Elovl4a participates in LC-PUFA biosynthesis and is regulated by PPARαβ in golden pompano Trachinotus ovatus (Linnaeus 1758)

The elongases of very long-chain fatty acids (Elovls) are responsible for the rate-limiting elongation process in long-chain polyunsaturated fatty acid (LC-PUFA) biosynthesis. The transcription factor, PPARα, regulates lipid metabolism in mammals; however, the detailed mechanism whereby PPARαb regulates Elovls remains largely unknown in fish. In the present study, we report the full length cDNA sequence of Trachinotus ovatus Elovl4a (ToElovl4a), which encodes a 320 amino acid polypeptide that possesses five putative membrane-spanning domains, a conserved HXXHH histidine motif and an ER retrieval signal. Phylogenetic analysis revealed that the deduced protein of ToElovl4a is highly conserved with the Oreochromis niloticus corresponding homologue. Moreover, functional characterization by heterologous expression in yeast indicated that ToElovl4a can elongate C18 up to C20 polyunsaturated fatty acids. A nutritional study showed that the protein expressions of ToElovl4a in the brain and liver were not significantly affected among the different treatments. The region from PGL3-basic-Elovl4a-5 (−148 bp to +258 bp) is defined as the core promoter via a progressive deletion mutation of ToElovl4a. The results from promoter activity assays suggest that ToElovl4a transcription is positively regulated by PPARαb. Mutation analyses indicated that the M2 binding site of PPARαb is functionally important for protein binding, and transcriptional activity of the ToElovl4a promoter significantly decreased after targeted mutation. Furthermore, PPARαb RNA interference reduced ToPPARαb and ToElovl4a expression at the protein levels in a time-dependent manner. In summary, PPARαb may promote the biosynthesis of LC-PUFA by regulating ToElovl4a expression in fish.

Interestingly, comparisons of the amino acid sequences for the Elovl4a protein from the above four species showed three conserved domains, which contained five putative membrane-spanning domains with a conserved HXXHH histidine motif and an ER retrieval signal (Fig. 1). The phylogenetic tree analysis indicated that ToElovl4a clustered with several other Elovl4a sequences from other osteichthyes, and more distantly, with avian (G. gallus) and mammalian (H. sapiens) Elovl4 (Fig. 2). ToElovl4a was grouped together with perciformes, such as O. niloticus.
Heterologous expression of the elongase ORF in Saccharomyces cerevisiae. The
Tissue distribution of ToElovl4a. Tissue distributions of ToElovl4a were delineated by qRT-PCR. The highest ToElovl4a mRNA levels were detected in the brain, followed by the stomach and intestine, whereas  FAs were extracted from yeast transformed with the pYES2 vector, including the ORF of the putative Elovl4a cDNA as an insert. Peaks 1-4 represent the main endogenous FAs of T. ovatus, namely, C16:0, C16:1 isomers, C18:0 and C18:1n -9, respectively. Based in the retention times, additional peaks were identified as 20: 3n-6 (B). Vertical axis, FID response; horizontal axis, retention time.
www.nature.com/scientificreports www.nature.com/scientificreports/ relatively low ToElovl4a expression levels were observed in the liver and spleen (Fig. 4). Notably, the expression of ToElovl4a in the brain was much higher than in other tissues (P < 0.05).
Nutritional regulation of ToElovl4a. The protein expression of ToElovl4a in the liver and brain fed with different levels of LNA or LA (18:3n-3 or 18:2n-6) through the diet was determined by a western blot. The GAPDH was used as an internal control for normalization. The express pattern of ToElovl4a protein levels in the liver and brain were uncorrelated with the fatty acid compositions (Fig. 5) (also Supplementary Fig. S1).
Promoter analysis of PPARαb regulation. The cloned candidate ToElovl4a promoter (1,057 bp) was an upstream non-transcribed sequence. To determine the binding region of PPARαb in the ToElovl4a promoter, a full length candidate promoter and several truncated mutants were constructed with a promoterless luciferase reporter vector, pGL3-basic. The promoter construct, Elovl4a-p5 (−148 bp to +258 bp), exhibited the highest promoter activity with PPARαb, suggesting that this region of the Elovl4a-p5 promoter sequence contained the PPARαb binding site (Fig. 6A).
