Introduction

Lipoprotein lipase (LPL, EC 3.1.1.34) is a member of the lipase gene superfamily, which initially consisted of three mammalian lipase (pancreatic lipase, PL; lipoprotein lipase, and hepatic lipase, HL) based on amino acid sequence similarity and gene organization [1]. Family size increased when several proteins were subsequently added based on amino acid homology, including PL-related proteins 1 and 2, phosphatidylserine phospholiase A1, and endothelial lipase (EL). All these enzymes would have evolved from a common ancestor and they play a critical role in the absorption and metabolism of lipids and lipoproteins. Lipases are water-soluble enzymes that hydrolyze ester bonds of water-insoluble substrates such as triglycerides, phospholipids, and cholesteryl esters [2, 3].

Mammalian mature LPL is a glycoprotein that is primarily synthesized in adipocytes, muscle cells, macrophages, and lactating mammary gland, but not in the adult liver [4, 5]. In human, LPL gene encoded 475 amino acids, including a 27 amino acids signal peptide. The mature LPL protein consists of 448 amino acids after resecting the signal peptide, which molecular weight is 55 kDa [3]. The enzyme is active as a non-covalent homodimer and is bound to the surface of endothelial cells via glycosaminoglycans, from which it can be released by heparin [6]. LPL plays a key role in lipid metabolism by hydrolyzing triglycerides in chylomicrons and very low-density lipoproteins (VLDL). It thereby generates non-esterified fatty acids and 2-monoacylglycerol, which are utilized either for storage or energy production [7]. The abnormalities in LPL function have been found to be associated with a number of pathophysiological conditions, including atherosclerosis, chylomicronaemia, obesity, Alzheimer’s disease, and dyslipidaemia associated with diabetes, insulin resistance, and infection [6].

LPL full-length cDNA has been cloned and characterized in a number of fish, including zebrafish (Danio rerio) [8], rainbow trout (Oncorhynchus mykiss) [9, 10], gilthead sea bream (Sparus aurata) [1], red sea bream (Pagrus major) [11–13], and European sea bass (Dicentrarchus labrax) [14]. Furthermore, LPL expression has been measured in a variety of tissues of rainbow trout [9], red sea bream [12, 15, 16] and gilthead sea bream [1]. In contrast to mammals, LPL activity and expression have also been detected in the liver of adult fish [1, 9, 15, 16]. The effects of feeding condition and nutritional state on LPL gene expression in the liver, visceral adipose tissue and muscle were investigated. The regulatory effect of dietary lipid level on LPL gene expression was tissue-specific and related to feeding conditions. In red sea bream, when fed high-fat diets tended increase the LPL mRNA level in the liver, but had no effect in the visceral mesenteric adipose tissues. Starvation (at 48 h post-feeding) drastically stimulated LPL gene expression in the liver, but decreased it in visceral mesenteric adipose tissues [13, 15, 17, 18]. In gilthead sea bream, when fed plant-protein-based diets, hepatic LPL expression was up-regulated whereas an opposite trend was found in the mesenteric adipose tissue [1]. Fasting for 2 weeks provoked a clear decrease in adipose tissue LPL activity and LPL mRNA levels, while no effects were observed in red muscle [19].

Common carp (Cyprinus carpio var. Jian) is one of the most popular cultured fishes in China. Some diseases associated with lipid metabolism have been found to increase year after year in artificial culture. The aim of the present work was to clone the full-length cDNA of LPL gene from liver of adult common carp and analyze its tissue distribution to explore its role in lipid metabolism. The results will facilitate further studies for the structure, function and regulation of the gene expression of LPL in omnivorous fish.

Materials and methods

Fish

Adult common carp (C. carpio var. Jian, Cyprinidae, Cypriniformes, Teleostei), which body weight ranged from 750 g to 1,000 g, were collected from a pond in Lianyungang, Jiangsu province in May, 2007, and transported to the indoor tanks in the Huaihai Institute of Technology, Jiangsu province, China. The fish were maintained in a 1000L tanks under natural conditions of photoperiod fed commercial diet for 2 weeks (May–June). The water temperature ranged from 23 to 27°C.

