Inhibiting nuclear factor erythroid 2 related factor 2-mediated autophagy in bovine mammary epithelial cells induces oxidative stress in response to exogenous fatty acids

In early lactation, bovine mammary epithelial cells undergo serious metabolic challenges and oxidative stress both of which could be alleviated by activation of autophagy. Nuclear factor erythroid 2 related factor 2 (NFE2L2), a master regulator of cellular redox homeostasis, plays an important role in the regulation of autophagy and oxidative stress. Thus, the objective of this study was to investigate the role of NFE2L2-mediated autophagy on oxidative stress of bovine mammary epithelial cells in response to exogenous free fatty acids (FFA). Exogenous FFA induced linear and quadratic decreases in activities of glutathione peroxidase (GSH-Px), catalase (CAT), and superoxide dismutase (SOD), and increases in the contents of reactive oxygen species (ROS) and malondialdehyde (MDA). Protein abundance of LC3-phosphatidylethanolamine conjugate (LC3-II) and the number of autophagosomes and autolysosomes decreased in a dose-dependent manner, while protein abundance of p62 increased in cells challenged with FFA. Activation of autophagy via pre-treatment with Rap attenuated the FFA-induced ROS accumulation. Importantly, FFA inhibited protein abundance of NFE2L2 and the translocation of NFE2L2 into the nucleus. Knockdown of NFE2L2 by siRNA decreased protein abundance of LC3-II, while it increased protein abundance of p62. Furthermore, sulforaphane (SFN) pre-treatment attenuated the FFA-induced oxidative stress by activating NFE2L2-mediated autophagy. The data suggested that NFE2L2-mediated autophagy is an important antioxidant mechanism in bovine mammary epithelial cells experiencing increased FFA loads.


