Reovirus Apoptosis and Virulence Are Regulated by Host Cell Membrane Penetration Efficiency

Abstract

Apoptosis plays an important role in the pathogenesis of reovirus encephalitis and myocarditis in infected animals. Differences in apoptosis efficiency displayed by reovirus strains are linked to the viral μ1-encoding M2 gene segment. Studies using pharmacologic inhibitors of reovirus replication demonstrate that apoptosis induction by reovirus requires viral disassembly in cellular endosomes but not RNA synthesis. Since the μ1 protein functions to pierce endosomal membranes during this temporal window, these findings point to an important role for μ1 in activating signaling pathways that lead to apoptosis. To understand mechanisms used by μ1 to induce apoptosis, a panel of μ1 mutant viruses generated by reverse genetics was analyzed for the capacities to penetrate host cell membranes, activate proapoptotic signaling pathways, evoke cell death, and produce encephalitis in newborn mice. We found that single amino acid changes within the δ region of μ1 reduce the efficiency of membrane penetration. These mutations also diminish the capacities of reovirus to activate proapoptotic transcription factors NF-κB and IRF-3 and elicit apoptosis. Additionally, we observed that following intracranial inoculation, an apoptosis-deficient μ1 mutant is less virulent in newborn mice in comparison to the wild-type virus. These results indicate a critical function for the membrane penetration activity of μ1 in evoking prodeath signaling pathways that regulate reovirus pathogenesis.

Apoptotic cell death is a common outcome of infections by many viruses (46). In some cases, apoptosis triggered by virus infection serves as a host defense mechanism to limit viral replication, while in others, apoptosis enhances viral infection by facilitating virus spread or allowing the virus to evade host inflammatory or immune responses (19). Virus-induced apoptosis of terminally differentiated, nonrenewing cell populations such as neurons is particularly detrimental to the host. Despite important consequences of apoptosis to viral pathogenesis, little is known about the nature of the viral intermediaries that initiate prodeath signals that contribute to disease.

Mammalian reoviruses are highly tractable models for studies of viral pathogenesis. Reoviruses are nonenveloped icosahedral viruses with a genome of 10 double-stranded RNA (dsRNA) segments (47). After infection of newborn mice, reoviruses disseminate systemically, producing injury to a variety of organs, including the central nervous system (CNS), heart, and liver (55). Infection of mice with type 3 reovirus strains results in fatal encephalitis (47), which is associated with extensive apoptosis at sites of viral replication (40, 43). Furthermore, pharmacologic interference with proapoptotic signaling following reovirus infection alleviates CNS disease (4, 42). In addition, reovirus encephalitis is diminished in mice lacking the NF-κB p50 subunit (40), an essential host cell factor required for apoptosis following reovirus infection (9, 13, 18). These data suggest that reovirus-induced apoptosis contributes to disease pathogenesis in the CNS.

Insight into mechanisms by which reovirus induces apoptosis has emerged from studies of strain-specific differences in the efficiency of apoptosis induction. These analyses pointed to the viral S1 gene as the primary determinant of apoptotic efficiency (11, 45, 52). Since the S1 gene encodes attachment protein σ1 (29, 56), such observations led to the idea that the efficiency of reovirus-induced apoptosis is determined by virus binding to cell surface receptors. However, in addition to the S1 gene, the M2 gene also contributes to differences in the efficiency of apoptosis induction by reovirus (45, 52). The M2 gene encodes outer-capsid protein μ1 (34, 36), which plays an essential role in the penetration of host cell membranes after viral disassembly. The μ1 protein is cleaved by endosomal proteases during viral entry into the δ and φ fragments, which remain associated with the newly generated infectious subvirion particles (ISVPs) (2, 5, 38, 49). The δ fragment undergoes conformational changes to facilitate viral penetration of endosomal membranes (7, 8, 32, 39, 50). Collectively, these data suggest that a μ1-mediated process that occurs during or after viral disassembly also influences apoptotic potential. Pharmacologic inhibitors of reovirus disassembly, but not RNA synthesis, block apoptosis (12, 14), highlighting a key role for disassembly steps in the biochemical pathway that leads to apoptosis induction.

Confirmation of an essential function of μ1 in apoptosis induction by reovirus was provided by experiments in which reovirus attachment to its cognate receptors was uncoupled from viral disassembly by providing an alternative means of viral entry (14). Antibody-mediated delivery of reovirus into normally nonpermissive cells engineered to express Fc receptors leads to productive infection and apoptotic cell death (14). These findings indicate that signaling pathways triggered by virus-receptor interactions are dispensable for reovirus-induced apoptosis and that initiation of prodeath signaling coincides with viral disassembly in cellular endosomes. Furthermore, the M2 gene segment is the sole determinant of strain-specific differences in the capacity of reovirus to induce apoptosis following delivery via Fc-mediated uptake (14). Although these studies point to an important role for the μ1 protein in apoptosis induction, the mechanism by which μ1 triggers apoptosis is undefined.

In this study, we took advantage of our newly developed reverse genetics strategy (26) to identify functional domains in μ1 that modulate proapoptotic signaling following reovirus infection. Using μ1 mutant viruses, we examined the relationships among the capacities of reovirus to penetrate cell membranes, elicit apoptosis, and cause encephalitis in newborn mice. We found that single amino acid substitutions in the δ region of μ1 decrease the capacity of viruses to effect membrane penetration without influencing replication efficiency in cell culture. In addition, these viruses have diminished capacities to activate the proapoptotic transcription factors NF-κB and IRF-3 and evoke apoptosis. These data suggest that the membrane penetration and apoptosis-inductive properties of μ1 are linked. Furthermore, we observed that an apoptosis-deficient μ1 mutant virus displays delayed growth kinetics in the brain following intracranial inoculation and is attenuated in its capacity to produce encephalitis. Thus, μ1-mediated membrane penetration is an essential step in initiating a proapoptotic signaling cascade that modulates the pathogenesis of reovirus disease.

MATERIALS AND METHODS

Cells.

HeLa and 293T cells were maintained in Dulbecco modified Eagle medium (DMEM) supplemented to contain 10% fetal bovine serum (FBS), 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 25 ng/ml amphotericin B (Invitrogen, Carlsbad, CA). L929 cells were maintained in Joklik's minimal essential medium supplemented to contain 10% FBS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 25 ng/ml amphotericin B.

Viruses.

