Social stress-induced serotonin dysfunction activates spexin in male Nile tilapia (Oreochromis Niloticus)

Significance

Spexin (SPX) is a novel neuropeptide involved in multiple biological roles. We recently reported that SPX1a gene expression is up-regulated in the midbrain area during social defeat stress in male Nile tilapia. We identified a strong relationship between serotonin and SPX1a gene expression in vitro and in vivo study during social stress. Our findings suggest that the serotoninergic system is an upstream regulator of SPX during chronic social defeat that reacts to stress-related physiological functions. The novel findings in this study unveil a potential 5-HT-SPX-related antidepressant pathway that will be important to understand the role of SPX in social stress-associated anxiety and psychiatric/behavioral disorders.

Abstract

Social disturbance in interpersonal relationships is the primary source of stress in humans. Spexin (SPX, SPX1a in cichlid), an evolutionarily conserved neuropeptide with diverse physiological functions, is up-regulated in the brain during chronic social defeat stress in teleost. On the other hand, repeated exposure to social stress can lead to dysregulation of the monoaminergic system and increase the vulnerability of developing depression. Since dysfunction of the serotonin (5-hydroxytryptamine, 5-HT) system is associated with social stress and the pathophysiology of depression, the present study investigated the regulatory relationship between the central 5-HT system and SPX1a in the male teleost, Nile tilapia (Oreochromis niloticus). To identify stress factors that regulate SPX1a gene expression, cortisol, dexamethasone (DEX), and 5-HT were used to treat tilapia brain primary cultures. Our study shows cortisol and DEX treatment had no effect on SPX1a gene expression, but SPX1a gene expression was down-regulated following 5-HT treatment. Anatomical localization showed a close association between 5-HT immunoreactive projections and SPX1a neurons in the semicircular torus. In addition, 5-HT receptors (5-HT2B) were expressed in SPX1a neurons. SPX1a immunoreactive neurons and SPX1a gene expression were significantly increased in socially defeated tilapia. On the other hand, citalopram (antidepressant, 5-HT antagonist) treatment to socially defeated tilapia normalized SPX1a gene expression to control levels. Taken together, the present study shows that 5-HT is an upstream regulator of SPX1a and that the inhibited 5-HT activates SPX1a during social defeat.

Epidemiology studies have shown that the primary source of stressors arises from a social disturbance in daily life events, termed social stress (1). Repeated exposure to social stress, such as conflict in a relationship or psychological abuse, can lead to increased vulnerability to developing major psychiatric disorders (2). Social stress exposure is known to affect key components of the hypothalamic-pituitary-adrenal axis (3), the plasma cortisol levels in defeated animals (4), and corticotrophin releasing hormone (CRH) gene expression in rodents (5, 6) and fish (7, 8).

Serotonin (5-hydroxytryptamine, 5-HT) is an evolutionarily conserved neurotransmitter across vertebrate species and is involved in the modulation of multiple social behaviors (9). The dysregulation of the monoaminergic 5-HT system during social stress is well documented. Numerous studies have shown elevated serotonergic activity in the brain of socially defeated animals (10–12) and in subordinate fish (13). A few lines of evidence suggest spexin (SPX) interactions with the 5-HT system in the neurophysiology of stress (14, 15).

SPX, also known as neuropeptide Q, is widely expressed in various brain regions and peripheral tissues in mammalian and non-mammalian vertebrates (16). The mature peptide sequence of SPX consists of 14 amino acids, which are highly conserved throughout the vertebrate species. A newly discovered orthologous SPX in non-mammalian vertebrates is phylogenetically homologous to SPX1 than SPX2. Hence, it is termed SPX1b, and the original SPX1 is termed SPX1a (17). SPX is associated with multiple physiological functions such as reproduction, metabolic homeostasis, and obesity (16, 18). Further, SPX has been implicated in stress. Mice exposed to chronic unpredictable mild stress can influence SPX levels (19). Chronically defeated fish in a social hierarchy have elevated plasma cortisol levels; they show increased levels of SPX1a in the midbrain area (20). However, the regulatory mechanism underlining the upregulation of SPX during social stress remains poorly understood.

