In dairy cows, the periparturient phase representing the time interval between 3 wk before to 3 wk after parturition is associated with strong metabolic adaptations [1]. Production of milk leads to a strong increase of the energy requirement, which however cannot be met as the food intake capacity is limited. Thus, during early lactation, dairy cows are typically in a negative energy balance, which is compensated by a stimulation of lipolysis in adipose tissue. This leads to strongly increased plasma concentrations of non-esterified fatty acids (NEFA), which are partially taken up into the liver. As the capacity of the liver for β-oxidation during the periparturient phase is insufficient, NEFA are incorporated into triacylglycerols (TAG). As very low-density lipoproteins (VLDL) cannot be produced at sufficient amounts due to a low synthesis of Apo B, TAG are stored in the liver leading to fatty liver syndrome. Moreover, there is commonly a strong stimulation of ketogenesis during early lactation, which can result in ketosis [1–3].

Previously, it has been shown, that increased plasma levels of NEFA, such as observed in dietary or genetic models of obesity or diabetes, are leading to stress of the endoplasmic reticulum (ER) in the liver [4–6]. ER stress is defined as an imbalance between the folding capacity of the ER and the protein load, resulting in the accumulation of unfolded or misfolded proteins in the ER lumen [5]. The disturbance of the ER homeostasis activates an adaptive response known as the unfolded protein response (UPR), which aims to restore ER homeostasis and functions by triggering three kinds of protective cellular responses: (i) up-regulation of ER chaperones, such as immunoglobulin heavy-chain binding protein (BiP), to assist in the refolding of proteins; (ii) attenuation of protein translation, and (iii) degradation of misfolded proteins by the proteasome by a process called ER-associated degradation (ERAD) [7, 8]. If ER stress-induced damage is too strong and homeostasis cannot be restored, the UPR can lead to cell death by the induction of apoptosis [9, 10]. Sensing of stress in the ER lumen is mediated by three ER stress transducers: inositol requiring 1 (IRE1), PKR-like ER kinase (PERK), and activating transcription factor 6 (ATF6) [5, 8]. Under non-stress conditions these transducers are bound to the abundant luminal chaperone BiP preventing them from activating downstream events. When misfolded proteins are accumulating in the ER lumen, BiP dissociates from the stress transducers in order to chaperone the misfolded proteins, which leads to an activation of ER stress transducers and an initiation of the UPR [9, 11]. Activation of PERK stimulates the phosphorylation of eukaryotic initiation factor (eIF) 2α, which attenuates protein translation [12]. IRE1 activation causes unconventional splicing of X-box binding protein 1 (XBP1) mRNA and translation into the transcription factor XBP1 [8]. XBP1 up-regulates ER chaperons, components of ERAD and stimulates phospholipid biosynthesis, which leads to an expansion of the ER membrane [5–7]. IRE1 activation moreover leads to an activation of nuclear factor kappa B (NF-κB), a transcription factor involved in inflammation, and an induction of pro-apoptotic genes [5, 13]. ATF6 is a transcription factor, which is activated by processing via site 1 and site 2 proteases in the Golgi. The activated ATF6 induces the expression of genes involved in ERAD, lipid biosynthesis, ER expansion and protein folding [5].

Interestingly, activation of the UPR, such as observed in models of obesity or diabetes or induced by application of chemical ER stress inducers, leads to a variety of symptoms in the liver, which are similar to those observed in periparturient dairy cows, such as the development of fatty liver [14–16], an induction of fibroblast growth factor (FGF) 21 [17], an enhancement of the antioxidant and cytoprotective capacity by activation of nuclear factor E2-related factor 2 (Nrf2) [18, 19], and an induction of inflammation [20, 21].

Performance parameters of the cows used in this experiment have recently been reported [34]. Milk yield of the cows, in average from wk 1 to wk 14 was approximately 33 kg/d, average feed intake was 18.5 kg DM/d. The onset of lactation led to a strong negative energy balance of about -65 MJ NEL/d in wk 1. At 14 wk postpartum, the cows had a slightly positive energy balance. Plasma NEFA concentrations showed their peak values at 1 wk postpartum while plasma β-hydroxybutyric acid (BHBA) concentrations were highest at 1 wk and 5 wk postpartum. Liver TAG concentrations were highest at 5 wk postpartum [34]. The cows considered in this study, moreover, showed increased mRNA concentrations of tumor necrosis factor α and acute phase proteins (C-reactive protein, haptoglobin, serum amyloid A) being indicative of a pro-inflammatory condition and increased mRNA concentrations of various Nrf2 target genes and of FGF21 in the liver at 1 wk postpartum [35, 36].

