Mouse Polyomavirus Enters Early Endosomes, Require
http://www.100md.com
病菌学杂志 2006年第9期
Department of Genetics and Microbiology, Faculty of Science, Charles University, Prague, Czech Republic
Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, Prague, Czech Republic
ABSTRACT
Mouse polyomavirus (PyV) virions enter cells by internalization into smooth monopinocytic vesicles, which fuse under the cell membrane with larger endosomes. Caveolin-1 was detected on monopinocytic vesicles carrying PyV particles in mouse fibroblasts and epithelial cells (33). Here, we show that PyV can be efficiently internalized by Jurkat cells, which do not express caveolin-1 and lack caveolae, and that overexpression of a caveolin-1 dominant-negative mutant in mouse epithelial cells does not prevent their productive infection. Strong colocalization of VP1 with early endosome antigen 1 (EEA1) and of EEA1 with caveolin-1 in mouse fibroblasts and epithelial cells suggests that the monopinocytic vesicles carrying the virus (and vesicles containing caveolin-1) fuse with EEA1-positive early endosomes. In contrast to SV40, PyV infection is dependent on the acidic pH of endosomes. Bafilomycin A1 abolished PyV infection, and an increase in endosomal pH by NH4Cl markedly reduced its efficiency when drugs were applied during virion transport towards the cell nucleus. The block of acidification resulted in the retention of a fraction of virions in early endosomes. To monitor further trafficking of PyV, we used fluorescent resonance energy transfer (FRET) to determine mutual localization of PyV VP1 with transferrin and Rab11 GTPase at a 2- to 10-nm resolution. Positive FRET between PyV VP1 and transferrin cargo and between PyV VP1 and Rab11 suggests that during later times postinfection (1.5 to 3 h), the virus meets up with transferrin in the Rab11-positive recycling endosome. These results point to a convergence of the virus and the cargo internalized by different pathways in common transitional compartments.
INTRODUCTION
Adsorption of mouse polyomavirus (PyV) on the host cell surface is mediated by the interaction of its major structural protein, VP1, with sialic acid. Recently, anionic glycosphingolipids GD1a and GT1b, which are heavily glycosylated gangliosides carrying sialic acid residues, were identified as specific receptors for PyV (37). Integrin 41 (also sialyated) has been implicated as a possible coreceptor in mouse cells (9). For simian virus 40 (SV40), another member of the Polyomaviridae, the major histocompatibility complex class I molecule was described as a receptor (8). However, it was later shown that the major histocompatibility complex class I molecule is not endocytosed together with the virus (2). Tsai et al. (37) previously demonstrated that ganglioside GM1 can serve as a functional receptor for SV40. This virus enters cells via caveola invaginations that fuse with larger peripheral organelles (called caveosomes) enriched by caveolin-1. In the steps that followed, SV40 was detected in tubular, caveolin-free membrane vesicles that move along microtubules and deliver virions to the smooth endoplasmic reticulum (ER) (29). The import of SV40 into the ER was found to be brefeldin A sensitive and thus mediated by the ER-Golgi-intermediate compartment represented by COPI-coated vesicles (25, 32). The endocytic pathway exploited by PyV is not completely understood and exhibits both similarities to and differences from that of SV40. PyV is internalized into smooth, monopinocytic vesicles, which fuse with larger peripheral endosomes, often found to contain caveolin-1 (33). However, expression of a dominant-negative mutant of dynamin-1 GTPase required for the formation of caveolae (but also of clathrin-coated vesicles) did not affect polyomavirus infection in some cell types, suggesting that caveolin-1 might not be necessary for PyV uptake (13).
In our previous studies, we observed only rare colocalization of PyV with Rab5 GTPase, which is involved in the regulation of early endosome fusion. At later times postinfection (p.i.), a subpopulation of the virus was found in the perinuclear area of 3T6 fibroblasts colocalizing with Rab11 GTPase and with transferrin, markers of recycling endosomes (20). The movement of the internalized virus is accompanied by transient disorganization of actin stress fibers (15, 33), and the importance of the microtubule cytoskeleton for virus trafficking has also been demonstrated previously (15, 18, 33). At later times after adsorption (approximately 3 h), the signal of PyV, similarly to SV40, was detected in ER cisternae, where it colocalized with ER-resident protein BiP/GRP78 (endoplasmic lumenal chaperone involved in the export of abnormal proteins from the ER to the cytosol) (10). Unlike SV40, PyV infection of 3T6 fibroblasts was not substantially inhibited by brefeldin A, suggesting that PyV does not exploit COPI vesicles employed in retrograde transport from the Golgi apparatus to the ER. Moreover, an alternative retrograde transport from the Golgi apparatus to the ER, used by other intracellular pathogens (e.g., Shiga toxin), also seems not to be utilized by PyV, as no significant colocalization of PyV and Rab6 GTPase (coordinating that retrograde pathway) has been detected. Because PyV bypasses late endosomes that are positive for Rab7 GTPase, it is apparent that the virus can escape degradation in lysosomes (20).
In this study, we applied selected inhibitors and confocal and electron microscopy approaches to further trace the endocytic route used by PyV to establish productive infection.
MATERIALS AND METHODS
Cell line cultivation and virus. Swiss albino mouse cells (NIH 3T6) and normal murine mammary gland (NMuMG) epithelial cells were grown at 37°C in a 5% CO2-air humidified incubator using Dulbecco's modified Eagle's medium (DMEM; Sigma) supplemented with 4 mM L-glutamine and 10% fetal calf serum (FCS). The culture medium of cells expressing green fluorescent protein (GFP) fusion proteins was further supplemented with 0.5 mg/ml G418 (Sigma). Jurkat cells (human leukemic T-lymphoblast cells, clone E6-1, TIB 152; kindly provided by V. Hoejí, IMG AS CR, Prague, Czech Republic) were cultivated in RPMI medium (Sigma) supplemented with 4 mM L-glutamine and 10% FCS. For virus infections, mouse polyomavirus (strain A3) was used at the indicated multiplicities of infection (MOIs).
GFP constructs and cell lines. A plasmid DNA construct expressing GFP-tagged caveolin-1 was a gift from Andre Le Bivic (IBDM, Marseille, France). Caveolin-1-GFP was produced by the fusion of GFP to the C terminus of caveolin-1 in the pEGFP-C2 cloning vector (Clontech). pCINeo/IRES-GFP/cav-1 DN (bicistronic expression vector for dominant-negative caveolin-1 [deletion of amino acids 1 to 81]) was a gift of Jan Eggermont (Katholieke Universiteit, Leuven, Belgium) (36). Constructs expressing GFP-fused mutant Eps15 (DIII2, DIII, and E95/295) cloned into pEGFP-C2 were gifts from Alexandre Benmerah (URA-CNRS, Paris Cedex, France) and were described elsewhere previously (5, 6). For the assay for evaluating the efficiency of PyV infection, vector EGFP-N1 (Clontech) was used as a positive control. Vectors for expression of GFP-fused wild-type Rab5 and a dominant-negative mutant of Rab5 S34N were kindly provided by P. D. Stahl (Washington University School of Medicine) (19). All GFP constructs were transfected to mouse 3T6 fibroblasts or NMuMG cell lines by electroporation. Briefly, exponentially grown cells (1 day after the passage) were trypsinized, and the homogenized cell suspension (1 x 107 cells) was pelleted. Cells were resuspended in 1 ml OPTIMEM-I medium (Gibco) with 10 to 20 μg DNA. Two hundred microliters of the suspension was electroporated using a Gene Pulser apparatus (Bio-Rad) set at 960 μF, 1,000 , and 300 V with a pulse length of 60 to 70 ms. Transfected cells were diluted 1:20 into fresh complete DMEM and grown for 48 h. Stable cell lines expressing GFP-tagged proteins were established by subcloning and maintaining cells upon G418 (Sigma) selection antibiotic in DMEM supplemented with 10% fetal calf serum (Gibco).
Virus. Mouse polyomavirus (A2 strain) was isolated from infected whole mouse embryo primary culture cells according to the standard protocol and purified to homogeneity by CsCl and sucrose gradient ultracentrifugation. The quality of preparation was confirmed by Coomassie blue-stained sodium dodecyl sulfate-acrylamide gel electrophoresis and electron microscopy (EM) (negative staining). The amount of virus particles was estimated by hemagglutination and by protein concentration analysis. For microscopy of living cells, virions were labeled with the red fluorescent marker Alexa-594 coupled with an amine-reactive probe (carboxylic acid succinimidyl ester [purchased from Molecular Probes]) according to the following labeling procedure: purified virus was dialyzed in 0.1 M carbonate buffer (pH 8.3), and 1 mg of the virus with 0.1 mg of the fluorescent reagent was incubated for 1 h at room temperature and then incubated overnight at 4°C. The separation of the conjugate from unreacted labeling reagent was made by extensive dialysis and subsequent purification of the virus on a 10 to 40% sucrose gradient. The virus was aliquoted and stored at –20°C before use. The optimal degree of labeling (ratio of virus to fluorescent marker) was assessed and improved to ensure that natural virus infectiveness was not affected. Briefly, Alexa-594-prestained virus was used to infect cells, and after fixation, coimmunolabeling with anti-VP1 antibody was performed, followed by green Alexa-488 secondary antibody staining. Colocalization of red and green signals proved that all viral particles were conjugated with red Alexa-594 dye, while the VP1 immunoepitope remained available for anti-VP1 antibody binding.
Virus tracking. For live microscopy, cells expressing enhanced GFP-caveolin were grown on 40-mm glass coverslips in phenol red-free DMEM. The cell cycle was synchronized by starving cultured cells in DMEM supplemented with 0.5% serum for 24 h before infection. Coverslips were mounted in an open metal chamber system maintained at 37°C and overlaid with fresh medium. To avoid rapid temperature changes and microtubule depolymerization at 4°C, all procedures were performed at 37°C with prewarmed media and solutions. Virus was diluted with serum-free culture medium and added to cells at an MOI of 100 to 1,000 particles/cell. Unbound virus was gently washed away after 30 min, and complete culture medium was added. Internalization of the virus by the host cells and cytoplasmic transport were monitored at the indicated times postinfection by time lapse live imaging using confocal microscopy. We used a Leica TCS SP2 AOBS confocal microscope operating with an argon laser (458, 476, 488, 496, and 514 nm; 10 mW) and with an HeNe laser (543 and 594 nm; 1 mW). Cells were examined with a 1.2-numerical-aperture water immersion objective (x60). According to the specific signal-to-noise ratio and GFP level of expression, we applied different sampling frequencies (T = 1 to 6 s). Sequential scanning between channels was used to separate fluorescence emission from different fluorochromes and to completely eliminate bleed-through channels. Leica confocal software was used for live microscopy. Video sequences and images were processed by Image J (NIH, Bethesda, MD) and Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA), respectively.
Relevance of caveolin for virus entry. Jurkat cells were infected with either PyV or SV40 (MOI of 3 x 103 virus particles per cell) or incubated with cholera toxin B subunit (CTb) (fluorescein isothiocyanate labeled and diluted to a final concentration of 0.5 μg/ml; Sigma) for 90 min at 37°C and then washed with RPMI-FCS and incubated another 2 h (PyV or SV40) or 1 h (CTb) before fixation. Cells were immunostained for PyV VP1, SV40 VP1, clathrin light-chain subunit, EEA1 marker of early endosomes, or -tubulin. DNA was stained with DAPI (4',6'-diamidino-2-phenylindole). For EM, PyV-infected cells were fixed with 3% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, at the appropriate time p.i., postfixed with 1% osmium tetroxide, dehydrated in graded ethanol solutions, and embedded in epoxy resin AGAR 100 (Grpl, Tuln, Austria) as described previously (33). Ultrastructural analysis was performed on ultrathin sections stained with uranyl acetate and lead citrate. The samples were examined with a JEOL JEM 1200EX electron microscope.
