Establishment and maintenance of persistent infection by Sindbis virus in BHK cells.

We have established a persistent infection of BHK cells with a preparation of Sindbis virus heavily enriched in defective interfering (DI) particles. The small fraction of cells that survived the initial infection grew out to form a stable population of cells [BHK(Sin-1) cells], most of which synthesized viral RNA and viral antigens. The presence of DI particles in this virus stock was required to establish this persistent state. BHK(Sin-1) cells released a small-plaque, temperature-sensitive virus (Sin-1 virus) as well as DI particles containing DI RNAs larger than those present in the original stock used to establish the persistent state. A cloned stock of Sin-1 virus, free of detectable DI particles, was able to initiate a persistent infection more quickly and with greater cell survival than the original stock of Sindbis virus containing DI particles. About 2 weeks after the Sin-1 virus-infected cells were cultured, DI RNAs arose and soon became the dominant viral RNA species produced by these cells.     

Defective interfering particles of Sindbis virus do not interfere with the homologous virus obtained from persistently infected BHK cells but do interfere with Semliki Forest virus.

Defective interfering particles derived from wild-type Sindbis virus no longer interfere with the infectious virus cloned from BHK cells persistently infected with Sindbis virus for 16 months. These particles do interfere with the replication of Semliki Forest virus.     

Defective Interfering Passages of Sindbis Virus: Nature of the Intracellular Defective Viral RNA

BHK cells infected with defective-interfering passages of Sindbis virus accumulate a species of RNA (20S) that is about half the molecular weight of the major viral mRNA (26S). We have performed competitive hybridization experiments with these species of RNA and have established that 20S RNA contains approximately 50% of the nucleotide sequences present in 26S RNA. Our further studies, however, demonstrate that 20S RNA is unable to carry out the messenger function of 26S RNA. We found very little of the defective RNA associated with polysomes in vivo. In addition, it was unable to stimulate protein synthesis in vitro under conditions in which 26S RNA was translated. We have also examined viral RNA synthesis in BHK cells infected with standard or defective-interfering passages of Sindbis virus. This comparison suggests that defective partioles do not synthesize a functional replicase.     

Heterologous interference in Aedes albopictus cells infected with alphaviruses.

Maximum amounts of 42S and 26S single-stranded viral RNA and viral structural proteins were synthesized in Aedes albopictus cells at 24 h after Sindbis virus infection. Thereafter, viral RNA and protein syntheses were inhibited. By 3 days postinfection, only small quantities of 42S RNA and no detectable 26S RNA or structural proteins were synthesized in infected cells. Superinfection of A. albopictus cells 3 days after Sindbis virus infection with Sindbis, Semliki Forest, Una, or Chikungunya alphavirus did not lead to the synthesis of intracellular 26S viral RNA. In contrast, infection with snowshoe hare virus, a bunyavirus, induced the synthesis of snowshoe hare virus RNA in both A. Ablpictus cells 3 days after Sindbis virus infection and previously uninfected mosquito cells. These results suggested that at 3 days after infection with Sindbis virus, mosquito cells restricted the replication of both homologous and heterologous alphaviruses but remained susceptible to infection with a bunyavirus. In superinfection experiments the the alphaviruses were differentiated on the basis of plaque morphology and the electrophoretic mobility of their intracellular 26S viral RNA species. Thus, it was shown that within 1 h after infection with eigher Sindbis or Chikungunya virus, A. albopictus cells were resistant to superinfection with Sindbis, Chikungunya, Una, and Semliki Forest viruses. Infected cultures were resistant to superinfection with the homologous virus indefinitely, but maximum resistance to superinfection with heterologous alphaviruses lasted for approximately 8 days. After that time, infected cultures supported the replication of heterologous alphaviruses to the same extent as did persistently infected cultures established months previously. However, the titer of heterologous alphavirus produced after superinfection of persistently infected cultures was 10- to 50-fold less than that produced by an equal number of previously uninfected A. albopictus cells. Only a small proportion (8 to 10%) of the cells in a persistently infected culture was capable of supporting the replication of a heterologous alphavirus.     

Common and distinct regions of defective-interfering RNAs of Sindbis virus

Defective-interfering (DI) particles are helper-dependent deletion mutants which interfere specifically with the replication of the homologous standard virus. Serial passaging of alphaviruses in cultured cells leads to the accumulation of DI particles whose genomic RNAs are heterogeneous in size and sequence composition. In an effort to examine the sequence organization of an individual DI RNA species generated from Sindbis virus, we isolated and sequenced a representative cDNA clone derived from a Sindbis DI RNA population. Our data showed that: (i) the 3' end of the DI RNA template was identical to the 50 nucleotides at the 3' end of the standard RNA; (ii) the majority (75%) of the DI RNA template was derived from the 1,200 5'-terminal nucleotides of the standard RNA and included repeats of these sequences; and (iii) the 5' end of the DI RNA template was not derived from the standard RNA, but is nearly identical to a cellular tRNAAsp (S. S. Monroe and S. Schlesinger, Proc. Natl. Acad. Sci. U.S.A. 80:3279-3283, 1983). We have also utilized restriction fragments from cloned DNAs to probe by blot hybridization for the presence of conserved sequences in several independently derived DI RNA populations. These studies indicated that: (i) a 51-nucleotide conserved sequence located close to the 5' end of several alphavirus RNAs was most likely retained in the DI RNAs; (ii) the junction region containing the 5' end of the subgenomic 26S mRNA was deleted from the DI RNAs; and (iii) the presence of tRNAAsp sequences was a common occurrence in Sindbis virus DI RNAs derived by passaging in chicken embryo fibroblasts.     

