Classification: Virology, Cell biology, Innate immunity
INTRODUCTION
Viruses of the Flaviviridae family comprise four genera: Flavivirus, Hepacivirus, Pestivirus, and Pegivirus [1]. They are characteristically similar in genome structure, virion morphology, and life cycle. They have unsegmented, single-stranded, and positive-sense RNA genomes of 9.6–12.3 kb. The genomes each encode a single polyprotein that is cleaved after translation, through the cooperation of host and viral proteases, into the following structural and nonstructural (NS) proteins: envelope (E), PrM, capsid, NS1, NS2a, NS2b, NS3, NS4a, NS4b, and NS5 [2, 3]. Whereas the structural proteins are components of viral particles, NS proteins are required for viral RNA replication and viral maturation. Flaviviruses are enveloped and contain capsids with icosahedral and spherical geometries.
Dengue virus (DENV) and Zika virus (ZIKV) belong to the genus Flavivirus of the family Flaviviridae. As many as 100 million cases of DENV infections in humans occur annually [4, 5]. ZIKV not only has caused a wide-ranging pandemic in recent years but also is associated with severe diseases, such as microcephaly and other neurological symptoms of maldevelopment of the brain in neonates, and Guillain-Barré syndrome in adults [2, 6–8]. Most cases of infection occur in tropical and subtropical areas, where mosquitoes of the genus Aedes are the major vector for the transmission of DENV and ZIKV [9–11]. Although a wide range of hosts have been recognized to be infected with DENV and ZIKV, the most important part of the life cycle is mosquito transmission among humans, or from nj non-human primates to humans.
DENV enters the mosquito body through a viremic blood meal, after the mosquito bites a DENV-shedding animal [12]. The virus then replicates in the epithelial cells of the mosquito mid-gut. Newly formed virion particles subsequently spread to various mosquito organs. DENV can replicate in multiple types of mosquito cells. For example, DENV can enter hemocytes, in which replicates to only a moderate level. Its replication is limited by multiple innate mechanisms associated with the differences in immune systems among mosquito species. The immune systems of Aedes aegypti and Aedes albopictus mosquitoes support DENV production at a sub-pathogenic level. This interaction between DENV and mosquitoes probably sustains viral production in organs, including the salivary glands, and allows the mosquitoes to live and subsequently transmit DENV to the next host (human or monkey) through subsequent bites [13]. Experimental studies remain lacking regarding whether ZIKV might use the same interaction to result in mosquito production of viruses and infection of subsequent hosts for viral transmission [14]. Therefore, we first investigated the interactions of DENV and ZIKV strains with mosquito cells to reveal the mechanisms underlying how DENV or ZIKV infection induces a sub-pathogenic state in mosquitoes.
Previous studies have shown that infection with DENV and other flaviviruses induces fusion of mosquito cells in cell culture systems [15–20]. Mutated DENV that does not cause cell-cell fusion has a lower replication phenotype, thus indicating that this induced cell fusion is important for viral replication [15, 16]. Whether the different types of DENV and different strains of ZIKV cause different pathogenic effects on mosquito cells remains unknown. Moreover, how cell fusion occurs during infection with DENV or other flaviviruses remains to be experimentally defined. In the present study, we demonstrated that different types and strains of DENV and ZIKV behave differently in inducing cell-cell fusion of C6/36 mosquito cells.
