Stressor controllability modulates the stress response in fish

Abstract

Background: In humans the stress response is known to be modulated to a great extent by psychological factors, particularly by the predictability and the perceived control that the subject has of the stressor. This psychological dimension of the stress response has also been demonstrated in animals phylogenetically closer to humans (i.e. mammals). However, its occurrence in fish, which represent a divergent vertebrate evolutionary lineage from that of mammals, has not been established yet, and, if present, would indicate a deep evolutionary origin of these mechanisms across vertebrates. Moreover, the fact that psychological modulation of stress is implemented in mammals by a brain cortical top-down inhibitory control over subcortical stress-responsive structures, and the absence of a brain cortex in fish, has been used as an argument against the possibility of psychological stress in fish, with implications for the assessment of fish sentience and welfare. Here, we have investigated the occurrence of psychological stress in fish by assessing how stressor controllability modulates the stress response in European seabass (Dicentrarchus labrax).

Results: Fish were exposed to either a controllable or an uncontrollable stressor (i.e. possibility or impossibility to escape a signaled stressor). The effect of loss of control (possibility to escape followed by impossibility to escape) was also assessed. Both behavioral and circulating cortisol data indicates that the perception of control reduces the response to the stressor, when compared to the uncontrollable situation. Losing control had the most detrimental effect. The brain activity of the teleost homologues to the sensory cortex (Dld) and hippocampus (Dlv) parallels the uncontrolled and loss of control stressors, respectively, whereas the activity of the lateral septum (Vv) homologue responds in different ways depending on the gene marker of brain activity used.

Conclusions: These results suggest the psychological modulation of the stress response to be evolutionary conserved across vertebrates, despite being implemented by different brain circuits in mammals (pre-frontal cortex) and fish (Dld-Dlv).

Keywords: Controllability; Cortisol; Dorsolateral pallium; Fish welfare; Immediate early genes; Stress.

Fig. 1

Behavioral and physiological responses of seabass expressed by fish during the test session (experimental treatments: CTR = stressor controllability; UnCTR = stressor uncontrollability; CTRUn = loss of stressor controllability): a Time in freezing (%), b frequency of escape attempts, c shoal cohesion, d Exploratory behavior, f plasma cortisol concentrations measured 30 min after the test session. Linear mixed models with planned comparisons are indicated; * p < .05; ** p < .01; *** p < .001; ns—non-significant. All descriptive statistics are mean ± SEM

Fig. 2

Expression (mean ± SEM) of the immediate early genes egr-1, c-fos, bdnf and npas4 in the Dm, Dld, Dlv and Vv regions of Sea bass brain under different experimental conditions (CTR = stressor controllability; UnCTR = stressor uncontrollability; CTRUn = loss of stressor controllability). Significant differences (planned comparisons) in expression levels between experimental conditions (i.e. CTR vs. UnCTR; CTR vs. CTRUn and UnCTR vs. CTRUn) are indicated by asterisks: * p < .05; ** p < .01; *** p < .001; ns—non-significant

Fig. 3

Neurogenomic states of seabass, as described by correlation (r) matrices of immediate early genes expression in the different brain nuclei (Dm, medial zone of the dorsal telencephalic area; Dl, lateral zone of the dorsal telencephalic area; Dld, dorsal lateral zone of the dorsal telencephalic area; Dlv, ventral lateral zone of the dorsal telencephalic area; Vv, ventral nucleus of the ventral telencephalic area) for each experimental treatment (CTR = stressor controllability; UnCTR = stressor uncontrollability; CTRUn = loss of stressor controllability); Color scheme represents r-values from − 1 (blue) to 1 (red); Asterisks indicate significant correlations after p-value adjustment for multiple correlations: *p < 0.05; **p < 0.01; ***p < 0.001; different capital letters indicate significantly different co-expression patterns among experimental treatments, and different small letters indicate significantly different co-expression patterns among brain nuclei, using the QAP correlation test

Fig. 4

Linear discriminant analysis of cortisol and egr-1, c-fos, bdnf and npas4 expressed in the candidate brain nuclei from seabass. The significant functions 1 and 2 highlight the three coping responses of fish under three experimental conditions: CTR = stressor controllability (circles); UnCTR = stressor uncontrollability; CTRUn = loss of stressor controllability (triangles). Discriminant scores for each individual are plotted and stars represent the centroid of each classified group