Diacylglycerol O-acyltransferase (DGAT) isoforms play a role in peridroplet mitochondrial fatty acid metabolism in bovine liver

Hepatocellular lipid accumulation characterizes fatty liver in dairy cows. Lipid droplets (LD), specialized organelles that store lipids and maintain cellular lipid homeostasis, are responsible for the ectopic storage of lipids associated with several metabolic disorders. In recent years, non-ruminant studies have reported that LD-mitochondria interactions play an important role in lipid metabolism. Due to the role of diacylglycerol acyltransferase isoforms (DGAT1 and DGAT2) in LD synthesis, we explored mechanisms of mitochondrial fatty acid transport in ketotic cows using liver biopsies and isolated primary hepatocytes. Compared with healthy cows, cows with fatty liver had massive accumulation of LD and high protein expression of the triglyceride (TAG) synthesis-related enzymes DGAT1 and DGAT2, LD syn-thesis-related proteins perilipin 2 (PLIN2) and perilipin 5 (PLIN5), and the mitochondrial fragmentation-related proteins dynamin-related protein 1 (DRP1) and fission 1 (FIS1). In contrast, factors associated with fatty acid oxidation, mitochondrial fusion and mitochondrial electron transport chain complex were lower compared with those in the healthy cows. In addition, transmission electron microscopy revealed significant contacts between LD-mitochondria in liver tissue from cows with fatty liver. Compared with isolated cytoplasmic mitochondria, expression of carnitine palmitoyl transferase 1A (CPT1A) and DRP1 was lower, but mitofusin 2 (MFN2) and mitochondrial electron transport chain complex was greater in isolated peridroplet mitochondria from hepatic tissue of cows with fatty liver. In vitro data indicated that exogenous free fatty acids (FFA) induced hepatocyte LD synthesis and mitochondrial dynamics consistent with in vivo results. Furthermore, DGAT2 inhibitor treatment attenuated the FFA-induced upregulation of PLIN2 and PLIN5 and rescued the impairment of mitochondrial dynamics. Inhibition of DGAT2 also restored mitochondrial membrane potential and reduced hepatocyte reactive oxygen species production. The present in vivo and in vitro results indicated there are functional differences among different types of mitochondria in the liver tissue of dairy cows with ketosis. Activity of DGAT2 may play a key role in maintaining liver mitochondrial function and lipid homeostasis in dairy cows during the transition period


INTRODUCTION
Negative energy balance (NEB) in high-yielding dairy cows is inextricably linked with decreased dry matter intake (DMI) and the increases in peripartal energy requirements to maintain pregnancy and lactation (Bauman andCurrie, 1980, Roche et al., 2017).Severe NEB in dairy cows is a consequence of inadequate metabolic adaptations (Sordillo and Raphael, 2013) and leads to the occurrence of subclinical ketosis, clinical ketosis, and often fatty liver disease, which is characterized by accumulation of lipid droplets (LD) (McFadden, 2020).
The LD are neutral lipid (cholesteryl esters and triglyceride) storage organelles composed of a single layer of phospholipids (Saheki et al., 2016).Accumulation of hepatic LD is an adaptive response to excess free fatty acids (FFA), which provides immediate protection against lipotoxicity (Lee et al., 2020).In contrast, mobilization of FA stored within cells as energy-rich triacylglycerols Diacylglycerol O-acyltransferase (DGAT) isoforms play a role in peridroplet mitochondrial fatty acid metabolism in bovine liver Shuang Wang, 1,2 * Bingbing Zhang, 3 * John Mauck, 4 Juan J. Loor, 4 Wenwen Fan, 2 Yan Tian, 2 Tianjiao Yang, 2 Yaqi Chang, 2 Meng Xie, 2 Ben Aernouts, 5 Wei Yang, 2 † and Chuang Xu 1,2 † in LD is one important way to help maintain the body's energy requirements during starvation (Wang et al., 2008, Zechner et al., 2012).Dysregulation of LD biosynthesis and degradation can affect the balance of intracellular lipid metabolism-related products, resulting in organelle dysfunction and damage, cell injury and malfunction (Lee et al., 1994).The contact and interaction between LD and various organelles (including mitochondria, endoplasmic reticulum, peroxisomes) are more conducive to its function, further regulating the generation, growth, fusion, and degradation of intracellular lipids (Binns et al., 2006, Rambold et al., 2015, Hugenroth and Bohnert, 2020).
Mitochondria, energy factories of cells, are organelles where many vital metabolic reactions occur (Chen et al., 2009, Gao et al., 2018a).In most mammalian cells, mitochondria generate the bulk of adenosine triphosphate (ATP) through the process of oxidative phosphorylation (OXPHOS), which involves 5 protein complexes (C I-V) located in the inner membrane of mitochondria.The interaction of LD and mitochondria promotes cellular metabolism through de novo lipogenesis (DNL) and β-oxidation (Freyre et al., 2019, Mancini et al., 2019).A mouse interscapular brown adipose tissue-related study reported that mitochondria connected with LD have unique bioenergetics characteristics (e.g., increased pyruvate oxidation, electron transport, and ATP synthesis capacities), composition, and kinetics to support LD expansion (Benador et al., 2018).Benador et al. isolated the mitochondria interacting with LD in mouse brown adipocytes by centrifugation and named it peridroplet mitochondria (PDM) (Benador et al., 2018).In addition, some mitochondria that interact with LD cannot be separated by ultracentrifugation and are referred to as LDanchored mitochondria (LDAM) (Cui et al., 2019, Cui andLiu, 2020).
Pulse chase assay tracking in starved mouse embryonic fibroblasts (MEFs) revealed a tight interaction between LD and mitochondria that promoted the direct influx of high concentrations of FA into mitochondria and prevented their accumulation in cytoplasm (Rambold et al., 2015).In contrast, a related study in mouse brown adipose tissue reported that PDM promoted the FFA flow into LD and lessened the toxic damage of FA on mitochondria (Benador et al., 2018).Ketosis or fatty liver in dairy cows are characterized by massive ectopic triglyceride (TAG) accumulation leading to varying degrees of damage (Strang et al., 1998, Sejersen et al., 2012).Mitochondrial dysfunction characterizes severe fatty liver (Gao et al., 2018b), and the livers of dairy cows with mild fatty liver have more mitochondria and produce more ATP (Du et al., 2018).Whether fatty liver or ketosis affect the close interaction between LD and mitochondria and the underlying mechanisms are unknown.
Diacylglycerol O -acyltransferase (DGAT), the final enzyme that catalyzes the synthesis of neutral lipids forming the core of LD, plays a pivotal role in the metabolism of FA during fatty liver (Yang et al., 2022).During starvation-induced autophagy, the inhibition of DGAT1 in mouse embryonic fibroblasts was not related to changes in diacylglycerol (DAG) and ceramide or the increases in reactive oxygen species (ROS) or endoplasmic reticulum stress, but was strongly linked with mitochondrial dysfunction (Nguyen et al., 2017).The changing characteristics and roles of different types of mitochondria (PDM and cytoplasmic mitochondria (CM)) on liver lipid accumulation and whether DGAT plays a direct or indirect role in LD-mitochondrial metabolic regulation of FFA use during states characterized by high concentrations of FFA are largely unknown.
This study aimed to elucidate the metabolic characteristics and role of LD-mitochondria in the liver of dairy cows with fatty liver and calf hepatocytes incubated with high concentrations of FA.A key objective was to further investigate whether and how DGAT could regulate FA in hepatocytes by tuning LD-mitochondrial metabolism.