MATERIALS AND METHODS
Cell lines, tissue culture, and viruses
Vero cells (ATCC® CCL-81™) were purchased from the ATCC and maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, penicillin (100 IU/ml)-streptomycin (100 μg/ml), and amphotericin B (2.5 μg/ml) [21]. Aedes albopictus clone C6/36 cells (ATCC® CRL-1660™) were purchased from the ATCC and maintained in RPMI1640 supplemented with 10% fetal calf serum, penicillin (100 IU/ml)-streptomycin (100 μg/ml), and amphotericin B (2.5 μg/ml). L-15 Medium (Leibovitz), purchased from Biological Industries USA (Cromwell, CT, USA), was buffered with free basic amino acids, phosphate buffers, and supplemented with galactose and sodium pyruvate to help maintain physiological pH control. The ZIKV strains MR766 [22], PRVABC59 [23], and DENV-2 (ATCC® VR-1584™) strain New Guinea C; the DENV-1 strain Western Pacific 74 (WP 74) [24]; the DENV-3 strain H87 [25]; and DENV-4 [26] were purchased from the ATCC and propagated in C6/36 cells for fewer than three passages to generate stocks. Other strains of ZIKV and DENV (Table 1) were from BEI Resources. The viral stocks were divided into aliquots and kept at −80°C for use.
Catalog | Serotype/Virus | Strain | Cytopathic effects |
---|---|---|---|
NR-83 | DENV-1 | TH-Sman | Syncytia |
NR-82 | DENV-1 | Hawaii | Syncytia |
NR-3786 | DENV-1 | 228690 (Jamaica) | Syncytia |
NR-49707 | DENV-1 | UIS1162 | Syncytia |
NR-44083 | DENV-1 | DENV-1/VN/BID-V1792/2007 | Syncytia |
NR-84 | DENV-2 | New Guinea C (NGC) | Syncytia |
NR-3795 | DENV-2 | BC141/96 (PR) | Syncytia |
NR-43279 | DENV-2 | US/BID-V1425/1999 | Syncytia |
NR-12220 | DENV-2 | ArA6894 | Syncytia |
NR-49750 | DENV-2 | R06-65-361 | Syncytia |
NR-43282 | DENV-3 | US/BID-V1043/2006 | Congregation |
NR-49753 | DENV-3 | PR06-65-544 | Congregation |
NR-80 | DENV-3 | Philippines/H87/1956 | Congregation |
NR-3798 | DENV-3 | BC14/97 | Syncytia |
NR-49757 | DENV-4 | PR06-65-740 | Syncytia |
NR-49724 | DENV-4 | UIS497 | Congregation |
NR-3803 | DENV-4 | BC123/97 | Congregation |
NR-86 | DENV-4 | H241 (tissue culture adapted) | Syncytia |
NR-48801 | DENV-4 | 703-4 | Syncytia |
ATCC®VR-84™ | ZIKV | MR766 | Congregation |
ATCC®VR-1843™ | ZIKV | PRVABC59 | Congregation |
NR-50066 | ZIKV | IbH30656 (Human/1968/Nigeria) | Congregation |
NR-50220 | ZIKV | H/PAN/2015/CDC-259249 | Congregation |
NR-50279 | ZIKV | MEXI-44 (Mosquito/2016/Mexico) | Congregation |
NR-50183 | ZIKV | FLR (Human/2015/Colombia) | Congregation |
NR-50723 | ZIKV | FSS13025 | Congregation |
Different strains of DENV and ZIKV were used to infect C6/36 cells. The CPE mode, either congregation or cell fusion, is indicated.
Antibodies
Mosquito antibodies were obtained from Bei Resources (MRA-249 Hybridoma 6D12 and MRA-253 Hybridoma 10C8 Anti-Aedes aegypti Salivary Glands). Other antibodies from BEI Resources included the following: 1) anti-DENV-1 envelope protein antibodies (NR-4751 and NR-9549), 2) NR-2556 monoclonal anti-Dengue virus Type 2 Envelope Protein, 3) clone 3H5-1 (produced in vitro), 4) NR-15511 monoclonal anti-Dengue Virus Type 3 Envelope Protein, Clone E2 (produced in vitro), and NR-15540 monoclonal anti-Dengue virus type 4 envelope protein, clone E23 (produced in vitro).