To detect an activation of the UPR due to ER stress, we determined mRNA concentrations of BiP (encoded by HSPA5) and 13 downstream genes of the three ER stress sensors by qRT-PCR and the mRNA concentration of the spliced variant of XBP1 (sXBP1) by standard RT-PCR. The mRNA concentrations of all the UPR target genes, with the only exception of WARS, were elevated from 3 wk antepartum to 1 wk postpartum (Table 2). Moreover, at 1 wk postpartum, mRNA concentration of sXBP1 was increased in comparison to 3 wk antepartum, indicating that splicing of XBP1 was induced following the onset of lactation (Figure 1). From 1 wk postpartum to later lactation, mRNA concentrations of all the genes involved in the UPR were declining. However, relative mRNA concentrations of some genes considered (ATF4, CASP3, EDEM1, WARS, XBP1) were also significantly increased at 5 wk postpartum in comparison to 3 wk antepartum (Table 2). The mRNA concentrations of most of the genes considered were not different between 14 wk postpartum and 3 wk antepartum, while mRNA concentrations of ATF4 and the 114 bp unspliced XBP1 were higher at 14 wk postpartum than at 3 wk antepartum.

Gene2	3 wk antepartum (n = 13)	1 wk postpartum (n = 13)	5 wk postpartum (n = 13)	14 wk postpartum (n = 13)
ATF4	1.00 ± 0.25a	3.14 ± 0.25c	2.08 ± 0.26b	1.65 ± 0.24b
BAK1	1.00 ± 0.29a	2.10 ± 0.30b	1.56 ± 0.29ab	1.59 ± 0.28ab
BAX	1.00 ± 0.36a	3.02 ± 0.39b	1.98 ± 0.35ab	1.52 ± 0.35a
CASP3	1.00 ± 0.31a	2.54 ± 0.27b	2.07 ± 0.26b	1.71 ± 0.26ab
CASP8	1.00 ± 0.25a	2.48 ± 0.24b	1.88 ± 0.23ab	1.73 ± 0.23ab
CASP9	1.00 ± 0.37a	2.45 ± 0.35b	1.89 ± 0.34ab	1.84 ± 0.36ab
DDIT3	1.00 ± 0.17a	1.86 ± 0.17b	1.58 ± 0.17ab	1.41 ± 0.17ab
DNAJC3	1.00 ± 0.15a	1.89 ± 0.17b	1.49 ± 0.17ab	1.43 ± 0.15ab
EDEM1	1.00 ± 0.21a	1.70 ± 0.21b	1.69 ± 0.20b	1.25 ± 0.22ab
HERPUD1	1.00 ± 0.45a	3.96 ± 0.37b	2.27 ± 0.43a	1.66 ± 0.38a
HSPA5	1.00 ± 0.29a	2.74 ± 0.27b	1.64 ± 0.26a	1.41 ± 0.24a
PDIA4	1.00 ± 0.25a	1.92 ± 0.28b	1.34 ± 0.24ab	1.23 ± 0.25ab
WARS	1.00 ± 0.35a	2.03 ± 0.31ab	2.31 ± 0.29b	1.93 ± 0.30ab
XBP1 (unspliced)	1.00 ± 0.25a	3.14 ± 0.25c	2.08 ± 0.26b	1.65 ± 0.24b

¹mRNA concentrations of genes are expressed relative to the mRNA abundance at 3 wk antepartum.

²Abbreviations of gene names: see Table 1.

Values are means ± SEM.

^a,b,cMeans without common superscripts differ significantly (P < 0.05).

In the present study, we observed that BiP, a chaperone which is considered as the master regulator of the UPR [5], and several downstream genes of the three ER stress transducers (IRE1, PERK, ATF6) are up-regulated in the liver of dairy cows at early lactation (Figure 2). It has been assumed that mRNA concentrations of ER chaperons, ERAD components, such as BiP, HERP, WARS, PDIA4, P58^IPK, EDEM1, XBP1, and ATF4, or genes, that are involved in the induction of apoptosis, such as Chop or caspases, are reliable markers of ER stress [22–24, 27, 37]. Thus, the present study strongly suggests the presence of ER stress in the liver of dairy cows during early lactation, which was associated with induction of the UPR. This suggestion is supported by the present finding, that the mRNA concentration of sXBP1 in the liver was increased at 1 wk postpartum, indicative of an increased XBP1 splicing, which is considered a hallmark of ER stress [8]. The observed activation of XBP1 by unconventional splicing of XBP1 mRNA 1 wk postpartum agrees with a recent study of Loor [38] who found an up-regulation of 39 target genes of XBP1 in the liver of dairy cows during the transition from late pregnancy to lactation by transcriptome analysis.

