Journal of Steroid Biochemistry and Molecular Biology

journal homepage: www.elsevier.com/locate/jsbmb
Journal of Steroid Biochemistry and Molecular Biology 211 (2021) 105893

Elucidation of the mechanism of NEFA-induced PERK-eIF2α signaling Image pathway regulation of lipid metabolism in bovine hepatocytes
Yan Huang 1, Chenxu Zhao 1, Yezi Kong, Panpan Tan, Siqi Liu, Yaoquan Liu, Fangyuan Zeng, Yang Yuan, Baoyu Zhao, Jianguo Wang *
College of Veterinary Medicine, Northwest A&F University, Yangling, Shaanxi, 712100, PR China

A R T I C L E I N F O

Keywords:
Non-esterified fatty acids
Protein kinase R-like endoplasmic reticulum kinase
Lipid metabolism Dairy cows
A B S T R A C T

During the periparturient transition period, negative energy balance (NEB) characterized by high concentrations of non-esterified fatty acids (NEFA) may cause fatty liver and ketosis in dairy cows. Previous studies have shown that the protein kinase R-like endoplasmic reticulum kinase (PERK) branch of the endoplasmic reticulum stress (ERS) response plays an important role in lipid metabolism in hepatocytes. This study, therefore, investigated the role of the PERK-branch in NEFA-induced fatty liver. Different concentrations of NEFA or GSK2656157 (a novel catalytic inhibitor of PERK) were used to treat hepatocytes isolated from calves. The NEFA treatment signifi-cantly increased the triacylglycerol (TG) content, the phosphorylation level of PERK and eukaryotic initiation factor 2α (eIF2α), and the abundance of glucose-regulated protein 78 (Grp78), C/EBP homologous protein (CHOP), sterol regulatory element-binding protein 1c (SREBP-1c), fatty acid synthase (FASN), peroxisome proliferator-activated receptor-α (PPARα), carnitine palmitoyltransferase 1A (CPT1A), apolipoprotein B (APOB), and the low-density lipoprotein receptor (LDLR). Compared with the 1.2 mM NEFA group, inhibition of PERKactivity further increased the TG content in hepatocytes, the very-low-density lipoprotein (VLDL) content in the
supernatant and the protein abundance of APOB while reducing the expression and nuclear levels of SREBP-1c and PPARα, as well as the expression of CPT1A and CPT2. In conclusion, the results showed that the NEFA- induced PERK-eIF2α signaling pathway promotes lipid synthesis, lipid oxidation, but inhibits the assembly and secretion of VLDL. Therefore, during the transition period, the activation of the PERK-eIF2α signaling pathway in the liver of dairy cows could defeat the acid-induced lipotoxicity and provide energy to alleviate NEB.

1. Introduction

The periparturient period has an important impact on the health and
excessive non-esterified fatty acids (NEFA) into the circulation [1,2]. These NEFAs are used partly for the synthesis of milk fat in the mam- mary gland, but many are also absorbed by the liver [3]. In the liver,milk production of dairy cows. During this period, cows experience the these NEFAs can enter the mitochondria for β-oxidation orchallenges of negative energy balance (NEB) and disturbed lipid meta- bolism, and lipolysis is necessary for metabolic adaptation and offsetting NEB. As a result, lipolysis in adipose tissues induces the release of
re-esterification back to triglycerides (TG) [4,5]. The livers of dairy cows have a low ability to assemble very-low-density lipoprotein (VLDL) [6]. As TG secretion in the form of VLDL particles is reduced, there is an

Abbreviations: NEB, negative energy balance; NEFA, non-esterified fatty acids; TG, triacylglycerol; PERK, protein kinase R-like endoplasmic reticulum kinase; ERS, endoplasmic reticulum stress; UPR, unfolded protein response; IRE-1α, inositol-requiring kinase-1α; ATF6, activating transcription factor 6; ATF4, activating tran- scription factor 4; APOB, apolipoprotein B; eIF2α, eukaryotic initiation factor 2α; Grp78, glucose-regulated protein 78; CHOP, C/EBP homologous protein; FBS, fetal
bovine serum; BSA, bovine serum albumin; SREBP-1c, sterol regulatory element-binding protein 1c; SCD1, stearyl-coenzyme a desaturase 1; ACC, acetyl-CoA carboxylase; ACL, ATP citrate lyase; FASN, fatty acid synthase; PPARα, peroxisome proliferator-activated receptor-α; CPT1A, carnitine palmitoyltransferase 1A; ACOX1, peroxisomal acyl-CoA oxidase 1; SIRT1, sirtuin 1; PGC-1α, peroxisome proliferator-activated receptor gamma coactivator-
1α; CD36, cluster of differentiation
36; L-FABP, liver fatty acid binding protein; LDLR, the low-density lipoprotein receptor; MTTP, microsomal triglyceride transfer protein; VLDL, very-low-density lipoprotein; DAPI, 4,6-diamidino-2-phenylindole; FGF21, fibroblast growth factor 21.
* Corresponding author.
E-mail address: [email protected] (J. Wang).
1 These authors contributed equally to this study.

https://doi.org/10.1016/j.jsbmb.2021.105893

Received 14 December 2020; Received in revised form 25 March 2021; Accepted 31 March 2021
Available online 2 April 2021
0960-0760/© 2021 Published by Elsevier Ltd.

accumulation of TG in the liver, which can result in fatty liver [7]. It is
estimated that during early lactation, more than 50 % of the cows have some degree of TG accumulation in the liver, of which 5–10% have se- vere fatty liver and 30–40% have moderate fatty liver [8]. Because of the serious economic losses caused by fatty liver understanding the mech-
anism of lipid metabolism in the liver is important for the prevention and treatment of fatty liver during the perinatal period.
Endoplasmic reticulum stress (ERS) is caused by the accumulation of unfolded or misfolded proteins in the ER. Accordingly, the unfolded protein response (UPR) alleviates ERS by inhibiting protein synthesis and enhancing the ability of the organelle to properly fold, transport, or degrade proteins [9,10]. Experiments in mice have indicated that ERS is associated with the development of insulin resistance and non-alcoholic fatty liver disease [11,12]. In the liver, tunicamycin, a pharmacological ERS inducer, contributes to lipid accumulation, or hepatic steatosis, indicating that ERS may participate in hepatic steatosis by reducing lipid transport or oxidative metabolism, increasing synthesis or uptake, or by some combination of these [13]. The UPR is initiated by the activation of three major pathways mediated by protein kinase RNA-like ER kinase
(PERK), inositol-requiring kinase-1α (IRE-1α), and activating tran-scription factor 6 (ATF6), respectively [10]. Of these, the PERK branch
plays an essential role in maintaining ER homeostasis through phos- phorylating the α subunit of the eukaryotic initiation factor 2 (eIF2) at
serine 51. This leads to a global inhibition of protein synthesis while enhancing the translation of certain mRNAs, including the activating transcription factor 4 (ATF4), that facilitate the adaptation and promote the survival of stressed cells [14,15]. It has been reported that glucosamine-induced PERK activation attenuates apolipoprotein B (APOB) synthesis, which suggests that the phosphorylation of PERK specifically impairs the ability of hepatocytes to assemble VLDL particles [16]. Besides, other studies have reported that dephosphorylation of
eIF2α can attenuate hepatic steatosis in mice and the livers of
liver-specific ATF4 knockout mice had lower TG contents compared with normal mice [17,18]. However, so far, the effects of activated PERK signaling on the lipid metabolism of hepatocytes in dairy cows remains unclear.
During the perinatal period, the NEFA concentrations in the plasma of dairy cows usually increase significantly, promoting the development of fatty liver. At the same time, studies with dairy cows reveal the ex- istence of ERS in the liver during early lactation, especially in dairy cows with fatty liver [19,20]. Besides, studies in cows using in vitro experi- ments showed that high concentrations of fatty acid can induce ERS [20]. Therefore, we hypothesized that the PERK branch may play a role in NEFA-induced lipid accumulation. The present study aimed to investigate this possibility and identify the underlying mechanism from multiple aspects of lipid metabolism (including lipid synthesis, lipid oxidation, and lipid transport).

