Medium-throughput zebrafish optogenetic platform identifies deficits in subsequent neural activity following brief early exposure to cannabidiol and Δ9-tetrahydrocannabinol
grid.17089.37Centre for Prions and Protein Folding Disease, University of Alberta, Edmonton, AB T6G 2M8 Canada
grid.17089.37Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2E9 Canada
grid.17089.37Department of Medical Genetics, University of Alberta, Edmonton, AB T6G 2H7 Canada
Neuroscience and Mental Health Institute, 4-120 Katz Group Centre, Edmonton, T6G 2E1 Canada
Abstract
In light of legislative changes and the widespread use of cannabis as a recreational and medicinal drug, delayed effects of cannabis upon brief exposure during embryonic development are of high interest as early pregnancies often go undetected. Here, zebrafish embryos were exposed to cannabidiol (CBD) and Δ9-tetrahydrocannabinol (THC) until the end of gastrulation (1–10 h post-fertilization) and analyzed later in development (4–5 days post-fertilization). In order to measure neural activity, we implemented Calcium-Modulated Photoactivatable Ratiometric Integrator (CaMPARI) and optimized the protocol for a 96-well format complemented by locomotor analysis. Our results revealed that neural activity was decreased by CBD more than THC. At higher doses, both cannabinoids could dramatically reduce neural activity and locomotor activity. Interestingly, the decrease was more pronounced when CBD and THC were combined. At the receptor level, CBD-mediated reduction of locomotor activity was partially prevented using cannabinoid type 1 and 2 receptor inhibitors. Overall, we report that CBD toxicity occurs via two cannabinoid receptors and is synergistically enhanced by THC exposure to negatively impact neural activity late in larval development. Future studies are warranted to reveal other cannabinoids and their receptors to understand the implications of cannabis consumption on fetal development.
Introduction
Cannabis is often portrayed as harmless, yet its health implications are not fully understood1. The extracts of the Cannabis sativa plant have been used for medical purposes for almost 5000 years, especially for the treatment of pain2. Despite controversy and incomplete knowledge of toxicities, cannabis has been legalized for recreational use in various jurisdictions. Cannabis is one of the most used illicit drugs during pregnancy, with an increase in consumption over recent years. Importantly, key phytocannabinoids such as Δ9-tetrahydrocannabinol (THC) and cannabidiol (CBD) can readily cross the placenta3–8. Moreover, CBD use is promoted due to its purported health benefits in various natural products including oils and food9.
The negative impacts of cannabis on embryonic development have been principally associated with the psychoactive compound THC10. Epidemiological and clinical studies associate maternal cannabis exposure with behavioural disturbances in the offspring and are linked to an increased risk of neuropsychiatric disorders11. Maternal exposure to THC in rats changes a series of behaviours in their offspring including water-induced grooming, increased light sensitivity, and altered exploratory behaviour12. Importantly, the negative impacts of cannabis may also occur via the non-psychotropic compound CBD, which disrupts motor-neuron development in zebrafish13. This latter study is of concern because it stands in contrast to reports that promote the positive health benefits of CBD, for example in the treatment of nausea during pregnancy14,15.
The mechanistic pathways by which THC and CBD are toxic during development remain elusive. Both of these cannabinoids have recently been linked to the sonic hedgehog signaling pathway in mice and zebrafish, and signal via the cannabinoid type 1 receptor (CB1R)16. THC binds to two distinct classes of G-protein coupled receptors, CB1R and CB2R, where it acts as a partial agonist17. Both receptor types are expressed in the nervous system, with CB1R being highly enriched in the CNS18–20. While CB2R is more commonly found in the peripheral nervous system21, it is also associated with the CNS22–24. CBD interacts with CB1R and CB2R, but with lower affinity relative to THC. CBD is also thought to antagonize cannabinoid-induced effects indirectly through other receptors25–29. Additional in vivo data on the signaling pathways of THC and CBD would highlight functionally important mechanisms.
