S-Phase-Dependent Enhancement of Dengue Virus 2 Replication in Mosquito Cells, but Not in Human Cells

Cells and viruses.

Cells were maintained in medium supplemented with 10% fetal calf serum, 10 mM HEPES, 100 U/ml penicillin, and 100 μg/ml streptomycin (complete medium). C6/36 Aedes albopictus cells (American Type Culture Collection [ATCC]) and BHK21 clone 15 cells (a gift of R. Childs, Davis, Calif.) were grown in L15 and α-minimal essential medium, respectively. Huh-7 cells (a gift of M. G. Katze, Seattle, Wash.) and primary human foreskin fibroblasts (HFF) and 293T cells (both gifts of D. A. Galloway, Seattle, Wash.) were maintained in Dulbecco's modified Eagle's medium. Mammalian cells were grown at 37°C in 5% CO₂, and mosquito cells were grown at 28°C without added CO₂. The prototype Thai DEN2 strain, 16681 (passage number unknown), was obtained from the Centers for Disease Control and Prevention, Ft. Collins, Colo. Isolation of Nicaraguan and Thai DEN2 strains has been described elsewhere (4, 47). Briefly, both viruses were isolated by seeding patient serum samples onto C6/36 cells, and viral stocks were produced in C6/36 cells as previously described (4). The Nicaraguan strain (N1042; a gift of A. Balmaseda, Managua, Nicaragua) was employed after three passages in mosquito cells, except for stocks produced in human cells (see below), and the Thai strain (C0477; a gift of R. Rico-Hesse, San Antonio, Tex.), was used after four passages. Virus stock titers were determined via plaque assay using BHK21 cells as previously described (17), and stocks were stored at −80°C. For production of DEN2 in human cells, monolayers of 293T human embryonic kidney cells were inoculated with N1042, previously passaged twice in mosquito cells. Cells were infected in Dulbecco's modified Eagle's medium supplemented with 2% fetal bovine serum for 2 h, and serum concentration was then brought up to 10%. The medium was collected and clarified by centrifugation at 2,000 × g at days 3, 4, and 5 postinfection, and virus titers were determined by plaque assay using BHK cells. Day 5 supernatants were used for subsequent infections.

Cell infection.

Duplicate wells of cells were used for each infection. Infection of C6/36 mosquito cells was carried out at a multiplicity of infection (MOI) of 0.5 or 1 as indicated. Huh-7 cells and HFF were infected at an MOI of 1. Cell number at the time of infection was estimated based on the number of cells plated, the plating efficiency, and the doubling time, which we have calculated for C6/36 cells, Huh-7 cells, and HFF. Cells were exposed to virus in growth medium containing 2% fetal bovine serum for 2 h. Virus was removed, and cells were washed five times with complete medium. Cells were then incubated in complete medium under the indicated conditions for manipulation of cell cycle status (see below).

Cell cycle manipulation.

For comparison of DEN2 titers from density-arrested and cycling mosquito cells, 6 × 10⁵ C6/36 cells were plated in 12-well plates and grown for 2 days. To ensure that cells were viable and not nutrient deprived at the time of infection, the growth medium was replaced the day after plating. Cells were infected with DEN2 N1042 the subsequent day, as described in “Cell infection” above. At this point, cells were greater than 90% confluent. After infection, cells were either maintained at confluence or were dislodged from the plate by pipetting. Cells recovered from one well were diluted into 6 new wells of a 12-well plate.

For analysis of DEN replication in S-phase cells, asynchronously cycling C6/36 cells, Huh-7 cells, or HFF in 12- or 24-well plates were infected with DEN2 N1042, C0477, or 16681 as described above in “Cell infection,” followed by incubation with complete medium containing either phosphate-buffered saline (PBS), 2 mM thymidine (THY; Sigma, St. Louis, MO), or 0.2 mM hydroxyurea (HU; Sigma). To ensure that cells remained subconfluent for the duration of the experiment, 1.5 × 10⁵ C6/36 cells or 4 × 10⁴ Huh-7 cells or HFF were seeded into 12-well plates for initial experiments. The human cells are significantly larger than C6/36 cells; hence, a smaller number were seeded to prevent them from reaching confluence. Subsequently, experiments were scaled down to 24-well plates (by a factor of 2) to conserve the limited stocks of low-passage DEN2. Cells were infected the day after plating unless otherwise noted. After the indicated time for all cell cycle experiments, the medium was collected for plaque assay analysis and cells were removed from the plate with trypsin (Invitrogen, Carlsbad, CA) for assessment of cell number via a hemocytometer and for flow cytometric analysis of DEN envelope (E) expression and cell cycle position (see below). Viability of C6/36 cells, Huh-7 cells, and HFF was confirmed by trypan blue exclusion.

