Aedes aegypti colony (Rockefeller strain) was obtained from Tulane University. Mosquitoes were maintained in rearing chambers in a secure insectary (arthropod containment level 3) ACL-3 in the NEIDL, following rearing conditions as described by Araujo et al. (2020).

For the transcriptome experiments, human blood from a donor was used (The Blood Center, New Orleans, LA, United States). Briefly, blood in EDTA was centrifuged, and plasma was separated and inactivated at 56^°C for 1 h. Red blood cells (RBCs) were washed three times with 0.01 M PBS and reconstituted with autologous plasma. For the infection, the blood mixture was then added in a 1:1 ratio to the supernatant from either the DENV-2 culture or uninfected cells. For the feeding experiments, 7–10 day old female A. aegypti mosquitoes were starved for 24 h before the feeding. Mosquitoes were fed for 30 min. Engorged females were sorted and kept in the incubator for 3 days with a 10% sucrose solution provided on day 3 after feeding, ovaries from all live mosquitoes were dissected in 0.01M PBS, and RNA was extracted as described below. Midguts were tested for the presence of infection with DENV using reverse transcription-quantitative PCR (RT-qPCR).

For RT-qPCR validation of genes and egg laying, pre-mated females at 7 days old were transferred to individual cups and fasted for 24 h prior to the DENV-2 blood meal or mock blood meal. Females were infected via feeding on a Hemotek Feeding System, and an infectious blood meal composed of mice blood and DENV-2 (2.7 × 10⁷ ffu/ml) supernatant at a 1:1 dilution was given to the mosquitoes (infected group). For the mock blood feeding (control group), females were fed with clean blood composed of Vero-E6 cell supernatant and mouse blood at a 1:1 dilution. The blood was freshly collected from Balb/C mice through cardiac puncture and placed in a heparin-coated tube according to standard procedures approved by IACUC PROTO201900005. The whole blood was centrifuged at 3,000 rpm for 10 min to separate plasma and RBCs. RBCs were washed three times in 0.01M PBS to remove anticoagulants, and plasma was heat-inactivated at 56^°C for 1 h. RBCs were resuspended with plasma and used immediately for mosquito blood feedings. Mosquitoes were allowed to feed for 30 min. Engorged females from each group were sorted into cups and provided with a 10% sucrose solution and returned to the incubator. Ovaries from females of each group were dissected in 0.01M PBS 3 days after blood feeding. Carcasses from the DENV-infected group were tested for DENV-2 using RT-qPCR, and a pool with the corresponding positive ovaries was used in the RT-qPCR for validation of the volcano plot genes. Individual egg-laying females were placed in individual tubes with 10% sucrose provided. Eggs were counted daily, from 3 to 7 days after receiving a blood meal.

The A. aegypti embryonic-derived cell line, Aag2 (gift from Dr. Michael J. Conway), was cultivated at 28^°C in an incubator with 5% CO₂ in high glucose Dulbecco’s Modified Eagle’s Medium, DMEM (Gibco), supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gemini), 1% Tryptose Phosphate Broth (Gibco), and Antibiotic-Antimycotic (Gibco). Aag2 cells were used for in vitro validation of gene expression from selected genes.

The dengue virus 2, New Guinea C (DENV-2-NGC) strain was obtained from BEI resources. The virus was passaged into Vero-E6 cells and cultured for 5 days in high glucose DMEM (Gibco) medium supplemented with 2% heat-inactivated FBS (Gemini) at 37^°C in an incubator with 5% CO₂. After visualization of cytopathic effect (CPE), the viral supernatant was centrifuged, spun down for 10 min at 4,000 rpm, and filtered using a 0.45 (μm) syringe filter. The virus was harvested and aliquoted in cryotubes and stored at -80^°C freezer to use for the mosquitoes’ infectious blood meal. DENV-2 was also passaged into Vero-E6 cells for focus forming assay (ffa) and the viral titer was obtained.

A pool of 100 ovaries was dissected in 0.01M PBS from females 3 days post DENV-2 blood feeding (DENV infected) and the 100 ovaries from the control group, 3 days post uninfected blood feeding (mock). This was performed in triplicate, and gene expression of infected ovaries was compared to the uninfected control group. Tissues were homogenized in 100 μl of RNA lysis buffer using a pestle and an additional 500 μl of buffer was added to each sample. Total RNA was extracted from infected and uninfected samples using the Qiagen RNeasy Plus Mini-Kit (Qiagen, 74136) per the manufacturer’s instructions. RNA sequencing was performed from mosquito samples in three biological replicates by outsourcing to Genewiz, which conducted poly(A) selection and standard strand-specific single-end RNA sequencing (RNA-Seq) library construction and sequencing on a NextSeq 500 instrument. A total of six FASTAQ files were generated and deposited at the NIH NCBI Sequence Read Archive (SRA) under the BioProject accession # PRJNA786000.

