We acquired undifferentiated normal human bronchial epithelial (NHBE) cells from Lonza, while murine fibroblasts—3T3-J2—were purchased from Kerafast. We used conditionally reprograming for a continuous expansion of NHBE cells in order to take the number of passages beyond the critical point of senescence (Palechor-Ceron et al., 2013). For such conditional reprogramming, undifferentiated NHBE cells were grown on a bed of mitomycin-treated 3T3-J2 fibroblast cells in the presence of the ROCK inhibitor, Y27632 (Tocris).

For 3T3-J2 cell expansion, we cultured them in 3T3 medium: DMEM (Thermo Fisher Scientific) supplemented with 10% bovine calf serum (Thermo Fisher Scientific) and 1% penicillin-streptomycin (Thermo Fisher Scientific). Cells were incubated at 37°C with 5% CO₂, and 95% relative humidity until confluency reached 70%. Cells were then seeded in new T-175 flasks (Corning) at a density of 3.5 × 10³ cells per cm². After 80 h, cell monolayers were washed with 1X PBS followed by adding fresh 3T3-J2 media supplemented with 2 μg/mL mitomycin C (Sigma Aldrich) to inactivate mitosis. After removing mytomicin-supplemented 3T3 medium, we carefully washed then the cell monolayer with 1X PBS thrice to remove mitomycin excess. Then, cell monolayers were detached from the surface using trypsin (Sigma Aldrich). After trypsin neutralization and cell washing and pelleting by centrifugation, cells were seeded at a density of 2 × 10⁴ cells/cm² in T75 or T175 flasks coated with collagen. After overnight incubation for cell adhesion, the mytomicin-treated 3T3-J2 cells were ready to support the expansion of epithelial cells.

Primary undifferentiated NHBE cells were seeded on mytomicin-treated 3T3-J2 bed in F medium supplemented with the ROCK inhibitor Y-27632 (7.5 μM), as previously described by Liu et al. (2012). Cultures were kept at 37°C with 5% CO₂ and 95% relative humidity. NHBE cells began to displace 3T3-J2 cells and at the same time formed round colonies. Once they reached 70% confluence, we incubated the culture in PBS + EDTA (Dulbecco’s 1X PBS; 100 mM EDTA) for 5 min to detach the 3T3-J2 fibroblasts. Then, NHBE cells were detached using a Trypsin-based commercial kit from Lonza following respective recommendations. Finally, NHBE cells were cryo-preserved in F medium supplemented with 10% FBS, 10% DMSO and 7.5 μM ROCK inhibitor.

For experimental assays, the undifferentiated NHBE cells were cultured and expanded in BEGM medium (Lonza) supplemented with 7.5 μM ROCK inhibitor in T75 flasks for 3 days. Then, NHBE cells were sub-cultured and seeded in 24-well plates at a density of 1.2 × 10⁵ cells/well in the presence of ROCK inhibitor. After overnight incubation, we removed Y27632 by changing the medium. After 24 h incubation in BEGM, cells reached 80% confluency and we used them for viral infection assays.

We evaluated whether temperature may have an impact on the actin cytoskeleton by staining F-actin with phalloidin. Briefly, NHBE cells were subcultured on rounded polylysine-coated glass coverslips (PDL, Neuvitro). At 70% confluency, cell cultures were incubated at different temperatures (4°C, 22°C or 37°C) for 1 h. A different set of cultures were used to explore the effect of a sudden temperature change on the F-actin cytoskeleton. This was conducted by exposing cultures to either 4°C or 22°C, which was then followed by incubation at 37°C for 10 min. After temperature treatment, we fixed NHBE cell cultures with 4% paraformaldehyde for 15 min at 4°C, which was followed by permeabilization with Triton X-100 for 15 min at room temperature. Finally, we incubated each cell culture with Alexa-488-labeled phalloidin (Thermo Fisher Scientific) in order to tag F-actin. We used FluorSafe mounting medium (Calbiochem). The cells were visualized at 63× (NA 1.4) using a confocal microscopy Axio observer Z1. Using Zen-Blue software, multiple Z-plane frames were serially acquired with a separation of 0.5 μm from each other along the Z-axis using the module Z-stack. Then, the images were merged and subjected to signal deconvolution using the module extended depth of focus.