Animals
The protocol for the current study was approved by the Ethics Committee for the Use and Care of Animals, Heilongjiang Bayi Agricultural University (Daqing, China).Holstein cows from a commercial dairy in Daqing City (Heilongjiang Province) milking 3,000 cows were used in this study.Throughout the trial, every cow was fed an identical diet and had unlimited access to feed (Table 1).Before the experiment's start, cows were checked to ensure there were no other comorbidities.Liver samples were obtained via biopsy, and the hepatic TAG content measured.Cows with hepatic TAG content below 1% were classified as healthy (Control group), whereas those with 5% or greater TAG were classified as having fatty liver.
Blood from 50 Holstein dairy cows with similar number of lactations (median = 3, range = 2 to 4) and days in milk (DIM, median = 9 d, range = 7 to 12) was obtained from the tail vein into a heparin-containing tube and blood β-hydroxybutyric acid (BHBA) immediately measured using a cow blood ketone meter Yicheng Bioelectronics Technology Co.,Ltd.,Beijing,China).Twenty cows were selected for liver biopsy based on blood BHBA levels (10 cows with blood BHB < 1.2 mM and 10 cows with blood BHB > 3.0 mM).Liver biopsies were conducted with an 18-gauge biopsy probe (EMENT-1815, Shenzhen Yiman Technology Co., Ltd., Shenzhen, China) before the morning feeding as Wang et al.: DGAT AND HEPATIC FATTY ACID METABOLISM described previously (Gerloff et al., 1986, Marcos et al., 1990, Swartz et al., 2021).In brief, following local anesthesia with subcutaneous injection of 2% lidocaine hydrochloride (L7780; Solarbio), the surgical areas (near the 11th or 12th right intercostal area) were shaved and disinfected with iodine scrub and 75% alcohol.A 3-mm incision was made in the skin using a sterile scalpel and the biopsy needle inserted through the incision and advanced into the liver.The liver tissue samples were removed and gently and thoroughly washed with sterile isotonic saline solution (IN9000; Solarbio) to remove blood contamination.Liver samples were immediately cut into small pieces using a scalpel and fixed in 4% paraformaldehyde at 4°C or frozen in liquid nitrogen and stored at −80°C until analysis.Lastly, according to clinical symptoms and liver TAG content, 6 dairy cows with fatty liver and 6 control cows were selected for subsequent experiments.Basic information of cows used in the study is provided in Table 2.

Liver Histology
Following a 24 h fixation of liver tissue in 4% paraformaldehyde (BL539A; Biosharp) without methanol, Oil Red O staining was performed.Liver tissue was equilibrated in 20% and 30% sucrose for 12 h and cryopreserved in optimal cutting temperature compound.The slices (6 µm thickness) were allowed to dry at room temperature for 20 min, rinsed with 60% isopropanol, and stained for 8 -10 min using Oil Red O solution (O0625; Sigma).After washing with water, tissues were counterstained with hematoxylin for 1 min and differentiated with acid ethanol for 10 s, stained slices were inspected under a light microscope, and images were captured.The captured images (with at least 3 biological replicates per group and 3 technical replicates per biological replicate) were quantified using Image J (1.52a version; National Institutes of Health).The analysis was performed as follows: the RGB image was converted to an 8-bit greyscale image, and the positive areas for oil red O staining were defined using color threshold function plug in Image J (default method; color space: red, 0-255; green, 0-255; blue, 0-117).Lastly, quantification was performed using the "analyse particles" function in Image J.