Total RNA extraction from liver

The fish were killed by a blow to the head and liver were rapidly excised and ground sample under liquid nitrogen to a fine power using a mortar. Total RNA was extracted from approximately 50 mg of liver using EZ-10 Spin Column RNA Purification Kit (BBI), according to instructions from the manufacturer. The total RNA was treated using an on-membrane DNase I digestion Kit (OMEGA). The integrity of isolated RNA was assessed by 1% w/v formaldehyde denaturing agarose gels electrophoresis, its quantity and purity was determined by absorbance measures at 260 nm and 280 nm with an UV-Visible spectrophotometer GeneQuant pro (GE), and diluted to 0.5 μg/μl with DEPC-water, then stored at −80°C for future use.

Synthesis of the first-strand cDNA and isolation of core cDNA fragments of LPL

Reverse transcription was performed with total RNA from the liver of adult common carp as template and oligo(dT)18 as primer using an AMV First Strand cDNA Synthesis Kit (BBI), according to the manufacturer’s instructions. 20 μl reaction volume mixture composed of 5 μl total RNA(0.5 μg/μl), 1 μl Oligo(dT)18 (0.5 μg/μl), 5 μl RNase-free ddH2O, 4 μl 5× Reaction Buffer, 1 μl RNase inhibitor (20 U/μl), 2 μl dNTP Mix (10 mM each), 2 μl AMV Reverse transcriptase (10 U/μl). The resulting first strand cDNA then be diluted to 1:10 with ddH2O, and used as a template for PCR. A 909 bp core cDNA fragment of LPL was amplified using LPL-F and LPL-R as specific primers (Table 1). Approximate locations of primers were shown in Fig. 1. A pair of primer was designed based on sequence information of zebrafish LPL cDNA obtained from GenBank (BC064296, NM_131127), using Primer Premier 5.0 software. PCR was set up in a 50 μl volume reaction mixture composed of 2 μl diluted first strand cDNA product, 5.0 μl 10× buffer (with 20 mM Mg2+), 1.0 μl dNTP (10 mM each), 1.0 μl each primer (25 μM), 0.4 μl AmpliTag® DNA polymerase (5U/μl BBI), and ddH2O to 50 μl. Amplification was carried out in a BIO-RAD thermal cycler (iCycler), using the reaction settings as follows: initial denaturation at 94°C for 4 min, followed by 30 cycles of denaturing for 40 s at 94°C, annealing for 40 s at 52°C, extension for 60 s at 72°C, and a final extension for 7 min at 72°C. Five PCR products were selected by 1% w/v agarose gels electrophoresis and delivered to Shanghai Sangon Biological Engineering and Technology and Service Co. Ltd (Shanghai, China) for sequencing. The sequence was conducted in both forward and reverse directions using the amplification primers by ABI prism™ 377 automatic DNA sequencer. The forward and reverse sequences were assembled using SeqMan II software in DNAStar Package version 5.01, and the core fragment of LPL was obtained. According to the sequence information of this fragment, gene-specific primers were designed for 3′ RACE and 5′ RACE.

Table 1 The primers used for LPL gene cDNA RT-PCR and RACE
Full size table
Fig. 1
figure 1

Primers and approximate locations in LPL gene cDNA used for RT-PCR and RACE

Full size image

Rapid amplification of the 3′ end (3′ RACE)

Rapid amplification of the 3′ end was performed using 3′-Full RACE Core Set (TaKaRa), according to the manufacturer’s instructions. The primers used for 3′ RACE are shown in Table 1. Firstly, reverse transcription of 2 μg of common carp total RNA was performed with Oligo(dT)16AP as primer, which containing an anchor sequence. The resulting first strand cDNA then be diluted to 1:10 with ddH2O and used as a template for PCR. Amplification was done with a gene-specific forward primer LPL3-F1 and a reverse primer AP that corresponds to the anchor sequence. After the first PCR, the product was diluted to 1:10 with ddH2O and used as template for the nested PCR, which was done with a specific forward primer LPL3-F2 and a reverse primer RACE3-R. The first and nested PCR were carried out under the same condition as above. The nested PCR product was separated by agarose gel electrophoresis and bands of the expected size was excised, and purified using the EZ-10 Spin Column DNA Gel Extraction Kit (BBI). The purified PCR product was subcloned into a PUCm-T vector (BBI) and five independent clones were sequenced in both forward and reverse directions using the universal M13 primers.