For quality control, 6 FASTQ files (3 from mock and 3 from DENV infected) were first inspected using FastQC (REF: Andrews S. FastQC: a quality control tool for high throughput sequence data, 2010) to assess any bias due to read quality and length and to confirm the linker adapters had been trimmed. Using the STAR aligner (Dobin A. STAR: ultrafast universal RNA-seq aligner, 2013), RNA-seq reads were then mapped to the reference genome of A. aegypti (strain Liverpool AGWG, assembly AaegL5) downloaded from VectorBase (2022), but an additional assignment of Drosophila gene names from FlyBase (Thurmond et al., 2019) was appended to enhance the identification of the mosquito gene’s machine-generated names.

As the standard analysis of DEGs based solely on p-values yielded an intractably large list of gene candidates, we modified our approach to filter for robustly affected genes. First, we looked at mosquito gene annotations that have gene ID, such as AAEL###### as described in the mosquito database (VectorBase, 2022) by appending the gene name of the Drosophila melanogaster ortholog from FlyBase (Thurmond et al., 2019).

After batch correction and initial filtering in R, overall relatedness can be observed within the mock and DENV-2 infected samples as seen by the clustering in the principal component analysis (PCA) plot (Supplementary Figure 1A). Using estimated read counts, the transcriptomes below the threshold value (est. read count < 1 across all samples) were further filtered using R. Correction was done for batch effects using DESeq2 (Love et al., 2014; Supplementary Figure 1A) for 2 PCAs before and after batch correction. Differential gene expression (DGE) analysis was then performed using DESeq2 as well as EdgeR (Robinson et al., 2009) to generate a robust list of DEGs. In each result set, the most significant DEGs were determined (FDR < 0.05), and a intersect analysis (Venn diagram) showed approximately 91% overlap of DEGs from DESeq2 results (Supplementary Figure 1B). Thus, DESeq2 results were utilized for downstream analysis.

A heatmap was generated to show the top 50 upregulated and downregulated DEGs. Furthermore, the significant DEGs (FDR < 0.05) were filtered for a base mean expression value of 150, and 12 highly expressed genes were picked for validation using RT-qPCR as highlighted in the volcano plot and the heatmap. Gene Ontology (GO) analysis was performed on significant DEGs (FDR < 0.05) using VectorBase’s annotation tool (Ashburner et al., 2000; VectorBase, 2022). The top entries for biological processes, molecular functions, and cellular components, were picked based on an FDR cutoff of 0.005 and excluding GO terms containing 1,000 genes or more to focus on specialized terms.

Mosquito egg laying assays were performed in two biological replicates. Two groups of pre-mated females, 7–10 days old, were used for this assay. DENV-2 infected and mock mosquitoes were blood-fed using a Hemotek Feeding System as previously described. The viral titer used in the infectious blood meal was 2.5 × 10⁷ focus forming units (ffu). Engorged females were sorted into individual 50 ml Falcon tubes for egg laying. Each tube was closed with a piece of mesh cloth and secured with an elastic band. A cotton ball with 10% sucrose was placed on the top of each tube for subsequent feeding of mosquitoes. At approximately 60 h post blood feeding, a strip of wet filter paper was placed inside each tube to allow oviposition. The egg counting started at 3 days post blood feeding and ended at 7 days post blood feeding when females were not laying any more eggs. At the end of the experiment, at 7 days post infection (7 DPI), each mosquito was homogenized for total RNA extraction using the QIAshredder (Qiagen, 79656) and the Qiagen RNeasy Plus Mini-Kit (Qiagen, 74136). Individual mosquitoes were tested for DENV-2 by RT-qPCR using the QuantiFast SYBR Green RT-PCR Kit (Qiagen, 204156). Viral burden from individual mosquitoes was obtained by using a standard curve from an in vitro transcription of standard DENV-2 RNA as described below. Only positive females were considered for the egg counting final graph.

