Search

Three structural forms of type 1 Lang reovirus (virions, intermediate subviral particles [ISVPs], and cores) have been examined by cryoelectron microscopy (cryoEM) and image reconstruction at 27 to 32-A resolution. Analysis of the three-dimensional maps and known biochemical composition allows determination of capsid protein location, globular shape, stoichiometry, quaternary organization, and interactions with adjacent capsid proteins. Comparisons of the virion, ISVP and core structures and examination of difference maps reveal dramatic changes in supra-molecular structure and protein conformation that are related to the early steps of reovirus infection. The intact virion (approximately 850-A diam) is designed for environmental stability in which the dsRNA genome is protected not only by tight sigma 3-mu 1, lambda 2-sigma 3, and lambda 2-mu 1 interactions in the outer capsid but also by a densely packed core shell formed primarily by lambda 1 and sigma 2. The segmented genome appears to be packed in a liquid crystalline fashion at radii < 240 A. Depending on viral growth conditions, virions undergo cleavage by enteric or endosomal/lysosomal proteases, to generate the activated ISVP (approximately 800-A diam). This transition involves the release of an outer capsid layer spanning radii from 360 to 427 A that is formed by 60 tetrameric and 60 hexameric clusters of ellipsoidal subunits of sigma 3. The vertex-associated cell attachment protein, sigma 1, also undergoes a striking change from a poorly visualized, more compact form, to an extended, flexible fiber. This conformational change may maximize interactions of sigma 1 with cell surface receptors. Transcription of viral mRNAs is mediated by the core particle (approximately 600-A diam), generated from the ISVP after penetration and uncoating. The transition from ISVP to core involves release of the 12 sigma 1 fibers and the remaining outer capsid layer formed by 200 trimers of rod-shaped mu 1 subunits that span radii from 306 to 395 A. In the virion and ISVP, flower-shaped pentamers of the lambda 2 protein are centered at the vertices. In the ISVP-to-core transition, domains of the lambda 2 subunits rotate and swing upward and outward to form a turret-like structure extending from radii 305 to 400 A, with a diameter of 184 A, and a central channel 84 A wide. This novel conformational change allows the potential diffusion of substrates for transcription and exit of newly synthesized mRNA segments.(ABSTRACT TRUNCATED AT 400 WORDS)

Oral inoculation of reovirus type 3 Abney (T3A) into neonatal mice induces hepatitis and the biliary atresia-associated oily fur syndrome (OFS), a phenotype previously linked to the S1 gene. We found that following oral inoculation, none of three T3A mutants, JH2, JH3, and JH4, containing different single amino acid substitutions in the M2 gene, induced the OFS or extensive liver necrosis. Similarly, reassortant viruses containing both a JH4-S1 and a JH4-M2 gene segment did not induce the OFS, whereas another reassortant containing a JH4-S1 gene and a M2 gene from reovirus type 3 Dearing fully recovered this capacity. Together, these results constitute the first evidence for the involvement of the M2 gene in the S1 gene-associated capacity of T3A to induce hepatobiliary disease in neonatal mice.

To better understand the mechanisms by which neurotropic viruses invade peripheral nerve pathways and produce CNS disease, we defined the type 3 (T3) reovirus genes that are determinants of the capacity of reovirus T3 strain Dearing (T3D) and T3 clone 9 (C9) to infect the spinal cord and kill mice after hindlimb injection. T3D and C9 viruses are both highly virulent (LD(50) < 10(1) PFU) after intracranial injection of neonatal mice. However, C9 is significantly more lethal than T3D after either intramuscular injection (LD(50) < 10(1) vs LD(50) 10(4) PFU) or peroral injection (LD(50) 10(3.4) vs LD(50) > 10(8.3) PFU). Using reassortant viruses containing different combinations of genes derived from T3D and C9, we found that the S1 gene, encoding the cell attachment protein sigma 1 and the nonstructural protein sigma 1s, and the L3 gene, encoding the core shell protein lambda 1 were the primary determinants of lethality after intramuscular injection. The L3 gene and the L2 gene encoding spike protein, lambda 2, determined differences in spinal cord titer after intramuscular injection. A C9 x T3D mono-reassortant containing all T3D genes except for the C9-derived L3 was lethal after peroral injection. These studies indicate that the S1, L2, and L3 genes all play a potential role in neuroinvasiveness and provide the first identification of a role in pathogenesis for the L3 gene.

Neonatal but not adult mice are vulnerable to reovirus invasion of the central nervous system after peripheral inoculation. After hindlimb injection, type 3 reovirus travels via the sciatic nerve to replicate in spinal cord motor neurons before spread to the brain and development of lethal encephalitis. Here we provide ultrastructural evidence for direct reovirus invasion of unmyelinated neonatal motor nerve terminals within 2 h and replication in spinal cord motor neurons within 14 h after hindlimb injection of 1-day-old mice. In adult mice, resistance to reovirus lethality after intracranial (IC) injection correlates with the restriction of virus growth in cortical neurons. We found that neuroinvasion also is age dependent after intramuscular injection. Virus lethality and CNS infection decreased sharply during the first postnatal week, while lethality after IC injection continued for 2 additional weeks. Mice inoculated at 7 days of age with high virus doses suffered paralysis of the injected limb, but significant brain infection was not lethal. These results suggest that reovirus invasion of the neonatal CNS is restricted by several progressive age-dependent mechanisms.

Reovirus type 3 clone 9 (T3C9) induces lethal encephalitis in neonatal, but not adult mice. Whether host factors that promote the development and/or functioning of nervous and gastrointestinal tissues could modulate the pathogenesis of this enteric virus was examined. The results showed that antibody specific for interleukin-3 or nerve growth factor antiserum, but not anti-interleukin-6 or anti-tumor necrosis factor-alpha/beta increased mice survival to T3C9 and decreased viral titers in nervous tissues early after infection. These data suggest that IL-3 and NGF are involved in the pathogenesis of T3C9 infection in neonatal mice.

