Respiratory depression and analgesia by opioid drugs in freely behaving larval zebrafish

When it comes to treating severe pain, a doctor’s arsenal is somewhat limited: synthetic or natural opioids such as morphine, fentanyl or oxycodone are often one of the only options available to relieve patients. Yet these compounds can make breathing slower and shallower, quickly depriving the body of oxygen and causing death. This lethal side-effect is particularly devastating as opioids misuse has reached dangerously high levels in the United States, creating an ‘opioid epidemic’ which has claimed the lives of over 80,000 Americans in 2020. It is therefore crucial to find safer drugs that do not have this effect on breathing, but this research has been slowed down by the lack of animal models in which to study the effect of new compounds.

Zebrafish are small freshwater fish that reproduce and develop fast, yet they are also remarkably genetically similar to mammals and feature a complex nervous system. However, it is not known whether the effect of opioids on zebrafish is comparable to mammals, and therefore whether these animals can be used to test new drugs for pain relief.

To investigate this question, Zaig et al. exposed zebrafish larvae to fentanyl, showing that the fish then exhibited slower lower jaw movements – a sign of decreased breathing. The fish also could also tolerate a painful stimulus for longer, suggesting that this opioid does reduce pain in the animals. Together, these results point towards zebrafish and mammals sharing similar opioid responses, demonstrating that the fish could be used to test potential pain medications. The methods Zaig et al. have developed to establish these results could be harnessed to quickly assess large numbers of drug compounds, as well as decipher how pain emerges and can be stopped.

The current North American opioid epidemic is staggering, with millions of emergency visits and over 45,000 deaths annually (Florence et al., 2016). Synthetic opioids like fentanyl and oxycodone, or natural opioids like heroin and morphine are highly addictive (Wilkerson et al., 2016) and can lead to respiratory depression (Dahan et al., 2010; Montandon et al., 2016a; Nagappa et al., 2017), that can be lethal with overdose (Gomes and Juurlink, 2016; Jones et al., 2013). Opioids are critical pain therapies and there are currently no other medications that offer effective relief of pain (IASP, 2018). However, the side-effects of opioids limit their effective use and may lead to sub-optimal pain treatments.

Opioid medications act on µ-, δ-, and κ- opioid receptors (Gutstein, 2001), which are expressed in discrete brain circuits. Although a wide range of opioids targeting these opioid receptors exist, the most potent opioids, such as fentanyl and oxycodone, are largely selective for the µ-opioid receptors (MORs) (Sora et al., 1997). Because MORs are expressed in neural circuits involved in breathing and nociception, opioids present effects beyond their intended therapeutic analgesic properties. MOR drugs induce side-effects such as addiction (Le Merrer et al., 2009), sedation (Montandon et al., 2016a), and hypoventilation (Dahan et al., 2010; Montandon and Slutsky, 2019; Nagappa et al., 2017). Hypoventilation is characterized by reduced respiratory rate (Macintyre et al., 2011), reduced airflow (Ferguson and Drummond, 2006), prolonged apneas (Mogri et al., 2008; Nagappa et al., 2017), and severe hypoxemia (Brown et al., 2006). Complete respiratory arrest and bradycardia are the main causes of death by opioid overdose (Pattinson, 2008). Importantly, respiratory depression and cessation of breathing with opioid overdose are due to the action of opioid drugs on brainstem respiratory network (Montandon and Slutsky, 2019). The only available treatment to reverse respiratory depression during an opioid overdose is the MOR antagonist naloxone which directly blocks the binding of opioid ligands to MORs. However, naloxone also blocks the analgesic properties of opioid drugs, so it cannot be used as an adjunct treatment. The discovery of novel opioid drugs or combinations of drugs with potent analgesic properties without the side-effects of respiratory depression has been hindered due to the lack of animal models allowing behavioral assessments of pain and breathing, while allowing for large-scale drug discovery.

Identification of new molecular targets and drugs is critical to the development of safe opioid therapies. Due to the complexity of the machinery regulating MOR inhibition, wide scale screening of drugs acting on the known mechanisms regulating MOR inhibition in rodent models is not feasible. It is therefore critical to identify amenable model systems where large-scale screening can be developed, while preserving the complexity of vertebrate central nervous system and behaviours. The challenge is to quantify complex phenotypes like respiratory neural activity and nociception on a large scale, which is normally limited by the complexity of the animal models. Rodents are not optimal because their large size, generation time, and drug availability limit screen scale. An alternative is the larval zebrafish, which shares anatomy, physiology, and large parts of the genome with humans. The small size of larval zebrafish, ease of production, and their external fertilization which allows easy gene knockdown, makes the zebrafish an attractive model organism to screen genes and drugs. Here, we propose to use the larval zebrafish as simple, yet powerful, animal model, so that the molecular mechanisms of MOR respiratory and nociceptive inhibition can be identified. Importantly, the zebrafish larvae can also be used for large-scale drug screening to identify new drug candidates.

The larval zebrafish has emerged as an ideal model system for drug and gene discovery since it combines the biological complexity of in vivo models with a nervous system homolog to humans including a respiratory neural network. Although fish use a different strategy to absorb oxygen and eliminate carbon dioxide than mammals, they rhythmically produce mandibular movements to move water through their gills. In lampreys, the respiratory neural network includes the paratrigeminal respiratory group (pTRG) and the vagal nucleus and this network generates rhythmic mandible movements (Bongianni et al., 2016). Because of their close evolutionary origins (Missaghi et al., 2016), the respiratory network shares similarities with the mammalian respiratory network (Cinelli et al., 2013) which generates breathing and regulates respiratory depression by opioids (Montandon et al., 2011). Interestingly, the pTRG presents similar functional properties than the mammalian respiratory network such as sensitivity to substance P (Mutolo et al., 2010) and to opioid ligands (Mutolo et al., 2007). Here, we propose that respiratory mandible movements can be used as an index of respiratory network activity, similar to respiratory activity of the trigeminal muscle in mammals (Jacquin et al., 1999). The zebrafish also possesses a nociceptive system encompassing spinal cord, brainstem and sub-cortical circuits (Taylor et al., 2017), with homology to mammals. Although pain circuit activity and its response to opioids cannot be directly assessed, the swimming escape response to nociceptive stimuli can be easily assessed in larval zebrafish. Nociceptive stimuli such as formalin (Magalhães et al., 2017) or (allyl)-isothiocyanate (AITC, also known as ‘mustard oil’) can be administered to larval zebrafish. AITC acts on transient receptor potential ankyrin 1 channels and transient receptor potential cation channel vanilloid receptor 1 (Oda et al., 2016) which are involved in nociception to chemical compounds (Bamps et al., 2021). Here, we propose to determine whether respiratory depression and analgesia can be measured in larval zebrafish.

Most opioid analgesics and drugs of abuse elicit their effects by binding to MORs (Gutstein, 2001). In zebrafish, the MOR has high homology to the mammalian MOR and shares 74% of its amino acid sequence with its mammalian counterpart (Sanchez-Simon and Rodriguez, 2008). The MOR also possesses similar binding properties to the mammalian MOR for morphine and DAMGO (Marron Fdez de Velasco et al., 2009; Mutolo et al., 2007). Because of the homologies of the respiratory, pain, and opioid systems in zebrafish and mammals, we propose that the larval zebrafish may mimic respiratory depression and analgesia by opioids and other phenotypes associated with opioid drugs in mammals. The objectives of this study are to demonstrate that larval zebrafish have evolutionary conserved opioid properties mimicking mammals and that they can be used to investigate opioid-induced respiratory depression and analgesia. We aim to demonstrate that larval zebrafish replicates the opioid pharmacology observed in mammals and humans and that current pharmacotherapies to block respiratory depression can be tested in larval zebrafish.

