Study Site

We collected mosquito egg samples in the municipality of Jojutla, located in the southern region of the state of Morelos, Mexico (18°36′N, 99°10′W). Arboviral infections are common in this municipality, which is usually hot and dry during October–April and warm and humid during May–September. During 2016, the incidence of Zika virus infections increased in Morelos; 269 cases were reported by the end of the year, of which 69 cases were from Jojutla. The reported cases in Jojutla were recorded as follows: 1 in August, 17 in September, 20 in October, 25 in November, and 6 in December (14).

Egg Collection and Hatching

Figure 1

Figure 1. Location of ovitraps in the municipality of Jojutla, Morelos, Mexico (red). The ovitraps were set according to the guidelines of the Vector Transmitted Diseases Program of the National Center of Preventive...

We collected Ae. aegypti eggs in collaboration with the National Center of Preventive Programs and Disease Control (CENAPRECE), following regulations for entomological surveillance with ovitraps (NOM-032-SSA2–2014). In 2016, 180 ovitraps were placed and maintained throughout the year in different locations in Jojutla (Figure 1). We determined the locations of the ovitraps on the basis of the incidence of human arbovirus infections. We collected the eggs using filter paper, which we then air dried for 2 days and stored in paper bags for further use.

We placed the egg papers in water containers and incubated them at 37°C for 1 wk under 24-h light/dark cycles. After the incubation period, we collected larvae in stages 2–3 and separated them into pools of 20–30 larvae. We macerated each of the pools in viral transport media (5% BSA, 100 U/mL penicillin, 100 μg/mL streptomycin, 2.5 μg/mL fungizone, and NaHCO₃ in Hank’s solution) using a pestle mixer (Thomas Scientific, https://www.thomassci.com) and stored pools at −80°C for further analysis.

Detection of Zika Virus in Larvae of Aedes aegypti

We incubated the macerated larvae pools with Trizol (Invitrogen, https://www.thermofisher.com) for 10 min, extracted total RNA from the lysates using Zymo-Spin RNA extraction columns (Zymo Research, https://www.zymoresearch.com), and quantified the RNA with a Nano Drop 2000 (Thermo Scientific, https://www.thermofisher.com). We detected the presence of Zika virus RNA by quantitative reverse transcription PCR (qRT-PCR) using the TaqMan system (https://www.thermofisher.com) with primers Zikv 1086 Fw 5′-CCGCTGCCCAACACAAG-3′ and Zikv 1162c Rv 5′-CCACTAACGTTCTTTTGCAGACAT-3′; and probe Zikv 1107-FAM probe 5′-AGCCTACCTTGACAAGCAGTCAGACACTCAA-3′, as previously described by Lanciotti et al. (15). We calculated the minimum infection rate (MIR) of the larvae pools by dividing the number of positive larvae pools by the total number of larvae tested and multiplying by 1,000.

Viral Isolation and Identification

We selected the larvae pools that tested positive for Zika virus RNA and displayed the lowest quantitation cycle (Cq) values (range 22.3–33.4) for viral isolation. In brief, we diluted macerated larvae pools in maintenance medium (EMEM medium with 1% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin) and filtered them through 0.22 μm membranes (Millex, http://www.emdmillipore.com). Twenty-four hours before infection, we seeded C6/36 cells in 24-well plates at 80% of confluence in growth medium (EMEM medium supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin) and incubated them at 28°C in a 5% CO₂ atmosphere. Cells were adsorbed with the clarified larvae macerates for 2 h at 28°C. After the incubation period, we added fresh maintenance medium to the cells and left the infection to proceed for 7 d at 28°C and a 5% CO₂ atmosphere until a cytopathic effect was observable. We tested the supernatants of the infected cells for the presence of Zika virus RNA by qRT-PCR, as described earlier in this section. We performed 4 passages of each isolate in C6/36 cells to increase viral titers for NGS.