Flow cytometry.

To assess cell cycle distribution, cells were fixed in ice-cold 70% ethanol, treated with 100 μg/ml RNase A (Sigma) for 20 min at 37°C, and stained with 50 μg/ml propidium iodide (Sigma) in PBS. DNA content was measured by flow cytometry using a Beckman-Coulter EPICS XL cytometer (Beckman-Coulter, Inc., Fullerton, CA) and Flow Jo software (Tree Star, Inc., Ashland, OR). G₀/G₁-phase, S-phase, and G₂/M-phase populations were delineated using the cell cycle feature of Flow Jo. Profiles of HU-treated C6/36 cells could not be fit to a model using the cell cycle feature and were therefore measured using gates drawn manually in Flow Jo.

Analysis of intracellular DEN2 envelope (E) protein expression was carried out as previously described (17), with some modifications. Briefly, mock- or DEN2-infected cells were removed from the plate by trypsinization, washed in PBS, and fixed for 10 min at room temperature with 4% paraformaldehyde (Fisher Scientific, Pittsburg, PA). Cells were permeabilized in HHSN (Hanks balanced salt solution, 10 mM HEPES, 0.1% saponin, and 0.02% sodium azide) for 15 min at 37°C and blocked in 10% fetal calf serum at 37°C for 30 min. Labeling of DEN E protein was carried out using the supernatant fraction from the 3H5 mouse hybridoma (ATCC) diluted 1:4 in HHSN plus 1% bovine serum albumin for 1 h at room temperature. Cells were washed and then incubated for 30 min at room temperature with Alexa 488-conjugated anti-mouse immunoglobulin G (IgG) (Molecular Probes, Eugene, OR) diluted 1:1,000 in HHSN-1% BSA. Cells were postfixed in 0.5% paraformaldelyde, and E expression was measured by flow cytometry and analysis with Flow Jo software.

Titers of cell-associated virus.

C6/36 cells infected with DEN2 N1042 as described above were harvested at 40 h postinfection (hpi). This time point was chosen to allow accumulation of progeny virions, particularly of the low-passage isolates, but to avoid multiple complete rounds of virus infection. The medium was collected for determination of extracellular virus titers. To obtain intracellular virus, cells were washed once in PBS, incubated for 3 min on ice with an alkaline/high-salt solution of 1 M NaCl plus 50 mM Na bicarbonate, pH 9.5, to remove surface-bound virus (18), washed two more times in PBS, and then lysed by freeze-thawing. Growth medium containing extracellular virus was also subjected to one cycle of freeze-thawing. Virus titers in the cell lysates and growth medium were measured by plaque assay.

Transfection of infectious DEN2 RNA.

Full-length DEN2 16681 RNA was generated by in vitro transcription as previously described (21) from a cDNA clone, pD2/IC-30P (39) (a gift of R. Kinney, Fort Collins, CO). Briefly, the cDNA construct was linearized with XbaI (New England Biolabs, Beverly, MA), and 1 μg was used for transcription. Transcription reactions were carried out using a Ribomax T7 in vitro transcription kit (Promega, Madison, WI) as per the manufacturer's instructions. Reactions were treated with RQ DNase I (Promega), and the RNA was passed through Nucaway spin columns (Ambion, Inc., Austin, TX) prior to transfection. RNA was transfected into C6/36 cells using Lipofectamine 2000 (Invitrogen). Each 50-μl reaction mixture was used for transfecting 6 wells of asynchronously cycling C6/36 cells (in 24-well plates). To ensure that treatment with THY did not impact transfection efficiency, cells were transfected in the absence of THY and then washed and treated with complete medium containing PBS or 2 mM THY at 3 h posttransfection. Viral supernatants were collected for plaquing, and cells were collected for propidium iodide staining and flow cytometry at 40 h posttransfection.

Reporter assays.

