condi-Targeting IRE1α and PERK in the endoplasmic reticulum stress pathway attenuates fatty acid-induced insulin resistance in bovine hepatocytes

Endoplasmic reticulum (ER) stress can be induced by various stimuli and triggers the unfolded protein response to activate intracellular signaling pathways that are mediated by 3 ER-resident sensors: inositol requiring protein-1α (IRE1α), PKR-like ER kinase (PERK), and activating transcription factor-6 (ATF6). In nonru-minants, ER stress plays a critical role in hepatic insulin resistance. However, whether ER stress plays a role in nonesterified fatty acid (NEFA)-induced hepatic insulin resistance in dairy cows is still unknown. Experiments were conducted using primary bovine hepatocytes isolated from 5 healthy calves (body weight: 30–40 kg; 1 d old). First, hepatocytes were treated with NEFA (1.2 m M ) for 0.5, 1, 2, 3, 5, 7, 9, or 12 h. Treatment with NEFA elevated abundance of phosphorylated IRE1α and PERK, and cleavage of ATF6, along with the ER stress-associated genes XBP1 , ATF4 , and DNAJC3 , resulting in both linear and quadratic effects. Furthermore, ER Tracker red staining and transmission electron microscopy results indicated that ER was dilated and degranulated in response to NEFA treatment, suggesting that ER stress was induced by NEFA treatment in bovine hepatocytes. Second, to assess the effect of ER stress on NEFA-induced insulin resistance, hepatocytes were treated with different concentrations of NEFA (0, 0.6, 1.2, or 2.4 m M ) for 5 h with or without taurourso-deoxycholic acid (TUDCA, a canonical inhibitor of ER stress). Here, NEFA induced insulin resistance by increasing the abundance of insulin receptor substrate-1 (IRS1) phosphorylation at the inhibitory residue Ser 307 (S307) and decreasing the abundance of phosphorylated protein kinase B (AKT) and glycogen synthase kinase-3β (GSK3β) in a dose-dependent manner. This was accompanied by upregulation of an abundance of gluconeogenic genes [phosphoenolpyruvate carboxykinase ( PEPCK ) and glucose-6-phosphatase ( G6-Pase )]. These detrimental effects of NEFA on insulin signaling could be reversed with TUDCA treatment, indicating a mechanistic link between ER stress and NEFA-induced insulin resistance. In a third experiment, pGPU6/GFP/ Neo vectors containing short hairpin RNA targeting IRE1α were used to silence IRE1α transcription, and GSK2656157 (PERK phosphorylation inhibitor) and 4-(2-aminoethyl) benzenesulfonyl fluoride (AEBSF; an inhibitor of ATF6) were used to block PERK and ATF6 branches, respectively. Notably, the silencing of the IRE1α branch improved NEFA-induced insulin resistance by decreasing phosphorylation of IRS1 (S307) and increasing phosphorylation of AKT and GSK3β, and reducing PEPCK and G6-Pase mRNA abundance, which was likely dependent on IRE1α kinase activity. Similarly, blockage of the PERK branch increased phosphorylation of AKT and GSK3β, and reduced PEPCK and G6-Pase mRNA abundance, but had no effect on phosphorylation of IRS1 (S307). However, re-sults showed that inhibition of the ATF6 branch had no effects on phosphorylation of IRS1, AKT, and GSK3β, and instead found increasing PEPCK and G6-Pase mRNA abundance. Taken together, data in the present study found that impeding IRE1α and PERK signaling might aid in relieving hepatic insulin resistance. However, the more detailed mechanisms of how IRE1α and PERK signaling contribute to hepatic insulin resistance in dairy cows remain to be determined


INTRODUCTION
During the transition period, most dairy cows undergo negative energy balance, a state caused by a low intake of dry matter and an increased demand for glucose to support fetal growth and milk synthesis (Herdt, 1988;Rukkwamsuk et al., 1999).Negative energy balance leads to various physiological changes and could, thus, give rise to the occurrence of pathologic condi-Targeting IRE1α and PERK in the endoplasmic reticulum stress pathway attenuates fatty acid-induced insulin resistance in bovine hepatocytes Zhiyuan Fang, 1 Wenwen Gao, 1 Qianming Jiang, 2 Juan J. Loor, 2 Chenchen Zhao, 1 Xiliang Du, 1 Min Zhang, 1 Yuxiang Song, 1 Zhe Wang, 1 Guowen Liu, 1 Xinwei Li, 1 and Lin Lei 1 * tions and disorders.Among key physiological changes, the increase in plasma concentrations of nonesterified fatty acids (NEFA) due to the gradual decrease in DMI and circulating insulin, both of which contribute to lipolysis, not only provides a source of energy for organs (Hayirli, 2006;Janovick et al., 2011) but may overwhelm the capacity of the liver to oxidize them and facilitate the development of fatty liver and ketosis (Grummer, 1993;Rukkwamsuk et al., 2000;Bobe et al., 2004).
