Vector competence of Aedes aegypti and Culex quinquefasciatus from the metropolitan area of Guadalajara, Jalisco, Mexico for Zika virus

Zika virus (ZIKV) is a mosquito-borne pathogen discovered in the late 40’s in Uganda during a surveillance program for yellow fever. By 2014 the virus reached Eastern Island in the Americas, and two years later, the virus spread to almost all countries and territories of the Americas. The mosquito Aedes aegypti has been identified as the main vector of the disease, and several researchers have also studied the vector competence of Culex quinquefasciatus in virus transmission. The aim of the present study was to evaluate the vector competence of Ae. aegypti and Cx. quinquefasciatus in order to understand their roles in the transmission of ZIKV in Guadalajara, Jalisco, Mexico. In blood feeding laboratry experiments, we found that Ae. aegypti mosquitoes showed to be a competent vector able to transmit ZIKV in this area. On the other hand, we found that F0 Cx. quinquefasciatus mosquitoes are refractory to ZIKV infection, dissemination and transmission.


Results
Mosquito oral infections. A total of 492 female mosquitoes were fed on eleven blood meal dates, and 412 survived for dissection and saliva collection at 14 dpi (some dissections were performed at 13 dpi and, in one case, 10 dpi) ( Table 1). From these 412 mosquitoes, 270 (65.5%) were Ae. aegypti and 142 (34.4%) Cx. quinquefasciatus. Blood meals provided to the mosquitoes were performed using ZIKV strains with different virus titers expressed as TCID 50 /mL, in three different orders of magnitude (10 4 , 10 5 , and 10 6 ) designated as low, medium and high virus concentrations (Table 1). From all 412 mosquitoes, saliva was collected, and each mosquito was dissected to measure transmission, infection and dissemination of the virus; mosquito body parts were ground with plastic pestles. All samples were mixed in the medium by pipetting, and supernatants were inoculated into individual wells from a 48 well plate containing confluent Vero cells cultures, looking for CPE caused by the presence of the virus in the mosquito homogenates. Of these, 289 from Ae. aegypti and one from Cx. quinquefasciatus inoculations showed CPE. More specifically, at medium virus concentration, a total of 82 Ae. aegypti mosquitoes showed CPE, while at the higher virus concentration 207 Ae. aegypti and 1 Cx. quinquefasciatus showed CPE. No CPE was observed in the lower virus concentration for both species (Table 2).
Susceptibility to ZiKV infection. The mosquito body homogenates were assessed for ZIKV infection in cell culture looking for CPE. The infection rate (IR) was calculated with the number of infected mosquito bodies divided by the total tested. In the lowest virus concentration, 37 bodies corresponding to Ae. aegypti and 22 to Cx. quinquefasciatus were assessed. None presented CPE, resulting in zero infection rates. In the medium virus concentration, out of 117 engorged Ae. aegypti, 45 bodies showed CPE resulting in an infection rate of 38.46%. For Cx. quinquefasciatus 37 bodies were evaluated and none produced CPE (0%). For the highest virus concentration, Ae. aegypti showed an infection rate of 93.10% (108 bodies), and for Cx quinquefasciatus 83 mosquitoes were evaluated without the appearance of CPE (zero infection rates) ( Table 2). Statistical analysis showed significant differences (p < 0.01) between the rates obtained with medium and high virus concentrations for Ae. aegypti mosquitoes (Fig. 1). For Cx. quinquefasciatus there was no infection, and therefore no differences were found in any of the viral titers. The calculation of minimum infection rate (MIR) was performed using the PooledInfRate v.4.0 program. Since this calculation is for mosquito pools ([number of positive pools/total specimens tested] x 1000) we use the mosquitoes from each feeding date as if they were mosquito pools; therefore, a total of 11 pools were formed and separated according to the viral titer used (10 4 , 10 5 , 10 6 ) to calculate MIR. For Ae. aegypti mosquitoes, MIR of 42.74 and 34.48 were found for the medium and high virus concentration, respectively ( Table 3).

Dissemination of ZiKV infection in mosquitoes.
