Vector competence of Culex mosquitoes (Diptera: Culicidae) in Zika virus transmission: an integrative review.

Objective
To identify studies on the competence of Culex mosquitoes as vectors for the transmission of Zika virus (ZIKV) around the globe.


Methods
We performed an integrative review to identify relevant articles on specific experiments to determine whether Culex mosquitoes are vectors for ZIKV. The sources we used for our research were the Brazilian Periódicos CAPES electronic portal (MEDLINE/PubMed, ScienceDirect Journals, Nature Publishing Group, SciELO, Springer Link, and 250 other databases) and gray literature.


Results
We identified 344 studies, of which 36 were considered for this review. In 8 studies, infection in salivary glands of Culex quinquefasciatus, Culex restuans, Culex tarsalis, and Culex coronator was detected. Cx. quinquefasciatus was the most studied among those confirmed as potential ZIKV vectors, and only strains of Asian lineages (THA/2014/SV0127-14; SZ01 (2016)) and American lineages (BRPE243 (2015); PRVABC59 (2015)) can infect the salivary glands of Culex mosquitoes. The tested African strains (MR766 and DAK AR 41525) were unable to infect salivary glands.


Conclusions
There is still a lack of compelling evidence that indicates Culex spp. are a competent ZIKV vector, but they should remain a target for further monitoring studies, especially regarding ZIKV transmission to other species. Furthermore, studies should not be limited to studying whether their salivary glands are infected.

Zika virus (ZIKV) is known to be transmitted among humans mainly through bites of Aedes aegypti (Linnaeus) and Aedes albopictus (Skuse) mosquitoes (1). The virus was initially isolated in a rhesus monkey in 1947. There was a second isolation from Aedes africanus (Theobald) in 1948 in an attempt to isolate yellow fever virus from mosquitoes in the Zika Forest of Uganda (2). Aedes mosquitoes are considered the only competent vectors for ZIKV transmission (3,4). Transmission can occur sexually (5), through blood transfusion and saliva, and from mother to child during pregnancy, birth, and breast-feeding (6).
ZIKV is a positively enveloped RNA virus member of the Flaviviridae family, genus Flavivirus (1,7). It was discovered in 1947 in the Zika Forest, in Uganda, and remained confined to some areas of Africa and Asia. In 2007, ZIKV emerged in the Yap Islands in the Federated States of Micronesia and also in the African country of Gabon. In addition, in 2013, the virus appeared in French Polynesia. By 2014, ZIKV had spread to other Pacific islands: New Caledonia, the Cook Islands, and Easter Island. In early 2015, the virus was identified in Brazil and then, later, throughout continental South America and Latin America (1,7). This fast and massive spread is worrisome because there are no available drugs or vaccines for the treatment of ZIKV infection, and a possible marked, severe outcome of ZIKV infection in pregnant women is microcephaly in newborns (6).
This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 IGO License, which permits use, distribution, and reproduction in any medium, provided the original work is properly cited. No modifications or commercial use of this article are permitted. In any reproduction of this article there should not be any suggestion that PAHO or this article endorse any specific organization or products. The use of the PAHO logo is not permitted. This notice should be preserved along with the article's original URL.
Within the genus Aedes, other anthropophilic species have been considered to be vectors, including Aedes hensilli (Farner) and Aedes vexans (Meigen). In addition, other mosquito genera have been investigated as vectors, especially Culex. (3,4,8,9). Culex spp. are already considered competent vectors to transmit such flaviviruses as human-like Japanese encephalitis virus (JEV), West Nile virus (WNV), and Saint Louis encephalitis (10). This fact makes them candidates for further investigations as a vector for other flaviviruses, such as ZIKV.
In a study performed in China, Culex quinquefasciatus (Say) was identified as a potential laboratory vector for ZIKV. In that research, mosquitoes were infected through a blood meal with a ZIKV strain (SZ01) isolated from a patient, and viral RNA was found in the salivary glands, midgut, and ovary. Importantly, Cx. quinquefasciatus mosquitoes were able to infect infant mice (11). In a session of the First International Workshop on Zika Virus, organized by the Oswaldo Cruz Foundation (FIOCRUZ) and held in Brazil in 2016, researchers from the Ageu Magalhães Institute (Recife, Pernambuco, Brazil) presented results of experiments with mosquitoes artificially fed with blood infected with the ZIKV. The virus was detected in the body and in salivary glands of Cx. quinquefasciatus 7 and 15 days after feeding, with a confirmed infection rate of 100% and 67%, respectively (12).
Currently, there are no vaccines or drugs against ZIKV (1,4,5), and the only precaution is to prevent mosquito bites by integrated vector control, including surveillance, biological and chemical control, source reduction, and environmental management (4,5). For example, the development of sanitization and management of urban water collection and vector control with DDT was responsible for vector Aedes mosquitoes disappearing after 1950, although the species has recently reinvaded European territory (4). Furthermore, the World Health Organization (WHO) also recommends the practice of safe sex for women living in areas of high virus transmission (13).
Therefore, considering that the genus Aedes is the principal target among the strategies for the prevention and control of ZIKV, we reasoned that it is essential to understand whether there are other vectors capable of or even competent in transmitting ZIKV. Thus, we aimed to identify studies on the competence of the Culex mosquitoes as vectors for ZIKV transmission through a review of the literature on this subject.

