Aquaculture has been practiced in Korea for several hundred years, with a significant contribution from mariculture farming. As such, it is important to maintain a sustainable mariculture industry in order to gain economic and environmental benefits. The mariculture field has encountered significant production losses in recent years due to the occurrence of numerous infectious diseases. Although there is a high demand for rock bream Oplegnathus fasciatus, especially in eastern Asia, culturing them in bulk has become a great challenge due to the infectious diseases primarily caused by Edwardsiella tarda (Mohanty and Sahoo, 2007) and rock bream iridovirus (RBIV) (Do et al., 2004). In this regard, exploring fish genes that are involved in growth related pathways and immunogenic mechanisms has become a trend in current marine genetic investigations as a primary step of developing a sustainable mariculture industry.

Follistatin (FST) is a monomeric glycoprotein that is considered an inhibitor of follicular stimulating hormone. FST has the ability to suppress other proteins in a manner similar to that of inhibin (Welt et al., 2002). Moreover, it can antagonize several growth factors, including activins, a group of TGF-β superfamily members (Tortoriello et al., 2001). FST is expressed in all types of tissues and exhibits a wide range of tissue distribution. This diversified tissue expression pattern also provides clues to understanding the primary function of FST, which can effectively modify the paracrine/autocrine roles of activin in a variety of tissues. To date, two major FST subfamilies have been identified (Clabaut et al., 1988). Subfamilies are classified based on the two major characteristics of FST-like proteins: (a) a relatively strong activin binding ability and (b) structural homology to other FST-like proteins. Members of the first subfamily have both properties while members of the second subfamily lack the ability to bind activin.

It is known that FST plays a major role in fish development, since significant levels of FST expression have been detected in a zebrafish model (Bauer et al., 1998). FST was found to be expressed at relatively high levels during advanced stages of embryonic development (Nakamura et al., 1990; Michel et al., 1993; Amthor et al., 2002). Moreover, according to previous studies on activin A, FST is involved in fish reproduction and muscle development. By antagonizing and interacting with the functions of members of the TGF-β superfamily, FST can regulate a variety of TGF-β superfamily proteins in growth- and development-related pathways. Early detection of FST transcripts in sea bream embryos also suggested a possible physiological role in fish development (Funkentein et al., 2009). In addition to activin A, extensive studies on FST suggest its involvement in antagonizing the action of several other members of the TGF-β superfamily. FST itself can also interact with bone morphogenetic proteins (BMPs) (Fainsod et al., 1997; Iemura et al., 1998; Amthor et al., 2002), growth differentiation factor-9 (GDF-9) (Lin et al., 2003), and myostatin [MSTN, or growth differentiation factor-8 (GDF-8)] (Zimmers et al., 2002) in order to abolish their functions during particular growth stages.

Recent studies on FST have provided new insight on its functions, including its involvement in immune-related pathways and developmental physiology. The induction of activin A causes the death of B cells, which are major components of the host defense mechanism (Yu et al., 1987). Additionally, differentiation and proliferation of cells from erythroid lineages was observed following the induction of activin A expression (Broxmeyer et al., 1988). These observations, and its ability to interact with other cytokines such as IL-1 and IL-6 at different regulatory levels, confirm the involvement of activin A in inflammatory pathways (Yu et al., 1987). The interaction of FST with members of the TGF-β superfamily, including activin A, suggests that FST can potentially participate in host immune/inflammatory responses. Moreover, studies under septic conditions (Broxmeyer et al., 1988) showed that because of this FST-activin A interaction, FST must also have a critical role in inflammatory disorders, as it is a stiff inhibitor of activin A. In this study, we have characterized the FST gene from rock bream (RbFST) at the molecular level and determined the basal mRNA expression levels of RbFST in selected tissues. The transcriptional modulation of RbFST was also examined in rock bream liver tissues, stimulated with E. tarda and Streptococcus iniae, in order to elucidate its putative involvement in the host immune response.

