|Home | About | Journals | Submit | Contact Us | Français|
Controlled sex-, stage- and tissue-specific expression of anti-pathogen effector molecules is important for genetic engineering strategies to control mosquito-borne diseases. Adult female salivary glands are involved in pathogen transmission to human hosts and are target sites for expression of anti-pathogen effector molecules. The Aedes aegypti 30K a and 30K b genes are expressed exclusively in adult female salivary glands and are transcribed divergently from start sites separated by 263 nucleotides. The intergenic, 5’- and 3’-end untranslated regions of both genes are sufficient to express simultaneously two different transgene products in the distal-lateral lobes of the female salivary glands. An anti-dengue effector gene, Mnp, driven by the 30K b promoter, expresses an inverted-repeat RNA with sequences derived from the premembrane protein-encoding region of the dengue virus serotype 2 genome and reduces significantly the prevalence and mean intensities of viral infection in mosquito salivary glands and saliva.
Genetically-modified mosquitoes are being developed as tools for both population suppression and population replacement to control vector borne diseases (Terenius et al., 2008; Fu et al., 2010). For the latter, it is hypothesized that replacing existing vector populations with ones that are unable to transmit specific pathogens will result in less disease and death. Significant advances have been made towards making transgenic mosquitoes carrying synthetic genes that confer resistance to dengue virus infection. Hammerhead ribozymes designed to fragment the DENV-2 genome are effective in reducing by >100-fold virus production in cultured cells (Nawtaisong et al., 2009). Female Ae. aegypti expressing in their midguts an RNAi-inducing inverted-repeat RNA derived from the pre-membrane (prM) protein-encoding region of the DENV-2 RNA genome have significant reductions in viral titers and a reduced ability to transmit the corresponding virus (Franz et al., 2006). However, expression of multiple effector molecules in more than one mosquito tissue likely is required to reduce intensities of infection to zero. Additionally, multiple effector molecules are necessary to avoid the potential selection for viral resistance to a particular anti-viral effector molecule or targeting mechanism.
The replication of dengue viruses in the midgut of Ae. aegypti after ingestion of a viremic blood meal is followed by pathogen dissemination and infection of multiple mosquito tissues, including the salivary glands, which are essential organs for transmission of viruses to a vertebrate host and the completion of the infection cycle. Previous efforts to characterize functionally the promoters of Ae. aegypti salivary gland genes, for example, those of Maltase-like I (MalI) and Apyrase (Apy) genes, achieved only low transgene expression indicative of weak promoter activity (Coates et al., 1999). Similar results were seen in transgenic Anopheles stephensi using an Apyrase gene promoter from An. gambiae (Lombardo et al., 2005). Abundant expression of a marker gene in mosquito salivary glands was first reported in transgenic An. stephensi using the promoter of the Anopheline anti-platelet protein (AAPP) gene, a member of the 30K gene family (Yoshida et al., 2006). Here we show tissue- and sex-specific expression of two different transgene products under the control of the functionally bi-directional Ae. aegypti 30K gene promoter. Expression of an inverted-repeat RNA based on sequences derived from the premembrane protein-encoding gene of DENV-2 (Franz et al., 2006) in the distal-lateral lobes of adult female salivary glands, reduces significantly virus titers in mosquito salivary glands.
The Ae. aegypti 30K genes encode 30kDa salivary proteins that were described first as antigens causing allergic reactions to mosquito bites in humans (Simon and Peng, 2001, Valenzuela et al., 2002). The 30K proteins are members of a family characterized by the presence of Gly/Glu (GE)-rich amino acid repeats. Transcriptome analyses revealed that these genes encode abundant secreted proteins in the salivary glands of mosquitoes (Valenzuela et al., 2002, Ribeiro et al., 2007). The 30K proteins in An. stephensi (AAPP) and one (Aegyptin) in Ae. aegypti have functional similarity as components of the mosquito saliva that inhibit platelet aggregation by binding collagen during blood feeding (Calvo et al., 2007; Yoshida et al., 2008). There are three members of the 30K gene family in the genome of Ae. aegypti, all located in the genomic supercontig 1.464 (http://aaegypti.vectorbase.org). Two of these, designated 30K a (AAEL010228) and 30K b (AAEL010235, Aegyptin), are transcribed divergently from transcription start sites separated by a DNA fragment 263 base pairs (bp) in length (Figure 1). A third gene belonging to the 30K family (AAEL010231) is located at a distance of ~14 kilobase-pairs (kb) from 30K b (AAEL010235).