Role of endosomal pH in PyV infection. The cell cycle of 3T6 or NMuMG cells was synchronized by starvation (24-h incubation in DMEM supplemented with 0.5% FCS). Cells were then treated with bafilomycin A1 (0.5 μM) or ammonium chloride (NH4Cl) (1 mM or 5 mM) for a total interval of 4 h, starting 2 h prior virus addition, immediately after adsorption, 2 h postadsorption, or 4 h postadsorption. Virus adsorption to cells was performed within a 30-min interval on ice. Nonadsorbed virus was washed out, DMEM (warmed to 37°C) with 10% FCS (with or without a drug) was added, and cells were then incubated at 37°C in a 5% CO2-air humidified incubator. In the case of drug treatment in the interval of –2 to +2 h postinfection, the adsorption of the virus was performed in the presence of drugs, but the time of adsorption was not included in the 4-h interval. At the end of drug treatment, the cells were washed, incubated for 24 h with freshly added complete DMEM, fixed, and immunostained with antibody against PyV early large T (LT) antigen. Numbers of infected cells were scored by immunofluorescence microscopy.
Immunofluorescence staining. At the indicated times postinfection (MOI of 102 to 103 virus particles per cell), cells grown on coverslips were washed three times with phosphate-buffered saline (PBS), fixed with 3% paraformaldehyde in PBS (30 min), and permeabilized with 0.5% Triton X-100 in PBS (10 min). After washing in PBS, cells were incubated with 0.25% bovine serum albumin and 0.25% porcine skin gelatin in PBS. Immunostaining with primary and secondary antibodies was carried out for 1 h and 30 min, respectively, with extensive washing with PBS after each incubation. The following primary antibodies were used: monoclonal rat anti-PyV LT (kindly provided by B. E. Griffin, Imperial College of Science, Technology, and Medicine at St. Mary's, London, United Kingdom), polyclonal rabbit anti-PyV VP1 (prepared in our laboratory), rabbit anti-SV40 VP1 (kindly provided by H. Kasamatsu, University of California—Los Angeles), mouse anti--tubulin (Exbio), mouse anti-transferrin antibody (Exbio), goat anti-EEA1 (Santa Cruz), mouse Con-1 antibody against clathrin light-chain subunit (24), rabbit polyclonal anti-caveolin-1 (Santa Cruz), rabbit anti-Rab11 (Zymed), and rabbit anti-GFP (Abcam). The following secondary antibodies were used: chicken anti-rabbit, goat anti-rabbit, donkey anti-rat, donkey anti-goat, and donkey anti-mouse conjugated with Alexa Fluor-488; goat anti-mouse conjugated with Alexa Fluor-546; and chicken anti-rabbit conjugated with Alexa Fluor-647 (all purchased from Molecular Probes). For fluorescent resonance energy transfer (FRET) analysis, Cy3-conjugated anti-mouse and Cy5-conjugated anti-rabbit secondary antibodies (purchased from Sigma) were used.
Colocalization assessment of confocal dual-color images. The following two requirements had to be satisfied to consider a given position as colocalization: (i) fluorescent signal-emitting molecules labeled with different fluorochromes should occupy the same pixel in the image, and (ii) the intensity of each component of the image should be within a certain range (in other words, the fluorochrome pairs with very different signal intensities should not be considered as colocalizing). To accommodate these conditions, we studied colocalization on the scatter plots of the signal intensities detected by two separate channels. In these plots, dimmer pixels are located close to the origin, brighter pixels are situated further out, and pure (red and green) pixels are clustered close to the axis. If colocalizing pixels are present, they are displayed (depending on the degree of colocalization) closer to the middle of the plot. We selected a field in the scatter plot that represents pixels in the image where colocalization in both channels occurs. We defined it as a region with its origin in the diagonal of the plot (maximum colocalization) and at coordinates that define the upper limit for the background intensity in the image. In this way, we could mark pixels that colocalize without taking into account pixels with a low intensity value, where the background intensity has high influence and pixels near the axis of the plot, which can represent bleed-through channels and false colocalization. In conclusion, in this mode, we considered pixels as colocalizing only when they were near the diagonal of the plot, and data pairs with very different signal intensities only in the case that both intensities were indeed high.
FRET by acceptor photobleaching method. 3T6 cells grown on coverslips were incubated with PyV (MOI of 102 to 103 virus particles per cell) or with both PyV and transferrin (1 ng per 105 cells) for 30 min on ice before a shift back to 37°C. Cells were fixed for 1.5 or 3 h after the shift and immunostained with anti-VP1- and anti-transferrin- or anti-Rab11-specific antibodies followed by secondary antibodies conjugated with Cy3 and Cy5, respectively. FRET efficiency (Ef) was calculated using the following formula: Ef = (Dpost – Dpre) x 100/Dpost, where Dpost and Dpre are the fluorescence intensities of the donor molecule after and before bleaching of the acceptor, respectively (4). This formula yields the increase in Cy3 fluorescence following Cy5 bleaching normalized by Cy3 fluorescence after the bleaching. We used a Leica SP2 AOBS confocal microscope operating with an HeNe laser tuned to lines at 543 nm (1.2 mW) to excite the Cy3 dye and at 633 nm (10 mW) to excite the Cy5 dye. Cells were examined with a x63 1.4-numerical-aperture oil immersion objective.
RESULTS
PyV utilizes caveolin-1-rich domains for internalization and trafficking. Our previous electron microscopic analysis of polyomavirus entry revealed ultramorphological similarities between PyV and SV40 uptake. We observed that PyV is not associated with clathrin-coated pits at the cell surface and that virion-loaded endocytic vesicles are devoid of the typical clathrin coat (33). In agreement with those previous findings, we now demonstrate that PyV does not colocalize with Con-1 (marker of clathrin-coated pits) at 20 min p.i. (Fig. 1A). The time of fixation and screening was chosen on the basis of a kinetic study of PyV internalization by cells (not shown).
We further tested the effects of epidermal growth factor receptor pathway substrate receptor clone 15 (Eps15) mutants that inhibit clathrin endocytosis. For this purpose, we prepared cell lines expressing the GFP-fused dominant-negative mutants DIII and E95/295 (5, 6), respectively, which compete for binding of the AP-2 clathrin adaptor with endogenous Eps15. The efficiency of infection in these cell lines was compared with that in control, nontransfected cells; cells expressing GFP alone (pEGFP-N1); or cells transfected with the control Eps15 mutant DIII2 (not binding AP-2 -adapting subunit and therefore not affecting the efficiency of clathrin endocytosis). The inhibition of clathrin endocytosis in the mutant cell lines was verified by the analysis of transferrin uptake (not shown). As expected, overexpression of GFP-fused dominant-negative mutant Eps15-DIII or Eps15-E95/295 did not reduce the efficiency of PyV infection (Fig. 1B). Moreover, the efficiency of infection was slightly enhanced in comparison with that assessed in all three control cell lines. A much more significant increase (approximately 100%) in the efficiency of PyV infection was observed in cells overexpressing the wild-type caveolin-1-GFP fusion protein (Fig. 1B).
To visualize a direct involvement of caveolin-1-positive membrane domains in living cells, we established a cell line of mouse epithelial cells expressing GFP-tagged caveolin-1. Live imaging of these cells showed that when caveolin was overexpressed, it still retained its ability to segregate into the surface membrane domains and was also present in intracellular mobile vesicles, pausing at the perinuclear area. Time lapse series have proven that the formation of caveolar invaginations and membrane fission and fusion events were not affected in this cell line (Fig. 2a). We also show here that caveolin-GFP membrane domains are highly dynamic.
When GFP-caveolin cells were infected with fluorescently labeled virus, we observed its movements in directions parallel to the cell surface at the peripheral (apical) cell confocal section. Virions were seen to bypass immobile plasma membrane-anchored caveolar domains (Fig. 2b). However, at least a fraction of endocytosed virions was internalized through caveolin-rich domains (Fig. 2d and e), and a virion(s) captured in the immobile caveolin-1-positive membrane compartment at the nuclear periphery was also observed (Fig. 2c). These observations are in agreement with our hypothesis that mouse PyV can follow alternative trafficking pathways in the host cell.
Caveolin is not necessary for PyV internalization. The evidence for the exploitation of caveolae and/or cavicles (vesicles derived from caveolae) for mouse PyV entry and trafficking is still controversial. Previously, Gilbert and Benjamin (13) did not detect a connection between caveola-mediated endocytosis and polyomavirus uptake in 3T3 fibroblasts and in BMK cells (baby kidney cells). However, later, they showed that when rat glioma C6 cells (which are deficient for complex gangliosides and thus poorly susceptible to PyV infection) were supplemented with GD1a ganglioside, they became infected with PyV with a much higher efficiency. Moreover, the virus entry pathway in these cells was dependent on functional caveola endocytosis (14).
In our previous experiments using different cell lines or different infection conditions, the extent of PyV and caveolin colocalization varied. To address the relevance of caveolae for PyV internalization and trafficking, we monitored the uptake of PyV by Jurkat cells, which do not express caveolin-1 and lack functional caveolae. The absence of caveolin-1 in Jurkat cells was confirmed by Western blot analysis of cell lysates using anti-caveolin-1 and anti-caveolin-2 antibodies (Fig. 3A). As shown on confocal sections of Jurkat cells labeled for tubulin and VP1, PyV virions entered cells, and their signal was found spread in the cytoplasm (Fig. 3B). Electron microscopy of ultrathin cell sections showed that the internalizing invaginations and monopinocytic vesicles were morphologically similar to those found in caveolin-positive fibroblasts or epithelial cell lines (Fig. 3C). In agreement with observations by EM, immunofluorescent staining of Jurkat cells for PyV VP1 and Con-1 (clathrin light-chain subunit) did not detect substantial colocalization of PyV with clathrin (Fig. 3D, panel a). CTb and SV40 (both with binding affinity to ganglioside GM1 as a receptor) were internalized by Jurkat cells, but in contrast to PyV, their signal was found concentrated in the perinuclear area colocalizing with microtubule-organizing center (MTOC) at the same time postinfection (Fig. 3B, panel b). Our experiments demonstrate that PyV, SV40, and cholera toxin, all exploiting caveolar endocytosis to enter cells and join their own specific trafficking route, can also be internalized by cells lacking caveolin through an alternative (clathrin coat-independent and caveolin-independent) endocytic pathway.
Transient expression of a dominant-negative mutant of caveolin-1 in NMuMG cells did not prevent infection by PyV (Fig. 4). The ratio of infected cells to uninfected cells was not significantly diminished for green (caveolin-1 mutant-expressing) cells in comparison with nontransfected control cells.
PyV enters early endosomes. To further investigate the character of compartments through which virions are transported during postendocytic steps, we monitored mutual colocalization of EEA1 (early endosome antigen), PyV, and caveolin-1 in 3T6 and NMuMG cells. While previous experiments performed with Rab5 GTPase (20) showed only rare colocalization with the major PyV capsid protein, VP1, we now observed a high extent of colocalization of EEA1 and VP1 beneath the cell membrane 30 min p.i. (Fig. 5A, panel a). Because a high proportion of entering PyV virions colocalized with GFP-caveolin-1 (Fig. 5A, panel b), we were interested in whether both EEA1 and caveolin-1 overlap, being components of the same early endosome. Figure 5A, panel c, shows a high proportion of colocalization of both signals. Thus, we conclude that after internalization of mouse PyV through membrane raft domains rich in caveolin-1, monopinocytic vesicles fuse with early endosomes (defined by the presence of EEA1) into a common endocytic compartment. Figure 5B demonstrates not only that the presence of EEA1 and caveolin-1 on the same compartments is the result of fusions induced by virus invasion but also that the compartments possessing both caveolin-1 and EEA1 markers occur in noninfected 3T6 fibroblasts. In enlarged details of the merged picture, small endosomes as well as large caveolin-1-positive compartments with distinct patches of EEA1-positive domains (similar to those observed previously by Pelkmans et al. in HeLa cells) (28) can be seen.