2. Materials and methods
2.1. Isolation, culture, and treatment of calf primary hepatocytes

Primary hepatocytes from five Holstein calves were isolated using the collagenase IV perfusion method. Briefly, the caudate liver lobes were obtained through surgical excision under sterile conditions. The livers were then perfused with perfusion solution A (140 mM NaCl, 6.7 mM KCl, 2.5 mM glucose, 10 mM HEPES, and 0.5 mM EDTA; pH
7.2–7.4, 37 ◦C) and perfusion solution B (140 mM NaCl, 6.7 mM KCl, 2.5
mM glucose, 30 mM HEPES, and 5 mM CaCl2; pH 7.2–7.4, 37 ◦C),
respectively, until the liquid became clear. The final perfusion was done with perfusion solution B containing 0.2 mg/mL collagenase IV (Gibco, Thermo Fisher, Waltham, MA, USA) for 15 min which aided in digestion. After 15 min, the digestion was terminated by adding 100 mL fetal bovine serum (FBS, Gibco), and the liver was cut into pieces. The un- digested tissue was removed and the remaining liver parenchymal
mixture was filtered through 100- (150 μm) and 200-mesh (75 μm) cellsieves. The hepatocyte filtrate was centrifuged for 10 min at 1000 rpm×
and the cell pellet was washed twice with RPMI-1640 (Gibco) basic medium. The hepatocytes so obtained were transferred to a 6-well tissue
×culture plate at 1 106 cells/mL using adherent medium (RPMI-1640 basic medium supplemented with 10 % FBS, 10—6 M of insulin, 10—6 M of dexamethasone, and 10 μg/mL of vitamin C). After incubation at 37

⦁ C in 5% CO2 for 4 h, the adherent medium was replaced with growth
medium (RPMI-1640 basic medium supplemented with 10 % FBS) and the growth medium was replaced every 24 h.
After 60 h of culture, the medium was replaced with RPMI-1640 basic medium containing 3.6 % Bovine Serum Albumin (BSA, Sigma,
St Louis, MS, USA). The hepatocytes were maintained in RPMI-1640 basic medium for 6 h and then treated with 0 or 1 μmol/L of GSK2656157 (HY-13820, MCE, Malta, NY, USA) for 1 h before treating
with 0, 0.6, 1.2, or 2.4 mM NEFA for 5 h. The NEFA stock (52.7 mM) solution included oleic acid (22.9 mM, O1008, Sigma), linoleic acid (2.6 mM, L1376, Sigma), palmitic acid (16.8 mM, P5585, Sigma), stearic acid (7.6 mM, S4751, Sigma), and palmitoleic acid (2.8 mM, 376910010, Thermo Fisher, Waltham, MA, USA), and the pH of the NEFA solution was adjusted to 7.4 using hydrochloric acid (1 M).

2.2. Triglyceride and very-low-density lipoprotein content determination

×Hepatocytes were collected as described above, washed twice with PBS, and centrifuged for 10 min at 1000 rpm at 4℃. The cell lysis was carried out with 1% Triton X-100 (ST795, Beyotime Institute of
Biotechnology, China) for 30 min. The TAG content of the hepatocytes was detected using a commercial kit (A110—1; Nanjing Jiancheng Bioengineering Institute, China) according to the manufacturer’s in-
structions. Total protein concentration was detected using a BCA Protein Assay kit (PC0020; Solarbio, China).
×The medium collected after treatment was centrifuged at 2000 rpm for 20 min. The VLDL content was determined using an ELISA kit (E11V0006; Shanghai Bluegene Biotech Co., Ltd., Shanghai, China) as
per the manufacturer’s instructions.

2.3. Oil red O staining

For cell staining, Oil Red O working solution was obtained by diluting the Oil Red O stock solution (O0625, Sigma) 3:2 with double- distilled water. Hepatocytes were fixed in 4% paraformaldehyde for 20 min and washed twice with PBS. Subsequently, hepatocytes were incubated in 60 % isopropanol for 2 min and stained with 1 mL of Oil Red O working solution for 30 min. Hepatocytes were again washed twice with PBS and counterstained with hematoxylin before microscopic analysis.

2.4. Protein extraction and western blotting

The total protein content was extracted from the cultured hepato- cytes and prepared for western blotting (P0013B, Beyotime Institute of Biotechnology). The total protein concentration was detected using a BCA Protein Assay kit (PC0020; Solarbio), which was used to stan- dardize the quantity of total protein loaded onto each gel. Proteins (30
μg per lane) were separated on 8–15 % SDS-polyacrylamide gels and
electrophoretically transferred onto polyvinylidine fluoride membranes. Following the transfer, membranes were blocked in 5% BSA/Tris- buffered saline/Tween at room temperature for 2 h. Blots were incu- bated with phospho-PERK antibody (1:1000; catalog no. 3179, Cell Signaling Technology, Danvers, MA, USA), glucose-regulated protein 78 (GRP78) antibody (1:1000; catalog BA2042, Boster Biological Tech- nology Co., Ltd, CA, USA), PERK antibody (1:1000; catalog no. 5683,
Cell Signaling Technology), phospho-eIF2α antibody (1:1000; catalog Ab32157, Abcam, MA, USA), eIF2α antibody (1:1000; catalog
Ab115822, Abcam), ATF4 antibody (1:1000; catalog OM108094, OmnimAbs, CA, USA), C/EBP homologous protein (CHOP) antibody(1:1000; catalog AC532, Beyotime Institute of Biotechnology), SREBP- 1c antibody (1:1000; catalog OM121949, OmnimAbs), stearyl- coenzyme a desaturase 1 (SCD1) antibody (1:1000; catalog Ab23331, Abcam), FASN antibody (1:1000; catalog D262701, Sangon Biotech Co. Ltd, China), ACC antibody (1:1000; catalog no. 3662, Cell Signaling Technology), ATP citrate lyase (ACL) antibody (1:1000; catalog

Table 1
List of primers used in this study.