Introduction
The rapid increase in milk production coupled with insufficient energy intake in early lactation drives cows into a period of negative energy balance [1]. Severe negative energy balance increases circulating concentrations of free fatty acids (FFA) by initiating fat mobilization of adipose tissue [2], which render bovine mammary gland highly-susceptible to metabolic stress [2,3]. In cows experiencing ketosis after calving, the high concentrations of FFA also are absorbed by the mammary gland and augments the metabolic stress within mammary cells [4,5]. High concentrations of FFA induce mitochondrial dysfunction and ROS within mammary epithelial cells and causes oxidative stress [5,6].
Autophagy is a conserved degradation process, which plays a cardinal role in the maintenance of cellular homeostasis via eliminating damaged cellular organelles and scavenging aging proteins [7][8][9]. Activation of autophagy, a major cellular defense process, alleviated metabolic stress and its associated pro-oxidant state in non-ruminants [10,11]. More importantly, work in dairy cows has suggested that enhanced autophagy is an adaptive mechanism to help mitigate metabolic stress in the mammary gland induced by hyperketonemia [5]. In vitro, activation of autophagy protected bovine mammary epithelial cells against H 2 O 2 -induced oxidative stress and apoptosis [4]. Thus, mitigation of oxidative challenges via autophagy might be an important defense mechanism in bovine mammary epithelial cells.
Nuclear factor erythroid 2 related factor 2 (NFE2L2) is a master transcription factor, which maintains redox state in the cell through regulating the transcription of numerous antioxidants [12][13][14]. Activation of NFE2L2 is essential to protect bovine mammary epithelial cells against H 2 O 2 -induced oxidative stress [15]. In addition, NFE2L2 is also a regulator of autophagy [13,16]. Work with nonruminants confirmed that NFE2L2 activates autophagy through the control of LAMP2A transcription, which alleviates oxidative stress by inhibiting accumulation of oxidized protein aggregates [17]. Interestingly, the regulation of autophagy via NFE2L2 also was confirmed in primary bovine mammary epithelial cells [18]. Thus, NFE2L2mediated autophagy might play a protective role against oxidative stress in bovine mammary epithelial cells.
Given the established role of NEF2L2 in the regulation of autophagy and oxidative stress in non-ruminants, we hypothesized that activation of NEF2L2-mediated autophagy protects bovine mammary epithelial cells against oxidative stress. In the present study, sulforaphane (SFN) and NFE2L2 small interfering RNA (Si-NFE2L2) were used to upregulate and downregulate NFE2L2 abundance in MAC-T cells. The main objective was to investigate effects of NFE2L2-mediated autophagy on oxidative stress in bovine mammary epithelial cells in response to exogenous FFA.
In Exp.1, to investigate the effects of FFA on oxidative stress and autophagy, different FFA concentrations were chosen to treat MAC-T cells to simulate the response observed in ketotic cows after calving [6,20]. The MAC-T cells were cultured in 2% fatty acid-free BSA with 0, 0.6, 1.2, or 2.4 mmol/L FFA for 24 h. Cells were cultured in serum-free medium for 12 h before the above treatments to induce autophagy as previous described [21,22]. The stock FFA mix was prepared as previously described [23]: palmitic acid (16.8 mmol/L), stearic acid (7.6 mmol/L), oleic acid (22.9 mmol/L), linoleic acid (2.6 mmol/L), and palmitoleic acid (2.8 mmol/L) were diluted in 0.1 mol/L KOH solution at 60°C, and the pH of the was adjusted to 7.4 with 1 mol/L hydrochloric acid.
In Exp.2, to investigate the effect of autophagy activation on FFA-induced oxidative stress, MAC-T cells were seeded into 6-well plates with lactogenic medium and grown until 80% to 90% confluence. Rapamycin (Rap) was diluted in dimethylsulfoxide (DMSO) to a final working concentration. There were 4 experimental treatments: BSA + DMSO, DMSO was used to treat cells for 18 h prior to addition of 2% BSA for an additional 24 h; 1.2 mmol/L FFA + DMSO, DMSO was used to treat cells for 18 h prior to challenge with 1.2 mmol/L FFA for 24 h; BSA + 100 nmol/L Rap, 100 nmol/L Rap was used to treat cells for 18 h prior to addition of 2% BSA for an additional 24 h; FFA + 100 nmol/L Rap, 100 nmol/L Rap was used to treat cells for 18 h prior to challenge with 1.2 mmol/L FFA for an additional 24 h.
In Exp.4, to investigate the role of NFE2L2 activation on autophagy and oxidative stress in MAC-T induced by FFA, cells were seeded into 6-well plates with lactogenic medium and grown until 80% to 90% confluence. SFN was diluted in DMSO to a final working concentration. The 4 experimental treatments included the following: BSA + DMSO, DMSO was used to treat cells for 24 h prior to addition of 2% BSA for an additional 24 h; 1.2 mmol/L FFA + DMSO, DMSO was used to treat cells for 24 h prior to challenge with 1.2 mmol/L FFA for 24 h; BSA + 10 μmol/L SFN, 10 μmol/L SFN was used to treat cells for 24 h prior to addition of 2% BSA for an additional 24 h; FFA + 10 μmol/L SFN, 10 μmol/L SFN was used to treat cells for 24 h prior to challenge with 1.2 mmol/L FFA for an additional 24 h.

RNA interference
In Exp.3, to explore the effect of NFE2L2 knockdown on autophagy in MAC-T cells, scrambled non-target negative control (Si-Control) or siRNA targeting NFE2L2 (Si-NFE2L2) were used to transfect MAC-T cells for 48 h. The Si-NFE2L2 and Si-Control were designed and synthesized by Shanghai Genechem Co., Ltd. (Shanghai, China). Sense and antisense primer sequences of the siRNA are reported in Table 1. MAC-T cells were seeded into 6-well plates at a density of 2.0 × 10 6 cells/ mL. The siRNA was mixed with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) in serum-free, antibiotic-free DMEM/F12 medium and then incubation for 20 min in room temperature. MAC-T cells were then treated with 50 nmol/L of Si-NFE2L2 or Si-Control/Lipofectamine solution in antibiotic-free DMEM/F12 medium for 12 h. Medium was then switched to DMEM/F12 medium supplemented containing 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin for 36 h.

