Negative impacts of social isolation on behavior and neuronal functions are recovered after short-term social reintroduction in zebra�sh.

Recently, social isolation measures were crucial to prevent the spread of the coronavirus pandemic. However, the lack of social interactions affected the population mental health and may have long-term consequences on behavior and brain functions. Here, we evaluated the behavioral, physiological, and molecular effects of a social isolation (SI) in adult zebra�sh, and whether the animals recover such changes after their reintroduction to the social environment. Fish were submitted to 12 days of SI, and then reintroduced to social context (SR). Behavioral analyses to evaluate locomotion, anxiety-like and social-related behaviors were performed after SI protocol, and 3 and 6 days after SR. Cortisol and transcript levels from genes involved in neuronal homeostasis (c-fos, egr, bdnf), and serotonergic (5-HT) and dopaminergic (DA) neurotransmission (thp, th) were also measured. SI decreased anxiety-related phenotype, and altered social behaviors in zebra�sh. Fish submitted to SI also presented changes in the transcript levels of genes related to neural activity, and 5-HT/DA signaling. Interestingly, most of the behavioral and molecular changes induced by SI were not found again 6 days after SR. Thus, we highlight that SR of zebra�sh to their conspeci�cs played a positive role in social behaviors and in the expression of genes involved in different neuronal signaling pathways that were altered after 12 days of SI. This study brings unprecedented data on the effects of SR in the recovery from SI neurobehavioral alterations, and reinforces the role of zebra�sh as a translational model for understanding the neurobiological mechanisms adjacent to SI and resocialization.


Introduction
The coronavirus pandemic led government leaders to adopt the social isolation as a main measure to prevent the spread of the virus [1].This position remains indispensable so that the infection does not advance in a way that overloads and collapses health systems [2].Meanwhile, it is evident that the lack of social interactions in this period dramatically increased stress, affecting the population mental health on a scale unprecedented in recent history.Increased anxiety, depression, panic disorders, fear, sleep disturbances, and alcohol consumption have been observed in individuals subject to social blockade conditions in different countries [3][4][5].It is suggested that at the present time the population is experiencing a new "pandemic" related to the catastrophic effects of social isolation on the mental health of individuals.Social isolation is not only associated with behavioral changes but also with increased sympathetic tone and activation of the hypothalamic-pituitary-adrenal (HPA) axis in the central nervous system (CNS) [6,7].Neural plasticity and intracellular communication are also affected by the lack of social interactions, altering the functionality of important neurotransmitter systems such as the serotonergic and the dopaminergic system [8,9]).Serotonin (5-HT) and dopamine (DA) are known to regulate anxiety, emotion and stress responses, and their signaling imbalance are often linked to the development of neuropsychiatric diseases [10][11][12].Therefore, it is likely that the challenges that occur during the period of con nement may have long-term consequences on brain functions, and may require medium/long term to be reversed.
Zebra sh (Danio rerio) is a translational model organism widely used to study social interactions and their disruptions [13][14][15].Like humans, zebra sh are predominantly diurnal, establish groups, form social hierarchies, and exhibit complex social interactions [16,17].Other advantages include complex, wellestablished and easily quanti able social behaviors [18], neuroanatomical characteristics [19][20][21], neurotransmitter systems, and highly conserved CNS mechanisms that regulate social interactions in zebra sh [22][23][24].Social behavior has a bene cial role on zebra sh, meanwhile the lack of social interactions can trigger complex pathological states of the zebra sh CNS [16].The effects are dependent on the type of social isolation protocol used (short or long term) and may vary depending on the age of the animal and the strain used in the experiment.Animals subjected to different protocols of social isolation can develop increased states of stress and anxiety, and depressive-like phenotypes [25][26][27].Interestingly, like humans, zebra sh recover better from stressful events in the presence of conspeci cs [28].
To obtain knowledge about the impacts of social isolation on neurobehavioral/neuro-molecular responses, and possible recovery approaches, herein we investigated 1) behavioral, physiological, and molecular changes caused by a 12-days social isolation protocol in adult zebra sh, and 2) whether the animals recover the changes after their reintroduction to the social environment.

