Experimental Infection with and Maintenance of Cell Fusing Agent Virus (Flavivirus) in Aedes aegypti

Abstract.

During the past two decades, there has been a dramatic increase in the recognition and characterization of novel insect-specific flaviviruses (ISFVs). Some of these agents are closely related to important mosquito-borne flavivirus pathogens. Results of experimental studies suggest that mosquitoes and mosquito cell cultures infected with some ISFVs are refractory to superinfection with related flavivirus pathogens; and it has been proposed that ISFVs potentially could be used to alter the vector competence of mosquitoes and reduce transmission of specific flavivirus pathogens, such as dengue, West Nile, or Zika viruses. In order for an ISFV to be used in such a control strategy, the virus would have to be vertically transmitted at a high rate in the target vector population to insure its continued maintenance. This study compared the vertical transmission rates of an ISFV, cell fusing agent virus (CFAV), in two Aedes aegypti colonies: one naturally infected with CFAV and the other experimentally infected but previously free of the virus. CFAV filial infection rates in progeny of female mosquitoes from both colonies were > 90% after two generations of selection, indicating the feasibility of introducing an ISFV into a mosquito population. This and other considerations for evaluating the feasibility of using ISFVs as an arbovirus control strategy are discussed.

INTRODUCTION

Cell fusing agent virus (CFAV) is the prototype of the insect-specific flaviviruses (ISFVs). CFAV was first isolated by Stollar and Thomas¹ who observed that culture fluid from the Peleg line of Aedes aegypti cells,² induced marked cytopathic effect with syncytium formation, when it was inoculated into a culture of Singh’s Aedes albopictus cells.³ Later studies^4,5 demonstrated that this new agent, designated the “cell fusing agent,” was a novel member of the genus Flavivirus, family Flaviridae. Subsequently, CFAV has been detected in or isolated from field-collected mosquitoes worldwide, predominantly Ae. aegypti.^6–12

The term “insect specific” refers to viruses that naturally infect hematophagous diptera (usually mosquitoes) and that replicate in mosquito cells in vitro, but do not replicate in vertebrate cells or infect humans or other vertebrates.¹³ This is in contrast to the classical arthropod-borne viruses of vertebrates (arboviruses) that are maintained principally, or to an important extent, through biological transmission between susceptible vertebrate hosts by hematophagous arthropods.¹⁴ The arboviruses are dual host (vertebrate and arthropod) viruses, whereas the insect-specific viruses appear to involve only a single host (hematophagous insects).

With advances in molecular tools for virus detection and the growing interest in mosquito microbiomes, there has been a recent explosion in the detection and description of new insect-specific viruses.¹³ As of November 2016, a total of 38 ISFVs have been reported.^13,15–19 Undoubtedly, this number will continue to grow as more metagenomics analyses are done on mosquitoes and other hematophagous insects.

In view of the limited host range of the ISFVs, compared with the vertebrate pathogenic flaviviruses, one obvious question is, “how are the ISFVs maintained in nature?” The most likely mechanism is vertical transmission in their respective insect hosts. This hypothesis is supported by reports of the isolation of or detection of six ISFVs (CFAV, Aedes flavivirus, Calbertado, Culex flavivirus [CxFV], Kamiti River, and Spanish Ochlerotatus viruses) from adult male mosquitoes or from immature forms (eggs, larvae, or pupae).²⁰ In addition, there is convincing evidence from both field and laboratory studies^21,22 of vertical transmission of CxFV by Culex pipiens mosquitoes.

We recently described an established laboratory colony of Ae. aegypti that was persistently infected with CFAV.¹¹ Subsequently, a series of experiments were carried out to 1) determine the efficiency of vertical transmission of CFAV in the naturally infected colony and 2) determine if we could introduce CFAV into female Ae. aegypti from a colony free of the virus, and if the experimentally infected females would also vertically transmit the virus to their offspring. This report describes our results.

MATERIALS AND METHODS

Mosquitoes.

