Zhipeng Tai, Shicheng Ling, Jie Cheng, Xiaoying Tan

Key Laboratory of Freshwater Animal Breeding, Ministry of Agriculture, Fishery College, Huazhong Agricultural University, Wuhan, 430070, China

Keywords:

Bmp family

Gene characterization

Tissue expression

Pelteobagrus fulvidraco

High fat diet

ABSTRACT

In this study, three members of the Bmp family were cloned and characterized in yellow cat fish Pelteobagrus fulvidraco, including Bmp2a, Bmp4 and Bmp9. The predicted amino acid sequences of P. fulvidraco Bmp2a, Bmp4 and Bmp9 showed the characteristic domains of the Bmp family, including an N-terminal signal peptide, Arg-X-XArg site, TGF-β family signature and seven conserved cysteines, indicating that function is likely to have been conserved during evolution. mRNAs of the three Bmp genes had a variable level of expression in tissues.Compared to the control diet, a high fat diet tended to down-regulate the mRNA expression of Bmp2a, Bmp4 and Bmp9 in mesenteric fat, liver and ovary, while it tended to up-regulate their mRNA levels in muscle and kidney.The responses to dietary lipid status and the potential role in lipid metabolism have not previously been reported and reinforces the idea of their multiple functions. Our findings provide the first data about the potential role of the Bmp family in lipid metabolism in teleost.

1.Introduction

Bone morphogenetic proteins (Bmps) belong to the transforming growth factor-β (TGF-β) superfamily. They were originally identified by their ability to induce the formation of bone and cartilage (Reddi &Reddi, 2009). At present, accumulating data suggests that Bmps play a role in various physiological processes, such as embryogenesis, muscle growth, adipogenesis and development (Chen et al., 2017; Wang et al.,2014; Zhang, Lan, Nie, Guan, & Gao, 2018). To date, more than 30 Bmps have been identified (Ducy & Karsenty, 2000; Zhang et al., 2018).Among these members,

Bmp2

,

Bmp4

and

Bmp9

effectively induced both adipogenic and osteogenic differentiation in mammals and their cell lines (Kang et al., 2009).

Bmp2

induces osteogenesis and chondrogenesis and prevents terminal differentiation of myogenic cells by inhibiting the transcription of the muscle-specific nuclear factors MyoD and myogenin(Katagiri et al., 1997).

Bmp4

plays a key role in regulating adipogenic precursor cell commitment and differentiation (Gustafson et al., 2015),and

Bmp9

has been correlated with liver disease (Herrera, Dooley, &Breitkopf-Heinlein, 2014).The Bmp family is relatively well studied in mammals, but largely unstudied in fish. Full-length cDNA sequences have been reported in several fish including, zebra fish

Danio rerio

(

Bmp2a

and

Bmp4

,Martı́nez-Barberá, Toresson, Da, & Krauss, 1997),

Sparus auratu

s (

Bmp2

,Rafael, Laizé, & Cancela, 2006), Senegalese sole

Solea senegalensis

(

Bmp2

and

Bmp4

, Marques et al., 2014) and blunt snout bream

Megalobrama amblycephala

(

Bmp2a

and

Bmp4

, Zhang et al., 2018). Regarding their function, Mowbray, Hammerschmidt, and Whit field (2001) described the mRNA expression patterns of

Bmp2b

and

Bmp4

in the developing zebra fish and provided evidence that Bmps were essential for the development of the vertebrate ear. Marques et al. (2014, 2016) reported that

Bmp2

and

Bmp4

were expressed in calcified tissues.Lipid is an essential nutrient which provides energy and essential fatty acids for the growth and development in fish (Izquierdo, Socorro,Arantzamendi, & Hernández-Cruz, 2000; Sargent et al., 1999). Recently,Zhong et al. (2018) found that some members of the Bmp family, such as

Bmp8a

, increased the expression levels of genes associated with the regulation of lipid metabolism in zebra fish, indicating its potential regulatory role in lipid metabolism. However, it is unknown if other members, such as

Bmp2

,

Bmp4

and

Bmp9

have a regulatory roles in lipid metabolism in fish. Furthermore, if nutritional status affects gene expression is unknown in fish.In the present study, the full-length cDNA sequences of

Bmp2a

,

Bmp4

and

Bmp9

were cloned and characterized, and their tissue expressions were explored in yellow cat fish

P. fulvidraco

, a widely cultured freshwater omnivorous fish in China and other Asian countries. The transcriptional response of

Bmp2a

,

Bmp4

and

Bmp9

to a high fat diet in different tissues were investigated in

P. fulvidraco

. The present study extends our understanding of the physiological role of these genes and provides the first reports for their roles in the regulation of lipid metabolism in fish.