The Nile tilapia (Oreochromis niloticus) is an ideal model to examine defeat stress because they exhibit social hierarchy, remarkable plasticity in behavior, and physiological function, which are influenced by their social environment (4, 21, 22). Our previous work with tilapia showed dysregulation of the serotonergic 5-HT system and SPX1a during social defeat stress (4, 20). The present study was designed to examine interactions between the 5-HT system and SPX1a during chronic social defeat stress. Therefore, in this study, we first examined the impact of cortisol, dexamethasone (DEX), and 5-HT treatment on SPX1a gene expression using primary cultures of the tilapia brain. The association between 5-HT fibers, 5-HT receptors, and SPX1a neurons was examined using immunohistochemistry. The percentage of SPX1a neurons co-expressing 5-HT2B receptors was compared between control and socially defeated tilapia. To evaluate changes in 5-HT during social stress, we examined the expression of 5-HT-related genes and the levels of 5-HT and its metabolite,5-hydroxyindoleacetic acid (5-HIAA). Finally, using chronically defeated tilapia, we examined the impact of selective 5-HT reuptake inhibitor (citalopram) treatment on SPX1a messenger ribonucleic acid (mRNA) expression levels and SPX1a expression during recovery from chronic social defeat stress.

Results

Effect of DEX, Cortisol, and 5-HT Treatment on SPX1a Gene Expression in Primary Cultures of the Brain of Tilapia.

Gene expression analysis of SPX1a was carried out in the primary culture from Area 2 (midbrain and hypothalamus) and 3 (cerebellum and hindbrain), where SPX1a neurons are localized in the tilapia brain (Fig. 1A). The presence of SPX1a neurons in the primary culture was further confirmed by immunostaining with the SPX1a antibody (Fig. 1B). Our data showed that DEX and cortisol treatment did not significantly impact SPX1a gene expression (Fig. 1 C and D). In contrast, 100μM of 5-HT treatment significantly down-regulated SPX1a gene expression, and the downregulation was further enhanced with 10μM of 5-HT treatment for 24 h (Fig. 1E). To further examine the involvement of 5-HT receptors, non-selective 5-HTR1 antagonist, metergoline, or 5-HTR2/3 antagonist, mirtazapine at 1µM was added prior to 5-HT treatment (10μM for 24 h). Our results indicated that both antagonists blocked the effect of 5-HT treatment (Fig. 1 F–H). However, the treatment with only the antagonist did not significantly impact SPX1a gene expression (Fig. 1 G–I).

Fig. 1.

Primary culture and gene expression study of the tilapia brain. (A) Macro-dissection of the tilapia brain into three areas. Area 1 (POA, preoptic area; TEL, telencephalon); Area 2 (OT, optic tectum; MB, midbrain; HYP, hypothalamus); Area 3 (C, cerebellum; HB, hindbrain); OB, olfactory bulb; PIT, pituitary. (B) i) Photomicrograph of the brain primary culture experiment. (Scale bars, 100 µ m.) ii). Immunocytochemistry staining of SPX1a neurons further confirms the presence of the SPX1a neurons in the brain primary culture. (Scale bars, 20 µm.) (C and D) DEX (10 nM and 100 nM for 24 h) and cortisol treatment (100 nM, and 1,000 nM for 24 h) do not impact the gene expression of SPX1a. (E) SPX1a gene expression was significantly down-regulated following treatment with 5-HT (10 µM and 100 µM for 24 h). (F) and (H) Pretreatment with 5-HT antagonist mirtazapine (1 µM) and metergoline (1 µM) 30 min prior to 5-HT treatment (10 µM for 24 h) blocked the action of 5-HT upon SPX1a. (G) and (I) 5-HT antagonist alone does not have any significant impact on the SPX1a gene expression.

Association between Serotonergic System and SPX1a Neurons.

The tilapia SPX1a antibody specificity was determined by pre-absorption of the SPX1a antibody with antigen peptide, which completely blocked the signal of SPX1a (Fig. 2 A–C). To further confirm the specificity of the SPX1a antibody, double-label immunofluorescence (SPX1a antibody) and in situ hybridization (ISH; SPX1a mRNA) were performed. Our results confirmed the antibody's specificity, as shown by the co-expression of the SPX1a antibody and SPX1a ISH (Fig. 2 D–F). To examine the antibody characterization, a western blot experiment was carried out. The western blot analysis showed a single band for SPX1a antibody with an expected band size (~13 KDa) which is preproprotein of tilapia SPX1a, in the extracted protein from the tilapia brain, along with a weak band with a size around 15KDa, which might be minor variant type of SPX1a protein.

Fig. 2.