The full-length RbFST cDNA sequence was identified from the previously established rock bream cDNA sequence database (Whang et al., 2011) using the Basic Local Alignment Search Tool (BLAST) (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

The open reading frame (ORF) and amino acid sequences of RbFST were derived using DNAssist 2.2. The FST protein sequences from other species were obtained using a BLAST search. These sequences were then used for pair-wise and multiple sequence alignments by utilizing EMBOSS needle (https://www.ebi.ac.uk/Tools/psa/emboss_needle/) and ClustalW2, respectively (Thompson et al., 1994). Phylogenetic relationships were determined with Molecular Evolutionary Genetics Analysis (MEGA) software version 5 (Tamura et al., 2011) using the Neighbor-Joining method and bootstrapping values taken from 1000 replicates. Characteristic protein signatures in the RbFST sequence were predicted using the ExPASy-prosite server (http://prosite.expasy.org) and the NCBI conserved domain database (CDD) (Marchler-Bauer et al., 2011). Some physicochemical properties of RbFST were determined using the ExPASy ProtParam tool (http://web.expasy.org/protparam).

O. fasciatus, with an average size of 50 g, were obtained from the Ocean and Fisheries Research Institute (Jeju special Self-Governing Province, Republic of South Korea). The fish were acclimated for one week in 400-L tanks at 22-24°C in a controlled environment (34 ± 1 practical salinity units, pH 7.6 ± 0.5). Subsequently, gills, liver, kidney, head kidney, spleen, skin, intestine, heart, brain, and muscle tissues were dissected from three healthy fish. Blood samples (~1 mL/fish) were taken from the caudal fin using a 22-G syringe. Samples were immediately centrifuged at 4°C (3,000 g) for 10 minutes to separate the blood cells, and then directly frozen in liquid nitrogen before storing at -80°C. Total RNA was isolated from all of the selected tissue samples, separately, using Tri ReagentTM (Sigma, St. Louis, MO, USA) according to the manufacturer’s protocol.

To examine the immune response of RbFST upon immune challenges, two immune challenge experiments were performed using E. tarda and S. iniae as pathogenic stimuli (Umasuthan et al., 2014). The fish were intraperitoneally injected with 100 μL of E. tarda (5 × 10³ CFU/mL) or S. iniae (1 × 10⁵ CFU/mL) that had been re-suspended in 1x phosphatebuffered saline (PBS). The control group was injected with the same volume of PBS. Next, liver tissues of at least three injected animals were collected at 3, 6, 12, 24, and 48 h postchallenge. This process was consistent between the challenged and control groups.

Total RNA was extracted from rock bream tissues that had been pooled from three fish, from both the control and challenged groups, using Tri Reagent^TM (Sigma). After quantification by UV spectrophotometry (optical density at 260 nm), the total RNA samples were diluted to 1 μg/μL and used to perform cDNA synthesis using the PrimeScript^TM cDNA Synthesis Kit (TaKaRa Bio, Japan), according to the manufacturer’s instructions. Finally, newly synthesized cDNA was diluted 40-fold (total volume: 800 μL) and stored at -20°C until use.

qPCR was performed using the Dice^TM TP800 Real-Time Thermal Cycler System (TaKaRa) in a 15-μL reaction volume containing 4 μL of diluted cDNA, 7.5 μL of 2 × TaKaRa ExTaq^TM SYBRpremix, 0.6 μL of sequence-specific primers (Table 1), and 2.3 μL of ddH₂O. The following thermal cycling conditions were used: 10 sec at 95°C; followed by 35 cycles of 5 sec at 95°C, 10 sec at 58°C, and 20 sec at 72°C; and a final cycle of 15 sec at 95°C, 30 sec at 60°C, and 15 sec at 95°C. The Dice^TM Real-Time System Software (version 2.00) automatically set the baseline. RbFST expression was determined using the Livak (2^-ΔΔCT) method (Livak and Schmittgen, 2001). The same qPCR profile was used for detection of the internal control gene expression, rock bream β-actin (Accession No. FJ975145), with gene-specific primers (Table 1). Expression levels were analyzed in triplicate and the data are presented as the mean ± standard deviation (SD) of relative mRNA expression. The relative expression levels of RbFST at the 0-h time-point (un-injected control) was used as the baseline for comparison. Expression levels were further normalized to the corresponding PBS-injected controls at each time point. The statistical significance of differences observed between the un-injected control (0 h) and challenged groups was determined with a two-tailed, unpaired t-test at a significance level of P < 0.05.

