Distinct Roles of Hemocytes at Different Stages of Infection by Dengue and Zika Viruses in Aedes aegypti Mosquitoes

Abstract

Aedes aegypti mosquitoes are vectors for arboviruses of medical importance such as dengue (DENV) and Zika (ZIKV) viruses. Different innate immune pathways contribute to the control of arboviruses in the mosquito vector including RNA interference, Toll and Jak-STAT pathways. However, the role of cellular responses mediated by circulating macrophage-like cells known as hemocytes remains unclear. Here we show that hemocytes are recruited to the midgut of Ae. aegypti mosquitoes in response to DENV or ZIKV. Blockade of the phagocytic function of hemocytes using latex beads induced increased accumulation of hemocytes in the midgut and a reduction in virus infection levels in this organ. In contrast, inhibition of phagocytosis by hemocytes led to increased systemic dissemination and replication of DENV and ZIKV. Hence, our work reveals a dual role for hemocytes in Ae. aegypti mosquitoes, whereby phagocytosis is not required to control viral infection in the midgut but is essential to restrict systemic dissemination. Further understanding of the mechanism behind this duality could help the design of vector-based strategies to prevent transmission of arboviruses.

Keywords: Aedes aegypti; Zika virus; cellular immunity; dengue virus; hemocytes; macrophage-like cells; vector mosquitoes.

Figure 1

Hemocyte accumulation in the midgut of Ae. aegypti mosquitoes in response to DENV and ZIKV. (A) Mosquitoes fed with sugar, blood or blood and virus were dissected at different times and their midgut was analyzed by confocal microscopy. Virus-infected mice were used as a source of blood. (B, C) Quantification of the number of midgut-associated hemocytes between mosquitoes fed with sugar, blood or blood and virus at 4 (B) and 8 (C) days post feeding. DENV and ZIKV were analyzed together. 2 independent experiments for each virus were pooled. Each dot represents an individual midgut. Total number of midguts tested is indicated below each box plot. Upper, middle and lower bars in the boxplot represent the 75th percentile, the median and the 25th percentile, respectively. Statistical analyses were performed using the Kruskal-Wallis test followed by Dunn’s test to correct for multiple comparisons. ns, non-significant. (D–G) Representative confocal microscopy images of mosquito midguts showing CM-DiL stained hemocytes in magenta, DNA in yellow, viral E proteins in green and actin in blue. Midguts from mosquitoes fed with sugar (D), blood (E), blood + DENV (F) and blood + ZIKV (G) are shown at 8 days post feeding.

Figure 2

Phagocytosis by hemocytes is not required to control DENV and ZIKV infection in the midgut of Ae. aegypti mosquitoes. (A) Mosquitoes injected with latex beads were fed 2 days later with blood + virus and dissected at different times to be analyzed. Virus-infected mice were used as a source of blood. (B, C) Percentage of total midgut infection area that shows staining for the viral protein at 4 and 8 days post infection was determined by immunofluorescence. Total number of midguts tested is indicated below each box plot. DENV (B) and ZIKV (C) were analyzed separately. 2 independent experiments for each virus were pooled. Each dot represents an individual midgut. (D–G) Representative confocal microscopy images of mosquito midguts showing CM-DiL stained hemocytes in magenta, DNA in yellow, viral E proteins in green and actin in blue. (D, E) Midguts from mosquitoes fed on blood + DENV. (F, G) Midguts from mosquitoes fed on blood + ZIKV. (D, F) Midguts from control mosquitoes injected with buffer; (E, G) Midguts from mosquitoes injected with latex beads. (H) DENV and (I) ZIKV RNA levels measured by RT-qPCR at 8 days post feeding. The number of positive midguts over the total tested is indicated below each boxplot. (J, K) Number of midgut-associated hemocytes in individual midguts from control and virus infected mosquitoes at 4 and 8 days post feeding. DENV (J) and ZIKV (K) were analyzed separately. Total number of midguts tested is indicated below each box plot. 2 independent experiments for each virus were pooled. (B, C, H–K) Each dot represents an individual midgut. Upper, middle and lower bars in the boxplot represent the 75th percentile, the median and the 25th percentile, respectively. Statistical analyses were performed using the Mann-Whitney-Wilcoxon test.

Figure 3

Systemic dissemination of ZIKV and DENV infection is controlled by hemocyte phagocytosis. (A) Mosquitoes injected with latex beads were fed 2 days later with blood and virus and dissected at 8 days post feeding to be analyzed. Virus-infected mice were used as a source of blood. (B, C) Prevalence of DENV (B) and ZIKV (C) infection in the mosquito carcass at 8 days post feeding. The number of positive mosquitoes over the total tested is indicated below each column. One representative experiment is shown. This experiment was repeated 3 times for DENV and once for ZIKV. Statistical analyses were performed using two-tailed Fishers exact test. (D, E) Viral RNA levels at 8 days post feeding for DENV (D) and ZIKV (E). One representative experiment is shown. This experiment was repeated 3 times for DENV and once for ZIKV. Each dot represents an individual mosquito. (F) Mosquitoes injected with latex beads were given an artificial blood meal with virus 2 days later and analyzed at 8 days post feeding. (G, H) Prevalence of DENV (G) and ZIKV (H) infection in mosquitoes injected with buffer or beads. The number of positive mosquitoes over the total tested is indicated below each column. One representative experiment is shown. This experiment was repeated twice for DENV and once for ZIKV. Statistical analyses were performed using two-tailed Fishers exact test. (I, J) DENV (I) and ZIKV (J) RNA levels in mosquitoes injected with buffer or beads. Each dot represents an individual mosquito. The number of positive mosquitoes over the total tested is indicated below each boxplot. One representative experiment is shown. This experiment was repeated twice for DENV and once for ZIKV. (D, E, I, J) Upper, middle and lower bars in the boxplot represent the 75th percentile, the median and the 25th percentile, respectively. Statistical analyses were performed using the Mann-Whitney-Wilcoxon test.

Figure 4

Phagocytosis by hemocyte is required to inhibit systemic replication of ZIKV and DENV. (A) Mosquitoes injected with latex beads were subsequently injected with virus 2 days later and samples were analyzed at different time points. (B) Viral RNA levels in mosquitoes injected with 5 or 50 PFU of DENV at 8 days post injection. (C) Viral RNA levels in mosquitoes injected with 5 PFU of ZIKV at 2, 4 and 8 days post injection. One representative experiment is shown. This experiment was repeated twice for DENV and once for ZIKV. (B, C) Each dot represents an individual mosquito. The number of positive mosquitoes over the total tested is indicated above each boxplot. Upper, middle and lower bars in the boxplot represent the 75th percentile, the median and the 25th percentile, respectively. Statistical analyses were performed using the Mann-Whitney-Wilcoxon test.