ICC assay
Immunostaining was performed on cells grown on coverslips after fixation with 1% paraformaldehyde (10 min at room temperature) and permeabilization in 0.2% Triton (20 min on ice) by sequential incubation with primary and Texas red-labeled secondary antibodies (Vector Laboratories, Burlingame, Calif.) in PBS for 30 min each. Finally, cells were equilibrated in PBS, stained for DNA with Hoechst 33258 (0.5 μg/ml), and mounted in Fluoromount G (Fisher Scientific, Newark, Del.).
Microscopy
Cells were examined with a Leica TCS SPII confocal laser scanning system. Two or three channels were recorded simultaneously and/or sequentially, and any possible breakthrough between the fluorescein isothiocyanate and Texas Red signals and between the blue and red channels was controlled for.
Cell membrane staining
To determine whether viral infection caused the fusion of C6/36 cells, we stained membranes with a CellMask™ Deep Red Plasma membrane Stain kit (Thermo Fisher, C10046) according to the manufacturer’s protocol. Briefly, the cells were incubated with the staining agent (1:1000 dilution) for 5 min at 37°C. The cells were then fixed with 4% formaldehyde and permeabilized with 0.2% Triton X-100. Then the cells were subjected to ICC assays with an anti-ZIKV antibody (anti-envelope).
RESULTS
Different types and strains of DENV and ZIKV cause various morphologic changes in C6/36 cells
Although DENV and ZIKV show different pathogenesis characteristics in humans, we wondered whether DENVs and ZIKV might cause different cytopathic effects (CPE) on mosquito cells. For that purpose, we selected C6/36 cells generated from Aedes albopictus, a vector for DENV and ZIKV. The C6/36 cells were infected with DENV or ZIKV at an MOI of 0.5, incubated at 28°C under 5% CO2, and photographed every other day under a DIC microscope with a 10× lens. The dynamic images are shown in Fig 1. Among the viruses (DENV-1, -2, -3, -4, and ZIKV), all DENV-1 and -2; most DENV-4; and two strains of DENV-3 infecting C36/6 cells resulted in syncytial morphology. At 0 hours post-infection (hpi), the cells were evenly distributed. The cells began to show fusion after 48 hpi, generating larger cells. At 4 dpi and later, the cell fusion expanded to form more fused cells, and the number of unfused cells clearly decreased. For ZIKV, and some DENV-3 and -4 infections, although the cells started to show congregation at 48 hpi, we did not observe any large cells caused by fusion. At 6 dpi and later, the congregation of cells became clearer; however, no cell fusion was seen.
To determine whether this phenomenon was unique to DENV-1 and -2, we examined multiple strains of DENVs including all four serotypes and seven strains of ZIKV. The strains of viruses are listed in Table 1. All strains of DENV-1 and -2, most strains of DENV-4, and only two strains of DENV-3 caused cell fusion. We tested seven strains of ZIKV, all of which resulted in cell congregation but no cell fusion. Therefore, infection of C6/36 with different DENVs and ZIKV strains caused different morphological CPE under the same cell culture conditions of 28°C and 5% CO2.
Network-like connections among fused cells containing vesicle-like structures in C6/36 cells infected with DENV-1 and -2
We next focused on the infection of DENV-1 and -2 in C6/36 cells. To show more details of the morphological change at the end of viral infection in C6/36 cells, we visualized C6/36 cells infected with DENV or ZIKV for 12 days in a larger microscopy field with a 10× lens. As shown in Fig 2A and 2B, DENV-1 or -2 infection caused cell fusion, and the large fused cells remained connected to other fused large cells, forming a network-like morphology, which was not observed in ZIKV-infected C6/36 cells (Fig 2C). Interestingly, most DENV-3 and DENV-4 and all ZIKV caused cell congregations, which were not connected (only ZIKV infection-caused cell congregations in C6/36 cells are shown).