Recombinant reoviruses were generated by plasmid-based reverse genetics (26). Monolayers of L929 cells at approximately 90% confluence (3 × 10⁶ cells) in 60-mm dishes (Costar, Cambridge, MA) were infected with rDIs-T7pol at a multiplicity of infection (MOI) of 0.5 50% tissue culture infective dose. At 1 h postinfection, cells were cotransfected with 10 plasmid constructs representing the cloned type 3 Dearing (T3D) genome—pT7-L1T3D (2 μg), pT7-L2T3D (2 μg), pT7-L3T3D (2 μg), pT7-M1T3D (1.75 μg), pT7-M2T3D (1.75 μg), pT7-M3T3D (1.75 μg), pBacT7-S1T3D (2 μg), pT7-S2T3D (1.5 μg), pT7-S3T3D (1.5 μg), and pT7-S4T3D (1.5 μg)—by using 3 μl of TransIT-LT1 transfection reagent (Mirus, Madison, WI) per μg of plasmid DNA. Following 5 days of incubation, recombinant virus was isolated from transfected cells by plaque purification with monolayers of L929 cells (54). For generation of μ1 mutant viruses, pT7-M2T3D was altered by QuikChange (Stratagene, Carlsbad, CA) site-directed mutagenesis. To confirm sequences of mutant viruses, viral RNA was extracted from purified virions and subjected to Onestep reverse transcription-PCR (QIAGEN) with M2-specific primers (primer sequences are available upon request). PCR products were analyzed following electrophoresis in Tris-borate-EDTA agarose gels or purified and subjected directly to sequence analysis. The presence of a noncoding signature mutation in the L1 gene of viruses generated by plasmid-based rescue also was confirmed by reverse transcription-PCR with L1-specific primers (26).

Purified reovirus virions were generated with second- or third-passage L-cell lysate stocks of twice-plaque-purified reovirus as previously described (16). Viral particles were Freon extracted from infected cell lysates, layered onto 1.2- to 1.4-g/cm³ CsCl gradients, and centrifuged at 62,000 × g for 18 h. Bands corresponding to virions (1.36 g/cm³) (48) were collected and dialyzed in virion storage buffer (150 mM NaCl, 15 mM MgCl₂, 10 mM Tris-HCl [pH 7.4]). The concentration of reovirus virions in purified preparations was determined from an equivalence of 1 U of optical density at 260 nm equals 2.1 × 10¹² virions (48). Viral titer was determined by plaque assay with L929 cells (54). ISVPs were generated by incubation of 2 × 10¹² virions with 200 μg/ml N-p-tosyl-l-lysine chloromethyl ketone-treated chymotrypsin in a total volume of 0.1 ml at 37°C for either 20 min (for ⁵¹Cr release assays) (21) or 1 h (for hemolysis and α-sarcin coentry assays) (37). Proteolysis was terminated by incubation of reaction mixtures on ice and addition of 2 mM phenylmethylsulfonyl fluoride. The generation of ISVPs was confirmed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and Coomassie brilliant blue staining.

Antibodies and plasmids.

Rabbit antisera specific for IRF-3 and p65/RelA were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse ascites specific for the PSTAIR peptide of Cdk1 was purchased from Sigma-Aldrich (St. Louis, MO). Horseradish peroxidase-conjugated anti-rabbit and anti-mouse secondary antibodies were purchased from Amersham Biosciences (Piscataway, NJ). Plasmid p-55C1BLuc (58) was obtained from Takashi Fujita. Plasmids pRenilla-Luc and pNF-κB-Luc (6) were obtained from Dean Ballard.

Virus replication.

Monolayers of L929 cells in 24-well plates (Costar) were adsorbed in triplicate with each reovirus strain at an MOI of 2 PFU/cell at room temperature for 1 h in serum-free medium, washed once with phosphate-buffered saline (PBS), and incubated in serum-containing medium for various intervals. Cells were frozen and thawed twice prior to viral titer determination by plaque assay with L929 cells (54). Viral yields were calculated according to the following formula: log₁₀ yield_tx = log₁₀ (PFU/ml)_tx − log₁₀ (PFU/ml)_t0, where tx is the time postinfection.

Assessment of viral infectivity by indirect immunofluorescence.

Monolayers of L929 cells (2 × 10⁵) in 24-well plates were adsorbed with reovirus at an MOI of 10⁴ particles/cell at room temperature for 1 h. Following removal of the inoculum, cells were washed with PBS and incubated in complete medium at 37°C for 18 h to permit the completion of a single cycle of viral replication. Monolayers were fixed with 1 ml of methanol at −20°C for a minimum of 30 min, washed twice with PBS, blocked with 2.5% immunoglobulin-free bovine serum albumin (Sigma-Aldrich) in PBS, and incubated with polyclonal rabbit anti-reovirus serum at a 1:1,000 dilution in PBS-0.5% Triton X-100 (TX-100) at room temperature for 1 h. Monolayers were washed twice with PBS-0.5% TX-100 and incubated with a 1:1,000 dilution of Alexa Fluor 546-labeled anti-rabbit immunoglobulin G. Monolayers were washed with PBS, and infected cells were visualized by indirect immunofluorescence with an Axiovert 200 fluorescence microscope (Carl Zeiss, New York, NY). Infected cells were identified by the presence of intense cytoplasmic fluorescence that was excluded from the nucleus. No background staining of uninfected control monolayers was noted. Reovirus antigen-positive cells were quantified by counting fluorescent cells in at least two random fields of view in triplicate wells at a magnification of ×20 or by counting the entire well for duplicate wells.

Assessment of ethanol sensitivity.

Absolute ethanol (Aaper Alcohol and Chemical Co., Shelbyville, KY) was diluted in PBS to a final concentration of 33%. Purified reovirus particles were incubated in PBS or PBS-33% ethanol at 37°C for 30 min. Treated virus was placed on ice and diluted 1:10 into PBS. Viral titers in the treatment mixtures were determined by plaque assay with L929 cells (54).

Hemolysis assay.

Citrated calf red blood cells (Colorado Serum Co., Denver, CO) were washed extensively with chilled PBS supplemented to contain 2 mM MgCl₂ (PBS-Mg) and resuspended at a concentration of 30% (vol/vol) in PBS-Mg. Hemolysis efficiency was analyzed by mixing a 3-μl aliquot of resuspended red blood cells with virion storage buffer, 1% TX-100 in virion storage buffer, or 5.4 × 10¹⁰ ISVPs in a total volume of 30 μl and incubating the mixture at 37°C for 1 h. Samples were placed on ice for 30 min to prevent further hemolysis, followed by centrifugation at 500 × g at 4°C for 3 min. The extent of hemoglobin release was quantified by measuring the A₄₁₅ of a 1:5 dilution of the supernatant in a microplate reader (Molecular Devices, Sunnyvale, CA). Percent hemolysis was calculated with the following formula: % hemolysis = 100 × [(A_sample − A_buffer)/(A_TX-100 − A_buffer)].