For dissemination, the mosquito head homogenates were analyzed in cell culture looking for CPE. The dissemination rate (DR) was calculated with the number of infected mosquito heads divided by the total tested, and the disseminated infection rate (DIR) was calculated with the number of infected mosquito heads divided by the infected mosquito bodies. In the lowest virus concentration, no dissemination was observed for both species. In the medium virus concentration, 28 heads of Ae. aegypti showed CPE resulting in 23.93% of DR and 62.22% of DIR. Again, no presence of CPE was observed for Cx. quinquefasciatus mosquitoes. For the high virus concentration in the case of Ae. aegypti, 71.55% of DR and 76.85% of DIR were obtained (83 positive heads). In Cx. quinquefasciatus the dissemination rates were once again 0% (Table 2). For Ae. aegypti, when statistically comparing the ranges of DR obtained in medium and high virus concentrations (10 5 and 10 6 ), significant differences were observed (p < 0.01) (Fig. 2a). On the other hand, no significant differences were observed in DIR between the titers of 10 5 and 10 6 (Fig. 2b). In the case of Cx. quinquefasciatus mosquitoes, no significant differences were observed for DR and DIR since both parameters were 0% (Fig. 2a,b).

ZiKV transmission.
Mosquito saliva was analyzed in cell culture looking for CPE in order to better understand the transmission of ZIKV. Estimation of transmission rate (TR) was calculated using the number of mosquitoes with infected saliva divided by infected mosquito heads, and transmission efficiency (TE) was calculated using the number of mosquitoes with infected saliva divided by the number of mosquitoes tested. In the lowest virus concentration, no transmission was observed for both species. In the medium virus concentration for Ae. aegypti, 9 positive saliva were found (TR of 32.14% and a TE of 7.69%) and for Cx. quinquefasciatus both rates were zero. For the high virus concentration in the case of Ae. aegypti, 16 saliva showed CPE (19.28% of TR and 13.79% of TE); in Cx. quinquefasciatus a single saliva showed CPE (TR of 0% and TE of 1.20%) ( Table 2). In the case of Ae. aegypti, statistical analysis showed no significant differences for both TR and TE in medium and high  www.nature.com/scientificreports www.nature.com/scientificreports/ virus concentrations (10 5 and 10 6 ) (Fig. 3a,b). On the other hand statistical differences were found (p < 0.05) when comparing the TE values between Ae. aegypti and Cx. quinquefasciatus, in the high virus concentration (10 6 ) (Fig. 3b).

Viral load in Saliva.
In order to estimate the ZIKV load and titer expectorated in mosquito saliva, RT-qPCRs were performed using RNA extractions directly from the original saliva tube. Standard curves were designed using RNA extraction from a ZIKV sample with a viral titer of (2.68 × 10 6 TCID 50 /mL) to compare titers in saliva. All PCR reaction were performed in triplicate. Out of the 26 saliva that yielded CPE in cell cultures, 23 nucleic acids extracted from the original tube were detected by the RT-qPCR reactions, including the single saliva that yielded CPE in Cx. quinquefasciatus. Using CT results of the saliva, the viral titer in both TCID 50 /mL and the    In DIR, no significant differences were found for Ae. aegypti. In the case of mosquitoes Cx. quinquefasciatus DR and DIR do not differ significantly. Statistical analysis was performed using Student's t-test.

Discussion
The viral loads for ZIKV on infected hosts varies through the time of infection 21 . Depending on viral titer in a bloodmeal, mosquito vectors will acquire different loads of virus that may or may not lead to infection 22 . ZIKV infected patients, usually present low viremia levels ranging from 10 3 to 10 6 RNA copies/mL 23 ; because of this, the viral titers used in this study were of three different orders of magnitude (10 4 , 10 5 , and 10 6 ) starting from the first viral load that is 100-fold higher than viremia reported for patients, trying to assure mosquito infections, and we designated them as low, medium and high virus concentrations in feeding experiments for both Ae. aegypti and Cx. quinquefasciatus mosquitoes; these virus concentrations were similar to those used in other studies. For instance, Fernandes et al. evaluated Cx. quinquefasciatus mosquitoes using a titer of 10 6 PFU/mL 15 ; Huang et al. also infected Cx. quinquefasciatus, but they used a higher order of magnitude, 10 7 TCID 50 /mL 24 . Chouin-Carneiro et al. used the same viral order of magnitude (10 7 TCID 50 /mL) and in this study they infected Ae. aegypti mosquitoes 12 . In another study carried out in Singapore, they analyzed both Cx. quinquefasciatus and Ae. aegypti mosquitoes, feeding them with 10 5 and 10 6 PFU/mL, and both species were infected. They then tested titers of different orders of magnitude in Ae. aegypti, (10 5 , 10 4 , 10 3 and 10 2 PFU/mL), and they proved that a viral load of 10 3 PFU/ mL was enough to infect the mosquito strains that they used for the experiments 25 . In our study, using a concentration of 10 4 TCID 50 /mL (low virus concentration) to engorge both mosquito species, resulted in no infection, dissemination or transmission. When we used the medium virus concentration (10 5 TCID 50 /mL), moderate to low rates of infection, dissemination and transmission were obtained for Ae. aegypti, where the highest rate was in the DIR, (dissemination in the mosquito bodies). For the high virus concertation (10 6 TCID 50 /mL), a change was observed in all the rates, where both, infection and dissemination rates were high, and transmission rates low. But only IR and DR presented statistical differences. In general, it could be observed that exposing mosquitoes to a higher virus concentration resulted in overall greater infection rates for Ae. aegypti. This conclusion was previously observed in a study using this mosquito species from Florida in a dose-response experiment 13 . For all the different virus concentrations used in the present study for Cx quinquefasciatus mosquitoes, there was no infection or dissemination of the virus; only a single saliva was able to generate CPE in cell culture. Thus, it can be concluded that this species is refractory to ZIKV infection in our laboratory assays, as many other studies have concluded 11,14,15,26 . In our study, MIR calculation was used for Ae. aegypti mosquitoes, and the obtained rates from medium and high virus concentration were higher (42.74 and 34.48, respectively) than those presented in a previous publication of our research group from wild caught Ae. aegypti mosquitoes where the MIR was 10.28, but were similar to other species evaluated in the same study 20 . In Cx. quinquefasciatus mosquitoes, all calculated MIRs were null.
