HAdV-D26 displays prolonged intracellular trafficking, resides in different cellular compartments during the first hour of internalization, and exhibits aberrant nuclear DNA delivery compared to HAdV-C5 in epithelial cells

The only study to date addressing the intracellular trafficking of HAdV-D26, indicates that this virus accumulates in the late endosomal compartment [6]. To investigate this behavior in more detail we performed confocal microscopy and transmission electron microscopy (TEM) analysis of HAdV-D26 intracellular trafficking in A549 cells. In contrast to well-characterized intracellular trafficking of HAdV-C5, which typically reaches the nucleus within approximately 60 min in vitro (S1A–S1C Fig), most HAdV-D26 particles remained dispersed throughout the cytoplasm 60 min after transduction. Even at 240 min p.i. (post-infection), there is no significant accumulation of HAdV-D26 at the nucleus, and most viral particles appear grouped on one side in the wider perinuclear area (Fig 1A–1C). To determine whether HAdV-D26 is free in the cytosol or associated with membrane-bound organelles, we investigated its intracellular trafficking in A549 cells 60 min after internalization using TEM. HAdV-D26 was observed at the plasma membrane, in several distinct cellular compartments including endosomes and degradative compartments, as well as free in the cytosol (Fig 1D). Quantitative analysis of viral localization revealed that the majority of the HAdV-D26 particles were found at the cell membrane, within membranous compartments and in the cytosol, with nearly 40% localized in the degradative compartment. ‘Degradative compartment’ is here used as a collective term that could include autolysosomes, late endosomes, lysosomes, amphisomes, and degradative autophagic vacuoles [14]. At the same time more than 60% of HAdV-C5 is found free in the cytosol (Figs 1E and S1D). These results suggest that after internalization, a distinct intracellular trafficking pattern of HAdV-D26 in A549 cells may contribute to its lower transduction efficiency compared to HAdV-C5 [4].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Intracellular trafficking of HAdV-D26 in epithelial cells.

(A) Intracellular trafficking of HAdV-D26 in A549 cells observed by confocal microscopy. Cells were incubated on ice with fluorescently labeled HAdV-D26 (green; MOI 10⁵ vp/cell) for 45 min, then 10 min at 37°C, followed by washing with fresh medium to remove unbound viruses. After additional incubation at 37°C for the indicated times, cells were fixed and stained with phalloidin-AF555 (red) and DAPI (blue). Representative confocal images are shown (scale bar = 10 μm). The lower panels show areas selected from the upper panels at larger magnification, with marked edges of the nucleus (blue, based on DAPI) and cell (red, based on phalloidin). (B) Kinetics of HAdV-D26 intracellular trafficking shown in (A). Intracellular localization of individual virions was quantified using the Relative Distance Index (Weighted). An index value of 0 represents particle position at the nucleus, while a value of 1 represents the plasma membrane. Green line represents median. The exponential weighting was applied to compensate for cellular asymmetry. (C) Subcellular distribution of HAdV-D26 shown in (A). Bars represent the percentage of virions (N = 283-817 per time point) localized within four mathematically defined zones based on their absolute distance (d) from the nucleus: nuclear fraction (d = 0 µm), perinuclear region (0 < d ≤ 3 µm), mid-cytoplasm (3 < d ≤ 7 µm) and the cortical periphery (d > 7 µm). (D) HAdV-D26 resides in different cellular compartments during the first hour of internalization and intracellular trafficking. Representative TEM images of intracellular trafficking of HAdV-D26 in A549 cells. Cells were incubated on ice with HAdV-D26 (MOI 2x10⁵ vp/cell) for 45 min, then 10 min at 37°C, followed by washing with fresh medium to remove unbound virus particles, and then incubated for 1 h at 37°C. Cells were then fixed and prepared for TEM. Scale bar = 500 nm. Endoplasmic reticulum (ER); Degradative Compartment (DG). White arrowheads denote virus particle. (E) Quantification of HAdV-D26 and HAdV-C5 localization in different cellular compartments in A549 cells shown in (D) and (S1D Fig). Number of cells for HAdV-D26: 56; number of viral particles for HAdV-D26: 262 total viral particles. Number of cells for HAdV-C5 42; number of viral particles for HAdV-C5: 503 total viral particles.

https://doi.org/10.1371/journal.ppat.1014028.g001

To gain better insight into the kinetics of genome delivery by HAdV-D26, we quantified its nuclear entry in A549 cells using qPCR with primers specific for the CMV promoter present in the viral genome. HAdV-D26 DNA was quantified at 0 h, 1 h, 4 h, and 8 h p.i. from both total cell lysate and isolated nuclei. HAdV-D26 genome delivery to nuclei was calculated as the ratio of HAdV-D26 DNA in a nuclear fraction to HAdV-D26 DNA in total cell lysate. Most of the HAdV-D26 DNA was delivered to the nucleus within 1 h (Fig 2A); however, at later time points, we observed a trend toward decreased nuclear delivery of HAdV-D26 DNA, which may reflect DNA degradation or impaired nuclear import. These data are in accordance with the modest transduction efficiency of HAdV-D26 compared to HAdV-C5 and -B35 [4]. Contrary to HAdV-D26, HAdV-C5 genome delivery to the nucleus slightly increased over time (Fig 2A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. HAdV-D26 shows impaired nuclear entry dynamics.