Zebrafish use gills to promote gas exchange with water. Flow of water through the gills is generated by a complex respiratory network in the brainstem which presents similar anatomy and properties to the mammalian respiratory network (Figure 1A). The respiratory network generates rhythmic movements and rhythmically activate muscles which promotes water flow through the gills. One of these muscles, the mandible muscle controls opening and closing of the mandible and is activated by the mandible nerve which originate from the trigeminal nerve. Mandible movements can be easily recorded using a high-definition camera positioned on top of the swimming zebrafish (Figure 1B,C), and is a robust index of respiratory network activity. Mandible movements can be quantified by counting the changes in pixel intensity around the mandible area (Figure 1B) and can be displayed over time (Figure 1D). The rate of mandible movements, identified as the number of mandible movement peaks (red circles, Figure 1D), indicates respiratory network rhythm, and is considered here as an index of respiratory rate. Using the convenience of multi-well plates, we administered combinations of drugs to embryo medium as described in Figure 1E. Combinations of drugs were administered to separate groups of fish. Each animal only received one combination of drugs and did not receive consecutive combinations of drugs because it would not be feasible to remove drugs from the water.

Figure 1

Download asset Open asset

To determine whether zebrafish can be used as a model of opioid-induced respiratory rate depression, we administered embryo water (control) or fentanyl - a commonly used opioid analgesic - to larval zebrafish while quantifying the rate of mandible movements (Figure 2A). Larval zebrafish under control conditions (embryo water) presented rates of mandible movements ranging from 15 to 115 movements per minute. Larvae were exposed to a concentration of fentanyl of 1 µM and respiratory rate was recorded over a 30 min time-period (Figure 2B). In the control group, respiratory rate initially increased compared to baseline but did not significantly change during the 30 min following baseline (Figure 2B). In the fentanyl group, respiratory rate was strongly reduced 4 min after fentanyl administration. Since the strongest respiratory rate depression was observed 5 min after exposure, we compared respiratory rates (measured over a one-minute time-period between minutes 5 and 6) of control and fentanyl groups (Figure 2C). Fentanyl (0.2, 0.4 µM) did not significantly decrease rate of mandible movements compared to control (p=0.473 and p=0.797, Figure 2C). Fentanyl (1 and 3 µM) significantly decreased rate compared to control groups (p=0.006 and p=0.038, Figure 2C). When data were normalized according to baseline rate, variability within groups was significantly reduced (Figure 2D). Fentanyl significantly decreased normalized rate at 1 µM (p<0.001) and 3 µM (p<0.001) compared to the control group (Figure 2E), but not at 0.2 µM (p=0.275) and 0.4 µM (p=0.200). In summary, fentanyl-induced respiratory rate depression was observed at 1 and 3 µM of fentanyl tested. We determined that such normalization best represented respiratory rate depression by fentanyl and we only presented normalized data for the subsequent figures with respiratory depression.

Figure 2

Download asset Open asset

Figure 2—source data 1

Respiratory rate depression by the opioid analgesic fentanyl.

https://cdn.elifesciences.org/articles/63407/elife-63407-fig2-data1-v2.xlsx

Download elife-63407-fig2-data1-v2.xlsx

To determine the pharmacology of opioid receptors in zebrafish, we compared groups of fish exposed to fentanyl with fish exposed to fentanyl and the MOR antagonists naloxone and CTAP (Figure 3). Exposure to fentanyl reduced the rate of mandible movements (Figure 3A). The MOR antagonist naloxone at 10 µM did not significantly reverse respiratory rate depression induced by fentanyl (p=0.178, Figure 3B). However, naloxone (20 µM) significantly reversed respiratory rate depression by fentanyl (p<0.001, Figure 3B). The selective MOR antagonist CTAP (4 µM) did not significantly reverse respiratory rate depression by fentanyl (p=0.077, Figure 3B), although a trend toward significance was observed. To determine whether naloxone or CTAP may increase respiratory rate when administered alone, we compared groups of fish exposed to CTAP (4 µM) or naloxone (5 and 20 µM) to control fish (Figure 3C). Neither CTAP nor naloxone showed effects on respiratory rate when administered alone (one-way ANOVA, p=0.122). In summary, fentanyl induced a significant respiratory rate depression which was blocked by the MOR antagonist naloxone but not CTAP.

Figure 3

Download asset Open asset

Figure 3—source data 1

Opioid receptor pharmacology regulating respiratory depression.

https://cdn.elifesciences.org/articles/63407/elife-63407-fig3-data1-v2.xlsx

Download elife-63407-fig3-data1-v2.xlsx

In the previous experiments, naloxone was administered concomitantly with fentanyl and prevented respiratory depression (Figure 3A,B). We then tested whether naloxone can completely reverse respiratory depression by fentanyl (1 µM) when administered after respiratory depression (Figure 3D). As expected, fentanyl significantly depressed respiratory rate (p=0.009), an effect not reversed by naloxone at 9 µM (p=0.111). Although a sequential approach may be of interest due to its repeated measure design, it presented two challenges. First, drugs were pipetted twice in the multi-well plate and affected the fish’s behavior, which may explain the large variability observed with naloxone. Second, due to the low volume of fish water in wells and volumes of stock solution administered, it was challenging to increase naloxone concentration in these experiments without diluting the concentration of fentanyl. We therefore concluded that sequential administration of drugs was not the most consistent and compared drug combinations using separate groups of fish.

Because zebrafish strains may present genetic variability and different sensitivities to opioid drugs (Marron Fdez de Velasco et al., 2009; Sanchez-Simon and Rodriguez, 2008), we compared respiratory depression by fentanyl in various zebrafish lines (Figure 3E). Fentanyl (1 µM) significantly reduced respiratory rate in AB fish as previously determined in previous experiments (p<0.001), but not in Tübingen (TU) (p<0.001) or AB x TU zebrafish crosses (p<0.001). To conclude, only the AB zebrafish strain presented sensitivity to fentanyl and was used for subsequent experiments. In summary, respiratory rate was depressed by fentanyl through activation of opioid receptors, an effect reversed by naloxone.

Morphine is a widely-used opioid drug that induces less severe respiratory depression than the highly potent opioids such as fentanyl (Gutstein, 2001). We tested whether morphine induces respiratory rate depression in zebrafish. There was no dose-dependent decrease in respiratory rate with increasing morphine dosages (1, 10, 20, 50 and 200 µM, p=0.088, Figure 3F). However, multiple comparison tests showed that morphine at 1 µM significantly depressed respiratory rate (p=0.021) when compared to controls, but not at the other concentrations tested (p>0.488). These results suggest that morphine depressed respiratory rate at low concentration, but that at higher concentrations, morphine may induce an excitatory effect.

To determine whether breathing in zebrafish larvae can be stimulated by specific excitatory drugs as observed in mammals, we administered two respiratory stimulants previously shown to stimulate breathing when depressed by opioids. We first administered the 5-HT_4A serotoninergic ligand BIMU8 (Manzke et al., 2003) with or without fentanyl (Figure 4A). BIMU8 (10 µM) combined with fentanyl did not show reversal of respiratory rate when compared to the fentanyl group (p=0.193). Fish exposed to BIMU8 alone showed significant higher respiratory rate compared to fentanyl alone (p=0.008) but was not different from the control group (p=0.294). BIMU8/fentanyl showed significantly lower respiratory rates compared to BIMU8 alone and control (p=0.018 and p=0.007). In summary, the addition of BIMU8 to fentanyl did not significantly reversed respiratory depression by fentanyl.

Figure 4

Download asset Open asset

Figure 4—source data 1

Respiratory depression and respiratory stimulants.

https://cdn.elifesciences.org/articles/63407/elife-63407-fig4-data1-v2.xlsx

Download elife-63407-fig4-data1-v2.xlsx

The AMPA positive allosteric modulator CX614 significantly reversed respiratory rate depression by fentanyl (p<0.001, Figure 4B). CX614/fentanyl showed higher respiratory rate than fentanyl alone, as well as higher respiratory rate than control (p=0.021), suggesting that CX614 at this dosage reversed respiratory depression by fentanyl. CX614 alone did not present higher respiratory rate when compared to control group (p=0.058) but showed a trend toward significance. Higher dosages of CX614 induced seizure and were not tested. To determine whether the respiratory depression observed with fentanyl was due to a direct effect of fentanyl on respiratory rate rather than an indirect effect on sedation (Montandon and Horner, 2019) or pain circuits (Jiang et al., 2004), we administered lidocaine, an analgesic not acting on MORs (Lopez-Luna et al., 2017). Lidocaine did not significantly change respiratory rate compared to control (p=0.267, Figure 4C). To dissolve drugs, we used DMSO as a solvent which may affect fish respiratory rate. DMSO administered at 0.0016% did not change respiratory rate (p=0.342).

