MG53 attenuates nitrogen mustard‐induced acute lung injury

Abstract Nitrogen mustard (NM) is an alkylating vesicant that causes severe pulmonary injury. Currently, there are no effective means to counteract vesicant‐induced lung injury. MG53 is a vital component of cell membrane repair and lung protection. Here, we show that mice with ablation of MG53 are more susceptible to NM‐induced lung injury than the wild‐type mice. Treatment of wild‐type mice with exogenous recombinant human MG53 (rhMG53) protein ameliorates NM‐induced lung injury by restoring arterial blood oxygen level, by improving dynamic lung compliance and by reducing airway resistance. Exposure of lung epithelial and endothelial cells to NM leads to intracellular oxidative stress that compromises the intrinsic cell membrane repair function of MG53. Exogenous rhMG53 protein applied to the culture medium protects lung epithelial and endothelial cells from NM‐induced membrane injury and oxidative stress, and enhances survival of the cells. Additionally, we show that loss of MG53 leads to increased vulnerability of macrophages to vesicant‐induced cell death. Overall, these findings support the therapeutic potential of rhMG53 to counteract vesicant‐induced lung injury.

to defective membrane repair. 22 We have shown that the recombinant human MG53 protein (rhMG53) when administered either intravenously (IV) or via aerosol has the ability to effectively mitigate lung ischaemia-reperfusion injury, lipopolysaccharide-induced inflammation and porcine pancreatic elastase (PPE)-induced emphysema in rodents and pigs. 23 In addition to membrane repair, recent studies demonstrate an anti-inflammatory function of MG53 in dampening NF-kB signalling, and knockdown of MG53 leads to hyper-inflammation in human macrophages. 24,25 Herein, we provide data to support the physiologic function of MG53 in lung protection and the potential therapeutic value of rhMG53 to treat vesicant-induced lung injury. When compared to wild-type littermates, mg53 −/− mice display an exacerbated lung injury with a more severe lung dysfunction following exposure to NM.
Additionally, we show that intravenous administration of rhMG53 to wild-type mice after exposure to NM can mitigate the adverse effects of NM-vesicant lung injury.

| Animals and animal treatment
All animal care and usage were done in accordance with federal policies and guidelines and approved by the Ohio State University's IACUC. mg53 −/− mice, tPA-MG53 and their wild-type littermates were bred and generated as previously described. 20,27 Mice were anaesthetized by intraperitoneal (IP) injection of ketamine and xylazine (80 mg/kg:10 mg/kg, respectively) and then were treated with PBS or NM (0.125 mg/kg) by intratracheal instillation following the protocol as described previously with minor modification. 2 NM was prepared immediately before administration. All procedures were performed in a designated room with chemical hood strictly following OSU Environment and Health and Safety guidelines. rhMG53 (2 mg/kg) in saline or saline alone was administered by daily IV injection right after NM treatment for five days.

| Cells, cell culture and stress treatment
Human bronchial epithelial cells (B2B) and THP-1 cells were purchased from the American Type Culture Collection (ATCC). The B2B and THP-1 cells were grown in RPMI 1640 medium supplemented with 10% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin at 37°C in the presence of 5% CO 2 . The sh-MG53-knockdown THP-1 cells were created and cultured as described previously. 24 Primary porcine aortic endothelial cells (PAoEC) were isolated as previously described 28 and cultured in MEM containing 10% FBS, 2 mM glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin (Lonza) with bovine brain extract at 37°C in the presence of 5% CO 2 .

| Apoptosis assay and ROS measurement
Cell apoptosis was investigated by dual staining with Alexa Fluor 488 annexin V and propidium iodide (PI) (Invitrogen Cat# V13241) following the manufacture protocol. Briefly, PAoEC and B2B cells were seeded in 6-well plates, cultured for 24 h and then incubated with 10 µM (for PAoEC)/20 µM (for B2B cells) NM or BSA control for 4 h, washed and incubated with BSA or rhMG53 (10 µg/ml) for another 20 h. Cells were detached by 0.25% trypsin-EDTA solution and washed with PBS for 1 time, and Annexin V and PI staining were performed for FACS analysis and analysed as described previously. 29 Cellular-reactive oxygen species (ROS) production was measured using a ROS Detection Assay Kit (Abcam Cat#ab113851) according to manufacture instructions. Briefly, cells were seeded in 6-well plates, cultured for 24 h and then incubated with 10 µM (for PAoEC)/20 µM (for B2B cells) NM or BSA control for 4 h, and washed with PBS, incubated with BSA or rhMG53 (10 µg/ml) and cultured for another 20 h. Cells were washed, stained and analysed with DCF staining. The intensity of red fluorescence was detected by Guava EasyCyte™ System, and images were taken by confocal microscope.

