Novelty stressor induces differential brain oxidative stress and antioxidant profiles between proactive and reactive stress coping styles

Abstract

Neuroendocrine stressors can disrupt the brain’s redox equilibrium by generating high levels of reactive oxygen species (ROS) that lead to oxidative stress. To what extent the relationship between neuroendocrine and oxidative stress is modulated by stress coping styles remain unclear. We subjected proactive and reactive zebrafish to an acute novelty stressor (i.e. neuroendocrine stressor), and quantified changes in behavior and whole brain biomarkers of oxidative stress and antioxidants (DNA damage, total glutathione (GSH), glutathione ratio, oxygen radical absorbance capacity (ORAC), and superoxide dismutase (SOD). Fish exposed to the novelty stressor had significantly higher total glutathione, trend for higher ORAC, DNA damage, and glutathione ratio, and trend for lower SOD levels than controls. Those with a reactive stress coping style exhibited significantly higher levels of SOD and glutathione ratio, and a trend for higher ORAC compared to proactive. Principal component analysis showed that reactive individuals that experienced the novelty stressor had significantly higher composite antioxidant scores than proactive. Our data show that exposure to a novelty stressor results in changes to antioxidant and oxidative stress levels in the brain that differs by stress coping style. Reactive individuals showed elevated antioxidant capacity. Overall, our findings suggest that elevated reactivity to neuroendocrine stressors commonly seen in reactive stress coping styles may be mitigated through the glutathione buffering system and other antioxidants.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-55780-7.

Introduction

Animals face a variety of stressors throughout their lives, and their behavioral and physiological responses to overcome these stressors differ among individuals^1,2. For example, interacting with a novel stimulus or environment (i.e., novelty stressor) reliably induces a neuroendocrine stress response across animals such as rodents, birds, and fish^3–10. Stress-induced increases in glucocorticoids and catecholamines can lead to oxidative stress¹¹. These hormones increase metabolic activity and excitatory signaling, which leads to higher mitochondrial production of reactive oxygen species (ROS) and could inhibit antioxidant responses^12–16. Oxidative stress arises from an imbalance between the production of reactive oxygen species (ROS) and the antioxidant defense system, disrupting redox homeostasis and promoting damaging effects on the structural and functional aspects of the cell¹⁷. The degree of disruption from neuroendocrine stress exposure depends on individual characteristics, including their stress-coping mechanisms¹⁸.

Stress-coping styles are characterized by correlated suites of behaviors and physiological responses to a neuroendocrine stressor that are consistent across time and contexts^19–22. For the proactive–reactive stress coping spectrum, those with a proactive stress coping style tend to be more aggressive, bold, active, and with lower neophobia. Physiologically, proactive coping styles are characterized by lower cortisol levels and neuroendocrine stress reactivity, and higher sympathetic reactivity. The reactive stress coping styles exhibit the direct opposite characteristics^19,23.

Despite neuroendocrine stress axis activity and reactivity directly altering oxidative stress and differing by stress coping style, the link between stress coping styles and oxidative balance is not well characterized²⁴. A study using selectively bred mouse lines that differed in attack latency showed higher levels of serum antioxidant capacity but not reactive oxygen metabolites in the long attack latency line (presumably a reactive stress coping style)^25–27. In contrast, highly neophobic birds (i.e., reactive stress coping style) had lower plasma antioxidant capacity and higher oxidative stress than those with lower neophobia²⁸. Altogether, oxidative stress and antioxidant activity may vary between stress coping styles, but this effect is complex and tissue-specific, and highly context-dependent (e.g., whether the animal experiences stress or not)^25,29–31. Studies to date investigating the relationship between stress coping styles and oxidative profiles have mainly used peripheral components such as blood and liver. However, the oxidative profiles between blood and brain are not consistent^32–34. Despite the brain being key to coordinating behavioral and physiological processes sensitive to glucocorticoids, and vulnerable to oxidative stress, little is known about neuroendocrine-induced oxidative stress in the brain between stress coping^35,36.

Therefore, this study aims to investigate how an individual’s stress coping style can influence changes in brain antioxidant and oxidative stress levels in response to a novelty stressor. We hypothesize that there would be differences in brain oxidative stress and antioxidant responses between the two coping styles to a novelty stressor. We predict that individuals with the reactive stress coping style would exhibit increased behavioral stress levels and increased levels of markers for oxidative stress. To test this prediction, we compared brain oxidative profiles of zebrafish selectively bred to display the proactive and reactive stress coping styles in both stressed and baseline conditions.