Genes Primers used for PCR Length of fragment (bp) For: ACCCATTCGAACGTCTGCCCTATT
18S 130
Rev: TCCTTGGATGTGGTAGCCGTTTCT
PERK For: CTCTTCCATCCTCATCCTCACA 296

BM4399, Boster Biological Technology Co. Ltd.), peroxisome
Rev: TTCACTTCTCGCATTACCTTCTC

proliferator-activated receptor-α (PPARα) antibody (1:1000; catalog PR- 4003, Zhenjiang Hope Biotechnology Co., Ltd, China), carnitine palmi-
toyltransferase 1A (CPT1A) antibody (1:1000; catalog Ab83862, Abcam), CPT2 antibody (1:1000; catalog AF2356, Beyotime Institute of Biotechnology), peroxisomal acyl-CoA oxidase 1 (ACOX1) antibody (1:1000; catalog 10957 1-AP, Proteintech, China), sirtuin 1 (SIRT1)
antibody (1:1000; catalog AF0282, Beyotime Institute of Biotech-
For GACCCTGACTCGGGCTAAAT Rev TGGACAGCGGCACCATATG

ATF4 130
CHOP 129
Grp78 243
eIF2α 166
SREBP-1c 239
For: CTCCTCCTCGGTATGTAATGACT Rev: AGTTCGGTCTCATCTGTATCTGT For: TGGTCTCAGACAACAGCAAG Rev: AGCTCATCTGGCATGGTTTC For: CCTGCTCTCCAGAGTCCAGTCA For: CCTGCTCTCCAGAGTCCAGTCAnology), Peroxisome proliferator-activated receptor Gamma Coac-
For: CTGACAGCTCCATTGACAAGGC
Rev: GGCTTCATGTAGGAATACCCTCtivator-1α (PGC-1α) antibody (1:1000; catalog Ab54481, Abcam), APOB antibody (1:1000; catalog PB1096, Boster Biological Technology Co.Ltd.), APOE antibody (1:1000; catalog AF1921, Beyotime Institute of Biotechnology), cluster of differentiation 36 (CD36) antibody (1:1000; catalog PB0398, Boster Biological Technology Co. Ltd.), liver fatty acid binding protein (L-FABP) antibody (1:1000; catalog PB0644, Boster Biological Technology Co. Ltd.), low-density lipoprotein receptor (LDLR) antibody (1:1000; catalog AF1438, Beyotime Institute of Biotechnology), microsomal triglyceride transfer protein (MTTP) anti-
body (1:1000; catalog D154124, Sangon Biotech Co., Ltd), or β-actin antibody (1:5000; AB0035; Abways, China) overnight at 4 ◦C, followed
by incubation with horseradish peroxidase-conjugated anti-mouse, or anti-rabbit antibody (Sangon Biotech Co. Ltd.) at room temperature for 2
h. Immunoreactive bands were visualized by enhanced chem- iluminescence solution (P0018FS, Beyotime Institute of Biotechnology) and imaged using a simple protein imager (ProteinSimple, Santa Clara, CA, USA). Each sample was run in triplicate. Protein gray intensity was quantified by Image-Pro Plus 6.0 (Media Cybernetics, Bethesda, MD,USA) and normalized to β-actin.

2.5. RNA extraction and qRT PCR assay

The RNA extraction from hepatocytes was carried out using TRIzol according to the manufacturer’s instructions (TaKaRa Biotechnology Co. Ltd., China). The purity and concentration of RNA were measured by
electrophoresis (1% agarose gels) and the ratio of UV absorbance at 260/280 nm. Then, 2 μg of total RNA in each sample was reverse- transcribed to cDNA (Tiangen Biotech Co. Ltd., China) according to the supplier’s protocol. The relative mRNA expression of the target genes was detected using the SYBR qPCR SuperMix Plus kit (Novopro-
tein, China) in the 7500 Real-Time PCR System (Applied Biosystems Inc., MA, USA). All the primer sequences are listed in Table 1. The conditions of qRT PCR were as follows: 95 ◦C for 1 min, followed by 40
cycles of amplification (95 ◦C for 20 s and 60 ◦C for 1 min). Each sample
was run in triplicate. The relative quantitation values normalized to the
gene cycle threshold (Ct) value were extrapolated using the 2—ΔΔCt method. 18S rRNA was stably expressed and was therefore chosen as a
reference used for normalization.

2.6. Immunofluorescence staining of SREBP-1c and PPARα protein
To study SREBP-1c and PPARα expression and localization, calf primary hepatocytes were seeded on coverslips were fixed in 4% para- formaldehyde for 30 min, then washed in 0.1 M PBS, twice for 3 min
each. Hepatocytes were permeabilized with 0.1 % Triton X-100 for 10 min and blocked with 10 % goat serum (C0265, Beyotime Institute of Biotechnology) for 1 h. The slides were then incubated with rabbit anti-
SREBP-1c and anti-PPARα antibody at 4 ◦C overnight and subsequently
with Alexa Fluor 488-conjugated donkey anti-rabbit secondary antibody (ab150073, Abcam) at room temperature for 1 h, respectively, followed by nuclear staining with 4,6-diamidino-2-phenylindole (DAPI) (C1002,
For: CCTGGTGTCCTGTTGTTGTG Rev: GTGTGGTGGTAGTTGTGGAAG
ACC 274
ACL 191
SCD1 251
FASN 110
PPARα 279
For: CAGCGACGTCAGCACACTGGATG Rev: GCATGGCATCTCTCAGGACCAC For: ATGAAGGCTGTGGTGATGGA Rev: TGGTGGTCTTGCTGAGTTGA For: GTCAACCTCACTCTGGATGGA Rev: TCGTGGTGGAACAGGACATAG For: ATCAGATGGCTCCGTTATTACAG
CPT1A 246
Rev: CAGTATTGGCACTTATTCCGATTC For: GGAATCTGTGAAGCCTCTTATGAA Rev: GCCTGGATGTGAGTCGGTAT
CPT2 219
ACOX1 189
For: CCTTCCTTCCTGTCTTGGTATG Rev: TTCAGAGGCACTCACAATGTTC For: TAAGCCTTTGCCAGGTATT Rev: ATGGTCCCGTAGGTCAG
SIRT1 186
PGC-1α 163
For: TGTGTCATAGGTTAGGTGGTGAA Rev: CTGAAGAATCTGGTGGTGAAGTT For: TGCTGCTCTGGTTGGTGAA
APOB 200
APOE 249
Rev: AGGCTCGTTGTTGTACTGATTAG For: GAACAGAATGAGCAAGTGAAGAAC Rev: AGGTCAAGTGATGGCAGAGAA For: GCCGCTTCTGGGATTACCT
Rev: GGTTCCGCAAGTCCTCCAT
MTTP 223
CD36 277
For: GAACAGGATATACCACCAGAATCG For: GAACAGGATATACCACCAGAATCG For: CTCATTGCTGGTGCTGTCATT
LDLR 275
L-FABP 149
Rev: CCTTGGCTAGATAACGAACTCTG For: GGAGGAGTGTGAGATGGAGTT Rev: CCTTCGTCATGGTACTGGTAAC For: AATGCGAGTGTGAAGAGG
Rev: GGTGTCGTAGGAGGAGAA

