ER–mitochondria contact sites regulate hepatic lipogenesis via Ip3r-Grp75-Vdac complex recruiting Seipin

Abstract

Background

Mitochondria and endoplasmic reticulum (ER) contact sites (MERCS) constitute a functional communication platform for ER and mitochondria, and they play a crucial role in the lipid homeostasis of the liver. However, it remains unclear about the exact effects of MERCs on the neutral lipid synthesis of the liver.

Methods

In this study, the role and mechanism of MERCS in palmitic acid (PA)-induced neutral lipid imbalance in the liver was explored by constructing a lipid metabolism animal model based on yellow catfish. Given that the structural integrity of MERCS cannot be disrupted by the si-mitochondrial calcium uniporter (si-mcu), the MERCS-mediated Ca²⁺ signaling in isolated hepatocytes was intercepted by transfecting them with si-mcu in some in vitro experiments.

Results

The key findings were: (1) Hepatocellular MERCs sub-proteome analysis confirmed that, via activating Ip3r-Grp75-voltage-dependent anion channel (Vdac) complexes, excessive dietary PA intake enhanced hepatic MERCs. (2) Dietary PA intake caused hepatic neutral lipid deposition by MERCs recruiting Seipin, which promoted lipid droplet biogenesis. (3) Our findings provide the first proof that MERCs recruited Seipin and controlled hepatic lipid homeostasis, depending on Ip3r-Grp75-Vdac-controlled Ca²⁺ signaling, apart from MERCs’s structural integrity. Noteworthy, our results also confirmed these mechanisms are conservative from fish to mammals.

Conclusions

The findings of this study provide a new insight into the regulatory role of MERCS-recruited SEIPIN in hepatic lipid synthesis via Ip3r-Grp75-Vdac complex-mediated Ca²⁺ signaling, highlighting the critical contribution of MERCS in hepatic lipid homeostasis.

Introduction

Numerous membrane-bound and membrane-less compartmentalized organelles in eukaryotic cells are involved in multiple vital biochemical reactions [1]. It has been corroborated by accumulating evidence that these organelles are interconnected rather than isolated, to facilitate the exchange of small molecules and metabolites via membrane contact sites (MCSs) [2, 3]. These MCSs are essential for the maintenance of organelle function and homeostasis, including hepatic lipid homeostasis [4, 5]. Emerging studies indicates that a disruption in effective inter-organelle communication is strongly associated with hepatic steatosis [6, 7], which is a condition characterized by an excessive buildup of neutral lipids that has rapidly spread around the world. In particular, the functional interaction between the two primary organelles (mitochondria and endoplasmic reticulum [ER]) responsible for neutral lipid metabolism has been highlighted in recent studies [8]. Palmitic acid (PA) is the most common saturated fatty acid in common diets. In the past few decades, PA-induced hepatic steatosis has been widely explored [9,10,11]. However, the precise molecular mechanisms through which communication between mitochondria and endoplasmic reticulum regulates PA-induced neutral lipid dysregulation and the mechanism of its involvement in hepatic steatosis have yet to be defined.

To achieve the functional interplay, ER and mitochondria communicate via mitochondria and endoplasmic reticulum contact sites (MERCS) [12], which can be regarded as the interface between the two organelles [13]. To date, the existence of MERCS has been established in various types of cells [14, 15]. These contacts have a specific structure and create microdomains that serve as a critical hub for a variety of intracellular processes [13, 16]. The first functional role associated with the MERCs compartment concerned its involvement in lipid synthesis and transfer [6, 14]. Due to the fact that MERCs are enriched in enzymes that contribute to lipid biosynthesis [16], the combined activity of endoplasmic reticulum and mitochondrial enzymes is also necessary, despite the fact that the endoplasmic reticulum is the primary organelle for lipid biosynthesis [17, 18]. Lipids must be transported from the endoplasmic reticulum to other organelles following synthesis, given these organelles are incapable of de novo phospholipid synthesis and depend exclusively on the endoplasmic reticulum as the only source [14, 16]. Consequently, the functional interplay of the endoplasmic reticulum and mitochondria during phospholipid biosynthesis and transfer is well established [19, 20]. In contrast to phospholipid metabolism, the underlying molecular mechanism for MERCS to control neutral lipid synthesis has not been sufficiently clarified, the focus of some emerging studies has been placed on the critical role of wrappER-mitochondria contacts in the very-low-density lipoprotein secretion and neutral lipid content in the liver [21, 22]. In addition, given several critical enzymes for neutral lipids synthesis are also enrich in MERCs, e.g. fatty acid CoA ligase 4 (Facl4) and diacylgycerol acyltransferase2 (Dgat2) [14, 23], implies that MERCs fulfil a great role in lipogenesis. This also suggests MERCs might be a promising target for the treatment of hepatic steatosis. Thus, additional in-depth mechanistic investigations are eagerly required.

The role of MERCs as a platform for Ca²⁺ transport is also among the most thoroughly described and proven to date [24, 25]. The endoplasmic reticulum is the principal Ca²⁺ storage organelle and has one main family of release channels to mitochondria, namely the Ip3r-Grp75-Vdac channel, which also function as a structural protein complex for MERCs [21, 26]. Briefly, Ca²⁺ efflux from the endoplasmic reticulum traverses the Ip3r-Grp75-Vdac channel across the freely permeable outer mitochondrial membrane, enters the inner mitochondrial membranes, and is accumulated by mitochondrial calcium uniporter (Mcu) [27]. As a fundamental intracellular second messenger, Ca²⁺ also plays an indispensable role in hepatic lipid homeostasis. Hence, it is necessary to examine the regulatory effect of MERCS on neutral lipid synthesis via its physical connection or relevant Ca²⁺ signaling, which may broaden the understanding of the role of MERCS in controlling neutral lipid synthesis.