Animals and ethical aspects
Subjects were short n wild-type zebra sh (Danio rerio) of mixed sexes (50:50 male:female ratio).Fish were obtained from a local supplier (Hobby Aquários, RS, Brazil) and acclimated in 40-L tanks for two weeks before any experimental procedure in a maximum density of four sh per liter.Tanks were lled with non-chlorinated water kept under constant mechanical, biological, and chemical ltration.Water temperature, pH, and conductivity were set at 26 ± 2ºC, 7.0-7,2, and 1500-1600 µS/cm, respectively.Animals were kept on a 14/10 light/dark photoperiod cycle (lights on at 7:00 am) and fed with a commercial ake sh food (Alcon BASIC®, Alcon, Brazil) twice daily.All protocols were approved by the Ethics Commission on Animal Use of the Federal University of Santa Maria (process number 2649140717).To ensure data reliability, we ran two independent batches for physiological-and neuralrelated markers analyses, and three independent batches for the behavioral investigation.

Experimental design -social isolation (SI) and social reintroduction (SR) protocols
After acclimatization period, sh were randomly divided into 3 groups: Control (CT -group housed), Social Isolation (SI -group housed and then isolated), and Social Reintroduction (SR -Group housed, isolated, and then returned to group housing).CT group consisted of sh living in normal social conditions for the species, i.e., housed in shoals and establishing social interactions.Fish of SI group were allocated in shoals for 12 days, and then, housed in individual tanks (15 cm length × 6 cm height × 12 cm width) for 12 days without any type of social interaction with the external environment (without access to visual, olfactory or tactile cues from other sh) [29].This time was chosen due to the translational relevance by simulating the period of con nement experienced as strategy to prevent coronavirus infection.Based on the life expectancy of humans and zebra sh, 6 months of social isolation in humans is equivalent to approximately 12 days in zebra sh.After the SI period, neurobehavioral analyzes were performed in the CT and SI groups, and the SR group was formed.The SR consisted of reintroducing animals to a social context after 12 days of isolation.Animals were reintroduced back to a single tank where they lived in groups before the isolation and were able to restore the social interactions for 12 days.The same neurobehavioral analyzes performed immediately after SI were performed in the SR group on 3 (SR3) and 6 (SR6) days after social reintroduction.

Behavioral measurements
The behavioral tests were performed between 10:00 am and 4:00 pm.Experimental tanks were lled with non-chlorinated in the same conditions that housing tanks and kept on a stable surface.Behaviors were recorded using a webcam connected to a laptop at 30 frames/s.The videos were analyzed using the videos-tracking software's ANY-maze™ (Stoelting CO, USA) and DeepLabCut, or manually depending on the test.The water was changed after each individual tested.

Novel tank test (NTT)
Locomotor and exploratory activities were analyzed using the NTT, which may re ect habituation to novelty stress [30].Zebra sh were individually placed in a novel apparatus (25 cm length × 15 cm height × 6 cm width) lled with 2 L non-chlorinated water for 6 min and their swimming behavior was recorded.The tank was virtually divided into two horizontal sections (bottom and top) to assess the vertical exploration using the following endpoints: transitions and time spent in top and latency to enter the top area.Distance traveled, absolute turn angle, and time immobile were used as locomotor measures.

Light-dark test (LDT)
The LDT consisted of a rectangular aquarium (30 cm length x 15 cm height x 10 cm width) divided into two equal parts colored by black or white self-adhesive lm lled with 2.5 L of non-chlorinated water (8 cm height).During the trial, uorescent lamps were used for illumination (approximately 250 lux above the tank).All procedures were performed based on a protocol previously described [31].After analysis, sh were removed from the home tank and gently placed in the white side of the apparatus.Then, during a session of 6 min, the following endpoints were determined: latency to enter in the dark area, time spent in lit area, shuttling, and number of risk assessment episodes.Risk assessment was counted when the sh enter in the white side and immediately (< 1 s) returned to the dark compartment, or just partially enter in the light area [31].Here, the number of risk assessment episodes was measured manually by three trained researchers (inter-rater reliability > 0.85).