Although Ae. aegypti is considered the main vector of the virus, previous studies of vector competence have produced contrasting results. Some have showed low to moderate percentages of virus competence 11,12,14,[27][28][29] , but others have shown that this species is highly competent 14,15,25,30,31 . The results in our study are more similar to those in which ranges of infection and dissemination are high to moderate, but transmission rates are low.
The estimated viral titers that we found in the saliva of Ae. aegypti and from a single saliva sample in Cx. quinquefasciatus were predominantly in a range of 500-1600 PFU / mL, and these titers are similar to those described by Dudley et al., where they found a ZIKV dose delivered by a mosquito to be 10 1.5 to 10 3.2 PFU per mosquito 32 . Some researchers have reported that in vitro mosquito salivation overestimates the amount of virus inoculated, compared to virus deposited at sites of in vivo blood feeding for other arboviruses 33 , while other researchers have reported that mosquitoes inoculate high doses of WNV as they probe and feed on peripheral tissues in animal models 34 . So, to determine precisely the amount of ZIKV transmitted by mosquitoes, more studies are required, but in general it can be considered that titers found in our study could be enough to transmit the virus during a blood feeding. www.nature.com/scientificreports www.nature.com/scientificreports/ Vector competence for ZIKV varies greatly. As has been published in our research, we found susceptibility to virus infection can be due to mosquito-virus match factors (also known as genotype-genotype interactions), including the source of blood used in experiments and viral load, among others. The role of immunity against the virus is not entirely clear. Although some mosquitoes are not considered highly competent vectors, they may be able to transmit the virus efficiently in the wild if they have an anthropophilic behavior.   www.nature.com/scientificreports www.nature.com/scientificreports/ In the case of saliva that were positive without presenting CPE in the body and head of the mosquitoes (one of Cx. quinquefasciatus and another of Ae. aegypti), this could be due to several factors involved to mosquito immunity, with a clarification phenomenon previously described. Some studies have described results in which the virus is not detectable in the midgut or the viral titers start to decrease, but the salivary glands of the same mosquito are positive 17,30 . In our study, ZIKV was found only in saliva but not in the head or body in one single pool from each species and we think this could be due to this phenomenon, nevertheless, another possibility could be related to a false positive result due to the high sensitivity of the assay. The clearance phenomenon has been described in both Cx. quinquefasciatus and in Ae. agypti, with ZIKV, DENV and WNV [35][36][37][38] . In the work of Liu et al., Cx. quinquefasciatus mosquitoes were found with infected midguts, where the viral titer was decreasing, and after 7 dpi, the virus was no longer detected. Although the clearance is not specifically mentioned, there is a possibility that it may be involved; the authors suggested that this phenomenon could have happened due to undigested infected blood. However, as little is known of vector competence of Cx. quinquefasciatus with ZIKV and the immune responses involved, these or other hypotheses may be involved here.
