|Home | About | Journals | Submit | Contact Us | Français|
Extracellular adenosine has tissue protective potential in several conditions. Adenosine levels are regulated by a close interplay between nucleoside transporters and adenosine kinase (ADK). Based on evidence of the role of ADK in regulating adenosine levels during hypoxia, we evaluated the effect of ADK on lung injury. Furthermore, we tested the influence of a pharmacological approach to blocking ADK on the extent of lung injury.
Prospective experimental animal study.
University based research laboratory.
In vitro cell lines, wildtype (Wt) and ADK+/− mice.
We tested the expression of ADK during inflammatory stimulation in vitro and in a model of lipopolysaccharide (LPS) inhalation in vivo. Studies using the ADK promoter were performed in vitro. Wt and ADK+/− mice were subjected to LPS inhalation. Pharmacological inhibition of ADK was performed in vitro, and its effect on adenosine uptake was evaluated. The pharmacological inhibition was also performed in vivo, and the effect on lung injury was assessed.
We observed the repression of ADK by pro-inflammatory cytokines and found a significant influence of NF-κB on regulation of the ADK promoter. Mice with endogenous ADK repression (ADK+/−) showed reduced infiltration of leukocytes into the alveolar space, decreased total protein and myeloperoxidase levels, and lower cytokine levels in the alveolar lavage fluid. The inhibition of ADK by 5-iodotubercidine increased the extracellular adenosine levels in vitro, diminished the transmigration of neutrophils and improved the epithelial barrier function. The inhibition of ADK in vivo showed protective properties, reducing the extent of pulmonary inflammation during lung injury.
Taken together, these data show that ADK is a valuable target for reducing the inflammatory changes associated with lung injury and should be pursued as a therapeutic option.
The acute respiratory distress syndrome (ARDS) and lung injury (ALI) are associated with high morbidity and mortality in the critically ill (1). Clinically, ALI is characterised by the acute onset of hypoxaemia (which requires oxygen support), pulmonary inflammation and changes in the pulmonary structure during radiological assessment. These changes are caused by the reduced function of the alveolar-capillary barrier, which leads to increased pulmonary vascular permeability, and by the infiltration of inflammatory cells into the alveolar space (2, 3). The severity of these pathophysiological changes determines the severity of the associated lung injury and the patient outcome (4).
Previous research has identified extracellular adenosine as a protective molecule in hypoxia (5), during ischaemia reperfusion (6) and in inflammatory conditions, including acute and chronic lung diseases (7, 8). Extracellular adenosine is generated by the conversion of extracellular ATP to ADP and AMP as a result of dephosphorylation by the enzymes ectonucleotidase triphosphate diphosphohydrolase (ENTPDase; CD39) and ecto-5′-nucleotidase (CD73), which are primarily located on the vascular endothelium (9). Signal transduction by extracellular adenosine then functions through the four purinergic G-protein-coupled adenosine receptors. These adenosine receptors are widely distributed throughout the organism, and all of these receptors are proposed to mediate tissue protection (10, 11). Following its generation, adenosine is mainly cleared from the extracellular space by the equilibrative nucleoside transporters (ENTs) and concentrative nucleoside transporters (CNTs), resulting in the movement of adenosine into cells. In the intracellular space, adenosine is then either degraded to inosine via adenosine deaminase (12) or converted to AMP and ADP by adenosine kinase (ADK) (13). Therefore, the intracellular enzyme ADK also regulates the extracellular adenosine levels. The inhibition of ADK results in elevated extracellular adenosine levels; therefore, this inhibition has a high therapeutic potential for protecting tissue in several pathological conditions (13, 14).
Because changes in the vascular barrier function and infiltration of leukocytes into the alveolar space are key processes during the initial stages of lung injury, we sought to describe the role of ADK in the development of this condition. In addition, we evaluated whether ITU can inhibit ADK, as a possible pharmacological intervention that acts on ADK to ameliorate the changes in lung injury.
All animal protocols were in accordance with the German guidelines for the use of living animals and were approved by the Institutional Animal Care and Use Committee of the Tübingen University Hospital and the Regierungspräsidium Tübingen. For experiments using materials from human blood samples, ethics committee approval was obtained, and each participant provided written informed consent.
