Methamphetamine effects in zebrafish (Danio rerio) depend on behavioral endpoint, dose and test session duration

Research using zebrafish ( Danio rerio ) has begun to provide novel information in many fields, including the behavioral pharmacology of drug use and misuse. There have been limited studies on the effects of methamphetamine in adult zebrafish and the parameters of exposure (dose, test session length) have not been well-documented. Behavior following drug exposure is generally measured during relatively short sessions (6-10 min is common) in a novel tank environment. Many procedural variables (isolation, netting, novel tank) elicit anxiety-like behavior that is most apparent during the initial portion of a test session. This anxiety-like behavior might mask the initial effects of methamphetamine. During longer test sessions, these anxiety-like responses would be expected to habituate and drug effects should become more apparent. To test this idea, we measured several locomotor activity responses for 50-min following a range of methamphetamine doses (0.1 – 3.0 mg/L via immersion in methamphetamine solution). Methamphetamine failed to alter swimming velocity, distance traveled, or freezing time. In contrast, methamphetamine produced a dose-dependent decrease in time spent in the bottom of the tank, an increase in the number of visits to the top of the tank, and an increase in the number of transitions along the sides of the tank. The


Introduction
Psychostimulants produce a transient increase in psychomotor activity and have a high potential for misuse due to an increase in dopaminergic neurotransmission (Ritz and Kuhar, 1993;Wise, 2004).Decades of preclinical research using primarily laboratory rodents indicate that the same mechanisms underlie psychomotor activation and misuse (Wise and Bozarth, 1987;Wise and Robble, 2020).Therefore, psychostimulant-induced increases in motor activity are often used as a proxy to understand factors underlying misuse.
Research describing the effects of psychostimulants in zebrafish (Danio rerio) is now accumulating.Zebrafish share orthologs of at least 70 % of human genes and most of the relevant mammalian neurochemical systems have been described in zebrafish (Friedrich et al., 2010;Levin et al., 2015;Schenk et al., 2020).Important advantages of using zebrafish include high throughput capacity, fecundity, short developmental period, ease of genetic manipulation and ease of drug administration (Kari et al., 2007), which contributes to their increased use in behavioral pharmacology research (Klee et al., 2012;Müller et al., 2020;Orger and de Polavieja, 2017;Schenk et al., 2020;Schneider, 2017;Stewart et al., 2011b).
Analysis of changes in patterns of zebrafish swimming behavior provide a means of assessing effects of drugs (Fontana et al., 2018;Idalencio et al., 2015;Kalueff et al., 2013Kalueff et al., , 2014;;Levin et al., 2015;Stewart et al., 2012).Results have suggested that test duration may be a critical determinant for measuring various effects of drugs on zebrafish behavior.When behavior was measured during a 6 min test, an initial bout of freezing was followed by a progressive decrease in freezing bouts and an increase in the number of transitions to and time spent in the upper half of the tank (Cachat et al., 2011).Using a 30-min test duration, distance travelled, immobility duration, and immobility bout duration were found to occur in a wave-like pattern with robust oscillations occurring every 5-7 min (Stewart et al., 2012), suggesting that longer duration tests might better capture drug effects on swimming patterns.
During typical test conditions, fish are netted, removed from a large, familiar housing tank that contains many fish and then placed into a novel tank for behavioral measurements.Under these conditions, baseline measures of swimming are consistent with an anxiety-like response, such as increased latency to enter the upper portion of the tank, decreased visits to the upper portion of the tank, decreased time spent in the upper half, increased erratic movements, and increased freezing bouts and duration (Cachat et al., 2011;Giacomini et al., 2020;Pullaguri et al., 2020).Because of the high level of anxiety-like behaviors at baseline, the effects of anxiolytic agents such as buspirone, nicotine, fluoxetine, and ethanol have been relatively easy to measure in zebrafish (Bencan et al., 2009;Giacomini et al., 2020;Idalencio et al., 2015;Levin et al., 2007;Moraes et al., 2021;Rosa et al., 2020;Serikuly et al., 2021;Stewart et al., 2011c;Tran et al., 2016).
Behavioral effects of psychostimulants routinely reported in rodents, non-human primates and humans have been more difficult to observe in zebrafish.Whereas psychostimulants increase velocity and distance travelled in rodents (Swerdlow et al., 1986), there is good consensus that there is no marked increase in either swimming distance or swimming speed following acute psychostimulant administration to zebrafish (Kyzar et al., 2013;López-Patiño et al., 2008;Stewart et al., 2011a).One possible explanation for the failure to observe these effects is that very short tests are conducted in zebrafish lasting usually only 6-10 min.In contrast, methamphetamine-induced hyperactivity in rats peaks at about 30 min following an intraperitoneal injection and persists for an additional 20-30 min (Brennan et al., 2007).This is consistent with the pharmacokinetic profile of methamphetamine and its primary metabolite, amphetamine, in rats (Rivière et al., 1999(Rivière et al., , 2000)).Shorter sessions as those that are conducted in zebrafish likely fail to capture the full effect of the drug.In the present study, we determined dose-effect curves for several behavioral effects of methamphetamine using 50 min duration session, to determine the full time-course of methamphetamine effects.

