Transgene-mediated suppression of the RNA interference pathway in Aedes aegypti interferes with gene silencing and enhances Sindbis virus and dengue virus type 2 replication

Authors

  • C. C. H. Khoo,

    1. Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology & Pathology, Colorado State University, Fort Collins, CO, USA
    Search for more papers by this author
  • J. B. Doty,

    1. Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology & Pathology, Colorado State University, Fort Collins, CO, USA
    Search for more papers by this author
  • M. S. Heersink,

    1. Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology & Pathology, Colorado State University, Fort Collins, CO, USA
    Search for more papers by this author
  • K. E. Olson,

    1. Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology & Pathology, Colorado State University, Fort Collins, CO, USA
    Search for more papers by this author
  • A. W. E. Franz

    Corresponding author
    • Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology & Pathology, Colorado State University, Fort Collins, CO, USA
    Search for more papers by this author

Correspondence: Alexander W. E. Franz, Arthropod-borne and Infectious Diseases Laboratory, Department of Microbiology, Immunology, and Pathology, Colorado State University, Fort Collins, CO 80523, USA. Tel.: +1 970 491 8611; fax: +1 970 491 8323; e-mail: afranz@colostate.edu

Abstract

RNA interference (RNAi) is the major innate antiviral pathway in Aedes aegypti that responds to replicating arboviruses such as dengue virus (DENV) and Sindbis virus (SINV). On the one hand, the mosquito's RNAi machinery is capable of completely eliminating DENV2 from Ae. aegypti. On the other, transient silencing of key genes of the RNAi pathway increases replication of SINV and DENV2, allowing the viruses to temporally overcome dose-dependent midgut infection and midgut escape barriers (MEB) more efficiently. Here we expressed Flock house virus B2 (FHV-B2) from the poly-ubiquitin (PUb) promoter in Ae. aegypti using the ΦC31 site-directed recombination system to investigate the impact of transgene-mediated RNAi pathway suppression on infections with SINV-TR339eGFP and DENV2-QR94, the latter of which has been shown to be confronted with a strong MEB in Ae. aegypti. FHV-B2 was constitutively expressed in midguts of sugar- and blood-fed mosquitoes of transgenic line PUbB2 P61. B2 over-expression suppressed RNA silencing of carboxypeptidase A-1 (AeCPA-1) in midgut tissue of PUbB2 P61 mosquitoes. Following oral challenge with SINV-TR339eGFP or DENV2-QR94, mean titres in midguts of PUbB2 P61 females were significantly higher at 7 days post-bloodmeal (pbm) than in those of nontransgenic control mosquitoes. At 14 days pbm, infection rates of carcasses were significantly increased in PubB2 P61 mosquitoes infected with SINV-TR339eGFP. Following infection with DENV2-QR94, midgut infection rates were significantly increased in the B2-expressing mosquitoes at 14 days pbm. However, B2 expression in PUbB2 P61 did not increase the DENV2-QR94 dissemination rate, indicating that the infection phenotype was not primarily controlled by RNAi.

