MicroRNAs Involved in the Regulation of LC-PUFA Biosynthesis in Teleosts: miR-33 Enhances LC-PUFA Biosynthesis in Siganus canaliculatus by Targeting insig1 which in Turn Upregulates srebp1

Post-transcriptional regulatory mechanisms play important roles in the regulation of LC-PUFA biosynthesis. Our previous study revealed that miR-33 could increase the expression of fatty acyl desaturases (fads2) in the rabbitfish Siganus canaliculatus, but the specific mechanism is unknown. Here, we confirmed that miR-33 could target the 3′UTR of insulin-induced gene 1 (insig1), resulting in downregulation of its protein level in the rabbitfish hepatocyte line (SCHL). In vitro overexpression of miR-33 inhibited the mRNA level of insig1 and increased the mRNA levels of Δ6Δ5 fads2 and elovl5, as well as srebp1. In SCHL cells, proteolytic activation of sterol-regulatory-element-binding protein-1 (Srebp1) was blocked by Insig1, with overexpression of insig1 decreasing mature Srebp1 level, while inhibition of insig1 led to the opposite effect. Srebp1 could enhance the promoter activity of Δ6Δ5 fads2 and elovl5, whose expression levels decreased with knockdown of srebp1 in SCHL. Overexpression of miR-33 also resulted in a higher conversion of 18:3n-3 to 18:4n-3 and 20:5n-3 to 22:5n-3, linked to desaturation and elongation via Δ6Δ5 Fads2 and Elovl5, respectively. The results suggested that the mechanism by which miR-33 regulates LC-PUFA biosynthesis in rabbitfish is through enhancing the expression of srebp1 by targeting insig1. The findings here provide more insight to the mechanism of miRNAs involvement in the regulation of LC-PUFA biosynthesis in teleosts.


Introduction
Long-chain (≥ C 20 ) polyunsaturated fatty acids (LC-PUFA) are highly bioactive forms of PUFA. The LC-PUFA with crucial physiological functions in humans and other animals (Janssen and Kiliaan 2014;Calder 2015) include arachidonic acid (ARA; 20:4n-6), eicosapentaenoic acid (EPA; 20:5n-3), and docosahexaenoic acid (DHA; 22:6n-3). As important components of cell membranes, LC-PUFA reduce membrane phase-change temperatures and enhance membrane fluidity, as well as playing crucial roles in maintaining the normal physiological function of biofilms (Xiao et al. 2001). LC-PUFA are also important in growth, survival, pigmentation, stress, and disease resistance of fish, as well as in the development of the brain, vision, and the nervous system (Tocher 2010).
The biosynthesis of LC-PUFA involves desaturation and chain elongation reactions that convert the C 18 PUFA precursors, linoleic acid (LA; 18:2n-6), and α-linolenic acid (ALA; 18:3n-3) into the physiologically important ARA, EPA, and DHA. Fish are important sources of n-3 LC-PUFA in the human diet and, consequently, LC-PUFA biosynthesis in fish has been investigated extensively in recent years (Castro et al. 2016; Monroig and Kabeya 2018). Generally, freshwater fish have the ability to convert LA and ALA to C 20-22 LC-PUFA (ARA, EPA, and DHA), with this process requiring enzymes such as fatty acyl desaturases (Fads) and elongation of verylong-chain fatty acids (Elovl) proteins. Fads and Elovl enzymes involved in LC-PUFA biosynthesis can also be found in marine fish, but lower activity and/or absence of key enzymatic capacities is associated with low LC-PUFA biosynthesizing capacity and thus dietary provision of EPA and DHA is required to sustain growth and development (Tocher et al. 2003). The rabbitfish Siganus canaliculatus is an exception to this pattern since this marine herbivore has the ability to biosynthesize LC-PUFA from C 18 PUFA Monroig et al. 2012). S. canaliculatus, in addition to its complement of elongases (Monroig et al. 2012), possesses two distinct fads2 genes encoding Δ6Δ5 and Δ4 desaturases enabling all desaturation reactions involved in the LC-PUFA biosynthesis pathway (Castro et al. 2016;Monroig and Kabeya 2018). Consequently, S. canaliculatus has become a valuable model for studying regulatory mechanisms of LC-PUFA biosynthesis in teleosts Monroig et al. 2012).
