities, reduced survival and basal heart rates, decreased synaptic activity and red-muscle-fiber thickness, alterations in the branching patterns of secondary motor neurons and Mauthner cells, changes in the expressions of postsynaptic nicotinic acetylcholine receptors in skeletal muscle, and reduced hatching rates [10,15,75]. In these studies, THC and CBD were used at concentrations believed to mimic the physiological range of cannabis use in humans (0.3–10 mg/L and 3–4 mg/L, respectively). In this regard, blood-plasma concentrations of THC and CBD caused by the consumption of a single cannabis cigarette have been found to reach peaks as high as 0.162 and 0.056 mg/L, respectively [105,106]. Table 1 summarizes studies on this topic.

Considering the deleterious effects of THC and CBD on developing embryos, the impact of these compounds on neural activity has recently been investigated through a novel in vivo assay based on a calcium-modulated photoactivatable ratiometric integrator (CaMPARI) system, which is able to provide a practical read-out of the neural activity in freely swimming larvae [3].

In acute regimens, both THC and CBD, if administered at high concentrations (6 and 3 mg/L, respectively), dramatically reduced the neural activity and locomotor activity of larvae at 4–5 dpf. Interestingly, the neuro-locomotor decrease was more pronounced when CBD and THC were combined. When treating embryos and 4 dpf larvae with low concentrations of CBD (up to 0.3–0.6 mg/L), no significant differences in the morphological parameters were observed, although the CBD significantly delayed the hatching of the embryos at the highest concentration used [32,51]. In most behavioral studies on the effects of cannabinoids in zebrafish, larvae were used at 5 dpf because, at this stage, they have fully developed digestive systems and inflated swim bladders, show mature swimming, and actively search for food [81,107]. In wild-type larvae at 5 dpf, the LD50 for THC, measured after chronic exposure (96 h beginning at age 24 hpf), was 3.37 mg/L [15]. In a study using zebrafish larvae with different characteristics and considering different drug-exposure times, a similar THC LD50 (3.65 mg/L) was found in fluorescent zebrafish of the Tg(fli1: EGFP) transgenic line at 4 dpf [86]. In acute regimens, the exposure of wild-type larvae to THC prompted a biphasic behavioral response consisting of increasing hyperactivity at concentrations ranging from 0.6 to 1.2 mg/L (2–4 μM), followed by the suppression of activity as the dose increased to 3.4 mg/L (10.8 μM) [15]. In line with these results, younger larvae (4 dpf) exposed to 0.3 mg/L THC exhibited a significantly increased duration of movement, while doses in the 0.6–1.25 mg/L range reduced the locomotor activity [32,86]. Evidence for the sedative effect of high doses of THC is also provided by Thornton et al. [14] and Amin et al. [75], who showed that THC at concentrations of 4–6 mg/L reduced swimming performances. These findings are consistent with results reported in rodents (i.e., dose-dependent hyperactivity followed by suppression at higher concentrations), as well as with the well-reported "stoning" action of THC in humans [45,108]. In chronic regimes, THC showed habituation, which is the development of tolerance to many of the acute effects in chronic exposition. Nevertheless, THC at 1.2 mg/L increased the distance traveled by fish [15]. This phenomenon has been associated with the downregulation of cannabinoid receptors after long-term exposure to cannabinoids [109]. In addition, the observation of reduced larval basal activity in response to exposure to THC at doses of up to 0.625 mg/L (2 μM) [31] suggests that THC produces a calming effect on larval locomotor activity up to this concentration, as opposed to hyperactivity at concentrations ranging

from 0.6 to 2.4 mg/L, and sedation at concentrations higher than 2.4 mg/L (see Figure 3). In this context, psychoactive drugs, such as THC or its analog WIN55,212-2, by activating cannabinoid receptors, can induce hypothermia and hypoactivity, increase tremors and startle behaviors, and, in severe cases, induce catalepsy-like immobilization [110,111].

**Figure 3.** Toxicological and behavioral effects of acute THC and CBD administration on wild-type zebrafish embryos and larvae.

Zebrafish treated with WIN55,212-2 at 0.5 and 1 μg/mL showed no activity, even in darkness, whereas this was lethal if applied at 10 μg/mL [111]. Chronic early-life treatment with THC (0.6 mg/L) did not affect the locomotor abilities in 30-month-old zebrafish, which suggests that this psychoactive cannabinoid has no long-term effects on swimming behavior if used at low doses [27].