Detection of TAG Content
Using an enzymatic assay kit (E1013; Applygen Technologies Inc.) and following the manufacturer's instructions, the content of TAG in liver tissue was measured.Approximately 50 mg of liver tissue was homogenized in 1 mL of lysis buffer.Fifty μL of homogenate were taken and placed in a new centrifuge tube for detecting protein content using the BCA Protein Assay kit (P0009; Beyotime Biotechnology) to normalize TAG content.The remaining tissue homogenate was heated at 70°C for 10 min and then centrifuged at 800 × g for 10 min at 4°C, and the supernatant was collected for the determination of tissue TAG content.

Calf Primary Hepatocyte Isolation and Culture
Newborn calves were purchased from a farm in Heilongjiang Province (China) (female, healthy, fasting).Six healthy 1-d-old Holstein female calves, weighing 35-45 kg, were anesthetized with sodium pentobarbital (50 mg/kg) followed by intravenous heparin (1,500 IU/kg) injection.Primary hepatocytes were isolated from each calf via a 2-step collagenase perfusion as we described previously and used independently as biological replicates in the experiments performed (Yang et al., 2022).Briefly, the caudate process obtained from hepatectomy was transferred into a clean beaker and quickly moved to an ultra-clean workbench.The subsequent work was strictly aseptic to ensure the separation of hepatocytes under sterile conditions.The perfusate A pre-warmed to 37°C (140 mM NaCl, 6.7 mM KCl, 10 mM HEPES, 2.5 mM glucose, and 0.5 mM EDTA, pH 7.4) was delivered into the caudate process via vascular cannula at a constant flow rate at 50 mL/min for 10-15min to flush the surface and internal blood.The liver was then perfused with the same flow rate of 37°C pre-warmed perfusion solution B (140 mM NaCl, 6.7 mM KCl, 30 mM HEPES, 2.5 mM glucose, and 5 mM CaCl 2 , pH 7.4; 37°C) for about 5 min until the liquid became clear.The digestive juice (0.1 g collagenase IV was dissolved in 500 mL perfusion solution B, pH 7.2-7.4)was perfused into the liver continuously at 20 mL/min for a period of 15 min, and fluid turbidity was observed to ensure that the liver structure was dissociated.Enzymatic digestion was stopped by addition of 4°C pre-cooled fetal bovine serum (FBS; SH30070.03;Hyclone Laboratories).After removing the liver tissue capsule, blood vessels and connective tissue with forceps, the minced tissue was resuspended in 4°C pre-cooled RPMI1640 medium (SH30809.01;Hyclone Laboratories), filtered through 100 mesh (150 μm) and 200 mesh (75 μm) cell sieves to purify the cell solution.The hepatocytes were washed twice in RPMI 1640 medium, counted in a trypan blue dye (93595; Sigma-Aldrich) and those with higher than 95% viability were used for experiments.The hepatocyte suspension was then centrifuged, washed, resuspended in adherent medium (RPMI-1640 basic medium supplemented with 10% FBS, 10 −6 mol/L of insulin, 10 −6 mol/L of dexamethasone, 10 μg/mL of vitamin C), and cells counted.Lastly, the hepatocyte suspension was seeded into a 6-well cell culture plate (2 mL per well, 1 × 10 6 cells/mL) and cultured at 37°C in 5% CO 2 .After 4 h, the adherent medium was replaced with growth medium (RPMI-1640 medium containing 10% FBS).The medium was replaced with fresh medium every 24 h until the hepatocytes were cultured for 44 h.

Identification of Primary Calf Hepatocytes
Identification of isolated calf hepatocytes by confocal immunofluorescence detection of cytokeratin-18 (CK-18) expression was as described previously (Jiang et al., 2013, Shabani Azandaryani et al., 2019).The hepatocytes from different calves were cultured individually for 44 h on coverslips in 24-well plates, and were washed with phosphate buffer saline (PBS) and fixed with 4% paraformaldehyde for 30 min at room temperature.After washing with PBS, the cells were incubated for 30 min at room temperature in blocking solution (3% Albumin Fraktion V, 5% normal goat serum, and 0.5% Triton in PBS) to permeabilize and block nonspecific binding, and subsequently incubated in CK-18 antibody (1:200, FITC-66187, Proteintech) diluted in blocking solution overnight at 4 °C.After washing with PBS, coverslips were mounted and imaged on a laser confocal microscope (LSM 5 Pascal, Zeiss, Oberkochen, Germany).