MicroRNAs (miRNAs) are a class of highly conserved, small non-coding,~22 nucleotides (nt) RNA molecules that are widespread in organisms. In land animals, miRNAs generally exhibit a negative regulatory effect on gene expression and are involved in a number of biological processes (Alvarezgarcia and Miska 2005;Carrington and Ambros 2003;Xu et al. 2003). Post-transcriptional regulatory mechanisms have been shown to play important roles in LC-PUFA biosynthesis and glycolipid metabolism, as well as growth, development, reproduction, and immune function in teleosts (Gong et al. 2015;Her et al. 2011;Siddique et al. 2016;Škugor et al. 2014;Tao et al. 2018;Zhang et al. 2014;Zhu et al. 2015). In rabbitfish, miR-17 was found to be involved in the regulation of LC-PUFA biosynthesis by targeting Δ4 fads2 (Zhang et al. 2014). Moreover, our previous study revealed that miR-33 was involved in the regulation of LC-PUFA biosynthesis by increasing the expression of fads2 in rabbitfish (Zhang et al. 2016b), although the underlying mechanism was not clearly established.
In mammals, miR-33 exists as two distinct isoforms, namely miR-33a and b, which differ from each other in two bases outside the seed region of the mature versions (Najafi-Shoushtari et al. 2010). Both miR-33a and miR-33b are located in the sterol regulatory element-binding protein (SREBP) intron region (Najafi-Shoushtari et al. 2010;Rayner et al. 2010;Goldstein et al. 2002). In rabbitfish, the miR-33 gene was identified within intron 16 of the gene encoding srebp1, and miR-33 overexpression suppressed the expression of insig1, which is predicted to be the target gene of miR-33 (Zhang et al. 2016b). However, whether this is a direct effect or not is unclear. Moreover, miR-33 overexpression led to an increase in the mRNA levels of fads2 and srebp1, thus, it is believed to be involved in the regulation of LC-PUFA biosynthesis, but the specific mechanism is unknown (Zhang et al. 2016b).
The Insulin-induced gene protein (Insig) is an important factor in the regulation of lipid metabolism (Jo et al. 2011). The two subtypes of Insig, namely Insig1 (Radhakrishnan et al. 2007) and Insig2 (Lee et al. 2005), combine with Srebp and Srebp cleavage-activating protein (SCAP) in the endoplasmic reticulum (ER) as an Insig-SCAP-Srebp complex. Insig1 has a stronger affinity for SCAP compared with Insig2 and is reported to block the proteolytic cleavage of Srebp proteins by retaining Srebp precursors in the ER membrane, consequently decreasing lipogenesis (Gong et al. 2006;Engelking et al. 2004). In particular, the mechanisms detailed above apply to mammals, and it remains unclear whether Insig1 can block Srebp proteolytic activation in teleost fish.
In mammals, miR-33 has been found within the intron of srebp genes and reported to function in cooperation with its host (Horie et al. 2013). Srebp1 is an important transcription factor involved in regulating the expression of key enzymes in LC-PUFA synthesis in the liver (Nara et al. 2002;Carmona-Antoñanzas et al. 2014). Srebp1 affects the synthesis of LC-PUFA via activation of acetyl CoA carboxylase (ACC), fatty acid synthetase (FAS), stearoyl-CoA desaturase (SCD), and other enzymes related to fatty acid metabolism in mouse liver (Shimomura et al. 1998). Srebp1c is also reported to promote the expression of Δ5 and Δ6 fads-like genes in the liver (Qin et al. 2009). In rabbitfish, our previous studies showed there might be potential interaction between Srebp1 and the key enzymes of LC-PUFA synthesis, especially Δ6Δ5 Fads2 and the PUFA elongase, Elovl5, and the sterol regulatory element (SRE) of Srebp protein predicted in the promoter region of Δ6Δ5 fads2 and elovl5 (Zhang et al. 2016a;Dong et al. 2018).