Beyotime Institute of Biotechnology). The samples were imaged using laser confocal microscopy (Carl Zeiss GmbH, Jena, Germany). Fluores- cent intensity from 3 randomly selected microscopic fields per group
was captured and analyzed. The fluorescence intensity of the specific (SREBP-1c or PPARα) staining co-localizing with DAPI was quantified using Image J software and divide this portion by the total fluorescence intensity of SREBP-1c or PPARα staining.

2.7. Statistical analysis

±The results were expressed as the mean standard error of the mean (SEM). Statistical analysis of the data from three independent experi- ments was carried out by GraphPad Prism 7 software (GraphPad InStat
Software, San Diego, CA, USA). Comparisons among groups were analyzed using repeated measurement ANOVA followed by Sidak’s multiple comparisons test. P < 0.05 was considered statistically signif- icant, and P < 0.01 was considered highly significant.

3. Results
3.1. Effect of NEFA on PERK-eIF2α signaling pathway and lipid accumulation in hepatocytes

To study effects of NEFA on the PERK signaling pathway protein expression, the effects of NEFA (0, 0.6, 1.2, or 2.4 mM) or 1 μM GSK2656157, as well as co-treatment with 1.2 mM NEFA and 1 μM
GSK2656157 on protein and relative mRNA levels of the PERK signaling pathway were investigated in calf primary hepatocytes. The protein abundance of Grp78, ATF4, and CHOP, and the phosphorylation levels
of PERK and eIF2α were found to be greater in hepatocytes treated with
0.6 (P < 0.05), 1.2, or 2.4 (P < 0.01) mM NEFA. Also, the mRNA levels of Grp78, PERK, eIF2α, ATF4, and CHOP were significantly higher in he- patocytes treated with 0.6, 1.2, or 2.4 mM NEFA in comparison with the
control group (P < 0.01). Pretreatment with GSK2656157 alleviated the NEFA-induced increase in protein abundance and mRNA levels of PERK, eIF2α, ATF4, and CHOP vis a vis treatment with 1.2 mM NEFA alone (P < 0.01, Fig. 1A, B and C). These results indicated that NEFA activated the PERK-eIF2α signaling pathway in calf primary hepatocytes. Further- more, the addition of 1.2 and 2.4 mM NEFA promoted TG accumulation in hepatocytes (P < 0.01, Fig. 1D). Oil Red O staining showed that upon treatment with 0.6, 1.2, or 2.4 mM NEFA, there was a significant
increase in the amount of lipid droplet (Fig. 1E). Besides, the treatment
with GSK2656157 significantly increased the NEFA-induced lipid accumulation (P < 0.01), which further showed a link between NEFA, the PERK-eIF2α signaling pathway, and lipid accumulation in
hepatocytes.

3.2. Effect of PERK-eIF2α signaling pathway on NEFA-regulating lipid synthesis
To study whether the NEFA-induced PERK-eIF2α signaling pathway in calf primary hepatocytes influences lipid synthesis, the effect of NEFA
and GSK2656157 on the protein and mRNA levels of SREBP-1c, SCD1, FASN, ACC, and ACL were investigated. Results show that both the protein and mRNA levels of SREBP-1c, SCD, FASN, ACC, and ACL were
increased in response to NEFA (P < 0.01, Fig. 2A, B and C). Additionally,
pretreatment with GSK2656157 alleviated the NEFA-induced increase in the protein abundance of SREBP-1c and the mRNA levels of SREBP-1c,
SCD1, FASN, ACC, and ACL (P < 0.01), compared to 1.2 mM NEFA
alone. As seen in Fig. 2D and E, co-treatment with NEFA and GSK2656157 inhibited the nuclear translocation of SREBP-1c (P < 0.01). These results indicate that the NEFA-induced PERK-eIF2α
signaling pathway mediates an increase in lipid synthesis.

Effects of NEFA or GSK2656157 on the PERK-eIF2α signaling pathway and lipid accumulation in hepatocytes. Hepatocytes were treated with various concentrations of NEFA (0, 0.6, 1.2, and 2.4 mM), or 1 μM of GSK2656157, or co-treatment. (A) Western blots analysis of Grp78, PERK, phospho-PERK (Thr980), eIF2α, phospho-eIF2α(Ser51), ATF4, and CHOP. (B) Relative protein expression levels of Grp78, phospho-PERK/PERK, phospho-eIF2α/eIF2α, ATF4, and CHOP. (C) Relative mRNA expression levels of Grp78, PERK, eIF2α, ATF4, and CHOP. (D) TG content in hepatocytes. (E) Representative Oil Red O staining in hepatocytes (400×). All experiments were done at least three times. Data are expressed as mean ± SEM *P < 0.05; **P < 0.01.

Effects of NEFA or GSK2656157 on lipogenic genes expression in hepatocytes. Hepatocytes were treated with various concentrations of NEFA (0, 0.6, 1.2, and
2.4 mM), or 1 μM of GSK2656157, or co-treatment. (A) Western blot analysis of SREBP-1c, SCD1, FASN, ACC, and ACL. (B) Relative protein expression levels of SREBP-1c, SCD1, FASN, ACC, and ACL. (C) Relative mRNA expression levels of SREBP-1c, SCD1, FASN, ACC, and ACL. (D) Immunofluorescence staining was
conducted to determine the effect of NEFA or GSK2656157 on SREBP-1c nuclear translocation in hepatocytes. Green, SREBP-1c staining; blue, DAPI nuclear staining. Scale bar: 20 μm. (E) Fluorescent intensity from 3 randomly selected microscopic fields per group in Fig. D was captured and analyzed. All experiments were done at least three times. Data are expressed as mean ± SEM *P < 0.05; **P < 0.01.