Localization of SPX1a neuron and interaction with the 5-HT system in the brain of Nile tilapia. (A) Photomicrographs of the coronal section showing the SPX1a-expressing cells in the the semicircular torus (TS). (B) Pre-absorption with the antigen peptide (10 µg/mL) completely blocked the SPX1a immunoreactivity. (C) Buffer control. (D–F) SPX1a-Ab cells co-express with SPX1a mRNA in the TS, verifying the specificity of the SPX1a antibody. (G) Western blot analysis of i) SPX1a showing an expected band size (~13 KDa) which is preproprotein of tilapia SPX1a in the extracted protein from the brain, along with a weak band with a size around 15 KDa, which most likely is the precursor protein, and ii) 5-HT2B antibody in tilapia brain homogenate. Lane 1 and 2 are protein samples extracted by NP-40, while Lane 3 is the protein extracted using Ripa buffer. (H–J) Double-labeled immunofluorescence showed close association between 5-HT fibers (red) and SPX1a neurons (fibers and cell bodies in green) in the TS. Fibers to fibers interaction between 5-HT and SPX1a was also observed (indicated by a white arrow). 5-HT fibers can be seen in close association with SPX cell bodies as well as with SPX fibers. (K) The close association between 5-HT fibers and SPX1a cell soma was further confirmed by 3D analysis of the confocal image. (L–N) Double staining with SPX1a and 5-HT2B antibody revealed that 5-HT2B-like immunoreactivity is expressed in the nucleoplasm of SPX1a neurons (indicated by white arrow). (O) The co-expression was further confirmed by confocal analysis using 3D reconstruction. Scale bars, 20 µm in (A–F, H–J), and (L–N); 10 µm in (K) and (O).

(Fig. 2Gi). The tilapia SPX1a antibody binds to the 14 amino acids in the C-terminal of the preproprotein of tilapia SPX1a. Thus, the 13 KDa band detected by the antibody in the western blot is the full size SPX1a protein. The tilapia full size preproprotein is 120 amino acids, which includes the mature 14 amino acids SPX1a protein. This gives the 13 KDa band in the western blot (NCBI Reference Sequence: XP_005475162.1; tilapia SPX1a).

For the 5-HT2B antibody, a single band with an expected band size (lane 3, ~55 KDa) along with two weak bands beneath it (<50 KDa; 50 to 55 KDa) was observed (Fig. 2Gii). Lane 1 and 2 are protein samples extracted by NP-40, while Lane 3 is the protein extracted using Ripa buffer. As NP-40 typically extracted proteins from the whole cell, and Ripa buffer extracted proteins from the nucleus, this indicates that the majority of 5-HT2B proteins are expressed in the cellular nucleus. The two weak bands beneath could potentially be due to several factors, such as protein degradation, or protein that share some percentage of similarity between mammalian 5-HT2B and tilapia 5-HT2B, for example, tilapia 5-HT2A (~54 KDa), which share about 70% homology with mammalian 5-HT2B. The double-labeled immunofluorescence showed close association between SPX1a neurons and fibers (green) in the semicircular torus (TS) (Fig. 2H) and 5-HT fibers (Red) (Fig. 2I). We also observed the close association between 5-HT fibers and SPX1a fibers (Fig. 2J). The close association between 5-HT fibers and SPX1a cell soma was further confirmed by 3D analysis of the confocal image (Fig. 2K). The 5-HT fibers are from cell bodies in the Raphe and predominantly from periventricular pretectal nuclei (PPV). 5-HT fibers can be seen in close association with SPX cell bodies as well as with SPX fibers.

Double-labeled immunofluorescence revealed that SPX1a neurons expressed 5-HT2B-like receptors in the cell soma (Fig. 2 L–O). Further 3D reconstruction analysis by confocal microscope confirmed that 5-HT2B-like receptors were expressed in the cell soma of SPX1a neuron (Fig. 2O).

Changes of the 5-HT System in the Brain during Chronic Social Defeat.

To investigate the changes in 5-HT during the chronic social defeat, gene expression analysis of tryptophan hydroxylase 2 (TPH2), serotonin transporter (SERT), and monoamine oxidase A (MAO-A) was carried out. No significant differences were observed in TPH2 and MAO-A gene expression under chronic social defeat (Fig. 3 A and B). However, a statistically significant upregulation of SERT was observed in Area 3 (cerebellum and hindbrain) (Fig. 3C).

Fig. 3.