[Table 1.] Oligomers used in this study

The RbFST cDNA sequence is a total of 2,419 bp that includes a 933-bp ORF encoding 311 amino acids. The 5’-untranslated region (UTR) is composed of 70 bp and the 3’-UTR consists of 1,416 bp. The molecular mass of RbFST was predicted to be 35 kDa and the theoretical isoelectric point was 4.8. According to the results obtained from NCBI-CDD and ExPASy-prosite servers, RbFST was an FST-like protein because it shares common structural properties with FST-family proteins, including the presence of a Kazal-type serine protease inhibitor domain, which extends from amino acid residues 51 to 98. An EF-hand, calcium-binding domain was also detected, expanding from amino acid residues 191 to 226 (Fig. 1).

[Fig. 1.] Nucleotide and amino acid sequences of RbFST. The nucleotide sequence (upper) and the deduced amino acid sequence (lower) are numbered. The start (ATG), stop (TAA) codons and poly (A) signal (AATAAA) are boxed. N-terminal signal peptide is underlined. The FST N-terminal domain-like domain is indicated with double headed arrows (Residues 28-51), Kazal type serine protease inhibitor domain is showed within “(Miyabe et al., 2014)”(Residues 51-96) and EF-hand calcium-binding domain profile is highlighted (Residues 191-226).

Pair-wise and multiple sequence alignments revealed that the putative RbFST protein shares characteristic features with FST protein orthologs from other species. Multiple sequence alignment (Fig. 2) showed the presence of a highly conserved Kazal domain, which is a feature of FSTs. The EF-hand, calcium binding domain is also identified in FSTs as a conserved domain among several species, including rock bream. Pairwise sequence comparisons showed that Siniperca chuatsi shares the highest identity (98.7%) and similarity (99.4%) with RbFST. The RbFST proteins from other fish species are also similar (Table 2). In phylogenetic analyses, RbFST clustered with Siniperca chuatsi FST within a fish clade. These data indicate that RbFST is a counterpart of the fish FST-family of proteins, likely originating from a common ancestor of vertebrates, with a characteristic domain architecture that is conserved among a number of fish species (Fig. 3).

[Fig. 2.] Clustal W multiple sequence alignment of the deduced amino acid sequence of RbFST in present study with known FSTs. The identical residues are indicated by asterisk (*) and semi conserved residues are indicated by semicolon (:) and dots (.). Kazal domain profile is indicated by thick double headed arrow and highly conserved regions in the domain are shaded. The EF-hand calcium-binding domain profile is indicated with thin double headed arrows and highly conserved residues are boxed.

[Fig. 3.] Phylogenetic analysis of RbFST with selected known full-length folliststin amino acid sequences of other species. Using MEGA v5.2.2 a phylogenetic tree was reconstructed employing the Neighbor Joining method. Branch numbers are the bootstrap values for 1000 replicates. Location of RbFST is indicated with a black triangle and clusters of fish, amphibians, birds and mammals are indicated accordingly.

[Table 2.] Amino acid identity and similarity of RbFST entire sequence to other homologues

In order to determine the transcriptional profile of RbFST under physiological conditions, qPCR was performed using gene-specific primers and cDNA was synthesized from different tissues obtained from healthy rock bream. Relative mRNA expression levels were calculated using the expression of rock bream β-actin as the reference gene and the expression in each tissue was further normalized to that in blood cells. The results demonstrate that some of the tissues express a considerably higher amount of RbFST transcripts, while other tissues show relatively lower levels of RbFST expression (Fig. 4). This suggests that RbFST functions in a tissue-selective manner. The lowest level of expression was observed in blood cells while the heart and liver showed the highest expression levels. The gill, spleen, kidney and brain showed moderate levels of RbFST expression.

[Fig. 4.] The tissue specific mRNA expression of RbFST determined by qPCR. Expression fold changes were shown relative to the mRNA expression level in blood cells. Error bars represent the SD (n=3).