Methods

Subjects

We used zebrafish line selectively bred to display, on average, the proactive and reactive stress coping style³⁷. The zebrafish used to generate these lines originated from wild caught zebrafish collected near the village of Gaighata, India. These lines are maintained through artificial selection of their behavioral response to a novelty stressor³⁷. These selectively bred lines have been extensively characterized and exhibit differences in stress-related responses across several anxiety-related behavioral assays and time, learning and memory performance, cortisol release rates, neural activity in the brain, responses to anxiogenic and anxiolytic compounds, escape-performance relayed to morphology, baseline neurotranscriptome profiles, but not gut microbiome diversity^{1,21,22,37–41}. Fish were selectively bred for 13 generations and were 4–8 months old at the time of testing. We randomly selected 24 zebrafish (12 baseline and 12 stressed fish). We chose this sample size based on our past studies using zebrafish lines, which demonstrated that sample size is the minimum to ensure sufficient statistical power to detect differences in behavioral and physiological responses (e.g., cortisol, brain gene expression, neural activity, pharmacodynamics)^{1,20–22,38,39,42,43}. Some samples were needed for oxidative stress and antioxidant assay optimizations to ensure the reliability of the protocol and were not included in the analysis. Our final sample sizes consisted of 9 baseline reactive (4 females, 5 males), 10 stressed reactive (3 females, 7 males), 12 baseline proactive (8 females, 4 males), and 10 stressed proactive (6 females, 4 males). Sex was determined upon visualization of the testes or ovaries on dissection. Prior to testing, all fish were housed in mixed-sex 40-L tanks on a custom-built recirculating system with solid filtration. Fish were maintained under consistent 14:10 light/dark cycle, temperature (26–27 °C), and water quality (pH: 6.8–7.4, conductivity: 600–900 μS), and fed twice daily with Tetramin Tropical Flakes (Tetra, USA).

Novelty stressor

We subjected a subset of fish to a novelty stressor by placing fish individually in a novel open field environment for 30 min. We quantified stress behaviors in this open field test (OFT) following established protocols^1,44. Prior studies using zebrafish show that novelty stressors through exposure to novel environments, such as novel diving tests and OFT, result in an elevated whole-body cortisol in zebrafish^6,8,9,21. In brief, fish were individually caught from their home tank and placed in a 31.75 cm × 31.75 cm acrylic tank with black walls and a white bottom filled with 4 L of water. We video recorded the trial for 30 min and later quantified behaviors using commercial automated tracking software (Ethovision XT, Noldus). To be consistent with other studies using OFT to measure response to a novelty, we only analyzed the first five minutes of the video recordings^1,42,45–47. We quantified time frozen (sec), total distance swam (cm), and average swim velocity (cm/sec). Freezing behavior was defined as the duration of time a fish moved slower than 0.5 cm/sec. At the conclusion of the open field test, we immediately euthanized the fish by decapitation with scissors, and placed the extracted brain on dry ice within 3 min of capture. Non-stressed (baseline) fish were caught directly from their home tank, euthanized by decapitation with scissors, and their brains were placed on dry ice within 3 min of capture. We did not use anesthesia during euthanasia and stored all the brains at -80 °C until tissue processing. All procedures were approved by the University of Nebraska Omaha Institutional Animal Care and Use Committee (17-070-09-FC).

Whole brain quantification of oxidative stress and antioxidant levels

In each brain sample, we measured various markers of oxidative stress and antioxidant capacity. We assessed total antioxidant capacity (TAC) as a measure of both enzymatic and non-enzymatic antioxidant capacity. We also measured glutathione since it is one of the most abundant small molecule antioxidants. We used an assay measuring both reduced and oxidized glutathione levels which also provides us with an assessment of oxidative stress levels. We evaluated oxidative DNA damage by measuring levels of 8-hydroxy-2-deoxyguanosine (8-OhDG). Below is a description of all the assays. For each assay, we ran samples in duplicate and most assays included inter-assay controls. All samples were processed by researchers that were blinded with respect to treatment. For all assays we used clear flat-bottom 96-well plates unless indicated otherwise.