5-HT system changes and citalopram treatment during the chronic social defeat. (A and B) No significant changes were observed in the gene expression of TPH2 and MAO-A. (C) There was a statistically significant upregulation of SERT in Area 3 of the tilapia brain after chronic social defeat. (*P < 0.05). (D) Changes in the 5-HT, 5-HIAA, and 5-HT turnover rate during the chronic social defeat. (E) Design of chronic stress and citalopram treatment experiment. Four experimental groups were utilized: CONV, CONCIT, CSDV, and CSDCIT. N=7/group. (F) Citalopram treatment does not impact the gene expression of SPX1a in the control group (CONV and CONCIT). SPX1a gene expression was up-regulated by chronic stress (CSDV). This upregulation of SPX1a was not observed in the citalopram-treated group (CSDCIT). (*P < 0.05)

To get a detailed picture of the changes in the 5-HT, the level of 5-HT and its metabolites, 5-HIAA in Area 2 and 3, were examined by LC-MS. The total 5-HT level was not changed during the chronic social defeat, while 5-HIAA was significantly increased by chronic social defeat stress exposure. The 5-HIAA/5-HT ratio was significantly increased with a corresponding increase in the 5-HIAA level, which signals an increase in 5-HT turnover during chronic social defeat (Fig. 3D).

Effect of Citalopram Treatment on SPX1a Gene Expression during Chronic Social Defeat.

Compared to our previous observation, in the present study, SPX1a gene expression remains up-regulated by chronic stress even after 5 d of recovery. This upregulation of SPX1a was not observed in the group exposed to similar chronic stress but treated with citalopram, suggesting that citalopram normalized the upregulation of SPX1a after chronic stress. In the control group without stress exposure, citalopram treatment does not impact the gene expression of SPX1a (Fig. 3F).

SPX1a Neurons Co-expressing 5-HT2B Receptors during Chronic Social Defeat.

Analysis and counting of the SPX1a-expressing neurons between control (Fig. 4A) and defeated (Fig. 4B) tilapia showed a significant increase in the number of SPX1a-expressing neurons (~55%) in the defeated tilapia compared to the controls (Fig. 4C). On the other hand, the percentage of SPX1a neurons co-expressing 5-HT2B receptors did not differ between the two groups. (Fig. 4D).

Fig. 4.

Numbers of SPX1a-expressing neurons and co-localization with 5-HT2B during the chronic social defeat stress. (A and B) Immunostaining of SPX1a and 5-HT2B during chronic social defeat. A white arrow indicated co-localization between SPX1a and 5-HT2B. (C) There was a significant increase in SPX1a-expressing neurons in the defeated tilapia compared to the control group. (D) The percentage of co-localization between control and defeat was not significantly different. (***P < 0.001) (Scale bar, 20 μm.)

Discussion

Our current study provides compelling evidence regarding the regulatory mechanism of SPX1a neurons during chronic social defeat stress. DEX and cortisol treatment had no effect on SPX1a gene expression, despite both doses were within the physiological range (20, 23). Our previous work demonstrated that SPX1a gene expression was up-regulated in the defeated tilapia, along with a parallel increase in plasma cortisol level. This indicates that SPX1a may be regulated via the feedback mechanism of cortisol. However, the lack of change of SPX1a gene expression under cortisol treatment observed in the current study suggests cortisol's delayed action to exert its impact (beyond 24 h), which was not examined in the current study. Alternatively, the action of cortisol may involve an intermediate factor, which was absent in the primary culture experiment, hence the lack of change. A possible regulatory relationship between CRH on SPX1 gene expression (24) and glucocorticoid receptor(GR) and CRH expression in the ventromedial nucleus of the TS (23, 25) has been suggested, but the presence of GR and CRH receptors on SPX1a neurons has not been determined. On the other hand, our previous work with acute social defeat stress showed CRH gene expression unchanged in the brain of defeated tilapia (26). Therefore, based on our findings, coupled with previous works, it is reasonable to assume that cortisol might not regulate SPX1a expression directly at the level of SPX1a neurons.

It is possible that cortisol's action could be mediated by the 5-HT system (5), which may explain cortisol's possible delayed action. And CRH can exert its effect on 5-HT neuronal activity via CRH receptors expressed on the neurons in the dorsal raphe (27–29). DEX treatment increases SERT gene expression (30) and SERT localization on 5-HT neurons (31). During chronic social defeat, cortisol can up-regulate SERT expression via GR (32). Taken together, these studies show that 5-HT could potentially be the mediatory factor linking cortisol action to SPX1a expression.

