Deguelin Attenuates Reperfusion Injury and Improves Outcome after Orthotopic Lung Transplantation in the Rat

The main goal of adequate organ preservation is to avoid further cellular metabolism during the phase of ischemia. However, modern preservation solutions do rarely achieve this target. In donor organs hypoxia and ischemia induce a broad spectrum of pathologic molecular mechanisms favoring primary graft dysfunction (PGD) after transplantation. Increased hypoxia-induced transcriptional activity leads to increased vascular permeability which in turn is the soil of a reperfusion edema and the enhancement of a pro-inflammatory response in the graft after reperfusion. We hypothesize that inhibition of the respiration chain in mitochondria and thus inhibition of the hypoxia induced mechanisms might reduce reperfusion edema and consecutively improve survival in vivo. In this study we demonstrate that the rotenoid Deguelin reduces the expression of hypoxia induced target genes, and especially VEGF-A, dose-dependently in hypoxic human lung derived cells. Furthermore, Deguelin significantly suppresses the mRNA expression of the HIF target genes VEGF-A, the pro-inflammatory CXCR4 and ICAM-1 in ischemic lungs vs. control lungs. After lung transplantation, the VEGF-A induced reperfusion-edema is significantly lower in Deguelin-treated animals than in controls. Deguelin-treated rats exhibit a significantly increased survival-rate after transplantation. Additionally, a downregulation of the pro-inflammatory molecules ICAM-1 and CXCR4 and an increase in the recruitment of immunomodulatory monocytes (CD163+ and CD68+) to the transplanted organ involving the IL4 pathway was observed. Therefore, we conclude that ischemic periods preceding reperfusion are mainly responsible for the increased vascular permeability via upregulation of VEGF. Together with this, the resulting endothelial dysfunction also enhances inflammation and consequently lung dysfunction. Deguelin significantly decreases a VEGF-A induced reperfusion edema, induces the recruitment of immunomodulatory monocytes and thus improves organ function and survival after lung transplantation by interfering with hypoxia induced signaling.


Introduction
Lung transplantation is, despite the highest mortality among all solid organ transplants, the only therapeutic option in patients with end-stage pulmonary disease [1]. The fragility and the poor tolerance against ischemia of this organ is responsible for the fact that only 20% of the candidate lungs are currently being transplanted [2]. Today, primary graft dysfunction (PGD) represents the most common acute complication during the first 30 days after lung transplantation [1,3]. PGD mainly develops on the soil of an ischemia-reperfusion (IR) reaction that is characterized by an overshooting inflammatory response, culminating in tissue edema and consecutive graft failure [4,5].
Despite downregulation of the cellular metabolism by modern preservation methods (i.e. preservation solutions, hypothermia etc.), hypoxia-induced gene transcription is not completely blunted. However, the complete blockade of the cellular metabolism for the time of ischemia might be of benefit. First, due to the loss of oxygen consumption, no metabolites are produced, that might have an adverse effect on cellular viability. Second, upon inhibition of cellular respiration, the intracellular oxygen concentration remains normal under ischemia, and thus oxygen is available for hydroxylation of the transcription factor Hypoxia Inducible Factor (HIF)-1, thus resulting in its decreased transcriptional activity [6,7,8].
HIF-1 occupies a central and critical role, involving a large number of genes [9]. Among these genes, especially Vascular Endothelial Growth Factor (VEGF) and genes that regulate cell adhesion to the endothelium such as Inter-Cellular Adhesion Molecule (ICAM)-1 are activated within minutes by HIF-1 [10,11,12,13,14,15]. VEGF is known to quickly increase vascular permeability about 50.000 times stronger than histamine, by enhancing the break-down of the endothelial barrier function and inducing the formation of interendothelial gaps [16]. This in turn leads to an increased tissue edema ending up in organ failure [14,17,18,19,20,21,22,23]. Together with ICAM-1, VEGF is highly chemotactic for inflammatory cells like monocytes [24,25].
Clinically relevant effects of hypoxia induced pathomechanisms are observed in high-altitude mountaineers, where low oxygen tension is highly correlated with lung and brain edema, as well as systemic inflammatory response syndrome [26,27,28,29]. Hypoxia thus triggers inflammation, and on the other hand, inflammation itself causes local hypoxia, combined with activation of the coagulation system [19,30], leading to a vicious circle as seen in acute lung injury [31,32].
After restoration of the normal blood flow (reperfusion), the resulting increased concentration of pulmonary VEGF leads to an increased vascular permeability resulting in early post transplantation dysfunction [33,34,35,36,37].
However, the main functions of the hypoxia induced genes are meant ''self-protective'' by preventing organ damage (i.e. increase of hemoglobin, increase of micro-vessel density through neoangiogenesis, increase of glycolysis to ensure cellular metabolism). Unfortunately, at least in endothelium-rich tissues, these effects seem to have a negative impact on the outcome after ischemic events i.e. transplantation. It has been shown that these negative effects are mainly triggered by HIF-1 and thus correlate with primary graft failure [14,15,22,38]. A dual blockade of the respiration chain and the hypoxia induced gene expression might therefore be of benefit in lung transplantation.
The rotenoid Deguelin is able to provide this dual inhibition. Deguelin acts as a mitochondrial inhibitor of the respiration chain via inhibition of the mitochondrial NADH dehydrogenase/ complex I [39]. The resulting reduced oxygen consumption leads to increased HIF-1 hydroxylation and thus inhibition of its transcriptional activity [7]. Therefore, a blockade of the deleterious effects of HIF-1 during lung transplantation might be beneficial to prevent PGD and improve short-term survival by reducing tissue edema.