Indirect Immunofluorescence Assay

For the indirect immunofluorescence assays, we grew Vero cells over glass coverslips and then infected them with the Zika virus isolate. At 48 hours postinfection, we fixed cells with 4% paraformaldehyde and then incubated them with ice-cold methanol for 10 min. We blocked the cells using 10% FBS in phosphate-buffered saline (PBS) for 1 h and then incubated them overnight at 4°C with a 1:500 dilution of the monoclonal antibody 4G2. After the incubation period, we washed the cells twice with PBS and then incubated them with an Alexa Fluor 488-conjugated goat antimouse IgG antibody (Jackson Immunoresearch, https://www.jacksonimmuno.com) for 1 h. We washed the cells twice with PBS and then mounted them with VECTASHIELD medium with DAPI (Vector Laboratories, https://vectorlabs.com) for confocal microscopy (Nikon, https://www.nikon.com).

Plaque Assays

For the plaque assays, we seeded Vero cells in 24-well plates until they reached a confluence of 80%. We used 10-fold serial dilutions of the Zika virus isolate to infect the cell monolayers and left them to adsorb for 1 h. After the adsorption period, we removed the virus and overlaid cell monolayers with 1 mL of DMEM (Invitrogen) supplemented with 2% carboxymethyl-cellulose (Sigma-Aldrich, https://www.sigmaaldrich.com), 2% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, and 2 mM L-glutamine (Gibco, https://www.thermofisher.com) and incubated at 37°C for 4 d. We fixed the cells with 4% paraformaldehyde in PBS and counterstained them with crystal violet-formaldehyde (Sigma-Aldrich).

Full Genome Sequencing and Assembly

For the complete genome sequencing of the Zika virus isolate, we depleted 200 ng of total RNA extracted from the supernatants of infected Vero cell monolayers of rRNA using the NEBNext rRNA Depletion Kit (human/mouse/rat) following the manufacturer’s instructions (New England Biolabs, https://www.neb.com). We constructed RNaseq Illumina shotgun libraries at the Unidad Universitaria de Secuenciación Masiva y Bioinformática-Instituto de Biotecnología and sequenced them using paired-end sequencing with a MySeq system (Illumina, https://www.illumina.com). We assessed the quality of reads using FASTQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc); low-quality positions and reads were eliminated using in-house scripts. We assembled the valid reads without reference using the program Trinity 2.8 (16) and evaluated the assembly using Qualimap (17). Finally, we verified the identity of the contig using blastn (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Phylogenetic Reconstruction

To determine the phylogenetic relatedness of the Zika virus isolate, we retrieved relevant Zika virus open reading frame (ORF) RNA sequences from the Virus Pathogen Resource (ViPR) database (18) (http://www.viprbrc.org). We removed identical sequences and those with undetermined bases from alignment. We applied near-identity clustering (0.999) to the remaining sequences in Cd-hit-test (19) and used 98 sequences for phylogenetic inference by maximum likelihood in RAxML (20) under a general time reversible plus gamma substitution model. We reconstructed the tree with 200 bootstrap replicas. We normalized branch lengths and condensed nodes with bootstrap values <50% to emphasize tree topology.

Variant Analysis

To determine the presence of variants in the larva-derived sequence, we used the alignment to identify positions that had mutations unique to our sequence or that were primarily shared with other mosquito-derived sequences and that were absent or rare in the human-derived genomes. To this end, we examined the alignment with the tool meta-CATS (21), which performs a χ² test to find positions with different polymorphism distribution between groups.

Zika Virus Isolation and Identification in Ae. aegypti Larvae

During the study period, we analyzed 151 larvae pools by qRT-PCR in search of Zika virus RNA. Only 17 (10.8%) of the pools tested positive: 9 (5.7%) from the larvae raised from the eggs collected in June and 8 (5.1%) from larvae raised from the eggs collected in November. To determine the proportion of infected larvae in the population, we calculated MIRs for the 2 collection periods (June and November 2016) and found an increase in the MIR observed from the larvae raised from the eggs collected in June (2.5) to the MIR from the eggs collected in November (6.9) (Table 1).