Chromium release assay.

L929 cells (2 × 10⁶) were incubated in 0.5 ml of serum-free medium supplemented to contain 100 μCi ⁵¹Cr-labeled sodium chromate at 37°C for 2 h. Cells were pelleted by centrifugation, washed twice with chilled PBS, and divided into aliquots of 6 × 10⁴ cells. Cells were adsorbed with each virus strain at 10⁵ ISVPs/cell at 4°C for 1 h. Infection was initiated following removal of unbound ISVPs by washing with chilled PBS and addition of 0.25 ml of prewarmed complete medium. To determine spontaneous and total ⁵¹Cr released, cells were mock infected or treated with 4% TX-100. Cells were transferred to 96-well plates and incubated at 37°C for 4 h. The plates were centrifuged at 1,000 × g at room temperature for 5 min to pellet unattached cells. The amount of ⁵¹Cr released into 25 μl of medium was quantified by liquid scintillation. Percent lysis was calculated with the following formula: % lysis = 100 × [(cpm_sample − cpm_mock)/(cpm_TX-100 − cpm_mock)].

α-Sarcin coentry assay.

Confluent monolayers of HeLa cells in 96-well plates were preincubated in serum-, cysteine-, and methionine-free DMEM (Invitrogen) at 37°C for 3 h. Cells were prechilled at 4°C for 1 h and adsorbed with each virus strain at 10⁶ ISVPs/cell at 4°C for 1 h. Infection was initiated following the removal of unbound ISVPs by washing with chilled PBS and the addition of prewarmed, cysteine- and methionine-free DMEM supplemented to contain 5% FBS, 22 μCi/ml ³⁵S-labeled methionine-cysteine (DuPont NEN, Wilmington, DE), and either 0 or 50 μg/ml α-sarcin (Calbiochem, San Diego, CA). After incubation for various intervals at 37°C, cells were lysed with 1× RIPA buffer (50 mM Tris [pH 7.5], 50 mM NaCl, 1% TX-100, 1% deoxycholate, 0.1% SDS, 1 mM EDTA) and proteins were precipitated following the addition of chilled trichloroacetic acid (TCA) to a final concentration of 30% by overnight incubation at 4°C. Precipitates were washed successively with 10% TCA and 95% ethanol, air dried, and solubilized in 50 mM unbuffered Tris containing 0.05% SDS. Acid-precipitable radioactive counts were quantified by liquid scintillation to assess the incorporation of ³⁵S-labeled amino acids into nascent proteins. Percent inhibition of protein synthesis at each interval was calculated with the following formula: 100 × (cpm incorporated following treatment with 50 μg/ml α-sarcin/cpm incorporated following treatment with 0 μg/ml α-sarcin).

Assessment of caspase 3/7 activity.

HeLa cells (10⁴) were seeded into black clear-bottom 96-well plates (Costar) and adsorbed with reovirus in serum-free medium at an MOI of 100 PFU/cell at room temperature for 1 h, followed by incubation for 24 h in serum-containing medium. Caspase 3/7 activity was quantified with the Caspase-Glo-3/7 assay system (Promega, Madison, WI) according to the manufacturer's instructions.

Quantitation of apoptosis by AO staining.

HeLa cells (5 × 10⁴) grown in 24-well plates were adsorbed with reovirus at an MOI of 100 PFU/cell at room temperature for 1 h. The percentage of apoptotic cells after 48 h of incubation was determined by acridine orange (AO) staining as previously described (52). For each experiment, >200 cells were counted and the percentage of cells exhibiting condensed chromatin was determined by epi-illumination fluorescence microscopy with a fluorescein filter set (Zeiss, Thornwood, NY).

Immunoblot assay.

HeLa cells were either adsorbed with reovirus at an MOI of 100 PFU/cell or mock infected in serum-free medium at 4°C for 1 h, followed by incubation in serum-containing medium at 37°C for 6 h. Nuclear extracts were prepared by hypotonic lysis of cells, followed by high-salt extraction of nuclear proteins as previously described (13). Extracts (10 to 20 μg of total protein) were resolved by electrophoresis in 4 to 20% polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were blocked for at least 1 h in blocking buffer (PBS containing 0.1% Tween 20 and 5% milk or 2.5% bovine serum albumin) and incubated with primary antibodies diluted 1:200 (for IRF-3), 1:500 (for p65/RelA), or 1:7,500 (for PSTAIR) in blocking buffer at room temperature for 1 h. Membranes were washed three times for 10 min each with washing buffer (PBS containing 0.1% Tween 20) and incubated with horseradish peroxidase-conjugated goat anti-rabbit (for IRF-3 or p65/RelA) or anti-mouse (for PSTAIR) immunoglobulin diluted 1:2,000 in blocking buffer. Following three washes, membranes were incubated for 1 min with chemiluminescent peroxidase substrate (Amersham Biosciences) and exposed to film.

Luciferase assays.

293T cells grown in 24-well plates were transfected at 0.18 μg/well with a reporter plasmid that expresses firefly luciferase under either IRF-3/7 or NF-κB control (p-55C1BLuc or pNF-κB-Luc, respectively) and at 0.02 μg/well with pRenilla-Luc, which expresses Renilla luciferase constitutively, by using FuGENE 6 (Roche, Indianapolis, IN) according to the manufacturer's instructions. After incubation for 24 h, transfected cells were adsorbed at 100 PFU/cell with reovirus in serum-free medium at room temperature for 1 h, followed by incubation at 37°C in serum-containing medium for 24 h. Luciferase activity in the cultures was quantified with the Dual-Luciferase Assay Kit (Promega) according to the manufacturer's instructions.

Infection of mice.

Two-day-old ND4 Swiss Webster mice (Harlan, Indianapolis, IN) were inoculated intracranially with 500 PFU of purified recombinant strain T3D (rsT3D) or I442V in a volume of 5 μl into the left cerebral hemisphere with a Hamilton syringe and a 30-gauge needle (51). For analysis of viral virulence, mice were monitored for weight loss and symptoms of disease for 21 days. For survival experiments, mice were euthanized when they were found to be moribund (defined by rapid or shallow breathing, lethargy, or paralysis). For analysis of virus growth, mice were euthanized at various intervals following inoculation and brains were collected into 1 ml of PBS and homogenized by freezing, thawing, and sonication. Viral titers in brain homogenates were determined by plaque assay (54). Animal husbandry and experimental procedures were performed in accordance with Public Health Service policy and approved by the Vanderbilt University School of Medicine Institutional Animal Care and Use Committee.

