A novel tilapia prolactin receptor is functionally distinct from its paralog

Prolactin (PRL) regulates several important physiological processes in fish, including salinity adaptation(Manzon, 2002; Sakamoto and McCormick, 2006),modulation of immune functions (Harris and Bird, 2000), pigment dispersion(Kitta et al., 1993),reproduction (Cavaco et al.,2003) and development (Nguyen et al., 2008). In euryhaline species, such as tilapia, PRL is known to promote freshwater adaptation. Relevance of PRL in freshwater adaptation has been shown since Pickford and Phillips demonstrated that hypophysectomized killifish required PRL to survive in fresh water(Pickford and Phillips, 1959). Recent studies also support a critical function of this hormone for teleost osmoregulation (Hirano, 1986; Sakamoto and McCormick, 2006). Two PRL variants, PRL₁₇₇ and PRL₁₈₈, are expressed in tilapia (Specker et al., 1985; Yamaguchi et al., 1988). However, based primarily on binding assays it was argued that only a single PRL receptor (PRLR) is present in fish(Auperin et al., 1994a). This argument has been consistently supported since the first PRLR was cloned from Oreochromis niloticus (Auperin et al., 1995; Sandra et al.,1995). At the same time the existence of only a single PRLR has been debated because of the different actions exerted by the two kinds of PRL(Auperin et al., 1994b; Specker et al., 1989). Shedding light on this controversy, recently it was reported that two PRLR genes exist in seabream, which conclusively demonstrated the occurrence of PRLR1 and PRLR2 in teleosts (Huang et al.,2007).

In a previous study aimed at the identification of genes up-regulated rapidly during osmotic stress in the euryhaline tilapia (Oreochromis mossambicus) we isolated an unidentified clone (SSH#828) transiently induced upon osmotic stress (Fiol et al.,2006a). In this work, we were able to identify clone SSH#828 as OmPRLR2 after sequencing the full-length mRNA. A partial sequence of a PRLR1 gene with high similarity to the O. niloticus PRLR1 had been cloned before in O. mossambicus(Prunet et al., 2000). In the present study we report the cloning of full-length cDNAs for O. mossambicus PRLR1 (OmPRLR1) and the novel O. mossambicus PRLR2 (OmPRLR2). Furthermore, we characterized the effect of environmental salinity, tissue type and sex on OmPRLR1 and OmPRLR2 expression at mRNA and protein levels in tilapia. In addition, we performed functional analyses of OmPRLR1 and OmPRLR2 using a heterologous expression system, observing a differential capability of activating downstream signaling pathways and responsiveness to tilapia PRL variants. Our studies revealed the presence of a truncated OmPRLR2 protein variant that is regulated in response to salinity stress and more highly expressed in males than in females. The implications of these findings for tilapia osmoregulation, in particular during the process of transitioning between plasma hypo-osmotic and plasma hyperosmotic milieus, are discussed.

Tilapia (Oreochromis mossambicus, Peters 1852) were maintained in large (4 ft, ∼1.22 m, diameter) tanks supplied with flow-through heated(26°C) Davis well water (fresh water, FW, Na⁺=28 mg l^–1, K⁺<5 mg l^–1,Ca²⁺=33 mg l^–1, Mg²⁺=36 mg l^–1, pH 8.0) at the Center for Aquatic Biology and Aquaculture (CABA) of the University of California, Davis. Five days prior to treatments, fish were transferred to 20 gallon (∼76 l) recirculation aquaria containing FW at 25–27°C. Fish were transferred from FW to sea water (SW) by making successive changes of a fraction of the water (about 20–30% of the volume) in order to minimize handling stress during experiments. Full-strength SW was reached over a period of ca. 90 min; then fish were kept with recirculation of SW. Control fish transferred from FW to FW were exposed to the same water changes but without altering the salinity. For long-term acclimation to SW fish were kept for at least 4 weeks in 29 gallon (∼110 l) recirculation SW aquaria. Osmotic stress was applied by making successive changes of a fraction of the water to reach the desired condition over a period of ca. 90 min. SW was collected at Bodega Bay, California (1000 mosmol kg^–1, ∼32 p.p.t.) or prepared with Instant Ocean sea salt to the same concentration. Animals were processed and sorted into batches based on sex and size to exclude these factors from compounding the analysis. Three all-male and three all-female batches were created and at least six animals from each batch were sampled at each point. Therefore, only same-sex animals were compared with PRLR expression values normalized to the respective FW controls within each batch. Relative abundance values for each batch were comparable regarding the effects of salinity acclimation and combined (averaged). Fish were sampled at the times indicated, gills were perfused and gill epithelium was collected by scraping it off the cartilage of individual gill arches as described in previous work (Kültz and Somero,1995). In addition, muscle (posterior to the dorsal fin), and the whole intestine, kidneys, heart, brain, liver and testes or ovaries were collected. All different tissues were kept at –80°C and subsequently utilized for RNA or protein isolation.

