RNA interference (RNAi) was shown over 18 years ago to be a mechanism by which arbovirus replication and transmission could be controlled in arthropod vectors. During the intervening period, research on RNAi has defined many of the components and mechanisms of this antiviral pathway in arthropods, yet a number of unexplored questions remain. RNAi refers to RNA-mediated regulation of gene expression. Originally, the term described silencing of endogenous genes by introduction of exogenous double-stranded (ds)RNA with the same sequence as the gene to be silenced. Further research has shown that RNAi comprises three gene regulation pathways that are mediated by small RNAs: the small interfering (si)RNA, micro (mi)RNA, and Piwi-interacting (pi)RNA pathways. The exogenous (exo-)siRNA pathway is now recognized as a major antiviral innate immune response of arthropods. More recent studies suggest that the piRNA and miRNA pathways might also have important roles in arbovirus-vector interactions. This review will focus on current knowledge of the role of the exo-siRNA pathway as an arthropod vector antiviral response and on emerging research into vector piRNA and miRNA pathway modulation of arbovirus-vector interactions. Although it is assumed that arboviruses must evade the vector’s antiviral RNAi response in order to maintain their natural transmission cycles, the strategies by which this is accomplished are not well defined. RNAi is also an important tool for arthropod gene knock-down in functional genomics studies and in development of arbovirus-resistant mosquito populations. Possible arbovirus strategies for evasion of RNAi and applications of RNAi in functional genomics analysis and arbovirus transmission control will also be reviewed.

Aedes aegypti transmit pathogenic arboviruses while the mosquito itself tolerates the infection. We examine a piRNA-based immunity that relies on the acquisition of viral derived cDNA (vDNA) and how this pathway discriminates between self and non-self. The piRNAs derived from these vDNAs are essential for virus control and Piwi4 has a central role in the pathway. Piwi4 binds preferentially to virus-derived piRNAs but not to transposon-targeting piRNAs. Analysis of episomal vDNA from infected cells reveals that vDNA molecules are acquired through a discriminatory process of reverse-transcription and recombination directed by endogenous retrotransposons. Using a high-resolution Ae. aegypti genomic sequence, we found that vDNAs integrated in the host genome as endogenous viral elements (EVEs), produce antisense piRNAs that are preferentially loaded onto Piwi4. Importantly, EVE-derived piRNAs are specifically loaded onto Piwi4 to inhibit virus replication. Thus, Ae. aegypti employs a sophisticated antiviral mechanism that promotes viral persistence and generates long-lasting adaptive immunity.

Background: Arthropod-borne viruses (arboviruses) can persistently infect and cause limited damage to mosquito vectors. RNA interference (RNAi) is a mosquito antiviral response important in restricting RNA virus replication and has been shown to be active against some arboviruses. The goal of this study was to use a recombinant Sindbis virus (SINV; family Togaviridae; genus Alphavirus) that expresses B2 protein of Flock House virus (FHV; family Nodaviridae; genus Alphanodavirus), a protein that inhibits RNAi, to determine the effects of linking arbovirus infection with RNAi inhibition. Results: B2 protein expression from SINV (TE/3'2J) inhibited the accumulation of non-specific small RNAs in Aedes aegypti mosquito cell culture and virus-specific small RNAs both in infected cell culture and Ae. aegypti mosquitoes. More viral genomic and subgenomic RNA accumulated in cells and mosquitoes infected with TE/3'2J virus expressing B2 (TE/3'2J/B2) compared to TE/3'2J and TE/3'2J virus expressing GFP. TE/3'2J/B2 exhibited increased infection rates, dissemination rates, and infectious virus titers in mosquitoes following oral bloodmeal. Following infectious oral bloodmeal, significantly more mosquitoes died when TE/3'2J/B2 was ingested. The virus was 100% lethal following intrathoracic inoculation of multiple mosquito species and lethality was dose-dependent in Ae. aegypti. Conclusion: We show that RNAi is active in Ae. aegypti cell culture and that B2 protein inhibits RNAi in mosquito cells when expressed by a recombinant SINV. Also, SINV more efficiently replicates in mosquito cells when RNAi is inhibited. Finally, TE/3'2J/B2 kills mosquitoes in a dose-dependent manner independent of infection route and mosquito species.