Zebrafish embryos have several advantages that complement mammalian models in attempting to disentangle these issues. Experiments are relatively economical because many embryos are available. Exposure studies are simple as compounds can be easily added and removed from the incubation media that bathes the developing embryos. Further, embryos develop externally and experimental results are not confounded by maternal physiology or variable transport to the fetus. Translucent larvae provide opportunities to implement cutting-edge fluorescing calcium sensors and to measure neural activity. The endocannabinoid system plays a role in zebrafish development30–32. CB1R and CB2R have similar expression profiles in the CNS of zebrafish compared to mammals, and CB1R shares a 70% protein sequence identity with its human homolog33. Therefore, the zebrafish is a favorable model organism to further illuminate evolutionary conserved biological mechanisms.
Considering the anticipated effects of THC and CBD on developing embryos, we sought to develop high-throughput quantification of neural activity following early drug exposures. Here we deploy the calcium modulated photoactivatable ratiometric integrator (CaMPARI)34 to document the effects of THC and CBD on neural activity. CaMPARI is expressed exclusively in neuronal tissue due to a pan-neural promoter (elavl3, a.k.a. HuC) and is a direct read-out of relative neural activity. Green fluorescing CaMPARI undergoes photoconversion (PC) to a red fluorescing protein only when in the presence of high Ca2+ levels occurring coincidentally with user-controlled intense 405 nm light. This photoconversion is irreversible and thus creates a temporal snapshot of neural activity in freely swimming larvae. The CaMPARI photoconversion is subsequently quantified as a ratiometric red/green output, and this mitigates many technical challenges of imaging neural activity that are imposed by issues such as variable expression, tissue mounting, or tissue depth. These advantages combine to make CaMPARI uniquely adaptable for high-throughput quantifications. We optimized the workflow and photoconversion of CaMPARI to make it practical for high-throughput well-plate-formats. Our novel in vivo assay is scalable to measure neural activity via screens of compound libraries. The results were supplemented with a behavioral assay performed on the same larvae, which offers an independent proxy measure that is related to neural activity.
Here, we investigate the actions of THC and CBD on neural activity and test whether the in vivo effects of CBD are mediated via CB1R and CB2R in zebrafish. We report that early embryonic exposure to THC and CBD (until 10 h post-fertilization) impacts neural activity later in larval development (at 4 and 5 days post-fertilization (dpf)). The concentrations of THC and CBD used mimic plasma levels of human subjects with high cannabis consumption35–37, though complexity arises from estimating diffusion through multiple barriers (e.g. incubation media through the chorion and then into the plasma and tissues of the zebrafish embryo). We document the effects of THC and CBD when applied individually and in concert, to partially mimic the combination of cannabinoids in recreational cannabis. We probe the roles of CB1R and CB2R in this signaling using established receptor inhibitors, towards revealing their in vivo requirement amongst the various receptor mechanisms that have been identified from in vitro studies.
Results
High-throughput assessment of CaMPARI is a reliable metric of neural activity
An objective of this study was to gain insight into the effects of THC and CBD on neural activity during early embryonic development. We measured neural activity with CaMPARI, which photoconverts irreversibly from a green to a red fluorescing protein only if user-applied 405 nm light application coincides with high calcium levels (Fig. 1a)34. Red and green fluorescing CaMPARI are quantified as a ratio (referred here as “CaMPARI activity"). We expressed CaMPARI through a pan-neural promoter, elavl3 exclusively in the CNS and thus it acts as a read-out of neural activity (Fig. 1b). Here, we optimized the use of CaMPARI zebrafish for an automated high-content INCell2000 plate-reader (Fig. 1c). We empirically determined that 4 dpf is a suitable age for quantifying CaMPARI, which provides an appropriate time for imaging (good CaMPARI expression and tissue clarity) coupled with a relatively late stage in CNS development.
