Identification of conserved ZIKV region divergent from Aedes
Aiming to design oligos optimized to minimize non-specific amplification in one-step qRT-PCR assay that allows the analysis of mosquito gene expression during ZIKV infection in the same samples, the ZIKV NS5 gene sequence of the PRVABC59 strain was submitted for sequence alignment analysis with Ae. aegypti and Ae. albopictus sequences from “EST”, “Assembled transcriptome” and “Transcripts” datasets from VectorBase, using the BLAST tool (https://www.vectorbase.org/blast). At that point, PRVABC59 NS5 sequences without similarity to mosquito genomic sequences were aligned with three ZIKV strains from the Asian phylogenetic lineage (Brazil, French Polynesia and Cambodia) and 2 ZIKV strains from African phylogenetic lineage (Uganda-MR766 and Senegal) using ClustalW. At the conclusion of the analysis, a region of 96 nucleotides from ZIKV NS5, highly conserved between ZIKV strains but divergent from Ae. aegypti and Ae. albopictus, was identified and utilized as template to design qRT-PCR primers (Fig. 1).
Confirmation of ZIKV detection in mammalian cells
We performed qRT-PCR analysis using the designed NS5-2362F and NS5-2457R primers with RNA template isolated from Vero cells infected with ZIKV MR766 and PRVABC59 strains. In Fig. 2, the amplification plots for reactions using RNA from cells infected with both strains are shown, demonstrating that the NS5 primers can amplify ZIKV strains from Asian and African phylogenetic lineages.
Assay sensitivity, reproducibility and specificity evaluation
The ZIKV 1086/ZIKV 1162c primers and the 1107-FAM probe  can be considered the gold standard to ZIKV RNA detection in mosquito samples since it has been preferentially used in most recent studies [31, 33–36, 38, 46, 49, 51]. The ZIKV 1086 and ZIKV 1162c have been successfully used to detect ZIKV RNA in cell lines, using SYBR Green as a dye instead the 1107-FAM  and could be a less expensive alternative to ZIKV detection in mosquitoes.
However, we observed nonspecific amplification in half of reactions using uninfected mosquito RNA as template, when using those primers in reaction with SYBR Green, while the reactions with NS5 primers did not present nonspecific detection (Table 2).
Once ZIKV 1086/ZIKV 1162c primers were designed to detect ZIKV RNA in human samples, the observed nonspecific amplifications could be caused by similarities between mosquito and viral sequences. Then, we performed the alignment of the sequence amplified by ZIKV 1086/ZIKV 1162c primers and Ae. aegypti sequences, using the BLAST tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The identities found in four of many significant alignments obtained between the primers annealing sites and mRNA sequences of Ae. aegypti are demonstrated in Fig. 3. The results shown in Table 2 and Fig. 3 demonstrate the importance of optimization with primers design to detect ZIKV RNA in mosquito samples.
Positive single-strand RNAs, transcribed in vitro, were used to create standard curves to assess the limit of detection, efficiency and reproducibility of the RT-qPCR assay using NS5-2362F and NS5-2457R primers and SYBR Green as fluorescent dye. The gold standard used in the literature (ZIKV 1086/ZIKV 1162c primers and 1107-FAM as the dye) was used in comparisons; referred to in the paper as “1107-FAM probe”. The detection of the RT-qPCR reactions was linear over six 10-fold dilutions (109 to 104 copies/reaction) using synthetic RNA with sequences from the MR766 and PRVABC59 strains (Fig. 4).
The coefficient of determination (R2) from synthetic RNA standard curves was calculated to assess the efficiency of reactions using NS5-2362F and NS5-2457R primers, compared to the gold standard (1107-FAM probe). The results show that reactions using the NS5 primers are highly efficient (R2 = 0.99) in the detection of RNA sequences from both ZIKV strains tested (MR766 and PRVABC59) (Fig. 4). The determination coefficients obtained in reactions using 1107-FAM probe were 0.99 and 0.97 to MR766 and PRVABC59 strains, respectively.
To further assess the efficiency and reproducibility of this assay, intra and inter assay coefficients of variation (CV) were calculated from quantification cycle (Cq) values obtained using 108 copies/reaction of in vitro transcribed RNA, from ZIKV MR766 and PRVABC59 strains. These data are shown in Fig. 5. The intra assay coefficients of variation for reactions using NS5 primers were 1.3 ± 0.1% (MR766 strain) and 1.5 ± 0.5% (PRVABC59 strain) while the values obtained using 1107-FAM probe were 1.0 ± 0.2% (MR766 strain) and 1.9 ± 1.8% (PRVABC59 strain). The inter assay variation for reactions using NS5 primers were 2.2 and 1.7% for sequences from MR766 and PRVABC59 strains, respectively. When using 1107-FAM probe, the inter assay variations were 1.3% for MR766 strain and 1.2% for PRVABC59 strain.
The data shown in Figs. 4 and 5 demonstrate that the developed assay is highly reproducible and efficient, and this assay has the same level of sensitivity as the gold standard used in literature (1107-FAM probe).
Assay detects same levels of ZIKV infection than the gold standard in mosquito samples
The NS5-2362F and NS5-2457R primers were also tested in RT-qPCR reactions using RNA template isolated from infected blood-fed (IBF) female Aedes mosquitoes. Total RNA from ZIKV-infected whole mosquitoes (PRVABC59 strain) was extracted 7 days post-blood-feeding. RT-qPCR using the NS5-2362F and NS5-2457R primers showed an infection rate of approximately 47% (24 of 51 females), with infection levels of 9.9 ´ 104 ± 8.1 ´ 104 ZIKV RNA copies/ng total RNA. The mosquito samples were also analyzed using the 1107-FAM (gold standard) and the results demonstrate a highly approximate ZIKV RNA detection level (9.6 ´ 104 ± 8.4 ´ 104 copies/ng total RNA), as well as the same infection rate, with that obtained using NS5 primers (Fig. 6).
Defensin A upregulation detected in orally infected Ae. aegypti
Since the Ae. aegypti antimicrobial peptide Defensin A has a role in the mosquito immune response to dengue, Chikungunya virus and Zika virus infection [64, 65], we chose to measure expression of this gene in our infected mosquitoes as confirmation that our assay allows for both ZIKV infection detection and analysis of gene expression in the same sample. RNA samples which tested positive for ZIKV were used in additional qRT-PCR analysis to evaluate the transcription of the gene Defensin A (data shown in Fig. 7). The results showed that from the ZIKV-positive mosquitoes Defensin A was upregulated in 69.6% (16 females), downregulated in 8.7% (2 females) and constitutive in 21.7% (5 females). In samples that displayed upregulation, the relative expression for Defensin A ranged from 2.2 to 50.3 in fold-change, with a mean of 10.3 ± 11.9 fold-change, corroborating results of Zhao et al.  that demonstrate Defensin A upregulation level of approximately 5 fold-change in pooled samples (10 females each) of ZIKV infected females.