Maresin 1 Attenuates Ventilator-induced Lung Injury by Reducing Oxidative Stress through the Nrf2/HO-1 and NF-κB Pathways

Mechanical ventilation could support the lives of patients with respiratory failure in a variety of ways, including maintaining airway patency and improving oxygenation, etc. However, mechanical ventilation itself could lead to lung damage, which is called mechanical ventilation-related lung injury (VILI). The incidence of VILI is high and the prognosis is poor, so clarifying the mechanism of VILI and seeking effective preventive and therapeutic measures are the urgent medical problem to be resolved. By constructing the VILI model, we studied the effect of Maresin1 on VILI and explored its possible mechanism at the animal level. We tested the related indicators of lung injury, the oxidative stress response, and the inammatory response. The results indicated that Maresin1 could inhibit the oxidative stress response and excessive inammation, thus ameliorating lung injury in VILI. We also detected the expression levels of the principal oxidative stress pathway Nrf2 / HO-1 and the key inammatory transcription factor NF-κB. We used the HO-1 inhibitor ZnPP to further conrm our conclusion. Our results suggested for the rst time that Maresin1 could promote the activation of the Nrf2 / HO-1 pathway and signicantly suppress the expression of NF-κB to exert a positive role in VILI.


Introduction
Mechanical ventilation is a vital means of life support for the patients who cannot rely on their spontaneous breathing function to meet the need of their normal life activities [1], such as the patients who are under general anesthesia and suffering from respiratory failure. However, while mechanical ventilation provides respiratory support and improves the oxygenation function of the body, mechanical ventilation itself could also cause or aggravate the damage of lung tissue, known as ventilator-induced lung injury (VILI). VILI could even evolve into acute respiratory distress syndrome (ARDS) and endanger the patient's life [2]. Though the protective lung ventilation strategies have made de nite progress, VILI still poses challenges for the clinical practice [3]. Therefore, the search for new and effective treatments for VILI is an important problem to be solved urgently [4].
The excessive stretch stimulation of large tidal volume mechanical ventilation on the alveolar wall may cause obvious pathological damage including the alveolar-capillary exudation, the neutrophil in ltration, the alveolar hemorrhage, and the alveolar wall thickening, etc. [5] Many studies suggested that in addition to the mechanical damage to VILI, the oxidative stress and the excessive in ammatory response could also signi cantly increase the severity of VILI [6]. Therefore, controlling the excessive oxidative stress response and the in ammatory responses provide targets for the prevention and treatment of VILI.
Nuclear factor erythroid 2 related factor2 (Nrf2) could promote the transcription of heme oxygenase-1 (HO-1), the indirect metabolite of which has a strong antioxidant capacity, so Nrf-2 / HO-1 plays an important role in cell adaptation and antioxidant stress [7]. HO-1 is the rate-limiting enzyme in the heme metabolism pathway in the human body. As a protective protein, HO-1 has numerous biological activities such as anti-oxidation, anti-in ammatory, anti-tracheal epithelial cell proliferation, anti-apoptosis, and regulation of microcirculation. HO-1 could negatively regulate the in ammatory response and the oxidative stress response in acute lung injury [8].
NF-κB, as a target for many in ammatory diseases, is regulated by multiple signals. ROS, a vital production of oxidative stress, is considered to be one of the important regulators of NF-κB activation and plays an important role in the occurrence and development of in ammation [9]. ROS is necessary for the objects to play a role in eliminating bacteria, but overexpressed ROS could induce in ammation and results in tissue damage. Studies have shown that the activation of Nrf2 / HO-1 could inhibit the production of ROS induced by lipopolysaccharides (LPS) and reduce the expression of in ammatory mediators to play a protective role in improving lung injury [10,11].
Specialized pro-resolving mediators(SPM) such as maresins and resolvins, which are derived from polyunsaturated fatty acids, play an important role in controlling in ammation and oxidative stress [12,13]. As a novel SPM [14], the effect of Maresin1 on improving lung injury through multiple mechanisms has been con rmed [15,16]. There was also a recent investigation shown that MaR1 could inhibit the oxidative stress in lung IRI [17]. However, whether MaR1 could inhibit the oxidative stress and the in ammation response to exert a protective role in VILI has not been reported.
Our previous studies have shown that Resolvin D1 could inhibit the expression of HMGB-1 via the Nrf2 / HO-1 pathway and thus play a positive role in VILI [18]. However, whether Maresin1 could reduce the oxidative stress response and the in ammatory response through the Nrf2 / HO-1 pathway and NF-κB have not been con rmed. Therefore, in this experiment, we chose the VILI model to verify whether Maresin1 could inhibit the oxidative stress response and the in ammatory response in VILI and whether its related effects are exerted through the Nrf2 / HO-1 and the NF-κB pathways.