Reovirus Type 3 clone 9 (T3C9)-induced lethal encephalitis is age dependent. We examined the effects of T3C9 inoculated into neonatal and adult mice by intracerebral, intramuscular, or peroral routes and the effect of lipopolysaccharide (LPS) on IL-1alpha levels in the blood and the brain. In parallel, we measured mice survival to T3C9 challenge, primary replication, and growth in and spread to the brain. The results show that T3C9 infection increased IL-1alpha only in the brain of neonatal mice, whereas LPS enhanced IL-1alpha in the brain and in the blood in both neonatal and adult mice. In neonatal mice, a T3C9-induced IL-1alpha increase coincided with viral replication-induced nervous tissue injury and preceded death. Anti-IL-1alpha antibody partially protected neonatal mice against T3C9 peroral challenge, further suggesting that this cytokine is involved in the mechanisms leading to lethal encephalitis. In adult mice, T3C9 was not lethal and did not modify IL-1alpha levels although it slowly replicated in nervous tissues when inoculated directly into the brain. Together, these results suggest that differences in nervous tissue response to T3C9 replication between newborn and adult mice could account in part for the age-dependent susceptibility to T3C9-induced lethal encephalitis., (Copyright 1999 Academic Press.)

The structure of mammalian orthoreovirus top component particles, which are profoundly deficient in the content of double-stranded RNA genome, was determined at 30 A resolution by transmission cryoelectron microscopy and three-dimensional image reconstruction. Previously undetected, ordered densities, appearing primarily as pentameric flowers in the reconstruction, were seen to extend 65 A inwardly from the inner capsid at the icosahedral fivefold axes. Identically positioned but lower density elements were observed in two types of partially uncoated top component particles obtained by limited proteolysis. The levels of three inner-capsid proteins-lamda 1, lamda 3, and mu 2-were reduced in concert with the internal densities during proteolytic uncoating. Since lamda 3 contains the catalytic regions of the viral RNA polymerase and since both lamda 1 and mu 2 appear to play roles in transcription or mRNA capping, the internal structures are concluded to be complexes of the viral transcriptase-related enzymes. The findings have implications for the mechanisms of transcription and mRNA capping by orthoreovirus particles.

The 144-kDa lambda2 protein is a structural component of mammalian reovirus particles and contains the guanylyltransferase activity involved in adding 5' caps to reovirus mRNAs. After incubation of reovirus T3D core particles at 52 degrees C, the lambda2 protein became sensitive to partial protease degradation. Sequential treatments with heat and chymotrypsin caused degradation of a C-terminal portion of lambda2, leaving a 120K core-associated fragment. The four other proteins in cores--lambda1, lambda3, mu2, and sigma2--were not affected by the treatment. Purified cores with cleaved lambda2 were subjected to transmission cryoelectron microscopy and image reconstruction. Reconstruction analysis demonstrated that a distinctive outer region of lambda2 was missing from the modified cores. The degraded region of lambda2 corresponded to the one that contacts the base of the sigma1 protein fiber in reovirus virions and infectious subvirion particles, suggesting that the sigma1-binding region of lambda2 is near its C terminus. Cores with cleaved lambda2 were shown to retain all activities required to transcribe and cap reovirus mRNAs, indicating that the C-terminal region of lambda2 is dispensable for those functions.

Reovirus type 3 Dearing (T3D) causes a prominent neutrophil influx, substantially greater than seen with reovirus type 1 Lang (T1L) in a rat model of viral pneumonia. We sought to measure reovirus-mediated increases in chemokine mRNA expression in pulmonary cells. We found that the neutrophilia induced by T1L and T3D infection in vivo correlated directly with increased levels of chemokine mRNA expression in T3D-infected compared with those of T1IL-infected lungs. In vitro, reovirus-infected normal alveolar macrophages (AMs) and the rat AM cell line NR8383 expressed greater levels of macrophage inflammatory protein 2, KC, and tumor necrosis factor alpha mRNA. A synergism between reovirus and lipopolysaccharide was also detected for macrophage inflammatory protein 2 and KC mRNA expression. Tumor necrosis factor protein secretion was also increased to a greater extent by T3D than by T1L in primary rat AMs and the NR8383 cells. We conclude that the virus-mediated inflammatory cytokine induction suggests a role for these cytokines in the neutrophil influx observed in the rat reovirus pneumonia model.

Interference of wild-type reovirus growth by some temperature-sensitive (ts) mutant viruses under non-permissive conditions or by other wild-type isolates has been demonstrated; however, the stage of the virus replication cycle at which interference occurs has not been defined. Examination of the time-course of the yields of T1 Lang (T1L) dsRNA in the progeny of mixed infections of T1L with T3 Dearing (T3D) or with a panel of T3D ts mutants at a non-permissive temperature revealed that interference takes place by 8-10 h post-infection and occurs prior to or at the same time as accumulation of reovirus dsRNA. Taken together with our previous results, these data indicate that interference occurs during a window between virus uncoating and synthesis of dsRNA in the reovirus replication cycle, probably at the stage of assembly of primary reovirus particles.

From pairwise mixed infections of different reovirus wild-type isolates (T3 Dearing plus T1 Lang or plus T2 Jones) the progeny virus is phenotypically mixed, i.e., progeny viral particles contain proteins derived from both parents but the corresponding genes derived from only one parent. Experiments with differential inactivation of virus progeny of mixed infections by monoclonal antibodies or by 33% ethanol reveal phenotypic mixing of two outer shell proteins, sigma 1 and probably mu 1. Phenotypic mixing of the mu 1 protein, the product of the M2 gene, is a possible explanation for the recent observation of M2 gene-linked dominant interference between reovirus isolates.

Reovirus intermediate subviral particles (ISVPs) but not intact virions or cores have been shown to possess the capacity to permeabilize mouse L cells as determined by a 51Cr release assay. We used monoclonal antibodies (MAbs) directed against proteins exposed on the ISVP surface (sigma 1, mu 1, and lambda 2) to probe the role(s) of these proteins in membrane interaction and penetration. One sigma 1-specific MAb (MAb-G5) and two mu 1-specific MAbs (MAb-10H2 and MAb-8H6) inhibited reovirus-induced 51Cr release when added pre- or post-ISVP attachment to L cells. MAb-G5 inhibits 51Cr release by interfering with ISVP attachment (via sigma 1) to L-cell receptor sites. The mu 1-specific MAbs (MAb-10H2 and MAb-8H6) inhibit 51Cr release by interfering with an undefined post-L-cell-attachment event that involves bivalent binding of the mu 1-specific MAbs to an epitope located in a central region of the mu 1 protein.

