Pathogenesis and Immunity

Immunogenicity of Cytopathic and Noncytopathic Viral Vectors

Gabriela Plesa, Philip M. McKenna, Matthias J. Schnell, Laurence C. Eisenlohr
DOI: 10.1128/JVI.00084-06

ABSTRACT

The impact of cytolytic versus noncytolytic viral infections on host responses is not well understood, due to limitations of the systems that have been used to address this issue. Using paired cytopathic and noncytopathic rabies viruses that differ by only two amino acids, we investigated several fundamental aspects of the immune response to these viral vectors. Greater cytopathic capacity translated into a greater degree of cross-priming to CD8+ T cells (TCD8+) and more-robust short-term humoral and cellular responses. However, long-term responses to the two viruses were similar, suggesting that direct priming drives the bulk of the TCD8+ antirabies response and that enhanced acute responses associated with greater virally mediated cellular destruction were balanced by other factors, such as prolonged antigen expression associated with noncytopathic virus. Such compensatory mechanisms may be in place to ensure comparable immunologic memories to various pathogens.

Viruses can be classified as either cytopathic, meaning that cells are killed during the course of infection, or noncytopathic. Some viruses (8) are inherently noncytopathic because their replication programs are relatively benign, while others actively maintain a noncytopathic state by shutting down proapoptotic mechanisms, by activating antiapoptotic mechanisms (7), or by inducing alternative transcription programs that preserve the viral genome without virion production (57). Similarly, some viruses are passively cytopathic as a consequence of a viral replication cycle that damages the cell due to high protein expression and virus assembly or actively cytopathic due to expression of proapoptotic molecules (7, 48).

Several aspects of an immune response are expected to be substantially impacted by cytopathogenicity. For example, humoral responses to cytopathic versus noncytopathic viruses have been reported to vary widely due to factors such as antigen release from dying cells, infection of antigen-specific B cells (61), and levels of CD4+ T-cell activation (33). CD4+ T-cell responses and the level of cross-presentation, a process in which antigen is transferred from the infected cell to a “professional” antigen-presenting cell (APC), can also influence antiviral CD8+ T-cell (TCD8+) responses. Cross-presentation is essential for the development of effective adaptive immunity when a virus does not infect APCs (10, 49). In vivo cross-priming has been shown to be more efficient upon apoptosis induction (1, 9, 11, 44, 47, 51), suggesting that cytopathic viruses will mediate greater levels of cross-presentation.

Previous studies of how cytopathogenicity influences the immune response (13, 21, 29) have compared viruses with substantial differences besides cytopathic potential, making it essentially impossible to attribute differences in immune response to cytopathogenicity versus other factors. In the studies reported here, our objective was to generate paired viruses that are minimally different other than being cytolytic and noncytolytic and to determine how this property alone influenced short- and long-term immune responses. Rabies virus (RV) is a noncytopathic member of the family Rhabdoviridae, while vesicular stomatitis virus (VSV), the prototype of the rhabdovirus family, is a cytopathic virus. The difference can be attributed largely to the matrix (M) protein expressed by VSV, which induces cell apoptosis either in the context of VSV infection (5, 14) or when it is expressed in isolation (23, 31). VSV M protein plays a major role in inhibition of host gene expression (4), inhibition of nucleocytoplasmic transport (22), and disruption of the cell cytoskeleton (35). However, the apoptosis of VSV-infected cells is most likely the result of both virus-induced cell disorganization and host antiviral response (34). Point mutation of VSV M protein at positions 33 and 51 from methionine to alanine (M33,51A) has been shown to decrease the cytopathic effect and to delay cell death (27). To create the desired system, we inserted into the RV genome a wild-type or a mutant (M33,51A) version of VSV M. These two viruses differing by only two amino acids are anticipated to be dramatically different with respect to cytopathogenicity. In order to track immune responses with greater precision, we inserted a well-described antigen into both viruses. The paired viruses were compared in terms of replication rate, cytopathic potential, ability to mediate cross-presentation, and, finally, induction of short- and long-term humoral and cellular responses.

MATERIALS AND METHODS

Plasmid construction.The plasmid encoding a recombinant RV vaccine vector (pSPBN) with the restriction sites BsiWI and NheI has been described previously (39). The NP/SIINFEKL fragment was amplified from the previously described pSC11 plasmid (59) by PCR using primers ATCGTACGCCACCATGGCGTCCCAAGGC and GCTCTAGACTACTAATTGTCGTA containing the BsiWI and XbaI restriction sites (underlined). The PCR fragment was gel purified, BsiWI and XbaI digested, and ligated into previously BsiWI- and NheI-digested pSPBN. The resulting plasmid was designated pRVnoM-NP/S. The VSV wild-type M protein (VSVwtM) (pBSM) has been described elsewhere (12). VSV*M was constructed by replacing the methionine at positions 33 and 51 in VSVwtM with alanine (M33,51A) by using a QuickChange multisite-directed mutagenesis kit (Stratagene) and the primers GAAGAGGACACTAGCGCGGAGTATGCTCCGAGC and TTGGAGTTGACGAGGCGGACACCTATGATCCG. VSVwtM and VSV*M were PCR amplified using primers containing BsiWI and NheI restriction sites (underlined) (GGGCGTACGAAAATGAGTTCCTTAAAGAAGATTC and AAAGCTAGCTCATTTGAAGTGGCTGATAGAA). The fragments were gel purified, digested with BsiWI and NheI, and ligated into pBNSP plasmid (38) (here named pRVnoM) previously digested with BsiWI and NheI. The new plasmids were designated pRVwtM and pRV*M. The plasmids pRVwtM and pRV*M were digested with NaeI and SnaBI. The fragments containing either VSVwtM or VSV*M were ligated into pRVnoM-NP/S, generating pRVwtM-NP/S and pRV*M-NP/S plasmids (see Fig. 1).

FIG. 1.
FIG. 1.