Peridroplet Mitochondrial Isolation and Identification
The PDM and CM of liver tissue and primary calf hepatocytes were isolated and collected according to a method used with liver of Wistar rats (Talari et al., 2023).Briefly, liver tissue (about 0.3 g) and hepatocytes (about 5 × 10 7 cells) were harvested and rinsed in PBS, and suspended in Sucrose-HEPES-EGTA buffer (250 mM sucrose, 5 mM HEPES, 2 mM EGTA, pH 7.2, precooling) supplemented with 2% fatty acid-free BSA.The suspension was then homogenized by 10-20 strokes with a glass-glass Dounce homogenizer.Subsequently, the homogenate was transferred to a 15 mL Eppendorf tube, and centrifuged in a swinging bucket rotor for 10 min at 900 × g at 4°C.The supernatant was carefully collected and transferred to a 15 mL Eppendorf tube, and buffer slowly added (20 mM HEPES, 100 mM KCl, 2 mM MgCl 2 , pH 7.4) to the upper layer without disturbing the below layer.It was then centrifuged for 40 min, 2000 × g at 4°C in a swinging bucket rotor to separate the fat layer.The fat layer was carefully aspirated into a new pre-chilled tube and resuspended in Sucrose-HEPES-EGTA buffer followed by centrifugation in a fixed-angle rotor at 10,400 × g for 10 min at 4°C to obtain PDM pellet at the bottom of the tube.In addition, the cytosolic fraction under the fat layer was aspirated and centrifuged at 10,400 × g for 10 min at 4°C in a fixed-angle rotor to obtain CM.Mitochondria samples were fully lysed with RIPA buffer (supplemented with PMSF) followed by determination of protein concentration using the BCA protein assay kit (Beyotime Biotechnology, Shanghai, China).
The fat layer pre-stripping and post-stripping and PDM and CM from liver and primary calf hepatocytes were stained with Mito-Tracker Red CMXRos (M7512; Invitrogen), BODIPY 493/503 (D3922; Invitrogen) and placed on a 1 mm glass slide and covered with cover glass.Imaging was performed using a laser scanning confocal microscope.

RNA Extraction and Real-Time Quantitative PCR
Total RNA was extracted from frozen liver tissue and hepatocytes using TRIzol (Invitrogen Corporation) according to the manufacturer's protocol.The concentrations of samples were measured using a K5500 microspectrophotometer (Beijing Kaiao Technology Development Ltd.).The quality of RNA was determined by the ratio of 260/280-nm absorbance.The RNA integrity was assessed by 1% agarose gel electrophoresis using a Gel-Loading Solution (AM8556; Life Technologies).The 18S and 28S ribosomal RNA bands were clearly visible, and the intensity of the 28S rRNA bands was approximately twice that of the 18S rRNA bands, indicat-ing that the RNA was intact.All the RNA samples in this study met the following RNA quality threshold: optical density (OD)260/280 = 1.9 to 2.2; OD260/230 ≥ 2.0; 28S:18S≈2.0.Equal amounts (1 μg) of RNA were reverse transcribed into cDNA using M-MLV reverse transcriptase (RNase H-, RR047A, TaKaRa Biotechnology Co. Ltd.).Real-time PCR (qPCR) was performed using an Applied Biosystems 7300 Real-Time PCR System (F.Hoffmann-La Roche AG) using Quantitect SYBR Green PCR Kit (4913914001; Roche).PCR thermal cycle conditions used were as follows: initial denaturation for 3 min at 95°C, 40 cycles of denaturation for 15 s at 95°C, annealing for 1 min at 60°C, extension for 30 s at 72°C and final extension for 5 min at 72°C.Because amplification can be affected by primer efficiency, we determined the PCR amplification efficiency and coefficient of determination (R 2 ) by the slope of the standard curves of using serial 10-fold dilutions of sample cDNA.The target gene amplifications used in this study had PCR efficiencies ranging from 95.8 to 101.3% (Table 3).Evaluated target genes were fatty acid synthase (FASN), stearoyl-CoA desaturase 1 (SCD1), diacylglyc erol acyltransferase 1 (DGAT1), diacylglycerol acyl transferase 2 (DGAT2), carnitine palmitoyltransferase 1A (CPT1A), carnitine palmitoyltransferase 1B (CPT1B), 3-Hydroxyacyl-CoA dehydrogenase (HADH), perilipin 2 (PLIN2), perilipin 3 (PLIN3), perilipin 5 (PLIN5), mitochondrial fusion protein 1 (MFN1), mitochondrial fusion protein 2 (MFN2), dynamin-1-like protein (DRP1), mitochondrial fission 1 protein (FIS1) and oxidative phosphorylation complexes I to V (C I-V) abundance was assayed in duplicate with an Applied Biosystems 7300 real-time quantitative PCR system using SYBR Premix Ex TaqI (F.Hoffmann-La Roche AG).Gene-specific primers (Table 3) were designed using Primer Express software (Applied Biosystems Inc.) according to the published mRNA sequence in GeneBank.The most stable reference genes of ACTB and GAPDH were identified and screened from ACTB, GAPDH, TATA-binding protein (TBP) and 18S rRNA (Supplemental Figure S1) by using the geNorm software (Vandesompele et al., 2002).The relative abundance values were normalized to the geometric mean of the quantification cycle of the reference genes (ACTB and GAPDH), and fold changes in gene expression relative to the mean of each treatment in the control group were calculated using the 2 −ΔΔCt method.The in vivo experiments were performed with 6 cows for each group, and the RT-qPCR reactions were performed in triplicate per cow.For in vitro experiments, each treatment set had 6 biological replicates, and each set of biological replicates included 2 technical replicates.