Deguelin Gavage is Well Tolerated
All animals received two daily gavages of Deguelin intragastrically at 4 mg/kg bodyweight for 3 days. All animals treated with Deguelin or solvent significantly gained weight during the trial (297.869.805 g and 335.662.064 g vs. 257.164.389 g, P = 0.0009, P,0.0001 resp.). However, animals receiving only solvent, were significantly heavier compared to the Deguelin treated animals (335.662.064 vs. 297.869.805, {P = 0.0196). No deleterious side effects were observed in all groups (Fig. 1).

Deguelin Effectively Blocks HIF-1 during Hypoxia in vitro
Deguelin effectively suppressed HIF-1 protein after 6 hours hypoxia in vitro at a concentration of 100 nM in human microvascular endothelial cells (HMECs) as well as in human lung epithelial derived cells (NCI-H460 and HTB-177). DMSO (solvent) with hypoxia and hypoxia (HOX) alone stabilized HIF-1( Fig. 2A). Deguelin effectively inhibits the hypoxia-induced expression of HIF target genes in a concentration dependent manner. Carbonic anhydrase IX (CAIX), vascular endothelial growth factor (VEGF)-A, lysyl oxidase (LOX), angiopoietin-related protein 4 (ANGPTL4), egl nine homolog 3 (EGLN3) and adrenomedullin (ADM) mRNA expression is significantly down-regulated in hypoxic cells (HOX) upon treatment (Fig. 2B). EGLN2, a HIF prolyl hydroxylase that is not a HIF target gene serves as negative control, to show that Deguelin is not a general transcription inhibitor.

Deguelin Effectively Blocks HIF-1 during Hypoxia in vivo
To test whether Deguelin has a sufficient in vivo activity, treatment was initiated 3 days prior to sacrification at 4 mg/kg BW twice/day. On day 4 the animals were killed and the lungs were excised and stored at 37uC for 1 h, simulating warm ischemia (w.i.). Sham lungs were immediately snap-frozen without ischemia (Fig. 3A). The readout was performed on these samples by measuring HIF-1 regulated genes using RT-PCR. Basal mRNA expression under normoxia (sham) was detectable for VEGF-A (100.0612. 28 (Fig. 3B). These results indicate that Deguelin potently interferes with the hypoxia induced gene expression (Fig. 2B). Especially VEGF-A, which regulates vascular permeability, and ICAM-1 as well as CXCR4, two genes regulating

Deguelin Prevents from Structural Changes and Cellular Edema in Ischemic Lungs
From the organs stored at 37uC for 1 h, samples were collected and histology was performed. All sections were stained with H&E and pictures with 106 and 406 magnification were taken. To quantify cellular edema and structural pulmonary changes, the area occupied by tissue (positive H&E staining) was determined planimetrically. For this purpose, 3 randomly taken pictures from each slice were analyzed (Fig. 4A, blue). The increase in cellular edema and structural changes directly correlate with the increase in area occupied by lung tissue per section. Lungs from animals that received Deguelin (w.i.D.) prior to ischemia (36.562.88% tissue/area), displayed significantly reduced cellular edema and structural changes than those from sham (50.561.34% tissue/area) or with warm ischemia (w.i.) (55.262.73% tissue/area) groups (Fig. 4B). In animals receiving Deguelin (w.i.D.), the microstructure was preserved. These results indicate that hypoxia/ischemia mediate the development of acute tissue edema.