A) HAdV-D26 genome delivery into the host nucleus. A549 cells were infected with HAdV-C5 or HAdV-D26, MOI 10⁴ vp/cell for 45 min on ice to ensure uniform virus entry into the cell. Five min after incubation at 37˚C unbound viruses were removed and fresh medium was added, followed by incubation at 37°C. Cells were harvested at defined time intervals. DNA was extracted from the total cells or isolated nuclei. Viral and cellular DNA were quantified by qPCR using primers for the CMV promoter present in the viral DNA or the cellular GAPDH gene. Results are presented as a ratio of viral DNA measured in the nuclear fraction to the total fraction, normalized to the value for 0 h. Data are presented as mean values ± SD from three experiments. B) HAdV-D26 spatial distribution in U2OS mCherry-α-tubulin cells. U2OS cell with stable expression of mCherry-α-tubulin (red) infected with fluorescently labeled HAdV-D26 (green), MOI 5x10⁴ vp/cell, representing fast movement of HAdV-D26 (white arrow) towards the nucleus. Scale bar represents 10 µm. C) Average velocity of rapid HAdV-D26 movements directed towards or away from the nucleus in U2OS cells (p = 0.12, unpaired t- test, 13 cells, 17 viral particles in both directions). D) Percentage of HAdV-D26 viral particles that reached nuclear surface of U2OS cells within an hour after infection (total of 206 viral particles from 14 cells). **, P < 0.01.

https://doi.org/10.1371/journal.ppat.1014028.g002

After endosomal escape adenoviruses use motor proteins for bidirectional trafficking on microtubules. In non-polarized cells, dynein-based minus-end directed transport prevails over the plus-end directed kinesin‐based transport, leading to the virus enrichment near the centrosome proximal to the nucleus [15]. Therefore, we tracked the movements of internalized fluorescently labeled HAdV-D26 and measured the velocity of its directed motions towards or away from the nuclear surface. For this assay we used U2OS cell line stably expressing mCherry-α-tubulin as a complementary model. The HAdV-D26 transduction efficiency in these cells was comparable to that in A549 (S2 Fig). HAdV-D26 virions usually followed the path of microtubule fibers (Fig 2B) and average HAdV-D26 velocity was 12 ± 6 µm/min towards the nucleus and 9 ± 5 µm/min away from the nucleus (Fig 2C). Observed velocities are comparable to those we measured for HAdV-C5 in the same model (12 ± 5 µm/min towards the nucleus and 13 ± 3 µm/min away from the nucleus) (S3 Fig). Although HAdV-D26 exhibits movements towards the nucleus, only an average of 34% viral particles among those with rapid movements reached the nuclear surface within one hour from infection (Fig 2D).

HAdV-D26 transiently visits the early endosome and resides in lysosomes up to 240 min post-infection

It is known from the literature that HAdV-C2 and -C5 visit early endosomes but are not transported to late endosomes [11]. In order to identify cytosolic HAdV-D26 particles and distinguish them from those inside membranous compartments, we carried out streptolysin O (SLO) assay. SLO-mediated perforation of the plasma membrane is used to introduce antibodies into the cytosol, and accessibility of Alexa Fluor 488-labeled virus to anti-Alexa Fluor 488 antibodies distinguishes cytosolic from endosomal viruses in the permeabilized cells [16]. At 45 min p.i. only 44% of HAdV-D26 particles were positive for anti-Alexa Fluor 488 antibody representing cytosolic particles, an intermediate value between those of HAdV-C5 and the entry-defective HAdV-C2_ts1 (Figs 3A and S4). These data allowed us to conclude that the majority of HAdV-D26 particles linger within some endosomal compartment, however at this point we did not know which one. Thus, we revisited HAdV-D26 localization within early and late endosomes. For that purpose, we observed co-localization of HAdV-D26 and EEA1, a marker of early endosomes, during 120 min p.i., as well as co-localization of HAdV-D26 and LAMP1, a marker of late endosomes and lysosomes, up to 240 min p.i. (Fig 3B). Our confocal microscopy data indicated that 30 min p.i. about 63% of the incoming HAdV-D26 co-localized with EEA1 (Fig 3C). At later time points, co-localization with EEA1 decreased. However, even at 120 min p.i. about 27% of HAdV-D26 was still localized in early endosomes. At 30 min p.i. about 30% of HAdV-D26 co-localizes with LAPM1, indicating that already at early points of infection HAdV-D26 is found in late endosomes or lysosomes. At 120 min p.i., nearly 80% of HAdV-D26 remains within early or late endosomes/lysosomes, and even at 240 min p.i., over 50% still co-localizes with the late endosomal/lysosomal marker LAMP1 (Fig 3C). Retention of HAdV-D26 in late endosomes/lysosomes is in line with our TEM results where we observed HAdV-D26 in degradative compartment.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. HAdV-D26 shows localization within lysosomes as long as 240 min p.i.

A) Membrane penetration of HAdV-C5, HAdV-D26 and HAdV-C2_ts1 in A549 cells measured by SLO penetration assay 45 min p.i. Quantification of cytosolic virus particles in cells infected with Alexa Fluor 488-labeled HAdVs. Intact cells were treated with SLO and then incubated with anti-Alexa Fluor 488 antibody which was detected by secondary Alexa Fluor 594-conjugated antibody. The plot shows the percentages of virus particles positive for anti-Alexa Fluor 488 antibody. One dot represents one cell. Data are presented as mean values ± SEM. B) Localization of HAdV-D26 within early and late endosomes/lysosomes upon internalization. Representative confocal images are shown. Cells were incubated on ice with fluorescently labeled HAdV-D26 (green; MOI 10⁵ vp/cell) for 45 min, then 10 min at 37°C, followed by washing with fresh medium to remove unbound viruses. After additional incubation at 37°C for the indicated times, cells were fixed and stained with an antibody against EEA1 or LAMP1 (magenta), phalloidin-AF555 (red), and DAPI (blue). The upper right panel shows examples of co-localization grading: white arrows indicate co-localization, whereas green arrows indicate lack of co-localization. C) Quantification of co-localization of HAdV-D26 with either EEA1 or LAMP1 at different time points shown in (B). Percentages in black above the plot show the total amount of viruses found in any endosome (either EEA1 or LAMP1 positive). Data are presented as mean values ± SD from two independent experiments. The cell number that was analyzed is as follows: 1) for EEA1: 21 cells at 0 min, 78 cells at 30 min, 46 cells at 60 min, 65 cells at 120 min; 2) for LAMP1: 43 cells at 0 min, 78 cells at 30 min, 50 cells at 60 min, 90 cells at 120 min. D) Co-localization of HAdV-D26 with LysoTracker in live A549 cells. Cells were incubated on ice with fluorescently labeled HAdV-D26 (green) for 45 min, then 10 min at 37°C, followed by washing with fresh medium to remove unbound viruses, and then incubated at 37°C for the indicated times. In the last 30 min of incubation, Lysotracker Deep Red solution was added to the medium. After incubation, cells were washed with fresh medium and immediately observed by a confocal microscope. Reflection Interference Contrast Microscopy (RICM) is shown in grey. Representative confocal images are shown (scale bar = 10 μm). N denotes number of analyzed cells, M denoted Manders’ coefficient.