Mitochondrial and Lipid Droplet Transmission Electron Microscopy Imaging
Ultrastructural changes of mitochondria and LD in the liver and hepatocytes were evaluated using transmission electron microscopy.Fresh liver tissue samples were collected, washed and the blood rinsed with pre-cold saline, dried with filter paper, and cut into small pieces.The hepatocytes were washed twice with cold PBS and collected from 6-well plates using trypsin.Subsequently, the collected liver and hepatocytes were fixed with 2.5% glutaraldehyde (G105907; Aladdin Biotechnology Co., Ltd.) at 4°C for more than 4 h, and post-fixed in 1% osmium tetroxide (1.24505; Merck) for 1 h.After sectioning the samples in Spurr resin, the samples were dried in ethanol, ultrathin sections (50 nm) were cut, stained with 0.2% lead citrate (A10701; Thermo Fisher Scientific) and 4% uranyl acetate (22400; Electron Microscopy Sciences), and photographed with an H-7650 transmission electron microscopy (Hitachi, Japan).Ultrastructural analysis was performed on 4-5 samples for each treatment and at least 5 random fields of view for each sample.

Mitochondrial Membrane Potential Assessment
Mitochondrial membrane potential (MMP) of hepatocytes was assessed by fluorescent probe 5,5′,6,6'-tetrachloro-1,1',3,3′-tetraethylbenzimidazo lylcarbocyanine iodide (JC-1; C2006; Beyotime) as described previously (Zhang et al., 2020).Hepatocytes were incubated with medium containing JC-1 (1:200) for 10 min at 37°C, washed 3 times with fresh medium, and immediately examined and analyzed the JC-1 staining signal by a laser scanning confocal microscope.The relative fluorescence intensity of captured images was quantified using Image J (1.52a version; National Institutes of Health).The analysis was performed as follows: the acquired images were imported into Image J, split into the 3 color RGB channels (red, green, and blue), and the positive staining thresholds were set separately.Lastly, the fluorescence signal intensity was calculated using the measurement function in image J. Quantification of MMP was expressed as a ratio of JC-1 aggregate to JC-1 monomer (red/green) fluorescence intensity.

Statistical Analysis
Results are reported as means ± standard error of the means.The data were analyzed with SPSS22.0 (IBM Corp.) and tested for homogeneity of variance before analyses.Data generated from liver tissue were compared using 2-tailed unpaired Student's t-test.Data from calf hepatocyte treatment comparisons were assessed by one-way ANOVA, and multiplicity for each experiment was adjusted with the Bonferroni procedure to control for type I error rate at 0.05.All results were expressed as means ± standard error of the means (SEM) and a P -value < 0.05 was considered statistically significant.

Changes in Lipid Accumulation in Liver
Numerous LD were visible in the liver of dairy cows with fatty liver, whereas lipid accumulation was barely detectable in the control liver (Figure 1A, B).The hepatic TAG content of dairy cows with fatty liver was greater than the control group (Figure 1C).

LD-Mitochondrial Localization and Mitochondrial Function in Liver
Ultrastructural analysis revealed a large number of accumulations of cytosolic LD in the hepatocytes of cows with fatty liver, and there were obvious contacts with mitochondria (Figure 3A).The PDM and CM were isolated from hepatocytes of cows with fatty liver by sucrose density gradient centrifugation, and were confirmed with fluorescence microscopy (Figure 3B).Compared with the CM, protein abundances of CPT1A and P-ACC/ACC were lower in PDM (Figure 3D, P = 0.0057, P = 0.0240, respectively).Protein abundance of MFN2 (Figure 3E, P = 0.0026) was greater in PDM than CM, whereas DRP1 protein abundances was lower (Figure 4E, P < 0.0001).Furthermore, compared with CM, protein abundance of OXPHOS (C I-V) was greater (P < 0.01) in PDM (Figure 3F, G).

Effects of DGAT Inhibition on LD Synthesis and Mitochondrial Function-Related Proteins and mRNA Abundance
The results of immunofluorescence staining of hepatocytes revealed that all cells were positive for CK 18 staining (Figure 4A).As expected, at the mRNA level, the addition of DGAT1 and DGAT2 inhibitors resulted in an approximately 64% and 72% reduction in the mRNA abundance of DGAT1 and DGAT2, respectively (Figure 4D), and also significantly reduced the protein abundance of DGAT1 and DGAT2 (Figure 4C, P = 0.0012, P < 0.0001, respectively).
Compared with the FFA group, protein abundance of PLIN2, PLIN5, and FIS1 (Figure 4F, G, J, P < 0.0001, P < 0.0001, P = 0.0002, respectively) was lower in cells in the DGAT1i+FFA group.In contrast, protein abundance of MFN2 (Figure 4H, P < 0.0001) was greater in cells with DGAT1i+FFA compared with the FFA group.Compared with the FFA group, protein abundance of PLIN2, PLIN5, DRP1 and FIS1 (Figure 4F, G and I, J, P < 0.0001) was lower in hepatocytes with DGAT2i + FFA.

Effects of DGAT Inhibition on Mitochondrial Membrane Potential and Intracellular ROS Concentration
The flow cytometry analysis and MMP results showed that FFA significantly increased the ROS production (Figure 5C, D, P < 0.0001) and reduced the MMP (Figure 5A, B, P < 0.0001) in hepatocytes compared with those in the control group.Compared with the control group, JC-1 green fluorescence was significantly increased and MMP was significantly decreased in the DGAT1i group (Figure 5A,B, P < 0.0001).In addition, ROS production increased significantly (Figure 5C, D, P < 0.0001).Importantly, DGAT2 inhibition alleviated the FFA induced increase of intracellular ROS (Figure 5C, D, P < 0.0001) as well as FFA-induced decrease of MMP (Figure 5A, B, P = 0.0245).