Animals
The 6-8 weeks old, weighing 20-22 g, SPF male C57BL/6 mice were purchased from the Experimental Animal Center of Wuhan University. The mice were randomly divided into 3 per cage and with free access to food and water. All the mice were housed in the animal experiment center of the SPF environment with the humidity of 60-65%, the light-dark cycles 12/12h, and the temperature of 22-24°C. All the mice were acclimatized for one week before the start of the experiment. The experiment was performed in accordance with the Chinese Animal Research Guidelines and was approved by the Laboratory Animal Management Committee of Tongji Medical College of Huazhong University of Science and Technology.

Experimental design
The construction method of the VILI model is the same as that described in our previous literature [18]. All experiments were repeated no less than three times. After that, we used Znpp, the speci c HO-1 inhibitor. The mice were randomly divided into the following six groups(n=7). (1) The Sham group(2) The MaR1 group (3) The ZnPP intervention control group the Znpp group : tracheal intubation was performed, but no mechanical ventilation was performed. The HO-1 antagonist ZnPP (10 mg/kg) was intraperitoneally injected and 200 μl of saline was injected into the tail vein 30 minutes before the tracheal intubation. (4) The HVT group (5) The MaR1 + HVT group (6) The HO-1 antagonist zinc protoporphyrin + Maresin1 group (the ZnPP + Maresin1 group): ZnPP (10 mg/kg) was intraperitoneally injected 30 minutes before the start of the tracheal intubation and the rest of the treatment was consistent with the MaR1 + HVT group. The mice in the sham group, the MaR1 group, the HVT group, and the MaR1 + HVT group were intraperitoneally injected additional saline(10 mg/kg) 30 minutes before starting the experiment. The respiratory rate of each group of mechanical ventilation is 80 times/min, the mechanical ventilation time is 4h, FiO2=2l%, I:E=1:2, PEEP=0 mmHg.

Histological analysis of lung tissues
The right upper lobe of the lung tissue was harvested for the pathological analysis. The lung tissue was made into slices about approximated 4 μm thick after being xed in 4% paraformaldehyde for 24h. The slices were stained with the hematoxylin and the eosin. Then the sections were observed under the ordinary optical microscope. The pathological injury score of the lung tissue was performed by using the lung tissue injury scoring criteria published by the American Thoracic Society [19].

Lung wet-to-dry weight ratio
The middle lobe of the right lung tissue from each group was collected to calculate the lung wet-to-dry weight ratio. After cleaning the surface of the fresh lung tissue with absorbent paper, weigh the initial weight of the lung tissue. Then put the lung tissue into the oven until the weight of the lung tissue becomes a constant weight. The weight of the lung tissue was weighed again. Calculate the lung wet-todry weight ratio and assess the extent of the lung tissue edema.

Analysis of arterial blood gas
At the end of the experiment, the blood (heparin anticoagulant) of the left carotid artery was collected and the arterial partial pressure of oxygen was immediately measured by a blood gas analyzer (Radiometer, Copenhagen, Denmark).

Measurement of the markers of the oxidative stress in lung tissues
According to the manufacturer's instructions, the malondialdehyde (MDA), superoxide dismutase (SOD), catalase (CAT), glutathione (GSH), myeloperoxidase(MPO) and 15-F2t-isoprostane were detected by their corresponding commercial assay kits (Jiancheng Bioengineering Institute, Nanjing, China). The activities of these indicators were assessed by a spectrophotometer

Bronchial alveolar lavage uid analysis
After the ligation of the right lung, 0.5 mL phosphate buffer saline (PBS) was utilized to lavage the left lung to obtain the BALF. The lavage process was repeated three times, and the recovery rate of more than 80% was considered effective.
The number of cells in the BALF was counted by a haematocytometer.
The protein concentration of the BALF was measured by using the BCA Protein Assay kit (Thermo Fisher Scienti c).
According to the manufacturer's instructions, the corresponding murine ELISA kits (Dakewe Bioengineering Co., Ltd., Shenzhen, China) were used to evaluate the levels of the related in ammatory mediators (IL-6, TNF-α, IL-β) in the BALF.