Full-length sequences for each gene were cloned using a SMART RACE cDNA amplification kit (Clontech, Mountain View, CA, USA). OmPRLR1fragment AAC31825 was used as a starting point and extended with 3′ RACE until completion of the full-length cDNA. OmPRLR2 full-length cDNA was obtained from the previously unknown sequence SSH#828 (GenBank DQ465388)using 5′ RACE. Novel full-length cDNA sequences were submitted to GenBank (EU999783, EU999784, EU999785).

Total RNA from mIMCD3 cells or tilapia tissues was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA) as specified by the manufacturer. RNA was treated with DNase (Turbo DNA free, Ambion, Austin, TX, USA) and purity was confirmed and quantity determined by measuring absorbance of the samples at 260 and 280 nm with a Nanodrop spectrophotometer. For cDNA synthesis, RNA(1 μg) was reverse-transcribed using Superscript III first-strand synthesis reagents (Invitrogen) with a random hexamer:oligo(dT) mix (1:1) as primers. Abundance of all transcripts was quantified with a PRISM 7500 real-time thermal cycler (Applied Biosystems, Foster City, CA, USA). Reactions were performed in duplicate in a 20 μl reaction volume using SYBR Green PCR Master Mix (Applied Biosystems) and 30 pmol of each primer. PCR conditions were 50°C/2 min and 95°C/10 min, followed by 40 cycles of 95°C/15 s, 60°C/1 min. Data were collected at 60°C. The ΔCtmethod was used to determine the relative amounts of different genes,normalized to the abundance of either β-actin (tilapia tissues) or ribosomal protein L32 (HEK293 cells), and expressed as normalized RNA level in arbitrary units. Normalization genes were selected based on constant levels of expression during all experimental conditions as determined previously(Fiol et al., 2006b). Efficiencies of individual PCR reactions were analyzed using LinRegPCR and were always ∼2. Gene-specific primer sequences were designed with Primer Express software (Applied Biosystems). The absence of unwanted by-products was confirmed by automated melting curve analysis and agarose gel electrophoresis of PCR products. Primers utilized in this study are shown in Table 1.

