RNA viruses can hijack vertebrate microRNAs to suppress innate immunity

Abstract

Currently, there is little evidence for a notable role of the vertebrate microRNA (miRNA) system in the pathogenesis of RNA viruses. This is primarily attributed to the ease with which these viruses mutate to disrupt recognition and growth suppression by host miRNAs. Here we report that the haematopoietic-cell-specific miRNA miR-142-3p potently restricts the replication of the mosquito-borne North American eastern equine encephalitis virus in myeloid-lineage cells by binding to sites in the 3' non-translated region of its RNA genome. However, by limiting myeloid cell tropism and consequent innate immunity induction, this restriction directly promotes neurologic disease manifestations characteristic of eastern equine encephalitis virus infection in humans. Furthermore, the region containing the miR-142-3p binding sites is essential for efficient virus infection of mosquito vectors. We propose that RNA viruses can adapt to use antiviral properties of vertebrate miRNAs to limit replication in particular cell types and that this restriction can lead to exacerbation of disease severity.

Extended Data Figure 1. EEEV 3′ NTR does not restrict translation in BHK-21 fibroblasts

a. WT EEEV and WT VEEV translation reporters encode the translational initiation control sequences fused to the fLuc gene. b. The EEEV 5′ Δ NTR and VEEV 5′ Δ NTR encode the truncated nsP1 gene and only the 3′ NTR of either EEEV or VEEV. c. The 3′ NTR of EEEV or VEEV was inserted into a host mRNA mimic reporter to generate the 5′ host 3′ EEEV or 5′ host 3′ VEEV reporters. All translation reporters contain a 5′ cap and a 3′ poly (A) tail. d. Translation of WT EEEV, EEEV 5′ Δ NTR, and 5′ host 3′ EEEV reporters in BHK-21 cells. Error bars represent mean ± S.D. and the data are averaged from three independent experiments performed in triplicate.

Extended Data Figure 2. VEEV 3′ NTR does not restrict translation in myeloid cells

Translation of WT VEEV, VEEV 5′ Δ NTR, and 5′ host 3′ VEEV reporters in RAW (a) and BHK-21 (b) cells. Error bars represent mean ± S.D. and the data are averaged from three independent experiments performed in triplicate.

Extended Data Figure 3. Removal of miR-142-3p binding sites in the 3′NTR of EEEV does not alter replication in BHK-21 fibroblasts

a. Red boxes indicate the four miR-142-3p binding sites in the 3′ NTR. Numbers represent nucleotide (nt) positions at the start and end of each miRNA binding site. b. Gray boxes correspond to the complimentary nts in the EEEV 3′ NTR and miR-142-3p. c. EEEV mutant 11337 contains a deletion in the 3′ NTR from nt 11337 to 11596. d. Replication of WT EEEV and 11337 in BHK-21 cells. n = 3 independent experiments. Error bars indicate geometric mean ± S.D., and asterisks indicate differences that are statistically significant (**p<0.01).

Extended Data Figure 4. miR-142-3p binding sites in EEEV restrict replication in human macrophage/monocyte cell lines and primary murine IFNAR^-/- BMDCs

a-b. Replication of WT EEEV and 11337 in human K562 (a) and THP-1 (b) cells. n = 2 (THP-1) and 3 (K562) independent experiments. c. Removal of type I IFN does not alleviate WT EEEV restriction in primary murine IFNAR^-/- BMDCs. n = three independent experiments. Data represent the geometric mean ± S.D., and asterisks indicate differences that are statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001).

Extended Data Figure 5. Relative expression of miR-142-3p in mouse and human cells

Quantitative RT-PCR on primary murine BMDCs, murine and human monocyte/macrophage cell lines, and BHK-21 cells expressing miR-142-3p (BHK-21 w/miR-142-3p). Fold increase in expression is calculated compared to expression of miR-142-3p in BHK-21 cells in which miR-142-3p expression was undetectable.

Extended Data Figure 6. Specific deletion of the miR-142-3p binding sites in the 3′ NTR of WT EEEV does not alter replication in BHK-21 fibroblasts

a. EEEV 142del virus contains four deletions corresponding to the complimentary nucleotides in the 3′ NTR that bind to miR-142-3p eliminating all four miR-142-3p binding sites. b. EEEV 142pm virus contains three point mutations in each of the miR-142-3p binding sites that correspond to the seed sequence of miR-142-3p. c. Replication of WT EEEV and 11337 in BHK-21 cells. n = three independent experiments. Error bars indicate geometric mean ± S.D.

