Photoperiod regulate gonad development via kisspeptin/kissr in hypothalamus and saccus vasculosus of Atlantic salmon (Salmo salar)

Atlantic salmon exhibit seasonal reproduction. However, the mechanisms governing this are still unclear. Generally speaking, kisspeptin has been recognized as a regulator of reproduction. Here, we report a relationship between kisspeptin, GnRH and photoperiod in Atlantic salmon. The results demonstrated that the expression of the Atlantic salmon kisspeptin-receptor (skissr) was not always consistent with the expression pattern of Atlantic salmon GnRH3 (sGnRH3) during all developmental processes. Kisspeptin may exert its influence primarily in the early and later stages of gonad development by promoting the secretion of sGnRH3. Meanwhile, the expression levels of kissr were higher in fish with gonads at stage II and stage V under the long-day photoperiod regime than under the short-day regime. In addition, both skissr and sGnRH3 were also expressed in the saccus vasculosus (SV), an organ only found in fish. The SV might be a seasonal sensor regulating reproduction in addition to the hypothalamus (Hyp).


Introduction
Seasonal reproduction is an important adaptive trait for animals living outside the tropics and photoperiod (day length) is undoubtedly one of the most effective environmental signals available to living organisms, including fish [1,2]. Photoperiod is the only environmental factor which provides a reliable indicator of the time of year and so enables reproduction and growth processes to be expressed at the most appropriate date [3]. It is now widely accepted that seasonally changing photoperiods provide the proximate environmental signal for the initiation and co-ordination of gonad development in most temperate fish [4].
In mammals, the synchronization of reproduction with photoperiod is mediated by melatonin which is secreted by the pineal organ [5,6]. Melatonin is thought to stimulate the production of kisspeptin and type 2 iodothyronine deiodinase in the pars tuberalis (PT) of the pituitary gland [7][8][9][10][11]. Teleost fish, do not possess a distinct PT, and the signal transduction pathway for their reproduction remains unclear [12]. Kisspeptin and its receptor GPR54 (kissr) have been identified as key factors in controlling the reproductive cycle by stimulating gonadotropin releasing hormone (GnRH) secretion in mammals. Loss of functional mutations in mouse and human GPR54 show that GPR54 is absolutely required for reproduction in mammals [13][14][15]. In teleost, the study of kisspeptin and its receptor is still in its infancy. Parhar and co-workers were the first to report the importance of kissr and GnRH in fish reproduction and found that kissr was involved in the sexual development of a cichlid fish [16]. Since then, kisspeptin and kissr have become the subject of active research in fish reproduction. It is believed that kisspeptin and kissr perform similar roles in fish to kisspeptin/GPR54 reported in mammals. Kisspeptin has been reportedly associated with the onset of puberty in some fish species [17][18][19], and has GnRH regulatory functions in several others [20]. In addition, Martinez-Chavez et al. found that a long photoperiod could delay the onset of puberty and suppress GPR54 expression in Nile tilapia by reducing the expression of kisspeptin [17]. This was the first study of teleost to suggest a possible connection between photoperiod and kisspeptin. Similar results have also been found in other fish species [21][22][23]. Furthermore, most research into photoperiods and GnRH have focused on the stage of puberty. Furthermore, there is very limited reference to the effects of kisspeptin/kissr on other stages of gonad development [24,25].
In mammals and birds, the PT of the pituitary gland is considered to play a key role in the regulation of seasonal reproduction [26][27][28]. However, fish differ from mammals and birds in not processing an anatomically distinct PT [12], and researchers have considered the hypothalamus to be the regulatory hub of photoperiodism in fish. Meanwhile, kisspeptin genes are expressed in the nucleus ventralis tuberis (NVT) and the nucleus posteriors periventricularis (NPPv) of the hypothalamus in the medaka [29,30]. This suggests that kisspeptin produced in the hypothalamus mediates seasonal reproduction [22].
A recent study of masu salmon (Oncorhynchus masou masou) found that some important factors involved in seasonal reproduction, such as thyroid-stimulating hormone (TSH), TSHreceptor and thyroid hormone-activating enzymes (DIO), are expressed in the saccus vasculosus (SV) [12]. The SV is an organ unique to fish, which is located posterior to the pituitary gland on the floor of the hypothalamus. This finding provides a new means of understanding the functions of the SV and the regulation of seasonal reproduction.
Atlantic salmon (Salmo salar L), are native to the North Atlantic and its surrounding rivers, and were introduced into Chinese using Recirculating Aquaculture Systems (RAS). Atlantic salmon are short-day seasonal breeders and are very sensitive to the photoperiod [31]. Furthermore, salmonids are characterized by their direct sensitivity to daylight and lack of endogenous rhythms found in other fish [32,33]. In a previous study, we found that photoperiod can significantly affect the gonadal development of Atlantic salmon reared in RAS. In order to enrich our knowledge of the functions of the SV in teleost fish, we set out to clarify the relationship between kissr and GnRH in the Hypothalamus and SV of Atlantic salmon under the different photoperiod regimes.