Our confocal images reveal close proximity between 5-HT projections and cell soma of SPX1a neurons in the TS. These results suggest a potential direct impact of serotonergic modulation on SPX1a neurons. In teleosts, the 5-HT neuronal population in the raphe nuclei send projections to various brain regions except to the optic tectum (33), and other 5-HT neuronal population in the PPV send projections to the TS (34). Therefore, the 5-HT fibers seen in close association with SPX1a neurons possibly originate from the 5-HT neuronal population in the PPV, which is a unique feature of the fish 5-HT system (35, 36). The abundance of 5-HT fibers in close association with SPX fibers suggests that 5-HT regulates SPX predominately at the level of SPX fibers and to a lesser at the level of SPX cell bodies. Here, we also speculate that the absence of 5-HT receptor expression on the cell membrane but their expression in the nucleus of SPX neurons suggests that 5-HT receptors are synthesized to be transported to the fibers under certain, yet to be identified, physiological state. The downregulation of SPX1a gene expression observed after 5-HT treatment could be an inhibitory action of 5-HT mediated via 5-HTRs coupled to Gi/Go protein. The modulatory effect of 5-HT upon SPX1a was blocked by non-selective 5-HTR1 antagonist metergoline and 5-HTR2/3 antagonist mirtazapine, which suggest that the regulatory action of 5-HT on SPX1a may be mediated via 5-HTR1 and 5-HTR2/3. However, further examination is needed to confirm the specific receptors involved.

Our novel finding of 5-HT2B-like immunoreactivity in the nucleus contradicts the expected expression in the cell membrane; however, nuclear localization of G protein coupled receptors is known (37). In fact, the presence of 5-HT2B and other 5-HT receptors (5-HT1A and 5-HT4) has been reported in the nucleus (37, 38). The presence of 5-HT2B in the cell body's cytoplasm is not an antibody cross-reaction with some unknown epitope since the human 5-HTR antibody used in this experiment has high sequence homology to tilapia 5-HTR. Conversely, the 5-HT2B nuclear receptors could represent a pool of inactive/dormant receptors in association with the endoplasmic reticulum (38), which are activated under certain yet unidentified milieu/states of SPX1a neurons. Although the molecular mechanisms of activation of nuclear 5-HT2B remain unknown, the presence of 5-HT2B as a nuclear receptor in the SPX1a neurons suggests that it may play a pivotal role in regulating the intracellular signaling mechanism of SPX1a neurons. Of note, when interpreting the results of the antagonist treatment experiment, one must be cautioned that the antagonists used are non-selective, broad action antagonists that do not provide a clear picture of the actual effect of individual 5-HT receptors (whether inhibitory or stimulatory) on SPX1a gene expression. There is still a gap to fill in to understand the mechanisms and pathway of serotonergic regulation of SPX1a expression. Pretreatment of antagonist or blocking of 5-HT receptors might be required to confirm the direct action of 5-HT on upregulation of SPX neurons in the brain. Although 5-HT2B receptors might be involved in the inhibition of SPX1b, the precise effect of other 5-HT receptors on SPX1a expression must be taken into consideration.

During chronic social defeat, a statistically significant upregulation of SERT was observed in Area 3, where major 5-HT neuronal populations are located. Our results are in agreement with previous studies where repeated social defeat increased SERT expression in the dorsal raphe nucleus in rats (32, 39). The upregulation of SERT gene expression during chronic social defeat could lead to increased reuptake of 5-HT from the synaptic cleft and decrease the extracellular 5-HT availability (40), which reduces the inhibitory signal upon SPX1a, leading to upregulation of SPX1a gene expression. Although the MAO-A gene expression did not change under chronic social stress, the elevated reuptake of 5-HT into the intracellular space could have increased the bioavailability of 5-HT for the catalytic conversion by MAO-A. Indeed, a high turnover rate of 5-HT into its metabolites, 5-HIAA, was observed. However, the increase in the 5-HT turnover rate alone may not be sufficient to account for the changes in SPX1a; instead, changes in the released levels of 5-HT and 5-HT receptors in SPX neurons must be studied.

Citalopram treatment normalized the upregulation of SPX1a in defeated but not in control fish. In this case, citalopram blocked SERT, which subsequently increased the extracellular 5-HT to restore SPX1a gene expression to normal. Therefore, the effect of 5-HT upon SPX1a gene expression is dependent on the balance of 5-HT receptors. The lack of change in SPX1a gene expression in the citalopram-treated control group could be due to the unchanged 5-HT. Furthermore, assuming that SPX1a containing neurons co-express several 5-HT receptors which are coupled with different G protein signaling; under normal condition, the 5-HT could simultaneously activate 5-HTR, which are coupled to Gi protein (such as 5-HTR1) or 5-HTR, which are coupled to Gq protein (such as 5-HT2B), transmitting both excitatory and inhibitory signals leading to no change in SPX1a gene expression. Such biphasic action of 5-HT has been previously reported on neuropeptides such as GnRH neurons (41).