The aim of the present study was to investigate the mechanism underpinning the regulation of LC-PUFA biosynthesis by miR-33. Firstly, dual-luciferase assay and Western blotting were performed to determine whether insig1 was a direct target gene of miR-33. Secondly, to explore the functional relationship between Insig1 and Srebp1 in rabbitfish, overexpression and inhibition of insig1 followed by Western blotting were used to detect the protein abundances of Insig1 and mature Srebp1. Subsequently, in order to further elucidate the roles of Srebp1 in the regulation of LC-PUFA biosynthesis in rabbitfish, changes in the expression, as well as the promoter activity, of Δ6Δ5 fads2 and elovl5, in response to changes in the in vitro expression level of srebp1, were explored. The data provide the basis for elucidating the mechanism of miR-33 involvement in the regulation of LC-PUFA biosynthesis in rabbitfish, as well as providing the theoretical basis for the participation of miRNA in the regulation of LC-PUFA biosynthesis in teleosts.

Plasmid Construction
In order to achieve overexpression of insig1, we constructed the pcDNA-Insig1 eukaryotic expression vector at the EcoRI and XhoI (New England Biolabs, Ipswich, USA) restriction sites, with full-length sequence of insig1 amplified using pcDNA-Insig1-F/R primers (Table 1). Similarly, we constructed the pcDNA3.1-SREBP overexpression vector at the XbaI and HindIII restriction sites, with full-length sequence of srebp1 amplified using LG-SREBP-F/R primers ( Table 1). The Δ6Δ5 fads2 and elovl5 promoter deletion dualluciferase reporter vectors were constructed previously in our laboratory (Dong et al. 2018). For heterologous expression of rabbitfish miR-33, a DNA fragment encompassing rabbitfish pre-miR-33 was digested with EcoRI and BamHI and inserted into the pEGFP-C3 plasmid. To construct the dual-luciferase reporter vectors, DNA fragments were inserted into pmirGLO dual-luciferase miRNA target expression vector (Promega, Madison, WI, USA) at the SacI and XbaI restriction sites. The recombinant vectors were as follows: (1) pmirGLO-Insig-3′UTR, pmirGLO including an insert consisting of a partial DNA fragment of the rabbitfish insig1 3′UTR, which includes the binding site of miR-33 in rabbitfish, amplified with Insig-3′UTR-F/R primers (Table 1); (2) pmirGLO -Insig-3′UTR-MU, pmirGLO including an insert consisting of the predicted binding site of miR-33 in insig1 3′UTR (5′-AATGCA-3′) mutated into 5′-TAAGGA-3′ to prevent complementarity of miR-33, and amplified with mutation primers Insig-3′UTR-Mu-F/R designed following the instructions of the Muta-direct TM site-Directed Mutagenesis Kit (SBS Genetech Co., Ltd., Beijing, China); (3) pmirGLO-R33 (positive control), pmirGLO including an insert consisting of a synthesized oligonucleotide containing a 100% match to miR-33 (Sangon Biotech, Shanghai, China). Sequences of primers and oligonucleotides used for cloning are provided in Table 1.

RNA Isolation and Quantitative Real-Time PCR
Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA) followed by determination of the concentration and quality of the total RNA on NanoDrop 2000 (Thermo Scientific, USA). cDNA was synthesized with 1 μg total RNA using the miScript II RT Kit (Qiagen, Hilden, Germany), and the expression of miR-33 determined using the miScript SYBR Green PCR Kit (Qiagen, Hilden, Germany) with miR-33 specific primer (qPCR-miR-33) ( Table 1) and universal primers. For the quantitative realtime PCR (qPCR), determination of the mRNA expression levels of insig1 (KU598855), srebp1 (JF502069.1), Δ6Δ5 fads2 (EF424276.2), and elovl5 (GU597350.1), LightCycler® 480 SYBR Green I Master (Roche, Germany) was used with rabbitfish gene-specific primers ( Table 1). The relative RNA level of each gene was normalized to that of 18s rRNA (AB276993), and calculated using the comparative threshold cycle method (Livak and Schmittgen 2012). All reactions were run on LightCycler® 480 thermocycler (Roche, Germany) using qPCR programs according to the manufacturer's specifications.

