Transplantation of Viable Allogeneic Mitochondria Improves Gas Exchange and Alveolar-Capillary Permeability in Rats with Endotoxin-Induced Acute Lung Injury


 Background: Acute lung injuries (ALI) cause disruption of the alveolar-capillary barrier and remains as the leading cause of death in critically ill patients. This study tested the hypothesis that the administration of freshly isolated viable allogeneic mitochondria can prevent injury of the alveolar-capillary barrier at the endothelial level, as mitochondrial dysfunction of the pulmonary endothelium is a critical aspect of ALI progression.Methods: ALI was induced by intratracheal lipopolysaccharide instillation (LPS, 1mg/kg) in anesthetized rats. Mitochondria (100 μg) were isolated from the freshly harvested soleus muscles of naïve rats and stained with a green fluorescence MitoTracker™ dyne. A mitochondria or placebo solution was randomly administered into the jugular veins of the rats at 2 h and 4 h after ALI induction. An arterial blood gas analysis was done 20 h later. The animals were then sacrificed, and lung tissues were harvested for analysis.Results: An IVIS Spectrum imaging system was used to obtain ex vivo heart-lung block images and demonstrated the enhancement of MitoTracker™ fluorescence in the lungs. Mitochondria transplantation significantly improved arterial oxygen contents (PaO2 and SaO2), while reduced CO2 tension in rats with ALI. Animals with mitochondrial transplants had significantly increased ATP concentration in lung tissues. Allogeneic mitochondria transplantation preserved alveolar-capillary barrier function, as shown by a reduction in protein levels in the bronchoalveolar lavage fluid, and decrease of extravasated Evans blue dyne and hemoglobin content in lung tissues. Relaxation responses to acetylcholine and eNOS expression were potentiated in injured pulmonary arteries, and inflammatory cells infiltration into lung tissue was also reduced following mitochondrial transplantation.Conclusions: Transplantation of viable mitochondria protects the integrity of endothelial lining of the alveolar-capillary barrier, thereby improving gas exchange during the acute stages of endotoxin-induced ALI. However, the longer-term effects of mitochondrial transplantation on pulmonary function recovery after ALI requires further investigation.


Introduction
Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) represent the most common causes of intensive care unit (ICU) admission, need for ventilatory support, and death in critically ill patients with a high mortality rate of up to 35-45% 1 . In severe ALI, intense pro-in ammatory responses and leukocyte in ltration into lung tissue destroy the integrity of the alveolar epithelial and pulmonary artery endothelial layers, leading to the breakdown of the alveolar-epithelial barrier. This in turn, promotes the formation of protein-rich pulmonary edema and impairs blood-gas exchange [2][3][4] . Since pulmonary vascular endothelium is the key modulator of ALI progression, pharmaceutic-and cell-based endothelialtargeting therapies have become emerging candidate therapies for ALI/ARDS in the past two decades 5,6 . Under normal conditions, the pulmonary vascular endothelium inhibits the majority of in ammatory and coagulation responses in the lung; but in ALI, these endothelial cells are activated and contribute to the pro-in ammatory and pro-coagulant responses 7 . Mitochondria serve as the most important sensors in regulating intra-and inter-cellular communications for the maintenance of healthy cellular homeostasis.
In order to maintain homeostasis, mitochondria are also extremely sensitive to environmental stressors 8 .
When the excessive physical stress is generated and the cellular environment becomes toxic or allostatic, defects in mitochondrial electron transport, intracellular Ca 2+ homeostasis, ion transport, mitochondrial biogenesis, cellular metabolism, redox regulation, and mitochondrial quality control result in mitochondrial dysfunction 8,9 . Recent evidence suggests that impaired mitochondrial homeostasis in pulmonary epithelial and vascular endothelial cells contributes to the progress of ALI 10,11 , and the loss of mitochondrial genomic integrity is the crucial initial step that drives lung injury propagation during ALI 12 .
Since viable exogeneous mitochondria transplantation has been shown therapeutic potential in various pre-clinical ischemia-reperfusion injury models [13][14][15] and in pediatric patients with myocardial dysfunction 16 , we conducted the rst pre-clinical study that tested the hypothesis that the administration of freshly-isolated viable allogeneic mitochondria can reduce endothelial injury at the alveolar-capillary barrier and improve gas exchange in an experimental model of endotoxin-induced ALI.