To demonstrate cell fusion, we stained the membrane with a CellMask™ Deep Red Plasma membrane Stain kit (Thermo Fisher, C10046). On day 12 after infection with DENV-1 (TH-Sman), the cells were incubated with the staining reagent (1:1000 dilution) and DAPI for 5 min at 28°C. Then the cells were washed with PBS and observed under a fluorescence microscope with a 40× lens. As shown in Fig 2D, the fused cells were linked by an extended cell membrane, which was stained red. Therefore, we hypothesized that the network-like morphology formed after DENV-1 or -2 infection caused the fusion of cellular membranes.
To confirm cell fusion after DENV-1 or -2 infection, we sought to demonstrate viral infection in the fused cells. To that end, we performed an immunofluorescence assay (IFA) with anti-DENV antibodies. As shown in Fig 3A, two fused cell clusters were connected by cell membranes. The C6/36 cells expressed viral envelope (Env) protein, shown in red, and NS3 protein, shown in green. We also observed the vesicle-like structures in the DENV-1 infected C6/36 cells (Fig 2A). Our IFA results indicated that the vesicle-like structures were actually empty (Fig 3B): no cells (on the basis of DAPI staining) were observed inside the structures.
To more directly demonstrate the cell fusion or congregation of C6/36 cells after DENV or ZIKV infection, we dynamically recorded DENV-2 (NGC strain), DENV-3 (NR-80), or ZIKV (MR766 strain) infections in C6/36 cells at 28°C under 5% CO2 (movies in Supplemental data), which clearly indicated cell fusion after DENV-2 infection (Movie 1), and cell congregation caused by DENV-3 and ZIKV (Movies 1 and 2). In Movie 1, we recorded the formation of cell fibrilization, thereby resulting in a net-like structure, and vesicle-like structures in DENV-2-infected C6/36 cells. Therefore, we discovered not only that DENV and ZIKV infection caused morphologically different CPE in C6/36 cells, but also that cell fusion of C6/36 induced by DENV-1 or -2 infection yields two types of structures: network-like and vesicle-like structures.
Cell fusion requires viral replication
We next sought to determine whether the cell fusion by DENV-1 or -2 might be independent of viral replication. For that purpose, we inactivated DENV-1 (NGC strain) by UV irradiation. The inactivation was validated by IFA at 24 hours after infection in C6/36 cells, as shown in the lower panel of Fig 4A, whereas the untreated DENV-2 infection was positive for envelope protein (shown in red in the upper panel of Fig 4A). Therefore, the UV treatment of DENV-2 effectively inactivated the virus.
We infected C6/36 cells with either UV-treated DENV-2 or untreated DENV-2, and captured images at 24 or 144 hpi. The inactivated DENV-2 did not cause any CPE, but the untreated DENV-2 caused cell fusion. Our experiments also indicated that higher titers of viral infection caused more rapid cell fusion (not shown). In addition, the cell fusion occurred only after 48 hpi. Therefore, viral replication is required for DENV-2 to cause fusion of C6/36 cells.
Dengue viruses cause syncytial effects on C6/36 cells within a range of pH values in the medium
Our aforementioned experiments were set up at 28°C under 5% CO2 to provide optimal growth conditions for C6/36 cells. The pH of the medium was 7.4–7.6, which is close to the physiological pH of cells. Another flavivirus, St Louis Encephalitis virus, has been found to induce cell fusion in low pH environments [27]. Another study has also indicated that an acidic environment is important for DENV-induced cell fusion [18], which differed from our experimental conditions. To determine whether DENV-2-induced cell fusion of C6/36 cells might be pH dependent, we used L-15 medium, which does not require CO2 to maintain the pH and therefore could be used to culture C6/36 cells without CO2.
We adjusted the pH of the L-15 from 5.0 to 9.0 (Fig 5A). We infected the cells with the same amount of DENV-2 simultaneously and observed the CPE daily under a microscope. As shown in Fig 5B, DENV-2 caused fusion of C6/36 cells at a pH range of 6.0 to 8.0. The optimal fusion pH ranged from 6.5 to 7.5. Interestingly, the highest yield of viral production was detected at pH 7.5. Therefore, we observed a correlation among the pH of the medium, syncytial CPE, and viral yield during DENV-2 infection in the C6/36 cell line.