Dual-Luciferase Assay
To determine whether insig1 was a direct target gene of miR-33, a dual-luciferase assay was performed using human embryonic kidney (HEK 293 T) cells (Chinese Type Culture Collection, Shanghai, China). The HEK 293T cells were seeded into 96-well cell culture plates in 100-μl high-glucose Dulbecco's modified Eagle medium (DMEM) (Gluta MAX) (Gibco, Life Technologies, USA) with 10% fetal bovine serum per well (FBS, Sijiqing Biological Engineering Material Company, China). The HEK 293T cells were grown for 24 h to 80% confluence, and then co-transfected with either pEGFP-miR-33 (50 ng) or pEGFP-empty (50 ng) with different recombinant dual-luciferase reporter vectors (50 ng) using Lipofectamine® 2000 Reagent (Invitrogen, Carlsbad, USA) according to the manufacturer's instructions. Firefly and Renilla luciferase activities were quantified after 48-h transfection using a microplate reader (Infinite M200 Pro, Tecan, Switzerland) and firefly luciferase activity was normalized to Renilla luciferase activity. Eight replicate wells were used for each treatment.

Western Blotting
In order to further study the potential relationship between miR-33 and the target gene Insig1 at the protein level, Western blotting was carried out. miR-33 was upregulated by transfection with miR-33 mimics (dsRNA oligonucleotides) and negative control (NC) oligonucleotides were obtained from Genepharma (Shanghai, China). The The underscore in the table indicates the restriction site in primers sequences were as follows: miR-33 mimic, sense, 5′-CGUGCAUUGUAGUUGCAUUG-3′; antisense, 5′-AUGCAACUACAAUGCACGUU-3′. SCHL cells were seeded onto 100 mm plates (2 × 10 6 cells per plate), grown for 24 h to 80% confluence, and then transfected with 300 pmol miR-33 mimics or NC using Lipofectamine® 2000 Reagent, in triplicate. Total protein was extracted at 48-h post-transfection using cell total protein extraction kit (Sangon Biotech, Shanghai, China) and concentrations quantified with non-interference protein assay kit (Sangon Biotech, Shanghai, China). Next, 30 μL of each sample was loaded and separated on a 12% SDS/PAGE, transferred onto PVDF membranes, and then incubated with anti-Insig1 rabbit pAb (Wanleibio, Shenyang, China) at 1:500 dilution. Actin level, determined using anti-actin antibody (Beyotime, Haimen, China) at 1:2000 dilution, was used for normalization. After three washes with Tris-Buffered Saline Tween (TBST), membranes were incubated with goat-anti-rabbit and goat-anti-mouse (Millipore, Bedford, MA, USA) secondary antibodies at a ratio of 1:15000. Membranes were washed three times with TBST, and the immunoreactive bands visualized using the Odyssey infrared imaging system 2.1 (LI-COR, USA) and analyzed by Image studio (ver 5.2) software with the quantity of Insig1 and Actin protein converted into intensity values.
Overexpression of miR-33 to Investigate the Functional Relationship Between miR-33 and LC-PUFA Biosynthesis-Related Genes To further investigate the potential role of miR-33 in the regulation of gene expression in LC-PUFA biosynthesis, miR-33 was upregulated by transfection with miR-33 mimics. SCHL cells were seeded into six-well plate (5 × 10 5 cells per well), grown for 24 h to 80% confluence, and then transfected with 100 pmol miRNA mimics or NC with Lipofectamine® 2000 Reagent, each process consisting of six replicates. At 24-h post-transfection, SCHL cells were harvested, RNA extracted and subjected to qPCR analysis of expression levels of insig1, srebp1, Δ6Δ5 fads2, and elovl5 mRNA.
Promoter Analysis to Explore the Functional Relationship Between srebp1, Δ6Δ5 fads2, and elovl5 To investigate the functional relationship between srebp1 and the promoters of Δ6Δ5 fads2 and elovl5, a dualluciferase assay was performed. The relationship between these constructs was investigated using HEK 293T cells, which were co-transfected with 100 ng pcDNA3.1-SREBP eukaryotic expression vector and 50 ng Δ6Δ5 fads2 or elovl5 promoter deletion dual-luciferase reporter vectors with Lipofectamine® 2000 Reagent (Invitrogen, Carlsbad, CA, USA). The Renilla reporter vector pGL4.75 plasmid was used as the internal reference. Firefly and Renilla luciferase activities were quantified after 48-h transfection by a microplate reader (Infinite M200 Pro, Tecan, Switzerland) and firefly luciferase activity was normalized to Renilla luciferase activity.