Rapid amplification of the 5′ end (5′ RACE)

Rapid amplification of the 5′ end was made following a protocol described by Dieffenbach [20], the primers used for 5′ RACE are shown in Table 1. Briefly, reverse transcription of 2 μg total RNA was performed with a specific reverse primer LPL5-R to obtain the first strand cDNA. After RNase H (MBI) treatment, an oligo(dA) tail at the 5′ end was added using terminal deoxynucleotidyl transferase (MBI). The resulting product then be diluted to 1:10 with TE buffer, and used as a template for the first PCR. Amplification was done with a universal forward primer Oligo(dT)16AP containing a poly-dT and an anchor sequence and a specific reverse primer LPL5-R1. The first PCR product was diluted to 1:10 with ddH2O, and used as a template for the nested PCR, which was done with a anchor forward primer AP and a specific reverse primer LPL5-R2. The first and nested PCR were carried out under the same condition as above. The nested PCR product was eluted from 1.5% w/v agarose gel. The purified PCR product was subcloned into a PUCm-T vector (BBI) and five independent clones were sequenced in both forward and reverse directions using the universal M13 primers.

Sequence analysis

The core fragment, 3′ end and 5′ end sequences were assembled using SeqMan II software in DNAStar Package to obtain full-length cDNA of LPL, The sequence was edited and analyzed using the program EditSeq of DNAStar package to search open reading frame (ORF) and then translate it into amino acid sequence using standard genetic codes. The amino acid sequence of LPL was tested for the presence of signal peptide with SignalP v3.0 at http://www.cbs.dtu.dk/services/signalP/ [21]. Putative transmembrane regions were predicted with HMMTOP v2.0 at http://www.enzim.hu/hmmtop/ [22]. The second structure of LPL protein was predicted at http://www.predictprotein.org/ and http://www.EMBL-Heidelberg.de/. The protein sequence alignment was made by program MegAlign of DNAStar package using the Clustal W method. The Phylogenetic tree was produced by Neighbor-Joining (NJ) methods (Kimura 2-parameter model, 10,000 replicates, boostrap phylogeny test) based on lipoprotein lipase amino acid sequences using MEGA software Version 3.1 [23].

Tissue expression of LPL mRNA

Tissue expression of the common carp LPL mRNA was determined by semi-quantitative RT-PCR using β-actin as internal control. Total RNAs were extracted from liver, head kidney, mesenteric adipose tissue, heart and white muscle of common carp as mentioned above. The integrity of isolated RNAs were assessed by 1% w/v denaturing agarose gels electrophoresis, whose quantity and purity were determined by absorbance measures at 260 nm and 280 nm, and diluted to 0.5 μg/μl with DEPC-water. Reverse transcriptions were performed with equal quantities of each total RNA (2 μg) as template and oligo(dT)18 as primer using AMV Reverse Transcriptase. The resulting first strand cDNA from each tissue then be diluted to 1:10 with ddH2O and used as template for PCRs. The PCRs were carried out using primers LPL-F1 and LPL-R (Table 1) for LPL cDNA amplification, Actin-F and Actin-R (Table 1), which were designed based on carp β-actin sequence information (M24113) and spanned a 93 bp intron in genomic DNA, for β-actin cDNA amplification as internal control in different tubes. The PCR products were analyzed by electrophoresis on 1.5% w/v agarose gels and the intensities of PCR bands were quantified using the Quantity one 4.6 system (BIO-RAD, USA). To avoid samples DNA contamination, the negative RT-PCRs control reactions were performed with every total RNA as templates.

Results

Molecular characterization of LPL

The full-length cDNA coding lipoprotein lipase (LPL) was completed from liver of adult common carp by assembling the core fragment, 3′ and 5′ end sequences, and submitted to GenBank (accession numbers FJ716101). The complete nucleotide sequence covered 2411 bp with an open reading frame (ORF) of 1,524 bp encoding 507 amino acids residues, including a putative signal peptide of 23 amino acids in length (Fig. 2). The 5′-untranslated region (UTR) covered 160 nt. An ATG initiation codon was found 160 nt downstream of the 5′-start, and a TAA stop codon was present 727 nt upstream of the 3′-end. The 3′-untranslated region (UTR) covered 727 nt containing one AACAAA motifs, which represent putative polyadenylation signals (nt 2,303–2,308).