METHODS
We performed a literature review in July 2019 to identify relevant articles on the vector competence of Culex spp. mosquitoes for ZIKV. The review was based on an advanced search of the Brazilian Periódicos CAPES electronic portal (www.periodicos. capes.gov.br), which includes MEDLINE/PubMed, ScienceDirect Journals, Nature Publishing Group, SciELO, Springer Link, and 250 other databases. Additional studies were identified by searching for gray literature, using the Google Scholar search engine, and with a manual search of the bibliographical references of the relevant identified publications.
The search strategy was drawn from the descriptors ("Culex" and "vector") AND (zik*), marking "any field" for the search field of the thesauri, in the advanced-search area of the Periódicos CAPES electronic portal. All studies from the literature search were analyzed to eliminate duplicates from the manual search strategies, specifically by comparing authors, titles, and name of the journal, along with their volume, number, and year of publication. After the duplicates were removed, the remaining studies were examined based on their titles and abstracts. At this stage, the eligibility criteria for the articles were: original studies; mentions of ZIKV and a mosquito from the genus Culex; establishes Culex spp. as a vector of ZIKV through specific results in experiments identifying the virus at their organisms; and in Portuguese, Spanish, or English. Reviews and opinion letters were discarded. The documents selected as potentially relevant went to the next step, where they were judged carefully from the reading of the full text. Only those that met all the eligibility criteria mentioned above were included in this review. The studies from each database were placed in spreadsheets in the Microsoft Excel program (version 2010), to eliminate duplicates and to create a database for references.

RESULTS
The study designs were classified as cohort studies (10 articles), case-control studies (26 articles), and a conference abstract (1 article). One of the documents presented two different study classifications, case-control and cohort.

Study selection
We identified 333 studies in the bibliographic search in the Periódicos CAPES website, 7 studies through gray literature, and 7 additional studies from the manual search of the bibliographical references of the relevant publications, for a total of 347 studies. Three duplicate papers were discarded, and then 278 articles were excluded after reading the title. We incorporated 66 studies whose titles were not clear concerning inclusion criteria into the next stage, which included reading the abstracts to avoid missing any article that matched our criteria (all the articles were written in English, so no study was excluded based on language). In the end, after discarding 26 articles by reading the abstract and 4 articles by reading the full text, we included 36 studies in this review. Figure 1 is a flow diagram of the methodology steps. Table 1 summarizes the articles used in this review.