Inhibition of srebp1 in Rabbitfish Hepatocytes
For srebp1 RNAi in SCHL cells, three pairs of siRNAs, namely siRNA-638, siRNA-1211, and siRNA-1303, were designed (numbers represented the location of the target site of the siRNA on the gene) and synthesized by a commercial company (Genepharma, Shanghai, China). SCHL cells were seeded onto six-well plates with 5 × 10 5 cells per well, grown for 24 h to 80% confluence, and then transfected with the siRNAs or NC using the RNAiMAX Reagent (Invitrogen, Carlsbad, CA, USA) in triplicate. At 48-h post-transfection, SCHL cells were harvested; RNA extracted and subjected to qPCR determination of the mRNA expression levels of srebp1.

Fatty Acid Analysis
SCHL cells were seeded onto 100-mm plates at a density of 2 × 10 6 cells per plate, grown for 24 h to 80% confluence, and then transfected with 300 pmol miRNA mimics or NC with Lipofectamine® 2000 Reagent triplicates. After 48-h incubation, cells were harvested for fatty acid composition analysis by gas chromatography (GC) after chloroform/methanol extraction, saponification, and methylation with boron trifluoride (Sigma-Aldrich, USA) as described previously Chen et al. 2016). For identification, the retention times of the fatty acids were compared with those of standard methyl esters (Sigma-Aldrich, USA), with quantification of each fatty acid in a certain number of cells being estimated using the signal of the internal standard 17:0 (heptadecanoic acid) (Sigma-Aldrich, USA). Fatty acid contents were expressed as percentage of total fatty acids ( Table 2).

Statistical Analysis
All the data are presented as means ± SEM. The dualluciferase assay and qPCR expression data were analyzed by one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test or Student's t test using Origin 7.0. A significance of P < 0.05 was applied to all statistical tests performed.

Overexpression of miR-33 Inhibits the Activity of Insig1 Protein in SCHL
Western blotting was used to further study the potential relationship between miR-33 and its target protein Insig1. After overexpression of miR-33, the protein level of Insig1 decreased by 40% compared with the NC group (Fig. 2). It was observed that miR-33 had an inverse expression pattern with Insig1 protein level in rabbitfish.

Insig1 Can Block the Formation of Mature Srebp1 Protein in Rabbitfish Hepatocytes
In order to provide direct proof for Insig1 being able to block Srebp proteolytic activation, Western blotting was used to further study the potential relationship between Insig1 and Srebp1. The protein levels of mature Srebp1 and Insig1 were detected after overexpression and inhibition of insig1. With overexpression of insig1, the protein level of Insig1 increased by 1.5-fold, while the protein level of mature Srebp1 decreased by 50% compared with the pcDNA3.1 group (Fig. 4a). With inhibition of insig1, the protein level of Fig. 1 a Scheme of miR-33 base pairing the 3′UTR of the rabbitfish insig1. b Rabbitfish miR-33 is overexpressed in HEK 293T cells by transfecting with pEGFP-miR-33. c Luciferase activity in HEK 293T cells cotransfected with pEGFP-miR-33 or pEGFP-empty with different recombinant dual-luciferase reporter vectors: pmirGLO-empty as negative control (lanes 1-2); pmirGLO-R33 as positive control (lanes 3-4); pmirGLO-Insig-3′ UTR containing 3′UTR of insig1 (lanes 5-6); pmirGLO-Insig-3′ UTR-MU with 4 nt site-directed mutation in 3′UTR of insig1 (lanes 7-8). The Renilla luciferase activity was used to normalize that of firefly luciferase. Data are shown as means ± SEM (n = 8) and asterisks represent significant differences (*P < 0.05; **P < 0.01; t test) Fig. 2 miR-33 is a negative regulator of rabbitfish Insig1 at the protein level. The SCHL cells were transfected with miR-33 or NC mimics, Western blotting detected protein expression levels of Insig1 and normalized by Actin. The numbers on the graph refer to the gray value ratio of the target gene to the reference gene, and the value indicated the relative expression level of the protein Insig1 decreased by 30%, while the protein level of mature Srebp1 increased by 1.5-fold compared with the si-NC group (Fig. 4b).
