1 INTRODUCTION

Seahorses utilize the charismatic reproduction strategy of male pregnancy (Wilson et al., 2001; Lin et al., 2016). The male seahorse possesses a brood pouch, in which the female seahorse deposits her eggs before they are fertilized and undergo embryonic development (Laksanawimol et al., 2006; Rosenqvist and Berglund, 2011). The brood pouch of seahorse is located along the ventral midline of the tail posterior to the anus. The outer epithelium of the pouch is composed of stratified cuboidal epithelial cells, whereas the lumen of the pouch is surrounded by three layers containing abundant blood vessels: two dermis layers and the pseudoplacenta. The outer dermis layer consists of a loose network of connective tissue and the inner dermis layer consists of tightly connected collagenous fibers. The pseudoplacenta is beneath the inner dermis layer and composed of connective tissue with a mesh-patterned structure and abundant blood vessels (Kawaguchi et al., 2017).

Seawater can directly flow into the brood pouch of seahorse during the non-pregnancy state (Whittington and Friesen, 2020). Previous studies had indicated that crucial immune molecules were expressed at high transcriptional levels in the seahorse brood pouch, such as secretion of C-type lectin (Melamed et al., 2005), expression of major histocompatibility proteins MHC IIa and IIb (Luo et al., 2016; Roth et al., 2020), and expression of interleukin IL-20RB (Whittington et al., 2015; Roth et al., 2020). Recent studies revealed that interleukins play crucial roles in the innate immune responses of mammals and teleosts (Kaiser et al., 2004; Secombes et al., 2011; Zou and Secombes, 2016). However, the immune responses of interleukin genes in the seahorse brood pouch are unclear.

Produced by activated monocytes and macrophages, interleukins are a highly diverse subgroup of cytokines present as more than 40 different types in mammals (Zou and Secombes, 2016). Different subfamilies of interleukins have different conserved domains and were defined in previous studies (Secombes et al., 2011; Zou and Secombes, 2016). These proteins are divided into pro-inflammatory and anti-inflammatory cytokines (Kaiser et al., 2004; Secombes et al., 2011). In teleosts, challenge with bacteria or lipopolysaccharide (LPS) can increase the transcription levels of pro-inflammatory cytokines. For example, the mRNA expression of interleukin (IL)-8 is rapidly upregulated after Vibrio parahaemolyticus challenge in the spleen, head kidney, and liver in Larimichthys crocea (Wang et al., 2019). Interleukin (IL)-1β can induce proinflammatory gene expression to further cause pathological inflammation involving activation of the nuclear factor (NF)-κB pathway in Miichthys miiuy (Hayden and Ghosh, 2011; Yang et al., 2017).

In this study, we identified and characterized members of the interleukin gene family at the genome and transcriptome levels in the lined seahorse (Hippocampus erectus) and built a phylogenetic tree with representative genes. Further, we evaluated the potential roles of il-1β, il-18, and il-8 in antibacterial immunity in the brood pouch of seahorses following challenge with LPS and V. parahaemolyticus.

2 MATERIAL AND METHOD 2.1 Ethics statement

All experiments were conducted in accordance with the guidelines and with approval from the animal research and Ethics Committees of the Chinese Academy of Sciences (approval number: SCSIOIACUC-2019-000140).

2.2 Interleukin gene family

To identify interleukin genes in lined seahorse, genome annotations files of Hippocampus comes, H. erectus, and Microphis manadensis were initially searched. The reference genomes of H. comes (LVHJ00000000) and M. manadensis (QODF00000000) and transcriptomic data of H. erectus (SRA392578, SRA392580) were searched with TBLASTN of parameter "tblastn.exe -query -db -out -evalue 0.000 01" using sequences of interleukin genes from humans and zebrafish obtained from the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/) as query sequences (Supplementary Table S1). We then confirmed the acquired interleukin gene of H. erectus from the conserved domain structures of Pfam (https://pfam.xfam.org/) (Supplementary Fig.S1).