To determine the analgesic properties of opioid drugs in larval zebrafish, we established a simple model of nociception combining exposure to formalin as a nociceptive stimulus and measurements of the subsequent swimming escape response. Using video recordings of larvae positioned in a multi-well plate, we tracked the individual movements of larvae exposed to embryo medium, formalin and a combination of formalin and fentanyl (Figure 5). We quantified two swimming behaviors (Figure 5A): swimming velocity (swimming distance in mm per second) and angular velocity (turn angle in degrees per second, Figure 5—figure supplement 1). In larvae exposed to formalin, swimming velocity was substantially increased by formalin (Figure 5B). Fentanyl combined with formalin showed considerable reduction in velocity compared to formalin alone (Figure 5B). Following exposure to formalin, swimming velocity significantly increased during the initial 3 min (Figure 5C, shaded blue), a response not observed in control larvae or formalin/fentanyl larvae (Figure 5C). We then compared the averaged swimming velocity during the first 3 min following drug exposure for the different groups of larvae. The formalin group presented significantly faster velocity compared to the control group (p<0.001, Figure 5D). Velocity in the formalin/fentanyl group was significantly slower than in the formalin group (p=0.037), suggesting that fentanyl reduced the escape response to formalin, which may indicate an analgesic effect. However, this effect may be due to the effects of fentanyl directly on swimming. Fentanyl administered alone did not decrease velocity when compared to control (p=0.221), showing that it did not affect swimming by itself. We then determined the effects of the µ-opioid receptor antagonists naloxone and CTAP (Figure 5D). In naloxone groups, velocities were not significantly higher than formalin/fentanyl (all p>0.122), suggesting that naloxone did not block the analgesic effects of fentanyl. In CTAP/formalin/fentanyl group, there was no difference compared to formalin/fentanyl or formalin alone (p=1.000 and p=1.000). Naloxone/formalin/fentanyl showed higher velocity than fentanyl or control groups (Figure 5D). These effects may be due to the impact of naloxone alone at 50 µM (Figure 5E). We then looked at the effect of formalin on angular velocity. The formalin group showed reduced angular velocity compared to control (p<0.001, Figure 5—figure supplement 1), an effect not reversed by fentanyl (p=0.151), which suggest angular velocity cannot be used to assess the escape response to formalin.

Figure 5 with 1 supplement see all

Download asset Open asset

Figure 5—source data 1

Nociception by formalin and analgesia by fentanyl.

https://cdn.elifesciences.org/articles/63407/elife-63407-fig5-data1-v2.xlsx

Download elife-63407-fig5-data1-v2.xlsx

Although naloxone did not prevent or reverse to analgesic effects of fentanyl, it may be due to the large variability in swimming response induced by formalin. In fact, swimming velocity in formalin/fentanyl/naloxone ranged from 0.3 to 4 mm/sec, which is substantially larger than the variability observed in control or fentanyl groups (Figure 5D). To induce more consistent nociceptive response, we administered (allyl)-isothiocyanate (AITC), a chemical acting AITC on transient receptor potential ankyrin 1 channels and transient receptor potential cation channel vanilloid receptor 1 (Oda et al., 2016). We used a similar experimental approach than we used with formalin by combining AITC, fentanyl and/or naloxone. AITC (100 µM) induced a strong increase in velocity compared to controls, an effect reduced by fentanyl (Figure 6A). AITC produced an initial increase in swimming velocity during the first 3 min following drug exposure, which is consistent with the formalin response. Conversely, in fish with a combination of fentanyl (6 µM) and AITC (100 µM), no swimming response was observed compared to AITC alone (Figure 6B), showing that fentanyl reduced the nociceptive response to AITC. Compared to the concentration of fentanyl used in the formalin assay (3 µM), a higher concentration (6 µM) was necessary to eliminate the effects of AITC. This concentration of fentanyl (6 µM) did not decrease velocity when administered alone compared to controls (Figure 6B). Averaged velocities in AITC fish were significantly higher than in controls (p<0.001, Figure 6C), and this increase was significantly lower in fish exposed to fentanyl and AITC (p<0.001). Angular velocity was increased by AITC, but not reversed by fentanyl (Figure 6—figure supplement 1). Interestingly, groups with naloxone at 20 and 50 µM combined with AITC and formalin presented higher swimming velocities than groups exposed to AITC/fentanyl (p=0.003 and p<0.001, respectively), whereas CTAP did not (p=0.342). In summary, AITC produced a fast increase in swimming velocity, an effect significantly reduced by fentanyl, which was reversed by naloxone.

Figure 6 with 1 supplement see all

Download asset Open asset

Figure 6—source data 1

Nociception by AITC and analgesia by fentanyl.

https://cdn.elifesciences.org/articles/63407/elife-63407-fig6-data1-v2.xlsx

Download elife-63407-fig6-data1-v2.xlsx

To determine whether larval zebrafish are sensitive to other types of analgesics, we exposed larvae to lidocaine, a non-opioid anesthetic (Figure 7A). As previously shown, formalin increased velocity compared to control (p<0.001, Figure 7B). In fish exposed to lidocaine and formalin, swimming velocity was significantly reduced compared to formalin group (p=0.002). Swimming velocity in lidocaine group was not different from control (p=0.693). Since we used a relatively high concentration of DMSO (1%) to dissolve lidocaine, we compared whether DMSO (1%) was different from control fish and we did not observed differences between the two groups (p=0.879, Figure 7A). Lidocaine did not affect angular velocity compared to formalin alone (Figure 7—figure supplement 1). In conclusion, two different types of analgesics (fentanyl and lidocaine) reduced the swimming response to formalin, therefore suggesting that these assays may properly assess analgesia.

Figure 7 with 1 supplement see all

Download asset Open asset

Figure 7—source data 1

Analgesic profiles of larval zebrafish.

https://cdn.elifesciences.org/articles/63407/elife-63407-fig7-data1-v2.xlsx

Download elife-63407-fig7-data1-v2.xlsx

The two stimulants BIMU8 and CX614 were administered to zebrafish to determine whether they can reverse respiratory depression by fentanyl. Only CX614 was able to significantly reverse respiratory depression. We therefore determine whether BIMU8 and CX614 were also reversing the analgesic properties of fentanyl. The combination of BIMU8 (20 µM)/formalin/fentanyl did not significantly change swimming velocity compared to the control (p=1.000) or formalin/fentanyl group (p=1.000, Figure 7B). However, CX614 (5 µM) /formalin/fentanyl group presented higher swimming velocities than the control group (p<0.001) and formalin/fentanyl (p=0.007), therefore showing that the ampakine CX614 reversed both respiratory depression and analgesia by fentanyl. Finally, Tübingen fish did not show respiratory depression by fentanyl. We therefore tested whether this strain is similarly insensitive to fentanyl in our analgesia assay. No difference in swimming velocity was observed between formalin and formalin/fentanyl group in TU fish (p=0.132), (Figure 7C), demonstrating that fentanyl did not reduce nociception elicited by formalin.

Opioid drugs are widely used as analgesics but present the severe side-effect of respiratory depression that can be lethal with overdose. No safe opioid therapy currently exists to treat severe and chronic pain. To identify new potent pain treatments without side-effects, opioid drug discovery needs to explore new research avenues. Novel approaches to drug discovery have been recently used including structure-based drug discovery using computer simulations (Manglik et al., 2016) or cell assays (Winpenny et al., 2016). Most drug discovery approaches use target-based approaches which exclude many drug candidates that do not specifically target proteins of interest. Here we propose phenotype-based drug discovery approaches in larval zebrafish to identify potent opioid pain therapies without the side-effects of respiratory depression. The larval zebrafish has emerged as a powerful model system for drug and gene discovery since it combines the biological complexity of in vivo models with a nervous system homolog to humans including the respiratory neural network. Such approach has the potential to validate many drug candidates without making any assumptions on their mechanisms of action or targets. Our drug discovery approach proposes to test whether drugs or combinations of drugs can be used as analgesics without the side-effects of respiratory depression and such strategy is of considerable interest (Dahan et al., 2018; Montandon and Slutsky, 2019). We combined several assays to test the respiratory depressant effects of pain killers and their analgesics properties. We demonstrated that larval zebrafish (days post-fertilization 12–14) can be used to test the analgesic properties of opioid drugs and to test the severity of respiratory depression. Our assays provide simple, yet effective, ways to quickly test potential drug candidates in a complex animal model while preserving its central nervous system and brain blood barrier. Due to the high production of fish embryos, we can use these phenotypic assays to perform high-throughput drug screening protocols that are otherwise not feasible in other animal models. Importantly, drug screening in rodents is not feasible as it would require extensive resources and time. Here, we propose new behavioral assays in larval zebrafish where live animals with intact central nervous system can be leveraged for phenotype-based drug discovery combining respiratory depression and analgesia by opioid drugs.