Fig. 2
figure 2

The nucleotide sequence of LPL gene full-length cDNA and the deduced amino acid sequence of Cyprinus carpio var. Jian. The grey bases indicate the primers used for RT-PCR. An asterisk represents a termination codon. The boxed amino acid sequences indicate the heparin-binding site. Catalytic triad is boxed in black squares. Four pairs of cysteine residues sites are double underlined. The underlined amino acid sequences WW indicate the lipid-binding site. Other functional sites, such as N-linked glycosylation sites, signal peptide, polypeptide lid, the boundary of N-terminus and C-terminus structural regions were showed on figure. The nucleotide sequence was submitted to the NCBI GenBank, accession no. FJ716101

Full size image

The LPL protein has a calculated molecular weight of 57.8-kDa and isolectric point (PI) of 8.531. This amino acid sequence contains two structural regions: N-terminus (24–354 residues) and C-terminus (355–507 residues). Before N-terminus, 1–23 residues is signal peptide, 6–23 residues is transmembrance helix. At N-terminus, some conversed functional sites were found, including two N-linked glycosylation sites Asn41 and Asn88; one catalytic triad Ser174, Asp198 and His283; one conserved heparin-binding site Arg321 to Arg324 (RKNR); eight cysteines residues Cys69 and Cys82, Cys258 and Cys281, Cys306 and Cys325, Cys317 and Cys320 which are involved in four disulfide bridges; one polypeptide “lid” that participates in substrate specificity. At C-terminus, Asn401 is another N-linked glycosylation site, and Trp434 and Trp435 (WW) is lipid-binding site. The amino acid sequence has a high similarity, and shows similar structural features to LPL of other species. Complete amino acid sequence alignment of LPL showed the best identity with zebrafish (the homology is 88.0%), while less identity with human (the homology is 58.3%).

Phylogenetic analysis based on LPL amino acid sequences

Phylogenetic analysis among six fish species and thirteen endotherms species based on lipoprotein lipase amino acids sequences was shown in Fig. 3. Tree topology showed that all fish species were in the same group forming a clade, which were distinguished from endotherms. In fish, red sea bream (Pagrus major) was first grouped with gilthead sea bream (Sparus aurata), which belong to the same family Sparidae with a high bootstrap value of 99%, and then formed a clade with European seabass (Dicentrarchus labrax), which belong to the same suborder Percoidei with a bootstrap value of 100%. Common carp (Cyprinus carpio Var. Jian) and zebrafish (Danio rerio), which belong to the same family Cyprinidae, formed another clade with a high bootstrap 100%. In endotherms, human (Homo sapiens), chimpanzee (Pan troglodytes), crab-eating macaque (Macaca fascicularis) and olive baboon (Papio anubis), which belong to primates, formed a clade with a bootstrap 100%. Whereas, sheep (Ovis aries), cattle (Bos taurus) and goat (Capra hircus), which belong to ruminants, formed another clade with a bootstrap 100%. Then all species of mammals were in the same group forming a clade, which were distinguished from chicken (Gallus gallus). The phylogenetic relationship based LPL amino acid sequences was agreed to traditional classification.

Fig. 3
figure 3

Phylogenetic tree based on lipoprotein lipase amino acid sequences made with MEGA 3.1 software using Neighbor-Joining (NJ) method. The distance matrix was calculated using the Kimara’s 2-parameter model. The numbers represent bootstrap percentages. The topology was tested using bootstrap analyses (10,000 replicates). GenBank accession numbers: Homo sapiens BAD96184, Pan troglodytes XP_001149804, Macaca fascicularis BAE90068, Papio anubis NP_001106082, Tupaia glis ABQ24000, Mesocricetus auratus BAD69621, Mus musculus BAE29770, Rattus norvegicus NP_036730, Felis catus NP_001036032, Ovis aries NP_001009394, Bos taurus NP_001068588, Capra hircus ABD04039, Gallus gallus NP_990613, Cyprinus carpio var. Jian FJ716101 which is boxed to show present study, Danio rerio NP_571202, Oncorhynchus mykiss NP_001118076, Dicentrarchus labrax CAL69901, Pagrus major BAE95413, Sparus aurata AAS75120