2.3 Phylogenetic analysis

At least one representative interleukin from each subfamily was selected (il-1β, il-18, il-17d, il-8, il-10, il-15, and il-6), and phylogenetic analysis was carried out using the amino acid sequences of interleukin genes from H. erectus and several representative fish (Supplementary Table S1). These sequences were downloaded from the NCBI for zebrafish (Danio rerio), tiger tail seahorse (H. comes), manado pipefish (M. manadensis), medaka (Oryzias latipes), fugu (Takifugu rubripes), and stickleback (Gasterosteus aculeatus). Multiple amino acid sequences were aligned using Muscle software (https://www.ebi.ac.uk/Tools/msa/muscle/) with default parameters (Edgar, 2004). We constructed phylogenetic trees using two methods. The phylogenetic tree determined by the maximum-likelihood method was constructed using MEGA7 and that based on the Bayesian method was constructed using Mrbayes (Heled and Drummond, 2010). Bootstrapping was conducted with 1 000 and 10 000 000 replications to evaluate the phylogenetic trees, respectively.

2.4 Animal husbandry

Fifty-six reproductively mature seahorses (average body weight, 13.9±0.5 g) were purchased from Zhangzhou Seahorse Breeding Center (Fujian, China) and maintained in a large tank with a 31–32 salinity and pH 8.2–8.3 at 25–27 ℃. The tank was connected to a central circulation system for 1 week prior to the experiments for acclimatization. The system provided mechanical and biological filtration, ultraviolet sterilization, and protein skimmer for the seahorse.

2.5 Tissue-specific expression analysis

To analyze the levels of interleukin gene transcription in various tissues under normal conditions, six male seahorses that were not injected were selected. Their tissues, including the liver, gill, kidney, intestine, muscle, gonad, and brood pouch, were collected in liquid nitrogen and stored at -80 ℃ until RNA extraction.

2.6 Immune challenge

Fifty seahorses were randomly divided into three groups/tanks for further injection: control (16 seahorses) and two experimental groups (17 seahorses for each group). Control seahorses were injected with 100-μL phosphate buffer saline (PBS) into the sidewall tissue of brood pouch. Experimental seahorses were equally subdivided into LPS or bacterial challenge groups and injected with 100-μL LPS (2 μg/μL) (Sigma, St. Louis, MO, USA, 0111: B4) or V. parahaemolyticus (10⁸ CFU/mL) (Oh et al., 2017; Tharuka et al., 2019), respectively.

The fish were euthanized prior to dissection by anesthetization with tricaine methanesulfonate (MS-222, Sigma, 20 mg/L) (Qin et al., 2016). The intact brood pouch, with both the outer epithelium and three inner layers (two dermis layers and the pseudoplacenta), was sampled at 6, 12, 24, and 48 h post-injection (hpi) from four individuals per time point per group.

2.7 RNA isolation, cDNA synthesis, and qRT-PCR assay

Total RNA was extracted from the tissues using TRIzol Reagent (Life Technologies, Carlsbad, CA, USA), and RNA quality was assessed by both agargel electrophoresis (three clear strips without degradation) and with a nucleic acid analyzer (A260/A280 ranged from 1.8 to 2.0). First-strand cDNA was synthesized using a Genome Erase cDNA Synthesis Kit with gDNA Eraser (TaKaRa, Shiga, Japan) following the manufacturer's protocol, and 1-μg RNA was used for each reaction. A real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed with a Roche Light-Cycler 480 real-time PCR system (Basel, Switzerland) according to the manufacturer's protocol.