DISCUSSION
Cell fusion is often used by viruses to spread progeny viral particles from one cell to another. Mosquitoes are the major vector for transmission of DENV and ZIKV among humans and non-human primates. Therefore, understanding the interaction of DENV or ZIKV with mosquito cells is extremely important. Here, we used both cell biological and virological approaches to discern the CPE of the two members of flaviviruses on C6/36 cells, an Aedes albopictus cell line established in 1967 from freshly hatched Aedes albopictus larvae of unspecified ancestry [28]. Together, our results indicated that DENV and ZIKV strains caused two types of CPE resulting in syncytial or congregative effects, and different strains cause different CPE. For example, in C6/36 cells DENV-1, -2, and most DENV-4 strains caused syncytial effects, whereas some DENV-4 strains, most DENV-3 strains, and all ZIKV strains caused congregative effects (Fig 1, Movies 1 and 2). To further support our findings, we tested several strains of each type of DENV and ZIKV (Table 1) selected from different countries and territories.
Whether DENV or ZIKV infection kills C6/36 cells remains arguable. In live mosquitoes, DENV or ZIKV infection should not cause cell death, even if the progeny viral particles have been produced within cells, because the infected mosquitoes remain alive, and able to both replicate and transmit viruses through biting. Indeed, several cell culture studies have revealed that mosquito cells are protected from DENV by the induced antioxidant defense as well as anti-apoptotic effects [29, 30]. Another study has also shown that DENV persistently infects C6/36 cells while allowing the cells to remain intact, and the virus is retained in cells for 20 weeks without CPE [31]. Therefore, these findings from a cell culture perspective may explain why mosquitoes live with the infected viruses, thus making mosquitoes a dangerous vector. However, other studies have shown that DENV causes cell fusion of C6/36 cells [17, 27, 32, 33]. Our current study also indicated the syncytial effects of DENVs on C6/36 cells. The syncytial effect of the virus is a type of CPE, because the consequence of cell fusion is usually cell death. Therefore, some DENV infections may kill infected cells in mosquitoes.
Herein, we undertook the first reported examination of the effects of different strains of DENV and ZIKV on cell fusion in C6/36 cells. All tested DENV-1 and DENV-2 strains caused syncytial effects on C6/36 cells; moreover, some DENV-4 strains, most DENV3 strains, and all ZIKV strains caused cell congregative effects, but no cell fusion was seen. Our results suggest that DENV may cause two distinct types of CPE: cell fusion and cell congregation. These results suggest that DENVs may have two modes of replication and spread within live mosquitoes: cell fusion and “release and entry” infection.
Early experimental results have suggested that cell fusion of C6/36 cells might be mediated by E protein, because anti-E antibody blocks virus-induced cell fusion [18]. However, anti-E antibody prevents the entry, infection, and replication of the viruses. Our experimental results showed that the UV-inactivated DENV did not cause cell fusion. Thus, the de novo produced viral proteins are important for DENV-caused cell fusion. We also found that the pH environment of the medium affects DENV-caused cell fusion. The viral replication assays demonstrated that cell fusion may facilitate viral production.
Our study has several limitations. First, because cell culture studies differ from in vivo experiments, many observations of DENV or ZIKV infection in mosquitoes cannot be completely explained by cell culture study results. An in vivo study of DENV or ZIKV infection in mosquitoes is needed to clarify whether different strains of DENV or ZIKV might have different replication mechanisms and routes of spread in mosquitoes. In addition, our results showed that different strains of DENV-3 and -4 caused different CPE in mosquito cells. However, the sample size might have been statistically insufficient to conclude which CPE is predominant in DENV-infected cells. We plan to analyze the genomic sequences to determine whether genomic consensus sequences might be associated with the DENV-induced CPE in mosquito cells.