3.3. Effect of PERK-eIF2α signaling pathway on NEFA-regulating lipid oxidation
The degree of lipid accumulation in calf primary hepatocytes can be partially determined by lipid oxidation. We found that NEFA treatment
increased the protein abundance of PPARα, CPT1A, CPT2, ACOX1, SIRT1, and PGC-1α (P < 0.01, Fig. 3A and B). In comparison with the
control group, the mRNA expression levels of CPT1A, CPT2, ACOX1, SIRT1, and PGC-1α were significantly increased in the 1.2 and 2.4 mM NEFA-treated groups (P < 0.01, Fig. 3C). Compared to 1.2 mM NEFA alone, GSK2656157 pretreatment reduced the NEFA-induced increases
in protein abundance of PPARα, CPT1A, CPT2, and SIRT1, as well as in the mRNA levels of CPT1A, CPT2, ACOX1, PPARα, SIRT1, and PGC-1α (P < 0.01, Fig. 3A, B and C). As seen in Fig. 3D and E, co-treatment with NEFA and GSK2656157 could inhibit the nuclear translocation of PPARα promoted by NEFA (P < 0.01). These results demonstrated that the PERK-eIF2α signaling pathway mediates the NEFA-induced increase in lipid oxidation.

Effects of NEFA or GSK2656157 on fatty acid oxidation genes expression in hepatocytes. Hepatocytes were treated with various concentrations of NEFA (0, 0.6, 1.2, and 2.4 mM), or 1 μM of GSK2656157, or co-treatment. (A) Western blot analysis of CPT1A, CPT2, ACOX1, SIRT1, and PGC-1α. (B) Relative protein expression levels of CPT1A, CPT2, ACOX1, SIRT1, and PGC-1α. (C) Relative mRNA expression levels of PPARα, CPT1A, CPT2, ACOX1, SIRT1, and PGC-1α. (D) Immunofluorescence staining was conducted to determine the effect of NEFA or GSK2656157 on PPARα nuclear translocation in hepatocytes. Green, PPARα staining;
blue, DAPI nuclear staining. Scale bar: 20 μm. (E) Fluorescent intensity from 3 randomly selected microscopic fields per group in Fig. D was captured and analyzed. All experiments were done at least three times. Data are expressed as mean ± SEM *P < 0.05; **P < 0.01.

3.4. Effect of PERK-eIF2α signaling pathway on NEFA-regulating lipid transport
The uptake of NEFA and the secretion of VLDL also play vital roles in the regulation of liver TG content. We found that NEFA treatment
increased the protein abundance of APOB, L-FABP, and LDLR (P < 0.01,
Fig. 4B and C). The protein abundance of APOE was greater in hepato-
cytes treated with 0.6 and 1.2 mM NEFA compared to the control group (P < 0.01). The protein abundance of MTTP was greater in hepatocytes treated with 0.6 (P < 0.05) and 1.2 (P < 0.01) mM NEFA compared to the control group. However, no significant change was observed in cells
treated with 2.4 mM NEFA. The mRNA expression levels of APOE,

MTTP, CD36, LDLR, and L-FABP were significantly increased in the 1.2 and 2.4 mM NEFA-treated groups compared to the control group (P < 0.01, Fig. 4D). GSK2656157 pretreatment increased the protein abun- dance of APOB (P < 0.01), but alleviated the NEFA-induced increases in mRNA levels of APOE, MTTP, CD36, LDLR, and L-FABP versus treatment
with 1.2 mM NEFA alone (P < 0.01, Fig. 4B, C and D). The VLDL content was lower upon 0.6 mM NEFA treatment compared to the control (P < 0.05, Fig. 4A). Additionally, the VLDL content was found to be greater
after co-treatment with NEFA and GSK2656157 compared to 1.2 mM NEFA alone (P < 0.05).

Effects of NEFA or GSK2656157 on lipid transport in hepatocytes. Hepatocytes were treated with various concentrations of NEFA (0, 0.6, 1.2, and 2.4 mM), or 1 μM of GSK2656157, or co-treatment. (A) VLDL content in the cell supernatant. (B) Western blot analysis of APOB, APOE, MTTP, CD36, L-FABP, and LDLR. (C) Relative protein expression levels of APOB, APOE, MTTP, CD36, L-FABP, and LDLR. (D) Relative mRNA expression levels of APOB, APOE, MTTP, CD36, L-FABP, and LDLR. All experiments were done at least three times. Data are expressed as mean ± SEM *P < 0.05; **P < 0.01.

4. Discussion

High levels of NEFA in the blood induced by NEB at the beginning of lactation is characteristic of dairy cows and can be used by other tissues for energy supply. However, elevated NEFA is also considered a critical pathogenic factor of metabolic diseases, including fatty liver and ketosis [21]. Gessner et al. reported that high plasma NEFA was a very critical inducement of the activation of ERS in early lactation [19]. Emerging evidence also indicates that high concentrations of fatty acids can induce ERS in the bovine hepatocytes in vitro [20]. The possible reason is that NEFAs include both saturated and unsaturated fatty acids, among these, saturated fatty acids are less likely to be converted to TG than unsatu- rated fatty acids [22]. Saturated fatty acids, therefore, may have toxic effects on the endoplasmic reticulum and induce endoplasmic reticulum dysfunction, which could induce ERS [22]. ERS causes the activation of UPR, which aims to attenuate ERS and restore ER homeostasis. Mounting evidence suggests that the UPR may mediate cell death by causing inflammation and activation of the inflammasome and apoptosis in cases of non-resolvable ER stress, which in turn may lead tothe induction of metabolic diseases such as fatty liver and diabetes [7, 23]. The UPR has three canonical branches. Compared to IRE-1α and ATF6, PERK is necessary for the regulation of protein translation and the
induction of UPR-related genes [24]. Phosphorylation of PERK can rapidly reduce protein translation to alleviate ERS [25]. Besides, ATF4 can activate Grp78 independent of the endoplasmic reticulum stress elements, which can alleviate ERS by increasing protein folding [26].

Studies in mouse mammary epithelial cells showed that PERK is essen- tial for lipogenic maturation and the deletion of PERK significantly reduced the lipid content of cells [27]. These studies have shown that the
PERK-eIF2α signaling pathway plays an active role in regulating lipo-
genesis. The present study aimed to investigate the potential effects of PERK signaling pathway on the NEFA-mediated lipid metabolism in bovine hepatocytes. Firstly, we can observe that NEFA treatmentinduced an increase in phosphorylation of PERK and eIF2α, and
up-regulation of Grp78, ATF4 and CHOP in hepatocytes, indicated that NEFA was able to activate the PERK pathway in dairy cows. Our data also showed that NEFA could induce lipid accumulation in calf hepa- tocytes, which agrees with other studies [28]. Interestingly, the accu- mulation of lipid in these NEFA-treated hepatocytes was significantly increased upon GSK2656157 treatment, indicating that the activation of the PERK branch could effectively alleviate the accumulation of TG in hepatocytes induced by NEFA. Therefore, we further explored the role of the PERK signaling pathway in NEFA-induced lipid accumulation in hepatocytes.
Lipid metabolism in hepatocytes includes multiple aspects, such as lipid uptake, fatty acid oxidation, lipogenesis, and lipid transport. First, the accumulation of TG and steatosis in hepatocytes can be partly attributed to the elevation of de novo lipogenesis [29]. However, lipo- genesis in the liver is important to alleviate the lipotoxicity of NEFA, facilitate the assembly of VLDL, and deliver energy to other tissues. SREBP-1c, a primary regulator of lipogenesis in the liver, regulates lipid synthesis by affecting the expression of the downstream components