Preference for conspeci cs test (PCT)
The preference for conspeci cs re ects a natural tendency of zebra sh to establish a group of interaction [32,33].The aggregation of sh and the formation of a shoal can be modi ed by different situations and compounds.Animals were placed individually on the central aquarium, situated between an empty tank and another tank with ve conspeci cs (stimulus group) allowing only visual contact between the sh.PCT was performed for 2 min and the apparatus was divided into two sections (A1 and A2), where A1 represented the area closer to the stimulus group.The entries, latency to enter and the time in which sh remained in A1 section were measured and expressed as a tendency of group approach.

Shoal test (ST)
After the experimental protocol, a zebra sh shoal (4 animals from the same group, n = 5) was placed in a novel apparatus (25 cm length × 15 cm height × 6 cm width) lled with 2 L non-chlorinated water for 5 min, and their shoal behavior was recorded [34].ST allows the animals exchange chemical, visual and tactile contacts.Animals tracking (X and Y coordinates) was obtained using DeepLabCut [35] considering three points (head, trunk and tail).Using ResNet 101, we trained a deep neural network on 180 frames (divided across 9 shoals/3 shoals per group), the network was then rede ned and improved by correcting outliers and including 750 additional frames.The X and Y coordinates for each sh trunk was then used to obtain the average (avg) inter-sh distance and the mean shoal area across the entire duration of the test (5 min).Calculations for the average inter-sh distance and shoal area were obtained using RStudio (v.2022.7.0.548).

Mirror-induced aggression (MIA) test
Aggressive behavior was assessed using the MIA test [32,36].After isolation period, sh were transferred individually to the experimental tank (25 cm length × 15 cm height × 6 cm width) and a mirror was placed inclined at 22.5° to the back wall of the tank so that the left vertical edge of the mirror was touching the side of the tank and the right edge was further away.Thus, when the experimental sh swam to the one side of the tank their re ected image appeared closer, evoking aggression-like responses.Tanks were virtually divided into three areas related to their proximity to mirror [32] and the following behaviors were determined in a single 5-min session: number of entries and time spent in mirror area, as well as the number and duration of aggressive episodes.Aggressive display was counted when sh presented erection of dorsal, caudal, pectoral, and anal ns, usually associated with undulating body movements and attack, and manually counted by two trained observers blind to the experimental condition (inter-rater reliability > 0.85) [37].

Cortisol extraction and analysis
Whole-body cortisol was extracted as described elsewhere [38,39].Immediately after the behavioral tests, animals were euthanized and frozen in liquid nitrogen for 10-30 s to cortisol extraction.Each sh was weighed and minced in 2 mL of phosphate-buffered saline (PBS pH 7.4).The content was transferred to a glass test tube, which 5 mL of laboratory-grade ethyl ether was added.The tube was vortexed for 1 min and centrifuged for 10 min at 3000 rpm.The sample was immediately frozen in liquid nitrogen, and the unfrozen portion (ethyl ether containing cortisol) decanted and transferred to a new tube to evaporate overnight.The cortisol content was resuspended in 1 mL of PBS pH 7.4 and used to quantify the amount of whole-body cortisol (EIAgen ™ CORTISOL test, BioChem ImmunoSystems).Both cortisol curve and samples were run in duplicate and the mean values obtained for each shoal were calculated and expressed as ng cortisol/g tissue [39].

Analysis of mRNA expression by quantitative reverse transcription PCR
Total RNA was extracted from samples in which 3 brains were pooled using Trizol Reagent (Invitrogen®, Brazil) and the RNA quantity and quality was analyzed in NanoDrop™ 2000/2000c (Thermo Scienti c) [40].After quanti cation, the total RNA was treated with DNase I (Invitrogen), and the cDNA was synthesized with M-MLV reverse transcriptase enzyme and oligo-dT primer using the manufacturer's protocol (Invitrogen).The qRT-PCRs were carried out with Platinum Taq DNA Polymerase (Invitrogen).In 20 µL reaction volumes each in 96-well plates with 10 µL of cDNA as template, 3.52 µL of H2O Milli-Q, 2µL of 10 x PCR buffer, 1.2 µL 50 mM MgCl2, 0.4µL of 10 mM dNTPs Mix, 0.4 µL of 10 µM Forward Primer, 0.4 µL of 10 µM Reverse Primer, 2µL of SYBR Green and 0.08 µL of 2U/rxm Taq DNA Polymerase.Threshold and baselines were manually determined using the StepOne Software v2.3 (Applied Biosystems, NY) and the CT (cycle threshold) value for each sample was calculated and recorded using formula 2 − ΔΔCT [41].Each well was analyzed in triplicate, and the ΔCT value was obtained by subtracting the β-actin CT value from the CT value of the gene of interest.Results were expressed as relative mRNA expression.All primer sequences used in this study were designed using NCBI tools, manufactured by Invitrogen, and are described and listed in Table S1 (supplementary material -S1).