Total RNA extraction and quantitative reversetranscription PCR analysis
Cells were washed twice with PBS before isolating total RNA with RNAiso Plus (AJG1895A, TaKaRa Biotechnology Co. Ltd., Dalian, China) according to the manufacturer's protocol. Concentration of total RNA was measured with a K5500 Micro Spectrophotometer (Beijing Kaiao Technology Development Ltd., Beijing, China). RNA integrity was determined by RNA integrity number with the Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). All samples had an RNA integrity number above 8.0. Subsequently, 1 μg of total RNA were reverse-transcribed into cDNA according to the instructions of the reverse transcription kit (TaKaRa Biotechnology Co. Ltd., Tokyo, Japan). The mRNA abundance was measured using a SYBR Green Plus Reagent Kit (TaKaRa Biotechnology Co. Ltd.) with the 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Calculated mRNA abundance levels was via the 2 -ΔΔCT method, and normalized to the mean of GAPDH and ubiquitin B. Primer sequences are reported in Table 2. Results were expressed as fold changes by normalizing the data to the control values.

Determination of intracellular ROS
After treating as indicated above, cells were detached with trypsin and incubated at 37°C for 5 min before collecting into centrifuge tubes. Cells were washed 3 times with PBS and incubated with 25 μmol/L 2′,7′-dichlorofluorescein diacetate (Beyotime Institute of Biotechnology) in serum-free DMEM/F12 medium at 37°C for 20 min in the dark. Cells were then washed 3 times with PBS before resuspending with bench-top solution. Subsequently, fluorescence of cells was determined using flow cytometry (FACSCalibur, Becton-Dickinson, Sunnyvale, CA, USA).

Detection of oxidative stress indicators
Activities of superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and catalase (CAT), and content of malondialdehyde (MDA) in MAC-T cells were measured by commercially available kits according to the manufacturer's protocol as previous described [24].

Statistical analysis
All experiments were repeated at least 3 times on different days. The data were analyzed using SPSS 23.0 software (IBM Corp., Armonk, NY) and results are presented as the means ± SEM. The Shapiro-Wilk and Levene tests were used to test the normality and variance homogeneity of data. Linear and quadratic comparisons were conducted to evaluate dose-dependent effects. Statistical significance was evaluated by paired ttest between the 2 groups (Exp. 3) via one-way ANOVA among the 4 groups (Exp. 1, 2 and 4) followed by Bonferroni correction. Differences with P < 0.05 were considered statistically significant.

FFA inhibit autophagy in MAC-T cells
Compared with 0 mmol/L FFA, protein abundance of p62 was greater in the 1.2 and 2.4 mmol/L FFA group (linear, P < 0.001, Fig. 2a and b and Table 3). In contrast, compared with 0 mmol/L FFA, protein abundance of LC3-II was lower with 1.2 and 2.4 mmol/L FFA (linear, P < 0.001, Fig. 2a and c and Table 3). Compared with 0 mmol/L FFA, the number of autophagosomes labeled with yellow puncta and autolysosomes labeled with red puncta was greater with 1.2 mmol/L FFA group (Fig.  2d).

FFA inhibit translocation of NNFE2L2 into the nucleus in MAC-T cells
Compared with 0 mmol/L FFA, nuclear protein abundance of NFE2L2 was lower with 1.2 or 2.4 mmol/L FFA (linear and quadratic, P < 0.001, Fig. 4a and b and Table  3). Consistent with alterations in nuclear protein abundance of NFE2L2, immunofluorescence staining results revealed that FFA treatment inhibited NFE2L2 translocation to the nucleus (Fig. 4c).

Knockdown of NFE2L2 inhibit autophagy in MAC-T cells
Compared with the Si-Control, mRNA abundance of NFE2L2 (P = 0.006, Fig. 5a) and protein abundance of NFE2L2 was lower in Si-NFE2L2 (P < 0.001, Fig. 5b and c). Compared with Si-Control, protein abundance of p62 was greater in Si-NFE2L2 (P < 0.001, Fig. 5b and d). In contrast, compared with Si-Control, protein abundance of LC3-II was lower in Si-NFE2L2 (P < 0.001, Fig. 5b and e). Consistent with alterations in protein abundance of LC3-II, knockdown of NFE2L2 decreased the number of autophagosomes labeled with yellow puncta and autolysosomes labeled with red puncta (Fig. 5f).