CaMPARI activity was quantified and displayed as a heatmap consisting of the red/green ratio in fluorescing intensities. At 4 dpf, CaMPARI outputs in the optic tectum and hindbrain regions of larvae showed robust neural activity, whereas without photoconversion CaMPARI activity was nearly zero. To establish a baseline of near-zero neural activity, larvae were anesthetized with MS-222, and a significant reduction (p < 0.01) in red/green ratio was obtained compared to untreated freely swimming larvae. In contrast, drug-induced neural activity, with the established convulsants pentylenetetrazole (PTZ) or 4-aminopyridine (4-AP), showed a consistent increase in CaMPARI activity, even at concentrations considered to be minimal when inducing seizures (Fig. 1d, e)38.
We benchmarked our CaMPARI outputs against locomotor activity of zebrafish larvae, which is an established metric of neural activity39. We chose 5 dpf as an optimal developmental timepoint for analyzing locomotor activity, because younger larvae are largely inactive. Larvae were left in the same 96-well plate for assessing both CaMPARI and locomotion. Exposing larvae to photoconversion with LED light at 4 dpf had no impact on the mean locomotor activity at 5 dpf (Fig. 1f). Larvae anesthetized with MS-222 did not display any detectable swimbouts (Fig. 1f) and the level of neural activity was well correlated with locomotor activity (see below). Overall, these results show that CaMPARI optogenetic measures deployed in zebrafish larvae offer a reliable high-throughput tool for measuring neural activity.
CBD and THC reduce subsequent neural activity
We assessed how embryonic exposure to cannabinoids impacts neural activity later in development. Here, the doses of CBD and THC aligned with our previous work13, reflecting high cannabis consumption in humans. Comparisons of our dosage to humans requires various considerations: (1) blood plasma concentrations of THC can peak up to 0.25 mg/l while smoking a single cigarette35; (2) the content of THC in cannabis has increased in the past 20 years; and (3) doses of intraperitoneally administered medical CBD can vary greatly, from 5 to 100 mg/kg, with a daily maximum of 1500 mg/kg36,37. The current study uses up to 6 µg/mL of THC and 3 µg/mL of CBD and absorption studies using Liquid Chromatography-Tandem Mass Spectrometry suggest that an estimated 0.1–10% of toxic compounds will pass through the chorion to reach the embryo40,41.
To gain insight into the effects of CBD and THC separately (Fig. 2a), we applied a range of concentrations of each compound. This validation was necessary as CaMPARI has not been used previously to study the effects of cannabinoids. Compounds were added to the embryonic bathing media soon after egg fertilization at 0.5 hpf and were then washed out towards the end of gastrulation at 10 hpf. CaMPARI was imaged in 4 dpf larva and locomotor activity was assessed at 5 dpf (Fig. 2b). Animals that were exposed to CBD at concentrations of 2 µg/ml and 3 µg/ml exhibited a dose-dependent reduction in CaMPARI activity (Fig. 2c, d). These reductions in CaMPARI output are substantial when compared with the ~ 50% reduction we observed in anesthetized larvae, where little neural activity is expected. Coordinated with this, locomotion was also reduced starting at 1.5 µg/ml and was significant (p < 0.01) at 3 µg/ml (Fig. 2e). THC had a similar effect as CBD and also reduced neural activity at higher doses, 4 µg/ml and 6 µg/ml, (Fig. 2f, g). Locomotor activity exhibited a more extensive reduction at 6 µg/ml (Fig. 2h). We compared neural activity and locomotion in the same individuals exposed to effective doses of cannabinoids and found significant correlations (Fig. S2; CBD r = 0.52 (p < 0.01) and THC r = 0.71 (p < 0.01)). Indeed, most larvae that displayed reduced neural activity also showed reduced locomotion. Together, our findings show that both CBD and THC reduced neural activity when embryos were exposed early in development.