We used an RSV engineered to express BlaM as was reported by De Ávila-Arias et al. (2024). Briefly, we had the sequence encoding the chimera P-BlaM synthesized by Integrated DNA Technologies as gBlocks. P stands for the RSV phosphoprotein, and BlaM corresponded to the optimized beta lactamase version (Y105W) (Wolf et al., 2009). By Gibson cloning (New England Biolabs), we inserted the gBlock at the sequence flanked by AvrII (nt. 2930) and PvuI (nt. 6573), using HC123 full-length RSV cDNA construct as a backbone. HC123 was derived from the green fluorescent protein (GFP) expressing cDNA MP224 (Hallak et al., 2000). The modified HC123 cDNA construct sequence has P-BlaM positioned after the RSV-P gene.

We used BHK cells constitutively expressing T7-RNA polymerase (BHK/T7 cells) to rescue the rgRSV-P-BlaM virus as described. BHK/T7 cells were subcultured in 6-well plates in DMEM-Glutamax medium (Thermo Fisher Scientific) supplemented with 2% FBS (Thermo Fisher Scientific) and 1% Penicillin-Streptomycin (Thermo Fisher Scientific). At 70% confluency, BHK/T7 cells were transfected with P-BlaM RSV cDNA and helper plasmids (Hotard et al., 2012) using Lipofectamine 3000 (Thermo Fisher Scientific)—in the following quantities: 800 ng of modified HC123 cDNA vector, 400 ng of N, P, and M2-1, and 200 ng of L. Transcription of each plasmid was driven by the T7 promoter. Two days later, transfected cells were subcultured in T25 flask (Corning), and syncytia formation was monitored for 48 h, followed by subculturing in a T75 flask (Corning). We proceeded to change medium each 48 h until we observed 50–60% syncytia in the monolayer. Cells were then scraped in a small fraction of the supernatant. Both scraped cells and supernatant were placed in the same Falcon tube, followed by a subsequent centrifugation at 2,141 g for 10 min at 4°C to pellet the cell debris. Magnesium sulfate at 0.1 M was added to stabilize the virus suspension. We snap-froze 1-mL aliquots in dry ice. Each aliquot was stored at −150°C (Panasonic).

We grew recombinant respiratory syncytial virus (RSV) in HEp-2 cell cultures. Briefly, HEp-2 cells were expanded in Opti-MEM^™ medium supplemented with Glutamax (Thermo Fisher Scientific), 10% FBS (Thermo Fisher Scientific) and 1% Penicillin-Streptomycin (Thermo Fisher Scientific) in T75 flasks for 3 days at standard culture conditions of 37°C, 5% CO₂, and 95% relative humidity. Subsequently, cells were transferred to T175 flasks at such cell density to reach 60% confluency the next day. After washing with Ca⁺⁺- and Mg⁺⁺-free PBS, HEp-2 cells were infected with rgRSV at a multiplicity of infection (MOI) of 1 infectious viral particle per 10 cells (MOI of 0.1). The infection proceeded in Opti-MEM (Thermo Fisher Scientific) supplemented with 2% FBS for 48 h while monitoring cytopathic effects and the formation of syncytia. At that time, we changed medium and left the culture proceed for 20 h more. Then, we transferred approximately ¾ of the medium to a 50 mL Falcon tube. We scraped the cells in the medium that was left in each flask and transferred them to the same 50-mL Falcon tube. The cell debris was pelleted by centrifugation at 2,141 g at 4°C for 10 min in a Sorvall Legend Match 1.6 centrifuge. The infectious supernatant was transferred to a new pre-chilled, labeled Falcon tube. We supplemented the supernatant to stabilize the virions with the following compounds at the indicated final concentrations: 0.1 M MgSO₄ (Sigma Aldrich), 0.1% human Albumin (Biotest) and 50 mM HEPES (Gibco). One-mL aliquots of supernatant in cryotubes were snap-frozen in dry ice. Subsequently, these were stored in a −150°C freezer (Panasonic).

We estimated the virus titer by preparing five-fold serial dilutions from stock in HEp-2 cell cultures, which were previously subcultured in 24-well plates. After inoculation for 2 h at 37°C, the inoculum was removed, replaced with fresh medium and incubated for 16 h at 37°C. Trypsin treatment was used to detach cells from the well surface. Using fluorescence-based flow cytometry, we quantified RSV-infected cells. Following Techaarpornkul et al. (2001) recommendations, we calculated the viral titer from that dilution that gave a MOI in the range of 0.05 to 0.1. The formula to calculate such MOI is: (% of infected cells × total number of cells × dilution factor)/(100 × volume of the infection aliquot).