Aedes aegypti from two established laboratory colonies (Galveston and Bangkok), maintained at the University of Texas Medical Branch, were used in these studies. The progenitors of the Galveston colony were collected as eggs and larvae in Galveston, Texas, in 2003 and have been maintained in continuous colony in our laboratory since that time.²³ In 2012 while testing all of our laboratory mosquito colonies for the presence of mosquito-specific viruses, it was observed that the Galveston colony was persistently infected with CFAV.¹¹ At the time, eight geographic strains of Ae. aegypti and Aedes albopictus were maintained in our insectary; but only the Galveston strain was found to be infected with the CFAV. The progenitors of the Bangkok Ae. aegypti colony were obtained from Bangkok, Thailand, in 2011; this colony is free of CFAV, as determined by culture in C6/36 cells, next-generation sequencing, and transmission electron microscopy.¹¹

Mosquitoes were reared in an insectary, maintained at 27°C with 80% relative humidity and a 16-hour light/8-hour dark photoperiod, as described previously.²⁴ Larvae were fed on Wardley shrimp pellets (Hartz Mountain Corp., Secaucus, NJ); adults were provided with cotton balls saturated with 30% sucrose solution.

Virus.

The strain of CFAV used in our experimental infections was originally isolated in a culture of C6/36 cells inoculated with a pool of 50 homogenized female Ae. aegypti from the Galveston colony.¹¹ The NCBI Reference sequence for the isolate is NC-001564. The sample used to experimentally infect the Ae. aegypti females (Bangkok colony) had been passaged three times in cultures of C6/36 cells.

Infection of mosquitoes.

Approximately 100 female mosquitoes of the Bangkok Ae. aegypti colony were inoculated intrathoracically²⁴ with approximately 0.15 uL of a C6/36 stock of CFAV (titer unknown). Infected mosquitoes were held in a 30.5 cm³ screened cage (BioQuip Produces, Gardena, CA) within a plastic glove box at 27°C and maintained on 30% sucrose solution. Four days after infection, mosquitoes were fed defibrinated sheep blood, using a Hemotek membrane feeding system (Discovery Workshops, Accrington, United Kingdom), as per manufacturer’s instructions. After feeding, approximately 12 blood-engorged females were removed from the cage and transferred into 48-mL individual polystyrene snap-cap plastic containers (Fisher Scientific, Pittsburgh, PA) with fine nylon netting on top and a moistened strip of paper toweling inside for oviposition. Cotton balls saturated with 30% sucrose solution were placed on the top of each vial. Four or 5 days later when eggs appeared on the moist toweling, the female parent was removed and frozen at −80°C for subsequent testing.

Female mosquitoes from the persistently infected Galveston Ae. aegypti colony were not inoculated with virus, but were simply fed on defibrinated sheep blood. After feeding, 12 engorged females were confined individually to plastic oviposition containers, as described earlier. When eggs appeared on the paper strip, the female was removed from the vial and frozen for subsequent testing. Egg papers were allowed to dry and were held at 27°C until the parent had been tested for CFAV infection.

Work with the Galveston and Bangkok colonies was done 2 months apart to avoid potential cross-contamination.

Rearing of F₁ and F₂ generation mosquitoes.

After the Galveston and Bangkok parent females had oviposited and have been tested for CFAV infection, egg papers from selected females positive for reverse transcription–polymerase chain reaction (RT-PCR) were hatched in deoxygenated water. Offspring from each female parent were reared in separate larval pans at 27°C. Pupae from each family were removed and placed in a 50-mL beaker with water, which was confined within a separate 473-mL cylindrical cardboard carton covered with fine mesh.

When the F₁ generation adults emerged, within 24 hours and before mating occurred, 10 or 20 males and females from each family were collected for virus assay. The remaining F₁ males and females in the family were left in the cage for 5–7 days to allow mating and subsequent blood feeding. After feeding, 12 engorged F₁ females were again confined individually in oviposition containers to obtain eggs of the F₂ generation; then the rearing and testing process described earlier was repeated. In this manner, the offspring from two consecutive generations of the naturally infected Galveston colony and two generations of the experimentally infected Bangkok colony mosquitoes were tested, and the CFAV filial infection rates were compared.

Virus assay of mosquitoes.

Mosquitoes (parent as well as F₁ and F₂ adults) were thawed and placed individually into 1.5-mL Eppendorf tubes containing 600 uL of phosphate-buffered saline, pH 7.4, with 10% fetal bovine serum. Subsequently, each insect was homogenized with a 3-mm steel ball, using a TissueLyser (Qiagen, Hilden, Germany). Total RNA was extracted using Trizol reagent (Ambien; Fisher Scientific) and RNeasy Mini Kit (Qiagen), following the manufacturer’s instructions.