Insulin resistance and hypoinsulinemia are key components of the homeorhetic adaptations in peripartal dairy cows.Although insulin resistance is clearly one aspect associated with predisposition of dairy cows to ketosis and liver dysfunction (Xu et al., 2014;Abuelo et al., 2016;Youssef et al., 2017), mechanisms whereby NEFA may lead to hepatic insulin resistance are not well known.By definition, insulin resistance refers to a condition in which insulin-sensitive tissues (skeletal muscle, liver, and adipose tissue) are less responsive to the action of insulin (Goldstein, 2002).As such, transduction of signals through insulin receptor substrate (IRS), phosphoinositol 3-kinase, protein kinase B (AKT), and glycogen synthase kinase-3β (GSK3β) are diminished (Cignarelli et al., 2019).In nonruminants, hepatic insulin resistance fails to suppress hepatocyte glucose production mediated by 2 key gluconeogenic enzymes, phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase (G6-Pase; Meshkani and Adeli, 2009).Although such mechanisms are clearly essential in ruminants, who must rely on hepatic gluconeogenesis for post-ruminal glucose supply, excessive concentrations of NEFA in vitro induced post-receptor insulin resistance and led to oxidative stress and lipid accumulation (Gao et al., 2018).
In hepatocytes, the endoplasmic reticulum (ER) is a critical site of protein synthesis, detoxification, lipid and glucose metabolism, and Ca 2+ regulation.The onset of ER stress occurs when ER homeostasis is perturbed by numerous disturbances such as lipotoxicity and inflammation, leading to the accumulation of unfolded proteins (Lebeaupin et al., 2018).Consequently, the ER elicits an evolutionarily conserved pathway termed the unfolded protein response (UPR; Walter and Ron, 2011), consisting of 3 signaling cascades: protein kinase RNA-like ER kinase (PERK), inositol requiring protein-1α (IRE1α), and activating transcription factor-6 (ATF6).In nonruminants, ER stress contributes to development of hepatic insulin resistance and progression of other liver diseases (Tabas and Ron, 2011;Cnop et al., 2012;Lei et al., 2016).In periparturient cows, ER stress-regulated genes in the liver, such as AMP-dependent transcription factor 4 (ATF4), 78 kDa glucose-regulated protein (GRP78), and X-box binding protein 1 (XBP1), are upregulated during early lactation relative to the dry period (Khan et al., 2015;Ringseis et al., 2015).Although NEFA lipotoxicity can induce ER stress (Zhu et al., 2019) and contribute to insulin resistance in bovine hepatocytes, it is unknown to what extent each of the 3 ER signaling cascades contributes to these effects.This prompted us to the hypothesis that the 3 branches of the ER stress pathway may contribute to hepatic insulin resistance to a different extent in peripartal dairy cows.Thus, we aim to determine the effects of all 3 branches of ER stress on NEFA-induced insulin resistance in bovine hepatocytes.To address this issue, isolated hepatocytes were treated with NEFA, signaling cascades inhibitors, or specific short hairpin RNA (shRNA) to assess aspects related to insulin sensitivity.