Fish are the most numerous orders of vertebrates on the planet. Yellow catfish Pelteobagrus fulvidraco, is a freshwater teleost widely distributed in China and other countries and its whole-genome sequence information have been obtained [28]. Moreover, yellow catfish have relatively high lipid accumulation under intensive farming, making them an attractive model for studying hepatic lipid metabolism [29, 30]. Previous studies indicated the similarities in lipid metabolic processes, and regulatory pathways between yellow catfish and mammals [31, 32], suggesting yellow catfish is a good model for application to human lipid metabolism. On the other hand, palm oil, rich in PA, is considered to be an available and affordable lipid source in aquatic feed [33]. However, the PA-induced hepatic lipid accumulation has an adverse effect on aquaculture. Thus, how to reduce PA-induced excess hepatic lipid deposition is an urgent issue for fish nutritionists. We often found that during dietary PA inducing hepatic lipid accumulation always accompanied by the considerably alteration of physical interactions between endoplasmic reticulum and mitochondria in yellow catfish in our previous studies. Given liver are enriched in MERCs, to understand the role of this interaction, proteomics was used to examine the liver fractions from yellow catfish. Our results clarify systemic control mechanisms for MERCs in neutral lipids synthesis by recruiting Seipin. Importantly, the essential role of Ip3r-Grp75-voltage-dependent anion channel (Vdac) complex and their relevant Ca²⁺ signaling for MERCs recruiting Seipin and subsequent lipogenesis are revealed. These findings revealed a previously unidentified but essential role of MERCs in regulating hepatic lipid homeostasis depending on both its physical connection and relevant Ca²⁺ signaling. Furthermore, our research confirmed the similar results in the HepG2 cell line, thereby establishing the viability of employing Ip3r-Grp75-Vdac-mediated MERCs as a preventive measure against hepatic steatosis.

Materials and methods

Ethical statement

Huazhong Agricultural University’s (HZAU) institutional ethical guidelines for the care and use of laboratory animals were followed throughout all investigations, and were approved by the Ethical Committee of HZAU (identification code: Fish-2022–07-12).

Expt. 1: Animals feeding and sampling (in vivoexperiment)

We formulated six experimental diets according to recent studies [29, 30], as shown in Table S1. Since PA is easy digestible for fish [31], thus PA was taken as the lipid source, and added at the concentrations of 0 g/kg (Control group), 30 g/kg, 60 g/kg, 90 g/kg, 120 g/kg, 150 g/kg, respectively. PA was added at the expense of cellulose. As shown in Table S1, dietary lipid concentrations range from 6.62% to 13.52%. Yellow catfish feeding was determined according to our previous studies [29]. All yellow catfish (1-year-olds) were obtained from a local fish farm (Wuhan, China) and transferred into 300-L circular tanks for 2 weeks to acclimatize. During the acclimation period, all the fish were fed with specialized commercial feed for yellow catfish. At the start of the study, each tank was supplied with thirty fish of uniform size (3.47 ± 0.11 g, mean ± SEM). Each experimental diet was randomly assigned to the three water tanks, and 18 tanks were used in the experiment. The diets were fed to the fish by hand to apparent satiation twice daily (08: 30 and 16: 00 h). Fish body weights were measured every two weeks. The amount of feed consumed by the fish in each tank adopted quantitative feeding based on fish body weight and was recorded daily. Care was taken to ensure that no uneaten feed remained in the tanks during feeding. Faecal matter was also quickly removed during the experiment. Daily mortality statistics were kept. The experiment lasted for 8 weeks. During the feeding experiment, the parameters in water quality were followed: dissolved oxygen 6.49 ± 0.13 mg/L; water temperature 28.0 ± 0.5 °C; pH 7.81 ± 0.21; NH4-N 0.062 ± 0.005 mg/L.

At the end of the feeding experiment, all yellow catfish were fasted for 24 h before sampling. MS-222 was used to euthanize all of the yellow catfish. Then all yellow catfish were counted and weighed to measure survival, weight gain (WG), specific growth rate (SGR) and feed conversion rate (FCR). The morphological parameters and growth performance were shown in Table S2. The liver tissues from three fish were sampled from each tank and fixed in 4% buffered formalin or 2.5% glutaraldehyde for the subsequent observations of histology and ultrastructure. Other fish liver samples were promptly frozen in liquid nitrogen for further examination, including the contents of triglyceride, MERCs extraction, proteomic analysis, gene and protein expression. The blood was drawn from the dorsal aorta by inserting the needle between the anal fin and the caudal fin. The collected blood (3–4 ml) was chilled in ice, centrifuged at 2 °C and 3000 g for 10 min, and a sample of plasma (1-ml) was processed immediately. The extraction of PA from plasma was carried out by the method of Bilinski et al. [32] and the PA were determined by gas chromatography-mass spectrometry (GC–MS)-based analytical methods, essentially as described by Ciucanu et al. [33].

