Intracellular Topology and Epitope Shielding of Poliovirus 3A Protein

Cells and viruses.

COS-1 cells were cultured as monolayers in Dulbecco's modified Eagle medium supplemented with 10% (vol/vol) calf serum, 100 U of penicillin per ml, and 100 U of streptomycin per ml at 37°C and 5% CO₂. Wild-type and M394T mutant poliovirus type 1 Mahoney stocks were prepared in HeLa cells as described elsewhere (4, 23). Viral titers were determined by plaque assays on COS-1 cells as described elsewhere (16). All infections were performed in 6- or 10-cm tissue culture dishes containing 3.5 × 10⁵ or 1 × 10⁶ COS-1 cells, respectively, at a multiplicity of infection (MOI) of 20 PFU/cell for 4.5 h at 37°C, unless otherwise indicated.

Plasmids and transfections.

The GFP-HO-2 plasmid construct was a generous gift from Tom A. Rapoport (Harvard University) (38). Briefly, it consists of the coding region of the green fluorescent protein (GFP) gene fused to the 5′ end of the coding region of the heme oxygenase 2 (HO-2) gene, under the transcriptional control of the cytomegalovirus promoter. The p5NC-α6-based DNA plasmid constructs used to express wild-type 3A and 3A-2 mutant protein under the transcriptional control of the simian virus 40 late promoter have been described (15). COS-1 cells were transfected in 10-cm tissue culture dishes by using Lipofectamine Plus reagents as described by the manufacturer, with 20 μl of Plus reagent and 30 μl of Lipofectamine reagent (Invitrogen Life Technologies, Carlsbad, Calif.). Either GFP-HO-2 plasmids (4 μg per transfection) or p5NC-α6-based plasmids (8 μg per transfection) were used for transfections. Cells were processed for either proteolysis or immunofluorescence 24 h following transfection.

Antibodies and reagents.

Anti-Grp94 monoclonal rat antibody was purchased from Stressgen Biotechnologies Corp. (Victoria, British Columbia, Canada) and used at a dilution of 1:50 for immunofluorescence and 1:200 for immunoblot analysis. Anti-GFP monoclonal mouse antibody was purchased from Clontech (Palo Alto, Calif.) and used at a dilution of 1:1,000 for immunoblot analysis. Anti-3A monoclonal mouse antibody was a hybridoma supernatant (15) used at a dilution of 1:30 for immunofluorescence and 1:50 for immunoblot analysis. Anti-mouse alkaline phosphatase (AP)-conjugated secondary antibody was purchased from Jackson ImmunoResearch Laboratories, Inc. (West Grove, Pa.), and used at a dilution of 1:10,000 for immunoblot analysis. Anti-rat AP-conjugated secondary antibody was purchased from Stressgen Biotechnologies Corp. and used at a dilution of 1:20,000 for immunoblot analysis. Fluorescein isothiocyanate (FITC)- and Texas Red-conjugated anti-mouse secondary antibodies were purchased from Vector Laboratories, Inc. (Burlingame, Calif.) and used at a dilution of 1:1,000 for immunofluorescence analysis. Prestained molecular weight markers and ECF reagent for the quantitative detection of AP-conjugated secondary antibodies were purchased from Amersham (Sunnyvale, Calif.). Digitonin and Triton X-100 were purchased from Sigma (St. Louis, Mo.). All reagents and antibodies were used and stored according to the manufacturer's suggestions.

Indirect immunofluorescence microscopy.

COS-1 cells were plated onto coverslips in 6- or 10-cm tissue culture dishes 24 h before transfection. Twenty-four hours posttransfection, cells were washed three times with phosphate-buffered saline (PBS), fixed with 4% formaldehyde in PBS for 10 min at room temperature, washed three more times with PBS, and permeabilized with either 0.5% Triton X-100 (Sigma) in PBS for 10 min at room temperature or 0.5 μg of digitonin per ml in buffer D (0.3 M sucrose, 2.5 mM MgCl₂, 0.1 M KCl, 1 mM EDTA, and 10 mM PIPES; pH 6.8) for 5 min on ice. Following permeabilization, cells were incubated in PBS containing 1% bovine serum albumin (BSA) for 30 min at room temperature, incubated in the appropriate primary antibody in PBS-1% BSA for 1 h at room temperature, washed three times for 15 min in PBS-1% BSA, incubated with the appropriate secondary antibody in PBS-1% BSA for 1 h at room temperature, and washed again in PBS three times for 15 min each time. The coverslips were mounted on slides by using 5 to 10 μl of Vectashield (Vector Laboratories, Inc.) as an antifading agent. The slides were viewed on an upright fluorescent microscope equipped with a 40× objective and photographed by using a digital camera and ImagePro software (Media Cybernetics, Silver Spring, Md.). Quantitation of fluorescence intensities was performed using Metamorph software (Universal Imaging Corporation, Downingtown, Pa.).