The DEN untranslated region (UTR) reporter construct has been previously described and contains the firefly luciferase (luc) gene flanked by the 5′ and 3′ UTRs of DEN2 (strain C0477) (21, 34). A reporter construct representing a cellular housekeeping gene, actin, was constructed by replacing the β-globin 5′ UTR in the 5′Bg-LUC-3′V268-A60 construct (34) with the 5′ UTR of actin. The actin 5′ UTR was generated using four overlapping oligonucleotides, Actin 1, 5′-CGCGTAAAATTTAATACGACTCACTATAACCGCCGAGACCGCGTCCGCCCCGCGA-3′; Actin 2, 5′-GCACAGAGCCTCGCCTTTGCCGATCCGCCGCCCGTCCACACCCGCCGCCAGCTCAC-3′; Actin 3, 5′-TCTGTGCTCGCGGGGCGGACGCGGTCTCGGCGGTTATAGTGAGTCGTATTAAATTTTA-3′; and Actin 4, 5′-CATGGTGAGCTGGCGGCGGGTGTGGACGGGCGGCGGATCGGCAAAGGCGAGGC-3′, that were annealed and then ligated into pGL3-5′Bg-LUC-3′V268-A60 that had been digested with MluI and NcoI (New England Biolabs). DEN and actin reporter mRNAs were transcribed from XbaI- and AseI-digested templates, respectively, using the Ribomax T7 in vitro transcription kit per the manufacturer's instructions. C6/36 cells were plated in 24-well plates and transfected the next day with 0.3 μg of reporter mRNA with Lipofectamine 2000 transfection reagent according to the manufacturer's instructions. After 3 h, the cells were washed and treated with complete medium containing either PBS or 2 mM THY. Cells were harvested in 1× cell culture lysis reagent (Promega) for luciferase assays, per the manufacturer's instructions, at 26 h posttransfection. This time point was determined empirically to detect luciferase activity before it plateaued. In parallel, RNA was harvested in TRIzol (Invitrogen) for assessment of reporter RNA levels by real-time reverse transcription-PCR (RT-PCR) as described previously (34).

Western blot assays. (i) Anti-NS1 blot assays.

Levels of DEN proteins were assessed in single-cycle replication assays. C6/36 cells were infected with DEN2 N1042 at an MOI of 1 and treated with THY as described above. At 24 hpi, prior to significant levels of virus secretion and spread to neighboring cells, cells were washed with PBS, detached from the plate with 3 mM EDTA, washed again in PBS, and counted. Approximately 2.5 × 10⁵ cells were pelleted and lysed in 50 mM Tris Cl, pH 8.0, 250 mM NaCl, 0.1% sodium dodecyl sulfate (SDS), 1% NP-40, 1 mM dithiothreitol, and protease inhibitor cocktail on ice for 15 min. Cellular debris was pelleted by microcentrifugation at 15,000 × g for 10 min at 4°C. Supernatants were boiled in Laemmli sample buffer, resolved by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to polyvinylidene difluoride membranes. Blots were probed with a cocktail of mouse hybridoma supernatants against DEN2 NS1 (clones 8H7, 3A5, and 2E9; P. R. Beatty and E. Harris, unpublished data) followed by rabbit-anti-mouse (IgG plus IgM)-horseradish peroxidase (HRP) conjugate (Jackson Immunochemicals, West Grove, PA) or with goat-anti-actin-HRP conjugate (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). NS1 was chosen because it is not contained in the incoming viral particles and, thus, indicates viral replication. Bands were visualized by Supersignal West Dura ECL substrate (Pierce, Rockford, IL) and exposure to Kodak BioMax XAR film (Eastman Kodak Co., Rochester, NY). For some experiments, the centrifugation step was excluded, and this had no impact on the levels of viral protein observed (data not shown).

(ii) Anti-E blot assays.

Infected C6/36 cells were washed at 24 hpi with an alkaline/high-salt solution to remove any surface-bound input virus, followed by two washes in PBS. Cells were lysed in buffer containing 20 mM HEPES, pH 7.4, 150 mM NaCl, 1% Triton X-100 (Fisher), 10% glycerol, and 1 mM EDTA. The samples, which were not boiled, were resolved by SDS-PAGE under nonreducing conditions. Blotting and probing were carried out as described for NS1, but with 3H5 hybridoma supernatant against DEN2 E protein as the primary antibody. Chemiluminescent bands were detected in the linear range using a Bio-Rad Chemidoc XRS imager with Quantity One software (Bio-Rad Laboratories, Hercules, CA).

Real-time RT-PCR. (i) Intracellular DEN2 RNA.