Antagonistic effects of CBD on neural activity are enhanced when combined with sub-effective doses of THC
Cannabis consumption during pregnancy exposes the fetus to CBD and THC in concert. In order to investigate whether CBD and THC have a combined effect that is different from either compound on its own, neural activity was measured when larvae were exposed to both cannabinoids simultaneously using the same timeline as the previous experiment (Fig. 2b). First, various sub-effective doses of CBD were applied from 0.5 to 1.5 µg/ml in the presence of 2 µg/ml THC (identified as sub-effective in Fig. 2). Preliminary dosage attempts revealed small, but significant differences with p < 0.01 and p < 0.05, respectively, on neural activity when combining 0.5 or 1.5 µg/ml CBD with 2.0 µg/ml THC (Fig. S3a, b). Locomotor activity also trended towards a dose-dependent reduction during concerted application of CBD and THC, which was significant at 1.5 µg/ml with p < 0.01 (Fig. S3c). Next, CBD and THC were added in a 1:1 ratio mixture at 1.0 and 2.0 µg/ml each. An enhanced antagonistic effect was revealed on neural activity when both CBD and THC were added at 2.0 µg/ml. CBD by itself at 2.0 µg/ml significantly (p < 0.05) reduced neural activity (Fig. 3d, e) consistent with results in Fig. 2d, but CBD and THC together, further reduced neural activity compared to CBD or THC alone (Fig. 3d, e; p < 0.01). The combined effects of CBD and THC seemed synergistic in that CBD by itself at 2.0 µg/ml, although significant, had a relatively small effect, and THC by itself at 2.0 µg/ml had no effect. However, when similar concentrations of CBD and THC were combined, neural activity was reduced to levels that were comparable with MS-222 anesthetized samples, suggesting an absence of neural activity. Locomotor activity was mainly affected by CBD with no additional reduction when compared to CBD and THC combined (Fig. 3f). No significant effect on neural activity or locomotor activity was obtained when 1.0 µg/ml of CBD and THC was applied (Fig. 3a–c).
Due to the effects on locomotion, we wanted to assess the integrity of neurons associated with locomotor networks. Reticulospinal neurons in the hindbrain were immunolabelled with the RMO44 antibody that targets NEFM, an established marker for neuronal damage42,43. Indeed, RMO44 immunolabelling was further reduced when CBD and THC were combined at 2.0 µg/ml suggesting that both CBD and THC affect neuronal health (Fig. S4). Together, these results suggest that CBD and THC are more potent in reducing neural activity when applied in combination.
CB1R and CB2R are both required for CBD-induced reduction of locomotor activity
Some reports suggest that CBD mediates its actions through CB1- and CB2-receptors25. To test whether these two receptors are involved in the neuronal effects mediated by CBD in vivo, we used the CB1R inverse agonist AM251 and the CB2R antagonist AM630. Previous findings support the specificity of CB1R and CB2R inhibitors in zebrafish44,45. For instance, a recent study showed that CB2Rs are required for the action of AM630 on zebrafish behaviour44. Importantly, the impact of AM630 on wildtype larvae was not detectable when applied to cnr2−/− mutant larvae (CB2R knockouts), thereby supporting drug specificity. Moreover, AM630 treated zebrafish showed a very similar photo-dependent response compared to the CB2R knockout44. Similarly, either knocking out CB1R or inhibiting CB1R with AM251 caused a similar response in rescuing locomotor activity when the activity was suppressed with the cannabinoid-receptor agonist, WIN55,212-245. In that study, the three metrics that were measured (acclimatization, dark-adaption, and recovery) for locomotion displayed arguably the same rescue-pattern in cnr1−/− mutant and AM251 treated larvae45. In the present study, both inhibitors were used in the nanomolar range, which is considered modest and which ensures a level of specificity towards their respective receptors (AM251, IC50 = 8 nM towards CB1R; AM630, IC50 = 31.2 nM towards CB2R). As described previously, the experimental timeline is as depicted in Fig. 2b, with CBD being added immediately after AM251 and AM630. When analyzing CaMPARI activity, our results revealed that the CBD-mediated reduction in neural activity could not be completely prevented with concentrations from 0.1 to 10 nM AM251 or AM630 in the presence of 3 µg/ml of CBD (Fig. 4a–d). In contrast, the CBD-mediated reduction of locomotor activity was prevented with a concentration of AM251 as low as 1 nM (Fig. 4e) or with 10 nM of AM630 (Fig. 4f). Importantly, neither inhibitor had an effect on locomotion when used on their own, in the absence of CBD (Fig. 4g, h). This finding excludes the possibility that the effects of AM251 or AM630 on locomotion were independent of CBD and rather specific towards CB1R and CB2R, respectively. Taken together, these findings suggest a role for both CB1R and CB2R in the CBD-mediated effects on locomotor activity and that inhibiting CB1R and CB2R can ameliorate the negative impacts of early CBD on late CNS function, at least with respect to locomotion.