Expt. 2: Cultures and treatments of yellow catfish primary hepatocytes or HepG2 cells (in vitro experiment) 

Primary hepatocyte culture and treatment: yellow catfish (3.47 ± 0.11 g, mean ± SEM) were obtained from the in vivo experiment after two weeks of acclimation. The primary hepatocytes were isolated and cultured from yellow catfish according to our recent publication [29, 30]. In order to further elucidate the underlying mechanisms of PA-enhanced MERCs mediating hepatic lipogenesis, according to our previous studies [30, 34], the cell viability in our pilot experiment (Fig. S1) we select no additional PA and 0.8 mM PA concentration incubated the primary hepatocytes. Meanwhile, based on the physiologically relevant concentrations of PA in plasma from yellow catfish (Table S3), the 0.8 mM PA concentration in vitro is closed to the physiologically relevant PA concentrations in plasma from yellow catfish after feed 150 g/kg PA diet in vivo. The cells were incubated at 28 ℃ for 48 h. Each treatment was performed in triplicate. For each cell culture, a pool of cells from three fish was used. In order to explore the role of Mcu, Grp75 and Seipin in MERCs-mediated hepatic lipid metabolism, we designed and transfected siRNA against mcu, grp75 and seipin, respectively, to hepatocytes of yellow catfish. The inhibition efficiency of siRNA on target protein expression in this analysis are given in Fig. S2. Meanwhile, in the present study, HepG2 cells were used to identify whether there are similar results in human cell line, with the transfection of siRNA against MCU, GRP75 and SEIPIN for human.

Sample analysis

Oil red O, Hematoxylin and Eosin (H&E), Bodipy 493/503 staining and transmission electron microscopy (TEM) observation

Oil red O and H&E staining tests were conducted according to the manufacture’s instruction, and followed the description in our publication [29, 34, 35]. The ten fields of each sample were quantified by the software Image J (NIH, Bethesda, MD, USA) to get the statistics of the relative areas of lipid droplets (LDs) in the Oil red O staining, the relative areas of vacuoles in the H&E staining.

Bodipy 493/503 staining for the LDs and TEM observation of hepatocytes were carried out according to the protocol which described in our previous publications [29, 34, 35]. The data analysis was conducted using FlowJo v.10 software.

Cell viability and determination of triglyceride

Cell viability was measured with the use of 3-(4,5-dimethylthia -zol-2-yl)-2, 5-diphenyltetrazolium bromide (V13154; Thermo Fisher Scientific) according to our previous publication [29, 34, 35]. The contents of triglyceride were determined by commercial kits (A110-1–1, Nanjing Jiancheng Bioengineering Institute), according to the manufacturer’s instructions. Soluble protein content was analyzed, based on protocols by Bradford [36].

Determination of enzymatic activities

The activities of several enzymes, such as fatty acid synthase (FAS), acetyl coa carboxylase (ACC), were measured as previously described [34, 35]. The liver samples were homogenized in 0.5 mL of ice-cold buffer (0.02 M Tris–HCl, 0.25 M sucrose, 2 mM EDTA, 0.1 M sodium fluoride, 0.5 mM phenylmethyl sulphonyl fluoride, 0.01 M β-mercapto-ethanol, pH 7.4), and centrifuged at 12,000 × g at 4 ℃ for 30 min. The supernatant was collected to assay the activities of two lipogenic enzymes. FAS activity was measured with 20 mM potassium phosphate buffer (pH 6.8), 0.1 mM acetyl-CoA, 0.2 mM malonyl-CoA, and 0.5 mM NADPH. Activities of lipogenic enzymes were determined by reading the change in absorbance at 340 nm/unit of time in the reaction media. ACC activity was measured by commercial kits (BC0415, Beijing Solarbio Science & Technology).

RNA isolation and quantitative real-time PCR analysis (qPCR)

Total RNA was isolated by using the Trizol reagent and then transcribed into the cDNA by using the Reverse Transcription Kit. Analyses on gene transcript levels were conducted through the real-time quantitative PCR (qPCR) method described before [29, 34, 35]. The primer sequences used in this analysis are given in Table S3. A set of nine housekeeping genes (gapdh, b2m, ef1α, 18 s rRNA, tuba, β-actin, hprt1, ubc9, tbp) were selected from our transcriptome database in order to test their transcription stability. Our pilot experiment indicated that gapdh and β-actin (M = 0.27) showed the most stable levels of expression across the experimental conditions, as suggested by geNorm. Thus, the relative expression levels were normalized to the geometric mean of the combination of gapdh and β-actin and calculated using the 2^−ΔΔCt method. For siRNA experiment, the primer sequences of siRNA on target protein are given in Table S4.

Hepatocellular MERCs isolation and sub-proteome analysis

The cortical MERCs were isolated according to a previously published protocol [37]. For MERCs isolation and sub-proteome analysis, iTRAQ sample preparation, LC–MS/MS and data analysis was performed according to the methods described in our previous study [34, 35]. Protein lysates were obtained using lysis buffer followed by centrifugation at 12,000 rpm for 60 min at 4 °C to remove cellular debris. Protein samples (250 μg) were digested with trypsin solution over night at 37 °C and labeled with iTRAQ reagents according to the manufacturer’s instructions. Then, typically peptides labeled with 4-plex iTRAQ tags were pooled and dried.

The peptides were separated on a Shimadzu LC-20AB HPLC Pump system coupled with a high pH RP column, which also been described in previous study [38]. The mass spectrometry analysis was performed as described in previous study [38], using Q-Exactive mass spectrometer (ThermoFisher Scientific, San Jose, CA), coupled with LC-20AD (Shimadzu). The software IQuant was used to quantify the labeled peptides with isobaric tags according to our previously published protocol [34, 35]. Two biological replicates were set for better coverage of the target proteome with reliable statistical consistency. All the proteins with a FDR < 1% proceeded with downstream analysis including GO, COG and Pathway. The overview of hepatocellular MERCS sub-proteome analysis was shown in Fig. S3.