The reovirus M2 gene is associated with the capacity of type 3 strain Abney (T3A) intermediate subviral particles (ISVPs) to permeabilize cell membranes as measured by chromium (51Cr) release (P. Lucia-Jandris, J. W. Hooper, and B. N. Fields, J. Virol. 67:5339-5345, 1993). In addition, reovirus mutants with lesions in the M2 gene can be selected by heating virus at 37 degrees C for 20 min in 33% ethanol (D. R. Wessner and B. N. Fields, J. Virol. 67:2442-2447, 1993). In this report we investigated the mechanism by which the reovirus M2 gene product (the mu 1 protein) influences the capacity of reovirus ISVPs to permeabilize membranes, using ethanol-selected T3A mutants. Each of three T3A ethanol-resistant mutants isolated (JH2, JH3, and JH4) exhibited a decreased capacity to cause 51Cr release relative to that of wild-type T3A. Sequence analysis of the M2 genes of wild-type T3A and the T3A mutants indicated that each mutant possesses a single amino acid substitution in a central region of the 708-amino-acid mu 1 protein: JH2 (residue 466, Tyr to Cys), JH3 (residue 459, Lys to Glu), and JH4 (residue 497 Pro to Ser). Assays performed with reovirus natural isolates, reassortants, and a set of previously characterized type 3 strain Dearing (T3D) ethanol-resistant mutants revealed a strong correlation between ethanol sensitivity and the capacity to cause 51Cr release. We found that ISVPs generated from the T3A and T3D mutants were stable when heated to 50 degrees C, whereas wild-type T3A ISVPs are inactivated under these conditions. Together, these data suggest that amino acid substitutions in a central region of the mu 1 protein affect the capacity of the ISVP to permeabilize L-cell membranes by altering the stability of the virus particle.

We undertook the present study to elucidate the pathogenesis of the pathologic response to reovirus infection in the lungs and further understand the interactions of reoviruses with pulmonary cells. We found that reoviruses were capable of causing acute pneumonia in 25- to 28-day-old Sprague-Dawley rats following intratracheal inoculation with the reoviruses type 1 Lang (T1L) and type 3 Dearing (T3D). The onset of the pneumonia was rapid, marked by type I alveolar epithelial cell degeneration, type II alveolar epithelial cell hyperplasia, and the infiltration of leukocytes into the alveolar spaces. More neutrophils were recruited into the lungs during T3D infection than during T1L infection, and the serotype difference in the neutrophil response was mapped to the S1 gene of reovirus. Viral replication in the lungs was required for the development of pneumonia due to T1L and T3D infections, and replication occurred in type I alveolar epithelial cells. T1L grew to higher titers in the lungs than did either T3D or type 3 clone 9, and the S1 gene was found to play a role in determining the level of viral replication. We propose that experimental reovirus infection in the lungs can serve as a model for the pathogenesis of viral pneumonia in which pulmonary inflammation results following direct infection of lung epithelial cells.

Reovirus adheres specifically to apical membranes of mouse intestinal M cells and exploits M-cell transepithelial transport activity to enter Peyer's patch mucosa, where replication occurs. Proteolytic conversion of native reovirus to intermediate subviral particles (ISVPs) occurs in the intestine, but it is not known whether conversion is essential for interaction of virus with M cells. We tested the capacity of native virions, ISVPs, and cores (that lack outer capsid proteins) to bind to intestinal epithelial cells in vivo and found that only ISVPs adhered to M cells. Thus, intraluminal conversion of native reovirus to ISVPs is a prerequisite for M-cell adherence, and outer capsid proteins unique to ISVPs (either sigma 1 or products of mu 1) mediate interaction of virus with M-cell apical membranes.

Many microorganisms gain access to the systemic circulation after entering the respiratory tract. The precise pathways used to cross the mucosal barriers of the lungs have not been clearly described. We have used the mammalian reoviruses in order to determine the pathway that a systemic virus uses to penetrate the mucosal barrier and enter the systemic circulation after entering the airways of the lungs. Reoviruses enter through pulmonary M cells, which overlie bronchus-associated lymphoid tissue, and subsequently spread to regional lymph nodes. Thus, the pathway through M cells represents a strategy by which viruses and probably other microorganisms can penetrate the mucosal surface of the respiratory tract and thereby enter the systemic circulation.

Following infection by pairs of reovirus isolates consisting of combinations of reovirus T1 Lang, T2 Jones, or T3 Dearing, we found that one of the isolates interfered with the yield of progeny RNA derived from the other parents. The most significant interference was produced by T2 Jones or T3 Dearing, when mixed with T1 Lang. Genetic analysis revealed that the presence of the M2 gene in the interfering parent (in the T1 Lang x T3 Dearing pair) was linked to interference. Studies on interference in infected cells indicated that interference occurs after adsorption and penetration.

A panel of serotype 3 (T3) reovirus strains was screened to determine their relative capacities to cause lethal infection and hepatobiliary disease following peroral inoculation in newborn mice. A wide range of 50% lethal doses (LD50s) was apparent after peroral inoculation of the different virus strains. Two of the strains, T3 Abney and T3 clone 31, caused mice to develop the oily fur syndrome associated with biliary atresia. The capacity to cause biliary atresia was not related to the capacity to cause lethal infection, however, because the LD50s of T3 Abney and T3 clone 31 were grossly disparate. Examination of liver and bile duct tissues revealed histopathologic evidence of biliary atresia and hepatic necrosis in T3 Abney-infected mice but not in mice inoculated with a T3 strain of similar virulence or with the hepatotropic T1 Lang strain. The consistency with which T3 Abney-infected mice developed biliary atresia-associated oily fur syndrome permitted us to determine the viral genetic basis of reovirus-induced biliary atresia. Analysis of reassortant viruses isolated from an in vitro coinfection with T3 Abney and T1 Lang indicated a strong association of the hepatobiliary disease-producing phenotype with the T3 Abney S1 gene, which encodes the viral cell attachment protein, sigma 1. Amino acid residues within the sigma 1 protein that were unique to disease-producing T3 strains were identified by comparative sequence analysis. Specific changes exist within two regions of the protein, one of which is thought to be involved in binding to host cell receptors. We hypothesize that changes within this region of the protein are important in determining the tropism of this virus for bile-ductular epithelium.