https://doi.org/10.1371/journal.ppat.1014028.g003

Since our TEM results showed significant accumulation of HAdV-D26 particles in structures consistent with degradative compartments, and because we saw co-localization of HAdV-D26 with LAMP1-positive endosomes, we wanted to further identify the compartment where HAdV-D26 is retained. Thus, we performed co-localization of HAdV-D26 with LysoTracker, a marker for lysosomes, using live confocal microscopy. A high degree of co-localization between HAdV-D26 and LysoTracker was observed at 120 and 240 min p.i. (Fig 3D), with mean Manders’ split coefficients (M1) of 0.962 ± 0.004 and 0.947 ± 0.008 respectively. These results further corroborate our data, indicating that HAdV-D26 resides within lysosomes for up to 240 min p.i., which might influence HAdV-D26 transduction efficiency.

Inhibiting lysosomal acidification, but not transport from early to late endosomes/lysosomes, modestly increases HAdV-D26 transduction efficiency

HAdV-D26 initially accumulates in early endosomes within 30 min p.i., then gradually shifts to late endosomes/lysosomes, where around 50% of the particles remain even at 240 min p.i. Since cargo directed to the lysosomes is typically destined for degradation, one can assume that prolonged residence of HAdV-D26 in late endosomes/lysosomes may lead to its degradation and may partly explain its low infectivity. To investigate whether disrupting lysosomal function affects HAdV-D26 infectivity, we measured its transduction efficiency in cells treated with the lysosomotropic agents such as bafilomycin A1 (BafA1), chloroquine, and ammonium chloride (NH₄Cl). These inhibitors elevate intralysosomal pH, thereby impairing lysosomal function and disrupting autophagic protein degradation [17]. Treatment with all three inhibitors modestly increased transduction efficiency of HAdV-D26, where only treatment with bafilomicin A1 was statistically significant (Fig 4A). Nevertheless, the obtained data imply that at least some HAdV-D26 particles trafficking to late endosomes/lysosomes cannot complete the infection cycle.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. HAdV-D26 transduction efficiency is unaffected by endosomal pH modulation, inhibition of endosomal maturation, or co-infection with HAdV-C5.

A) Increasing intralysosomal pH or inhibiting early-to-late endosome transition did not markedly affect HAdV-D26 transduction efficiency. Cells were pretreated with inhibitors (30 min, 37˚C) - bafilomycin A1 (10 nM), chloroquine (50 µM), NH₄Cl (5 mM) or EGA (15 µM) and subsequently incubated with HAdV-D26 for one hour (MOI 10⁴ vp/cell) in medium with the specific inhibitor, after which the medium was changed and fresh medium was added. After 24 h of incubation, the fluorescence intensity of eGFP encoded by HAdV-D26 and the fluorescence intensity of cellular DNA labeled by Hoechst was measured. The results were calculated as the ratio of eGFP/Hoechst values and normalized to infected cells that were not treated (n.t.). B) Co-infection with HAdV-C5 does not influence HAdV-D26 transduction efficiency. Cells were infected either with HAdV-D26 (MOI 10⁴ vp/cell) alone or with a mixture of HAdV-D26 (MOI 10⁴ vp/cell) and HAdV-C5 at different MOIs (10², 10³, 10⁴ vp/cell) (1 h, 37˚C), and then the medium was changed, and fresh medium was added. After 24 h of incubation, the fluorescence intensity of eGFP encoded by HAdV-D26 and the fluorescence intensity of cellular DNA labeled by Hoechst was measured. The results are presented as the ratio of eGFP/Hoechst values and normalized to cells infected with HAdV-D26. Data are presented as mean ± SD from two or three independent experiments in duplicates. *, P < 0.05; **, P < 0.01.

https://doi.org/10.1371/journal.ppat.1014028.g004

We also investigated whether HAdV-D26 transport from early endosomes to late endosomes/ lysosomes would affect its transduction efficiency. For that purpose, we used EGA, an inhibitor of endosomal trafficking that specifically blocks transport to LAMP1-positive compartments, thereby impacting lysosomal function without disrupting other recycling pathways [18]. The late endosome trafficking inhibitor EGA only slightly decreased transduction efficiency with HAdV-D26 (Fig 4A), further corroborating our assumption that for efficient transduction, HAdV-D26 escapes from endosomes before reaching the lysosomes. Overall, these results indicate that trafficking from the early endosome to the lysosome is not essential for efficient HAdV-D26 transduction.

So far, our data suggest that HAdV-D26 might behave similarly to HAdV-C2_ts1, which does not efficiently escape from endosomes [19]. On the other hand, HAdV-C2 is liberated from the early endosome rather quickly and very efficiently. When HAdV-C2_ts1 was co-infected with HAdV-C2 its normally increased co-localization in early endosomes was significantly reduced [11], implying that HAdV-C2 facilitates the release of HAdV-C2_ts1 from early endosomes. Based on a similar concept, we performed a co-infection experiment using HAdV-D26 (encoding an eGFP reporter gene) and HAdV-C5 (encoding a β-galactosidase reporter gene) at increasing multiplicities of infection (MOI), and assessed HAdV-D26 transduction efficiency by quantifying eGFP expression. Co-infection of HAdV-D26 with HAdV-C5 at a high MOI of 10⁴ increased HAdV-D26 transduction efficiency by only 18%, while co-infection at lower MOIs had no significant effect on HAdV-D26 transduction efficiency (Fig 4B). These data indicate that HAdV-C5 endosome penetration does not liberate HAdV-D26 competent for productive transduction or that HAdV-D26 particles do not traffic through the same endosomal compartments as HAdV-C5.

