Chemicals and reagents were purchased from Fisher Scientific (Pittsburgh, PA, USA), unless otherwise stated. RNA polymerase, RNase-free DNase, and pGEM-T easy vector were purchased from Promega (Madison, WI, USA). Taq DNA Polymerase, Phusion High-Fidelity DNA Polymerase, and restriction endonucleases were purchased from New England BioLabs (Beverly, MA, USA). M-MLV reverse transcriptase, oligo(dT)_12–18 primers, and TRIzol^® Reagent were purchased from Invitrogen Life Technologies, Inc. (Carlsbad, CA, USA). SMART™ RACE cDNA Amplification Kit was purchased from Clontech Laboratories, Inc. (Mountain View, CA, USA). mMESSAGE mMACHINE kit was purchased from Applied Biosystems/Ambion (Austin, TX, USA). IGF peptides were purchased from GroPep (Adelaide, SA, Australia) and IGFBP-3 and BMP-2 were from R&D systems (Minneapolis, MN, USA). The anti-GFP antibody was purchased from Torrey Pines Biolabs Inc. (East Orange, NJ, USA). Anti phospho-Smad1/5/8 antibody was purchased from Cell Signaling Technology Inc. (Danvers, MA, USA). Total anti-Smad1/5/8 antibody was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). Digoxigenin and anti-digoxigenin-POD antibodies were purchased from Roche (Indianapolis, IN, USA).

The sea lamprey (P. marinus) tissues were provided by Dr. Weiming Li, Michigan State University. Total RNA was extracted from the adult liver and used for cDNA synthesis. Wild-type zebrafish (Danio rerio) were maintained on a 14-h light/10-h dark cycle and fed twice a day. Embryos were obtained by natural cross. Fertilized eggs were raised at 28.5°C and staged according to Kimmel et al. (15). 2-Phenylthiourea [0.003% (wt/vol)] was added to the embryo rearing solution in some experiments to inhibit embryo pigment formation. This study was approved by the Institutional Committee on the Use and Care of Animals, University of Michigan. It was carried out in accordance with the recommendation of the Institutional Committee on the Use and Care of Animals, University of Michigan. The protocol was approved by the Institutional Committee on the Use and Care of Animals, University of Michigan.

The human IGFBP-3 (NM_000598) and zebrafish Igfbp-3 (NM_205751) amino acid (aa) sequences were used as queries to tblastn search the sea lamprey genome database.¹ The best BLAST hits were collected, and their scaffold locations were determined. The genome sequence surrounding these best reads of short sequence was retrieved and used to predict the sea lamprey IGFBP-3 sequence using GENESCAN.² The predicted protein sequence was searched by blastp.³ The full-length sequence of sea lamprey IGFBP-3 cDNA was determined by 5′- and 3′-rapid amplification of cDNA ends (RACE) using the SMART™ RACE cDNA Amplification Kit. The gene specific primers used were 5′GSP1 and 5′NGSP1 for 5′ end amplification, and 3′GSP2 and 3′NGSP2 for 3′ end amplification (Table S1 in Supplementary Material). The genomic structure of lamprey igfbp-3 was determined by the Blat program⁴ using the cloned full-length cDNA sequence and lamprey Assembly WUGSC 3.0.

The aa sequence of full-length lamprey IGFBP-3 and that of major IGFBPs (Table S2 in Supplementary Material) was aligned using BioEdit 7.0 (16). The same sequence information was used for subsequent phylogenetic tree construction. The following three bootstrap-supported tree construction methods were used: maximum likelihood (ML), Bayesian posterior probability (BayPP), and neighbor joining (NJ). The ML analysis was performed using ML 3.0 (17). ProtTest 2.4 was used to select the optimal model of aa substitution (18). The robustness of the ML was estimated by 1,000 bootstrap replications using the Jones–Taylor–Thornton (JTT) probability model. BayPP analysis was performed using MrBayes v.3.1 (19), assuming a four-category γ among site rate variation distribution, with uniform priors over trees, branch lengths (0.5), and the ASRV α parameter (0.05–10). A total of 1,000,000 generations were performed with four chains (Markov chain Monte-Carlo) and sampled for every 100 generations. The first 250 samples from each run, a point well past stationarity, were discarded as burn-in. All analyses converged on the same tree and found that the Jones protein model had 100% posterior probability. The NJ tree was constructed using MEGA4 (20) with JTT matrix in NJ method. The reliability of each tree node was assessed by the bootstrap method with 1,000 replications. The constructed trees were then visualized by TreeView (21).