Respiratory depression by opioids has been widely studied in mice, rats, dogs, and sheep (Krause et al., 2009; Montandon et al., 2011; Montandon et al., 2016b). The use of complex animal models to assess respiratory variables comes with many challenges. To accurately assess respiratory depression by opioid drugs, it is critical to quantify respiratory variables without the use of anesthetics which interact with the respiratory system, drug metabolism, and mechanisms of action of opioid drugs. Respiratory assessments in non-anesthetized rodents is the gold standard but it raises concerns due to the stress associated with animal handling, opioid injection, and changes due to altered arousal states and behaviours (Montandon et al., 2016a; Montandon and Slutsky, 2019), which may ultimately affect research outcomes. In larval zebrafish, we demonstrated, for the first time, that respiratory network activity can be assessed, that larval zebrafish present respiratory rate depression by opioid drugs in a similar fashion than humans, and that opioid pharmacology observed in humans is preserved in zebrafish. Most fishes have a complex respiratory system to move water through their gills. Although fishes use a different strategy to absorb oxygen and eliminate carbon dioxide than mammals, they rhythmically produce mandibular movements to move water through their gills. In lampreys, the paratrigeminal respiratory group (pTRG) generates rhythmic mandible movements (Bongianni et al., 2016). Because of their close evolutionary origins (Missaghi et al., 2016), the pTRG shares similarities with the mammalian respiratory network (Cinelli et al., 2013) which generates breathing and regulates respiratory depression by opioids (Montandon et al., 2011). The pTRG also presents similar functional properties than mammals (Gray et al., 1999) such as sensitivity to substance P (Mutolo et al., 2010) and to opioid drugs (Mutolo et al., 2007). Here, we propose that respiratory mandible movements controlled, at least in part, by the pTRG can be used as an index of respiratory network activity, similarly to respiratory activity of the trigeminal muscle in mammals (Jacquin et al., 1999). We showed that the rate of respiratory movements presented dose-dependent decreases in response to increasing dosages of the MOR ligand fentanyl. Although it is difficult to compare administration of fentanyl through fish water in zebrafish and systemic injection in mammals, the dosages used in zebrafish corresponds to those used in rodents (Yassen et al., 2008). Respiratory depression by fentanyl was substantially blocked by the MOR antagonist naloxone, which is consistent with the fact that fentanyl is acting mostly through MORs (James and Williams, 2020). However, the highly selective MOR antagonist CTAP did not prevent respiratory depression by fentanyl, which is similar to the partial antagonism observed in some rodent studies (Zhang et al., 2007). These results suggest that either other opioid receptors are involved in the effect of fentanyl or that CTAP did not cross the blood brain barriers in zebrafish due to its large molecular weight. Interestingly, morphine reduced significantly respiratory rate in larval zebrafish at the low concentration of 1 µM, which is consistent with clinical studies (Montandon et al., 2016a), but not at higher concentrations. The lack of respiratory depression at high morphine concentrations may be due to disinhibition of excitatory networks regulating breathing in the zebrafish. In rodents, morphine disinhibits the ventral tegmental area (Chen et al., 2015), a brain area involved in sleep-wake states and arousal (Venner et al., 2019). Since increased arousal state reduces the severity of respiratory depression by opioids (Montandon et al., 2016a; Montandon and Horner, 2019), it may lead to a reduced respiratory depression at high concentrations of morphine.

The two major strains of wild-type zebrafish, AB and Tübinger (TU), showed different sensitivities to fentanyl. AB fish present pronounced respiratory depression, an effect not observed in TU fish, regardless of the concentration used (data not shown). This strain-specificity suggests that the mechanisms of action of opioid drugs differ between zebrafish strains which could be due to polymorphisms of the MOR gene as it can be found in humans (Oertel et al., 2006) or genetic differences in various genes involved in MOR inhibition (Bian et al., 2012). To determine whether respiratory stimulants can reverse respiratory depression by opioids in larval zebrafish, we tested two types of agonists targeting excitatory receptors: ampakines which comprise drugs acting on AMPA receptors (Ren et al., 2006) and serotoninergic agents (Manzke et al., 2003). CX-614, an ampakine allosteric modulator of AMPA receptors, reversed respiratory depression in our larval zebrafish models, which is consistent with its effects in rodents (Ren et al., 2006) or humans (Dahan et al., 2018). BIMU8, a 5-HT_4A agonist, was administered in combination with fentanyl and compared with fentanyl alone. BIMU8 did not significantly reverse respiratory depression at the concentrations tested, which is consistent with the low efficacy of this treatment in humans (Lötsch et al., 2005). Other serotonin agents such as 5-HT₁ and 5-HT₃ agonists (van der Schier et al., 2014) can be easily tested using our zebrafish models.

Zebrafish larvae were previously used as animal models to study pain and analgesia. Acetic or citric acids were added to fish water to induce pain (Lopez-Luna et al., 2017; Steenbergen and Bardine, 2014). Our models did not use acids as nociceptive stimuli as they change water pH, to which zebrafish are sensitive (Avdesh et al., 2012; Steenbergen and Bardine, 2014). Instead, we used formalin‚ a nociceptive stimulus widely used to induce pain in rodent models (Yoon et al., 2015), and showed that formalin increased swimming velocity. Here, we propose that increased swimming velocity in response to formalin represents the fish escape response to nociceptive stimuli. In fact, lidocaine, a non-opioid analgesic, applied with formalin showed a significant reduction compared to formalin alone. Lidocaine induces analgesia by inactivating voltage-gated sodium channels in neurons, without activating opioid receptors (Tetzlaff, 2000), and these results support the concept that formalin induces pain in larval zebrafish. To promote analgesia in larval zebrafish, fentanyl was combined with formalin and it reduced the swimming response to formalin, therefore suggesting that it reduced pain. It could be suggested that the escape response observed in our assays is due to the effect of formalin on locomotor activity (not nociception) and that the reduced response with fentanyl may be due to inhibition of motor activity. However, fentanyl alone did not reduce swimming velocity compared to control larvae. To complement the nociceptive assay with formalin, we used AITC, also known as ‘mustard oil’, which acts on cation channel transient receptors potential ankyrin 1one and transient receptor potential vanilloid 1 (Oda et al., 2016), two receptors involved in nociception to chemical compounds (Bamps et al., 2021). Like formalin, the swimming response to AITC was reduced by fentanyl. To determine the role of µ-opioid receptors in the analgesic responses to fentanyl, we administered the antagonist naloxone. In the formalin assay, naloxone did not block the effects of formalin, but blocked it in the AITC assay, which is consistent with the effects of naloxone in humans (Gutstein, 2001). Although naloxone did not significantly block analgesia when formalin was used, it is clear in Figure 5d that naloxone increased substantially velocity in some animals, but also reduced it in other animals. In summary, the fact that fentanyl reduced the nociceptive responses of AITC and formalin, two chemicals promoting nociception through different mechanisms, suggests that our approaches correctly assess opioid analgesia in larval zebrafish, while opioid pharmacology remains unclear in the formalin/fentanyl assays. As observed with the respiratory the CX-614 reversed analgesia which is consistent with their effects in rodents (Dahan et al., 2018). BIMU-8 did not reverse analgesia which is consistent with its lack of effects in clinical trials (Dahan et al., 2018). In conclusion, we established opioid analgesia assays that can be used to quantify pain and the analgesic properties of new opioid therapies that can easily be used for large scale drug discovery using larval zebrafish.