PyV infection is dependent upon acidic pH of endosomes. In order to determine whether a transient presence of PyV in the acidic endosomes is necessary for productive infection, we examined the effects of ammonium chloride and bafilomycin A1. NH4Cl penetrates into endosomes and increases endosomal pH, whereas bafilomycin A1 (a specific inhibitor of vacuolar H+ ATPase) prevents acidification of endosomes. Bafilomycin A1 (at a concentration as low as 0.5 μM) completely abolished PyV infection of 3T6 and NMuMG cells when applied 2 h prior to infection and left in medium for an additional 2 h after the addition of the virus (Fig. 6). Treatment by NH4Cl (at a concentration of 1 mM) inhibited infection of 3T6 and NMuMG cells by approximately 85% and 65%, respectively, compared to untreated controls. Virus infectivity diminished with increasing NH4Cl concentrations. The inhibition effect of an elevated endosomal pH was also obvious when NH4Cl or bafilomycin A1 was added to cells immediately after adsorption of the virus. However, no effect on the efficiency of PyV infection was observed when cells were exposed to the agent 2 h (and later) after virus adsorption, at the time when the virus left early endosomal compartments and was transported further to the nuclear periphery.
Furthermore, we tried to determine the particular step in which virus trafficking is affected when acidification of endosomes is blocked. Substantial colocalization of virus and EEA1 was observed in both treated and nontreated cells 20 min postinfection by confocal microscopy (Fig. 7A). However, 4 h postinfection, when the virus exited early endosomes in control, untreated cells, a substantial colocalization of VP1 and EEA1 was still observed in cells treated with bafilomycin A1 or NH4Cl (Fig. 7B).
These data indicate that the acidic milieu of endosomes is necessary for productive PyV infection of the host cells. When pH is increased (or acidification is blocked), the virus is still able to enter early endosomal compartments, but its exit and further trafficking are suppressed. Accordingly, overexpression of the dominant-negative mutant of Rab5 (Rab5S34N) reduced PyV infection markedly, while overexpression of wild-type Rab5 had no inhibition effect (not shown).
PyV and transferrin meet in Rab11-positive endosomes. As we have shown previously (20), PyV VP1 colocalizes with both transferrin and Rab11 GTPase between 1 and 3 h postadsorption. At the same time, partial colocalization of PyV VP1 with the BiP/GRP78 marker of the ER was demonstrated.
Various endosomal compartments, including caveosomes and recycling endosomes, accumulate in the perinuclear space near the MTOC (22, 27). The resolution of a confocal microscope (200 nm) is not sufficient to distinguish whether PyV and transferrin are present in the same endosomal compartment, particularly in such a crowded area of the cytoplasm. To circumvent this limitation, we used FRET, a method to detect protein colocalization at a 2- to 10-nm resolution. The range over which FRET between the donor (Cy3) and acceptor (Cy5) fluorescent molecule occurs is given by the spectral parameter R0, i.e., the distance at which the FRET efficiency is 50%. R0 for the Cy3-Cy5 system is 5 nm (4). Cells were bleached in the Cy5 channel by scanning a region of interest (ROI) using the 633-nm HeNe laser line at 100% intensity. We performed FRET in 15 different cells and bleached more then 60 different ROIs. Before and after the bleaching, Cy3 images were collected to assess changes in donor fluorescence. Figure 8A, panel a, shows the images of the donor (transferrin, Cy3) and the acceptor (VP1 PyV, Cy5) before and after photobleaching in the ROI marked in the figure. Fig. 8A, panel b, presents the intensity value in the ROIs marked on Fig. 8A, panel a.
As a control, we performed a similar calculation in the same number of nonbleached regions of the specimen to evaluate background FRET signals. We calculated the histogram distribution of FRET efficiency for the bleached and nonbleached regions. Distribution of FRET efficiency for bleached regions is positive-shifted from the pseudo-FRET efficiency observed in the nonbleached regions (histogram in Fig. 8A, panel c). The average FRET efficiency values between transferrin and PyV were 9.3 ± 6.9 for the bleached ROIs, which represent the true FRET, and an average value of 2.3 ± 3.2 for the nonbleached ROIs, which represent the background (or false) FRET.
We were further interested in whether the membrane compartments carrying both PyV and transferrin are recycling endosomes. Therefore, we repeated the experiments with Cy3-labeled Rab11 and measured the FRET efficiency of selected ROIs (Fig. 8B). Fig. 8B, panel b, represents intensity values in the ROIs marked on the images (Fig. 8B, panel a), and the FRET efficiency distribution is shown on the histogram (Fig. 8B, panel c). We calculated the FRET efficiency in 60 different ROIs from 14 different cells. An average value of 4.8 ± 2.8 for the bleached ROIs (representing the true FRET) has been assessed. An average value of 2.21 ± 3.6 for background FRET, or false FRET (from the nonbleached ROIs), has been obtained. We obtained very similar background FRET (false FRET) values for both transferrin and Rab11 under the same sample conditions (type of cells, virus, fluorescence donor, and acceptor) and the same acquisition conditions (objective and laser lines). Moreover, in both experiments, we obtained a positive-shifted distribution of FRET efficiency in respect to the nonbleached regions and average values of FRET efficiency higher than those of background FRET. The conclusion can be made that both Rab11 and transferrin exhibit a true FRET and that PyV and transferrin are in intimate contact in a Rab11-positive compartment.
DISCUSSION
Despite intensive studies on the early events of mouse polyomavirus infection, the mechanism of virus uptake and trafficking in the host cell still remains puzzling. Even closely related viruses such as SV40 and human JC virus differ in their modes of entry and subsequent trafficking. JC virus uses the classical endocytic pathway via clathrin-coated pits to enter early endosomes (30), while SV40 is internalized by caveolae (1). Also, another human polyomavirus, BK virus, is dependent on an intact caveolin-1 scaffolding domain to enter the host cell (11). The role of lipid raft microdomains in PyV entry is also unclear. Treatment of cells with cholesterol-depleting agents such as methyl--cyclodextrin, which disrupts detergent-resistant lipid rafts, revealed controversial results, most likely due to the different PyV strains and cell lines used (13, 15, 33). We have previously shown the presence of caveolin-1 on monopinocytic vesicles carrying PyV virions (33). Observations of interactions of PyV with caveolin-1 domains and compartments in living cells revealed that some PyV particles enter cells through caveolin-1-rich domains, while others bypass them, and at later times postinfection, the signal of PyV can be seen in large caveolin-1-rich compartments in the perinuclear space (Fig. 2; see movies at http://www.natur.cuni.cz/molbio/virology/suppl.html). Is caveolin-1 indeed necessary for PyV infection Many toxins that are transported from the cell surface to the ER utilize ganglioside receptors similar to those of polyomaviruses (37). CTb exploits the same GM1 type of ganglioside as SV40 (16). The CTb enters cells via caveolae but can also be efficiently internalized in cells lacking caveolin-1 by a cholesterol-dependent process (26). We show here that in Jurkat cells that do not express caveolin, not only CTb but also SV40 (Fig. 3Bb) was efficiently endocytosed, and both antigens exhibited similar pericentriolar localization. The uptake of PyV occurred as well, but the virus signal remained spread within the cytoplasm. Nevertheless, no colocalization of clathrin with PyV was observed in Jurkat cells, and ultrastructural analysis proved that PyV virions were internalized in smooth, tightly fitting vesicles morphologically similar to those of caveolin-1-positive mouse fibroblast or epithelial cells (33). Moreover, expression of a dominant-negative mutant of caveolin-1 in NMuMG epithelial cells did not prevent their productive infection by PyV.
In mouse fibroblasts or in epithelial cells, virus-loaded monopinocytic vesicles were often shown to fuse with peripheral endosomes early after internalization (33). Therefore, we further examined the characteristics of these compartments. Colocalization of PyV with the early endosome antigen EEA1 has shown that the majority of the virus entered early endosomes. The uptake of virions was slower than that of transferrin, and both cargos rarely met in early endosomes when added simultaneously to the cells. However, at later times postinfection (1.5 to 3 h p.i.), PyV VP1 colocalized with transferrin and Rab11 mostly in perinuclear areas, suggesting that recycling endosomes can be the compartments where they can encounter each other (20). Thus, PyV internalized in caveola-derived vesicles, or perhaps also in other raft-derived vesicles, later enters compartments of the classical clathrin-dependent pathway (e.g., early and/or recycling endosomes) and gets together with a cargo delivered there in a clathrin-dependent manner. Sharma et al. (34) previously demonstrated that glycosphingolipids could be internalized via caveola-related endocytosis and could then rapidly merge with the clathrin pathway in early endosomes. Glycosphingolipids can form distinct microdomains within the early endosomal membrane, which behave differently with respect to their subsequent intracellular trafficking and might play a role in cargo sorting (34).
We further examined the effect of the elevation of endosomal pH on PyV infection. Bafilomycin A1 disrupts the H+ gradient that exists in vesicles of the vacuolar system (7, 39). We proved that, at least in mouse 3T6 fibroblasts and NMuMG epithelial cells, this agent prevents PyV productive infection. Moreover, exposition of these cells to NH4Cl (which rapidly elevates endosomal pH) also inhibited infection by PyV when the cells were exposed to the drugs during the initial steps of infection. At later times postinfection, since the virus signal was detected in the perinuclear space, infection became pH insensitive. Remarkably, previous findings of other groups showed that neither SV40 (35, 38) nor PyV (13) infection was affected by NH4Cl treatment. On the other hand, infection by human JC virus, but also BK virus (internalized by caveolae), was found to be sensitive to the elevation of endosomal pH (3, 11). Ashok and Atwood (3) previously found that JC virus infection is, in contrast to that of SV40, decreased by approximately 70% by NH4Cl, while infection by both SV40 and JC virus can be completely inhibited by bafilomycin A1. The different sensitivities of cells to NH4Cl treatment and the time of exposition could account for the discrepancy between our finding and previous (13) findings. Since under the conditions described previously by Ashok and Atwood (3), our cells died before they could have been analyzed, we shortened the time of NH4Cl treatment by half.
Although both SV40 and cholera toxin bind the same ganglioside GM1 receptor, intoxication by CTb, in contrast to SV40 infection, depends on exposure to an acidic milieu (17). Recently, Pelkmans et al. (28) described a Rab5-dependent pathway in which caveolar vesicles are targeted to early endosomes, where they form distinct but transient membrane domains. When they monitored the SV40 and CTb cargo, the low pH of early endosomes selectively induced a conformational change of CTb (21) and its lateral diffusion into a surrounding membrane subdomain of the compartment, while SV40 remained trapped in the caveolin-positive subdomain from where it was subsequently sorted out into caveosomes. They also found that when caveolin-1 was down-regulated, virions of SV40 diffused into the lumen of endosomes, but that did not lead to infection. This is in agreement with the previous observations that the pathway through caveosomes and the ER is necessary for productive SV40 infection (29). We found out that cells expressing a dominant-negative caveolin-1 mutant can be productively infected by PyV, and in cells with an elevated endosomal pH, the majority of PyV was retained in early endosomes, similar to the CTb that was found to be restricted in its mobility (28). The traffic of PyV is also similar to that of CTb (and differs from SV40) with respect to the dependence on Rab5 function.