As for CBD, embryonic exposure to concentrations of up to 0.15 mg/L did not cause notable morphological abnormalities [32]. The LD50 values for CBD, calculated in zebrafish, are 4.4 mg/L at 2 dpf, 3.7 mg/L at 3 dpf [112], and 0.53 mg/L at 4 dpf [86]. In this latter study, larvae chronically exposed to low concentrations of CBD showed a biphasic locomotor response pattern, similar to that previously reported for THC [15]. In detail, 0.07 mg/L CBD produced a significantly increased duration of larval movement, while concentrations of 0.1–0.3 mg/L had a hypolocomotor effect. The acute administration of CBD at doses of up to 0.3 mg/L did not alter the locomotor behavior of 5 dpf zebrafish larvae, whereas higher concentrations caused larval hyperactivity [31]. In support of these findings, a study using auditory/mechanical tests to evaluate fish behavioral responses to unexpected sound and touch stimuli showed that THC and CBD concentrations of 6 mg/L and 3 mg/L, respectively, reduced their responses to sound [10]. An inhibitory effect on locomotion of low doses of CBD, ranging from 0.5 to 10 μg/mL, has been reported, but without a dose-dependent mechanism [111]. The same research evaluated larval responses to CBD after an initial exposure to WIN55,212-2. The results indicated that CBD could attenuate the WIN55,212-2-induced abnormal immobilization. Differences between the control and CBD-treated groups were no longer detected after 24 h of recovery in clean water, and this recovery trend was observed even after exposure to toxic levels of WIN55,212-2. Another study tested the analgesic properties of THC and CBD in a zebrafish larval model of nociception [25]. In detail, larvae, while recovering from acute exposure to low levels (0.1–0.5%) of acetic acid (nociception stimulus), were exposed to low levels of

THC or CDB (0.15 mg/L). The THC-exposed larvae showed reduced activity compared with that of both the acetic acid-treated and control groups, which is in line with the proposed calming effect of THC at doses of up to 0.6 mg/L (Figure 3). Notably, however, CBD appeared to increase the larval locomotor activity after acetic acid exposure, and it also had a nominal effect on the control-group locomotion, seemingly confirming its nociceptive properties. In other research analyzing both the immediate and long-term effects of THC (up to 0.6 mg/L) and CBD (up to 0.15 mg/L) on larval locomotor behavior, it was observed that THC exposure reduced the swimming behavior in the treated larvae (F0), as previously reported, whereas the locomotor parameters in their offspring (F1) were increased in comparison with the controls. Instead, CBD had no effect on F0 larvae, and it decreased activity in unexposed F1 larvae [32]. Furthermore, in 3 dpf larvae, 1.25 mg/L of CBD extract accelerated the caudal-fin regeneration and reduced apoptosis after amputation [112].

Several different cannabinoids have been tested on 5 dpf wild-type zebrafish larvae. In particular, exposure to CBN and CBDV at concentrations higher than 0.75 mg/L led to malformations and bradycardia, and the calculated LD50 for CBN was 1.12 mg/L [14,30]. A behavioral analysis suggested that the locomotion of the treated larvae remained unaltered up to 0.043 mg/L of CBN [14], but was significantly reduced at higher concentrations, in both dark and light conditions, which also affected their anxiety status. Conversely, CBDV administration had no significant effect on zebrafish [30]. In another study, a novel dihydrophenanthrene derivative, isolated from commercial cannabis, exhibited behavioral dose effects similar to those previously described with CBD [12]. Evaluating the toxicity and antitumor effects of abnormal CBD and its analog O-1602 (which have no or only little affinity for CB1 and CB2), Tomko et al. [24] found that both atypical cannabinoids significantly reduced tumor growth, but concentrations greater than 0.8 mg/L caused higher levels of toxicity to the larvae. Finally, data from another study indicated that THCV and THCV −OH have significant effects on the skeletal ossification of larvae at 8 dpf [37].

Recently, two similar behavioral studies, conducted independently in Canada and Italy, evaluated the effects induced by full-spectrum cannabis extracts, as opposed to purified major cannabinoids, on the zebrafish model. Research data on these extracts are scarce, and because cannabis consumers use the entire inflorescences, more scientific evidence is needed to clarify the bioactivity of all the cannabinoids, including their simultaneous interactions. In the study by Nixon et al. [92], acute exposure to the extracts produced similar complex concentration-dependent activity patterns to those observed by the group when using pure THC and CBD in a previous study [31]. However, distinct concentration-dependent differences were found both between the extracts (characterized by different ratios of THC:CBD) and versus the purified THC and CBD, which suggests that these differences might be related to the activity of other minor cannabinoids (specifically CBC, CBG, and CBDA). In the study by Licitra et al. [91], an excitatory effect on the locomotor activity was observed in larvae exposed to cannabis extract derived from CBD-rich-strain plants (containing about 0.5 and 7 μg/L of THC and CBD, respectively), without leading to toxicity effects. These studies underlined that the precise bioactivity of the single compounds in cannabis extracts and their interaction with the ECS pathway are highly complex issues that require further work.

