Ammonia-Induced Energy Metabolism Disorder and Autophagy via AMPK/mTOR/ULK1 Pathway in Chicken Livers

Ammonia (NH 3 ) is a well-known environmental pollution gas, threatening human health. NH 3 is also the most harmful gas to poultry for many years. Some studies have found NH 3 can damage eyes, respiratory system, and digestive system. However, molecular mechanism of NH 3 toxicity on chicken livers remains unclear. In this study, we selected chicken liver as research object and successfully duplicated NH 3 poisoning model of chickens. The ultrastructure of chicken livers was observed. The activities of ATPases (Na + K + ATPase, Mg ++ -ATPase, Ca ++ -ATPase, and Ca ++ Mg ++ -ATPase) and the expression of energy metabolism-related genes (HK1, HK2, PK, PFK, PDHX, CS, LDHA, LDHB, AMPK, SDHA, SDHB, and avUCP) and autophagy-related genes (PI3K, LC3I, LC3II, Beclin1, SQSTM1, mTOR, ULK1, ATG5, ATG12, and ATG13) were measured to explore the effect of NH 3 on energy metabolism and autophagy in chicken livers. Our results showed that excess NH 3 caused liver tissue damage. Meanwhile, ATPases activities were inhibited during NH 3 treatment. Moreover, we found that NH 3 exposure altered the expression of energy metabolism-related and autophagy-related genes. NH 3 -induced compensatory increase of AMPK activated autophagy process through inhibiting mTOR and promoting ULK1. In addition. there was dose-dependent and time-dependent effects on all detected indexes in NH 3 -caused chicken liver damage. Excess NH 3 induced energy metabolism disorder and autophagy via AMPK/mTOR/ULK1 pathway in chicken livers.

Excess NH 3 emission in the atmosphere can not only disrupt the global greenhouse balance but also cause air pollution [7]. For example, it can form ammonium salt aerosol with acid gas to participate in PM2.5 formation and contribute to the production of acid rain [8,9]. It was reported that about 75% of NH 3 emission in European atmosphere came from livestock production [10]. Some researches found that NH 3 in the poultry house is mainly generated from two aspects: In the process of protein being digested by enzymes in poultry digestive tracts [8] and microbial degradation in chicken manure and bedding [11] Chickens are sensitive to NH 3 . Chickens growing in high concentration of NH 3 for a long time appear tearing and in ammation of cornea and conjunctiva in chicken eyes [12,13]. In addition, highconcentration NH 3 can damage respiratory system and digestive system of chickens. Two studies showed that excess NH 3 caused immune disorders and in ammatory damage [14], and necro-injury in chicken tracheas [15]. Zhang J et al. found that NH 3 exposure interfered with nutrient absorption in small intestinal mucosae of broilers [16]. NH 3 exposure can also damage immune function in chickens [17][18][19].
Liver is the largest gland in organism and plays important roles in metabolism, detoxi cation, and immunity. Schaerdel AD et al. demonstrated that NH 3 can be absorbed into blood through the lungs of rats, resulting in the increase of blood NH 3 [20]. The increase of NH 3 in blood can aggravate hepatic detoxi cation burden on NH 3 and cause liver NH 3 poisoning. A study found that long-term exposure to high-concentration NH 3 triggered chronic liver injury in broilers [21]. Excess NH 3 also induced hepatic injury via promoting apoptosis in rats [22] and chickens [23]. Therefore, it is of great signi cance to explore the effect of NH 3 on poultry health, especially on livers.
Energy metabolism is the process of material metabolism, accompanied by the release, transfer, storage, and utilization of energy. During energy metabolism process, adenosine triphosphate (ATP) cannot penetrate cell membrane, and the absorption and utilization of nutrients depend on the electrochemical proton gradient produced by ATPase [24]. ATPases, a group of membrane-binding enzymes, are necessary for normal cell function and help to maintain membrane potential and osmotic balance of cells [25,26]. Adenosine 5'-monophosphate-activated protein kinase (AMPK), as a whole-body energy sensor, can inhibit energy consumption process and activate energy production process through integrating different signaling pathways to meet energy needs of cells [27]. NH 3 neurotoxicity can induce brain energy metabolism disorder in patients with chronic and acute hepatic encephalopathy [28].
Selenium de ciency inhibited the expression of genes related to energy metabolism and lead to energy metabolism disorder in broiler erythrocytes [29]. Another study showed that excess NH 3 can lead to metabolic disorder in chicken thymuses [30]. Under cell starvation condition, AMPK can trigger autophagy by inhibiting mammalian target of rapamycin (mTOR) and activating unc-51 like autophagy activating kinase 1 (ULK1) to meet cell energy needs [31]. AMPK/mTOR/ULK1 signaling pathway is one of classical pathways connecting energy metabolism and autophagy [31,32]. Shear stress induced autophagy in vascular smooth muscle cells via activating AMPK/mTOR/ULK1 pathway [33]. Autophagy is a selfdegradation process of cellular components through autophagosome-lysosome pathway [34]. Autophagy can serve as a cellular defense mechanism to respond to external harmful stimuli by degrading protein aggregates, damaged organelles, and even pathogens in cells [35]. Autophagy initiation depends on the participation of a series of autophagy-related genes (ATG), such as mTOR, ULK1, Phosphatidylinositol 3kinase (PI3K), microtubule associated protein 1 light chain 3 (LC3), Beclin1, sequestosome 1 (SQSTM1), ATG5, ATG12, and ATG13. Zhang et al. (2008) studied the changes of autophagy after traumatic brain injury in rats and found that double membrane structure of autophagy increased, and the protein levels of LC3 and Beclin1 increased in neurons and astrocytes, which indicated that autophagy was activated after brain injury [36]. Harmful substance cadmium exposure increased mRNA and protein expression of Beclin1, LC3I, LC3II, and ATG5, and induced autophagy in chicken kidneys [37]. Another substance NH 3 exposure increased the expression of classical autophagy markers LC3 and Beclin1, and decreased SQSTM1, indicating that excess NH 3 caused autophagy in skeletal muscle of patients with liver cirrhosis [38]. In this study, we successfully established the model of NH 3 poisoning in chickens, observed chicken liver morphology, and measured ATPases (Na + K + -ATPase, Mg ++ -ATPase, Ca ++ -ATPase, and Ca ++ Mg ++ -ATPase) activities and the expression of energy metabolism-related genes (HK1, HK2, PK, PFK, PDHX, CS, LDHA, LDHB, AMPK, SDHA, SDHB, and avUCP) and autophagy-related genes (PI3K, LC3I, LC3II, Beclin1, SQSTM1, mTOR, ULK1, ATG5, ATG12, and ATG13) to explore molecular mechanism of energy metabolism disorder and autophagy induced by NH 3 in chicken livers.