Primary Bovine Hepatocyte Isolation and Culture
All animal procedures in this study protocol were approved by the Ethics Committee on the Use and Care of Animals at Jilin University (Changchun, China;No. 202203001).In the present study, a total of 5 healthy newborn female Holstein calves (BW: 30-40 kg; 1 d old) were purchased from a 1,000-cow dairy farm (Changchun, Jilin Province, China) and housed in a temperature-controlled environment.All calves were fasted for 12 h before liver tissue acquisition.After that, euthanasia was administered by a veterinarian using a barbiturate overdose via the jugular vein, following the American Veterinary Medical Association guidelines for the euthanasia of animals (AVMA, 2020).Animals' deaths were confirmed through checking for the absence of eye reflexes and heartbeat and appearance of glazed or glassy eyes.
Primary bovine hepatocytes were isolated as previously described, with modifications (Gao et al., 2021).Briefly, the liver caudate lobe was obtained through surgical excision.The resective caudate lobe was immediately perfused with perfusion solution A (140 mM NaCl, 6.7 mM KCl, 10 mM HEPES, 2.5 mM glucose, 0.5 mM EDTA; pH 7.2-7.4,37°C) through exposed blood vessels at a flow rate of 50 mL/min for 12 min, and solution B (140 mM NaCl, 6.7 mM KCl, 30 mM HEPES, 2.5 mM glucose, 5 mM CaCl 2 ; pH 7.2-7.4,37°C) at a flow rate of 50 mL/min until the returned liquid became clear.Next, the liver was digested with digestion solution [solution B containing 0.02% collagenase type IV (cat.no.17104019; Gibco); pH 7.2-7.4,37°C] for 20 min.Enzymatic digestion was stopped by covering with 100 mL of cold RPMI-1640 basic medium (Hyclone Laboratories) with 10% fetal bovine serum (Hyclone Laboratories) when the tissue was visually and tactilely deemed digested (mean digestion time of 18 min).Afterward, the liver capsule was peeled off, and the liver was disrupted mechanically by shaking and using tweezers to dissociate cells from the remaining scaffold structures.The resulting cell suspension was then filtered through 100 (150 µm) and 200 mesh (75 µm).To eliminate cell debris and non-parenchymal cells, the resulting cell suspension was washed twice with RPMI-1640 basic medium by centrifugation at 50 × g for 5 min at 4°C.The hepatocyte pellets were resuspended, and the cell suspension was gently mounted over a cushion of 25% Percoll (Sigma-Aldrich).Hepatocytes were purified by centrifugation through the Percoll layer.Cell number and viability were measured using Trypan Blue (Sigma-Aldrich).Only hepatocyte preparations with viability >90% were used.After resuspension in adherent medium (RPMI-1640 basic medium; 10% fetal bovine serum; 10 −6 M insulin; 10 −6 M dexamethasone; 10 µg/mL vitamin C; 1% penicillin/streptomycin), hepatocytes were seeded in 6-well culture plates at a density of 1 × 10 5 cells/cm 2 , and incubated at 37°C in 5% CO 2 .Medium was replaced with growth medium (RPMI-1640 basic medium; 10% fetal bovine serum; 1% penicillin/streptomycin) after 4 h of seeding.Growth medium was replaced with fresh medium every 24 h during the cultivation period.All cells were harvested 72 h after isolation.No splitting of cells was performed in any experiment.
To assess insulin signaling in hepatocytes after treatments, hepatocytes were subsequently stimulated with 100 nM insulin (Sigma-Aldrich) for 10 min, and detailed groupings are described in the figure legends.Cell culture experiments were performed in triplicate.

