The role of a new insulin-like peptide in the pearl oyster Pinctada fucata martensii

Pinctada fucata martensii, is an economically important marine bivalve species cultured for seawater pearls. At present, we know little about the molecular mechanisms of the insulin signalling pathway in this oyster. Herein, we cloned and analysed an insulin-like peptide (PfILP) and its signalling pathway-related genes. We detected their expression levels in different tissues and developmental stages. Recombinant PfILP protein was produced and found to significantly increase primary mantle cell activity and induce the expression of the proliferating cell nuclear antigen (PCNA) gene. PfILP could also regulate the 293T cell cycle by stimulating the S phase and inhibiting the G1 and G2 phases. Recombinant PfILP protein induced the expression of its signalling pathway-related genes in mantle cells. In vitro co-immunoprecipitation analysis showed that PfILP interacts with PfIRR. PfILP activated expression of the pfIRR protein, and also activated the mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K) pathways by stimulating phosphorylation of MAPK and AKT. Further analysis showed that PfILP up-regulated glycogen synthesis-related genes glycogen synthase kinase-3 beta (GSK-3β), protein phosphatase 1 (PP1) and glucokinase (GK) at the mRNA level, as well as the expression of the PP1 protein, and phosphorylation of GSK-3β. These results confirmed the presence of a conserved insulin-like signalling pathway in pearl oyster that is involved in cell activity, glycogen metabolism, and other physiological processes.

of gene products and that more functional data are needed. Regarding ILP receptors, a unique pfIR receptor (PfIRR) was identified in P. fucata martensii and found to be mainly expressed in the testis and adductor muscle 35 . Moreover, an IGF binding protein gene (Pfigfbp) from P. fucata martensii was characterised, and found to be transcribed mainly in the foot 36 . However, no insulin-like analogues or peptides were found in P. fucata martensii, and effectors of the insulin-like signalling pathway remain poorly documented. In addition, it is unknown whether PfILPs has similar functions to vertebrates insulins or IGFs. Furthermore, whether PfILP can act like vertebrate insulin/IGFs and activate similar downstream signalling pathways in P. fucata martensii remains unknown.
To investigate these questions, in the present work we identified the first ILP and six effectors in the pearl oyster P. fucata martensii. We analysed their phylogenetic relationships with other homologous proteins and determined their spatio-temporal expression by quantitative real-time PCR (qPCR). We probed the interactions between recombinant PfILP and pfIRR, and the effects of PfILP-mediated pfIRR stimulation in the activation of the MAPK and PI3K signalling cascades in primary P. fucata martensii mantle cells. We further examined the biological activity of PfILP using in vitro phosphorylation assays and investigated its modulatory roles in glycogen metabolism and the cell cycle.

Results
Identification of seven ILP signalling pathway genes. Seven potential effectors related to ILP signalling (based on sequence homology)-Pfirs1, Pfpik3r1, Pfakt, Pfsos2, Pfrap-1, Pfraf and PfILP-were identified from P. fucata martensii(Supplementary Table S1). BLASTp analyses showed high sequence homology between these six genes (except the PfILP) and genes known to be involved in the insulin signaling pathway (Supplementary  Table S1). Sequence alignment demonstrated the levels of high identity; From an evolutionary perspective, six ILP signalling pathway genes are thought to have evolved from an ancestral gene, respectively, which explains the common features . For PfILP genes, it has four exons and three introns in genomic (Fig. 1A). PfILP protein comprises a signal peptide followed by a B-chain, a connecting peptide (C-peptide), and an A-chain (Fig. 1B). PfILP includes six cysteines in the A and B chain, allowing the formation of disulphide bridges, and two putative disulphide bonds were assumed to form two inter-chain bridges across the two chains, and one intra-chain bond on the A-chain (Fig. 1B) that is essential for tertiary folding. The 3D structure of PfILP is consistent with this phenomenon (Fig. 1C). Multiple sequence alignment of PfILP showed that their sequences are poorly conserved, especially among invertebrates, except for the six cysteine sites, which are strictly conserved ( Supplementary Fig. 7). The GenBank number is in Supplementary Table S2. The phylogenetic tree shows that PfILP sequence identified in P. fucata martensii is orthologous to the previously identified mollusks ILPs in Crassostrea gigas, Crassostrea virginica, Mizuhopecten yessoensis, Sinonovacula consricta (Fig. 1D). The GenBank number is in Supplementary Table S3. The identified ILP signalling-related components provide valuable information for future studies of this pathway in P. fucata martensii. Predicted three-dimensional model of PfILP based on human insulin-like growth factor 1 (IGF1; PDB id 2gf1). Conserved cysteine residues and the resulting disulphide bridges are marked. The model was constructed using Swiss-Model and edited using PyMOL Viewer. (D) Maximum likelihood phylogenetic tree of PfILP from P. fucata martensii with homologs from other species. PfILP from P. fucata martensii is marked with a green circle. Numbers at tree nodes indicate the bootstrap percentage after 1000 replicates.