For protein extraction, cells were lysed in a buffer that contained 50 mmol l^–1 Tris HCl, pH 7.4, 1% nonidet P-40, 0.25% sodium deoxycholate, 150 mmol l^–1 NaCl, 1 mmol l^–1EDTA, 1 tablet of minicomplete protease inhibitor mixture (Roche,Indianapolis, IN, USA) per 10 ml, 1 mmol l^–1 activated Na₃VO₄ and 1 mmol l^–1 NaF. Protein concentrations were determined by BCA protein assay according to the manufacturer's instructions (Pierce Chemical, Rockford, IL, USA). Proteins were separated by SDS-PAGE. Equal amounts of protein (25 μg) were loaded in each lane of 10% Tris-glycine SDS-PAGE gels. Samples were electrophoresed at 120 V, the gels briefly rinsed in transfer buffer (25 mmol l^–1 Tris, 200 mmol l^–1 glycine, 20%methanol), and proteins blotted onto PDVF membranes (Millipore, Bedford, MA,USA) at 1 mA cm^–2 for 90 min using a TransBlot SD semidry transfer cell (Bio-Rad, Hercules, CA, USA). Membranes were blocked for 1 h at room temperature in a solution containing 137 mmol l^–1 NaCl,20 mmol l^–1 Tris, pH 7.6 (HCl), and 5% (w/v) non-fat dry milk. They were then incubated for 2 h in blocking buffer containing anti-PRLR2 antibody at 1:500 dilution (custom-made, raised in rabbits using a KLH-conjugated peptide SESSSEESSEKTKSSQ; Sigma-Genosys, Woodlands, TX, USA). Goat anti-rabbit secondary antibody conjugated to horseradish peroxidase was used at 1:2000 dilution (Pierce catalogue no. 31460). Blots were developed with SuperSignal Femto (Pierce) and imaged with a ChemiImager (Alpha Innotech,San Leandro, CA, USA).

Details of antibody production are available from D.K. on request.

PRLR1 and PRLR2 open reading frames (ORFs) were amplified with the respective ATG forward and END reverse primers(Table 1). PCRs were performed using the Advantage HF-2 PCR kit (Clontech) in a MasterCycler (Eppendorf,Hamburg, Germany) using the following cycling parameters: 94°C/1 min, 32 cycles of 94°C/30 s, 60°C/30 s, 72°C/1.5 min, and then 72°C/5 min. PCR products were extracted from agarose gels using the GeneClean system(Q-BIO gene, Carlsbad, CA, USA) and then double-pass sequenced on an ABI 3730 automated DNA sequencer (Foster City, CA, USA). pcDNA5/FRT/PRLR1 and pcDNA5/FRT/PRLR2 constructs were created by cloning the respective PCR products into pcDNA5/FRT/TO TOPO TA expression vector (Invitrogen). The constructs were then propagated in Escherichia coli strain DH5α(Invitrogen). Endotoxin-free plasmid mega-preps were performed using a kit as described by the manufacturer (Qiagen GmbH, Hilden, Germany). Stable cell lines were established by transfecting HEK293 T-Rex cells (Invitrogen) with 2μg of a 1:9 mix of pcDNA/FRT/PRLR1 or pcDNA5/FRT/PRLR2 plasmid DNA:pOG44 plasmid DNA and 4 μl of LipofectAMINE 2000 reagent (Invitrogen). Twenty-four hours after transfection cells were exposed to selection in medium containing 0.15 μgml^–1 hygromycin (Invitrogen). After 2 weeks individual colonies were picked, expanded, and tested for expression of OmPRLR1 or OmPRLR2 by RT-PCR and quantitative real-time PCR analysis.

Cells were cultured in 96-well plates and exposed to hypersaline media. Cell survival was analyzed with ViaLight cytotoxicity assay (Lonza,Walkersville, MD, USA) following the manufacturer's indications. The method incorporates bioluminescence detection of cellular ATP as a measure of viability; alive cells, which contain higher amounts of intracellular ATP than dead or dying cells, will generate after cell lysis a higher amount of luminescence. Thus, a higher value in relative luminescence units correlates directly with a higher number of alive cells. Similar results were obtained when viable cells were counted in Neubauer hemocytometer chambers using 0.2%Methylene Blue as an indicator of dead cells.

Multiple sequence alignments were performed using AlignX, a multiple sequence alignment application of Vector NTI Advance 9.0 (InforMax Software,Invitrogen), which uses a modified ClustalW algorithm.

Differences between pairs of data were analyzed by F-test and Student's unpaired t-test. Differences in time series data sets were statistically evaluated using ANOVA. Significance threshold was set at P<0.05 and data are presented as means ± s.e.m.

