MicroRNA-31-5p Exacerbates Lipopolysaccharide-Induced Acute Lung Injury via Inactivating Cab39/AMPKα Pathway

Acute lung injury (ALI) and the subsequent acute respiratory distress syndrome remain devastating diseases with high mortality rates and poor prognoses among patients in intensive care units. The present study is aimed at investigating the role and underlying mechanisms of microRNA-31-5p (miR-31-5p) on lipopolysaccharide- (LPS-) induced ALI. Mice were pretreated with miR-31-5p agomir, antagomir, and their negative controls at indicated doses for 3 consecutive days, and then they received a single intratracheal injection of LPS (5 mg/kg) for 12 h to induce ALI. MH-S murine alveolar macrophage cell lines were cultured to further verify the role of miR-31-5p in vitro. For AMP-activated protein kinase α (AMPKα) and calcium-binding protein 39 (Cab39) inhibition, compound C or lentiviral vectors were used in vivo and in vitro. We observed an upregulation of miR-31-5p in lung tissue upon LPS injection. miR-31-5p antagomir alleviated, while miR-31-5p agomir exacerbated LPS-induced inflammation, oxidative damage, and pulmonary dysfunction in vivo and in vitro. Mechanistically, miR-31-5p antagomir activated AMPKα to exert the protective effects that were abrogated by AMPKα inhibition. Further studies revealed that Cab39 was required for AMPKα activation and pulmonary protection by miR-31-5p antagomir. We provide the evidence that endogenous miR-31-5p is a key pathogenic factor for inflammation and oxidative damage during LPS-induced ALI, which is related to Cab39-dependent inhibition of AMPKα.


Introduction
Acute lung injury (ALI) and the subsequent acute respiratory distress syndrome are devastating diseases manifested as severe refractory hypoxemia and multiple organ failure, which cause high mortality rates and poor prognoses among patients in the intensive care units. Despite the advances in mechanical ventilations and symptomatic therapies, no specific and effective management strategies are available for ALI patients [1,2]. Lipopolysaccharide (LPS) is a major component of the outer membranes in Gram-negative bacteria and functions as a key pathogenic factor to induce sepsisrelated ALI. Upon LPS exposure, the downstream nuclear factor kappa-B (NF-κB) is activated to trigger the synthesis of multiple inflammatory mediators, including interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α). These cytokines in turn recruit leukocytes (e.g., neutrophils and macrophages) to infiltrate into lung tissue and further amplify sepsis-induced inflammation and lung injury [3,4]. Nucleotide-binding domain-like receptor protein 3 (NLRP3) inflammasome acts as a molecular scaffold for the maturation of various procytokines that contributes to inflammatory injury during sepsis-related ALI [3,4]. Besides, these leukocytes and excessive inflammation also promote an overproduction of reactive oxygen species (ROS) and elicit oxidative damage to pulmonary cells. In contrast, oxidative stress augments leukocyte chemotaxis and NLRP3 activation, thereby accelerating proinflammatory cytokine production and pulmonary injury [5]. To scavenge these free radicals, many antioxidant enzymes are synthesized under the control of a redox-sensitive transcription factor, named nuclear factor-erythroid 2 related factor 2 (NRF2) [6]. Therefore, inhibiting inflammation and oxidative stress may provide an effective method for the prevention and treatment of ALI.
AMP-activated protein kinase α (AMPKα) is a highly conserved serine/threonine protein kinase among eukaryotic organisms and has diverse beneficial functions on energy modulation, mitochondrial homeostasis, autophagic flux, fibrotic remodeling, and cell death [7][8][9]. Recent studies indicate that AMPKα also plays critical roles in the pathogenesis of sepsis-induced ALI via regulating inflammation and oxidative stress. Lv et al. found that AMPKα activation suppressed intrapulmonary inflammation and oxidative damage, thereby preventing LPS-induced ALI, while conversely, AMPKα inhibition exacerbated lung injury in mice [10,11]. Taken together, these findings provide a basis for targeting AMPKα as the promising strategy to treat sepsisrelated ALI.
MicroRNAs (miRs) are a class of evolutionarily conserved, single-stranded short noncoding RNAs that regulate gene expression at posttranscriptional levels through binding to the 3 ′ -untranslated regions (UTR) of targeted messenger RNAs [12,13]. A number of researches have proved the necessity of microRNAs in modulating sepsis-induced inflammation, oxidative stress, and ALI [14,15]. miR-31-5p is well accepted as an oncogenic microRNA and participates in the proliferation, migration, invasion, and chemosensitivity of cancer cells [16,17]. Yet, Kim et al. found that miR-31-5p was elevated in TNF-α-treated human endothelial cells and defined it as a NF-κB-responsive microRNA with inflammation-modulating actions [18]. Results from Toyonaga et al. indicated that miR-31-5p was upregulated in inflammatory bowel disease and was associated with colonic epithelial cell integrity and function [19]. Besides, a recent study reported the role of miR-31-5p on ROS accumulation in hepatocellular carcinoma [16]. Based on these data, we supposed that miR-31-5p might be involved in the pathogenesis of LPS-induced ALI.