Subjects and husbandry
Zebrafish (Danio rerio) were bred in the Otago Zebrafish facility on 20 August 2020.Fish were a cross between the AB line and petshop fish.They were reared in a temperature-(25.2-26.1 • C), conductivity-(390-458 μS) and pH-(7-7.8)controlled Tecniplast facility that was maintained on a 0800-2200 light cycle (lights on at 0800).Fish were fed two dry feeds (000-400 ZM Zebrafish Management Ltd) and one artemia (brine shrimp) feed per day.When the fish were approximately 5 months old, they were tagged dorsally with visible implant elastomer for identification and separated into single sex tanks.Approximately two weeks later, the fish were moved to the Department of Zoology where they were housed in a 27-tank (Tecniplast tanks; 284 × 169 × 114 mm) recirculating system maintained at 25 • C. The laboratory was on a 0700-2230 h light cycle with lights on at 0700.Fish were provided with dry feed (Zebrafish Management Ltd) two times per day in the mornings and afternoons.Female fish were available and used in this study and behavioral testing began when the fish were approximately 8 months of age.All procedures were carried out under standard operating procedures for zebrafish husbandry and with the approval of the University of Otago Ethics Committee (protocol 20-136).

Drugs
Methamphetamine HCl was obtained from BDG Synthesis (Porirua, New Zealand).The drug was mixed fresh daily and was dissolved in filtered tap water that was aerated overnight to dechlorinate (the same water that is used for water changes on the fish system).All doses refer to the salt weight.

Behavioral testing
Fish were tested individually in one of 4 separate novel 2 L glass tanks (20.5 × 20.5 × 7.5 cm).Each of the 4 tanks had two sheets of translucent white acrylic on the back outside of the tank and were lined on the inside with white film on two sides and the bottom to eliminate glare and reflection from house lights.Tanks were lit from behind and above.The tanks were videoed front-on using a Basler acA1300-60/gc GigE camera with a 4.4-11 mm lens that was positioned 112 cm away from the test tanks.
On the test day, fish were netted and immersed individually in a mL beaker containing methamphetamine (100 mL of 0.0 (control), 0.1, 0.3, 1.0, or 3.0 mg/L; n = 9, 7, 10, 6 and 7, respectively).Each fish was only exposed to one dose of methamphetamine.Assignment to dose and to test arena was randomly determined.The fish remained in the methamphetamine solution for 10 min after which they were retrieved with a slotted spoon, rinsed in tank water and then transferred to one of the 4 testing arenas for 50 min.Doses and exposure duration were determined in preliminary studies.Higher doses of 40 mg/L were often disruptive, with fish spending much of their time on the bottom of the arena.
Tests were carried out beginning at 1000 h and continued until h each day, allowing a maximum of 16 fish (4 groups of 4 fish) to be tested each day.The water in the testing tank was replaced after each test to avoid additional stress and variability in behavior that can occur in the absence of water changes (Fontana et al., 2021).At the end of the test session, the fish were placed in a holding tank where they remained until all tests for the day were completed.Subsequently, fish were returned to the home tank where they were fed.
A 3 × 3 grid was superimposed on the video image for behavior analysis providing 9 equal sized zones (within the behavioral software EthoVision XT; Noldus); 3 zones represented the top third of the tank, zones represented the middle third of the tank and 3 zones represented the bottom third of the tank.The software returned data for each 5 min bin of the 50 min test.Behaviors recorded were swimming speed, swimming distance, the number of visits to the top 3 zones of the tank, time spent in the bottom 3 zones of the tank, time spent freezing and number of transitions up and down the right and left walls of the tank.