Full size image

Tissue expression of the common carp LPL mRNA

Tissue expression of LPL mRNA in liver, head kidney, mesenteric adipose tissue, heart and white muscle of adult common carp was determined by semi-quantitative RT-PCR using β-actin as internal control. The results showed that the amplification yielded 478 bp bands for LPL and 438 bp bands for β-actin. RT-PCR analysis revealed that the expressions of LPL gene were detected in all examined tissues of adult common carp (Fig. 4a), while the negative control exhibited no products (data not shown). The expression levels of LPL in the mesenteric adipose tissue were highest among these tissues, following in liver and head kidney, and the lowest expression was found in heart and white muscle. The levels of LPL mRNA in mesenteric adipose tissue were 1.28, 1.25, 2.79 and 2.92 fold higher than in liver, head kidney, heart and white muscle, respectively. (Fig. 4b).

Fig. 4
figure 4

Expression analysis of LPL mRNA in different tissues of Cyprinus carpio performed by semi-quantitative RT-PCR with β-actin as internal control. a RT-PCR (30 cycles) products analyzed by agarose gel electrophoresis. b Relative expression profile of LPL normalized with β-actin level. Data points represent the mean values ± SE of the three replicates with different letters indicate significant different (P < 0.05). M, marker; L, liver; HK, head kidney; MA, mesenteric adipose tissue; H, heart; WM, white muscle

Full size image

Discussion

Sequence comparisons to LPL from endotherms

In the present study, we cloned and characterized the full-length cDNA coding LPL from liver of adult common carp, and analyzed its distribution patterns to investigate its possible roles in lipid metabolism. The cDNA encoded 507 amino acids residues, including a putative signal peptide of 23 amino acids in length, which sequence is very different from other species. The entire mature protein (484 amino acids in length) has a high similarity, and shows similar structural features to LPL of other species. The amino acid sequence of common carp LPL was compared with the corresponding sequences from human, chicken, zebrafish, rainbow trout, European seabass, gilthead sea bream, and red sea bream. The result was shown in Fig. 5.

Fig. 5
figure 5

Multiple alignment of LPL amino acid sequence. Common carp (car, FJ716101), human (hum, BAD96184), chicken (chk, NP_990613), zebrafish (zbf, NP_571202), rainbow trout (rbt, NP_001118076), European seabass (esb, CAL69901), gilthead sea bream (gsb, AAS75120), and red sea bream (rsb, BAE95413). Sites with an amino acid identical to that on the top line (human) indicated with a dot. Alignment gaps indicated by dashes. Amino acid substitutions indicated by the respective amino acid. The signal peptide, conserved cysteine residues ( ), active site residues (▲), putative heparin binding domains (underline) and a potential conserved N-linked glycosylation site in all sequences (N), a N-linked glycosylation site in all fish (n), and a N-linked glycosylation site only found in carnivorous fish (N) are indicated on figure

Full size image

LPL is a noncovalent dimer of two identical subunits that are held together in a head-to-tail arrangement [3]. In human, three amino acid residues Ala203, Gly215 and Gly222 have been proven to be critical to produce dimer, and to be found very conserved among endothermy species [5]. These amino acid residues are also found in common carp LPL (Ala218, Gly230, Gly237). The LPL subunit contain several functional sites: the catalytic triad (Ser, Asp, His), the lipid binding site(s), the heparin binding site(s), and the site for interaction with the activator apolipoprotein CII (apoCII). In human, the N-terminal region contains the conserved serine, aspartic acid and histidine catalytic triad (Ser159, Asp183 and His268, human numbering), where the triacylglycerol hydrolysis is carried out [5]. All of these residues are conserved in all species including common carp LPL. In human, tryptophan clusters at Trp417, Trp420 and Trp421 (human numbering) play an important role in binding to lipid substrates and lipoprotein. The Lys434 as a site for interaction with low-density lipoprotein receptor-related protein (LRP) [5]. All of these residues are conserved in all species including common carp. In human, the putative heparin binding located some positively charged basic residues rich regions, including from Arg306–Arg309, Arg321–Lys327, and the discontinuous sequence of Lys346 and Lys430–Lys441 [5]. The corresponding regions in the common carp LPL also contained the positively charged basic residues except for one (Lys440, human numbering). Especially, the region Arg306 to Arg309 (RKNR), and positively charged basic residues Arg321, Arg324, Lys327, Lys346, Arg432, and Lys434 (human numbering) are very conserved in all species including common carp. LPL requires a specific cofactor, apoC-II, to hydrolyze triglycerides in chylomicrons. In human, the study of chimeric lipases suggested that two regions, Val92–Leu95 and Tyr100–Ser106 in the N-terminal domain of LPL, were responsible for cofactor activation [24, 25]. The amino acid residues Val92–Leu95 (VPKL, human numbering) is conserved in all species including common carp, but the residues Tyr100–Ser106 (human numbering) is very different between endotherms and fish species.