Experimental design
Atlantic salmon with an average body mass of 1071.70 ± 155.54g were purchased from the Shandong Oriental Ocean Sci-Tech Co., Ltd, Shandong province, China. The fish were then allowed to acclimate for four weeks in a RAS under 24L:0D photoperiod, after which they were distributed between the experimental RAS tanks (130 cm height ×200 cm diameter). Each experimental group contained 60 fish. During the experimental period, the water temperature was maintained at 16.27 ± 0.54˚C, pH 7.2-7.5, total ammonia-nitrogen <0.25mg/ L, and salinity 24-26 over the course of the experiment.
The fish were divided into six photoperiod treatment groups. Four of the photoperiod treatments remained constant throughout the experiment 24L:0D, 18L:6D, 12L:12D, 16L:8D. The remaining two photoperiod groups had varying photoperiods during the experiment. In the first treatment, the photoperiod changed from 24L:0D to 8L:16D (the LL-SL treatment), and in the second the photoperiod changed from 8L:16D to 24L:0D (the SL-LL treatment), the photoperiod being changed by five minutes per day in both cases. There were three replicate tanks (initially n = 60 fish/tank). The experiment was performed over a seven month period from September to the following March, a period spanning the first reproductive period. Fish were sampled every month. Nine female fish were anesthetized until death in seawater with 0.05% MS-222 (3-Aminobenzoic acid ethyl ester methanesulfonate). Body weight, length and gonad weight were recorded, and the brains were immediately, weighed, frozen in liquid nitrogen, and stored at -80˚C. The gonads and brains were placed in Bouin's solution for 24h and then stored in 70% ethanol for later histological examination. Blood was centrifuged and plasma was stored at -80˚C.
Furthermore, the gonadal development in this paper were determined based on GSI, external morphological and histology observation as follow: stage II (beginning of vitellogenesis with primary yolk vesicles,GSI = 0.15%±0.05%),stage III (secondary yolk stage, GSI = 0.31% ± 0.18%), stage IV (accumulating of yolk happened, GSI = 11.87% ± 8.85%), stage V(cytoplasm of oocyte was filled with yolk granules, GSI = 17.57%±3.01%). All of the procedures described in this study were reviewed and approved by the ethical committee of the Institute of Oceanology, Chinese Academy of Sciences.

RNA extraction, preparation of first strand cDNA and Quantitative realtime PCR
Total RNA was extracted from the different regions of the Atlantic salmon brains (telencephalon, diencephalon, Hyp, mesencephalon and the SV) using a fast 200 RNA extraction kit (Fastagen, Shanghai, China), according to the manufacturer's instructions. The total RNA was then dissolved into 20 μL RNase free water. After that, 1μg of total RNA was reverse transcribed to first-strand cDNA by a First-Strand cDNA Synthesis SuperMix (TransGen, Beijing, China) according to the manufacturer's instructions. The reaction system contained 2 × TS Reaction Mix, 0.5μL Oligo dT Primer, 1μL genomic DNA remover and RNase free water up to a 20μL volume.
The primers used to amplify sGnRH3, skissr and β-actin ( [34] were described in Table 1, which were designed from conserved regions of fish GnRH3 and kiss2r in the GenBank database. The gene expression of sGnRH3 and skissr were quantified using SYBR TransStart Top Green qPCR SuperMix Kit (TransGen, Beijing, China) in an eppendorf Mastercycler ep realplex real-time PCR instrument (Eppendorf, German), using the standard curve method with βactin as a reference gene. The primers used to amplify sGnRH, skissr, and β-actin are listed in Table 2. Amplification was performed in a 20 μL reaction volume according to the manufacturer's instructions, using 10 μL 2×Top Green qPCR SuperMix, 0.4μL (4μM) forward and reverse primers, 0.4μL Passive Reference Dye, 1μL cDNA and ddH 2 O up to a 20 μL final volume.