Effects of DGAT inhibition on mitochondrial function and dynamics
The results of PDM revealed that protein abundance of MFN2 was lower (Figure 6C, P = 0.0052) in hepatocytes with DGAT1i + FFA compared with the FFA group, while it was greater in the DGAT2i + FFA group (P = 0.0121).The CM results revealed that protein abundance of MFN2 was lower in hepatocytes with DGAT1i + FFA compared with the FFA group (P = 0.0022).In addition, the protein abundance of MFN2 was greater in PDM than in CM in hepatocytes with DGAT2i + FFA (P = 0.0015).In contrast, the protein abundance of DRP1 was lower in PDM than in CM in hepatocytes with FFA and DGAT1i + FFA (Figure 6D, P = 0.0010, P = 0.0071, respectively).Mitochondrial function-related results indicated that compared with the FFA group, the protein abundance of C II, C III and C V (Figure 6G, H, J, P = 0.0144, P < 0.0001, P = 0.0011, respectively) was greater in PDM with DGAT2i + FFA.Furthermore, the protein abundance of C II, C III and C V (Figure 6G, H, J, P = 0.0022, P = 0.0002, P = 0.0102, respectively) was greater in PDM than in CM in hepatocytes with DGAT2i + FFA.

DISCUSSION
Hepatocellular lipid accumulation characterizes ketosis and fatty liver, and results partly from an impairment in the capacity for lipid transport to extrahepatic tissues (Mills et al., 1986, Liu et al., 2014).To alleviate cytotoxic effects, excess FFA are esterified into TAG and stored as LD in the cell (Marchese et al., 2022).The LD-Mitochondrion contact contributes to lipid synthesis and LD expansion, which constitute an important energy supply source during starvation stress (Valm et al., 2017, Benador et al., 2018).The obvious contact between LD and mitochondria in hepatocytes of dairy cows with fatty liver may further promote LD expansion in response to excess FFA availability.The enzyme DGAT2, the endoplasmic reticulum (ER) resident transmembrane domain enzyme, was also found to exist in the mitochondriarelated membrane.In non-ruminant cells, its interaction with mitochondria depends on its N-terminal amino acid to regulate fatty acid metabolism (Stone et al., 2009).
Although our previous study reported that inhibition of DGAT, especially DGAT1, led to a reduction of the number of LD and an increase in mitochondrial ROS in hepatocytes (Yang et al., 2022), the effect of DGAT on the biosynthesis of LD and mitochondrial function in hepatocytes under high concentration of FFA is still unknown.In this study, we collected liver samples from cows with clinical ketosis to determine the changes of factors related to LDs biosynthesis, mitochondrial dynamics and function.In addition, DGAT1 and DGAT2 inhibitors were added to the ketosis model simulated by high-concentration FFA stimulation of isolated calf primary hepatocytes in vitro to explore the mechanism of the effect of LDs-mitochondria on lipid metabolism in the liver of cows with fatty liver.
The enzymes DGAT1 and DGAT2 catalyze the esterification of fatty acid acyl-CoA and 3-phosphoglycerol, which is the last step of TAG biosynthesis (Zhang et al., 2018).Our data and a previous study revealed that the expression of DGAT1 and DGAT2 in liver tissue of dairy cows with fatty liver was significantly increased and aggravated the accumulation of TAG in liver tissue (Gross et al., 2013).Therefore, we attempted to alleviate hepatocyte lipotoxicity injury by adding inhibitors of DGAT1 and DGAT2 enzymatic activities.The interesting finding was that the addition of DGAT inhibitors significantly re- duces the transcription and translation levels of the genes, which may be closely related to its impact on the nuclear localization of SREBF1, a key factor in the de novo lipid synthesis in hepatocyte (Li et al., 2015).That is, the reduction of enzyme activity affects lipid homeostasis, and hepatocytes inhibit de novo lipid synthesis through a negative feedback regulatory mechanism, thereby reducing their transcription and translation levels.This finding is consistent with the results of studies on the effects of DGAT1 inhibitors on human prostate cancer cells (Nardi et al., 2019).Other evidence from non-ruminant animals indicated that TAG storage in LD, in addition to exporting hepatocytes in the form of very low-density lipoprotein (VLDL), is an important compensatory mechanism to reduce the risk of cellular lipotoxicity (Sinha et al., 2017, Miranda et al., 2019, Liu et al., 2022).