Experimental Models.
Male C57BL/6 mice (8-10 weeks old) received a single intratracheal injection of LPS (5 mg/kg) for 12 h to induce ALI in vivo, while an equal volume of saline was used as the negative control. In a separate study, the mice were intratracheally injected with a lethal dose of LPS (25 mg/kg) for survival analysis [3,14]. For the treatment of miR-31-5p duplexes, the mice were intravenously treated with miR-31-5p agomir, antagomir, and their negative controls at indicated doses for 3 consecutive days before LPS injection according to a previous study [14]. To inhibit AMPKα in mice, CpC (20 mg/kg) was intraperitoneally injected every two days from 1 week before miR-31-5p manipulation [20]. For Cab39 knockdown in lung tissue, the mice were exposed to a single intravenous injection of 2 × 10 7 PFU shCab39 or shCtrl as a control [21]. All experimental procedures were in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines and also approved by the Animal Ethics Committee of Renmin Hospital of Wuhan University (approval no. WDRM 20190114).

Pulmonary Function Evaluation.
Pulmonary function was calculated from the continuous respiratory data using the Buxco system (Sharon, CT, USA). Airway resistance, dynamic lung compliance, and pulmonary ventilation were carried out in anesthetized mice after a brief acclimation to the chamber.

Blood Gas Analysis.
A heparinized PE10 polyethylene catheter was used to collect blood samples from the right common carotid artery of the mice, and then the blood gas parameters were analyzed by an IL GEM® Premier 3000 blood gas analyzer.

Lung
Wet-to-Dry Ratio Calculation. After they were excised, the lungs were blotted dry and weighed immediately to obtain the wet weight. Next, the lungs were desiccated in an oven at 80°C for 4 days until constant weight to get the dry weight [3]. The lung wet-to-dry ratio was calculated to assess pulmonary edema.

Bronchoalveolar Lavage Fluid (BALF) Measurement.
After euthanasia, the mice received intratracheal injections of 1 mL ice-cold phosphate-buffered saline (PBS, pH = 7:4) for 3 times to collect the BALF samples, which were then centrifuged at 1500 rpm for 5 min at 4°C to pellet the cells [3,4]. The cell-free supernatants were stored at -80°C for detecting cytokines and total proteins. The sedimented cell pellets were resuspended in 0.5 mL PBS and counted by Wright-Giemsa's staining and a hemocytometer. Total protein concentrations were directly measured using a BCA protein assay kit, while the inflammatory cytokines were measured using the enzyme-linked immunosorbent assay (ELISA) kits.
2.9. Cytokine Detection. The concentrations of IL-6, TNF-α, IL-1β, and IL-18 in BALF, lung tissue, or cell medium were measured using the available kits according to the manufacturer's instructions. The data were calculated by comparing the optical density with the standard curve.
2.10. Enzymatic Activity Measurement. The activities of LDH, MPO, caspase-1, SOD, CAT, and Gpx in lung tissue or macrophages were determined by the commercial kits following the manufacturer's instructions.

Detection of Intracellular ROS and Oxidative Products.
The lung homogenates or cell lysates were incubated with DCFH-DA (50 μmol/L) at 37°C for 30 min in the dark [29,30]. DCF fluorescence intensities were then measured using a Bio-Tek microplate reader at the excitation and emission wavelengths of 485 nm and 535 nm, respectively. The levels of oxidative products for lipids (MDA and 4-HNE) or proteins (3-NT) were detected using the commercial kits according to the manufacturer's instructions.
2.12. Dual-Luciferase Reporter Gene Assay. The wild type or mutant 3′-UTR of Cab39 was cloned into the pGL3 plasmid (Promega; Madison, WI, USA) containing a luciferase report gene, which was then cotransfected with the miR-31-5p agomir or AgNC using the Lipofectamine 6000 Reagent (Beyotime; Shanghai, China). The cells were cultured for 48 h and then collected to detect the firefly and Renilla luciferase activities by a Dual-Luciferase Reporter Assay System (Promega; Madison, WI, USA) [31][32][33].
2.13. Statistical Analysis. All data were presented as the mean ± standard deviation (SD) and analyzed using the SPSS 22.0 software. An unpaired two-tailed Student t-test was performed to compare the significance between two groups, whereas the differences among three or more groups were calculated by one-way ANOVA analysis followed by Tukey's post hoc test. Statistical significance was defined as P < 0:05.

