Chikungunya virus (CHIKV) is a reemerging alphavirus that has caused epidemics of fever, arthralgia, and rash worldwide. There are currently no licensed vaccines or antiviral therapies available for the prevention or treatment of CHIKV disease. We conducted a high-throughput, chemical compound screen that identified digoxin, a cardiac glycoside that blocks the sodium-potassium ATPase, as a potent inhibitor of CHIKV infection. Treatment of human cells with digoxin or a related cardiac glycoside, ouabain, resulted in a dose-dependent decrease in infection by CHIKV. Inhibition by digoxin was cell type-specific, as digoxin treatment of either murine or mosquito cells did not diminish CHIKV infection. Digoxin displayed antiviral activity against other alphaviruses, including Ross River virus and Sindbis virus, as well as mammalian reovirus and vesicular stomatitis virus. The digoxin-mediated block to CHIKV and reovirus infection occurred at one or more postentry steps, as digoxin inhibition was not bypassed by fusion of CHIKV at the plasma membrane or infection with cell surface-penetrating reovirus entry intermediates. Selection of digoxin-resistant CHIKV variants identified multiple mutations in the nonstructural proteins required for replication complex formation and synthesis of viral RNA. These data suggest a role for the sodium-potassium ATPase in promoting postentry steps of CHIKV replication and provide rationale for modulation of this pathway as a broad-spectrum antiviral strategy.

We performed experiments to determine the effect of PKR activation on respiratory syncytial virus (RSV) replication. We first determined that RSV infection activates PKR which induces the phosphorylation of eIF2α, resulting in the formation of host stress granules. We used RNA interference to decrease endogenous PKR levels. RSV replication was not altered in cells deficient for PKR expression. However, RSV-mediated stress granule formation was significantly reduced in PKR-knockdown cells. As an alternative method to block PKR activation, we used treatment with the kinase inhibitor 2-aminopurine (2-AP). We observed that 2-AP treatment significantly reduced viral replication. We also treated PKR-knockdown cells with 2-AP and inoculated with RSV. Under these conditions, 2-AP treatment diminished viral replication in the absence of PKR expression. These results suggest that PKR activation has a minimal effect on RSV replication and that the antiviral effect of 2-AP during RSV infection likely occurs via a PKR-independent mechanism.

Receptors expressed on the host cell surface adhere viruses to target cells and serve as determinants of viral tropism. Several viruses bind cell surface glycans to facilitate entry, but the contribution of specific glycan moieties to viral disease is incompletely understood. Reovirus provides a tractable experimental model for studies of viral neuropathogenesis. In newborn mice, serotype 1 (T1) reovirus causes hydrocephalus, whereas serotype 3 (T3) reovirus causes encephalitis. T1 and T3 reoviruses engage distinct glycans, suggesting that glycan-binding capacity contributes to these differences in pathogenesis. Using structure-guided mutagenesis, we engineered a mutant T1 reovirus incapable of binding the T1 reovirus-specific glycan receptor, GM2. The mutant virus induced substantially less hydrocephalus than wild-type virus, an effect phenocopied by wild-type virus infection of GM2-deficient mice. In comparison to wild-type virus, yields of mutant virus were diminished in cultured ependymal cells, the cell type that lines the brain ventricles. These findings suggest that GM2 engagement targets reovirus to ependymal cells in mice and illuminate the function of glycan engagement in reovirus serotype-dependent disease.

Various sensors that detect double-stranded RNA, presumably of viral origin, exist in eukaryotic cells and induce IFN-responses. Ongoing IFN-responses have also been documented in a variety of human autoimmune diseases including relapsing-remitting multiple sclerosis (RRMS) but their origins remain obscure. We find increased IFN-responses in leukocytes in relapsing-remitting multiple sclerosis at distinct stages of disease. Moreover, endogenous RNAs isolated from blood cells of these same patients recapitulate this IFN-response if transfected into naïve cells. These endogenous RNAs are double-stranded RNAs, contain Alu and Line elements and are transcribed from leukocyte transcriptional enhancers. Thus, transcribed endogenous retrotransposon elements can co-opt pattern recognition sensors to induce IFN-responses in RRMS.