A significant increase in the number of SPX1a-expressing neurons was observed in the defeated group; however, the percentage of neurons co-expressing 5-HT2B did not differ between the control and the defeated group. We speculate that the inhibitory signal was reduced, while the excitatory signal was maintained or increased during the defeat, which led to the upregulation of SPX1a gene expression and an unchanged percentage of SPX1a neurons co-expressing 5-HT2B. On the other hand, the unchanged number of co-localizations may also indicate a potential reduction in other inhibitory 5-HTR expressions (such as 5-HTR1), which was not examined in the current study. Given that 5-HT inhibits SPX1a mRNA expression in brain cell cultures, the rise in 5-HT turnover induced by social stress is expected to reduce instead of elevating SPX1a expression in the brain. However, the action of 5-HT on SPX1a is biphasic, which probably depends on the balance of the signals mediated via the 5-HT receptors. Even though we did not examine gene expression changes of 5-HT receptors during chronic social stress, the transcription factor of 5-HT receptors (REST) is activated during chronic social stress (42), which could influence the basal expression level of 5-HT receptors as well as their physiological functions.

During chronic social defeat, the increased turnover rate of 5-HT activates SPX1a neurons, which justifies the association of 5-HT in the suppression of aggression (43). The TS area is part of the mechanosensory lateral line system that is responsible for the integration of visual-sensory inputs and relay them to the social decision-making network by sending projections to various brain regions, including the anterior tuberal nucleus (ATn) (44). Increased SPX1a neuronal activity after chorionic defeat may then relay signals to downstream target sites (such as the ATn) and dispose the animal to behavioral inhibition, together with reduced locomotor activity and aggressive behavior and an increase in the perception of stimuli in order to escape from danger (45). Since behavioral inhibition was observed after chronic activation of the 5-HT system, and the fact that SPX1a gene expression remains elevated following 5 d of recovery, suggests that SPX1a is involved in chronic rather than acute behavioral inhibition. Some studies have suggested that SPX in the brain has been implicated in stress response (20, 24) and emotional control via SPX receptors, galanin receptor type2 and type3 (15, 46, 47). Treatment of SPX-based galanin receptor type2 agonist decrease anxiety-like and depressive-like symptoms in mice (47, 48) and in zebrafish (15). Further study to identify SPX’s action targets, such as galanin receptor type2 and type3 positive neurons downstream, is needed for a deep understanding of the possible function of SPX on stress-related physiological functions.

In conclusion, our study provides evidence of the close association between 5-HT fibers and 5-HT2B receptors co-expressed in SPX1a neurons, which shows that 5-HT is an upstream regulator of SPX1a. The downregulation of SPX1a gene expression observed after 5-HT treatment could be an inhibitory action of 5-HT mediated via 5-HTR1 and 5-HTR2/3 in vitro study. In the in vivo study, the upregulation of SERT gene (5-HT transporter) expression during chronic social defeat could lead to increased reuptake of 5-HT from the synaptic cleft and decrease the extracellular 5-HT availability which reduces the inhibitory signal upon SPX1a, leading to upregulation of SPX1a gene expression. In addition, citalopram (5-HT reuptake inhibitor) treatment normalized the upregulation of SPX1a in defeated fish but not in control fish. In this case, citalopram blocked SERT, which subsequently increased the extracellular 5-HT to restore SPX1a gene expression to control levels. Our finding raises the possibility of a reciprocal relationship between SPX1a and 5-HT in the tilapia brain and thus provides novel insights into the inhibitory role of 5-HT in the regulation of SPX1a neurons.

Material and Methods

Animals.

Sexually mature male Nile tilapia (4 to 8 mo old) were kept in tanks (size: 90 × 45 × 45 cm) at 28 ± 0.5 °C. The fish were exposed to 14hL:10hD photoperiod and fed three times per day with commercial fish pellets (Star Feedmills (M) Sdn Bhd). All experimental procedures were approved and carried out in accordance with guidelines given by the Animal Ethics Committee of Monash University (MUM/2018/02).

Gene Expression Analysis.

