Expression patterns of kiss2 and gpr54‐2 in Monopterus albus suggest these genes may play a role in sex reversal in fish

Due to its exceptionally small genome size and protogynous hermaphroditism, Monopterus albus has been proposed as a model for vertebrate sexual development. The Kiss/GPR54 system is a central regulator of sexual development in most vertebrates, but its role in sex reversal remains hypothetical. In contrast to mammals, fishes often possess more than one copy of the kiss and gpr54 genes. Our objectives were to identify all kiss/gpr54 genes in the genome of M. albus and to assess their involvement in sex reversal via their expression patterns (qPCR) in females, males, and intersex specimens. We identified only two genes: kiss2 and gpr54‐2. kiss2 expression was extremely high in the gonads of males, intermediate in females, and low in intersex; and reduced in all tissues of intersex. gpr54 expression was also extremely high in the gonads of males, high in intersex, but low in females. gpr54 expression in brain was high in all three sexes. In conclusion, (a) kiss1 has been functionally replaced in M. albus; (b) the functions of gpr54‐2 in brain are not sex‐specific; (c) kiss2 appears to undergo a ‘reset’ in the expression during the sex change; and (d) sex‐specific expression patterns in the gonads indicate that these two genes may play a role in sex reversal in fish.

Due to its exceptionally small genome size and protogynous hermaphroditism, Monopterus albus has been proposed as a model for vertebrate sexual development. The Kiss/GPR54 system is a central regulator of sexual development in most vertebrates, but its role in sex reversal remains hypothetical. In contrast to mammals, fishes often possess more than one copy of the kiss and gpr54 genes. Our objectives were to identify all kiss/gpr54 genes in the genome of M. albus and to assess their involvement in sex reversal via their expression patterns (qPCR) in females, males, and intersex specimens. We identified only two genes: kiss2 and gpr54-2. kiss2 expression was extremely high in the gonads of males, intermediate in females, and low in intersex; and reduced in all tissues of intersex. gpr54 expression was also extremely high in the gonads of males, high in intersex, but low in females. gpr54 expression in brain was high in all three sexes. In conclusion, (a) kiss1 has been functionally replaced in M. albus; (b) the functions of gpr54-2 in brain are not sexspecific; (c) kiss2 appears to undergo a 'reset' in the expression during the sex change; and (d) sex-specific expression patterns in the gonads indicate that these two genes may play a role in sex reversal in fish.
Whereas the sexual fate of a majority of vertebrates is irreversibly determined during embryonic development, many fish species exhibit remarkable sexual plasticity [1]. In some species, this plasticity manifests itself in an intriguing phenomenon of sex reversal during their adult life [1,2]. Among these is a freshwater teleost species Asian swamp eel Monopterus albus (Zuiew, 1793; Synbranchiformes: Synbranchidae) native to South-East Asia. It is protogynous: It starts its sexual life history as a female, undergoes an intersex stage, and then develops into the final male stage [3,4]. Although sex reversal of this fish has received some scientific attention in recent years, for example, [3,5,6], the control and molecular mechanisms of this process remain poorly understood. As this species also has an exceptionally small genome, it has been proposed as a model system for sexual development in vertebrates [3,4]. It is also an economically important cultured fish in many Asian countries [7], but the sex reversal causes low fecundity and presents a major obstacle for largescale breeding [8]. Therefore, our research unit is currently undertaking a research project that aims to improve our understanding of the molecular control of sex reversal in M. albus.
KisspeptinGPR54/system is a component of the hypothalamus-pituitary-gonadal (HPG) axis with a central role in the development of sexual maturity (and sex reversal) in vertebrates. Precisely, Kisspeptin interacts with GPR54 to regulate the secretion of gonadotropin-releasing hormone from hypothalamus [9]. The ancestral state for vertebrates is believed to be the existence of two kiss (kiss1 and kiss2) genes (encoding the Kisspeptin protein) and two Kisspeptin receptors (gpr54, also known as kissr gene), but the kiss2 system was lost in all mammals [10]. As opposed to this, in fish, the number of paralogues of kiss and gpr54 varies from only one to several (four) [10][11][12][13]. As there is evidence that kiss1 and kiss2 paralogues have different functions [12,14,15], this variability in the number of kiss genes is rather intriguing.
Although fishes possess a number of characteristics that make them a suitable model for studying the Kisspeptin/GPR54 system [16,17], the complexity of the system, the existence of compensatory mechanisms, and species specificity of the system hamper the progress of these studies, so the exact roles of kisspeptin signaling genes in sexual development in fish remain debated [18,19]. Whereas the Kisspeptin/GPR54 expression has been studied extensively in fish in the context of brainpituitary axis, early development, onset of puberty, spawning cycles, and annual gonadal stages [18,20], there appears to be only one report of the expression of these genes in the context of sex reversal [12]. As a result, although the association between Kisspeptin signaling and sex reversal has been proposed [1,12], the role of this system in sex reversal remains unknown.
As Kisspeptin system has not been studied in M. albus at all, the objectives of this study were as follows: (a) to identify all extant kiss and gpr54 genes in the genome of M. albus; (b) to indirectly corroborate their roles in the sex reversal in this fish by studying their expression patterns in different tissues of male, female, and intersex (undergoing the sex reversal) specimens.