A total of 12 genes were chosen for RT-qPCR validation. At 3 days after infected blood feeding (DENV-2 infected) and uninfected blood feeding (mock), ovaries of 10 females (DENV-2 infected) and ovaries of 10 females (mock) were dissected in 0.01M PBS and homogenized in RLT plus buffer for total RNA extraction per manufacturer’s instructions using the QIAshredder (Qiagen, 79656) and RNeasy Plus Mini Kit (Qiagen, 74136). Individual carcasses of DENV-2 blood-fed mosquitoes were homogenized as described and RNA extracted to be tested for DENV-2, and only the ovaries from positive carcasses were considered for the RT-qPCR assay as described below:

The in vitro transcription of standard DENV-2 RNA (ssRNA) was obtained following the methods described by Araujo et al. (2020), using 5 μg of total RNA from DENV-2 NGC infected Vero-E6 cells. The first strand was generated using SuperScript III (Invitrogen, 18080051) with random hexamers, and the cDNA generated was used in PCR reaction with the T7 promoter sequence added to the forward primer DENV2_T7-5′-TAATACGACTCACTATAGGGAGAGCAGATCTCTGATGAA TAACCAACG-3′ and the reverse primer DENV2_Env_RC: 5′-CATTCCAAGTGAGAATCTCTTTGTCA-3′. PCR cycles of 95^°C for 3 min, followed by 35 repeated cycles of 95^°C for 30 s, 62^°C for 30 s, and 72^°C for 1 min. PCR products were used as a template for the in vitro transcription using the MEGAscript T7 Transcription Kit (AMB13345).

Absolute quantitation of viral burden per mosquito was obtained using the RNA from carcasses as a template in the one-step RT-qPCR reaction using the QuantiFast SYBR Green RT-PCR Kit (Qiagen) and a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, California, United States) and calculated using a 10-fold dilution series of DENV2 ssRNA as a standard.

The relative gene expression for all 12 genes selected for validation of RNA-Seq and the immune pathway genes were normalized against the A. aegypti actin gene using the 2^–ΔΔCT method. Primers used for these assays were designed using the Primer-BLAST tool available online (Ye et al., 2012). All of the primers used in this study can be found in Supplementary Table 1.

Transfection of siRNA on Aag2 cells was performed using Lipofectamine 3000 (Thermo Fisher Scientific) according to the manufacturer’s instructions. The siRNA sequences can be found in Supplementary Table 1. Briefly, 5 × 10⁴ cells were seeded into 24-well plates 1 day before transfection. After cells achieved 90% confluency, 10 pmol of gene-targeting siRNA or scrambled siRNA control was mixed with Lipofectamine 3000 (Thermo Fisher Scientific) and Opti-MEM^® (Gibco), which was then used to transfect Aag2 cells. At 72 h post transfection, the siRNA-transfected cells were inoculated with DENV-2 at an multiplicity of infection (MOI) of 1 for 2 h at 30°C and 5% CO₂ incubator. At 2 h post infection (hpi), the inoculum was removed, and the cells were washed once with high glucose DMEM (Gibco). Following washing, DMEM (Gibco) supplemented with 2% FBS was then added to the cells. Time points (cell lysates and viral supernatant) were collected at 24 hpi. Cell lysates were harvested in the RLT buffer by adding beta-mercaptoethanol (β-ME), RNA was extracted using the RNeasy Plus Mini Kit (74106, Qiagen), and RNAi efficiency was validated using qRT-PCR. Cell culture supernatants were collected, and viral titers were obtained by ffa as described. Viral RNA was extracted using the QIAamp Viral RNA Kit (52906, Qiagen). DENV load was then quantified by RT-qPCR using SYBR Green and DENV-2 ssRNA as standard, as we previously described.

Following dengue infection, cell supernatants from siRNA-transfected Aag2 cells were collected to measure infection by a focus-forming assay. Vero-E6 cells were seeded in 96-well plates at a density of 10,000 cells per well and cultured in DMEM (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gemini) and 1% penicillin/streptomycin (Gibco) at 37^°C in an incubator with 5% CO₂. At 24 h after seeding, Vero-E6 cells were treated with 100 μl of supernatants from infected cells in a serial 10-fold dilution from undiluted to 10^–7 diluted in DMEM (Gibco) for 2 h at 37^°C in an incubator with 5% CO₂. After this time, the supernatant was removed, the cells were washed with PBS, and 100 μl of overlay consisting of a 1:1 ratio of DMEM:2% FBS media and 5% carboxymethyl cellulose was added to the cells. Infection was allowed to progress for 5 days, after which the overlay was aspirated, and cells were washed with 0.01M PBS, then fixed with 4% paraformaldehyde (Thermo Fisher Scientific) supplemented with 0.1% Triton X-100 (Thermo Fisher Scientific) for 30 min at room temperature. The fixing solution was then aspirated, and cells were washed with 0.01M PBS again, then blocked with 1% bovine serum albumin (BSA) (Millipore Sigma B6917) in 0.01M PBS at room temperature for 1 h. Following this time, the blocking buffer was removed and cells were incubated with flavivirus group antigen-antibody [D1-4G2-4-15 (4G2)] (Novus Biologicals, NBP-52709, Centennial, CO., United States) diluted 1:1,000 in 1% BSA in 0.01M PBS for 2 h at room temperature. After this time, the antibody was removed, cells were washed 3 times with 0.01M PBS, and cells were incubated with IRDye^® 680RD Goat anti-Mouse IgG Secondary Antibody (LI-COR^® Biosciences, Lincoln, NE, United States) diluted 1:1,000 in 1% BSA in 0.01M PBS for 2 h at room temperature. After this time, the antibody was removed, and cells were washed 3 times with 0.01M PBS. PBS was aspirated and cells were allowed to dry. Focus forming assays were performed in duplicate. Foci were identified by the detection of DENV E-specific signal in the 700 nm channel using the Odyssey CLX300 Near-Infrared Fluorescence Imaging System (LI-COR^® Biosciences, Lincoln, NE, United States) and counted and reported as focus forming units (FFU).