Introduction

The yellow fever mosquito, Aedes aegypti, is a vector of medically important arboviruses including the flaviviruses (Flaviviridae; Flavivirus) dengue virus type 1–4 (DENV1−4) and yellow fever virus and alphaviruses (Togaviridae; Alphavirus) such as chikungunya virus and under laboratory conditions Sindbis virus (SINV) (Sanchez-Vargas et al., 2004; Weaver & Reisen, 2010). Arboviruses often cause severe pathology in their vertebrate hosts but are usually not pathogenic to their invertebrate vectors (Blair, 2011). In the mosquito, arboviruses are confronted with several innate immune pathways, such as RNA interference (RNAi), Toll, immune deficiency (Imd), and Janus kinase-signal transducer and activator of transcription (JAK-STAT) (Sanchez-Vargas et al., 2004; Xi et al., 2008; Souza-Neto et al., 2009; Ramirez & Dimopoulos, 2010; Blair, 2011). Of these, RNAi is the major antiviral innate immune pathway in Ae. aegypti (Franz et al., 2006; Campbell et al., 2008; Sanchez-Vargas et al., 2009; Khoo et al., 2010; Mathur et al., 2010; Scott et al., 2010). RNA silencing is divided into three distinct, highly conserved small RNA pathways, micro RNA (miRNA), small interfering RNA (siRNA), and piwi-interacting RNA (piRNA) (Ghildiyal & Zamore, 2009). Based on the origin of the RNA substrate the siRNA pathway is further divided into endogenous RNAi and exogenous RNAi, the latter of which has a major impact on virus infections in mosquito vectors. RNAi can eliminate DENV2 from transgenic mosquitoes expressing an inverted-repeat RNA to trigger the RNAi pathway against the virus (Franz et al., 2006; Mathur et al., 2010). However, arboviruses are able to persistently infect vectors despite being targeted by the RNAi machinery as shown by the presence of 21 nt virus-derived small interfering RNAs in arbovirus-infected, transmission-competent mosquito vectors (Scott et al., 2010; Hess et al., 2011). Moreover, temporally induced silencing of the RNAi machinery in Ae. aegypti led to significantly increased SINV and DENV2 titres combined with increased midgut infection and dissemination rates and a shortened extrinsic incubation period without compromising insect survival (Campbell et al., 2008; Sanchez-Vargas et al., 2009; Khoo et al., 2010). Arboviruses must have evolved strategies to cope with the mosquito's RNAi pathway. One such strategy could be a nucleic acid-based decoy mechanism (Siu et al., 2011). Plant viruses and several animal viruses encode proteins that suppress RNAi to protect their viral RNA from homology-dependent degradation (Burgyan & Havelda, 2011; Song et al., 2011). Effects of RNAi suppression in Ae. aegypti by a recombinant SINV have been recently investigated (Myles et al., 2008; Cirimotich et al., 2009). A recombinant SINV strain was engineered to express the Flock house virus [Nodaviridae; (FHV)] B2 protein, a potent suppressor of RNAi. The 12 kDa B2 protein is essential for the replication of FHV (Chao et al., 2005; Li & Ding, 2005; Wang et al., 2006). B2 protein is a homodimer and indiscriminately binds to double-stranded RNA (dsRNA) molecules independent of their nucleotide sequences and sizes such as siRNAs duplexes and long dsRNAs, thereby protecting dsRNA from being accessed and processed by dicer2 of the RNAi machinery. When mosquitoes were orally infected with SINV-B2, virus titres, midgut infection and escape rates were significantly increased compared to the control virus. Strikingly, SINV-B2 caused high mortality amongst the mosquitoes at 4–6 days post-infection, suggesting that RNAi has the potential to protect the vector from pathogenic effects of replicating arboviruses (Myles et al., 2008; Cirimotich et al., 2009).

In previous studies key genes of the RNAi machinery in Ae. aegypti were temporarily and/or tissue specifically silenced. As a consequence, the effects on arbovirus replication were transient and did not affect mosquito fitness. The aim of this study was to generate transgenic mosquitoes expressing FHV-B2 as tools to study effects of constitutive RNAi suppression in trans on unrelated arboviruses such as SINV (recombinant strain: TR339eGFP) and DENV2 (strain: QR94). The QR94 strain belongs to the American genotype of DENV2 and has been shown to inefficiently disseminate from the mosquito midgut following oral acquisition (Salazar et al., 2010). SINV-TR339eGFP has been previously shown to encounter a dose-dependent midgut escape barrier (MEB) in Ae. aegypti, which was diminished when the virus titre in the midgut was increased (Kramer et al., 1981; Khoo et al., 2010). In this study, we wanted to assess whether both viruses would respond similarly to B2-mediated RNAi suppression in Ae. aegypti by increasing their replication efficiencies, eventually reaching concentrations high enough to become detrimental for the mosquito. We also wanted to observe whether stable RNAi suppression would allow SINV-TR339eGFP and DENV2-QR94 to disseminate more efficiently from the mosquito midgut, which would be an indication that the viruses are confronted with a dose-dependent MEB (Kramer et al., 1981).

Here we describe the generation of transgenic Ae. aegypti that constitutively express FHV-B2 under control of the Ae. aegypti poly-ubiquitin (PUb) promoter (Anderson et al., 2010). To obtain a predictable transgene expression phenotype, transgenic mosquitoes were generated using the ΦC31 site-directed integration system, which has recently been adapted for this purpose (Franz et al., 2011). SINV-TR339eGFP and DENV2-QR94 infection patterns in the RNAi impaired mosquitoes were evaluated and compared with those in nontransgenic control mosquitoes.