To assess the relative expression levels of ADK, we tested various tissues from WT and ADK+/− mice. For details, please see the online supplement.
In a separate experiment, human lung epithelial A549 cells were grown to confluence and stimulated with TNF-α (100 ng/ml), IL-6 (20 ng/ml), or IL-1β (20 ng/ml) for 2, 4, 8 and 24 hours. For further details, please see the supplement material.
Cell culture and mouse tissue samples were normalised for the protein levels before being applied to SDS containing poly-acrylamide gels and were processed according to standard protocols.
Lung epithelial A549 cells and murine tissues for immunofluorescence staining of the lungs were harvested and processed according to standard protocols. For further details, please see the supplement material.
Peripheral blood was taken from healthy donors into Sarstedt monovettes containing sodium citrate (Sarstedt) and neutrophils were isolated as previously described (17). For transendothelial migration studies, A549 cells were grown on the basolateral aspect of permeable transwell inserts (0.4 μm pore size, 6.5 mm diameter) until they were confluent. Migration assays were performed as described elsewhere (17). In a subset of experiments, ADK A549 cells were inhibited by ITU for 30 min before the transepithelial migration assay was started.
Human A549 cells were seeded and grown to a monolayer on the permeable membrane as described above. Flux assays were performed as previously described (18).
Human lung epithelial A-549 cells were cultured for 2 days in 6-well plates to confluence. At the start of the experiment, the cell media were replaced with HBSS+ containing 10 or 25 μm/ml 5-iodotubericine (ITU). As a control, we used cells without ITU treatment, and to control adenosine reuptake, we administered 25 μm/ml ITU plus 10 μm/ml dipyridamole, which is a cellular adenosine uptake inhibitor. For details, please see the online supplement.
Promoter analysis and identification of both the transcription start site and the potential NF-κB-binding sites were evaluated using MatInspector by Genomatix. The pGL4.17-expressing sequence vector, which corresponds to the full-length ADK promoter region, was purchased from GeneArt. For further details, please see the supplement material.
The ChIP assay (CHP1; Sigma-Aldrich) was performed according to the manufacturer’s instructions. As the ChIP-qualified antibody of interest, we used the NF-κB antibody Anti-NF-κB p105/p50 – ChIP Grade Abcam (ab7971). For the PCR detection of ADK, we used the following primers: ttt cct agg ctg agg ctt ccc (forward) and tca gct ccc tgt aac agc act (reverse).
After animals were sacrificed, bronchoalveolar lavage (BAL) was collected by performing a tracheotomy and flushing the lungs three times with 0.6 ml of PBS. For further details, please see the supplement material.
We performed statistical analysis using one-way analysis of variance (ANOVA) to determine the differences between the groups using Dunnetts post-hoc analysis. Unpaired Student t test was used where appropriate. A value of P-value < 0.05 was considered to be statistically significant.
In an initial experiment, we sought to evaluate whether the ADK levels change during lung injury in mice. For this, we exposed mice to a lung injury model involving LPS inhalation. Following this, we found significant repression of ADK at the transcriptional level in pulmonary tissue (Figure 1A). The protein levels of ADK also reflected this finding (Figure 1B). To illustrate the repression of ADK, we performed immunofluorescence staining of the pulmonary tissue. The ADK staining in pulmonary tissue sections showed reduced ADK-specific immunofluorescence following LPS exposure. An epithelial marker, cytokeratin, was used for co-immunostaining (Figure 1C, Supplemental Figure E1).