| Lung function measurement
Mice were anaesthetized by intraperitoneal (IP) injection of diazepam (17.5 mg/kg) followed by ketamine (80 mg/kg), and lung mechanics were assessed as previous described. 30 Briefly, mouse was mechanically ventilated on a computer-controlled flexiVent FX piston ventilator (SciReq; Montreal, Canada), with a tidal volume of 10 ml/kg at a frequency of 200 breaths/minute, against 2-3 cmH 2 O PEEP, as in our previous studies. Following two total lung capacity manoeuvres to standardize volume history, basal airway resistance and dynamic lung compliance were measured by the forced oscillation technique. 31

| Antibodies and western blotting
Primary antibodies used in this study are as follows: anti-cleaved caspase-3 (Cell Signaling Technologies) and anti-GAPDH (Santa Cruz Biotechnology). Total protein extractions were prepared and subjected to immunoblot analysis as described previously. 32,33 Briefly, after blocking, membranes were incubated with relevant antibodies and probed with corresponding HRP-conjugated secondary antibodies (Cell Signaling Technologies). All films were developed with ECL-Plus regents (GE healthcare) and imaged using ChemiDoc TM Gel Imaging System (Bio-Rad).

| Cell membrane injury assay and confocal microscopy
For membrane repair assay, B2B cells were transfected with GFP-MG53 and then subjected to microelectrode penetrationinduced acute cell membrane injury, and the data were analysed as previously described. 20

| Histology and immunofluorescent staining
Histology and immunofluorescent staining were performed as previously described. 32,33 Briefly, tissues were dissected from experimental animals and then fixed in 4% paraformaldehyde (PFA) overnight at 4°C. After fixing, samples were washed three times for 5 min with 70% ethanol. Washed samples were processed, embedded in paraffin. 4 μm thick paraffin sections were cut. Cells were fixed with 4% PFA.

| Statistical analysis
All data are expressed as means ± standard error of mean (SEM).
Statistical evaluation was conducted using the student's t test and by ANOVA for repeated measures. A value of p < 0.05 was considered statistically significant.

| Mg53 −/− mice are more susceptible to NMinduced lung injury
To understand the physiological role of MG53 in protection against vesicant-induced lung injury, we administered NM (0.125 mg/kg) intratracheally (IT) to mg53 −/− mice and wild-type littermate controls, according to the protocol developed by Laskin and colleagues. 2 At 5 days post-NM exposure, more severe lung damage was observed in mg53 −/− mice than in wild-type mice ( Figure 1A). While carotid arterial oxygen saturations (SaO 2 ) only showed a marginal difference between wild-type and mg53 −/− mice ( Figure 1B), significant elevations of airway resistance (R rs , Figure 1C) and compromised lung dynamic compliance (C dyn , Figure 1D) were observed in the mg53 −/− mice.
Histological analysis revealed massive parenchymal necrosis with increased infiltration of immune cells in the mg53 −/− lung ( Figure 1E).
These observations are consistent with our recent findings that mg53 −/− mice experience worsened morbidity and delayed recovery compared to wild-type mice in a non-lethal infection with influenza virus, correlating with increased inflammatory pathology in the lungs of the mg53 −/− mice. 24 Together, these data demonstrate that genetic ablation of MG53 renders the mice more susceptible to vesicantinduced lung dysfunction and inflammation.

| rhMG53 mitigates NM-induced lung injury in mice
We recently reported that recombinant human MG53 protein (rhMG53) protects mice from lethal influenza virus infection and could function as a therapeutic to treat inflammation-driven infectious diseases. 25 To determine whether rhMG53 has similar protective effects against chemical lung injury, C57BL/6J mice (10 weeks age) were exposed to NM (0.125 mg/kg, IT) to induce lung injury.
Mice were divided into two groups, one receiving tail vein administration of 2 mg/kg rhMG53 protein right after NM exposure (and then daily thereafter), and the others receiving saline as control at the same frequency with analysis conducted in a blinded manner.
Mice receiving saline showed progressive decline of body weight, which was mitigated by rhMG53 treatment during the 5-day observation period (Figure 2A, left). Out of the 4 mice in the control group, one died on day 2, while all rhMG53-treated mice survived the 5 days observation period.
We have generated a transgenic mouse model with sus- At the gross level, lung injury appeared less severe in rhMG53treated mice compared with the control group 5 days after NM exposure ( Figure 2B). Even with the limited number of animals, we observed that rhMG53 treatment led to improved oxygen saturation (SaO 2 , Figure 2C) and trended towards increased dynamic lung compliance (C dyn , Figure 2D) and reduced airway resistance (R rs , Figure 2E).

