Fig 1.
Inference of intercellular spread of DENV by using a 1% methylcellulose-containing overlay.
A-C) In the presence of 1% methylcellulose, a few infections appeared in the monolayer of C6/36 cells at 24 hpi while gradually enlarged aggregates of infected cells were observed at 48 and 72 hpi, respectively. D-F) Scattered infected cells are shown in the monolayer of C6/36 cells cultured in the medium only even when the virus was inoculated for 72 h. green: cells positive to antibodies against viral E protein; blue: cells stained with DAPI.
Fig 2.
Cell-to-cell transmission of DENV in C6/36 cells demonstrated in transwell and co-culture systems.
A) Diagram showing transwell (0.4 μm pore size) or co-culture systems used to investigate virus spread between donor (D) and recipient (R) cells, either with or without the presence of neutralizing antiserum (NeuAb). For more detailed information, please see the description in the Materials and Methods. Cells were observed with an immunofluorescence assay (IFA) after treatment with respective antibodies. B) Mock-infection donor cells were used as controls while recipient cells were transfected with rhodamine fluorescent protein or RFP (red) prior to the investigation. In the transwell system, donor cells infected by DENV for 24 h at MOI = 1 were identified by polyclonal antibodies against the viral NS3 protein (green). Meanwhile, some recipient cells were actually infected by the virus, indicating the possibility of diffuse movement by viruses released by donor cells (white arrow). As no infection occurred in recipient cells in the presence of NeuAb, it seemed that intercellular virus spread may have been blocked by adding NeuAb. Results from co-cultures revealed that a number of recipient cells were infected and thus positive to NS3 antibodies even NeuAb has been added to the culture. This suggested an intercellular spreading mode by DENV, most likely by cell-to-cell transmission. C) Recipient cells sorted by flow cytometry were collected for viral RNA analysis by reverse transcriptase (RT)-PCR. No viral RBA can be detected in mock-infected cells (mock). A small amount of viral RNA was detected in cells collected from the transwell system without NeuAb (transwell), but not in the presence of antiserum (transwell+antiserum). However, a large amount of viral RNA was amplified in the co-culture group even in the presence of NeuAb (co-culture+antiserum), further implying the possibility of cell-to-cell transmission of DENV. The RNA level of 18S was used as the internal control in this experiment.
Fig 3.
Quantitative assessment for cell-to-cell transmission of DENV.
A) Transwell and co-culture systems were further used to quantify cell-to-cell transmission of DENV; the study design followed that shown in the Fig 2A, except for the replacement of RFP by eGFP to label recipient cells. In the experiment, eGFP-transfected recipient cells from both systems at 2 or 3 dpi were gated for detection of viral NS3 protein, which was mostly detected in recipient cells collected from the co-culture system. B) The infection rate of recipient cells from the transwell system was significantly lower than the co-culture system at either time points; i.e., 2 or 3 dpi (Student’s t-test; p<0.01). C) Viral NS3 was not shown in cell with mock-infection while a proportion of recipient cells were infected without treatment of NeuAb in the transwell system. Nevertheless, no infected recipient cells appeared in the same system treated with NeuAb, indicating antiserum may restrict the virus diffusion movement. More, recipient cells in the co-culture system in the presence of NeuAb, was still easily infected by the virus due to a large proportion of cells containing viral proteins can be detected. D) Among the two groups containing infected recipient cells, the proportion of cells positive to viral NS3 in the NeuAb-free transwell system (41.40%) was significantly lower (63.02%) than in the co-culture system in the presence of antiserum (Student’s t-test, p<0.05). Mock infection was used as the negative control in this experiment. It seemed that the intercellular spread of the DENV via cell-to-cell transmission was more efficient than diffuse movement of cell-free viruses.
Fig 4.
Upregulation of tetraspanin C189 and its association with cell-to-cell transmission in C6/36 cells.
A) At the protein level, endogenous C189 was up-regulated at 18 hpi and persisted until 48 hpi in C6/36 cells infected by DENV, consistent with our previous detection of C189 at the RNA level. Actin was used as the internal control in the experiment. B) Under confocal microscopy, C189 appeared as red spots in response to infection by DENV. It was largely co-localized with viral E protein (green) in the cytoplasm of infected cells, implying that they have an intimate association in C6/36 cells infected by DENV.
Fig 5.
The role of C189 in cell-to-cell transmission of DENV.
A) Real-time (RT)-PCR C6/36 cells untransfected or transfected with an unrelated sequence (miN) both increased the expression of C189 in response to DENV infection. On the other hand, its expression level was significantly reduced when C189 was knocked down by a microRNA-based knockdown system (miC189) (Student’s t-test, p<0.01). B) Based on the overlay of 1% methylcellulose, C189 knockdown restricted the aggregation of infected cells that were either untransfected or transfected with an unrelated sequence. When a recovery system using a C189-overexpressing plasmid was applied to miC189 cells (miC189/C189), an increase in infected cell aggregates appeared. C) The average diameter of aggregated cells in the C189 knockdown group (miC189) was 170 μm, which was significantly smaller than controls, either C6/36 cells (290 μm) or miN (280 μm), and even the recovered group (270 μm) (Student’s t-test, p<0.05). This indicates that C189 is essential during cell-to-cell transmission of DENV in C6/36 cells.
Fig 6.
C189-containing membrane-bound vacuoles were formed to deliver DENV via cell-to-cell transmission.
A) Confocal microscopy revealed that membrane-bound vacuoles containing C189 (C189-VCs) (green) were formed in DENV-infected C6/36 cells overexpressing eGFP-tagged C189, within which viral E protein (red) or the virus were contained. B) An ultrastructural photograph showing that vacuoles containing numerous virions (arrow) were usually formed in the cytoplasm of C6/36 cells infected by DENV, usually at 24 hpi. C) Immunocytochemical study using colloidal gold particles to label C189 showing vacuoles containing maturing virions (thick arrow) were peripherally positive to the staining of C189 antibodies (thin arrow). This indicated that C189-VCs might be where viral proteins and even virus particles are stored. D) Further sucrose gradient density ultracentrifugation to separate intracellular components from cell lysates of DENV-infected C6/36 cells transfected with HA-tagged plasmid containing the C189 insert (HA-tagged C189) revealed that envelope (E) and capsid (C) proteins of the virus were both detected simultaneously with C189 (represented by positive to antibodies against HA) in fractions containing higher concentrations of sucrose. However, no viral proteins were detected in cell lysate of controls that were transfected with eGFP only. E) Immunoprecipitation (IP) performed with constructed HA-tagged C189 showed that viral proteins (E, NS3, and C) did not directly interact with C189, although a weak band (*) appeared at the position below the E protein. It seemed that C189 in C189-VCs formed in DENV-infected C6/36 cells in response to DENV infection may not have a direct interaction although they coexist spatially in the cell.
Fig 7.
Intercellular translocation of C189-containing membrane-bound vacuoles containing DENV in C6/36 cells.
A) In culture for C189eGFP-expressing cells infected by DENV2, viral E protein (red and arrow) co-localizes with C189 (green and arrow) at 24 hpi, which is frequently seen in filopodia extended by donor cells. B) For further clarification, C189eGFP-expressing donor cells infected by DENV were co-cultured with RFP-expressing recipient cells in the presence of neutralizing antibodies (NeuAb) for 6 h. It turns out that viral E protein and C189 were still delivered simultaneously from the donor cell to the recipient cell (arrow) via filopodia (arrow head). D: the donor cell; R: the recipient cell. Scare bar = 10 μm.