Discussion
In some jurisdictions, including Canada, cannabis has been legalized despite the lack of mechanistic understanding and health implications. Cannabis consumption has increased in the past decade46,47 and recent reports associate prenatal cannabis consumption with stillbirths and the neurodevelopmental syndrome, autism spectrum disorder48,49. Yet, cannabis and CBD oil have been used to treat labor pain50. In animal models, such as rats and chicken, cannabis has been shown to have detrimental effects on developing embryos10,51. Recent evidence reveals that early exposure to CBD and THC impacts the overall development in zebrafish13. Interestingly, the effects of CBD and THC were delayed and were observed much later in development after the exposure had ended.
In vitro, CBD and THC can function through CB1R and CB2R, which are expressed in neuronal tissues17–25. A goal of this study was to investigate the effects of CBD and how it might interact with THC during early development to impact neural activity in older larvae. We implemented CaMPARI transgenic zebrafish as a read-out of neural activity and supplemented this approach by measuring the locomotor activity of the same larvae. CaMPARI has proven to be a relatively straightforward and economical neural integrator that can be used in a well-format with the zebrafish animal model. We found that neural activity was reduced in zebrafish larvae following brief exposure to THC and CBD early in development. Surprisingly, CBD had similar impacts on neural activity at lower concentrations compared to THC. Notably, the impacts of CBD were increased by sub-effective doses of THC. Our data quantifying locomotion suggested that these CBD-mediated effects involve CB1R and CB2R.
CaMPARI assay
We optimized the workflow and photoconversion of CaMPARI such that this innovation can be scalable to a large number of compounds in future studies. Zebrafish larvae are often used in a 96-well plate format for quantifying locomotion39,52–56. Nevertheless, locomotor activity is not a direct measurement of neural activity. Calcium sensors, such as the GCaMP series57,58, the more recently introduced GECO59,60, and CaMPARI, offer opportunity for more direct measures of neural activity. Importantly, CaMPARI provides a temporal-snapshot of neural activity in the freely swimming animal, which can be anesthetized and imaged after the experiment34. Furthermore, the ratiometric red/green signals are imaged together and mitigate inter-individual differences arising from different expression levels or variances due to tissue depth. We optimized conditions for a well-plate set-up analyzing neural activity in the optic tectum and hindbrain. Future work might refine the method to quantify less defined brain areas or specific neurons.