Western blotting analysis
According to the manufacturer's protocol, the protein of the lung tissue was extracted by using the protein extraction reagent kit (KeyGEN BioTECH, Nanjing, China) and the nucleoprotein was extracted using the Nuclear and Cytoplasmic Protein Extraction Kit(Beyotime, China). The protein concentration was measured with the BCA Protein Assay kit(Thermo Fisher Scienti c). The western blotting analysis was performed as mentioned in our previous research [20]. The membrane was rst incubated with the primary antibody: Nrf2 (Cell Signaling Technology, Danvers, MA, USA), HO-1(Cell Signaling Technology, Danvers, MA, USA), NF-κB p65(Cell Signaling Technology, Danvers, MA, USA), Histone 3 (Cell Signaling Technology, Danvers, MA, USA), GAPDH Antgene, China at 4℃ overnight. Then the membrane was incubated with the corresponding secondary antibody (Cell Signaling Technology, Danvers, MA, USA). Finally, the enhanced chemiluminescence reagent (Beyotime Institute of Biotechnology, Shanghai, China) and the UVP 160 imaging system (Upland, CA, USA)was used for exposure imaging. The grayscale value of the images was analyzed using the Image J software.

Statistical Analysis.
The whole data were expressed as means±SEM. The processes of the statistical analysis process and the mapping were performed by the Graphpad Prism 6.0 software. One-way analysis of variance (ANOVA) was used for comparison among groups, and the differences between groups were compared using Newman-Keuls. P < 0.05 was considered to be statistically signi cant.

Marsin1 attenuated ventilator-induced lung injury and signi cantly improves pulmonary function.
The pathological changes of the lung tissue were assessed by HE staining (Figure 1A-D). The content of the lung injury in the sham group and the MaR1 group didn't show a signi cant difference, which indicated Maresin1 itself does not have obvious damage to the lung tissue. Excessive mechanical ventilation could cause evident lung injury and Maresin1 could attenuate the histological lung injury. The pathological injury score( Figure 1E) was in line with the results observed by He staining.
The detection results of the arterial partial pressure of oxygen ( Figure 1F), the W/D ratio( Figure 1G), and the protein concentration in BALF( Figure 1H) were consistent with the results of the pathological changes. Maresin1 itself had no signi cant effect on the partial arterial oxygen pressure, the W/D ratio, and the protein concentration in BALF. The PaO2 of the HVT group was signi cantly lower than the sham group(P < 0.01) while the W/D ratio and the protein concentration in BALF were signi cantly higher than the normal mice(P < 0.01), indicating that excessive mechanical ventilation could contribute to the lung injury. Compared with the HVT group, the PaO2 in the MaR1 + HVT group was signi cantly increased(P < 0.01), while the W/D ratio and the protein were signi cantly reduced(P < 0.01), suggesting that Maresin1 could reduce lung injury and improve lung oxygenation function.

Maresin inhibited the excessive in ammatory response in VILI
The excessive in ammation is the main cause that leads to the occurrence of the VILI. Our results Figure  2 showed that the cell count in BALF Figure 2A , the MPO activity Figure 2B and the levels of proin ammatory factors Figure 2C-E in the HVT group were signi cantly increased compared to the normal mice(P < 0.01), while Maresin1 signi cantly reduced the levels of pro-in ammatory factors, the cell count, and the MPO activity. These results suggested that excessive mechanical ventilation could cause a signi cant in ammatory response that leads to lung injury and Maresin1 had a signi cant role in antiin ammatory.

Maresin1 inhibited the oxidative stress response in VILI
Oxidative stress plays an indispensable role in the initiation and progression of the VILI. As shown in Figure 3A-F, compared with the sham group, the HVT group showed low expression levels in GSH-PX and CAT but a higher level of the ROS, MDA, SOD, and15-F2t-isoprostane(P < 0.01). These tests revealed that excessive oxidative stress was indeed involved in VILI. Compared with the HVT group, the GSH-PX and CAT levels of the MaR1 + HVT group were higher(P < 0.01), and the ROS, MDA, SOD, and15-F2tisoprostane levels were lower(P < 0.01), clearing that Maresin could signi cantly inhibit the oxidative stress response.