Rat model of acute lung injury
All experimental procedure were performed in accordance with the ARRIVE Guideline and the guidelines for animal research issued by the E-Da Hospital and I-Shou University. The study was approved by the Institutional Animal Care and Use Committee of the E-Da Hospital, Kaohsiung, Taiwan (IACUC approval number 109024). Adult male Sprague-Dawley rats were anesthetized using intraperitoneal Zoletil cocktail anesthetic (30 mg/kg) injections. Following mini-open incision, a placebo solution (0.9% saline, 0.5 ml) or Escherichia coli lipopolysaccharide (LPS, 1 mg/kg in 0.5 ml of saline; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) was instilled through glottis with a catheter for sham operations and ALI induction, respectively 17 .

Isolation and administration of mitochondria
Mitochondria were freshly isolated from health naïve rats within 1h before delivery. Rats were asphyxiated with CO 2 and their bilateral soleus muscles were harvested. A commercially available mitochondria isolation kit (Thermo Scienti c, Waltham, MA) was used to isolate viable mitochondria 18 .
The freshly isolated soleus muscles were brie y washed with a phosphate buffer solution (PBS, containing 0.1 M sodium phosphate and 0.15 M sodium chloride at pH 7.2) then incubated with a tryptase solution (0.3 mg/ml) for 3 minutes. The muscles were homogenized using a grinder and the resulting tissue samples were centrifuged at 11,752 g for 3 minutes at 4ºC. After three cycles of suspension and centrifugation, the mitochondrial pellets were collected, and the cytosolic components were discarded using mitochondria isolation reagents (A to C solutions).

Allocation of Experimental Groups and Study Protocol
Animals were randomly allocated to sham (0.9% saline) or ALI group (LPS). Rats in the ALI group were then randomly re-assigned to either the control or mitochondrial group. The control group received intravenous PBS (200 µL) and the mitochondria group received skeletal muscle-derived mitochondria in PBS (100 µg in 200 µL PBS). Solutions in both groups were delivered via the jugular veins. Each animal in the control or mitochondrial group received the treatment solution at 2 h and 4 h after intratracheal instillation of LPS and they recovered from anesthesia under a warm blanket. A total of 11 rats received the sham operation, 23 rats received LPS treatment (including 10 controls and 13 mitochondrial transplantation), and 6 rats were served as naïve animals for tissue harvesting and ex vivo image scanning. The experimental animals were sacri ced under deep anesthesia 24 h after instillation of saline or LPS to collect blood samples and lung tissues for analysis.

Identi cation of the Distribution of Exogenous Mitochondria following systemic delivery
In some experiments, the isolated mitochondria were stained with a green MitoTracker™ uorescence dyne (catalog #M7514, Invitrogen™, Carlsbad, CA) for the ex vivo imaging under an in vivo imaging system (IVIS, Caliper IVIS Spectrum System, PerkinElmer, Waltham, MA). Heart and lungs were harvested en bloc at 30 min after intravenous delivery of MitoTracker uorescence-labelled mitochondria and scanned under the 3-dimensional tomography imaging system.

Artery blood gas analysis (ABG)
Arterial blood was drawn from the femoral artery of anesthetized rats under room air condition. Oxygen saturation (SaO 2 ), partial pressure of oxygen (PaO 2 ) and CO 2 (PaCO 2 ), and other ABG parameters in the arterial blood were analyzed using a blood gas analyzer (ABL 520, Radiometer, Copenhagen, Denmark).