2.Materials and methods

Two experiments were conducted. The first experiment was the cloning and tissue expression profile analysis of

Bmp2a

,

Bmp4

and

Bmp9

.The second experiment was designed to evaluate the changes of mRNA levels of

Bmp2a

,

Bmp4

and

Bmp9

genes in different tissues when fish received a high fat diet. The experiments performed on animals followed the ethical guidelines of Huazhong Agricultural University and all experimental protocols were approved by Huazhong Agricultural University.

2.1.Experiment 1: cloning of Bmp2a, Bmp4 and Bmp9 genes and their mRNA tissue expression

P. fulvidraco

(23.5 ± 3.3 g, mean ± SEM) for cloning and exploring mRNA tissue expression of the

Bmp2a

,

Bmp4

and

Bmp9

genes were obtained from a local fishpond (Wuhan, China). The culture and management of yellow cat fish were similar to those described in our recent publications (Wei et al., 2017). Tissues (liver, brain, intestine, heart,spleen, kidney, gill, muscle, ovary and mesenteric fat) were immediately prepared for RNA isolation and expression profile analysis. RNA isolation, synthesis of cDNAs and cloning of

Bmp2a

,

Bmp4

and

Bmp9

genes were similar to that previously described (Chen et al., 2016; Wei et al.,2017). Degenerate primers (Supplementary Table 1), based on the most conserved regions of fish Bmp2,

Bmp4

and

Bmp9

sequences available in the GenBank and the Ensembl database, were used to amplify partial cDNA fragments. The 3and 5UTR sequences were obtained through nested 3and 5RACE PCR performed with a SMART RACE cDNA Amplification Kit (Clontech, USA) using the manufacturer manual.Gene-specific primers were designed for 3and 5RACE PCR (Supplementary Table 1).

Table 1The sequence information of Bmp2a, Bmp4 and Bmp9 from P. fulvidraco.

The core fragment, 3end and 5end sequences were assembled using SeqMan II software in DNASTAR PACKAGE to obtain the fulllength cDNA sequences of

Bmp2a

,

Bmp4

and

Bmp9

. The sequences were edited and analyzed using the program EDITSEQ of the DNASTAR and the open reading frame (ORF) identified and the amino acid sequence of each Bmp determined. The nucleotide sequence for each Bmp were compared with DNA sequences present in GenBank using the BLAST network service at the NCBI (http://blast.ncbi.nlm.nih.gov/).Sequence alignments and percentage of amino acid conservation were assessed with the Clustal-W multiple alignment algorithm. Domains were analyzed using the SMART program (http://www.smart.embl-he idelberg.de/) and online CDD tool at NCBI (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi). The phylogenetic trees were generated through the neighbor-joining (NJ) method with MEGA 5.0 (Tamura et al., 2011) using a JTT +G model (Jones, Taylor, & Thornton, 1992),and the best-fit model for sequence evolution was obtained with the ML model selection. Bootstrap sampling was reiterated 1000 times.

2.2.Experiment 2: transcriptional response of Bmp2a, Bmp4 and Bmp9 in different tissues to a high fat diet

For this experiment, two experimental diets were formulated with dietary lipid levels of 11.3% (control) and 15.4% (high fat diet, HFD), as described in Ling, Wu, Zhang, and Luo (2019) and Chen et al. (2019).Each diet was assigned to three tanks in a completely randomized design, with 6 tanks for the experiment. At the initiation of the feeding study, 30 similarly-sized fish (mean initial weight: 3.79 ±0.16 g) were stocked in each fiberglass tank, and were fed to satiation twice daily and managed as described above.

At the termination of the 8-wk feeding study, all fish were fasted for one day before sampling. Fish were euthanized with tricaine methane sulphonate (MS-222 at 100 mg/L) and three fish per aquarium were collected. Five tissues (mesenteric fat, liver, ovary, muscle and kidney)were quickly dissected out and frozen in liquid nitrogen, and stored at−80 °C for until analysis.