To determine whether ADK expression is regulated at the transcriptional level in response to pro-inflammatory stimuli, we exposed confluent pulmonary epithelial cell monolayers (A549) to TNF-α (100 ng/ml), IL-6 (20 ng/ml) or IL-1β (20 ng/ml) for 0, 2, 4, 8 and 24 hours. In addition, we measured ADK transcriptional expression following exposure to different concentrations of these cytokines. We found significant repression of ADK at the transcriptional level, and we detected an early onset of repression, within 2 hours of the start of exposure (Figure 2A and B, Supplemental Figure E2A and B). To verify our results, we examined ADK expression at the protein level in A549 cells using different cytokine concentrations. The protein analysis confirmed the results that were observed at the transcriptional level, showing repression of ADK protein expression after pro-inflammatory stimulation with TNF-α, IL-6 or IL-1β (Figure 2C, Supplemental Figure E2C). To visualise the repression of ADK, we used immunofluorescence staining of alveolar A549 cells. Following 24 hours of stimulation with TNF-α, there was reduced ADK immunofluorescence (Figure 2D, Supplemental Figure E1). This finding could be confirmed by measurement of ADK through ELISA (Supplemental Figure E3).
To gain further insight into the regulatory mechanisms controlling ADK at the transcriptional level, we performed a transcription factor binding site analysis of the ADK promoter region using the available public databases. We identified one NF-κB binding site at −309 bp (Figure 3A). To identify the functional role of this NF-κB responsive element, we first tested the binding activity of NF-κB to the ADK promoter in A549 cells by using chromatin immunoprecipitation. There was a strong binding response of NF-κB after stimulation with TNF-α, which indicated an NF-κB-driven mechanism that regulated ADK production at the transcriptional level (Figure 3B). To assess the functional role of the NF-κB binding site, we designed luciferase reporter constructs that expressed the putative full-length ADK (ADK-FL) promoter (Figure 3C). In addition, we manufactured site directed mutations in the NF-κB response element regions to prevent NF-κB binding. We transfected A549 cells with the promoter constructs and stimulated the cells with 100 ng/ml TNFα for 24 h. The luciferase activity of the unmodified ADK-FL promoter was strongly reduced in response to the inflammatory stimulus. Following mutation of the binding site for NF-κB (Δ309), this repression of luciferase activity was reversed.
Because high extracellular adenosine levels are known to be involved in tissue protection during ischaemic or inflammatory events, we first tested a potential increase in intracellular adenosine by blocking ADK through the addition of ITU to pulmonary epithelial (A549) cells. Eight hours after incubation with either 25 μM or 10 μM ITU, the intracellular adenosine levels increased approximately 3-fold compared with the starting adenosine level (Figure 4A). Compared with the controls, the intracellular adenosine concentration was approximately 2-fold higher 8 hours after blocking ADK with ITU. To determine the impact of adenosine uptake after adenosine enhancement as a result of the ITU-mediated inhibition of intracellular conversion of adenosine, we added adenosine to the extracellular space and detected the changes in extracellular adenosine concentration over time. We used untreated cells as a negative control and treatment with the adenosine uptake inhibitor dipyridamole as a positive control. We found that the loss of extracellular adenosine from the supernatant was significantly attenuated after ADK inhibition by ITU. The extracellular adenosine levels in all ITU treated samples were elevated in a concentration dependent manner, but these levels were influenced by ENT-dependent adenosine uptake inhibition with dipyridamole (Figure 4B).
A hallmark of inflammation is the recruitment of PMNs to injured tissue, where they cross the endothelial and epithelial barrier in a transmigration process (19). Therefore, we sought to evaluate the influence of ADK inhibition on this process. For this, we used a transepithelial migration assay that was previously described (18). We found a significant reduction in the transepithelial migration of PMNs when they were treated with 25 μM ITU (Figure 5A). We were also able to detect a reduction in paracellular permeability when testing FITC-Dextran flux following preincubation of the epithelium with ITU (Figure 5B).
As homozygous ADK deficient mice are not viable, we exposed heterozygous ADK+/− mice to a model of lung injury induced through LPS inhalation. Before the start of the experiment, we validated the diminished ADK expression in several organs of these mice (Supplemental Figure E43 A and B) (15). Following exposure to LPS, ADK+/− mice had significantly reduced cell infiltration, total protein content and MPO amount in bronchoalveolar lavage fluid (BALF) compared with WT controls (Figure 6A–C). Diminished TNF-α, IL-1β and IL-6 cytokine levels corroborated our findings of an attenuated inflammatory response in ADK+/− mice (Figure 6D–F). The decreased inflammation in ADK+/− mice was also associated with the decreased infiltration of PMNs into pulmonary tissue compared with WT controls (Figure 6G, Supplemental Figure E5). We found an increased expression of the adenosine receptor 2B in these animals (Supplemental Figure E6).