The detection of CFAV in extracted RNA from the insects was assessed by RT-PCR assay, using CFAV-specific primers designed to amplify a segment of the E gene encoded by the viral RNA. The forward and reverse primers used for CFAV were 5′AATGAGACCTGTTCGCTTAG-3′ and 5′CGTTTGTCAATCAAGGCAG-3′, respectively. Amplification was performed with each oligonucleotide primer at a final concentration of 0.5 μM with 1.0 U of avian myeloblastosis virus (AMV) reverse transcriptase (Roche, Mannheim, Germany), which contained 1.5 mM MgCl₂ and 0.2 mM (each) deoxynucleoside triphosphates in a final reaction volume of 50 μL. Thermocycling conditions for the first round of amplification were 50°C for 30 seconds and 94°C for 4 minutes, followed by 38 cycles of denaturation (94°C, 30 seconds), annealing (55°C, 30 seconds), and extension (68°C, 1 minute), and a final extension at 72°C for 7 minutes. The amplification product was visualized in a 2% agarose-1× TAE (tris-acetate-ethylenediaminetetraacetic acid) gel by ethidium bromide staining and UV transillumination. The expected size for amplification (PCR product) was of ∼340 bp (Figure 1).

Figure 1.

Molecular detection of cell fusing agent virus partial sequences of E gene for envelope protein in Aedes aegypti from Galveston colony (L: 50 pb DNA ladder; C+: positive control; C−: negative control); lanes 1–4: female; 5–10: male F₂ offspring from parents.

RESULTS

Galveston colony.

Figure 2 shows the CFAV filial infection rates among Ae. aegypti taken from the persistently infected Galveston colony. Initially, 12 females from the colony (parental or F₀ generation) were blood fed, confined for oviposition, and their eggs were collected. When tested after oviposition, 12 of the 12 F₀ females (100%) were RT-PCR positive for CFAV. The F₁ generation eggs from three of these CFAV-positive female parents (designated A, B, and C) were hatched and 10 male and 10 female offspring of each parent were tested. F₁ filial infection rates for the parents were parent A, 8/20 (40%); parent B, 19/20 (95%); and parent C, 20/20 (100%) (Figure 2).

Figure 2.

Diagram showing selection method and cell fusing agent virus infection rates in three generations of Aedes aegypti mosquitoes from the naturally infected Galveston colony.

Seven additional F₁ female offspring from F₀ parent B were blood fed, confined, and their eggs (F₂ generation) were collected. Eggs from two of the seven RT-PCR positive F₁ parent females (designated S and T) were hatched, and a sample of their F₂ generation adult offspring was tested for CFAV. A total of 24 F₂ offspring from F₁ parents S and T, were tested; and all 24 of their F₂ offspring were CFAV-positive (Figure 2). These results confirmed that CFAV was maintained by vertical transmission in the Galveston Ae. aegypti colony.

Bangkok colony.

Four days after inoculation, F₀ female mosquitoes from the CFAV-free Bangkok colony were blood fed and confined individually in containers for oviposition. Six of these F₀ generation parents were subsequently tested for CFAV and five of the six were positive. Eggs from three of the infected F₀ parents (designated parents E, F, and G) were hatched and 10 of each of their F₁ offspring were tested for CFAV. The filial infection rates in their F₁ progeny were parent E, 5/10 or 50%; parent F, 4/10 or 40%; and parent G, 0/10 or none infected (Figure 3).

Figure 3.

Diagram showing selection method and cell fusing agent virus infection rates in three generations of Aedes aegypti from the experimentally infected Bangkok colony.

Accordingly, the remaining F₁ adult female offspring of parent F were blood fed and confirmed for oviposition. After eggs were laid, 10 of these F₁ females were tested for CFAV infection, but only two of the 10 F₁ females (20%), designated parents X and Y, were infected. The eggs from F₁ parents X and Y were hatched, reared to adults, and 35 or their F₂ offspring were tested. Twelve of 20 F₂ offspring (60%) from parent X were positive for CFAV; and 14 of 15 offspring (93.3%) from parent Y were infected. At this point the experiment was terminated.