Opioid analgesics constitute essential pain therapies that present the lethal side-effect of respiratory depression therefore limiting their effective use in the clinical and at-home settings. There is currently no effective safe pain therapy due to the difficulty at identifying new drug combinations with potent analgesia but reduced respiratory side-effects (Dahan et al., 2018). Using larval zebrafish, we propose models allowing phenotype-based drug discovery approaches, i.e. permitting drug testing without assumptions related to mechanisms of action and targets. Our novel drug discovery models allow high-throughput drug screening in a simple and amenable animal model presenting similar pharmacological and genetic profiles than humans (MacRae and Peterson, 2015). Although zebrafish has been used to identify new anesthetics and their mechanisms of actions (McGrath et al., 2020; Yang et al., 2019), our study is the first to assess respiratory depression by opioid drugs in larval zebrafish combined with analgesia. In addition to high-throughput screening approaches, our models can also be used in combination of transgenic or knockout zebrafish to better understand the mechanisms opioid inhibition, analgesia and respiratory depression, as well as live microscopy of neural circuits of pain and respiration (Ahrens et al., 2013).

All data generated during this study are included in the manuscript. Source files are available.

1. 1. Ahrens MB
   2. Orger MB
   3. Robson DN
   4. Li JM
   5. Keller PJ

   (2013) Whole-brain functional imaging at cellular resolution using light-sheet microscopy

   Nature Methods 10:413–420.

   https://doi.org/10.1038/nmeth.2434

   • PubMed
   • Google Scholar
2. 1. Avdesh A
   2. Chen M
   3. Martin-Iverson MT
   4. Mondal A
   5. Ong D
   6. Rainey-Smith S
   7. Taddei K
   8. Lardelli M
   9. Groth DM
   10. Verdile G
   11. Martins RN

   (2012) Regular care and maintenance of a zebrafish (Danio rerio) laboratory: an introduction

   Journal of Visualized Experiments 1:e4196.

   https://doi.org/10.3791/4196

   • Google Scholar
3. 1. Bamps D
   2. Vriens J
   3. de Hoon J
   4. Voets T

   (2021) TRP channel cooperation for nociception: therapeutic opportunities

   Annual Review of Pharmacology and Toxicology 61:655–677.

   https://doi.org/10.1146/annurev-pharmtox-010919-023238

   • PubMed
   • Google Scholar
4. 1. Bian JM
   2. Wu N
   3. Su RB
   4. Li J

   (2012) Opioid receptor trafficking and signaling: what happens after opioid receptor activation?

   Cellular and Molecular Neurobiology 32:167–184.

   https://doi.org/10.1007/s10571-011-9755-5

   • PubMed
   • Google Scholar
5. 1. Bongianni F
   2. Mutolo D
   3. Cinelli E
   4. Pantaleo T

   (2016) Neural mechanisms underlying respiratory rhythm generation in the lamprey

   Respiratory Physiology & Neurobiology 224:17–26.

   https://doi.org/10.1016/j.resp.2014.09.003

   • PubMed
   • Google Scholar
6. 1. Brown KA
   2. Laferrière A
   3. Lakheeram I
   4. Moss IR

   (2006) Recurrent hypoxemia in children is associated with increased analgesic sensitivity to opiates

   Anesthesiology 105:665–669.

   https://doi.org/10.1097/00000542-200610000-00009

   • PubMed
   • Google Scholar
7. 1. Chen M
   2. Zhao Y
   3. Yang H
   4. Luan W
   5. Song J
   6. Cui D
   7. Dong Y
   8. Lai B
   9. Ma L
   10. Zheng P

   (2015) Morphine disinhibits glutamatergic input to VTA dopamine neurons and promotes dopamine neuron excitation

   eLife 4:e09275.

   https://doi.org/10.7554/eLife.09275

   • PubMed
   • Google Scholar
8. 1. Cinelli E
   2. Robertson B
   3. Mutolo D
   4. Grillner S
   5. Pantaleo T
   6. Bongianni F

   (2013) Neuronal mechanisms of respiratory pattern generation are evolutionary conserved

   Journal of Neuroscience 33:9104–9112.

   https://doi.org/10.1523/JNEUROSCI.0299-13.2013

   • PubMed
   • Google Scholar
9. 1. Dahan A
   2. Aarts L
   3. Smith TW

   (2010) Incidence, reversal, and prevention of Opioid-induced respiratory depression

   Anesthesiology 112:226–238.

   https://doi.org/10.1097/ALN.0b013e3181c38c25

   • PubMed
   • Google Scholar
10. 1. Dahan A
    2. van der Schrier R
    3. Smith T
    4. Aarts L
    5. van Velzen M
    6. Niesters M

    (2018) Averting Opioid-induced respiratory depression without affecting analgesia

    Anesthesiology 128:1027–1037.

    https://doi.org/10.1097/ALN.0000000000002184

    • PubMed
    • Google Scholar
11. 1. Ferguson LM
    2. Drummond GB

    (2006) Acute effects of fentanyl on breathing pattern in anaesthetized subjects

    British Journal of Anaesthesia 96:384–390.

    https://doi.org/10.1093/bja/ael011

    • PubMed
    • Google Scholar
12. 1. Florence CS
    2. Zhou C
    3. Luo F
    4. Xu L

    (2016) The economic burden of prescription opioid overdose, abuse, and dependence in the united states, 2013

    Medical Care 54:901–906.

    https://doi.org/10.1097/MLR.0000000000000625

    • PubMed
    • Google Scholar
13. 1. Gomes T
    2. Juurlink DN

    (2016) Opioid use and overdose: what we've learned in Ontario

    Healthcare Quarterly 18:8–11.

    https://doi.org/10.12927/hcq.2016.24568

    • PubMed
    • Google Scholar
14. 1. Gray PA
    2. Rekling JC
    3. Bocchiaro CM
    4. Feldman JL

    (1999) Modulation of respiratory frequency by peptidergic input to rhythmogenic neurons in the preBötzinger complex

    Science 286:1566–1568.

    https://doi.org/10.1126/science.286.5444.1566

    • PubMed
    • Google Scholar
15. Book

    1. Gutstein HB

    (2001) Opioid analgesics

    In: Hardman J. G, Limbird L. E, Gilman A. G, editors. Goodman and Gilmans's the Pharmacological Basis for Therapeutics. New-York: McGraw-Hill. pp. 569–618.

    https://doi.org/10.1021/jm058286b

    • Google Scholar
16. Software

    1. IASP

    (2018) IASP Statement on Opioids

    IASP Statement on Opioids.

    https://www.iasp-pain.org/Advocacy/OpioidPositionStatement
17. 1. Jacquin TD
    2. Sadoc G
    3. Borday V
    4. Champagnat J

    (1999) Pontine and medullary control of the respiratory activity in the trigeminal and facial nerves of the newborn mouse: an in vitro study

    European Journal of Neuroscience 11:213–222.

    https://doi.org/10.1046/j.1460-9568.1999.00420.x

    • PubMed
    • Google Scholar
18. 1. James A
    2. Williams J

    (2020) Basic opioid pharmacology - An update

    British Journal of Pain 14:115–121.

    https://doi.org/10.1177/2049463720911986

    • PubMed
    • Google Scholar
19. 1. Jiang M
    2. Alheid GF
    3. Calandriello T
    4. McCrimmon DR

    (2004) Parabrachial-lateral pontine neurons link nociception and breathing

    Respiratory Physiology & Neurobiology 143:215–233.

    https://doi.org/10.1016/j.resp.2004.07.019

    • PubMed
    • Google Scholar
20. 1. Jones CM
    2. Mack KA
    3. Paulozzi LJ

    (2013) Pharmaceutical overdose deaths, united states, 2010

    Jama 309:657–659.

    https://doi.org/10.1001/jama.2013.272

    • PubMed
    • Google Scholar
21. 1. Krause KL
    2. Neumueller SE
    3. Marshall BD
    4. Kiner T
    5. Bonis JM
    6. Pan LG
    7. Qian B
    8. Forster HV