We tested a range of concentrations to determine the neutralization profile of palivizumab (AbbVie). The rgRSV (recombinant green fluorescent protein expressing RSV) virus was incubated with palivizumab for 1 h at 37°C before adding the mixture to NHBE cell monolayer. We tested the following antibody concentrations: 200, 12.5, 0.78, 0.195, and 0.048 μg/mL. These antibody solutions were prepared in BEGM medium. After inoculation, as described above, infection was allowed to proceed for 16 h at 37°C. Then, cell cultures were processed for fluorescence-based flow cytometry. Infected cells were identified by GFP expression.

After washing with HEPES-based saline solution (HBSS), NHBE cells at 80% confluency were infected with rgRSV-P-BlaM at different infectious doses (please see respective results section) suspended in BEBM. The virus adsorption proceeded at either 22°C or 4°C for 1 h to synchronize viral attachment.

After synchronization at the chosen temperature, unbound virions were removed and cells were rinsed with HBSS, and the plates were placed in the incubator at 37°C. At different time intervals (0, 10, 30, 45, 60, 90 and 120 min), we stopped the infection either by using palivizumab (antibody block) or by placing a selected plate at 4°C (temperature block, TB).

When we used BlaM activity on CCF2-AM to identify RSV-infected cells, we stopped the infection by either antibody (200 μg/mL palivizumab) neutralization or TB (4°C) treatment at the respective interval. When the last time interval was reached, we changed the medium on all plates to HBSS supplemented with 2 μM CCF2-AM and 200 μg/mL palivizumab in all culture plates. CCF2-AM (GeneBLAzer in vivo Detection Kit, Thermo Fisher Scientific) was loaded into cells by incubating all cultures for 3 h at 17°C. This approach is known as “time-of-addition BlaM” assay and is the standard procedure as was reviewed by Jones and Padilla-Parra (2016). Finally, cells were prepared for fluorescence-based flow cytometry. CCF2 is a compound made up of a hydroxycoumarin moiety linked by a beta-lactam ring bridge to fluorescein. In the absence of BlaM, the FRET mechanism is present: excitation at 405 nm causes emission at 520 nm. In the presence of BlaM, the FRET mechanism is lost as BlaM cleaves the beta-lactam ring releasing the hydroxycoumarin from the fluorescein: excitation at 405 nm causes emission at 447 nm.

We also ran a TB-chase assay in which after synchronizing rgRSV-P-BlaM adsorption at 4°C for 1 h, we added 200 μg/mL palivizumab to each culture when switching the temperature to 37°C. After each abovementioned time interval, we placed the respective plate at 4°C to prevent virus entry via the endosome. When the last time interval was reached (150 min), cells were processed as above for determining BlaM activity using CCF2-AM.

When we used GFP fluorescence as the readout, we stopped the infection by changing the rgRSV-containing medium to BEGM supplemented with the RSV-neutralizing antibody (200 μg/mL palivizumab) to neutralize active virions adsorbed to the plasma membrane. Then, we allowed the infection to proceed for 16 h before preparing the cells for fluorescence-based flow cytometry, using a BD FACS CANTO II.

NHBE cells were grown on Millicell EZ slides (Millipore). Cells were infected with rgRSV-GFP (MOI = 2), and incubated at 22°C for 2, 5 and 10 h. Unattached virions at the cell membrane were removed using a cold trypsin-EDTA solution (0.95 mg/mL). Then, cells were fixed with 4% paraformaldehyde and permeabilized with 0.05% saponin. Endogenous peroxidase was inactivated by treating the cells with 2% hydrogen peroxide for 45 min.

Cells were incubated for 1 h at room temperature with the following primary antibodies: mouse anti-RSV-N, rabbit anti-EEA1, rabbit anti-RAB5, rabbit anti-RAB7, and rabbit anti-RAB11. Cells were incubated for 1 h with Alexa-488-conjugated goat anti-rabbit antibodies to identify the primary rabbit antibodies targeting the endosomal markers. After washing with 1X PBS to remove the antibody excess, cells were incubated for 1 h with horseradish peroxidase (HRP)-conjugated goat anti-mouse antibody. Following a final wash with 1X PBS, cells were incubated with Alexa 555-labeled tyramide for 10 min, whose amplification buffer was supplemented with hydrogen peroxide. Nuclei were labeled with DAPI (NucBlue, Thermo Scientific). FluorSave (Millipore) mounting media was used for image visualization and long-term preservation.