Histology of mouse brains.

Two-day-old ND4 Swiss Webster mice were inoculated intracranially with 500 PFU of either rsT3D or I442V. Mice were euthanized, brains were resected, and brain tissues were fixed overnight in 10% formalin, followed by 70% ethanol. Fixed organs were embedded in paraffin, and 6-μm histological sections were prepared. Consecutively obtained sections were stained with hematoxylin and eosin (H&E) for evaluation of histopathologic changes or processed for immunohistochemical detection of reovirus protein or activated caspase 3 (40).

RESULTS

Viruses with mutations in the δ region of μ1 are viable and display resistance to inactivation by ethanol.

Reovirus variants selected for resistance to inactivation by 33% ethanol (21, 57) or high temperature (35) have mutations in the δ region of the μ1-encoding M2 gene segment (Fig. 1A). These viruses also display a diminished capacity to mediate chromium release from preloaded cells (21) or elicit hemolysis (8, 35). The reduced capacity of mutant μ1 viruses to permeabilize membranes is linked to an increase in viral particle stability (21), which likely alters conformational changes that facilitate membrane penetration (8). These data suggest that residues within the δ region of μ1 regulate membrane penetration by affecting the overall stability of μ1. Structural studies of μ1 indicate that residues in δ responsible for membrane penetration efficiency also are in a region that mediates interaction between μ1 and σ3 molecules in the viral capsid (30, 59).

FIG. 1.

Single amino acid changes in membrane penetration protein μ1 alter the ethanol sensitivity of reovirus. (A) Crystal structure of a μ1 monomer (Protein Data Bank accession no. 1JMU) (30). The μ1 δ and φ domains are in gray and red, respectively. The locations of amino acids P233, A319, V425, Q440, and I442, which are altered in ethanol-resistant reovirus mutants (57), are shown. (B) rsT3D or the indicated μ1 mutant virus was treated with 0 or 33% ethanol at 37°C for 30 min. Titers of the treatment mixtures were determined by plaque assay. Results are expressed as mean viral titers of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D.

To determine the relationship between membrane penetration and apoptosis induction, we used reverse genetics (26) to rescue wild-type rsT3D and viruses with mutations present in the μ1 protein of ethanol-resistant mutants 3a1 (P233S and I442V), 3a5 (A319E), 3b4 (Q440L), and 3d5 (V425F) (57) in an otherwise rsT3D background (Fig. 1A). Viruses with these substitutions were found to be viable and were rescued with equivalent efficiency (data not shown). The particle-to-PFU ratios of the viruses used in this study are shown in Table 1. To examine whether these mutations confer resistance to viral inactivation by ethanol, we quantified residual infectivity of rsT3D and each μ1 mutant strain following treatment with 33% ethanol (Fig. 1B). Incubation of wild-type rsT3D and the P233S mutant with ethanol decreased viral infectivity by 10⁵-fold. In contrast, incubation of A319E, V425F, Q440L, and I442V resulted in a 10³-fold decrease in viral infectivity. These data indicate that the changes at amino acids 319, 425, 440, and 442, respectively, contribute to the relative ethanol resistance of mutant viruses 3a5, 3d5, 3b4, and 3a1.

TABLE 1.

Infectious potentials of wild-type and mutant reovirus strains

Virus straina	Particle/PFU ratiob	Particle/FFU ratioc on:
		L929 cells	HeLa cells
rsT3D	8.3 × 102	3.9 × 104	3.2 × 105
P233S	1.2 × 103	2.6 × 104	1.6 × 105
A319E	3.8 × 103	1.1 × 105	1.3 × 106
V425F	7.2 × 102	2.1 × 104	4.9 × 105
Q440L	1.2 × 103	3.3 × 104	5.7 × 105
I442V	1.3 × 103	2.8 × 104	2.8 × 105

^aViral concentration (in virions per milliliter) was determined by measuring the absorbance of the purified viral preparation at 260 nm.

^bViral titer (in PFU per milliliter) was determined by plaque assay with L929 cells. The particle-to-PFU ratio was determined by using the following formula: viral concentration/viral titer on L929 cells.

^cViral infectivity (in fluorescent-focus units [FFU] per milliliter) was determined by indirect immunofluorescence with L929 or HeLa cells. The particle-to-FFU ratio was determined by using the following formula: viral concentration/viral infectivity on L929 or HeLa cells.

Mutations in δ diminish membrane penetration efficiency.

To ascertain the membrane penetration properties of the μ1 mutant viruses, we first tested ISVPs generated from each virus for the capacity to lyse erythrocytes (Fig. 2A). The capacity of reovirus ISVPs to perturb erythrocyte membrane integrity and cause hemolysis correlates with endosomal membrane penetration (1, 8). Incubation of bovine erythrocytes with either rsT3D or P233S caused efficient hemolysis, whereas the remaining μ1 mutants, A319E, V425F, Q440L, and I442V, failed to elicit significant levels of erythrocyte lysis. To determine whether these residues also affect membrane penetration in cells permissive for reovirus infection, we next tested the capacities of the wild-type and μ1 mutant viruses to mediate the release of ⁵¹Cr from preloaded L929 cells (Fig. 2B). Incubation of L929 cells with A319E, V425F, Q440L, and I442V resulted in significantly less ⁵¹Cr release in comparison to rsT3D and P233S. These data suggest an important role for δ residues A319, V425, Q440, and I442 in membrane perforation.

FIG. 2.

μ1 mutants exhibit altered membrane penetration capacities. (A) A 3% (vol/vol) solution of bovine erythrocytes was incubated with 5.4 × 10¹⁰ ISVPs of rsT3D or the indicated μ1 mutant at 37°C for 1 h. Hemolysis was quantified by determining the absorbance of the supernatant at 415 nm. Hemolysis following treatment of an equal number of cells with virion storage buffer or virion storage buffer containing 1% TX-100 was considered to be 0 or 100%, respectively. Results are expressed as the mean percent hemolysis of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D. (B) L929 cells preincubated with ⁵¹Cr-labeled sodium chromate were adsorbed with rsT3D or the indicated μ1 mutant at 10⁵ ISVPs/cell at 4°C for 1 h and incubated at 37°C for 4 h following addition of complete medium. The amount of ⁵¹Cr released into the medium was determined by liquid scintillation. ⁵¹Cr release following treatment of an equal number of cells with virion storage buffer or by addition of 4% TX-100 to the medium was considered to be 0 or 100%, respectively. Results are expressed as the mean percent lysis of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D. (C and D) HeLa cells starved of cysteine and methionine were adsorbed with rsT3D or the indicated μ1 mutant at 10⁶ ISVPs/cell at 4°C for 1 h. Infection was initiated in medium containing ³⁵S-labeled cysteine and methionine in the presence or absence of α-sarcin. Cells were lysed following incubation at 37°C for either 3 h (C) or the indicated times (D). Proteins were precipitated with TCA, and acid-precipitable radioactivity was quantified by scintillation counting. Protein synthesis following infection by each virus in the absence of α-sarcin was considered to be 100%. Results are expressed as mean protein synthesis in the presence of α-sarcin of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D.