On the other hand, another hypothesis is that the blood source could also be involved in the results presented herein regarding the Cx. quinquefasciatus refractory tendencies to ZIKV infection. If mosquitoes had been fed with blood obtained directly from infected patients, the results could be different. It could be interesting to study this hypothesis, since in a previous publication from our group, ZIKV was isolated from salivary glands of wild Cx. quinquefasciatus mosquitoes and also from male mosquitoes from this species. In animal models for instance, Dudley et al. reported Ae. aegypti mosquitos acquiring ZIKV infection from a macaque infected by mosquito bites but not from macaques infected with a virus administrated by syringe, concluding that virus with different passage histories could result in genotypic and phenotypic differences 32 . Therefore, differences in the viral stocks we used for oral infection experiments could favor the infection in one mosquito species but not the other. A similar case using an animal model is mentioned in the article by Roundy et al., where they compared natural and artificial blood feeds. They observed that murine blood had greater infectivity than blood that was given in artificial feeds. They concluded that this difference in competence after blood exposure undoubtedly contributed to the underestimation of Ae. aegypti mosquitoes as a vector of ZIKV in previous studies 18 . Therefore, more studies are needed to understand the role of the cell lines used for virus amplification in the vector competence studies depending on the mosquito species.
In addition, factors such as environment, mosquito density, feeding patterns, mosquito survival rates (known as vector capacity), and others also influence the transmission of the virus. For example, Tabachnick et al. and Smartt et al. mention that there are environmental and biological conditions where Cx. quinquefasciatus mosquitoes are probably competent for ZIKV, and these mosquitoes may play a role in the transmission of the virus to humans 37,39 . Richard et al. showed that two Aedes species from French Polynesia did not show sufficient competence to transmit ZIKV and suggested that another mosquito species could contribute to the spread of the virus in that region. They also mentioned that Cx. quinquefasciatus is an abundant species that could have a role in the transmission of the virus 40 .
In our study, increasing the viral titers for the oral infections in Ae. aegypti mosquitoes showed increasing numbers in all rates. Although transmission rates were not significantly different, the increase of mosquitoes with ZIKV in the saliva could be observed, suggesting that this phenomenon could happen in the wild.
Even though Cx. quinquefasciatus mosquitoes can be found infected in nature (despite that reported viral titers were not according to those for transmitting the virus to humans), laboratory studies do not always manage to reproduce wild conditions, as happened in our study. However, there are publications where infection has been achieved, but the factors that influence the susceptibility in Cx. quinquefasciatus are not clear. Therefore, several analyses are needed in order to elucidate the exact role of this species in the cycle of transmission of the virus.
In conclusion, even when we detected the virus in salivary glands of wild Cx. quinquefasciatus mosquitoes in a previous study 20 , surprisingly, in the present work using F0 mosquitoes for oral blood feeding laboratory experiments, we found this species refractory to ZIKV infection, dissemination and transmission. Further analysis trying to explain this contrast should therefore be performed. On the other hand, in the case of Ae. aegypti mosquitoes, this species showed to be a competent vector to transmit ZIKV in the Guadalajara Jalisco metropolitan area.

Methods
Biosafety and ethical approval. This study was approved by the Centro de investigación y Asistencia en Three different ZIKV strains were used in this study for the oral exposure experiments, or as positive control on RT-qPCR reactions; two of them were isolated from wild caught Ae. aegypti and Cx. quinquefasciatus mosquitoes from the Guadalajara, Jalisco, metropolitan area in 2016, and these isolates were designated as Ae. G6, and Cx. The third strain used was isolated in CIATEJ from a serum sample previously confirmed as ZIKV by RT-qPCR at the Central Laboratory of Epidemiology, Mexican Institute of Social Security. The serum was inoculated into a 25 cm 2 flask containing a confluent monolayer of C6/36 cells, and adsorbed for 1 hr. at 28 °C, rocking the flask every 15 min to distribute the inoculum; then L-15 medium containing 2% FBS was added, and incubated at 28 °C; the culture was observed under an inverted microscope daily until cytopathic effect (CPE) appeared. In order to confirm virus isolation, supernatant from the culture that showed CPE was filtered with 0.22um membrane disk and re-inoculated in fresh cells to confirm virus isolation by re appearance of CPE. The identity of all virus strains used in the study was confirmed at CIATEJ by RT-qPCR using the primers and probes previously reported by Lanciotti et al. 23 .
Virus titration. In order to quantify the virus for the experiments, the tissue culture infectious dose 50 (TCID 50 ) was calculated. Vero cells were seeded into 96-well plate and incubated overnight for adherence. 10-fold serial dilutions of viral supernatant from the second passage were prepared in L-15 medium containing 2% FBS. Then, 100 µL of these dilutions was inoculated per well using 8 replicates per dilution. After 5 days, the plate was analyzed looking for CPE. The viral titer was quantified as described by Reed and Muench 41 . Mosquitoes. Ae. aegypti and Cx. quinquefasciatus mosquito species were used for the vector competence experiments using the adults (F0) direclty from eggs or larvae collected from areas where no human confirmed cases were recored by health autorities, trying to reduce the posibility to emerged vertical ZIKV infected mosquitoes which could bias the results. Emerged mosquitoes were mantained in the CIATEJ insectary with the following conditions: 28 ± 1 °C with a light cycle: darkness of 12 h: 12 h and a relative humidity of 70%, approximately, until mosquito oral experimental infections. In the case of Ae. aegypti, eggs, larvae and adults of this species were provided by the Entomological Research Unit of the Department of Public Health of the State of Jalisco; for Cx. quinquefasciatus, larvae collections were carried out with personnel of this same department along with CIATEJ personnel.