The expression patterns of seven ILP signalling pathway genes in different tissues and developmental stages. Seven tissues from pearl oysters or different developmental stages were selected to determine the basal transcriptional levels by real-time PCR. Most of them are highly expressed in the foot or in D-shaped larvae whereas, in contrast in adductor muscle or 32-cells embryos there is apparently contrasting expression patterns (Fig. 2). The results showed that the expression patterns of the seven genes in tissues or developmental stages were totally same, which suggests that they might be involved in same physiological activities of in P. fucata martensii.
Production of recombinant PfILP. The expression vector containing a cDNA encoding the mature PfILP polypeptide (no signal peptide) fused to a His-tag was constructed and transformed into E. coli Transetta (DE3) cells, and expression of recombinant PfILP protein was induced by IPTG. The results showed that most of the protein was present in insoluble inclusion bodies (Fig. 3A). Negative control cells containing vector alone did not yield overexpression bands. The His-tagged PfILP protein was retrieved from inclusion bodies via solubilisation before purification. Refolding PfILP protein sample was filtered to remove impurities from the resin by repetitive washing. And then a single protein band of 26 kDa was greatly enriched in the final column eluate (Fig. 3B). Subsequent immunoblotting performed with anti-His antibody revealed that the purified protein was specifically labelled (Fig. 3C).
Cell viability following treatment with recombinant PfILP. A large quantity of cells were migrated out from explant at day 7. Cells almost covered plates at 2 weeks (Fig. 4A). Trypan blue staining demonstrated that more than 90% of cells survived up 2 weeks (Fig. 4B). The mitogenic effect of PfILP was examined by measuring the cell viability of primary cells using a Cell Counting Kit-8. The results clearly showed that PfILP improved mantle cell viability at a concentration of 0.25 μg/ml till top at 1 μg/ml (Fig. 4C). Additionally, the PCNA gene was up-regulated maximally by PfILP at a dose of 1 μg/ml (Fig. 4D).
PfILP regulates the expression of pfIRR. To verify the interaction between PfILP and pfIRR, in vitro Co-IP analysis was performed using P. fucata martensii mantle cells. The results showed that PfILP was co-immunoprecipitated by pfIRR (Fig. 5A), and pfIRR was co-immunoprecipitated by PfILP (Fig. 5B). Additionally, we investigated the expression of pfIRR in mantle cells following treatment with PfILP. The pfirr transcript was significantly up-regulated following treatment with 1.0 μg/ml or 2.0 μg/ml PfILP (Fig. 5C). The pfIRR protein expression was examined by determining the ratio of pfIRR/GAPDH. As shown in Fig. 5D, PfILP

PfILP activates MAPK and PI3K/Akt signalling pathways.
To study the effects of PfILP on activation of the MAPK and PI3K signal transduction pathways, the seven genes (Pfirs1, Pfpik3r1, Pfakt, Pfsos2, Pfrap-1, Pfraf and Pferk) associated with the insulin-like signalling pathway of P. fucata martensii were examined. Additionally, phospho-p44/42 MAPK, p44/42/MAPK, phospho-Akt and Akt also were examined. As the treatment concentration of PfILP increased, Pfirs1, Pfpik3r1 and Pfakt transcripts were upregulated, and levels peaked at a dose of 2.0 μg/ml. By contrast, Pfsos2 and Pferk transcripts were not significantly changed. The Pfrap-1 transcript was upregulated at a dose of 0.25 μg/ml, and the Pfraf transcript was upregulated at a dose of 2.0 μg/ml (Fig. 6A).
Phosphorylation of Akt was examined by determining the ratio of phosphorylated to total Akt. As shown in Fig. 6B, PfILP induced the phosphorylation of Akt (on residue T308) from 5 min to 240 min, and phospho-Akt levels peaked at 60 min. PfILP increased Akt (T308) phosphorylation with increasing dose from 0.25 μg/ml to 1.0 μg/ml, but this declined at a dose of 2.0 μg/ml (Fig. 6C). PfILP also induced MAPK phosphorylation within 5 min, and the induction lasted for 240 min (Fig. 6D). Again, activation time-dependently increased from 0 to 240 min. When treated for 30 min, PfILP dose-dependently induced phosphorylation of p44/42 MAPK at a concentration ranging from 0 to 1.0 μg/ml (Fig. 6E). These results clearly show that recombinant PfILP activates both MAPK and PI3K/Akt signalling pathways.