Construction of cytopathic and noncytopathic recombinant RV vectors expressing different model antigens. VSVwtM or mutated VSV M was introduced between the N and P genes of RVnoM, resulting in RVwtM (cytopathic phenotype) or RV*M (noncytopathic phenotype). The Ova257-264 epitope embedded within the influenza nucleoprotein (NP/S) was inserted between rabies G and L genes (RVnoM-NP/S, RVwtM-NP/S, or RV*M-NP/S).

Recovery of recombinant viruses.For recovery of the recombinant RVs, a previously described RV recovery system was used (16, 53). Briefly, BSR T7 cells, BSR cells which stably express T7 RNA polymerase, were transfected with 5 μg of full-length RV cDNA (pRVnoM, pRVwtM, pRV*M, pRVnoM-NP/S, pRVwtM-NP/S, or pRV*M-NP/S) in addition to plasmids encoding the rabies N, P, L, and G proteins by using a Ca2PO4 transfection kit (Stratagene) according to the vendor's instructions. Three days posttransfection, supernatants were transferred onto fresh BSR cells. Infectious virus was detected 3 days later by immunostaining with fluorescein isothiocyanate (FITC)-labeled anti-RV N antibody (Centacor, Inc.). The correct sequences of the VSVwtM and VSV*M genes in the recovered viruses were confirmed by sequencing.

Cells.BSR cells, clonally derived from BHK-21 cells, were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS). L-Kb (L929 transfected with H-2Kb) and MC57G (H-2Kb) cells were maintained in DMEM with 5% FBS. P815 cells were maintained in RPMI with 10% FBS.

Mice.C57BL/6 (H-2Kb) female, 6- to 8-week-old mice were purchased from The Jackson Laboratory and maintained at the Thomas Jefferson University Animal Facilities (Philadelphia, PA). The animal protocols were approved by the Thomas Jefferson University Institutional Animal Care and Use Committee.

One-step growth curve.MC57G cells were plated in 60-mm-diameter dishes until 80% confluence (18 h) and infected at a multiplicity of infection (MOI) of 10 with the respective recombinant RVs. After incubation at 37°C for 1 h, the medium was aspirated, cells were washed three times with phosphate-buffered saline (PBS) to remove any unabsorbed virus, and 3 ml of DMEM with 10% FBS was added back. At indicated time points (see Fig. 3), 0.1 ml of supernatant was removed and stored at 4°C. The aliquots were titrated in duplicate on BSR cells as described previously (38, 39).

MTT assay.MC57G cells were incubated with recombinant viruses in duplicate at an MOI of 1 at 37°C for 3 days in a 120-μl volume. MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide) (Sigma) was diluted to 5 mg/ml in sterile PBS and filtered once through a 45-μm filter. MTT solution (25 μl) was added to each well, and the plates were incubated for an additional 4 h at 37°C. At that point, 0.1 ml of solubilization solution (10% Triton X-100 diluted in 0.1 N HCl in anhydrous isopropanol) was added to each well and incubation at 37°C was continued for an additional 2 h. The absorbance value was determined by an enzyme-linked immunosorbent assay (ELISA) reader at 590 nm.

Trypan blue exclusion.MC57G cells were plated in six-well plates and infected with recombinant viruses at an MOI of 1. At different time points, cells were collected and mixed with trypan blue to exclude dead cells and the live cells were counted.

Annexin staining.MC57G cells were plated in six-well plates and infected with recombinant viruses at an MOI of 1. At different time points, cells were collected, stained for annexin and propidium iodide (PI) (Pharmingen) as indicated by the manufacturer, and analyzed using an Epics Profile flow cytometer in the Kimmel Cancer Center Flow Cytometry Facility (Thomas Jefferson University).

Presentation assay.MC57G cells were plated in six-well plates and infected the next day with recombinant viruses at an MOI of 1. At different time points, cells were collected, washed three times with FACS buffer (PBS plus 5% fetal calf serum) and counted, and 106 cells were incubated for 1 h with 25.D1.16 culture supernatant (monoclonal antibody specific for residues 257 to 264 of ovalbumin [Ova257-264] complexed to H-2Kb) (45) at 4°C. Cells were washed three times in FACS buffer and incubated for 30 min at 4°C with 3 μg/ml FITC-labeled horse anti-mouse immunoglobulin G (Vector Laboratories). Cells were washed with FACS buffer, resuspended in 1% paraformaldehyde, and analyzed using an Epics Profile flow cytometer in the Kimmel Cancer Center Flow Cytometry Facility (Thomas Jefferson University).

ELISA.RV G protein was isolated from purified virions and RV N as ribonucleoprotein from infected cells as described previously (32, 54). RV N or G was resuspended in coating buffer (50 mM Na2CO3, pH 9.6) at a concentration of 200 μg/ml, and 100 μl/well was plated in 96-well ELISA MaxiSorp plates (Nunc). After overnight incubation at 4°C, plates were washed three times (PBS [pH 7.4], 0.1% Tween 20), blocked with blocking buffer (PBS [pH 7.4], 5% dry milk powder) for 30 min at room temperature, and incubated with serial dilutions of mouse sera for 1 h. Plates were washed three times, followed by the addition of horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (heavy plus light chains) secondary antibody (1:5,000; Jackson ImmunoResearch Laboratories). After 30 min of incubation at 37°C, plates were washed three times and 200 μl of OPD (o-phenylenediamine dihydrochloride) (Sigma) substrate was added to each well. The reaction was stopped by the addition of 50 μl of 3 M H2SO4 per well. Optical density was determined at 490 nm.