Deguelin Treatment Reduces Reperfusion Edema by Suppression of VEGF-A Protein in Lung Grafts and Improves Short-term Survival after Orthotopic Lung Transplantation
To test the hypothesis whether Deguelin might reduce tissue edema in vivo via modulation of the VEGF-A pathway, we performed transplantation experiments. For this purpose we transplanted the left lung from either pre-treated donors to pretreated recipients (Deguelin) or the left lung from non-treated donors to non-treated recipients (control). Analysis of the transplanted organs revealed that lungs from animals, which were treated with Deguelin, had significantly lower levels of VEGF-A mRNA and protein when compared to untreated subjects (Fig. 5A).
A small part of the transplanted left lung was used to determine the wet-to-dry ratio, directly reflecting the extent of reperfusion edema in the lung tissue. Animals, which received Deguelin, had significantly less tissue water and thus less edema compared to lungs that did not receive any treatment ( Fig. 5B left graph). Together with these findings, animals that had lower posttransplantation-edema also had a significant better survival rate. In corresponding micrographs, lungs that have a high degree of edema show similar structural changes compared to lungs from the preliminary experiment (Fig. 5C) that underwent warm ischemia, without administration of Deguelin. These data suggest, that hypoxia directly induces reperfusion edema via the upregulation of VEGF-A expression.
Besides, Deguelin also has immunomodulatory effects on the donor lung by favoring the recruitment of anti-inflammatory macrophages (CD 163+) involving the CCL2/MRC-1/IL4 pathway. In corresponding micrographs, the amount of CD 163+ cells is significantly higher in animals that have been treated with Deguelin vs. controls (36.263.7 vs. 8.763.1; P,0.0001). This directly correlates with the better outcome of these animals. A CD 68 staining reveals that mature macrophages are mainly participating in this anti-inflammatory and HIF-1 inhibitory effect. In Deguelin treated animals the amount of CD 68+ cells is significantly higher than in non-treated animals (37.865.9 vs. 8.562.2; P,0.0001) and thus correlates directly with the amount of CD163+ cells. However, in a pan-macrophage/dendritic cell staining (RM-4), the total macrophage count is significantly higher in untreated than in Deguelin treated animals (67.461.9 vs. 34.062.3; P,0.0001). Furthermore, the amount of RM-4 positively stained macrophages in the Deguelin group corresponds to the amount of CD-163+ cells from the same group (Fig. 7).

IL6 Protein Downregulation and Interferon (IFN) c Protein
Upregulation is Co-associated with Adverse Outcome IL6 and IFNc are known to play an important role in inflammatory reactions especially in accordance to rejection reactions. As the recruitment of M2 macrophages also requires to some extent a pro-inflammatory reaction, we aimed to test whether IL6 or IFNc might be regulated in our experimental setting. Therefore we performed ELISA on serum samples derived from sham, Deguelin and Perfadex-treated (control) rats. Interestingly, animals that have been treated with Perfadex only, have significantly lower levels of serum IL6 as seen in Deguelin or sham animals (81.5614.3 pg/ml vs. 200.0632.8 pg/ml resp.178.9622.0 pg/ml; P,0.004). Consistent with this, the serum protein expression of IFNc is significantly upregulated in controls vs. Deguelin treated animals (6.160.5 pg/ml vs. 3.960.3 pg/ml; P = 0.0025). These data show that both IL6 and IFNc might play an important role in orchestrating these inflammatory reactions and showing that a lack of IL6 and an overexpression of IFNc are co-associated with the M1 or proinflammatory macrophage phenotype (Fig. 8).