Activation of NFE2L2 attenuated FFA-induced autophagy inhibition and oxidative stress
Compared with BSA + DMSO, protein abundance of nuclear NFE2L2 was lower with 1.2 mmol/L FFA + DMSO (P = 0.001, Fig. 6a and b). However, pretreatment with 10 μmol/L SFN, an activator of NFE2L2, increased protein abundance of nuclear NFE2L2 (P < 0.001) and attenuated the down-regulation of protein abundance of nuclear NFE2L2 induced by FFA (P = 0.002, Fig. 6a and b). Pretreatment with 10 μmol/L SFN decreased protein  abundance of p62 (P = 0.034) and attenuated the upregulation of protein abundance of p62 induced by FFA (P < 0.001, Fig. 6a and c), while it increased protein abundance of LC3-II (P < 0.001, Fig. 6a and d) and attenuated the down-regulation of protein abundance of LC3-II induced by FFA (P < 0.001, Fig. 6a and d). Consistent with alterations in protein abundance of LC3-II, the number of autophagosomes labeled with yellow puncta and autolysosomes labeled with red puncta were greater with 1.2 mmol/L FFA + 10 μmol/L SFN (Fig. 6e). Pretreatment with 10 μmol/L SFN decreased content of ROS (P = 0.042, Fig. 6f) and attenuated the increase in content of ROS induced by FFA (P < 0.001, Fig. 6f), while it decreased the activity of SOD and attenuated the increase in activity of SOD induced by FFA (P = 0.032, Fig. 6g).