Double color fluorescence in situ hybridization for Atlantic salmon brains
Antisense digoxigenin (DIG) probes and antisense fluorescein isothiocyanate (FITC) were transcribed for salmon sGnRH gene and skissr gene by using a DIG and FITC-labeling Kit (Roche, US), Sense probe for salmon sGnRH and skissr were transcribed as negative control ( Table 3).
The brains from Atlantic salmon reared in the different photoperiod treatments were fixed in 4% paraformaldehyde in 0.1 M PBS (phosphate buffered saline, pH 7.4) at 4˚Covernight. The samples were then dehydrated using a graded methanol series. After that, the samples were mixed with warm paraffin to embed. Sections of paraffin embedded brains were prepared on 5μM glass slides coated with 0.1% Poly-L-lysine solution. The partial CDS of sGnRH3 and skissr were cloned into pGEM-T vectors for preparing sense and antisense RNA probes from a T7 or SP6 promoter using a FITC or digoxigenin (DIG) RNA Labeling Kit (Roche). The sections were hybridized with the sense or antisense probes at 66˚C for 18 hours. After hybridization, the samples were incubated overnight at 4˚C with horseradish peroxidase (POD)conjugated anti-FITC-antibody (Roche) at a 1:2000 dilution in the blocking solution to detect the FITC signal. After three washes in PBST, the samples were incubated for one hour in TSA-Fluorescein at a 1:150 dilution in TSA Amplification Buffer. The samples were then subjected to detect the DIG signal. They were incubated overnight at 4˚C with POD-conjugated anti-DIG antibody (Roche) at a 1:2000 dilution in blocking buffer with 1% H 2 O 2 . Following  PerkinElmer). The nuclei were stained using 4'-6-Diamidino-2-phenylindole (DAPI) and embedded in ProLong 1 Gold Anti-fade reagent (Invitrogen). The slides were then mounted and photographed using a Nikon Eclipse 50i fluorescence microscope. In this procedure, two antisense RNA probes were co-incubated in a single sample during the hybridization step, and developed red and green fluorescence.

Histology
The fixed specimens were dehydrated in graded series of alcohol and embedded in paraffin. Paraffin samples were cut in series of sagittal and cross-sections (5 μm), then stained hematoxylin and eosin (H&E) for histological observation under a light microscope (NikonYS-100, Tokyo, Japan) and the pictures were taken with a digital camera (Nikon coolpix-4500, Tokyo, Japan).

Statistical analyses
All statistical analyses were performed using SPSS 20.0. The results were presented as means ± SD and compared using a one-way ANOVA followed by Tukey's test. All assays were performed in independent triplicates.

Results
The location of skissr and sGnRH3 in the brain of Atlantic salmon First, the mRNA expressions of skissr and sGnRH3 were detected in the different regions of the Atlantic salmon brain using qPCR (Quantitative real-time PCR) with β-actin mRNA as a reference gene. The results showed that the both sGnRH3 and skissr transcripts were primarily expressed in the diencephalon. The transcription levels of sGnRH3 and skissr were higher in the SV than in other parts of the brain except the diencephalon (Fig 1a.). Second, the diencephalon and SV of Atlantic salmon were isolated and used to perform in situ hybridization in order to confirm the precise location of sGnRH3 and skissr transcripts. The results showed that sGnRH3 and skissr transcripts were mainly expressed in the Hyp of the diencephalon. In the SV, the sGnRH3 and skissr also showed the same expression pattern. Both sGnRH3 and skissr were found in the cells close to the ventricles (Fig 1b).

Changes in skissr in the Hyp and SV during gonad development
The experiment ran throughout virtually every stage of development of the Atlantic salmon, from stage II to maturity. The expression levels of kissr were taken as the mean value in each photoperiod treatment. In the early and late stages of gonad development, when the ovaries were at stage II and stage V, the expression levels of kissr transcripts were significantly higher than at stage III and stage IV (Fig 2a and 2c). A similar phenomenon was also observed in the SV (Fig 2b and 2c).