Protease assay.

For each proteolysis experiment, four 10-cm tissue culture dishes of COS-1 cells were transfected with GFP-HO-2 plasmid and then infected with poliovirus 24 h later for various times. Cells were harvested by the semipermeabilized-cell method of Pind et al. (35). Briefly, cells were washed three times with ice-cold 50/90 H/KOAc buffer (50 mM HEPES adjusted to pH 7.2 with KOH and 90 mM potassium acetate) and hypotonically swollen in 10/18 H/KOAc buffer (10 mM HEPES adjusted to pH 7.2 with KOH and 18 mM potassium acetate) for 10 min on ice. The hypotonic buffer was then aspirated, and the cells were scraped into 1 ml of ice-cold 50/90 H/KOAc buffer. Then 4 ml of scraped cells was added to 8 ml of ice-cold 50/90 H/KOAc buffer and collected by centrifugation at 800 × g for 4 min at 4°C to remove soluble cytosolic proteins. The cells were then resuspended in 2 ml of ice-cold PBS, split into two microcentrifuge tubes, and collected by centrifugation at 4°C for 4 min at 800 × g. The cell pellet in one tube was resuspended in ice-cold PBS and the other tube was resuspended in ice-cold PBS containing 0.5% Triton X-100 (Sigma). Both tubes were placed on a rotator at 4°C for 10 min; 100-μl aliquots were incubated with various concentrations of proteinase K (Sigma) for 15 min at 37°C. Proteolysis was stopped by the addition of 20 μl of a solution containing 1 mM phenylmethylsulfonyl fluoride (Sigma) and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer. Samples were incubated at 100°C for 5 min, and 35 μl of each sample was loaded onto each of three SDS-PAGE gels.

SDS-PAGE gels and Western blotting.

To analyze samples for 3AB and 3A, lysates were electrophoresed on a 16.5% Tricine-SDS gel (40). To analyze samples for Grp94 or GFP-HO-2, lysates were electrophoresed on an 8% glycine-SDS-PAGE gel (24). Following SDS-PAGE, samples were transferred to polyvinylidene difluoride membranes (Millipore, Billerica, Mass.) for 1 h at 800 mA, using a Hoefer tank transfer system (Hoefer Pharmacia Biotech, San Francisco, Calif.). The efficiency of transfer was verified by using prestained molecular weight markers (Amersham). After transfer was complete, membranes were blocked by rocking in PBS-T (PBS that contained 0.1% [vol/vol] Tween-20 [Sigma]) and 5% (wt/vol) Carnation nonfat dry milk for 1 h at room temperature. Membranes were rocked for 1 h at room temperature with the appropriate primary antibody diluted in PBS-T. Membranes were then washed four times in PBS-T for 10 min each time, rocked for 1 h at room temperature with the appropriate AP-conjugated secondary antibody, diluted in PBS-T, washed four times in PBS-T for 10 min each time, developed in ECF reagent (Amersham) for 5 min, viewed on a PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.), and analyzed with ImageQuant software.

Molecular modeling.

Modeling was performed with the Swiss PDB Viewer (http://www.expasy.ch/spdbv) (20), and rendered with POV-RAY (http://www.povray.org). Coordinates for the unit cell of the three-dimensional structure of poliovirus RNA-dependent RNA polymerase (21) were provided by J. Hansen (Yale University) and S. Schultz (Diné College) and can be obtained from the National Center for Biotechnology Information library under PDB identification number 1RDR.

Immunofluorescence of 3A protein in transfected and infected cells.

During studies of the effects of poliovirus infection and the expression of poliovirus 3A protein in isolation, we compared the amounts of accumulated protein and the immunofluorescence signals observed from 3A-containing polypeptides using a monoclonal antibody that recognizes an epitope between amino acids 23 and 59 (Fig. 1A) (15). As shown by the immunoblot in Fig. 1B, less 3A protein accumulated in COS-1 cells transfected with the 3A-expressing plasmid than in cells that underwent 4.5 h of poliovirus infection. To determine the average amount of 3A in each cell, the difference between the transfection efficiency (30%) and the infection efficiency (100%) was taken into account. Nonetheless, it was striking to observe that the immunofluorescence signal obtained from transfected cells was much stronger than that from the infected cells, even at 4.5 h postinfection, when approximately 10-fold more 3A-containing proteins were present per cell in the infected population (Fig. 1C). Furthermore, no increase in fluorescent signal was observed between 3.5 and 4.5 h of poliovirus infection, although the amount of 3A and 3AB proteins increased 20-fold.

