The potential roles of m6A modification in regulating the inflammatory response in microglia

Abstract

Background: Microglia are key regulators of the inflammatory response in the brain. Adenosine in RNAs can be converted to m⁶A (N⁶-methyladenosine), which regulates RNA metabolism and functions as a key epitranscriptomic modification. The m⁶A modification pattern and m⁶A-related signatures under pro-inflammatory and anti-inflammatory conditions of microglia remain unclear.

Methods: Primary rat microglia were differentiated into pro-inflammatory M1-like (M1-L), anti-inflammatory M2-like (M2-L), and resting, unstimulated (M0-L) phenotypes. m⁶A mRNA and lncRNA epitranscriptomic microarray analyses were performed, and pathway analysis was conducted to understand the functional implications of m⁶A methylation in mRNAs and lncRNAs. The m⁶A methylation level and gene expression of mRNAs and lncRNAs were subsequently verified by m⁶A Me-RIP and qRT-PCR.

Results: A total of 1588 mRNAs and 340 lncRNAs, 315 mRNAs and 38 lncRNAs, and 521 mRNAs and 244 lncRNAs were differentially m⁶A methylated between M1-L and M0-L (M1-L/M0-L), M2-L and M0-L (M2-L/M0-L), M2-L and M1-L (M2-L/M1-L), respectively. Furthermore, 4902 mRNAs, 4676 mRNAs, and 5095 mRNAs were identified distinctively expressed in M1-L/M0-L, M2-L/M0-L, and M2-L/M1-L, respectively. Pathway analysis of differentially m⁶A methylated mRNAs and lncRNAs in M1-L/M0-L identified immune system, signal transduction, and protein degradation processes. In contrast, the distinct m⁶A methylated mRNAs in M2-L/M0-L were involved in genetic information processing, metabolism, cellular processes, and neurodegenerative disease-related pathways. We validated m⁶A methylation and the expression levels of five mRNAs and five lncRNAs, which were involved in upregulated pathways in M1-L/M0-L, and five mRNAs involved in upregulated pathways in M2-L/M0-L.

Conclusions: These findings identify a distinct m⁶A epitranscriptome in microglia, and which may serve as novel and useful regulator during pro-inflammatory and anti-inflammatory response of microglia.

Keywords: Bioinformatics analysis; Inflammatory response; Methylated m6A RNA immunoprecipitation; Microarray; Microglia; Phenotype; Polarization; m6A lncRNA; m6A mRNA.

Fig. 1

mRNA and lncRNA m⁶A modification profile changes in different primary rat microglia phenotypes. Hierarchical clustering of all samples revealed the non-random partitioning of samples into three major groups: M1-L vs M0-L; M2-L vs M0-L; and M2-L vs M1-L. Each column represents one sample and each row represents one mRNA (A) or lncRNA probe set (B)

Fig. 2

mRNA expression analysis of M0-L, M1-L, and M2-L phenotypes in microglia. A Volcano plot analysis of 3627 upregulated and 1275 downregulated mRNAs (M1-L vs M0-L, P < 0.05); 4360 upregulated and 316 downregulated mRNAs (M2-L vs M0-L, P < 0.05); 1896 upregulated and 3199 downregulated mRNAs (M2-L vs M1-L, P < 0.05). Red boxes represent ≥ 1.5-fold change difference, P < 0.05. Green boxes represent ≤ 0.7-fold change difference, P < 0.05. B Expression of markers specific to different brain and immune cell types in M0-L, M1-L, and M2-L samples. Data represent mean ± SD of three biological replicates. C Association analysis of m⁶A methylation and mRNA expression: M1-L vs M0-L: 39 Hyper-Up mRNAs, 319 Hyper-Down mRNAs, 515 Hypo-Up mRNAs, 138 Hypo-Down mRNAs (P < 0.05); M2-L vs M0-L: 26 Hyper-Up mRNAs, 0 Hyper-Down mRNAs, 177 Hypo-Up mRNAs, 11 Hypo-Down mRNAs (P < 0.05); M2-L vs M1-L: 95 Hyper-Up mRNAs, 32 Hyper-Down mRNAs, 48 Hypo-Up mRNAs, 49 Hypo-Down mRNAs (P < 0.05)