Success in controlling the AIDS epidemic is as likely to arise from unrelated areas of research as from AIDS-directed programmes.

Since blood-borne viruses often interact with endothelial cells before tissue invasion, the interaction between viruses and endothelial cells is likely to be important in viral pathogenicity. Two reovirus isolates (type 1 Lang and type 3 Dearing) differ in their capacity to grow in cultured bovine aortic endothelial cells. The mammalian reoviruses have 10 double-stranded RNA gene segments in their genome. By using 24 reassortant viruses, observed differences in the capacity of different strains to grow in cultured endothelial cells were mapped to the M1 gene (P = 0.00019), which encodes the viral core protein mu 2. No differences were detected in binding or proteolytic processing of viral outer capsid proteins of parental virions between the two reovirus isolates. Northern blot analysis showed a decreased production of viral mRNA in endothelial cells infected with type 3 Dearing reovirus, but not type 1 Lang. Thus, we have identified a viral gene (the M1 gene) responsible for determining the difference in growth capacity of the two reovirus isolates in cultured endothelial cells. Reovirus is an attractive model in which to study the interaction of viruses with endothelial cells at a molecular genetic level.

Mechanisms by which nonenveloped viruses penetrate cell membranes as an early step in infection are not well understood. Current ideas about the mode for cytosolic penetration by nonenveloped viruses include (i) formation of a membrane-spanning pore through which viral components enter the cell and (ii) local breakdown of the cellular membrane to provide direct access of infecting virus to the cell's interior. Here we report that of the three viral particles of nonenveloped mammalian reoviruses: virions, infectious subvirion particles, and cores (the last two forms generated from intact reovirus virions by proteolysis), only the infectious subvirion particles induced the formation of anion-selective, multisized channels in planar lipid bilayers under the experimental conditions used in this study. The value for the smallest size conductance varied depending on the lipid composition of the bilayer between 90 pS (Asolectin) and 300 pS (phosphatidylethanolamine:phosphatidylserine) and was found to be voltage independent. These findings are consistent with a proposal that the proteolytically activated infectious subviral particles mediate the interaction between virus and the lipid bilayer of a cell membrane during penetration. In addition, the findings indicate that the "penetration proteins" of some enveloped and nonenveloped viruses share similarities in the way they interact with bilayers.

In this study, we investigated the interaction of reovirus particles with cell membranes by using a 51Cr release assay. We confirmed prior observations (J. Borsa, B. D. Morash, M. D. Sargent, T. P. Copps, P. A. Lievaart, and J. G. Szekely, J. Gen. Virol. 45:161-170, 1979) that intermediate subviral particles (ISVPs) of reovirus type 3 strain Abney (T3A) induced the release of 51Cr from preloaded L cells and showed that the intact virion and core forms did not. Reovirus type 1 strain Lang (T1L) ISVPs were found to be less efficient at 51Cr release than T3A ISVPs. Reassortants between these strains indicated that the 51Cr release phenotype segregates with the M2 gene segment. Biochemical studies indicated that the ISVPs' acquisition of the capacity to induce 51Cr release followed the cleavage of the viral M2 gene product mu 1/mu 1C to fragments delta and phi during virion conversion to ISVP but did not directly correlate with this cleavage. These studies suggest that the reovirus M2 gene product (in its cleaved form) plays a role in interacting with cell membranes.

We examined pathogenic characteristics of plaque-purified reoviruses isolated from persistently infected L-cell cultures (PI viruses) after intracranial inoculation into newborn mice. The PI viruses were isolated from independent cultures initiated with high-passage stocks of the wild-type (wt) strain, type 3 Dearing. The virulence of most PI viruses was equivalent to that of the wt strain. However, replication of PI viruses in the central nervous system of infected mice was prolonged to 25 (but not 50) days postinoculation. Thirty-eight percent (n = 186) of mice inoculated with the PI viruses had residual virus detectable in brain tissue 25 days after inoculation, in contrast to only 16% (n = 57) of mice inoculated with wt virus (P = 0.009). Mean residual brain titers were more than 20-fold higher in mice inoculated with PI viruses compared with wt virus (4.3 x 10(4) versus 2.1 x 10(3); P = 0.006). Tropism of PI virus within the brain resembled that of wt virus, and the distribution of PI virus antigen in the brain did not change over time. The extent of necrosis in the brains of mice harboring PI virus 25 days after inoculation was minimal, despite continued presence of high titers of infectious virus. The latter observation resembles the absence of cytopathicity seen in L-cell cultures persistently infected with reovirus. These observations suggest that the interaction of PI viruses with cells can be altered in vivo as well as in cell culture, but virus is eventually cleared from the infected animal.

We used a recently isolated and characterized panel of monoclonal antibodies (MAbs) specific for cross-reactive determinants on reovirus outer capsid proteins to define mechanisms of antibody-mediated protection in vivo. We studied the capacities of MAbs to protect against lethal infection with reoviruses which differ in site of primary replication, route of spread, and central nervous system tropism. We found the following. (i) MAbs specific for each of the viral outer capsid proteins (sigma 1, sigma 3, and mu 1) and the core spike protein (lambda 2) were protective under certain circumstances. (ii) In vitro properties of MAbs, including isotype, neutralization of viral infectivity, inhibition of virus-induced hemagglutination, and avidity of binding, were poorly predictive of the capacities of MAbs to protect in vivo. (iii) MAbs did not act at a single stage during pathogenesis to mediate protection; instead, protective MAbs were capable of altering a variety of stages in reovirus pathogenesis. (iv) MAbs protective against one reovirus also protected against other reoviruses that utilized different pathogenetic strategies, suggesting that the viral epitope bound by an antibody rather than the pathogenetic strategy employed by the virus is a critical determinant of antibody-mediated protection in vivo. (v) A prominent mechanism of protective MAb action is inhibition of viral spread through nerves from a site of primary replication (e.g., the intestine or muscle tissue) to the central nervous system.