Encouraged by the detected lung tissue protection in ADK+/− mice, we evaluated the pharmacological inhibition of ADK in WT animals as a potential therapeutic approach to ALI. For this, we administered the specific ADK inhibitor ITU via inhalation over a period of 30 minutes at concentrations of 0.1 mg/kg ( 3.33 μg/kg·min) and 0.01 mg/kg ( 0.33 μg/kg·min) immediately after NaCl or LPS inhalation. As shown in the in vitro experiments using lung epithelial cells, ITU blocks ADK; as a result, the extracellular adenosine levels increase. Following the inhalation of ITU, we found significantly attenuated pulmonary inflammatory changes (Figure 7A–C). Low cytokine levels of TNF-α, IL-1β and IL-6 in the BALF confirmed that ITU was protective during pulmonary inflammation (Figure 7D–F). Reduced numbers of neutrophils infiltrating into the lung tissue were confirmed by immunohistochemistry (Figure 7G). Taken together, our findings provide evidence that the pharmacological inhibition of ADK decreases the pulmonary damage in ALI.
To date, lung injury significantly contributes to in-hospital mortality (20). A crucial component of lung injury is the loss of alveolar-capillary function, which results in the trafficking of neutrophils from the vascular space into the alveolar space and in the development of tissue oedema (21). Based on previous observations that adenosine kinase has the potential to protect barrier function during hypoxia, we evaluated the role of ADK during ALI (22). We report here that ADK is repressed during inflammation both in vitro and in vivo. This repression translates into a protective response during pulmonary inflammation, as observed in animal experiments involving genetic repression of ADK. Moreover, pharmacological studies using ITU have demonstrated that the in vivo inhibition of ADK attenuates tissue injury during pulmonary inflammation (for mechanistic insight see Figure 8).
In the initial experiments, we observed both the transcriptional and translational repression of ADK during an acute inflammatory process in vitro. This observation is in line with our previous finding that ADK repression was induced in endothelial cells and intestinal epithelial cells during tissue hypoxia (22). To gain more insight into the regulation of ADK, we focused on NF-κB because it is one of the transcription factors that initiates inflammation (23). For this approach, we first analysed the promoter region of ADK and found one putative response element region for NF-κB binding in the promoter region. As proof of concept, we first checked for NF-κB binding during in vitro inflammatory conditions in lung epithelial cells, by using chromatin immunoprecipitation. Our results indicated strong binding of NF-κB to the putative ADK promoter. Following this, we measured the expression level of ADK in vivo. After exposure to LPS, we also observed the repression of ADK in vivo. To further investigate the functional effects of the ADK repression, we evaluated ADK+/− mice, which had significant repression of ADK within the pulmonary tissue. We exposed these animals to LPS inhalation and observed significantly reduced numbers of infiltrating cells and levels of cytokines and other inflammatory markers in BALF from these animals. This observation is in line with the rapid increase in extracellular adenosine and the transcriptional repression of ADK during hypoxia (22, 24), and it might be a protective adaptation to the detrimental effects of pulmonary inflammation. The ADK repression increases the extracellular adenosine concentration, which has been reported to protect tissue in several conditions (25). Extracellulae adenosine signals through its known adenosine receptors (A1, A2A, A2B, A3) to ameliorate inflammation (26). We have demonstrated here that especially the adenosine 2b receptor is induced in our model of lung injury. This is in line with previous work showing the protective potential of the A2B adenosine recptor (27). The protective role of high extracellular adenosine levels and adenosine receptor signalling were also described during myocardial reperfusion injury, several neurological conditions, arthritis and colitis (26, 28–32). To the best of our knowledge, repression of ADK during inflammatory tissue conditions, as present during lung injury, has not been shown before. This observation is novel and describes and endogenous anti-inflammmtory adaptation to an acute inflammatory process.