    (2009) Micro-opioid receptor agonist injections into the presumed pre-Botzinger complex and the surrounding region of awake goats do not alter eupneic breathing

    Journal of Applied Physiology 107:1591–1599.

    https://doi.org/10.1152/japplphysiol.90548.2008

    • PubMed
    • Google Scholar
22. 1. Le Merrer J
    2. Becker JA
    3. Befort K
    4. Kieffer BL

    (2009) Reward processing by the opioid system in the brain

    Physiological Reviews 89:1379–1412.

    https://doi.org/10.1152/physrev.00005.2009

    • PubMed
    • Google Scholar
23. 1. Lopez-Luna J
    2. Al-Jubouri Q
    3. Al-Nuaimy W
    4. Sneddon LU

    (2017) Reduction in activity by noxious chemical stimulation is ameliorated by immersion in analgesic drugs in zebrafish

    The Journal of Experimental Biology 220:1451–1458.

    https://doi.org/10.1242/jeb.146969

    • PubMed
    • Google Scholar
24. 1. Lötsch J
    2. Skarke C
    3. Schneider A
    4. Hummel T
    5. Geisslinger G

    (2005) The 5-hydroxytryptamine 4 receptor agonist mosapride does not antagonize morphine-induced respiratory depression

    Clinical Pharmacology & Therapeutics 78:278–287.

    https://doi.org/10.1016/j.clpt.2005.05.010

    • PubMed
    • Google Scholar
25. 1. Macintyre PE
    2. Loadsman JA
    3. Scott DA

    (2011) Opioids, ventilation and acute pain management

    Anaesthesia and Intensive Care 39:545–558.

    https://doi.org/10.1177/0310057X1103900405

    • PubMed
    • Google Scholar
26. 1. MacRae CA
    2. Peterson RT

    (2015) Zebrafish as tools for drug discovery

    Nature Reviews Drug Discovery 14:721–731.

    https://doi.org/10.1038/nrd4627

    • PubMed
    • Google Scholar
27. 1. Magalhães FEA
    2. de Sousa C
    3. Santos S
    4. Menezes RB
    5. Batista FLA
    6. Abreu ÂO
    7. de Oliveira MV
    8. Moura L
    9. Raposo RDS
    10. Campos AR

    (2017) Adult zebrafish (Danio rerio): An alternative behavioral model of Formalin-Induced nociception

    Zebrafish 14:422–429.

    https://doi.org/10.1089/zeb.2017.1436

    • PubMed
    • Google Scholar
28. 1. Manglik A
    2. Lin H
    3. Aryal DK
    4. McCorvy JD
    5. Dengler D
    6. Corder G
    7. Levit A
    8. Kling RC
    9. Bernat V
    10. Hübner H
    11. Huang XP
    12. Sassano MF
    13. Giguère PM
    14. Löber S
    15. Duan D
    16. Scherrer G
    17. Kobilka BK
    18. Gmeiner P
    19. Roth BL
    20. Shoichet BK

    (2016) Structure-based discovery of opioid analgesics with reduced side effects

    Nature 537:185–190.

    https://doi.org/10.1038/nature19112

    • PubMed
    • Google Scholar
29. 1. Manzke T
    2. Guenther U
    3. Ponimaskin EG
    4. Haller M
    5. Dutschmann M
    6. Schwarzacher S
    7. Richter DW

    (2003) 5-HT4(a) receptors avert opioid-induced breathing depression without loss of analgesia

    Science 301:226–229.

    https://doi.org/10.1126/science.1084674

    • PubMed
    • Google Scholar
30. 1. Marron Fdez de Velasco E
    2. Law PY
    3. Rodríguez RE

    (2009) Mu opioid receptor from the zebrafish exhibits functional characteristics as those of mammalian mu opioid receptor

    Zebrafish 6:259–268.

    https://doi.org/10.1089/zeb.2009.0594

    • PubMed
    • Google Scholar
31. 1. McGrath M
    2. Hoyt H
    3. Pence A
    4. Jayakar SS
    5. Zhou X
    6. Forman SA
    7. Cohen JB
    8. Miller KW
    9. Raines DE

    (2020) Competitive antagonism of etomidate action by diazepam: in vitro GABAA receptor and in vivo zebrafish studies

    Anesthesiology 133:583–594.

    https://doi.org/10.1097/ALN.0000000000003403

    • PubMed
    • Google Scholar
32. 1. Michel MC
    2. Murphy TJ
    3. Motulsky HJ

    (2020) New author guidelines for displaying data and reporting data analysis and statistical methods in experimental biology

    Journal of Pharmacology and Experimental Therapeutics 372:136–147.

    https://doi.org/10.1124/jpet.119.264143

    • PubMed
    • Google Scholar
33. 1. Missaghi K
    2. Le Gal JP
    3. Gray PA
    4. Dubuc R

    (2016) The neural control of respiration in lampreys

    Respiratory Physiology & Neurobiology 234:14–25.

    https://doi.org/10.1016/j.resp.2016.08.007

    • PubMed
    • Google Scholar
34. 1. Mogri M
    2. Khan MIA
    3. Grant BJB
    4. Mador MJ

    (2008) Central sleep apnea induced by acute ingestion of opioids

    Chest 133:1484–1488.

    https://doi.org/10.1378/chest.07-1891

    • PubMed
    • Google Scholar
35. 1. Montandon G
    2. Qin W
    3. Liu H
    4. Ren J
    5. Greer JJ
    6. Horner RL

    (2011) PreBotzinger complex neurokinin-1 receptor-expressing neurons mediate opioid-induced respiratory depression

    Journal of Neuroscience 31:1292–1301.

    https://doi.org/10.1523/JNEUROSCI.4611-10.2011

    • PubMed
    • Google Scholar
36. 1. Montandon G
    2. Cushing SL
    3. Campbell F
    4. Propst EJ
    5. Horner RL
    6. Narang I

    (2016a) Distinct cortical signatures associated with sedation and respiratory rate depression by morphine in a pediatric population

    Anesthesiology 125:889–903.

    https://doi.org/10.1097/ALN.0000000000001303

    • PubMed
    • Google Scholar
37. 1. Montandon G
    2. Ren J
    3. Victoria NC
    4. Liu H
    5. Wickman K
    6. Greer JJ
    7. Horner RL

    (2016b) G-protein-gated inwardly rectifying potassium channels modulate respiratory depression by opioids

    Anesthesiology 124:641–650.

    https://doi.org/10.1097/ALN.0000000000000984

    • PubMed
    • Google Scholar
38. 1. Montandon G
    2. Horner RL

    (2019) Electrocortical changes associating sedation and respiratory depression by the opioid analgesic fentanyl

    Scientific Reports 9:14122.

    https://doi.org/10.1038/s41598-019-50613-2

    • PubMed
    • Google Scholar
39. 1. Montandon G
    2. Slutsky AS

    (2019) Solving the opioid crisis: respiratory depression by opioids as critical end point

    Chest 156:653–658.

    https://doi.org/10.1016/j.chest.2019.05.015

    • PubMed
    • Google Scholar
40. 1. Mutolo D
    2. Bongianni F
    3. Einum J
    4. Dubuc R
    5. Pantaleo T

    (2007) Opioid-induced depression in the lamprey respiratory network

    Neuroscience 150:720–729.

    https://doi.org/10.1016/j.neuroscience.2007.09.028

    • PubMed
    • Google Scholar
41. 1. Mutolo D
    2. Bongianni F
    3. Cinelli E
    4. Pantaleo T

    (2010) Role of neurokinin receptors and ionic mechanisms within the respiratory network of the lamprey

    Neuroscience 169:1136–1149.

    https://doi.org/10.1016/j.neuroscience.2010.06.004

    • PubMed
    • Google Scholar
42. Book

    1. Nagappa MW
    2. Montandon TN
    3. Sprung G
    4. Chung F

    (2017) Opioids, Respiratory Depression and Sleep-Disordered Breathing

    In: Mulier J, editors. Best Practive & Research Clinical Anaesthesiology. Elsevier. pp. 469–485.

    https://doi.org/10.1016/j.bpa.2017.05.004

    • Google Scholar
43. 1. Oda M
    2. Kurogi M
    3. Kubo Y
    4. Saitoh O

    (2016) Sensitivities of two zebrafish TRPA1 paralogs to chemical and thermal stimuli analyzed in heterologous expression systems