ELISPOT.Groups of three C57BL/6 mice, 6-week-old females, were inoculated intraperitoneally (i.p.) with 104 focus-forming units of RVwtM, RVwtM-NP/S, RV*M, and RV*M-NP/S viruses. Mice were bled retroorbitally at 8 or 28 days postimmunization and sacrificed, and spleen-derived single-cell suspensions were prepared. Red blood cells were removed with ACK lysing buffer (BioWhittaker), cells were resuspended in RPMI supplemented with 10% FBS, and 105 cells per well were plated in triplicate. Plates were previously coated overnight with 10 μg of anti-mouse gamma interferon (IFN-γ)/ml in sterile PBS and blocked for 2 h with RPMI with 10% FBS. Splenocytes added to the plates were incubated with MC57G cells (105 cells per well) either fresh (as a negative control) or pulsed with Ova257-264 peptide (10−6 M final) for 1 h at 37°C or infected with RV*M for 18 h. The plates were incubated for 18 to 20 h at 37°C and washed 10 times with PBS containing 0.25% Tween 20. Wells were incubated with 5 μg of biotinylated rat anti-mouse IFN-γ antibody/ml for 2 h at room temperature, washed five times, treated with 1 mg of horseradish peroxidase-conjugated streptavidin/ml in PBS containing 1% FBS, incubated for 1 h at room temperature, and washed four times. IFN-γ-secreting, spot-forming cells were detected using an AEC substrate reagent set for an enzyme-linked immunospot (ELISPOT) assay (Pharmingen). The plates and reagents were purchased from Pharmingen. For cross-priming experiments, a modified version of the procedure developed by Norbury et al. (42) was used. P815 cells were infected with recombinant viruses at a previously detected optimum MOI of 3 for 2 days, UV irradiated (312 nm) for 10 min, and counted, and 2 × 106 cells per 200 μl were injected i.p. into C57BL/6 mice. Seven days after immunization, mice were sacrificed, and spleen cell suspensions were prepared and analyzed by ELISPOT assay as described above. L-Kb cells were used to assay the Ova257-264-specific responses, as they stably express higher levels of H-2Kb molecules than MC57G cells.

RESULTS

Construction of cytopathic and noncytopathic rabies viruses.The genes encoding VSV wtM and the M33,51A (*M) double mutant protein of the noncytopathic VSV phenotype were cloned between the N and P genes of the highly attenuated RV strain pBNSP (38) (here designated pRVnoM), resulting in pRVwtM or pRV*M (Fig. 1). Because one goal was to study epitope- specific TCD8+-cell immune responses within this system and no immunodominant RV-specific epitopes have been identified, we inserted the influenza A nucleoprotein gene that we previously modified to encode the extensively studied Ova257-264 epitope (also known as SIINFEKL) (59). The region between the G and L genes of pRVnoM, pRVwtM, or pRV*M was the target for introduction of the heterologous NP/S gene. The resulting plasmids were entitled pRVnoM-NP/S, pRVwtM-NP/S, and pRV*M-NP/S (Fig. 1). The recombinant RVs were recovered by standard methods (53).

Cytopathic characteristics of recombinant RVs.To determine whether the insertion of VSV M protein into the RV genome confers cytopathogenicity, we infected MC57G cells with RVs (RVnoM, RVwtM, and RV*M) and counted the live cells at successive days by trypan blue exclusion. As shown in Fig. 2a, the number of live cells decreased continuously after infection with RVwtM. Cells infected with RVnoM or RV*M proliferated 2 days after infection. After that point, the numbers decreased but only modestly compared to cells infected with RVwtM. We confirmed these results using the MTT assay (Fig. 2b), which measures the activity of mitochondrial dehydrogenases in living cells, an early indicator of cell viability. The extent of cell death induced by RVwtM was substantially higher than that induced by the other viruses. RVnoM- and RV*M-infected cells died much more slowly, suggesting more passive mechanisms of death induction. As a final assessment of cytopathogenicity, we measured the proportions of dead and apoptotic cells by PI and annexin staining (Fig. 2c). These data indicate higher proportions of early apoptotic (PI− annexin+) and late apoptotic (PI+ annexin+) cells after infection with RVwtM than after infection with RV*M viruses 2 days postinfection. At day 3 postinfection, we observed lower percentages of apoptotic cells, most likely due to proliferation of uninfected cells, as also seen in Fig. 2a. The patterns of cell survival were similar regardless of whether NP/S was inserted (data not shown), indicating no influence of the heterologous protein with respect to cell death. These experiments indicate that expression of the VSVwtM protein converted the phenotype of RV from noncytopathic to cytopathic and confirmed that residues 33 and 51 are essential for lytic activity, as previously reported (27).

FIG. 2.
FIG. 2.

Analysis of cytolytic capacities of recombinant RVs. (a) Live cells counted by trypan blue exclusion. MC57G cells were infected with recombinant RVs at an MOI of 1, and trypan-excluding cells were counted at different times postinfection. Data are representative of three independent experiments, each of them done in triplicate. (b) MTT assay. Three days after the MC57G cells were infected with recombinant RVs at an MOI of 1, cell viability was measured by MTT conversion as described in Material and Methods. Data are representative of four independent experiments done in duplicate. The numbers represent P values calculated using Student's t test. (c) Annexin and PI staining of MC57G cells infected with recombinant viruses at an MOI of 1 and analyzed at the indicated time points. The numbers represent percentages of cells corresponding to each quadrant. The experiment was done twice. OD, optical density; PI, propidium iodide.

One-step growth curves of recombinant RVs.The observed cytopathogenicity of VSVwtM-expressing RVs could have been the consequence of a faster replication rate of the RVwtM virus and, consequently, a more rapid loss of cell viability. In order to address this possibility, we infected cells with an MOI of 10 to ensure synchronous infection of all cells. As shown in Fig. 3, recombinant RVs have similar replication rates. At 72 h postinfection, the RVnoM titer is about 0.5 log higher, but the M-expressing viruses (RVwtM and RV*M) grow to the same titers. Thus, it is the lytic potential of RVwtM virus, not the replication rate, that influenced survival of infected cells as seen in Fig. 2.

FIG. 3.
FIG. 3.

One-step growth curves of recombinant RVs. MC57G cells were infected with recombinant viruses at an MOI of 10. Aliquots of culture supernatants were collected at different time points, and viral titers were determined in duplicate. Data are representative of four independent experiments.