As a surrogate measurement of the capacity of these viruses to deliver viral cores into the cytosol, we tested the capacities of rsT3D and each μ1 mutant to allow α-sarcin coentry in the wake of virus infection (Fig. 2C). Inhibition of cellular protein synthesis following coentry of α-sarcin is determinative of efficient membrane penetration by reovirus (8, 33). Quantification of protein synthesis inhibition at 3 h following infection indicated that, with the exception of the A319E mutant, all of the viruses used in this study efficiently inhibited protein synthesis in the presence of α-sarcin. These data suggest that most of the μ1 mutant viruses analyzed are capable of breaching cellular membranes at an interval of 3 h after initiation of infection.

To determine whether the discrepancy in the assessment of membrane penetration efficiency of the μ1 mutants by cytolysis assays (Fig. 2A and B) and α-sarcin coentry assays (Fig. 2C) is related to an alteration in the kinetics of membrane penetration, we compared the capacities of rsT3D and μ1 mutant viruses A319E and I442V to inhibit cellular translation in the presence of α-sarcin over a time course (Fig. 2D). Infection of cells with rsT3D resulted in rapid and efficient inhibition of host cell protein synthesis. In contrast, infection with either A319E or I442V resulted in dampened inhibition of cellular translation. Notably, the A319E mutation reduced the membrane penetration efficiency of reovirus to a greater extent than I442V. These data indicate that although both A319E and I442V penetrate membranes with diminished efficiency, subtle differences exist in the capacities of these μ1 mutant viruses to breach cellular membranes.

Viral growth and infectivity are affected when membrane penetration capacity is severely compromised.

To evaluate whether differences in viral membrane penetration efficiency affect viral growth over a single cycle of reovirus replication, we infected L929 cells with rsT3D and each μ1 mutant virus at an MOI of 2 PFU/cell and quantified viral titers at 0, 12, and 24 h following infection (Fig. 3A). At 24 h postinfection, all of the viruses produced an ∼1,000-fold increase in infectious progeny, except A319E, which produced only an ∼100-fold increase. These data indicate that introduction of most of the μ1 mutations that compromise viral membrane penetration capacity do not affect overall viral yield. Interestingly, the A319E mutation, which we found to have the most profound effect on membrane penetration (Fig. 2), also diminished viral replication.

FIG. 3.

Growth and infectivity of membrane penetration-inefficient μ1 mutants. (A) L929 cells were adsorbed with rsT3D or the indicated μ1 mutant at an MOI of 2 PFU/cell. Titers of virus in cell lysates at the indicated intervals postinfection were determined by plaque assay. Results are expressed as viral yields of triplicate samples. Error bars indicate standard deviations. (B) L929 cells were adsorbed with rsT3D or the indicated μ1 mutant at an MOI of 10⁴ particles/cell. After incubation at 37°C for 18 h, cells were fixed with methanol. Infected cells were visualized by immunostaining with a polyclonal rabbit anti-reovirus serum, followed by incubation with Alexa Fluor 546-labeled anti-rabbit immunoglobulin G. Reovirus-infected cells were quantified by counting fluorescent cells. Results are expressed as the mean number of fluorescent-focus units (FFU) per field of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D.

To determine whether the reduced yields of A319E are a function of its capacity to establish an initial round of infection, we quantified the number of L929 cells expressing viral protein following infection with rsT3D and each mutant virus (Fig. 3B). We found that infection of L929 cells with A319E produced ∼70% fewer infected cells in comparison to infection with rsT3D and each of the other μ1 mutant viruses. Concordantly, the infectivity of A319E on a per-cell basis was substantially less than that of the other viruses tested in this study (Table 1). These data suggest that the decrease in viral yield following infection with A319E is related to diminished infectivity, which is likely attributable to its reduced membrane penetration efficiency.

Membrane penetration-inefficient μ1 mutants possess a diminished capacity for apoptosis induction.

Our previous studies indicate that reovirus-induced apoptosis does not require de novo synthesis of viral RNA and protein (12, 14), suggesting that the proapoptotic stimulus is contained within infecting viral capsids. Genetic analyses have pointed to the μ1-encoding M2 gene segment as a critical viral determinant of apoptosis efficiency (14, 52, 53). However, the lack of otherwise isogenic wild-type and μ1 mutant viruses has precluded a formal evaluation of the relationship between μ1-mediated membrane penetration and apoptosis induction. To address these questions, we quantified the capacities of rsT3D and each μ1 mutant virus to stimulate effector caspase activity following infection of host cells (Fig. 4A). Infection of HeLa cells with either rsT3D or P233S resulted in an ∼2.5-fold increase in caspase 3/7 activity 24 h following reovirus infection in comparison to that in mock-infected control cells. In contrast, infection with each mutant virus produced an ∼1.5-fold increase in caspase 3/7 activity, suggesting that residues within the δ domain of μ1 affect proapoptotic signaling.

FIG. 4.

Membrane penetration-inefficient μ1 mutants are apoptosis defective. (A) HeLa cells were adsorbed with rsT3D or the indicated μ1 mutant at an MOI of 100 PFU/cell. Following 24 h of incubation, caspase 3/7 activity in cell lysates was determined. Results are expressed as the mean ratio of caspase 3/7 activity from infected cell lysates to that from mock-infected cells of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to cells infected with rsT3D. (B) HeLa cells were infected with rsT3D or the indicated μ1 mutant at an MOI of 100 PFU/cell. Following 48 h of incubation, the percentage of apoptotic cells was determined by staining with AO. Results are expressed as the mean percentage of apoptotic cells of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D.