DISCUSSION
In our work, we found 36 relevant studies. This was 18 more works than the already identified articles in five reviews in the literature that examined the capacity of Culex spp. to transmit ZIKV (3,4,8,9,48). Cx. quinquefasciatus was identified in 6 articles, while Cx. restuans, Cx. tarsalis, and Cx. coronator were identified as possibly competent in 1 study each.
As stated before, researchers from the Ageu Magalhães Institute have detected ZIKV in salivary glands of Cx. quinquefasciatus at 7 and 15 days after feeding, and confirmed an infection rate of 100% and 67%, respectively (12). Additionally, one Chinese paper (11) was the first to demonstrate the presence of ZIKV (Strain SZ01) in the salivary glands of Cx. Quinquefasciatus; it was also the only study to show the transmission capacity to another species. This work reported a transmission rate of 89% to mice; these animals had viral RNA in their brain at 10 days post-engorgement, with a titer of 7.85 RNA log 10 copies/ml. On the other hand, this study did not provide experiments with the same methodology with Ae. aegypti as a positive control.
One study demonstrated the presence of ZIKV in the salivary glands of Cx. quinquefasciatus mosquitoes in concentrations similar (P > 0.05) to those found in Ae. aegypti, both fed with 1 × 10 6 PFU/ml (log 10 ) blood meal titers (43). Even with a minimum blood meal titer of 1 × 10 4 PFU/ml (log 10 ), Cx. quinquefasciatus salivary glands were infected. These findings indicate that Cx. quinquefasciatus can produce virus in the salivary gland even when fed with low titers of viral particles. This ability better mimics what occurs in nature, where mean human viremia is lower than 2.5 log 10 PFU/ml (49). Likewise, the study in the Brazilian state of Pernambuco (43) investigated wild-caught mosquitoes from the city of Recife. They observed no traces of recent feeding and used electron microscopy to detect ZIKV in the salivary glands. That paper did not perform experiments to analyze ZIKV transmission from Cx. quinquefasciatus to another species.
Dibernardo et al. (44) detected ZIKV in 3 of 58 (5%) of salivary glands of Cx. restuans just after intrathoracic inoculation. Cx. tarsalis was refractory using the same methodology. The authors believe that Cx. restuans could transmit by bite but also suggested the presence of salivary and midgut barriers for Cx. restuans and Cx. tarsalis. They concluded that Cx. restuans is not a competent ZIKV vector due to its feeding behavior.
Elizondo-Quiroga et al. (45) detected ZIKV in wild-caught Cx. tarsalis, Cx. coronator, and Cx. quinquefasciatus. However, this study was not conclusive regarding the ZIKV infection titers in the saliva of the collected mosquitoes because the mosquitoes that presented the lowest maximum titer of infection were Ae. aegypti, the primary ZIKV vector. Unfortunately, for none of those mosquitoes did the authors indicate whether the mosquitoes had any trace of recent feeding or which ZIKV strains were involved. Such deficiencies impair a more detailed assessment of the competence of these Culex spp.
RNA analysis was performed in saliva from Cx. quinquefasciatus eluted from filter paper at 14 days postinfection (DPI) in a first experiment and then at 16 DPI in a second experiment (46). The ZIKV titers were 5.6 ± 4.5 log 10 ZIKV PFUe/ml and 5.02 log 10 ZIKV PFUe/ml, respectively. This same research group (46) neither investigated transmission to another species nor applied the same methodology for Ae. aegypti (as a positive control). The lack of positive controls diminishes the reliability of the results. On the other hand, this paper intriguingly showed that viruses isolated from Culex saliva can form plaques in Vero cells. These data prove that biologically active virus can be obtained from the saliva of those mosquitoes.
A Thai study investigated vertical transmission of ZIKV to larvae in Ae. aegypti, Ae. albopictus, and Cx. quinquefasciatus, with positive results for all the species (47). The ZIKV strain virus/H. sapiens-tc/THA/2014/SV0127-14 infected the salivary glands of these mosquitoes fed with 1.7×10 5 FFU/ml blood meal titers. These Thai investigators (47) did not demonstrate any kind of transmission to other species but applied the same methodology for Ae. aegypti as a positive control.
In two studies (43,46), the infection of salivary glands declined over time, findings that are different from the Chinese study mentioned earlier (11). Thus, there are inconsistent results with Cx. quinquefasciatus, with a marked decrease from day 8 to day 12 postexposure, but an apparent increase again at days 16 and 18 postexposure. Another aspect in these three articles is that they all used different strains: the Chinese study (11) used SZ01, from an infected patient who had returned from Samoa to China; the Brazilian paper (43) used BRPE243/2015, from Pernambuco; and the United States paper (46) used PRCABC59, from Puerto Rico (2015). Another study about vector competence of Aedes mosquitoes argued that in Ae. albopictus and Ae. aegypti, ZIKV transmission can be relatively dependent on the virus strain (49). Thus, it is important to note that diverse strains, especially strains isolated from patients, can present distinct behaviors. This phenomenon may represent one factor responsible for the varied results found in the literature.
With regards to blood meal titers, only one work (43), which used 4.0 log 10 ZIKV PFU/ml, reached an IR of 36% at 7 DPI and 10.53% at 15 DPI. The majority of studies that reached some IR used blood meal titers greater than 5.0 log 10 ZIKV PFU/ ml, 5.0 log 10 ZIKV RNA copies/ml, or 1 × 10 6.7 TCID 50 /ml. The results were refractory at lower blood meal concentrations. As described above, 13 studies did not detect the infection capacity, dissemination, or transmission of any Culex spp., even with higher blood meal titers. Some authors (28,38,44) suggested that the random inability to transmit in Culex mosquitoes may be linked to a gut barrier of some Culex spp., a place where viral particles attack and initiate penetration and replication. However, Amraoui et al. (28) did not demonstrate that inoculation of viral particles into the hemocoel tissue of Cx. quinquefasciatus favored viral ZIKV dissemination or transmission. A study from the United States (46) warned about the existence of specific populations with regard to variability in transmission competence. The authors concluded that some Cx. quinquefasciatus populations may be capable of salivating ZIKV under environmental and other unknown conditions. This statement is noteworthy. It cannot be ignored that some mosquito populations may be more prone to ZIKV infection and dissemination. All these aspects definitely merit further studies that could reveal new intervention approaches.
Another particular condition revealed by Ciota et al. (49) in a study on the vector competence of Aedes mosquitoes was that there were significant differences in the proportion of infected mosquitoes with equivalent ZIKV titers but two different types of meals. Fresh blood meals resulted in a significantly higher IR than did stocked meals frozen and stored at -80°C and then thawed before preparation (P < 0.0001). Thus, considering the documents of Table 1 that indicated the transmission capacity, the Brazilian work in Pernambuco (43) stored the viral stocks at -80°C and subsequently thawed them to prepare the blood meal. However, another paper (11) was unclear about storage conditions and used a stock of virus that had been passaged twice in C6/36 cells prior to the infectious feed. This method suggests the use of a recently prepared blood meal.
Roundy et al. (50) noted that the criterion (iii) proposed by Barnett (51) for incrimination of an arthropod vector (repeated demonstration of natural infection of the vector) has only been fulfilled for Ae. aegypti and Ae. albopictus. Furthermore, as can be seen in our review, only one study (11) demonstrated ZIKV transmission to other species (criteria (iv) for incrimination of an arthropod (51)). However that study used a ZIKV strain not tested in any other work found for this review, and it did not include Ae. aegypti positive control tests. Thus, the defining evidence for Culex spp. as a ZIKV vector is still lacking.
Some authors (32,38,40) agree that the focus on prevention of ZIKV disease should remain on population control of the genus Aedes. Indeed, the probability of a Culex mosquito biting two humans in a sequence and transmitting the Zika virus is small, according to its preference for feeding on avian hosts (52). Nevertheless, Culex spp. are widespread in urban centers and also feed on human blood. Considering the results of studies with collected mosquitoes, when the Culex spp. was infected, the Aedes species was infected (30,42,43,45). Only one study (15) presented data that showed neither Culex spp. nor Aedes mosquitoes infected with ZIKV. Conversely, in four papers (16,24,25,26), there was no Culex mosquito infected while the official vector Ae. aegypti was infected. Two articles (25,26) used the same collection of mosquitoes from the field with different methodologies to investigate viral ZIKV RNA. All these results with field-caught mosquitoes showed that Culex spp. and Ae. aegypti may use identical hosts (but not always). In fact, in a study with mosquitoes from field collections in Thailand (53), in Cx. quinquefasciatus there were mixed blood meals, with 7.84% from humans or monkeys, 47.06% from dogs, and 33.33% from others hosts. Comparatively, in Ae. aegypti, there were also several blood meals: 70.0% from humans plus monkeys or 13.33% only from monkeys and 10.0% from other kinds of hosts. These data demonstrate that variations in favorite hosts from place to place influence the infection rates of Culex spp. and Aedes mosquitoes. Additionally, as argued by Kauffman et al. (8), especially regarding the work from Pernambuco, Brazil (43), Cx. quinquefasciatus may serve as a secondary vector in places with abundant ZIKV infection in humans.
Additional experimental studies that use identical strains, experimental conditions, methodologies, with positive controls in Ae. aegypti and/or Ae. albopictus, and that utilize tests to prove the possibility of ZIKV transmission to other species from Culex spp., could be very decisive for discarding or confirming the contribution of these mosquitoes as a competent or incompetent ZIKV vector. According to the studies we have investigated, we feel that, besides well-implemented sanitization, the main strategies for the prevention and control of ZIKV should remain on the genus Aedes.