Functional analysis of the 30K a and 30K b putative cis-regulatory DNA was performed using mosquito transgenesis. The primary structure of the genes was established by aligning ESTs available at NCBI and Vectorbase with the published genomic DNA sequence (Nene et al., 2007) (Figure 1). In addition to the 263 bp intergenic region, cis-regulatory sequences of the 30K a gene may be located in the 43 bp 5’-end untranslated region (UTR) or the 68 bp 3’-end UTR. Similarly, 30K b gene cis-regulatory elements may reside in the 56 bp 5’-end UTR or 67 bp 3’-end UTR. A 35 bp deletion was discovered in the 5’UTR of the 30K b gene of Higgs white-eye strain mosquitoes when compared to the published sequence of the Liverpool strain (Nene et al., 2007), making the 5’-end UTR 21 bp in length. The transgene-based results obtained here support the conclusion that the tissue- and sex-specific expression profiles of the 30K a and 30K b genes in these two mosquito strains are identical despite this sequence difference. A TATA box and putative arthropod initiation factor sequence (INR, based on the consensus TCAKTY) for 30K a are present 66 bp and 42 bp, respectively, to the 5’-end of the start codon (Arnosti, 2003; Butler and Kadonaga, 2002) (Figure 1). The TATA box and INR for 30K b are 48 bp and 17 bp, respectively, to the 5’end of the start codon. Putative transcription factor binding sites, including AP-1 and a heat-shock factor site, were identified by bioinformatic searches of the Transfac database (http://www.cbrc.jp/research/db/TFSEARCH.html) (Heinemeyer et al., 1998), however their functionality remains to be investigated.
Three different promoter-reporter constructs based on the Higgs strain gene structure were tested in mosquitoes (Figure 2). pMos-30K consists of the 30K intergenic DNA sequence, 30K a 5’-end UTR followed by a truncated 30K a gene fused to the SV40 polyadenylation signal sequence, and the 30K b 5’-end UTR followed by a truncated 30K b gene fused to Enhanced Green Fluorescent Protein (EGFP) open reading frame (ORF) and the bovine growth hormone (BGH) polyadenylation signal. Two additional transformation constructs, 30KExGM and GFP30KMnp, consist of the intergenic sequence, 30K a 5’- and 3’-end UTR, full length 30K a ORF fused to the EFGP ORF, the 5’ and 3’UTR for 30K b and an anti-dengue effector molecule (a modified prM protein-encoding viral RNA, Mnp; Franz et al., 2006). Flanking genomic DNA of 365 bp adjacent to 30K b 3’UTR and 602 bp adjacent to 30K a 3’UTR also were incorporated. Additionally, 30KExGM includes the first exon for 30K b and the inverted repeat sequence targeting the DENV-2 prM gene separated by a 54 bp linker sequence. The constructs were cloned into the Mos1 mariner-based plasmid, pMos3XP3DsRed, containing the Discosoma sp. red fluorescent protein (DsRed) marker gene (Horn et al., 2002). Each construct was mixed separately with Mos1 mariner helper plasmids (Horn et al., 2002) and injected into preblastoderm embryos from the Ae. aegypti Higgs white-eye recipient strain (Jasinskiene et al., 1998; Coates et al., 1998; Higgs et al., 1996).
A total of 1200 embryos were injected with pMos-30K and 72 males and 53 females survived. The G0 adults were outcrossed with the Higgs recipient strain, G1 larvae screened visually for DsRed fluorescence in eyes, and one transgenic line, P2, was recovered. Similarly, 822 and 800 embryos were injected with constructs 30KExGM and GFP30KMnp, respectively. Following injection of 30KExGM, 105 males and 73 females survived, and 128 males and 75 females were recovered from injections of GFP30KMnp. Five lines, 1, 15, 22, 27 and P4, containing the transgene 30KExGM and one line, P6, carrying GFP30KMnp were generated. Southern blot analyses support the presence of one transgene integration in pMos-30K line P2, two inserts in 30KExGM lines 1, 22, 27, P4 and GFP30KMnp P6, and three in 30KExGM line 15 (Supplemental Figure 1). Independent assortment of multiple inserted transgenes can occur during the initial out-crossing of the transgenic mosquitoes. Because pools of mosquitoes were used for Southern blot analyses, low representation of one particular insert can result in the difference in the intensity of the signals seen in lines 27 and P6 (Supplemental Figure 1).
