Introduction

As euthanasia is often required in both clinical veterinary practice and scientific research conducted in fish, finding a protocol efficient to induce a reliably rapid and painless death is of critical importance. In a context where no Food and Drug Administration (FDA)-approved drug is available for fish euthanasia by immersion, buffered tricaine methanesulfonate (MS-222) is one of the solutions recommended by the American Veterinary Medical Association Guidelines for the Euthanasia.¹ Immersion in MS-222 is currently the most commonly used method for fish euthanasia,^2,3 as immersion does not require contention, decreasing patient stress. This anesthetic agent acts as a voltage-sensitive sodium channel blocker and is approved by the FDA for fish and amphibian temporary immobilization.¹ Personal protective equipment should be worn during powder handling to prevent retinal lesions,⁴ skin and pulmonary irritation. Due to its 21-day withdrawal time, MS-222 is not adapted to euthanize food fish species. Prior to euthanasia, fish are usually transferred to an aerated bucket filled with water from their original tank.² Doses ranging from 250 to 500 mg/L for a minimum duration of immersion of 10 minutes have been historically recommended for most fish species.⁵ Durations of immersion are likely to vary but some authors have recommended keeping fish in the immersion solution for 30 minutes after cessation of opercular movements.^1,6–8 A review article even suggested that this duration can be reduced to 5 minutes after apnea onset in some fish species.⁹ However, a recent study has shown that this method was inefficient to euthanize hypoxia-tolerant species such as cyprinidae and suggests that immersion in buffered MS-222 up to 1000 mg/L for 15 minutes should be followed by a cerebral disrupting method,¹⁰ or KCl injection.¹¹

Death can be defined by cardiovascular arrest and/or cerebral death, but its least controversial definition is a lack of reversibility.¹² It has been reported that MS-222 induces death in fish by cardiovascular dysfunction.¹³ During deep anesthesia, MS-222 induces a depression of the medullary respiratory center¹⁴ leading to hypoxic water in the opercular chambers, which causes bradycardia and increased resistance to blood flow in the gill lamellae,¹⁵ ultimately leading to death. However, heart beats can persist for more than 4 hours in goldfish after decapitation,¹⁰ due to a high tolerance to cardiac hypoxia.¹⁶ Thus, cardiovascular monitoring is not a sensitive indicator to detect death in goldfish, while a persistent lack of cardiac activity is a specific death criteria. Thus, death was defined as an irreversible apnea and persistent cardiac arrest when fish were placed in freshwater for more than 3 hours, similar to a previous study.¹⁰

The objective of this study was to determine the lowest dose and duration of immersion in buffered MS-222 that would induce reliable euthanasia in goldfish. The hypothesis was that a dose of buffered MS-222 of 1000 mg/L for an hour would be efficient to euthanize goldfish by immersion, meaning that fish would not resume operculation after withdrawal from the solution and no heartbeat would be detected by Doppler after the end of the observation period.

Materials and Methods

This project was approved by an Institutional Animal Care and Use Committee (protocol 19-Rech-2004). The study group consisted of 24 goldfish acquired from a local pet shop, with body weight ranging from 3 to 10 g. Goldfish were of unknown sex and age. Fish were kept in 32-L rectangular aquarium tanks. Water was dechlorinated as directed by the manufacturer (Prime), salt was added at a concentration of 2 g/L as in a previous study¹⁰ and the temperature was maintained at 20 °C. A U2 filter (Fluval Hagen) was installed in each tank prior to fish transfer, with addition of bacteria (Cycle, Nutrafin Hagen) to stimulate the nitrogen cycle. Goldfish were then randomly allocated to each of the 4 groups of 6 fish acclimated for 10 days. Standard quarantine treatments were performed, including chloroquine (Galenova) by immersion at 10 mg/L for 24 hours and praziquantel (Fishman Chemical) immersion twice at 2 mg/L for 3 days 1 week apart. Individual tank water quality monitoring included 4 parameters (pH, total ammonia, nitrite, nitrate; Master test kit; Fluval Hagen). Alkalinity was 160 mg/L before sodium bicarbonate addition, hardness was 30 mg/L, pH ranged from 7.2 to 7.6, ammonia ranged from 0 to 3.7 mg/L, nitrite concentration ranged from 0 to 0.1 mg/L, and nitrate concentration ranged from 0 to 50 mg/L. Physical examinations, including weighing and skin scrapes, conducted after quarantine were unremarkable. Fish were fed once a day with commercial fish flakes.

