Figure 1.
Generation of transgenic Ae. aegypti expressing an IR-RNA targeting DENV2 in midguts of bloodfed females.
(A) Schematic representation of the transgene based on the mariner Mos1 TE [16], [17]. The endonuclease restriction sites used for the Southern blot analysis are indicated. The cDNA probe of the Southern blot corresponded to the left arm of Mos1. Abbreviations: ma.right or ma.left = mariner (Mos1) right or left arm; 3xP3 = eye-specific synthetic promoter; svA = SV40 poly A signal; Mnp = M (no protein) and includes DENV2 prM-M coding sequence preceded with stop codons to prevent translation; Mnp (−) or (+) = orientation of sequence to transcribe IR-RNA (with respect to the AeCPA promoter); i = intron sequence; AeCPA = Ae. aegypti carboxypeptidase A promoter; EGFP = enhanced green fluorescent protein. (B) Flowchart showing the establishment of transgenic mosquito lines based on pooling of individual families containing transgenic founders. HWE = Higgs white-eye Ae. aegypti line.
Table 1.
Overview of transgenic Ae. aegypti (pools and lines) generated containing the pMos- carb/Mnp/i/Mnp/svA transgene.
Figure 2.
NextGen small RNA sequencing to reveal processing of the IR-RNA into siRNAs in Carb109M mosquitoes.
(A) Alignment of small RNA sequences derived from the IR-RNA to the DENV2-Jamaica1409 genome (10,723 nucleotides). (B) Alignment of small RNA sequences to the 568 nt region of DENV2-Jamaica1409 RNA targeted by the IR (Mnp) effector gene. (C) Size distribution of small RNA reads. Y-axis: Read Count = number of small RNAs detected by NextGen sequencing. X-axis: Distribution of small RNA sequences mapped to the DENV2-Jamaica1409 genome. Red bars, negative-sense small RNAs; blue bars, positive-sense small RNAs.
Figure 3.
Characterization of transgene integration in Carb109 mosquitoes.
(A) Southern blot analysis to detect transgene integration sites among transformed Ae. aegypti. (B) Southern blot analysis to detect transgene integration sites in Carb109M, Carb109F, Carb109M/GDLS.BC6, and Carb109F/GDLS.BC6. Lane 1: HWE; lane 2: Carb109F G9; lane 3: Carb109M G9; lane 4: Carb109F/GDLS.BC6; lane 5: Carb109M/GDLS.BC6. Total DNA was digested with PstI. (C) Physical mapping of a transgene integration site in Carb109M mosquitoes. In bold: mariner Mos1 TA target sequence motif; in bold and underlined: duplication of the TA target sequence as a consequence of mariner Mos1 integration; highlighted in green and blue: partial sequences of the left and right arms of the TE.
Figure 4.
Analysis of Carb109M DENV2-resistance phenotype over 33 intercrossed generations.
HWE (control) and Carb109M mosquitoes received bloodmeals containing 106 pfu/ml DENV2-Jamaica1409. Virus titers in the mosquitoes were assessed at 7 and 14 dpi. Each data point represents the virus titer of a single female. Mean values and standard errors are indicated.
Figure 5.
Carb109M resistance phenotype when challenged with different Mexican DENV2 genotypes.
HWE (control) and Carb109M (G9, G13, G14) mosquitoes were infected with DENV2 C-932/Acapulco 97, Mex96 Merida, QR94 Quintana Roo, and 14757 Yucatan, representing the Asian 2, Cosmopolitan, American, and Asian-American genotypes, respectively. Mosquitoes received DENV2-containing bloodmeals with the following titers: 1.3×106 pfu/ml (C-932/Acapulco 97), 2.0×105 pfu/ml (Mex96), 4.0×106 pfu/ml (QR94), and 2.0×106 (14757). Virus titers in the mosquitoes were assessed at 7 and 14 dpi. Each data point represents the virus titer of a single female. Mean values and standard errors are indicated.
Figure 6.
Analysis of resistance to DENV2 infection in Carb109M/GDLS.BC5 and Carb109F/GDLS.BC5 mosquitoes.