Transcripts encoding ADAR1, a double-stranded, RNA-specific adenosine deaminase involved in the adenosine-to-inosine (A-to-I) editing of mammalian RNAs, can be alternatively spliced to produce an interferon-inducible protein isoform (p150) that is up-regulated in both cell culture and in vivo model systems in response to pathogen or interferon stimulation. In contrast to other tissues, p150 is expressed at extremely low levels in the brain and it is unclear what role, if any, this isoform may play in the innate immune response of the central nervous system (CNS) or whether the extent of editing for RNA substrates critical for CNS function is affected by its induction. To investigate the expression of ADAR1 isoforms in response to viral infection and subsequent alterations in A-to-I editing profiles for endogenous ADAR targets, we used a neurotropic strain of reovirus to infect neonatal mice and quantify A-to-I editing in discrete brain regions using a multiplexed, high-throughput sequencing strategy. While intracranial injection of reovirus resulted in a widespread increase in the expression of ADAR1 (p150) in multiple brain regions and peripheral organs, significant changes in site-specific A-to-I conversion were quite limited, suggesting that steady-state levels of p150 expression are not a primary determinant for modulating the extent of editing for numerous ADAR targets in vivo.

Bloodstream spread is a critical step in the pathogenesis of many viruses. However, mechanisms that promote viremia are not well understood. Reoviruses are neurotropic viruses that disseminate hematogenously to the central nervous system. Junctional adhesion molecule A (JAM-A) is a tight junction protein that serves as a receptor for reovirus. JAM-A is required for establishment of viremia in infected newborn mice and viral spread to sites of secondary replication. To determine how viruses gain access to the circulatory system, we examined reovirus infection of polarized human brain microvascular endothelial cells (HBMECs). Reovirus productively infects polarized HBMECs, but infection does not alter tight junction integrity. Apical infection of polarized HBMECs is more efficient than basolateral infection, which is attributable to viral engagement of sialic acid and JAM-A. Viral release occurs exclusively from the apical surface via a mechanism that is not associated with lysis or apoptosis of infected cells. These data suggest that infection of endothelial cells routes reovirus apically into the bloodstream for systemic dissemination in the host. Understanding how viruses invade the bloodstream may aid in the development of therapeutics that block this step in viral pathogenesis.

Many viruses attach to target cells by binding to cell-surface glycans. To gain a better understanding of strategies used by viruses to engage carbohydrate receptors, we determined the crystal structures of reovirus attachment protein σ1 in complex with α-2,3-sialyllactose, α-2,6-sialyllactose, and α-2,8-di-siallylactose. All three oligosaccharides terminate in sialic acid, which serves as a receptor for the reovirus serotype studied here. The overall structure of σ1 resembles an elongated, filamentous trimer. It contains a globular head featuring a compact β-barrel, and a fibrous extension formed by seven repeating units of a triple β-spiral that is interrupted near its midpoint by a short α-helical coiled coil. The carbohydrate-binding site is located between β-spiral repeats two and three, distal from the head. In all three complexes, the terminal sialic acid forms almost all of the contacts with σ1 in an identical manner, while the remaining components of the oligosaccharides make little or no contacts. We used this structural information to guide mutagenesis studies to identify residues in σ1 that functionally engage sialic acid by assessing hemagglutination capacity and growth in murine erythroleukemia cells, which require sialic acid binding for productive infection. Our studies using σ1 mutant viruses reveal that residues 198, 202, 203, 204, and 205 are required for functional binding to sialic acid by reovirus. These findings provide insight into mechanisms of reovirus attachment to cell-surface glycans and contribute to an understanding of carbohydrate binding by viruses. They also establish a filamentous, trimeric carbohydrate-binding module that could potentially be used to endow other trimeric proteins with carbohydrate-binding properties.