We identified OmPRLR2 from a sequence previously classified as unknown in a study aimed at identifying genes that are rapidly upregulated during osmotic stress in tilapia (Fiol et al., 2006a). To enable characterization of prolactin receptors in tilapia, we performed full-length cloning of both OmPRLR1, which had previously been partially cloned, and OmPRLR2. OmPRLR1 and OmPRLR2 full-length cDNA sequences were obtained by RACE-PCR using GenBank IDs AAC31825 and DQ465388 as starting sequences. The nucleotide and deduced amino acid sequences of OmPRLR1 and OmPRLR2 are shown in supplementary material Fig. S1. OmPRLR1 and OmPRLR2 share only 28.3% identity of sequence overall. When separately comparing sequences in the transmembrane region, the extracellular domain (ECD) of the amino-terminus and the intracellular domain (ICD) of the carboxy-terminus, we noticed that the degree of sequence conservation was much higher in the ICD than in the ECD (Table 2). A comparison of tilapia PRLRs and the human PRLR long variant showed higher similarity between OmPRLR1 and hPRLR than between the two tilapia PRLRs (Table 2). We were able to identify conserved features of the cytokine class I receptor (CKR1) family,including two pairs of cysteine residues, N-glycosylation sites and the functional WS, and Box 1 and Box 2 domains in both tilapia PRLRs. However,ubiquitination and 14-3-3 binding domains were only present in hPRLR and OmPRLR1 but not in OmPRLR2. The location of these domains in human and tilapia PRLRs is depicted and summarized in Fig. 1 and Table 3. Furthermore, three tyrosine phosphorylation sites were identified in OmPRLR1 and four in OmPRLR2. Analysis of the amino acids surrounding the tyrosine residues revealed that three of them (I, II and IV) are conserved and present in both receptors while site III is only present in OmPRLR2(Table 2; Fig. 1). An alignment of selected PRLR1 and all reported PRLR2 sequences is shown in supplementary material Fig. S2.

Levels of expression of OmPRLR1 and OmPRLR2 mRNAs were analyzed in different organs of FW-acclimated tilapia. Both OmPRLR1 and OmPRLR2 showed highest expression levels in gills, kidney,intestine and muscle, although expression was also detected in brain, liver and heart for both receptors (Fig. 2).

Using specific OmPRLR2 antibodies, we analyzed the protein levels in the different organs of fish acclimated to FW. Two different isoforms were detected: the expected full-length protein of ∼55 kDa and a smaller peptide of about 35 kDa (Fig. 3). OmPRLR2 was expressed widely but with organ-specific expression profiles.

We investigated the response of OmPRLR1 and OmPRLR2 mRNAs after acute hyperosmotic and hyposmotic stress in gills and confirmed that OmPRLR2 transcript increased in gills 6 h after transfer of fish from FW to SW compared with the handling control (FW to FW transfer; Fig. 4). Comparison of OmPRLR1 and OmPRLR2 transcript levels in fish that were long-term acclimated to FW and SW showed that steady-state levels of OmPRLR1 and OmPRLR2 were higher in FW gills than in SW gills(FW–FW versus SW–SW in Fig. 4). We also compared OmPRLR2 protein abundance in fish that were long-term acclimated to FW versus SW. While no differences were detected in any organ for the 55 kDa isoform, expression of the 35 kDa truncated protein was significantly higher in fish acclimated to FW in gills, kidney, muscle, heart and brain(Fig. 5). These results suggest that OmPRLR2 regulation during salinity stress is complex and consists of multiple phases at transcript and protein levels.

Interestingly, expression of the 35 kDa truncated variant of OmPRLR2 was significantly higher in gills of males than females in FW-acclimated fish (Fig. 6).