Dissection and examination of the salivary glands of transgenic females from pMos-30K line P2 show that EGFP fluorescence due to the fusion protein 30K b-EGFP is localized exclusively in the distal-lateral lobes of the salivary glands (Figure 3). Similarly, EGFP fluorescence resulting from the 30K a-EGFP fusion protein accumulation in all five 30KExGM transgenic lines and GFP30KMnp line P6 was detected in the distal-lateral lobes of female salivary glands (Figure 3; data not shown). The transgenic mosquitoes from 30KExGM line 27 were all males and this line was not included in gene expression analysis. No fluorescence was detected in the proximal-lateral and medial lobes of the salivary glands of females from any of the transgenic lines. The absence of the 30K b first exon in GFP30KMnp had no apparent influence on the expression of 30K a-EGFP protein in the salivary glands of transgenic females.
Reverse transcription polymerase chain reaction (RT-PCR) performed to determine expression profiles of the truncated 30K a-SV40 and 30K b-EGFP fusion transgenes in pMos-30K line P2 demonstrated the presence of 30K b-EGFP transcripts in the head, thorax and abdomen of the female transgenic mosquitoes (Supplemental Figure 2). Amplicons also were detected in male transgenic mosquitoes. No expression of 30K a-SV40 was detected by RT-PCR. Interestingly, immuno-blot analyses of salivary glands from transgenic and control females, carcasses (whole body excluding salivary glands) from transgenic and control females, and transgenic and control males detected EGFP protein only in the salivary glands of transgenic females consistent with the fluorescence patterns seen in whole glands (Supplemental Figure 3). These results provide evidence to support the conclusion that the control sequences used in pMos-30K are sufficient to drive expression of the 30K b-EGFP transgene. However, its transcript accumulation in tissues other than salivary glands might be due to the absence of regulatory DNA responsible for tissue-specific control of gene expression and/or transcript stability and localization. Alternatively, the ectopic expression could be due to a position-effect resulting from the insertion site of the transgene. The lack of expression of the truncated 30K a is consistent with the interpretation that some of the regulatory DNA sequences are missing in this transgene.
Mosquitoes carrying the 30KExGM and GFP30KMnp transgenes had tissue- and sex-specific expression of reporter and effector gene products consistent with the expression profile of the 30K a and 30K b endogenous genes. Northern blot analyses of mRNA isolated from adult males, adult female salivary glands and female carcasses (whole bodies without salivary glands) hybridized with a probe specific to the ORF of the EGFP reporter gene show signal only in the salivary gland samples (Figure 4). These data confirm the ability of the 30K a transgene promoter to drive sex- and tissue-specific expression. Similarly, Northern blot analyses performed with small RNAs isolated from whole adult females (Figure 4), males, adult female salivary glands and female carcasses (Figure 5), and hybridized with a probe specific to the Mnp gene, detected a moiety 21-22 nucleotides in length, consistent with the RNAi-mediated processing of the Mnp anti-dengue effector gene. However, the full-length effector molecule was not detected (Supplemental Figure 4), most likely due to its rapid degradation to the smaller molecular-weight species as a result of RNAi activity (Franz et al., 2006). Salivary gland-specific expression of the EGFP reporter gene from the 30K a promoter and the tissue-specific localization of siRNAs expected from the 30K b-directed expression and subsequent processing of the Mnp anti-dengue effector gene show that transgene constructs based on the 30K a and b cis-regulatory sequences can be used to express two different effector molecules in the adult female salivary glands. These results also define DNA fragments sufficient for this specific expression pattern (Supplemental Figure 5).