Fig. 3

Expression patterns of m⁶A modification-associated genes among the M0-L, M1-L and M2-L phenotypes. Venn diagrams showing unique and common A Hyper-Up and Hypo-Down, B Hyper-Down and Hypo-Up, C Hypo-Up and Hyper-Down, D Hypo-Down and Hyper-Up genes between “M1-L vs M0-L” (purple) and “M2-L vs M1-L” (blue). (P < 0.05)

Fig. 4

Biological function predictions of the hypo-upregulated mRNAs in M1-L versus M0-L by pathway analysis. A–C Pathway analysis was applied to 46 upregulated hypo-methylated mRNAs and revealed that 16 upregulated pathways were involved in three biological processes (P < 0.05). Enrichment scoring = −log10 (P value)

Fig. 5

Schematic overviews of signaling pathways associated with 46 upregulated and m⁶A hypo-methylated mRNAs in M1-L/M0-L. The yellow squares represent the location of 46 hypo m⁶A and upregulated mRNAs in the related pathways and white squares represent other mRNAs of the pathways which were not be regulated by m⁶A modification. The red letters represent the m⁶A hyper-methylated mRNAs in M2-L/M1-L. Detailed descriptions of the proteins, m⁶A mRNAs, and pathways are presented in Table 1

Fig. 6

Functional predictions of the hyper-upregulated mRNAs in M1-L versus M0-L based on pathway analysis. A–C Pathway analysis was applied to 9 mRNAs and revealed that 10 upregulated pathways were involved in 3 biological processes (P < 0.05). Enrichment scoring = −log10 (P value). D Nine m⁶A hyper-methylated mRNAs and their proteins are in the 10 upregulated signaling pathways. The green circles represent proteins, the orange arrows show the mRNA and protein pair, and blue arrows indicate signaling pathways that proteins involved. The mRNAs using red font represent the m⁶A mRNAs that were hypo-methylated in M2-L vs M1-L

Fig. 7

Functional predictions of the upregulated mRNAs with hypo-methylation or hyper-methylation in M2-L versus M0-L based on pathway analysis. Pathway analysis was applied to 65 upregulated mRNAs in M2-L/M0-L that were hypo- or hyper-methylated, and revealed that 36 upregulated pathways were involved in four biological processes (P < 0.05). Enrichment scoring = −log10 (P value)

Fig. 8

Biological function predictions of the m⁶A lncRNAs based on pathway analysis of their adjacent coding-genes within 300 kb in the genome. A Subgroup analysis of 87 altered m⁶A lncRNAs that were hyper-methylated in M1-L/M0-L and hypo-methylated in M2-L/M1-L in relation to their nearby coding genes. B Subgroup analysis of three altered m⁶A lncRNAs that were hypo-methylated in M1-L/M0-L and hyper-methylated in M2-L/M1-L in relation to their nearby coding genes. C Pathway analysis was applied to 13 mRNAs that were adjacent to 11 m⁶A-modified lncRNAs and revealed that 14 upregulated pathways were associated with two biological processes. Enrichment scoring = −log10 (P value)

Fig. 9

Schematic overviews of the signaling pathways in which 10 m⁶A lncRNAs are probably involved. Ten m⁶A methylated lncRNAs, their regulated mRNAs and proteins are in the 14 upregulated signaling pathways. The red circles represent proteins, the green arrows represent the regulatory relationship of lncRNA to mRNA, the orange arrows show the mRNA and protein pair, and blue arrows indicate signaling pathways that proteins are involved

Fig. 10

The m⁶A methylation level and expression analysis of the lncRNAs and mRNAs in different phenotypes of primary rat microglia. The m⁶A levels of A five mRNAs and B five lncRNAs were analyzed by MeRIP-qPCR; the expression level of C five mRNAs and D five lncRNAs were analyzed using qRT-PCR in M1-L vs M0-L and M2-L vs M1-L. E The m⁶A level and expression level of five mRNAs were analyzed by MeRIP-qPCR and qRT-PCR, respectively in M2-L vs M1-L. qRT-PCR was performed using the GAPDH gene as an internal control. Error bars represent the standard errors of independent samples. n = 3 per group, *P < 0.05