To further confirm the interaction of ToPPARαb with ToElovl4a, the influence of ToPPARαb overexpression on ToElovl4a transcription was determined. PPARαb overexpression increased the promoter activity of ToElovl4a-5 at all tested time points in heterologous HEK 293 T cells, and the maximum difference occurred at 12 h posttransfection, which was 5.6-fold higher in the PPARαb-overexpressing cells than that in the controls (Fig. 6B). These results indicated that constitutively expressed PPARαb positively regulated ToElovl4a expression in HEK 293 T cells.
To identify the PPARαb binding sites in the Elovl4a promoter, the predicted binding sites were mutated (Fig. 7, Table 2). The effects on promoter activity were investigated in 293 T cells that were transfected with each mutant and PPARαb. The results revealed that mutation of the M2 binding site (+209 bp to +223 bp) caused significant reduction in promoter activity (Fig. 7), showing that M2 was the PPARαb binding site in the Elovl4a promoter. Notably, three other predicted binding sites did not induce luciferase activity with PPARαb, suggesting that these three sites were not required for triggering ToElovl4a expression with PPARαb.  Supplementary Fig. S2A). When ToPPARαb expression was reduced, the protein levels of ToElovl4a were considerably depleted compared with the control    Supplementary Fig. S2B). These results suggested an active regulatory role of ToPPARαb on ToElovl4a expression in the TOCF cells.

Discussion
The present study sought to gain insights into the mechanisms underlying the transcriptional regulation of LC-PUFA biosynthesis in T. ovatus. To achieve this, sequence and functional characterization, tissue expression patterns and transcriptional regulation of ToElovl4a were investigated. The ToElovl4a ORF encodes a protein that is 81%-96% identical to Elovl4 proteins from other teleosts. These isolated ToElovl4a proteins contain three classic structural motifs, including transmembrane domains, a conserved histidine box (HXXHH), and an ER retrieval signal (RXKXX) in the canonical C-terminal, indicating its specific role is in LC-PUFA biosynthesis 26 . These three conserved boxes were also found to be present in other species Elovl4 proteins [11][12][13][14][15] . The ToElovl4 sequence is positioned within the teleost Elovl4a clade together with O. niloticus, and the teleost Elovl4 clade is outgrouped by the tetrapod Elovl4 clade containing the Elovl4 sequences from avians and mammals.
Three members of the fatty acid elongases protein family, Elovl2, Elovl4 and Elovl5, have been described as crucial enzymes involved in the biosynthetic pathway of LC-PUFA in teleosts 1,27 . For marine fish, yeast heterologous expression systems indicated that Elovl5 can effectively elongate both C18 and C20 PUFA, whereas Elovl4 is mainly involved in the elongation of C20-22 LC-PUFA producing polyenes up to 36 carbons 1,13 . However, it  www.nature.com/scientificreports www.nature.com/scientificreports/ was also revealed that Elovl4 proteins are able to utilize all assayed C18-22 PUFA substrates [13][14][15] . In this study, the functional characteristics of ToElovl4a via heterologous expression in S. cerevisiae showed that the T. ovatus putative elongase is Elovl4a, which can only elongate C18 (18:3n-6) substrates to C20 (20:3n-6) PUFA. In agreement with the functional data obtained for some marine fish, such as the 7.6% low activity in Scatophagus argus 13 , the 4.6% in Acanthopagrus schlegelii 14 , and the 6.1% in Larimichthys crocea 15 , ToElovl4 also showed low activity (1.05%) towards PUFA substrates, which confirmed its role in the biosynthesis of VLC-PUFA. However, until now, unlike the present study, Elovl4a was also found to effectively convert C18-C22 PUFA to longer polyenoic products up to C36 in other carnivorous fish 1,[11][12][13][14][15] , suggesting that marine fish Elovl4 exhibited high elongation efficiency towards C18-C22 PUFA substrates, except ToElovl4a. It is inferred that ToElovl4a solely elongates omega-6 C18 fatty acids, and this has been hypothesized as an adaptive strategy to supplement for Elovl5 in T. ovatus 27 . For T. ovatus, yeast heterologous expression systems showed that Elovl5 can effectively transform C18-C20 PUFA and ToFads6 that possess Δ4/Δ5/Δ8 Fad desaturation activity 27,28 . In addition to the present study, thus far, the complete classical pathways of LC-PUFA biosynthesis have not been elucidated for T. ovatus 29 .  Table 2. All values are presented as the means ± SD (n = 3). Asterisks indicate that the values are memorably different from the individual controls (*p < 0.05 and **p < 0.01). Bars on the same group with different letters are statistically significant from one another.   www.nature.com/scientificreports www.nature.com/scientificreports/ In the present study, the highest ToElovl4a mRNA expression was detected in the brain, showing that essential fatty acid metabolism occurs in the brain 30 . However, relatively moderate ToElovl4a mRNA expression levels were detected in the stomach, intestine and gonad. Interestingly, these are the first tissues exposed to dietary lipids, and they are the main lipid metabolism tissues in the body 30 . Moreover, the liver is the main site for LC-PUFA synthesis 31 . These studies indicate that lower levels of hepatic Elovl4a transcripts in carnivorous marine fish, like T. ovatus, may correlate with their limited LC-PUFA biosynthetic abilities 3 .