Tissue preparation

We homogenized brain sample and prepared different aliquots of brain homogenate for total antioxidant capacity (TAC), superoxide dismutase (SOD), glutathione, and oxidative DNA damage assay³⁴. Brains were homogenized in a homogenization buffer (50 mM potassium phosphate buffer, 0.5 mM EDTA, pH 7.0) plus protease inhibitor cocktail (Sigma-Aldrich P2714, dissolved in 10 mL water, diluted 1:10 in homogenization buffer) and divided into separate aliquots for further analyses using previously described protocols^34,48. Homogenization was carried using a Cole-Parmer Battery-Operated Pestle Motor Mixer (Cole-Parmer North America, Vernon Hills, Illinois) at a concentration of 10 µL homogenization solution per 1 mg tissue. Once homogenized, the brain sample homogenates were split into three separate aliquots (40 µL, 60 µL, 100 µL). Aliquots of brain homogenate were created for each assay as follows: The 100 µL homogenates were centrifuged at 4 °C and 14,000 g for 10 min. Following centrifugation 60 µL of supernatant was removed and stored at − 80 °C until SOD analysis, while the remaining supernatant was discarded and the pellet was stored for quantifying DNA damage. The 40 µL aliquot was diluted 1:1 with lysis buffer (20 mM Tris–HCl, 137 mM NaCl, 1% NP-40, 10% glycerol, 2 mM EDTA) and centrifuged at 4 °C and 10,000 g for 10 min. After centrifugation, 65 µL of the supernatant was aliquoted into a 0.5 mL centrifuge tube. From this aliquot, 14 µL was transferred to a new tube for usage in a bicinchoninic acid (BCA) assay. Both aliquots were stored at -80 °C until analysis of oxygen radical absorbance capacity (ORAC) to assess TAC. The 60 µL aliquot was centrifuged at 4 °C and 10,000 g for 10 min. Once centrifuged, 45 µL of supernatant was aliquoted into another 0.5 mL tube and diluted 1:1 with 5% 5-sulfo-salicylic acid dihydrate (SSA) solution (Sigma-Aldrich 247,006) which was again centrifuged at 4 °C and 10,000 g for 10 min. Upon centrifugation, 70 µL of the supernatant was aliquoted into a 0.5 mL centrifuge tube. From this aliquot, 14 µL was transferred to a new tube for a BCA assay. Both aliquots of supernatant were stored at − 80 °C until analysis of glutathione.

Protein quantification

Protein concentration of each prepared brain homogenate (excluding DNA damage) was measured with a bicinchoninic acid (BCA) Protein Assay kit (Pierce, Rockford IL) following the manufacturer’s protocol. Frozen supernatant was thawed on ice and diluted 1:4 with buffer used in their respective sample preparation. We used 10 µL of this diluted supernatant in the assay. Absorbance was read by a plate reader (Epoch2T, Biotech Instruments, Winooski, VT, USA).

Superoxide dismutase (SOD)

Superoxide Dismutase (SOD), an enzymatic antioxidant, was measured via a modified competitive assay (Dojindo Molecular Technologies, Inc.) as previously described⁴⁸. We used 20 µL of prepared brain sample homogenate in duplicate in a 96-well plate at a range of different concentrations and mixed with 200 µL WST-1 working solution and xanthine oxidase. We used 4 or 5 different concentrations of prepared brain sample homogenate, ranging in protein concentration from 0.03 to 0.12 ng/ µL. The reaction rate was read over five minutes and analyzed for the IC50, or where 50% of the reaction is inhibited by SOD at a set concentration of sample and reported per unit of SOD (one unit of SOD is defined as the amount of enzyme in 20 µL of sample solution that inhibits the reduction reaction of WST-1 with superoxide anion by 50%). Absorbance was read by a plate reader (Epoch2T, Biotech Instruments, Winooski, VT, USA).

Total antioxidant capacity

Total antioxidant capacity (TAC) was measured in 20 µL of prepared brain sample homogenates via an Oxygen Radical Absorbance Capacity (ORAC) assay as previously described⁴⁹. Due to (i) our aim being to access overall oxidative status rather than compartment-specific antioxidant capacity and (ii) total antioxidant capacity not mapping cleanly onto biological oxidative stress dynamics in the brain^33,34, we did not differentiate between water- or lipid-soluble antioxidant activities. We standardized the samples to a protein concentration of 150 µg/mL. TAC is reported as µmol Trolox equivalent (TE)/µg protein. Absorbance was read by a plate reader (Spectramax M3, Molecular Devices, Sunnyvale, CA, USA). The intra-assay and inter-assay coefficient of variation was 1.4% and 4.6%, respectively.