Discussion
Today, the number of patients on the waiting lists for organ transplantation is steadily increasing; in contrast, the number of available organs is not. This leads to the fact that currently organs are considered for transplantation even when potentially subjected to longer ischemia periods or poor hemodynamics of the donor. Some years ago, they probably would have been rejected [40]. Given these circumstances, pharmacologic protection of the available organs is of increasing interest. Ischemia and reperfusion (IR) injury to organs represents one of the most common clinical problems in transplantation medicine. IR may lead to primary or secondary organ failure after restoration of the blood flow [41,42,43]. The theory that hypoxia triggers inflammation is not new. There is sufficient evidence that hypoxia leads to activation of several deleterious pathways like the occurrence of edema due to enhanced vascular leakage and accumulation of inflammatory cells in organs, increasing the risk of early and late graft rejection after transplantation [27]. Hypoxia or ischemia are known triggers of a vicious circle, enhancing inflammation and organ damage. Simplified, the paradigm -tissue hypoxia leads to inflammation, tissue inflammation leads to hypoxia-can be applied [44].
During hypoxia or ischemia, HIF-1 is mainly responsible for the initiation of gene transcription in many lung diseases like gutinduced lung injury following IR, hypoxic pulmonary hypertension, COPD and fibrous airway remodeling [45,46,47,48]. Hypoxia stabilizes HIF-1 which in turn directly and quickly upregulates target gene's expression within minutes [49,50,51,52]. Usually, the maximum transcription rate culminates at 2 hours of hypoxia. In the present study, however, we could demonstrate that HIF-1 induces transcription in ischemic lung tissue already much earlier (within 1 hour), however with lower efficacy. The reason for this quick response upon hypoxia might be explained by the fact that hypoxic organs and cells have to react immediately to prevent damage and ensure survival of the organism. Therefore we suppose that HIF-1 has two major actions, a deleterious fast and pro-edematous one, and a longer lasting one, namely angiogenesis and endothelial proliferation [17]. In transplantation medicine, VEGF has been shown to exert pro-inflammatory and proapoptotic actions during the early phase of reperfusion [53]. Our own data concerning patients who underwent lung or heart transplantation could correlate increased HIF-1 binding activity and high VEGF-A levels with more pronounced edema and rejection rates [22,23,37]. Therefore these findings led to the hypothesis that the inhibition of VEGF-A pathways, as well as other hypoxia triggered pro-inflammatory genes such as ICAM-1 and CXCR4 could reduce early complications such as edemacaused PGD after lung transplantation. Several recent publications also describe protective effects of the latter genes. Namely the work by Jiang et al. demonstrated that a therapeutic enhancement of the vascular integrity via HIF-1 activation during acute rejection promotes tracheal graft health and prevents from chronic rejection [54]. This work stands in direct contrast to the work of Krebs et al. who demonstrated that VEGF is responsible for the occurrence of bronchiolitis obliterans [55]. These data indicate that the sum of the hypoxia induced effects might be protective. One possible explanation might be that the work by Jiang et al. does not include short-term survival results reflecting early complications. As the trachea is not participating in the gas exchange, animals might show early inflammation, but they will not die of it, as both lungs are still functional and gas exchange remains unchanged. To test our hypothesis that the summative effects of hypoxia negatively influence the graft fate during the early reperfusion phase by increasing tissue edema and inflammation, we used the model of orthotopic lung transplantation in the rat. This model has the advantage that a read out of early complications by providing short term survival data is possible. In the present work, we could demonstrate that the summative effect of hypoxia induced gene expression is deleterious. We could show for the first time that solely 1 hour of hypoxia leads to an increase of pro-edematous VEGF-A mRNA expression and in turn leading to an edematous tissue alteration as shown in figure 4. Transplanted and reperfused lungs without Deguelin treatment had a high water content as depicted in the wet-to-dry ratio (Fig. 5B). Together with this, the survival was significantly increased in the Deguelin group, underlining the beneficial effect of HIF-1 Inhibition in vivo. The mechanisms by which Deguelin exerts the anti-inflammatory properties might rather directly result from inhibition of the HIF-1 pathway than from interference with the Pi3K/Akt pathway, which is not addressed by the concentration of