To better understand the mechanism(s) by which viruses respond to chemical or physical treatments, we isolated a series of mutant strains of reovirus type 3 Dearing that exhibit increased ethanol resistance. Following exposure to 33% ethanol for 20 min, the parental strain exhibited a 5 log10 decrease in infectivity. The mutant strains, however, exhibited a 2 to 3 log10 decrease in titer following identical treatment. Through the use of reassortant viruses, we mapped this increased ethanol resistance mutation to the M2 gene segment, which encodes a major outer capsid protein, mu1C. Sequence analysis of mutant M2 genes revealed that six of seven unique mutants possessed single-point mutations in this gene. In addition, the change in six of seven mutants caused a predicted amino acid change in a 35-amino-acid region of the gene product between amino acids 425 and 459. The identification of ethanol resistance mutations within a discrete region of this outer capsid protein identifies that portion of the protein as important in reovirus stability. The presence of viral particles possessing altered stability also suggests that subpopulations of viruses may possess altered environmental stability, which, in turn, could affect viral transmission.

Previous studies demonstrated that both cellular and viral mutants are selected during maintenance of persistent infections established in murine L cells with high-passage stocks of mammalian reoviruses. In particular, when one culture was cured of persistent infection, the resulting cells were found to support the growth of viruses isolated from persistently infected cultures (termed PI viruses here) better than that of wild-type (wt) viruses (R. Ahmed, W. M. Canning, R. S. Kauffman, A. H. Sharpe, J. V. Hallum, and B. N. Fields, Cell 25:325-332, 1981). To address the nature of cellular and viral mutations selected during maintenance of persistent reovirus infections, we established independent, persistently infected cultures with L cells and high-passage stocks of wt reovirus. These cultures served as sources of new PI viruses and cured cells for study. We found that although wt viruses grew poorly in cured cells when infection was initiated with intact virions, they grew well in cured cells when infection was initiated with infectious subvirion particles generated from virions by in vitro treatment with chymotrypsin. This finding indicates that the block to growth of wt viruses in cured cells involves an early step that is unique to infection by virions, such as proteolytic processing in an endocytic compartment. We also found that PI viruses grew better than wt viruses in L cells treated with ammonium chloride, a weak base that inhibits the pH decrease in endosomes and lysosomes. Because ammonium chloride blocks an early step in infection by intact virions, probably the proteolytic processing of viral outer capsid proteins by acid-dependent cellular proteases in late endosomes or lysosomes, this finding indicates that PI viruses differ from wt viruses with respect to viral entry into cells. Therefore, these results indicate that both cells and viruses evolve mutations that affect one or more early steps in the viral growth cycle during maintenance of L-cell cultures persistently infected with reoviruses.

Penetration of a cell membrane as an early event in infection of cells by mammalian reoviruses appears to require a particular type of viral particle, the infectious subvirion particle (ISVP), which is generated from an intact virion by proteolytic cleavage of the outer capsid proteins sigma 3 and mu 1/mu 1C. Characterizations of the structural components and properties of ISVPs are thus relevant to attempts to understand the mechanism of penetration by reoviruses. In this study, a novel, approximately 13-kDa carboxy-terminal fragment (given the name phi) was found to be generated from protein mu 1/mu 1C during in vitro treatments of virions with trypsin or chymotrypsin to yield ISVPs. With trypsin treatment, both the carboxy-terminal fragment phi and the amino-terminal fragment mu 1 delta/delta were shown to be generated and to remain attached to ISVPs in stoichiometric quantities. Sites of protease cleavage were identified in the deduced amino acid sequence of mu 1 by determining the amino-terminal sequences of phi proteins: trypsin cleaves between arginine 584 and isoleucine 585, and chymotrypsin cleaves between tyrosine 581 and glycine 582. Findings in this study indicate that sequences in the phi portion of mu 1/mu 1C may participate in the unique functions attributed to ISVPs. Notably, the delta-phi cleavage junction was predicted to be flanked by a pair of long amphipathic alpha-helices. These amphipathic alpha-helices, together with the myristoyl group at the extreme amino terminus of mu 1/mu 1N, are proposed to interact directly with the lipid bilayer of a cell membrane during penetration by mammalian reoviruses.

The mammalian reoviruses serve as a model for understanding both how viruses multiply within cells and, at a higher level of complexity, how the virus is assembled to "deliver" the virus to the cells in the host in which they can multiply and cause disease. They thus serve as a model for the molecular determinants of viral pathogenesis.

Pathogenesis studies in animals can uncover details concerning viral replication, growth, and access to target organs, in vivo. This, in turn, reveals opportunities for antiviral intervention that may be otherwise missed by limiting analysis to growth of virus in tissue culture. In this report, reovirus infection of mice is used as a model. Three general aspects of reovirus behavior in mice are presented and each demonstrates a property of the virus that could easily have been missed by studies in tissue culture.

Thirteen newly isolated monoclonal antibodies (MAbs) were used to study relationships between reovirus outer capsid proteins sigma 3, mu 1c, and lambda 2 (core spike) and the cell attachment protein sigma 1. We focused on sigma 1-associated properties of serotype specificity and hemagglutination (HA). Competition between MAbs revealed two surface epitopes on mu 1c that were highly conserved between reovirus serotype 1 Lang (T1L) and serotype 3 Dearing (T3D). There were several differences between T1L and T3D sigma 3 epitope maps. Studies using T1L x T3D reassortants showed that primary sequence differences between T1L and T3D sigma 3 proteins accounted for differences in sigma 3 epitope maps. Four of 12 non-sigma 1 MAbs showed a serotype-associated pattern of binding to 25 reovirus field isolates. Thus, for reovirus field isolates, different sigma 1 proteins are associated with preferred epitopes on other outer capsid proteins. Further evidence for a close structural and functional interrelationship between sigma 3/mu 1c and sigma 1 included (i) inhibition by sigma 3 and mu 1c MAbs of sigma 1-mediated HA, (ii) enhancement of sigma 1-mediated HA by proteolytic cleavage of sigma 3 and mu 1c, and (iii) genetic studies demonstrating that sigma 1 controlled the capacity of sigma 3 MAbs to inhibit HA. These data suggest that (i) epitopes on sigma 3 and mu 1c lie in close proximity to sigma 1 and that MAbs to these epitopes can modulate sigma 1-mediated functions, (ii) these spatial relationships have functional significance, since removal of sigma 3 and/or cleavage of mu 1c to delta can enhance sigma 1 function, (iii) in nature, the sigma 1 protein places selective constraints on the epitope structure of the other capsid proteins, and (iv) viral susceptibility to antibody action can be determined by genes other than that encoding an antibody's epitope.