To determine whether the observed differences in activation of effector caspase activity are sufficient to affect viral apoptotic potential, we quantified apoptotic cells 48 h following infection with each virus strain (Fig. 4B). Following infection of HeLa cells with either rsT3D or P233S, ∼40 to 50% of the cells showed altered nuclear morphology characteristic of apoptotic cells. However, infection with the A319E, V425F, Q440L, or I442V mutant virus produced apoptosis in only ∼15 to 25% of the cells. Collectively, these data indicate that membrane penetration-inefficient μ1 mutants are defective in apoptosis induction and suggest a functional link between these two μ1 properties.

Membrane penetration-inefficient μ1 mutants are diminished in the activation of proapoptotic transcription factors NF-κB and IRF-3.

The activity of transcription factor NF-κB is required for induction of apoptosis following infection with reovirus (13, 18). To test whether differences in the apoptotic potentials of μ1 mutants are related to the activation of NF-κB, we examined these viruses for nuclear translocation of the NF-κB RelA/p65 subunit following infection. For these experiments, HeLa cells infected with each viral strain were fractionated into nuclear and cytoplasmic compartments 6 h postinfection and nuclear extracts were analyzed for RelA by immunoblotting (Fig. 5A). While mock-infected cells showed minimal RelA in the nuclear fraction, infection with rsT3D and P233S led to the accumulation of substantially more nuclear RelA. The amount of nuclear RelA following infection with each apoptosis-defective μ1 mutant was appreciably less than that observed following infection with rsT3D and P233S.

FIG. 5.

μ1 mutants activate NF-κB and IRF-3 with diminished efficiency. (A and C) HeLa cells were adsorbed with rsT3D or the indicated μ1 mutant virus at an MOI of 100 PFU/cell. Nuclear extracts were prepared at 6 h postinfection, resolved by SDS-polyacrylamide gel electrophoresis, and immunoblotted with an antiserum specific for RelA/p65 (A, top) or IRF-3 (C, top). As a control for protein concentration, the blots were stripped and reprobed with an antibody specific for the PSTAIR peptide of Cdk1 (A and C, bottom). (B and D) 293T cells were cotransfected with either pNF-κB-Luc (B) or p-55C1BLuc (D) and pRenilla-Luc 24 h prior to adsorption with the indicated virus at an MOI of 100 PFU/cell. Luciferase activity in cell lysates was determined at 24 h postinfection. Results are presented as the ratio of normalized luciferase activity of infected cell lysates to that of mock-infected lysates of triplicate samples. Error bars indicate standard deviations. *, P < 0.05 as determined by Student's t test in comparison to rsT3D.

To determine whether differences in nuclear RelA levels affect NF-κB-dependent gene expression, we transfected 293T cells with a reporter plasmid that expresses firefly luciferase under the control of an NF-κB-regulated promoter. 293T cells were chosen for these experiments because of the ease of transfection. To normalize for transfection efficiency, we used a plasmid that constitutively expresses Renilla luciferase. The capacity of each virus to stimulate NF-κB activity was quantified by assessment of luciferase activity in lysates of infected cells (Fig. 5B). In concordance with the biochemical data, infection of 293T cells with rsT3D and P233S led to an approximately threefold increase in NF-κB-dependent gene expression. In contrast, NF-κB reporter gene activity was enhanced no more than twofold following infection with each apoptosis-defective μ1 mutant. These findings suggest that NF-κB activation is dependent on efficient membrane penetration by μ1. Moreover, by linking μ1 function to the activation of NF-κB, these data provide a possible explanation for the apoptosis defect observed following infection with these mutant viruses.

In addition to NF-κB, transcription factor IRF-3 is activated following reovirus infection and contributes to efficient apoptosis induction (20). To examine whether mutations in μ1 affect the capacity of reovirus to activate IRF-3, virus-induced nuclear translocation of IRF-3 was monitored by immunoblotting 6 h following infection of HeLa cells (Fig. 5C). Infection with rsT3D and P233S resulted in substantially greater nuclear IRF-3 levels in comparison to those in mock-infected cells. However, infection with the other μ1 mutants resulted in diminished accumulation of IRF-3 in the nucleus. To determine the functional consequences of these differences in IRF-3 localization, we used the p-55C1BLuc reporter plasmid (58), which expresses firefly luciferase under the control of tandem IRF-binding sites (Fig. 5D). Infection of reporter plasmid-transfected 293T cells with rsT3D, P233S, and Q440L resulted in an approximately fourfold increase in IRF-3 reporter gene activity 24 h following infection. However, infection with the other μ1 mutants caused only an approximately twofold increase in IRF-3 activity, indicating that the level of IRF-3 translocated to the nucleus following infection by these viruses is not sufficient to mediate maximal IRF-3-dependent gene expression. The apparent incongruity in IRF-3 activation by nuclear localization and reporter gene expression following infection with rsT3D, P233S, and Q440L may be attributable to kinetic differences in the nuclear localization of IRF-3 following infection with these viruses. Since the times at which nuclear localization (6 h postinfection) and reporter gene activity (24 h postinfection) were assayed vary considerably, a small kinetic difference in nuclear translocation of IRF-3 that is detectable early may not manifest itself at a later time when transcriptional activity of IRF-3 is assessed. Nonetheless, these data suggest that a μ1 function associated with membrane penetration also leads to activation of IRF-3.

A membrane penetration-inefficient, apoptosis-defective μ1 mutant is attenuated in mice.

To determine whether in vitro differences in the apoptosis-inducing capacities of μ1 mutants affect reovirus disease, we inoculated newborn mice with rsT3D and apoptosis-defective mutant I442V. This mutant was selected from the panel of apoptosis-defective viruses since our cell culture data indicated that the isoleucine-to-valine substitution at amino acid 442 in δ alters the apoptotic potential of reovirus without affecting its replication efficiency (Fig. 3 and 4). To assess the virulence of these viral strains, newborn mice were inoculated intracranially with 500 PFU of either rsT3D or I442V and weighed daily for 21 days as a marker of reovirus disease progression (Fig. 6A). Weight gain in rsT3D-infected mice ceased at ∼7 days postinfection. In contrast, I442V-infected mice continued to gain weight until ∼10 days following infection. These data suggest a delay in reovirus disease following infection with the I442V mutant virus.

FIG. 6.

An apoptosis-defective μ1 mutant displays attenuated virulence. ND4 Swiss Webster mice were inoculated intracranially with 500 PFU of rsT3D or I442V mutant virus. Mice (n = 18 to 20) were monitored for weight gain (A) and survival (B). *, P < 0.001 as determined by log rank test in comparison to rsT3D. (C) Brains of infected mice were resected at the times shown and homogenized by freeze-thawing and sonication. Viral titers in brain homogenates were determined by plaque assay. Results are expressed as the viral titer in the brain of each infected animal indicated by a circle. Mean viral titers are indicated by horizontal black lines. *, P < 0.05 as determined by Student's t test in comparison to rsT3D at the same time after infection.