For synteny analysis, lamprey, zebrafish, and human IGFBP-3/igfbp-3 gene, and their neighboring genes were extracted from Ensembl (Table S3 in Supplementary Material), and a schematic diagram was constructed to show locations of genes on each chromosome or scaffold.

The lamprey igfbp-3 open reading frame sequence was amplified by PCR using primers F1 and R1 (Table S1 in Supplementary Material) and cloned into pGEM-T easy vector and sequenced. It was then subcloned into pCS2+ and pCS2 + EGFP vectors using primers F2/R2 and F2/R3 (Table S1 in Supplementary Material), respectively. The lamprey IGFBP-3 IGF-binding domain (IBD) mutant was generated by changing R⁸⁶, P⁸⁷, L⁸⁸, L⁹¹, and L⁹² to S, A, S, Q, and G using the primers IBDF and IBDR (Table S1 in Supplementary Material), according to Imai et al. (22). The nuclear localization signal (NLS) mutant was generated by mutating ²⁴³RGRQR²⁴⁷ into MDGEA, the corresponding sequence of human IGFBP-1 (23), using primers NLSF and NLSR (Table S1 in Supplementary Material). The transactivation domain (TAD) mutants were generated by changing E²⁵/D²⁸/E⁴⁸/P⁴⁹/E⁷⁰ into the corresponding residues of human IGFBP-1 (24) by three rounds of mutagenesis using primers TAD1F/TAD1R, TAD2F/TAD2R, and TAD3F/TAD3R (Table S1 in Supplementary Material).

To analyze their possible TA activity, lamprey IGFBP-3 and its N-domain were fused to Gal4 DNA-binding domain (DBD). Briefly, DNA fragments encoding lamprey IGFBP-3 and its N-domain were amplified by PCR using primers shown in Table S1 in Supplementary Material. The PCR products were cloned into the BamHI site of the pBind vector, thereby fusing the fragments in frame to the C-terminus of Gal4DBD.

Human embryonic kidney (HEK) 293T cells, U2 osteosarcoma (U2OS) cells, and HeLa cells, purchased from the American Type Culture Collection (Manassas, VA, USA), were cultured as reported previously (25). HEK293 cells were used in cell proliferation and one-hybrid assays because they are fast growing and easy to transfect. HeLa cells are flattened cells with a large nucleus, and they were used in the subcellular localization experiment. U2OS cells have been previously used to study BMP2 signaling (26).

For one-hybrid transcription activation assay, 500 ng of pBind or pBind-IGFBP DNA and 500 ng of pG5-luc DNA were co-transfected into HEK293T cells as reported previously (24). Twenty-four hours after transfection, cells were washed with PBS and lysed in lysis buffer. The firefly and Renilla luciferase activities were determined using the Dual-Luciferase Reporter assay system (Promega). For subcellular localization studies, EGFP, lamprey IGFBP-3, and its various mutants–EGFP fusion constructs were transfected into HeLa cells. Transfected cells were fixed by PFA, stained by DAPI, and then imaged as previously reported (26).

To obtain the CM, lamprey IGFBP-3, its various mutants, or human IGFBP-3 plasmid DNA (2 µg) were transfected into HEK 293T cells. Twenty-four hours later, cells were washed three times with PBS and changed to serum-free medium. After incubation for 48 h, the CMs were collected and centrifuged to remove cellular debris. These CMs were subjected to western immunoblotting and ligand blotting analyses as described previously (26). Human IGF-I was labeled with digoxingenin following the published methods (27).

To determine the possible IGF-dependent action of lamprey IGFBP-3, HEK293T cells were plated into 96-well plates in DMEM supplemented with 10% fetal bovine serum. After incubated for 1 day, cells were rinsed three times with serum-free DMEM. One hundred microliters of CMs collected from cells transfected with lamprey IGFBP-3, its IBD mutant, or human IGFBP3 were mixed with 100 µl of DMEM with or without IGF-I (with a final concentration of 100 ng/ml) and incubated for 30 min at room temperature. The mixed media were added to cells. Forty-eight hours later, cell numbers were assayed using a cell counting kit from ZOMANBIO (Beijing) following the manufacturer’s instruction. To study the effect of lamprey IGFBP-3 on IGF-I-induced pAkt signaling, HEK293T cells were plated into 24-well plates and treated as described above. The cells were harvested after 3 h of treatment. The cell lysates were analyzed by immunoblotting using antibodies against pAkt and total Akt.