Changes in sGnRH3 in the Hyp and SV during gonadal development
The expression of sGnRH3 in the Hyp and SV increased as the gonads developed. The expression of sGnRH3 transcripts was lowest when the ovaries were at stage II. The expression of sGnRH3 mRNA was relatively stable when the ovaries were at stage III and stage IV. The mRNA level of sGnRH3 increased to its highest when the ovaries were at stage V (Fig 3a). The level of sGnRH3 transcripts in the SV showed the same expression pattern as in the Hyp, while the expression level of sGnRH3 in the SV was lower than in the Hyp (Fig 3b) The expression pattern of skissr under different photoperiod treatments in the Hyp and SV of Atlantic salmon Due to the skissr was mainly expressed during the early and later stages of gonad development in Atlantic salmon. So, we examined the expression of skissr in the Hyp and SV under the different photoperiod treatments, when fish ovaries at stage II and stage V. The results showed that photoperiod can affect the expression of skissr both in the Hyp and SV (Fig 4). When the ovaries were at stage II, skissr transcript levels were highest in the 24L:0D photoperiod treatment, followed by the LL-SL treatment (Fig 4a) these being the two treatments with the longest photoperiods. There were no significant differences between the other photoperiod groups. When the ovaries were at stage V, the highest kissr transcript levels were detected in the 24L:0D group followed by the LL-SL group (Fig 4c). There were no significant differences between other photoperiod treatments. The expression pattern of skissr in the SV was similar to that found in the Hyp, and the level of skissr transcripts in the SV was lower than in the Hyp (Fig 4b and 4d).

The expression pattern of sGnRH3 under different photoperiods in the Hyp and SV of Atlantic salmon
The expression levels of sGnRH3 were also affected by the photoperiod. When the ovaries were at stage II, the highest expression levels of sGnRH3 were detected in the 24L:0D photoperiod treatment, followed by the LL-SL treatment. The lowest expression levels were observed in the 8L:16D treatment, followed by 12L:12D and SL-LL (Fig 5a). However, when the ovaries were at stage IV, there were no significant differences between the treatments. In addition, when the   ovaries were at stage V, the expression levels of sGnRH3 transcripts were once more affected by the photoperiod. The highest expression levels of sGnRH3 transcripts were detected in the 24L:0D and SL-LL treatments, and there were no significant differences between the others (Fig 5c). The expression patterns of sGnRH3 in the SV were similar to those in the Hyp (Fig 5b  and 5d).
The co-expression of sGnRH3 and skissr in the Hyp and SV of Atlantic salmon In order to investigate the possible relationship between GnRH neuron and kisspeptin in Atlantic salmon, the co-expression of sGnRH3 and skissr transcripts were examined in both the Hyp and SV. The results showed that the cells co-express both sGnRH3 and skissr transcripts both in the Hyp and SV. Meanwhile, this co-expression pattern generally appeared when the ovaries were at stages II and V, there was no detectable co-expression in the other two stages. (Figs 6 and 7).