Intracranial inoculation of 2-day-old mice with type 3 reovirus strains such as T3D results in 100% mortality at doses as low as 10 PFU (51). To determine whether the I442V mutation alters lethality, infected mice were monitored for survival for an interval of 21 days (Fig. 6B). In concordance with the weight data, there was a significant difference in the survival of mice infected with rsT3D and I442V. The median survival time of rsT3D-infected mice was 11 days, and all infected mice succumbed to infection by 13 days postinfection. In contrast, the median survival time of I442V-infected mice was 13 days and ∼17% of the mice survived infection with I442V. These data indicate that the μ1 I442V mutant virus is attenuated in comparison to rsT3D.

To determine whether the delay in disease onset following infection with I442V results from delayed viral replication kinetics, mice infected with rsT3D and I442V were euthanized at 4, 8, and 10 days after infection and viral titers in resected brains were quantified by plaque assay (Fig. 6C). rsT3D produced significantly higher titers in the brains of infected animals at 4 and 8 days postinfection in comparison to those produced by I442V. Although the titer of rsT3D also was significantly higher than that of I442V at 8 days postinfection, the magnitude of this difference was considerably less. By 10 days postinfection, I442V had attained a titer equivalent to that of rsT3D. These data indicate that I442V replicates with slower kinetics in vivo but eventually reaches the same peak titer as the wild-type virus.

To evaluate the extent of histopathologic injury of the CNS in rsT3D- and I442V-infected mice, we examined H&E-stained coronal brain sections from mice euthanized at 10 days following intracranial viral inoculation (Fig. 7 and 8). Mice infected for 10 days were selected for these histological analyses since rsT3D and I442V reached equivalent titers at that time point. Sections from rsT3D-infected mice showed a striking meningoencephalitis, with inflammatory infiltrates and extensive neuronal death primarily in the cerebral cortex, hippocampus, thalamus, and hypothalamus. In the rsT3D-infected hippocampus, extensive damage to the pyramidal cell layer was evident, often with a sharp demarcation between areas of viable tissue and regions of dead and dying cells (Fig. 7A). Prominent reactive blood vessels and mononuclear inflammatory cells were present in the damaged hippocampus. The dying neurons in the affected brain regions showed condensed, pyknotic, or fragmented nuclei, consistent with an apoptotic mechanism of cell death. Brain sections from I442V-infected mice showed a distribution of injury that was identical to that in brains from animals infected with rsT3D. However, the severity of the injury was reduced in I442V-infected mice, with only focal cell death in the hippocampus and fewer inflammatory infiltrates (Fig. 7B). Similarly, the hypothalamus from infected mice showed evidence of greater damage following infection with rsT3D than the I442V mutant (Fig. 8). Furthermore, enhanced vacuolation of the hypothalamic tissue was evident in rsT3D-infected brains. These findings suggest that the μ1 mutant virus produces less pathological injury than the wild-type virus in the murine CNS.

FIG. 7.

Histopathologic analysis of the hippocampi of mice infected with wild-type or μ1 mutant reovirus. ND4 Swiss Webster mice were inoculated intracranially with 500 PFU of rsT3D or I442V mutant virus. At 10 days postinoculation, brains of infected mice were resected, fixed, and embedded in paraffin. Brain sections were stained with H&E, polyclonal reovirus-specific antiserum, or activated caspase 3-specific antiserum. The hippocampus regions of consecutive sections of the brains of rsT3D-infected (A) and I442V-infected (B) mice are shown. Original magnifications, ×100 (top part of each panel) and ×400 (bottom part of each panel).

FIG. 8.

Histopathologic analysis of the hypothalami of mice infected with wild-type or μ1 mutant reovirus. ND4 Swiss Webster mice were inoculated intracranially with 500 PFU of rsT3D or I442V mutant virus. At 10 days postinoculation, brains of infected mice were resected, fixed, and embedded in paraffin. Brain sections were stained with H&E, polyclonal reovirus-specific antiserum, or activated caspase 3-specific antiserum. The hypothalamus regions of consecutive sections of the brains of rsT3D-infected (A) and I442V-infected (B) mice are shown. Original magnifications, ×100 (top part of each panel) and ×400 (bottom part of each panel).

To determine whether the observed differences in severity of injury result from alterations in viral tropism, sections of infected mouse brains consecutive to those used in the H&E analyses were stained with reovirus-specific antiserum (Fig. 7 and 8). Numerous immunoreactive neurons were detected in the brains of mice infected with either rsT3D or I442V. Antigen-positive neurons were observed in a pattern recapitulating the pathology following H&E staining; the cerebral cortex, hippocampus, thalamus, and hypothalamus were primarily involved. Despite markedly less cell death and inflammation, a similar localization of viral antigen was observed in brains infected with the I442V mutant virus. Moreover, consistent with the brain titers at 10 days postinfection (Fig. 6C), the intensity of reovirus-specific staining was equivalent following infection with either strain of virus in both the hippocampus and the hypothalamus. These data indicate that mutations in μ1 that affect viral membrane penetration and apoptosis induction in cultured cells do not alter reovirus tropism.

To directly test whether rsT3D and I442V vary in the capacity to elicit apoptosis in vivo, sections from infected mouse brains were stained with an antiserum specific for the activated form of caspase 3 as a biochemical marker of apoptosis (Fig. 7 and 8). We found that in the brains of mice infected with rsT3D, regions displaying extensive viral replication also were positive for the presence of activated caspase 3, consistent with previous observations (40, 43). In sharp contrast, reovirus antigen-positive regions of I442V-infected brain showed minimal staining for activated caspase 3. These data suggest that the attenuated virulence of I442V is a consequence of a reduction in its capacity to evoke apoptosis in vivo.

DISCUSSION

The reovirus μ1 protein functions to pierce cellular membranes after proteolytic viral disassembly in cellular endosomes (7, 8, 32, 39, 50). In addition to its membrane penetration function, the μ1-encoding M2 gene segment is linked to differences in apoptosis efficiency displayed by reovirus strains type 1 Lang and T3D (14, 45, 52, 53). Furthermore, plasmid-driven expression of the μ1 protein results in apoptotic cell death (10). To determine whether the membrane penetration and apoptosis induction functions of μ1 are linked, we generated mutant viruses that incorporate single amino acid changes in the δ region of μ1 and characterized them for the capacity to penetrate membranes and activate proapoptotic signaling pathways. We found that viruses with reduced membrane penetration efficiency also display a diminished capacity to activate IRF-3 and NF-κB and elicit apoptosis. These findings suggest that membrane penetration by μ1 is a critical step in the reovirus replication cycle that regulates the activation of signaling cascades that culminate in cell death.