RNA Extraction and Quantitative Real-Time PCR
Total RNA from harvested hepatocytes was extracted with TRIzol reagent (cat.no.9108; TaKaRa Biotechnology Co. Ltd.) according to the manufacturer's recommendations.The concentration of total RNA was determined with a Gene Quant II RNA/DNA Calculator (Pharmacia Biotech).The 260/280 nm ratio of all samples was 1.8 to 2.0.Then, RNA integrity was tested by electrophoresis on 1% agarose gels.One microgram of total RNA was reverse-transcribed to cDNA in 20 µL of reaction mixture using PrimeScript Reverse Transcriptase Kit (cat.no.6110B; TaKaRa Biotechnology Co. Ltd.).The mRNA abundance was evaluated with the SYBR Green Plus Reagent Kit (cat.no.DRR041A; TaKaRa Biotechnology Co. Ltd.) in a 7500 Real-Time PCR System (Applied Biosystems).Real-time PCR was conducted with an initial denaturation at 94°C for 2 min, followed by 35 cycles of denaturation (94°C, 10 s), annealing (60°C, 15 s), and extension (72°C, 30 s).The final cycle of amplification was carried out at 72°C for 5 min.The PCR amplification efficiency and coefficient of determination (R 2 ) were calculated by the slope of the standard curves of using serial 10-fold dilutions of sample cDNA.Target mRNA was normalized relative to ACTB, and differences in relative abundance determined with the 2 −ΔΔCT method.All primers listed in Table 1 were designed using Primer Express software (Applied Biosystems) and synthesized by Sangon Biotech Co. Ltd.The cycles-to-threshold values of β-actin were not affected by different treatments (Shi et al., 2014;Zhu et al., 2019;Huang et al., 2021; Supplemental Figure S1).Thus, β-actin was deemed suitable as internal control for normalization.

ER Tracker Staining and Image Analysis
For ER detection, we used ER Tracker red dye (cat.no.C1041; Beyotime), an ER-specific probe, according to the supplier's instructions (https: / / figshare .com/articles/ online _resource/ C1041 _ER -Tracker _Red _Endoplasmic _Reticulum _Red _Fluorescent _Probe _/ 14812803); this was followed by nuclear staining with 33258 (Beyotime) for 20 min.Cells were observed using laser confocal microscopy (Fluovie FV1200; Olympus).The image analysis method was performed as reported elsewhere (Hart et al., 2012).Briefly, mean fluorescence intensity and area of the ER (red) and nucleus (blue) were quantified in binary format with Image-Pro Plus 6.0 (Media Cybernetics), values were corrected for background, volume was determined by multiplying the area by intensity, and final ER volumes were divided by nucleus volumes to obtain the relative ER volume value for each cell.

Transmission Electron Microscopy
Hepatocytes were cultured in 6-well plates, washed with serum-free medium in situ, and scraped off gently with a cell scraper, and cell pellets were then collected by centrifugation at 150 × g for 10 min at 4°C. Cell pellets were fixed with 2.5% glutaraldehyde and 2% paraformaldehyde at 4°C for 2 h and washed in PBS 3 times, post-fixed in 1% osmium tetroxide for 1 h at 4°C, and then dehydrated in an ethanol series (70, 80, 90, and 100%) for 2 min each on ice.Cells were rinsed once and infiltrated with Spurrs resin.Ultrathin sections of

Statistical Analysis
All experiments were repeated at least 3 times using independently prepared cell preparations from different animals.Data are expressed as mean ± SEM.Linear and quadratic contrasts were conducted to evaluate time-and dose-dependent effects.Comparisons among groups were analyzed using one-way ANOVA with a Bonferroni post-test, and P < 0.05 was considered statistically significant.All analyses were performed using GraphPad Prism 5 (GraphPad InStat Software, https: / / www .graphpad.com/scientific -software/ instat/ ) and SPSS 20.0 software package (IBM Inc.).