PfILP influences the expression of glycogen-related genes and proteins.
To investigate the participation and conservation of carbohydrate/glycogen metabolism in response to PfILP induction, we analysed three glycogen-related genes involved in PfILP signalling. The results showed that the relative expression levels of GSK-3β, GK and PP1 were significantly increased in mantle cells after PfILP stimulation (Fig. 7A).

Discussion
The insulin/insulin-like signalling pathway is evolutionarily conserved in invertebrate and vertebrate animals 37,38 .
Several key insulin pathway components including PfILP, Pfirs1, Pfpik3r1, Pfakt, Pfsos2, PfRap-1 and PfRaf have been cloned from P. fucata martensii. PfILP shares the typical characteristics of ILP proteins, including six conserved cysteine residues that form disulphide bridges, and a similar 3D structure (Fig. 1). Previous studies showed that the insulin-relaxin superfamily have a similar modular organisation as their precursor, including an N-terminal signal peptide, A, B and C domains 2,13 . Post-translational modification leads to cleavage of the signal peptide and the C-peptide, resulting in a mature hormone consisting of the A and B chains. The six conserved cysteine residues in ILPs form one intra-chain (within the A-chain) and two inter-chain disulphide bridges that play a key role in protein function by cross-linking peptides 19,32 . Herein, six potential effectors of signalling, sharing varying degrees of conservation, were investigated. In vertebrates, these effectors are related to the MAPK pathway (Pfsos2, PfRap-1 and PfRaf) or the PI3K pathway (Pfirs1, Pfpik3r1 and Pfakt), suggesting both pathways may be conserved in molluscan species.
Analysis of the expression of Pfirs1, Pfpik3r1, Pfakt, Pfsos2, Pfrap-1, Pfraf and PfILP revealed that they were expressed during different developmental stages and in various tissues (Fig. 2). These results indicate that these proteins encoded by these genes may have broad functions. ILP signalling pathway genes display similar expression patterns in oyster tissues. It should be noted that PfILPs (including Pfirs1, Pfpik3r1, Pfakt, Pfsos2, PfRap-1 and PfRaf) were highly expressed in the foot. In oysters, the neuroendocrine system is relatively developed in the foot. In mollusc species, growth, reproduction, and their associated metabolic processes are known to be subjected to neuroendocrine control mechanisms, and nervous ganglia are crucial centres for the production of regulatory molecules 3 . However, they were expressed primarily in D-shape larvae in the present study. The D-shape larval stage is a period in which embryonic development and organ formation is relatively fast. Furthermore, the D-shape larval stage is crucial for soft tissue and shell growth. High expression of ILP signalling pathway genes during this developmental stage implies that they may play a crucial role. www.nature.com/scientificreports www.nature.com/scientificreports/ In this study, we explored a mantle primary cell culture as an in vitro model to investigate the effect of the insulin-like peptides PfILP on P. fucata martensii mantle cell development and metabolism. The recombinant mature PfILP peptide (without the signal peptide) increased mantle cell viability (Fig. 4A), and upregulated the PCNA gene. This phenomenon show that PfILP may promote cell proliferation. In different gastropod species, neuroendocrine factors participate in soft growth and shell growth by stimulating cell proliferation and protein synthesis [39][40][41][42][43] .
In brain, ILPs are specifically synthesised and released into the hemolymph, and then transported to target cells, where they interact with insulin receptors, triggering downstream signalling pathways [44][45][46] . Using Co-IP assays, PfILP and pfIRR were shown to interact with each other (Fig. 5A,B). Moreover, we investigated if the binding of PfILP to ILP receptors (pfRR) could activate the receptors and elicit downstream signalling cascades in P. fucata martensii. Early research shows that the activation of IGF receptors can cause the MAPK and PI3K/Akt pathways cascades 47,48 . In our previous study, human IGF-1 was shown to interact with P. fucata martensii pfIRR and to activate MAPK and PI3K/Akt pathways in P. fucata martensii oocytes 49 . Furthermore, in Sparus aurata, insulin signalling pathway-related transcripts were affected by IGF-I or IGF-II 50 . In our study, most ILP signalling pathway-related genes in P. fucata martensii was up-regulated in response to PfILP. Moreover, we demonstrate that PfILP also induced the expression of pfRR (Fig. 5C-E) and activated PI3K/Akt and MAPK pathways by stimulating phosphorylation of Akt and MAPK, although their threshold concentrations and durations differed (Fig. 6). Previous studies reported that Akt activation stimulates myogenic differentiation, whereas MAPK activation stimulates mitogenesis 48,51,52 . These results showed that PfILP can activate downstream intracellular signalling pathways in P. fucata martensii.