DEN RNA levels were assessed in single-cycle infection assays. C6/36 cells were infected with DEN2 N1042 (MOI = 0.5) and treated with PBS or THY as described above. At 24 hpi, prior to significant levels of virus secretion and spread to neighboring cells, cells were washed in PBS, treated with alkaline/high salt as described above, and washed two more times in PBS. Total RNA was harvested in TRIzol, purified, and analyzed by real-time RT-PCR using the Superscript III Platinum One-Step qRT-PCR system (Invitrogen) with primers to DEN2 NS1 and to actin (the internal positive control) (Invitrogen). Primer sequences were as follows: 6-carboxyfluorescein (FAM) fluorophore-labeled NS1 forward primer, 5′-cacaaCCATGAAGAGGGCATTTG-FAM-G-3′; unlabeled NS1 reverse primer, 5′-TTTGTTTCCACATCAGATTCCCA-3′. Nucleotides indicated in lowercase letters are non-DEN sequence. Reactions were carried out in duplicate on an ABI Prism 7700 sequence detection system (Applied Biosystems, Foster City, CA). A four-point standard curve was generated using DEN2 RNA isolated from a known quantity of virus as previously described (60), using SDS 1.9.1 software (Applied Biosystems) for the analysis. The amount of total DEN2 RNA from THY-treated cells was determined relative to that from PBS-treated cells.

(ii) Extracellular DEN2 RNA.

C6/36 cells were infected with DEN2 N1042 (MOI = 0.5). At 2 hpi, virus was removed and, in addition to washes in complete medium, cells were subjected to an alkaline/high-salt wash before treatment with PBS or THY. At 40 hpi the growth medium was collected for analysis of virus titers by plaque assay and for viral RNA by real-time RT-PCR as described above (except that actin primers were not used). Extracellular virus titers and RNA levels were displayed relative to those of the PBS-treated controls.

Analysis of purified virions.

Infected C6/36 cell supernatants were harvested at 40 hpi and clarified at 1,000 × g for 3 min at 4°C. Cells were harvested for Western blot analysis as described above. Virus was pelleted at 64,000 × g for 2 h at 4°C in an SW60Ti rotor. Pellets were resuspended in 20 mM Tris-Cl, pH 7.4, 150 mM NaCl, and 1 mM EDTA (TNE) and loaded onto 15-to-55% sucrose-TNE (wt/wt) gradients, which were centrifuged in an SW60Ti rotor for 17 h at 144,000 × g. The bottom of the tube was punctured with a needle to collect 700-μl fractions. Aliquots were taken for measurement of PFU/ml by plaque assay, and the remaining fractions were trichloroacetic acid precipitated and resolved by nonreducing SDS-PAGE as described above for analysis of E protein.

Statistical analysis.

Calculations of the means, standard deviations, P values, and 95% confidence intervals were carried out using Microsoft Excel (Microsoft, Redmond, WA). P values were determined by Student's t test.

DEN2 infection does not alter the cell cycle status of mosquito cells, but cell cycle status influences virus titer.

To determine if the growth state of mosquito cells influences DEN2 titers, C6/36 Aedes albopictus cells were grown to greater than or equal to 90% confluence but maintained in fresh growth medium to induce a G₀/G₁ arrest via contact inhibition rather than by nutrient deprivation. Cells were then infected with DEN2 strain N1042, a clinical isolate from Nicaragua with a short tissue culture passage history. At 2 hpi, cells were washed extensively and were either maintained in density arrest with the addition of fresh growth medium or released into the cell cycle by replating at a lower density. To allow enough time for one viral replication cycle but fewer than two complete replication cycles, the experiment was harvested at 40 hpi. Cultures were trypsinized to ensure collection of all of the cells for accurate counting. Cells were washed in PBS, counted via hemocytometer, and analyzed by flow cytometry for intracellular expression of DEN2 E protein and for cell cycle status. In parallel, the growth medium was collected, centrifuged to remove cellular debris, and analyzed by plaque assay for virus titer. As comparisons were between arrested and cycling cell populations, measurements of virus output were adjusted for cell number.