We aimed to examine the impact of DENV infection on mosquito fecundity. For this, we fed mated 7–10-day old A. aegypti females with uninfected blood (mock) vs. blood containing DENV-2 (Figure 1A). We observed that during the first gonadotrophic cycle, the females in the DENV-infected group laid significantly fewer eggs compared to females in the mock blood-fed group. In our fecundity assay, we counted eggs daily from 3–7 days post blood feeding (mock or DENV), and we did not observe significant differences in the timing of oviposition, with most mosquitoes laying most of their eggs 3 days post blood meal (Figure 1B). In the mock group, the average number of eggs laid per female was 93 eggs/female from a total of 30 females analyzed. In the DENV-2 infected group, the average number of eggs laid was 77 eggs/female from a total of 36 females. These data show a reduction of 22.5% in fecundity from the DENV-2-infected mosquitoes (Figure 1C). After egg laying, we confirmed that each female mosquito was infected with DENV using RT-qPCR to detect from 10¹ to 10⁸ viral copies. The ratio of DENV-2 infection in the mosquitoes was 50% in the first gonadotrophic cycle alone (Figure 1D).

Since blood feeding triggers changes in complex gene expression in females (Giraldo-Calderón et al., 2020), we wondered if the impact that DENV had on the first gonadotrophic cycle could be linked to an alteration of gene expression profiles during blood feeding. Therefore, we performed RNA-seq and transcriptomic analysis of differentially expressed ovarian genes in A. aegypti females at 3 days after DENV-2 blood feeding compared to females at 3 days after uninfected blood feeding (mock). As the standard analysis of DEGs based solely on p-values yielded an intractably large list of gene candidates, we modified our approach to filter for robustly affected genes. First, we looked at mosquito gene annotations that have gene IDs, such as AAEL###### as described in the mosquito database (VectorBase, 2022) by appending the gene name of the Drosophila melanogaster ortholog from FlyBase (Thurmond et al., 2019).

After batch correction and initial filtering in R, overall relatedness can be observed within the mock and DENV-2 infected samples as seen by the clustering in the PCA plot (Supplementary Figure 1A). Using DESeq2 from a total of 14,537 DEGs, we identified that 1,744 (12%) genes were significant at an FDR cutoff of 0.05, of which 878 (6%) were upregulated, while 866 (6%) genes were downregulated in the DENV-2 infected pool (Figures 2A,B). The top 50 DEGs can be observed in the heatmap (Figure 2C). From the volcano plot and heatmap, we chose 12 DEGs for the validation of gene expression by qRT-PCR since this technique is powerful in validating gene expression due to its sensitivity and precision. From the list of 12 genes, seven were upregulated, namely, amd-AAEL022306, CG5958-AAEL001293, ImpL2-AAEL002130, Oatp58Dc-AAEL021099, CG15506-AAEL026430, scyl-AAEL020092, CG17684-AAEL007764, and five were downregulated, namely, cep290-AAEL025983, shd-AAEL023509, CG31826-AAEL010344, LManII-AAEL005763, and CG31869-AAEL019708. These 12 genes (Figure 2D) are shown to be between the most upregulated or most downregulated (basemean > 150), highlighted in red in the heatmap. We also searched for the ortholog using FlyBase to find the predicted function of those 12 genes. The Drosophila ortholog gene function of each of the 12 genes can be found in Table 1. The table illustrates the utility of appending the Drosophila gene ortholog name to the mosquito gene ID from Vectorbase because 8/12 of these genes lacked a functional description in the 2021 edition of Vectorbase. However, all of these genes had a clear Drosophila ortholog for which Flybase contained a rich functional annotation for the gene. A trend of metabolic and physiology-regulation functions is apparent in this list of genes, with functions such as carboxy-lase for amd-AAEL022306, membrane transport in Oatp58Dc-AAEL021099, dipeptidyl-peptidase in CG17684-AAEL007764, and lysosomal alpha-mannosidase in LMannII-AAEL005763.

