Juvenile three-spined stickleback (Gasterosteus aculeatus) were caught at Vikhög harbor, Sweden (55°43’N 12°57’E) on 26 February 2017. The animals were initially kept at the Department of Zoology, Stockholm University, and were transferred to the Biomedical Center at Uppsala University on 20 July 2017 where they were kept in two 500L holding tanks with flow-through of tap water at ~16°C and under constant aeration. Animals were fed daily with frozen artemia (Gibbon, Sweden). Two weeks before the start of the experiment, fish were individually tagged with visible implant elastomer tags (Northwest Marine Technology, USA).
We used a static system for oxazepam treatment, consisting of 12 aerated 22L glass aquaria with 4 oxazepam concentrations (0, 1, 10 or 100 μg L-1). On the first day of the experiment, behavioral testing was performed using a 15-minute long novel tank diving test, after which fish were transferred to exposure tanks until each treatment tank contained 20 fish. During the 8 days of treatment, animals were fed with frozen artemia daily, after which tanks were cleaned of feces.
On the 7th day of treatment, animals were tested for behavior in the novel tank diving test (15 min), and subsequent scototaxis test (15 min). On day 8 of the experiment, animals were individually caught from the treatment tank and exposed to confinement stress for 30 minutes in plastic beaker containing 100 mL water. Immediately thereafter, animals of were euthanized with an overdose of MS-222 and a cut through the spinal cord. For the animals from the tanks containing 0 and 100 μg L-1, the brain was removed from the skull and dissected into 5 parts (telencephalon, olfactory bulbs, cerebellum, brain stem and the remaining midbrain), and the remainder of the head was cut off. These brain parts as well as the body were frozen on dry ice and stored at -80°C. Brain dissection was not performed for the animals exposed to 1 and 10 μg L-1, instead the head was removed and the body was frozen on dry ice and stored at -80° C.
Fish samples were stored at -20 °C and taken out to thaw. 50 mL falcon tubes were used to each individual fish, and were weighed before insertion of fish sample. Samples were cut with scissors to small pieces (>1mm) before homogenisation in PBS buffer using a tissue homogenisator VWR CDI12 with S12N-7S rotor (VWR International, Sweden). Cortisol was extracted from the tissue homogenate by adding ethyl acetate, after which samples were centrifuged at 1000g for 5 minutes. The supernatant was transferred into a borosilicate glass tube (VWR, Sweden) and the ethyl acetate was removed by evaporation using a nitrogen gas sparge with heat block at 40 °C (Bibby Scientific, USA). Ultimately samples were reconstituted in 2400 µl ethyl acetate.
For detailed method, see attached protocol in appendix. The technique that was used to analyze cortisol concentration of each fish sample was radioimmunoassay (RIA). RIA is a common method to measure concentration of substances, cortisol in this case. By using a standard curve using a dilution series of cortisol we can determine the cortisol concentration in analyzed sample. A scintillaion machine (Tri-Carb 2910TR®, PerkinElmer, USA) was used to perform RIA and data was registered by a software (Quantasmart™, PerkinElmer, USA). Count time of scintillation machine was 5 min/sample.
To determine if cortisol response is influenced by different concentrations of oxazepam, we analyzed the data using R (R Development Core Team, 2018), using packages ‘lme4’ (Bates et al., 2015; ) and ‘lmerTest’ (Kuznetsova A, Brockhoff PB, Christensen RHB, 2017) to fit a linear mixed-effects model (LMM) to the data. We included oxazepam concentration, time of day caught, weight of the fish before dissection and all interactions as fixed factors, and sample and tank as random factors. Because of the high complexity of this LMM, the model is simplified using backward step-wise selection (step function in lmerTest; ref), which resulted in a minimum adequate model (MAM) with oxazepam concentration and weight as fixed factor, and sample as random factor.
As shown in table 1 and figure 1, the minimum adequate LMM revealed a close to significant effect of oxazepam treatment on whole-body concentrations of cortisol in stickleback. However, there is a significance effect of the weight of fish after dissection.
Visual inspection of the differences between replicate tanks revealed that the groups exposed to 1 ug/L differed strongly between replicates (figure 2). If this concentration is removed from the model and only control and fish exposed to 100 μg/L is considered, oxazepam treatment had a significant lowering effect on whole-body cortisol concentration after confinement stress (table 2, p=0.032). On average, oxazepam treated fish have whole body cortisol concentrations 0.114 ng/mg fish which is lower than in control treated animals.
Table of contents :
2. Materials and method