In 2006, we reported a mariner (Mos1)-transformed Aedes aegypti line, Carb77, which was highly resistant to dengue-2 virus (DENV2). Carb77 mosquitoes expressed a DENV2-specific inverted-repeat (IR) RNA in midgut epithelial cells after ingesting an infectious bloodmeal. The IR-RNA formed double-stranded DENV2-derived RNA, initiating an intracellular antiviral RNA interference (RNAi) response. However, Carb77 mosquitoes stopped expressing the IR-RNA after 17 generations in culture and lost their DENV2-refractory phenotype. In the current study, we generated new transgenic lines having the identical transgene as Carb77. One of these lines, Carb109M, has been genetically stable and refractory to DENV2 for >33 generations. Southern blot analysis identified two transgene integration sites in Carb109M. Northern blot analysis detected abundant, transient expression of the IR-RNA 24 h after a bloodmeal. Carb109M mosquitoes were refractory to different DENV2 genotypes but not to other DENV serotypes. To further test fitness and stability, we introgressed the Carb109M transgene into a genetically diverse laboratory strain (GDLS) by backcrossing for five generations and selecting individuals expressing the transgene's EGFP marker in each generation. Comparison of transgene stability in replicate backcross 5 (BC₅) lines versus BC₁ control lines demonstrated that backcrossing dramatically increased transgene stability. We subjected six BC₅ lines to five generations of selection based on EGFP marker expression to increase the frequency of the transgene prior to final family selection. Comparison of the observed transgene frequencies in the six replicate lines relative to expectations from Fisher's selection model demonstrated lingering fitness costs associated with either the transgene or linked deleterious genes. Although minimal fitness loss (relative to GDLS) was manifest in the final family selection stage, we were able to select homozygotes for the transgene in one family, Carb109M/GDLS.BC₅.HZ. This family has been genetically stable and DENV2 refractory for multiple generations. Carb109M/GDLS.BC₅.HZ represents an important line for testing proof-of-principle vector population replacement. Expression of a DENV2 sequence-derived IR RNA in the mosquito midgut initiates an antiviral intracellular RNAi response that efficiently blocks DENV2 infection and profoundly impairs vector competence for that virus in Aedes aegypti. DENV2-specific IR RNA expression in the Carb109M strain has maintained the RNAi-based, refractory phenotype for 33 generations in laboratory culture. The two transgene integration sites were stable after multiple generations and following introgression into a genetically-diverse (GDLS) Ae. aegypti population. Introgression of the transgene into the GDLS genetic background changed GDLS from a DENV2 susceptible phenotype to a DENV2 refractory phenotype. The DENV2 refractory homozygous line, Carb109M/GDLS.BC₅.HZ, exhibits (relative to GDLS) minimal fitness loss associated with the transgene. This line could be a potential candidate for proof-of-principle field studies.

Aedes aegypti is the primary mosquito vector of dengue viruses (DENV; serotypes 1–4). Human-mosquito transmission cycles maintain DENV during epidemics but questions remain regarding how these viruses survive when human infections and vector abundance are minimal. Aedes mosquitoes can transmit DENV within the vector population through two alternate routes: vertical and venereal transmission (VT and VNT, respectively). We tested the efficiency of VT and VNT in a genetically diverse laboratory (GDLS) strain of Ae. aegypti orally infected with DENV2 (Jamaica 1409). We examined F1 larvae from infected females generated during the first and second gonotrophic cycles (E1 and E2) for viral envelope (E) antigen by amplifying virus in C6/36 cells and then performing an indirect immunofluorescence assay (IFA). RT-PCR/nested PCR analyses confirmed DENV2 RNA in samples positive by IFA. We observed VT of virus to larvae and adult male progeny and VNT of virus to uninfected virgin females after mating with males that had acquired virus by the VT route. We detected no DENV2 in 30 pools (20 larvae/pool) of F1 larvae following the first gonotrophic cycle, suggesting limited virus dissemination at 7 days post-infection. DENV2 was detected by IFA in 27 of 49 (55%) and 35 of 51 (68.6%) F1 larval pools (20 larvae/pool) from infected E2 females that received a second blood meal without virus at 10 or 21 days post-infection (E2-10d-F1 and E2-21-F1), respectively. The minimum filial infection rates by IFA for E2-10d-F1 and E2-21d-F1 mosquitoes were 1:36 and 1:29, respectively. The VNT rate from E2-10d-F1 males to virgin (uninfected) GDLS females was 31.6% (118 of 374) at 8 days post mating. Twenty one percent of VNT-infected females receiving a blood meal prior to mating had disseminated virus in their heads, suggesting a potential pathway for virus to re-enter the human-mosquito transmission cycle. This is the first report of VNT of DENV by male Ae. aegypti and the first demonstration of sexual transmission in Aedes by naturally infected males. Our results demonstrate the potential for VT and VNT of DENV in nature as mechanisms for virus maintenance during inter-epidemic periods. Aedes aegypti mosquitoes are the primary vectors of dengue viruses (DENV) and other medically important arthropod-borne viruses (yellow fever, Zika, chikungunya). DENVs replicate in Ae. aegypti and humans in a transmission cycle that maintains the virus in nature. However, adverse conditions (i.e. dry seasons and low temperatures, vector populations, and host susceptibility) can suppress the transmission cycle. How does the virus survive in nature when the mosquito-human epidemic cycle subsides? Alternative DENV transmission routes can allow the virus to remain in the vector population. DENV can vertically transmit from mother to offspring and venereally transmit from infected male offspring to females during mating. In this report, we show that DENV2 efficiently pass from one generation to the next after the infected female mosquito acquires a second blood meal. Furthermore, infected male progeny arising from vertical transmission can then infect females during mating. Vertical transmission of DENV2, possibly in combination with venereal transmission, can potentially maintain DENV2 for sufficient time for the vector-human transmission cycle to resume.