    Chemical Senses 41:261–272.

    https://doi.org/10.1093/chemse/bjv091

    • PubMed
    • Google Scholar
44. 1. Oertel BG
    2. Schmidt R
    3. Schneider A
    4. Geisslinger G
    5. Lötsch J

    (2006) The mu-opioid receptor gene polymorphism 118A>G depletes alfentanil-induced analgesia and protects against respiratory depression in homozygous carriers

    Pharmacogenetics and Genomics 16:625–636.

    https://doi.org/10.1097/01.fpc.0000220566.90466.a2

    • PubMed
    • Google Scholar
45. 1. Pattinson KTS

    (2008) Opioids and the control of respiration

    British Journal of Anaesthesia 100:747–758.

    https://doi.org/10.1093/bja/aen094

    • Google Scholar
46. 1. Ren J
    2. Poon BY
    3. Tang Y
    4. Funk GD
    5. Greer JJ

    (2006) Ampakines alleviate respiratory depression in rats

    American Journal of Respiratory and Critical Care Medicine 174:1384–1391.

    https://doi.org/10.1164/rccm.200606-778OC

    • PubMed
    • Google Scholar
47. 1. Sanchez-Simon FM
    2. Rodriguez RE

    (2008) Developmental expression and distribution of opioid receptors in zebrafish

    Neuroscience 151:129–137.

    https://doi.org/10.1016/j.neuroscience.2007.09.086

    • PubMed
    • Google Scholar
48. 1. Sora I
    2. Funada M
    3. Uhl GR

    (1997) The mu-opioid receptor is necessary for [D-Pen2,D-Pen5]enkephalin-induced analgesia

    European Journal of Pharmacology 324:R1–R2.

    https://doi.org/10.1016/S0014-2999(97)10016-4

    • PubMed
    • Google Scholar
49. 1. Steenbergen PJ
    2. Bardine N

    (2014) Antinociceptive effects of buprenorphine in zebrafish larvae: an alternative for rodent models to study pain and nociception?

    Applied Animal Behaviour Science 152:92–99.

    https://doi.org/10.1016/j.applanim.2013.12.001

    • Google Scholar
50. 1. Taylor JC
    2. Dewberry LS
    3. Totsch SK
    4. Yessick LR
    5. DeBerry JJ
    6. Watts SA
    7. Sorge RE

    (2017) A novel zebrafish-based model of nociception

    Physiology & Behavior 174:83–88.

    https://doi.org/10.1016/j.physbeh.2017.03.009

    • PubMed
    • Google Scholar
51. 1. Tetzlaff JE

    (2000) The pharmacology of local anesthetics

    Anesthesiology Clinics of North America 18:217–233.

    https://doi.org/10.1016/S0889-8537(05)70161-9

    • Google Scholar
52. 1. van der Schier R
    2. Roozekrans M
    3. van Velzen M
    4. Dahan A
    5. Niesters M

    (2014) Opioid-induced respiratory depression: reversal by non-opioid drugs

    F1000Prime Reports 6:79.

    https://doi.org/10.12703/P6-79

    • PubMed
    • Google Scholar
53. 1. Venner A
    2. Todd WD
    3. Fraigne J
    4. Bowrey H
    5. Eban-Rothschild A
    6. Kaur S
    7. Anaclet C

    (2019) Newly identified sleep-wake and circadian circuits as potential therapeutic targets

    Sleep 42:zsz023.

    https://doi.org/10.1093/sleep/zsz023

    • PubMed
    • Google Scholar
54. 1. Wilkerson RG
    2. Kim HK
    3. Windsor TA
    4. Mareiniss DP

    (2016) The opioid epidemic in the united states

    Emergency Medicine Clinics of North America 34:e1–e23.

    https://doi.org/10.1016/j.emc.2015.11.002

    • PubMed
    • Google Scholar
55. 1. Winpenny D
    2. Clark M
    3. Cawkill D

    (2016) Biased ligand quantification in drug discovery: from theory to high throughput screening to identify new biased μ opioid receptor agonists

    British Journal of Pharmacology 173:1393–1403.

    https://doi.org/10.1111/bph.13441

    • PubMed
    • Google Scholar
56. 1. Yang X
    2. Jounaidi Y
    3. Mukherjee K
    4. Fantasia RJ
    5. Liao EC
    6. Yu B
    7. Forman SA

    (2019) Drug-selective anesthetic insensitivity of zebrafish lacking γ-Aminobutyric acid type A receptor β3 subunits

    Anesthesiology 131:1276–1291.

    https://doi.org/10.1097/ALN.0000000000002963

    • PubMed
    • Google Scholar
57. 1. Yassen A
    2. Olofsen E
    3. Kan J
    4. Dahan A
    5. Danhof M

    (2008) Pharmacokinetic-pharmacodynamic modeling of the effectiveness and safety of buprenorphine and fentanyl in rats

    Pharmaceutical Research 25:183–193.

    https://doi.org/10.1007/s11095-007-9440-z

    • PubMed
    • Google Scholar
58. 1. Yoon SY
    2. Woo J
    3. Park JO
    4. Choi EJ
    5. Shin HS
    6. Roh DH
    7. Kim KS

    (2015) Intrathecal RGS4 inhibitor, CCG50014, reduces nociceptive responses and enhances opioid-mediated analgesic effects in the mouse Formalin test

    Anesthesia & Analgesia 120:671–677.

    https://doi.org/10.1213/ANE.0000000000000607

    • PubMed
    • Google Scholar
59. 1. Zhang Z
    2. Xu F
    3. Zhang C
    4. Liang X

    (2007) Activation of opioid mu receptors in caudal medullary raphe region inhibits the ventilatory response to hypercapnia in anesthetized rats

    Anesthesiology 107:288–297.

    https://doi.org/10.1097/01.anes.0000270760.46821.67

    • PubMed
    • Google Scholar

Acceptance summary:

Although opioids, e.g., morphine, remain the major pharmacotherapy for severe, chronic pain, adverse side effects, especially respiratory depression, significantly limit their utility. Here, Zaig et al. established a zebrafish model for mammalian opioid-induced respiratory depression. This model will enable future work to efficiently and effectively screen for novel compounds that provide opioid receptor-mediated analgesia but not respiratory depression.

Decision letter after peer review:

Thank you for submitting your article "Respiratory depression and analgesia by opioid drugs in freely-behaving larval zebrafish" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Michael Taffe as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Jan Marino Ramirez (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data.

Our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife, either of which would be linked to the original paper.

Essential revisions:

In this report Zaig describe a zebrafish model for mammalian opioid induced respiratory depression, as a potential screening approach for novel, selective opioids. All reviewers found the manuscript of significant interest, however, there are several concerns that need to be addressed and a limited requirement for some additional experimental information. Certainly the conclusion needs to be modified. Specifically, the reviewers believe that you have overstated the similarities between mammals and zebrafish. The mandibular rhythm in zebrafish is most certainly not generated by the preBötC, but more likely is by an oscillator in the pons. Thus, the fact that both movements are opioid sensitive doesn't mean that they are sensitive in the same way.

A major concern is that the same data are recycled throughout the paper. This is not how the experiments are described in the Materials and methods and this is also not how the supplementary data is annotated. Different animals are annotated to have the same frequency of mandibular movements. If there are other controls that were actually used, then the correct comparisons are not being performed. This is something the authors can address or correct without new data.

An experiment that should be reported in the course of revision relates to the results described in Figure 5: Here, CTAP does not show a rescue (perhaps a trend), while naloxone does. Therefore, a naloxone only control is required to conclude that the fentanyl rescue is not due to antagonizing endogenous opioid signaling on δ or κ receptors. Additionally, a fentanyl only control in Figure 5B and C would be important to show.

Also, please list the concentration of naloxone used in these experiments. Related to this point: All of the fentanyl data for Figure 5 have been collected, but not plotted in Figure 5B and C.

Additional comments to be addressed:

There are discrepancies between the methods description and source data provided.