Enveloped viruses make use of endosomal acidic pH for inducing the fusion of their envelopes with the vesicle membrane and for nucleocapsid escape to free cytosol. Some nonenveloped viruses, e.g., adenoviruses or human rhinovirus serotype 2, also exploit the acidic milieu of early endosomes to trigger conformational changes in their capsids followed by partial disassembly to evade the barrier of the endosomal membrane (23, 31).
It became obvious that the majority of PyV virions (or their partially disassembled complexes) are further transported in membranous compartments to the perinuclear space. Time lapse microscopy (Fig. 2) showed that PyV was internalized in large caveolin-rich compartments located in the perinuclear space. We showed previously (20) that the signal of major capsid protein VP1 colocalized there with transferrin and Rab11 but also with BiP/GRP78, a marker of the ER. We suggest that similar to cholera or Shiga toxin, a vast proportion of PyV is carried from early endosomes, via recycling endosomes, to the ER. In contrast to bacterial toxins, PyV seems to bypass the Golgi compartment (our unpublished results). Recycling endosomes are mildly acidic compartments containing components of membrane rafts such as glycosylphosphatidylinositol-anchored proteins, sphingomyelin, cholesterol, and caveolin-1 (12). In this regard, they resemble caveosomes utilized by SV40. Moreover, in some cells, both caveosomes and recycling endosomes were found to share an overlapping distribution. Caveosomes immunolabeled with anti-caveolin-1 antibody merged with the transferrin receptor signal (present in recycling endosome membranes) in the perinuclear space near the MTOC of CHO cells. We suggest that caveosomes and recycling endosomes are intertwined rather than that they represent an identical membrane system (22). Nevertheless, positive FRET between VP1 and transferrin and between VP1 and Rab11 provides strong support for the presence of PyV VP1 and transferrin within the same compartment. Further research is required to define functional and biochemical differences between caveosomes and recycling endosomes and to reveal whether both compartments can communicate in a way similar to that of caveola-derived vesicles with early endosomes.
The most intricate question, from where and by what mechanism the PyV genome is delivered into the nucleus, remains unclear and is currently under vigorous investigation. A summary of results obtained by our studies of endocytic pathways exploited by mouse polyomavirus to deliver its genome into the cell nucleus is schematically drawn in Fig. 9.
ACKNOWLEDGMENTS
This work was generously supported by the Grant Agency of the Czech Republic (grant 204/03/0593), by the Centre of Functional Organization of Cells (LC545), by the Centre for New Antivirals and Antineoplastics (1M6138896301), by the programs of the Ministry of Education, Youth, and Sport of the Czech Republic, and by project no. AVOZ50520514.
We are grateful to all researchers (names are given in Materials and Methods) who provided Rab5, Eps15, and caveolin-A mutants or antibodies and to . Takáová for assistance in preparation of the manuscript.
REFERENCES
Anderson, H. A., Y. Chen, and L. C. Norkin. 1996. Bound simian virus 40 translocates to caveolin-enriched membrane domains, and its entry is inhibited by drugs that selectively disrupt caveolae. Mol. Biol. Cell 7:1825-1834.
Anderson, H. A., Y. Chen, and L. C. Norkin. 1998. MHC class I molecules are enriched in caveolae but do not enter with simian virus 40. J. Gen. Virol. 79:1469-1477.
Ashok, A., and W. J. Atwood. 2003. Contrasting roles of endosomal pH and the cytoskeleton in infection of human glial cells by JC virus and simian virus 40. J. Virol. 77:1347-1356.
Bastiaens, P. I. H., and T. M. Jovin. 1998. Fluorescence resonance energy transfer microscopy, p. 136-146. In J. E. Celis (ed.), Cell biology: a laboratory handbook, 2nd ed., vol. 3. Academic Press, New York, N.Y.
Benmerah, A., M. Bayrou, N. Cerf-Bensussan, and A. Dautry-Varsat. 1999. Inhibition of clathrin-coated pit assembly by an Eps15 mutant. J. Cell Sci. 112:1303-1311.
Benmarah, A., C. Lamaze, B. Begue, S. L. Schmid, A. Dautry-Varsat, and N. Cerf-Bensussan. 1998. AP-2/Eps15 interaction is required for receptor-mediated endocytosis. J. Cell Biol. 140:1055-1062.
Bowman, E. J., A. Siebers, and K. Altendorf. 1988. Bafilomycins: a class of inhibitors of membrane ATPases from microorganisms, animal cells, and plant cells. Proc. Natl. Acad. Sci. USA 85:7972-7976.
Breau, W. C., W. J. Atwood, and L. C. Norkin. 1992. Class I major histocompatibility proteins are an essential component of the simian virus 40 receptor. J. Virol. 66:2037-2045.
Caruso, M., L. Belloni, O. Sthandier, P. Amati, and M.-I. Garcia. 2003. 41 integrin acts as a cell receptor for murine polyomavirus at the postattachment level. J. Virol. 77:3913-3921.
Deeks, E. D., J. P. Cook, P. J. Day, D. C. Smith, L. M. Roberts, and J. M. Lord. 2002. The low lysine content of ricin A chain reduces the risk of proteolytic degradation after translocation from the endoplasmic reticulum to the cytosol. Biochemistry 41:3405-3413.
Eash, S., W. Querbes, and W. J. Atwood. 2004. Infection of Vero cells by BK virus is dependent on caveolae. J. Virol. 78:11583-11590.
Gagescu, R., N. Demaurex, R. G. Parton, W. Hunziger, L. A. Huber, and J. Gruenberg. 2000. The recycling endosome of Madin-Darby canine kidney cells is a mildly acidic compartment rich in raft components. Mol. Biol. Cell 11:2775-2791.
Gilbert, J. M., and T. L. Benjamin. 2000. Early steps of polyomavirus entry into cells. J. Virol. 74:8582-8588.
Gilbert, J. M., J. Dahl, J. You, C. Vui, R. Holmes, W. Lencer, and T. L. Benjamin. 2005. Ganglioside GD1a restores infectibility to mouse cells lacking functional receptors for polyomavirus. J. Virol. 79:615-618.
Gilbert, J. M., I. G. Goldberg, and T. L. Benjamin. 2003. Cell penetration and trafficking of polyomavirus. J. Virol. 77:2615-2622.
Griffiths, S. L., R. A. Finkelstein, and D. R. Critchley. 1986. Characterization of the receptor for cholera toxin and Escherichia coli heat-labile toxin in rabbit intestinal brush borders. Biochem. J. 238:313-322.
Janicot, M., F. Fouque, and B. Desbuquois. 1991. Activation of rat liver adenylate cyclase by cholera toxin requires toxin internalization and processing in endosomes. J. Biol. Chem. 266:12858-12865.
Krauzewicz, N., J. tokrová, C. Jenkins, M. Elliott, C. F. Higgins, and B. E. Griffin. 2000. Virus-like gene transfer into cells mediated by polyoma virus pseudocapsids. Gene Ther. 7:2122-2131.
Li, G., and P. D. Stahl. 1993. Structure-function relationship of the small GTPase rab5. J. Biol. Chem. 268:24445-24480.
Mannová, P., and J. Forstová. 2003. Mouse polyomavirus utilizes recycling endosomes for a traffic pathway independent of COPI vesicle transport. J. Virol. 77:1672-1681.
McCann, J. A., J. A. Mertz, J. Czworkowski, and W. D. Pickin. 1997. Conformational changes in cholera toxin B subunit-ganglioside GM1 complexes are elicited by environmental pH and evoke changes in membrane structure. Biochemistry 36:9169-9178.
Mundy, D. I., T. Machleidt, Y. S. Ying, R. G. Anderson, and G. S. Bloom. 2002. Dual control of caveolar membrane traffic by microtubules and the actin cytoskeleton. J. Cell Sci. 115:4327-4339.
Nakano, M. Y., K. Boucke, M. Suomalainen, P. Stidwell, and U. G. Grebe. 2000. The first step of adenovirus type 2 disassembly occurs at the cell surface, independently of endocytosis and escape to the cytosol. J. Virol. 74:7085-7095.
Nathke, I. S., J. Heuser, A. Lupas, J. Stock, C. W. Turck, and F. M. Brodsky. 1992. Folding and trimerization of clathrin subunits at the triskelion hub. Cell 68:899-910.
Norkin, L. C., H. A. Anderson, W. A. Scott, and A. Oppenheim. 2002. Caveolar endocytosis of simian virus 40 is followed by brefeldin A-sensitive transport to the endoplasmic reticulum, where the virus disassembles. J. Virol. 76:5156-5166.
Orlandi, P. A., and P. H. Fishman. 1998. Filipin dependent inhibition of cholera toxin: evidence for toxin internalisation and activation through caveolae like domains. J. Cell Biol. 141:905-915.
Pasqualato, S., F. Senic-Matuglia, L. Renault, B. Goud, J. Salamero, and J. Cherfils. 2004. The structural GDP/GTP cycle of Rab11 reveals a novel interface involved in the dynamics of recycling endosomes. J. Biol. Chem. 279:11480-11488.
Pelkmans, L., T. Burli, M. Zerial, and A. Helenius. 2004. Caveolin-stabilized membrane domains as multifunctional transport and sorting devices in endocytic membrane traffic. Cell 118:767-780.
Pelkmans, L., J. Kartenbeck, and A. Helenius. 2001. Caveolar endocytosis of simian virus 40 revealed a new two step vesicular transport pathway to the endoplasmic reticulum. Nat. Cell Biol. 3:473-483.
Pho, M. T., A. Ashok, and W. J. Atwood. 2000. JC virus enters human glial cells by clathrin-dependent, receptor-mediated endocytosis. J. Virol. 74:2288-2292.
Prchla, E., C. Plank, E. Wagner, D. Blaas, and R. Fuchs. 1995. Virus-mediated release of endosomal content in vitro: different behavior of adenovirus and rhinovirus serotype 2. J. Cell Biol. 131:111-123.
Richards, A. A., E. Stang, R. Pepperkok, and R. G. Parton. 2002. Inhibitors of COP-mediated transport and cholera toxin action inhibit simian virus 40 infection. Mol. Biol. Cell 13:1750-1764.
Richterová, Z., D. Liebl, M. Horák, Z. Palková, J. tokrová, P. Hozák, J. Korb, and J. Forstová. 2001. Caveolae are involved in the trafficking of mouse polyomavirus virions and artificial VP1 pseudocapsids toward cell nuclei. J. Virol. 75:10880-10891.
Sharma, D. K., A. Choudhury, R. D. Singh, C. L. Wheatley, D. L. Marks, and R. E. Paganos. 2003. Glycosphingolipids internalized via caveolar-related endocytosis rapidly merge with the clathrin pathway in early endosomes and form microdomains for recycling. J. Biol. Chem. 278:7564-7572.
Shimura, H., Y. Umeno, and G. Kimura. 1987. Effects of inhibitors of the cytoplasmic structures and functions on the early phase of infection of cultured cells with simian virus 40. Virology 158:34-43.
Trouet, D., D. Hermans, G. Droogmans, B. Nilus, and J. Eggermont. 2001. Inhibition of volume-regulated anion channels by dominant-negative caveolin-1. Biochem. Biophys. Res. Commun. 284:461-465.
Tsai, B., J. M. Gilbert, S. Stehle, W. Lencer, T. L. Benjamin, and T. A. Rapoport. 2003. Gangliosides are receptors for murine polyoma virus and SV40. EMBO J. 22:4346-4355.
Upcroft, P. 1987. Simian virus 40 infection is not mediated by lysosomal activation. J. Gen. Virol. 68:2477-2480.