ZnPP inhibited the protective effect of Maresin1 on lung tissue in VILI
As shown in Figure 5, there were no obvious pathological changes that occurred in the sham group, the MaR1 group, and the ZnPP group, indicating that Maresin1 and ZnPP do not cause lung injury by themselves. Pathological sections of lung tissue in the HVT group showed noticeable damage to lung tissue compared with the sham group(P < 0.01). The pathological changes in the lung caused by HVT became not obvious after the treatment of Maresin1. However, the histological changes in the lungs of the ZnPP + Maresin1 group were obviously more severe than those in the MaR1 + HVT group(P < 0.01). The lung injury score accorded with the pathohistological change. These results indicated that Maresin1 decreased the pathological damage of lung tissue caused by mechanical ventilation in part by increasing the expression of HO-1 3.6 ZnPP weakened the inhibition effect of Maresin1 on the oxidative stress and in ammatory response.
As shown in Figure 6, the MPO, the IL-1β, the TNF-α, the ROS, the MDA, and the 15-F2t-isoprostane in the HVT group were signi cantly higher than those in the sham group (P < 0.01). Maresin1 could inhibit the expression of the factors referred to above(P < 0.01). When the expression of HO-1 was inhibited by ZnPP, the MPO, the IL-1β, the TNF-α, the ROS, the MDA, and the 15-F2t-isoprostane showed a signi cant increase compared with the MaR1 + HVT group(P < 0.01). These phenomena indicated that ZnPP could weaken the inhibitory effect of Maresin1 on the in ammation response and the oxidative stress response.

Discussion
As a common clinical treatment method, mechanical ventilation was widely used in clinical critical illness such as acute lung injury (ALI) and acute respiratory distress syndrome (ARDS). However, the current reports suggested that the prevalence of VILI in surgical patients with normal lung function is between 6.2% and 24% [21,22]. For patients with pulmonary dysfunction, the incidence of VILI is higher. Various known pathophysiological factors including inappropriate in ammatory response, excessive destruction of airway barrier, alveolar edema, cell apoptosis, etc. could contribute to the occurrence of the VILI, but the precise pathological mechanism of VILI has not yet been elucidated.
Although the research on the prevention and treatment of VILI has made certain progress, such as small tidal volume ventilation, the application of related drugs and biological agents [23,24], the incidence of VILI remains remarkable and exploring the exact and effective treatment for VILI are still medical problems to be solved urgently.
There were studies have shown that Maresin 1 could ameliorate the lung ischemia/reperfusion injury by suppressing the oxidative stress via activating the Nrf-2-mediated HO-1 signaling pathway [17,25]. However, there is no research indicating whether Maresin1 could mitigate lung injury by suppressing the oxidative stress in VILI. So we constructed the VILI model to clarify the role of the Maresin1 exerts in VILI and explore and the possible mechanism and treatment methods of VILI. Our previous research showed that Maresin1 could inhibit the excessive in ammation caused by sepsis in mice [20,26]in a dose-dependent manner, the 1ng dose of Maresin1 can signi cantly protect mice and is a very suitable dose choice. There was also related study shown that 1ng of Maresin1 can effectively inhibit the oxidative stress response in lung tissue [17]. What's more, we chose two doses of Maresin1 of 0.1ng and 1ng for pre-experiment. The pathological results of the lung showed that the protective effect of 1 ng of Maresin1 on VILI was more stable and effective, so we chose 1 ng of Maresin for research.
We rst veri ed the protective effect of Maresin1 in VILI. The results of the pathological section showed that Maresin1 could signi cantly ameliorate the extent of the pathological damage of lung tissue in VILI.
Then we used the number of the cells in BALF to re ect the overall in ammation level, the protein concentration in BALF to assess the vascular permeability, and the changes in lung tissue W / D ratio to measure the severity of the pulmonary edema. The large exudation of capillary uid and protein is the main cause of pulmonary edema [27]. Our results showed that Maresin1 could mitigate the pathological damage of lung tissue, reduce the permeability of microvascular, and extenuate the degree of pulmonary edema in VILI. In addition, Maresin1 could dramatically improve lung oxygenation function.
Considering the important role of in ammatory response in VILI, we tested the relevant in ammatory indicators. The in ammatory mediators closely related to VILI are mainly MPO, TNF-α, IL-1β, IL-6, etc [28]. MPO is a speci c marker of PMN [19]. Our results showed that mice could develop severe in ammation after hyperventilation. The VILI mice showed a huge increase in related in ammatory indicators, such as cell counts in BALF and the levels of related in ammatory factors (eg IL-1β, IL-6, TNF-α). Maresin1 could markedly inhibit the in ammation response.
As we mentioned earlier, oxidative stress and in ammation are closely related, both played a vital role in the start and development of VILI. So we also explored the oxidative stress response in VILI. When oxidative stress occurs, the imbalance between oxidation and antioxidants of the body could be re ected by the measurement of the levels of a number of related indicators. Superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GSH-PX) could prevent lung injury from ROS exposure [29].
MDA could re ect the degree of lipid peroxidation in the body, 15 F 2t isoprostane is a highly speci c and sensitive indicator of lipid peroxidation products in the body. Therefore, we tested the above indicators in lung tissue, and the results showed that mechanical ventilation could signi cantly worsen the oxidative stress response, and Maresin1 could inhibit the oxidative stress response.
We next studied the classical pathway of oxidative stress Nrf2 / HO-1 and the important in ammation response transcription factor NF-κB. Under normal physiological conditions, Nrf2 combined with keap1 and localized in the cytoplasm. When the oxidative stress response occurs, Nrf2 dissociates from the conjugate, translocates into the nucleus, and induces HO-1 transcription, thereby exerting antioxidant effects [30]. Uncontrolled oxidative stress response could also contribute to the activation of the NF-κB [31]. Under normal physiological conditions, the p65 subunit of NF-κB in the cytoplasm combines with IκBα to form the NF-κB -IκBα complex [32]. When cells are stimulated by oxidative stress, IκBα is phosphorylated and bound to ubiquitin to degrade. The degradation of IκBα results in that NF-κB is activated to bind to speci c DNA sequences in the nucleus, thereby participating in the process of the in ammation responses [33]. So we examined the protein expression of HO-1 and the expression of Nrf2 and NF-κB in the nucleus after the Maresin1 treatment. The results showed that Maresin1 could promote the Nrf2 to translocate into the nucleus and signi cantly inhibit the translocation of NF-κB. To con rm our ndings, we further used the HO-1 antagonist ZnPP. After the treatment of ZnPP, the expression levels of HO-1 and Nrf2 were drastically reduced, while the expression of NF-ΚB was signi cantly increased, indicating that Maresin1could indeed inhibit excessive oxidative stress to improve lung injury in VILI by activating Nrf2 /HO-1pathway.
To further verify that Maresin1 does exert antioxidant and anti-in ammatory effects through the Nrf2 / HO-1 pathway and NF-κB, we continued to use ZnPP to validate our hypothesis. As expected, when ZnPP was used to inhibit the expression of HO-1, the degree of pathological damage in the lung tissue was signi cantly increased, and the blood oxygen partial pressure was considerably reduced. At the same time, we re-examined the relevant indicators of the oxidative stress response and the in ammation response. The results showed that when HO-1 was inhibited, the oxidative stress response and the in ammation response were signi cantly worsened. All these results indicated that Maresin1 signi cantly inhibited the production of pro-in ammatory factors via the suppression of the NF-κB and exerted the antioxidative effects via promoting the activation of the HO-1/Nrf2 pathway.