Two distinct hypotheses could account for the reduced immunofluorescence signal from poliovirus-infected cells compared to that from cells that expressed 3A protein in isolation. First, since permeabilization was performed with digitonin, which is known to permeabilize plasma but not intracellular membranes, it was possible that the amino-terminal epitope of 3A protein and its precursors was found within the lumen of an intracellular compartment in infected, but not transfected, cells. A second possibility was that the N terminus of 3A is cytosolic in both transfected and infected cells, but in infected cells the epitope is blocked from interacting with the monoclonal antibody. To distinguish between these possibilities, we tested the membrane topology of poliovirus 3A protein by protease digestion and immunofluorescence analysis under conditions in which the integrity of intracellular membranes was demonstrated.

Immunofluorescence analysis of 3A- and 3A-2-transfected COS cells.

To test whether the N-terminal sequences of 3A protein were cytosolic in cells that expressed 3A in isolation, COS cells were transfected with a plasmid that encoded either wild-type 3A protein or 3A-2 protein, a mutant version of 3A that displays a reduced ability to disrupt ER-to-Golgi traffic (15). After 24 h, the cells were fixed and treated with either Triton X-100, to permeabilize both plasma and intracellular membranes, or with digitonin, to leave the intracellular membranes intact. As a control for the integrity of intracellular membrane topology, the immunofluorescence signal obtained from Grp94, known to reside within the lumen of the ER, was also tested. As shown in Fig. 2, Grp94 staining was readily detected in the Triton X-100-permeabilized, but not the digitonin-permeabilized, samples, arguing that the topology of the membranes was undisturbed during sample preparation. The immunofluorescence signal obtained using the anti-3A antibody, on the other hand, was visible despite the method of permeabilization, supporting the hypothesis that at least part of the population of 3A proteins was exposed to the cytosol.

When expressed in isolation, poliovirus 3A protein inhibits host protein traffic from the ER to the Golgi apparatus at a step that has not yet been identified (13-15). The 3A-2 mutation reduces the ability of 3A protein in isolation, or of viruses that contain the mutation, to inhibit ER-to-Golgi traffic, but does not substantially interfere with viral growth (6, 14, 15). As shown in Fig. 2, 3A protein that contains the 3A-2 mutation also displayed cytosolic localization of its amino-terminal residues.

Immunofluorescence analysis of poliovirus-infected cells.

During poliovirus infection, 3A protein sequences are found both in precursors such as 3AB and in the processed 3A product. To test whether the amino-terminal domains of 3A and larger 3A-containing proteins are localized on the cytosolic surface of membranes during poliovirus infection, and to test whether the N terminus of 3A is cytosolic during infection as it is following transfection, we performed immunofluorescence analysis on poliovirus-infected cells. Grp94 staining of poliovirus-infected cells could be observed following Triton X-100 but not digitonin permeabilization, arguing that intracellular membranes were intact. However, the immunofluorescence signal obtained with anti-3A antibody did not depend on the permeabilization detergent used, arguing that 3A epitopes were exposed to the cytosol, as shown in Fig. 3.

Proteolytic digestion of 3A protein.

Although the immunofluorescence analysis showed that some proportion of the 3A-containing proteins display cytosolic localization of their amino-terminal sequences, such analysis could not reveal the potential presence of a luminal subpopulation if there were heterogeneity in the topology of the protein or its precursors. To test the potential existence of a luminal subpopulation of 3A-containing proteins, we performed limited proteolysis on semipermeabilized cells. The plasma membranes of transfected and infected cells were mechanically permeabilized by hypotonic swelling, followed by scraping of the cells from the tissue culture dish. This technique lyses the cells while leaving the ER membranes intact, in the proper orientation, and functional to support anterograde transport (2, 5, 35). Semipermeabilized cells were washed to remove cytosolic contents and subjected to digestion with various concentrations of proteinase K in the absence or presence of Triton X-100. The degree of proteolysis of control and experimental proteins was determined by immunoblot analysis.

To determine the efficiency of plasma membrane lysis and the integrity of the ER, the luminal protein Grp94 and the fusion protein GFP-HO-2 (38) were monitored. As shown in Fig. 4A and D, the luminal ER protein Grp94 was susceptible to proteinase K in the presence, but not in the absence, of detergent, arguing that the ER remained intact during the permeabilization procedure. The cytosolic GFP-HO-2 protein, on the other hand, was 80% digested in the absence of detergent and 100% digested in the presence of detergent. The simplest explanation of this result is that the gentle technique used to render the cells semipermeable resulted in the lysis of only 80% of the cells.