Ova257-264 presentation by recombinant RVs.In preparation for the immunogenicity experiments reported below, we tested the capacities of cytopathic (RVwtM) and noncytopathic (RVnoM and RV*M) virus recombinants to express the Ova257-264 epitope. MC57G cells (H-2Kb) were infected in vitro and at different time points were stained with an antibody specific for the Kb/Ova257-264 complex (45) and analyzed by flow cytometry (Fig. 4). Staining intensity correlates with the density of Kb/Ova257-264 complexes on the cell surface (59). One day postinfection, we observed no differences in the level of epitope expressed by various viruses. Two days postinfection, noncytopathic RV*M and cytopathic RVwtM expressed similar epitope levels. At day 3, RV*M and RVwtM maintained the same expression level, which was threefold lower than that of RVwtM. It seems likely that the lower epitope expression of viruses containing the extra VSV M gene is due to transcriptional attenuation, a feature of all rhabdoviruses, which results in greater expression of upstream genes (3, 25). The key point is that levels of protein and epitope expression by RVwtM and RV*M are comparable.

FIG. 4.
FIG. 4.

Recombinant RVs express variable levels of Ova257-264 epitope. MC57G cells were infected with RVnoM-NP/S, RVwtM-NP/S, or RV*M-NP/S at an MOI of 1 in triplicate, and at different times postinfection, the cells were stained with an anti-Kb/Ova257-264 antibody followed by a secondary FITC-labeled antibody and analyzed by flow cytometry for the mean fluorescence intensity (MFI). Three separate experiments showed similar results. P values (shown above bars) were calculated using Student's t test.

Cross-priming capacities of recombinant RVs.As discussed, cytolysis is expected to enhance cross-priming capacity. To test this, we infected P815 cells (that do not express H-2Kb) with each recombinant virus for 2 days and injected them into C57BL/6 mice (H-2Kb background). Prior to injection, the cells were UV irradiated in order to inactivate any infectious viruses that might directly infect professional APCs in vivo. The efficacy of inactivation was tested by viral titration after UV treatment, and no infectious virus was detected. Seven days after transfer, Ova257-264-specific TCD8+-cell responses were assessed by ELISPOT assay for IFN-γ-producing cells (Fig. 5). Our results show that the Ova257-264 epitope is cross-presented by both types of viruses but that the cytopathic virus induces significantly higher levels of cross-priming.

FIG. 5.
FIG. 5.

Cross-priming of recombinant RVs. P815 cells were infected with the recombinant viruses at an MOI of 3 for 2 days and then UV irradiated. Cells were injected i.p. into C57BL/6 mice, and 7 days later, the Ova257-264-specific TCD8+ responses were assayed by ELISPOT assay for IFN-γ-positive cells. The data represent three experiments with two mice per group. For each experiment, the maximal response was set at 100%, and the other responses were related to the maximal value. The variances for the two groups were sufficiently similar (0.039 and 0.037) to allow application of a paired Student t test assuming equal variance.

In vivo rabies and Ova257-264-specific immune responses.The immunogenicity of cytopathic and noncytopathic viruses to RV-specific determinants was investigated by infecting mice and analyzing the humoral and TCD8+-cell responses at 8 and 28 days postinfection. We measured both antiglycoprotein (RV G)- and antinucleoprotein (RV N)-specific antibody responses by ELISA. Eight days postinfection, RV N-specific antibodies were not detected. In contrast, RV G-specific antibodies were detectable with cytopathic RVs inducing higher titers than the noncytopathic counterparts (Fig. 6a). By day 28, both anti-RV N and anti-RV G responses were detectable. At this point, responses induced by the cytopathic and noncytopathic viruses to both antigens were similar (Fig. 6b and c).

FIG. 6.
FIG. 6.

In vivo rabies and Ova257-264-specific immune responses induced by recombinant RVs. C57BL/6 mice were infected i.p. with 104 focus-forming units of recombinant RVs as indicated. Mice were bled 8 (a) and 28 (b and c) days after infection, and antibody titers specific for RVG (a and b) and RVN (c) were determined by ELISA. (d) The frequency of rabies-specific, IFN-γ-producing TCD8+ cells was determined at day 8 and 28 postinfection by ELISPOT assay using MC57G cells infected with recombinant RV*M. (e) Ova257-264-specific TCD8+ responses were analyzed by ELISPOT assay at 28 days postinfection by using MC57G cells pulsed with Ova257-264 peptide. For each graph, the bars represent the means for three mice per group. Data are representative of three identical experiments. OD, optical density; Neg, negative.

The kinetics of the RV-specific TCD8+-cell response was also analyzed by ELISPOT assay, which quantifies the number of antigen-specific IFN-γ-secreting cells (Fig. 6d). As with humoral responses, the cytopathic virus RVwtM induced a higher TCD8+-cell response at 8 days postinfection, but by day 28, the numbers of RV-specific TCD8+ cells generated by infections with cytopathic and noncytopathic viruses were similar. The same was true when Ova257-264-specific immune responses at day 28 were analyzed in the same way. The results presented in Fig. 6e showed that infection with cytopathic virus RVwtM activated the same number of epitope-specific TCD8+ cells as the noncytopathic virus RV*M. Thus, degree of cross-priming, as seen in Fig. 5, was not predictive of the level of TCD8+-cell priming (Fig. 6e). Additionally, even though the kinetics of immune responses induced by cytopathic and noncytopathic viruses were different, the long-term responses were comparable.

DISCUSSION

We describe here the development of paired cytopathic and noncytopathic viruses and the consequences of cytopathogenicity for various aspects of the host immune responses. Our results suggest that humoral and cellular long-term immune responses to these viruses are similar but that the dynamics are distinct, with the responses to cytolytic viruses being significantly faster. Further, the cytolytic virus has a greater capacity for cross-presentation, although cross-presentation appears to contribute minimally to TCD8+-cell priming.