Immunoprecipitation and western blot

Immunoprecipitation was performed to identify the interaction among Ip3r, Vdac1, Grp75 and Seipin based on the protocols by Zhao et al. [39]. Briefly, in order to conduct the immunoprecipitation analysis, we lysed the cells in NP-40 buffer (Beyotime, Nantong, China, p0013F) with the addition of a protease inhibitor cocktail (Beyotime, Nantong, China, P1010). The cell lysate was then mixed with anti-Ip3r (ab264281; abcam, Cambridge, UK), anti-Grp75 (ab2799; abcam, Cambridge, UK), anti-Vdac2 (ab14734; abcam, Cambridge, UK), anti-Mcu (ab272488; abcam, Cambridge, UK) or anti-Seipin (ab106793; abcam, Cambridge, UK), respectively, as the bait protein, at 4℃ overnight. The incubation was sustained for 3 h, followed by the addition of protein A/G agarose (P2012; Beyotime, Nantong, China). The immunocomplexes were washed 5 times by PBS supplemented with PMSF protease inhibitor. Finally, the western blot analysis was performed with anti-Ip3r (ab264281; abcam, Cambridge, UK), anti-Grp75 (ab2799; abcam, Cambridge, UK), anti-Vdac2 (ab14734; abcam, Cambridge, UK), anti-Mcu (ab272488; abcam, Cambridge, UK) or anti-Seipin (ab106793; abcam, Cambridge, UK), as the prey protein. To identify the protein levels of Accα (PA5-17,564, Thermo Fisher Scientific, MA, US), Fas (ab128870; abcam, Cambridge, UK), Ip3r, Grp75, Vdac2, Mcu, Seipin, Cpt1, Srebp, Facl4 and Gapdh western blot analysis was performed according to our previous study [35]. Image J was used to measure the densitometry of these bands (Fig. S4).

Immunofluorescence and 2D/or 3D confocal analysis

Immunofluorescence and confocal analysis were utilized to analyze the colocalization of Seipin, Facl4, mitochondria, endoplasmic reticulum, lipid droplets, Ca²⁺ in the hepatocytes of yellow catfish, and the colocalization of Seipin, Facl4, mitochondria, endoplasmic reticulum of yellow catfish or HepG2 cells according to our most recent publications [35, 39]. For immunofluorescence analysis, the hepatocytes of yellow catfish or HepG2 cells were cultured in twelve-well plates and incubated with the corresponding treatments for 48 h, washed thrice with PBS and fixed in 4% paraformaldehyde at room temperature for 10 min, then centrifuge at 850 g for 10 min to collect the cells. Following this, they were incubated with particular primary antibodies, anti-Seipin (ab106793; abcam, Cambridge, UK), anti-Facl4 (ab227256; abcam, Cambridge, UK) (1:200), overnight at 4 °C after being blocked in 5% BSA for 2 h. Following three 5-min PBST washes, the cells were incubated for 60 min at room temperature and in the dark with a secondary anti-body made of goat anti-rabbit IgG H&L (AlexaFluor® 647, 1:500, 150,079; Abcam, MA, USA). The pictures were captured by a laser scanning confocal microscope (Leica Wetzlar, Germany). For the 3D imaging, the images were recorded by employing the Plan-Apochromat × 40 objective with a triple zoom. For mitochondria, endoplasmic reticulum, lipid droplets and Ca²⁺ confocal analysis, hepatocytes were stained by corresponding fluorescent probes, respectively, and then were captured by the laser scanning confocal microscope. The CytoFlex flow cytometer (Beckman Coulter, Miami, FL, USA) was used to measure the LDs and Ca²⁺, which were identified as the green dots. The software FlowJo v.10 was used to analyze the data.

Statistical analysis

All data were expressed as mean ± standard error of means (SEM). The normality of data distribution and the homogeneity of variances were analyzed using the Kolmogorov–Smirnov test and Bartlett’s test, respectively. Then, data were subjected to one-way ANOVA and Duncan’s multiple range test using SPSS 19.0 software, and the minimum significance level was set at P < 0.05. Values without the same letter indicated significant difference among different treatments (P < 0.05); differences between si-NC and si-seipin, grp75 or mcu groups were analyzed using a Student’s t-test for independent samples, *P < 0.05, **P < 0.01.

Results

Excessive dietary PA intake caused hepatic neutral lipid accumulation mainly through de novolipogenesis

Growth performance, morphological parameters and feed utilization were shown in Table S2. The survival was 100% among the six treatments. Compared to other treatments, 120 g/kg PA and 150 g/kg PA diets significantly decreased WG and SGR, also significantly increased FCR.

In the present study, along with dietary PA intake increasing, especially in 150 g/kg PA diet group, the vacuoles in H&E, lipid droplets in Oil Red O, and triglyceride content were significantly aggravated (Fig. 1A-E). This finding suggests that excessive dietary PA intake caused hepatic lipid accumulation and dysregulation. The relative areas for triglyceride staining in the liver of the group fed 150 g/kg PA exceeded 50 percent, demonstrating the establishment of a model for hepatic steatosis. Interestingly, further studies indicated excessive dietary PA intake up-regulated the expression of genes involved in lipogenesis (enzyme: g6pd, 6pgd, acca, fas and dgat2; transcription factor: pparγ, srebp1) and seipin, especially for 150 g/kg PA diet, but not lipolysis (atgl, hsl, cpt1, pparα) (Fig. 1F), which was further proven by the the ACC, FAS activity, the protein expression of Srebp1 and Cpt1 (Fig. 1G), implying lipogenesis might be the main driving force behind excessive dietary PA intake inducing hepatic lipid accumulation. The most prominent phenotype observed among these crucial genes for lipogenesis is Seipin, which localizes at MERCs.