To confirm the presence of a truncated variant of OmPRLR2 at the mRNA level we investigated the presence of splice variants. Using RT-PCR and primers including the initiation codon and stop codons, we were able to detect two transcripts. One of them comprised a 1590 bp ORF and the other a 1050 bp ORF (Fig. 7A). The shorter transcript was found to be expressed ubiquitously as is illustrated in Fig. 7A for heart, brain and eye. Isolation and sequencing of the shorter (1050 bp) band verified that it indeed represented a mRNA coding for a OmPRLR2 variant. A graphical representation of both OmPRLR2 variants, indicating key residues and domains, is depicted in Fig. 7B. Synteny analysis comparing all available genomic data for the fish PRLR2 locus (Danio rerio, Fugu rubripes, Gasterosteus aculeatus,Oryzias latipes and Tetraodon nigroviridis) permitted identification of a conserved exon structure and high nucleotide conservation at splice sites. At least nine main exons are present in PRLR2 genes; exon I corresponding to the 3′ UTR and the coding region being composed of exons II–IX. The presence of additional exons spanning the 5′UTR is expected but cannot be predicted because the 5′ UTR was not included in the known or predicted orthologous sequences. Nevertheless, we were able to predict the exon composition of OmPRLR2 variants in the coding region. Full-size OmPRLR2 showed a typical exon composition with exons II–IX comprising the coding region and a high degree of conservation of exon sizes. On the other hand, the short isoform lacked exons IV, V, VI and part of exon VII (Fig. 7C). Thus, the short OmPRLR2 variant seems to result from alternative splicing between exon III and exon VII.

Stably transfected cell lines expressing OmPRLR1 and OmPRLR2 were established in HEK293 T-Rex cells to functionally characterize tilapia PRLRs. Quantification of changes in the expression of PRLR target genes was used to monitor responsiveness to hormone treatment in cell lines expressing either OmPRLR1 or OmPRLR2. Multiple conserved genes whose expression was previously shown to be modulated by prolactin were analyzed; c-Fos, p53, p21, Bcl-xl, Socs2, c-Myc and Spi2.1 expression levels were quantified by quantitative real-time PCR after 4 h treatment with either tilapia PRL₁₇₇ or tilapia PRL₁₈₈ at 20 ng ml^–1.

Based on their expression profiles in response to prolactin treatment, PRLR target genes could be separated into three different groups. First, c-Myc expression was induced for both kinds of receptor, when cells were treated with PRL₁₇₇ but not with PRL₁₈₈(Fig. 8A). The second group comprises most of the genes analyzed, including c-Fos, p53, p21,Bcl-xl and Socs2, which were induced only in cells expressing OmPRLR2 when treated with PRL₁₇₇ but not PRL₁₈₈(Fig. 8B shows c-Fosas a representative example of this group). The third group includes Spi2.1, which was induced only in cells expressing OmPRLR1 when treated with either PRL₁₇₇ or PRL₁₈₈(Fig. 8C).

When cell lines expressing OmPRLR1 and OmPRLR2 were challenged with hyperosmotic media, a significant increase in osmotic stress tolerance was observed for OmPRLR2-expressing cells compared with control cells transfected with the empty vector(Fig. 9). On the other hand, OmPRLR1-expressing cells showed no difference in osmotic stress tolerance. When exposed to water of 550 mosmol kg^–1 for 24 h, OmPRLR2-expressing cells showed >80% survival, while OmPRLR1-expressing and empty vector controls showed <20% survival(Fig. 9). When cells were exposed to similar osmotic stress conditions but with heat-inactivated fetal bovine serum (FBS) instead of regular FBS in the medium, no increase in osmotic stress tolerance was observed (data not shown). This result suggests that a PRLR2-activating substance such as prolactin or a related hormone is present in FBS and is necessary to stimulate PRLR2 downstream signaling pathways.