Glutathione

We measured glutathione using a glutathione fluorescence kit (Arbor Assays, Ann Arbor, MI, USA) per manufacturer’s protocol. Samples were first diluted 1:2.5 with kit assay buffer to reduce SSA concentration to 1%. Fifty microliters of treated samples, standards, and blanks were plated in duplicate in a black half-area 96-well plate, and 25 μL of Thiostar detection reagent was added to each well. Plates were incubated at room temperature for 15 min before endpoint fluorescent emission at 510 nm with excitation at 390 nm was read by a plate reader (Tecan M200 Infinite) to quantify the concentration of free reduced glutathione (fGSH). Next, the concentration of oxidized glutathione dimers (GSSG) was quantified by adding 25 μL kit reaction mixture to each well followed by a second 15-min incubation to convert all GSSG in the sample to fGSH before a second fluorescence reading to find the total concentration of reduced glutathione (tGSH) in each sample. GSSG was calculated as (tGSH—fGSH)/2. We used the ratio of fGSH:GSSG as an overall marker of oxidative stress with higher ratios corresponding to less oxidized glutathione and hence lower levels of oxidative stress. All samples are reported standardized to protein concentration in the sample (total GSH / protein concentration). The intra-assay coefficient of variation was 2.4% (samples were measured using one plate).

Oxidative DNA damage

We measured oxidative DNA damage (8-OHdG) following a previously described protocol with modifications to allow for measurements of small tissue samples⁴⁸. DNA was extracted from homogenized brain tissue samples using a DNA extraction kit (Zymo quick-DNA miniprep plus kit, Irvine, CA, USA) per manufacturer’s protocol. Purified DNA samples were extracted from the columns by adding 100 µL of DNA elution buffer and incubating at room temperature for 5 min, followed by centrifugation at 17,000 × g for 1 min. This step was then repeated with an additional 100 µL of DNA elution buffer, again incubating for 5 min at room temperature and centrifuged at 17,000 × g for 1 min. This two-step elution procedure was used to increase the amount of extracted DNA. We then used ethanol precipitation to increase DNA concentration in each DNA sample through addition of 19 µL 3 M sodium acetate (pH 5.2) and 475 µL cold ethanol. Samples were then stored at − 20 ℃ overnight before undergoing 30-min centrifugation at top speed at 4 ℃. All ethanol was then removed, and the DNA was reconstituted in 15 µL elution buffer. After ethanol precipitation, the total DNA concentrations were determined using a BioTek take3 reader in an Epoch2T microplate reader using the Nucleic Acid Quantification software provided. We standardized all DNA samples to 100 ng/µL and stored sample at 4 ℃ until digestion.

DNA samples were digested following a modified protocol from Quinlivan and Gregory⁵⁰. Extracted DNA were added to a digest mix [173.61 Units benzonase, 208.33 mUnits phosphodiesterase, and 138.89 Units phosphatase in a tris–HCl buffer (20 mM, pH 7.9) containing 100 mM NaCl and 20 mM MgCl₂] at a concentration of 4 µL digest mix per 0.8 ng DNA and incubated at 37 ℃ overnight to ensure complete hydrolysis of DNA samples. Following digestion, samples were stored at − 20 ℃ until use in DNA damage assay. DNA damage (8-OHdG) was measured with a DNA damage ELISA kit (StressMarq, Biosciences Inc., Victoria, BC, Canada) following manufacturer’s protocol. Results are reported as ng damage (8-OHdG) / ng DNA.