Greater than half of the density-arrested and cycling populations expressed E protein by the time of harvest (data not shown); however, DEN2 replication had no impact on C6/36 cell cycle distribution compared to mock-infected cells (Fig. 1A, compare left and right panels of each set). Flow cytometric analysis of propidium iodide-stained cells demonstrated that, regardless of DEN2 infection, nearly 80% of the density-arrested population were in G₀/G₁-phase (Fig. 1A, top panels), with approximately 6 to 7% and 10% of cells in S-phase and G₂/M-phase, respectively. In contrast, less than half of the cycling populations were in G₀/G₁ phase, while approximately 33 to 35% and 14 to 15% were in S-phase and G₂/M-phases, respectively (Fig. 1A, bottom panels). Overall production of low-passage DEN2 N1042 varied somewhat among experiments; however, adjustment of virus titer for population size revealed a consistent threefold increase (P < 0.01, 95% CI = 2.0 to 4.4) in virus output from cycling C6/36 cells compared to density-arrested cells (Fig. 1B). Moreover, this increase was also apparent when titers were adjusted for the number of cells expressing E protein, with a fourfold difference in virus output from cycling cells (P < 0.001) (data not shown). These data indicate that the growth state of C6/36 cells can modulate DEN2 titers but that DEN2 appears not to alter the cell cycle of C6/36 cells under our experimental conditions.

DEN2 titers are increased in S-phase mosquito cells.

The previous results demonstrated that cycling C6/36 populations produced higher DEN2 titers per cell than did density-arrested C6/36 populations (Fig. 1). Given that the largest change in cell cycle position between cycling and density-arrested cultures was the percentage of cells in S-phase (Fig. 1A, compare top and bottom panels), it followed that the effect on DEN2 titers may occur during S-phase. Cycling mosquito cells spend 4 to 6 h in S-phase during each round of the cell cycle (3, 27) (data not shown). If virus production were enhanced during S-phase, then a cell that is held in S-phase over the course of the experiment should produce more virus than an actively cycling cell that spends several shorter intervals in S-phase over the course of the experiment. This prediction was tested by infecting asynchronously cycling C6/36 cells with DEN2, followed by imposition of a THY-induced S-phase block. Treatment with excess THY is commonly employed to synchronize mammalian cells in S-phase (10, 11); however, the efficacy of THY in arresting mosquito cells was unknown. Therefore, a dose-response experiment was carried out with C6/36 cells, and all THY doses tested were effective (1 to 10 mM THY) (data not shown). THY was used at a 2 mM final concentration for subsequent experiments. Based on trypan blue exclusion and synchronization and release experiments, this dose did not impact cell viability over the duration of the experiment (data not shown).

To determine the influence of S-phase on DEN2 titers, asynchronously cycling C6/36 cells were infected with DEN2 strain N1042, with a low-passage DEN2 Thai isolate (C0477), or with a high-passage DEN2 strain (16681), washed at 2 hpi, and then treated with PBS or THY. The low-passage isolates were predicted to be more sensitive to cellular growth state than a highly tissue culture-passaged virus whose replication is less restricted in cell culture (17, 21). Cells and medium were harvested at 40 hpi for measurement of cell number, cell cycle distribution, and virus titers and for assessment of E protein expression (see below). Flow cytometric analysis of propidium iodide-stained cells revealed that approximately 35% of PBS-treated cells were in S-phase at the time of harvest, while over 60% of THY-treated cells were stalled in S-phase (Fig. 2A, top and bottom panels, respectively). In response to the THY-imposed S-phase block, titers of DEN2 N1042 were increased by approximately 27-fold (P < 0.001) and DEN2 C0477 increased by 29-fold (P < 0.0001), after accounting for cell number (Fig. 2B). Normalization of DEN2 N1042 and C0477 titers to the number of cells expressing E protein also revealed over a log increase in virus output from THY-treated C6/36 cells (data not shown). The response of the more highly tissue culture-adapted DEN2 strain, 16681, to arrest of C6/36 cells in S-phase was less robust, with a ninefold increase, on average, in titers relative to the PBS-treated control (P < 0.05) (Fig. 2B). Finally, similar increases in virus production were observed when virus stocks were produced in the human cell line 293T, indicating that the results were not dependent upon the cellular source of DEN2 stocks (data not shown).

It was possible that the increased DEN2 titers in Fig. 2B were due to more efficient egress from S-phase cells, rather than an increase in virus production. To address this question, titers of both intracellular and extracellular virus were measured from DEN2 N1042-infected C6/36 cells treated with either PBS or THY. At 40 hpi, the growth medium was collected for quantifying extracellular virus titers, and the cells were harvested for measuring intracellular virus titers. To remove surface-bound virions, the cells were washed and then incubated in an alkaline/high-salt solution (18, 21). Cells were then lysed by freeze-thawing. Extracellular virus was also subjected to a cycle of freeze-thawing. Titers of intracellular and extracellular DEN2 from THY-treated cultures were both increased by approximately 30-fold relative to titers from the PBS-treated control cells (data not shown). These data indicate that increased DEN2 titers during S-phase were due to increased virus production, not to more efficient viral egress from S-phase cells.