In general, there seems to be a good correlation between neural activity measured by CaMPARI and locomotor activity61,62. However, some findings may be more sensitive towards one assay vs another, as described above. Furthermore, when interpreting our data, it is important to consider that when neural activity is very low, for instance when anesthetized, a clear red to green CaMPARI ratio can be still measured. This is unlikely due to autofluoresence because larvae that were not photoconverted yielded zero CaMPARI activity. We cannot exclude the possibility that due to longer photoconversion times some CaMPARI was converted into the red fluorescence version even when anesthetized. It is well established that when anesthetized with MS-222, neurons are barely active63,64. Indeed, the behavioural assay shows no locomotion in the presence of MS-222 and we used this assay as a supplement in this study. Thus, we believe this method is a suitable technique to measure substantial changes in neural activity in the larger parts of the zebrafish CNS. For future studies in connection to epilepsy this innovation may be particularly helpful as some studies use convulstants, PTZ or 4-AP, to simulate seizures62,65. Indeed, we found that both, PTZ and 4-AP, induce CaMPARI activity. Altogether, the advantages with this method outweigh its limitations and could serve for future studies, especially in connection with epilepsy.
Several competing technologies were considered prior to designing this study, in the hope that we could build upon and complement these potent approaches. Other than behavioural assays described above, the GCaMP series has been of interest57,58. However, the larvae have to be analyzed with high-resolution live microscopy during the experiment and movement of the larvae would add considerable variability. The assays would be data-intensive and technically challenging, particularly where large sample sizes and different experimental groups are difficult to obtain (especially in vivo). Additionally, we were interested in measuring outputs from freely behaving animals because constraining them for imaging would likely impact the results and would be time-consuming. Other widely used markers for neural activity include the immediate-early gene c-fos and phosphorylation of ERK65,66. However, CaMPARI is more temporally specific due to its immediate photoconversion. In addition, both markers (c-fos and ERK) are not only less tissue specific, but also respond to stress and other factors. The elavl3 promoter that drives CaMPARI ensures CNS-specific expression. Overall, CaMPARI is a practical direct output of neural activity and thus the chosen innovation for this study.
Delayed implications after early embryonic exposure of CBD and THC
CBD and THC have both been implicated in negatively impacting embryonic neurodevelopment. Later abnormalities following acute early exposure to these cannabinoids have only recently been investigated13. Here, early exposure of CBD and THC during the first 10 h of embryonic development reduced neural activity in a dose-dependent manner when measured 4 days later. These later effects may be due to abnormal neurodevelopment and/or the slow release of the cannabinoids from the tissue due to their lipophilic nature. Consistent with this finding, zebrafish larvae treated with a similar exposure (5–10 hpf) to cannabinoids have shown effects that may be due abnormal development (at 5 dpf). Those effects were not exclusive to neurons, but also included morphological abnormalities, decreased survival rate, decreased mEPCs activity (may relate to muscle development), decreased heart rate and delayed hatching13. Considering the specific time-window of exposure (first 10 h in development), an abrupt effect on the developmental program including neurons seems plausible. Interestingly, CBD seems to impact neural activity at a lower concentrations than THC, as observed herein and during our previous study13. It is important to acknowledge that CBD consumption has been documented to help with seizures and cancer by reducing pain67,68. Nevertheless, our study suggests caution regarding its use during pregnancies.
This study reveals some complexity when assessing the impacts of CBD and THC together. Their combined effects were more potent in decreasing later neural activity, locomotion and immunolabelling of reticulospinal neurons, suggesting there might be an impact on neuronal integrity42,69. Overall, these findings were surprising because CBD has been found to negate the effects of THC on CB1R27. One potential explanation is that at higher concentrations both CBD and THC alter the developmental program, which could underlie the abnormal morphologies reported previously13. In this study, we observed similar morphologies, which may have affected our results. However, because there were decreases in neural activity and locomotion at concentrations that did not show major morphological changes13, it is unlikely that the effects of CBD and THC are due to abnormal morphologies alone. Moreover, it is possible that the impact of higher concentrations of CBD and THC on developing brains is different compared to lower concentrations on fully developed brains. Finally, CBD and THC serve as ligands for several types of receptor with different affinities25, and another possibility is that at high concentrations, CBD and THC affect receptors other than CB1R and CB2R.