FASN, SCD1, ACL, and ACC [30]. In the present study, the higher abundance of SREBP-1c, FASN, SCD1, ACL, ACC in hepatocytes treated with NEFA suggested that high concentrations of NEFA possibly increased the expression of the lipogenic gene, thereby promoting lipid synthesis. Moreover, our results revealed that the increased expression of SREBP-1c and the mRNA levels elevation of its target genes (FASN, ACC, SCD1, and ACL) induced by NEFA was significantly attenuated by inhibiting PERK activation in calf primary hepatocytes. SREBP synthe- sized on ER membranes is reported to bind to SREBP cleavage-activating protein (SCAP) and is escorted into the Golgi complex where the SREBP is cleaved so that it can enter the nucleus. Insulin-induced gene (INSIG-1), as an endoplasmic reticulum protein, binds to SCAP and in- hibits the transport of SREBP to the Golgi complex [31]. However, the
phosphorylated eIF2α could attenuate global translation, causing the
depletion of INSIG-1, which promotes the activation of SREBP-1c [27]. Similarly, studies in mice pointed out that Grp78 overexpression reduced ERS markers and inhibited SREBP-1c cleavage, resulting in adecrease in hepatic triglycerides and cholesterol [32]. Dephosphoryla- tion of eIF2α inhibits the expression of FASN, ACC, and SCD1, whichmight contribute to reducing hepatic steatosis in mice [17]. These studies substantiate the observations of the present study and alsoindicate that activation of the PERK-eIF2α signaling pathway stimulates
the expression of lipogenic genes, thereby exacerbating the lipid accu- mulation in calf hepatocytes induced by NEFA. Our results suggest that PERK inhibitor can inhibit the activity of SREBP-1c, thereby inhibiting lipid synthesis. However, this result contradicts the increase in TG caused by the inhibition of the PERK signaling pathway (Fig. 1D). The TG content in hepatocytes is not only determined by lipid synthesis, but also by lipid oxidation and transport. Therefore, the increasing TG content upon the inhibition of PERK may be caused effects on lipid oxidation and transport in the hepatocytes.

The energy requirements of cows increase two- to three-fold in the period between three weeks antepartum and three weeks postpartum [4]. Therefore, NEFA, produced by lipolysis, enters the bloodstream and acts as a fuel to provide energy to the tissues throughout the body [1]. More than 25 % NEFA is metabolized by the liver and subsequently oxidized or re-esterified into TG [33]. Increased liver lipid oxidation not only provides energy but also alleviate lipid accumulation. In the liver,the target genes of PPARα, a member of the nuclear receptor family of
ligand-activated transcription factors, include CPT1A, CPT2, andACOX1, of which CPT1A is a critical molecule involved in lipid oxida- tion [34]. PGC1α, as a PPARα coactivator, is also involved in fatty acid oxidation in hepatocytes [35]. SIRT1 can directly deacetylate PGC1α, thereby enhancing the activity of PGC1α [36]. We found that the expression levels of PPARα and its target genes CPT1A, CPT2, and ACOX1 were markedly increased by NEFA treatment in calf primaryhepatocytes. One reason might be that NEFA also acts as signaling molecules to activate PPARα and participate in regulating lipolytic geneexpression to modulate lipid oxidation [37]. Meanwhile, other experi- ments have found that dairy cows with mild fatty liver display increased lipid oxidation in the liver [38]. Li et al. indicated that NEFAs could
increase lipid oxidation by activating the AMPKα signaling pathway[37]. However, Dong et al. have found that cows with moderate fatty liver have a lower abundance of PPARα and its target molecule CPT1A. This resulted in high concentrations of NEFA which could inhibit fatty
acid oxidation in vitro. This report is in contrast to our observations [28]. We speculate that the effect of NEFA on lipid oxidation of hepatocytes may depend on the time of NEFA treatment. Therefore, in the early stages of the development of fatty liver in dairy cows, liver lipid oxidation is enhanced, which assists dairy cows in obtaining more en- ergy during NEB. When the NEFA concentration continues to increase
and exceeds the liver’s metabolic capacity, TG will accumulate in theliver. This might impair the normal function of the liver, thereby affecting the lipid oxidation ability and aggravating the development of the fatty liver. Interestingly, we also showed that the increased expres-
sion of PPARα and its target genes (CPT1A, CPT2, and ACOX1) inducedby NEFA was alleviated after the inhibiting activity of PERK in hepa- tocytes. This indicates that activation of the PERK-eIF2α signaling pathway can promote lipid oxidation in hepatocytes. Similarly, Flister
et al. found that mice exposed to a high-sucrose diet for 30 days developed liver ERS with increased gene expression of PPARα and CPT1A [39]. Moreover, Frank et al. showed that the PERK/ATF4pathway induced the expression of fibroblast growth factor 21 (FGF21) in isolated rat hepatocytes and FGF21 was shown to stimulate hepatic lipid oxidation [40,41]. These studies further validate that the activation of the PERK signaling pathway has a positive effect on lipid oxidation. The results of PERK signaling pathway promoting lipid oxidation could explain the increase in intracellular TG with GSK2656157 treatment. We consider that the PERK signaling pathway has an adaptive role in the early stages of ERS, which is activated to re-establish cell homeostasis and promote oxidation.

Fatty acids are secreted from hepatocytes in the form of VLDL and the decreased ability of VLDL assembly and secretion will also lead to lipid accumulation in perinatal dairy cows [42]. An important reason why dairy cows are susceptible to fatty liver disease during the perinatal period is that the ability of the liver for assembling and secreting VLDL is inherently insufficient [6]. However, high concentrations of plasma NEFA and the accumulation of TG in the liver may further impair the secretion of VLDL during the perinatal period [8]. APOB and APOE are the most important obligate structural components of VLDL, whereas the MTTP is a cofactor that appears to be involved in stages of VLDL biogenesis [43]. The LDLR can degrade the dense APOB-containing li- poproteins, thereby inhibiting the secretion of VLDL in the liver [44]. In our experiments, 0.6 and 1.2 mM NEFA treatment of hepatocytes in vitro for 5 h was found to significantly promote the expression of APOB, APOE, and MTTP which indicated that NEFA treatment could stimulate hepatocytes to enhance VLDL assembly and transport more TG out of hepatocytes. However, from the amount of VLDL in the cell supernatant, it could be inferred that NEFA inhibited the secretion of VLDL at low concentrations, partly due to the increased expression of LDLR [44]. Moreover, inhibition of PERK activity increased the protein abundance of APOB and the content of VLDL in the cell supernatant, which indi-
cated that the activation of the PERK-eIF2α signaling pathway couldinhibit the assembly and secretion of VLDL. Wei et al. reported that the glucosamine-mediated PERK-eIF2α pathway decreased APOB synthesis; the study also suggested that, although phosphorylation of eIF2α could
attenuate global protein translation, APOB was extremely sensitive to the regulation of the PERK branches [16]. Other studies pointed out that fatty acids were found to promote APOB secretion after only 3 h of exposure at moderate concentrations; however, extending the time to 16 h led to a reduced APOB secretion, suggesting that the relationship be- tween lipid-induced ERS and VLDL secretion was found to be parabolic in nature [45]. Overall, the present study confirmed that the lipid accumulation in hepatocytes is partly due to the reduction of TG excreted in the form of VLDL caused by the activation of the PERK signaling pathway.