Functional Relationship Between srebp1 and Δ6Δ5 fads2 and elovl5 Our previous studies predicted that the SRE binding element of Srebp protein was present in the promoter regions of Δ6Δ5 fads2 and elovl5 in rabbitfish (Fig. 5) (Dong et al. 2018). Thus, the pcDNA3.1-SREBP eukaryotic expression plasmid, Δ6Δ5 fads2, and elovl5 promoter deletion dual-luciferase reporter vectors were co-transfected into HEK 293T cells. The dual-luciferase reporter assay results showed that overexpression of srebp1 resulted in significantly increased fluorescence activity of Δ6Δ5 fads2 promoter deletion D2 and D4 (P < 0.05 and P < 0.01), while the fluorescence activity of D1 and D3 was not significantly different from the control group (P > 0.05) (Fig. 6). At the same time, when srebp1 was overexpressed, the fluorescence activity of elovl5 promoter deletion D1, D2, and D3 significantly increased (P < 0.05) (Fig. 7).

Srebp1
Regulates the Expression of Δ6Δ5 fads2 and elovl5 In order to knockdown srebp1, the silencing efficiency of three pairs of siRNA, namely siRNA-638, siRNA-1211, and siRNA-1303, was determined. The results showed that siRNA-638 and siRNA-1211 had significant knockdown effects on srebp1 mRNA expression (P < 0.05) (Fig. 8), with about 16% and 18% knockdown efficiency, respectively, while there was no significant difference between siRNA-1303 and the control group (P > 0.05). Thus, siRNA-1211 was used for the RNAi experiments. Following successful srebp1 knockdown with siRNA-1211 in SCHL cells, the expression levels of Δ6Δ5 fads2 and elovl5 were also significantly decreased (P < 0.05) (Fig. 9).

Upregulation of miR-33 and Biosynthesis of LC-PUFA in Rabbitfish Hepatocytes
The effects of miR-33 on LC-PUFA biosynthesis in SCHL cells was assessed by determining the effects of overexpressing miR-33 on fatty acid composition. It was observed that miR-33 overexpression resulted in a higher conversion of 18:3n-3 to 18:4n-3, and 20:5n-3 to 22:5n-3, as well as higher levels of ARA, EPA, and DHA in rabbitfish hepatocytes. Compared with the NC group, the conversion of 18:3n-3 to 18:4n-3 increasing significantly (P < 0.05) ( Table 2).
Discussion miR-33 is highly conserved in animals and plays a crucial role in the regulation of lipid metabolism, such as reverse cholesterol transport and fatty acid oxidation (Gerin et al. 2010;Horie et al. 2010;Najafi-Shoushtari et al. 2010;Rayner et al. 2010). We reported previously that miR-33 has a potential role in the regulation of LC-PUFA biosynthesis in fish (Zhang et al. 2016b). Hence, the present study aimed to further explore the mechanisms underpinning this finding. In our previous research, we found that overexpression of miR-33 Fig. 4 a Effects of insig1 overexpression on the protein level of Insig1 and Srebp1. b Effects of insig1 inhibition on the protein level of Insig1 and Srebp1. The numbers on the graph refer to the gray value ratio of the target gene to the reference gene, and the value indicated the relative expression level of the protein Fig. 3 Effects of miR-33 overexpression on the mRNA level of insig1, srebp1, Δ6Δ5 fads2, and elovl5 in SCHL cells. The gene expression was determined by qPCR in SCHL cells transfected with miR-33 mimics or NC mimics for 24 h. Data are means ± SEM (n = 6). Asterisks represent significant differences (P < 0.05; ANOVA, Tukey's test) upregulated the transcription of Δ4 fads2 and Δ6Δ5 fads2, but suppressed the expression of insig1. In mammals, INSIG1 could bind to the sterol-sensing domain of SCAP (SREBP cleavage activating protein), which makes the SCAP/SREBP complex reside longer in the ER, ultimately blocking SREBP entry into the nucleus and preventing its action as a transcription factor (Gong et al. 2006;Engelking et al. 2004). Accordingly, it was necessary to demonstrate the direct effect of miR-33 on insig1, including whether Insig1 can block Srebp proteolytic activation, and whether Srebp1 directly upregulates the expression of genes involved in LC-PUFA biosynthesis. We herein report compelling evidence corresponding to each of these points. Our data confirmed that miR-33 downregulated the protein abundance of Insig1 in rabbitfish through direct targeting of the 3′UTR of insig1. Generally, miRNAs depend on the Bseed sequence^to identify and partially combine with the 3′UTR of target genes, thereby inducing target mRNA degradation or inhibiting protein translation, manifested by the reduction of function or activity of target genes. The dual-luciferase assay revealed that miR-33 could repress the 3′UTR luciferase activity of insig1, which provided evidence that insig1 was a direct target gene of miR-33. In rabbitfish hepatocytes, the overexpression of miR-33 decreased the protein abundance of Insig1, therefore indicating that a negative regulation of miR-33 occurred at the translational level, in agreement with the general mechanism of animal miRNA (Bartel 2009;Pillai et al. 2005).