miR-31-5p Antagomir Alleviates LPS-Induced ALI in
Mice. We first explored whether miR-31-5p expression was altered during ALI, and the data identified an upregulation of miR-31-5p levels in lung tissue upon LPS injection (Figure 1(a)). As shown in Figure 1(b), miR-31-5p antagomir normalized the aberrant miR-31-5p expression in LPSinjured lungs to a physiological level at the dose of 100 mg/kg; therefore, we selected this dose of miR-31-5p antagomir in our further study. Intriguingly, miR-31-5p antagomir treatment decreased airway resistance and increased lung compliance and pulmonary ventilation in mice with LPS-induced ALI (Figure 1(c)). Accordingly, the reduced partial pressure of arterial oxygen (PaO 2 ) was also prevented by miR-31-5p antagomir (Figure 1(d)). LPS injection caused severe pulmonary edema and protein leakage, which were attenuated in mice with miR-31-5p antagomir treatment, as confirmed by the decreased wet-to-dry ratio of lung tissue and BALF protein concentrations (Figures 1(e) and 1(f)). LDH is a critical marker of cellular damage, and the increased LDH activity in lung tissue after LPS injection was remarkably decreased by miR-31-5p antagomir ( Figure 1(g)). Moreover, miR-31-5p antagomir treatment could improve the survival status of LPS-challenged mice ( Figure 1(h)). Collectively, we conclude that miR-31-5p antagomir alleviates lung injury and enhances the respiratory function of LPS-treated ALI mice.

miR-31-5p Agomir Aggravates LPS-Induced ALI in Mice.
We then investigated whether the increased miR-31-5p expression could accelerate LPS-induced ALI in mice. As depicted in Figure 2(a), miR-31-5p agomir elicited almost two times expression of miR-31-5p in lung tissue upon LPS stimulation at the dose of 50 mg/kg; thus, we used this dose in the next study. As mentioned above, the mice with LPS injection had lower PaO 2 that was further decreased by miR-31-5p agomir treatment (Figure 2(b)). Besides, the airway resistance was further decreased, and the lung

miR-31-5p Antagomir Mitigates Intrapulmonary
Oxidative Damage in LPS-Treated Mice. Oxidative stress is the other feature of LPS-induced ALI [3,14]. As shown in Figures 4(a)-4(c), the lungs with LPS stimulation had increased ROS generation and oxidative products for lipids (MDA and 4-HEN) and proteins (3-NT) that were significantly decreased by miR-31-5p antagomir. SOD, CAT, and Gpx are three key intracellular antioxidant enzymes to scavenge the excessive free radicals for redox homeostasis. We observed that miR-31-5p antagomir restored total SOD, CAT, and Gpx activities in lungs with LPS injury (Figure 4(d)). NRF2 is an important transcription factor in regulating the expression of numerous antioxidant enzymes, and we thus detected whether miR-31-5p antagomir affects the NRF2 pathway. As depicted in Figure 4(e), LPS significantly decreased NRF2 expression and nuclear accumulation in lung tissue, yet to a lesser extent compared to those in lung tissue with miR-31-5p antagomir protection. Altogether, these data demonstrate an antioxidant role of miR-31-5p antagomir in LPS-induced ALI.