To explore the functional annotation trends more broadly beyond the 12 validated DEGs, we subjected all the significant DEGs (FDR < 0.05) to a Gene Ontology (GO) analysis, which was aided by the Drosophila ortholog names. At Benjamini-Hochberg corrected p-values < 0.05, we identified enriched biological processes, molecular functions, and cellular components among the significant mosquito ovary DEGs impacted by DENV-2 blood meal (Figure 3). A trend of physiological and metabolic functions was also apparent in these GO terms, such as peptide metabolic and translation processes, as well as ribosome and translation activity functions.

The general overview of the GO results showed a trend of downregulated genes within the GO biological process in the mosquitoes’ ovaries after DENV-2 blood meal (Figure 3 upper panel). This lowering of the biological process may be related to a decreased number of eggs observed in the first gonadotrophic cycle of the infected mosquitoes.

From the gene expression results shown by the heatmap, we selected 12 DEGs for validation by qRT-PCR (Figure 4A). We chose 12 genes to have a better representation of the transcriptome and focus on a search for virally regulated genes. We performed two independent biological replicates of groups of mosquitoes that were fed uninfected blood (mock) vs. a DENV2-infected blood meal. At 3 days post blood feeding, we dissected the ovaries and tested the female carcasses to confirm DENV-2 infection. We used pools of DENV-2 positive ovaries for the RT-qPCR validation assay. As anticipated, all 12 DEGs we tested exhibited similar directional gene expression changes between the RNAseq experiment and the RT-qPCR experiment. Interestingly, two DEGs (amd-AAEL022306 and Oatp58Dc-AAEL021099) after qRT-PCR assay were confirmed to be most upregulated, with over a 3-fold increase, among the 12 selected genes (Figure 4B). However, these two genes were the best candidates for further investigation.

We decided to compare the gene expression of these two genes (amd-AAEL022306 and Oatp58Dc-AAEL021099) with that of an A. aegypti eggshell organization factor 1 (EOF1) (Isoe et al., 2019). The EOF1 was recently identified and plays a role in the formation and melanization of the eggshell. Our results using DENV-2 blood-fed females and mock-fed females, 3 days after feedings, showed no differences in the DENV-2-infected mosquitoes compared to the mock-infected ones (Supplementary Figure 2).

We hypothesize that DEGs from the A. aegypti ovary transcriptome may regulate host responses to DENV-2 infection and modulate fecundity. Using an in vitro approach, we chose to investigate the roles of the most upregulated genes from the RT-qPCR validation assays using the Aag2 cell line. This cell line has been shown to recapitulate the immune responses found in adult mosquitoes (Zhang et al., 2017; Li M. J. et al., 2020; Rosendo Machado et al., 2021). Aag2 cells have been used to experimentally validate mosquito immune pathways (Russell et al., 2021) and to elucidate the mechanisms of antimicrobial peptides (AMP) (Zhang et al., 2017). The advantage of using Aag2 cells to study antiviral immunity genes is that they are much simpler to infect and perform small interfering RNA (siRNA) gene knockdowns than in vivo methods. In our Aag2 cell assays, we observed the ideal peak knockdown at 3 days post siRNA transfection of Oatp58Dc-AAEL021099 and amd-AAEL022 in Aag2 cells measured by RT-qPCR (Figure 5A). Following 3 days after knockdown, we infected the cells with DENV-2 at an MOI of 1. At 24 h post infection (24 hpi), cell culture supernatant was collected, and the viral titers were determined by ffa (Figure 5B). We also checked the viral loads by RT-qPCR (Figure 5C). We observed an increase in DENV-2 RNA copies in Aag2 cells after siRNA-mediated knockdown of either Oatp58Dc-AAEL021099 or amd-AAEL022306 compared to cells with the control scrambled siRNA, indicating a potential role of these genes in the mosquito antiviral response (Figure 5D). The same phenotype was observed by Ramirez and Dimopoulos (2010) when silencing MyD88 in laboratory strains of A. aegypti, DENV titers were increased in the mosquito’s midgut.