BackgroundInfants and young children are particularly susceptible to viral encephalitis; however, the mechanisms are unknown. We determined the age-dependent contribution of innate and adaptive immune functions to reovirus-induced encephalitis in mice.MethodsNewborn wild-type mice, 2-20 days of age, were inoculated with reovirus or diluent and monitored for mortality, weight gain, and viral load. Four- and fifteen-day-old IFNAR-/- and RAG2-/- mice were inoculated with reovirus and similarly monitored.ResultsWeight gain was impaired in mice inoculated with reovirus at 8 days of age or less. Clinical signs of encephalitis were detected in mice inoculated at 10 days of age or less. Mortality decreased when mice were inoculated after 6 days of age. Survival was ≤15% in wild type (WT), RAG2-/-, and IFNAR-/- mice inoculated at 4 days of age. All WT mice, 92% of RAG2-/- mice, and only 48% of IFNAR-/- mice survived following inoculation at 15 days of age.ConclusionsSusceptibility of mice to reovirus-induced disease decreases between 6 and 8 days of age. Enhanced reovirus virulence in IFNAR-/- mice relative to WT and RAG2-/- mice inoculated at 15 days of age suggests that maturation of the type-I interferon response contributes to age-related mortality following reovirus infection.

The function of the mammalian orthoreovirus (reovirus) σNS nonstructural protein is enigmatic. σNS is an RNA-binding protein that forms oligomers and enhances the stability of bound RNAs, but the mechanisms by which it contributes to reovirus replication are unknown. To determine the function of σNS-RNA binding in reovirus replication, we engineered σNS mutants deficient in RNA-binding capacity. We found that alanine substitutions of positively charged residues in a predicted RNA-binding domain decrease RNA-dependent oligomerization. To define steps in reovirus replication facilitated by the RNA-binding property of σNS, we established a complementation system in which wild-type or mutant forms of σNS could be tested for the capacity to overcome inhibition of σNS expression. Mutations in σNS that disrupt RNA binding also diminish viral replication and σNS distribution to viral factories. Moreover, viral mRNAs only incorporate into viral factories or factory-like structures (formed following expression of nonstructural protein μNS) when σNS is present and capable of binding RNA. Collectively, these findings indicate that σNS requires positively charged residues in a putative RNA-binding domain to recruit viral mRNAs to sites of viral replication and establish a function for σNS in reovirus replication. IMPORTANCE Viral replication requires the formation of neoorganelles in infected cells to concentrate essential viral and host components. However, for many viruses, it is unclear how these components coalesce into neoorganelles to form factories for viral replication. We discovered that two mammalian reovirus nonstructural proteins act in concert to form functioning viral factories. Reovirus μNS proteins assemble into exclusive factory scaffolds that require reovirus σNS proteins for efficient viral mRNA incorporation. Our results demonstrate a role for σNS in RNA recruitment to reovirus factories and, more broadly, show how a cytoplasmic non-membrane-enclosed factory is formed by an RNA virus. Understanding the mechanisms of viral factory formation will help identify new targets for antiviral therapeutics that disrupt assembly of these structures and inform the use of nonpathogenic viruses for biotechnological applications.

The mammalian orthoreovirus double-stranded (ds) RNA-binding protein σ3 is a multifunctional protein that promotes viral protein synthesis and facilitates viral entry and assembly. The dsRNA-binding capacity of σ3 correlates with its capacity to prevent dsRNA-mediated activation of protein kinase R (PKR). However, the effect of σ3 binding to dsRNA during viral infection is largely unknown. To identify functions of σ3 dsRNA-binding activity during reovirus infection, we engineered a panel of thirteen σ3 mutants and screened them for the capacity to bind dsRNA. Six mutants were defective in dsRNA binding, and mutations in these constructs cluster in a putative dsRNA-binding region on the surface of σ3. Two recombinant viruses expressing these σ3 dsRNA-binding mutants, K287T and R296T, display strikingly different phenotypes. In a cell-type dependent manner, K287T, but not R296T, replicates less efficiently than wild-type (WT) virus. In cells in which K287T virus demonstrates a replication deficit, PKR activation occurs and abundant stress granules (SGs) are formed at late times post-infection. In contrast, the R296T virus retains the capacity to suppress activation of PKR and does not mediate formation of SGs at late times post-infection. These findings indicate that σ3 inhibits PKR independently of its capacity to bind dsRNA. In infected mice, K287T produces lower viral titers in the spleen, liver, lungs, and heart relative to WT or R296T. Moreover, mice inoculated with WT or R296T viruses develop myocarditis, whereas those inoculated with K287T do not. Overall, our results indicate that σ3 functions to suppress PKR activation and subsequent SG formation during viral infection and that these functions correlate with virulence in mice.