The procedure for gene expression study has been described previously (20). Briefly, the brain tissue was harvested and macro-dissected into three areas (Area 1: telencephalon, preoptic area; Area 2: optic tectum, midbrain, hypothalamus; Area 3: cerebellum, and hindbrain). Total RNA was isolated from each area, and 1,000 ng RNA was synthesized into cDNA. The quantitative real-time PCR reaction was performed using the SensiFAST SYBR Hi-ROX Kit (Bioline). The cycling condition was holding stage (95°C) for 2 min and 40 cycles at 95°C for 5 s and 63°C for 30 s. A standard curve was generated by the plasmid containing the targeted gene at a concentration of 10⁹ to 10³ to calculate the absolute copy number of mRNA and normalized to the housekeeping gene. The tilapia primer sequences were as follows: SPX1a: forward primer, 5′-AAGGGCTCATTCCAGCGAAG-3′ and reverse primer, 5′-CGAGTCTCTAGGTGAAGAGTGTC (MN_399812); TPH2: forward primer, 5′-ATGATGATGTTCTCCAGCAAGT-3′ and reverse primer: 5′-CTCATTGATGGGACAGAAGGA -3′ (XM_005473325.3); SERT: forward primer, 5′-GAAAGCTGCTGGAGACCAGT-3′ and reverse primer: 5′-GAAAGAGCCAAGAGGACCCC-3′ (XM_005473325.3). MAO-A: forward primer, 5′– AGCTGCTGGAGGGCTACTAA-3′ and reverse primer: 5′-TTCCATCCAACACGGTCTGC-3′ (XM_005473325.3). Elongation Factor 1-Alpha gene was chosen as the housekeeping gene (forward primer: 5′ -GGACTGGCTTATGCTGATT-3′; reverse primer: 5′-ACTGAGAAGAGGCACTGT-3′; NM_001279647.1).

Cortisol, DEX, and 5-HT Treatment on Primary Cell Cultures of the Tilapia Brain.

The brain tissues (n=42, 6 per treatment group) were dissected into three brain regions: Area 1, Area 2, and Area 3. Macro-dissected brain samples (Area 2 and Area 3) were washed into D-PBS (Nacalai Tesque) and treated by Trypsin-EDTA solutions (Gibco). The pellet was resuspended in Leibovitz L-15 medium (Gibco) supplemented with 10% Fetal Bovine Serum (FBS) and 0.1% Penicillin/Streptomycin antibiotics (Gibco) and cultured to a pre-coated 12-well plate (BD, New Jersey, USA). After 24 h of incubation, cortisol (100 nM and 1,000 nM, Nacalai Tesque), DEX phosphate (10 nM, and 100 nM, Sigma), and 5-HT (10 μM, and 100 μM, Sigma) were supplemented for 24 h. 5-HT receptor antagonist Mirtazapine (5-HT receptor 2 and 3 antagonists) (Tocris Bioscience) and Metergoline (5-HT receptor 1 antagonist) (Sigma) at 1 µM were added 30 min before the 5-HT treatment (10 μM) for 24 h. The treatment dose and duration were referred to recent studies (49–51). The cell viability was determined using the Muse Count & Viability kit (Luminex). The average cell viability was 80.5%.

Generation of Tilapia SPX1a Antibody and the Specificity Characterization.

The polyclonal antiserum for tilapia SPX1a was raised against a synthetic peptide QGRRFISEDRKEGD in the C-terminal of preproprotein tilapia SPX1a; an expected band size of the preproprotein tilapia SPX1a is ~13 KDa in western blot. (GenBank accession number QGP74752.1; Antibody ID: PAS 23095 Apical Scientific Sdn Bhd, Malaysia). The specificity of the tilapia SPX1a antibody was determined by the pre-absorption test with 10 µg/mL of the antigens and double labeling of immunofluorescence and ISH (SPX1a mRNA). Furthermore, western blot analysis was conducted with tilapia brain protein extract.

Combination of ISH and Immunofluorescence.

The brain tissue (n=12) was fixed in 4% paraformaldehyde/PBS, cryoprotected in 20% sucrose/PBS, and then embedded in the OCT compound (Leica). Coronal sections were cut at 15 μm using a cryostat machine (Leica CM 1860). Sections were permeabilized in 0.2M HCL and then digested with proteinase K (1 μg/mL). Hybridization was then performed with a DIG-labeled SPX1a RNA probe (500 ng/mL) (20). After washing in saline sodium citrate and blocking with 2% normal goat serum (NGS), the section was incubated with peroxidase-conjugated anti-DIG antibody (1:500, Roche Diagnostics, Germany) and developed with the TSA Cyanine 3 system (PerkinElmer). To proceed with double labeling, TSA-labeled sections were incubated with tilapia SPX1a antibody (1:1,000) and visualized with Alexa Fluor 594 anti-rabbit IgG (1:400; RRID: AB_2534079, Thermo Scientific).