Yochimori, T., A. Yamamoto, Y. Moriyama, M. Futai, and Y. Tashiro. 1991. Bafilomycin A1, a specific inhibitor of vacuolar type H+ ATPase, inhibits acidification and protein degradation in lysosomes of cultured cells. J. Biol. Chem. 266:17707-17712.(David Liebl, Francesco Di)
Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, Prague, Czech Republic
ABSTRACT
Mouse polyomavirus (PyV) virions enter cells by internalization into smooth monopinocytic vesicles, which fuse under the cell membrane with larger endosomes. Caveolin-1 was detected on monopinocytic vesicles carrying PyV particles in mouse fibroblasts and epithelial cells (33). Here, we show that PyV can be efficiently internalized by Jurkat cells, which do not express caveolin-1 and lack caveolae, and that overexpression of a caveolin-1 dominant-negative mutant in mouse epithelial cells does not prevent their productive infection. Strong colocalization of VP1 with early endosome antigen 1 (EEA1) and of EEA1 with caveolin-1 in mouse fibroblasts and epithelial cells suggests that the monopinocytic vesicles carrying the virus (and vesicles containing caveolin-1) fuse with EEA1-positive early endosomes. In contrast to SV40, PyV infection is dependent on the acidic pH of endosomes. Bafilomycin A1 abolished PyV infection, and an increase in endosomal pH by NH4Cl markedly reduced its efficiency when drugs were applied during virion transport towards the cell nucleus. The block of acidification resulted in the retention of a fraction of virions in early endosomes. To monitor further trafficking of PyV, we used fluorescent resonance energy transfer (FRET) to determine mutual localization of PyV VP1 with transferrin and Rab11 GTPase at a 2- to 10-nm resolution. Positive FRET between PyV VP1 and transferrin cargo and between PyV VP1 and Rab11 suggests that during later times postinfection (1.5 to 3 h), the virus meets up with transferrin in the Rab11-positive recycling endosome. These results point to a convergence of the virus and the cargo internalized by different pathways in common transitional compartments.
INTRODUCTION
Adsorption of mouse polyomavirus (PyV) on the host cell surface is mediated by the interaction of its major structural protein, VP1, with sialic acid. Recently, anionic glycosphingolipids GD1a and GT1b, which are heavily glycosylated gangliosides carrying sialic acid residues, were identified as specific receptors for PyV (37). Integrin 41 (also sialyated) has been implicated as a possible coreceptor in mouse cells (9). For simian virus 40 (SV40), another member of the Polyomaviridae, the major histocompatibility complex class I molecule was described as a receptor (8). However, it was later shown that the major histocompatibility complex class I molecule is not endocytosed together with the virus (2). Tsai et al. (37) previously demonstrated that ganglioside GM1 can serve as a functional receptor for SV40. This virus enters cells via caveola invaginations that fuse with larger peripheral organelles (called caveosomes) enriched by caveolin-1. In the steps that followed, SV40 was detected in tubular, caveolin-free membrane vesicles that move along microtubules and deliver virions to the smooth endoplasmic reticulum (ER) (29). The import of SV40 into the ER was found to be brefeldin A sensitive and thus mediated by the ER-Golgi-intermediate compartment represented by COPI-coated vesicles (25, 32). The endocytic pathway exploited by PyV is not completely understood and exhibits both similarities to and differences from that of SV40. PyV is internalized into smooth, monopinocytic vesicles, which fuse with larger peripheral endosomes, often found to contain caveolin-1 (33). However, expression of a dominant-negative mutant of dynamin-1 GTPase required for the formation of caveolae (but also of clathrin-coated vesicles) did not affect polyomavirus infection in some cell types, suggesting that caveolin-1 might not be necessary for PyV uptake (13).
In our previous studies, we observed only rare colocalization of PyV with Rab5 GTPase, which is involved in the regulation of early endosome fusion. At later times postinfection (p.i.), a subpopulation of the virus was found in the perinuclear area of 3T6 fibroblasts colocalizing with Rab11 GTPase and with transferrin, markers of recycling endosomes (20). The movement of the internalized virus is accompanied by transient disorganization of actin stress fibers (15, 33), and the importance of the microtubule cytoskeleton for virus trafficking has also been demonstrated previously (15, 18, 33). At later times after adsorption (approximately 3 h), the signal of PyV, similarly to SV40, was detected in ER cisternae, where it colocalized with ER-resident protein BiP/GRP78 (endoplasmic lumenal chaperone involved in the export of abnormal proteins from the ER to the cytosol) (10). Unlike SV40, PyV infection of 3T6 fibroblasts was not substantially inhibited by brefeldin A, suggesting that PyV does not exploit COPI vesicles employed in retrograde transport from the Golgi apparatus to the ER. Moreover, an alternative retrograde transport from the Golgi apparatus to the ER, used by other intracellular pathogens (e.g., Shiga toxin), also seems not to be utilized by PyV, as no significant colocalization of PyV and Rab6 GTPase (coordinating that retrograde pathway) has been detected. Because PyV bypasses late endosomes that are positive for Rab7 GTPase, it is apparent that the virus can escape degradation in lysosomes (20).
In this study, we applied selected inhibitors and confocal and electron microscopy approaches to further trace the endocytic route used by PyV to establish productive infection.
MATERIALS AND METHODS
Cell line cultivation and virus. Swiss albino mouse cells (NIH 3T6) and normal murine mammary gland (NMuMG) epithelial cells were grown at 37°C in a 5% CO2-air humidified incubator using Dulbecco's modified Eagle's medium (DMEM; Sigma) supplemented with 4 mM L-glutamine and 10% fetal calf serum (FCS). The culture medium of cells expressing green fluorescent protein (GFP) fusion proteins was further supplemented with 0.5 mg/ml G418 (Sigma). Jurkat cells (human leukemic T-lymphoblast cells, clone E6-1, TIB 152; kindly provided by V. Hoejí, IMG AS CR, Prague, Czech Republic) were cultivated in RPMI medium (Sigma) supplemented with 4 mM L-glutamine and 10% FCS. For virus infections, mouse polyomavirus (strain A3) was used at the indicated multiplicities of infection (MOIs).
GFP constructs and cell lines. A plasmid DNA construct expressing GFP-tagged caveolin-1 was a gift from Andre Le Bivic (IBDM, Marseille, France). Caveolin-1-GFP was produced by the fusion of GFP to the C terminus of caveolin-1 in the pEGFP-C2 cloning vector (Clontech). pCINeo/IRES-GFP/cav-1 DN (bicistronic expression vector for dominant-negative caveolin-1 [deletion of amino acids 1 to 81]) was a gift of Jan Eggermont (Katholieke Universiteit, Leuven, Belgium) (36). Constructs expressing GFP-fused mutant Eps15 (DIII2, DIII, and E95/295) cloned into pEGFP-C2 were gifts from Alexandre Benmerah (URA-CNRS, Paris Cedex, France) and were described elsewhere previously (5, 6). For the assay for evaluating the efficiency of PyV infection, vector EGFP-N1 (Clontech) was used as a positive control. Vectors for expression of GFP-fused wild-type Rab5 and a dominant-negative mutant of Rab5 S34N were kindly provided by P. D. Stahl (Washington University School of Medicine) (19). All GFP constructs were transfected to mouse 3T6 fibroblasts or NMuMG cell lines by electroporation. Briefly, exponentially grown cells (1 day after the passage) were trypsinized, and the homogenized cell suspension (1 x 107 cells) was pelleted. Cells were resuspended in 1 ml OPTIMEM-I medium (Gibco) with 10 to 20 μg DNA. Two hundred microliters of the suspension was electroporated using a Gene Pulser apparatus (Bio-Rad) set at 960 μF, 1,000 , and 300 V with a pulse length of 60 to 70 ms. Transfected cells were diluted 1:20 into fresh complete DMEM and grown for 48 h. Stable cell lines expressing GFP-tagged proteins were established by subcloning and maintaining cells upon G418 (Sigma) selection antibiotic in DMEM supplemented with 10% fetal calf serum (Gibco).
Virus. Mouse polyomavirus (A2 strain) was isolated from infected whole mouse embryo primary culture cells according to the standard protocol and purified to homogeneity by CsCl and sucrose gradient ultracentrifugation. The quality of preparation was confirmed by Coomassie blue-stained sodium dodecyl sulfate-acrylamide gel electrophoresis and electron microscopy (EM) (negative staining). The amount of virus particles was estimated by hemagglutination and by protein concentration analysis. For microscopy of living cells, virions were labeled with the red fluorescent marker Alexa-594 coupled with an amine-reactive probe (carboxylic acid succinimidyl ester [purchased from Molecular Probes]) according to the following labeling procedure: purified virus was dialyzed in 0.1 M carbonate buffer (pH 8.3), and 1 mg of the virus with 0.1 mg of the fluorescent reagent was incubated for 1 h at room temperature and then incubated overnight at 4°C. The separation of the conjugate from unreacted labeling reagent was made by extensive dialysis and subsequent purification of the virus on a 10 to 40% sucrose gradient. The virus was aliquoted and stored at –20°C before use. The optimal degree of labeling (ratio of virus to fluorescent marker) was assessed and improved to ensure that natural virus infectiveness was not affected. Briefly, Alexa-594-prestained virus was used to infect cells, and after fixation, coimmunolabeling with anti-VP1 antibody was performed, followed by green Alexa-488 secondary antibody staining. Colocalization of red and green signals proved that all viral particles were conjugated with red Alexa-594 dye, while the VP1 immunoepitope remained available for anti-VP1 antibody binding.
Virus tracking. For live microscopy, cells expressing enhanced GFP-caveolin were grown on 40-mm glass coverslips in phenol red-free DMEM. The cell cycle was synchronized by starving cultured cells in DMEM supplemented with 0.5% serum for 24 h before infection. Coverslips were mounted in an open metal chamber system maintained at 37°C and overlaid with fresh medium. To avoid rapid temperature changes and microtubule depolymerization at 4°C, all procedures were performed at 37°C with prewarmed media and solutions. Virus was diluted with serum-free culture medium and added to cells at an MOI of 100 to 1,000 particles/cell. Unbound virus was gently washed away after 30 min, and complete culture medium was added. Internalization of the virus by the host cells and cytoplasmic transport were monitored at the indicated times postinfection by time lapse live imaging using confocal microscopy. We used a Leica TCS SP2 AOBS confocal microscope operating with an argon laser (458, 476, 488, 496, and 514 nm; 10 mW) and with an HeNe laser (543 and 594 nm; 1 mW). Cells were examined with a 1.2-numerical-aperture water immersion objective (x60). According to the specific signal-to-noise ratio and GFP level of expression, we applied different sampling frequencies (T = 1 to 6 s). Sequential scanning between channels was used to separate fluorescence emission from different fluorochromes and to completely eliminate bleed-through channels. Leica confocal software was used for live microscopy. Video sequences and images were processed by Image J (NIH, Bethesda, MD) and Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA), respectively.
Relevance of caveolin for virus entry. Jurkat cells were infected with either PyV or SV40 (MOI of 3 x 103 virus particles per cell) or incubated with cholera toxin B subunit (CTb) (fluorescein isothiocyanate labeled and diluted to a final concentration of 0.5 μg/ml; Sigma) for 90 min at 37°C and then washed with RPMI-FCS and incubated another 2 h (PyV or SV40) or 1 h (CTb) before fixation. Cells were immunostained for PyV VP1, SV40 VP1, clathrin light-chain subunit, EEA1 marker of early endosomes, or -tubulin. DNA was stained with DAPI (4',6'-diamidino-2-phenylindole). For EM, PyV-infected cells were fixed with 3% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, at the appropriate time p.i., postfixed with 1% osmium tetroxide, dehydrated in graded ethanol solutions, and embedded in epoxy resin AGAR 100 (Grpl, Tuln, Austria) as described previously (33). Ultrastructural analysis was performed on ultrathin sections stained with uranyl acetate and lead citrate. The samples were examined with a JEOL JEM 1200EX electron microscope.