Cell preparations were imaged using a motorized, inverted Axio-observer Z1 fluorescence microscopy (Zeiss). Images were analyzed using the COLOC-2 plugin of ImageJ software. The colocalization of viral particles with the endosomal markers was determined by the Manders’ coefficient being ≥0.5.

Primary cultures of undifferentiated bronchial epithelial cells grown at 80% confluency in 24-well plates were exposed to different concentrations of drugs that were chosen based on their capability to interfere with cell processes associated to endocytosis. Each drug was administered to the cell cultures before adding the viral inoculum (rgRSV at an MOI of 0.2) and kept in the culture media during infection for 2 h at 37°C. After washing with HBSS to remove unabsorbed virions, a solution of cold trypsin (0.0025 mg%) was added to the cell cultures for 8 min at 4°C to remove unfused virions. Trypsin solution was withdrawn and after HBSS washing, fresh media was added to cell cultures. Infection was allowed to proceed for 16 h at 37°C. GFP positive cells were quantified by flow cytometry.

In order to confirm that positive hits targeted just the early steps conducive to RSV infections, cell cultures grown to 80% confluency were inoculated with rgRSV for 2 h at 37°C before adding the respective positive hit, which was kept in the culture medium for 3 h at 37°C. After withdrawing the respective inhibitor and washing the cell culture with HBSS, fresh media was added back to the cell cultures and the infection was allowed to proceed for 16 h at 37°C. GFP positive cells were quantified by flow cytometry.

The cell cytotoxicity of each of the positive hits was evaluated in primary cultures of undifferentiated bronchial epithelial cells grown at 80% confluency. Each drug was kept in the culture medium for 3 h at 37°C. Then, the medium containing the respective inhibitor was withdrawn and the cell culture was washed with HBSS. Fresh BEGM medium was added to the cell culture for 30 min. Cells were detached from the plate using trypsin resulting in a single cell suspension prepared in binding buffer. Cell suspension was incubated with Alexa-647 labeled annexin V and SYTOX blue, which was immediately analyzed by flow cytometry. Cells were classified as viable (double negative), early apoptosis (positive annexin V), late apoptosis (positive annexin V and SYTOX blue) or dead cells (positive SYTOX blue).

Primary cultures of undifferentiated bronchial epithelial cells grown at 80% confluency in 24-well plates were infected with either rgRSV-P-BlaM or rgRSV to evaluate the role that each drug played during RSV entry or infection, respectively. The infectious aliquot dose was an MOI of 2. Virions were incubated at 22°C for 1 h to synchronize attachment and infection was initiated by placing the plates at 37°C. At different times (0, 10, 30, 45, 60, 90, and 120 min), either the drug or palivizumab was added to the respective plate. The time at which the plate was placed at 37°C was labeled as T0 (time zero).

For virus entry, cells were loaded with CCF2-AM for 3 h at 17°C, and the drug was maintained in the loading buffer. For virus infection, all plates received palivizumab at the end of the assay before allowing the infection to proceed at 37°C for 16 h. Flow cytometry was used to quantify BlaM (+) or GFP (+) cells.

All experiments were performed at least in triplicate and are presented as normalized values with the mean ± standard deviation (SD). For the different assays, an ANOVA was performed to determine differences between treatments. A subsequent one way t-test was used to identify specific treatments showing statistically significant differences. All analyses were conducted using GraphPad Prism (GraphPad Software 9, La Jolla, CA, United States). The level of significance was set at p < 0.05. ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001.

Attaching virus at 4°C for 1 h is one of the most common methods to synchronize virus entry and infection. We used Alexa-488-labeled phalloidin as a probe for F-actin to determine whether incubation at different temperatures affects the cell cytoskeleton in NHBE cell cultures. For illustration, we showed each of the F-actin-associated structures in Supplementary Figure S1.

At 37°C, the actin cytoskeleton of NHBE cells showed stress fibers, cortical actin, and perinuclear actin (Figure 1A). When NHBE cells were incubated at 22°C for 1 h, they showed stress fibers and cortical actin (Figure 1B). However, incubation at 4°C for the same time interval caused a collapse of F-actin structures, leaving only perinuclear actin (Figure 1C).

