Bioinformatics insights into acute lung injury/acute respiratory distress syndrome

There are several publications reviewed the application of genomics in the study of ALI. In 2002, Marshall and his colleagues provide the first evidence of a genetic influence in ARDS, suggesting an important role for angiotensin converting enzyme (ACE) [16]. In this study, a total of 358 patients were enrolled including 96 patients fulfilling American/European Consensus Committee criteria for ARDS, 174 coronary artery bypass graft (CABG) controls and 88 non-ARDS but require mechanical ventilation patients in the intensive care unit (ICU). DNA was extracted from whole blood samples and ACE genotype was determined by three-primer polymerase chain reaction amplification. This study detected that the frequency of the DD genotype and the D allele was markedly increased in patients with ARDS compared with both the CABG and ICU comparison groups and the healthy population sample [16].

Similarly, Gong, et al demonstrated that -308GA promoter polymorphism in the TNF-a gene was associated with the mortality of ARDS [27]. SNPs of 43 Toll like receptor-related genes were identified and one of these genes was demonstrated to be associated with susceptibility to organ dysfunction and death in patients with sepsis [28]. Later, pre–B-cell colony-enhancing factor (PBEF) was found significantly increased in the expression profiling in animal models of ALI and in human ALI. In this study, multiple technologies including PCR, microarray analysis and some proteomic measurement methods were used. This combination of various technologies in bioinformatics study is important and helpful [29].

However, different from these traditional epidemiological studies in at risk populations, Marzec.et al first applied positional cloning and gene sequencing to identify candidate susceptibility genes and SNP polymorphism. They demonstrated that the -617 A SNP of Nrf2 (NF-E2 related factor 2) had a significantly higher risk for developing ALI after major trauma [30]. Then a large-scale candidate gene genotypings were carried out in a cohort of critically ill subjects with trauma to identify ALI risk variants after major trauma. Five SNPs were found to have a significant association with ALI, of which two SNPs in ANGPT2 (rs1868554 and rs2442598) replicated the significant association with ALI. And furthermore, the top associated SNPs were tested in a multicenter case-control population [31]. In a two-stage discovery/validation study, the commonly occurring FAS haplotype including FAS21541T and FAS9325A was significantly associated with ALI susceptibility and increased FAS mRNA expression in peripheral leukocytes in response to innate immune stimulation [32]. A complete summary of all published applications of genetics in the study of ALI/ARDS is far beyond the scope of this review. Table 1 provides a summary, with references, of some of the most recently or interesting applications of genetics in the diagnosis, susceptibility or prognosis of ALI/ARDS [31‐37].

Among these identified genes, some of them were well demonstrated to play critical roles in the development of ALI/ARDS. NADPH oxidase contributed to the production of reactive oxygen species, which is able to induce a cell death response in oxidative stress. NADPH oxidase 1 was reported to act as an important mediator of acute lung injury mediated by oxidative stress, particularly in alveolar cells. NOX1-generated ROS are largely responsible for alveolar cell death and subsequent lung injury through JNK and ERK pathways [38]. Angiopoietin-2 (ANGPT2) has been implicated in pulmonary vascular leak syndromes including ALI and sepsis in both animal and human studies [39, 40], while recently, two ANGPT2 SNPs (rs2442598 and rs1868554) were found to be strongly associated with the development of ALI in patients with major trauma [31]. However, studies are warranted to further elucidate the characterization of ANGPT2 genetic variation and expression. Fas pathway has been investigated as a potential contributor to the inflammation and alveolar epithelial cell apoptosis observed in the lungs of patients with ALI [41, 42]. SP-B enhances the rate of phospholipid absorption to the alveolar air/water interface [43], decreases surface tension by interfering with the attractive forces acting between water molecules [44] and has anti-inflammatory properties [45]. However, the others were seldom reported and futher studies are needed to validate their functions in ALI/ARDS.

Gene expression profiles represent only a small fraction of the complexity in an organism. On the other hand, gene expression data obtained by the microarray analysis are usually moderately correlated with the expression levels of corresponding protein products due to the post-transcriptional regulation. Thus, extrapolation of protein expression on the basis of microarray-based gene expression measurements must be validated by direct measurement of the protein [15]. Then proteomics comes.

Proteomics captures a nearly comprehensive set of expressed proteins in a cell or organism, which determine how a cell or organism functions. Proteomics become more practical and feasible for researchers in the field of pulmonary medicine using BALF, lung tissue, and exhaled breath condensates as the source of protein analysis [46]. In 2003, Bowler et al first applied the proteomic approach in the BALF, plasma and lung edema fluid in ALI/ARDS. In this paper, the author reported that although many of the proteins were demonstrated to have increased or decreased relative intensity in the plasma and EF of ALI patients compared with normal subjects, a limitation to the 2-DE proteomics approach is that it is not easy to quantify all changes in relative protein expression among a large number of samples. The author mentioned that the proteomics approach may be complementary to microarray studies [47].