To address the possibility that the increased virus titers were related to unanticipated cell cycle-independent effects of THY treatment, a different reagent was employed to stall C6/36 cells in S-phase following infection with DEN2 strain N1042. Treatment of C6/36 cells with 0.2 mM HU, previously shown to synchronize mosquito cells in S-phase (26), resulted in an accumulation of cells in early S-phase (Fig. 3A) and a concomitant increase in titers of progeny virions (Fig. 3B). The increase in titers ranged from 10- to 70-fold, with an average of 35-fold, and was statistically significant (P < 0.02).

DEN2 production is not enhanced in S-phase human cells.

Experiments were also carried out with human cells to determine the effect of cell cycle on DEN2 titers in the other natural DEN host. Asynchronously cycling Huh-7 human hepatoma cells were infected with DEN2 strain N1042, C0477, or 16681 at an MOI of 1. Hepatoma cells were employed because the liver is thought to be a site of DEN replication (6) and liver damage has been observed in association with severe dengue (50). At 2 hpi, cells were washed and then incubated with complete medium containing PBS or 2 mM THY. The experiment was harvested at 30 hpi to allow enough time for one complete round of viral replication but not for two complete rounds and because virus titers from the PBS-treated control Huh-7 cells were similar at this time point to titers from the PBS-treated C6/36 cells in the previous experiment (Fig. 2B). Cells were collected via trypsinization for assessment of cell number by hemocytometer and for flow cytometric analysis of DNA content and intracellular E protein expression. In parallel, virus titers in the growth medium were determined by plaque assay. Analysis of DNA content revealed that the cell cycle distribution of PBS-treated Huh-7 cells (Fig. 2C, top panels) was similar to that of C6/36 cells (Fig. 2A, top panels). In addition, a similar accumulation of cells in S-phase resulted following THY treatment (60 to 70% of the population) (Fig. 2C, bottom panels, and A). In contrast to results in C6/36 cells, S-phase arrest of Huh-7 cells did not result in increased titers of any of the DEN2 strains (Fig. 2D). To determine if this lack of effect was due to the time of harvest, additional experiments were harvested at 20 and 45 hpi and also revealed no increase of DEN2 titers in S-phase hepatoma cells (data not shown). Moreover, no S-phase-dependent increase in titers occurred in an alternative cell type, primary human fibroblasts, infected with DEN2 strain 16681(data not shown). Thus, the S-phase-dependent enhancement of DEN2 production in mosquito cells was not recapitulated in either of the human cell types tested.

To compare the percentage of cells infected in PBS- and THY-treated mosquito and human cell cultures, flow cytometric analysis of E protein expression was performed with cells from the above experiments. DEN2 N1042- or C0477-infected C6/36 cells treated with THY exhibited an approximately twofold increase in the number of E-positive cells relative to the infected, PBS-treated controls at 40 hpi (Fig. 4A). There was a smaller, non-statistically significant increase in the percentage of E-positive C6/36 cells in 16681 infections (Fig. 4A). Similar experiments with human cells revealed no difference in the percentage of cells expressing DEN E protein (Fig. 4B and data not shown). Because cells were treated with THY after infection was initiated, no difference in the number of infected cells between PBS- and THY-treated populations was expected. Two potential explanations for the increased number of E-positive C6/36 cells observed following THY treatment were that (i) THY-treated cells were more permissive for an early step of infection, such as genome uncoating or initial translation, or that (ii) virus was released earlier from the THY-treated cells than from the PBS-treated cells, with DEN undergoing translation in the newly infected neighboring cells.

The DEN2 replication cycle is accelerated in S-phase mosquito cells.

To determine if the DEN2 replication cycle is accelerated in S-phase mosquito cells, as suggested by the twofold increase in E-expressing C6/36 cells at 40 hpi in THY-treated cultures (Fig. 4A), C6/36 cells were infected with DEN2 N1042 or C0477 and treated with PBS or THY as in the previous experiments. Medium was collected at various hpi, and virus titers were measured by plaque assay. For both DEN2 strains, progeny virus was first detectable by 28 hpi in THY-treated cultures but was not consistently detectable until after 31 hpi in PBS-treated cultures (Fig. 5). Titers of DEN2 N1042 and C0477 in THY-treated cultures in the representative experiments shown were 1.4 × 10⁵ and 1.3 × 10⁵ PFU/ml, respectively, at the last time point. In comparison, titers of N1042 and C0477 from PBS control cultures were approximately 8 × 10³ and 6 × 10³ PFU/ml, respectively (Fig. 5).