Measurement of vascular reactivity
Pulmonary artery rings (approximately 2 mm long) were freshly isolated and mounted in organ chambers containing 25 ml of Krebs solution. The chambers were maintained at 37ºC and aerated continuously with 94% O 2 /6% CO 2 . Changes in isometric force were recorded continuously using an isometric forcedisplacement transducer (Grass IT30; Grass Instrument, West Warwick, RI). Each ring was gradually stretched to 2.0 g. After a 45-min equilibration period, the rings were induced to contract using KCl (40 mM) and slow phenylephrine titration (PE, 10 −9 to 10 −5 M). Concentration-response curves were charted by recording acetylcholine (10 −9 to 10 −5 M; Sigma-Aldrich, Burlington, MA) during contraction to a median PE effective concentration (EC50).
Assessment of pulmonary capillary permeability and lung wet-to-dry ratio The permeability of alveolar-capillary barrier was assessed by measuring protein content in the bronchoalveolar lavage uid (BALF), the Evans blue assay and lung wet-to-dry ratio (LWDR) 19 . The trachea in the lab animals were dissected and lungs were lavaged with 6 ml/kg chilled PBS. The recovered lavage uid was analyzed for total protein content using a protein assay kit (Bio-Rad Laboratories, Hercules, CA). Evans blue dye (2% v/v, 20 mg/kg, Sigma Aldrich) was administered via the jugular vein at 30 min before euthanization and lung tissue was perfused free of blood with normal saline. Evans blue was extracted from the freshly isolated lung tissues by incubating in 99.5% formamide (4 ml/g tissue; Sigma Aldrich) for 48h and the tissue concentrations were determined by a spectrophotometry at the wavelength of 620 nm. To assess lung water content, the left lung was excised and weighed immediately. Lung tissues were dried in an oven at 60°C for 24 h and reweighed. LWDR was obtained by dividing the mass of the initial specimen by the mass of the dried specimen.

Measurement of lung hemoglobin content and myeloperoxidase activity assay
The hemoglobin colorimetric assay was performed using a hemoglobin colorimetric assay kit (catalog #700540, Cayman Chemical, Ann Arbor, MI). Lungs were excised and homogenized in 5 ml PBS solutions (pH 7.4, with 0.16 mg/ml heparin, per gram weight of tissue). Homogenized lung tissues were centrifuged at 10,000 g for 15 minutes at 4°C. After the supernatant was removed, the tissue samples were added to assay wells and incubated at room temperature for 15 minutes. The absorbance at 560-590 nm was read and analyzed. The enzymatic activity of myeloperoxidase (MPO) in the homogenized lungs was measured by a commercially available MPO assay kit (Cell Biolabs Inc., San Diego, CA).

Measurement of Tissue adenosine triphosphate (ATP)
Tissue ATP concentrations were analyzed using an ATP assay kit (catalog #ab83355, Abcam, Cambridge, MA). The right lungs of lab animals were excised and homogenized in the ATP assay buffer. The homogenized tissue was centrifuged at 13,000 g for 5 minutes at 4°C. Insoluble components were removed and the solution was deproteinized using a sample preparation kit. The supernatant was then collected for ATP colorimetric assay.

Western Blot
Pulmonary artery trunk and lung tissues were minced and homogenized in lysis buffer. Equal amount of proteins (100 mg) were loaded into polyacrylamide gels and transferred to nitrocellulose membranes using the wet transferring method. The membranes were incubated overnight at 4°C with primary antibodies (endothelial nitric oxide synthase (eNOS), phosphorylated eNOS-S1177, CD11b and iNOS) in 1:1000 dilutions. All primary antibodies were purchased from BD Biosciences (New Jersey, USA), except CD11b was obtained from Abcam (Cambridge, MA). After washing with PBS, the membranes were incubated with horseradish peroxidase-linked secondary antibodies for 1 hour at room temperature.
Bands were visualized using enhanced chemiluminescence and quanti ed with scanning densitometry (the ImageJ; 1.48v, National Institutes of Health, Bethesda, Md).
Lung tissues were immersed in 10% buffered formal saline and xed for 24 h. Biopsies were then processed through increasing grades of alcohol and embedded in para n wax. Hematoxylin and eosin (H&E)-stained lung sections were examined under a light microscope and photographed. Lung injury was assessed by an investigator blinded to the treatment groups according to a scale of 0-2 for each of the following criteria: (1) in ltration of neutrophils in the alveolar and interstitial space, (2) amount of hyaline membranes and proteinaceous debris, (3) degree of lung hemorrhage and (4) extent of parenchymal consolidation 20 .