| NM exposure impairs MG53-mediated membrane repair in lung epithelial cells
To understand the mechanism that underlies MG53's role in protection against NM-induced lung injury, we conducted in vitro membrane repair assay as described previously. 20 The repair patch remained stable over the 2 min observation period in control cells not exposed to NM ( Figure 3B, green). Remarkably, B2B cells treated with NM displayed dysfunctional GFP-MG53 movement following microelectrode induced membrane injury ( Figure 3B, red). While there was an initial small accumulation of GFP-MG53 at the injury site within the first 30 s after injury, the repair patch did not remain stable as this decreased within the 2 min recording period.
Using FM1-43 dye entry as a direct measure of cell membrane integrity, 20,26 we found that exposure of B2B cells to NM caused more entry of FM1-43 dye, which could be reduced by treatment with rhMG53 (5 µg/ml) ( Figure 3C,D). These findings provide direct evidence that vesicant exposure of lung epithelial cells leads to disruption of cell membrane repair machinery.

| rhMG53 protects lung epithelial as well as endothelial cells against NM-induced injury
Studies show that SM/NM exposure leads to elevation of cellular oxidative stress. 6,10,11,35,36 We have demonstrated before that MG53's membrane repair function is impaired when cells are subjected to chronic oxidative stress. 37,38 We next conducted a series of studies to quantify the potential protective role of rhMG53 in mitigating NM-induced oxidative stress and injury to B2B cells. We Our laboratory has developed a protocol to isolate primary porcine aortic endothelial cells (PAoEC), which are widely used as an in vitro system to evaluate endothelial cell function (of either aortic F I G U R E 1 Knockout of MG53 enhances NM-induced lung injury. Mice were treated by intratracheal instillation of either 50 µl NM (0.125 mg/kg in saline) or 50 µl saline alone. After 5 days of exposure, the lung function was examined, and mice were euthanized for lung tissue section preparation. (A) Representative images of wild-type (upper panels) and mg53 −/− (low panels) lung subjected NM exposure. Quantification of the lung function with blood oxygen saturation level (SaO 2 ) (B), and resistance (Rrs) (C) and dynamic compliance (C dyn ) (D). (E) Histological analyses revealed massive necrosis with increased infiltration of immune cells in the mg53 −/− lung (right panel). These data demonstrate that ablation of MG53 renders the mice more susceptible to NM-induced lung injury. *p < 0.05 for the indicated group or pulmonary tissues). 28 Similar findings were also observed with PAoEC cells, where rhMG53 treatment ameliorated NM-induced ROS generation ( Figure 4D). To further quantify the degree of NMinduced cell death, we conducted FACS analysis of PAoEC stained with propidium iodide (PI) and annexin V ( Figure 5A). PAoEC cells that were annexin V-positive and PI-negative were defined as undergoing apoptotic cell death. As shown in Figure 5B, rhMG53 treatment improved survival of the PAoEC cells and reduced apoptotic cell death.
These findings provide a novel mechanism for NM-induced tissue injury that involves oxidative stress-mediated disruption of the cell membrane repair machinery. It also lays the foundation for the use of exogenous rhMG53 to boost the defence mechanism of these cells against vesicant-induced tissue injury.