Research on acute exposure to the cannabis receptor agonists WIN55,212-2 and CP55,940 indicated that both compounds reduce the locomotor activity in a dose-dependent fashion, in both light and dark phases, while the specific CB2 agonists HU-910 and JWH-133 had no effect on locomotion, in any circadian phase [87]. Using cnr1−/− larvae, the authors found no inhibitory effect of WIN55,212-2 or CP55,940 on the average swimming velocity. The CB1 antagonist AM251 did not affect locomotor activity, but blocked the effect of WIN55,212-2, which indicates that these endocannabinoids are not active in regulating the locomotor activity in zebrafish larvae at 5 dpf.

Another gene-expression analysis, performed on 4 dpf fluorescent larval zebrafish exposed at 96 hpf to THC or CBD, focused on the differential expressions of 10 key morphogenic or neurogenic genes [86]. The authors found the c-fos expression to be differen-

tially upregulated in a concentration-dependent manner following both THC (1.25 and 2.5 mg/L) and CBD (0.07 and 0.1 mg/L) exposure, and it was correlated with increased neural activity and hyperlocomotor behavior in the zebrafish. In addition, the same concentrations of THC resulted in deleted in azoospermia-like (dazl)-gene upregulation, while the expressions of vasa, sox2, sox3, sox9a, bdnf, reln, krit1, and the CB1-expressing gene cnr1 were similar to the control values. Along the same lines, during the key developmental stages (14, 24, 48, 72, and 96 hpf), THC and CBD caused the differential expressions of c-fos, bdnf, and dazl [32]. Contrary to the findings on cnr1 gene expression reported by Carty et al. [86], treatment with a full-spectrum cannabis extract (THC-poor strain) induced the overexpression of both cnr1 and cnr2 cannabinoid receptors in 5 dpf zebrafish larvae [91]. Additionally, CBD was found to reduce the gene- and protein-expression levels of cxcl8, tnfα, and il-1β, and of IL-1β, caspase 3, and PARP [112]. Pandelides and colleagues observed that treatment with cannabinoids can alter the expression of proinflammatory cytokines in aged fish, which suggests a possible reduction in inflammation over the course of the lifespan [27]. In particular, exposure to low levels of THC during zebrafish development led to a significant reduction in tnfα and il-1β at an advanced age, but this was not observed at higher doses, which indicates a biphasic or hormetic effect. Furthermore, the differential effect on the pparγ expression of exposure to cannabinoids in adult male vs. female zebrafish suggests that cannabinoid exposure could have long-term effects on reproduction, growth, and survival during early development [27].

In addition to larval locomotor activity, Achenbach et al. [31] assessed the uptake kinetics of THC and CBD, and their possible metabolism by larvae, suggesting that both cannabinoids are bioaccumulated in the living organism, but at concentrations that are substantially lower than their levels in test media. Studies involving liquid chromatography– tandem mass spectrometry analysis have shown that, when a test compound is dissolved in the embryo water, only 0.1–10% of it typically crosses the chorion and actually reaches the embryo [3,10], which limits the effectiveness of the treatment. In support of this, Carty et al. [87] found that, despite the best laboratory efforts, the actual THC concentrations in water corresponded to between 64% and 88% of the expected values at time 0, and the THC detection rate fell to between 16% and 32% at 96 hpf. Similarly, the actual CBD concentrations were only 33–40% of the nominal at time 0, and decreased to either not detected or 3% of baseline after 96 h. Indeed, in pharmacological and toxicological research with aquatic species, where the test compounds are usually diluted in the incubation water, it is essential to consider the relationship between the drug concentration in the medium and its adsorption and degradation rates.