Statistics
Normality of data and homogeneity of variances were analyzed by Kolmogorov-Smirnov and Bartlett's tests, respectively.The effects of isolation protocol on behaviors and physiological and molecular parameters were analyzed by one-way analysis of variance (ANOVA) followed by Tukey's post-hoc test.For the number of entries per zone in the PCT test, two-way ANOVA was performed using zones (3 levels -close, middle and far) and groups (3 levels -CT, SI and SR6) as factors.Data were expressed as mean ± standard error of the mean (S.E.M.) and the signi cance level was set at p < 0.05.
Figure 2C shows the effects of SI and SR in the ST.A signi cant effect for shoal area (F (2,12) = 4.501, p = 0.0348) and distance between sh (F (2,12) = 4.563, p = 0.0336) was observed.Here, there was a decrease in the distance between sh and shoal area for SI in relation to the CT (p* < 0.05), but not compared to SR6.No effect was found for minimum (F (2,12) = 0.7831, p = 0.4790) and maximum (F (2,12) = 3.605, p = 0.0594) distance reached between sh.

Social reintroduction reverses several effects of SI on neural-and physiological-related markers
Cortisol levels and the mRNA expression levels of key neural-related markers are showed in Fig. 3.A signi cant ANOVA effect was observed for whole-body cortisol levels (F (2,18) = 7.762, p = 0.0037; Fig. 3A).
Similarly, fosaa (F (2,47) = 11.77,p < 0.0001), fosab (F (2,41) = 18.69, p < 0.0001), egr (F (2,50) = 10.41,p = 0.0002), and bdnf (F (2,45) = 21.82,p < 0.0001) mRNA levels were altered in zebra sh brain (Fig. 3C).SI decreased cortisol levels and the mRNA levels of fosaa, fosab, egr and bdnf compared to CT group (p* < 0.05).No effect was found between CT and SR for stress-related markers, but a signi cant effect between SI and SR was found for fosaa, fosab, egr and bdnf (p* <0.05).A heatmap summarizing the differences for the expression of neural markers is illustrated in Fig. 3B.3.4 SI alters the mRNA expression of important genes involved in dopaminergic and serotonergic signaling which are recovered after social reintroduction Figure 4 shows the effects of SI on the mRNA expression levels of genes involved in dopaminergic (4A and B) and serotoninergic (4C and D) responses.There was a signi cant effect relative to mRNA levels of th (F (2,28) = 5.060, p = 0.0133) and th2 (F (2,24) = 4.138, p = 0.0286), where SI increased th and th2 levels compared to the CT (p* < 0.05).SR6 did not differ from the CT group.The mRNA relative levels of drd1a, drd2a, drd2b, and slc6a3 did not differ among the groups, and their average expression is summarized on Fig. 4A.
There was also a signi cant effect of SI on serotonergic system related-gene thp1a (F (2,30) = 7.913, p = 0.0017) and thp2 (F (2,32) = 9.205, p = 0.0007).The mRNA relative levels of thp1a and thp1b were increased in the animals from SI group (p* <0.05), meanwhile no signi cant differences were found between SR and CT animals.In SR6, the mRNA relative levels of thp1a and thp2 were signi cantly different from SI animals (p* < 0.05).The mRNA relative levels of htr1aa, htr1b, and htr2aa did not differ among the groups (Fig. 4C).