Kinetics of NEFA-Induced ER Stress
Upon treatment with NEFA for different durations, phosphorylation of IRE1α and PERK, and the cleavage of ATF6 (cleaved ATF6), had both linear and quadratic increases (P < 0.01), with peaks at 5 h and 7 h, respectively, followed by a decrease at later time points (Figure 1A-D; Supplemental Table S1, https: / / figshare .com/articles/ figure/ Fang _et _al _Supplemental _material/ 20106476; Lei, 2022).Compared with the 0.5-h group, abundance of XBP1 and ATF4 increased over NEFA treatment duration, reaching their highest level at 5 h (Figure 1E and F; P < 0.01), followed by a decrease at subsequent time points, displaying quadratic trends (Supplemental Table S1; P < 0.01).Abundance of DNAJC3 increased gradually and displayed the highest abundance at 9 h (Figure 1G; P < 0.01), decreasing at 12 h, resulting in both linear and quadratic trends (Supplemental Table S1; P < 0.01).The significant increase of ER/nucleus value (Figure 1H, P < 0.01) after ER Tracker red immunostaining revealed that the ER was enlarged with NEFA treatment.Furthermore, transmission electron microscopy revealed that the rough ER was dilated and degranulated in response to NEFA treatment, compared with the control (Figure 1I).

PERK and its Link with NEFA-Induced Insulin Resistance
Upon NEFA treatment, blocking PERK using GSK2656157 (3 µM) had no significant effect on the level of phosphorylated IRS (S307; Figure 4A and B).However, hepatocytes in which PERK was inhibited displayed higher phosphorylation levels of AKT (S473) and GSK3β (Ser9) compared with hepatocytes treated with NEFA alone (Figure 4A, C and D; P < 0.01).In addition, blocking PERK suppressed mRNA abundances of PEPCK (P < 0.05) and G6-Pase (P < 0.05) in bovine hepatocytes after NEFA challenge (Figure 4 E and F).Additionally, we found that blocking PERK (Supplemental Figure S5A-C, https: / / figshare .com/articles/ figure/ Fang _et _al _Supplemental _material/ 20106476; Lei, 2022) in bovine hepatocytes treated with NEFA had no effects on the levels of phosphorylated of IRE1α (Supplemental Figure S5A and D) and its downstream components TRAF2 (Supplemental Figure S5A and E) and phosphorylated JNK (Supplemental Figure S5A and F).It was noteworthy that PERK inhibition significantly reduced NEFA-induced cleavage of ATF6 (Supplemental Figure S5G; P < 0.01) in bovine hepatocytes.

DISCUSSION
The role and the underlying mechanisms whereby ER stress participates in various pathological conditions that afflict dairy cows are receiving increased attention (Chiappisi et al., 2017;Zhu et al., 2019;Sharmin et al., 2020).Current evidence from dairy cow studies indicates that increased lipotoxic stress due to high levels of NEFA triggers the ER stress response (Shi et al., 2015;Gao et al., 2018).Indeed, ER stress signaling can be activated in bovine mammary epithelial cells and hepatocytes by exogenous NEFA in a dose-dependent manner (Zhu et al., 2019;Sharmin et al., 2020;Huang et al., 2021).The present results further confirmed that NEFA activates ER stress transducers (PERK, IRE1α, and ATF6) as well as their downstream ER stressrelated genes (XBP1, ATF4, and DNAJC3) in primary bovine hepatocytes from 1-d-old calves.Intriguingly, and similar to data from adipocytes challenged with NEFA (Jiao et al., 2011), we uncovered a phenomenon in which ER stress signaling tended to be diminished after peak activation with the duration of NEFA treatment.This negative-feedback mechanism may relieve suppression of protein translation in the course of a persistent ER stress event.Along with these effects, ultrastructural morphologic alterations detected after NEFA challenge suggest that these molecules are inducers of ER stress in bovine hepatocytes.