Results

Generation of ΦC31 recombinants of Aedes aegypti expressing FHV-B2 from the PUb promoter

As shown in Fig. 1A, FHV-B2 expression was under control of the constitutive Ae. aegypti PUb promoter (Anderson et al., 2010). The gene-of-interest expression cassette was inserted into the attB site-containing donor for site-specific integration into docking strain attP26 as described before (Franz et al., 2011). Following co-injection of 1828 preblastoderm embryos with in vitro transcribed ΦC31 integrase mRNA (600 ng/μl) and the modified attB donor plasmid (300 ng/μl), 196 G0 individuals (100 females, 96 males) survived. After out-crossing to the Higgs white eye (HWE) recipient strain, 32 male and four female pools were established. One pool, P61, consisting of 29 G0 females, produced offspring that showed both enhanced cyan fluorescent protein (ECFP) and DsRed eye marker expression indicating recombination between the attP site of the docking strain and the attB site of the donor plasmid (Fig. 1A). The estimated transformation frequency was 1%.

Figure 1.

Molecular characterization of ΦC31-generated recombinants of Aedes aegypti over-expressing FHV-B2 from the constitutive poly-ubiquitin (PUb) promoter. (A) Diagram of the recombination event between the attP site of docking strain attP26 and the attB site of the donor pattB-DsRed2/PUb/v5B2/svA. (B) Confirmation of site-directed donor integration in PUbB2 P61 mosquitoes by conventional PCR. Annealing sites of PCR primers are shown. pB left, pB right: left and right arm of piggyBac transposable element; ECFP, enhanced cyan fluorescent protein; 3xP3, eye-tissue specific 3xP3 promoter; attP, phage attachment site; SVA, transcription termination signal of Simian Virus 40; attB, bacterial attachment site; v5B2, v5-tagged FHV-B2; PUb promoter, Aedes aegypti poly-ubiquitin promoter; attR, attachment site right; attL, attachment site left.

Molecular characterization of line PUbB2 P61

To confirm the ΦC31-based recombination event amongst PUbB2 P61 mosquitoes at the molecular level, we PCR amplified genomic DNA of regions encoding the recombination sites, which were expected to be converted from attP and attB into attR and attL. Previously, it was reported that docking strain attP26 contains two attP site integrations (Franz et al., 2011). As shown in Fig. 1B, PCR data revealed that only one of the two attP sites in PUbB2 P61 mosquitoes was converted into attR and attL by recombination with the attB site of the donor whereas the other attP site remained intact. This indicates that there was a single donor integration event in PUbB2 P61 mosquitoes even though these mosquitoes could have potentially accommodated two ΦC31-mediated donor integrations.

FHV-B2 expression in PUbB2 P61 mosquitoes

v5-tagged B2 protein was readily detected by Western blots in midguts of sugar-fed PUbB2 P61 females at 5–7 days post-eclosion but not in the HWE (negative) control (Fig. 2A). In carcasses, however, B2 protein was only faintly detected at similar time points. Similarly, B2 was readily detected in midguts but not in carcasses of mosquitoes that had received a bloodmeal before (data not shown). Presence of B2 antigen in midguts of sugar- or blood-fed females was confirmed in immunofluorescence assays. As shown in Fig. 2B, v5-tagged B2 antigen was detected in the entire midgut epithelium at different time points following sugar-feeding, blood-feeding or DENV2 infection. These results confirm that FHV-B2 was present in midguts of PUbB2 P61 mosquitoes regardless of sugar- or blood-feeding, whereas in carcasses B2 was barely detectable.

Figure 2.

FHV-B2 over-expression in midguts of transgenic mosquitoes of line PUbB2 P61. (A) Western blot analysis of PUbB2 P61 females to detect FHV-B2 expression. Total protein extracted from five midguts or carcasses of sugar-fed PubB2 P61 and HWE (negative control) mosquitoes was electrophoretically separated in 10% SDS-PAGE. B2 over-expression was detected with an IgG antibody specific for the v5 epitope tag. Total protein of SINV-TE3′2J-B2 infected cell culture medium was used as positive control. Lane 1: HWE midguts, 5 days post-eclosion; lane 2: PUbB2 P61 midguts, 5 days post-eclosion; lane 3: PUbB2 P61 midguts, 7 days post-eclosion; lane 4: SINV-TE3′2J-B2 culture supernatant, 36 h post-infection; lane 5: molecular size marker; lane 6: PUbB2 P61 carcasses, 1 day post-eclosion; lane 7: PUbB2 P61 carcasses, 5 days post-eclosion; lane 8: PUbB2 P61 carcasses, 7 days post-eclosion. (B) Detection of FHV-B2 protein in midguts by immunofluorescence assay using anti-v5 IgG as detection antibody. (1) Sugar-fed HWE (negative control); (2) sugar-fed PUbB2 P61 female; (3) PUbB2 P61 female at 5 days post-bloodmeal (pbm); (4) DENV2-QR94 infected female at 5 days pbm; (5) close-up image (400×) of midgut epithelium over-expressing FHV-B2 in a sugar-fed PUbB2 P61 female at 7 days post-eclosion; (6) close-up image (400×) of midgut epithelium of a sugar-fed HWE female at 7 days post-eclosion.