Statistical analysis
All data sets were tested for the normality assumption using the Shapiro-Wilk test before statistical procedures. The values of continuous variables were compared using the Mann-Whitney U test was used.
Two-way repeated measures (RM) ANOVA was used to compare differences in vasomotor reactivity between the control and mitochondrial transplantation groups at the different concentrations of vasoactive drugs. A Dunn's post hoc procedure was used for multiple comparisons. Results are presented as the median and interquartile range (IQR). Statistical signi cance was accepted at a level of P< 0.05. All of the statistical analyses were performed using the SigmaPlot 14.0 (Systat Software Inc., San Jose, CA).

Uptake of allogenic mitochondria following intravenous administration
Compared with naïve animals, MitoTracker™ Green uorescence densities were more signi cantly enhanced in the ex vivo lung lobes of rats with ALI that received skeletal muscle-derived allogeneic mitochondria transplantation ( Figure 1A), indicating that exogenous mitochondria in the pulmonary circulation are retained following intravenous administration. Tissue ATP concentration in the lung homogenates of animals with ALI was restored following allogeneic mitochondrial transplantation ( Figure 1B).  Table 1). Saturation of oxygen content (SaO 2 ) in arterial blood was also signi cantly improved in mitochondrial group (Table 1). Three different means were used to measure the integrity of the alveolarcapillary barrier in rats with ALI. Compared to shams, the LWDR was reduced in the lungs of ALI animals (Figure 2A), and transplantation of mitochondria did not change the lung weight ratio (Figure 2A). The BALF protein content and extravasated Evans blue dye content were signi cantly reduced in the lung tissues of mitochondrial-transplanted animals to levels similar to the sham-operated group ( Figures 2B  and 2C), suggesting improvement in pulmonary endothelial permeability following mitochondrial transplantation. Blood samples were collected via femoral artery and analyzed for ABG at 24 h after induction of acute lung injury (ALI); ALI+mito: rats with ALI received allogeneic mitochondrial transplantation; PaO 2 and PaCO 2 : partial pressures of oxygen and carbon dioxide in artery blood; SaO 2 : oxygen saturation in the artery blood. Data were analyzed using the Mann-Whitney U test and are presented as mean±SD. *P<0.005 compared with sham group and † P<0.05 compared with ALI+mito group.
Transplantation of mitochondria potentiates pulmonary vascular endothelial function Transplantation of mitochondria did not affect the contraction responses to KCl depolarization and α1adrenergic stimulation in the isolated pulmonary artery of rats with ALI ( Figures 3A and 3B). However, the maximal tension of relaxation induced by 10 −5 M acetylcholine was signi cantly restored in the mitochondrial transplanted group (55.9±12.0 vs 73.6±5.6%, ALI vs mitochondrial groups, respectively; P=0.006; Figure 3C). However, the concentration-response curves of endothelial-dependent relaxation of pulmonary arteries were superimposed in the two ALI groups ( Figure 3C). eNOS is one of the most important regulator of maintaining normal cellular function in the vascular endothelium. 21 Protein expressions of eNOS and its phosphorylated form (p-eNOS-S1177) in the pulmonary artery were suppressed in rats with ALI, but they were normalized following mitochondrial administration ( Figure 3D).

Transplantation of mitochondria attenuates damage of lung parenchyma
The severity of lung damage was assessed using a lung injury score, pulmonary hemorrhage, and tissue in ammation. Lung injury was precipitated by increased neutrophil in ltration and red blood cell accumulation in the alveolar and interstitial spaces at 24 h after intratracheal lipopolysaccharide instillation ( Figure 4A). Consolidation, alveolar ectasia and hyaline membrane formation were noted in the lung parenchyma of rats with ALI ( Figure 4A). Overall lung injury scores were signi cantly reduced in ALI animals received mitochondrial transplantation (1.57±0.11 vs 1.11±0.17, ALI vs ALI+mito, respectively; P=0.004 ( Figure 4B). Tissue hemoglobin content was reduced in animals with mitochondria transplants compared to placebo rats with ALI, which indicates pulmonary hemorrhage improvement in ALI ( Figure 4C). In ammatory cell in ltration was determined by the protein levels of iNOS and CD11b in the lung homogenates 17 . Mitochondrial transplantation signi cantly suppressed expressions of iNOS and CD11b in the lungs of endotoxin-induced ALI ( Figure 5A). However, no differences in the lung myeloperoxidase activity were found between ALI animals that received placebo or mitochondrial treatment ( Figure 5B).