Fig. 1

Excessive dietary PA intake caused hepatic neutral lipid accumulation mainly through lipogenesis. A Representative images of yellow catfish liver tissues stained by H&E, scale bars, 50 μm, and oil red O, scale bars, 50 μm and hepatocellular ultrastructure (TEM). 200 × magnification; HE, hepatocytes; VA, vacuoles; LD, lipid droplet. B Relative areas for hepatic vacuoles in H&E staining. C Relative areas for hepatic vacuoles in oil red O staining. D Lipid droplet mean perimeter in TEM. E Hepatic triglyceride content. F Accα activity. G Fas activity. H Western blot analysis of Seipin, Srebp1 and Cpt1, quantitative analysis for western blot bands shown in Fig. S4. I The mRNA levels of genes involved in hepatic lipid homeostasis. Values are mean ± SEM (n = 3 represents 3 independent tanks and at least six fish were sampled from each tank). P value was calculated by one-way ANOVA and further post hoc Duncan’s multiple range testing. Different letters (a-e) indicate significant differences among the six groups (P < 0.05)

Dietary PA intake enhanced hepatic endoplasmic reticulum and mitochondrion contacts and MERCs structural join

Given the vital role of MERCS on hepatic lipid homeostasis, it is necessary to further identify whether dietary intake of PA can cause the structural and functional alteration of MERCS. As depicted in Fig. 2A-E, the TEM analysis results revealed that dietary intake of PA significantly decreased the contact point distance between the ER and mitochondria, while simultaneously elongating the ER-mitochondrion contact and normalizing the mitochondria perimeter. This finding suggested that dietary intake of PA promoted hepatic ER-mitochondrion contacts. Besides, dietary intake of PA significantly increased the number of extracted MERCS in the liver, indicating the activation of the structural join of MERCS.

Fig. 2

Dietary PA intake enhanced hepatic endoplasmic reticulum and mitochondrion contacts and MERCs structural join. A Representative TEM images for hepatocellular ultrastructure and isolated MERCs. MERCs, mitochondria and endoplasmic reticulum contact sites; MT, mitochondria. B The contact point distance between the endoplasmic reticulum and mitochondrion in panel A. C The length between the endoplasmic reticulum and mitochondrion in panel A. D The percentage of the endoplasmic reticulum-mitochondrion contact length normalized mitochondrion perimeter. E The amount of MERCs number. F Proteins that are involved in lipid metabolism related to MERCS from the hepatocyte MERCs sub-proteome. G Proteins that are involved in MERCS tethering related to MERCS from the hepatocyte MERCs sub-proteome. H Proteins that are involved in MERCs Ca.²⁺ signaling related to MERCS from the hepatocyte MERCs sub-proteome. I Western blot analysis of Ip3r-Grp75-Vdac1 axis and Seipin. Quantitative analysis for western blot bands shown in Fig. S4. J The diagram of differentially expressed MERCs protein in hepatocellular MERCs sub-proteome. Values are mean ± SEM (n = 3 represents 3 independent tanks and at least six fish were sampled from each tank). P value was calculated by one-way ANOVA and further post hoc Duncan’s multiple range testing. Different letters (a-e) indicate significant differences among the six groups (P < 0.05)

On the other hand, the potential and pivotal role of MERCs-relevant protein on hepatic lipid metabolism has been revealed in emerging evidence [4, 6, 7]. Then, using hepatocellular MERCs sub-proteome analysis, the differentially expressed MERCS proteins were analyzed. First, overall, 13,433 unique peptides corresponding to 2325 proteins were identified. Good coverage (> 95% of the total proteins) was obtained for proteins larger than 10 kDa. MS raw files have been submitted to ProteomeXchange (accession number: 1–20,230,530-123367). Second, MERCs proteins were highly specialized in MERCs tethering, MERCS-related lipid metabolism and MERCs Ca²⁺ signaling, which is consistent with their known functions (Fig. 2F-H). Among these differentially expressed MERCs proteins involved in lipid metabolism, excessive PA diet upregulated Seipin protein expression by more than twofold, which shown the most prominent phenotype. Concerning Ca²⁺ signaling, an increased PA diet resulted in upregulation of Ip3r, Grp75, and Vdac1 by more than 1.5-fold, suggesting the activation of Ip3r/Grp75/Vdac-controlled Ca²⁺ signaling. The alteration of MERCs proteins was further verified by Western blotting (Fig. 2I).

The formation of Ip3r-Grp75-Vdac complexes but not their relevant Ca²⁺ signaling determine the integrity of MERCs

It was demonstrated that excessive dietary intake of PA caused the enrichment of SEIPIN and the activation of Ip3r/Grp75/Vdac signaling at MERCS. On that basis, the regulatory relationship between MERCS-related Ip3r/Grp75/Vdac signaling and SEIPIN was further examined. The IP analysis results confirmed that Ip3r/Grp75/Vdac formed protein complexes (Fig. 3A-C; Fig. S5A-C). Interestingly, it was further indicated that Ip3r/Grp75/Vdac complexes had an interaction relationship with SEIPIN, which interconnected with Grp75 and SEIPIN (Fig. 3D; Fig. S5D). The subcellular localization analysis results revealed that si-grp75 reduced the enrichment of SEIPIN induced by excessive dietary intake of PA at MERCS (Fig. 3E), suggesting the potential role of Ip3r/Grp75/Vdac signaling in the recruitment of SEIPIN by MERCS.