5. Conclusion
This study provides evidence that the NEFA-mediated PERK-eIF2α signaling pathway can promote lipid de novo synthesis, stimulate lipol-
ysis, and inhibit the assembly and secretion of VLDL, thereby regulating hepatocyte lipid metabolism on treatment with high NEFA concentra- tions (Fig. 5). Therefore, during the transition period, activation of the PERK signaling pathway has been deemed necessary for the stability of lipid metabolism in the liver of dairy cows. Our study contributes to an
improved understanding of the relationship between the PERK-eIF2α
signaling pathway and fatty liver in dairy cows. Although our research focuses on the effect of the PERK branch on the lipid metabolism of dairy cow liver cells, the mechanisms by which the other two branches of the UPR regulate liver lipid metabolism in dairy cows are still unclear. Future research will be directed to unraveling the underlying Schematic diagram of the proposed mechanism of PERK-eIF2α signaling pathway affecting lipid metabolism in hepatocytes. Phosphorylation of PERK and eIF2α promote SREBP-1c and PPARα to enter the nucleus, thereby promoting lipid synthesis and lipid oxidation in hepatocytes. On the other hand, the phos- phorylation of PERK a inhibitory effect on translation of APOB mRNA, causing the defect in VLDL assembly in hepatocytes. In general, the activation of the PERK-eIF2α signaling pathway could alleviate the NEFA-induced lipid accumulation in hepatocytes.mechanism of regulation of liver lipid metabolism in cows.

Authors contributions

YH, CZ and JW conceived and designed the experiments. YH, KY, TP and SL performed the experiments. LY, ZF and YY analyzed the data. YH, ZB and JW wrote the paper. All authors read and approved the final manuscript.
Author statement

No conflict of interest exits in the submission of this manuscript, and manuscript is approved by all authors for publication. All authors declared that the work described was original research that has not been published previously, and not under consideration for publication elsewhere, in whole or in part.

Declaration of Competing Interest

The authors report no declarations of interest.

Acknowledgments

The work was supported by the National Natural Science Foundation of China (grant number 31873032).
References
[1] A.A. Adewuyi, E. Gruys, F.J. van Eerdenburg, Non esterified fatty acids (NEFA) in dairy cattle. A review, Vet. Quart. 27 (2005) 117–126.
[2] E. Peripolli, C. Reimer, N.T. Ha, et al., Genome-wide detection of signatures of selection in indicine and Brazilian locally adapted taurine cattle breeds using whole-genome re-sequencing data, BMC Genomics 21 (2020) 624.
[3] A.W. Bell, Regulation of organic nutrient metabolism during transition from late pregnancy to early lactation, J. Anim. Sci. 73 (1995) 2804–2819.

[4] J.H. Han van der Kolk, J.J. Gross, V. Gerber, et al., Disturbed bovine mitochondrial lipid metabolism: a review, Vet. Q. 37 (2017) 262–273.
[5] A.E. Brickner, J.A. Pires, T.F. Gressley, et al., Effects of abomasal lipid infusion on liver triglyceride accumulation and adipose lipolysis during fatty liver induction in
dairy cows, J. Dairy Sci. 92 (2009) 4954–4961.
[6] S. Oikawa, H. Saitoh-Okumura, M. Tanji, et al., Relevance of serum concentrations of non-esterified fatty acids and very low-density lipoproteins in nulli/primiparous
and multiparous cows in the close-up period, J. Vet. Med. Sci. 79 (2017) 1656–1659.
[7] R. Ringseis, D.K. Gessner, K. Eder, Molecular insights into the mechanisms of liver- associated diseases in early-lactating dairy cows: hypothetical role of endoplasmic
reticulum stress, J. Anim. Physiol. An. N. 99 (2015) 626–645.
[8] G. Bobe, J.W. Young, D.C. Beitz, Invited review: pathology, etiology, prevention, and treatment of fatty liver in dairy cows, J. Dairy Sci. 87 (2004) 3105–3124.
[9] Y. Kawaguchi, D. Hagiwara, T. Miyata, et al., Endoplasmic reticulum chaperone BiP/GRP78 knockdown leads to autophagy and cell death of arginine vasopressin neurons in mice, Sci. Rep. 10 (2020) 19730.
[10] P. Walter, D. Ron, The unfolded protein response: from stress pathway to homeostatic regulation, Science 334 (2011) 1081–1086.
[11] X. Gao, S. Guo, S. Zhang, et al., Matrine attenuates endoplasmic reticulum stress and mitochondrion dysfunction in nonalcoholic fatty liver disease by regulating SERCA pathway, J. Transl. Med. 16 (2018) 319.
[12] H.L. Hu, X.P. Fan, Q. Guo, et al., Silicon dioxide nanoparticles induce insulin resistance through endoplasmic reticulum stress and generation of reactive oxygen species, Part. Fibre Toxicol. 16 (2019) 41.
[13] J.S. Lee, R. Mendez, H.H. Heng, et al., Pharmacological ER stress promotes hepatic lipogenesis and lipid droplet formation, Am. J. Transl. Res. 4 (2012) 102–113.
[14] P.P. Yao, M.J. Sheng, W.H. Weng, et al., High glucose causes apoptosis of rabbit corneal epithelial cells involving activation of PERK-eIF2 alpha-CHOP-caspase-12
signaling pathway, Int. J. Ophthalmol. 12 (2019) 1815–1822.
[15] T. Rzymski, M. Milani, D.C. Singleton, et al., Role of ATF4 in regulation of
autophagy and resistance to drugs and hypoxia, Cell Cycle 8 (2009) 3838–3847.
[16] W. Qiu, Q. Su, A.C. Rutledge, et al., Glucosamine-induced endoplasmic reticulum
stress attenuates apolipoprotein B100 synthesis via PERK signaling, J. Lipid Res. 50 (2009) 1814–1823.
[17] S. Oyadomari, H.P. Harding, Y. Zhang, et al., Dephosphorylation of translation
initiation factor 2alpha enhances glucose tolerance and attenuates hepatosteatosis in mice, Cell Metab. 7 (2008) 520–532.
[18] K. Li, Y. Xiao, J. Yu, et al., Liver-specific gene inactivation of the transcription factor ATF4 alleviates alcoholic liver steatosis in mice, J. Biol. Chem. 291 (2016)
18536–18546.
[19] D.K. Gessner, G. Schlegel, R. Ringseis, et al., Up-regulation of endoplasmic reticulum stress induced genes of the unfolded protein response in the liver of periparturient dairy cows, BMC Vet. Res. 10 (2014) 46.