Animals and NH 3 treatment
All procedures used in our experiment were performed in accordance with the requirements of the Northeast Agricultural University's Institutional Animal Care and Use Committee with the approval number SRM-06. All chickens were housed in environmental control cabins at the Laboratory Animal Center, College of Veterinary Medicine, Northeast Agricultural University (Harbin, China). One hundred and eight healthy 1-day-old chickens for fattening were randomly divided into three groups (3 replicates per group and 12 chickens per replicate): the low NH 3 group (≤ 5 mg/m 3 during day 1-42), the middle NH 3 group (10 ± 0.5 mg/m 3 during day 1-21, 15 ± 0.5 mg/m 3 during day , and the high NH 3 group (20 ± 0.5 mg/m 3 during day 1-21, 45 ± 0.5 mg/m 3 during day . A cylinder of compressed anhydrous NH 3 (Dawn Gas Co. Ltd., Harbin, China) and a Photoacoustic Field Gas-Monitor Innova-1412 (Lumasense Technologies, Inc., Santa Clara, CA, USA) were used to produce and control NH 3 concentration in each environmental control cabin. Feed and water were provided ad libitum. On the 14th, 28th, and 42nd days, the chickens were euthanized (200 mg/kg sodium pentobarbital) and livers were collected.

Ultrastructural observation
Liver tissues were cut into about 0.5-1 mm 3 tissue blocks. The tissue blocks were xed in 2.5% (v/v) glutaraldehyde solution, were rinsed with phosphate buffered saline (pH = 7.2), and then were xed in 1% (v/v) osmium tetroxide. After being dehydrated with gradient ethanol solution, the tissue blocks were impregnated with epoxy resin and were cut into ultrathin sections. Obtained ultrathin sections were stained with uranyl acetate and lead citrate. The observation and record of liver ultrastructure were performed using a transmission electron microscope (GEM-1200ES, Japan Electron Optics Laboratory Co., Ltd, Tokyo, Japan).

