Potential regulatory role of miRNA and mRNA link to metabolism affected by chronic intermittent hypoxia

Abstract

Aim: Intermittent hypoxia (IH) is the prominent feature of obstructive sleep apnea (OSA) pathophysiology, which is an in dependent risk factor of cardiovascular complications. The effects of IH on adipocyte metabolism were explored by high-throughput sequencing technology. Methods: Plasma was collected from OSA patients and control group to perform mRNA sequencing. 3T3-L1 cells were differentiated into adipocytes then subjected to a 5%-21% O₂ hypoxic environment (IH) for 24 h. High-throughput sequencing method was used to determine differential mRNA and miRNA patterns in fat cells exposed to IH. We then performed Gene Ontology (GO) analysis, identified relevant KEGG pathways and miRNA-target-pathways. Results: Sequencing data showed that OSA affected the expression of 343 mRNAs in the plasma. At the same time, we found that IH affected the expression of 3034 mRNAs in the adipocytes. In addition, 68 differentially expressed mRNAs were overlapped in plasma from OSA patient and IH-induced adipocyte model. We observe that 68 differential genes could be connected to 49 reciprocally expressed miRNAs. We showed that IH significantly reduced the expression of miR-182-5p and miR-30c-2-3p. KEGG predicted that the function of expressed miR-182-5p and miR-30c-2-3p was enriched to AKT signaling pathway. Notably, IH activated PI3K/AKT pathway in fat cells. Conclusion: Our results demonstrated that IH might induce adipocyte metabolism by regulating miR-182-5p and miR-30c-2-3p.

Keywords: adipocyte; chronic intermittent hypoxia; mRNA; miRNA; obstructive sleep apnea.

FIGURE 1

mRNA expression profile in plasma from OSA patients. (A) Differential expression of mRNA; (B). Upregulated genes. (C) Downregulated genes. (D) Upregulated significant pathways. (E) Downregulated significant pathways. The y-axis shows the GO category and the x-axis shows a negative log-transformed p-value -lgP where a larger -lgP indicates a smaller p-value.

FIGURE 2

IH-exposed adipocyte gene expression profile. (A) Differential expression of mRNA; (B). Upregulated genes. (C) Downregulated genes. (D) Upregulated significant pathways. (E) Downregulated significant pathways. The y-axis shows the GO category and the x-axis shows a negative log-transformed p-value -lgP where a larger -lgP indicates a smaller p-value.

FIGURE 3

miRNA-Gene-Network and miRNA-Pathway-Network. (A) Venn diagram showed the number of overlapped genes of differentially expressed in adipocyte model and plasma from OSA patient. (B) Heatmap showed the 49 differentially expressed miRNAs related to expressed mRNAs by miRNA sequence analysis. (C,D) miRNAs are represented by Box nodes, and the predicted target genes/target pathways are represented by cycle nodes. The inhibitory effect of miRNA on its predicted targets is showed by Edges. The contribution of one miRNA to the surrounding genes or the contribution of one gene to the surrounding miRNAs are showed by Degrees. The miRNAs and genes/pathways always have the largest degrees in the network. (E,F) qRT-PCR analysis of key miRNAs in the effect of IH (**p value < 0.05).

FIGURE 4

Bioinformatics analysis of verified miRNAs and the effect from IH-exposed fat cells on pathway. (A) The overlapped predicted pathways of mir-182-5p and mir-30c-2-3p is showed by venn diagram. (B) KEGG Pathway analysis of 2 differentially expressed miRNAs. (C) The protein expression of p-Akt in the effect of IH (**p value < 0.05).

FIGURE 5

Both miR-182-5p and miR-30c-2-3p are involved in proliferation and metabolism of adipocyte via PI3K-AKT pathway. The levels of miR-30c-2-3p (A) and miR-182-5p (B) in adipocytes. (C,D) CCK8 assay demonstrated that miR-182-5p and miR-30c-2-3p overexpression promoted adipocytes growth and viability. (E) Concentrations of adiponectin from the culture supernatant in adipocytes were determined using ELISA kits. (F) The protein expression of p-Akt in the IH-induced adipocytes of transfecting miR-182-5p and miR-30c-2-3p mimic. n = 4–10/group, **p < 0.01 vs. IH group.