Aside from the impact of CBD and THC together, an additional interesting finding is that the most prominent action of early exposure to CBD is on subsequent locomotion. It remains unclear if that is due to CBD impacting muscle or other developmental aspects. It is also likely that the intertwined effects of CBD and THC are dose-dependent and past observations from higher CBD and THC doses suggest that the development of additional morphological deficits may be important for future considerations.
Potential mechanism involved in the delayed effect of CBD and THC
At this stage, it is unclear what mechanisms account for the reduction in neural activity later in development when embryos are exposed briefly to CBD and THC in the first hours after fertilization. We were especially interested in the less widely studied CBD. The potential mechanisms of acute cannabinoid exposure on later neuromuscular outputs are diverse. Using HEK cells several receptor candidates have been identified that were shown to interact with CBD25. A recent study supports a functional role for the actions of CBD through CB1R and CB2R in zebrafish32. Herein, CB1R and CB2R were found to be required for the impact of CBD on locomotor activity. Nanomolar concentrations of receptor inhibitors partially rescued CBD effects. Previous studies have implied that CBD may bind to both CB1R and CB2R in vitro25 and the data in this study support these interactions in vivo. CBD has also been shown to cause abnormal craniofacial and brain development in mice and zebrafish, which seems to be mediated through CB1R and the hedgehog signaling pathway16. CB1R and CB2R are expressed very early in development on stem cell progenitors70. Therefore, developmental programs may be impacted such that changes persist until later, even when cannabinoids are no longer present in the circulation. Such changes may not be specific to the CNS as CBRs are not only expressed in neuronal progenitor cells, but also on stem cells forming other types of tissues70. This could explain our previous findings where CBD affected heart-rate and overall morphology after 5 dpf when exposed only briefly during gastrulation11. This study shows that CB1R and CB2R inhibition had little detectable effect on CBD impacting later neural activity. This might be an artifact of our CaMPARI method measuring general CNS activity in large areas, rather than resolving motor axons in detail, which have been shown to be affected previously13. We also acknowledge that pharmacological inhibitors may not completely block the CB1Rs and CB2Rs, or that they may have off-target effects. Finally, it is also possible that other CBD receptors such as GPR55 and TRPV1 are involved26, but this needs further investigation.
Conclusion
In conclusion, both CBD and THC exposure during early embryonic stages had a negative impact on later neural activity, which was synergistic when the cannabinoids were combined. This suggests that the impact of cannabis on early development is likely greater than when using isolated compounds. Our findings also support an in vivo mechanism of CBD functioning through CB1R and CB2Rs. This study opens up a new path for investigating detailed mechanisms involving these and other receptors, especially in neurons that are affected by cannabinoid exposure early in development. Considerable evidence is accumulating that warns prenatal cannabis consumption can have negative consequences on neurodevelopmental disorders3 and further studies are warranted to fully understand the impacts on human development.
Methods
Animal ethics and exposure to drugs
Zebrafish maintenance was approved by the Animal Care & Use Committee: Biosciences at the University of Alberta and operated under the guidelines of the Canadian Council of Animal Care. The fish were maintained in the University of Alberta fish facility at 28 °C under a 14/10 light/dark cycle as previously described71 and all authors complied with ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines72.
CaMPARI transgenic zebrafish were outcrossed with Casper strains not carrying the transgene. For the experimental setup, embryos were randomly collected and placed in egg water (60 µg/ml Instant Ocean) in groups of 20 embryos as early as practically possible within 0.5 hpf. Stock solutions of CBD and THC were obtained from Sigma at 1 mg/ml dissolved in methanol. Receptor inhibitors AM251 (Selleck Chemicals, Houston, TX, USA) and AM630 (Adooq Bioscience, Irvine, CA, USA) were dissolved in DMSO. Accordingly, zebrafish embryos were treated with vehicles or compounds as close to 1 hpf as practically possible and removed towards the end of gastrulation (10 hpf). Within experiments all treatments contained the same amount of methanol or DMSO, specific percentages are indicated in the figure legends. Compounds were removed with several washes and egg water was replaced every day until CaMPARI analysis. Not all larvae carried the transgene and were screened for green fluorescence. An increase in mortality and abnormal morphology was observed when treated with higher doses of CBD or THC, as reported previously13.