As a protein directly associated with hepatic LD, PLIN2 is only stable when bound to LD and is ubiquitinated and degraded in the absence of droplets (Libby et al., 2016, Nguyen et al., 2019).PLIN5 is also an important LD-surface protein that mediates the recruitment of mitochondria to the surface of LD, thereby altering bioenergetics in mitochondria (Khodayari et al., 2021).Thus, the greater expression of PLIN2 and PLIN5 in the liver tissue of cows with fatty liver and primary hepatocytes challenged with high concentrations of FFA confirmed their involvement in formation of LD due to excess FFA.In addition to this response, the previous worked also reported greater numbers of mitochondria and de novo lipid synthesis during fatty liver (Du et al., 2018).Although LD and mitochondria have some ability to regulate and adapt to excess intracellular FFA metabolism, there is a limit.For example, exposure of mouse primary hepatocytes and HepG2 to saturated FFA resulted in increased mitochondrial depolarization, cytochrome c release and ROS production, exacerbating liver injury (Li et al., 2008).In addition to tissues with high oxidation capacity such as liver, heart and skeletal muscle, saturated fatty acids can also affect the dynamics and functions of mitochondria in cultured neurons (Sajic et al., 2021).It is obvious that mitochondrial dynamics were imbalanced in the liver tissue of cows with fatty liver and in hepatocytes exposed to high concentrations of FFA in this study.The disruption of mitochondrial dynamics may lead to impaired mitochondrial function, thereby affecting mitochondrial respiration and ROS production (Li et al., 2018).It is noteworthy than in the current study, DGAT2 inhibition alleviated the FFA-induced imbalance in mitochondrial dynamics and, thus, reduced intracellular ROS production.Whether such response was also accompanied by partial ER damage could not be determined.Combined with our previous findings (Yang et al., 2022), the inhibition of DGAT2 may affect LDmitochondrial contact, promote the expansion of LDs, affect mitochondrial dynamics and function, and thus alleviate hepatocyte lipid toxicity.
Metabolic homeostasis refers to the achievement of energy balance through the use and storage of energy (Kim et al., 2021).As a metabolic hub, the liver is more dependent on mitochondria to meet its energy needs in the form of ATP, which in turn influences and regulates energy metabolism (Talari et al., 2023).With the deepening of research, the functional differences between PDM and CM precipitated by high-speed centrifugation have been gradually discovered and explored.For example, murine PDM isolated from interscapular brown adipose tissue compared with CM had enhanced pyruvate oxidation, electron transport and ATP synthesis capacity, while β-oxidation capacity was reduced (Benador et al., 2018).Thus, the lower expression of FFA oxidation-related proteins and the greater mitochondrial respiratory chain complexes in PDM from liver tissue of cows with fatty liver and in hepatocytes challenged with high concentrations of FFA confirmed previous data.In contrast, PDM isolated from liver tissue of Wistar rats fed a standard diet had increased fatty acid oxidation and decreased respiration (Talari et al., 2023), which was attributed to hepatic steatosis and LD accumulation caused by high concentrations of FFA.These data further emphasized the physiological significance of PDM in energy stress states.
Mitochondria are known to undergo continuous cycles of fusion and fission to balance mitochondrial content in cells.Thus, compared with CM, the greater mitochondrial fusion kinetics in PDM isolated from liver tissue of keotic cows and hepatocytes challenged with a highconcentration FFA was consistent with results from liver of Wistar rats (Talari et al., 2023), In addition, DGATs, especially the inhibition of DGAT2, can elicit a negative regulation of PDM fusion and respiration, suggesting that these enzymes also have direct or indirect effects on the functional separation between PDM and CM.
Because lipid metabolism of cells is influenced by a variety of factors such as age (Hahn et al., 2017), duration of in vitro culture, and the basic composition of the culture medium (Witte et al., 2019), and despite the fact that most published studies attempting to determine the mechanisms regulating lipid metabolism in mature bovine liver have been conducted using calf hepatocytes, caution should be exercised when attempting to extrapolate regulatory mechanisms from hepatocytes from 1-dold calves (Graulet et al., 1998, Mashek and Grummer, 2004, Zhang et al., 2024).Moreover, RT qPCR is often applied to quantify gene expression levels, and the 2 -ΔΔCT data analysis method is widely used in relative quantitative studies; however, the method relies heavily on the amplification efficiency of the primers, and there are still some limitations in the analysis of the results, and thus the relevant data need to be further validated by protein expression assays or other methods.Stable internal reference screening is important, but 18S rRNA is considered unsuitable for standardized analysis due to its extremely high abundance.Furthermore, 18S rRNA does not contain a poly-A tail.The cDNA synthesis kits contain random primers should be considered.In addition, despite the importance of mitochondrial function during fatty liver in dairy cows, to date, the functional segregation between PDM and CM has not been studied in dairy cows.Further in-depth research is still needed to ascertain if in states of energy stress, the PDM can promote the sequestration of FFA to prevent lipotoxic damage, while CM could still maintain its normal function.