deguelin used in this model (Fig. 9). Our data show for the first time that treatment with Deguelin significantly reduces hypoxia induced deleterious effects in lung grafts and thereby significantly lowers reperfusion edema and improves survival. These results indicate that systemic perioperative inhibition of the respiratory chain in mitochondria in donor organs might be a beneficial therapeutic option. It was shown recently that ICAM-1 and CXCR4 are involved in lung injury via recruitment of inflammatory cells [56,57]. We could now demonstrate that Deguelin application in vitro and in vivo leads to a consecutive downregulation of ICAM-1 and CXCR4. This attenuation consequently leads to less inflammation in the grafts. When looking at the micrographs, we found however that Deguelin treatment leads to occupation of the alveolar spaces by monocytic cells. This however appears not logical, as the lung inflammation markers ICAM-1 and CXCR4 were downregulated. Immunohistochemical staining revealed that these monocytic cells highly expressed the surface markers CD 163, meaning that these monocytes are anti-inflammatory macrophages. The marker CD 68, which represents mature macrophages, is poorly expressed in healthy lungs, however in the transplanted lung CD 68 is upregulated indicating that the anti-inflammatory macrophages are fully differentiated. Whether these macrophages are recruited as precursor cells or whether these cells are host derived remains unclear at this moment. The RM-4 staining (pan-macrophage marker) shows that transplanted lungs have in total fewer monocytic cells. Interestingly, nearly 100% of these cells are anti-inflammatory. As we used an allo-isogenic model, a mild immunologic reaction is accepted, simulating a good cross-match. Interestingly, IL10, a potent chemoattractant for anti-inflammatory macrophages is downregulated in Deguelin-treated animals. However, a downregulation of the anti-inflammatory cytokine IL10 is not contradictory. First, IL10 is mainly produced by antigen presenting cells (APCs). Immunocytochemistry showed that the infiltrating immune cells in the HIF-1 inhibitor-treated group are mainly anti-inflammatory. Therefore few APCs are recruited to the lung, resulting in a downregulated IL10 expression. Secondly, recent works demonstrated that IL10 must already be upregulated before the event to act as anti-inflammatory agent. Should it be upregulated after the inflammatory event, it has a pro-inflammatory effect [58]. Thirdly, IL4 is known to downregulate IL10 expression [59]. As IL4 is upregulated in Deguelin treated rats, the downregulation of IL10 is plausible. On the other hand, serum analysis of IL6 and IFNc revealed, that control animals have significantly lower levels of IL6 and significantly higher levels of IFNc. It is known that IL6 is not only a simple pro-inflammatory cytokine. Together with IL4, IL6 induces M2 polarization of macrophages, which reflects our findings [60]. The rather high level of IL6 in the sham group is plausible, as the thoracotomy leads to a rather large wound, which is known to upregulate IL6. Higher IL6 levels in Deguelin treated animals thus directly correlate with the better outcome. The low IL6 levels in the controls might also reflect the effect of IFNc or might be explained by a different IL6 kinetics in this group. Comparatively IFNc, that is known to inhibit M2 polarization in vivo, is upregulated in control animals, which show low levels of M2 macrophages [61]. Thus, this cytokine profile is plausible and reflects the pathways involved in M2 differentiation in this animal model. We think that for the initiation of a M2 polarization, some extent of inflammation must precede this reaction. Once macrophages are recruited to the site of inflammation, IL4, MRC-1 and CCL2 initiate M2 polarization. In the case of the controls, the initial IL6 upregulation is missing and thus, M2 differentiation is missing. Together with this, IL10, CXCR4 and ICAM-1 upregulation is leading to an overshooting inflammatory response. However, the exact mechanisms underlying this complex reaction remain unclear. We think it is rather a consequence of the cells infiltrating the grafts and the following specific reactions initiated by target-oriented cytokines. Therefore Deguelin is not likely to directly stimulate IL6, CCL2 or MRC-1.
Taken together, our data clearly indicate that in the model of orthotopic lung transplantation, a systemic treatment with Deguelin prior to transplantation is beneficial and reduces early complications such as PGD. The suppression of hypoxia induced pathomechanisms leads to an attenuated VEGF-A activity and therefore significantly reduces reperfusion edema. Deguelin treatment also prevents from pro-inflammatory phenotypes by suppressing ICAM-1 and CXCR4 expression. The exact mechanisms, by which Deguelin influences the CCL2, MRC-1 and IL4 signals are yet unknown and remain subject of future studies.