The mRNA levels were assayed by qPCR as previously described(Chen et al., 2016). The primer sequences of each analyzed gene are given in Supplementary Table 2. A set of ten reference genes (18S rRNA,RPL7, β-actin, HPRT, TUBA, B2M, TBP, GAPDH, ELFA and UBCE) were selected from the literature (Zhao, Gul, Li, & Wang, 2011) in order to test their transcription stability. The relative expression of genes was calculated using the 2method (Livak & Schmittgen, 2001) and normalized to the geometric mean of the best combination of two genes suggested by geNorm (Vandesompele et al., 2002). Prior to the analysis,experiments were performed to check the stability of the reference genes and ELFA and β-actin were used as the reference for the tissue distribution of Bmp. UBCE and β-actin were used as the reference for the high fat diet experiment.

Table 2Amino acid sequence identity of Bmp2a, Bmp4 and Bmp9 from P. fulvidraco and other species (%).

2.3.Statistical analysis

Quantitative data were expressed as the means ±SEM. Prior to statistical analysis, all data were tested for normality of distribution using the Kolmogorov-Smirnov test. The homogeneity of variance among the different tissues was tested using the Barlett's test. Data were then subjected to One-Way ANOVA and Tukey's multiple range tests.Differences between the control and HFD group were analyzed by a Student's t-test for independent samples. All statistics were performed using SPSS 19.0 for Windows (SPSS, Chicago, IL, USA).

3.Results

3.1.Molecular characterization of Bmp2a, Bmp4 and Bmp9 cDNA sequences

The isolated full-length cDNA sequences of

Bmp2a

,

Bmp4

and

Bmp9

were 1970 bp, 1982 bp and 1691 bp, respectively (Table 1). Sequence analysis revealed the ORF of

Bmp2a

,

Bmp4

and

Bmp9

encoded proteins of 418 amino acids, 408 amino acids, and 395 amino acids, respectively.Alignment of predicted polypeptide sequences showed that the amino acid sequences of

P. fulvidraco Bmp2a

,

Bmp4

and

Bmp9

were similar to those from other fish and mammals, exhibiting 41.0—89.0%,72.1—95.3% and 39.5—86.8% amino acid sequence identities, respectively (Table 2). The amino acid sequence of

P. fulvidraco Bmp2a

,

Bmp4

and

Bmp9

possessed the characteristic features of BMPs, including N-terminal signal peptide, Arg-X-X-Arg site, TGF-β family signature and seven conserved cysteines (Fig. 1).

Fig. 1.ClustalX alignment of the deduced amino acid sequence of Bmp2a, Bmp4 and Bmp9 from P. fulvidraco. Accession numbers are as follows (Bmp2a, KY930914;Bmp4, KY930915 and Bmp9:, KY930916). The identical residues are shaded in dark gray. The N-terminal signal peptide is underlined. The proposed Arg-X-X-Arg sites after which the mature protein is cleaved is in boxed in dotted lines; the TGF-β family signature is boxed in with solid lines. Black arrowheads indicate the conserved cysteines in the mature protein.

The phylogenetic analysis showed that

Bmp2a

clade was closer to the

Bmp4

clade, which was clearly separated from the

Bmp9

clade. In each clade, the branching of the phylogenetic tree fit with established taxonomic relationships. All teleost

Bmp2a

,

Bmp4

and

Bmp9

genes formed an independent cluster, whereas amphibian and mammalian proteins formed another cluster. In all clades,

P. fulvidraco

was grouped with

Ictalurus punctatus

,

Astyanax mexicanus

and

Danio rerio

, which are all members of the Ostariophysi. They then formed a clade with the species of Acanthomorphata (

Poecilia formosa

,

Xiphophorus maculates

, and

Lepisosteus oculatus

) (Fig. 2).

Fig. 2.Phylogenetic tree based on the protein sequences of Bmps from P. fulvidraco and other vertebrate species using the neighbor-joining (NJ) method in MEGA 5.0(Tamura et al., 2011) based on the JTT +G model (Jones et al., 1992). Branch support values represent a percentage of 1000 bootstrap replicates.

3.2.mRNA tissue expression of Bmp2a, Bmp4 and Bmp9

The mRNA expressions of

Bmp2a

,

Bmp4

and

Bmp9

genes from

P. fulvidraco

were detected in all sampled tissues, but their mRNA abundance varied (Fig. 3).