Shotgun proteomics in BALF was analyzed and 870 different proteins were identified, including surfactant, proteases, and serum proteins. However, because of the several limitations of the 2DE method to detect particular classes of proteins [48], it can only be treated as an excellent screening tool to initially characterize a sample of mostly unknown protein composition, and additional methods, including isolation of subpopulations of proteins and further refinement of the methodologies must be cooperated to identify specific proteins [49]. Then, through the combined application of SELDI-TOF, 2-DE and western blot analysis, apolipoprotein A1, and S100 calcium-binding proteins A8 and A9 were detected to increase in the inflammatory BALF [50].

Following these studies, quantitative proteomic analysis was then used to profile the changes in protein expression in the lungs at the onset and during the course of acute lung injury. Computational analysis as used to map complex protein interactions in the lung fluids and study how these interactions changed during the course of ARDS. This approach to protein network analysis identified novel mediators of acute lung injury and protein pathways were redundant and involved in multiple biological processes [51].

Table 2 provides a summary of the proteomic study in ALI/ARDS from 2009-2011 [52‐58]. Caveolin-1 is involved in PMN adhesion, chemotaxis, and epithelial and endothelial cell apoptosis/senescence which accelerated lung injury in the early stage. However, cav-1 may be beneficial in the late stage through anti-fibrosis [59]. MMPs are a family of zinc-dependent proteinases which expressed by all cell types relevant to ARDS pathogenesis including alveolar epithelial cells, PMNs, macrophages, and fibroblasts. Many studies implicate MMPs in the injurious process as well ALI/ARDS, which may mainly contribute to its capability of degrading all components of the extracellular matrix [60, 61]. Vascular endothelial growth factor (VEGF) and its receptors are critical in the regulation of both vascular permeability and endothelial cell survival. However, the role of VEGF and related molecules in ALI/ARDS is controversial [62]. PAI-1 is an endogenous inhibitor of the plasminogen activators. Studies reported that elevated level of PAI-1 is an independent risk factor for mortality and adverse clinical outcomes for ALI patient [63]. RAGE is a multi-ligand-binding receptor that can bind advanced glycation end products, amyloid β-peptide, S100 proteins, and high-mobility group box-1. RAGE-ligand interaction results in intracellular signaling, which leads to activation of the proinflammatory transcription factor nuclear factor-κB (NF-κB) [64, 65]. Study indicated that RAGE may be useful as a biological marker of alveolar type I cell injury [66]. TNF-alpha and IL-8 are also well-validated inflammatory cytokines in ALI/ARDS [57]. Figure 2 integrated these factors together to show their potential roles in the pathogenesis of ALI/ARDS.

Although the amplification of quantitative metabolomics in clinic bioinformaticsis in its infancy, it has been increasingly used recently. It is a unique tool of discovery because it identifies multiple endogenous markers, and it changes over time which can be predictive of clinical outcome and therapeutic response [67, 68]. The ability of magnetic resonance imaging (MRI) and metabolic NMR were investigated to detect and quantify inflammation-mediated lung injury [69]. Recently, proton nuclear magnetic resonance ⁽¹ H-NMR) based quantitative metabolomics with subsequent computational analysis was applied to investigate the plasma metabolic changes associated with sepsis-induced ALI. Several metabolites including glutathione, adenosine, phosphatidylserine, and sphingomyelin were found different between patients with ALI and healthy subjects. Additionally, computational network analysis was carried out to investigate a distinct metabolic pathway for each metabolite [24]. Moreover, there is another case-control study conducted on plasma samples from six albumin-treated and six saline-treated patients to measure the metabolic changes after albumin treatment. This study showed that blood albumin levels normalized and oxygenation improved with albumin treatment compared to saline treatment. In this study, 1 H-NMR spectroscopy was introduced to characterize systemic metabolic responses and identify patients’ unique response to pharmacologic therapy and appealed to play a potential role in the guiding treatment and reducing healthcare cost [68].

In summary, ALI/ARDS is a complex disease which can be initiated by a diverse array of precipitating factors and involving the interplay of both the environment and genetic factors. However, numerous environmental agents can induce common pathological changes, including damage to endothelial and epithelial cells, infiltration of inflammatory cells, cytokine release and so on. Those common biological pathways may be controlling generalized responses to injury and repair which cannot distinguish the type of lung injury across different models. On the other hand, susceptibility to ALI/ARDS is also different among patients with same conditions. Clinical models have identified risk factors and laboratory testing to evaluate severity, guide clinical therapy and tell prognosis. However, these strategies remain imprecise. The application of bioinformatics enables us to identify the individual expression patterns as well as the injury-specific pattern, thus may facilitate therapy and predict prognosis. Besides, in the absence of significant insights into disease pathogenesis, comprehensive bioinformatics technologies may help us to get deep understanding of such complex and life-threatening illness thus may provide novel therapeutic targets. Finally, there are several available samples that can be used in the bioinformatics study of the lung, such as exhaled breath gas, blood, lung tissue, pleural effusion and BALF, which may largely facilitate the biomarker research.