Ten minutes after shifting from 22°C to 37°C, no changes in the cytoskeleton were observed (Figure 1D vs. Figure 1B). In contrast, 10 min after shifting from 4°C to 37°C, actin rings under the plasma membrane were visible in most cells (Figure 1E). Quantitative evaluation of 10 different fields (20–38 cells per field) showed that 40–60% of all cells showed actin rings under the plasma membrane after warming from 4°C to 37°C. However, such rings were not visible after warming from 22°C to 37°C (Figure 1F). Actin rings are one of the features of membrane blebbing, the outward ballooning of the plasma membrane (Charras et al., 2006; Chikina et al., 2019). Consequently, cells might take up virions as the blebs retract.

We used an engineered RSV as described by De Ávila-Arias et al. (2024), which carries the enzyme beta-lactamase (BlaM) to assess viral entry. This method was previously used to explore HIV entry (Cavrois et al., 2002; Miyauchi et al., 2009). The RSV P gene was fused to BlaM, and the resulting P-BlaM gene was inserted downstream of the P gene in the RSV genome, which also contains the gene encoding GFP (rgRSV-P-BlaM) (Figure 2A).

Undifferentiated bronchial epithelial cells were inoculated with rgRSV-P-BlaM and virus entry was allowed to occur for 2 h at 37°C. The cells were then loaded with CCF2-AM, a fluorochrome with a beta-lactam ring linking a hydroxycoumarin and a fluorescein moiety. The BlaM protein delivered by the virion cleaves the beta-lactam ring, destroying the FRET signal and releasing the fluorescent hydroxycoumarin (Figure 2B), allowing us to identify those cells in which viral entry had occurred. To prevent further virus entry, CCF2 loading and concomitant cleavage by BlaM was performed at 17°C for 3 h (Figure 2B). We titrated rgRSV-P-BlaM, and measured BlaM activity at 3 h or GFP expression at 16 h. The percentage of cells with BlaM activity (73.6%) mirrored the percentage of cells expressing GFP (72.7%) (Figure 2C). Virus titer estimated by beta-lactamase activity, was 6.30 × 10⁵ FFU, while the titer estimated through GFP expression was 1.01 × 10⁶ FFU, as previously reported in De Ávila-Arias et al. (2024).

In general, two different approaches have been used to synchronize viral entry or infection: (1) spinoculation; and (2) attaching the virus to the cell membrane at temperatures not conducive to fusion.

Regarding spinoculation, Guo et al. (2011) cautioned against using spinoculation as a means to synchronize virus binding to cells. Spinoculation itself exerts biophysical and biochemical effects on cells that make them more susceptible to infection. Moreover, Guo et al. (2011) suggested that “some of the spin-induced cellular permissiveness may be beyond the natural capacity of an infecting virus.” In addition Jones et al. (2017), who demonstrated that Dynamin-2 stabilizes the HIV-1 fusion pore at the plasma membrane, avoided spinoculation because this technique induces changes in small GTPase activity that could disrupt actin cytoskeletal regulation with a potential knock-on effect on endocytosis.

Therefore, we followed the general approach of binding virions to the cell membrane at temperatures that either greatly slow or prevent the triggering of the fusion mechanism. Although the standard protocol binds virions to cells at 4°C, there are reports in the literature of synchronizing entry by binding virions at higher temperatures. Herold et al. (2014) synchronized binding of Vpr BlaM-HIV-1 at 22°C to assess HIV-1 entry into primary T cells. In addition, Fuller and Spear (1987) bound HSV-1 at 29°C.

A two-pronged approach made it possible to determine which cell membrane fused with the virus envelope. Virus attachment to the cell surface was carried out by incubation at 4°C or 22°C for 1 h. Then, unbound virions were washed away with HBSS. After warming to 37°C, virus entry was blocked at different times by the addition of palivizumab (AB) or by temperature (TB), placing the cultures at 4°C. Palivizumab neutralizes RSV infection by preventing the triggering of the conformational change from the prefusion to the post-fusion conformation of the RSV-F protein required for exposure of the fusion peptide (McLellan et al., 2011; Zhao et al., 2017). As a result, it prevents fusion with the plasma membrane. The TB protocol would block fusion of the virus envelope with any of the cell membranes (Miyauchi et al., 2009). Cold temperature reduces the lipid membrane fluidity.