Fig. 3

The formation of Ip3r-Grp75-Vdac complexes and their correlations with Seipin and Mcu in MERCs. A IP analysis for Ip3r-Grp75 interaction. B IP analysis for Grp75-Vdac1 interaction. C IP analysis for Vdac1-Ip3r interaction. D IP analysis for Grp75-Seipin interaction. E Western blot analysis for subcellular localization of Ip3r, Grp75 and Vdac1 in MERCs, Mcu in mitochondrion in hepatocyte of yellow catfish with or without mcu, grp75 or seipin knockdown. Quantitative analysis for western blot bands shown in Fig. S4. F The diagram of Ip3r-Grp75-Vdac complexes and their correlations with Seipin and Mcu. Three independent biological experiments for all data

The structural integrity of MERCS is the cornerstone for maintaining its homeostasis and function [7,8,9]. Mcu, which is situated in the inner mitochondrial membrane, controlled Ca²⁺ signaling but not the structural integrity of MERCs [23, 27]. Thus, si-mcu was designed to further examine if Ip3r/Grp75/Vdac complexes regulated MERCs recruitment Seipin by its Ca²⁺ signaling or simply by its involvement in maintaining the structural integrity of MERCs. First, si-mcu transfection would not influence the subcellular localization of Ip3r/Grp75/Vdac complexes into MERCs (Fig. 3E). Second, additional in vitro confocal analysis revealed that si-grp75 compromised the structural integrity of MERCs, but si-mcu and si-seipin had no discernible impact (Fig. 4A, B), which was further confirmed by the TEM observation (Fig. 4C-G).

Fig. 4

Ip3r-Grp75-Vdac complex but not their relevant Ca²⁺ signaling determine the integrity of MERCs. A Representative confocal images of MERCS in hepatocyte staining with MT (red) and ER (green). B Representative 3D confocal images of MERCs in hepatocyte staining with mitochondrion (red) and endoplasmic reticulum (green). C Representative TEM images for hepatocellular ultrastructure in yellow catfish. scale bars, 1 μm. D The fusion rate between mitochondrion (red) and endoplasmic reticulum (green) in panel A. E The contact point distance between the endoplasmic reticulum and mitochondrion in panel C. F The length between the endoplasmic reticulum and mitochondrion in panel C. G The percentage of the endoplasmic reticulum-mitochondrion contact length normalized mitochondrion perimeter in panel C. All data were expressed as mean ± SEM (n = 3 represents three independent experiments, and three replicates were used for each independent experiment). P value was calculated by Student’s t tests. Asterisk (*) indicates significant differences between two treatments (P < 0.05)

Ip3r-Grp75-Vdac complex and their relevant Ca²⁺ signaling is essential for MERCs recruiting Seipin

Following our previous findings that Ip3r-Grp75-Vdac complexes can regulate the structural integrity of MERCs, the mechanism by which Ip3r/Grp75/Vdac mediates MERCs recruiting Seipin was examined. Initially, subcellular localization analysis revealed, as depicted in Fig. 5A, that the down-regulation of Seipin’s subcellular localization into MERCs was caused by the inhibition of both mcu and grp75. This demonstrated the critical role of the Ip3r-Grp75-Vdac complex and their respective Ca²⁺ signaling in MERCs recruiting Seipin. Co-staining with MERCS and Seipin demonstrated that the colocalization between MERCs and Seipin had diminished significantly following si-mcu and si-grp75 transfection (Fig. 5B). Altogether, these results indicated both Ip3r-Grp75-Vdac complex-maintained MERCs’s structural integrity and their relevant Ca²⁺ signaling are essential for MERCs recruiting Seipin.

Fig. 5

Ip3r-Grp75-Vdac complex and their relevant Ca²⁺ signaling is essential for MERCs recruiting Seipin. A Western blot analysis for subcellular localization of Seipin in mitochondrion, endoplasmic reticulum or MERCS in hepatocyte of yellow catfish with or without mcu, grp75 or seipin knockdown. Quantitative analysis for western blot bands shown in Fig. S4. B Representative confocal images of co-staining between Seipin (red) and MERCs (Facl4, green) in hepatocyte of yellow catfish with or without mcu, grp75 or seipin knockdown. Three independent biological experiments for all data

Ip3r-Grp75-Vdac/Mcu axis-controlled Ca²⁺ signaling and their role on the MERCs-recruited Seipin activating lipogenesis

After confirming that the Ip3r-Grp75-Vdac complex and its associated Ca²⁺ signaling play a crucial role in MERCs recruiting Seipin, more research was conducted to examine the effect of Ip3r-Grp75-Vdac/Mcu axis-controlled Ca²⁺ signaling on lipogenesis activated by MERCs-recruited Seipin. First, co-staining with Ca²⁺ and mitochondria clearly shown both si-mcu and si-grp75 transfection decreased the mitochondrial Ca²⁺ content, suggesting the inhibition of MERCs-related Ca²⁺ signaling (Fig. 6A). Then, both si-mcu and si-grp75 reduced the accumulation of lipid droplets in hepatocytes (Fig. 6B-F) in this study, indicating that the Ip3r-Grp75-Vdac complex and its associated Ca²⁺ signaling play an upstream regulatory function in hepatic lipogenesis.

Fig. 6

Ip3r-Grp75-Vdac/Mcu axis-controlled Ca²⁺ signaling mediated MERCs-recruited Seipin to activate lipogenesis. A Representative confocal images of co-staining between mitochondrion (purple) and Ca²⁺ (green) in hepatocyte of yellow catfish with or without mcu, grp75 or seipin knockdown. B Representative confocal images of co-staining between Seipin (red) and lipid droplet (green) in hepatocyte of yellow catfish with or without mcu, grp75 or seipin knockdown. C Relative fusion fluorescence intensity for co-staining between MT and Ca²⁺ in panel A. D Relative fluorescence intensity of lipid droplet. E The number of lipid droplet. F Images of hepatocytes stained by Bodipy 493/503 (lipid droplet) output by flow cytometry analysis. All data were expressed as mean ± SEM (n = 3 represents three independent experiments, and three replicates were used for each independent experiment). P value was calculated by Student’s t tests. Asterisk (*) indicates significant differences between two treatments (P < 0.05)

On the other hand, the aforementioned experiment was replicated using the HepG2 cell line. As illustrated in Fig. 7, the structural integrity of MERCs was not significantly impacted by si-MCU and si-SEIPIN transfection. However, si-GRP75 transfection caused substantial damage to MERCs’s structural integrity, consistent with the yellow catfish findings. Meanwhile, both si-MCU and si-GRP75 transfection significantly reduced the MERCs-recruited SEIPIN. Significantly, the deposition of lipid droplets was also reduced in the down-regulated SEIPIN recruitment, further validating the central role of the Ip3r-Grp75-Vdac complex and its relevant Ca²⁺ signaling in facilitating lipogenesis through MERCs-recruited SEIPIN.