Early research found that exogenous insulin regulate glycogen metabolism in invertebrates such as the white shrimp Penaeus vannamei 53 and the lobster Panulirus argus 54 . HrIGF-I through coupled PfIRR to activate the MAPK and PI3K signaling pathways, and then regulate glycogen metabolism in P. fucata martensii 51 . GSK3 can regulate the activity of GS to synthesis Glycogen 55,56 . Insulin can inhibit GSK3 to activate GS by, and then activate PP1 57 . PP1 is an important regulator in blood glucose levels as well as glycogen metabolism in the liver 57 . And GK act as a glucose sensor by catalysing the phosphorylation of glucose to glucose-6-phosphat in carbohydrate metabolism 58 . In our work, PP1, GSK-3β and GK mRNA levels were upregulated significantly by PfILP (Fig. 7). www.nature.com/scientificreports www.nature.com/scientificreports/ These results suggest that PfILP regulates glycogen metabolism pathways like vertebrate ILP. Furthermore, cell cycle analysis showed that PfILP influenced 293 T cell growth by promoting cell cycle progression (Fig. 8).
In summary, our discovery of PfILPs and six effectors in a mollusc species strengthens the hypothesis that the insulin signaling pathway has a common ancestor between vertebrates and invertebrates. Our findings can help us better understand the regulatory mechanisms and biological roles of insulin-like signalling pathways in P. fucata martensii and other invertebrates. www.nature.com/scientificreports www.nature.com/scientificreports/

Materials and Methods
Pearl oysters (2-year-old) were obtained from the Daya Bay Marine Biology Research Stations of the Chinese Academy of Sciences (Shenzhen, Guangdong, P.R. China) and maintained in an aerated indoor cement ponds for 1 week under controlled temperature (20 ± 2 °C) and fed Chlorella vulgaris every day. Animal experimentation in this study was approved by Experimental Animal Ethics of Chinese Academy of Sciences. RNA extraction. Total RNA were extracted using a TRIzol (Magen, Guangzhou, China). The integrity of RNA was determined by a 1% agarose gel electrophoresis. The quality and quantity of RNA were measured with a Quawell Q5000 (Thermo, California, USA). First-strand cDNA synthesis was performed using a SMART RACE cDNA Amplification Kit (Clontech, Palo Alto, CA, USA).
For analysis of gene expression in different tissues, mantle, digestive gland, gonad, adductor muscle, foot, heart, and gill were extracted as described above. The embryo development stages samples-the polar body, 2-cell, 32-cell, blastocyst, trochophore, D-shaped larvae, and umbo larvae stage were also extracted as above. Nine pearl oysters were randomly divided into three replicates, and equal quantities (500 ng) of total RNA from different tissues were reverse-transcribed into cDNA templates for ReverTra Ace qPCR using RT Master Mix with gDNA Remover (Toyobo, Osaka, Japan) according to the manufacturer's instructions.
Cloning of ILP signalling elements. To clone ILP signalling elements from P. fucata martensii, multiple pairs of primers (Supplementary Table S4) were designed according to partial fragments from the transcriptome and used to amplify full-length cDNAs. PCR products were cloned into the pGM-T Fast Vector (TIANGEN, Beijing, China) and the resulting constructs were transformed into Escherichia coli DH5α cells for DNA sequencing.