Behavioral data from adult wild-type zebrafish indicate that a sedative effect was evoked following acute exposure to high doses of THC (30 and 50 mg/L) [45]. Moreover, reduced top swimming behavior was observed during the THC exposure, which indicated an anxiogenic effect. In another study, low doses of THC (up to 0.6 mg/L) did not cause significant behavioral effects in treated adults, but a significant reduction in thigmotaxis was seen in F1-generation fishes [32]. Other authors, however, have found these THC doses (0.3–0.6 mg/L) to induce repetitive swimming patterns in adult zebrafish [2]. In the same study, N-methyl-D-aspartate (NMDA), GABA antagonist pentylenetetrazol (PTZ), selective CB2 inverse agonist AM630, and sulpiride (an antipsychotic) attenuated a THCinduced behavioral stereotypy, while the selective CB1 inverse agonist AM251 did not. These results support a possible role for CB2 as a mediator of abnormal behavioral patterns induced by THC [2]. In terms of cognitive abilities, it has been reported that the acute administration of tiny doses of THC (0.03 mg/L) did not lead to any observable effect on color-discrimination learning, but heavily impaired the fish spatial-memory retrieval [113]. Conversely, in studies of possible CBD effects, acute exposure to 40 mg/L reduced the swimming speed and distance [7], while no changes in these parameters were reported when using concentrations ranging from 0.1 to 10 mg/L [48]. These latter lower doses showed an anxiolytic effect on zebrafish in the novel tank test, which is in line with the findings in acute regimens in mammalian models [114]. However, CBD at 5 mg/L caused

memory impairment in an avoidance task, while the same dose did not affect aggressive behavior and social interaction [48].

Studies exploring the reproductive effects of cannabinoids sugges<sup>t</sup> that developmental exposure to THC can cause persistent sex-specific alterations to the reproductive system, particularly in male fish [32], and even across generations [27]. Similarly, THC treatment significantly reduced the ATP levels in mammal spermatozoa [115], and altered ECS signaling is linked to infertility in human semen [116]. Thus, the reproductive effects could be a result of altered metabolism [27]. In rodents, greater tolerance to THC in female rats than in male rats has been observed; this is probably due to the presence of hormones that are able to modulate the THC effects [117]. However, both in rats and zebrafish, maternal exposure to THC has been linked to altered locomotor and exploratory behavior in the offspring [32,118]. Unexpectedly, treating embryos with a low dose of THC (0.024 mg/L) increased the survival in aged males (30 months old), while, in aged females, the same dose improved egg production and reduced body mass [27].

#### *3.2. Effects of Phytocannabinoids in Zebrafish Models of Neurological Disorders*

The observed neuromodulatory effects of cannabinoids on the CNS have led neuropharmacological researchers to increasingly focus on the clinical potential of these molecules for use in the treatment of neurological disorders. Several zebrafish lines characterized by neuro-hyperactivity, seizures, bipolar disorder, and anxiety/stress and addiction behaviors have already been developed [17,31,119]. Epilepsy is a common neurological disorder that affects over 70 million people worldwide [120]. Approximately one-third of patients show multidrug resistance [121]. Therefore, research efforts are aimed at developing new drug treatments. Seizure treatment is one of the oldest reported uses of cannabis, and recently, the use of pure cannabinoids has been suggested as a means to treat severe forms of refractory childhood epilepsy (i.e., Dravet syndrome) [122,123]. Several zebrafish models of epilepsy, and more generally of psychiatric and muscular disorders linked to neuro-hyperactivity, have already been created and offer several specific key advantages, as explained below [14,17,119]. A number of small molecules targeting different receptors or ion channels can be used to induce seizures or neural hyperactivity in zebrafish larvae. The best characterized chemically induced model is PTZ exposure. Zebrafish larvae exposed to PTZ show a concentration-dependent abnormal pattern of behavior: increased locomotion followed by fast darting activity, and finally, clonic convulsions accompanied by a loss of posture [119]. In addition, PTZ administration leads to electrophysiological changes in the zebrafish optic tectum [124]. For instance, homozygous scn1Lab−/− mutants display significant phenotypic similarity to humans with Dravet syndrome, including spontaneous seizures, resistance to many available antiepileptic drugs, and early death [14,17]. The zebrafish knock-out model of neuro-hyperactivity, obtained by loss-of-function mutations in the GABA receptor subunit alpha 1 (gabra1−/−), offers a unique advantage for drugscreening purposes because seizures (in addition to the sporadic ones) can be triggered by exposure to light [119]. In this context, CBD and THC significantly reduced the seizureinduced total distance moved, both in chemically induced and genetic models [14,119]. Although the exact mechanisms by which cannabinoids exert their antiseizure effects are not well understood, a number of molecular targets are known to be modulated by cannabinoids. Because CBD is a positive allosteric modulator of GABA receptors, it could, for example, be capable of reducing seizure events through this mechanism. This might hold true despite the fact that THC has been associated with GABA-release inhibition [14], as, even in this case, the THC properties depend, at least in part, on the seizure-model characteristics and cannabinoid dose. Furthermore, as previously indicated, the ability of these phytocannabinoids to reduce seizures could also be mediated by the transient receptor potential vanilloid 1 (TRPV1) channel: THC, CBD, CBN, and CBDV are all transient receptor potential cation channel subfamily A member 1 agonists. Finally, NMDA receptor, glycolysis, and fatty acid amide hydrolase may be potential cannabinoid targets, participating in seizure-effect modulation [14]. Recently, a commercially available library containing