If the viral escape rates of the TB protocol and the AB protocol were to overlap, this would indicate that the viral envelope was fusing with the plasma membrane. If the viral escape rate of the TB protocol were to be delayed relative to the AB protocol, this would indicate that the viral envelope is fusing with an endosomal membrane because it would take some time for the virus to reach the endosomal compartment where fusion occurs (Miyauchi et al., 2009).

Palivizumab at the highest concentration tested (200 μg/mL) neutralized 99.7% of rgRSV-P-BlaM virions in suspension (Figure 3A) and 96.6% of cell surface-bound rgRSV-P-BlaM virions (Figure 3B). At this concentration, palivizumab was able to efficiently neutralize an infectious dose of 2 FFU/cell (Figure 3).

A schematic diagram of the protocol to compare the effects of is shown in Figure 4A. When RSV was attached to the cell surface at 4°C, the virus escape rate observed in the TB protocol was lower than that in the AB protocol (Figure 4B), indicating that attachment at 4°C leads to virus uptake by endocytosis. The fraction of internalized virions that were not fused was estimated by subtracting the percentage of cells showing BlaM activity in AB and TB at each time point. The percentage of internalized but unfused virions ranged from 25 to 30% at different times, as evidenced by TB.

When virions were bound at 22°C, the viral escape rates overlapped in both protocols (Figure 4C). This indicated that the cell surface-bound virions had fused their envelope to the plasma membrane. Therefore, the temperature chosen for synchronization influenced whether RSV was taken up by endocytosis or fused to the plasma membrane.

It is possible that RSV may have reached a temperature-arrested intermediate state (TAS) at 22°C, facilitating its viral escape from palivizumab upon warming to 37°C. Using an experimental design similar to the one we describe in Figures 4A–C, De La Vega et al. (2011) found that HIV pseudoviruses acquired resistance to the fusion inhibitor more rapidly than those preincubated at 4°C. However, we found that rgRSV-P-BlaM bound at 4°C escaped much faster than virions adsorbed at 22°C (Figure 4B vs. Figure 4C); consequently, this temperature did not induce TAS for RSV.

Incubation at 22°C might prevent endocytosis in primary cultures of NHBE cells. Consequently, we evaluate general endocytosis processes at 22°C. pHrodo-labeled transferrin and dextran were used to evaluate clathrin-dependent endocytosis and pinocytosis/macropinocytosis, respectively. pHrodo is a fluorescent dye that shows little to no fluorescent signal at neutral pH but fluorescent brightly in an acidic environment such as late endosomes or lysosomes. This allows us to conduct the assays without preincubation at 4°C or the requirement of a negative control. Cell cultures were incubated with the respective tracers while kept at 22°C. The baseline was established by incubating cultures with the tracers at 37°C for 30 min. The 37°C incubation should not be considered a positive control, only a reference point. Both clathrin-dependent endocytosis (Figure 5A) and pinocytosis/macropinocytosis (Figure 5B) remained active at 22°C.

Since there are other endocytic pathways in addition to clathrin-dependent endocytosis or macropinocytosis, we addressed the question of whether cells could internalize virions at 22°C.

RSV endocytosis and endosomal trafficking in cultured cells was assessed after virus adsorption at 22°C. Virions were probed by a mouse monoclonal anti-RSV antibody whose specific binding was detected by Tyramide Signal Amplification (TSA). The endosomes were identified using mouse monoclonal antibodies that recognize early endosomes (EEA1), primary endosomes (Rab5), recycling endosomes (Rab11), and late endosomes (Rab7). RSV virions (red) were detected inside all endosomal compartments (green) (Figure 6A). In Supplementary Figure S2, the corresponding negative controls for the TSA are shown to demonstrate the absence of any false labeling.

According to Heider and Metzner (2014), the ratio of total to infectious viral particles can reach 1,000 total particles per 1 infectious viral particle for RSV. Consequently, the total number of viral particles can range from 10 to 1,000 per infectious particle, and for the purposes of this assay, each cell of the monolayer may have been exposed to between 20 and 2,000 viral particles during the incubation period. Trypsin/EDTA was used to remove bound but unfused virions prior to labeling. A low number of RSV-N-associated signals per cell supports the success of the trypsin/EDTA treatment in removing unfused virions from the cell surface (Figure 6A).