Real-time quantitative PCR (qPCR) analysis. Gene expression levels was determined by qPCR using a Roche LightCycler480 instrument (Roche, Basel, Switzerland). The 18S rRNA gene was used as an internal control, and primers were used for qPCR are listed in Supplementary Table S4. QPCR was performed using SYBR Green (Toyobo) and the expression levels were calculated based on the 2 −ΔΔCT method. Each sample was tested with three replicates. The primers used for qPCR are listed in Supplementary Table S4. Plasmid construction, and expression and purification of recombinant PfILP. The cDNA encoding PfILP mature polypeptide was amplified using sequence specific primers (Supplementary Table S4) containing EcoR I and Hind III restriction sites. After double digestion with EcoR I and Hind III, the products was cloned in-frame into the EcoR I/Hind III sites of the pET-28α expression vector to generate an N-terminal polyhistidine-tagged version of PfILP lacking its signal peptide (amino acids 1-32). The recombinant plasmid were transformed into E. coli Transetta (DE3) cells which grown overnight in liquid Luria-Bertani (LB) medium supplemented with 100 mg/ml kanamycin and 100 mg/ml penicillin under constant shaking (220 rpm) at 37 °C. The cultures was then diluted 1:50 into fresh LB with penicillin and kanamycin, and grown until an OD 600 of 0.5-0.7 under the same conditions. The bacterial was then induced with 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) and incubation continued at 37 °C or a further 12 h. The bacterial cells were pelleted by centrifugation at 8,000 × g for 5 min. www.nature.com/scientificreports www.nature.com/scientificreports/ Primary culturing of mantle cells. Oysters were maintained in a sterile seawater tank containing 0.024 g/l ampicillin and 0.02 g/l gentamicin for two days prior to experimentation. The outside of the shell was rapidly wiped with 72% ethanol. Shells were opened and the mantle was dissociated. The primary mantle cells were cultured according to a previous study 60 . Cell Counting Kit-8 assays and PCNA expression analysis. Cell Counting Kit-8 assays were performed to test the effects of recombinant PfILP on mantle cell growth. Mantle cells (2000 cells/well) were isolated and seeded in 96-well plastic plates and cultured with 2.5% fetal calf serum. After culturing for 24 h, various concentrations (0, 0.25, 0.5, 1, 1.5, and 2 μg/ml) of recombinant PfILP was added and culturing continued for a further 24 h. Cell proliferation was then assayed using the Cell Counting Kit-8 following the manufacturer's instructions. PCNA (Proliferating cell nuclear antigen, from pearl oyster) gene expression levels was measured as an indicator of cell proliferation status.
Effects of PfILP on PfIRR and Western blot assays. Co-immunoprecipitation (Co-IP) analysis of the interaction between PfILP and PfIRR was carried out using the Pierce Classic IP Kit (#26146; Pierce, Rockford, IL, USA) according to the manufacturer's instructions. Mantle cells were treated with 1 mM PfILP overnight at 4 °C, and Mantle cells were lysed and immunoprecipitated using a Pierce Classic IP Kit (#26146; Pierce, Rockford, IL, USA). Lysates were co-immunoprecipitated using anti-His-PfILP antibodies, subjected to SDS-PAGE (10%) and detected by western blotting with an anti-pfIRR antibody (10 μg) from our lab. Co-IP lysates obtained using anti-pfIRR antibodies (10 μg) were detected by western blotting with anti-His-PfILP (10 μg). Relative expression levels of Pfirr were analysed after a 24 h incubation with PfILP at various concentrations. pfIRR protein expression levels were analysed after a 1 h incubation with PfILP at various concentrations, or with 1 μg/ml PfILP for various times, by western blotting according described previously 60 . Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a reference.
Cell cycle analysis. Because mantle primary cells display weak cell proliferation, we used human 293 T cells instead, and these cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum at 37 °C with 5% CO 2 . After 3 days, 293 T cells (5 × 10 4 cells/well) were seeded in 24-well plates (Falcon, Corning, NY, USA). When cells had grown to 80%, the medium was discarded, and different concentrations of recombinant PfILP protein were added with 2% serum medium and cultured at 37 °C with 5% CO 2 for 36 h. Phosphate-buffered saline (PBS) was used instead of recombinant PfILP protein in the negative control group. Samples were collected and washed twice with PBS, resuspended in 1.5 ml of pre-cooled 80% ethanol, mixed, and incubated overnight at 4 °C in the dark. Samples were collected and washed twice with PBS, then mixed with 0.5 ml FxCycle PI/RNase Staining Solution (Thermo Fisher Scientific) and incubated at room temperature for 30 min in the dark. Samples were collected and filtered through a 40 μm cell strainer and analysed by FACScan flow cytometry (Philadelphia, Pennsylvania, USA) at 488 nm. Statistical analysis. All data graph were analysed by GraphPad Prism 6 (GraphPad Software, La Jolla, CA, USA). Values were displayed the means ± SEM and analyzed by Student's t-test. SPSS software (version 18.0) (SPSS, Chicago, IL, USA) was used for the statistical analysis. A p-value < 0.05 or <0.01 was considered statistically significant.

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.