370 synthetic cannabinoids (compounds engineered to bind cannabinoid receptors with high affinity) was screened [17] in 5 dpf homozygous scn1Lab−/− zebrafish larvae in order to identify molecules with the ability to reduce seizure-like behaviors. Five compounds exerting significant antiseizure activity during acute exposure were identified. It is essential to note that synthetic cannabinoids are not FDA-approved "for human or veterinary use", and substantial evidence of serious adverse effects has been reported for some of them [17]. Further research using the above models could be of grea<sup>t</sup> help in discerning the true therapeutic potential of various cannabinoids for the treatment of epilepsy.










*Biomedicines* **2022**, *10*, 1820

#### **4. Pointers on Behavioral Analysis**

Behavioral analysis was performed in 18 of the 21 studies dealing with zebrafish larvae. The age of the larvae ranged from 1 to 7 dpf. Two tests were applied: the VMR test and the mechanical escape response; the latter was used in three of the 18 studies and was combined with an auditory stimulus in only one of them [10]. The VMR test normally involves several phases of light–dark succession, and it aims to stimulate an unconscious defensive response initiated by a drastic change in lighting [126]. In wild-type larvae without sight impediments, the locomotor activity increases at light onset, before decreasing to the baseline level after ca. 30 s. The wild type also shows increased locomotor activity at light offset, but they need more time (ca. 30 min.) to return to the baseline level of locomotion [127]. During embryogenesis, mechanical stimulus to the tail of the zebrafish embryo can be used to elicit the coiling behavior (touch response) [128]. Similarly, the escape response can be stimulated in larvae using mechanical, acoustic, electrical, or optical stimuli [129]. The escape response mimics predator-avoidance behavior, which is usually mediated by the Mauthner cells [130] located in the hindbrain [125,131].

Of the five studies carried out in zebrafish adults, four evaluated locomotion, one of these also explored social behavior and memory [48], and the other was conducted on color-discrimination learning and spatial cognition [113].

Overall, the results on the locomotion, both in larvae and adult fish, showed significant differences between studies. Cannabinoids, depending on the concentrations used, could either increase or decrease locomotor activity. As we stated in a recent systematic review on social-preference tests in zebrafish [101], the lack of a standardized approach to behavioral assessment makes it difficult to compare studies. Furthermore, in view of the heterogeneity in terms of the administered cannabinoids, doses, and exposure times of the current research, the standardization of behavioral tests could help to allow inferences to be drawn from findings in zebrafish species and provide more consistent data for translationalmedicine purposes. The age of larvae used to perform the VMR test could, ideally, be set at 5 dpf: at this age, the larvae show limited (but sufficient) physiological development [132], but they are not ye<sup>t</sup> independently feeding and are therefore subject to the EU directive on the protection of animals used for scientific purposes (Directive 2010/63/EU). Moreover, the exposure time could be set at 24 hpf, or 120 hpf to evaluate the effects of prolonged exposure. It should be highlighted that the daily replacement of the drug was performed in only one of the studies reviewed [37]. Although the adsorption of the medication can be considered minimal, and especially in the case of cannabinoids [3,10], we still believe that an approach that keeps the drug concentration constant over time, and that also considers the possible evaporation of egg water or medical compound, if volatile, will be the most accurate. The approach could be further standardized by introducing a standard duration of locomotor experiments and choosing the preferred drug-administration route for studies in adult zebrafish. We think the duration should be 30 min, and that drugs could be optimally administered through food. Furthermore, with regard to the method used to analyze the behavioral effects of cannabinoid treatments in adults, it may be useful to elect the novel tank test as the major read-out, considering that behavioral experiments should ideally last 10 min, after 5 min of habituation time.