FHV-B2 expression in PUbB2 P61 mosquitoes suppresses induced silencing of AeCPA-1 in midguts of blood-fed females

The ability of v5B2 to suppress RNAi in midguts of PUbB2 P61 females was analysed in a functional assay in which induced silencing of the midgut-specific (Ae. aegypti) AeCPA-1 gene was monitored over time. Females were intrathoracically injected with 1 μg dsRNA targeting AeCPA-1 (Isoe et al., 2009). Two days later, females received a non-infectious bloodmeal to induce AeCPA-1 transcription. Abundance of AeCPA-1 transcripts in midguts was monitored between 7–48 h post-bloodmeal (pbm). Within 24 h pbm, AeCPA-1 transcript copy numbers per 100 ng of total midgut RNA increased fourfold in phosphate-buffered saline (PBS)-injected or non-injected HWE and PUbB2 P61 females (Fig. 3). The highest AeCPA-1 transcript copy number (7.0 × 106) was detected at 24 h pbm amongst midgut RNA of non-injected HWE. After 24 h pbm, AeCPA-1 transcription levels steadily declined until reaching a base level of <10 000 copies by 48 h pbm. Following dsRNA injection, AeCPA-1 transcript copy numbers in midguts of HWE mosquitoes decreased at least 10-fold during the 48 h observation period, indicating robust silencing of the gene (Fig. 3). In midguts of similarly treated PUbB2 P61 females, AeCPA-1 transcription was reduced less than twofold at 24 h pbm, which indicates RNAi suppression through functional activity of B2.

Figure 3.

Functional assay to confirm suppression of induced silencing of the endogenous Aedes aegypti carboxypeptidase A-1 (AeCPA-1) gene in midguts of PUbB2 P61 females. HWE and PUbB2 P61 females were intrathoracically injected with 1 μl of 1 μg/μl dsRNA targeting AeCPA-1 or with phosphate-buffered saline (PBS; control). Twenty-four hours post-injection mosquitoes were given a bloodmeal. Following total RNA extraction from pools of 20 midguts at 7–48 h post-bloodmeal, AeCPA-1 transcript copy number equivalents were quantified by quantitative RT-PCR. NI, non-injected.

The results show that in midguts of PUbB2 P61 females, expression of v5B2 suppressed silencing of an endogenous gene with an efficiency of >50% during the observation period. This is a strong indication that in principle, B2 in midguts of PUbB2 P61 females was functional, which was confirmed in a second independent experiment (data not shown). We speculate that the inability of B2 to completely suppress silencing of an endogenous gene in PUbB2 P61 mosquitoes is an effect of B2 expression in these mosquitoes rather than being caused by the intrinsic nature of the B2 protein. It is possible that slight imbalances in the dsRNA : B2 protein ratio could affect the efficiency of gene silencing or silencing suppression.

Challenging of PUbB2 P61 mosquitoes with SINV and DENV2

After oral challenge, SINV-TR339eGFP mean titres were significantly increased at 7 days pbm in midguts of PUbB2 P61 females (P-value = 0.0457; ANOVA, Tukey−Kramer test) as compared to the HWE control (Fig. 4A). Infection rates and mean virus titres of carcasses did not differ significantly at 7 days pbm. At 14 days pbm, viral infection rates were significantly increased in carcasses (P-value = 0.0471; Chi-square test) but not in midguts (Fig. 4B), indicating that B2-mediated suppression allowed SINV-TR339eGFP to overcome the dose-dependent MEB more efficiently.

Figure 4.

Intensity of SINV-TR339eGFP and DENV2-QR94 infections in PUbB2 P61 and HWE mosquitoes. (A) Virus titres and infection rates at 7 days post-bloodmeal (pbm); (B) virus titres and infection rates at 14 days pbm. One week post-eclosion, females were orally challenged with SINV-TR339eGFP (titre in the bloodmeal: 2.0 × 107 pfu/ml) or DENV2-QR94 (titre in the bloodmeal: 1.0 × 106 pfu/ml). Each data point represents the virus titre (pfu/ml) in an individual midgut or carcass as analysed by plaque assays. P-values (α = 0.05) for infection rates and mean virus titres are shown in the tables.