Functional Rab5, Rab7, and Rab9 are involved in HAdV-D26 transduction

After endocytosis, cargos are sorted into endosomes and directed to various destinations. Rab GTPases, localized to distinct endosomes, regulate these processes by controlling membrane budding, vesicle formation, motility, tethering, and fusion through effector recruitment [10]. It has been shown that HAdV-C2 and -C5 reach the cytosol in a Rab5-independent manner, while HAdV-C2_ts1 depends on Rab5 for endocytosis and transport to late endosomes and lysosomes. At the same time, sorting of HAdV-C2_ts1 to late endosomes was independent of Rab7 [11]. To test if functional Rab proteins are involved in HAdV-D26 intracellular trafficking, we determined HAdV-D26 transduction efficiency in cells transfected with wild-type (WT) or dominant-negative (DN) form of Rab5, Rab7, Rab9, or Rab11. Dominant-negative mutants work in cells by competing with endogenous Rabs for binding to GEFs (Guanine Nucleotide Exchange Factor) and since they cannot interact with downstream target proteins, they form ‘dead-end’ complexes further preventing the activation of endogenous Rabs [20]. The localization of transfected WT or DN forms and their transfection efficiencies are shown on S5 and S6 Figs respectively; note that WT proteins retain a punctate endosomal localization closely mimicking the physiological staining pattern of endogenous Rabs, whereas DN mutants display an altered, more diffuse distribution as previously reported [21]. Transduction efficiency of HAdV-D26 in cells transfected with DN form of Rab5, Rab7, or Rab9 measured by flow cytometry (Fig 5A) and confocal microscopy (Fig 5B) was partially decreased, while the Rab11 DN did not influence HAdV-D26 transduction efficiency. The most pronounced effect was observed in cells expressing Rab9 DN, where HAdV-D26 transduction efficiency was decreased by approximately 40% compared to cells transfected with Rab9 WT. These results suggest that HAdV-D26 transduction efficiency partially depends on functional Rab5, Rab7, and Rab9 proteins.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. HAdV-D26 transduction efficiency partially depends on functional Rab5, Rab7 and Rab9 proteins.

A549 cells were transfected with WT or DN Rab5, Rab7, Rab9 or Rab11 plasmids (48 h, 37°C), and subsequently incubated with HAdV-D26 for one hour, after which the medium was changed, and fresh medium was added. After 24 h, transduction efficiency was assessed by flow cytometry or confocal microscopy. A) HAdV-D26 transduction efficiency corresponding to the geometric mean of the fluorescence intensity of eGFP signal encoded by HAdV-D26 (MOI 10⁴ viral particles per cell) shown as a value relative to the corresponding WT sample. Data are presented as mean ± SD from three independent experiments in duplicates. B) Panels show representative cells from one experiment, where cells expressing a certain Rab protein (DsRed positive cells; red) were scored for HAdV-D26 infection (eGFP positive cells; green). Cell nuclei are shown in blue (DAPI). Yellow arrows indicate cells that are red/green positive (transfected and transduced cells), while white arrows indicate cells that are red positive (transfected cells) and green negative (non-infected cells). Scale bar, 10 μm. HAdV-D26 transduction efficiency (MOI 10⁵ vp/cell) is quantified as a percentage of the eGFP expressing cells transfected with Rab5, Rab7, Rab9 or Rab11 DN compared to eGFP expressing cells transfected with WT plasmids (right hand graphs). N represents the number of analyzed cells. *, P < 0.05; ***, P < 0.001.

https://doi.org/10.1371/journal.ppat.1014028.g005

Rab9 knockdown decreases HAdV-D26 localization in the lysosome

To investigate further the role of Rab proteins in HAdV-D26 intracellular trafficking, we analyzed HAdV-D26 co-localization with EEA1 in cells transfected with siRNA specific for Rab5, Rab7, or Rab9 (Rab5, Rab7 or Rab9 knockdown efficiency is shown in S7 Fig). The results (Fig 6A) show that knockdown of Rab5 protein in A549 cells increases the localization of HAdV-D26 in the early endosome. Namely, at 30 min p.i., 13% more HAdV-D26 localized to EEA1-positive endosomes in Rab5-knockdown cells compared to those transfected with scrambled siRNA. However, by 120 min p.i., HAdV-D26 localization in EEA1-positive endosomes was similar between Rab5-knockdown and control cells (si(-)). Rab7 knockdown in A549 cells did not affect HAdV-D26 localization in early endosomes at 30 min p.i., but during the following 90 min, it increased retention of HAdV-D26 in early endosomes compared to control cells.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Rab5 and Rab7 knockdown increase HAdV-D26 localization within early endosomes, while Rab9 knockdown decreases HAdV-D26 localization within late endosomes/lysosomes.

Cells were transfected with specific siRNA at the final concentration of 10 nM for Rab5 and Rab7, and 5 nM for Rab9, and 48 h later incubated with fluorescently labeled HAdV-D26 (green; MOI of 10⁵ vp/cell) for 45 min on ice to ensure uniform virus entry into the cell and subsequently transferred to 37˚C for 10 min. Unbound viruses were washed, fresh medium was added, and cells were returned to 37˚C for 30 min or 120 min, after which they were fixed and stained with antibody against (A) EEA1 (magenta) or (B) LAMP1 (magenta). F-actin is shown in red, and the nucleus in blue. Representative confocal images are shown. Scale bar = 10 μm. Rab5 and Rab7 knockdown experiments were performed in parallel using a shared si(-) control, whereas Rab9 experiments were conducted independently with a batch-matched si(-) control. Graphs on the right show quantification of HAdV-D26 co-localization with EEA1 or LAMP1 at 30 and 120 min following knockdown of the indicated Rab protein. Co-localization was quantified by counting the number of virus particles overlapping with EEA1- or LAMP1-positive structures per cell. Data represent the mean ± SEM from at least 30 cells analyzed per condition. *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

https://doi.org/10.1371/journal.ppat.1014028.g006

At specific time points, increased co-localization of HAdV-D26 with EEA1 in Rab5 or Rab7 knockdown cells suggests that, under normal conditions, HAdV-D26 might be liberated from Rab5 and Rab7 positive early endosomes. Knockdown of Rab9 does not change the percentage of HAdV-D26 co-localization with EEA1 marker, indicating that Rab9 does not interfere with HAdV-D26 trafficking within early endosomes.