CONCLUSIONS
Significant functional differences between PDM and CM in the liver tissue of cows with fatty liver and hepatocytes challenged with high-concentrations of FFA.The PDM is more important in promoting mitochondrial fusion.Inhibition of DGAT2 expression in hepatocytes further promotes functional separation between PDM and CM, thereby regulating mitochondrial function and fatty acid metabolism.

NOTES
This work was supported by Heilongjiang Province Natural Science Foundation Excellence Youth Project (YQ2022C029), and China Agriculture Research System (CARS-36).The authors extend their sincere thanks to the members of China Agricultural University and Heilongjiang Provincial Key Laboratory for Prevention and Control of Bovine Diseases for their efforts in these experiments.The authors declare no conflicts of interest.
Wang et al.: DGAT AND HEPATIC FATTY ACID METABOLISM

Figure 1 .
Figure 1.Liver histology and triglyceride (TAG) content in control cows and dairy cows with fatty liver.(A) Oil-red O (200 × magnification, scale bar = 50 µm) staining of liver sections.(B) Quantification of oil-red O staining.(C) Hepatic triacylglycerol (TAG) content in healthy dairy cows (n = 6) and dairy cows with fatty liver (n = 6).Data were analyzed with unpaired t-tests and expressed as mean ± SEM.

Figure 2 .
Figure 2. Hepatic lipid metabolic status and changes in mitochondrial dynamics in control cows and dairy cows with fatty liver.(A) Hepatic protein levels of DGAT1, DGAT2, CPT1A, PLIN2, PLIN5, MFN2, DRP1 and FIS1 in control cows and dairy cows with fatty liver.Representative blots are shown.(B) Quantification of hepatic protein levels of DGAT1, DGAT2, CPT1A.(C) Quantification of hepatic protein levels of PLIN2, PLIN5.(D) Quantification of hepatic protein levels of MFN2, DRP1 and FIS1.(E) Relative hepatic mRNA levels of FASN, SCD1, DGAT1 and DGAT2.(F) Relative hepatic mRNA levels of CPT1A, CPT1B, and HADH.(G) relative hepatic mRNA levels of PLIN2, PLIN3 and PLIN5.(H) Relative hepatic mRNA levels of MFN1, MFN2, DRP1 and FIS1.(I) Western blot analysis of complexes I to V (C I-V) in control cows and dairy cows with fatty liver.(J) Relative protein abundance of complexes I to V (C I-V).(K) Relative hepatic mRNA levels of complexes I to V (C I-V).The data of the control were used to normalize other treatments.Comparisons among groups were calculated using unpaired t-tests and expressed as mean ± SEM.

Figure 3 .
Figure 3. Ultrastructure analysis, mitochondria extraction, identification and functional detection of liver sections.(A) Hepatic ultrastructural analysis of in control cows and dairy cows with fatty liver.(B) Peridroplet mitochondria (PDM) and cytoplasmic mitochondria (CM) were stripped and pelleted from liver tissue of dairy cows with fatty liver by high-speed centrifugation on a sucrose gradient.Fluorescent staining to identify PDM and CM, scale bar = 50 μm.(C) Mitochondria protein levels of CPT1A, P-ACC, ACC, DRP1, MFN2, and FIS1 in PDM and CM.Representative blots are shown.(D) Quantification of mitochondria protein levels of ACC, P-ACC, P-ACC/ACC and CPT1A.(E) Quantification of mitochondria protein levels of MFN2 and DRP1.(F) Western blot analysis of complexes I to V (C I-V) in PDM and CM from hepatocytes of dairy cows with fatty liver.Representative blots are shown.(G) Relative protein abundance of complexes I to V (C I-V).Comparisons among groups were calculated using unpaired t-tests and expressed as mean ± SEM.

Figure 4 .
Figure 4. Identification of primary calf hepatocytes and evaluation of inhibitor effects, and the effect of diacylglycerol acyltransferase 1 (DGAT1) and diacylglycerol acyltransferase 2 (DGAT2) inhibitor on lipid droplet and mitochondrial dynamics in primary calf hepatocytes.(A) Identification of cytokeratin-18 immunofluorescence.(B) Western blot analysis of DGAT1 and DGAT2 with addition of diacylglycerol acyltransferase 1 (DGAT1) and diacylglycerol acyltransferase 2 (DGAT2) inhibitors in hepatocytes, respectively.(C) Relative protein abundance of DGAT1 and DGAT2.(D) Relative mRNA abundance of DGAT1 and DGAT2.Comparisons among groups were calculated using unpaired t-tests.The data presented are the mean ± SEM; ** P ≤ 0.01 indicate differences from control.(E) Hepatocytes protein level of PLIN2, PLIN5, MFN2, DRP1 and FIS1.Representative blots are shown.(F -J) Relative protein abundance of PLIN2, PLIN5, MFN2, DRP1 and FIS1.Data were analyzed using a one-way ANOVA with subsequent Bonferroni correction.The data presented are the mean ± SEM; * P ≤ 0.05, ** P ≤ 0.01 indicate differences from control.# P ≤ 0.05, ## P ≤ 0.01 indicate differences from FFA alone.

Figure 5 .
Figure 5.The effect of DGAT1 and DGAT2 inhibition on mitochondrial membrane potential (MMP), and reactive oxygen species (ROS) in primary calf hepatocytes.(A) MMP staining in primary calf hepatocytes.(B) MMP staining quantification results are presented as the JC-1 aggregates/ JC-1 monomers.(C-D) The level of ROS was analyzed by flow cytometry.A one-way ANOVA with a Bonferroni correction was used to calculate group comparisons.The data presented are the means ± SEM; * P ≤ 0.05, ** P ≤ 0.01 indicate differences from control.# P ≤ 0.05, ## P ≤ 0.01 indicate differences from FFA alone.

Figure 6 .
Figure 6.The effect of DGAT1 and DGAT2 inhibition on mitochondrial function and dynamics in PDM and CM from primary calf hepatocytes.(A) Mitochondria protein levels of CPT1A, MFN2 and DRP1 in PDM and CM.Representative blots are shown.(B -D) Quantification of mitochondria protein levels of CPT1A, MFN2 and DRP1.(E) Western blot analysis of complexes I to V (C I-V) in PDM and CM from primary calf hepatocytes.Representative blots are shown.(F -J) Relative protein abundance of complexes I to V (C I-V).Two-way ANOVA with a Bonferroni correction was used to calculate group comparisons.The data presented are the means ± SEM; * P ≤ 0.05 indicates a significant difference between the 2 groups, **, P ≤ 0.01 indicates an extremely significant difference between the 2 groups.
Wang et al.: DGAT AND HEPATIC FATTY ACID METABOLISM

Table 1 .
Wang et al.: DGAT AND HEPATIC FATTY ACID METABOLISM Formulation and Composition of Basal Diet

Table 3 .
Wang et al.: DGAT AND HEPATIC FATTY ACID METABOLISM Sequences of Primers Used for Real-Time PCR Amplification