Recheck physical examinations were performed on the day of experiment and were unremarkable. Individuals were identified based on their variable coloration pattern. Powdered MS-222 was weighed with a precise scale (10 mg) and used within 30 minutes. Testing of each group was performed in a predetermined sequence so that all fish belonging to a group were not tested together. Each goldfish was caught with an atraumatic net and placed in the immersion solution of its group (Table 1) for 15 or 60 minutes depending on the group. Fish were immersed in buffered MS-222 (Aqualife TMS; Syndel) at doses and for the durations indicated in Table 1, then recovered in freshwater. No air stone was placed in experimental buckets to prevent water movement. Buffering¹⁷ was performed with sodium bicarbonate (Arm and Hammer) 1:1 weight to weight ratio as previously described¹⁰ and the pH was confirmed to be neutral using pH indicator paper prior to fish immersion. Initial respiratory rate, and time to apnea, defined as loss of visible gill movements, was recorded for each goldfish. In a subset of group 1 (n = 3), heart rate was monitored every 5 minutes by Doppler (fish were momentarily brought at the water surface for monitoring), while the other 3 fish of group 1 were left undisrupted in the immersion tank. At the end of the immersion period, fish were removed from the MS-222 solution and placed in the freshwater tank. In groups 1 (G1), 2 (G2) and 4 (G4), fish were observed every 5 minutes for 3 hours to detect any individual that would resume operculation (observation period). In group 3 (G3), fish were observed for 18 hours to detect any fish that would resume operculation. If they did so, time to re-operculation was recorded.

Death was defined as a lack of re-operculation and lack of heartbeat at the end of the observation period. If the goldfish did not resume operculation at the end of the observation period, a Doppler (Ultrasonic Doppler Flow Detector Model 811-B; Parks Medical Electronics) was placed on the heart to assess the presence of a heartbeat and record heart rate. If a heartbeat was detected in a persistently apneic fish, brain pithing was performed. Full necropsy was performed on each pithed fish. After opening the coelomic cavity ventrally, macroscopic examination of organs was performed. Tissues were left in anatomic position, were fixed in neutral-buffered 10% formalin, processed routinely for histopathology with hematoxylin, eosin, and saffron and Ziehl-Neelsen stain, and examined by light microscopy. Any goldfish operculating at the end of the observation period was recovered and adopted out by veterinarians.

Data were analyzed using a statistical analysis software (R version 3.4.1; R Core Team). Median times to apnea and re-operculation were calculated.

Results

Goldfish from all groups stopped operculating when placed in the buffered MS-222 solution. Median time to apnea was 35 seconds at 1000 mg/L, and 65 seconds at 500 mg/L. During the observation period, re-operculation occurred in 5 out of 6 individuals of G1 (Table 2). Time to re-operculation was quicker in the group monitored every 5 minutes with the Doppler than in the other subgroup (mean of 37.5 min versus 81 min respectively). None of the goldfish-from G2, G3, and G4 (n = 18), resumed operculation. At the end of the 3-hour observation periods, all fish from G1 (100%), 3 fish from G2 (50%), and 1 fish from G4 (17%) had a remaining heartbeat with heart rate ranging from 15 to 80 beats per minute. At the end of the 18-hour observation period, 0 fish of G3 had a remaining heartbeat (Table 3). Histopathology results were unremarkable, and no acid-fast organisms were observed. All recovered fish were still alive a year after the experiment.