Transgenic mosquitoes from 30KExGM lines 1, 15, 22, and P4, GFP30KMnp line P6 and the control Higgs strain were fed a blood meal containing dengue virus serotype 2 (DENV-2). Mosquito carcasses (without salivary glands) were assayed 15 days post infection (dpi) for virus titer and showed a high prevalence of infection (>50%) in both transgenic and control mosquitoes (Figure 9A). Prevalence of infection (range 27-43%) in the salivary glands was significantly reduced (p < 0.01) in 1, 15, 22, and P6 transgenic mosquitoes lines, compared to the control Higgs strain (62.3%). Mean intensities of infection measured as viral titers (plaque forming unit, pfu) were similar among carcasses of all mosquito lines (Figure 9A). In contrast, virus titers in dissected salivary glands of all transgenic mosquitoes lines were significantly lower (p < 0.01, with mean values ranging from 200-1520 pfu/ml) than in the control Higgs group (3100 pfu/ml; Figure 9B). These experiments support the conclusion that the salivary glands of the mosquitoes expressing the RNAi effector gene are less susceptible to the DENV-2 infection. Assays to analyze the transmission potential of transgenic and control mosquitoes were conducted at 14 dpi by allowing them to probe a feeding solution that was subsequently collected and assayed for viral titers. None of the mosquitoes from lines 1 and 15 transmitted detectable numbers of viable DENV-2 virus to the artificial feeding solution (Figure 9C). Only 10%, 11.1% and 12.5% (p < 0.001) of saliva samples from transgenic mosquito lines 22, P6 and P4, respectively, infected the feeding solution compared to a 40% saliva samples from control Higgs females. A significant decrease (p < 0.05) in the average viral titer (0-7.3 pfu/ml) from the five transgenic mosquito lines was observed when compared to the control Higgs group (51.8 pfu/ml). To confirm that the effect on viral infection is a result of the expression of the Mnp effector gene and not due to the presence of EGFP in the salivary glands of transgenic mosquitoes, females from pMos-30K line P2 were injected intrathoracically with DENV-2 at 10 pfu/mosquito (Figure 9D). Infected females were assayed for virus at 7 dpi and no statistically significant difference was observed in viral replication in the salivary glands and carcasses indicating that the EGFP transgene had no effect on virus infection in the transgenic mosquitoes. To bypass the midgut barrier, transgenic and Higgs control females also were intrathoracically injected with 10 or 100 pfu of DENV-2 (Supplemental Figure 5). As observed from the blood-feeding experiment, both the prevalence of infection and the viral titers at 8 dpi in the carcasses of intrathoracically-injected transgenic and control mosquitoes were similar (Supplemental Figure 5A). The prevalence of infection was significantly reduced (p < 0.001) in salivary glands of all transgenic mosquito lines (range 53-78%) compared to the control group (90%) (Supplemental Figure 5B). Viral titers in dissected salivary glands were significantly lower (p < 0.002) in all transgenic mosquito lines with mean values ranging from 300 to 450 pfu/ml compared to Higgs with a mean value of 2000 pfu/ml. Finally, injection with 100 pfu of DENV-2 showed a similar significant difference between Higgs salivary glands and either lines 1 or 15 (Supplemental Figure 5C). These results support the conclusion that the DENV-2 transmission potential of transgenic mosquitoes was reduced strongly as a result of the expression of RNAi effector gene against the virus in the salivary glands.
We report here for the first time a bi-directional promoter in mosquitoes driving expression of two different transgenes. The promoter will be instrumental in the development of genetically-modified Ae. aegypti mosquitoes expressing two different anti-dengue effector genes in the female salivary glands. The expression of two different effector genes in one construct are expected to mitigate issues associated with the loss of linkage through meiotic recombination between the two transgenes. Transgenic mosquitoes expressing the anti-dengue effector gene not only show reduced viral infection and replication in the salivary glands of infected females, they also demonstrated reduced DENV-2 transmission potential. While indirect immunofluorescence assays provide evidence that dengue viruses preferably enter salivary glands through the distal-lateral lobes (Salazar et al., 2007), work is in progress to find additional promoters to express genes in the proximal and medial lobes of the salivary gland as well. We expect that expression of effector genes in multiple tissues will be required to achieve complete resistance of Ae. aegypti to dengue viruses.
