3.4 Immune and inflammation-related miRNA-mRNA regulatory network
The empirically validated mouse PPI interactions in the IntAct, MIPS, and MINT databases were collected, and a total of 4,853 PPI interactions were obtained after the union of these three sets. A total of 544 verified miRNA-mRNA pairs data were collected from TarBase. Furthermore, the 3827 miRNA-mRNA pairs obtained in Section 3.3 were integrated with the 4853 and 544 regulatory relationships to obtain a global miRNA-mRNA regulatory network (Figure 4A). Remarkably, among the differentially expressed genes, there were 117 genes for which the GO function or the GO function of the parent node was related to the immune system process. Then, based on the global miRNA-mRNA regulatory network, we further selected edges that contained at least one edge of these 117 genes from the global network (Figure 4A) to construct an immune and inflammation-related network. This network had 595 edges and was further divided into three subnetworks. The first subnetwork had 479 edges (Figure 4B), the second subnetwork had 115 edges (Figure 4C), and the third subnetwork had only 1 edge (Protein interaction between H2-Aa and H2 Ab1).
Interestingly, most miRNAs in the first subnetwork were upregulated, whereas most miRNAs in the second subnetwork were downregulated. Further enrichment analysis was performed on 106 genes in the first subnetwork. These 106 immune-related genes were largely involved in positively regulating the immune system: position regulation of NF kappaB transcription, position regulation of natural killer cell mediated cytotoxicity, position regulation of T cell promotion, and position regulation of interleukin-6 production (E-Table 9). There were 51 miRNAs involved in this subnetwork, including miR-21, let-7f, mir-431, and mir-205.
Functional enrichment analysis of 38 genes in the second subnetwork revealed that these 38 immune-related genes were involved in negatively regulating the immune system: negative regulation of cell promotion, negative regulation of plasma activation, negative regulation of response to stimulus, and negative regulation of immune system process (E-Table 10). However, these genes played a positive role in vascular proliferation and oxygen metabolism: positive regulation of angiogenesis and positive regulation of oxygen and reactive oxygen species metabolic process (E-Table 10). There were 30 miRNAs involved in this subnetwork, including mir-29b, mir-295 and mir-290-5p.
DISCUSSION
Neonatal hyperoxia not only disrupts lung development, but also reprograms key immunoregulatory molecules in the lung.29 However, at present, neonatal immunology is a poorly researched field, with only a handful of studies devoted to the pulmonary immune response;3 thus, our report considerably expands knowledge in this field. In this study, we performed traditional bioinformatics methods to analyze the raw data and to identify miRNA-mRNA regulatory networks associated with immune and inflammation processes, which play key roles in the development of BPD.
The downregulated genes were mainly related with immune response and inflammatory processes, which emphasized their roles in BPD pathogenesis. There are studies indicate that the classic proinflammatory cytokines, including Il1α, Il8, Il1β and TNF-α are overexpressed during the saccular stage in neonates developing BPD, and their overexpression significantly disrupts lung development, both structurally and functionally.3,4,8,30,31 While there are also reports demonstrate that some parameters of the early inflammatory response (neutrophils, cytokines such as Il1, TNF-α) may not be detectable after days to weeks of exposure to noxious stimuli; they have already initiated the signaling pathways of immune and inflammatory processes that can affect the BPD lung development.30,32 The seemingly contradictory results may be due to the different stages of lung development or related with the complex inflammatory signal pathways during BPD pathogenesis. Therefore, a better understanding of the mechanisms underlying the dynamic immune and inflammatory responses during BPD development could reveal potential therapeutic targets to attenuate lung injury.
The upregulated genes were mainly related with ECM remodeling, and these data indicated that abnormal ECM remodeling was implicated in the pathogenesis of BPD, which is consistent with other reports.33,34 Our previous studies also found that excessive production and accumulation of collagen appeared around the airways and blood vessels and in the alveolar regions of BPD mice, and a number of genes associated with ECM remodeling and fibrosis, such as fibronectin 1(FN1) and collagen 1α, were significantly dysregulated with the development of BPD,33,34 these observations were also consistent with those of our current study. Moreover, these ECM components are believed to be involved in inflammation and immune-related signal pathways,33-35 which supports the notion that the ECM process is central to the development of BPD.
Based on the miRNA-mRNA co-expression networks associated with immune and inflammation processes, we have found many important and interesting interactions between miRNAs and mRNAs. Some targeted genes are regulated by a single miRNA, whereas others can be regulated by multiple miRNAs which can interact with each other and form an interesting regulatory network. Of note, miR-205 was the most highly upregulated miRNA with 39 candidate targeted genes. These genes (such as H2-D1, Sh2d1a, Fcer2a, Lbp, Il18) are mostly enriched in immunity-related functions: positive regulation of lymphocyte mediated immunity, positive regulation of lymphocyte mediated immunity, positive regulation of immune response, positive regulation of lymphocyte mediated immunity, and induction of apoptosis. For example, Lbp has been investigated as a biomarker for inflammation.36 It is an acute-phase protein essential for the response to bacterial lipopolysaccharides or endotoxins in gram-negative bacteria, and can also increase its expression in response to subclinical infections.36 Taking miR-21 as another example, we found that it had 42 candidate targeted genes, and the functions of these genes were also related to immunity and inflammation, such as Il12a,37 Cr2,38Lbp,36 and Fasl.39 Furthermore, these targeted genes may also be regulated by other miRNAs, such as miR-205 and miR-155,18,40 or may be involved in other inflammatory pathways through signal interaction. Thus, these miRNAs and mRNAs construct regulatory networks associated with immune and inflammation processes by interacting with each other directly or indirectly. Further research will be needed to validate these miRNA-mRNA regulatory networks in detail, and we believe this will lead to additional significant findings.
In summary, this study characterized the miRNA and mRNA expression profiles of lungs under normoxic and hyperoxic conditions, providing insights into the immune and inflammatory responses to hyperoxia exposure during late BPD stages. Our findings further emphasize that the immune and inflammation processes play a significant role in BPD development and may provide relevant information for the development of new biomarkers and therapeutic targets not only to treat BPD, but also to facilitate normal adaptive immune responses, which will have broader implications in the management of premature infants.