The standard curve was constructed using a series of 10-fold dilutions of quantified pMD18-T vector (TaKaRa) with the target gene of interest. Two qRT-PCR experiments (tissue-specific expression and injection) were conducted, and β-actin was used as the reference gene for each experiment because of its stable expression in different seahorse tissues (Cq value ranges from 15–18). The sequences and locations of all primers are shown in Supplementary Table S2 and the amplification sequence of each pair of primers spans at least an intron-exon boundary to avoid the amplification of genomic DNA. No primers had non-specific amplification for the following three reasons: 1) There is only one specific peak in the melt curve; 2) one clear strip in the agar-gel electrophoresis; 3) both no template control and no reverse transcriptase control were conducted per run with three replicates, and they both showed no specific amplification (Cq>35) (Bustin et al., 2009). The final sample for qRT-PCR (10 μL) contained 1 μL of cDNA template, 5 μL of SYBR^Ⓡ Green Realtime PCR Master Mix (Roche), 0.4 μL of each primer (10 pmol/μL), and 3.6 μL of nuclease-free H₂O. The qRT-PCR thermal profile comprised an initial cycle at 95 ℃ for 3 min, followed by 40 cycles of 95 ℃ for 30 s, 58 ℃ for 30 s, and 72 ℃ for 20 s.

The relative expression of interleukin mRNAs was calculated using the 2^-ΔΔCt method (Livak and Schmittgen, 2001). For tissue-specific expression analysis, the liver tissue was used for normalization and for the immune challenge experiment, and the sample from the 6-h PBS-injected control was used to normalize the expression level. Three technical replicates were evaluated.

2.8 Statistical analysis

All data are presented as the mean relative mRNA expression ±SD and analyzed with SPSS software (version 22.0, SPSS, Inc., Chicago, IL, USA). Analysis of variance (one-way ANOVA) followed by Duncan's multiple range test was used to determine the statistical significance of mRNA expression in healthy and challenged tissues.

3 RESULT 3.1 Phylogenetic analysis of seven interleukin sequences in H. erectus

We identified 13 interleukin genes in the H. erectus genome (Supplementary Table S1), seven of which were further confirmed by phylogenetic analysis of the lined seahorse, tiger tail seahorse, manado pipefish, and other close species. As shown in Fig. 1, seven interleukin proteins of the lined seahorse were clustered with their respective homologs in other bony fishes. A conserved phylogenetic relationship was identified within syngnathids: for each selected interleukin member, H. erectus showed the closest relationship with H. comes, and the two seahorses were then clustered with the pipefish M. manadensis and other tested teleosts (Fig. 1). Similar results were obtained from the phylogenetic tree constructed using Bayesian method (Supplementary Fig.S2).

3.2 Expression analysis of interleukin genes in the brood pouch of H. erectus

The expression level of 13 interleukin genes was quantified in the seahorse brood pouches (Fig. 2a). The maximum expression was observed for il-8 and the minimum was observed for il-6, showing a difference of approximately 300-fold (Fig. 2a). To further detect the expression of key interleukin genes in the brood pouch of seahorse, we conducted tissuespecific expression analysis of four genes (Fig. 2b–e). First, we selected the top three genes showing the highest expression levels in seahorse brood pouch (il-8, il-15, and il-18) (Fig. 2a). Second, il-1β was chosen because it not only showed significant responses after LPS and bacterial challenge in Ctenopharyngodon idella (Bo et al., 2015) and Micropterus salmoides (Ho et al., 2016), but also showed relatively high expression in the brood pouch of seahorse (Fig. 2a). Our results showed that il-8, il-15, il-18, and il-1β were expressed at relatively high levels in the brood pouch; il-18 in the brood pouch showed the highest expression level among all selected tissues (Fig. 2b–e).

3.3 Immune responses of il-1β, il-8, and il-18 in seahorse brood pouch

We evaluated the potential immune responses of interleukins in the seahorse brood pouch after challenge with LPS and V. parahaemolyticus. In addition to il-15, significantly changes in il-1β, il-8, and il-18 all were detected after injection with V. parahaemolyticus or LPS (Fig. 3a–c and Supplementary Fig.S3). il-1β mRNA was significantly upregulated at 6 hpi and then gradually downregulated following V. parahaemolyticus challenge but there was no significant response to LPS stimulation (Fig. 3a). Both il-1β and il-18 belong to the same il-1 subfamily but responded differently to exposure to LPS and V. parahaemolyticus. The expression of il-18 was unchanged within the first 24 hpi and then significantly upregulated at 48 hpi in response to LPS (P < 0.05); however, no significant response to V. parahaemolyticus was observed (Fig. 3b).