Lowering the protein abundance of Insig1 by miR-33 indirectly increased the level of Srebp1 mature protein because the inhibitory effect of Insig on Srebp activation was reduced. Our previous studies showed that, in SCHL cells, miR-24 enhanced the expression of srebp1 mRNA and the production of mature Srebp1 protein by targeting insig1, while opposite Fig. 6 Effects of srebp1 overexpression on activity of Δ6Δ5 fads2 promoter deletion. Here is a dual-luciferase report experiment in HEK 293T cells, with 0.04-ng pGL4.75 plasmid as the internal reference, and pGL4.10 as the negative control. The black ellipse in the figure represents the position of the SRE element predicted on the promoter sequence. Data are shown as means ± SEM and asterisks represent significant differences (*P < 0.05; **P < 0.01; t test) results were observed with knockdown of miR-24 in rabbitfish hepatocytes suggesting that insig1 may inhibit the formation of mature Srebp1 (Chen et al. 2019). In the present study, we confirmed Insig1 can block Srebp proteolytic activation, with overexpression of insig1 decreasing abundance of mature Srebp1 protein, while inhibition of insig1 leads to an increase in abundance of mature Srebp1 protein. This suggested that Insig proteins dissociated from the Insig-SCAP-SREBP complex, so that the SCAP-SREBP complex could transfer to the Golgi (Gong et al. 2006;Engelking et al. 2004). It is reported that, under the shear processing of S1P and S2P, the mature protein of Srebp1 is formed, which then enters the nucleus and exerts its regulatory functions (Yang et al. 2002;Yabe et al. 2003). In rabbitfish hepatocytes, overexpression of miR-33 decreased the protein abundance of Insig1, coupled with decreased mRNA expression of insig1 and increased expression of srebp1 mRNA. With an increase in the mRNA expression level of srebp1, there would be a corresponding increase in the protein abundance of Srebp1. It is known that miR-33 is found within an intron of the srebp gene in both mammals and teleosts (Horie et al. 2013;Zhang et al. 2016a). In rabbitfish, increased transcription of srebp1 elevated the abundance of miR-33 in primary hepatocytes (Zhang et al. 2016b). This was interesting as it suggests that transcription of srebp1 helps to bring about self-proteolysis. In addition, it was observed in the present study that, when miR-33 was overexpressed, the level of srebp1 also increased. The question here is how did miR-33 enhance the expression of srebp1 to participate in LC-PUFA biosynthesis? Research in mice has shown that srebp1c can promote the expression of the Δ5 and Fig. 8 Relative expression of srebp1 after silenced by different siRNAs. Mock means adding only transfection reagent groups; NC means adding siRNA negative control group; SiRNA was the experimental group. Data are means ± SEM (n = 3) and different superscripts indicate significant differences (P < 0.05) Fig. 7 Effects of srebp1 overexpression on activity of elovl5 promoter deletion. Here is a dual-luciferase report experiment in HEK 293T cells, with 0.04-ng pGL4.75 plasmid as the internal reference and pGL4.10 as the negative control. The black ellipse in the figure represents the position of the SRE element predicted on the promoter sequence. Data are shown as means ± SEM and asterisks represent significant differences (*P < 0.05; **P < 0.01; t test) Δ6 fads-like genes (Qin et al. 2009). Similarly, it has been reported that Srebp1 can be activated by TO901317, a liver X receptor (lxr) agonist, that activates the Lxr-Srebp1 pathway and further promotes the expression of Δ5 and Δ6 fads in Atlantic salmon (Salmo salar) (Minghetti et al. 