A BLAST search of the Ae. aegypti genome (genebuild AaegL1.1) was performed based on the 30K a and b sequence on the VectorBase website (http://www.vectorbase.org) (Altschul et al., 1990). This search revealed that these genes are located on the supercontig 1.464. A third member of the family also was found on the same supercontig.
All oligonucleotide primer sequences for gene constructions and primary sequence validation are listed in Supplemental Table 1. A schematic representation of the final constructs utilized to generate transgenic mosquitoes is shown in Figure 2. The 30K putative promoter was amplified using genomic DNA extracted from Higgs mosquitoes as template and primers 30 K P FP and 30 K P RP that anneal in the first exon of both genes. The generated amplicon was cloned into the pGlow-topo plasmid (Invitrogen) at the 5’-end of the EGFP ORF and the BGH polyadenylation signal. The SV40 signal was amplified from pDsRed-monomer-N1 (Clontech) with primers SpeIsv40 FP and SpeIsv40 RP and cloned at the 3’-end of the 30K a gene in the SpeI site. The construct was amplified from pGlow-topo with primers AscI SV40 F and EcoRV BGH R and cloned into blunted FseI and AscI sites in the pMos3XP3DsRed plasmid (Horn et al, 2002) to generate pMos30K.
30KExGM and GFP30KMnp were constructed by initially cloning the Mnp sequence from plasmid pMos-carb/Mnp/i/Mnp/svA (Franz et al., 2006) into the NotI and BamHI sites of the pSLfa plasmid (Horn et al., 2002). The putative 30K promoter with the 5’UTRs for both 30K a and b genes and complete ORF for 30K a gene was amplified from Higgs genomic DNA, and separate fragments with and without the first 30K b exon were made using primer pairs 30K PstI for and 30K NotI rev exon and 30K PstI for and 30K NotI RP N, respectively. These amplified DNA fragments then were cloned to the 5’-end of Mnp in pSLfa. A 54 bp linker sequence formed by the annealing of NotI sense and antisense oligonucleotides also was cloned between the 30K b first exon and Mnp in the construct 30KExGM. EGFP ORF, 30K a 3’UTR and 30K b 3’UTR were amplified using primer pairs GFP SacII for and GFP PstI rev, 30Ka 250 3utr speI F and 30K a 3UTR sacII RP and 30KbutrBamHIF and 30Kb 3340 3utr EcorI R, respectively, and cloned into the pSLfa plasmid. The constructs were cloned into AscI site in the pMos3XP3DsRed plasmid.
Preblastoderm embryos from Aedes aegypti Higgs white-eyed strain (Higgs et al., 1996) were injected with the pMos30K, 30KExGM or GFP30KMnp constructs and Mos1 mariner helper plasmids, following established protocols (Jasinskiene et. al. 1998, 2007). Lines are maintained by intercrossing transgenic siblings.
Surviving G0 males were mated individually or in groups of three with 10-15 or 45 Higgs females, respectively. G0 females were mated in pools of 7-8 with equal numbers of Higgs males. G1 larvae were screened for red fluorescence in the eyes. Transformed animals were raised and mated with Higgs mosquitoes.
Genomic DNA was extracted from six transgenic or Higgs wild-type females or fourteen transgenic males using the Promega wizard genomic DNA extraction kit. Genomic DNA was digested overnight with the restriction enzymes EcoRI or XhoI. DNA fragments were separated on a 1% agarose gel followed by overnight transfer to a BioRad Zeta probe nylon membrane. The blot was probed with a 700bp DsRed DNA fragment labeled with 32P. The probe was hybridized at 65°C overnight, the blot washed and exposed to x-ray film or a phosphor screen.
Total RNA was extracted from the salivary glands of twelve transgenic females, carcasses (whole-body excluding salivary glands) of six transgenic females, six whole transgenic males, six whole Higgs females or six whole Vg40 transgenic females using Trizol (Invitrogen). RNA was separated on a 1.2% formaldehyde gel for 3 hrs at 70V. The Millenium markers (Ambion) were used as molecular weight standards. The RNA was transferred to Zeta probe nylon membrane overnight. The blot was probed with EGFP or Mnp fragment labeled with 32P. The probes was hybridized at 65°C overnight, washed and exposed to a phosphor screen.