The S2 gene nucleotide sequences of prototype strains of the three reovirus serotypes were determined to gain insight into the structure and function of the S2 translation product, virion core protein sigma 2. The S2 sequences of the type 1 Lang, type 2 Jones, and type 3 Dearing strains are 1,331 nucleotides in length and contain a single large open reading frame that could encode a protein of 418 amino acids, corresponding to sigma 2. The deduced sigma 2 amino acid sequences of these strains are very conserved, being identical at 94% of the sequence positions. Predictions of sigma 2 secondary structure and hydrophobicity suggest that the protein has a two-domain structure. A larger domain is suggested to be formed from the amino-terminal three-fourths of sigma 2 sequence, which is separated from a smaller carboxy-terminal domain by a turn-rich hinge region. The carboxy-terminal domain includes sequences that are more hydrophilic than those in the rest of the protein and contains sequences which are predicted to form an alpha-helix. A region of striking similarity was found between amino acids 354 and 374 of sigma 2 and amino acids 1008 and 1031 of the beta subunit of the Escherichia coli DNA-dependent RNA polymerase. We suggest that the regions with similar sequence in sigma 2 and the beta subunit form amphipathic alpha-helices which may play a related role in the function of each protein. We have also performed experiments to further characterize the double-stranded RNA-binding activity of sigma 2 and found that the capacity to bind double-stranded RNA is a property of the sigma 2 protein of prototype strains and of the S2 mutant tsC447.

A crucial event in the pathogenesis of systemic enteric virus infections is entry of virus into the nervous system. Whether enteric virus spreads from the intestinal tract to the central nervous system through nerves or through the bloodstream was examined using a serotype 3 reovirus strain. After peroral inoculation of newborn mice with reovirus, serial histologic sections of small intestine, brain and spinal cord were prepared and stained by immunoperoxidase to detect viral antigen. Three days after inoculation, viral antigen was observed in mononuclear cells of ileal Peyer's patches and in neurons of the adjacent myenteric plexus. Infection first appeared in the central nervous system 1-2 days later in neurons of the dorsal motor nucleus of the vagus nerve. Endothelial cells, meninges, choroid plexus, hypothalamus, and area postrema were not infected, indicating neural rather than bloodborne spread from the intestine. Staining of neurons in the dorsal motor nucleus of the vagus nerve depended on the route of virus inoculation and was independent of the amount of virus in the bloodstream. These results demonstrate that an enteric virus entering a host from the intestinal lumen can spread to the central nervous system through nerve fiber innervating the intestine.

Viral growth in specific tissue is usually required in order to lead to pathology. Two reovirus isolates (type 1 Lang and type 3 Dearing) differ in their capacity to grow in cultured mouse heart cells. The mammalian reoviruses contain a genome of 10 double-stranded RNA gene segments. By the use of 37 reassortant viruses (consisting of viruses with different combinations of genes derived from the two parents), difference in capacity of different strains to grow in heart cells was mapped to three different genes, all of which encode viral core proteins: the M1 gene (P less than 0.000044); the L1 gene (P = 0.00094); and the L3 gene (P = 0.019). Using the same set of reassortant viruses, the L1 (P = 0.00015) and L3 (P = 0.0065) genes were involved in differences of the ability of viral strains to grow in mouse L cells (fibroblasts), but the M1 gene (P = 0.12) was not. These findings suggest that the M1 gene plays an important and specific role in determining the relative capacity of certain viral strains to grow in the heart. Thus, we have identified viral genes responsible for differing growth capacity in heart muscle cells in culture. These findings provide a novel system for studies of viral myocarditis at a molecular genetic level.

The structural protein mu 1 of mammalian reoviruses was noted to have a potential N-myristoylation sequence at the amino terminus of its deduced amino acid sequence. Virions labeled with [3H]myristic acid were used to demonstrate that mu 1 is modified by an amide-linked myristoyl group. A myristoylated peptide having a relative molecular weight (Mr) of approximately 4,000 was also shown to be a structural component of virions and was concluded to represent the 4.2-kDa amino-terminal fragment of mu 1 which is generated by the same proteolytic cleavage that yields the carboxy-terminal fragment and major outer capsid protein mu 1C. The myristoylated 4,000-Mr peptide was found to be present in reovirus intermediate subviral particles but to be absent from cores, indicating that it is a component of the outer capsid. A distinct large myristoylated fragment of the intact mu 1 protein was also identified in intermediate subviral particles, but no myristoylated mu-region proteins were identified in cores, consistent with the location of mu 1 in the outer capsid. Similarities between amino-terminal regions of the reovirus mu 1 protein and the poliovirus capsid polyprotein were noted. By analogy with other viruses that contain N-myristoylated structural proteins (particularly picornaviruses), we suggest that the myristoyl group attached to mu 1 and its amino-terminal fragments has an essential role in the assembly and structure of the reovirus outer capsid and in the process of reovirus entry into cells.