Role of endosomal pH in PyV infection. The cell cycle of 3T6 or NMuMG cells was synchronized by starvation (24-h incubation in DMEM supplemented with 0.5% FCS). Cells were then treated with bafilomycin A1 (0.5 μM) or ammonium chloride (NH4Cl) (1 mM or 5 mM) for a total interval of 4 h, starting 2 h prior virus addition, immediately after adsorption, 2 h postadsorption, or 4 h postadsorption. Virus adsorption to cells was performed within a 30-min interval on ice. Nonadsorbed virus was washed out, DMEM (warmed to 37°C) with 10% FCS (with or without a drug) was added, and cells were then incubated at 37°C in a 5% CO2-air humidified incubator. In the case of drug treatment in the interval of –2 to +2 h postinfection, the adsorption of the virus was performed in the presence of drugs, but the time of adsorption was not included in the 4-h interval. At the end of drug treatment, the cells were washed, incubated for 24 h with freshly added complete DMEM, fixed, and immunostained with antibody against PyV early large T (LT) antigen. Numbers of infected cells were scored by immunofluorescence microscopy.
Immunofluorescence staining. At the indicated times postinfection (MOI of 102 to 103 virus particles per cell), cells grown on coverslips were washed three times with phosphate-buffered saline (PBS), fixed with 3% paraformaldehyde in PBS (30 min), and permeabilized with 0.5% Triton X-100 in PBS (10 min). After washing in PBS, cells were incubated with 0.25% bovine serum albumin and 0.25% porcine skin gelatin in PBS. Immunostaining with primary and secondary antibodies was carried out for 1 h and 30 min, respectively, with extensive washing with PBS after each incubation. The following primary antibodies were used: monoclonal rat anti-PyV LT (kindly provided by B. E. Griffin, Imperial College of Science, Technology, and Medicine at St. Mary's, London, United Kingdom), polyclonal rabbit anti-PyV VP1 (prepared in our laboratory), rabbit anti-SV40 VP1 (kindly provided by H. Kasamatsu, University of California—Los Angeles), mouse anti--tubulin (Exbio), mouse anti-transferrin antibody (Exbio), goat anti-EEA1 (Santa Cruz), mouse Con-1 antibody against clathrin light-chain subunit (24), rabbit polyclonal anti-caveolin-1 (Santa Cruz), rabbit anti-Rab11 (Zymed), and rabbit anti-GFP (Abcam). The following secondary antibodies were used: chicken anti-rabbit, goat anti-rabbit, donkey anti-rat, donkey anti-goat, and donkey anti-mouse conjugated with Alexa Fluor-488; goat anti-mouse conjugated with Alexa Fluor-546; and chicken anti-rabbit conjugated with Alexa Fluor-647 (all purchased from Molecular Probes). For fluorescent resonance energy transfer (FRET) analysis, Cy3-conjugated anti-mouse and Cy5-conjugated anti-rabbit secondary antibodies (purchased from Sigma) were used.
Colocalization assessment of confocal dual-color images. The following two requirements had to be satisfied to consider a given position as colocalization: (i) fluorescent signal-emitting molecules labeled with different fluorochromes should occupy the same pixel in the image, and (ii) the intensity of each component of the image should be within a certain range (in other words, the fluorochrome pairs with very different signal intensities should not be considered as colocalizing). To accommodate these conditions, we studied colocalization on the scatter plots of the signal intensities detected by two separate channels. In these plots, dimmer pixels are located close to the origin, brighter pixels are situated further out, and pure (red and green) pixels are clustered close to the axis. If colocalizing pixels are present, they are displayed (depending on the degree of colocalization) closer to the middle of the plot. We selected a field in the scatter plot that represents pixels in the image where colocalization in both channels occurs. We defined it as a region with its origin in the diagonal of the plot (maximum colocalization) and at coordinates that define the upper limit for the background intensity in the image. In this way, we could mark pixels that colocalize without taking into account pixels with a low intensity value, where the background intensity has high influence and pixels near the axis of the plot, which can represent bleed-through channels and false colocalization. In conclusion, in this mode, we considered pixels as colocalizing only when they were near the diagonal of the plot, and data pairs with very different signal intensities only in the case that both intensities were indeed high.
FRET by acceptor photobleaching method. 3T6 cells grown on coverslips were incubated with PyV (MOI of 102 to 103 virus particles per cell) or with both PyV and transferrin (1 ng per 105 cells) for 30 min on ice before a shift back to 37°C. Cells were fixed for 1.5 or 3 h after the shift and immunostained with anti-VP1- and anti-transferrin- or anti-Rab11-specific antibodies followed by secondary antibodies conjugated with Cy3 and Cy5, respectively. FRET efficiency (Ef) was calculated using the following formula: Ef = (Dpost – Dpre) x 100/Dpost, where Dpost and Dpre are the fluorescence intensities of the donor molecule after and before bleaching of the acceptor, respectively (4). This formula yields the increase in Cy3 fluorescence following Cy5 bleaching normalized by Cy3 fluorescence after the bleaching. We used a Leica SP2 AOBS confocal microscope operating with an HeNe laser tuned to lines at 543 nm (1.2 mW) to excite the Cy3 dye and at 633 nm (10 mW) to excite the Cy5 dye. Cells were examined with a x63 1.4-numerical-aperture oil immersion objective.
RESULTS
PyV utilizes caveolin-1-rich domains for internalization and trafficking. Our previous electron microscopic analysis of polyomavirus entry revealed ultramorphological similarities between PyV and SV40 uptake. We observed that PyV is not associated with clathrin-coated pits at the cell surface and that virion-loaded endocytic vesicles are devoid of the typical clathrin coat (33). In agreement with those previous findings, we now demonstrate that PyV does not colocalize with Con-1 (marker of clathrin-coated pits) at 20 min p.i. (Fig. 1A). The time of fixation and screening was chosen on the basis of a kinetic study of PyV internalization by cells (not shown).
We further tested the effects of epidermal growth factor receptor pathway substrate receptor clone 15 (Eps15) mutants that inhibit clathrin endocytosis. For this purpose, we prepared cell lines expressing the GFP-fused dominant-negative mutants DIII and E95/295 (5, 6), respectively, which compete for binding of the AP-2 clathrin adaptor with endogenous Eps15. The efficiency of infection in these cell lines was compared with that in control, nontransfected cells; cells expressing GFP alone (pEGFP-N1); or cells transfected with the control Eps15 mutant DIII2 (not binding AP-2 -adapting subunit and therefore not affecting the efficiency of clathrin endocytosis). The inhibition of clathrin endocytosis in the mutant cell lines was verified by the analysis of transferrin uptake (not shown). As expected, overexpression of GFP-fused dominant-negative mutant Eps15-DIII or Eps15-E95/295 did not reduce the efficiency of PyV infection (Fig. 1B). Moreover, the efficiency of infection was slightly enhanced in comparison with that assessed in all three control cell lines. A much more significant increase (approximately 100%) in the efficiency of PyV infection was observed in cells overexpressing the wild-type caveolin-1-GFP fusion protein (Fig. 1B).
To visualize a direct involvement of caveolin-1-positive membrane domains in living cells, we established a cell line of mouse epithelial cells expressing GFP-tagged caveolin-1. Live imaging of these cells showed that when caveolin was overexpressed, it still retained its ability to segregate into the surface membrane domains and was also present in intracellular mobile vesicles, pausing at the perinuclear area. Time lapse series have proven that the formation of caveolar invaginations and membrane fission and fusion events were not affected in this cell line (Fig. 2a). We also show here that caveolin-GFP membrane domains are highly dynamic.
When GFP-caveolin cells were infected with fluorescently labeled virus, we observed its movements in directions parallel to the cell surface at the peripheral (apical) cell confocal section. Virions were seen to bypass immobile plasma membrane-anchored caveolar domains (Fig. 2b). However, at least a fraction of endocytosed virions was internalized through caveolin-rich domains (Fig. 2d and e), and a virion(s) captured in the immobile caveolin-1-positive membrane compartment at the nuclear periphery was also observed (Fig. 2c). These observations are in agreement with our hypothesis that mouse PyV can follow alternative trafficking pathways in the host cell.
Caveolin is not necessary for PyV internalization. The evidence for the exploitation of caveolae and/or cavicles (vesicles derived from caveolae) for mouse PyV entry and trafficking is still controversial. Previously, Gilbert and Benjamin (13) did not detect a connection between caveola-mediated endocytosis and polyomavirus uptake in 3T3 fibroblasts and in BMK cells (baby kidney cells). However, later, they showed that when rat glioma C6 cells (which are deficient for complex gangliosides and thus poorly susceptible to PyV infection) were supplemented with GD1a ganglioside, they became infected with PyV with a much higher efficiency. Moreover, the virus entry pathway in these cells was dependent on functional caveola endocytosis (14).
In our previous experiments using different cell lines or different infection conditions, the extent of PyV and caveolin colocalization varied. To address the relevance of caveolae for PyV internalization and trafficking, we monitored the uptake of PyV by Jurkat cells, which do not express caveolin-1 and lack functional caveolae. The absence of caveolin-1 in Jurkat cells was confirmed by Western blot analysis of cell lysates using anti-caveolin-1 and anti-caveolin-2 antibodies (Fig. 3A). As shown on confocal sections of Jurkat cells labeled for tubulin and VP1, PyV virions entered cells, and their signal was found spread in the cytoplasm (Fig. 3B). Electron microscopy of ultrathin cell sections showed that the internalizing invaginations and monopinocytic vesicles were morphologically similar to those found in caveolin-positive fibroblasts or epithelial cell lines (Fig. 3C). In agreement with observations by EM, immunofluorescent staining of Jurkat cells for PyV VP1 and Con-1 (clathrin light-chain subunit) did not detect substantial colocalization of PyV with clathrin (Fig. 3D, panel a). CTb and SV40 (both with binding affinity to ganglioside GM1 as a receptor) were internalized by Jurkat cells, but in contrast to PyV, their signal was found concentrated in the perinuclear area colocalizing with microtubule-organizing center (MTOC) at the same time postinfection (Fig. 3B, panel b). Our experiments demonstrate that PyV, SV40, and cholera toxin, all exploiting caveolar endocytosis to enter cells and join their own specific trafficking route, can also be internalized by cells lacking caveolin through an alternative (clathrin coat-independent and caveolin-independent) endocytic pathway.
Transient expression of a dominant-negative mutant of caveolin-1 in NMuMG cells did not prevent infection by PyV (Fig. 4). The ratio of infected cells to uninfected cells was not significantly diminished for green (caveolin-1 mutant-expressing) cells in comparison with nontransfected control cells.
PyV enters early endosomes. To further investigate the character of compartments through which virions are transported during postendocytic steps, we monitored mutual colocalization of EEA1 (early endosome antigen), PyV, and caveolin-1 in 3T6 and NMuMG cells. While previous experiments performed with Rab5 GTPase (20) showed only rare colocalization with the major PyV capsid protein, VP1, we now observed a high extent of colocalization of EEA1 and VP1 beneath the cell membrane 30 min p.i. (Fig. 5A, panel a). Because a high proportion of entering PyV virions colocalized with GFP-caveolin-1 (Fig. 5A, panel b), we were interested in whether both EEA1 and caveolin-1 overlap, being components of the same early endosome. Figure 5A, panel c, shows a high proportion of colocalization of both signals. Thus, we conclude that after internalization of mouse PyV through membrane raft domains rich in caveolin-1, monopinocytic vesicles fuse with early endosomes (defined by the presence of EEA1) into a common endocytic compartment. Figure 5B demonstrates not only that the presence of EEA1 and caveolin-1 on the same compartments is the result of fusions induced by virus invasion but also that the compartments possessing both caveolin-1 and EEA1 markers occur in noninfected 3T6 fibroblasts. In enlarged details of the merged picture, small endosomes as well as large caveolin-1-positive compartments with distinct patches of EEA1-positive domains (similar to those observed previously by Pelkmans et al. in HeLa cells) (28) can be seen.