The Carb109 transgene was introgressed via five consecutive backcrosses into GDLS. HWE (control), GDLS (control), Carb109M, Carb109F, Carb109M/GDLS.BC5, Carb109F/GDLS.BC5 were challenged with DENV2-Jamaica1409 (titer in the bloodmeal: 106 pfu/ml). BC5-neg mosquitoes were used as additional controls and represented mosquitoes not positive for the eye-marker after five backcrosses. Virus titers of mosquitoes were assessed at (A) 7 and (B) 14 dpi. Each data point represents the virus titer of a single female. Mean values and standard errors are indicated.
Figure 7.
Analysis of resistance to DENV2 infection of homozygous Carb109M/GDLS.BC5.HZ mosquitoes.
GDLS, Carb109M, Carb109M/GDLS.BC5.HZ mosquitoes were challenged with DENV2-Jamaica1409 (titer in the bloodmeal: >106 pfu/ml). Virus titers of mosquitoes were assessed at 7 and 14 dpi. Each data point represents the virus titer of a single female. Mean values and standard errors are indicated.
Figure 8.
Proportions of EGFP-expressing Carb109M and Carb109F larvae during five consecutive backcrosses to GDLS.
Carb109F and Carb109M were crossed reciprocally with GDLS mosquitoes. Wild-type individuals were culled and remaining EGFP expressing transgenic individuals (Carb109F/GDLS.BC1 and Carb109M/GDLS.BC1) were backcrossed further to GDLS for four additional generations. Proportions of EGFP-expressing larvae in each of the five backcross generations in Carb109F/GDLS (filled circles) and Carb109M/GDLS (open circles) are shown. Proportions were compared by estimating the 95% HDI (error bars) with WinBUGS and the Credible Intervals for Proportions script [44].
Figure 9.
Transgenic allele frequencies among Carb109/GDLS backcrossed mosquitoes over five generations (F1–F5) in absence of selection for the transgenic phenotype.
Initial frequencies (p0) of the Carb109 transgene were either (A) 0.5 (transgenic heterozygote x transgenic heterozygote) for Carb109F/GDLS.BC1, Carb109M/GDLS.BC1, Carb109F/GDLS.BC5, and Carb109M/GDLS.BC5 or (B) 0.25 (transgenic heterozygote x GDLS) for Carb109F/GDLS.BC1, Carb109M/GDLS.BC1, Carb109F/GDLS.BC5, and Carb109M/GDLS.BC5. Fifteen lines were established for each of the eight experiments. Proportions of EGFP-expressing offspring were estimated by examining ∼150 larvae from each of the 15 lines over five successive generations (F1–F5) of inter-breeding without selection for the transgenic phenotype. Bars around mean proportions represent Bayesian 95% Highest Density Intervals (95% HDI). Proportions showing non-overlapping 95% HDI are credibly different.
Table 2.
Fitness coefficients estimated by Maximum Likelihood search for coefficients most consistent with the observed frequencies of EGFP expressing larvae following A) backcrossing for one or five generations without selection for the transgene or B) backcrossing for five generations with selection for the transgene after each backcrossing step.
Figure 10.
Comparison of observed transgenic allele frequencies over five generations (F1–F5) of Carb109F/GDLS.BC5 and Carb109M/GDLS.BC5.
Three replicate crosses (F.1, F.2, F.3) between Carb109F/GDLS.BC5 heterozygotes (200 individuals in each cross) and three replicate crosses (M.1, M.2, M.3) between Carb109M/GDLS.BC5 heterozygotes (200 individuals/cross) were performed in separate cages. All resulting wild-type larvae were culled. Remaining transgenic individuals were allowed to inter-mate. This procedure was followed for five generations (F1–F5). Results were compared to values expected under Fisher's Selection Model. (A) Observed frequencies of EGFP- expressing Carb109F/GDLS.BC5 larvae of replicates F.1, F.2, F.3 compared to values predicted under Fisher's Selection model (redline). (B) Observed frequencies of EGFP-expressing Carb109M/GDLS.BC5 larvae of replicates M.1, M.2, M.3 in comparison to values predicted under Fisher's Selection model (redline). Proportions were compared by estimating the 95% HDI (error bars) with WinBUGS and the Credible Intervals for Proportions script [44].