To further test the hypothesis that the greater number of E-expressing cells in THY-treated cultures observed at 40 hpi resulted from more rapid virus production and spread, C6/36 cells were infected with DEN2 N1042 (MOI = 1) and treated with PBS or 2 mM THY as in previous experiments, and the number of E-expressing cells was determined in single-cycle infections harvested at 24 hpi. This time point was chosen because it is late during the first cycle of virus replication, but before detectable viral egress (Fig. 5). Flow cytometric analysis revealed little difference in the percentage of E-expressing cells at 24 hpi, with a fold increase of 1.18 ± 0.15 following THY treatment, relative to PBS-treated cells. Thus, our results demonstrate an increased rate of virus replication in S-phase mosquito cells that is not due to an increase in the number of cells initially infected with DEN2.

Early events in DEN2 replication are not responsible for increased titers during S-phase.

Because infection of C6/36 cells in the previous experiments was initiated prior to THY treatment, the increased DEN2 titers were not likely due to increased virus adsorption. To formally test whether increased virus titers were due to enhanced entry or genome uncoating, these processes were bypassed by transfection of full-length, positive-strand DEN2 RNA in vitro transcribed from pD2/IC-30P, a cDNA construct of strain 16681 (43). Infectious clones of DEN2 N1042 and C0477 are not yet available. Asynchronously cycling C6/36 cells were transfected, washed at 3 h posttransfection, and treated with PBS or THY. Growth medium was collected at 40 h posttransfection, and virus titers were assessed by plaque assay. Cell cycle status was determined in parallel to confirm the THY-induced S-phase arrest (data not shown). DEN2 titers from the THY-treated transfected cells were about 11-fold higher than from PBS-treated cells (Fig. 6), similar to the DEN2 16681 results when infection was initiated with intact virus (Fig. 2B). Similar results were obtained when experiments were analyzed at an earlier time posttransfection (27 h [data not shown]). These results demonstrate that the S-phase-responsive step of DEN2 replication follows genome uncoating.

S-phase responsiveness of DEN2 is not due to enhanced viral translation.

The efficiency of DEN2 translation has been shown to regulate the extent of virus replication, with translation efficiency being dependent upon both virus strain and cell type (20, 21). However, initial results suggested that DEN2 translation was not substantially enhanced during S-phase. Measurements of mean fluorescence intensity (MFI) of DEN2-infected, E-positive C6/36 cells were carried out with the flow cytometric data shown in Fig. 4A, and the MFI of THY-treated cells was, on average, less than twofold greater than that of PBS-treated cells at 40 hpi (fold increases for DEN2 N1042, C0477, and 16681 were 1.7 ± 0.4, 1.6 ± 0.2, and 1.3 ± 0.1, respectively). Based on the time of viral egress from C6/36 cells (Fig. 5), viral proteins detected at 40 hpi were likely derived from both the initial round of DEN infection as well as initiation of a second round of infection, particularly in THY-treated cultures, which exhibited more rapid production of viral progeny. Therefore, to examine viral protein expression in single-cycle infections, C6/36 cells were infected with DEN2 N1042 (MOI = 1), treated with PBS or THY as in previous experiments, and harvested at 24 hpi. Flow cytometric analysis demonstrated that the MFI of E-positive PBS- and THY-treated cells was similar, with a fold increase following THY treatment of 1.24 ± 0.20, relative to PBS-treated controls. Similar results were obtained by Western blot analysis of E (Fig. 7A) and NS1 (Fig. 7B), the latter being present only after the incoming virus undergoes translation (48). Thus, while differences in virus titers from PBS- and THY-treated cultures were apparent at the earliest times of detectable viral egress (Fig. 5), levels of viral protein appeared to be similar.