Human U2OS cells were seeded into 24-well plates. After 24 h, the cells were washed and incubated in serum-free medium for 3 h. Then, 100 µl of DMEM containing human BMP2 (20 ng/ml) were mixed with 100 µl of CMs as described and added to the cells. After 3 h, cells were harvested in RIPA buffer containing protease and phosphatase inhibitors (10 µg/ml aprotinin, 10 µg/ml leupeptin, 10 µg/ml pepstatin, 1 mM PMSF, 1 mM sodium orthovanadate, and 1 mM sodium fluoride) and subjected to immunoblotting using antibodies against pSmad1/5/8 and total Smad1/5/8.

Capped mRNA was synthesized using a commercial mMESSAGE mMACHINE kit and linearized plasmid DNA as template. Two hundred picograms of mRNA encoding EGFP, lamprey IGFBP-3, and its various mutants, 50 pg of mRNA encoding zebrafish Chordin, and 2.5 pg of mRNA encoding zebrafish Bmp2b were microinjected into zebrafish embryos at one to two cell stages, respectively. Intact embryos were used as controls. After microinjection, embryos were raised in embryo rearing medium (28) and kept at 28.5°C. Live embryos were imaged, and body length and somite number were determined at 24 hpf.

For qRT-PCR analysis, total RNA was extracted using TRIzol^® Reagent following the manufacturer’s instruction. After DNase treatment, RNA (1 µg) was reverse transcribed with oligo(dT)₁₈ primers and M-MLV reverse transcriptase. qRT-PCR was performed in an iCycler iQ Multicolor real-time PCR detection system (Bio-Rad Laboratories, Inc., Hercules, CA, USA) using SYBR^® Premix Ex Taq™ (Takara Bio Inc., Dalian, Liaoning). Primer pairs used for qPCR were as follows: tp63F/tp63R for tp63 and actinF/actinR for β-actin (Table S1 in Supplementary Material). Each sample was measured in duplicate. Target gene mRNA levels were calculated using the 2^−ΔΔCT method (29) and presented as relative (-fold) values of the control group after being normalized by β-actin mRNA levels.

Values are presented as means ± SD. Differences among groups were analyzed using one-way ANOVA followed by the Turkey’s multiple comparison test. All analyses were performed using GraphPad Prism version 5.01 (San Diego, CA, USA). Significance was accepted at P < 0.05 or higher.

Our search of the sea lamprey genome database identified an igfbp-like gene (ENSPMAG00000004139) on scaffold GL476421 (previously known as Contig6676.1). Using RACE, the full-length cDNA was obtained and sequenced. The genomic structure was determined. This lamprey gene contains four exons and three introns (Figure 1A). There are two polyadenylation signal sequence (Figure 1A). Thus, the full-length cDNAs are 2,111 bp (KX904869) and 1,894 bp (KX904870), respectively. The putative signal peptide is 32 aa and mature protein is 278 aa with a predicted molecular mass of 29.8 kDa. There are 12 cysteine residues in the highly conserved N-terminal domain (residues 1–101) and 6 cysteine residues in the conserved C-terminal domain (residues 200–278) (Figure 1B). A conserved IGFBP motif is present in its N-domain and a thyroglobulin type 1 repeat in the C-domain (Figure 1B). This protein has the highest sequence identities with IGFBP-3/Igfbp-3 in the IGFBP family (Table S4 in Supplementary Material). Phylogenetic analyses have placed this lamprey IGFBP in the IGFBP-3 clade (Figure 1C; Figures S1 and S2 in Supplementary Material). Therefore, we named this gene as lamprey igfbp-3. The lamprey igfbp-3 is located on scaffold GL476421 (Figure 1D). In human and zebrafish genomes, IGFBP-3/igfbp-3 gene is located next to IGFBP-1/igfbp-1, arranged with a tail-to-tail orientation (Figure 1D). We were unable to find any igfbp-1-like gene in lamprey scaffold GL476421. An ACDY1 orthologous gene was found in GL476421 in this scaffold (Figure 1D).

