MicroRNA profiling of human primary macrophages exposed to dengue virus identifies miRNA-3614-5p as antiviral and regulator of ADAR1 expression

Abstract

Background: Due to the high burden of dengue disease worldwide, a better understanding of the interactions between dengue virus (DENV) and its human host cells is of the utmost importance. Although microRNAs modulate the outcome of several viral infections, their contribution to DENV replication is poorly understood.

Methods and principal findings: We investigated the microRNA expression profile of primary human macrophages challenged with DENV and deciphered the contribution of microRNAs to infection. To this end, human primary macrophages were challenged with GFP-expressing DENV and sorted to differentiate between truly infected cells (DENV-positive) and DENV-exposed but non-infected cells (DENV-negative cells). The miRNAome was determined by small RNA-Seq analysis and the effect of differentially expressed microRNAs on DENV yield was examined. Five microRNAs were differentially expressed in human macrophages challenged with DENV. Of these, miR-3614-5p was found upregulated in DENV-negative cells and its overexpression reduced DENV infectivity. The cellular targets of miR-3614-5p were identified by liquid chromatography/mass spectrometry and western blot. Adenosine deaminase acting on RNA 1 (ADAR1) was identified as one of the targets of miR-3614-5p and was shown to promote DENV infectivity at early time points post-infection.

Conclusion/significance: Overall, miRNAs appear to play a limited role in DENV replication in primary human macrophages. The miRNAs that were found upregulated in DENV-infected cells did not control the production of infectious virus particles. On the other hand, miR-3614-5p, which was upregulated in DENV-negative macrophages, reduced DENV infectivity and regulated ADAR1 expression, a protein that facilitates viral replication.

Fig 1. Differentially expressed miRNAs in MDMs challenged with DENV.

(A) MDMs obtained from three different blood donors (D28, D29 and D30) were treated as follows: 1) mock-infected and non-sorted, 2) treated with UVi-DENV, 3) challenged with GFP-DENV (DENV-challenge), 4) mock-infected and passed through the FACS sorter, 5) challenged with GFP-DENV and sorted for GFP positive cells (DENV-plus), 6) challenged with GFP-DENV and sorted for GFP negative cells (DENV-neg). (B) Hierarchical unsupervised Pearson correlation of miRNAs showing differentially expressed miRNAs across the samples (ANOVA, p<0.05). (C) Differentially expressed miRNAs across the samples after the Benjamini and Hochberg correction for multiple testing (p<0.05).

Fig 2. Validation of the miRNA expression profile.

(A) miR-3960 and miR-4508 expression was determined by real-time PCR using specific primers for the mature form of the miRNAs. Fold changes of the respective comparisons were determined through the comparative Ct method taking into account the PCR efficiency and miRNA-16-5p as reference miRNA. Data is presented as mean ± SEM from three different blood donors. (B) Spearman correlation between the fold changes obtained by RNA sequencing and real-time PCR.

Fig 3. Effect of miR-181a, miR-4301, miR-3960 and miR-4508 on DENV infection of Huh7 cells.

Huh7 cells were transfected with a mimic negative control (NC) and with the indicated miRNA mimics. At 24 hpt, cells were infected at MOIs 1, 5 and 10. The percentage of E-positive cells (grey bars) and infectious virus particle production (black bars) were determined at 24 hpi. The percentages of infection of cells transfected with the NC and infected at MOIs 1, 5 and 10 were 6.57±2.07; 13.9±3.48 and 18.53±2.64 respectively; while the viral titers were 1.2x10⁵±5x10⁴; 1.08x10⁶±2x10⁵ and 7.06 x10⁶±6x10⁴ PFU/ml respectively. Data is presented as the percentage relative to the NC and shows mean ± SEM from three independent experiments. Statistical differences were assessed with Student’s t-test.

Fig 4. Effect of miR-3614-5p on DENV and WNV infection of Huh7 cells.

