Identification of significant alteration genes, pathways and TFs induced by LPS in ARDS via bioinformatical analysis

Acute respiratory distress syndrome (ARDS) or acute lung injury (ALI) is an acute hypoxemic respiratory failure, characterized by lung tissue oedema and injury, inflammatory responses, and compromised gas exchange following macrophage activation, surfactant dysfunction, and epithelial destruction [1, 2]. It has been widely recognized as a clinical problem worldwide, accompanied by high morbidity and mortality [3, 4]. According to a recent international multi-centre research in 50 countries, the prevalence of ARDS was 10.4% of ICU admissions [5]. According to a cross-sectional study, the mortality of ARDS from 2012 to 2013 in Chinese 20 ICUs is about 34% [6]. Especially in the last 2 years, the incidence and mortality of COVID-19-associated ARDS have worsened the outcome. For the patients of COVID-19 ARDS in ICU, mortality ranged between 26 and 61.5%, especially for those who received mechanical ventilation, the mortality ranged between 65.7 and 94% [7]. Despite a variety of basic and clinical research held, there is still no effective pharmacotherapy for it. Currently, the treatment remains primarily with ventilation and conservative fluid management. Therefore, it is critical to study ARDS’s pathogenesis and explore specific biomarkers for this condition.

The development process of ARDS is complicated, and the specific mechanism is not yet fully understood. Multiple studies have confirmed that ARDS is related to the damage and disruption of the epithelial and endothelial cells, as well as dysregulated inflammation [8‐10]. A breakdown in endothelial junctions or the injury of endothelial cells can aggravate lung vascular permeability. Cheng et al. [11] reported that the severe endothelial pyroptosis caused by bacterial endotoxin lipopolysaccharide (LPS) was mediated by the inflammatory caspases. IL-1β could impair CREB-mediated VE-cadherin transcription to induce endothelial injury [12]. Besides, lung epithelial permeability alteration is also an important factor in ARDS pathogenesis. Short et al. [13] showed that the alveolar barrier could be damaged by influenza by disrupting epithelial cell tight junctions, specifically with loss of tight junction protein claudin-4. Related gene mutations or expression alterations in the ARDS might be suitable to serve as diagnostic or therapeutic targets. Microarrays have been used to quantify the high-throughput expression of genes for many species quickly [14]. As a result, the data produced from microarrays were stored in some public databases. We could explore lots of valuable clues from these raw data for further experimental research. Some different bioinformatic studies have been exploited in the past few years, which provided us with abundant integrated bioinformatical methods for studies [15].

To identify the better potential diagnosis or therapeutic targets for ARDS, firstly, we performed a transcriptome analysis of mice lung tissues. The tissues were treated with LPS and the raw data was acquired from Gene Expression Omnibus (GEO) to explore differentially expressed genes (DEGs) and pathways. Once characterized hub genes, we could evaluate their expression in human lung tissues. Finally, the hub DEGs above mentioned were processed further to find the common TFs.

In the present study, a total of 39 genes related to ARDS were identified. GBP2, IFIT1 and IFIT3 were identified as common hub genes. They were mainly involved in Salmonella infection, cytokine–cytokine receptor interaction, TNF signalling pathway, Toll-like receptor signalling pathway and so on and were confirmed expressed in varied organs, including the lung tissue. STAT1, E2F1, IRF1, IRF2, and IRF9 were identified as the main TFs and were predicted to regulate these hub genes in ARDS.

IFIT1 and IFIT3 belong to the interferon-induced protein with the tetratricopeptide repeats (IFIT) protein family. They are involved in regulating immune responses and restrict viral infections through a variety of mechanisms, including the restriction of viral RNA translation [27]. Recent studies showed that IFIT3 could modulate IFIT1 RNA Binding specificity and protein stability [28, 29]. Xu et al. reported that IFIT3 transcription was dependent on NF-κB activation [30], while NF-κB played a vital role in ARDS [31, 32]. Exome-wide analysis showed that IFIT3 mutation was associated with COPD and airflow limitation [33]. All of them suggested that IFIT1 and IFIT3 mutation might be involved in the occurrence of ARDS and supported our hypothesis.

According to our findings, STAT1, E2F1, IRF1, IRF2, and IRF9 were screened as TFs according to iRegulon, which is the plugin of Cytoscape. STAT1 is a member of the STAT family of 7 cytoplasmic proteins. It had essential effects on innate immunity via defending the host from different infections [34]. Sevoflurane could reduce LPS-induced ARDS via modulating STAT1 [35]. In hepatocellular carcinoma, IFIT3 could bind signal transducer and activator of transcription 1 (STAT1) and STAT2 to enhance STAT1–STAT2 heterodimerization and nuclear translocation upon IFN-α treatment, thus promoting IFN-α effector signalling [36]. It suggested that the interaction between STAT1 and IFIT3 might play a significant role in ARDS progression. IRF1, IRF2, and IRF9 belong to the interferon regulatory factor (IRF) family. IRF-1 deficiency played a key role in the classical ROS-dependent release of NETs, which might serve as a novel target in ARDS [37]. In the recent COVID-19 studies, NETs contributed to COVID-19 related ARDS [38, 39] by contributing to excessive thrombosis. It suggested that IRF-1 might play a role in COVID-19 related ARDS. Wang et al. reported that LncRNA XIST could aggravate LPS-induced ARDS in mice by upregulating IRF2 [40]. The above reports were entirely consistent with our findings.

Despite our findings supported by some studies, we did not conduct further animal experiments and clinical data analysis to verify it. It gave us a hint for further study direction. Next, we will implement some animal experiments to develop more sensitive biomarkers and drugs, followed by some related clinical trials.

In summary, our bioinformatics analysis study identified three DEGs (GBP2, IFIT1 and IFIT3) in ARDS pulmonary tissues according to two different microarray datasets (GSE2411 and GSE130936). Results suggested that these three genes could be targets for the study of ARDS, and might be regulated by TFs, STAT1, E2F1, IRF1, IRF2, or IRF9. Anyway, these predictions would be verified by a series of experiments in the future. These studies have opened up new research directions for the diagnosis and treatment of ARDS.

In this review, we used a bioinformatics approach to help explore genes, pathways and related TFs that might be involved in the occurrence of ARDS. We speculated GBP2, IFIT1 and IFIT3 might play important roles. The findings might provide novel insights into the development of promising targets for the diagnosis and treatment of ARDS in the future.