Penetration of the nervous systems of suckling mice by prototype strains of the three mammalian reovirus serotypes was studied after footpad inoculation of a dose (10(7) PFU) representing 3.5 x 10(3) 50% lethal doses (LD50) for reovirus type 3 Dearing and less than 1 LD50 for reoviruses type 1 Lang and type 2 Jones. Type 3 Dearing entered both motor and sensory neurons; infected neurons were clearly detectable by immunohistochemical staining 19 h after inoculation. By day 2, a second cycle of infection had occurred, and by day 4, several hundred motor and sensory neurons and interneurons were infected. By this time, infection also involved large areas of the brain stem and brain. There was evidence of both retrograde and anterograde movement of viral antigen within axons and dendrites. Unexpectedly, reovirus type 1 Lang followed neuronal pathways as well as being disseminated in the bloodstream. Reovirus type 2 Jones also entered neurons. While the number of motor neurons and interneurons infected with type 1 Lang or type 2 Jones remained limited within the first 4 days after inoculation, infection of sensory neurons increased with time and reached a level by day 4 comparable to that observed after infection with type 3 Dearing. Viral antigen was also found in the brain stem and brain, but this infection was limited. These three strains multiplied in nonneuronal tissues. Connective tissue in the footpad was massively infected by all three strains 19 h after inoculation. By this time, foci of infection were also present in muscle and skin. Viral antigen was repeatedly observed in the endothelium of blood vessels and in the meninges after infection with type 1 Lang. The titer of type 1 Lang increased in the blood with time, which was not observed after infection with strains of the other two serotypes. In this study, we found that prototype strains of the three reovirus serotypes exhibited different degrees of neurotropism, all being capable of entering neurons. Transmission of the infection occurred through synapses rather than from cell body to cell body. Thus reovirus, like herpesvirus and rabies virus, is a good marker for the identification of neuronal pathways, although its capacity to grow in neurons, unlike that of herpesvirus and rabies virus, is restricted to newborn animals.

Hemagglutination (HA) by the mammalian reoviruses is mediated by interactions between the viral sigma 1 protein and sialoglycoproteins on the erythrocyte surface. Three serotype 3 (T3) reovirus strains were identified that do not agglutinate either bovine or type O human erythrocytes (HA negative): T3 clone 43 (T3C43), T3 clone 44 (T3C44), and T3 clone 84 (T3C84). These three strains also showed a diminished capacity to bind the major erythrocyte sialoglycoprotein, glycophorin, in an enzyme-linked immunosorbent assay. To determine the molecular basis for these findings, we examined the deduced sigma 1 amino acid sequences of the three HA-negative T3 strains and four HA-positive T3 strains. The limited number of sequence differences in the sigma 1 proteins of these seven strains allowed us to identify single unique amino acid residues in each of the HA-negative strains (aspartate 198 in T3C43, leucine 204 in T3C44, and tryptophan 202 in T3C84) that cluster within a discrete region of the sigma 1 tail. The identification of sigma 1 residues important for HA and glycophorin binding suggests that tail-forming sequences are exposed on the virion surface, where they interact with carbohydrate residues on the surface of cells.

The S1 gene nucleotide sequences of 10 type 3 (T3) reovirus strains were determined and compared with the T3 prototype Dearing strain in order to study sequence diversity in strains of a single reovirus serotype and to learn more about structure-function relationships of the two S1 translation products, sigma 1 and sigma 1s. Analysis of phylogenetic trees constructed from variation in the sigma 1-encoding S1 nucleotide sequences indicated that there is no pattern of S1 gene relatedness in these strains based on host species, geographic site, or date of isolation. This suggests that reovirus strains are transmitted rapidly between host species and that T3 strains with markedly different S1 sequences circulate simultaneously. Comparison of the deduced sigma 1 amino acid sequences of the 11 T3 strains was notable for the identification of conserved and variable regions of sequence that correlate with the proposed domain organization of sigma 1 (M.L. Nibert, T.S. Dermody, and B. N. Fields, J. Virol. 64:2976-2989, 1990). Repeat patterns of apolar residues thought to be important for sigma 1 structure were conserved in all strains examined. The deduced sigma 1s amino acid sequences of the strains were more heterogeneous than the sigma 1 sequences; however, a cluster of basic residues near the amino terminus of sigma 1s was conserved. This analysis has allowed us to investigate molecular epidemiology of T3 reovirus strains and to identify conserved and variable sequence motifs in the S1 translation products, sigma 1 or sigma 1s.

Core particles of reovirus type 3 Dearing (T3D) crystallized in the face-centered cubic space group F432 with dimensions of 1270 A along each edge of the unit cell. Core particles of reovirus type 1 Lang (T1L) did not crystallize. Experiments with core particles derived from 27 different T1L x T3D reassortant viruses indicated that the L2 genome segment determined the capacity of cores to crystallize. This finding indicates important differences in the surface topography of the L2-translation product, the lambda 2 protein, of these two isolates, and suggests that important crystal contacts are mediated by this protein. These data are used to generate a model of the packing of reovirus core particles within the unit cell.

This report describes a model for the structure of the reovirus cell-attachment protein sigma 1. S1 gene nucleotide sequences were determined for prototype strains of the three serotypes of mammalian reoviruses. Deduced amino acid sequences of the S1-encoded sigma 1 proteins were then compared in order to identify conserved features of these sequences. Discrete regions in the amino-terminal two-thirds of sigma 1 sequence share characteristics with the fibrous domains of other cellular and viral proteins. Most of the amino-terminal one-third of sigma 1 sequence is predicted to form an alpha-helical coiled coil like that of myosin. The middle one-third of sigma 1 sequence appears more heterogeneous; it is predicted to form a large region of beta-sheet that is followed by a region which contains two short alpha-helical coiled coils separated by a smaller region of beta-sheet. The two beta-sheet regions are each proposed to form a cross-beta sandwich like that suggested for the rod domain of the adenovirus fiber protein (N. M. Green, N. G. Wrigley, W. C. Russell, S. R. Martin, and A. D. McLachlan, EMBO J. 2:1357-1365, 1983). The remaining carboxy-terminal one-third of sigma 1 sequence is predicted to form a structurally complex globular domain. A model is suggested in which the discrete regions of sigma 1 sequence are ascribed to morphologic regions seen in computer-processed electron micrographic images of the protein (R. D. B. Fraser, D. B. Furlong, B. L. Trus, M. L. Nibert, B. N. Fields, and A. C. Steven, J. Virol. 64:2990-3000, 1990.