Next, we analyzed HAdV-D26 co-localization with LAMP1 in cells transfected with siRNA specific for Rab5, Rab7, or Rab9 (Fig 6B). Knockdown of Rab5 protein in A549 cells modestly increased the localization of HAdV-D26 to late endosome/lysosome at 30 min p.i., with 10% more virus localized in LAMP1-positive endosomes compared to cells transfected with scrambled siRNA. At 120 min p.i., HAdV-D26 localization in LAMP1-positive endosomes was similar between Rab5-knockdown and control cells. In contrast, Rab7 knockdown in A549 cells reduced the proportion of HAdV-D26 in late endosomes/lysosomes at 120 min p.i. from 52% to 42%. This decrease in localization in LAMP1 positive endosomes is even more obvious in cells with Rab9 knockdown, where localization of HAdV-D26 120 min p.i. decreased from 60% to 24%. Decreased co-localization of HAdV-D26 with LAMP1 in Rab7- or Rab9-knockdown cells suggests that Rab7 and Rab9 might promote HAdV-D26 delivery to late endosome/lysosome. Knockdown of Rab5 does not change the percentage of HAdV-D26 co-localization with LAMP1 marker, indicating that Rab5 does not interfere with HAdV-D26 trafficking within late endosomes/lysosomes.

Rab9 knockdown increases HAdV-D26 transduction efficiency

Since we saw that functional Rab proteins are involved in HAdV-D26 transduction and that knockdown of Rab proteins influences HAdV-D26 transport through the endosomes, we wanted to investigate if depleting different Rab proteins would also play a role in the transduction efficiency of HAdV-D26. Thus, we downregulated Rab5, Rab7, or Rab9 by using specific siRNAs and measured the transduction efficiency of HAdV-D26. As shown in Fig 7A, a decrease of approximately 30% in HAdV-D26 transduction efficiency was observed upon Rab5 downregulation, while downregulation of Rab7 did not influence HAdV-D26 transduction efficiency. On the other hand, downregulating Rab9 increased HAdV-D26 transduction efficiency by 62%. Increased adenovirus transduction efficiency can be a consequence of increased internalization or increased adenovirus genome delivery to the host nucleus [22]. Therefore, we first investigated whether decreased expression of Rab9 influenced HAdV-D26 internalization. We quantified the number of internalized fluorescently labeled HAdV-D26 30 and 120 min p.i. and observed that downregulation of Rab9 increased HAdV-D26 internalization in average by more than 30% (Fig 7B). Downregulating Rab9 did not alter the expression levels of αvβ3 integrin, CAR or CD46 (S8 Fig). Knowing that Rab9 knockdown increases HAdV-D26 escape from lysosomes, we also wanted to investigate if Rab9 knockdown would change the amount of HAdV-D26 genome imported into the host nucleus. For that purpose, we measured HAdV-D26 genome delivery to the nucleus in A549 cells transfected with scrambled or Rab9 siRNA and observed an increase in HAdV-D26 DNA nuclear import in cells with Rab9 knockdown (Fig 7C). All these results indicate that Rab9 influences HAdV-D26 transduction efficiency on several levels including virus internalization and genome delivery.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Knockdown of Rab9 increases internalization and transduction efficiency of HAdV-D26.

A) Transduction efficiency of HAdV-D26 in A549 cells downregulated for Rab5, Rab7, or Rab9. Cells were transfected with specific siRNA at a final concentration of 10 nM for Rab5 and Rab7, and 5 nM for Rab9, and 48 h later infected with HAdV-D26 at an MOI of 10⁴ vp/cell. Transduction efficiency was measured as eGFP fluorescence 48 h p.i. The results are obtained by flow cytometry and presented as mean ± SD from three independent experiments normalized to control, i.e., cells transfected with scrambled siRNA. B) Internalization of HAdV-D26 in A549 cells downregulated for Rab9. Cells were transfected with specific Rab9 siRNA, 5 nM, and 48 h later infected with fluorescently labeled HAdV-D26, MOI 10⁵ vp/cell, for 45 min on ice to ensure uniform virus entry into the cell and subsequently transferred to 37˚C for 10 min. Unbound viruses were washed, fresh medium was added, and cells were returned to 37˚C for 30 min or 120 min, after which they were fixed and imaged by confocal microscopy. Quantification is presented as the number of viral particles per cell with mean ± SD. C) HAdV-D26 genome delivery into the host nucleus in Rab9 downregulated cells. Cells were transfected with scrambled siRNA or Rab9 siRNA at a final concentration of 5 nM, and 48 h later infected with HAdV-D26, MOI 10⁴ vp/cell for 45 min on ice to ensure uniform virus entry into the cell. Five min after incubation on 37˚C unbound viruses were removed, and fresh medium was added, followed by incubation at 37°C. Cells were harvested at defined time intervals. DNA was extracted from the total cells or isolated nuclei. Viral and cellular DNA were quantified by qPCR using primers for the CMV promoter present in the viral DNA or the cellular GAPDH gene. Results are presented as a ratio of viral DNA measured in the nuclear fraction to the total fraction, normalized to the corresponding si(-) value. Data are presented as mean values ± SD from three experiments. Subsequently, the same experimental setup was performed as described in Fig 2A. *, P < 0.05; **, P < 0.01, ***, P < 0.001.