The CKR1 family includes growth hormone receptor, prolactin receptor,leptin receptor, erythropoietin receptor, and receptors for several leukins(Bazan, 1989). Several common features have been described in these proteins, such as conserved domains and key residues (Bole-Feysot et al.,1998). A single-pass transmembrane region divides the amino-terminal ECD from the carboxy-terminal ICD. The ECD contains two pairs of conserved disulfide-linked cysteines and a conserved penta-peptide named the WS domain, which are present in all members of the family and were demonstrated to be involved in ligand binding(Bole-Feysot et al., 1998). A number of N-glycosylation sites are commonly present in the ECD and are involved in positioning the receptor on the cell surface(Buteau et al., 1998). Several conserved domains have also been characterized in the ICD, including Box 1(Edery et al., 1994; Kelly et al., 1991), Box 2(Murakami et al., 1991) and multiple tyrosine phosphorylation sites(Endo et al., 2003). Additionally, some less ubiquitous motifs have been described, including the ubiquitination domain (Li et al.,2004) and the 14-3-3 binding domain(Olayioye et al., 2003). Both tilapia PRLRs contain all the above-mentioned ECD features and the single-pass transmembrane region. Furthermore, Box 1 and Box 2 could be also identified in the ICD of both tilapia PRLRs. But several important differences between OmPRLR1 and OmPRLR2 were observed in the ICD: while three tyrosine phosphorylation residues were present in OmPRLR1, four were identified in OmPRLR2. Moreover, the ubiquitination and 14-3-3 binding domains were only present in OmPRLR1. The lack of these regulatory motifs in OmPRLR2 suggests considerable differences in the post-translational regulation of both tilapia receptors. These differences between tilapia prolactin receptors were found to be ubiquitous for all published teleost sequences.

We were able to immunologically detect the presence of a shorter 35 kDa OmPRLR2 variant, in addition to the expected full-length 55 kDa peptide. The corresponding transcript of the shorter variant was cloned and sequenced and shown to represent a splice variant (short OmPRLR2). It is possible that additional splice variants of OmPRLR2 are expressed that were not detected in this study because they lack the antigenic region detected by our OmPRLR2 antibody. The presence of shorter PRLR variants is a common feature in mammals but was only recently described in fish when alternatively spliced PRLR1 forms having a truncated ICD were detected in carp (San Martin et al.,2007). Carp variations resemble human PRLR splice variants but are different from the short OmPRLR2 variant found in this study, where the ICD remains constant compared with the full-length variant.

The exon composition of both OmPRLR2 variants was deduced based on synteny analysis of the PRLR2 loci of five fish with fully sequenced genomes(Danio rerio, Fugu rubripes, Gasterosteus aculeatus, Oryzias latipesand Tetraodon nigroviridis). The exon composition of the coding region of the long OmPRLR2 variant showed the conserved pattern of exon number and size. In contrast, the short variant represents a truncated protein that lacks a region flanked by exons III and VII, which is likely to be a result of alternative splicing.

Given the required dimerization step of PRLR during receptor activation,the presence of distinct PRLR2 variants may allow the formation of hybrid receptors with altered PRL binding and downstream pathway activation capabilities. A schematic model illustrating signaling consequences of OmPRLR1 and OmPRLR2 activation as well as a possible hybrid between long and short OmPRLR2 is presented in Fig. 10. According to this model, transient up-regulation of the short OmPRLR2 variant observed during salinity stress may constitute a regulatory event that sequesters long OmPRLR2 variants and inhibits PRL binding. Of interest, higher expression of the short OmPRLR2 variant in tilapia males than in females may indicated that prolactin signaling is gender specific and inhibited at the receptor level in males. Recently, gender-specific differences were also found in the expression of a variety of other tilapia receptors, including growth hormone receptor and several estrogen receptors(Davis et al., 2008).

Both PRLRs showed ubiquitous expression in all analyzed tissues, consistent with a multifunctional role of PRL. High expression levels were especially detected in gills, kidney, muscle and intestine. Earlier studies in tilapia, Sparus aurata and Takifugu rubripes reported comparable expression profiles for prolactin receptors(Lee et al., 2006; Pierce et al., 2007; Santos et al., 2001). The two tilapia PRLRs are both more highly expressed in gills of long-term FW-acclimated tilapia than in SW-acclimated fish. However, upon acute transfer from FW to SW mRNA levels of OmPRLR2 increase significantly although transiently for up to 12 h, confirming our previous data(Fiol et al., 2006a). The reason for such complex regulation of OmPRLR2 abundance at transcript and protein levels is currently not understood.