Previous studies have indicated that Fad enzymatic activity and gene expression vary with dietary LNA/ LA (18:3n-3/18:2n-6) ratio 13,32 . Upregulation of Δ6 Fads2 gene expression was detected in Siganus fuscescens, Maccullochella peelii, Oncorhynchus mykiss and Scatophagus argus that were fed high dietary ratios of LNA/LA [32][33][34][35] . Unlike desaturases, there is a lack of data on the influence of dietary LNA/LA ratio on elongase expression. Xie et al. 13 showed that the expression of Elovl4 and Elovl5 is significantly affected by dietary fatty acid composition, and they showed the highest expression of mRNA in the liver and eye of fish fed a diet a LNA/LA ratio of 1.7:1 in Scatophagus argus. Unfortunately, in the present nutritional experiment, no pattern was found between the expression of ToElovl4a and fatty acid composition.
In general, mRNA levels of some genes in eukaryotic cells are dependent on transcription factors and RNA polymerases binding to specific sequences in gene promoters 36 . Consequently, the integrity and activity of a promoter can affect the gene expression. Moreover, PPARs are ligand-activated transcription factors that are necessary for regulating gene expression in the PUFA biosynthesis pathway 18 . Dual luciferase reporter assays were conducted to clarify regulatory mechanisms whereby PPARαb is believed to modulate Elovl4a expression. Analysis of the truncated mutants indicated that ToElovl4a reporter activity was induced by the overexpression of PPARαb. The core binding region in the ToElovl4a promoter is −148 bp to +258 bp (Fig. 6A). This was the first evidence showing that the transcription of Elovl4a may be upregulated by PPARαb. In a previous study, PPARαb interacted with the binding site of the ToElovl5 and ToFads6 promoter region to positively regulate ToElovl5 and ToFads6 transcription, respectively 27,28 . Obviously, PPARαb plays a key regulatory role in the LC-PUFA biosynthesis in T. ovatus. Furthermore, the deletion of the PPARαb M2 binding site (+209 bp to +223 bp) results in significantly reduced promoter activity (Fig. 7). To further confirm whether PPARαb is a transcription factor implicated in ToElovl4a function, the effects of PPARαb knockdown on ToElovl4a protein expression were investigated by western blotting in TOCF cells. These data showed that PPARαb upregulated ToElovl4a protein levels.
In summary, the functional studies presented here show that ToElovl4a may effectively extend 18:3n-6 substrates. Moreover, the proposed synthesis pathway of LC-PUFA was for T. ovatus 27,28 . Furthermore, we demonstrated clear associations between PPARαb and the ToElovl4a promoter and the positive regulatory functions of PPARαb in ToElovl4a transcription. These results provide new insights into the regulation and function of Elovl4a in fish and further reveal the complexity of the associated regulatory mechanisms.

Materials and Methods
Ethics statement. All experiments in this study were approved by the Animal Care and Use Committee of South China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences (No. SCSFRI96-253) and were performed according to the regulations and guidelines established by this committee. To minimize suffering of the fish, all surgeries were implemented with 0.01% 2-phenoxyethanol (Sigma-Aldrich) anaesthesia.