Engineering CaMPARI transgenic zebrafish for integrative calcium imaging
CaMPARI photoconversion
For photoconversion, bright green Tg[elavl3:CaMPARI (W391F + V398L)]ua3144 larvae at 4 dpf were placed in a 96-well plate containing 150 µl egg water (made as previously decribed71) and acclimatized for 2 h. The central 48 wells of the plate were exposed to 405 nm LED array (Loctite) for 300 s at a distance of 10 cm. LED array illuminated the plates entirely and evenly. The plates were floating in water at room temperature to ensure that the larvae did not overheat. For imaging, the larvae were anesthetized in 0.24 mg/mL tricaine methanesulfonate (MS-222, Sigma) following photoconversion. MS-222 was washed out and zebrafish were rested for a day prior to analysis of locomotor activity. Previously, CaMPARI was not optimized for well-plate-formats because a 10 s photoconversion did not yield sufficient red fluorescing CaMPARI to be detected reliably. The method used in this study does not require a multi-photon confocal microscope where the larvae need to be carefully embedded in agarose and placed against the microscope slide. Instead, the larvae can be imaged directly from the well after MS-222 treatment. This assay also complements more straightforward and more economical fluorescing microscopes for labs with a lower dedicated budget.
CaMPARI analysis
The ratio of red to green fluorescent emissions (red photoconverted CaMPARI in ratio to green CAMPARI) was interpreted as relative neural activity, as previously defined34. Images were objectively obtained with an InCell 2000 microscope (GE Healthcare, US). 96-well plates were run using FITC (1 s exposure) and dsRed (2 s exposure) channels. Blinded experimentalists used ImageJ to quantify the fluorescence mean of the red (dsRed) and green (FITC) channels. The area of the optic and hindbrain area were quantified as those regions were most consistent in the view. A mask was applied to exclude noise outside of the CNS.
RMO44 staining
The zebrafish used in this experiment were of the Tuebingen Longfin strain. Set-up was similar as described above. Following the experiment, larvae were stained with anti-RMO44 (Thermo Fisher Scientific, Waltham, MA, USA) and imaged using a Zeiss LSM 710 confocal microscope (Oberkochen, Germany) as described elsewhere76.
Statistics
GraphPad Prism Software (Version 7, GraphPad, San Diego, CA) was used to analyze statistics from data obtained from our CaMPARI assay or tracking locomotor activity. All the data were presented as mean ± SEM (standard error of the mean). Statistical significance of p < 0.05 was determined using a non-parametric t-test between two groups followed by a Mann–Whitney analysis, if appropriate. Multiple groups were compared using One-Way ANOVA with Dunnett's multiple comparisons test. Pearson r from correlation analysis was also determined where indicated.
Supplementary Information
Untitled section
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-021-90902-3.
Acknowledgements
This study was supported by grants to W.T. Allison from Alberta Prion Research Institute-Alberta Innovates BioSolutions and the Alzheimer Society of Alberta and the Northwest Territories, and to D.W. Ali from the Natural Sciences and Engineering Research Council of Canada and Alberta Innovates mCannabis. L.F. Locskai was supported by undergraduate studentships from Natural Sciences and Engineering Research Council of Canada and from Alberta Innovates Health Solutions. R. Kanyo was supported by a SynAD postdoctoral fellowship funded via Alzheimer Society of Alberta and Northwest Territories through their Hope for Tomorrow program and the University Hospital Foundation. We would like to also extend our thanks to members of the Allison and Ali lab. In addition, we would also like to acknowledge the Sciences Animal Support Services at the Unversity of Alberta.
Competing interests
The authors declare no competing interests.