Compared with il-1β and il-18, il-8 showed remarkable responses following stimulation by LPS and V. parahaemolyticus. il-8 transcripts were upregulated starting at 6-h post-LPS injection, reaching a maximum at 12 hpi (18-fold above the control; P < 0.05). After 12 hpi, il-8 expression decreased but remained significantly higher than that in the control (P < 0.05) until 48 hpi. In contrast, il-8 transcripts were significantly higher at 6-h post V. parahaemolyticus injection than in the control (P < 0.05). After 6 hpi, il-8 expression decreased but remained higher than that in the control until 48 hpi (Fig. 3c).

4 DISCUSSION

Interleukin genes play important roles in the innate immune response (Secombes et al., 2011; Zou and Secombes, 2016). In mammals and teleosts, research into the functional roles of interleukins has focused on immune organs, such as the blood and gills (Herath et al., 2016), but little is known about the seahorse brood pouch. In this study, we identified 13 interleukins in the lined seahorse and investigated the immune responses of il-1β, il-18, and il-8 following challenges with LPS and V. parahaemolyticus directly to the brood pouch.

The independent cluster pattern of each of the seven selected interleukin genes indicated by phylogenetic analysis confirmed the accuracy of the identified interleukin sequences in the lined seahorse (Fig. 1). Additionally, compared to the inconsistent tree topology of different interleukin genes among other teleosts, species within syngnathids (H. comes, H. erectus, and M. manadensis) showed conserved phylogenetic relationships in all 7 selected interleukins. Discrepancies between trees of a single gene and species-level tree may result from incomplete lineage sorting (Nichols, 2001; Wang et al., 2018) and the conserved tree topology of interleukins within syngnathids and between syngnathids and other teleosts indicate the unique role of interleukins in syngnathids with a brood pouch.

il-8, il-18, and il-1β encode pro-inflammatory cytokines that mediate immune responses and induce inflammatory reactions to infections in mammals and teleosts (Biet et al., 2002; Bird et al., 2002; Li and Yao, 2013; Wang et al., 2016). Previous studies reported that these genes are highly expressed in typical immune tissues, such as the blood and gills, in zebrafish (Liao et al., 2018) and large yellow croaker (Li et al., 2013). Our results showed that il-8, il-15, il-18, and il-1β were highly expressed in seahorse brood pouch (Fig. 2a & b). As the brood pouch has important immune functions (Melamed et al., 2005; Whittington et al., 2015), high expression of these ils in the brood pouch indicates their importance in the immune response of brood pouches against pathogens. Additionally, il-1β and il-8 showed high expression levels in the liver (Fig. 2b & e), indicating a role for these genes in regulating immune responses in this organ (Xiao et al., 2019).