miR-31-5p Antagomir Prevents LPS-Induced
Inflammation, Oxidative Stress, and ALI via Activating AMPKα In Vivo and In Vitro. We next tried to clarify whether AMPKα was required for the protective effects of miR-31-5p antagomir against LPS-induced ALI. As depicted in Figure 6(a), LPS-induced AMPKα inactivation was prevented by miR-31-5p antagomir, while it was further exacerbated by miR-31-5p agomir (Figures 6(a) and 6(b)). To verify   Figure 6: miR-31-5p antagomir prevents LPS-induced inflammation, oxidative stress, and ALI via activating AMPKα in vivo. (a and b) The mice were pretreated with miR-31-5p agomir (50 mg/kg) or antagomir (100 mg/kg) for 3 consecutive days and then received LPS (5 mg/kg) stimulation for an additional 12 h. Representative western blot images and the statistical data (n = 6). (c and d) CpC (20 mg/kg) was intraperitoneally injected every two days from 1 week before miR-31-5p manipulation to inhibit AMPKα in mice. IL-6, TNF-α, and the number of inflammatory cells were measured in BALF (n = 6).  Oxidative Medicine and Cellular Longevity the levels of ROS content, MDA, 4-HNE, and 3-NT formation in macrophages, but not in those with CpC treatment (Figures S3C-S3E). These studies define AMPKα as a potential molecular target for the protective effects of miR-31-5p antagomir against LPS-induced ALI.
3.8. miR-31-5p Antagomir Activates AMPKα via Increasing Cab39 Expression. We finally investigated the possible pathway through which miR-31-5p antagomir activated AMPKα. TargetScan software was used to predict the potential target of miR-31-5p, and we observed a putative binding site of miR-31-5p in the 3′-UTR of Cab39 that serves as a scaffold protein for AMPKα activation (Figure 7(a)) [35]. Besides, LPS-elicited Cab39 suppression in lungs was preserved by miR-31-5p antagomir (Figure 7(b)). To examine whether miR-31-5p can directly bind to the 3′-UTR of Cab39, a dual-luciferase reporter gene assay was performed. As shown in Figure 7(c), luciferase activities were significantly inhibited when miR-31-5p was cotransfected with the luciferase plasmid harboring wild type Cab39 3 ′ -UTR, yet it failed to do the same when the binding site was mutated. To strengthen the role of Cab39 in AMPKα activation by miR-31-5p antagomir, we knocked down Cab39 expression in lung tissue using lentiviral vectors (Figure 7(d)). As expected, Cab39 silence blocked AMPKα activation in miR-31-5p antagomir-treated murine lungs, accompanied by increased levels of BALF IL-6 and TNF-α as well as ROS generation (Figures 7(e)-7(g)). Accordingly, the beneficial effects of miR-31-5p antagomir against LPS-induced pulmonary edema and dysfunction were also retarded after Cab39 silence (Figures 7(h) and 7(i)). Consistent with the in vivo data, AMPKα activation by miR-31-5p antagomir was counteracted in macrophages with shCab39 infection, and the inhibitory effects on inflammation and ROS generation were also abolished (Figures 7(j) and 7(k)). Therefore, we summarize that Cab39 is required for miR-31-5p antagomirmediated AMPKα activation and the subsequent pulmonary protection against LPS-induced ALI.