Identification and characterization of kiss and gpr54 genes
We relied on the published (GenBank Acc. No. AONE00000000) complete genome of M. albus [4], and transcriptome data for different developmental stages (accession number SRX1007627) [5], to identify all extant kiss and gpr54 genes (including paralogues) using BLASTn suite. To identify any putative splice variant isoforms as well [10], we searched proteomes of three gonad types during sex reversal [21]. To indirectly corroborate the identity of the two genes, we conducted phylogenetic analyses using datasets comprising 25-30 vertebrate homologues (we selected four different major taxa: fish, amphibians, mammals, and Monotremata; GenBank accession numbers are shown in the figures). For this, we used predicted amino acid sequences and neighbor-joining method implemented in MEGA7 [22] with 1000 bootstrap replicates.

Samples
Specimens used for experiments were hatched and cultured at the breeding base of the Hubei Provincial Engineering and Technology Research Center for Asian swamp eel. Female swamp eels usually enter the intersex stage immediately after spawning (laying eggs); the intersex period, characterized by simultaneous ovarian degeneration and testicular development, lasts for about 2-3 months, after which the transition into males is complete. Generally, 1-year-old swamp eels are all females; among the 2-yearolds, 70% are females and 30% are males; and among the 3-year-olds, 90% are males and 10% are females. For the experiment, we selected three (apparently) healthy specimens of each sex: females aged~12 months, intersex speci-mens~18 months, and males~36 months ( Table 1). Specimens were euthanized in buffered MS-222 (400 mgÁL À1 concentration), immediately dissected, and the following tissues collected: liver, muscle, spleen, kidney, intestines, heart, gonads, brain, and pituitary gland. Tissues were immediately macrodissected into several~0.1 g sections (to be used for RNA extraction directly), flash-frozen in liquid nitrogen, and stored at À80°C until RNA extraction. To corroborate the sex, a section of gonads was taken for histological examination, conducted under an electronic microscope as described [23,24]. Under the electron microscopy, only oocytes can be seen in females, oocytes and sperm cells can be seen in intersexuals, and only sperm cells can be seen in males (Fig. 1). All animals were handled and experimental procedures conducted in accordance with the guidelines for the care and use of animals for scientific purposes set by the EU Directive 2010/63/EU. Animal experiments for this study have been reviewed and approved by the Ethics Committee of the Yangtze University.

RNA, cDNA, and RACE
Samples (~0.1 g) were ground in liquid nitrogen with mortar and pestle, and RNA was extracted following the standard TRIzol extraction protocol. Quality and quantity of the extracted RNA were assessed by electrophoresis on 1% agarose gel (Fig. 2) and NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). cDNA library was synthesized using RevertAid TM First Strand cDNA Synthesis Kit with DNase I (#K1621 and #EN0521, respectively, both Thermo Scientific) following the manufacturer's protocol, and cDNA stored at À80°C. As we could identify only one kiss/gpr54 gene using the methodology described in the first section, to further confirm that there are no other expressed kiss/gpr54 paralogues present in the transcriptome of M. albus (most importantly, to corroborate that kiss1 and gpr54-1 are absent), we designed degenerate primers for kiss1, kiss2, and gpr54 genes (  [25,26], but we also tested the expression of  Inter is intersex, H is heart, L is liver, S is spleen, K is kidney, B is brain, G is gonad, M is muscle, P is pituitary gland, and I is intestine. Discarded samples are highlighted with a red X sign above. another commonly used reference gene, GAPDH, in different tissues. Stability of Ct values of these two genes was analyzed using the comparative cycle threshold method implemented in the GeNorm software [27]. Although both were suitable, b-actin exhibited less variability (Table S1). Gene expression levels were calculated using the 2 ÀDDCT method [28]. As we could not obtain results for some samples (difficulties during RNA extraction), some results are based on three and some on two biological replicates (Table S1). Blanks were included in each run, and technical replicates were triplicate. qPCR experiments were conducted by the Bio-Transduction Lab (Wuhan, China). MIQE checklist is available as Table S2.