Fig. 7

The conservative mechanisms of MERCs recruiting Seipin to activate lipogenesis via Ip3r-Grp75-Vdac complex and relevant Ca²⁺ signaling from fish to mammals. A HepG2 cells stained have been stained for MERCS in representative 3D confocal images with or without mcu, grp75 or seipin knockdown (nucleus, blue; MT, red; ER, green). B Up row: representative confocal images of co-staining between Seipin (green) and MERCs (red) in HepG2 cells with or without treatment with mcu, grp75 or seipin knockdown; Bottom row: images of hepatocytes stained by Bodipy 493/503 for lipid droplet with or without mcu, grp75 or seipin knockdown. Three independent biological experiments for all data

Discussion

Palmitic acid (PA) is the most common saturated fatty acid in common diets, is the most abundant saturated fatty acid. It has also emerged as a prevalent cause of hepatic steatosis [10, 40, 41], a condition primarily caused by the buildup of hepatic neutral lipid [11, 38]. In fish, as the palm oil has been widely used in aquatic feed [42], PA-induced hepatic lipid accumulation has become an adverse factor for aquaculture. Thus, understanding the induction mechanism of PA for lipid deposition also represents an urgent demand for fish nutritionists. Mitochondria and endoplasmic reticulum (ER) contact sites (MERCS) have emerged as central hubs involved in hepatic lipid homeostasis [41]. Nevertheless, the precise function of MERCs in hepatic neutral lipid production remains a subject of extensive research. Furthermore, since coordinating the Ca²⁺ transfer between the endoplasmic reticulum (ER) and the myocardium is one of the most thoroughly studied roles of the MERCs [43], it is critical to determine if the MERCS regulates hepatic lipid homeostasis through its related Ca²⁺ signaling. This study found a novel function of MERCs in hepatic neutral lipid synthesis by recruiting Seipin. Significantly, an unanticipated and crucial function for Ca²⁺ signaling associated with the Ip3r-Grp75-Vdac complex in facilitating hepatic lipogenesis activation by MERCs was discovered.

Consistent with our prior findings [30, 34], it is unsurprising that our current study unequivocally demonstrated that an excessive consumption of dietary PA resulted in the accumulation of neutral lipids in the liver, as indicated by the substantially elevated triglyceride content. This increase may be primarily attributed to increased de novo lipogenesis, not just simply up-regulation of esterification process. Amongst all dietary PA groups, 150 g/kg PA-induced hepatic lipid accumulation appears to be the most prominent phenotype, which the relative areas for triglyceride staining in liver exceeded 50%, suggesting the establishment of dietary PA-induced hepatic steatosis model in this group. MERCs are involved in many essential cellular processes and their alteration can lead to pathological conditions [44, 45]. In this regard, increasing evidence reports the dysregulation of MERCs has been associated with hepatic lipid metabolism disorder [11]. Thus, the extent to which MERCs contributes to PA-induced hepatic lipid accumulation was subsequently determined. Our research unequivocally demonstrates that liver endoplasmic reticulum-mitochondrion interactions, namely the constant, uniform distance between endoplasmic reticulum and mitochondrion, increased significantly in tandem with the increase in dietary PA concentration, thereby establishing the presence of MERCs. This observation revealed the involvement of MERCs in PA-induced hepatic neutral lipid accumulation, which is further supported by our in vitro study.

In the exploration into the mechanism of MERCS regulating hepatic lipid metabolism, our proteomics analysis clearly uncovers Seipin located at MERCs and its enrichment of Seipin at MERCs was significantly stimulated by excessive dietary PA intake in liver. As reported in previous studies suggested Seipin is a master regulator of lipid droplets biogenesis and is needed for the delivery of lipid and protein cargo to nascent lipid droplets [46]. Our further studies have shown that Seipin was mainly recruited to MERCs both in vivo and in vitro. Importantly, along with the impaired capability of MERCs recruiting Seipin, the PA-induced hepatic lipid droplets accumulation was effectively alleviated, supporting the notion that Seipin localization at MERCs is functionally relevant to PA-induced hepatic neutral lipid accumulation. In a similar manner, a recent study by Combot et al. [47] revealed that lipid loading enriches a subset of Seipin at MERCs in human and mouse cells. Additionally, the authors hypothesized that diminished mitochondrial ATP production compromises adipocyte properties in Seipin-deficient adipose tissue, thereby contributing to the pathogenesis of lipodystrophy. Recently, it has been determined that the recruitment of Seipin to MERCs-lipid droplets interactions is critical for lipid droplets biogenesis [44, 48]. According to the accumulating data, PA caused hepatic lipid metabolism disturbance by directly regulating the subcellular location of Seipin at MERCs to increase hepatic neutral lipid synthesis. In conjunction with a recent publication, these results offer unique insights into the function of MERCs-mediated Seipin [47, 48].