https://doi.org/10.1371/journal.ppat.1014028.g007

Cells, viruses, and antibodies

HEK293 (human embryonic kidney; ATCC CRL-1573) and A549 (human lung carcinoma; ATCC CCL-185) cells were obtained from the ATCC Cell Biology Collection and were cultured in antibiotic-free Dulbecco’s Modified Eagle’s Medium- high glucose (DMEM) supplemented with 10% (vol/vol) Fetal Bovine Serum (FBS) if not stated differently. Osteosarcoma U2OS cell line (female) stably expressing mCherry-α-tubulin was a kind gift from Marin Barišić (Danish Cancer Society Research Center, Copenhagen, Denmark) and Helder Maiato (Institute for Molecular Cell Biology, University of Porto, Portugal). U2OS cells were maintained in DMEM supplemented with 10% (vol/vol) FBS and 50 µg/mL geneticin. All cells were grown at 37°C in a humidified incubator with a 5% CO₂ atmosphere. Replication-incompetent recombinant adenoviral vectors based on adenovirus types 2, 5 or 26 were previously constructed [43–45]. Viruses were propagated on HEK293 cells and purified by CsCl gradients. They carried an enhanced green fluorescent protein (eGFP) gene driven by the cytomegalovirus (CMV) promoter as a reporter gene. Fluorescent labeling of adenoviruses was previously described [4]. Primary antibodies used for immunofluorescence, co-localization analyses, western blot or flow cytometry were as follows: anti-EEA1 (#2411, Cell Signaling, SAD), anti-LAMP1 (ab24170, Abcam, UK), anti-Rab5 (#3547, Cell Signaling, SAD), anti-Rab7 (#9367, Cell Signaling, SAD), anti-Rab9A (#5118, Cell Signaling, SAD), anti-Alexa Fluor 488 recombinant F(ab)-rabbit monoclonal antibody (15H9L93, Invitrogen), anti-CAR (RcmB, Merck Millipore), anti-CD46 (MEM-258, Thermo Fisher Scientific) and anti-αvβ3 integrin (LM609, Merck Millipore).

Transduction efficiency assay

Transduction efficiency was measured either by flow cytometry as previously described [5] or by spectrophotometer. For assaying by spectrophotometer, cells were seeded in 96-well plates at a density of 4x10³ cells per well, and 48 h later incubated with HAdV vectors at MOI of 10⁴ vp/cell. After 1 h at 37°C in a humidified incubator with a 5% CO₂ atmosphere the medium was removed and replaced with of antibiotic-free DMEM supplemented with 10% (vol/vol) FBS and cultured for additional 24 h. Cells were then washed twice with PBS and eGFP expression was measured by spectrophotometer at 515 nm. Subsequently, after washing, cellular DNA was stained with Hoechst 33342 (5 µM) and the fluorescence intensity was measured at 455 nm. The results were calculated as the ratio of eGFP/Hoechst values.

For measuring transduction efficiency after interfering with endosome pH cells were pretreated with following inhibitors (30 min, 37˚C): bafilomycin A1 (10 nM), chloroquine (50 µM), NH₄Cl (5 mM) or EGA (15 µM) and subsequently incubated with HAdV-D26 for one hour (MOI 10⁴ vp/cell) in medium with the specific inhibitor, after which the medium was changed and fresh medium was added. After 24 h of incubation, the fluorescence intensity of eGFP was measured and subsequently the fluorescence intensity of cellular DNA labeled by Hoechst.

For studying role of Rab proteins in HAdV-D26 transduction efficiency cells were transfected with DN or WT form of Rab5, Rab7, Rab9 or Rab11 purchased from Addgene: DsRed-Rab5 WT, #13050; DsRed-Rab5 DN, #13051; DsRed-Rab7 WT, #12661; DsRed-Rab7 DN, #12662; DsRed-Rab9 WT, #12677; DsRed-Rab9 DN, #12676; DsRed-Rab11 WT, #12679; DsRed-Rab11 DN, #12680. Plasmids were transfected into A549 cells using Lipofectamine (Thermo Fisher Scientific) according to the manufacturer’s protocol, and 48 h after cells were transduced with HAdV-D26.

Transfection with Rab WT or DN coding plasmids and transduction efficiencies after infection of transfected cells with HAdV-D26 were measured by flow cytometry using a FACSCalibur flow cytometer (BD Biosciences, USA). Data acquisition was performed using the BD CellQuest software package (BD Biosciences, USA). The number of acquired events per sample was 10,000. Data were analyzed using FCS Express 3 (De Novo Software, USA), and the gating strategy was established by first identifying the baseline cell population and eliminating cell debris using linear Forward Scatter (FSC) versus Side Scatter (SSC) dot plots. Within this viable cell gate, viral transduction efficiency was quantified based on the green fluorescence intensity of the eGFP signal encoded by the adenoviral vectors, recorded in the FL1 channel (excitation 488 nm, emission 530/30 nm filter). Results are expressed as relative transduction efficiency, calculated from the geometric mean of fluorescence intensity (MFI) of the eGFP signal and normalized to the corresponding control sample. To evaluate transfection efficiencies, the expression of DsRed-tagged Rab constructs (WT and DN) was analyzed in the FL2 channel relative to untransfected control cells. Transfection rates were determined as the percentage of positive cells positioned beyond the autofluorescence threshold of the control population.

SLO penetration assay

This assay has been previously described [16]. Briefly, A549 cells were seeded on Alcian blue-coated glass coverslips in 24-well dishes (40,000 cells/well) and grown for 2 days. Alexa Fluor 488-labeled viruses were bound to cells at 0°C for 60 min. The unbound virus was washed away and the cells were placed into a 37°C water bath for 45 min. After virus internalization, cells were placed on ice, washed twice with streptolysin O (SLO) binding buffer and processed for SLO-mediated perforation of the plasma membrane. Cells were washed twice with SLO binding buffer and incubated with SLO solution (6.2 µg/mL) for 10 min on ice. Unbound SLO was removed by two washes with SLO binding buffer. Cells were then incubated in SLO binding buffer for 5 min at 37°C in a water bath to induce pore formation and immediately returned to ice. After being washed with SLO internalization buffer cells were incubated with anti-Alexa Fluor 488 antibody for 1 h at 0°C. Next, cells were and fixed with 3% paraformaldehyde, permeabilized with 0.5% Triton X-100 in PBS for 5 min and washed once with PBS. Blocking was performed with 10% goat serum in PBS for 5 min at room temperature. Then, cells were incubated with anti-rabbit Alexa Fluor 594 antibody and DAPI for 30 min at room temperature, followed by three 4-min washes in PBS. A post-fixation step was performed with 3% PFA in PBS for 12 min at room temperature. Cells were washed once with PBS, quenched with 25 mM ammonium chloride in PBS for 10 min, washed once more with PBS, and mounted for imaging.