Statistical analysis
All data were analyzed using R v 4.3.0(R Core Development Team, 2023).First, we determined the influence of dose on behavior across the entire 50-minute session.Distance moved, velocity, freezing time and bottom time were all modelled using linear models (lme4 package, lm function).The number of transitions and the number of visits to the top zone are count data and modelled using a negative binomial generalized linear model (MASS package, nb.glm function).Second, time course data for each of the behaviors as a function of dose were analyzed using separate mixed models with dose and time as fixed factors and fishID included as a random factor to account for the repeated measures over time (lme4 package, lmer function).The number of transitions and visits to the top zone were log-transformed.Significant effects were followed up using Dunnett's post hoc tests to compare effects of the doses of methamphetamine to the control group (emmeans package and function).

Results
When analyzing the data across the entire 50-minute sessions, there was a significant effect of dose on the number of transitions up and down the sides of the tank (z = − 3.63, p = 0.0003; Fig. 1A), the time spent on the bottom of the tank (t = 3.19, p = 0.003; Fig. 1C) and the number of visits to the top of the tank (z = − 4.26, p < 0.0001; Fig. 1E).Doses of 0.1 and 0.3 mg/L increased the number of transitions (Fig. 1A), decreased time in the bottom of the tank (Fig. 1B) and increased the number of visits to the top of the tank (Fig. 1C).Higher doses failed to produce these effects.When analyzing the data over session time, there were significant dose x time interactions for the number of transitions up and down the sides of the tank and the time spent in the bottom of the tank, though these effects were mainly driven by the 3.0 mg/L dose, as the behaviors produced following administration of any of the other doses were not significantly different from the control responses (Table 1; Fig. 1B, D, E, respectively).Nonetheless, within 10-20 min of exposure to the 0.1 or 0.3 mg/L doses of methamphetamine, there was a clear trend for an increase in the number of transitions, a decrease in time spent in the bottom and an increase in the number of visits to the top zones.These effects persisted throughout the session.In contrast, following administration of 3.0 mg/L, the majority of time during the 50 min session was spent in the bottom of the tank.
Methamphetamine did not significantly alter the remaining behaviors (distanced moved (t = − 0.29, p = 0.77); velocity (t = − 0.06, p = 0.95); freezing time (t = − 0.22, p = 0.83; Fig. 2).There was a tendency for velocity and distance moved to increase during the initial 20 min of the session following exposure to the 0.1 and 0.3 mg/L methamphetamine doses, but these increases were not significantly different from the behavior in the control group (Fig. 2, Table 1).

