Antiviral RNAi Response in Culex quinquefasciatus-Derived HSU Cells

Abstract

Culex spp. mosquitoes are important vectors of viruses, such as West Nile virus, Eastern equine encephalitis virus and Rift valley fever virus. However, their interactions with innate antiviral immunity, especially RNA interference (RNAi), are not well known. Most research on RNAi pathways in mosquitoes is focused on the tropical vector mosquito Aedes aegypti. Here, we investigated the production of arbovirus-specific small RNAs in Cx. quinquefasciatus-derived HSU cells. Furthermore, by silencing RNAi-related proteins, we investigated the antiviral role of these proteins for two different arboviruses: Semliki Forest virus (SFV) and Bunyamwera orthobunyavirus (BUNV). Our results showed an expansion of Ago2 and Piwi6 in Cx. quinquefasciatus compared to Ae. aegypti. While silencing Ago2a and Ago2b increased BUNV replication, only Ago2b showed antiviral activity against SFV. Our results suggest differences in the function of Cx. quinquefasciatus and Ae. aegypti RNAi proteins and highlight the virus-specific function of these proteins in Cx. quinquefasciatus.

Keywords: Aedes aegypti; Culex quinquefasciatus; RNA interference; RNAi; antiviral immunity.

Figure 1

BUNV (A) and SFV (B) growth kinetics in Cx. quinquefasciatus-derived HSU cells. HSU cells were inoculated either with BUNV-NLuc (MOI 0.1) or with SFV-NLuc (MOI 0.1 or MOI 10) expressing NLuc as a reporter protein. The mean of three independent experiments performed in technical duplicates is shown with standard error.

Figure 2

Small RNA analyses of BUNV-infected Cx. quinquefasciatus-derived HSU cells. Small RNAs of HSU cells were mapped to the BUNV genome and antigenome. (A) Distribution of the 21 nt small RNAs or (B) 26–29 nt small RNAs along the genome and antigenome of the three segments of BUNV (S, M, L). (C) Length distribution of BUNV-specific small RNAs. Positive numbers are RNAs mapping to the antigenome of BUNV (green), while negative numbers indicate RNAs mapping to the genome of BUNV (pink). Y-axis: absolute count of small RNAs. (D) Relative nucleotide frequency and conservation per position of 26–29 nt small RNAs mapping to the BUNV genome or antigenome. (E) The overlap z-score indicating the probability of overlap between the genome and antigenome of 26–29 nt BUNV-specific small RNAs was calculated. Two independent experiments were carried out, and the results of one representative experiment are shown here (Figure S2).

Figure 3

BUNV-specific viral (v)DNA production in infected Cx. quinquefasciatus-derived HSU cells. DNA of BUNV- and mock-infected HSU cells were collected at 6 dpi. PCR using primer set 3 (Table S3) was performed. cDNA of BUNV-infected cells was used as a positive control.

Figure 4

Silencing of RNAi transcripts and its effect on arbovirus infection in HSU cells. HSU cells were transfected either with gene-specific dsRNAs (Ago2a, Ago2b, Ago3, Piwi1/3, Piwi4, Piwi5, Piwi6a, Piwi6b) or with control dsRNA (eGFP-specific). The following day, cells were infected with BUNV-NLuc (MOI 0.1) or SFV-NLuc (MOI 10). To confirm the silencing of the transcripts (A), mRNA targets were quantified using gene-specific primers and GADPH mRNA as a housekeeper transcript. Fold change in mRNA targets was calculated using the 2^−ΔΔCT method with eGFP dsRNA samples as the control. The mean of at least 3–5 replicates and the standard error of the mean are shown. To investigate the effect on arbovirus infection (B), Nanoluciferase activity was measured and normalised to samples transfected with eGFP dsRNA as a control (*: p < 0.05, **: p < 0.001, ***: p < 0.01). Boxplots show the median of 4 independent experiments performed in technical duplicates.