Differentially expressed miRNAs after HFMD virus infection
We first analyzed the GSE85829 and GSE94551 datasets and found some common differentially expressed miRNAs (DE-miRNAs). In 16HBE cells based on the GSE85829 dataset, EV71 infection induced 25 DE-miRNAs and CA16 infection induced 13 DE-miRNAs (Fig. 1 A and B), among which 7 common DE-miRNAs were found. In the HUVEC line according to the GSE94551 dataset, 39 DE-miRNAs were observed after EV71 infection and 99 DE-miRNAs were found after CA16 infection (Fig. 2 A and B), and they also shared 7 common DE-miRNAs (Fig. 2B). However, no intersection was found between two groups of 7 common DE-miRNAs. The common DE-miRNAs among four groups of cells were presented by a Venn diagram in Fig. 2C. Similarly, miRNAs in serum exosomes were analyzed based on the GSE52780 dataset, and a total of 258 DE-miRNAs were found (Fig. 3A), including 85 up-regulated and 173 down-regulated ones; the top 10 DE-miRNAs were listed in Fig. 3B. Afterwards, we screened those DE-miRNAs which appeared three times in the above samples (cell lines or exosomes) in different datasets and regarded them as key miRNAs in HFMD. Ultimately, five key miRNAs were acquired: miR-100-3p (appeared 4 times), miR-125a-3p (appeared 3 times), miR-1273g-3p (appeared 3 times), miR-5585-3p (appeared 3 times), and miR-671-5p (appeared 3 times).
GO and KEGG enrichment
Based on above five key miRNAs, GO functional and KEGG pathway enrichment were performed using the miRPath V3 database. First, GO and KEGG enrichment were acquired by the intersection targets of the five key miRNAs (only four miRNAs were included in this database) (Fig. 4A and B) Further, this was verified by the union of key miRNA related GO terms or KEGG pathways, and heatmaps were shown in Fig. 4 C and D. Overall, this result was consistent with Fig. 4 A and B. Parallelly, we applied the mRNA profiles in SH-SY5Y cells infected by EV71 from the GSE45589 dataset and performed the enrichment analysis. The enriched GO functions of up-regulated and down-regulated genes were listed in Fig. 4 E and F, respectively. A total of 29 up-regulated GO functional terms were identified (meiotic nuclear division, response to radiation and regulation of microtubule-based process, etc.), accomplished by 43 down-regulated terms (positive regulation of transcription, DNA-templated, negative regulation of apoptotic process, and anatomical structure morphogenesis, etc.). Ten enriched KEGG pathways were shown in Fig. 4 G, including 2 up-regulated ones (cell cycle and spliceosome), and 8 down-regulated pathways (cytokine-cytokine receptor interaction, hematopoietic cell lineage, and intestinal immune network for IgA production, etc.). In comparison, there were three common enriched GO terms between miRNA-derived prediction and mRNA-derived (dataset GSE45589) analysis: biosynthetic process, cytosol, and nucleoplasm. And one common KEGG pathway, namely cell cycle, was shared between miRNA-based and mRNA-based enrichment.
Common differential mRNAs and PPI network
Using TarBase V8 in DIANA tools, we acquired 1520 potential targets (mRNA) from the 5 key DE-miRNAs, among which 11 DEGs were also included by the159 DE-mRNAs in the GSE45589 dataset: MACF1, MARS, SF3B3, SMARCC1, BRMS1L, SMC1A, SPHK2, LIG1, CSF3, CYR61 and FGFR1OP (Fig. 5A). Theoretically, these genes were the most likely ones influenced by HFMD virus infection. GO functional analysis showed three terms might be enriched according to these DEGs: positive regulation of cell proliferation, anatomical structure morphogenesis, and ATP binding (Fig. 5B). These common DEGs showed a PPI network mainly connected by SMC1A, SMARCC1, SF3B3, LIG1 and BRMS1L (Fig. 5C), and this network locates at a core place in the PPI network constructed by the 159 DE-mRNAs in the GSE45589 dataset (Fig. 5D, the isolated nodes were removed). Together, changes in 5 key miRNAs and 11 key mRNAs may play crucial roles in HFMD virus induced pathological changes and count be used as diagnostic markers for the HFMD.