Previous studies investigating the response to bacteria in other immune tissues (Zou et al., 2004; Xu et al., 2016; Wang et al., 2019) showed a trend toward a stronger, localized, pro-inflammatory response at the site of infection or injury (Herath et al., 2016). il-1β and il-18 belong to the il-1 subfamily which, in teleosts, mediates the host response to microbial invasion, inflammation, and immune reactions (Akira, 2000; Herath et al., 2016). Interestingly, il-1β and il-18 responded to different challenges (V. parahaemolyticus and LPS, respectively) in the brood pouch of seahorse. LPS is a main component of the cell walls in gram-negative bacteria, comprising lipid A, core polysaccharides, and O-specific chains (Wang and Quinn, 2010). Previous studies showed that lipid A is responsible for the major bioactivity of endotoxins (Wang and Quinn, 2010), In addition to LPS, the outer membrane protein of Helicobacter pylori (a gram-negative bacterium) can also induce characteristic inflammatory responses (Sugimoto et al., 2009). The different sensitivities of il-18 and il-1β post-LPS and post-V. parahaemolyticus challenge may have resulted from their protective functions against different bacterial components. For example, il-18 may only act on lipid A of LPS, whereas il-1β may have to be induced by the whole cell wall of V. parahaemolyticus, including LPS and outer membrane protein. A similar tendency was found in mammals, whereas the outer membrane protein of H. pylori (a gram-negative bacterium) induced characteristic inflammatory responses (Sugimoto et al., 2009). In humans, lipid A compounds can strongly induce IL-18 over IL-1B (Shimoyama et al., 2011). In addition, the bacterial outer membrane protein (OipA) plays an important role in the expression of Il-1b, Il-17, and Tnf-a in mice (Sugimoto et al., 2009).

Il-1β and Il-18 were previously shown to be involved in acute and chronic inflammatory responses in mammals, respectively (Shimoyama et al., 2011). For example, Il-1β mRNA was significantly upregulated at 3 h in blood cells and at 6 h in the spleen after Streptococcus iniae challenge (Herath et al., 2016). Our data showed that il-1β mRNA was increased at 6 h following V. parahaemolyticus infection in the brood pouch, whereas the il-18 mRNA response was slower and was increased at 48 h in response to LPS stimulation, supporting the results of previous studies. Unlike il-1β and il-18, the expression of il-8 mRNA, a chemokine gene, rapidly increased at 6 h and was sustained from 6 to 48 h following LPS and V. parahaemolyticus stimulation (Fig. 3c). This suggests its involvement in a powerful, acute, inflammatory reaction, which is consistent with previous reports of chemokines. In Hippocampus abdominalis, mRNA expression of the CXC chemokine gene (ShCXCL) was upregulated in the blood and kidney tissues after immune stimulation by live bacteria such as S. iniae and Edwardsiella tarda (Oh et al., 2017). Additionally, rapid upregulation of il-8 expression was observed in the immune tissues in L. crocea after V. parahaemolyticus injection (Li and Yao, 2013; Wang et al., 2019) and LPS stimulation (Li and Yao, 2013). In addition, a previous report in mammals illustrated that IL-18 and IL-1β can stimulate IL-8 production via the NF-κB signaling pathway (Biet et al., 2002; Geisert et al., 2012; Herath et al., 2016). NF-κB family transcription factors have been shown to regulate tissue immune function and inflammatory responses in humans (Ghosh and Hayden, 2008; Hayden and Ghosh, 2011). In addition to direct bacterial stimulation, our result may indicate that upregulated il-8 mRNA followed il-1β and il-18 induction by activating the NF-κB signaling pathway (Fig. 3d, Biet et al., 2002; Geisert et al., 2012; Herath et al., 2016), further indicating that il-8 plays an important role in the antibacterial immunity of the seahorse brood pouch.

5 CONCLUSION

In conclusion, we demonstrated that il-1β, il-18, and il-8 play key roles in the antibacterial immune defense of the brood pouch in the lined seahorse (H. erectus). il-1β and il-18 responded to V. parahaemolyticus and LPS, and they could induce acute and chronic inflammatory responses, respectively. il-8 may be involved in a more powerful antibacterial immune responses and may itself be induced by IL-1β and IL-18, involving the NF-κB signaling pathways. These findings provide valuable information that will be helpful for further analyzing antibacterial immune and initial inflammatory responses.

6 DATA AVAILABILITY STATEMENT

The authors declare that all data supporting the findings of this study are available within the article and its supplementary files.

Electronic supplementary material

Supplementary material (Supplementary Tables S1–S2 and Figs.S1–S3) is available in the online version of this article at https://doi.org/10.1007/s00343-021-0310-z.