Dr. Mc Cully and his colleagues reported the rst clinical application of mitochondrial transplantation in
pediatric patients with ischemia-reperfusion-associated myocardial dysfunction 16 . Mitochondria transplantation have since been considered as a revolutionary approach for regenerative medicine 22 . Mitochondrial transplantation has the advantage of rapid isolation and puri cation to meet the clinical needs 23 . Mitochondria are unlikely to induce alloreactivity and damage-associated molecular pattern molecules reaction. Therefore, mitochondria have potential for both syngeneic and allogeneic transplants 24 . Furthermore, exogenous mitochondria can be conveniently administrated through intravenous route 14,18 .
Intratracheal lipopolysaccharide instillation destroys lung parenchyma through the generation of proteases and reactive oxygen and nitrogen species produced by activated polymorphonuclear cells in the interstitial and alveolar compartments that is comparable to the pathophysiological changes in patients with ALI/ARDS 25 . Previous studies found that the intra-alveolar in ammatory reaction composed of a neutrophilic exudate at the initial 6 to 12 h, a monocytic exudate peaking at 24 h, and then followed by a lymphocytic exudate 26 . After lipopolysaccharide instillation, we detected hypoxemia and hypercapnia, indicating the impaired gas exchange due to development of alveolar dead space and ventilation-to-perfusion mismatch in these animals 27 . The alveolar-capillary barrier is organized by a network of collagen and laminin that separates the epithelium and endothelium 28 . The capillary endothelium is a semipermeable barrier to uid exchange, whereas the alveolar epithelium is a tight layer that restricts the passage of water, electrolytes and hydrophilic solutes to the air space 29 . Increased protein contents in the BALF, pulmonary hemorrhage, in ammatory cell in ltration, and parenchymal consolidation further con rmed alveolar-capillary disruption.
Under IVIS, we detected the engraftment of transplanted mitochondria into the pulmonary circulation was signi cantly more enhanced in lungs of rats with ALI. The uptake of these exogenous mitochondria was further con rmed by the increased ATP concentrations in the lungs. Consistent with our previous report 18 , these ndings suggest that intravenous delivery of exogenous mitochondria can selectively locate and be taken up by areas of the pulmonary arterial system with mitochondrial dysfunction. After two mitochondrial transplants, the protein content recovered in BALF and extravasated Evans blue dye was signi cantly reduced, suggesting improved integrity of the monolayer capillary endothelium in preventing leakage of plasma proteins and albumin into the air-space 19,30 . However, lung water content measured by LWDR was not differ signi cantly among the three treatment groups. We speculate that the development of extensive alveolar ectasia, lung consolidation and predominant in ltration of lymphocytes at 24 h after ALI 31 , might obscure the formation of pulmonary edema.
Pulmonary endothelial function was also assessed using vasomotor function tests and through the expression of eNOS in the pulmonary artery. The isometric tension analysis found no differences in the concentration response curves of phenylephrine and acetylcholine of ALI animals treated with PBS or mitochondria. However, the relaxation response to maximal concentration of acetylcholine, endothelialdependent relaxation induced by stimulating the muscarinic receptors, was signi cantly potentiated in animals received mitochondrial treatment. The increased isometric tension in the pulmonary artery of ALI rats that received PBS at high acetylcholine concentrations might be due to vasoconstriction reaction mediated by direct stimulation of the muscarinic receptors on the vascular smooth muscle cells where the endothelial layer is damaged or denuded 5,32 . Since the lung injury primarily originated from the bronchoalveolar site in this model, it was therefore reasonable that fewer changes in the vasoreactivity tests were detected in comparison to experimental models of direct pulmonary endothelial injury, such as intravenous administration of oleic acid 5 . Endothelial NOS is an important enzyme in the endothelial cells that synthesizes optimal amount of NO in order to maintain a normal endothelial homeostasis 21,33 .