Discussion
To our knowledge, this is the rst study that evaluated the effects of social reintroduction after a 12-days SI protocol in adult zebra sh, and brings unprecedented data on the effects of reintroducing animals to the social environment, and their recovery from behavior and molecular alterations.The exposure to SI for 12 days resulted in a decreased anxiety-related phenotype associated with alterations in social behavior.At basal levels, cortisol of SI exposed animals was decreased, and the mRNA expression of genes related to neural activity, serotonergic, and dopaminergic systems was disturbed.SR had positive impact in animals' behaviors of SI.The aggression-like behavior disappeared 3 days after SR.After 6 days from the SR, all social behavior, physiological and molecular analyzed parameters returned to the baseline.
Conserved genetic of neural circuits and easily quanti able complex social behaviors traits make zebra sh an excellent model to elucidate the neural mechanisms involved in normal social interactions and their disruptions [15].Developmental and adult social isolation protocols using zebra sh have brought translatable data in this eld [29,27,42,43].Here, we demonstrated that sh submitted to SI spent more time in the top of the tank, and had decreased risk assessment, behavioral phenotypes related to a decreased anxiety but also increased risk-taking behavior.Interestingly, we also found a decrease in the levels of cortisol, effect that could be correlated with the reduction of anxiety-like behaviors observed herein [44,45].The exposure of animals to a novel environment can be enough to increase anxiety and stress-related measurements [30] in sh that are living with their conspeci cs.However, changes in housing conditions such as social isolation can increase adaptability to the environment [46].For example, studies demonstrate that sh housed individually for 2 weeks [45,47] or 3 months [43] are less anxious and show lower baseline cortisol levels than their group-housed counterparts.Here, we hypothesized that zebra sh adapts to the 12-days SI and this fact reduced the levels of stress and anxiety.Decreased anxiety is not necessarily bene cial for zebra sh in nature, because lower stress reactivity and anxiety, may indicate altered perception of danger and/or risk situations increasing the chances of animals to be preyed [47,46,17].
Like humans, zebra sh have a predominantly diurnal activity, establish groups, form social hierarchies, and display complex social interactions [16,17].In our study, we observed that animals subjected to SI showed changes in different behavioral parameters associated with social behavior.While isolated sh submitted to the conspeci c preference test (PCT -only visual cues) preferred to remain further away from conspeci cs, sh challenged to shoal test (ST -visual and olfactory cues) preferred swim closer to each other.Interestingly, in the mirror-induced aggression test (MIA test), isolated animals increased aggressiveness, a non-a liative social interaction.In nature, zebra sh social preference and hierarchy depend on sh size, stripe pattern, behavior phenotype/trait depending highly on olfactory stimuli [48].However, when facing a single individual, zebra sh may show aggressive traits especially by dominant individuals in an attempt to dispute for territory and resources [49].The increased aggression of SI could indicate that these sh become more territorial and it could also be related with sh decreased assessment of risk, therefore exposing itself more in potential negative interactions.When looking at PCT, the lack of interactions could be triggered as a defense mechanism once animals that remained 12 days without visual and olfactory contact may have impaired their social recognition memory.This type of memory is a critical ability for territorial defense, establishment of dominance hierarchies, pair bonding, and mating in zebra sh [50].Although these data can go against the fact that an increase shoaling was observed, it's important to highlight that in the PCT test, animals do not share water with their conspeci cs and could therefore limit their social recognition memory.Another hypothesis underlying increase shoaling could be associated to the animals social craving, which reports that even a brief acute period of SI in rodents induces an increased motivation for subsequent social interaction [51].This increased sociability depends on midbrain DA neurons [52], similar to other kinds of craving (e. g. hunger, sleep, drug addiction).Taken together, these disruptions in animals' social behavior after SI can directly affect their well-being and survival in the wild.