Extended Data Figure 7. Type I IFN attenuates 11337 and 71-77/11337

Survival curves in IFNAR^-/- mice. n = 8 and 10 (71-77/11337) mice per virus from two independent experiments.

Extended Data Figure 8. EEEV 11337 and 71-77/11337 infect myeloid lineage cells in the PLN

a. Percent virus-infected cells in PLN in naïve, WT VEEV, WT EEEV, 71-77, 11337, and 71-77/11337 infected mice. Plots are representative of n = 4 (naïve), 5 (71-77), or 6 mice from two independent experiments. b-c. WT VEEV, 11337 and 71-77/11337 infect myeloid lineage cells in the PLN. b. Representative flow plot frome 1 mice of CD11b (y-axis) and CD11c (x-axis) expression on virus-infected cells. n = 4 (naïve), 5 (71-77) or 6 mice from 2 independent experiments. c. Summary of CD11b and CD11c expression on virus-infected cells from WT VEEV, 11337, and 71-77/11337 infected PLNs. Only mice with responses above naïve mice background levels were used to determine CD11b and CD11c expression.

Figure 1. EEEV restriction in myeloid cells is due to miR-142-3p binding sites in the 3′ NTR

a. The EEEV 3′ NTR restricts translation in RAW cells. Errors bars represent mean ± standard deviation (S.D.), from 3 independent experiments. b-c. Replication of WT EEEV and 11337 in RAW cells (b), and C57BL6BMDCs (c). d. Overexpression of miR-142-3p in BHK-21 cells blocks EEEV infection compared to overexpression of the control miR-124. Data is represented as the ratio of the percentage of cells co-expressing the microRNA, miR-142, (eGFP) and virus-infected (mCherry) to cells co-expressing miR-124 and virus-infected. Data are averaged (mean ± S.D.) from two independent experiments. e. Replication of EEEV 142del and 142pm in RAW cells. f. Ablation of the miR-142-3p binding sites in EEEV increases translation in RAW cells. g. Replication of WT EEEV and 11337 are similar in miR-142^-/- BMDCs. Data represent the geometric mean ± S.D. from three independent experiments unless indicated. Asterisks indicate differences that are statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 2. miR-142-3p binding sites in EEEV 3′ NTR decrease virus replication in the lymph node and enhance disease progression

a. Survival curves of CD-1 mice. n = 9 and 10 (11337) mice/group, three independent experiments. b. Mice were monitored daily for clinical signs of disease. Y-axis represents number of mice exhibiting each sign of disease on each day. c. Serum levels of IFN-α/β in infected CD-1 mice. n = 8 mice, two independent experiments. Limit of detection of IFN-α/β ranged from 1.9 IU/ml to 3.9 IU/ml. d. Quantification of viral replication in the PLNs using nLuc-reporter viruses. n = 8 mice (CD-1) and 6 mice (IFNAR1^-/-) per time point, two independent experiments. e. Visualization of mCherry-reporter virus-infected whole PLN from CD-1 mice 12 h.p.i.. Images are representative of one PLN from two mice. Arrows indicates virus-infected cell(s) in WT EEEV, 71-77, and 11337–infected PLN. f. Quantification of virus-infected cells in PLN harvested 12 h.p.i.. n = 4 (naïve), 5 (71-77), or 6 mice, two independent experiments. Error bars for all experiments represent geometric mean ± S.D. Asterisks indicate differences for all experiments that are statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 3. EEEV sequences containing the miR-142-3p binding sites in EEEV are required for efficient mosquito infection

a. Replication of WT EEEV, 11337, 142del and 142pm in C6/36 mosquito cells. n = 3 or 4 (11337) independent experiments. Error bars represent geometric mean ± S.D. and asterisks indicate differences that are statistically significant using a two-tailed unpaired t test comparing WT EEEV to all other viruses(** p < 0.01, ***p <0.001). b. Infection rates of A. taeniorhynchus after ingestion of infectious bloodmeals(n = 20 mosquitoes per virus). # Indicates 0 of 20 mosquitoes infected.