Although the engineered mutations in the δ region of μ1 diminished the membrane penetration capacity and apoptotic potential of mutant viruses, most of these amino acid substitutions (with the exception of A319E) had no effect on viral yield in cell culture. We hypothesize that, as opposed to viral infection, which requires membrane breach by a single infectious virion, apoptosis induction by reovirus requires endosomal penetration by many virions. It is possible that inefficient membrane penetration leads to quantitative or qualitative differences in the capacity of μ1 or its proteolytic fragments to engage cellular sensors. These differences would, in turn, result in weaker death signals.

We envision four possibilities to explain how μ1-mediated membrane penetration might evoke proapoptotic signaling. First, μ1-mediated endosomal disruption could lead to release of proteases such as cathepsins, which might subsequently damage mitochondria and stimulate death signaling (15, 17, 44). This idea is supported by the observation that mitochondrial injury following reovirus infection occurs in a time frame consistent with reovirus entry (27, 28). Second, analogous to cell entry by Sindbis virus (24), membrane rupture mediated by μ1 might activate sphingomyelinases and result in release of the proapoptotic second messenger ceramide. Third, since the δ fragment of μ1 generated during viral disassembly is known to gain access to the cytoplasm following successful penetration of endosomal membranes (8), it is possible that δ initiates proapoptotic signaling, perhaps via its interaction with heat shock factors (23) or other as-yet-unidentified cellular sensors. Fourth, membrane penetration regulated by the δ fragment may allow the delivery of other viral components, such as the μ1 φ domain, into the relevant postendosomal compartment from which proapoptotic signaling is initiated. In support of this idea, ectopic expression of φ in transfected cells leads to apoptosis (10).

Our data indicate that the membrane penetration function of μ1 regulates the activation of both IRF-3 and NF-κB. While we found that efficient membrane penetration by μ1 is prerequisite for NF-κB activation, at least one membrane penetration-inefficient μ1 mutant, Q440L, retained its capacity to activate IRF-3. These data suggest that IRF-3 and NF-κB are activated by distinct mechanisms. This conclusion is fully consistent with our previous studies of proapoptotic signaling with genome-deficient reovirus particles. We previously reported that while reovirus genomic dsRNA is dispensable for NF-κB activation (13), it is required for activation of IRF-3 (20). Furthermore, we have shown that activation of IRF-3 but not NF-κB following reovirus infection requires the function of the cytoplasmic dsRNA sensor RIG-I (20). We therefore hypothesize that the effect of μ1 on IRF-3 activation may be independent of its effect on membrane penetration. Instead, the μ1 protein may regulate RIG-I-mediated recognition of genomic dsRNA by affecting genomic dsRNA release through its effects on overall capsid stability (35, 57). Such genome release would presumably occur as a consequence of aberrant viral disassembly in the endocytic pathway (20).

In addition to identifying μ1-mediated membrane penetration as an important event in eliciting cell death signaling pathways, we also found that an apoptosis-deficient μ1 mutant is attenuated in its capacity to cause encephalitis. Variation in reovirus virulence following intracranial inoculation also has been previously attributed to the M2 gene segment (22). In comparison to prototype strain T3D, which has a 50% lethal dose of 30 PFU, a type 3 field isolate strain, T3C8, was found to have a 50% lethal dose of 3 × 10⁴ PFU. Reassortant analyses indicate that the avirulence of T3C8 is associated with the μ1-encoding M2 gene segment (22). Although the mechanism by which the M2 gene affects virulence was not elucidated in that study, the M2-associated decrease in virulence was associated with diminished efficiency of reovirus replication in the brain without alteration in its tropism (22), similar to our results.

Although the attenuated μ1 mutant I442V reaches a titer equivalent to that of rsT3D in cultured cells, it replicates with slower kinetics in the brains of infected mice in comparison to rsT3D. The decrease in virulence of I442V may be a consequence of slower replication kinetics, diminished apoptosis in target organs, or both effects. It also is possible that apoptosis induction is required for efficient dissemination of reovirus. Thus, a reovirus mutant such as I442V, which possesses a diminished capacity to induce apoptosis, may display a lower replication efficiency due to a defect in cell-to-cell spread within the CNS. Alternatively, it is possible that the slower growth rate of the I442V mutant is a manifestation of the reduced membrane penetration capacity of its mutant μ1 protein. However, since membrane penetration efficiency and apoptotic potential are governed by the same region of μ1, our current structure-function analysis cannot discriminate between these possibilities. Furthermore, previous studies to inhibit neuronal apoptosis following reovirus infection also do not definitively uncouple viral growth and apoptotic potential since these modalities exhibit differential effects on reovirus replication (4, 40, 42). Our ongoing studies with mice that lack critical mediators of apoptosis and additional μ1 mutants will enhance an understanding of the precise role of apoptosis in reovirus replication in vivo. These studies also will help clarify how apoptosis contributes to the pathogenesis of reovirus disease.

In this study, we identified a critical role for μ1-mediated membrane penetration in activating innate immune response signaling pathways that lead to the induction of cell death. Interestingly, other viruses such as murine coronavirus, Sindbis virus, and vaccinia virus also have been reported to require postattachment cell entry events in endosomes to induce apoptosis (25, 31, 41). Our study therefore highlights a potentially conserved, heretofore unknown role for host cell membrane penetration in initiating innate immune responses and prodeath signaling following detection of virus infection. Timely sensing of pathogens during infection is essential for activation of innate immune responses or induction of apoptosis to allow host cells to more efficiently limit viral spread. Sensing of viral genomes or capsid components by Toll-like receptors (TLRs) present within the endosomes is a relatively well-characterized mechanism by which early detection of pathogens is achieved by host cells without the need for viral replication (3). However, not all viruses may be sensed at this stage of infection since their components are either not capable of being recognized by TLRs or may escape TLR detection. Thus, an alternative early pathogen detecting system such as the one we think operates to sense reovirus may serve as an important complementary mechanism by which host cells detect intrusion by viruses. Understanding the precise mechanisms that underlie penetration-induced proapoptotic signaling will provide broader insight into events at the pathogen-host interface that govern disease.