We then challenged PUbB2 P61 females with the QR94 strain of DENV2, which belongs to the American genotype of DENV2 (Lorono-Pino et al., 2004). DENV2 strains of the American genotype have been shown to disseminate inefficiently from the midgut in Ae. aegypti (Armstrong & Rico-Hesse, 2003; Anderson & Rico-Hesse, 2006; Salazar et al., 2010). RNAi suppression in PUbB2 P61 females did not enable DENV2-QR94 to overcome the MEB because mean virus titres and infection rates in carcasses remained low for the transgenic and control (HWE) mosquitoes at both time points (Fig. 4A, B). However, similar to SINV-TR339eGFP, mean titres of DENV2-QR94 were significantly increased at 7 days pbm (P-value = 0.0172; ANOVA, Tukey−Kramer test) in midguts of PUbB2 P61 females when compared to the HWE control, even though mean DENV2-QR94 titres in midgut tissue of PUbB2 P61 mosquitoes remained 1–2 logs plaque forming units (pfu)/ml lower at both time points than mean titres of SINV-TR339eGFP. At 14 days pbm, DENV2-QR94 midgut infection rates were significantly increased in the transgenic mosquitoes (P-value = 0.0321; Chi-square test). In summary, FHV-B2 expression in line PUbB2 P61 potentially increased the vector competence for SINV-TR339eGFP but not for DENV2-QR94. SINV-TR339eGFP mean titres were significantly increased in midgut tissue at 7 days pbm, which then led to increased disseminated infection rates at 14 days. In contrast, significantly increased DENV2-QR94 mean titres in mosquito midguts at 7 days pbm did not affect virus dissemination rates at 14 days pbm.

Discussion

The underlying hypothesis of this study was that FHV-B2-mediated suppression of the RNAi pathway in Ae. aegypti would strongly increase arboviruses titres and infection rates. As a result the viruses would become pathogenic to the vector if not kept in check by the antiviral immune pathway. Previously, it was shown that transgene-mediated silencing of Aa-dcr2 in the midgut of Ae. aegypti led to statistically significant higher SINV-TR339eGFP titres as well as midgut infection and dissemination rates (Khoo et al., 2010). However, because of the promoter used to drive effector gene expression in these transgenic mosquitoes, the RNAi impairment was temporal and tissue-specific. As a consequence, arboviruses did not produce titres in the mosquitoes high enough to affect overall vector fitness. We optimized our strategy regarding transgene design and expression in three ways to increase transgene-mediated suppression of RNAi in Ae. aegypti. (1) We chose to over-express a strong RNAi suppressor, FHV-B2, which had been shown before to inhibit siRNA biogenesis in Ae. aegypti (Myles et al., 2008; Cirimotich et al., 2009); (2) we used the Ae. aegypti PUb promoter to constitutively over-express B2 in mosquito midguts and secondary tissue; (3) we generated transgenic mosquitoes using the ΦC31 site-directed integration system to ensure predictable transgene expression in each mosquito generation.

Constitutive over-expression of B2 in PUbB2 P61 mosquitoes suppressed the RNAi pathway in the mosquitoes as shown by the reduced efficiency of dsRNA targeting AeCPA-1 to silence its mRNA in female midguts after blood-feeding. Although its original function is to render the replication intermediate (dsRNA) of FHV inaccessible to dicer2, the B2 protein is able to inhibit the activity of any dsRNA species present in the cytoplasm of an invertebrate cell (Lingel et al., 2005). As the PUb promoter drives constitutive gene expression in midguts and to a lesser extent in secondary tissues as shown earlier and also in this study, B2-mediated suppression of RNA silencing of midgut-specific AeCPA-1 was not surprising (Anderson et al., 2010). Furthermore, we showed that two unrelated arboviruses, SINV-TR339eGFP and DENV2-QR94, responded similarly to RNAi suppression in midgut tissue of PUbB2 P61 mosquitoes by producing significantly increased mean virus titres at 7 days pbm. At 14 days pbm, PUbB2 P61 females infected with DENV2-QR94 had significantly increased midgut infection rates, whereas carcass infection rates of this virus were not affected. At the same time point, however, SINV-TR339eGFP had significantly increased carcass infection rates in the B2-expressing mosquitoes, indicating that the virus overcame the dose-dependent MEB more readily. It is possible that DENV2-QR94 failed to overcome the MEB in PUbB2 P61 mosquitoes because the overall virus dose in midgut tissue of the transgenic mosquitoes was still low at 7 days pbm, not exceeding 420 pfu/ml, ∼fivefold higher than in the HWE control. Thus, the relatively low virus titre DENV2-QR94 generated in Ae. aegypti appeared to be a phenomenon that was primarily not controlled by the mosquito's RNAi response.