Functional analysis of OmPRLR1 and OmPRLR2 in HEK293 cells allowed us to characterize the activation of major target genes that are involved in downstream signaling pathways in response to ligand binding. The two tilapia PRLRs responded differently to treatment with the two tilapia prolactins (PRL₁₇₇ or PRL₁₈₈). Based on the responses of a number of genes previously shown to be activated by prolactin, we were able to distinguish targets that are activated only via OmPRLR1 by both PRL₁₇₇ and PRL₁₈₈ (Spi2.1), targets that are activated via both receptors by PRL₁₇₇ only(c-Myc) and targets that are only activated via OmPRLR2 by PRL₁₇₇ only (c-Fos, p53, p21, Bcl-xl and Socs2). Similar results were obtained for Spi2.1 activation by seabream PRLR1 and PRLR2 when treated with salmon prolactin(Huang et al., 2007). However,these authors found that a c-Fos promoter reporter gene was induced by activation of both prolactin receptors. The discrepancy with our results may be attributed to the different expression systems employed (transient versus stably inducible in our study), the hormone concentrations used (300 ng ml^–1versus 20 ng ml^–1in our study), and the use of heterologous receptor–hormone combinations versus utilization of all receptors and hormones from the same species (O. mossambicus) in our study. Differential effects of PRL variants were observed previously for transiently expressed O. niloticus PRLR1. In this case, a higher affinity for PRL₁₈₈was observed and a correspondingly lower concentration of hormone was required to activate the Jak2/Stat5 pathway (Sohm et al., 1998). In addition, different potencies were reported for the two prolactins with regard to their effect on pigment dispersion in cultured xanthophores and erythrophores(Oshima et al., 1996). Pigment dispersion was affected by nanomolar concentrations of tilapia PRL₁₇₇ but only by micromolar concentrations of tilapia PRL₁₈₈. Specific activation of c-Fos, p21, p53, Bcl-xl, and other key components of cellular stress response networks and cell survival signaling by OmPRLR2 binding of PRL₁₇₇ can explain the increased osmotolerance of HEK293 cells over-expressing OmPRLR2. However, we cannot exclude the possibility that OmPRLR2 activates additional survival pathways that are not activated via OmPRLR1 and lead to increased osmotolerance. Nonetheless, a transient increase in osmotolerance, whether mediated via c-Fos and the other OmPRL2 target genes identified in this study or by as yet unidentified PRLR2 targets, may be beneficial during the transient `crisis period' encountered by tilapia upon transfer to SW. During this period, which lasts about 1–3 days, major restructuring of the gill epithelium and changes in the direction and activity of ion transport take place. The transient increase in PRLR2 expression may serve to support cell survival during this time.

In summary, we have identified a novel prolactin receptor in tilapia(OmPRLR2) and compared its expression at mRNA and protein levels to that of OmPRLR1. We identified a short OmPRLR2 splice variant, which is differentially expressed by males and females and regulated by environmental salinity. Functional characterization of both OmPRLRs revealed differences in their responsiveness to tilapia prolactins (177 or 188) and several shared but also unique downstream targets. In particular, activation of c-Fos via transient induction of OmPRLR2 and/or inhibition of prolactin signaling viaformation of non-functional receptor hybrids may contribute to osmoprotection of gill cells during the period of active restructuring of gill epithelium in response to salinity stress. Further dissection of prolactin signaling pathways is necessary to discern how each variant of this hormone(PRL₁₇₇ and PRL₁₈₈) contributes to physiological osmoregulation during steady-state conditions and during periods of acute osmotic stress.