Discussion
Using a range of methamphetamine doses and a longer testing period than is typically used for zebrafish, we show that effects of methamphetamine were dependent on dose and the specific behavior measured.Behaviors that were altered by methamphetamine (visits to the top of the tank, time in the bottom of the tank and number of transitions along the sides of the tank) became apparent by 10-20 min following drug exposure and generally these changes in behavior persisted throughout the remainder of the session.Clearly, these behavioral alterations in response to methamphetamine would not have been observed if shorter test sessions had been used.
In the current study, behavioral effects of methamphetamine were observed over more than a 10-fold dose range (from 0.1 to 3.0 mg/L).For behavioral responses that were significantly altered by methamphetamine, the lower doses of 0.1 and 0.3 mg/L increased the number of visits to the top of the tank, decreased time spent in the bottom of the tank and increased the number of transitions along the sides of the tank.Higher doses failed to produce these effects.These findings are consistent with data derived from other laboratory animals; the dose effect curve for locomotor activity is an inverted U-shaped curve with both low and high doses producing limited increases in hyperactivity (Bhimani et al., 2021;Camp et al., 1994).For example, an increase in locomotor activity in rats was observed following methamphetamine doses also across a 10-fold dose range (0.2-2.0 mg/kg, IP), and locomotor activity decreased following a higher dose (5.0 mg/kg) (Bhimani et al., 2021).In another example from the literature, methamphetamine increased locomotor activity in rats following administration of 1.0-3.75mg/kg and then activity decreased after 5.0 mg/kg (Tuv et al., 2021).Thus, the similar pattern of behavioral responses in zebrafish from the current study compared with the pattern in rodents supports the conclusion that the appropriate doses and dose range were evaluated in zebrafish in the current study.
The increase in the number of visits to the top of the tank and decrease in time spent in the bottom of the tank are consistent with an anxiolytic effect of methamphetamine (Kalueff et al., 2013).The transitions along the walls of the tank have been suggested to reflect thigmotaxis, the tendency to avoid open spaces (Champagne et al., 2010;Egan et al., 2009), and is a characteristic behavioral response to psychostimulants administered to rodents (Clayman and Connaughton, 2022;Simon et al., 1994).The time-course data from our study are consistent with data obtained in other laboratory animals, most notably rodents (Ago et al., 2016;Brennan et al., 2007;Carati and Schenk, 2011;Fujii et al., 2007;G J Rivière et al., 1999), and support the suggestion that longer test sessions should be used to evaluate drug effects in zebrafish.
A common effect of methamphetamine in other laboratory animals is an increase in locomotor activity, as indicated by increased distance travelled and increased speed of movement.It was therefore surprising that we failed to demonstrate a similar increase in swimming speed or distance moved.There was a trend in this direction following administration of the lower doses of methamphetamine, but these effects were not significantly different from the behavior in the control group.One possible explanation is that the species differences (fish versus rodents) prevent comparable testing methods from being used.In rodents, there is usually an initial habituation period prior to psychostimulant injection (Brennan et al., 2009;Camp et al., 1994;Fujii et al., 2007).During this period, locomotor hyperactivity elicited by the novel environment gradually declines (Bolivar, 2009) so that increases produced by psychostimulants administered following the habituation period are more readily observed.A similar procedure is difficult to replicate in fish.Anxiety-like behavior produced by immersion in a beaker (Kalueff et al., 2014) and by the netting procedures used to transfer fish from one tank to another (Tran and Gerlai, 2015) might have prevented observations of further increases in swimming distance or speed (Aponte and Petrunich-Rutherford, 2019;Tran et al., 2016).Some studies have suggested that anxiety-like responses of control zebrafish habituate during a session (Tackie-Yarboi et al., 2020;Wong et al., 2010).We also observed habituation in some of the indicators of anxiety-like behaviors.One possibility is to introduce repeated testing prior to the methamphetamine test.Repeated testing can result in habituation as indicated by a decrease in the magnitude of a response to various stimuli (Grissom and Bhatnagar, 2009;Thompson and Spencer, 1966).Repeated exposure to the test condition prior to methamphetamine tests might, therefore, be a means of decreasing the magnitude and/or duration of anxiety-like effects in future studies (Gaspary et al., 2018;Kirshenbaum et al., 2019;Wong et al., 2010).Additionally, testing fish in shoals rather than in isolation might also reduce anxiety-like behaviors and increase the ability to observe methamphetamine-produced behaviors.In support of this idea, a transient increase in methamphetamine-produced swimming distance was observed in fish treated with a high dose of methamphetamine (40 mg/L) and tested in shoals of six to reduce the anxiogenic effects of isolation (Bedrossiantz et al., 2021).