Bmp2a

mRNA levels were the highest in gill,followed by liver and kidney, and showed no significant differences among other tissues (such as brain, intestine, heart, spleen, muscle,ovary and mesenteric fat).

Bmp4

mRNA levels were highest in the ovary,followed by mesenteric fat, heart, muscle, spleen, intestine and kidney,but was not significantly different among other tissues. The ranked order of expression of

Bmp9

mRNA was ovary

＞

muscle

＞

mesenteric fat

＞

brain

＞

liver =intestine =kidney =heart, and were the lowest in spleen and gill.

Fig. 3.qPCR analysis for Bmp2a, Bmp4, Bmp9 in liver (L), brain (B), intestine (I), heart (H), spleen (S), kidney (K), gill (G), muscle (M), ovary (O) and mesenteric fat(F). Data (mean ± SEM, n =6) were expressed relative to the geometric mean of the reference genes ELFA and β-actin. Expression of genes in the liver was considered to be 1 and gene expression in other tissues are relative to the liver. Bars that do not share a common letter are significantly different among the tissues (p ＜0.05).

3.3.Transcriptional responses of Bmp2a, Bmp4 and Bmp9 in several tissues of yellow cat fish to high fat diet (HFD)

In the mesenteric fat, HFD down-regulated the mRNA expression of

Bmp2a

,

Bmp4

and

Bmp9

(

p ＜

0.05) (Fig. 4A). In the liver, compared to the control, HFD down-regulated the expression of

Bmp2a

and

Bmp4

but showed no significant effects on

Bmp9

expression (

p ＜

0.05) (Fig. 4B). In the ovary, the mRNA expression of

Bmp2a

and

Bmp9

(

p ＜

0.05) were significantly lower in the HFD than those in the control, and the expression of

Bmp4

did not show significant differences (Fig. 4C). In the muscle, HFD up-regulated the mRNA expression of

Bmp2a

,

Bmp4

and

Bmp9

(

p ＜

0.05) (Fig. 4D). In the kidney, the expression of

Bmp4

and

Bmp9

were up-regulated in the HFD compared with the control, and mRNA expression of

Bmp2a

showed no significant differences between the two groups (Fig. 4E).

Fig. 4.Effect of dietary fat levels on mRNA expression of Bmp2a, Bmp4 and Bmp9 in different tissue of P. fulvidraco (A: mesenteric fat; B: liver; C: ovary; D: muscle; E:kidney). Data (mean ± SEM, n =3) are expressed relative to the geometric mean of the reference genes (UBCE and β-actin). An asterisk indicates significant differences between two groups (t-test, p ＜0.05). CD =control diet; HFD =high fat diet.

4.Discussion

In the present study, three genes previously shown to have a pivotal role in the development of vertebrates were cloned and characterized in the yellow cat fish for the first time. The amino acid sequence of

P. fulvidraco Bmp2a

,

Bmp4

and

Bmp9

possessed the characteristic features of Bmps, including an N-terminal signal peptide, Arg-X-X-Arg site,TGF-β family signature and seven conserved cysteines. The Arg-X-X-Arg sequence is a putative proteolytic processing site, which generates the mature protein was cleaved off (Cui, Jean, Thomas, & Christian, 1998).Seven conserved cysteines, which was regarded as a common feature of TGF-β superfamily (Martı́nez-Barberá et al., 1997; Miyashita et al.,2008; Zhang et al., 2018), was detected in all mature proteins of

Bmp2a

,

Bmp4

and

Bmp9

in yellow cat fish. Such structural similarity suggests they may have a similar function in vertebrates. Phylogenetic analysis suggested that

P. fulvidraco Bmp2a

,

Bmp4

and

Bmp9

peptides were closely related to other fish, which is consistent with the classical taxonomic classes.The mRNA expressions of

Bmp2a

,

Bmp4

and

Bmp9

genes from

P. fulvidraco

were detected at varying abundance in all sampled tissues,which indicates they may have a diversity of function as previously reported in vertebrates including fish (Rafael et al., 2006; Zhang et al.,2018; Bragdon et al., 2011). The present study indicated that

