lncRNA AK054386 Functions as a ceRNA to Sequester miR-199 and Induce Sustained Endoplasmic Reticulum Stress in Hepatic Reperfusion Injury

Hepatic ischemia-reperfusion injury (IRI) is a very complex pathological process that is often associated with liver trauma and surgery, especially liver transplantation surgery. Although endoplasmic reticulum stress (ERS) plays a role in this process, the posttranscriptional regulators and the underlying mechanisms are still unclear. Here, we report that the lncRNA AK054386 was increased in hepatic IRI models. Furthermore, AK054386 can act as a “competing endogenous RNA (ceRNA)” and regulate ERS-related factors by binding and sequestering miR-199, which was shown to inhibit ERS in our previous report. Increased expression of AK054386, which might be mediated by activated NF-κB, resulted in sustained ERS and increased cell apoptosis and death in hepatic IRI mouse and cellular models. In contrast, AK054386 inhibition had protective effects on these models. Our data indicate that AK054386 and miR-199 are critical players in hepatic IRI, and we broadened the scope regarding ceRNA mechanisms. We hope that our results will improve the understanding of hepatic IRI and may provide potential therapeutic targets.

(A). Representative hematoxylin-and-eosin (HE) stained sections of liver tissues from the sham-operated controls (Control), the mice suffering from hepatic ischemia without reperfusion (Ischemia) and the hepatic IRI mouse model (IRI). More serious necrosis was observed in the IRI group compared with that in the other two groups. AK054386 functions as a ceRNA to sequester miR-199 and induce sustained ER stress in hepatic IRI. The unfolded protein-induced ER stress response could result in the elevation of ER mediator mRNAs. In normal hepatocytes, miR-199 could negatively regulate ER stress by targeting the mRNAs of ER-related factors, which 4 promotes hepatocyte survival. In hepatic IRI, over-activated UPR causes sustained ERS, which induces the activation of Nuclear Factor-κB (NFKB1). NF-κB can bind to the AK054386 promoter and induce its transcription. This LncRNA then sequesters miR-199, resulting in the up-regulation of GRP78, ATF6 and IRE1a, which promotes aggravated and sustained ER stress. This positive feedback response causes hepatocyte apoptosis and cell death.

Supplementary Methods
Mice, cell lines and reagents. C57BL/6 mice (Female at 9-11 weeks) were purchased from the SMMU Laboratory Animal Center and were used in accordance with the Institutional guidelines for animal care. The cell lines were cultured following a common protocol that was described in our previous manuscript (Dai BH, et al., 2013). The mouse hepatocyte line BNL-CL2 was obtained from the American

Induction of hepatic IRI model in mice
Hepatic IRI was induced following a common protocol that is described previously (Abe et al., 2009, Kim J.Y. et al., 2017. Briefly, mice underwent a midline laparotomy after anesthetization with chloral hydrate, and the hepatic hilum was dissected. A microvascular clamp was applied to the Glisson system for constructing 6 the hepatic IRI model. The clamp was removed and the abdominal wall was closed after 1 h of hepatic ischemia. After a designated period (1 h/3 h/6 h) of reperfusion, the whole blood or liver samples of the mice were collected following anesthesia for further study (histological analysis, TUNEL assay, and other molecular analyses). The hepatic ischemia models underwent only 1 h of ischemia without reperfusion.
Sham-operated controls underwent the same procedure except for vascular occlusion.
For all the mouse experiments, 5 mice were in each group (Sham-operated controls, ischemia and IRI).

Induction of hepatic IRI model in cells.
A three gas incubator was set with the following gas concentrations: (1) (2) 5.0% CO 2 and (3) 94% N 2 . The next day, the normal medium on the cultured mouse hepatocyte line BNL-CL2 was changed to ischemia-mimic medium(10.0 mmol/L KCl, 98.5 mmol NaCl, 0.9 mmol/L NaH 2 PO 4 , 20.0 mmol/L HEPES, 6.0 mmol/L NaHCO 3 , 1.8 mmol/L CaCl 2 , 1.2 mmol/L MgSO 4 , and 40.0 mmol/L sodium lactate pH 6.8). Then, the cells were placed in to the three gas incubator and cultured for 8 h for ischemia simulation. To simulate reperfusion, the cells were cultured under normal conditions at 5% CO 2 and 37°C for 3 h and the ischemia-mimic medium was then replaced with normal culture medium.

Serum measurements. Mice serum levels of Aspartate aminotransferase (AST) and
alanine aminotransferase (ALT) were measured using a commercial AST Activity Assay Kit (Biovision) and ALT Activity Assay Kit (Biovision) according to the manufacturer's instructions at the SMMU Animal experiment center.
The liver tissues were homogenized with ice-cold PBS. ELISA kits were applied for measurement of IL-6 and CRP levels in liver tissue homogenates in accordance with 7 instructions (Xitang Biotechnology, Shanghai, China).

Subcellular fractionation. Cytosolic fractions and nuclear fractions of BNL-CL2 cell
were isolated and collected with a PARIS kit AM1921 (AmTX) according to the manufacturer's instructions to investigate the subcellular distribution of AK054386.
Then, the total RNA was extracted from the cytosolic fractions and nuclear fractions respectively, followed by cDNA synthesis and qRT-PCR .

Gene over-express and knock-down. Lentiviral vectors were purchased from
Genepharma (Shanghai, China). For miR-199 overexpression, the lentivirus contained the premiR-199 sequences, whereas for AK054386 overexpression, the whole sequence was included in the lentivirus. The lentiviral infection experiments were performed according to the manufacturer's instructions and was described in our previous manuscript (Dai BH, et al., 2013). For the in vivo experiments, the lentivirus was injected intravenously 3 days before hepatic IRI modeling surgery. The Apoptosis assays. Apoptosis was analyzed by flow cytometry analysis following a common protocol that was described in our previous manuscript (Dai BH, et al., 2013). Cells from each sample were tested using Annexin V-PE apoptosis detection kit (BioVision, Inc., USA) by a BD FACSCalibur flow cytometer (BD Bioscience, San Jose, CA, USA) according to the manufacturer's instructions. Triplicate samples (30000 events/ sample)were acquired and analyzed using the FACSDiva software (version 4.1.2; BD Biosciences). Apoptosis was also analyzed by TUNEL which evaluated DNA fragmentation as reported in a previous protocol (Mosbah et al., 2012;Zaouali et al., 2013).
Cytotoxicity assays. Hepatocyte damage was determined by analyzing the LDH in cell culture supernatants using semi-automated and routine clinical methods.

RNA preparation, reverse transcription and quantitative real-time-PCR
(qRT-PCR). Methods in this part are also following a common protocol that was described in our previous manuscript (Dai BH, et al., 2013).. Briefly, total RNA was extracted using TRIzol reagent (Invitrogen). Taqman probes and primer sets (AB, Foster City, CA, USA) were used for testing the miRNA levelsaccording to the manufacturer's instructions. For mRNA analysis, the Reverse Transcription System Kit (Promega, Madison, WI, USA) was used for the first-strand cDNA generatation .
ThePower SYBR Green PCR Master Mix (Applied Biosystems)was used for real-time PCR testing on a StepOne Plussystem (Applied Biosystems). GAPDH mRNA levels or U6 snoRNA were used as internal normalization controls. Student's t-test were performed for statistical comparisons between experimental groups, and P <0.05 was considered statistically significant.