Results
Identification and characterization of kiss2 and gpr54 genes in M. albus We identified only one copy of each of the two genes in the genome (and transcriptome) of M. albus. Kiss2 was identified within the scaffold ID NW_018127972.1 (positions 3063234-3063734, identity 100%). Gpr54 was identified against the annotated sequence of this gene (XM_020591446, LOC109955350, identity 99%: 1131/1134 bp). We did not find any indications of the existence of splice variants in our sequencing results. Phylogenetic analyses corroborated that they are orthologues of kiss2 and gpr54-2 (or kissr2) genes (Fig. 3). The kiss2 mRNA sequence is composed of 656 nucleotides (CDS = 351 nucleotides), encoding a protein of 116 amino acids. It contains putative Kisspeptin-10 (FNLNPFGLRF) and Kisspeptin-12 (SKFNLNPFGLRF) fragments [10]. gpr54-2 is composed of 1134 bases, encoding protein of 377 amino acids. kiss CDS comprised two exons and gpr54 CDS comprised five exons (Fig. 4). Sequences are deposited in GenBank under the accession numbers MF085053 (kiss2) and MF085054 (gpr54).
Expression patterns of gpr54 and kiss2 genes RNA extraction from pituitary gland was not successful (Fig. 2), so the expression was studied in liver, muscle, brain, spleen, kidney, small intestine, heart, and gonads of male, intersex, and female specimens ( Fig. 5; Table S1). Gpr54 was very highly expressed in the brain of all three sexes (> 10 000-fold), but M-3 sample (male-3) was a minor outlier, and I-3 was a major outlier (Fig. 5A). The highest expression among all samples was observed in male gonads: > 100 000 000-fold in the M-3. Expression was also very high in the intersex gonads (~10 000-fold), while expression in female gonads was much lower (< 500fold). Its expression levels were comparatively low in all other tissues, with the following minor exceptions: male spleen (M-3 sample~3000-fold, but very low in the M-1 sample), intersex liver (intermediate expression of 700-to 1000-fold, but very low in the I-1 sample), and male muscle (low in M-2 and M-3, but M-1 was an outlier with > 1000-fold).
Kiss2 was exceptionally highly expressed in the gonads of males (both samples~10 000-fold; Fig. 5B). Expression in female gonads was also comparatively high (> 1000-fold, but with rather high variability among technical replicates), but in the gonads of intersex specimens, its expression was much lower (< 600fold). Expression in all other tissues was comparatively low, with a minor exception of the female muscle (~600-700-fold), and two individual outlier samples: M-2 spleen and M-1 kidney (both > 1000-fold). Intriguingly, expression in the brain tissue of all three sexes was also very low:~300-600-fold in females, 60-180-fold in males, and ≤ 1-fold in intersex.