Statistical analysis

All statistical analyses were performed in SPSS (Version 24). We used a general linear model to assess for the main effects of treatment, stress coping styles, and sex, and treatment by stress coping style by sex interaction effect on each oxidative stress and antioxidant marker (SOD, ORAC, total GSH, glutathione ratio, and DNA damage). For variables where there were no significant main effect of sex or treatment by stress coping style by sex interaction effect (Supplementary Table 1), we removed sex from the model and reported the main effects of treatment, stress coping style, and treatment by stress coping style interaction effect. We only found sex to have a significant main effect on ORAC levels and reported results with sex both included and excluded. We also conducted a principal component analysis (PCA) on the five oxidative stress and antioxidant markers to obtain principal component scores for evaluation in a general linear model as described above. As we only had behavioral measures for stressed fish, we used a general linear model to assess for main effects of stress coping styles and sex, and a stress coping style by sex interaction effect on total distance swam, average swim velocity, and time frozen. There was no significant main effect of sex or stress coping style by sex interaction effect (Supplementary Table 2). Thus, we removed sex from the model and report the main effect of stress coping style on each behavior. To investigate relationships between individual variation of behavior and brain oxidative stress and antioxidant quantities, we ran linear mixed models to examine for effects of redox biomarker, stress coping style, and redox biomarker by stress coping style interaction effect. All the plots were generated in RStudio (R version 4.5.0) using the ggplot2 package.

Results

Effect of novelty stress and coping styles on oxidative stress and antioxidant markers

Oxidative stress markers (DNA damage and glutathione ratio)

We found that stressed fish had a trend for higher DNA damage than control fish (stress treatment effect: =3.560, p = 0.068, η²ₚ = 0.092) (Fig. 1a). This effect of stress treatment on brain DNA damage did not vary by stress coping style (strain by stress treatment interaction = 0.447, p = 0.508, η²ₚ = 0.013). DNA damage levels were not different between stress coping styles (strain effect: =1.798, p = 0.189, η²ₚ = 0.049). For the glutathione ratio, we found that stressed fish tended to have a higher glutathione ratio level than control fish (stress treatment effect: = 3.291, p = 0.078, η²ₚ = 0.084) (Fig. 1b). This effect of stress treatment on brain glutathione ratio did not differ by stress coping style (strain by stress treatment interaction: = 0.150, p = 0.700, η²ₚ = 0.004). Glutathione levels did not vary between stress coping styles (strain effect: = 0.869, p = 0.358, η²ₚ = 0.024).

Fig. 1.

Effects of acute stress on oxidative stress and antioxidant markers across stress coping styles. The boxplot shows (a) DNA damage, (b) glutathione ratio, (c) ORAC, (d) total glutathione, and (e) SOD levels. The blue box indicates the baseline group of fish, while the yellow box represents the stressed group of fish. (*) represents p < 0.05 and (†) represents 0.05 < p < 0.10. Each dot represents a fish.

Antioxidant markers (ORAC, SOD and total glutathione)

We found that females had significantly higher levels of ORAC than males (sex effect: =9.384, p = 0.004, η²ₚ = 0.227). ORAC levels were significantly higher in the reactive relative to the proactive (strain effect: = 6.742, p = 0.014, η²ₚ = 0.174). With sex removed from the model, we observed that the reactive line had a trend for higher levels of ORAC than the proactive line (strain effect: = 3.766, p = 0.060, η²ₚ = 0.095) (Fig. 1c). This effect of stress coping styles on the brain ORAC did not differ by treatment levels (strain by stress treatment interaction: = 0.356, p = 0.555, η²ₚ = 0.010). ORAC levels did not vary between treatment levels (stress treatment effect: =0.746, p = 0.394, η²ₚ = 0.020). For total GSH, we observed that the stressed fish had significantly higher total glutathione levels (Fig. 1d) relative to control fish (stress treatment effect: =4.453, p = 0.042, η²ₚ = 0.110). We found that the reactive fish had significantly higher total glutathione levels compared to the proactive fish (strain effect: =15.261, p = 0.001, η²ₚ = 0.298). This effect of stress coping styles did not vary between treatment level (strain by stress treatment interaction: =0.067, p = 0.797, η²ₚ = 0.002). We observed that the stressed fish tended to have higher SOD levels compared to the baseline fish (stress treatment effect: =3.711, p = 0.062, η²ₚ = 0.020) (Fig. 1e). Also, the stressed reactive fish had higher SOD levels compared to the proactive fish (strain effect: =6.881, p = 0.013, η²ₚ = 0.049). This stress coping style effect did not vary between treatment levels (strain x stress treatment interaction: = 0.850, p = 0.363, η²ₚ = 0.010).