Discussion
Our present study indicates that miR-31-5p is upregulated in murine lungs upon LPS stimulation and that this upregulation is instrumental for the provocation of inflammation and oxidative damage both in mice and in cultured macrophages. miR-31-5p antagomir attenuates, while miR-31-5p agomir exacerbates pulmonary injury and dysfunction in LPS-treated mice. Besides, we report that miR-31-5p antagomir reduces proinflammatory cytokine secretion, ROS generation, and lung injury via activating AMPKα, and conversely, AMPKα inhibition by CpC blocks the pulmonary protection in vivo and in vitro. Further data identify Cab39 as a direct target of miR-31-5p, and miR-31-5p antagomir prevents LPS-induced Cab39 downregulation and thus activates AMPKα. We report here for the first time that endogenous miR-31-5p is a key pathogenic factor for inflammation and oxidative damage during LPS-induced ALI.
Excessive inflammation and oxidative stress are thought to play critical roles in the initiation and progression of ALI [3,36]. Inflammatory cells and the proinflammatory cytokine notably trigger free radical overproduction, and ROS in turn activates the inflammatory programs, which create a vicious cycle to accelerate inflammation and oxidative damage and provoke the occurrence of ALI [5]. Upon LPS stimulation, NF-κB is phosphorylated and translocated from the cytoplasm to the nucleus, where it binds to its consensus sequence on the promoter-enhancer region of targeted genes and drives the transcription of inflammatory cytokines [37]. Besides, the local inflammation also recruits circulating leukocytes into the lung tissue and further amplifies the inflammatory response. Inflammasomes function as the molecular scaffolds for inflammatory response and are essential for the maturation of multiple procytokines [38][39][40]. Emerging studies demonstrate a central role of NLRP3 inflammasome in the pathological process of ALI. Upon stimulation, the ASC adaptor interacts with the NLRP3 scaffold to activate caspase-1, which then proteolytically cleaves the precursors of multiple proinflammatory cytokines and releases the mature forms, including IL-1β and IL-18 [3,4,14]. Consistently, LPS stimulation activated NF-κB and NLRP3 inflammasome in the present study, and elicited increases of multiple proinflammatory cytokines in lung tissues and cultured alveolar macrophages, which were notably blunted by miR-31-5p antagomir treatment. Oxidative stress is also implicated in the development of ALI. Free radicals, such as hydrogen peroxide and superoxide anion, are increased in lung tissue in response to LPS injury, which directly cause lipid and protein peroxidation, resulting in the injury and death of lung cells [3]. Besides, ROS overproduction by LPS challenge in lung tissue also triggers the dissociation of thioredoxin interacting protein (TXNIP) from thioredoxin, which then binds to and activates NLRP3 inflammasome [39,41]. The redox sensor NRF2 is physiologically retained in the cytoplasm by Kelch-like ECH-associated protein 1 (KEAP1), but it detaches from KEAP1 upon oxidative stress and subsequently translocates to the nucleus to trigger the antioxidant response [22]. Herein, we observed that miR-31-5p antagomir restored NRF2 expression and nuclear accumulation in LPS-induced ALI, thereby preventing ROS overproduction and oxidative damage.
miR-31-5p is a well-known tumor-associated microRNA and participates in the progression of various tumors, including lung cancer, colorectal cancer, and hepatocellular carcinoma [16,17]. Emerging studies indicate that miR-31-5p also plays indispensable roles in maintaining the pathophysiological homeostasis in noncancerous tissues. Liu et al. proved that miR-31-5p could repress the proliferation and differentiation of tongue myoblasts [42]. Results from Ji et al. indicated that miR-31-5p silence significantly alleviated the doxorubicin-induced myocardial apoptosis and cardiac dysfunction in mice [43]. A very recent study by Toyonaga et al. demonstrated that miR-31-5p was required for colonic epithelial cell integrity and could predict the clinical outcomes in patients with Crohn's disease [19]. We herein for the first time identified the pathogenic role of miR-31-5p in LPS-induced inflammation, oxidative stress, and ALI. Cab39 functions as a scaffold protein of liver kinase B1 (LKB1), an upstream kinase of AMPKα, and stabilizes the 10 Oxidative Medicine and Cellular Longevity Putative binding site of miR-31-5p in Cab39 3′-UTR

11
Oxidative Medicine and Cellular Longevity LKB1's activity through forming a heterotrimeric complex with the STE20-related kinase adaptor in the cytoplasm [35,44]. In the current study, we observed that miR-31-5p directly bound to the 3 ′ -UTR of Cab39 to inhibit its expression, while miR-31-5p antagomir restored the Cab39 protein level, accompanied by an increased AMPKα phosphorylation and the improvement on LPS-induced inflammation, oxidative stress, and pulmonary dysfunction. Despite being famous as an energy sensor, AMPKα also plays indispensable roles in regulating inflammation and oxidative stress. Its activation was reported to reduce ROS generation and prevent diabetes-, sepsis-, or doxorubicin-induced cardiac injury [45][46][47]. In line with our findings, some investigators proved that AMPKα notably enhanced the expression and nuclear translocation of NRF2 to scavenge the excessive free radicals [3,10]. Besides, a recent study found that AMPKα activation effectively decreased p47phox expression and phosphorylation, thereby decreasing the generation of free radicals [48]. Previous data indicated that the inhibition of oxidative stress by AMPKα significantly blocked NLRP3 inflammasome activation and inflammatory damage [3,10,48]. NF-κB is essential for NLRP3 inflammasome activation, and the findings from us and other laboratories demonstrated that NF-κB inhibition was sufficient to alleviate the activation of NLRP3 inflammasome [34]. Moreover, results from Chen et al. revealed that AMPKα regulated dynamin-related protein 1-mediated mitochondrial fission and thereby restrained NLRP3 inflammasome activation [49]. Collectively, our data defined miR-31-5p as a promising therapeutic target for the treatment of ALI.
Of note, there exist some limitations in the current study. First, an invasive respiratory mechanics as previously described would be more appropriate to evaluate pulmonary function [50]. Second, whether miR-31-5p affects the production of free radicals need to be confirmed. Besides, the role of miR-31-5p in other pulmonary cells, e.g., the lung epithelial cells, during the progression of ALI should be investigated in further studies.

Data Availability
The data that support the findings of this study are available from the corresponding authors upon reasonable request.