A major criticism of previous studies with similar objectives (13, 21, 29) is the use of model cytopathic and noncytopathic viruses that are markedly different with respect to natural host, cell tropism, virulence, and replication rate (41), making it impossible to attribute differences in immune responses to cytopathogenicity alone. In our system, we have eliminated many of these concerns through the use of paired vectors that were minimally different (two amino acids in total). Our system appears to be quite feasible for investigating various aspects of immunity related to viral cytopathogenicity. VSV cytopathogenicity is due in great measure to expression of M protein. Point mutation of VSV M protein (M33 51A) resulted in a virus with reduced and delayed kinetics of the cytopathic effect (27). Consequently, for our analysis we chose a highly attenuated RV strain (38) that is neither cytolytic nor apoptogenic in our hands. Several options were open to us in the engineering of the paired viruses, particularly with respect to positioning of inserted genes. Nonsegmented, negative-stranded RNA virus genomes are transcribed progressively from the 5′ end. A transcription attenuation effect occurs at each gene junction, resulting in progressively lower levels of gene expression (3, 25). We positioned the VSV M genes upstream in the RV genome, between the N and P genes (see Fig. 1), to ensure high expression levels and, consequently, efficient induction of apoptosis by the wild-type product. This strategy met our expectations since the VSVwtM-expressing viruses displayed a much stronger cytopathic phenotype (Fig. 2a and b), which was due, at least in part, to apoptosis (Fig. 2c). Ova257-264 epitope was inserted downstream between the G and L genes. An attenuation effect caused by the presence of an extra upstream gene (VSV M) was expected (25) and evident at the level of surface epitope expression; at day 3, epitope expression from RVwtM and RV*M was ∼3-fold lower than from RVnoM (Fig. 4).

Cytolytic RVs have been generated previously and the impact on the acute humoral response investigated (15, 46). In those cases, cytotoxicity was achieved through expression of cytochrome c (46) or insertion of a second rabies glycoprotein gene (15). Despite these differences and distinct routes of immunization, it is notable that the levels of RV G-specific antibodies in the acute phase of the immune response followed similar patterns, being higher when induced by the cytopathic virus. With our extended studies of immunogenicity, we observed that both humoral (Fig. 6a, b, and c) and TCD8+ responses (Fig. 6d and e) followed similar kinetics, with short-term responses being higher when induced by the cytopathic virus but long-term responses being similar. Even though the RV-based cytopathic and noncytopathic viruses showed similar in vitro growth curves (Fig. 3), it is possible that in vivo the cytopathic virus has a growth advantage due to the potential for VSV M-mediated suppression of antiviral responses (4). This is a possible explanation for the faster response to RVwtM. Interestingly, a comparison of distinct biotypes of bovine viral diarrhea virus indicated higher humoral responses and T-cell proliferation to the noncytopathic biotype at more than 90 days after inoculation. However, in contrast to our system, the bovine viral diarrhea virus isotypes display different patterns of protein expression, possibly influencing their immunogenicity (33).

It has been shown previously that the magnitude of the acute TCD8+ responses directly correlates with the antigen dose when all other variables are held constant (59) and that the acute clonal size influences the magnitude of the long-term responses (24). In our system, the cytopathic and noncytopathic viruses expressed similar epitope levels (Fig. 4), and, consequently, the magnitudes of Ova257-264-specific TCD8+ long-term responses were similar (Fig. 6e). We were unable to carry out the same analysis with natural rabies epitopes, since none has yet been defined. However, the overall response showed higher numbers of rabies-specific TCD8+ cells following cytopathic virus immunization during the initial phase, with levels becoming comparable to those of the noncytopathic virus by day 28 (Fig. 6d). This result supports the concept of immune compensatory mechanisms. Such mechanisms have been found to contribute to similar immune responses developed in wild-type and selectively immunocompromised (26, 28) and knock-out mice (56) but have not been observed with other systems (58).

What might be the basis for such compensation? We propose two mutually inclusive possibilities to explain these differences in response dynamics. First, early after infection, cells infected with cytopathic viruses are expected to drive higher levels of “danger” signaling via release of heat shock proteins, double-stranded RNA, and other factors (20, 30, 37). Second, antigen release would stimulate greater cellular and humoral responses via innate and antigen-specific mechanisms, such as cross-presentation. Although the levels of damage and antigen release mediated by noncytopathic viruses are less severe during the acute phase, the cumulative effects of a more protracted infection may be comparable, ultimately resulting in similar levels of long-term responses.

The nature of the cells providing the antigenic material for cross-presentation is still a matter of debate, experimental evidence suggesting that the source may be live (36), apoptotic (2, 6, 52, 55), or necrotic cells (50). We did not investigate necrosis induction in our system, as VSVwtM-related cytopathogenicity has been documented to occur through apoptosis (5, 14, 23, 31). However, it is possible that different cytopathic viruses induce different levels of cell apoptosis and/or necrosis that may influence the cross-priming outcome. The mechanism of cytopathogenicity may also be an important factor in determining the degree of cross-presentation as well as other outcomes that were investigated here. Our results (Fig. 5) nevertheless support the notion that viral cytopathogenicity enhances cross-presentation.

Notably, the responses to Ova257-264 epitope (Fig. 6e) did not mirror the cross-priming results (Fig. 5), suggesting that, even with the lytic virus, a substantial portion of RV priming occurs through direct presentation. A report that RV infects human dendritic cells (DC) supports this notion (17). Direct priming has been suggested for other viruses as well (18, 19, 43). If viruses have access to professional APCs, direct priming should be more efficient, as cross-priming requires an additional step of antigen transfer (60). However, several groups have reported that apoptotic DC directly expressing antigens can be internalized by live DC that cross-prime TCD8+ cells at early time points postinfection (6, 40) or transfection (47). The balance between direct and cross-presentation is likely unique for each virus, and much more work will be needed to determine how a particular balance is determined.

There is still uncertainty about which host defense mechanisms are more protective against cytopathic and noncytopathic viruses. We did not investigate this question here, but our paired-RV system should be of great utility for exploration of this critical issue. One of the areas most impacted is virus-based therapy, where the host response to the vector is always a concern, whether the goal is induction of protective immune response, modulation of an inefficient response, or gene replacement therapy. Kinetics of an immune response and degree of variability are certainly two factors that will determine the success of a particular application, suggesting that viral cytopathogenicity is a variable that should be considered in designing viral vector-based therapy.

ACKNOWLEDGMENTS

We thank Martin Koser for excellent technical help. We also acknowledge the contributions of James McGettigan, Mary-Ellen Smith, and Gene Tan through numerous discussions about this work.