Total RNA was extracted from the heads, thoraces and abdomens of the transgenic females, whole transgenic males or whole Higgs females using Trizol (Invitrogen). 500ng of RNA from each sample were treated with DNAseI (Invitrogen) for 15 minutes at room temperature. The reaction was terminated by addition of 25mM EDTA followed by inactivation of DNAse at 65°C for 10 minutes. One-step RT-PCR was performed using the Qiagen One step RT-PCR kit and 100ng of DNAse-treated RNA as template. Primers AscI SV40 F and 30KA ORF F were used to amplify the 30K a transgene at an annealing temperature of 69°C, and primers EcoRV BGH R and 30KB ORF F were used to amplify transgene 30K b. Actin FP and RP were used to amplify the Aedes actin1 gene sequences (Ibrahim et al., 1996). The annealing temperature of 62°C was used to amplify 30K b and the actin control.
Salivary glands were dissected from 15 transgenic or Higgs mosquitoes and homogenized in 15μL of lysis buffer, which is prepared by dissolving complete mini (Roche) protease inhibitor cocktail (serine, cysteine and metalloprotease inhibitors) and Peflabloc SC (Roche) in water. Two carcasses reserved after the dissection of salivary glands and two males from transgenic or Higgs females were homogenized in 100μL of lysis buffer. An equal volume of Laemlli buffer (BioRad) with 0.1M dithiothreitol (DTT) was added to all samples. Samples were heated at 95°C for 5 min and centrifuged for 5 minutes before loading on a 12% Tris-HCl Ready gel (BioRad). 10μL of BioRad Dual color precision marker were used to calculate the Mr of proteins. Following electrophoresis, proteins were transferred to PVDF membrane using semidry transfer system for 1.5 hr at 33V. Proteins were stained with Coomassie Stain solution (BioRad) and the membrane blocked overnight in 5% nonfat dry milk at 4°C. The membrane was incubated in 1:1000 anti-GFP antibody (Santa Cruz) for 1 hour at room temperature followed by incubation in 1:50,000 alkaline phosphate-conjugated goat anti-rabbit IgG (Jackson Immunosearch). Alkaline phosphatase activity was detected by incubating the membrane in NBT/BCIP (Roche) substrate solution for 30 minutes. All washes and antibody dilutions were made in 1X TBS (150 mM NaCl, 100 mM Tris) with 1% Tween-20.
Salivary glands were dissected from 5-day old transgenic or Higgs wild-type females in 1 X PBS and examined for EGFP fluorescence under a dissecting microscope fitted with UV optics.
Total RNA was extracted from whole mosquitoes, males or females, dissected salivary glands or carcasses (whole body excluding salivary glands) using Trizol (Invitrogen). RNA samples were submitted to electrophoresis on a 15% TBE-urea polyacrylamide gel. After electrophoresis, RNA was transferred to a neutral Hybond Nx membrane for one hour at 20V at 4°C and crosslinked using carbodiimide for 1.5 hrs at 60°C (Pall et al., 2007). The membrane was incubated in UltraHyb hybridization buffer (Ambion) with salmon sperm DNA at 42°C for 1 hour. The 500nt sense fragment of Mnp with T7 promoter on the 3’end was used to prepare in vitro-transcribed 32P labeled RNA probe (MEGAscript kit, Ambion). A total of 7μg of labeled probe was hydrolyzed using carbonate buffer for 2.5 hrs and added to the prehybridized blot in UltraHyb buffer for overnight incubation at 42°C. The membrane was washed in 2 X SSC, 0.1% SDS twice for five minutes, in 0.1 X SSC, 0.1% SDS twice for fifteen minutes each time at 55°C, and exposed to the phosphor screen for image detection. A 22nt commercially-synthesized (Sigma-Aldrich) RNA oligonucleotide 5’AUUUAACCACACGUAAUGGAGA 3’ corresponding to a segment of the sense-strand of Mnp was used as the positive control.