Although the state of physiological insulin resistance helps prioritize the insulin-independent uptake of glucose by the mammary gland (De Koster and Opsomer, 2013), it can render cows more susceptible to fatty liver and ketosis (Youssef and El-Ashker, 2017).Previous work indicated that high concentrations of NEFA lead to insulin resistance in primary bovine hepatocytes and are associated with the progression of fatty liver disease (Rico et al., 2015;Gao et al., 2018).The greater IRS1 serine phosphorylation and lower AKT and GSK3β phosphorylation levels, along with the upregulation of gluconeogenic genes in hepatocytes treated with NEFA, underscored the link between insulin sensitivity and fatty acids.Furthermore, the reversal of impaired insulin signaling following mitigation of ER stress induced by NEFA with use of the ER stress inhibitor TUDCA also underscored the role of ER stress in insulin signaling.Together, the available results suggest that high concentrations of NEFA impair hepatic insulin signaling at least partly by inducing ER stress in bovine hepatocytes.
Available data from model organisms indicate that during sustained ER stress the 3 canonical pathways of the UPR are associated with alterations in insulin signaling in the liver (Özcan et al., 2004;Flamment et al., 2012).The reduction in IRS1 serine phosphorylation, increased AKT and GSK3β phosphorylation levels, and downregulation of gluconeogenic genes (PEPCK and G6-Pase) after IRE1α knockdown in the present study underscore that the IRE1α arm contributes to NEFA-induced insulin resistance.IRE1α is a bifunctional RNase/kinase enzyme (Iwawaki et al., 2001).Furthermore, our data support a model where the kinase activity but not RNase activity of IRE1α is the component of ER stress that mediates NEFAinduced insulin resistance in bovine hepatocytes.Upon ER stress, IRE1α phosphorylation activates IRE1α kinase activity and activates the recruitment protein TRAF2 to indirectly activate c-Jun N-terminal kinase (JNK) and inhibitor of kappa-B kinase (IKK), both of which affect insulin signaling by phosphorylating IRS1 on specific serine residues and reducing its interaction with other signaling proteins (PI3K/AKT and MAP kinase), leading to insulin resistance (Urano et al., 2000;Guerrero-Hernández et al., 2014).Consistently, IRE1α knockdown in bovine hepatocytes suppressed NEFA-induced TRAF2 and activation of JNK.However, further investigation is needed to determine whether NEFAinduced JNK and IKK activations participate in ruminant hepatic insulin sensitivity.Previous studies have shown that direct crosstalk occurs between the arms of UPR (Verfaillie et al., 2013;Brewer, 2014).However, in this study we did not observe significant effect of IRE1α knockdown on the other 2 arms of UPR, PERK and ATF6, indicating that the improvement of IRE1α knockdown on insulin signaling may not depend on the crosstalk between UPR signal transduction cascades.
Apart from IRE1α signaling, activated PERK phosphorylates eukaryotic initiation factor 2 α subunit (eIF2α), which leads to rapid attenuation of translation and, thus, alleviates the ER workload by preventing the production of newly synthesized proteins (Kim et al., 2015).In that context, the fact that blocking of the PERK branch with a high-selectivity inhibitor, GSK2656157, can reconstruct phosphorylation of both AKT and GSK3β and reduce G6-Pase and PEPCK mRNA abundance suggested a functional role in bovine hepatocytes.However, unlike the IRE1α arm, inhibition of PERK did not rescue the NEFA-induced IRS1 serine phosphorylation.It is noteworthy that in previous studies with nonruminants, PERK activation led to increased expression of tribbles-like protein 3 (TRB3), a pseudokinase that promotes insulin resistance by inhibiting AKT, a downstream target of IRS1 (Ozcan et al., 2013;Villalobos-Labra et al., 2019).Indeed, in the present study, the mRNA abundance of TRB3 was attenuated by blocking PERK in NEFA-treated bovine hepatocytes.Thus, our findings indicated that, upon NEFA challenge, inhibition of the PERK