PyV infection is dependent upon acidic pH of endosomes. In order to determine whether a transient presence of PyV in the acidic endosomes is necessary for productive infection, we examined the effects of ammonium chloride and bafilomycin A1. NH4Cl penetrates into endosomes and increases endosomal pH, whereas bafilomycin A1 (a specific inhibitor of vacuolar H+ ATPase) prevents acidification of endosomes. Bafilomycin A1 (at a concentration as low as 0.5 μM) completely abolished PyV infection of 3T6 and NMuMG cells when applied 2 h prior to infection and left in medium for an additional 2 h after the addition of the virus (Fig. 6). Treatment by NH4Cl (at a concentration of 1 mM) inhibited infection of 3T6 and NMuMG cells by approximately 85% and 65%, respectively, compared to untreated controls. Virus infectivity diminished with increasing NH4Cl concentrations. The inhibition effect of an elevated endosomal pH was also obvious when NH4Cl or bafilomycin A1 was added to cells immediately after adsorption of the virus. However, no effect on the efficiency of PyV infection was observed when cells were exposed to the agent 2 h (and later) after virus adsorption, at the time when the virus left early endosomal compartments and was transported further to the nuclear periphery.
Furthermore, we tried to determine the particular step in which virus trafficking is affected when acidification of endosomes is blocked. Substantial colocalization of virus and EEA1 was observed in both treated and nontreated cells 20 min postinfection by confocal microscopy (Fig. 7A). However, 4 h postinfection, when the virus exited early endosomes in control, untreated cells, a substantial colocalization of VP1 and EEA1 was still observed in cells treated with bafilomycin A1 or NH4Cl (Fig. 7B).
These data indicate that the acidic milieu of endosomes is necessary for productive PyV infection of the host cells. When pH is increased (or acidification is blocked), the virus is still able to enter early endosomal compartments, but its exit and further trafficking are suppressed. Accordingly, overexpression of the dominant-negative mutant of Rab5 (Rab5S34N) reduced PyV infection markedly, while overexpression of wild-type Rab5 had no inhibition effect (not shown).
PyV and transferrin meet in Rab11-positive endosomes. As we have shown previously (20), PyV VP1 colocalizes with both transferrin and Rab11 GTPase between 1 and 3 h postadsorption. At the same time, partial colocalization of PyV VP1 with the BiP/GRP78 marker of the ER was demonstrated.
Various endosomal compartments, including caveosomes and recycling endosomes, accumulate in the perinuclear space near the MTOC (22, 27). The resolution of a confocal microscope (200 nm) is not sufficient to distinguish whether PyV and transferrin are present in the same endosomal compartment, particularly in such a crowded area of the cytoplasm. To circumvent this limitation, we used FRET, a method to detect protein colocalization at a 2- to 10-nm resolution. The range over which FRET between the donor (Cy3) and acceptor (Cy5) fluorescent molecule occurs is given by the spectral parameter R0, i.e., the distance at which the FRET efficiency is 50%. R0 for the Cy3-Cy5 system is 5 nm (4). Cells were bleached in the Cy5 channel by scanning a region of interest (ROI) using the 633-nm HeNe laser line at 100% intensity. We performed FRET in 15 different cells and bleached more then 60 different ROIs. Before and after the bleaching, Cy3 images were collected to assess changes in donor fluorescence. Figure 8A, panel a, shows the images of the donor (transferrin, Cy3) and the acceptor (VP1 PyV, Cy5) before and after photobleaching in the ROI marked in the figure. Fig. 8A, panel b, presents the intensity value in the ROIs marked on Fig. 8A, panel a.
As a control, we performed a similar calculation in the same number of nonbleached regions of the specimen to evaluate background FRET signals. We calculated the histogram distribution of FRET efficiency for the bleached and nonbleached regions. Distribution of FRET efficiency for bleached regions is positive-shifted from the pseudo-FRET efficiency observed in the nonbleached regions (histogram in Fig. 8A, panel c). The average FRET efficiency values between transferrin and PyV were 9.3 ± 6.9 for the bleached ROIs, which represent the true FRET, and an average value of 2.3 ± 3.2 for the nonbleached ROIs, which represent the background (or false) FRET.
We were further interested in whether the membrane compartments carrying both PyV and transferrin are recycling endosomes. Therefore, we repeated the experiments with Cy3-labeled Rab11 and measured the FRET efficiency of selected ROIs (Fig. 8B). Fig. 8B, panel b, represents intensity values in the ROIs marked on the images (Fig. 8B, panel a), and the FRET efficiency distribution is shown on the histogram (Fig. 8B, panel c). We calculated the FRET efficiency in 60 different ROIs from 14 different cells. An average value of 4.8 ± 2.8 for the bleached ROIs (representing the true FRET) has been assessed. An average value of 2.21 ± 3.6 for background FRET, or false FRET (from the nonbleached ROIs), has been obtained. We obtained very similar background FRET (false FRET) values for both transferrin and Rab11 under the same sample conditions (type of cells, virus, fluorescence donor, and acceptor) and the same acquisition conditions (objective and laser lines). Moreover, in both experiments, we obtained a positive-shifted distribution of FRET efficiency in respect to the nonbleached regions and average values of FRET efficiency higher than those of background FRET. The conclusion can be made that both Rab11 and transferrin exhibit a true FRET and that PyV and transferrin are in intimate contact in a Rab11-positive compartment.
DISCUSSION
Despite intensive studies on the early events of mouse polyomavirus infection, the mechanism of virus uptake and trafficking in the host cell still remains puzzling. Even closely related viruses such as SV40 and human JC virus differ in their modes of entry and subsequent trafficking. JC virus uses the classical endocytic pathway via clathrin-coated pits to enter early endosomes (30), while SV40 is internalized by caveolae (1). Also, another human polyomavirus, BK virus, is dependent on an intact caveolin-1 scaffolding domain to enter the host cell (11). The role of lipid raft microdomains in PyV entry is also unclear. Treatment of cells with cholesterol-depleting agents such as methyl--cyclodextrin, which disrupts detergent-resistant lipid rafts, revealed controversial results, most likely due to the different PyV strains and cell lines used (13, 15, 33). We have previously shown the presence of caveolin-1 on monopinocytic vesicles carrying PyV virions (33). Observations of interactions of PyV with caveolin-1 domains and compartments in living cells revealed that some PyV particles enter cells through caveolin-1-rich domains, while others bypass them, and at later times postinfection, the signal of PyV can be seen in large caveolin-1-rich compartments in the perinuclear space (Fig. 2; see movies at http://www.natur.cuni.cz/molbio/virology/suppl.html). Is caveolin-1 indeed necessary for PyV infection Many toxins that are transported from the cell surface to the ER utilize ganglioside receptors similar to those of polyomaviruses (37). CTb exploits the same GM1 type of ganglioside as SV40 (16). The CTb enters cells via caveolae but can also be efficiently internalized in cells lacking caveolin-1 by a cholesterol-dependent process (26). We show here that in Jurkat cells that do not express caveolin, not only CTb but also SV40 (Fig. 3Bb) was efficiently endocytosed, and both antigens exhibited similar pericentriolar localization. The uptake of PyV occurred as well, but the virus signal remained spread within the cytoplasm. Nevertheless, no colocalization of clathrin with PyV was observed in Jurkat cells, and ultrastructural analysis proved that PyV virions were internalized in smooth, tightly fitting vesicles morphologically similar to those of caveolin-1-positive mouse fibroblast or epithelial cells (33). Moreover, expression of a dominant-negative mutant of caveolin-1 in NMuMG epithelial cells did not prevent their productive infection by PyV.
In mouse fibroblasts or in epithelial cells, virus-loaded monopinocytic vesicles were often shown to fuse with peripheral endosomes early after internalization (33). Therefore, we further examined the characteristics of these compartments. Colocalization of PyV with the early endosome antigen EEA1 has shown that the majority of the virus entered early endosomes. The uptake of virions was slower than that of transferrin, and both cargos rarely met in early endosomes when added simultaneously to the cells. However, at later times postinfection (1.5 to 3 h p.i.), PyV VP1 colocalized with transferrin and Rab11 mostly in perinuclear areas, suggesting that recycling endosomes can be the compartments where they can encounter each other (20). Thus, PyV internalized in caveola-derived vesicles, or perhaps also in other raft-derived vesicles, later enters compartments of the classical clathrin-dependent pathway (e.g., early and/or recycling endosomes) and gets together with a cargo delivered there in a clathrin-dependent manner. Sharma et al. (34) previously demonstrated that glycosphingolipids could be internalized via caveola-related endocytosis and could then rapidly merge with the clathrin pathway in early endosomes. Glycosphingolipids can form distinct microdomains within the early endosomal membrane, which behave differently with respect to their subsequent intracellular trafficking and might play a role in cargo sorting (34).
We further examined the effect of the elevation of endosomal pH on PyV infection. Bafilomycin A1 disrupts the H+ gradient that exists in vesicles of the vacuolar system (7, 39). We proved that, at least in mouse 3T6 fibroblasts and NMuMG epithelial cells, this agent prevents PyV productive infection. Moreover, exposition of these cells to NH4Cl (which rapidly elevates endosomal pH) also inhibited infection by PyV when the cells were exposed to the drugs during the initial steps of infection. At later times postinfection, since the virus signal was detected in the perinuclear space, infection became pH insensitive. Remarkably, previous findings of other groups showed that neither SV40 (35, 38) nor PyV (13) infection was affected by NH4Cl treatment. On the other hand, infection by human JC virus, but also BK virus (internalized by caveolae), was found to be sensitive to the elevation of endosomal pH (3, 11). Ashok and Atwood (3) previously found that JC virus infection is, in contrast to that of SV40, decreased by approximately 70% by NH4Cl, while infection by both SV40 and JC virus can be completely inhibited by bafilomycin A1. The different sensitivities of cells to NH4Cl treatment and the time of exposition could account for the discrepancy between our finding and previous (13) findings. Since under the conditions described previously by Ashok and Atwood (3), our cells died before they could have been analyzed, we shortened the time of NH4Cl treatment by half.
Although both SV40 and cholera toxin bind the same ganglioside GM1 receptor, intoxication by CTb, in contrast to SV40 infection, depends on exposure to an acidic milieu (17). Recently, Pelkmans et al. (28) described a Rab5-dependent pathway in which caveolar vesicles are targeted to early endosomes, where they form distinct but transient membrane domains. When they monitored the SV40 and CTb cargo, the low pH of early endosomes selectively induced a conformational change of CTb (21) and its lateral diffusion into a surrounding membrane subdomain of the compartment, while SV40 remained trapped in the caveolin-positive subdomain from where it was subsequently sorted out into caveosomes. They also found that when caveolin-1 was down-regulated, virions of SV40 diffused into the lumen of endosomes, but that did not lead to infection. This is in agreement with the previous observations that the pathway through caveosomes and the ER is necessary for productive SV40 infection (29). We found out that cells expressing a dominant-negative caveolin-1 mutant can be productively infected by PyV, and in cells with an elevated endosomal pH, the majority of PyV was retained in early endosomes, similar to the CTb that was found to be restricted in its mobility (28). The traffic of PyV is also similar to that of CTb (and differs from SV40) with respect to the dependence on Rab5 function.
Enveloped viruses make use of endosomal acidic pH for inducing the fusion of their envelopes with the vesicle membrane and for nucleocapsid escape to free cytosol. Some nonenveloped viruses, e.g., adenoviruses or human rhinovirus serotype 2, also exploit the acidic milieu of early endosomes to trigger conformational changes in their capsids followed by partial disassembly to evade the barrier of the endosomal membrane (23, 31).