Several barriers protect the central nervous system (CNS) from pathogen invasion. Yet viral infections of the CNS are common and often debilitating. Understanding how neurotropic viruses co-opt host machinery to overcome challenges to neuronal entry and transmission is important to combat these infections. Neurotropic reovirus disseminates through neural routes and invades the CNS to cause lethal encephalitis in newborn animals. To define mechanisms of reovirus neuronal entry and directional transport, we used primary neuron cultures, which reproduce in vivo infection patterns displayed by different reovirus serotypes. Treatment of neurons with small-molecule inhibitors of different endocytic uptake pathways allowed us to discover that the cellular machinery mediating macropinocytosis is required for reovirus neuronal entry. This mechanism of reovirus entry differs from clathrin-mediated endocytosis, which is used by reovirus to invade non-neuronal cells. Analysis of reovirus transport and release from isolated soma or axonal termini of neurons cultivated in microfluidic devices indicates that reovirus is capable of retrograde but only limited anterograde neuronal transmission. The dynamics of retrograde reovirus movement are consistent with fast axonal transport coordinated by dynein along microtubules. Further analysis of viral transport revealed that multiple virions are transported together in axons within non-acidified vesicles. Reovirus-containing vesicles acidify after reaching the soma, where disassembly of virions and release of the viral core into the cytoplasm initiates replication. These results define mechanisms of reovirus neuronal entry and transport and establish a foundation to identify common host factors used by neuroinvasive viruses. Furthermore, our findings emphasize consideration of cell type-specific entry mechanisms in the tailored design of neurotropic viruses as tracers, oncolytic agents, and delivery vectors.

Mammalian orthoreoviruses (reoviruses) are nonenveloped viruses that replicate in cytoplasmic membranous organelles called viral inclusions (VIs) where progeny virions are assembled. To better understand cellular routes of nonlytic reovirus exit, we imaged sites of virus egress in infected, nonpolarized human brain microvascular endothelial cells (HBMECs) and observed one or two distinct egress zones per cell at the basal surface. Transmission electron microscopy and 3D electron tomography (ET) of the egress zones revealed clusters of virions within membrane-bound structures, which we term membranous carriers (MCs), approaching and fusing with the plasma membrane. These virion-containing MCs emerged from larger, LAMP-1-positive membranous organelles that are morphologically compatible with lysosomes. We call these structures sorting organelles (SOs). Reovirus infection induces an increase in the number and size of lysosomes and modifies the pH of these organelles from ∼4.5-5 to ∼6.1 after recruitment to VIs and before incorporation of virions. ET of VI-SO-MC interfaces demonstrated that these compartments are connected by membrane-fusion points, through which mature virions are transported. Collectively, our results show that reovirus uses a previously undescribed, membrane-engaged, nonlytic egress mechanism and highlights a potential new target for therapeutic intervention.

The gp41 membrane-proximal external region (MPER) is a target for broadly neutralizing antibody responses against human immunodeficiency virus type 1 (HIV-1). However, replication-defective virus vaccines currently under evaluation in clinical trials do not efficiently elicit MPER-specific antibodies. Structural modeling suggests that the MPER forms an α-helical coiled coil that is required for function and immunogenicity. To maintain the native MPER conformation, we used reverse genetics to engineer replication-competent reovirus vectors that displayed MPER sequences in the α-helical coiled-coil tail domain of viral attachment protein σ1. Sequences in reovirus strain type 1 Lang (T1L) σ1 were exchanged with sequences encoding HIV-1 strain Ba-L MPER epitope 2F5 or the entire MPER. Individual 2F5 or MPER substitutions were introduced at virion-proximal or virion-distal sites in the σ1 tail. Recombinant reoviruses containing heterologous HIV-1 sequences were viable and produced progeny yields comparable to those with wild-type virus. HIV-1 sequences were retained following 10 serial passages in cell culture, indicating that the substitutions were genetically stable. Recombinant viruses engineered to display the 2F5 epitope or full-length MPER in σ1 were recognized by purified 2F5 antibody. Inoculation of mice with 2F5-containing vectors or rabbits with 2F5- or MPER-containing vectors elicited anti-reovirus antibodies, but HIV-1-specific antibodies were not detected. Together, these findings indicate that heterologous sequences that form α-helices can functionally replace native sequences in the α-helical tail domain of reovirus attachment protein σ1. However, although these vectors retain native antigenicity, they were not immunogenic, illustrating the difficulty of experimentally inducing immune responses to this essential region of HIV-1. IMPORTANCE Vaccines to protect against HIV-1, the causative agent of AIDS, are not approved for use. Antibodies that neutralize genetically diverse strains of HIV-1 bind to discrete regions of the envelope glycoproteins, including the gp41 MPER. We engineered recombinant reoviruses that displayed MPER epitopes in attachment protein σ1 (REO-MPER vectors). The REO-MPER vectors replicated with wild-type efficiency, were genetically stable, and retained native antigenicity. However, we did not detect HIV-1-specific immune responses following inoculation of the REO-MPER vectors into small animals. This work provides proof of principle for engineering reovirus to express antigenic epitopes and illustrates the difficulty in eliciting MPER-specific immune responses.