Monkey kidney cells (LLC-MK2; ATCC CCl-7.1) and Aedes albopictus C6/36 cells were obtained from the American Tissue Culture Collection (Prince William County, Virginia). Vertebrate and insect cells were cultured in Modified Eagle's medium (MEM) supplemented with 8% heat inactivated (56°C for 30 minutes) fetal bovine serum (FBS), 1% penicillin/streptomycin, 1% glutamine, 1% non-essential amino acids and maintained at 37° C and 28° C, respectively, in an incubator with 5% CO2. High-passage DENV-2 stock (JAM 1409) was used to propagate virus in C6/36 cells. Briefly, monolayers of C6/36 cells were inoculated with DENV-2 at a MOI of 0.01 and held at 28° C. The supernatant was collected on the day with the highest cytopathic effect (6-8 days post-inoculation). Concentrations of DENV-2 particles were determined by plaque assay.
Seven to eight days after eclosion, female mosquitoes (100–150/one-pint carton) were starved of sucrose and deprived of water for 24 hours prior to blood feeding. The mosquitoes were challenged with an artificial infectious blood meal consisting of 40% (v/v ) defibrinated sheep blood (Colorado Serum Co., Boulder, CO), DENV2 Jamaica 1409-infected C6/36 cell suspension (60 % v/v), and 1 mM ATP. Blood meals were maintained at a constant temperature of 37°C. Mosquitoes were allowed 1 hour to feed. Blood meals were titered post-feed to verify virus concentration. Fully-engorged mosquitoes were selected and held for 14 days at a constant temperature of 28°C and a relative humidity of 80% in an insectary with a 12-hour photoperiod.
Fourteen days after receiving a DENV2-containing blood meal, females were put in small cartons in groups of 7-8 mosquitoes. Sugar and water were removed one day before the assay. Females were allowed to probe and feed on 350 μl of a feeding solution (50 % FBS, 50 % MEM, 0.2 mM ATP, ~50 μg sucrose/phenol red, pH 7) that was placed between two parafilm membranes stretched over a glass feeder. After probing, mosquitoes and feeding solutions were collected and plaque assays were performed to determine virus titer.
The method for intrathoracic inoculation of mosquitoes is derived from Gubler and Rosen (1976). Adults female mosquitoes (5 days post-eclosion) were infected by injecting 69 nL of DENV-2 Jamaica 1409 (100 or 10 pfu) using a Nanojet II (Drummond Scientific Company, Broomall, PA). All injections were performed under a dissecting microscope using glass needles prepared with a vertical pipette puller (P-30, Sutter Instrument Co., Novato, CA). Mosquitoes were incubated for 7 days at 28°C, 80% relative humidity. The salivary glands were dissected and triturated in 0.5 ml of MEM medium. After centrifugation, the supernatant fluid was filtered (Acrodisc Syringe filters with 0.2 μm HT Tuffryn membrane) and virus titer determined by plaque assay.
LLC-MK2 cells were grown in confluent monolayers in 24-well plates. Ten-fold serial dilutions of salivary gland or whole-mosquito supernatants were added for 1 hour and cells were overlaid with agar. After 7 days of incubations at 37°C cells were stained with a solution of 3mg/ml MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) added directly to the plate and incubated for 4 hours (Sladowski et al., 1993, Takeuchi et al., 1991). Viral titers were determined by plaque counting.
To determine viral titer and prevalence of DENV-2 infection in carcass and salivary glands, data were subjected to analysis of variance (ANOVA) and Pearson chi-square test, respectively, by using SAS (SAS User's Guide, Cary, NC: Statistical Analysis System Institute, Inc., 1987). Sources of variation were mosquito lines (transgenic lines; 30KExGM [1, 15, 22, and P4] and GFP30KMnp P6, and as controls Higgs and pMos-30K strains), replicate (two) and their interaction. For the intrathoracic injections, the two viral concentrations tested did not differ in the magnitude of viral titers in either carcasses or salivary glands, and no significant interaction between viral concentration and mosquito lines was detected; therefore they were considered as replicates. Similarly, the differences in viral prevalence between transgenic and control mosquitoes was identical between viral concentrations. Due to the lack of virus in saliva in most transgenic mosquitoes in the transmission assay, the averages of the five transgenic lines were pooled for comparison with the control group. If variances were not homogeneous, data were subjected to log-10 transformations. When differences among treatment means were detected, they were separated using least significant difference (LSD) procedure adjusted by Tukey-Kramer.