Huh7 cells were transfected as described in the legend to Fig 3. (A) Infectious virus particle production at 24 hpi. In cells transfected with NC and infected at MOI 1, 5 and 10 the viral titers were 1.2x10⁵±5x10⁴; 1.08x10⁶±2x10⁵ and 7.06x10⁶±6x10⁴ PFU/ml respectively. (B) Number of released DENV genome equivalent particles at 24 hpi. For NC-transfected cells infected at MOIs 1, 5 and 10 the DENV genome equivalent particles per ml were: 2.33x10⁷±1x10⁷; 6.77x10⁷±3x10⁷ and 1.72x10⁸±5x10⁷ respectively. (C) Infectious DENV production following infection at MOI 5 and harvesting times of 24, 36, 48 hpi. The corresponding viral titers of cells transfected with NC were: 1.08x10⁶±2x10⁵; 2.59x10⁶±1x10⁶ and 2.6x10⁶±1x10⁶ PFU/ml. (D) Infectious WNV production at 12 hpi; in cells transfected with the NC the viral titers were 1.82x10⁵±1x10⁴ and 4.14x10⁵±1.32x10⁵ PFU/ml in upon infection at MOIs 0.5 and 1, respectively. Data is presented as the percentage relative to the NC and shows mean ± SEM from at least three independent experiments. Statistical differences were assessed with Student’s t-test.

Fig 5. Overexpression of miR-3614-5p regulates protein expression in Huh7 cells.

Huh7 cells were transfected with a mimic NC and with the mimic of miR-3614-5p. At 24 hpt changes in protein expression were analyzed by LC/MS. (A) Clustered heat map displaying differentially expressed proteins between the groups (two independent experiments per group). Underlined proteins were also predicted by bioinformatics tools. (B) Gene set enrichment analysis of the proteins regulated by the miR-3614-5p. (C) Enriched KEGG pathways upregulated by the miR-3614-5p (PR: positive regulation, NR: negative regulation).

Fig 6. Overexpression of miR-3614-5p downregulates ADAR1 expression in Huh7 cells.

(A) Huh7 cells were transfected with the mimic of miR-3614-5p or a siRNA against ADAR1 (siADAR). The correspondent negative control (NC) mimic and NC siRNA were also used. At 24 hpt, total protein was extracted and ADAR1 expression was detected by western blot. (B) Huh7 cells were transfected with the mimic of miR-3614-5p and NC. At 24 hpt, cells were infected with DENV at MOI 1 and 10. At 24 hpi, total protein was extracted and ADAR1 expression was detected by western blot. (A, B) The expression of ADAR1 was normalized to that of GAPDH and it is expressed as the percentage of the cells transfected with the correspondent NC. Data shows mean ± SEM from three independent experiments. Differences were assessed with Student’s t-test.

Fig 7. DENV infection induces ADAR1 expression in MEFs.

MEFs were infected with DENV at the indicated MOIs. The methodology is as described in the legend to Fig 6. The expression of ADAR1 was normalized to that of GAPDH and it is expressed as the percentage of the mock-infected cells. Data shows mean ± SEM from three independent experiments.

Fig 8. ADAR KO MEFs are less permissive to DENV than wild-type MEFs.

Wild-type (WT) MEFs, p53 KO MEFs (p53^-/-) and p53/ADAR double KO MEFs (ADAR^-/-) were infected with DENV at MOIs 1 and 5. At the indicated time points, the percentage of infection (upper panels) was determined by flow cytometry and the infectious virus particle production (lower panels) by plaque assay. Data shows mean ± SEM from three independent experiments.

Fig 9. miR-3614-5p impairs DENV production in WT MEFs but not in ADAR KO cells.

WT and p53^-/-ADAR1^-/- cells were transfected with the mimic of miR-3614-5p or the correspondent negative control (NC). At 24 hpt, cells were infected with DENV at MOI 5. At 24 hpi, cell supernatants were titrated by plaque assay. Data shows mean ± SEM from three independent experiments. Differences were assessed with Student’s t-test.