In conjunction with earlier observations described below, the results of this study suggest that the efficiency of RNAi suppression depends on the concentration of B2 molecules being expressed in the cell. Following this scenario, the ratio amongst dicer2, B2 and viral RNA molecules in a cell would dictate how virus replication becomes affected. In an earlier germline transformation experiment we generated a transgenic mosquito line, Carb/B2-133, which over-expressed FHV-B2 from the midgut-specific, bloodmeal-inducible AeCPA promoter (Edwards et al., 2000; Moreira et al., 2000). As this line was produced by using a transposable element (TE) instead of the ΦC31 site-directed recombination system, we had no control over the number of transgene integration events per cell. In early generations Carb/B2-133 showed an unusually strong eye marker expression level and when challenged with SINV-TR339eGFP, the virus produced significantly higher titres and infection rates in the transgenic line than in the HWE control (see Fig. S1). However, following three generations of outcrossing to the recipient strain, mosquitoes of this line displayed weak eye-marker expression levels and diminished effects on SINV-TR339eGFP titres and infection rates. We speculate that line Carb/B2-133 originally contained multiple transgene copies, which substantially enhanced B2 expression levels and as a consequence strongly increased virus infection rates and titres. However, the multi-copy transgene integration pattern was unstable in line Carb/B2-133 and therefore lost after three generations.

Previous reports showed that FHV-B2 expression from recombinant SINV-TE3′2J caused a dramatic increase in virus titres associated with significant mortality of infected HWE mosquitoes (Myles et al., 2008; Cirimotich et al., 2009). The authors showed that complete RNAi impairment owing to in cis expression of the RNAi suppressor caused the recombinant arbovirus to become pathogenic for the mosquito. When we infected PUbB2 P61 mosquitoes with SINV or DENV2, we did not observe any increased levels of mortality amongst the mosquitoes (see also Fig. S2). The mode of B2 production in the two systems could account for the observed differences. B2 expression from a recombinant virus leads to RNAi pathway suppression in each virus-infected cell because each viral RNA generating cell also produces equivalent quantities of B2 (Myles et al., 2008; Cirimotich et al., 2009). However, when expressing B2 in trans using PUbB2 P61 mosquitoes it is possible that not every cell that is infected by a virus also produces sufficient concentrations of B2 to completely suppress RNAi, thereby enhancing virus replication.

The only other example of transgene-mediated FHV-B2 expression in insects is that of Drosophila, in which B2 was expressed from the Daughterless, actin5C or hsp promoter (van Rij et al., 2006; Chou et al., 2007; Berry et al., 2009). Injection of the transgenic flies with Drosophila C virus (DCV; Dicistroviridae) lowered the 50% lethal dose of the virus by ∼280-fold as compared to the similarly infected, nontransgenic control. The virus is lethal when injected into flies but not when orally acquired (Lautie-Harivel, 1992). In addition, DCV encodes a potent suppressor of RNAi, DCV-1A at the N-terminus of ORF 1 (van Rij et al., 2006). Thus, the presence of both virus-encoded and heterologous RNAi suppressors in the insect cell could have increased the pathogenic potential of DCV. Previous investigations suggest that arboviruses do not encode strong suppressors of RNAi (Myles et al., 2008; Cirimotich et al., 2009). Thus, effects of RNAi suppression in the Drosophila−DCV system might not be directly comparable to those in the mosquito−arbovirus system.

Experimental procedures

Transgene construction and generation of transgenic mosquitoes

The template of the v5-tagged FHV-B2 cDNA originated from recombinant SINV-TE3′2Jv5B2 (Cirimotich et al., 2009). In plasmid pSLfa1180fa-AeCPA/EGFP/svA (Franz et al., 2011), the AeCPA promoter and EGFP encoding fragments were exchanged for PUb and v5B2. The resulting PUb/v5B2/svA expression cassette was then transferred into modified pattB-DsRed2 (Nimmo et al., 2006) using the AscI sites to generate plasmid pattB-DsRed2/PUb/v5B2/svA. Transgenic Ae. aegypti were generated using the ΦC31 site-directed integration system and the HWE-based docking strain attP26 as described before (Franz et al., 2011). The transgenic Ae. aegypti line was named PUbB2 P61.