Which MERCs protein or protein complexes are involved in recruiting Seipin to MERCs sites? Indeed, the cross-talk between endoplasmic reticulum and mitochondrion is complex and modulated in different ways and by numerous protein complexes [13, 16, 49]. Despite the fact that numerous resident proteins at MERCs and their biological functions have been found [50,51,52], additional research is necessary to better comprehend the mechanism by which Seipin is recruited to MERCs. In the recent research, Yan et al. [53] discovered through resolving the structure of human Seipin that Seipin forms a scaffold that helps maintain membrane phospholipid homeostasis and ER surface tension. What’s more, the integrity of ER-mitochondria contact sites is crucial for ORP5/8 function in regulating Seipin-mediated LD biogenesis [43]. These lines of evidence all pointed out that Seipin is being associated with mitochondrial and ER functions. In the current study, MERCs was improved by the PA treatment, and our present proteomics analysis identified a clear enrichment of Ip3r-Grp75-Vdac complexes at MERCs in the liver from the excessive dietary PA group. Recent studies have highlighted the increasing importance of Ip3r-Grp75-Vdac complexes for maintaining MERCs integrity and Ca²⁺ signaling [21, 24, 25]. Our current TEM and confocal data provide further support for this notion, demonstrating that inducible deletion of Grp75 results in extended contact between the endoplasmic reticulum and mitochondrion in nearly all hepatocytes. Given the critical nature of integrity in preserving the homeostasis and functionality of MERCs, our inquiry pertains to the potential involvement of Ip3r-Grp75-Vdac complexes in the recruitment of Seipin to MERCs sites. According to our subcellular localization study and Co-IP data, Ip3r-Grp75-Vdac complexes interact directly with Seipin via Grp75-connected MERCs sites. Notably, further research proposed that the inability to subcellularly localize Seipin at MERCs sites was the consequence of Grp75 knockdown, indicating that the this Ip3r-Grp75-Vdac/Seipin interaction is crucial for MERCs recruiting Seipin. The underlying molecular mechanism of MERCs recruiting Seipin is still limited, although the recent study by Combot et al. [47] have indicated the MERCs-enriched Seipin localizes in the vicinity of calcium regulators SERCA2, IP3R, and VDAC, further partially corroborating our data. Thus, this was the first direct evidence of MERCs recruiting Seipin via its Ip3r-Grp75-Vdac complexes.

Not only the MERCs integrity is important for MERCs recruiting Seipin, Ca²⁺ signaling also the crucial function of the Ip3r-Grp75-Vdac complexes in MERCs. The importance of MERCs in regulating Ca²⁺ transport from the endoplasmic reticulum to the mitochondrion [21, 25] has been well-documented; this is predominantly maintained through the interaction of Ip3r-Grp75-Vdac complexes. Thus, another unanswered question is Ip3r-Grp75-Vdac complexes control MERCs recruiting Seipin just though its structural functions, or also depend on its Ca²⁺ signaling? If the latter, given dysregulation of hepatic Ca²⁺ homeostasis has been linked to hepatic steatosis [54], how do Ca²⁺ signaling-mediated Seipin subcellular localization at MERCs affect hepatic lipid homeostasis? In order to address the aforementioned questions, we used the mitochondrial calcium uniporter (Mcu), a pore-forming and Ca²⁺-conducting subunit of the uniporter holocomplex reside in the inner mitochondrial membrane (IMM), and the MERCs integrity protein Grp75. We creatively inhibited Grp75 and Mcu expression, resulting in impaired Ca²⁺ signaling and MERCs structural functions, respectively. Our confocal and TEM analyses initially validated the viability of establishing the aforementioned two distinct models: Mcu-deficient hepatocytes preserve the structural integrity of MERCs but exhibit Ca²⁺ signaling abnormalities. Notably, our study additionally discovered that inducible Mcu deletion disrupts MERCs recruiting Seipin, which eliminates the buildup of LDs. Recent research indicated that Seipin was found to physically interact with the ER calcium pump SERCA in Drosophila, acting as a positive SERCA regulator [55]. Further research [47] found that Seipin interacted with MERCs-enriched proteins critical for ER calcium homeostasis: SERCA2 and IP3R, and decreased mitochondrial calcium import when Seipin was acutely removed. This study linked function of Seipin to MERCs-mediated Ca²⁺ signaling. Similarly, recent studies also indicated that MERCs-recruited Seipin may play a regulatory role in regulating mitochondrial Ca²⁺ import [48], and promoted lipid storage via calcium-dependent mitochondrial metabolism [56]. What’s more, Li et al. [57] also proved that Seipin regulates lipid metabolism by affecting intracellular calcium distribution. Taken together, these results reveal a novel Ca²⁺ signaling-dependent manner for MERCs recruiting Seipin and regulating hepatic lipid homeostasis, broadening and detailing the role of MERCS in lipogenesis. Significantly, the current study obtained similar results in human cell line, which further confirmed that Ip3r-Grp75-Vdac-mediated MERCs could potentially serve as a viable therapeutic target for hepatic steatosis.

In conclusion, the results of this study demonstrated that excessive consumption of dietary PA induced the recruitment of Seipin by MERCs-related Ip3r-Grp75-Vdac complexes, which then triggered lipid droplets biogenesis and ultimately resulted in hepatic steatosis. Significantly, our study also revealed an unusual method by which MERCs recruits Seipin and regulates hepatic lipid homeostasis in a manner dependent on Ca²⁺ signaling, as opposed to relying just on physical contact. A working model related to the aforementioned mechanism is shown in Fig. 8.

Fig. 8

Graphical conclusions for the mechanism of ER–mitochondria contact sites (MERCs) regulate hepatic lipogenesis via Ip3r-Grp75-Vdac complex recruiting Seipin in yellow catfish