branch alleviated insulin resistance in bovine hepatocytes, probably in a manner independent of IRS1.Interestingly, blocking PERK did not alter IRE1α signaling cascades but increased the cleavage level of ATF6 in NEFA-treated bovine hepatocytes, which may be mediated through ATF4, a transcription activator of PERK downstream during NEFA-induced ER stress (Teske et al., 2011).However, future studies will be needed to better understand the role of this crosstalk in hepatic insulin resistance in dairy cows.Unlike IRE1α and PERK, which transduce UPR signals via their kinase activity, ATF6 is proteolytically activated by site 1 protease (S1P) and S2P, and the cleaved ATF6 acts as a transcription factor that induces genes encoding ER chaperones and ER-associated degradation proteins to alleviate ER stress (Salvadó et al., 2015).The rapid and robust increase of cleaved ATF6 as early as 1 h after NEFA treatment suggested that ATF6 functions as a fast response mechanism to counter ER stress induced by NEFA in bovine hepatocytes.Evidence in humans and rodents has shown that ATF6 is implicated in hepatic lipid and glucose homeostasis (Yamamoto et al., 2010;Usui et al., 2012).For instance, ATF6 can enhance fatty acid oxidation by increasing transcriptional activity of PPARα (Chen et al., 2016).Furthermore, liver-specific knockout of ATF6 in mice resulted in severe hepatic steatosis and glucose intolerance (Sun et al., 2018).Contrary to our expectation, inhibition of ATF6 did not significantly affect phosphorylation of IRS1, AKT, or GSK3β in NEFA-treated hepatocytes.In addition, inhibition of ATF6 further enhanced the NEFA-induced increase of gluconeogenic genes PEPCK and G6-Pase abundance rather than inhibiting them.Studies with rodents have reported that ATF6 elicits a suppressive effect on gluconeogenesis through disruption of the interaction between CREB and the CREB-regulated transcription coactivator 2 (Wang et al., 2009).If such a mechanism exists in bovine hepatocytes, it would support our results.
The acknowledged limitation of this study is that the primary hepatocytes isolated from 1-d-old calves might not be representative of adult bovine hepatocytes: at birth all 4 compartments of the ruminant stomach, except the abomasum, are nonfunctional, undeveloped, and small in size (Tamate et al., 1962;Longenbach and Heinrichs, 1998).Furthermore, the preruminant liver in calves is not yet mature, and, thus, various functional disparities in hepatocytes may exist between calves and periparturient cows (Rosa et al., 2018).Although previous studies have confirmed that calf hepatocytes in culture express and produce the marker of mature hepatocytes, albumin (Gao et al., 2021;Lei et al., 2021), and possess the ability to respond to insulin, glucagon, and glucose stimuli (Donkin and Armentano, 1995;Li et al., 2016;Gao et al., 2018), as well as the capacity of exogenously added fatty acid catabolism and anabolism (Li et al., 2014), the differences of functions between hepatocytes from calf liver and adult dairy cow liver should be considered and still need to be further studied.

CONCLUSIONS
High NEFA influx into hepatocytes induces ER stress and causes insulin resistance.Both the IRE1α and PERK branches of the UPR, but not the ATF6 branch, are the main pathways controlling ER stressmediated NEFA-induced insulin resistance.Our studies contribute to the understanding of the role of ER stress in hepatic insulin resistance, and provide new targets for the prevention and treatment of metabolic disorders in dairy cows.
Fang et al.: IRE1Α, PERK, AND HEPATIC INSULIN RESISTANCE Fang et al.: IRE1Α, PERK, AND HEPATIC INSULIN RESISTANCE 50 nm were cut and stained with 4% uranyl acetate and 0.2% lead citrate.Sections with the cells were observed under a transmission electron microscope with 80 kV (H-7650; Hitachi) at 6,800× magnification.At least 5 random views in each sample were collected.

Table 1 .
Fang et al.: IRE1Α, PERK, AND HEPATIC INSULIN RESISTANCE List of primers used for quantitative real-time PCR (F = forward; R = reverse) 1