Integrated effect of novelty stress and coping styles on redox profiles

The PCA revealed two major axes of variation across all five biomarkers (Fig. 2a). Total GSH, SOD, and ORAC all loaded positively on the first principal component and accounted for 39.99% of the variation (antioxidant axis). GSH ratio and DNA damage loaded positively on the second principal component and accounted for an additional 22.5% of the variation (oxidative stress axis). We found that the stressed fish tended to have higher PC1 scores compared to the baseline fish (stress treatment effect: = 3.008, p = 0.092, η²ₚ = 0.081). The reactive fish had significantly higher PC1 scores than proactive fish (strain effect: = 14.396, p = 0.001, η²ₚ = 0.297) (Fig. 2b). Also, the stressed fish had significantly higher PC2 scores compared with the baseline fish (stress treatment effect: =9.116, p = 0.005, η²ₚ = 0.211) (Fig. 2c). For the behavior variables, stress coping style did not have an effect on time frozen (strain effect: = 0.428, p = 0.521, η²ₚ = 0.023), average swimming speed (strain effect: = 2.685, p = 0.119, η²ₚ = 0.130), or distance swam (strain effect: = 2.424, p = 0.137, η²ₚ = 0.119). We found a trend for main effects of DNA damage on time frozen only (DNA damage: = 3.538, p = 0.080). There were no significant relationships between levels of other redox biomarker (discrete or composite), stress coping style, or redox biomarker by stress coping style interaction effect on variation of any behavior (Supplementary Table 3).

Fig. 2.

Composite measures (principal component scores) of oxidative stress and antioxidant biomarkers across stress coping styles. (a) Plots of principal components (b) SOD, total glutathione, and ORAC loaded onto principal component 1 (antioxidant axis), whereas (c) DNA damage and glutathione ratio loaded onto principal component 2 (oxidative stress axis). (*) represents p < 0.05 and (†) represents 0.05 < p < 0.10. Each dot represents a fish.

Discussion

This study investigated the effects of novelty stressor on brain oxidative profiles and how these effects are influenced by individual differences in stress coping style. Our findings show that stress coping style and responses to a novelty stressor alters discrete and composite measures of brain oxidative stress and antioxidant markers. More specifically, higher levels of antioxidants (SOD, total glutathione, ORAC, and PC1) in reactive individuals suggest that an individual’s stress-coping style can directly modulate the antioxidant system response to novelty stressors to potentially mitigate cellular oxidative damage in the brain.

Our study showed that exposure to a novelty stressor increases markers of oxidative stress in the brain. Stressed fish tended to have higher amounts of DNA damage and glutathione ratio in both coping styles, suggesting increased molecular damage and altered redox balance in the system via dysregulation of glutathione redox status (Fig. 1). Similarly, the composite measure of oxidative stress (PC2) significantly increased in stressed fish (Fig. 2). Novelty stressors have been well established to induce a neuroendocrine stress response in a variety of animals^{3,5,6,8,21,22,51}. More specifically, in zebrafish, prior studies (including one using the same selectively bred line as the current study) have shown that novelty stressors through exposure paradigms such as the novel diving test and open field test result in elevated whole body cortisol levels^6,8,21,51,52. While a limitation of the current study is that we did not measure cortisol levels, it has been well established that exposure to novelty will reliably induce a glucocorticoid stress response. Our observations of changes in brain oxidative stress levels by a presumed neuroendocrine stress response are consistent with other studies showing similar neuroendocrine stress-induced oxidative stress in zebrafish¹³, birds⁵³, and rodents⁵⁴. Given the evidence that activation of the neuroendocrine stress axis can alter redox balance^15,55,56, we speculate that an excessive release of glucocorticoids altered oxidative stress and antioxidant levels in the brain by increasing ROS production, leading to higher glutathione ratios and DNA damage⁵³.

The onset of oxidative stress is not only indicated by increases in pro-oxidants but can also be a result of decreases in antioxidants^57,58. We found that across both coping styles, relative to control, there was a significant decrease in antioxidant biomarker levels as well as in the composite measures (PC1 scores) (Figs. 1 and 2). This suggests that a novelty stressor and presumed neuroendocrine stress can also suppress the antioxidant defense system. Our findings are consistent with other studies showing that acute neuroendocrine stressors suppress antioxidant defenses in the brain of zebrafish¹³ as well as in rodents showing a decrease in antioxidant levels in both plasma and the brain^55,59,60. One possible explanation posits that acute neuroendocrine stress increases oxidative demand, which can deplete or shift antioxidant pools, thereby reducing measurable antioxidant capacity^59,60.