ATPase activity detection
Page 5/22 The activities of Na + K + -ATPase, Ca ++ -ATPase, Mg ++ -ATPase, and Ca ++ Mg ++ -ATPase were detected with a kit (Kit number: A016-2) following the manufacturer's instructions (Nanjing Construction Bioengineering Research). All samples were repeatedly detected three times in a single assay using a spectrophotometer (ELX800, BioTek Instruments Inc., Winooski, USA). Quantitative Real-time PCR (qRT-PCR) mRNA expression of energy metabolism-related and autophagy-related genes was evaluated using qRT-PCR method. Total RNA was isolated from chicken liver tissues with RNAiso Plus (Takara, Japan).
Complementary DNA was synthesized in a 20 µL system with FastKing RT Kit (With gDNase) (Kit number: KR116, Tiangen Biotech Co., Ltd., Beijing, China). All primers ( Table 1) used in our experiment were synthesized by Invitrogen Biotechnology Co. Ltd. (Shanghai, China). The qRT-PCR was performed in a 20 µL reaction mixture (Roche, Switzerland) with a QuantStudio 3 real-time PCR system (Applied Biosystems, Foster City, CA, USA). Table 1 Primer sequences for qRT-PCR

Gene
Forward Primer Reverse Primer Western blot analysis

Statistical analysis
The data analysis was performed using GraphPad Prism 6.0 (GraphPad Software, Inc., USA). One-way and two-way analysises of variance with Tukey's multiple comparison test were used to analyze statistical signi cance. Data were given as mean ± standard deviation (SD). Different lowercase letters represented signi cant differences (P < 0.05) among different groups at the same time point, and different uppercase letters represented signi cant differences (P < 0.05) in the same group among different time points.

Ultrastructural changes
The ultrastructure of chicken liver cells was examined using a transmission electron microscope, and ultrastructural changes were shown in Fig. 1. There was no autophagosome and autolysosome in the low NH 3 group (Fig. 1a). Autophagosomes and autolysosomes enclosing a few organelles were observed in the middle NH 3 group (Fig. 1b). The ultrastructure of chicken liver cells in the high NH 3 group (Fig. 1c1-c2) showed typical autophagy characteristics such as more autophagosomes and autolysosomes enclosing a large number of organelles.

ATPase activity
The activities of Na + K + -ATPase, Mg ++ -ATPase, Ca ++ -ATPase, and Ca ++ Mg ++ -ATPase were presented in Fig. 2. On the 14th, 28th, and 42nd days, the activities of above four ATPases decreased signi cantly (P < In addition, we found that protein expression of the four genes decreased signi cantly (P < 0.05) with NH 3 treatment concentration.
mRNA and protein expression of autophagy-related genes In our experiment, qRT-PCR and Western blot analysis were used to detect mRNA expression and protein expression of autophagy-related genes, respectively. The results were shown in Fig. 4. On the 14th, 28th, and 42nd days, compared with the low NH 3 group, mRNA expression of LC3I (b), LC3II (c), Beclin1 (d), ULK1 (g), ATG5 (h), ATG12 (i), and ATG13 (j) in the middle NH 3 group and the high NH 3 group increased signi cantly (P < 0.05), and mRNA expression of above genes was the highest signi cantly (P < 0.05) in the high NH 3 group. However, mRNA expression of PI3K (a), SQSTM1 (e), and mTOR (f) in the middle NH 3 group and the high NH 3 group was signi cantly lower than those in the low NH 3 group (P < 0.05), and mRNA expression of above genes was the lowest signi cantly in the high NH 3 group (P < 0.05). Protein expression of PI3K and SQSTM1 decreased signi cantly, and protein expression of ULK1 and ATG12 increased signi cantly with the increase of NH3 concentration on day 42 (P < 0.05).
There was no signi cant difference in mRNA expression of detected autophagy-related genes among different time points in the low NH 3 group (P > 0.05). Regarding the middle NH 3 group and the high NH 3 group, mRNA expression of LC3I, LC3II, Beclin1, ULK1, ATG5, ATG12, and ATG13 increased signi cantly with the increase of NH 3 exposure time (P < 0.05) except for there was no signi cant difference in Beclin1 between on the 14th day and on the 28th day (P > 0.05). mRNA expression increased with the increase of NH 3 exposure time, but there was no signi cant difference between the 14th day and the 28th day (P > 0.05). Besides, mRNA expression of PI3K, SQSTM1, and mTOR decreased signi cantly with the increase of NH 3 exposure time (P < 0.05).