Figure 2

Figure 2. Phenotypic analysis of Zika virus isolate 31N from an Aedes aegypti larval pool, Jojutla, Morelos, Mexico. A) Cytopathic effect of the Zika virus isolate 31N in C6/36 and Vero cells. The...

To confirm the presence of infectious Zika virus in the larvae pools, we used 11 of the 17 pools that tested positive for Zika virus RNA that displayed the lowest Cq values to attempt viral isolation. We were able to isolate infectious Zika virus from only 1 of the larvae pools (31N), as determined by the presence of cytopathic effect in C6/36 cells characterized by monolayer detachment (Figure 2, panel A) and the detection of Zika virus RNA in culture supernatants by qRT-PCR with a Cq value of 23.73.

The ability of the 31N isolate to infect a mammalian cell line was confirmed by the appearance of the characteristic cytopathic effect of Zika virus in infected Vero cell cultures, characterized by cell rounding and detachment (Figure 2, panel A). Moreover, immunofluorescent detection of the viral envelope (E) protein using the monoclonal antibody 4G2 revealed perinuclear staining (Figure 2, panel B). The plaque assays carried out in Vero cells revealed that the 31N isolate has the ability to produce lytic plaques; thus, this isolate can be considered cytopathic in mammalian cell culture (Figure 2, panel C).

Complete Genome Sequencing of Zika Virus Isolate 31N

We isolated viral RNA from the supernatant of Vero cells infected with the fourth passage of Zika virus isolate 31N and processed it by RNA sequencing by the NGS MiSeq Illumina protocol. Around 40% of the total reads (6,535,816) were assembled in a single contig of 10,795 nt in length, with a mean depth of 42,000 reads. BLAST results of the assembled contig revealed that the strain ZIKV/Aedes.sp/MEX/MEX_I-7/2016, which belongs to a Zika virus isolated from Aedes mosquitoes obtained in the state of Chiapas, Mexico, had the highest identity (98.8%) with our isolate.

Variant Analysis

We identified single-nucleotide variants (SNVs) in the Zika virus isolate 31N genome by aligning the sequence with other mosquito- and human-derived sequences obtained from the ViPR database and running the meta-CATS analytic tool, as described in the Materials and Methods section. To carry out this analysis, we grouped the sequences by the host of origin. We identified SNVs in positions 3176, 3286, and 5636 that were unique to the larva genome. On the other hand, the SNVs in positions 2071 and 3333 were found in human-derived sequences but were not present in other mosquito genomes. The SNV in position 2071 was shared only with a sequence from French Polynesia, likely an example of parallel evolution, whereas the polymorphism in position 3333 was common with 2 sequences from Mexico and may correspond to a local variant. Finally, we found 8 SNVs that were also present in other human and mosquito sequences but were more common in mosquitoes. Only 1 SNV was found in the E gene, whereas the rest were found in nonstructural (NS) genes: 5 in NS2A, 4 in NS3, 2 in NS5, and 1 in NS1 (Table 2).

Phylogenetic Reconstruction of Zika Virus

Figure 3

Figure 3. Condensed phylogenetic reconstruction of selected complete Zika virus genomes. This tree depicts the phylogenetic relationships between Zika virus isolate 31N from an Aedes aegypti larval pool, Jojutla, Morelos, Mexico (bold), and...

To characterize the evolutionary relationship between the 31N isolate and other Zika virus genomes that have been previously reported, we performed a phylogenetic analysis. The phylogenetic reconstruction of the complete genomes of 98 Zika virus sequences, including the 31N isolate, revealed that the 31N isolate belongs to the Asian-American lineage of Zika virus and clusters together with other sequences of human and mosquito origin from Mexico (Figure 3).