Conclusions
This work demonstrated the accumulation of evidence to prove the capacity of the ZIKV to infect Culex spp. However, only 7 studies out of the 36 identified for this review demonstrated the infection of Cx. restuans, Cx. quinquefasciatus, Cx. tarsalis, and Cx. coronator salivary glands. Furthermore, only 1 study showed the capacity of transmission to mice. Considering the records found here, Cx. quinquefasciatus remains the most widely studied species with confirmed salivary glands infected by ZIKV.
Additionally, only Asian or American ZIKV strains were able to infect the salivary glands of Culex mosquitoes: THA/2014/ SV0127-14, SZ01, BRPE243, and PRVABC59. The MR766 and DAK AR 41525 African strains were unable to infect Culex spp. Further experimental studies that utilize the same strains, experimental conditions, use a positive Aedes control, and test ZIKV transmission to other species via Culex spp. are still needed to confirm the contribution of the Culex mosquitoes in ZIKV transmission. We believe that strategies for ZIKV control should stay focused on the genus Aedes, but responsible authorities should continue to monitor Culex spp. mosquitoes, especially regarding their ability to transmit ZIKV to other species. This surveillance should not be limited to determining whether their salivary glands are infected.
Author contributions. All the authors (SGVR, EGR, and WCS) conceived the original idea, collected the data, analyzed the data and interpreted the results, and wrote the paper. All the authors reviewed and approved the final version.

Conflicts of interests. None declared.
Disclaimer. Authors hold sole responsibility for the views expressed in the manuscript, which may not necessarily reflect the opinion or policy of the RPSP/PAJPH and/or PAHO.