Bmp2a

mRNA levels were highest in gill, followed by liver and kidney, and showed no significant differences in other tissues (such as brain, intestine, heart, spleen, muscle, ovary and mesenteric fat). High mRNA expressions of

Bmp2

in the liver have been reported in several fish(Marques et al., 2014; Rafael et al., 2006; Zhang et al., 2018), suggesting an involvement in liver cell trans-differentiation in fish (Rafael et al.,2006). In the yellow cat fish,

Bmp4

mRNA levels were the highest in the ovary, followed by the mesenteric fat, heart, muscle, spleen, intestine and kidney, but was not significantly different among other tissues. Su and Dong (2018) reported that the expression of

Bmp4

mRNA was the highest in muscle of common carp. Regarding its mRNA expression and function in the ovary,

Bmp4

is reported to increase follicle-stimulating hormone-induced estrogen production and reduce FSH-induced progesterone production in cultured rat granulose cells (Shimasaki et al.,1999). Juengel et al. (2006) reported that

Bmp4

was found in the theca cells of the bovine antral follicles. Other study indentified relatively high levels of

Bmp4

gene expression in adult spleen and heart (Marques et al.,2016; Zhang et al., 2018). It is believed that the oocytes may be one of the sources of Bmp production, which may regulate follicle growth and development (Li & Ge, 2011). Chen et al. (2017) found that most of the Bmp genes were widely expressed, but with a relatively high expression level in gill, intestine, liver, spleen and brain in common carp.The function of Bmps in regulating adipogenesis is little studied in fish. However,

Bmp3

and

Bmp3b

expression were increased in adipose tissue of diet-induced obese mice as compared with that in control mice,and siRNA-mediated

Bmp3b

knockdown in 3T3-L1 cells enhanced adipogenesis (Hino, Miyazawa, Miyazato, & Kangawa, 2012). Conversely,overexpressing

Bmp3b

inhibited adipocyte differentiation (Hino et al.,2012). In the present study a HFD down-regulated the mRNA expression of

Bmp2a

,

Bmp4

and

Bmp9

in the mesenteric fat. Previous studies indicate the effects of Bmps on adipogenesis appear to depend on the stage of cell development and the dosage of different Bmp ligands (Schulz &Tseng, 2009). In the present study HFD down-regulated the expression of

Bmp2a

and

Bmp4

in the liver, the main organ of lipid metabolism(Zheng et al., 2015).Members of Bmp family have been reported to regulate gonadotropin receptors in chicken (Haugen & Johnson, 2010). Bmps produced by the oocytes bind to their receptors located in the follicle cells along with gonadotropin receptors. It has been suggested that Bmps from the oocytes may signal the follicle cells in a paracrine manner to regulate gene expression in these cells, including follicle-stimulating hormone receptor and luteinizing receptor (Li, Zhou, & Ge, 2012). The up-regulation of

Bmp2a

,

Bmp4

and

Bmp9

in the muscle and the expression of

Bmp4

and

Bmp9

in the kidney of HFD fish highlight the complexity and multiplicity of their response to diet. Thus, we speculated that the differential mRNA expressions of these genes might be attributable to the HFD-induced changes in lipid metabolism in the tissues of

P. fulvidraco

.In summary, we characterized the full-length cDNA sequences of

Bmp2a

,

Bmp4

and

Bmp9

and explored their tissue expression profiles in

P. fulvidraco

. The differential mRNA expression patterns of these genes in response to HFD in various tissues indicated a differential tissue response to a high fat diet. Our finding, suggest that

Bmp2a

,

Bmp4

and

Bmp9

may be involved in lipid metabolism in teleosts and form the basis for future investigation into the molecular mechanism for their regulatory roles in lipid metabolism.

CRediT authorship contribution statement

Zhipeng Tai: Investigation, Methodology, Formal analysis, Writing -original draft. Shicheng Ling: Investigation, Writing - review & editing.Jie Cheng: Formal analysis, Methodology, Writing - review & editing.Xiaoying Tan: Funding acquisition, Project administration, Supervision, Writing - review & editing.

Declaration of competing interest

All the authors stated no conflict of interest.

Acknowledgments

The study was funded by National Natural Science Foundation of China (Grant nos.: 31572605, 31001101) and Project of Innovative Group for Excellent Young Scientists in Universities of Hubei Province from Hubei Provincial Department of Education (T201933).

Appendix A.Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.aaf.2019.12.008.