As a final means of testing whether DEN2 translation is increased in S-phase C6/36 cells, the translational activity of the regulatory viral 5′ and 3′ UTRs was measured using RNA-based UTR-reporter gene chimeras (Fig. 7C, top right panel) (21, 34). Asynchronously cycling C6/36 cells were transfected with in vitro-transcribed reporter mRNA, washed after 3 h, and treated with PBS or THY. Cells were harvested for measurement of luciferase activity, for real-time RT-PCR analysis of reporter RNA levels, and for cell cycle analysis at 26 h posttransfection, when luciferase expression from our reporter constructs in C6/36 cells was still increasing (data not shown). Cells arrested in S-phase in response to treatment with THY (Fig. 7C, left panel). However, luciferase activity from neither the DEN2 reporter nor a control reporter containing the β-actin 5′ UTR and a polyadenylated tail changed in response to THY treatment (Fig. 7C, bottom right panel). Together, these results indicate that the S-phase-responsive step of DEN2 replication follows viral translation.

Viral RNA levels do not account for increased DEN2 titers during S-phase.

Real-time RT-PCR was carried out on total RNA isolated from N1042-infected or PBS- or THY-treated C6/36 cells in single-cycle infections (24 hpi) to determine if the THY-induced increase in DEN2 titers was due to higher levels of viral RNA, resulting either from increased RNA synthesis or from increased RNA stability. Primers to amplify both positive- and negative-strand RNA were directed at the NS1 region of the genome. The overall level of viral RNA was approximately twofold higher in THY-treated C6/36 cultures compared to PBS-treated cultures (Fig. 8). This result suggests that the level of viral RNA in S-phase mosquito cells may contribute to, but does not fully account for, the increase in DEN2 titers.

Viral assembly is enhanced in S-phase mosquito cells.

Given our results indicating that increased production of DEN2 from S-phase C6/36 cells is not attributable to early steps of infection (Fig. 6), to enhanced viral translation (Fig. 7), or to higher intracellular levels of viral RNA (Fig. 8), we hypothesized that increased production of DEN2 could result from an S-phase-dependent enhancement in virion assembly and/or infectivity. To test this hypothesis, we compared the amount of viral RNA and E protein secreted into the growth medium to the amount of PFU released. If viral assembly is facilitated in S-phase mosquito cells, then higher levels of E protein and viral RNA should be released into the growth medium in THY-treated cultures. Alternatively, if virions produced in S-phase cells are more infectious, due either to more efficient virion maturation or to other factors influencing infectivity, then levels of E protein and viral RNA released into the medium should be similar in PBS- and THY-treated cultures.

To compare the amount of viral RNA released to PFU, C6/36 cells were infected with DEN2 N1042 (MOI = 0.5) for 2 h, followed by stringent washing, which included incubation with an alkaline/high-salt solution to remove residual surface-bound particles (18). Cells were then treated with PBS or THY. Clarified viral supernatants were generated at 40 hpi, a time point chosen to facilitate analysis of secreted virions. DEN2 RNA in the supernatants was measured by real-time RT-PCR using primers recognizing the NS1 region. In parallel, virus titers in the medium were measured by plaque assay. Extracellular levels of viral RNA were approximately 10-fold higher, and titers were 20-fold higher, in the THY-treated cultures compared to PBS-treated cultures (Fig. 9A). Relative to PBS-treated cells, THY-treated cells released approximately 10-fold-higher levels of DEN RNA and 20-fold-higher virus titers (Fig. 9A). This result suggests an enhancement of DEN2 assembly in S-phase mosquito cells. However, based on the discrepancy between increases in RNA level and increases in PFU, the data do not eliminate the possibility that there is a small enhancement in overall virion infectivity.

To compare the amount of E protein and PFU released into the growth medium, C6/36 cells were infected with DEN2 N1042, washed, and treated as described for the previous experiment. Virions were concentrated from clarified supernatants and then purified via sucrose gradient centrifugation. Fractions were collected from the bottom of the tube and analyzed by plaque assay for infectious titers and by quantitative Western blot for levels E protein. Consistent with an earlier study of DEN assembly in mosquito cells (61), E protein was detected only in gradient fractions containing infectious particles (data not shown). In contrast to the similar intracellular levels of E observed in PBS- and THY-treated cultures, levels of extracellular E increased in parallel with virus titers following THY treatment, with a 40-fold increase in titers and an approximately 30-fold increase in E protein compared to that from PBS-treated cultures (Fig. 9A). Shown are the gradient fractions containing peak virus titers (60% of the total virus from either PBS- or THY-treated cultures). However, similar results were obtained when all fractions were included in the measurements and when virus was not purified over sucrose gradients before analysis (data not shown). These results indicate that more efficient virion assembly accounts for the majority of increased virus production in S-phase mosquito cells.