Human metapneumovirus (HMPV), a member of the Paramyxoviridae family, is a leading cause of lower respiratory illness. Although receptor binding is thought to initiate fusion at the plasma membrane for paramyxoviruses, the entry mechanism for HMPV is largely uncharacterized. Here we sought to determine whether HMPV initiates fusion at the plasma membrane or following internalization. To study the HMPV entry process in human bronchial epithelial (BEAS-2B) cells, we used fluorescence microscopy, an R18-dequenching fusion assay, and developed a quantitative, fluorescence microscopy assay to follow virus binding, internalization, membrane fusion, and visualize the cellular site of HMPV fusion. We found that HMPV particles are internalized into human bronchial epithelial cells before fusing with endosomes. Using chemical inhibitors and RNA interference, we determined that HMPV particles are internalized via clathrin-mediated endocytosis in a dynamin-dependent manner. HMPV fusion and productive infection are promoted by RGD-binding integrin engagement, internalization, actin polymerization, and dynamin. Further, HMPV fusion is pH-independent, although infection with rare strains is modestly inhibited by RNA interference or chemical inhibition of endosomal acidification. Thus, HMPV can enter via endocytosis, but the viral fusion machinery is not triggered by low pH. Together, our results indicate that HMPV is capable of entering host cells by multiple pathways, including membrane fusion from endosomal compartments.

We explored the role of Peyer's patch (PP) dendritic cell (DC) populations in the induction of immune responses to reovirus strain type 1 Lang (T1L). Immunofluorescence staining revealed the presence of T1L structural (sigma1) and nonstructural (sigmaNS) proteins in PPs of T1L-infected mice. Cells in the follicle-associated epithelium contained both sigma1 and sigmaNS, indicating productive viral replication. In contrast, sigma1, but not sigmaNS, was detected in the subepithelial dome (SED) in association with CD11c(+)/CD8alpha(-)/CD11b(lo) DCs, suggesting antigen uptake by these DCs in the absence of infection. Consistent with this possibility, PP DCs purified from infected mice contained sigma1, but not sigmaNS, and PP DCs from uninfected mice could not be productively infected in vitro. Furthermore, sigma1 protein in the SED was associated with fragmented DNA by terminal deoxy-UTP nick-end labeling staining, activated caspase-3, and the epithelial cell protein cytokeratin, suggesting that DCs capture T1L antigen from infected apoptotic epithelial cells. Finally, PP DCs from infected mice activated T1L-primed CD4(+) T cells in vitro. These studies show that CD8alpha(-)/CD11b(lo) DCs in the PP SED process T1L antigen from infected apoptotic epithelial cells for presentation to CD4(+) T cells, and therefore demonstrate the cross-presentation of virally infected cells by DCs in vivo during a natural viral infection.

Reovirus infection leads to apoptosis in both cultured cells and the murine central nervous system (CNS). NF-kappaB-driven transcription of proapoptotic cellular genes is required for the effector phase of the apoptotic response. Although both extrinsic death-receptor signaling pathways and intrinsic pathways involving mitochondrial injury are implicated in reovirus-induced apoptosis, mechanisms by which either of these pathways are activated and their relationship to NF-kappaB signaling following reovirus infection are unknown. The proapoptotic Bcl-2 family member, Bid, is activated by proteolytic cleavage following reovirus infection. To understand how reovirus integrates host signaling circuits to induce apoptosis, we examined proapoptotic signaling following infection of Bid-deficient cells. Although reovirus growth was not affected by the absence of Bid, cells lacking Bid failed to undergo apoptosis. Furthermore, we found that NF-kappaB activation is required for Bid cleavage and subsequent proapoptotic signaling. To examine the functional significance of Bid-dependent apoptosis in reovirus disease, we monitored fatal encephalitis caused by reovirus in the presence and absence of Bid. Survival of Bid-deficient mice was significantly enhanced in comparison to wild-type mice following either peroral or intracranial inoculation of reovirus. Decreased reovirus virulence in Bid-null mice was accompanied by a reduction in viral yield. These findings define a role for NF-kappaB-dependent cleavage of Bid in the cell death program initiated by viral infection and link Bid to viral virulence.