Mosquito oral infections.
In order to determine vector competence, different mosquito batches were fed with human blood containing virus. Five to seven day old mosquito females were deposited in feeding boxes; 50 to 80 females extracted from maintenance cages of mosquito colonies by mechanical aspiration were placed in each box. Mosquitoes were kept without food for 48 hours, with a source of water (swabs with water) which was removed from the feeding boxes 24 hours before oral infections in order to ensure that a large proportion of females were fed. The food source in the feeding experiments was a mixture containing 1 mL of blood and 1 mL of viral suspension with a known viral titer, supplemented with adenosine triphosphate (ATP) as a phagostimulant, in a final concentration of 10 mM. The blood was obtained by venous puncture in vacutainer tubes with heparin as anticoagulant from volunteers involved in the project at CIATEJ. Infectious blood meals were performed using glass feeders connected to the water recirculation system at 37 °C, covered with a parafilm membrane (plastic film). Exposure time to bloodmeals was limited from 60 to 90 min. Engorged females were transferred to new feeding boxes and kept with swabs with 10% sucrose and water in the insectarium at 28 ± 1 °C, a light cycle: darkness of 12 h: 12 h and 70% humidity.
infection, dissemination and transmission analysis. In order to analyze infection, dissemination and transmission of ZIKV, batches of 10 to 60 mosquitoes were analyzed 14 days after infection (dpi). Only in some cases, when many mosquitoes died before reaching 14 dpi, was the dissection performed a few days earlier (most of the time 13 dpi and in one case 10 dpi) on the still alive mosquitoes. The number of mosquitoes per group depended on how many mosquitoes had fed in that batch. For each individual female mosquito, 3 different samples were collected: the body (B) (abdomen and thorax) to estimate the infection, the head (H) for virus dissemination and saliva (S), for virus transmission. Each of these samples was deposited into individual 1.5 tubes containing viral maintenance medium (200 μL of D-MEM medium supplemented with 20% FBS and 1% penicillin -streptomycin -amphotericin cocktail).
To obtain the three different samples per mosquito, subsets of mosquitoes per batch were knocked down by cold shock (−20 °C) for approximately 40-50 sec. or until they were fully anesthetized. Then, to collect saliva, wings and legs were removed and discarded immediately after the cold shock and the mosquito proboscis was inserted into a 30 μL microcapilar tube (Microcaps, Drummond Scientific Company, Broomall, PA, USA) or 25 μL microcapilar (Kimble ® microcapillary pipettes, Sigma-Aldrich, St. Louis, MO, USA), containing viral maintenance medium. After 45 minutes, microcapillary tubes were placed into a tube with the medium. Subsequently, heads were separated from the body and each of these parts was placed into a different tube containing the same medium. Next, all tubes were centrifuged at 14,000 rpm for 1 min. and stored at −80 °C for further analysis. In order to identify the infection in the different samples, the tubes were thawed, the mosquito parts were ground in the same medium in which they were stored, and the resultant homogenates were centrifuged at 10,000 rpm for 10 min, including tubes with saliva. All centrifugations were performed in an eppendorf microcentrifuge 5417R (Eppendorf, Hamburg, Germany). Next, 50 μL of each supernatant (B, H, S) were inoculated into individual wells of a 48 well plate with Vero cells and incubated for 1 h at 37 °C. After this time, cell maintenance medium was added. The plates were incubated at 37 °C and examined daily for evidence of viral CPE for a maximum period of 21 days. The remaining homogenates were stored at −80 °C for further analysis. Viral analysis in positive samples. Once mosquito homogenate parts showed CPE in the cell culture, they were considered as positive to the virus; next, nucleic acid extraction was carried out directly from each homogenate stored at −80, using a QiAmp Viral RNA Mini Kit (Qiagen ™ , Hilden, Germany). In order to determinate presence of the virus and viral load in the original homogenates, RT-qPCRs were carried out in a Light Cycler 480 II PCR instrument (Roche Diagnostics, Penzberg, Germany) using the next 2 kits: Aedex³ Kit (Genes2life, Irapuato, Mexico) following the manufacturer's protocol, and the primers and probe for viral load quantification