Ethics Statement
The study was approved by the Animal Care and Use Committee of the state of Hesse (Regierungsprä sidium Darmstadt), Darmstadt, Germany (V54-19c20/15-F91/56). Surgery and animal care was performed in accordance with the ''Guide for the care and use of laboratory animals'' (National Institutes of Health, volume 25, no. 28, revised 1996), EU Directive 86/ 609 EEC and German Protection of Animals Act.

Deguelin Preparation and Administration
The HIF-1 inhibitor Deguelin was obtained from Enzo Life Sciences (Lörrach, Germany). Stock solution was prepared by dissolving Deguelin in 100% DMSO at 25 mg/ml and stored at 220uC until further use. For the use in animals, Deguelin was dissolved in corn oil at a final concentration of 10 mg/ml. Deguelin was applied via gavage twice a day at a dose of 4 mg/kg BW. Treatment started 3 days before transplantation in donors and recipients and ended 48 hours after transplantation (in recipients) (Fig. 10).

Animals
Male Sprague Dawley rats (SD, Janvier, St. Berthevin, France) weighing 225-250 g were housed in the central research establishment of the Goethe-University Frankfurt. At the beginning of the experiment, rats were randomized to form pairs of donor and recipient rats, which were kept together in approved plastic cages (2 animals per cage), had water and food ad libitum and were housed in rooms equipped with a 12 h light cycle. The rats were individualized using a unique color code printed on the tail with water-resistant paint.

Organ Transplantation and Harvest
In general the experiment was divided into two stages. The first stage comprised harvesting lungs from animals either treated or not treated with Deguelin under different ischemic conditions (Table 1).
To harvest the lungs the rats were anesthetized with Ketamin-Xylazine (75/10 mg/kg body weight) and intubated. Anesthesia was continued with Isoflurane (1.5 vol %) using a pressure controlled small animal ventilator (Föhr Medical Instruments, Seeheim-Ober Beerbach, Germany) at 85 bpm and pmax = 25 mbar. Before removal, the lungs were flushed with 20 mL perfusion solution (PerfadexH solution, kept at 4uC) until the lungs turned  pale. The left and right lungs were dissected from the heart-lung complex, one half of the obtained organs were snap-frozen in liquid nitrogen, the other half was stored in 4% formaldehyde. In the second stage of the experiments, an orthotopic allo-isogenic (SD-SD) unilateral lung transplantation was performed. These transplantation experiments compromised 2 groups: Deguelintreated animals vs. control animals without treatment ( Table 2, Fig. 10).

Donor Procedure
Donor animals were anesthetized, intubated and ventilated as described earlier. Donor organs were obtained after perfusion of the heart-lung complex using an antegrade perfusion via the pulmonary artery. Lungs from the Deguelin group and control lungs were perfused each with 20 ml of PerfadexH (Vitrolife, Kungsbacka, Sweden), a solution with low potassium, not exceeding 20 cm H 2 O hydrostatic pressure. After removal, the heart-lung complex was wrapped in sterile gauze and kept on ice until backtable preparation.

Backtable
For this purpose the left lung artery, vein and bronchus were dissected under sterile conditions from enclosing tissue. The loose ends of the three structures were drawn through cuffs (made from venous catheter), everted and fixed using a 6-0 prolene thread. The prepared left lung grafts were wrapped in sterile gauze and stored on ice until transplantation.

Recipient Procedure
The recipient rats were anesthetized as described before. The thoracic cavern was opened using a left antero-lateral thoracotomy. The left lung lobe was carefully prepared, exposing the pulmonary artery, vein and bronchus. Then, the vessels and bronchus were selectively ligated using 6-0 prolene vessel loops to obstruct air and blood flow. Artery and vein were incised and flushed with a solution of Heparin/saline (150 U/mL). Then the bronchus was incised and the prepared donor lung was placed on top of the recipients left lung lobe. Connections between the structures of the donor organ and the recipient organism were established by inserting the corresponding donor structures into the incisions of the recipient vessels and bronchus. The generated connections were secured using 5-0 silk ligation (Fig. 11). The reperfusion of the vein and artery and the reestablishment of airflow through the bronchus was accomplished in the order mentioned. After successful transplantation the recipient's left lung lobe was removed, the donor organ was placed into the thorax and the wound was closed. After successful suturing, local anesthesia was instilled (Ropivacain 0,2%, 1 ml) to ensure post-operative pain relief. Enrofloxacin was used as a post-surgical antibiotic and pain was controlled with Piritramid administered via drinking water (1,5 mg/50 ml). The recipient rats were observed for an additional 48 h and then killed under deep anesthesia. The Lungs were harvested, dissected and stored as described earlier.