Discussion
Cells can show a series of morphological changes to response to stress after being stimulated by harmful substances. For example, hydrogen sul de exposure can lead to nuclear membrane contraction, chromatin aggregation and marginalization, mitochondrial cristae swelling and vacuolation [39]. Previous studies have shown that NH 3 toxicity can cause pathological changes in cell morphology and lead to tissue damage. A study found that the morphology of human hepatic stellate cells changed dramatically, with spindle shaped broblast phenotype and signs of abnormal lumen of lysosome after NH 3 treatment [40]. There are many autophagic bodies with double vesicles in skeletal muscle cells of mice treated with high NH 3 [38]. Similar to the above studies, in our experiment, high-concentration NH 3 resulted in tissue damage in chicken livers. The ultrastructure showed that a large number of mitochondrial cristae broke, and autophagy bodies and autophagy lysosomes wrapped with a large number of organelles appeared, indicating that the toxicity of NH 3 resulted in the destruction of mitochondrial structure and obvious autophagy damage.
Some researchers found that harmful stimulation can damage mitochondrial function, destroy energy metabolism balance, and lead to the disorder of energy metabolism. Chi Q et al. found that harmful gas H 2 S exposure led to the damage of chicken spleen mitochondrial structure, decreased ATPase activity and the expression of energy metabolism related genes (HK2, PK, PDHX, SDHB and avUCP), which indicating that energy metabolism disorder occurred in chicken spleens [41]. Mitochondrial dysfunction and energy metabolism disorder were related to NH 3 toxicity [42]. Another report demonstrated that H 2 S can also reduce the expression of AMPK, HK1, PFK, LDHA and LDHB, leading to the disorder of energy metabolism in chicken livers [43]. NH 3 toxicity inhibited SDH activity, leading to mitochondrial membrane potential collapse, mitochondrial swelling, and ATP depletion, which con rmed that mitochondrial dysfunction and energy metabolism disorder occurred in mice livers and brains under NH 3 toxicity [44]. In order to explore the effect of NH 3 on energy metabolism of chicken livers, we measured the expression of energy metabolism-related genes (HK1, HK2, PK, PFK, PDHX, CS, LDHA, LDHB, SDHA, SDHB, and avUCP) and found that NH 3 inhibited above energy metabolism-related genes and caused energy metabolism disorder in chicken livers. ATPase can catalyze the process of ATP decomposition and release energy to supply cells to complete various biological reactions, which plays a key role in the energy conversion and utilization of biological cells. Accordingly, it is very important to maintain ATPase activity in energy metabolism process. Cao C et al. reported that selenium de ciency inhibited the activities of Na + K + -ATPase, Ca ++ -ATPase, and Ca ++ Mg ++ -ATPase in chicken arteries and veins [45]. Lead exposure reduced the activities of Na + K + -ATPase, Mg ++ -ATPase, and Ca ++ -ATPase in chicken kidney tissues [46]. Hence, in our present experiment, we further determined the activities of Na + K + -ATPase, Mg ++ -ATPase, Ca ++ -ATPase, and Ca ++ Mg ++ -ATPase, and found that high-concentration NH 3 decreased the activities of above four ATPases, meaning that excess NH 3 inhibits energy metabolism in chicken livers. In addition, we found that the decrease of the expression of energy metabolism-related genes and ATPases activities was dose-dependent and time-dependent. Our above results suggested that NH 3 exposure destroyed energy metabolism balance, which may lead to energy de ciency. AMPK plays an important role in signal pathway of energy and substrate metabolism, and is a central energy sensor to maintain metabolic homeostasis [47]. Interestingly, AMPK mRNA and protein expression increased signi cantly under NH 3 exposure. This may be due to that upon activation, AMPK can inhibit energy requiried for cellular activities and meanwhile stimulate catabolic pathways [48]. The decrease of energy level in liver tissues activated AMPK. AMPK compensatively satis ed energy demand through inhibiting energy consumption process, activating energy production process, and even activating autophagy process. Hence, the compensatory increase of AMPK level further re ected energy metabolism disorder induced by NH 3 in chicken livers.