siRNA experiments

To downregulate specific molecules, we used following predesigned MISSION esiRNA: Rab5 (EHU053901), Rab7 (EHU053221), Rab9 (EMU171281), universal negative control #1 (SIC001), all from Sigma-Aldrich. Cells were transfected at a confluence of 30–50% using Lipofectamine RNAiMax reagent (Invitrogen) according to the manufacturer’s protocol. Final esiRNA concentrations were 10 nM for Rab5 and Rab7, and 5 nM for Rab9. The efficiency of silencing was verified 48 h after transfection by Western blot. Briefly, cells transfected with specific or negative control siRNA were lysed with Laemmli buffer heated to 95°C and sonicated. Proteins were separated by SDS-PAGE and transferred to nitrocellulose membrane. After blocking, the membrane was probed with appropriate primary antibodies, followed by incubation with appropriate horseradish peroxidase-conjugated IgG secondary antibody. Detection was performed with Pierce ECL Western Blotting Substrate using ChemiDoc Imaging System. Densitometry was performed with ImageJ software. Proteins were normalized to the total proteins stained with amido black. The results are presented as the relative expression of proteins compared to cells transfected with control siRNA.

Transmission electron microscopy

A549 cells grown on glass cover slides were infected with HAdV-D26 or HAdV-C5 at an MOI of 2x10⁵ vp/cell, incubated on ice for 45 min, and then shifted to 37°C for 10 min. After removal of unbound virus particles by washing with fresh medium, cells were incubated for 1 h at 37 °C, after which they were washed twice with PBS and fixed with a mixture of 2% glutaraldehyde and 1% tannic acid in 0.4 M HEPES buffer (pH 7.2). Dehydration was performed using dry ethanol, and heavy metal staining and embedding in EPON resin were carried out on the monolayer according to the protocol described in [46]. Serial sections 70 nm thick were collected from the first 2 µm of the basal cell surface and sections were stained with uranyl acetate and lead citrate for image acquisition, which was performed using a JEOL JEM 1400 transmission electron microscope operated at 120 kV. The number of viral particles present in different cellular compartments was quantified in 56 cells for HAdV-D26 and 42 for HAdV-C5, each one presenting an area greater than 50 µm². The distribution of viral particles among the different compartments was normalized to the total number of quantified particles (262 HAdV-D26 particles, 503 HAdV-C5 particles).

Confocal microscopy

Cells (2x10⁴ per coverslip) were seeded on coverslips in 24-well plates. Two days later, fluorescently labeled adenoviruses were added to the cells at an MOI 5x10³ vp/cell for HAdV-C5 and 10⁵ for HAdV-D26. Because HAdV-26 infects less efficiently A549 cells than HAdV-C5, confocal microscopy was performed under MOIs that yield comparable detectable signals (resulting in up to 50 viruses per cell which allowed for accurate assessment of intracellular localization and trafficking patterns). Cells were than incubated on ice for 45 min to allow viral binding. Subsequently, the cells were transferred to 37°C for the indicated time. The cells were fixed with 2% paraformaldehyde (PFA) in PBS for 12 min at room temperature. Nuclei were labeled with DAPI (4 = ,6-diamidino-2-phenylindole). Coverslips were slide mounted using Fluoromount G (Southern Biotech). Confocal laser scanning microscopy analyses were performed using a Leica TCS SP8 X inverted confocal microscope (Leica Microsystems, Wetzlar, Germany) with 63x/1.40 oil-immersion objective. The images showing intracellular trafficking of Alexa Fluor 488-labeled HAdVs are maximum projections of at least 7 confocal stacks and were processed with the Leica Application Suite X (LAS X) software platform, Adobe Photoshop CC software (Adobe Systems), and ImageJ.

To quantify the intracellular trafficking kinetics of viral particles during the time-course infection (30, 60, 120, 180, and 240 min), a shape-corrected spatial index, termed the Relative Distance Index (Weighted), was calculated for each individual virion (N, number of analyzed virions). First, the absolute Euclidean distances from each viral particle to the nucleus (d1, defined by DAPI staining) and to the plasma membrane (d2, defined by actin staining) were measured in micrometers (µm) using the Euclidean Distance Map (EDM) function in Fiji. To compensate for the characteristic morphological asymmetry of A549 cells and to biologically weigh particles residing in close physical proximity to the nucleus, a non-linear exponential weighting was applied. The Relative Distance Index (Weighted) (I) for each particle was determined using the following formula: I = (d1/ (d1 + d2)) * (1 - EXP(-d1/ 2)). This index strictly constrained the values between 0 (at the nucleus) and 1 (at the plasma membrane). The exponential term served as a mathematical buffer to maintain lower index values for virions within the perinuclear region, preventing local membrane geometry from artificially inflating the distance metrics. For categorical population analysis, virions were assigned to four distinct subcellular fractions based on their absolute distance from the nucleus (d): (i) Nuclear fraction (d = 0 µm), (ii) Perinuclear fraction (0 < d ≤ 3 µm, corresponding to the endoplasmic reticulum and Golgi zones), (iii) Mid-cytoplasmic fraction (3 < d ≤ 7 µm, corresponding to the main cytosol and microtubule transit tracks), and (iv) Cortical/Plasma membrane fraction (d > 7 µm, capturing the cortical actin network and cell periphery). For each time point, the distribution of virions within these four fractions was expressed as a percentage of the total detected viral particles.