Diets, fish, feeding trial and sampling. Eight isonitrogenous and iso-lipidic diets were formulated with 45% crude protein and 12% crude lipid with different lipid sources (Supplementary Table 1). Diet 1 contained fish oil (FO) as the control, and diets 2-8 contained different proportions of fish oil, krill oil, soybean oil and corn oil. The dietary formulations, proximate and fatty acid compositions are shown in Supplementary Table 1. T. ovatus juvenile fish (body weight: 82.9 ± 2.4 g) were collected from Linshui Marine Fish Farm in Hainan Province, China. The fish were raised on commercial feed (Hengxin, crude protein >37%, crude fat >7%) according to standard feeding schemes 2 weeks before the feeding trial and maintained in fresh seawater at 29 ± 1 °C, a salinity of 35‰, and with dissolved oxygen >6 mg/L in a recirculating aquaculture system. The feeding experiment was conducted in 32 cages (1 m × 1 m × 1.5 m) in a corresponding environment with each cage including 20 fish that were randomly allocated. The fish were anaesthetized using MS222 (0.1 g/L; Sigma, Alcobendas, Spain); then, the liver and brain were sampled, flash frozen in liquid nitrogen, and stored at −80 °C until further use.
To determine the tissue expression profile of ToElovl4a, healthy fish tissue (n = 6) containing small intestine, liver, white muscle, brain, spleen, fin, gill, head kidney, stomach, blood, males and female gonads were sampled, flash frozen in liquid nitrogen, and stored at −80 °C until further use.
Gene cloning and bioinformatics of ToElovl4a. Total RNA (1 μg) was extracted from T. ovatus brain by TRIzol Reagent (Takara, Japan). The quality and quantity (concentration) of isolated RNA were determined using a NANODROP 2000 spectrophotometer (Thermo Scientific). Subsequently, cDNA was synthesized using the PrimeScript TM RT reagent kit (Takara, Kyoto, Japan), according to the manufacturer's instructions. A putative ToElovl4a number was derived from the annotation file of T. ovatus. Subsequently, a putative ToElovl4a sequence was obtained based on CDS data of T. ovatus. (https://doi.org/10.6084/m9.figshare.7570727.v1 (2019)).
To determine the veracity of the putative Elovl4a sequence, gene-specific primers were designed (Supplementary Table 2). The PCR protocol used has been previously described 37 . The amplified products were purified by a DNA purification kit (Tiangen, China), ligated into the pEASY-T1 vector (TransGen Biotech, China), and sequenced (Invitrogen, Guagnzhou, China). Validated plasmids were transformed into competent Trans1-T1 cells (TransGen Biotech, China). A Blast search on the putative Elovl4a ORF sequence further confirmed the accuracy and validity.

Preparation of the Elovl4a polyclonal antibody and western blotting analysis.
To prepare the polyclonal anti-Elovl4a antibody, a specific domain (Elovl4a aa 92-106 ) of Elovl4a was compounded from Genecreate (Wuhan, China). The resulting PCR product was inserted into the pET-B2M vector using Nde I/Xho I sites. To express recombinant T. ovatus Elovl4a protein (rToElovl4a), the recombinant plasmid was transformed into Escherichia coli BL21 (DE3) (Novagen, Germany). The rToElovl4a was purified as previously described 45 . To generate a polyclonal antibody, purified rToElovl4a protein was injected into white New Zealand rabbits using standard methods 46 . Once generated, the polyclonal antibody was pre-adsorbed using E. coli lysate supernatants to eliminate inhomogeneous antibodies and was depurated on a HiTrapTM Protein A HP column on a AKTAprime ™ Plus system (GE Healthcare, USA).
To confirm specificity of the rabbit anti-Elovl4a antibody, human embryonic kidney (HEK293T) cells were transfected with pcDNA3.1 and pcDNA3.1-Elovl4a for 48 h. After this period, cells were harvested by centrifugation at 160 g for 10 min at 4 °C. The total protein was extracted using ProteoPrep ® Total Extraction Sample Kit (Sigma-Aldrich). Then, the total protein were electrophoresed on 12% SDS-PAGE and electrophoretically transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, USA) using the PierceG2 Fast Blotter (25 V for 10 min, Pierce, Rockford, IL, USA). Western blotting analyses was executed according to a previously described protocol 47 .