Specifically, the Materials and methods state that "To determine whether various drugs induce respiratory rate depression, we compared the rates of mandible movements between several groups of larvae: control group (embryo medium), fentanyl group, fentanyl/naloxone group (combination of fentanyl and naloxone), fentanyl/other drugs (fentanyl and potential drug candidates). All these groups were tested simultaneously in 12-well plates (4 larvae per group if 3 groups were tested). By combining all groups in the same multi-well plate, we ensured that all animals experienced the same conditions therefore allowing comparisons between groups." In fat, the number of control fentanyl experiments used in each experiment is different from the fentanyl+drug numbers (for example, Figure 4B). How can this be if these are done all as replicates with similar number per 12 well plate?

Additionally, the source data show that the fentanyl control data are exactly the same between experiments, Figures 3E, 4A, B, and probably 3B (based on the figure). Also, these data are mostly the same as Figure 2E. However, in the source data, each of these exact same data points has a different fish number annotated. Are these the same fish or different fish? There is an error somewhere and this is absolutely critical to explain since the comparison cohort impacts every conclusion made about mandibular movement. Conclusions shown to be statistically significant or "tending that way" may no longer be valid conclusions.

A central hypothesis is that mandibular movement produced by the paratrigeminal group represents the respiratory rhythm created by the preBötC. However, two successful pharmacological approaches that rescue opioid depression of the preBötC in rodents, 5HT4a (BIMU8, Mazke et al.) and ampakine therapy (CX614, Lorier et al.), do not significantly rescue fentanyl suppression of mandibular jaw movement (Figure 4). Please address these findings.

Although the data may trend towards significance, there is a concern about the comparison cohort used. The 1µM fentanyl data used in Figure 3B, E and Figure 4A, B appears to be the same, but different than or partially overlapping with the fentanyl cohort in Figure 2E. Since the "% baseline change" in Figure 2E is less, if these data are included in Figures 3 and 4, these trending rescues would become less obvious. A clearer experiment might be to attempt rescue of mandibular movement in the same animal after measuring its mandibular depression.

Essential revisions:

In this report Zaig describe a zebrafish model for mammalian opioid induced respiratory depression, as a potential screening approach for novel, selective opioids. All reviewers found the manuscript of significant interest, however, there are several concerns that need to be addressed and a limited requirement for some additional experimental information. Certainly the conclusion needs to be modified. Specifically, the reviewers believe that you have overstated the similarities between mammals and zebrafish. The mandibular rhythm in zebrafish is most certainly not generated by the preBötC, but more likely is by an oscillator in the pons. Thus, the fact that both movements are opioid sensitive doesn't mean that they are sensitive in the same way.

We agree with the reviewers. Our study only shows that the respiratory network involved in breathing in fish is depressed by opioid drugs in a similar way than in mammals. The precise circuits inhibited by opioids are still unknown in zebrafish. We have changed the text in the Introduction related to pTRG and preBötC.

A major concern is that the same data are recycled throughout the paper. This is not how the experiments are described in the Materials and methods and this is also not how the supplementary data is annotated. Different animals are annotated to have the same frequency of mandibular movements. If there are other controls that were actually used, then the correct comparisons are not being performed. This is something the authors can address or correct without new data.

The reviewers are right. Some data have been used in different figures when the experiments should have been separately performed. To avoid any confusions, we have decided to repeat all these experiments highlighted by the reviewers. We repeated experiments as described above.

An experiment that should be reported in the course of revision relates to the results described in Figure 5: Here, CTAP does not show a rescue (perhaps a trend), while naloxone does. Therefore, a naloxone only control is required to conclude that the fentanyl rescue is not due to antagonizing endogenous opioid signaling on δ or κ receptors.

We agree. One of the challenges using formalin is that it substantially increases variability in swimming. Although it is visible that naloxone/formalin/fentanyl substantially increase swimming, it also increases variability of the data which makes the average velocity not significant different. However, when using AITC, the variability doesn’t increase as much and naloxone was able to reverse the analgesia induced by fentanyl. It is now addressed in the Discussion. We also performed additional experiments with naloxone and CTAP alone at various concentrations.

Additionally, a fentanyl only control in Figure 5B and C would be important to show.

We have now included fentanyl in Figure 5 showing time course and representative tracings.

Also, please list the concentration of naloxone used in these experiments.

It has been included.

Related to this point: All of the fentanyl data for Figure 5 have been collected, but not plotted in Figure 5B and C.

We performed new experiments and now provide new data.

Additional comments to be addressed:

There are discrepancies between the methods description and source data provided.

Specifically, the Materials and methods state that "To determine whether various drugs induce respiratory rate depression, we compared the rates of mandible movements between several groups of larvae: control group (embryo medium), fentanyl group, fentanyl/naloxone group (combination of fentanyl and naloxone), fentanyl/other drugs (fentanyl and potential drug candidates). All these groups were tested simultaneously in 12-well plates (4 larvae per group if 3 groups were tested). By combining all groups in the same multi-well plate, we ensured that all animals experienced the same conditions therefore allowing comparisons between groups." In fat, the number of control fentanyl experiments used in each experiment is different from the fentanyl+drug numbers (for example, Figure 4B). How can this be if these are done all as replicates with similar number per 12 well plate?

We have modified the Materials and methods to make it clearer. We tried to have all the conditions tested at the same time in the 12-well plate. However, there were more groups and animals per group than the number of wells per plate (12). We therefore had to combine recordings from various 12-well plates. Also, due to the criteria we used to exclude fish, the number of fish measured for each condition varies considerably. After revisions of the data and new experiments performed, no recordings have been used more than once. All experimental data are now unique for each panel in the figures.

Additionally, the source data show that the fentanyl control data are exactly the same between experiments, Figures 3E, 4A, B, and probably 3B (based on the figure). Also, these data are mostly the same as Figure 2E. However, in the source data, each of these exact same data points has a different fish number annotated. Are these the same fish or different fish? There is an error somewhere and this is absolutely critical to explain since the comparison cohort impacts every conclusion made about mandibular movement. Conclusions shown to be statistically significant or "tending that way" may no longer be valid conclusions.

We repeated the experiments, carefully checked all the data and updated the spreadsheet accordingly. Data points of each condition are unique and have never been used more than once.

A central hypothesis is that mandibular movement produced by the paratrigeminal group represents the respiratory rhythm created by the preBötC. However, two successful pharmacological approaches that rescue opioid depression of the preBötC in rodents, 5HT4a (BIMU8, Mazke et al.) and ampakine therapy (CX614, Lorier et al.), do not significantly rescue fentanyl suppression of mandibular jaw movement (Figure 4). Please address these findings.

We agree with the reviewers. Although BIMU-8 was successful in some animal studies it had limited success in humans. In our experiments, BIMU-8 did not reverse respiratory depression, which may be due to differences in the serotoninergic system of zebrafish or differences in drug diffusion through the fish brain blood barrier. We address it in the Discussion.

Although the data may trend towards significance, there is a concern about the comparison cohort used. The 1µM fentanyl data used in Figure 3B, E and Figure 4A, B appears to be the same, but different than or partially overlapping with the fentanyl cohort in Figure 2E. Since the "% baseline change" in Figure 2E is less, if these data are included in Figures 3 and 4, these trending rescues would become less obvious. A clearer experiment might be to attempt rescue of mandibular movement in the same animal after measuring its mandibular depression.

We have performed additional experiments to answer these questions. We do not perform sequential experiments with several conditions in the same fish, because drug pipetting and movement of the multi-well plate affect mandible and behavioral recordings.

Author details

1. Shenhab Zaig

   Keenan Research Centre for Biomedical Sciences. St. Michael’s Hospital Unity Health Toronto, Toronto, Canada

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft

   Competing interests

   No competing interests declared

2. Carolina da Silveira Scarpellini

   Department of Medicine, Faculty of Medicine, University of Toronto, Toronto, Canada

   Contribution

   Formal analysis, Validation, Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5576-3468

3. Gaspard Montandon

   Keenan Research Centre for Biomedical Sciences. St. Michael’s Hospital Unity Health Toronto, Toronto, Canada

   Contribution

   Conceptualization, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration

   For correspondence

   gaspard.montandon@utoronto.ca

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3587-4472

• 5,282

  Page views

• 269

  Downloads

• 11

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.