We imaged 320 cells per experimental condition inquiring the colocalization of RSV with the endosomal marker. At 2, 5, and 10 h, the number of RSV-N-associated signals per cell was 1.61 ± 0.47, 1.61 ± 0.36, and 1.68 ± 0.37, respectively. After 2 h incubation at 22°C, RSV-N signal colocalized with the respective endosomal compartment as follows: 57.45 ± 13.61% of the RSV signal colocalized with early endosomes; 36.7 ± 7.84% of the RSV signal colocalized with primary endosomes; 19.44 ± 5.21% of the RSV signal colocalized with recycling endosomes, and 7.09 ± 2.06% of the RSV signal colocalized with late endosomes (Figure 6B and Supplementary Table S1). Overall, cells incubated at 22°C take up virions and maintain endosomal trafficking. We estimated the p-value of Costes included in Coloc-2 to determine whether the colocalization of RSV-N signals with the endosomal markers was fortuitous. If Costes’ p-value is >0.95, then, the probability of a spurious colocalization signal is less than 5%. A set of random images was selected to represent the three time points (2 h, 5 h, and 10 h) for each endosomal marker. We established the RSV-N signals as regions of interest (ROI). The RSV-N associated red signal was randomized using a point spread function (PSF) of 1, following Costes’ significance test. We found that the Costes’ p-value was 0.98 ± 0.02. In addition of the statistical analysis, the following reasons give additional support:

A functional endocytosis assay was performed according to the guidelines of Herold et al. (2014) (Supplementary Figure S3). Palivizumab (200 μg/mL) was used to neutralize the virions that had not reached palivizumab-inaccessible compartments after 2, 5, and 10 h of incubation at 22°C. After the respective incubation time at 22°C, the cell cultures were warmed up to 37°C and the infection was allowed to progress for 18 h. The expression of GFP was used as a reporter for the infection. Palivizumab neutralization was expected to be as effective as trypsin treatment. Although there was a consistent low number of RSV-N signals inside cells, the fraction of GFP-expressing cells increased over time in each of the time points tested. This suggests that virions that were able to reach palivizumab-inaccessible compartments fused their envelopes after warming the cultures to 37°C. Thus, endocytosis was functional at 22°C.

As shown in Figure 4C, BlaM signal was detected in 7 to 10% of the cells that took up CCF-2 at the time of warming to 37°C. Either these virions fused with the plasma membrane even at the suboptimal temperature of 22°C, or endocytosed virions may have fused to deliver BlaM to the cytosol.

Given that there is basal endocytosis at 22°C and that we found virions inside the endosomes, how much do these virions contribute to the percentage of cells showing a positive signal due to BlaM? Hence the pulse-chase design; the virions were adsorbed to the plasma membrane for 1 h at 22°C. At the end of adsorption, all cultures were placed at 37°C. After 10 min, the set of cultures was divided into 2 arms, one following the AB block protocol (Figure 7, red line) and the other following the pulse-chase design (Figure 7, blue line). In the pulse-chase arm, all cultures received palivizumab 10 min after synchronization. A set of 3 cultures from the pulse-chase branch were immediately refrigerated to 4°C. This procedure was then repeated every 10 min for sets of 3 cultures from the pulse-chase arm. Once 2 h has lapsed, all cultures from both protocols were loaded with CCF2-AM at 17°C (Miyauchi et al., 2009).

If endocytosis is the preferred pathway for RSV to deliver its contents to the cytosol, an increase in BlaM signal would be expected at any time after neutralization of cell surface-adsorbed virions with palivizumab during the chase assay. However, there was no increase in BlaM signal even 120 min after warming to 37°C (Figure 7B).

We evaluated several drugs known to interfere with endocytosis for their effects on RSV infection using dose-response inhibition assays. The following drugs targeting different endocytic pathways were used in dose-response assays: Dynasore (dynamin inhibitor), latrunculin A (actin depolymerization), cytochalasin D (actin depolymerization), jasplakinolide (stabilizes preformed actin filaments in vitro and inhibits their disassembly), Y27632 (RhoA inhibitor), NSC23766 (Rac1 inhibitor), Casin (Cdc42 inhibitor), Wiskostatin (N-Wasp inhibitor), LY294002 (PI3K inhibitor), Wortmannin (PI3K inhibitor), Edelfosine (PI-PLC inhibitor), U73122 (PI-PLC inhibitor), EIPA (Na⁺/H⁺ exchanger inhibitor), and Ciliobrevin A (dynein inhibitor).