FOOTNOTES

    • Received 12 January 2006.
    • Accepted 7 April 2006.
  • ↵*Corresponding author. Mailing address: Thomas Jefferson University, Kimmel Cancer Center, 233 S. 10th Street, BLSB 730, Philadelphia, PA 19107. Phone: (215) 503-4540. Fax: (215) 923-4153. E-mail: Laurence.Eisenlohr{at}mail.jci.tju.edu.

REFERENCES

  1. ↵
    Albert, M. L., B. Sauter, and N. Bhardwaj. 1998. Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs. Nature 392:86-89.
    OpenUrlCrossRefPubMedWeb of Science
  2. ↵
    Arrode, G., C. Boccaccio, J. Lule, S. Allart, N. Moinard, J. P. Abastado, A. Alam, and C. Davrinche. 2000. Incoming human cytomegalovirus pp65 (UL83) contained in apoptotic infected fibroblasts is cross-presented to CD8+ T cells by dendritic cells. J. Virol. 74:10018-10024.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Barr, J. N., S. P. Whelan, and G. W. Wertz. 2002. Transcriptional control of the RNA-dependent RNA polymerase of vesicular stomatitis virus. Biochim. Biophys. Acta 1577:337-353.
    OpenUrlCrossRefPubMed
  4. ↵
    Black, B. L., and D. S. Lyles. 1992. Vesicular stomatitis virus matrix protein inhibits host cell-directed transcription of target genes in vivo. J. Virol. 66:4058-4064.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Blondel, D., G. G. Harmison, and M. Schubert. 1990. Role of matrix protein in cytopathogenesis of vesicular stomatitis virus. J. Virol. 64:1716-1725.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Bosnjak, L., M. Miranda-Saksena, D. M. Koelle, R. A. Boadle, C. A. Jones, and A. L. Cunningham. 2005. Herpes simplex virus infection of human dendritic cells induces apoptosis and allows cross-presentation via uninfected dendritic cells. J. Immunol. 174:2220-2227.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Boya, P., A. L. Pauleau, D. Poncet, R. A. Gonzalez-Polo, N. Zamzami, and G. Kroemer. 2004. Viral proteins targeting mitochondria: controlling cell death. Biochim. Biophys. Acta 1659:178-189.
    OpenUrlPubMed
  8. ↵
    Buchmeier, M. J., R. M. Welsh, F. J. Dutko, and M. B. Oldstone. 1980. The virology and immunobiology of lymphocytic choriomeningitis virus infection. Adv. Immunol. 30:275-331.
    OpenUrlCrossRefPubMed
  9. ↵
    Candido, K. A., K. Shimizu, J. C. McLaughlin, R. Kunkel, J. A. Fuller, B. G. Redman, E. K. Thomas, B. J. Nickoloff, and J. J. Mule. 2001. Local administration of dendritic cells inhibits established breast tumor growth: implications for apoptosis-inducing agents. Cancer Res. 61:228-236.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Chen, W., K. A. Masterman, S. Basta, S. M. Haeryfar, N. Dimopoulos, B. Knowles, J. R. Bennink, and J. W. Yewdell. 2004. Cross-priming of CD8+ T cells by viral and tumor antigens is a robust phenomenon. Eur. J. Immunol. 34:194-199.
    OpenUrlCrossRefPubMedWeb of Science
  11. ↵
    Chen, Z., T. Moyana, A. Saxena, R. Warrington, Z. Jia, and J. Xiang. 2001. Efficient antitumor immunity derived from maturation of dendritic cells that had phagocytosed apoptotic/necrotic tumor cells. Int. J. Cancer 93:539-548.
    OpenUrlCrossRefPubMedWeb of Science
  12. ↵
    Chong, L. D., and J. K. Rose. 1993. Membrane association of functional vesicular stomatitis virus matrix protein in vivo. J. Virol. 67:407-414.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Christensen, J. E., J. P. Christensen, N. N. Kristensen, N. J. Hansen, A. Stryhn, and A. R. Thomsen. 2002. Role of CD28 co-stimulation in generation and maintenance of virus-specific T cells. Int. Immunol. 14:701-711.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Desforges, M., G. Despars, S. Berard, M. Gosselin, M. O. McKenzie, D. S. Lyles, P. J. Talbot, and L. Poliquin. 2002. Matrix protein mutations contribute to inefficient induction of apoptosis leading to persistent infection of human neural cells by vesicular stomatitis virus. Virology 295:63-73.
    OpenUrlCrossRefPubMed
  15. ↵
    Faber, M., R. Pulmanausahakul, S. S. Hodawadekar, S. Spitsin, J. P. McGettigan, M. J. Schnell, and B. Dietzschold. 2002. Overexpression of the rabies virus glycoprotein results in enhancement of apoptosis and antiviral immune response. J. Virol. 76:3374-3381.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Finke, S., and K. K. Conzelmann. 1999. Virus promoters determine interference by defective RNAs: selective amplification of mini-RNA vectors and rescue from cDNA by a 3′ copy-back ambisense rabies virus. J. Virol. 73:3818-3825.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Foley, H. D., M. Otero, J. M. Orenstein, R. J. Pomerantz, and M. J. Schnell. 2002. Rhabdovirus-based vectors with human immunodeficiency virus type 1 (HIV-1) envelopes display HIV-1-like tropism and target human dendritic cells. J. Virol. 76:19-31.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Freigang, S., D. Egger, K. Bienz, H. Hengartner, and R. M. Zinkernagel. 2003. Endogenous neosynthesis vs. cross-presentation of viral antigens for cytotoxic T cell priming. Proc. Natl. Acad. Sci. USA 100:13477-13482.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Freigang, S., H. C. Probst, and M. van den Broek. 2005. DC infection promotes antiviral CTL priming: the ‘Winkelried’ strategy. Trends Immunol. 26:13-18.
    OpenUrlCrossRefPubMed
  20. ↵
    Gallucci, S., M. Lolkema, and P. Matzinger. 1999. Natural adjuvants: endogenous activators of dendritic cells. Nat. Med. 5:1249-1255.