Methamphetamine is highly lipophilic drug and lipophilic drugs are generally well absorbed through respiration from the gills as well as through the skin in adult zebrafish (Bedrossiantz et al., 2021;Yin et al., 2019Yin et al., , 2021)).Toxicokinetic studies have determined the bioaccumulation of methamphetamine in zebrafish after 1-14 days of continuous exposure to either environmentally relevant methamphetamine concentrations (10 ng/L and 1000 ng/L) or to a pharmacologically relevant methamphetamine concentration (100 μg/ L) (Yin et al., 2019(Yin et al., , 2021)).For instance, female zebrafish were continuously exposed for 14 days to a mixture of methamphetamine and ketamine, another aquatic contaminant, followed by a 7-day elimination period (Yin et al., 2021).Parent drug and metabolite levels varied among specific body organs (brain, liver, intestine, ovary and muscle), with brain having the highest concentrations (ng/g) of methamphetamine and amphetamine across time.Generally, methamphetamine was considered to have relatively rapid uptake and elimination kinetics (t 1/ 2 ).The kinetic rate constant for absorption (Ku) revealed that uptake of methamphetamine was slower with increasing methamphetamine exposure concentration, whereas the kinetic rate constant for elimination (Ke) was greater with increasing methamphetamine exposure concentration.When zebrafish are exposed to the higher pharmacologically relevant methamphetamine dose (100 μg/L), uptake would be expected to be slower, but elimination more rapid relative to the environmentally relevant, low exposure methamphetamine concentrations.
Results showed that the t 1/2 for methamphetamine and amphetamine in zebrafish brain following exposure to the 100 μg/L for 14 days was 9.4 and 28.8 h, respectively.These results in zebrafish are consistent with findings from human studies in which slow-release oral methamphetamine was administered for 13 days (continuous administration) and the t 1/2 for methamphetamine elimination was 10.1 h (Cook et al., 1992).In contrast with zebrafish, methamphetamine distribution into human brain is lower than in most other body organs, however, clearance from brain is slow, resulting in long lasting brain exposure likely contributing to its neurotoxicity (Volkow et al., 2010).Taken together, methamphetamine is clearly bioavailable in zebrafish when exposure occurs via the aqueous environment, methamphetamine is widely distributed throughout various organs of the zebrafish body, and metabolism clearly plays a major role in methamphetamine pharmacokinetics in zebrafish.Future pharmacokinetic studies should obtain samples at shorter time intervals following acute dosing that correspond to the elimination half-life of methamphetamine so that more reliable estimates of its kinetic parameters can be obtained and related to its pharmacodynamic effects (Fowler et al., 2008;Rivière et al., 1999).Nevertheless, the current findings suggest that short behavioral testing periods, as typically conducted in zebrafish, do not likely capture the complete pharmacodynamic effects of methamphetamine on behavior.Our data show that the dose-related effects of methamphetamine on behavior (number of transitions along the sides of the tank, time spent in the bottom and the number of visits to the top zone of the test tank) emerge 10 min after exposure and continue to be apparent throughout the 50 min test session.Although measures of velocity and distance travelled were not significant in zebrafish, the trend for increased velocity and distance also persisted across the 50-min session.Moreover, behaviors in zebrafish that were significantly increased in response to methamphetamine can be used as a rapid, relatively inexpensive preclinical screen to evaluate compounds that block methamphetamine effects and thus may have potential as pharmacotherapies to treat methamphetamine use disorder.

FFig. 1 .
Fig.1.Mean (± standard error) changes in zebrafish behavior (transitions up and down the sides of the arena, time spent in the bottom zones, and visits to the top zones) following exposure to various doses of methamphetamine over a 50-min assay.A, C and E show the sum of the behavioral measure across the 50-min sessions.B, D, and F show the behavior data broken down by 5-min time bins across the session.Asterisks indicate doses that are significantly different from the control (0 mg/ L).Sample sizes for each dose (mg/L): 0 (n = 9), 0.1 (n = 7), 0.3 (n = 10), 1.0 (n = 6), 3.0 (n = 7).

Table 1
Mixed model results for each of the six behaviors measured.F-ratios, degrees of freedom and p-values are presented (p-values in bold are significant).