From the present study, the metabolic reasons for the production of ER stress during early lactation cannot be explained. However, it has been shown that an increased load of the liver with fatty acids induces ER stress, with saturated fatty acids being more deleterious in this respect than unsaturated fatty acids [6, 16, 37]. Saturated fatty acids are less readily converted into TAG than unsaturated fatty acids, and thus travel to the ER in the free form, where they may disrupt ER morphology and function [16]. Thus, it is likely that the elevated concentrations of NEFA, consisting mainly of saturated fatty acids and oleic acid [41], in the blood of the cows in early lactation could contribute to the induction of ER stress in the liver. ER stress induction in the liver of cows during early lactation could also be due to the occurrence of a pro-inflammatory condition. Periparturient cows exert commonly an inflammation-like condition in the liver, induced by various events such as injuries and trauma during calving stress, mammary gland oedema, uterus involution, infectious or metabolic diseases, parasites or endotoxins from the gut [42–44]. A pro-inflammatory condition during early lactation has also been observed in the cows considered in this study [35]. As inflammation strongly induces ER stress [18], activation of the UPR in the liver of the cows could have been triggered by the pro-inflammatory condition. The finding that the expression of UPR target genes was declining from 1 wk postpartum to later lactation supports the hypothesis that ER stress in early lactation was caused mainly by high plasma levels of NEFA and the inflammatory condition. However, the finding that some of the UPR target genes remained up-regulated even at 14 wk postpartum – when the cows were already in a positive energy balance – suggests that a certain degree of ER stress could also be induced by the high metabolic activity of the liver during lactation, without being burdened by high NEFA concentrations or inflammation.

Many biochemical alterations in the liver induced by UPR under pathophysiologic conditions such as obesity, diabetes or chronic inflammation are similar to those observed in the liver of periparturient dairy cows, with fatty liver development being one important example. ER stress-induced fatty liver is caused by an increased expression of genes involved in lipogenesis, a reduced expression of genes involved in fatty acid oxidation and lipolysis, and an impairment of the production of VLDL required for the export of TAG from the liver [14, 45–47]. As these biochemical alterations leading to fatty liver are very similar to those observed in the liver of periparturient dairy cows [1, 2, 48], it is likely that the ER stress-induced UPR is involved in the development of fatty liver in periparturient cows. Other similarities between various biochemical alterations induced by ER stress and those observed in the liver of periparturient dairy cows are the induction of inflammation, an activation of Nrf2 pathway and an up-regulation of FGF21. As mentioned above, inflammation can directly induce ER stress. However, ER stress and the concomitant UPR also enhance the inflammatory process [18], suggesting that ER stress could contribute to the induction of the pro-inflammatory condition in the liver of periparturient cows. Nrf2 is a transcription factor which regulates the transcription of a great number of genes with antioxidative and cytoprotective functions [49, 50]. Activation of this transcription factor during the periparturient phase in the liver of the cows has been considered as a compensatory means to protect the liver against the deleterious effects of pro-inflammatory cytokines and reactive oxygen species (ROS) [36]. The fact that ER stress also causes an activation of Nrf2, probably as a means to counteract oxidative stress provoked under ER stress conditions [18, 19] suggests that the observed activation of Nrf2 in the liver of dairy cows at early lactation might also be caused by the UPR.

FGF21 is a hormonal regulator which stimulates hepatic lipid oxidation, ketogenesis and gluconeogenesis during energy deprivation [51–53]. Recently, it has been found that the expression of FGF21 in the liver and plasma levels of FGF21 are increased in dairy cows during the periparturient phase, and that there is even a relationship between hepatic TAG content and plasma FGF21 concentration in dairy cows [34, 39, 40]. More recently, it has been observed that FGF21 is directly induced by ER stress, mediated by an activation of the PERK cascade [17]. Thus, it is likely that the up-regulation of FGF21 in the liver of periparturient cows is mediated by an ER stress-induced UPR. The fact that FGF21 stimulates ketogenesis indicates that ER stress present in the liver of periparturient cows might enhance the development of ketosis via an up-regulation of FGF21.

Based on the similarities between various biochemical alterations induced by ER stress and those observed in the liver of dairy cows during the periparturient phase, it is probable that the induction of ER stress and the concomitant UPR contribute to pathophysiologic conditions during this phase, such as the development of fatty liver, ketosis, and hepatic inflammation. The existence of ER stress in the liver of dairy cows, moreover, might be of relevance for glucose homeostasis as it has been shown that ER stress impairs gluconeogenesis [54, 55], a pathway, which is of extraordinary relevance in high yielding dairy cows.

It should be noted that the induction of ER stress gene network expression in dairy cows has been also observed in the mammary gland during the transition from pregnancy to lactation [56]. The physiological relevance of that pathway, however, might be different between liver and mammary gland. While ER stress induced UPR in the liver might be regarded as a means to maintain ER homeostasis and liver function, ER stress signalling in the mammary gland might be involved in mammary lipogenesis and milk protein synthesis [56].

In overall, the present study reveals the existence of ER stress in the liver of dairy cows during early lactation. Since ER stress causes many biochemical adaptations and symptoms similar to those observed in the liver of periparturient cows, such as the development of fatty liver, ketosis or inflammation, it is assumed that the ER stress-induced UPR might contribute to the pathophysiologic conditions commonly observed in the liver of periparturient cows.