Molecular characterization

Site-directed donor integration by recombination between attB of the donor plasmid and attP of docking strain attP26 was confirmed in a conventional PCR assay. Total DNA was extracted from single females of lines attP26 and PUbB2 P61 using the DNeasy Blood & Tissue kit (Qiagen, Valencia, CA, USA). For each PCR reaction 500 ng of total DNA was used. The (nonrecombined) attP site was detected using primers attP FWD: 5′CAAATGTGTTCTGTGATGACCTG3′ and attP REV: 5′CTCCCTTGCTACTGACATTATGG3′. Recombined attL and attR sites were detected using primer pairs attL FWD 5′GAGGTCGACGATGTAGGTCAC3′, attL REV 5′ACCTTTTCTCCCTTGCTACTGAC3′ and attR FWD 5′TCAAACTAAGGCGGAGTGG3′, attR REV 5′GATGGGTGAGGTGGAGTACG3′, respectively (Nimmo et al., 2006).

Transient silencing of AeCPA-1

Total RNA was extracted from 10 Ae. aegypti females using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). A 490 bp cDNA template of the coding region of the AeCPA-1 gene (VectorBase accession no. AAEL010782) was amplified from total RNA by reverse-transcription PCR (RT-PCR) using primers CPA FWD (5′GACTTATGGAAACGAAGCCGCT3′) and CPA REV (5′ACAAATATGCCTGGATTTCCCGTTTTG3′) and the SuperScript One-Step RT-PCR kit with Platinum Taq (Invitrogen). The amplicon was cloned into pCR II-TOPO (Invitrogen). cDNA obtained from this cloning step was PCR amplified using primers that contained T7 promoter sequences at their 5′ ends (T7-CPA FWD: 5′-TAA TAC GAC TCA CTA TAG GGG ACT TAT GGA AAC G-3′ and T7-CPA REV: 5′-TAA TAC GAC TCA CTA TAG GGA CAA ATA TGC CTG G-3′; T7 sequences are underlined). The resulting amplicon was in vitro transcribed using the Ambion MEGAscript Kit (Invitrogen) to produce AeCPA-derived dsRNA. Seven-day-old HWE and PUbB2 P61 females were intrathoracically injected with 1 μl of 1 μg/μl AeCPA-1 dsRNA each. Controls were either not injected or injected with PBS. Mosquito midguts were dissected at 7, 12, 24, 36 and 48 h post-injection. Total RNA was extracted from pools of 20 midguts using TRIzol Reagent and used for quantitative RT-PCR (qRT-PCR).

qRT-PCR

Using the AeCPA-1-specific primers CPA qRT FWD (5′CAT ACC GGT GAG CAC GCC TC3) and CPA qRT REV (5′GTG TAG GCA ATC TTT ACG TCC AAA G3′), a 185 bp cDNA fragment of the gene was PCR amplified from total RNA of Ae. aegypti, which differed from the cDNA fragment used as template for dsRNA production. Following pCR II-TOPO cloning of the amplicon, the resulting plasmid was serially diluted and used as the standard curve to calculate transcript copy number equivalents per ng of total RNA. Using the QuantiFast SYBR Green RT-PCR Kit (Qiagen), qRT-PCR reactions contained the above primers at 10 μM concentration and 100 ng of total RNA from mosquito midguts. qRT-PCR reactions were conducted in a IQ5 Real-Time PCR Detection system (Bio-Rad, Hercules, CA, USA) according to the following protocol: 10 min at 50 °C, 5 min at 95 °C and 35 cycles each of 10 s at 95 °C and 30 s at 60 °C.

Immunofluorescence assay

Midguts of PUbB2 P61 and HWE mosquitoes were dissected and stored in 4% formaldehyde (Vector, Burlingame, CA, USA) for 3 days at 4 °C. Fixed midguts were washed in PBS and permeabilized with PBT (PBS with 1% 10× bovine serum albumin and 2% Triton X-100) for 1 h at room temperature. The v5B2 protein was detected by incubating midguts with the mouse anti-v5 IgG antibody (Invitrogen) at a dilution of 1:200 for 1.5 h at 37 °C. Midguts were subsequently incubated with a biotin-labelled anti-mouse secondary antibody (Sigma-Aldrich, St Louis, MO, USA) at a dilution of 1:200 for 1.5 h at 37 °C. Midguts were incubated with streptavidin-fluorescein at a dilution of 1:100 in PBS before being washed in PBS and mounted onto glass slides.