Discussion
Although the ability to manipulate the timing of puberty and reproductive development in farmed fish is of major commercial relevance [16], the understanding of sex reversal phenomenon in fishes is important and fascinating from a much wider biological, ecological, and evolutionary perspective [2]. A number of gene (and genome) duplications (and subsequent gene losses) in the evolutionary history of fish [10,11,29,30] and the existence of compensatory mechanisms [19] make the studies of kiss/gpr54 system in fish very complex. Both genes (kiss2 and gpr54-2) identified in the genome of M. albus are comparable to those characterized in other teleost fishes in terms of size [12,14,18]. There is evidence that kiss1 and kiss2 have different functions in other fish species [10,11,14]; for example, in medaka (Oryzias latipes), only the kiss1 system, but not the kiss2 system, shows expression dynamics strongly indicative of its direct involvement in the HPG axis regulation [17]. As opposed to this, in some other fishes, which includes M. albus, the kiss1 gene is missing, and only kiss2 is expressed [17]. As the evolution of Kisspeptin and its receptor in fish, aside from gene duplications and loss, involves appearance of splice variant isoforms as well [10], we hypothesized that the decreased size of the genome of M. albus may have resulted in a 'space-saving' functional adaptation: replacing a paralogue with an isoform. However, after searching the genome, transcriptome, and proteome of M. albus, we could not find evidence for the existence of paralogues or isoforms, which indicates that kiss1 may have been functionally replaced in M. albus with kiss2 or some other compensatory mechanism. Constitutive expression of kiss2 and/or gpr54-2 in all studied tissues has been observed in several fish species [12,14,16]. Although we detected both genes in all studied tissues, their expression levels were generally low, and in some cases barely detectable. An intriguing observation regarding the constitutive expression pattern of kiss2 is a notably lower expression in intersex specimens (barely detectable in most tissues apart from gonads, where it was also much lower than in males and females). This indicates that kiss2 appears to undergo a 'reset' in the expression as M. albus undergoes the sex change. This might be in correlation with the proposed important role of the kisspeptin system in modulating gonadal sex differentiation, pubertal timing, and reproduction in teleost fishes [17,18]. In the light of these findings, it would be of interest to assess the constitutive expression of kiss2 in juvenile (prepubescent) M. albus in future studies.
Both genes are generally highly expressed in fish brain [12,14,16,18,31,32], where they probably have a broad range of functions [33][34][35]. The expression of gpr54-2 in the brain of M. albus is in full agreement with this, with similar (high) expression levels among all three sexes, which suggests that functions in brain do not vary between sexes. This may be a reflection of the notorious promiscuousness of kiss receptors, which is believed to be an evolutionary guarantee of functional robustness of the HPG axis regulation system in cases of gene loss or neofunctionalization [17]. Kiss2, however, was not highly expressed in any of the studied brain samples, which may be an indication of functional diversification. However, as the expression of these two genes varies strongly among different parts of the brain [12,14,36], so this may also be a reflection of the fact that we sampled (almost) entire brains, and failed to extract DNA from the pituitary gland, which may have skewed the overall expression results downwards.
Expression patterns of all of these genes (including paralogues) are generally extremely variable, both among tissues and paralogues themselves [18], but it remains unclear whether this is merely a reflection of molecular noise [37], or of species-specific functions of different paralogues. For example, patterns differed among all four genes (two sets of paralogues) in chub mackerel, with the kiss2 gene expressed only in the brain, and neither of the two kiss-receptor genes expressed in the ovary [18], which is notably different from our results. This is not surprising, as they are likely to be functionally distinct: Multiple studies have shown that in fish kiss1 appears to be more effective than kiss2 at inducing puberty [18], but kiss2 has a stronger stimulatory effect on the testicular development during nonbreeding period [38]. This is in agreement with high kiss2 expression in the gonads of male M. albus. Overall, studies suggest that kisspeptin peptide modulates gonadotropin secretion and influences gonadal development in fish, but the effects on ovarian development appear to be slower than the effects on testicular development [18]. This may bear importance for the interpretation of our results, as the expression in female gonads was much lower than the expression in male gonads.
In zebrafish, where two gpr54 paralogues have been identified, gpr54-1 was expressed in the brain and gonads, while gpr54-2 was constitutively expressed in most tissues [32]. The expression pattern of gpr54-2 in M. albus therefore appears to resemble that of zebrafish gpr54-1, but only in males. Most importantly, gpr54-2 expression pattern in gonads (extremely high in males, high in intersex, low in females) presents a strong evidence of sex-specific expression pattern of this gene in the gonads of M. albus.
Unfortunately, the interpretation of our data is hampered by technical difficulties that we experienced during the RNA extraction, which resulted in us having to discard some samples. This in turn resulted in a small number of biological replicates (fish specimens) used in the study and prevented us from conducting meaningful statistical analyses. Furthermore, some of our samples also exhibited signs of RNA degradation (Fig. 2), so it will be necessary to have our findings corroborated on a larger number of samples by a future study. Difficulties in the extraction of RNA from fish gonads have been discussed in detail before [39], so future studies should also probably try to tinker with their RNA extraction protocols in order to obtain a higher-quality RNA.

Conclusions
Using M. albus as a model, this study aimed to improve the understanding of Kisspeptin/GPR54 axis functioning in the sex reversal in fish (and vertebrates in general). The main conclusions that can be inferred from our results are summarized as follows: Kiss1 system (present in many other fish species) has been functionally replaced in M. albus with Kiss2 system or some other compensatory mechanism; expression of gpr54-2 in brain does not appear to vary between sexes; kiss2 expression appears to undergo a 'reset' in expression as M. albus undergoes the sex change; and we found a strong evidence of sex-specific expression pattern of both kiss2 and gpr54-2 in the gonads of M. albus. Although these findings indicate that Kisspeptin/GPR54 system might indeed play a role in the sex reversal of M. albus, interpretation of these data and comparisons between species are rendered exceptionally complex by the existence of multiple paralogues in other fish species, by their species-specific functions, diverse reproductive strategies, and differences in experimental approaches between studies [10,16]. Additionally, as targeted mutations of kiss and gpr54 do not disrupt gonadal development and reproductive performance in zebrafish, this is a strong indication of the existence of a compensatory mechanism that ensures reproductive success [19,40,41]. This suggests that, as opposed to mammals, kisspeptin may not be a central and absolute upstream regulator of the HPG axis in fish [18]. Regardless of these limitations, in future studies we aim to explore the relationship between the regulation of these genes and downstream sex reversal and gonadal differentiationassociated genes and hormones, and thereby improve the understanding of Kisspeptin/GPR54 axis functioning in sex reversal in fish using M. albus as a model.

Author contributions
TLY and DQY conceived and designed the project; TLY and MTP acquired, analyzed, and interpreted the data; and TLY and DQY wrote the paper.

Data accessibility
The two sequenced genes are deposited in GenBank under the accession numbers MF085053 (kiss2) and MF085054 (gpr54).

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Table S1. qPCR data. Table S2. MIQE checklist.