Discussion

None of the 18 goldfish exposed to 1000 mg/L of buffered MS-222 for more than 15 minutes resumed operculation in the present study. This result was unexpected, as the initial expectation was that only goldfish exposed to MS-222 for 1 hour would be successfully euthanized. Thus, the result of a previously published study¹⁰ was not repeatable in the conditions used in the present study. In addition, after 18 hours of observation in freshwater at 20 °C, none of the goldfish had remaining heart beats, confirming part of the study hypothesis that euthanasia would be achieved by immersion in buffered MS-222. Overall, a dose of 1000 mg/L of buffered MS-222 for 15 minutes was sufficient to euthanize juvenile goldfish in the water conditions used in this study.

It is challenging to determine if persistently apneic fish displaying a heartbeat at the end of the observation period were in irreversibly comatose or in a stage of cerebral death. To comply with current AVMA guidelines,¹ pithing of these fish was performed prior to necropsy for ethical reasons, even if MS-222 has been documented to block sensory nerve activity.^18,19 In human medicine, cerebral death is defined by 4 criteria: unresponsivity, apnea, lack of reflexes, and flat electroencephalogram.¹² In such cases, a heartbeat can be present while the person is considered legally dead.¹² In amphibians, MS-222 has been shown to alter electroencephalographic tracing by severely depressing cerebral activity,²⁰ while cardiac activity was maintained for at least 2 hours, preventing the recording of an electroencephalographic flat line signal. At temperatures close to 0 °C, fish of the family cyprinidae, such as goldfish, can survive in anoxic condition for months.²¹ Tolerance to cerebral hypoxia in cyprinidae is achieved through the suppression of certain cerebral function, such as vision and hearing, increased extracellular neurotransmitters, and conversion of lactate to ethanol in muscles.²¹ Thus, electroencephalograms were not used in the present study as a lack of cerebral activity is not a specific indicator of death in goldfish. As a result, death was defined based on circulatory criteria in the present study, as reported in comparative medicine.⁹ Based on this definition, fish with a remaining heartbeat were not considered dead. However, it is impossible to apply the human definition of cerebral death to fish anesthetized with MS-222 due to its intrinsic effect on cerebral activity. Regardless, since none of the fish from G3 recovered, it can be questioned if pithing is always necessary prior to necropsy of juvenile goldfish exposed to 1000 mg/L of buffered MS-222 for 15 minutes, especially if apneic fish remain immersed in MS-222 solution and unresponsive until necropsy.

Because 5 out of 6 goldfishes resumed operculation at a 500 mg/L of MS-222 for 15 minutes, this protocol can be considered ineffective to euthanize goldfish, as previously described.¹⁰ In addition, the median time to induction of apnea was 65 seconds in this group versus 35 seconds after exposure to 1000 mg/L of MS-222. To limit stress associated with induction, the use of immersion in 500 mg/L of buffered MS-222 followed by pithing for goldfish euthanasia should be discouraged. Instead, the use of 1000 mg/L for 15 minutes appears more appropriate in juvenile goldfish kept in the same water conditions. An additional advantage of this protocol is the possibility to collect brain tissue for histopathology or molecular testing.

When comparing median time to apnea to the one observed in the previous study,¹⁰ it was quicker in the present study (65 s versus 150 s at 500 mg/L, and 35 s versus 60 s at 1000 mg/L MS-222). Similarly, the median re-operculation time of fish exposed to 500 mg/L for 15 minutes was much longer in this study (75 min for G1) than in the previous study (6.5 min).¹⁰ This suggests a deeper plane of anesthesia in the present study and could explain the successful euthanasia of individuals from other groups.

Powdered anesthetic agent was used in the present study while a stock solution was used in a previous study.¹⁰ Stock solutions of MS-222 have been reported to remain efficient for 3 months when kept in a dark and cool environment.²² Since the stock solution was replaced daily in the previous study,¹⁰ this is unlikely to be the cause of the different results between the 2 studies.

Another parameter to take into consideration was the use of heart rate Doppler monitoring. A quicker time to re-operculation was observed in the group monitored every 5 minutes with the Doppler, suggesting that this technique might decrease the anesthetic effect of MS-222 by stimulating the fish, and may partly explain the failure of euthanasia in previous protocols conducted this way. However, no statistical evaluation was performed due to small sample size and low statistical power. A higher number of individuals would be needed in further studies to re-assess statistical significance and test this assumption.