Wet-to-dry Ratio
A small part of the collected lungs was minced and stored in a reaction tube and the weight was determined. Then the tube was placed in an oven at 60uC for 72 h. Thereafter, the tube was reweighed and the ratio between wet and dry weight was calculated. Comparison of the wet-to-dry ratios of different organ samples allowed drawing conclusions regarding the dimension of lung edema as reported elsewhere [22].

Protein Isolation, Western Blot and ELISA
Briefly, protein from left lungs was isolated as described earlier [23,62]. 10% and 7.5% SDS-gels were loaded with 50 mg protein.
Proteins were immunodetected on Hybond C supermembrane (Amersham Pharmacia Biotech, Buckinghamshire) with Spectra broad range marker (Fermentas, Germany) as a standard. The blots were probed with antibodies as follows: HIF-1 alpha -mouse monoclonal antibody (Abcam, Cambridge, UK), VEGF -mouse monoclonal antibody (Abcam, Cambridge, UK). Digitalization and evaluation of the blots was performed with a Kodak Imager (Carestream, Stuttgart, Germany). For serum sampling, blood was withdrawn and clotting was performed for 15 minutes followed by centrifugation with 3000 RPM for 10 minutes at 4uC. Serum supernatant was transferred to a new tube and then stored at 280uC for further processing. ELISA on rat IL6 and rat Interferon-c (Thermo Fisher, Braunschweig, Germany) were performed according to the manufacturer's protocol. Detection of the concentration was performed using a microplate reader (Bio-Tek Instruments, Bad Friedrichshall, Germany).

Histology and Immunohistochemistry
One half of the samples from the left lung lobe was fixed in formaldehyde 4% for 48 h and dehydrated afterwards in solutions  Figure 11. Technique of lung transplantation. (A) For perfusion, a venous catheter is placed through an incision in the right ventricle into the pulmonary artery. After perfusion, the heart-lung complex is removed in toto, suspended on the trachea that is ligated in maximal inspiration to avoid atelectasis. B) Cuffs are prepared using venous catheters. The vessel is slipped through the lumen and fixed with a 6-0 polypropylene suture. C) Donor (cuffed) blood vessels and bronchus were slipped (sleeved) into the recipient's artery, bronchus and vein. When no twisting of the corresponding structure was observed, the anastomosis was secured with a 5-0 Silk suture placed around the cuff and sleeved recipient's structure. doi:10.1371/journal.pone.0039265.g011 of increasing alcohol concentrations and then kept in methylbenzoate-I for 16 h and methylbenzoate-II for 8 h. After washing in liquid Paraffin (Merck, Darmstadt, Germany) samples were embedded in paraffin and cut to 4 mm slides. Samples were stained with haematoxylin and eosin (AppliChem, Darmstadt, Germany) according to the manufacturer's protocol. Immunohistochemistry was performed with DAB-staining, using RM-4 pan-macrophage/ dendritic cell marker (TransGenic Inc., Kobe, Japan), CD 163 (AbD Serotec, Düsseldorf, Germany), CD 68 (Millipore, Schwalbach, Germany), ICAM-1 (Labomics, Nivelles, Belgium) and CXCR4 (Abcam, Cambridge, UK) antibodies [62]. Briefly, after deparaffinization, the slides were rehydrated in decreasing alcohol concentrations. After blocking with 6% serum (from the species of the secondary antibody), the primary antibody was incubated for 2 h at room temperature. For detection, a secondary antibody labelled with HRP (anti mouse IgG, DAKO, Hamburg, Germany) (for CD 163, CD 68 and ICAM-1) was incubated for 1 h at room temperature. For detection of CXCR4 a biotinylated anti rabbit antibody was incubated for 1 h prior to detection with streptavidin-HRP (AXXORA, Lörrach, Germany). Development was performed with DAB incubation for 10 minutes and counterstaining was done with haematoxylin for 10 min. Images were taken with the Leica DM5000B microscope and analyzed using an automatized MatLab algorithm. The software determines the number of pixel in each RGB channel, thus measuring the relative area occupied by positive staining per image. Quantifica-tion was done setting brown pixels proportioned to total number of pixel.

Statistical Analysis
Statistical analysis was performed with GraphPad PrismH 5.02 software (GraphPad Software, Inc., San Diego, California). Results are expressed as means 6 standard error of the mean (SEM). Statistical significance was calculated using one-way ANOVA followed by Student's t test. Statistical significance was set to p,0.05.