Live cell microscopy

Day before live-cell microscopy, 6x10⁴ U2OS-mCherry-α-tubulin cells were seeded per chamber on 4-chamber glass-bottom dishes (#1.5 glass thickness, iBL). Viral infection was performed on ice with 5x10⁴ MOI and the cells were kept on ice until live cell imaging. The microscopy was performed on Expert Line easy3D STED microscope system (Abberior Instruments) with a 100 x/1.4NA UPLSAPO100x oil objective (Olympus, Tokio, Japan) and avalanche photodiode detector. During imaging, the cells were kept at 37°C and a 5% CO₂ atmosphere within a heating chamber (Okolab). Images were acquired every ten seconds using Imspector software. Imaging was performed within an hour after infection. Image analysis was performed in Fiji/ImageJ (National Institutes of Health). Quantification and data analysis were performed in GraphPad Prism (GraphPad Software). Tracking of viral particles was done manually by using the Multi-point tool in Fiji. Upon inspection of viral particle movement within the cells, viral particles which could be followed for at least 20 s (three time- frames) were taken into account. Only fast movements towards or from the nucleus were tracked. For evaluation of virus redistribution, virus particles were counted as reaching the nucleus if they were ≤ 1 µm from the nuclear surface.

For co-localization of HAdV-D26 with LysoTracker in live A549 cells, 2x10⁴ cell were seeded on coverslips in 24-well plates. Two days later, fluorescently labeled adenoviruses were added to the cells (MOI 5x10⁴ vp/cell) and incubated on ice for 45 min to allow binding, then 10 min at 37°C, followed by washing with fresh medium to remove unbound viruses, and then incubated at 37°C for the indicated times. In the last 30 min of incubation, LysoTracker Deep Red solution (Invitrogen) was added to the medium to final concentration of 50 nM. After incubation, cells were washed with fresh medium and immediately observed by a Leica TCS SP8 X inverted confocal microscope (Leica Microsystems, Wetzlar, Germany) as stated before. Reflection Interference Contrast Microscopy (RICM) was used to determine cell shape. The co-localization analysis was performed using digital images processed with a co-localization plug-in in Fiji. Manders’ split co-localization coefficient was used to quantify pixel-wise co-localization between HAdV-D26 and LysoTracker signals. Specifically, the coefficient represents the proportion of the HAdV-D26 signal overlapping with the LysoTracker signal, indicating the extent of viral localization within lysosomal compartments. For each time point (120 and 240 min p.i.), a total of 4 independent fields of view containing N = 38 and N = 41 individual cells, respectively, were analyzed. Data are presented as mean ± standard deviation (SD).

Adenovirus genome delivery to the host nucleus

A549 cells (2x10⁴ per well) were seeded in 6-well plates, and 24 hours after seeding, viral infection was performed on ice with 1x10⁴ MOI in DMEM supplemented with 0.2% (vol/vol) FBS, and incubated on ice for 45 min to synchronize viral entry. After 5 min at 37°C in a humidified incubator with a 5% CO₂ atmosphere, the medium was removed and replaced with antibiotic-free DMEM supplemented with 10% (vol/vol) FBS. The cells were then kept at 37 °C for defined time points (0, 1, 4, and 8 h).

To monitor viral DNA delivery dynamics in A549 cells after Rab9 silencing, Rab9 knockdown was performed using DharmaFECT 4 on the same day as cell seeding, according to the manufacturer’s protocol. After 48 hours of incubation, viral infection was performed as mentioned above. At the defined time points, cells were washed with PBS, trypsinized and after washing with PBS divided into two fractions. The cell pellet from first fraction was used for total DNA isolation, while the second cell pellet was used for nuclear fraction isolation. For nuclear fraction isolation, 100 µL of hypotonic solution (10 mM HEPES (pH 7.5), 10 mM NaCl, 1 mM DTT, 1 mM EDTA, 2 mM MgCl₂x6H₂O) was added to the pellet and incubated for 5 min at room temperature, followed by the addition of 100 µL of 0.05% NP-40 detergent for 10 min. The mixture was vortexed and centrifuged for 2 min at 13,400 × g. The resulting pellet was washed four times in PBS. The final slightly translucent pellet represented the nuclear fraction. Total cellular and viral DNA were extracted using the QIAamp DNA Blood Mini Kit (Qiagen) according to the manufacturer’s instructions. DNA concentration was measured spectrophotometrically at 260 nm, and purity was assessed at 280 nm. The isolated DNA served as a template for qPCR using primers targeting the CMV promoter (forward primer: 5’-TGGGCGGTAGGCGTGTA-’3; reverse primer: 5’-CGATCTGACGGTTCACTAAACG-3’) or the cellular GAPDH gene (forward primer: 5’- AGAACATCATCCCTGCCTCTACTG-3’; reverse primer: 5’ TGTCGCTGTTGAAGTCAGAGGAGA-3’).

Flow cytometry

Flow cytometry was used to determine the cell surface expression of CAR, CD46 and αvβ3 integrin after transfecting cells with si(Rab9). For analysing cell surface expression of CAR, CD46 and αvβ3 integrin transfected cells were grown in tissue culture dishes, detached with trypsin, and washed twice with PBS. Subsequently, the cells were incubated on ice with the specific primary antibodies that recognized CAR, CD46 and αvβ3 integrin. The binding of unlabelled primary antibodies was revealed by using FITC-conjugated anti-mouse IgG as a secondary reagent. Flow cytometry was done on FACSymphony A1 flow cytometer (BD Biosciences, USA).

Statistical analysis

All experiments were performed in at least two biological replicas in duplicate or triplicate. All analyses and graphs were created in GraphPad Prism (GraphPad Software Inc., USA). To analyse the intracellular trafficking kinetics of viral particles during the time-course infection, due to the high frequency of absolute zero-values (nuclear fraction) and the inherently skewed, non-normal distribution of the spatial tracking datasets across all time points (N = 261–817 particles per condition), non-parametric methods were implemented. The global effect of infection time on the Relative Distance Index (Weighted) was evaluated using the Kruskal-Wallis test. Upon confirming global statistical significance, Dunn’s multiple comparison post-hoc test with Bonferroni-adjusted p-values was applied to determine specific differences between individual time-point pairs. Other data were analysed by unpaired Student’s t test or Mann-Whitney U test and expressed as mean ± SD or SEM. ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.