Discussion
Reproduction process is achieved through a precise synchronization of gonadal development and environmental signals [4]. So far, many environmental factors have been found to affect  (Fig 4a and 4c) and saccus vasculosus (Fig 4b and  4d) in fish with gonads at stage II (Fig 4a and 4b) and stage V (Fig 4c and 4d) under different photoperiod treatments. Data are shown as mean ±SD. One-way ANOVA is performed to determine the significant differences between means Columns sharing different letters show significant difference (p < 0.05). reproduction, including photoperiod, temperature, nutritional status, rainfall, population level and lunar phase [35]. Of these factors, it is widely accepted that photoperiod and temperature are the only environmental factors that provide a consistent guide to the timing of reproduction, and that photoperiod provides the most precise and reliable signal to entrain the reproduction process [36]. In our previous study on Atlantic salmon (in press), we also found that photoperiod could affect gonadal development, and that a long photoperiod could promote gonadal development.
Kisspeptin has been identified as playing a key role in the initiation of puberty and the regulation of seasonal breeding in mammals [11,37,38]. Research into kisspeptin in teleost fish is still in its infancy, and it is speculated that kisspeptin performs similar roles in fish species as in mammals.
In this study, we first investigated the relationship between sGnRH3 and skissr. sKissr was mainly expressed in fish during the early and late stages of gonad development.. However, the  (Fig 5a and 5c) and saccus vasculosus (Fig 5b and 5d) in fish with gonads at stage II (Fig 5a and 5b) and stage V (Fig 5c and 5d) under the different photoperiod treatments. Data are shown as mean ±SD. One-way ANOVA is performed to determine the significant differences between means Columns sharing different letters show significant difference (p < 0.05).
doi:10.1371/journal.pone.0169569.g005 sGnRH3 transcript levels did increase as gonad development progressed. Research on the effect of kisspeptin on gonad development after puberty is very limited. Only a few studies shown that kisspeptin can affect ovulation during the reproductive stages [21,39]. However, results regarding the role of kisspeptin in gonad development are still contradictory. Some researchers believe that kisspeptin can promote the secretion of luteinizing hormone (LH) [40]. On the other hand, another study found that a similar dose of kisspeptin cannot elicit LH secretion [41]. In light of the results of this study, we speculated that kisspeptin can promote GnRH release only in the early and late stages of gonadal development. In order to confirm this hypothesis, the co-expression of sGnRH3 and skissr mRNA in the brain at the different gonad stages is examined using double color fluorescence in situ hybridization. The results showed that the skissr is expressed in the sGnRH3 neurons, indicating that kisspeptin might affect GnRH secretion directly. The skissr transcripts mainly appeared at stage II and stage V. We therefore speculated that kisspeptin initiates GnRH release at two stages, the early stage including puberty, and the later stage of gonadal development. Both of these periods are critical to reproduction, and we believe that these two stages need more GnRH to ensure that the gonads develop. Zmora et al. reported that kisspeptin regulated GnRH release is stage dependent especially at the pre-spawning phase in striped bass (Morone saxatilis) [19,39]. In seasonally breeding mammals, some researchers implied that the photoperiodic control of reproduction may involve indirect/direct regulation of the kisspeptin/kissr system [10,42,43]. In teleost, there are very few studies of photoperiod and the kisspeptin/kissr system. Martinez-Chavez reported that a long photoperiod could inhibit the kisspeptin receptor expression and then delay the onset of puberty [17]. However, another study in medaka found that a long photoperiod induced higher numbers of kiss neurons than short photoperiods [29]. There is still no direct evidence as to whether photoperiod can regulate kisspeptin expression. In this study, we found that kissr expression was controlled by the photoperiod when fish had gonads at stages II and V. When fish had gonads at stage II, the higher levels of skissr transcripts were mainly apparent in the LL-SL and 24L:0D treatments. At this time, the photoperiod of the LL-SL group was about 21L:3D, also a long-day photoperiod. Fish with gonads at stage V mainly showed the higher level of kissr transcripts in the 24L:0D and in the SL-LL photoperiod treatments, while the SL-LL photoperiod was changing from short-day to long-day. These results indicate that a long photoperiod can promote the expression of kisspeptin and kissr. In addition, we detected changes in GnRH transcripts under different photoperiods when the fish gonads were at stage II and stage V. The sGnRH3 transcript levels were also higher under the long-day photoperiod, similar to the changes in skissr transcript levels. Combined with the results from double-color in situ hybridization, we conclude that long-day photoperiods can promote sGnRH3 secretion via the kisspeptin/kissr system only at both early and later stages of gonad development.
The SV is a circumventricular organ of the Hyp of fish, and the functions of the SV have not yet been entirely clarified [44]. A recent study of masu salmon (O. masou masou) found that some elements controlling seasonal reproduction are expressed in the SV, and that ablation of the SV prevents photoperiodically-induced gonadal development. This suggests that the SV plays a key role as a seasonal sensor in fish [12,45]. In order to investigate whether the SV is an organ that can regulate seasonal reproduction via the kisspeptin/kissr system in Atlantic salmon, the changes in expression of skissr and sGnRH3 under different photoperiod treatments were examined. The SV is composed of coronet cells, supporting cells and cerebrospinal fluid contacting cells (Fig 8) [46]. We found that both skissr and sGnRH3 were expressed in cells close to the ventricles, which might be the CSF-c cells (not in the coronet cells) of the SV. This indicates that SV might function to regulate gonad development. The changes in sGnRH3 and skissr transcript levels in the SV under the different photoperiod treatments are similar to the changes in the Hyp. However, the expression levels of both skissr and sGnRH3 are lower than those in the Hyp. This phenomenon was apparent both under different photoperiods and the different gonad development stages. Due to physiological activity of fish is too easily influenced by environment cues, we suggest that in the Atlantic salmon, the SV might be an organ which assists in regulating reproduction via photoperiodic signals to maintain normal physiological activity.

S1 File. Contains original figures and data.
(RAR)