Autophagy is closely related to energy metabolism disorder. Under energy de ciency state, AMPK can directly phosphorylate ULK1, thus further promoting autophagy [49]. Phosphoserine phosphatase can inhibit autophagy in hepatocellular carcinoma cell via the AMPK/mTOR/ULK1 signaling pathway [50]. ULK1 complex is a bridge between upstream energy receptor AMPK/mTOR and downstream autophagy [49]. Wang S et al. demonstrated that cadmium exposure led to energy metabolism disorder, meanwhile the expression of autophagy-related genes LC3, ATG5, and Beclin1 increased signi cantly, and mTOR expression decreased, suggesting the occurrence of energy metabolism disorder and autophagy in chicken ovaries [51]. Cadmium exposure inhibited mTOR expression by inducing AMPK expression, and increased the expression of LC3I, LC3II and Beclin1, which triggered BNIP3 dependent autophagy pathway [52]. Under normal conditions, PI3K can phosphorylate mTOR to maintain the rich activity of mTOR in tissues, thus inhibiting ULK1 and ATG13 complex and preventing them from starting autophagy [53]. Autophagy signal can inhibit mTOR and activate ULK1. Once activated, ULK1 can phosphorylate ATG13 and induce Beclin1 activation and other autophagy activating molecules to recruit autophagy proteins to participate in the formation of autophagy bodies [54]. Therefore, AMPK/mTOR/ULK1 pathway plays an important role in regulating autophagy process. ATG5, a key protein involved in membrane extension of phagocytes in autophagic vesicles, forms a complex with ATG12 and locates on the outer membrane of autophagy, which is crucial for autophagic precursors formation [55]. ATG5/ ATG12 complex can promote the binding of LC3I to autophagic membrane. The membrane binding LC3II is an important marker molecule of autophagy, which increases with the increase of autophagic membrane [34]. SQSTM1, as an autophagic substrate marker, can connect LC3 and ubiquitination substrates, and then be integrated into autophagy and degraded in autophagy lysosomes [56]. Deng X et al. reported that PM2.5 increased the level of LC3, ATG5 and Beclin1 in a time-and concentration-dependent manner, suggesting that PM2.5 induced autophagy in A549 cells [57]. Puerarin reduced autophagic bodies formation in hippocampal CA1 area after cerebral ischemia-reperfusion injury, decreased the expression of Beclin1, the ratio of LC3II/LC3I, AMPK and ULK1, and increased SQSTM1 expression, indicating that puerarin alleviated autophagy through inhibiting the activation of AMPK/mTOR/ULK1 signaling pathway [58]. NH 3 toxicity induced autophagy by inhibiting mTOR expression and increasing expression of LC3I, LC3II, Beclin1, and ATG5, which con rmed that NH 3 toxicity induced cardiac autophagy in chickens [59].
In this study, we measured the expression of autophagy-related genes (PI3K, LC3I, LC3II, Beclin1, SQSTM1, mTOR, ULK1, ATG5, ATG12 and ATG13) in chicken livers. The results were consistent with above mechanisms and researches. NH 3 exposure triggered AMPK/mTOR/ULK1 signaling pathway through inhibiting mTOR and activating ULK1 and induced autophagy, and energy metabolism disorder midiated autophagy caused by NH 3 in chicken livers.

Conclusion
Taken together, our results demonstrated that NH 3 exposure damaged liver tissue, inhibited ATPases activities, and altered the expression of energy metabolism-related and autophagy-related genes. There was dose-dependent and time-dependent effect on all detected indexes in NH 3 -caused chicken liver damage. Excess NH 3 induced energy metabolism disorder and autophagy. Energy metabolism disorder mediated NH 3 -induced autophagy by triggering AMPK/mTOR/ULK1 pathway (Fig. 5  The activities of Na+K+-ATPase, Mg++-ATPase, Ca++-ATPase, and Ca++Mg++-ATPase in chicken livers. Each value represented mean ± SD (n = 12). Different lowercase letters represented signi cant differences (P < 0.05) among different groups at the same time point, and different uppercase letters represented signi cant differences (P < 0.05) in the same group among different time points.