    OpenUrlCrossRefPubMedWeb of Science
  21. ↵
    Gray, P. M., G. D. Parks, and M. A. Alexander-Miller. 2003. High avidity CD8+ T cells are the initial population elicited following viral infection of the respiratory tract. J. Immunol. 170:174-181.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Her, L. S., E. Lund, and J. E. Dahlberg. 1997. Inhibition of Ran guanosine triphosphatase-dependent nuclear transport by the matrix protein of vesicular stomatitis virus. Science 276:1845-1848.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    Hobbs, J. A., G. Hommel-Berrey, and Z. Brahmi. 2003. Requirement of caspase-3 for efficient apoptosis induction and caspase-7 activation but not viral replication or cell rounding in cells infected with vesicular stomatitis virus. Hum. Immunol. 64:82-92.
    OpenUrlCrossRefPubMed
  24. ↵
    Hou, S., L. Hyland, K. W. Ryan, A. Portner, and P. C. Doherty. 1994. Virus-specific CD8+ T-cell memory determined by clonal burst size. Nature 369:652-654.
    OpenUrlCrossRefPubMed
  25. ↵
    Iverson, L. E., and J. K. Rose. 1981. Localized attenuation and discontinuous synthesis during vesicular stomatitis virus transcription. Cell 23:477-484.
    OpenUrlCrossRefPubMedWeb of Science
  26. ↵
    Jacobson, E. B., Y. Baine, Y. W. Chen, T. Flotte, M. J. O'Neil, B. Pernis, G. W. Siskind, G. J. Thorbecke, and P. Tonda. 1981. Physiology of IgD. I. Compensatory phenomena in B lymphocyte activation in mice treated with anti-IgD antibodies. J. Exp. Med. 154:318-332.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Jayakar, H. R., and M. A. Whitt. 2002. Identification of two additional translation products from the matrix (M) gene that contribute to vesicular stomatitis virus cytopathology. J. Virol. 76:8011-8018.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Jonjic, S., I. Pavic, P. Lucin, D. Rukavina, and U. H. Koszinowski. 1990. Efficacious control of cytomegalovirus infection after long-term depletion of CD8+ T lymphocytes. J. Virol. 64:5457-5464.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    Kagi, D., P. Seiler, J. Pavlovic, B. Ledermann, K. Burki, R. M. Zinkernagel, and H. Hengartner. 1995. The roles of perforin- and Fas-dependent cytotoxicity in protection against cytopathic and noncytopathic viruses. Eur. J. Immunol. 25:3256-3262.
    OpenUrlCrossRefPubMedWeb of Science
  30. ↵
    Katze, M. G., Y. He, and M. Gale, Jr. 2002. Viruses and interferon: a fight for supremacy. Nat. Rev. Immunol. 2:675-687.
    OpenUrlCrossRefPubMedWeb of Science
  31. ↵
    Kopecky, S. A., and D. S. Lyles. 2003. Contrasting effects of matrix protein on apoptosis in HeLa and BHK cells infected with vesicular stomatitis virus are due to inhibition of host gene expression. J. Virol. 77:4658-4669.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    Koser, M. L., J. P. McGettigan, G. S. Tan, M. E. Smith, H. Koprowski, B. Dietzschold, and M. J. Schnell. 2004. Rabies virus nucleoprotein as a carrier for foreign antigens. Proc. Natl. Acad. Sci. USA 101:9405-9410.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Lambot, M., A. Douart, E. Joris, J. J. Letesson, and P. P. Pastoret. 1997. Characterization of the immune response of cattle against non-cytopathic and cytopathic biotypes of bovine viral diarrhoea virus. J. Gen. Virol. 78:1041-1047.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    Lyles, D. S. 2000. Cytopathogenesis and inhibition of host gene expression by RNA viruses. Microbiol. Mol. Biol. Rev. 64:709-724.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Lyles, D. S., and M. O. McKenzie. 1997. Activity of vesicular stomatitis virus M protein mutants in cell rounding is correlated with the ability to inhibit host gene expression and is not correlated with virus assembly function. Virology 229:77-89.
    OpenUrlCrossRefPubMed
  36. ↵
    Maranon, C., J. F. Desoutter, G. Hoeffel, W. Cohen, D. Hanau, and A. Hosmalin. 2004. Dendritic cells cross-present HIV antigens from live as well as apoptotic infected CD4+ T lymphocytes. Proc. Natl. Acad. Sci. USA 101:6092-6097.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Matzinger, P. 2002. The danger model: a renewed sense of self. Science 296:301-305.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    McGettigan, J. P., R. J. Pomerantz, C. A. Siler, P. M. McKenna, H. D. Foley, B. Dietzschold, and M. J. Schnell. 2003. Second-generation rabies virus-based vaccine vectors expressing human immunodeficiency virus type 1 Gag have greatly reduced pathogenicity but are highly immunogenic. J. Virol. 77:237-244.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    McGettigan, J. P., S. Sarma, J. M. Orenstein, R. J. Pomerantz, and M. J. Schnell. 2001. Expression and immunogenicity of human immunodeficiency virus type 1 Gag expressed by a replication-competent rhabdovirus-based vaccine vector. J. Virol. 75:8724-8732.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Motta, I., F. Andre, A. Lim, J. Tartaglia, W. I. Cox, L. Zitvogel, E. Angevin, and P. Kourilsky. 2001. Cross-presentation by dendritic cells of tumor antigen expressed in apoptotic recombinant canarypox virus-infected dendritic cells. J. Immunol. 167:1795-1802.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Mullbacher, A., M. Regner, Y. Wang, E. Lee, M. Lobigs, and M. Simon. 2004. Can we really learn from model pathogens? Trends Immunol. 25:524-528.