Phosphorylation of eNOS at Ser1177 through Akt/Protein kinase B or AMP-activated protein kinase is a critical requirement for eNOS activation 34 . The Western blot analysis con rmed that mitochondria transplantation restored eNOS and p-eNOS-S1177 levels in the pulmonary artery, further supporting the engraftment of exogenous mitochondria into the pulmonary circulation can improve endothelial function in rats with ALI. In addition, the increase in capillary endothelium integrity also reduces trans-endothelial migration of immune cells towards the in ammatory cascade in lung tissues 35 . iNOS and CD11b are considered as the cell markers of activated M1 macrophages during the pro-in ammatory phase of ALI 36 . The signi cant suppression of iNOS and CD11b expressions in the injured lung following mitochondrial transplantation might imply the reduced pro-in ammatory cell in ltration in rats with ALI.
The exact mechanisms underlying mitochondrial transplantation in tissue regeneration remains undetermined, but three potential mechanisms have been proposed 37 . The Ca 2+ buffering capacity of mitochondria may attenuate the environmental Ca 2+ overload during cellular stress through the opening of voltage-dependent anion channels 38 . The second theory proposes that the internalization of exogenous mitochondria in recipient cells improves the mitochondrial function of target cells by increasing ATP generation and mitochondrial oxygen consumption 39 . Finally, the viable exogenous mitochondria may release ATP into the extracellular environment and salvage the dysfunctional cells 37,40 . However, there is currently insu cient evidence to support the direct anti-in ammatory potential of transplanted mitochondria during tissue injury. It is then expected that we did not nd any signi cant effects of mitochondrial transplant on the myeloperoxidase activity assay in the lung tissue exposed to endotoxin instillation.
There are a number of limitations in our study. First, the intracellular localization of these exogenous mitochondria was not determined in this study. Secondly, this study recorded very few ALI-related mortality throughout the study period. The effects of mitochondrial transplantation on other outcome measurements (e.g. overall mortality rate and lung parenchymal repair) were not studied. Thirdly, the optimal dose of mitochondria and the frequency of treatment were rather arbitrary, as there is still no general consensus on the standard dosing for mitochondrial transplantation 41 . Fourthly, this was a preclinical, proof-of-concept research for the potential application of mitochondrial transplantation in ALI/ARDS and more mechanistic analysis are still under investigation in our laboratory.
In conclusion, this is the rst report demonstrating that intravenous transplantation of viable allogeneic mitochondria in rats with endotoxin-induced ALI signi cantly improves gas exchange by reducing alveolar-capillary barrier endothelial disruption. However, the effects on survival outcomes and long term recovery of pulmonary function in subjects with ALI after mitochondrial transplants require further investigation.  Changes in the alveolar-capillary permeability were assessed by degree of lung edema (lung wet-to-dry ratio, LWDR), protein content in the broncho-alveolar lavage uid (BALF) and extravasated Evan blue dyne in the lung tissue. There were not differences in the LWDR among the three treatment groups (A), but total protein discovered from the BALF (B) and Evan blue dyne (C) in the lung homogenates was signi cantly   (A). Gross ndings in lungs with acute lung injury (ALI) were heterogeneous, consisted of patchy consolidation and hemorrhage on the pleural surface of the lungs. The injured lungs were in unevenly bluish-red color, as compared with the normal pink aerated appearance in the lungs of sham-operated rats. The gross ndings of lung lobes are presented as dorsal views. Representative lung tissue sections stained with hematoxylin and eosin show alveolar wall thickening and consolidation, neutrophilic in ltration in the alveolar and the interstitial spaces (yellow arrows), extravasation of erythrocytes in the lung interstitium (black arrowheads) and lling of proteinaceous debris or formation of hyaline membrane in the airspaces (red arrows) in the rats with ALI. (B). Severity of lung injury was scored using a semiquantitative histopathology score system 19