Each drug was added to cell cultures 1 h before RSV inoculation and maintained during the first 2 h of infection. Infection was allowed to continue for 16 h, and RSV-infected cells were identified by GFP expression and counted. We have summarized the data by the lowest and highest concentrations of treatments that had inhibitory and no inhibitory effects on infection, respectively (Figure 8A). We used a cut-off of 50% reduction in RSV infection to identify hits in our drug screening. Drugs that inhibited infection are grouped to the left of the dotted line. Latrunculin A, jasplakinolide, EIPA, U73122, and edelfosin reduced the infection by at least 50% of that observed in DMSO-exposed cell cultures (Figure 8A). Treatments that failed to inhibit infection are grouped to the right of the dotted line.

To rule out long-term drug-associated effects that interfere with virus replication, each drug was administered after the first 2 h of RSV infection and left in the medium for 3 h. The drug was then removed, and fresh medium was added to the cell cultures. None of the drugs reduced RSV infection by 50% of that observed in DMSO-treated cell cultures (Figure 8B) indicating they do not affect viral replication. In addition, there was no cell cytotoxicity due to late apoptosis or necrosis (Figure 8C).

Although these drugs are known to interfere with endocytosis, this does not necessarily mean that they prevent RSV infection in this way. Therefore, we performed ToA assays to investigate whether these drugs reduce RSV infection by targeting endocytosis. If the plasma membrane is where RSV fuses its membrane, ToA assays would show the same escape kinetics as the AB protocol because palivizumab interferes with viral fusion. On the other hand, if a drug’s ToA shows delayed escape kinetics compared to the AB protocol, it would imply that the drug is preventing the endosomal uptake required for RSV to reach the site where fusion would occur. A schematic representation of the procedure is shown (Figure 9A), which was conducted following the Miyauchi et al. (2009) approach.

We assessed viral escape by BlaM activity or GFP expression. EIPA showed very high fluorescence at the same wavelength as hydroxycoumarin. Therefore, we could not assess BlaM activity when cells were exposed to EIPA. None of the ToA of the drugs showed the same escape kinetics as was seen with the ToA of palivizumab (Figures 9B–D). All the drugs, except EIPA, rapidly lost their ability to prevent RSV escape; at least 50% of the cells showed BlaM activity or GFP expression when the cultures were shifted to 37°C, and they were no longer able to neutralize viral escape 40 min later (Figures 9B–D). In the case of EIPA, 35% of cells showed GFP expression at the time of warming to 37°C. However, the loss of sensitivity to EIPA followed a linear trend (Figure 9D). Palivizumab administered 40 min after temperature shift to 37°C failed to neutralize 50% of infection events as estimated by GFP expression (Figure 9C).

Although some of the drugs would not reach complete inhibitory effect shortly after administration, ToA approach would still be useful since it would serve as an introductory way to determine if the kinetics of RSV escape from these drugs would be similar to that determined by the AB protocol. There was no difference in the kinetic rates suggesting that all of them may be impairing fusion at the cell membrane (Supplementary Table S2; p = 0.305).

Since these drugs were expected to reduce the RSV entry kinetics, we assessed the impact of these drugs on the actin cytoskeleton as a proxy for detecting changes on the actin-dependent plasma membrane morphology. Actin dynamics underlies plasma membrane curvature and tension. We used Alexa-488-tagged phalloidin to detect F-actin. All drugs that impair RSV infection were found to reduce cell membrane deformations, such as filopodia and lamellipodia (Figure 10) which correlates with a reduction in actin availability for the formation of these structures and would be indirectly associated with changes in membrane curvature. Particularly, an increase in the stress fibers was observed in cell cultures exposed to U73122 and edelfosine. As expected, F-actin collapsed when cell cultures were treated with latrunculin or jasplakinolide. Cultures treated with EIPA showed a marked increase in the F-actin distributed along the edges of the cell, suggesting an increase in the cortical actin. When we probed F-actin in cell cultures exposed to drugs that did not block RSV infection, membrane deformations—filopodia and lamellipodia—were preserved. Actually, dynasore-treated cell cultures showed a marked increase in filopodia (Figure 11).