microRNA-guided immunity against respiratory virus infection in human and mouse lung cells

Abstract

Viral infectivity depends on multiple factors. Recent studies showed that the interaction between viral RNAs and endogenous microRNAs (miRNAs) regulates viral infectivity; viral RNAs function as a sponge of endogenous miRNAs and result in upregulation of its original target genes, while endogenous miRNAs target viral RNAs directly and result in repression of viral gene expression. In this study, we analyzed the possible interaction between parainfluenza virus RNA and endogenous miRNAs in human and mouse lungs. We showed that the parainfluenza virus can form base pairs with human miRNAs abundantly than mouse miRNAs. Furthermore, we analyzed that the sponge effect of endogenous miRNAs on viral RNAs may induce the upregulation of transcription regulatory factors. Then, we performed RNA-sequence analysis and observed the upregulation of transcription regulatory factors in the early stages of parainfluenza virus infection. Our studies showed how the differential expression of endogenous miRNAs in lungs could contribute to respiratory virus infection and species- or tissue-specific mechanisms and common mechanisms could be conserved in humans and mice and regulated by miRNAs during viral infection.

Keywords: Lungs; MicroRNA; RNA silencing; RNA–RNA interaction; Respiratory virus.

Fig. 1.

Endogenous miRNAs that are expressed in human or mouse lungs. (A) Small RNA-seq data were obtained from a public database, and miRNA expression profiles of human or mouse lungs were analyzed to extract miRNAs expressed in human or mouse lungs from all miRNAs registered on miRBase (Filter 1). The miRNAs that were assumed to be effective were extracted in terms of nucleotide preference at the 5′ end of the mature miRNA (Filter 2). The miRNAs that could interact with viral RNAs via the seed region in positions 2–7 from the 5′ end of the miRNA were extracted (Filter 3). Using the miRNAs that passed all filters, we performed the visualization of “viral RNA versus endogenous miRNA”. The number of human and mouse miRNAs registered on miRBase and that of miRNAs that passed each filter is shown on the right side of the flowchart. (B) The Venn diagram of the miRNAs that are expressed in the lungs and exhibited an effective sequence: 148 human miRNAs and 131 mouse miRNAs. (C) The XY plot of the expression levels (TPM) of common effective miRNAs in human or mouse pulmonary lungs. (D) The violin plots of the expression levels (TPM) of common or species-specific miRNAs in human or mouse lungs.

Fig. 2.

The possible interaction between parainfluenza virus RNAs and endogenous miRNAs in human and mouse lungs. (A) The miRNA interaction sites on the genome and its complementary strand of mouse parainfluenza virus (MPIV). (B) The miRNA interaction sites on the genome of the human parainfluenza virus (HPIV) and its complementary strand. The upper half shows the number of human or mouse miRNAs that interact with each position of the complementary strand of the viral genome (positive stand). The lower half shows the number of human or mouse miRNAs that interact with each position of the viral genome (negative strand). The positions with blue lines indicate the position with higher interaction counts in humans than in mice (human–mouse ≥2) and those with orange lines indicate the position with higher interaction counts in humans than in mice (mouse–human ≥2). The black asterisks in A show the example of positions demonstrating a similar tendency for interactions with miRNAs in humans and mice. The blue asterisk shows the example of positions with higher interaction counts in humans than in mice, and the orange asterisk shows the example of positions with higher interaction counts in mice than humans.

Fig. 3.

Normalized number of target sites of pulmonary miRNAs on each transcript encoding MPIV or HPIV genes. The sequences underwent 3-mer shuffling 10 times, and the averaged values of these shuffled sequences from 1 to 10 iterations were used as the results of the shuffled sequences. The numbers were normalized to the length of the transcripts. Statistical analysis was performed using a one-tailed, two-sample t-test with equal variances. (A) MPIV (B) HPIV.

Fig. 4.

Common or species-specific miRNAs that possess multiple interaction sites on viral RNA. (A) The interaction sites of miRNA with multiple target sites (≥10) on the genome of MPIV (the lower half) and its complementary strand (the upper half). (B) The interaction sites of miRNA with multiple target sites (≥10) on the genome (the lower half) of HPIV and its complementary strand (the upper half). miRNAs were distinguished into three types: common miRNAs in humans and mice, human-specific miRNAs, and mouse-specific miRNAs.

Fig. 5.

The sponge effect and upregulated genes induced by interaction between viral RNAs and endogenous miRNAs. (A) The gene ontology (GO) analysis of candidate genes that can be dysregulated in MPIV-infected human or mouse lungs. (B) The GO analysis of candidate genes that can be dysregulated in HPIV-infected human or mouse lungs. The genes that can be targeted by more than three miRNAs among miRNAs with multiple target sites (≥10) on the viral genome and its complementary strand were defined as the target genes that can be dysregulated.

Fig. 6.

Gene expression profiling of MPIV-infected A549 cells. XY plot of normalized read counts (log10, TPM) of all transcripts (gray) and upregulated genes (SeV/mock >2, pink) in A549 cells collected at 1 h (A), 2 h (B), or 6 h (C) following MPIV infection and the gene ontology (GO) analysis of upregulated genes in MPIV-infected cells. (D) Rate of genes targeted by three or more pulmonary miRNAs with multiple interaction sites (≥10). Each bar represents the rate of genes targeted by three or more pulmonary miRNAs with multiple interaction sites (≥10) in all transcripts (gray bars) or in upregulated transcripts at 1, 2, or 6 h, respectively (pink bars). Statistical analysis was performed using a Fisher's test.