Supplemental Figure 1. Southern blot analysis for detection of transgene integration and copy number in transgenic lines. A) Genomic DNA extracted from six transgenic (P2) or control Higgs (H) recipient line females was digested with EcoRI. DNA fragments separated on agarose gel were transferred to nylon membrane and probed with DsRed 32P-labeled probe. B) Genomic DNA extracted from transgenic (lines 1, 15, 22, P4, P6) and Higgs (H) females or transgenic males (line 27) was digested with XhoI. DNA fragments on membranes were hybridized with DsRed-specific 32P-labeled probe. The inset shows an enlargement of the signal image associated with line 15. Relative locations of the DNA fragments used as probes and restriction endonuclease cleavage sites (EcoRI and XhoI) are shown at the bottom of the figure.
Supplemental Figure 2. RT-PCR analysis to determine tissue- and sex- specific transgene expression in pMos30K line P2. Gene amplification of RNA samples derived from transgenic females and males shows no expression of the reporter construct in the 30K a construct and promiscuous expression for the 30K b construct. Total RNA was isolated from the head, thorax (Th) or abdomen (Ab) of 5 day old transgenic females, males or Higgs females. Aedes Actin1 gene primers were used as a control for RNA preparations. Approximate sizes of amplicons are listed on the right of the Figure.
Supplemental Figure 3. Immuno-blot analysis to determine tissue-specific expression of EGFP protein. Salivary glands were dissected from 15 3-day old line P2 transgenic or Higgs (H) wild-type females. Proteins from salivary glands, carcass (whole body with salivary glands removed) and males from transgenic or Higgs wild type were separated on 12% SDS-PAGE, transferred to PVDF membrane and probed using 1:1000 rabbit anti-GFP (Santa Cruz) primary antibody, 1:50000 goat anti-rabbit AP conjugate secondary antibody (Jackson Immunosearch) followed by alkaline phosphatase detection. The asterisk (*) corresponds to the 30K b-EGFP protein.
Supplemental Figure 4. Detection of Mnp in transgenic mosquitoes A) Northern blot detection of Mnp in salivary glands of transgenic mosquitoes using a 32P-labeled Mnp probe. Total RNA from 12 salivary glands (SG) of transgenic females, carcasses (C) of six females after dissecting salivary glands and six transgenic males (M) was extracted and separated on a 1.2% formaldehyde agarose gel. SG and C from Higgs (H) females was used as negative control and total RNA from six 24 hrs post blood fed females from transgenic line Vg40-Mnp were used as positive control (+C). B) Ethidium bromide-stained gel before transfer.
Supplemental Figure 5. 30K a and 30K b cis-regulatory DNA sufficient to drive transgene expression. The red, blue and green arrows represent DNA sequences included in the pMos30K, 30KExGM and GFP30KMnp transgene constructs, respectively. The table indicates whether those sequences were sufficient (Yes) or not (No) to replicate endogenous 30K a and 30K b expression characteristics.
Supplemental Figure 6. Challenge infections of mosquitoes by intrathoracic injection of DENV-2. Transgenic lines (P6, P4, 22), Higgs (H) and pMos-30K control mosquitoes were injected intrathoracically with 10 pfu of DENV-2 and viral titers were determined 8 dpi in A) carcasses and B) salivary glands. C) DENV-2 titers determined 8 dpi in salivary glands of transgenic (1 and 15) and Higgs (H) control mosquitoes injected intrathoracically with 100 pfu of DENV-2. The number of mosquitoes assayed from each group (n) is indicated on the figure. Horizontal bars show mean titer values.
Supplemental Table 1. Oligonucleotide primers used in gene constructs and primary sequence verification
We thank Dr. Alexander Franz for providing the plasmid pMos-carb/Mnp/i/Mnp/svA, transgenic line Vg40-Mnp, and for discussions regarding the detection of siRNA, and Lynn Olson for help in compiling the manuscript. This work was supported by a grant from the Foundation for the National Institutes of Health through the Grand Challenges in Global Health Initiative.