Conclusion
Our study showed that Maresin1 could reduce the oxidative stress response and the in ammation response through the Nrf-2 / HO-1 and the NF-κB signaling pathway, thereby signi cantly improving VILI.   Maresin1 inhibited the excessive in ammation response in VILI. A: The cell count in BALF B: The MPO activity C: IL-1β D: IL-6 E: TNF-α. Data was presented as means ± SEM, n = 7. **P < 0.01 versus the sham group; ##P < 0.01 versus the HVT group.  Maresin1 activated the Nrf2/HO-1 pathway and suppressed the NF-κB in VILI. A and E are the expression level of Nrf2, Histone 3, HO-1, GAPDH, and NF-κB p65 proteins of the lung tissues. B-D and F-H are the statistical tables corresponding to A and E respectively. Data was presented as means ± SEM, n = 7. *P < 0.05 versus the sham group; **P < 0.01 versus the sham group; #P < 0.05 versus the CLP group; ##P < 0.01 versus the CLP group;&& P < 0.01 versus the MaR1 + HVT group.  ZnPP weakened the inhibition effect of Maresin1 on the in ammation response and the oxidative stress response. A: MPO activity B: IL-1β C: TNF-α D: ROS E: MDA F: 15-F2t-isoprostane. Data was presented as means ± SEM, n = 7. **P < 0.01 versus the sham group; ##P < 0.01 versus the HVT group;&& P < 0.01 versus the MaR1 + HVT group.