Double Immunofluorescence and Image Analysis.

After immunohistochemistry with anti-tilapia SPX1a (1:1,000), the tissue sections were incubated with primary monoclonal mouse anti-5-HT (1:400, Abcam, ab16007) or primary monoclonal mouse anti-5-HT2B (1:400, Santa Cruz Biotechnology). The specificity of the 5-HT antibody in teleosts has been determined by previous studies (52, 53), while the specificity of the 5-HT2B antibody was determined by western blot analysis using tilapia brain protein extract, following the same protocol (54). The signal was developed with Alexa Fluor 594 goat anti-mouse IgG (1:400, Thermo Scientific). After staining, the sections were viewed under a confocal laser microscope (C1s; Nikon) with confocal imaging software (NIS elements AR, version 4; Nikon).

The 3D construction and visualization of close associations between SPX1a and 5-HT fibers were analyzed with a 60-× water immersion objective lens at 0.225 µm Z-steps, at the laser wavelength of 488 and 543 nm in 8 to 10 sections to include all the SPX1a neurons. The counting of all SPX1a neurons with a distinct nucleus in the TS and the percentage of co-localization with 5-HT2B between the control and socially defeated tilapia (6/group; approx. 40 sections/brain sample) was performed at 20× objective lens magnification.

Experimental Procedure for Social Defeat Stress.

The procedure for chronic social defeat experiments and subsequent sample preparation have been described previously (20). Briefly, the dominant fish (length: 15 to 1 cm) was transferred to the experimental tank to establish dominance territory. Next, the subject fish (length: 11 to 12 cm) was introduced for interaction with the dominant fish. A partition was inserted into the tank once the number of attacks from the dominant fish hit 200. The experiment ended 2 h later after the initial social interaction, and the subject fish was shifted to another housing tank. The procedure was performed for five consecutive days. The control fish experienced a similar procedure as the subject fish, without the social defeat stress.

Measurement of 5-HT and 5-HIAA.

5-HT and 5-HIAA measurement in male tilapia brain area 2 and 3 (control = 8, defeat = 8) was carried out using liquid chromatography-mass spectrometry (the Agilent Technologies 6410 Triple Quad Liquid chromatography-mass spectrometry equipped with a ZORBAX SB- C18 column). The procedures were as described previously (4). Briefly, the brain tissues were homogenized with 350 μL of 20 ng/mL isoproterenol (internal standard, Nacalai Tesque) in 50% acetonitrile (Fisher Scientific) with 0.1% formic acid (Sigma). The supernatant was filtered by Cosmospin filter G (Nacalai Tesque). A standard solution of 5-HT, 5-HIAA (Sigma), 5-HIAA, and 20 ng/mL isoproterenol was prepared in 50% acetonitrile with 0.1% formic acid in the range of 1.25 to 160 ng/mL. The production of 5-HT, 5-HIAA, and isoproterenol was observed for further analysis (m/z 177.1–160 for 5-HT, m/z 192.07–146 for 5-HIAA, and m/z 212.1–194 for isoproterenol). Data acquisition software Agilent Masshunter Quantitative Analysis software (RRID: SCR_015040, Agilent Technologies) was used for data analysis.

Citalopram Treatment and Chronic Social Defeat.

In this study with citalopram hydrobromide (Sigma), four experimental groups were utilized: control with vehicle (CONV), control with citalopram (CONCIT), chronic social defeat with vehicle (CSDV), and chronic social defeat with citalopram (CSDCIT) (n=7 in each group). CSDCIT and CSDV were exposed to chronic social defeat protocol for 5 d, while CONCIT and CONV acted as the corresponding control group and underwent a similar procedure as the subject fish without the social defeat stress. On day 6 to day 10, CSDCIT and CONCIT were administered with citalopram (10 mg/kg body weight) in 40μL water via intraperitoneal injection. The dose for citalopram was determined based on the effective dose previously used in rodents (55) and SSRI treatment in zebrafish (56). CSDV and CONV were administered with water. Treatment continued for a period of 5 d. On day 10, the brain samples were collected 5 h after the last dose of citalopram (Fig. 3E). SPX1a gene expression analysis by quantitative real-time PCR was performed as described above.

Statistical Analysis.

All data were analyzed using SPSS version 23. One-way ANOVA was used for multiple group comparisons. All data are expressed as the mean ± SEM per group, and P value less than 0.05 was considered statistically significant.

Data, Materials, and Software Availability

All study data are included in the main text.