The IGF-binding domain (IBD), TAD, and NLS are found in lamprey IGFBP-3 (Figure 1B). To determine whether these motifs are functional, they were mutated and tested functionally (Figure S3 in Supplementary Material). Ligand blotting assay showed that lamprey IGFBP-3, but not its IBD mutant, was able to bind human IGF-I (Figure 2A). The NLS and TAD mutants were also able to bind IGF-I (Figure 2A). The expression levels of these fusion proteins were similar (Figure 2A). When expressed in HeLa cells, the lamprey IGFBP-3-GFP signal was observed exclusively in the nucleus (Figure 2B). The IGFBP-3 IBD mutant and TAD mutant had similar nuclear. Mutation of NLS, on the other hand, abolished the nuclear localization (Figure 2B), suggesting that the lamprey IGFBP-3 NLS is functional. Next, one-hybrid TA assay was performed. Lamprey IGFBP-3 N-domain (BP3-N) caused a 16-fold increase in GAL4-dependent TA activity (Figure 2C). As the case in human IGFBP-3 (30), the full-length lamprey IGFBP-3 had no such activity, indicating that its activity is masked by its N-domain (Figure 2C). Next, the lamprey TAD sequence was mutated to corresponding sequence in human IGFBP-1, which had no TA activity (24). The TAD mutant had significantly reduced TA activity (Figure 2C). These results indicate that the IBD, NLS, and TAD in lamprey IGFBP-3 are functional.

We tested whether lamprey IGFBP-3 has IGF-dependent action in cultured human cells because it is technically difficult to perform functional studies in lamprey. Lamprey IGFBP-3 and its IBD mutant expression plasmids were transfected into HEK293T cells. Human IGFBP-3 plasmid and empty vector transfected cells were used as controls. Addition of IGF-I significantly stimulated HEK293T cell proliferation (Figure 3A). Co-incubation with lamprey IGFBP-3 inhibited IGF-I-induced cell proliferation (Figure 3A). The inhibitory effect was comparable to that of human IGFBP-3 (Figure 3A). Likewise, both lamprey and human IGFBP-3 inhibited IGF-I-induced Akt phosphorylation (Figure 3B). The lamprey IGFBP-3 IBD mutant had no effect on IGF-I-induced cell proliferation or pAkt signaling (Figures 3A,B), suggesting that this action of lamprey IGFBP-3 requires its IBD domain.

We next investigated whether lamprey IGFBP-3 has IGF-dependent action in vivo in zebrafish embryos by mRNA injection experiments. As shown in Figure 4A, the EGFP mRNA-injected embryos were indistinguishable from the WT embryos. Forced expression of lamprey IGFBP-3 significantly decreased body length (Figures 4A,B) and somite number (Figures 4A,C). Similar effects were seen with the TAD and NLS mutants (Figures 4A–C). Mutation of the IBD abolished this action (Figures 4A–C), suggesting that this action is primarily IGF-binding dependent.

We have previously reported that human and zebrafish IGFBP-3/Igfbp-3 can antagonize BMP action, and this action is independent of its IGF-binding function (26). To investigate whether lamprey IGFBP-3 has this activity, co-incubation experiments were carried out. Addition of human BMP-2 strongly induced Smad1/5/8 phosphorylation in cultured human U2OS cells (Figure 5A). Co-incubation with CMs containing lamprey IGFBP-3, zebrafish Igfbp-3, or human IGFBP-3 strongly inhibited BMP-2-induced Smad1/5/8 phosphorylation (Figure 5A). The CMs from the empty vector transfected cells had no such effect (Figure 5A).

Next, mRNA injection experiments were carried out. Injection of Bmp2b mRNA resulted in an eightfold and highly significant increase in the mRNA levels of tp63 (Figure 5B), a known Bmp target gene (31). Injection of Bmp2b with Chordin mRNA, a well-known Bmp antagonist, significantly reduced tp63 mRNA expression (Figure 5B). Likewise, injection lamprey IGFBP-3 mRNA resulted in a significant decrease in tp63 mRNA expression (Figure 5B). This effect was comparable to those of human IGFBP-3 and zebrafish Igfbp-3 (Figure 5B). Importantly, the lamprey IGFBP-3 IBD mutant had similar effect (Figure 5B). These results suggest that lamprey IGFBP-3 has the ability to antagonize Bmp signaling and this action is IGF independent.