The receptor-recognition interaction that initiates reovirus infection is mediated by the sigma 1 protein, located at the vertices of the icosahedral virion. We have applied computer-based image-averaging techniques to electron micrographs of negatively stained preparations of sigma 1 purified from virions (serotype 2 Jones). Combining these results with inferences based on the amino acid sequence has led to a molecular model in which the overall folding of the chains is described; its conformation embodies motifs, coiled-coil alpha-helices and nodular multichain elements rich in beta-sheets, previously detected in the corresponding proteins of other viruses, but with some novel variations. Sigma 1 is a filamentous lollipop-shaped molecule with an overall length of approximately 48 nm; it has a flexible "tail," approximately 40 nm long by 4 to 6 nm wide, terminating at its distal end in a globular "head," approximately 9.5 nm in diameter. The purified protein is a tetramer (4 by 50 kilodaltons) consisting of two similarly oriented dimers bonded side by side and in register. For each chain, a cluster of hydrophobic residues at its amino terminus resides at the proximal end of the tail; next, an alpha-helical domain (residues 25 to 172) participates in a two-chained coiled coil, 22 nm long, with two such coiled coils pairing laterally to form the proximal half of the tail. The remainder of the tail (residues 173 to approximately 316) is less uniform in width and is expected to be rich in beta-sheet; the interdimer bonding is evidently sustained through this portion of the molecule. Finally, the globular head consists of the carboxy-terminal domains (which contain the receptor-binding sites) folded into compact globular conformations; in appropriate side views, the head is resolved into two subunits, presumably contributed by the respective dimers. This model for how the four sigma 1 polypeptide chains are threaded in parallel through the fiber is supported by the observed match between an empirical curvature profile, which identifies the locations of relatively flexible sites along the tail, and the flexibility profile predicted on the basis of the model. Appraisal of the interactions that stabilize the coiled coils suggests that (i) the alpha-helices are individually only marginally stable, a property that may be of significance with regard to the retracted conformation in which sigma 1 is accommodated in the intact virion, and (ii) the predominant interactions between the two coiled coils are likely to involve hydrogen bonding between patches of uncharged residues.

Oral inoculation of suckling mice with reovirus serotype 1 (strain Lang) results in the conversion of intact virions to intermediate subviral particles (ISVPs) in the intestinal lumen. Digestion of virus in vitro with chymotrypsin or trypsin reveals two distinct forms of ISVPs, while the predominant species of ISVPs found in the small intestinal lumen appears to be identical to the chymotrypsin product. The in vivo conversion of virions to ISVPs was blocked by pretreatment of mice with protease inhibitors, resulting in inefficient replication of reovirus in intestinal tissue. The early inhibition of viral replication in suckling mice pretreated with protease inhibitors was not observed when suckling mice were inoculated with ISVPs generated by in vitro digestion with either chymotrypsin or trypsin. However, replication was decreased during secondary rounds of replication in mice receiving repeated doses of protease inhibitors, suggesting that luminal proteolytic digestion is important in rendering progeny virions infectious in the gut.

To explore how bacteria and their products may modulate viral infection, we investigated the effect of a well-characterized and highly purified product of Bordetella pertussis, a pertussis toxin, also known as lymphocytosis promoting factor (LPF), on enteric reovirus infection. LPF is known to have a variety of effects, including modulation of circulation and homing of lymphoid cells. When adult mice are inoculated with reovirus type 1 perorally, reovirus first enters the Peyer's patches (PP) through M cells, and then spreads to mesenteric lymph nodes (MLN) and spleen with minimal dissemination to other peripheral tissues. In view of the profound effect of LPF on lymphoid tissues, we evaluated whether LPF might influence the early stages of type-1 reovirus infection following peroral inoculation. Pretreatment of adult BALB/c mice with LPF significantly inhibited the spread of reovirus in a manner dependent upon the route of inoculation; LPF inhibited the extra-intestinal spread of virus from PP to MLN after intragastric inoculation; in contrast there was enhancement of the spread of blood-borne viruses to MLN after intravenous inoculation. This result, together with the fact that the efferent lymph from PP reaches MLN, suggests that a proportion of reoviruses were conveyed from PP to MLN in association with lymphoid cells along the lymphatic channels and that LPF affects reovirus, in part, by blocking cell movement.

A series of 94 isolates of reovirus from humans, cattle, and mice, showed extensive variability in the patterns of migration of the ten double-stranded RNA genome segments. This variation was found in all three serotypes, and involved all ten genome segments, including the segment responsible for serological specificity. Although a single pattern was present among several samples isolated from individuals and collected at a single time and place, there were often multiple genetic variants of a single serotype present in a population. Samples isolated from widely different geographic origins or different mammalian hosts showed different patterns; samples from a single species from the same area over a period of time showed more limited variations. Among most isolates, the migration of the slowest S segment, the segment that encodes the hemagglutinin and is responsible for serological specificity in laboratory strains, was similar to reference strains for type 1 and type 3 isolates. However, the type 2 isolates showed considerable variation in this segment.

We studied the ability of chymotrypsin to activate the transcriptases of the three serotypes of reovirus. When we used conditions that reproducibly caused the activation of type 3 transcriptase by chymotrypsin alone, type 2 transcriptase was sometimes activated, and type 1 transcriptase was never activated. Using intertypic recombinants containing various combinations of genome segments from reovirus types 3 and 1, we showed that the M2 segment determined this difference. Biochemical experiments indicated that the digestion of reovirus type 1 by chromotrypsin was blocked at an intermediate stage in uncoating. We found conditions which reproducibly activated the transcriptases of all three serotypes. This allowed us to compare the biochemical properties of the three transcriptases. Although the monovalent cation preferences, divalent cation preferences and optima, and temperature optima of type 1, 2, and 3 transcriptases were indistinguishable, the pH activity curves were reproducibly different. The largest difference was between type 2 and 3 transcriptases; the pH optimum of type 2 transcriptase was lower than the pH optimum of type 3 transcriptase. Using intertypic recombinants containing various combinations of genome segments from reovirus types 2 and 3, we demonstrated that the L1 segment specified this difference.

The pathogenesis of infection of mice by the mammalian reoviruses involves several discrete steps. Each of the three viral outer capsid proteins has a highly distinct and specialized role: one protein (sigma 1) binds to cell surface receptors; a second protein (mu 1C) determines the capacity for viral growth at mucosal surfaces; and the third protein (sigma 3) is responsible for inhibiting cell macromolecular synthesis. A detailed picture of the molecular basis of reovirus virulence and attention is now emerging.