DISCUSSION

The original objectives of this project were 2-fold: 1) to determine the efficiency of vertical transmission of CFAV in a naturally infected colony of Ae. aegypti (Galveston strain) and 2) to test the feasibility of introducing CFAV into another Ae. aegypti colony (Bangkok strain) that was free of the virus. Recent reports^25–29 indicate that mosquitoes and mosquito cell cultures infected with some ISFVs are refractory to superinfection with a related flavivirus pathogen. It has been suggested that if one could introduce such an ISFV into a mosquito vector population, it might be possible to alter the insects’ vector competence and thus reduce transmission of specific flavivirus pathogens.^13,27–29 Thus answers to the two questions (objectives) posed earlier are important in evaluating the feasibility of such an arbovirus control strategy.

The level of vertical transmission of CFAV in the naturally infected Galveston Ae. aegypti colony was high as shown in Figure 2. All (12/12) of the F₀ parent females from the colony were infected. The overall CFAV filial infection rate among 60 F₁ offspring from three of these parents (A, B, and C, respectively) was 78.3%. In the F2 generation, the CFAV filial rate was 100% (20/20). These results are similar to those reported by Saiyasombat and others²² with a colony of Cx. pipiens mosquitoes naturally infected with CxFV. The CxFV filial infection rate among F₁ generation progeny of naturally infected, wild caught Cx. pipiens was 97.4%.²² These very high filial infection rates indicate that CFAV, CxFV, and possibly other ISFVs are maintained in their mosquito hosts by vertical transmission. This heritable transmission model (or stabilized infection) has also been demonstrated with Sigma virus (Rhabdoviridae) in Drosophila melanogaster and for La Crosse with (Bunyaviridae) in Aedes triseriatus.²⁰

Our second objective, namely to test the feasibility of introducing CFAV into an Ae. aegypti colony (Bangkok) that was free of the virus, was successful; but F₁ filial infection rates were relatively low and some selection was needed to obtain higher CFAV filial infection rates. In the F₁ generation progeny of the experimentally infected (inoculated) Bangkok F₀ generation females, the filial infection rate was 30% (12/40) (Figure 3). In contrast, the CFAV filial infection rate in F₂ progeny from two CFAV-positive F₁ females was 90% (27/30), suggesting that with further selection, one could obtain females with a “stabilized infection,” as observed in the Galveston colony.

The lower CFAV filial infection rate (30.0%) in the F₁ generation progeny of the experimentally infected (inoculated) Ae. aegypti females from the Bangkok colony was comparable to other results reported by Saiyasombat and others²² in their experiments with CxFV in Cx. pipiens. The latter studies also found much lower CxFV filial transmission rates in F₁ offspring of needle-infected Cx. pipiens, than in the naturally infected mosquitoes.²² However, only the F₁ generation progeny were tested in this study, and no attempt was made to select for higher rates. So it may be feasible to develop mosquito colonies (lines) that are persistently infected with ISFVs.

The availability of ISFV-infected mosquito lines would allow in-depth studies of the effect of ISFV infection on the vector competence of mosquitoes for specific flavivirus pathogens. There has been considerable speculation that ISFVs can alter a mosquito’s vector competence for certain mosquito-borne flavivirus pathogens, due to heterologous interference or by altering the insect’s basal innate immunity.^13,24–29 However, to date most of the experimental studies to test this hypothesis have been done in vitro, using the C6/36 mosquito cell line. But the C6/36 cell line has a dysfunctional antiviral RNA interference response and obviously does not have the full innate immune response of a live insect.^30–32 The availability of ISFV-infected mosquito lines would also permit more realistic proof-of-concept studies of whether ISFVs could be used for control of certain mosquito-borne flavivirus pathogens.

Another important consideration in evaluating the feasibility of an ISFV control strategy is the mosquito host range of a candidate ISFV. Susceptibility of a mosquito to ISFV infection does not necessarily mean that the virus will be vertically transmitted in the insect.²⁰ But vertical transmission is essential, if the candidate ISFV is going to persist in the target mosquito population. Thus CxFV would not qualify as a candidate control agent for Dengue virus 1-4 or Zika virus, since it is probably not vertically transmitted in their Aedes vectors. Likewise, CFAV could not be used as a control agent for West Nile virus (WNV), since it is unlikely to be vertically transmitted in the Culex vectors of WNV. The ability of an ISFV to be vertically transmitted and its filial infection rate can only be determined by in vivo studies in the targeted vector species. The concept of using ISFVs to alter the vector competence of mosquitoes for selected flavivirus pathogens is valid and has potential, but much more information is needed before its feasibility can be determined.