The goldfish used in this study weighed 3 to 10 g which corresponds to a juvenile stage of goldfish. In comparison, goldfish used in the previous study weighed 20 to 40 g. The effect of fish size on the efficiency of anesthetic drugs has been studied but no consensus has been achieved to date.¹⁷ It is possible that juvenile goldfish are more susceptible than adults to MS-222 as reported in cods.²³ A previous study showed a positive relationship between cod body weight, and respectively induction and recovery times.²³ Similarly, smaller medaka (Oryzias dancena) are more sensitive to anesthesia with lidocaine.²⁴ This could explain why times to apnea were shorter in the present study than in another study conducted on adult goldfish.¹⁰ Conversely, MS-222 has been shown less efficient in larval than in adult zebrafish.³ In addition, MS-222 has been documented to affect certain physiologic parameters: it induces increased hematocrit, hyperglycemia, and causes various electrolyte changes overtime, such as increased sodium, potassium, chloride and ionized calcium blood concentrations.²⁵ It is likely that fish intrinsic physiologic parameters, including renal and gill metabolism, also affect MS-222 efficacy. After acetylation, non-polar metabolites of MS-222 are typically excreted via the gills while polar metabolites and native MS-222 are excreted via the kidneys.²⁶ While renal function is challenging to assess in fish,²⁷ their developmental stage influences renal physiology, with a correlation of the mesonephric nephrons number with body mass and variation of glomeruli distribution overtime, as demonstrated in zebrafish.²⁸ Thus, the different age and body mass of goldfish used in the present study could explain that euthanasia was efficient with MS-222 at 1000 mg/L for 15 minutes, while it was reported inefficient in adults.¹⁰

Certain water quality parameters from the previous study were also unknown, including alkalinity and hardness.¹⁰ These parameters could impact the response of the fish to MS-222, as it is more efficient when neutrally buffered,¹³ at higher temperature and at lower hardness.²² The effect of these environmental parameters can be questioned. The preferred temperature zone recommended in goldfish is 15 °C to 22 °C.²⁹ At higher temperatures, the acid-base status of the fish is modified.⁹ This can lead to hypercapnia and secondary hyperventilation. These metabolic changes enhance the assimilation of the drug through the gills. The present study was conducted at 20 °C while a temperature of 23 °C was used in the previous study.¹⁰ Thus, a milder effect of MS-222 would have been expected in the present study based on the temperature difference. Therefore, ambient temperature is unlikely to explain the different findings of the 2 studies. In the present study, the same buffering ratio of sodium bicarbonate was chosen as in the previous study from Balko et al¹⁰ (ie, buffering with an equal weight of sodium bicarbonate). In addition, a neutral pH of 7 was confirmed with pH-indicator paper as various buffering ratios have been reported for fish anesthesia.³⁰ The amount of sodium bicarbonate needed for neutral buffering depends on the initial alkalinity of the source water. In acidic environments, a portion of MS-222 is positively charged, limiting its diffusion through the gills.¹³ Therefore, it is possible that the reduced effect of MS-222 and failure to obtain euthanasia in the previous study¹⁰ was associated with a lower alkalinity of the source water. In addition, water bicarbonate ions stimulate chloride efflux by the gills, independently of the pH.³¹ Thus, a different environmental bicarbonate concentration could have had an effect on MS-222 gill excretion.

Limitations of the present study include the small number of fish included and the lack of randomization of treatment order. In addition, the authors were unaware of the previous water quality parameters used by the goldfish breeders.

To conclude, it was possible to induce juvenile goldfish euthanasia with an immersion in 1000 mg/L of neutrally buffered MS-222 at 20 °C for 15 minutes. Clinicians and researchers using this euthanasia technique should pay particular attention to water quality parameters as these can influence the effect of MS-222. Further studies are needed to evaluate the effect of fish age, weight and other physiologic parameters on MS-222 efficacy.