    OpenUrlCrossRefPubMed
  42. ↵
    Norbury, C. C., S. Basta, K. B. Donohue, D. C. Tscharke, M. F. Princiotta, P. Berglund, J. Gibbs, J. R. Bennink, and J. W. Yewdell. 2004. CD8+ T cell cross-priming via transfer of proteasome substrates. Science 304:1318-1321.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    Norbury, C. C., D. Malide, J. S. Gibbs, J. R. Bennink, and J. W. Yewdell. 2002. Visualizing priming of virus-specific CD8+ T cells by infected dendritic cells in vivo. Nat. Immunol. 3:265-271.
    OpenUrlCrossRefPubMedWeb of Science
  44. ↵
    Nowak, A. K., R. A. Lake, A. L. Marzo, B. Scott, W. R. Heath, E. J. Collins, J. A. Frelinger, and B. W. Robinson. 2003. Induction of tumor cell apoptosis in vivo increases tumor antigen cross-presentation, cross-priming rather than cross-tolerizing host tumor-specific CD8 T cells. J. Immunol. 170:4905-4913.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    Porgador, A., J. W. Yewdell, Y. Deng, J. R. Bennink, and R. N. Germain. 1997. Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 6:715-726.
    OpenUrlCrossRefPubMedWeb of Science
  46. ↵
    Pulmanausahakul, R., M. Faber, K. Morimoto, S. Spitsin, E. Weihe, D. C. Hooper, M. J. Schnell, and B. Dietzschold. 2001. Overexpression of cytochrome c by a recombinant rabies virus attenuates pathogenicity and enhances antiviral immunity. J. Virol. 75:10800-10807.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    Racanelli, V., S. E. Behrens, J. Aliberti, and B. Rehermann. 2004. Dendritic cells transfected with cytopathic self-replicating RNA induce crosspriming of CD8+ T cells and antiviral immunity. Immunity 20:47-58.
    OpenUrlCrossRefPubMedWeb of Science
  48. ↵
    Roulston, A., R. C. Marcellus, and P. E. Branton. 1999. Viruses and apoptosis. Annu. Rev. Microbiol. 53:577-628.
    OpenUrlCrossRefPubMedWeb of Science
  49. ↵
    Rovere-Querini, P., and I. E. Dumitriu. 2003. Corpse disposal after apoptosis. Apoptosis 8:469-479.
    OpenUrlCrossRefPubMedWeb of Science
  50. ↵
    Sauter, B., M. L. Albert, L. Francisco, M. Larsson, S. Somersan, and N. Bhardwaj. 2000. Consequences of cell death: exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells. J. Exp. Med. 191:423-434.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    Schaible, U. E., F. Winau, P. A. Sieling, K. Fischer, H. L. Collins, K. Hagens, R. L. Modlin, V. Brinkmann, and S. H. Kaufmann. 2003. Apoptosis facilitates antigen presentation to T lymphocytes through MHC-I and CD1 in tuberculosis. Nat. Med. 9:1039-1046.
    OpenUrlCrossRefPubMedWeb of Science
  52. ↵
    Scheffer, S. R., H. Nave, F. Korangy, K. Schlote, R. Pabst, E. M. Jaffee, M. P. Manns, and T. F. Greten. 2003. Apoptotic, but not necrotic, tumor cell vaccines induce a potent immune response in vivo. Int. J. Cancer 103:205-211.
    OpenUrlCrossRefPubMedWeb of Science
  53. ↵
    Schnell, M. J., H. D. Foley, C. A. Siler, J. P. McGettigan, B. Dietzschold, and R. J. Pomerantz. 2000. Recombinant rabies virus as potential live-viral vaccines for HIV-1. Proc. Natl. Acad. Sci. USA 97:3544-3549.
    OpenUrlAbstract/FREE Full Text
  54. ↵
    Siler, C. A., J. P. McGettigan, B. Dietzschold, S. K. Herrine, J. Dubuisson, R. J. Pomerantz, and M. J. Schnell. 2002. Live and killed rhabdovirus-based vectors as potential hepatitis C vaccines. Virology 292:24-34.
    OpenUrlCrossRefPubMed
  55. ↵
    Subklewe, M., C. Paludan, M. L. Tsang, K. Mahnke, R. M. Steinman, and C. Munz. 2001. Dendritic cells cross-present latency gene products from Epstein-Barr virus-transformed B cells and expand tumor-reactive CD8+ killer T cells. J. Exp. Med. 193:405-411.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    Tanabe, M., T. Matsumoto, K. Shibuya, K. Tateda, S. Miyazaki, A. Nakane, Y. Iwakura, and K. Yamaguchi. 2005. Compensatory response of IL-1 gene knockout mice after pulmonary infection with Klebsiella pneumoniae. J. Med. Microbiol. 54:7-13.
    OpenUrlAbstract/FREE Full Text
  57. ↵
    Thorley-Lawson, D. A., and G. J. Babcock. 1999. A model for persistent infection with Epstein-Barr virus: the stealth virus of human B cells. Life Sci. 65:1433-1453.
    OpenUrlCrossRefPubMedWeb of Science
  58. ↵
    Vazquez-Torres, A., B. A. Vallance, M. A. Bergman, B. B. Finlay, B. T. Cookson, J. Jones-Carson, and F. C. Fang. 2004. Toll-like receptor 4 dependence of innate and adaptive immunity to Salmonella: importance of the Kupffer cell network. J. Immunol. 172:6202-6208.
    OpenUrlAbstract/FREE Full Text
  59. ↵
    Wherry, E. J., K. A. Puorro, A. Porgador, and L. C. Eisenlohr. 1999. The induction of virus-specific CTL as a function of increasing epitope expression: responses rise steadily until excessively high levels of epitope are attained. J. Immunol. 163:3735-3745.
    OpenUrlAbstract/FREE Full Text
  60. ↵
    Zinkernagel, R. M. 2002. On cross-priming of MHC class I-specific CTL: rule or exception? Eur. J. Immunol. 32:2385-2392.
    OpenUrlCrossRefPubMedWeb of Science
  61. ↵
    Zinkernagel, R. M., O. Planz, S. Ehl, M. Battegay, B. Odermatt, P. Klenerman, and H. Hengartner. 1999. General and specific immunosuppression caused by antiviral T-cell responses. Immunol. Rev. 168:305-315.
    OpenUrlCrossRefPubMedWeb of Science
View Abstract
Back to top
Download PDF
Citation Tools
Print

Alerts
Email

Related Articles

Cited By...