| rhMG53 protects THP-1 cells against NMinduced injury
In addition to oxidative stress, vesicant exposure is known to cause sustained inflammation that contributes to the exacerbated and delayed lung injury. Thus, controlling inflammation is also important to combat vesicant-induced lung injury. [2][3][4][5]7,8,14,[16][17][18]39 We recently reported that MG53 has an anti-inflammatory role associated with tissue injury. 24 We found that THP-1 human macrophages express MG53 and that loss of MG53 leads to hyper-inflammation due to activation of NF-ĸB.
We transfected THP-1 cells with shRNA against MG53 to generate a MG53-knockdown cell line (sh-MG53) ( Figure 6). As shown in Figure 6A, sh-MG53 cells were more susceptible to NM-induced cell F I G U R E 2 rhMG53 mitigates NMinduced lung injury. Mice were treated by intratracheal instillation of either 50 µl NM (0.125 mg/kg in saline) or 50 µl saline alone. After exposure, the mice were weighted daily for up to 5 days postexposure (A). (B) Representative images (left panel) and HE staining (right panel) of saline control and rhMG53-treated (IV, 2 mg/kg) lungs subjected to NM exposure. After 5 days of exposure, the lung function was examined, and mice were euthanized for lung tissue section preparation. Quantification of the lung function with blood oxygen saturation level (SaO 2 ) (C), and resistance (Rrs) (D) and dynamic compliance (C dyn ) (E).

| DISCUSS ION
MG53 plays an essential role in cell membrane repair and pulmonary protection. 20,21 MG53 also has anti-inflammatory function associated with chronic injury, sepsis and viral infection. 24,25,40 In this study, we demonstrated that knockout of MG53 causes more severe lung injury and lung dysfunction following exposure to NM. Previously, we have demonstrated pulmonary pathology with the mg53 -/mice, which may be the underlying contributor to the increased susceptibility of the lung to NM-induced injury. 22  Targeting alveolar membrane injury repair may offer an effective means to rescue the damaged lung in the acute phase of injury and to prevent the progression into the prolonged injury phase.
Through live-cell imaging, we demonstrated that, in lung epithelial cells exposed with NM, GFP-MG53 becomes immobilized Accumulating evidence suggests that the cytotoxic mechanism associated with mustard exposure contributes to oxidative stress, which induces damage to the lung. 6,10,11,35,36 Previously, we have demonstrated that MG53's membrane repair function is impaired when cells are subjected to chronic oxidative stress. 37,38 The preservation of cell integrity and mitigation of oxidative stress can provide added benefits to pulmonary protection. Using cultured lung epithelial (B2B) and endothelial (PAoEC) cells, we found that rhMG53 treatment can improve survival of the cells and reduce ROS levels upon exposure to NM. These findings provide evidence that NM-induced tissue injury and oxidative stress can all be mitigated by MG53, laying the foundation for the use of exogenous rhMG53 to boost the defence mechanism of the lungs against vesicant-induced injury.
In addition to oxidative stress, mustard gas exposure is known to induce infiltration of inflammatory cells and cytokines release in the lung that contributes to the exacerbated and delayed lung injuries. 2,8,9,41 Macrophages are known to play a role in both acute and chronic pulmonary pathologies. An imbalance of macrophagereleased pro-inflammatory and anti-inflammatory cytokines will aggravate acute lung injury and promote the development of lung toxicity. 2,3,5,7,8,15,41 By flow cytometric analysis, we showed that THP-1 cells with knockdown of MG53 are more susceptible to NMinduced cell death, while rhMG53 treatment enhances cell survival after NM exposure. Preservation of macrophage integrity by MG53 can add to the defence mechanism of the lungs during the early phase of NM exposure. In addition to maintenance of macrophage integrity, we recently showed that repetitive administration F I G U R E 4 rhMG53 attenuates NM-induced oxidative stress in human bronchial epithelial (B2B) porcine aortic endothelial (PAoEC) cells. The B2B cells and primary porcine aortic endothelial (PAoEC) cells were cultured for 24 h and treated with 10 µM (for PAoEC)/20 µM (for B2B cells) NM or BSA control for 4 h, and washed, incubated with BSA or rhMG53 (10 µg/ml) and cultured for another 20 h. Cells were stained and with DCF staining for ROS measurement. In rodents and humans, MG53 is secreted by muscle cells and is present at low levels in blood in normal physiologic conditions. 26,42 Thus, a therapeutic approach that modulates endogenous MG53 levels/function or involves systemic administration of rhMG53 protein is potentially a safe biologic means to treat and prevent tissue damage, including vesicant-induced multi-organ injury. We have performed toxicological studies in rodents and dogs, 33 which demonstrate that rhMG53 has broad safety, underscoring its promise as a potential therapeutic to treat multi-organ injuries. Future studies testing the safety and efficacy of rhMG53 in large animal models of vesicant-induced pulmonary injury represent an essential component for our effort in translating the basic findings with MG53 into human applications.