Evidence also shows that social behavior disruptions differently alter the brain expression of genes involved in neuronal activity, synaptic plasticity, and neurogenesis [8,53].A remarkable decrease in brain gene expression of immediate early genes c-fos and egr, and neurotrophin bdnf was found in zebra sh brain after SI protocol.In rodents, studies demonstrate that SI stress also down-regulated brain c-fos, egr, and bdnf gene transcription, and it has been implicated in SI-induced behavioral abnormalities [54][55][56].Immediate early genes have been shown to be also involved in the development, regulation, and integrity of different neurotransmitter systems, such as dopaminergic and serotoninergic pathways [57,58].Although different protocols show con icting results about the effects of SI on these NT pathways, a decrease in serotonin (5-HT), dopamine (DA) and their metabolites levels in the brain of zebra sh exposed to different SI protocols was previously observed [59,29,60].Therefore, it is evident that the neurobehavioral effects of SI on DA and 5-HT pathways vary greatly depending on the protocol used, animal model used, age of the animals, isolation conditions, among other environmental conditions [61,62,10].In our study, the increased transcription of th and thp genes involved in dopamine and serotonin metabolization, respectively, suggest that SI can alter DA and 5-HT homeostasis.Based on this, SI could directly increase the levels of these NT in the brain, or this result re ects a compensatory effect exerted by the metabolizing enzymes in response to DA and 5-HT altered levels [8].Further studies looking at DA and 5-HT levels would be necessary to elucidate this hypothesis.Considering the role of DA and 5-HT in behavior, we believe that SI-inducing decreased anxiety, changes in sociability, and increased aggressiveness are, at least in part, correlated with reduction of neuroplasticity-related genes and the modulation of the dopaminergic and serotoninergic systems in the brain of adult zebra sh.
Regarding the behavior and neural harmful effects of SI found here, we considered relevant to assess whether these effects could be reversed after reintroducing these animals into the social environment containing visual and olfactory cues (SR protocol).Surprisingly, in a short period of time (3-6 days), the SR was able to driven social behaviors to baseline levels, reversing decreased expression of neural markers genes, and the altered expression of 5-HT-and DA-related genes.Evidence suggest that close conspeci cs social interactions can provide social support and buffer the stress response improving the health state of the individual [63-65].Mammals' studies have shown that the presence of conspeci cs after a stressful experience helps the animals to recover behavior and physiological homeostasis more quickly and reduce the long-term effects of negative social interactions [66-68].Similarly, zebra sh respond to perceived social support of conspeci cs from a familiar shoal and recover from stressful events better in the presence of conspeci cs [28].Social buffering is modulate by oxytocin hormone [69], and can reduces stress-induced c-fos expression and the potentiation of glutamate synapses [70][71][72].Based on presented data, we suggest that through chemosensory cues the SR protocol acted such as a "stress buffer".SR protocol acted at molecular level and recovered, at least in part, the neuronal plasticity and activity (observed by c-fos, egr, and bdnf restored expression), and DA and 5-HT systems homeostasis (observed by th, and thp restored expression).Thus, SR effects recovering neural homeostasis could be crucial for restoring social and cognitive behavioral abilities of zebra sh impaired after SI and one of the mechanisms underlying animals' recovery.
In conclusion, we observed that SR of zebra sh to their conspeci cs played a positive role in social behaviors and in the transcription of genes involved in different neuronal pathways that were altered after 12 days of SI.We emphasize that this is the rst study carried out on zebra sh that brings results of the effects at the behavioral and molecular level of a social isolation protocol followed by the reintroduction of the animals to a social environment.Considering the conditions recently experienced by the population due social restrictions imposed during the COVID-19 pandemic, these data highlight that resocialization of individuals after SI can have a positive impact on brain and reinforces the role of zebra sh as a translational model for understanding how reintroduction to a conspeci c group act as a positive buffer.Although further studies are needed to verify the involvement of other biochemical and neurochemical pathways involved in responses to social isolation, we suggest that this SI protocol has potential to be used in the screening of pharmacological or non-pharmacological interventions that may play neuroprotective roles in this context.

3. 2
Figure 2A demonstrates the effects of SI and SR in the MIA test.An ANOVA effect for number of aggressive episodes was observed (F (3,40) = 4.909, p* = 0.0054), where SI animals had a higher number of aggressive episodes compared to CT (p** = 0.0079), SR3 (p* = 0.0492) and SR6 (p* = 0.0162).Interestingly, no signi cant effect was observed among the CT, SR3 (p = 0.8890) and SR6 (p = 0.9930) groups in this parameter.No signi cant time spent displaying aggressive episodes (F (3,40) = 0.036, p = 0.9903) nor latency to the rst aggressive episodes (F (3,40) = 2.812, p = 0.0515) was changed by SI and SR.Number of entries, time spent and latency to rst enter the area closest to the mirror (A1) were also not affected by SI and SR (data not shown).

Figures Figure 1
Figures

Figure 3 Effects
Figure 3