In addition to overall stress treatment effects on oxidative stress and antioxidant levels, we observed coping-style specific differences. Fish with the reactive stress coping style had significantly higher brain antioxidant activity than the proactive fish across all three biomarkers (total glutathione, total antioxidant capacity, SOD) and the composite measure (PC1 score) (Figs. 1 and 2). Our findings are consistent with prior studies showing higher serum antioxidant capacity in reactive-like mice^25,61,62. However, opposite patterns have been reported in reactive-like birds, which showed lower plasma antioxidant levels²⁸. Notably, while previous studies have largely used peripheral tissues like serum and plasma, our study used whole-brain tissue to measure redox biomarkers. This distinction is important because, relative to other organs, the brain is highly susceptible to oxidative stress due to its high metabolic demand and lipid-rich composition^35,63–65. Further, the redox profiles between peripheral tissue and brain are not always correlated^32–34. With the brain’s critical role in modulating behavioral displays and physiological responses, we hypothesize that changes in its oxidative stress and antioxidant states may facilitate the biases seen in a stress coping style.

SOD and glutathione are essential for neutralizing ROS, and their levels can be a proxy for antioxidant capabilities^66,67. ORAC provides an overall measure of antioxidant capacity, including small molecule and enzymatic antioxidants^68,69. We hypothesize that elevated antioxidant capabilities and capacities in the brain of those with a reactive stress coping style are related to their heightened sensitivity to neuroendocrine stressors. Many studies have shown that the reactive stress coping style have higher baseline, faster release rate, or higher peak cortisol levels than proactive individuals^{21,51,70–72}. Reactive individuals’ higher antioxidant capacity and capabilities may help mitigate the impact of showing a more frequent cortisol-induced pro-oxidant production in their brains (i.e., oxidative stress). We detected no relationship between individual variation in stress behavior and oxidative stress or antioxidant brain marker levels (Supplementary Table 3). While there may be no linear relationship between stress behaviors and redox biomarkers, we cannot rule out that the lack of differences in presumed neuroendocrine-induced stress behavior between the stress coping styles could be a confound. The inconsistency of the behavioral observations of the current study with prior studies using these selectively bred lines^21,44 could be due to natural variation within the lines.

Overall, this study addresses an important knowledge gap related to whether presumed neuroendocrine stress from exposure to novelty stressor alters brain oxidative profiles between stress coping styles. Our results revealed that a key distinction between coping styles is their antioxidant capability and capacity in response to presumed neuroendocrine stressors. Individuals with the reactive stress coping style may have higher brain antioxidant capabilities and capacities to mitigate the presumably more frequent neuroendocrine stress-induced oxidative stress during daily activities. It is unclear how stressor duration may impact the relationship between stress coping styles. Future studies should explore the impacts of chronic neuroendocrine stress on brain redox balance differences between the stress coping styles. Also, quantification of cortisol would allow for more thorough analyses of the relationship between individual variation of neuroendocrine stress activity and brain oxidative stress antioxidant biomarkers across all fish and within a stress coping style. Ultimately, our study highlights that the brain’s redox profile may be a key trait that is part of the correlated suite of behaviors and physiological responses to neuroendocrine stressors that characterize a stress coping style.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

PSJ: Data analysis and interpretation; manuscript drafting. RJF: Data collection and analysis; manuscript drafting. BJB: Data collection and analysis; manuscript drafting. PD: Data analysis and interpretation; funding acquisition; manuscript drafting. RYW: Project conceptualization and oversight; data collection; funding acquisition; data interpretation; manuscript drafting.

Funding

This project was funded by the National Institutes of Health (R15MH113074 & 5P20GM103427 to RYW; R15GM150286 & R35GM161778 to PDD) and AALAS Grants for Laboratory Animal Science (to PDD).

Data availability

The data supporting this study are included in the electronic supplementary document.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

Ethical approval for this study was obtained from the University of Nebraska at Omaha’s Institutional Animal Care and Use Committee, IACUC (17-070-09-FC). The present study followed national and institutional guidelines for humane animal treatment and complied with relevant legislation from IACUC.