LPL is an N-linked glycoprotein, and the consensus sequence for an N-glycosylation site is given as N-X-S/T (X is not P). There are two N-glycosylation sites, Asn70 and Asn386 in human LPL [5]. However, the numbers and positions of potential N-linked glycosylation sites are different among species. Three potential N-glycosylation sites are contained in the common carp LPL, in which, the first site Asn41 (common carp numbering) is conserved among teleost fish which is not found in human and chicken. The second site Asn88 (common carp numbering) was found only in common carp. The third site Asn401 (common carp numbering) in C-terminus region is completely conserved among all species including human and chicken. Whereas, one N-linked glycosylation site Asn480 (rainbow trout numbering) was found only in carnivorous fish. We postulate that this site may be related to the carnivorous of fish, but need to be verified.

Human LPL contain ten cysteines residues (Cys54 and Cys67;Cys243 and Cys266;Cys291 and Cys310;Cys302 and Cys305;Cys445 and Cys465, human numbering) involving in five disulfide bridges. One of the disulfide bridges is located in the C-terminal folding region (Cys445 and Cys465) [5]. Common carp LPL contains eight cysteines residues in the N-terminal region, but excludes any cysteines in the C-terminal region. This case was found in other fish, such as rainbow trout and red sea bream [9, 13]. This pair of residues in the C-terminal region are conserved in LPL from all mammals, and has been shown to be important for the interaction of LPL with emulsified lipids and also for the stability at 37°C of the LPL dimer [26]. The stability of fish LPL at 37°C was markedly low in comparison with bovine and human LPL. A possible reason for the difference might be that fish LPL does not have the C-terminal disulfide. Fish are poikilothermal vertebrates resulting in a body temperature lower than 37°C. This may have been a necessary adaptation for function at higher body temperatures [9].

Tissue expression of LPL mRNA

In the present study, tissue expression of LPL mRNA in liver, head kidney, mesenteric adipose tissue, heart and white muscle of adult common carp, was determined by semi-quantitative RT-PCR approach. The result showed that this gene is ubiquitously expressed in all tissues tested, even in liver (Fig. 4a). In mammals, Hepatic LPL expression is terminated soon after birth [27]. But in contrast to what has been found in mammals, adult fish, such as rainbow trout, gilthead sea bream, and red sea bream, not only visceral adipose tissue but also liver showed substantial LPL gene expression [1, 9, 13, 16]. This indicates LPL may have a function in fish liver that in higher vertebrates may be fulfilled by the related hepatic lipase. Our present data showed that the expression of LPL is tissue-specific, in which the highest expression was found in mesenteric adipose tissue, following in head kidney and liver, and the lowest expression was found in heart and white muscle. A graded tissue expression of LPL (adipose tissue > liver > white skeletal muscle) was also found in rainbow trout and gilthead sea bream [1, 9]. In gilthead sea bream, the expression of LPL in mesenteric adipose tissue was several times higher than in liver and skeletal muscle.

In summary, in this study we have cloned and characterized the cDNA encoding LPL from the liver of the omnivorous common carp. The main domains (catalytic site, disulfide bridge, N-linked glycosylation site, heparin binding domain, lipid binding site, and site of dimer formation) are basically conserved between the common carp and other vertebrates. Our results demonstrate that the liver is one of the main LPL producing tissue in adult fish.