Immunoblot analysis

Immunoblot analysis was conducted according to Cirimotich et al. (2009). Briefly, midguts (n = 5) of PUbB2 P61 and HWE mosquitoes were dissected and homogenized in 10 μl of sample buffer [NuPAGE LDS Sample Buffer with NuPAGE Reducing Agent (Invitrogen)] using a handheld tissue grinder with pestle. Ten microlitres of the total protein (∼180 ng/lane) was electrophoretically separated in a 10% SDS-PAGE gel (Invitrogen) and transferred onto a nitrocellulose membrane at 30 volts for 1 h. Membranes were blocked for 1 h at room temperature in blocking buffer (PBS with 0.05% Tween 20 and 5% nonfat dry milk) and incubated at 4 °C overnight with a mouse anti-v5 IgG antibody at a dilution of 1:5000 in blocking buffer. Membranes were then incubated with a horseradish peroxidase-conjugated goat anti-mouse IgG secondary antibody (KPL Inc., Gaithersburg, MD, USA) diluted 1:1000 in blocking buffer for 1 h at room temperature. The Vector VIP substrate Kit for Peroxidase (Vector) was used to develop the membrane.

Virus propagation

SINV-TR339eGFP stocks were generated according to Khoo et al. (2010). SINV was grown in Vero cells at 0.01 multiplicity of infection (m.o.i.) and harvested 36–38 h post-infection by removing the cell culture medium. DENV2-QR94 is an American genotype DENV2 strain, which was collected in 1994 from a patient in Quintana Roo State, Mexico (Lorono-Pino et al., 2004). Stocks of the virus were generated in Aedes albopictus (C6/36) cells by infecting cells at 0.01 m.o.i., changing the cell culture medium 6 days post-infection and maintaining the cells until 12 days post-infection. Virus was harvested by removing the cells’ culture medium and scraping off the cells.

Oral virus infection of mosquitoes

Until blood-feeding, adult mosquitoes were reared on raisins and water. Each 2.5 L carton contained 100 females and five males. Raisins and water were removed from cartons 36 and 5 h, respectively, prior to blood-feeding. To infect one-week-old females with SINV-TR339eGFP or DENV2-QR94, defibrinated sheep blood was mixed at a 1:1 ratio with freshly harvested virus. The SINV titre in bloodmeal was 2.0 × 107 pfu/ml and that of DENV2 was 1.0 × 106 pfu/ml. Female mosquitoes were fed for 1 h with 2 ml of bloodmeal at 37 °C using one glass feeder per carton. After blood-feeding, only those females that were three quarters or fully engorged were selected. These females were reared in 470 ml cartons and fed with sucrose and water until further analysis.

Determination of virus titres by plaque assay

Virus titres from individual midguts and carcasses were determined by plaque assay at 7 and 14 days pbm. SINV-TR339eGFP- and DENV2-QR94-infected samples were homogenized in 0.5 ml 7% fetal bovine serum (FBS)-complemented MEM and DMEM medium, respectively. Homogenized samples were filtered with Acrodisc HT Tuffryn 0.2 μm syringe filters (Pall Life Sciences, East Hills, NY, USA). Vero cells (for SINV) or LLC-MK2 (Rhesus monkey kidney) cells (for DENV2) were seeded into 24-well plates and left for 3 days to achieve confluence. Cells were infected with 10-fold serial dilutions of individual midgut or carcass homogenates. Cells were incubated for 1 h at 37 °C before being overlaid with an agarose nutrient mixture [1× Medium 199 (Sigma-Aldrich, St Louis, MO, USA), 10% FBS, 4% NaHCO3, 0.5% MEM vitamins, 0.5% MEM amino acids (Mediatech Inc., Manassas, VA, USA); 1% Hanks−DMEM medium (only used for DENV2)]. For SINV, plates were incubated at 37 °C for 4 days and for DENV2, plates were incubated at 37 °C for 12 days. Cells were then stained with 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (Sigma-Aldrich, St Louis), incubated at 37 °C for 24 h and the number of plaques was counted for each sample. Virus titres of individual mosquitoes were calculated as pfu/ml.

Statistical analysis

Statistical analyses were performed using SAS Statistical Analysis Software (SAS Institute Inc., Cary, NC, USA). The MIXED procedure was used for restricted maximum likelihood parameter estimation with incomplete data followed by the Tukey−Kramer test. SINV-TR339eGFP and DENV2-QR94 infection levels were normalized by log10 transformation. Infection rates were analysed with the Chi-square test.

Acknowledgements

We thank J. zumBrunnen for help with statistical analyses and C. Meridith for providing stocks of HWE eggs. This work was funded by R01 grant AI073298-01 from the National Institutes of Health.

Ancillary