During poliovirus infection, the patterns of proteolytic digestion of 3A and 3AB were identical to each other and to that of the cytosolic marker GFP-HO-2 (Fig. 4B to F). Specifically, in the absence of detergent, approximately 80% of 3A and 3AB were digested, consistent with the hypothesis that there is no protected subpopulation of either 3A or 3AB in productively infected cells (Fig. 4F). In the presence of detergent, as expected, 100% of the 3A and 3AB populations are susceptible to digestion with proteinase K. The results from these analyses are consistent with the results from the immunofluorescence analyses (Fig. 2), supporting the hypothesis that the amino termini of both 3AB and 3A are cytosolic. They also support the hypothesis that it is not the topology of 3AB or 3A but rather another factor that decreases the intensity of the 3AB or 3A immunofluorescence signal in infected cells.

Epitope shielding of 3A protein in wild-type poliovirus-infected but not vaccinia virus-infected cells.

To test whether 3A-containing sequences are relatively inaccessible to antibody probes when present in a poliovirus RNA replication complex, we expressed poliovirus 3A in the context of vaccinia virus infection as well as during poliovirus infection (13). A vaccinia virus vector was chosen because, as with poliovirus infection, almost all of the cells on a tissue-culture plate could be infected in a uniform way, making it more straightforward to use immunoblot analysis to compare the amounts of 3A-containing proteins made in each cell under the two different conditions. As shown in Fig. 5, although no 3A-containing proteins were detectable by immunoblotting at 3 h after infection with poliovirus or 4 h after infection with rVV-3A,GFP virus, equivalent immunofluorescence signals were seen with the anti-3A antibody at both time points. However, by 4 h after infection with poliovirus, both 3AB and 3A were detectable by immunoblotting, giving an amount of 3A-containing proteins greater than that seen by 8 h after infection with rVV-3A,GFP. Nonetheless, the immunofluorescence signal seen in the poliovirus-infected cells was much reduced relative to that in the VV-3A-infected cells. At later time points, such as 5 h after infection with poliovirus, the shielding of the 3A immunofluorescence signal was not so pronounced (Fig. 5A). Therefore, 3A-containing polypeptides, definitely 3AB and probably 3A as well, are shielded from probing with antibodies by an intracellular process, concurrent with the formation of the poliovirus RNA replication complex, which forms between 3 and 4 h postinfection.

Immunofluorescence of 3A protein during wild-type and M394T mutant poliovirus infections.

Poliovirus protein 3AB is known to bind directly to the viral RNA-dependent RNA polymerase, 3D (7, 8, 11, 31, 39, 41, 42, 44, 45), making 3D polymerase a likely candidate for the protein that directly shields the 3A epitope. A well-characterized mutant virus, 3D-M394T, contains a single mutation in the polymerase coding region, displays a temperature-sensitive RNA replication defect in infected cells, and shows a specific defect in the initiation of RNA replication in cell-free assays (4). Met394 in 3D polymerase is adjacent to the defined 3AB-binding site, and the purified M394T polymerase is defective in the addition of uridylyl residues to 3B (VPg), the protein primer for RNA synthesis (34). To test whether the 3A and 3AB epitope shielding observed during wild-type poliovirus infection would be less pronounced in the presence of a 3D polymerase potentially defective in these interactions, the immunofluorescence signals of 3A-containing proteins in wild-type and M394T infections were compared. Further incubation was continued at 39.5°C in the presence of 2 mM guanidine to inhibit RNA replication of the wild-type as well as the temperature-sensitive virus.

As shown in Fig. 6, the immunofluorescence signals at 0 and 75 min after the temperature shift were comparable between the wild-type and mutant viruses. However, the immunofluorescence signal of 3A during wild-type poliovirus infection at 45 min after the temperature shift was significantly less than during M394T mutant poliovirus infection. The difference in the signals was not due to a greater amount of 3A and 3AB proteins during wild-type poliovirus infection than during M394T mutant poliovirus infection, because immunoblot analysis demonstrated that 3A-containing proteins were more abundant during the wild-type infection than during the M394T infection (data not shown).

To quantify the fluorescence signal from 3A-containing proteins during wild-type and M394T mutant virus infection, the fluorescence intensities of 35 to 70 randomly chosen cells from each condition were measured. As can be seen in Fig. 6B, the temperature-sensitive virus showed substantially increased fluorescence, and therefore reduced epitope shielding, at the nonpermissive temperature. Therefore, the increased signal from the M394T infection at the nonpermissive temperature was due to increased availability of 3A-containing proteins to antibody. Parenthetically, this was probably also the case at the “permissive” temperature for the M394T mutant virus. As can be seen in Fig. 6 at the 0-min point, the M394T mutant-infected cells yielded a fluorescent signal similar to that of the cells infected with wild-type virus, in which 3A-containing proteins were more abundant. The reduction in shielding caused by the M394T mutation in the viral polymerase was more dramatic, however, after the temperature shift.