2011). Studies have also found that SRE elements are present in the human FADS2 (Δ6) promoter (Nara et al. 2002). In Atlantic salmon, Atlantic cod (Gadus morhua L.), and zebrafish (Danio rerio), SRE elements have been predicted in the Δ6 fads2 promoter (Zheng et al. 2009). In rabbitfish, our previous studies characterized the upstream promoter sequences of Δ6Δ5 fads2 and elovl5 genes (Dong et al. 2018) and predicted SRE elements in the upstream promoter sequences. These data indicated that srebp1 might participate in the regulation of LC-PUFA biosynthesis by regulating Δ6Δ5 fads2 and elovl5. When further exploring the functional relationship between srebp1 and the promoters of Δ6Δ5 fads2 and elovl5, dual-luciferase reporter assays showed that overexpression of srebp1 significantly enhanced the fluorescence activity of Δ6Δ5 fads2 and elovl5 promoter containing SRE element. These results demonstrated that srebp1 can activate the expression of the rabbitfish Δ6Δ5 fads2 and elovl5 depending on the SRE element. We previously revealed that lxr agonist (TO901317) could significantly increase the mRNA expression level of srebp1, together with an increase in the expression of fads2 in rabbitfish (Zhang et al. 2016a). Here, when the expression of srebp1 was depleted by RNAi, the expression levels of Δ6Δ5 fads2 and elovl5 also decreased. Therefore, miR-33 enhances LC-PUFA biosynthesis by upregulating srebp1, which acting on Δ6Δ5 fads2 and elovl5 in rabbitfish.
Analysis of the differences in the fatty acid composition reflects the ability of C 18 PUFA to be converted into LC-PUFA through the desaturation and elongation pathway (Brown 2005). We observed that overexpression of miR-33 resulted in higher production of C 20-22 LC-PUFA compared with the NC group, as the level of ARA, EPA, and DHA increased in rabbitfish hepatocytes. These findings indicated that miR-33 promoted LC-PUFA biosynthesis. Moreover, overexpression of miR-33 increased the conversion of 18:3n-3 to 18:4n-3 and 20:5n-3 to 22:5n-3. Specifically for rabbitfish, the conversion of 18:3n-3 to 18:4n-3 requires the action of the Δ6Δ5 Fads2, while the conversion of 20:5n-3 to 22:5n-3 requires that of Elovl5; hence, these results suggested an increase in the enzymatic activities of Δ6Δ5 Fads2 and Elovl5. These results have therefore shown that miR-33 promotes LC-PUFA biosynthesis at the physiological level, Fig. 9 Relative expression of srebp1, Δ6Δ5 fads2, and elovl5 gene after silenced by siRNA-1211 in SCHL cells. NC means adding siRNA negative control group; SiRNA-1211 was the experimental group. Data are means ± SEM (n = 3) and asterisks represent significant differences (P < 0.05; ANOVA, Tukey's test) Fig. 10 Diagrammatic representation of the roles of miR-33 in regulation of LC-PUFA biosynthesis. The putative mechanism diagram is proposed for effects of miR-33 overexpression on target genes, host genes, and LC-PUFA biosynthesis-related genes. Red arrows denote suppression; green arrows indicate promotion which might be achieved by regulating the Δ6Δ5 Fads2 and Elovl5, key enzymes in LC-PUFA biosynthesis.
In summary, we propose a putative mechanism (Fig. 10), suggesting that miR-33 inhibits the translation of Insig1 protein, thereby indirectly increasing the level of mature Srebp1 protein. Since Srebp1 upregulates the expression of key enzymes in LC-PUFA biosynthesis (i.e., Δ6Δ5 Fads2 and Elovl5), so miR-33 can increase the activity of LC-PUFA biosynthesis in rabbitfish. As miR-33 is expressed with transcription of srebp1, the interaction between Srebp1 and miR-33 might be a self-promoting mechanism, which helps to enhance the function of Srebp1.