Discussion
The concentrations of FFA in plasma of dairy cows after calving increase due to severe negative energy balance [25], and can induce oxidative stress in the mammary gland [3]. Work with vitro experiments has confirmed that exogenous FFA induce ROS production in bovine mammary epithelial cells, resulting in oxidative stress [6]. Thus, the greater ROS and MDA content when cells were challenged with FFA in the present study underscored the positive effect of this metabolite on free radical production and levels of lipid peroxidation products.
In fact, greater ROS and MDA content coupled with lower activities of SOD, GSH-Px, and CAT confirmed that exogenous FFA induced oxidative stress in mammary epithelial cells in a dose-dependent manner. Thus, we speculated that mammary uptake of FFA when circulating concentrations increase markedly is conducive to a state of oxidative stress that could impact cell function. In fact, the degree of oxidative stress in mammary tissue of ketotic cows was correlated positively with the circulating concentrations of FFA [26]. Autophagy serves to reduce oxidative stress through removal of protein aggregates and damaged organelles [27,28]. The microtubule-associated protein light chain 3 (LC3) is processed from the cytosolic form (LC3-I) to the LC3-II during autophagy, the protein abundance of which is positively correlated with autophagic activity [29,30]. The protein p62 is an autophagy receptor involved in the recognition of ubiquitin-labeled substrates targeted for autophagy, the protein abundance of which is inversely correlated with autophagic activity [31]. In the present study, the lower protein abundance of LC3-II along with greater protein abundance of p62 underscored the negative effect of FFA on autophagy in mammary cells. Work with mice has demonstrated that high circulating concentrations of FFA due to metabolic stress impair autophagy and leads to oxidative stress [32]. Thus, inhibition of autophagy due to metabolic During autophagy, cellular contents are engulfed by double-membrane vesicles called autophagosomes and subsequently delivered to lysosomes for degradation [33]. Thus, down-regulation of autophagosomes (yellow puncta) and autolysosomes (red puncta) further indicated that FFA inhibited autophagic flux in mammary epithelial cells. Work by Sun et al. [4] reported that activation of autophagy alleviated H 2 O 2 -induced oxidative stress in bovine mammary epithelial cells. In nonruminants, Rap, an activator of autophagy, can enhance autophagy through upregulation of LC3-II and downregulation of p62, and promotion of fusion between the autophagosomes and lysosomes [34,35]. In this study, the increased protein abundance of LC3-II and the decreased protein abundance of p62 confirmed the role of Rap on autophagy in mammary epithelial cells. The decrease in intracellular ROS levels due to autophagy activation in the present study agreed with previous work demonstrating a cytoprotective role of autophagy against oxidative stress [36].
At least in non-ruminants, the transcription factor NFE2L2, besides its role as a master regulator of cellular homeostasis, plays important roles in autophagy [13,37]. In the present study, the decrease in nuclear protein abundance of NFE2L2 in response to FFA treatment was consistent with alterations in autophagy, and agrees with similar work in HepG2 cells and bovine hepatocytes [38,39]. Thus, we speculate that alterations in protein abundance of NFE2L2 are closely related to inhibition of autophagy induced by metabolic stress.
Knockdown of NFE2L2 deceased LC3-II protein abundance in PC12 cells [40]. In agreement with those data, the fact that knockdown of NFE2L2 via siRNA increased protein abundance of p62 and decreased both protein abundance of LC3-II and number of autophagosomes and autolysosomes suggested that NFE2L2 directly regulated autophagy. Thus, promoting NFE2L2 translocation into the nucleus in mammary cells to activate the autophagy pathway may be a potential treatment strategy for oxidative stress.
The specific mechanisms whereby FFA inhibit NFE2L2 transport to the nucleus in bovine mammary epithelial cells remains to be elucidated. From a regulatory standpoint, it is noteworthy that Keap1, a component of the E3 ubiquitin ligase complex, can directly regulate the abundance and nuclear translocation of NFE2L2 [41]. We speculated that decreased nuclear translocation of NFE2L2 was partly due to inhibition of the detachment of Keap1-NFE2L2. In fact, previous research confirmed that protein abundance of Keap1 is reduced in adipose tissue of dairy cows with ketosis [42].
Enhanced autophagy is thought to be a response to mitigate oxidative stress. NFE2L2 modulated autophagic flux in addition to upregulating antioxidant defenses via the p62-NFE2L2 feedback pathway [43]. Activation of NFE2L2 by SFN increased autophagy activity via inducing LC3-II elevation and vesicle formation in lens cells [44]. In the present study, the upregulation of LC3-II protein abundance and downregulation of p62 protein abundance in response to activation of NFE2L2 indicated that it plays a similar role in bovine mammary epithelial cells in promoting autophagic flux. Furthermore, the downregulation of protein abundance of LC3-II, decrease in number of autophagosomes and autolysosomes, decrease in intracellular ROS, and the increase of SOD activity in FFA-challenged cells in which NFE2L2 was activated with SFN underscored the positive effect of NFE2L2-mediated autophagy on oxidative stress. Thus, we speculate that autophagy was activated through NFE2L2 translocation into the nucleus to alleviate oxidative stress. Our data validated previous work demonstrating that activation of the NFE2L2 pathway upregulates autophagy [45], and activation of autophagy is a regulatory mechanism to attenuate oxidative stress [46].

Conclusions
The present findings indicated that FFA induces oxidative stress and inhibits NFE2L2-mediated autophagy in bovine mammary epithelial cells. Activation of NFE2L2mediated autophagy via SFN alleviated FFA-induced oxidative stress. Taken together, the present study confirmed that NFE2L2-mediated autophagy may be a promising therapeutic target for reducing FFA-induced oxidative stress of bovine mammary gland during periods such as the transition into lactation.  b Protein abundance of NFE2L2. c Protein abundance of p62. d Protein abundance of LC3-II. e MAC-T cells were transfected with the recombinant adenovirus mRFP-GFP-LC3 for 36 h and with/without SFN for another 12 h, before treated with/without 1.2 mmol/L FFA for another 24 h. Representative images of autophagosomes (yellow puncta) and autolysosomes (red puncta), scale bar = 25 μm. f ROS content. g SOD activity. Comparisons among groups were calculated using a one-way ANOVA with subsequent Duncan correction. The data presented are the mean ± SEM. Different superscript lowercase letters in bar charts represent significant difference (P < 0.05) DAPI: 4′,6-diamidino-2-phenylindole dihydrochloride; TBST: Tris-buffered saline; ECL: Enhanced chemiluminescence solution