To test the therapeutic potential of ADK inhibition for treating lung injury, we evaluated the pharmacological inhibition of this enzyme. The intracellular enzyme ADK catalyses the conversion of adenosine and ATP to AMP and ADP; as such, it plays a key role in controlling the cellular concentration of adenosine (33). Consequently, an increase in the intracellular adenosine levels as a result of ADK inhibition enhances the extracellular adenosine levels (7). To test this, we first evaluated the inhibition of ADK in vitro in epithelial cells by using ITU. In this experiment, we found a significant increase in the extracellular adenosine levels by inhibiting intracellular ADK. To evaluate the potential functional consequences we next investigated whether the increased extracellular adenosine levels, which are mediated through ADK inhibition by ITU, interfere with the migration of neutrophils across the epithelial barrier. We observed improved epithelial barrier function, as well as diminished neutrophil migration, in response to ADK inhibition. To test whether ADK inhibition has a therapeutic effect in LPS-induced acute lung injury, and whether it dampens the inflammatory response to LPS inhalation, we performed in vivo experiments using 5-iodotubercidin (ITU) administered at two concentrations after the onset of inflammation. We identified a strong inhibition of inflammation at the low concentration of ITU (0.01 mg/kg), as well as the diminished infiltration of cells, and lower protein and MPO content and lower proinflammatory cytokine levels in the BALF. Other groups found similar tissue protection and decreased inflammation after the administration of ADK inhibitors. During experimental sepsis the adenosine kinase inhibitor GP-1-515 showed to reduce the infiltration of neutrophils and the cytokine production, this was associated with improved outcome of experimental animals (34). These results were confirmed in another study of using a model of paw inflammation, where again GP-1-515 demonstrated anti-inflammatory potential through adenosine kinase inhibition (35). We used the pharmacological agent ITU here in this study through inhalation to test a lung targeted intervention with ITU which also has a low IC50, that is comparable to GP-1-515. As such our results are in parts comparable to these previous studies, although genetic deletion of adenosine kinase and in vitro models for the role of ADK were not included in this work done previously. Further studies have also confirmed our findings and have shown that the tissue protective potential of ADK inhibition is also present during cerebral stroke (36, 37) and myocardial ischaemia reperfusion injury (38). As such, our findings extend the previously known protective role of ADK inhibition to the field of pulmonary inflammation.
In summary, the results of our study are in line with previous investigations that demonstrated the tissue protective properties of extracellular adenosine. In this study, inflammation promoted the transcriptional and translational repression of ADK both in vitro and in vivo. The active inhibition of ADK by ITU increased adenosine levels. The naturally initiated process of ADK repression seems to be a protective tissue adaptation to an inflammatory insult. Therefore, this study improves our understanding of the mechanisms underlying the regulation of extracellular adenosine by ADK and extends our understanding of the role of ADK during lung injury.
We gratefully acknowledge Dr. Claudia Bernardo de Oliveira Franz, Michaela Hoch-Gutbrod, Edgar Hoffmann and Alice Mager for excellent technical assistance.
Grant Support: This work was partially supported by a Grant from the Deutsche Forschungsgemeinschaft (DFG) grant DFG-MO 2252/1 to D.K and by a grant from the Deutsche Forschungsgemeinschaft (DFG) DFG-RO 3671/6-1 to P.R
Copyright form disclosures:
Dr. Boison received royalties from Springer (Book on Adenosine), is employed by Legacy Research Institute, and received support for article research from the National Institutes of Health (NIH). His institution received grant support from the NIH, US Department of Defense, and CURE Foundation. The remaining authors have disclosed that they do not have any potential conflicts of interest.
Conflict of interest: None.
Author Contribution Statement:David Köhler: designed research, performed experiments, analysed data, and wrote the manuscript
Ariane Streißenberger: designed research, performed experiments, and analysed data
Julio C. Morote-García: designed research, performed experiments, and analysed data
Mariella Schneider: performed experiments and analysed data
Tiago Granja: performed experiments
Andreas Straub: wrote the manuscript
Detlef Boison: checked the manuscript and provided animals
Peter Rosenberger: designed research and wrote the manuscript