It became obvious that the majority of PyV virions (or their partially disassembled complexes) are further transported in membranous compartments to the perinuclear space. Time lapse microscopy (Fig. 2) showed that PyV was internalized in large caveolin-rich compartments located in the perinuclear space. We showed previously (20) that the signal of major capsid protein VP1 colocalized there with transferrin and Rab11 but also with BiP/GRP78, a marker of the ER. We suggest that similar to cholera or Shiga toxin, a vast proportion of PyV is carried from early endosomes, via recycling endosomes, to the ER. In contrast to bacterial toxins, PyV seems to bypass the Golgi compartment (our unpublished results). Recycling endosomes are mildly acidic compartments containing components of membrane rafts such as glycosylphosphatidylinositol-anchored proteins, sphingomyelin, cholesterol, and caveolin-1 (12). In this regard, they resemble caveosomes utilized by SV40. Moreover, in some cells, both caveosomes and recycling endosomes were found to share an overlapping distribution. Caveosomes immunolabeled with anti-caveolin-1 antibody merged with the transferrin receptor signal (present in recycling endosome membranes) in the perinuclear space near the MTOC of CHO cells. We suggest that caveosomes and recycling endosomes are intertwined rather than that they represent an identical membrane system (22). Nevertheless, positive FRET between VP1 and transferrin and between VP1 and Rab11 provides strong support for the presence of PyV VP1 and transferrin within the same compartment. Further research is required to define functional and biochemical differences between caveosomes and recycling endosomes and to reveal whether both compartments can communicate in a way similar to that of caveola-derived vesicles with early endosomes.
The most intricate question, from where and by what mechanism the PyV genome is delivered into the nucleus, remains unclear and is currently under vigorous investigation. A summary of results obtained by our studies of endocytic pathways exploited by mouse polyomavirus to deliver its genome into the cell nucleus is schematically drawn in Fig. 9.
ACKNOWLEDGMENTS
This work was generously supported by the Grant Agency of the Czech Republic (grant 204/03/0593), by the Centre of Functional Organization of Cells (LC545), by the Centre for New Antivirals and Antineoplastics (1M6138896301), by the programs of the Ministry of Education, Youth, and Sport of the Czech Republic, and by project no. AVOZ50520514.
We are grateful to all researchers (names are given in Materials and Methods) who provided Rab5, Eps15, and caveolin-A mutants or antibodies and to . Takáová for assistance in preparation of the manuscript.
REFERENCES
Anderson, H. A., Y. Chen, and L. C. Norkin. 1996. Bound simian virus 40 translocates to caveolin-enriched membrane domains, and its entry is inhibited by drugs that selectively disrupt caveolae. Mol. Biol. Cell 7:1825-1834.
Anderson, H. A., Y. Chen, and L. C. Norkin. 1998. MHC class I molecules are enriched in caveolae but do not enter with simian virus 40. J. Gen. Virol. 79:1469-1477.
Ashok, A., and W. J. Atwood. 2003. Contrasting roles of endosomal pH and the cytoskeleton in infection of human glial cells by JC virus and simian virus 40. J. Virol. 77:1347-1356.
Bastiaens, P. I. H., and T. M. Jovin. 1998. Fluorescence resonance energy transfer microscopy, p. 136-146. In J. E. Celis (ed.), Cell biology: a laboratory handbook, 2nd ed., vol. 3. Academic Press, New York, N.Y.
Benmerah, A., M. Bayrou, N. Cerf-Bensussan, and A. Dautry-Varsat. 1999. Inhibition of clathrin-coated pit assembly by an Eps15 mutant. J. Cell Sci. 112:1303-1311.
Benmarah, A., C. Lamaze, B. Begue, S. L. Schmid, A. Dautry-Varsat, and N. Cerf-Bensussan. 1998. AP-2/Eps15 interaction is required for receptor-mediated endocytosis. J. Cell Biol. 140:1055-1062.
Bowman, E. J., A. Siebers, and K. Altendorf. 1988. Bafilomycins: a class of inhibitors of membrane ATPases from microorganisms, animal cells, and plant cells. Proc. Natl. Acad. Sci. USA 85:7972-7976.
Breau, W. C., W. J. Atwood, and L. C. Norkin. 1992. Class I major histocompatibility proteins are an essential component of the simian virus 40 receptor. J. Virol. 66:2037-2045.
Caruso, M., L. Belloni, O. Sthandier, P. Amati, and M.-I. Garcia. 2003. 41 integrin acts as a cell receptor for murine polyomavirus at the postattachment level. J. Virol. 77:3913-3921.
Deeks, E. D., J. P. Cook, P. J. Day, D. C. Smith, L. M. Roberts, and J. M. Lord. 2002. The low lysine content of ricin A chain reduces the risk of proteolytic degradation after translocation from the endoplasmic reticulum to the cytosol. Biochemistry 41:3405-3413.
Eash, S., W. Querbes, and W. J. Atwood. 2004. Infection of Vero cells by BK virus is dependent on caveolae. J. Virol. 78:11583-11590.
Gagescu, R., N. Demaurex, R. G. Parton, W. Hunziger, L. A. Huber, and J. Gruenberg. 2000. The recycling endosome of Madin-Darby canine kidney cells is a mildly acidic compartment rich in raft components. Mol. Biol. Cell 11:2775-2791.
Gilbert, J. M., and T. L. Benjamin. 2000. Early steps of polyomavirus entry into cells. J. Virol. 74:8582-8588.
Gilbert, J. M., J. Dahl, J. You, C. Vui, R. Holmes, W. Lencer, and T. L. Benjamin. 2005. Ganglioside GD1a restores infectibility to mouse cells lacking functional receptors for polyomavirus. J. Virol. 79:615-618.
Gilbert, J. M., I. G. Goldberg, and T. L. Benjamin. 2003. Cell penetration and trafficking of polyomavirus. J. Virol. 77:2615-2622.
Griffiths, S. L., R. A. Finkelstein, and D. R. Critchley. 1986. Characterization of the receptor for cholera toxin and Escherichia coli heat-labile toxin in rabbit intestinal brush borders. Biochem. J. 238:313-322.
Janicot, M., F. Fouque, and B. Desbuquois. 1991. Activation of rat liver adenylate cyclase by cholera toxin requires toxin internalization and processing in endosomes. J. Biol. Chem. 266:12858-12865.
Krauzewicz, N., J. tokrová, C. Jenkins, M. Elliott, C. F. Higgins, and B. E. Griffin. 2000. Virus-like gene transfer into cells mediated by polyoma virus pseudocapsids. Gene Ther. 7:2122-2131.
Li, G., and P. D. Stahl. 1993. Structure-function relationship of the small GTPase rab5. J. Biol. Chem. 268:24445-24480.
Mannová, P., and J. Forstová. 2003. Mouse polyomavirus utilizes recycling endosomes for a traffic pathway independent of COPI vesicle transport. J. Virol. 77:1672-1681.
McCann, J. A., J. A. Mertz, J. Czworkowski, and W. D. Pickin. 1997. Conformational changes in cholera toxin B subunit-ganglioside GM1 complexes are elicited by environmental pH and evoke changes in membrane structure. Biochemistry 36:9169-9178.
Mundy, D. I., T. Machleidt, Y. S. Ying, R. G. Anderson, and G. S. Bloom. 2002. Dual control of caveolar membrane traffic by microtubules and the actin cytoskeleton. J. Cell Sci. 115:4327-4339.
Nakano, M. Y., K. Boucke, M. Suomalainen, P. Stidwell, and U. G. Grebe. 2000. The first step of adenovirus type 2 disassembly occurs at the cell surface, independently of endocytosis and escape to the cytosol. J. Virol. 74:7085-7095.
Nathke, I. S., J. Heuser, A. Lupas, J. Stock, C. W. Turck, and F. M. Brodsky. 1992. Folding and trimerization of clathrin subunits at the triskelion hub. Cell 68:899-910.
Norkin, L. C., H. A. Anderson, W. A. Scott, and A. Oppenheim. 2002. Caveolar endocytosis of simian virus 40 is followed by brefeldin A-sensitive transport to the endoplasmic reticulum, where the virus disassembles. J. Virol. 76:5156-5166.
Orlandi, P. A., and P. H. Fishman. 1998. Filipin dependent inhibition of cholera toxin: evidence for toxin internalisation and activation through caveolae like domains. J. Cell Biol. 141:905-915.
Pasqualato, S., F. Senic-Matuglia, L. Renault, B. Goud, J. Salamero, and J. Cherfils. 2004. The structural GDP/GTP cycle of Rab11 reveals a novel interface involved in the dynamics of recycling endosomes. J. Biol. Chem. 279:11480-11488.
Pelkmans, L., T. Burli, M. Zerial, and A. Helenius. 2004. Caveolin-stabilized membrane domains as multifunctional transport and sorting devices in endocytic membrane traffic. Cell 118:767-780.
Pelkmans, L., J. Kartenbeck, and A. Helenius. 2001. Caveolar endocytosis of simian virus 40 revealed a new two step vesicular transport pathway to the endoplasmic reticulum. Nat. Cell Biol. 3:473-483.
Pho, M. T., A. Ashok, and W. J. Atwood. 2000. JC virus enters human glial cells by clathrin-dependent, receptor-mediated endocytosis. J. Virol. 74:2288-2292.
Prchla, E., C. Plank, E. Wagner, D. Blaas, and R. Fuchs. 1995. Virus-mediated release of endosomal content in vitro: different behavior of adenovirus and rhinovirus serotype 2. J. Cell Biol. 131:111-123.
Richards, A. A., E. Stang, R. Pepperkok, and R. G. Parton. 2002. Inhibitors of COP-mediated transport and cholera toxin action inhibit simian virus 40 infection. Mol. Biol. Cell 13:1750-1764.
Richterová, Z., D. Liebl, M. Horák, Z. Palková, J. tokrová, P. Hozák, J. Korb, and J. Forstová. 2001. Caveolae are involved in the trafficking of mouse polyomavirus virions and artificial VP1 pseudocapsids toward cell nuclei. J. Virol. 75:10880-10891.
Sharma, D. K., A. Choudhury, R. D. Singh, C. L. Wheatley, D. L. Marks, and R. E. Paganos. 2003. Glycosphingolipids internalized via caveolar-related endocytosis rapidly merge with the clathrin pathway in early endosomes and form microdomains for recycling. J. Biol. Chem. 278:7564-7572.
Shimura, H., Y. Umeno, and G. Kimura. 1987. Effects of inhibitors of the cytoplasmic structures and functions on the early phase of infection of cultured cells with simian virus 40. Virology 158:34-43.
Trouet, D., D. Hermans, G. Droogmans, B. Nilus, and J. Eggermont. 2001. Inhibition of volume-regulated anion channels by dominant-negative caveolin-1. Biochem. Biophys. Res. Commun. 284:461-465.
Tsai, B., J. M. Gilbert, S. Stehle, W. Lencer, T. L. Benjamin, and T. A. Rapoport. 2003. Gangliosides are receptors for murine polyoma virus and SV40. EMBO J. 22:4346-4355.
Upcroft, P. 1987. Simian virus 40 infection is not mediated by lysosomal activation. J. Gen. Virol. 68:2477-2480.
Yochimori, T., A. Yamamoto, Y. Moriyama, M. Futai, and Y. Tashiro. 1991. Bafilomycin A1, a specific inhibitor of vacuolar type H+ ATPase, inhibits acidification and protein degradation in lysosomes of cultured cells. J. Biol. Chem. 266:17707-17712.(David Liebl, Francesco Di)