The main role of FST in heart tissues remains unclear and studies are being conducted to identify the involvement of FST in cardiac conditions (Ogura et al., 2012; Miyabe et al., 2014). Even though the expression of FST under disease conditions, such as myocyte hypertrophic growth, the loss of ventricular performance in response to pressure overload (FST-like 1), and ischemic injury (FST-like 3) (Shimano et al., 2011), has been detected, FST’s expression and involvement in heart tissues under physiological conditions has not yet been extensively studied. As detected in our expressional analysis, transcript levels of RbFST were prominent in liver tissues. The liver is known to be involved in host-immune responses (Racanelli and Rehermann, 2006). FST consistently has the ability to interact with most of the cytokines and components of cell signaling pathways, which in turn can trigger immune responses (Murakami et al., 2012). The presence of FST transcripts at high levels under physiological conditions in the liver may indicate the importance of FST as an agent of immuno-modulation. Similarly, moderate levels of FST mRNA have been ascertained from the tissue distribution analysis of sea bream Sparus aurata (Funkenstein et al., 2009).

Prominent expression levels of FST differed among species under physiological conditions. In Sparus aurata, higher expression levels were detected in the gill filaments, eye, and brain tissues (Funkenstein et al., 2009). Although broad tissue distribution profiles of FST have been detected in mammals (Michel et al., 1991; Tuuri et al., 1994) and chickens (Davis and Johnson, 1998), only a few fish species exhibiting a wide range of FST tissue-specific expression have been documented, including sea bream (Funkenstein et al., 2009). Moreover, prominent expression of FST has been detected in mammalian muscle tissue (Tuuri et al., 1994) as well as in chickens and other fish species, including sea bream (Funkenstein et al., 2009) and zebrafish (Bauer et al., 1998). Studies conducted on the ability of FST to bind with myostatin (MSTN) have shown considerable levels of FST expression in muscle tissues from mammals and chickens. Evidence has been gathered on the possible association between FST and MSTN, similar to that observed in mammalian and chicken muscles, in fish muscles from studies on sea bream FST (Funkenstein et al., 2009).

In both vertebrates and invertebrates, FST is known to be involved in growth and immunological pathways (Murakami et al., 2012). The liver is a key component of both the adaptive immune and innate immune systems (Racanelli and Rehermann, 2006). The components of the innate immune system play a major role since they act as the first line of defense against infectious pathogens. In order to examine the immune responsive expression of FST under pathogenic conditions, liver tissue was collected from immune-challenged rock bream fish.

As shown in Fig. 5A, after healthy rock breams were challenged with S. iniae, the transcript levels of RbFST at 6 h post injection (p.i.) were significantly up-regulated (P < 0.05) (fold ~ 2.5 ) in the liver. At the 12 h p.i. time-point, there was not a significant (P > 0.05) elevation in the expression levels from the basal level. A considerable augmentation in RbFST mRNA levels was observed at 24 h and 48 h p.i. (P < 0.05), reflecting the positive regulation of RbFST upon S. iniae infection. After injection with E. tarda, RbFST expression levels were up-regulated (fold ~ 3.2) at 6 h p.i. Following a significant down-regulation at 24 h p.i., RbFST mRNA levels reached the basal level. These data suggest the plausible involvement of RbFST in immune-related mechanisms. Recent studies on FST-like proteins have also confirmed the role of FST in innate immunity via CD14 and toll-like, receptor 4-related signaling pathways (Murakami et al., 2012). Collectively, these observations may further confirm the importance of FST in innate immunity and possibly in the expressional modulation of FST mRNA levels upon pathological challenges.

[Fig. 5.] Transcriptional regulation of RbFST in liver tissue upon pathogen challenges. Expression of RbFST after Streptococcus iniae (A) and Edwardsiella tarda (B) challenges. Error bars represent the SD (n=3). Significantly up-regulated or down-regulated time points are marked with “*”.

We have characterized a member of the FST-family of proteins from rock bream (RbFST) at the molecular level and determined its expression levels in different tissues under physiological conditions, revealing ubiquitous expression. Moreover, RbFST expression was modulated upon pathogen stress mounted by two different live pathogens (S. iniae and E. tarda). Altogether, findings from this study suggest the putative involvement of RbFST in immune regulation in rock bream, further validating its important role in rock bream physiology. Nevertheless, further studies are needed to confirm the involvement of FST in the immune regulation of teleosts.