Cloning of the Elovl4a promoter and construction of deletion mutants. Genomic DNA was extracted from the muscle tissue of T. ovatus as described previously 48 and used as a template for candidate promoter cloning. The sequence upstream of the Elovl4a gene was obtained from genomic sequencing data of T. ovatus. To identify the role of PPARαb in the transcriptional regulation of ToElovl4a, five different promoter regions of ToElovl4a were amplified by specific primers (Supplementary Table 2) and subcloned into the Kpn I and Xho I restriction sites of the pGL3-basic luciferase reporter plasmid (Promega, USA). Five recombinant plasmids, denoted pGL3-basic-Elovl4a-1 (−148 to +56), pGL3-basic-Elovl4a-2 (−500 to +56), pGL3-basic-Elovl4a-3 (−1001 to +56), pGL3-basic-Elovl4a-4 (−148 to +155) and pGL3-basic-Elovl4a-5 (−148 to +258), were www.nature.com/scientificreports www.nature.com/scientificreports/ constructed (Fig. 6A). The truncated mutants were amplified using PrimeSTAR Master Mix (Takara, Japan). The programme parameters were 95 °C for 4 min, followed by 30 cycles of 95 °C for 40 s, 56 °C for 40 s, and 72 °C for 1 min. A general DNA purification kit (Tiangen, China) was used to purify the PCR products. All purified PCR products and the pGL3-basic (Promega, USA) vector were digested with Kpn I and Xho I and concatenated by T4 DNA ligase (Takara, Japan) overnight at 16 °C. Recombinant plasmids were extracted using the EndoFree Plasmid Giga Kit (Tiangen, China), and constructs were confirmed by sequencing as described above.

Construction of truncated mutants for the identification of predicted transcription factor (TF) binding sites in the Elovl4a promoter.
To determine the potential function of the PPARαb binding sites on the core Elovl4a promoter, four truncated mutations of recombinant plasmids were established. The transcription factor binding site prediction (TFBS)-JASPAR database (http://jaspar.genereg.net/), TRANSFAC ® , and MatInspector ® were used to search for potential binding sites in the Elovl4a promoter sequence with PPARαb.
According to the manufacturer's protocol, truncated mutants were designed and produced with a Muta-direct TM site-directed mutagenesis kit (SBS Genetech, Shanghai, China) from the deletion mutant pGL3-basic-Elovl4a-5, which was wild-type. The prediction of four binding sites (M1, M2, M3, and M4) were directly deleted, and the corresponding TF binding site sequences are shown in Fig. 7A. Furthermore, to acquire the TF binding site mutations, we used the method of PCR augmentation referred to a previous study 49 . The influence of TF binding site mutations on the promoter activity of ToElovl4a were determined by a dual luciferase assay as described below.
Expression analysis of ToElovl4a with ToPPARαb. The ORF of T. ovatus PPARαb (ToPPARαb) (GenBank accession number: MH321826) was amplified with primers incorporating restriction sites for Nhe I and Hind III at the 5′ and 3′ ends, respectively (Supplementary Table 2). The DNA fragment was digested with the same restriction endonucleases (Nhe I and Hind III; Takara, Japan) and ligated into a correspondingly restricted pCDNA3.1-Flag vector (Invitrogen, USA). Transcription factors ToPPARαb and pGL3-basic-Elovl4a-5 of the promoter segment were chosen to determine the regulatory relationship between ToPPARαb and ToElovl4a. Detection of promoter activities were at specific time points (0 h, 3 h, 6 h, 12 h, 24 h, 48 h and 72 h). The siRNA for PPARαb (PPARαb-si) and the negative control (si-NC) were purchased from Genecreate (Wuhan, China). The PPARαb siRNA sequence is listed in Supplementary Table 2. After transfection with TOCF cells, the total protein was isolated at specific time points (0 h, 6 h, 12 h, and 24 h) as described above.
Statistical analysis. SPSS 19.0 software (IBM, USA) was used to conduct the statistical analyses. The data were analysed by the Duncan test using one-way ANOVA. All data from the relative expression represented at least three replications along with means ± standard error of the mean (SE). Differences were considered significant at the p < 0.05 level.