Viruses require specific cellular receptors to infect their target cells. Angiotensin-converting enzyme 2 (ACE2) is a cellular receptor for two divergent coronaviruses, SARS coronavirus (SARS-CoV) and human coronavirus NL63 (HCoV-NL63). In addition to hostcell receptors, lysosomal cysteine proteases are required for productive infection by some viruses. Here we show that SARS-CoV, but not HCoV-NL63, utilizes the enzymatic activity of the cysteine protease cathepsin L to infect ACE2-expressing cells. Inhibitors of cathepsin L blocked infection by SARS-CoV and by a retrovirus pseudotyped with the SARS-CoV spike (S) protein but not infection by HCoV-NL63 or a retrovirus pseudotyped with the HCoV-NL63 S protein. Expression of exogenous cathepsin L substantially enhanced infection mediated by the SARS-CoV S protein and by filovirus GP proteins but not by the HCoV-NL63 S protein or the vesicular stomatitis virus G protein. Finally, an inhibitor of endosomal acidification had substantially less effect on infection mediated by the HCoV-NL63 S protein than on that mediated by the SARS-CoV S protein. Our data indicate that two coronaviruses that utilize a common receptor nonetheless enter cells through distinct mechanisms.

Gag proteins direct the process of retroviral particle assembly and form the major protein constituents of the viral core. The matrix region of the HIV-1 Gag polyprotein plays a critical role in the transport of Gag to the plasma membrane assembly site. Recent evidence indicates that Gag trafficking to late endosomal compartments, including multivesicular bodies, occurs prior to viral particle budding from the plasma membrane. Here we demonstrate that the matrix region of HIV-1 Gag interacts directly with the delta subunit of the AP-3 complex, and that this interaction plays an important functional role in particle assembly. Disruption of this interaction eliminated Gag trafficking to multivesicular bodies and diminished HIV particle formation. These studies illuminate an early step in retroviral particle assembly and provide evidence that the trafficking of Gag to late endosomes is part of a productive particle assembly pathway.

Mice infected with reovirus develop abnormalities in glucose homeostasis. Reovirus strain type 3 Abney (T3A) was capable of systemic infection of nonobese diabetic (NOD) mice, an experimental model of autoimmune diabetes. Reovirus antigen was detected in pancreatic islets of T3A-infected mice, and primary cultures of pancreatic islets from NOD mice supported T3A growth. Significantly fewer T3A-infected animals compared to uninfected controls developed diabetes. However, despite the alteration in diabetes penetrance, insulitis was evident in T3A-infected mice. These results suggest that viral infection of NOD mice alters autoimmune responses to beta-cell antigens and thereby delays development of diabetes.

Chikungunya virus (CHIKV) and related alphaviruses cause epidemics of acute and chronic musculoskeletal disease. To investigate the mechanisms underlying the failure of immune clearance of CHIKV, we studied mice infected with an attenuated CHIKV strain (181/25) and the pathogenic parental strain (AF15561), which differ by five amino acids. Whereas AF15561 infection of wild-type mice results in viral persistence in joint tissues, 181/25 is cleared. In contrast, 181/25 infection of μMT mice lacking mature B cells results in viral persistence in joint tissues, suggesting that virus-specific antibody is required for clearance of infection. Mapping studies demonstrated that a highly conserved glycine at position 82 in the A domain of the E2 glycoprotein impedes clearance and neutralization of multiple CHIKV strains. Remarkably, murine and human antibodies targeting E2 domain B failed to neutralize pathogenic CHIKV strains efficiently. Our data suggest that pathogenic CHIKV strains evade E2 domain-B-neutralizing antibodies to establish persistence.