[20] Y. Zhu, Y. Guan, J.J. Loor, et al., Fatty acid-induced endoplasmic reticulum stress
promoted lipid accumulation in calf hepatocytes, and endoplasmic reticulum stress existed in the liver of severe fatty liver cows, J. Dairy Sci. 102 (2019) 7359–7370.
[21] X. Li, W. Huang, J. Gu, et al., SREBP-1c overactivates ROS-mediated hepatic NF- kappaB inflammatory pathway in dairy cows with fatty liver, Cell. Signal. 27
(2015) 2099–2109.
[22] C.L. Gentile, M.A. Frye, M.J. Pagliassotti, Fatty acids and the endoplasmic reticulum in nonalcoholic fatty liver disease, Biofactors 37 (2011) 8–16.
[23] C. Lebeaupin, D. Vallee, Y. Hazari, et al., Endoplasmic reticulum stress signalling and the pathogenesis of non-alcoholic fatty liver disease, J. Hepatol. 69 (2018)
927–947.
[24] E. Lauressergues, E. Bert, P. Duriez, et al., Does endoplasmic reticulum stress participate in APD-induced hepatic metabolic dysregulation? Neuropharmacology
62 (2012) 784–796.
[25] O. Galindo-Hernandez, I. Cordova-Guerrero, L.J. Diaz-Rubio, et al., Protein translation associated to PERK arm is a new target for regulation of
metainflammation: a connection with hepatocyte cholesterol, J. Cell. Biochem. 120 (2019) 4158–4171.
[26] D.P. Ha, A.S. Lee, Insulin-like growth factor 1-receptor signaling stimulates GRP78 expression through the PI3K/AKT/mTOR/ATF4 axis, Cell. Signal. 75 (2020), 109736.
[27] E. Bobrovnikova-Marjon, G. Hatzivassiliou, C. Grigoriadou, et al., PERK-dependent regulation of lipogenesis during mouse mammary gland development and
adipocyte differentiation, Proc. Natl. Acad. Sci. USA 105 (2008) 16314–16319.
[28] J.H. Dong, J.J. Loor, R.K. Zuo, et al., Low abundance of mitofusin 2 in dairy cows
with moderate fatty liver is associated with alterations in hepatic lipid metabolism, J. Dairy Sci. 102 (2019) 7536–7547.
[29] V.T. Samuel, K.F. Petersen, G.I. Shulman, Lipid-induced insulin resistance:
unravelling the mechanism, Lancet 375 (2010) 2267–2277.
[30] X.W. Li, Y. Li, W.T. Yang, et al., SREBP-1c overexpression induces triglycerides
accumulation through increasing lipid synthesis and decreasing lipid oxidation and VLDL assembly in bovine hepatocytes, J. Steroid Biochem. 143 (2014) 174–182.
[31] D. Guo, Y. Wang, J. Wang, et al., RA-XII suppresses the development and growth of liver Cancer by inhibition of lipogenesis via SCAP-dependent SREBP supression, Molecules 24 (2019) 1829.
[32] H.L. Kammoun, H. Chabanon, I. Hainault, et al., GRP78 expression inhibits insulin and ER stress-induced SREBP-1c activation and reduces hepatic steatosis in mice, J. Clin. Invest. 119 (2009) 1201–1215.

[33] H. Sejersen, M.T. Sorensen, T. Larsen, et al., Liver protein expression in dairy cows with high liver triglycerides in early lactation, J. Dairy Sci. 95 (2012) 2409–2421.
[34] S. Qin, J. Yin, K. Huang, Free fatty acids increase intracellular lipid accumulation
and oxidative stress by modulating PPARalpha and SREBP-1c in L-02 cells, Lipids 51 (2016) 797–805.
[35] G. Haemmerle, T. Moustafa, G. Woelkart, et al., ATGL-mediated fat catabolism
regulates cardiac mitochondrial function via PPAR-alpha and PGC-1, Nat. Med. 17 (2011) 1076–1085.
[36] C. Canto, J. Auwerx, PGC-1 alpha, SIRT1 and AMPK, an energy sensing network
that controls energy expenditure, Curr. Opin. Lipidol. 20 (2009) 98–105.
[37] X. Li, X. Li, H. Chen, et al., Non-esterified fatty acids activate the AMP-activated
protein kinase signaling pathway to regulate lipid metabolism in bovine hepatocytes, Cell Biochem. Biophys. 67 (2013) 1157–1169.
[38] X. Du, T. Shen, H. Wang, et al., Adaptations of hepatic lipid metabolism and mitochondria in dairy cows with mild fatty liver, J. Dairy Sci. 101 (2018)
9544–9558.
[39] K.F.T. Flister, B.A.S. Pinto, L.M. Franca, et al., Long-term exposure to high-sucrose diet down-regulates hepatic endoplasmic reticulum-stress adaptive pathways and
potentiates de novo lipogenesis in weaned male mice, J. Nutr. Biochem. 62 (2018) 155–166.
[40] F.G. Schaap, A.E. Kremer, W.H. Lamers, et al., Fibroblast growth factor 21 is
induced by endoplasmic reticulum stress, Biochimie 95 (2013) 692–699.
[41] M.K. Badman, P. Pissios, A.R. Kennedy, et al., Hepatic fibroblast growth factor 21 is
regulated by PPAR alpha and is a key mediator of hepatic lipid metabolism in ketotic states, Cell Metab. 5 (2007) 426–437.
[42] M. Alves-Bezerra, D.E. Cohen, Triglyceride metabolism in the liver, Compr.
Physiol. 8 (2017) 1–8.
[43] S. Wang, Z. Chen, V. Lam, et al., IRE1alpha-XBP1s induces PDI expression to increase MTP activity for hepatic VLDL assembly and lipid homeostasis, Cell
Metab. 16 (2012) 473–486.
[44] S.L. Larsson, J. Skogsberg, J. Bjorkegren, The GSK2656157 low density lipoprotein receptor prevents secretion of dense apoB100-containing lipoproteins from the liver, J. Biol.
Chem. 279 (2004) 831–836.
[45] T. Ota, C. Gayet, H.N. Ginsberg, Inhibition of apolipoprotein B100 secretion by lipid-induced hepatic endoplasmic reticulum stress in rodents, J. Clin. Invest. 118 (2008) 316–332.