Sequence analysis of cDNA''s derived from the RNA of Sindbis virions and of defective interfering particles

Sindbis virus generates defective interfering (DI) particles during serial high-multiplicity passage in cultured cells. These DI particles inhibit the replication of infectious virus and can be an important factor in the establishment and maintenance of persistent infection in BHK cells. In an effort to understand how these DI particles are generated and how they interfere with the replication of standard virus, we performed a partial sequence analysis of the RNA obtained from two independently isolated populations of DI particles and from two Sindbis virus variants and compared these with the RNA of the parental wild-type virus. The 3'-terminal regions of the RNAs were sequenced by the dideoxy chain terminating method. Internal regions of the RNA were examined by restriction endonuclease digestion of cDNA's made to the various RNAs and by direct chemical sequencing of 5' end-labeled restriction fragments from cDNA made to the DI RNAs. One of the variant viruses examined was originally derived from cells persistently infected with Sindbis virus for 16 months and is resistant to interference by the DI strains used. In the 3'-terminal region of the RNA from this variant, only two base changes were found; one of these occurs in the 20-nucleotide 3'-terminal sequence which is highly conserved among alphaviruses. The DI RNA sequences were found to have been produced not by a single deletional event, but by multiple deletion steps combined with sequence rearrangements; all sequences examined are derived from the plus strand of Sindbis virion RNA. Both DI RNAs had at least 50 nucleotides of wild-type sequence conserved at the 3' terminus; in addition, they both contained conserved and perhaps amplified sequences derived from the non-26S region of the genome which may be of importance in their replication and interference ability.     

Studies of defective interfering RNAs of Sindbis virus with and without tRNAAsp sequences at their 5'' termini.

Three of six independently derived defective interfering (DI) particles of Sindbis virus generated by high-multiplicity passaging in cultured cells have tRNAAsp sequences at the 5' terminus of their RNAs (Monroe and Schlesinger, J. Virol. 49:865-872, 1984). In the present work, we found that the 5'-terminal sequences of the three tRNAAsp-negative DI RNAs were all derived from viral genomic RNA. One DI RNA sample had the same 5'-terminal sequence as the standard genome. The DI RNAs from another DI particle preparation were heterogeneous at the 5' terminus, with the sequence being either that of the standard 5' end or rearrangements of regions near the 5' end. The sequence of the 5' terminus of the third DI RNA sample consisted of the 5' terminus of the subgenomic 26S mRNA with a deletion from nucleotides 24 to 67 of the 26S RNA sequence. These data showed that the 5'-terminal nucleotides can undergo extensive variations and that the RNA is still replicated by virus-specific enzymes. DI RNAs of Sindbis virus evolve from larger to smaller species. In the two cases in which we followed the evolution of DI RNAs, the appearance of tRNAAsp-positive molecules occurred at the same time as did the emergence of the smaller species of DI RNAs. In pairwise competition experiments, one of the tRNAAsp-positive DI RNAs proved to be the most effective DI RNA, but under identical conditions, a second tRNAAsp-positive DI RNA was unable to compete with the tRNAAsp-negative DIs. Therefore, the tRNAAsp sequence at the 5' terminus of a Sindbis DI RNA is not the primary factor in determining which DI RNA becomes the predominant species in a population of DI RNA molecules.     

Packaging signals in alphaviruses.

Alphaviruses synthesize large amounts of both genomic and subgenomic RNA in infected cells, but usually only the genomic RNA is packaged. This implies the existence of an encapsidation or packaging signal which would be responsible for selectivity. Previously, we had identified a region of the Sindbis virus genome that interacts specifically with the viral capsid protein. This 132-nucleotide (nt) fragment lies within the coding region of the nsP1 gene (nt 945 to 1076). We proposed that the 132-mer is important for capsid recognition and initiates the formation of the viral nucleocapsid. To study the encapsidation of Sindbis virus RNAs in infected cells, we designed a new assay that uses the self-replicating Sindbis virus genomes (replicons) which lack the viral structural protein genes and contain heterologous sequences under the control of the subgenomic RNA promoter. These replicons can be packaged into viral particles by using defective helper RNAs that contain the structural protein genes (P. Bredenbeek, I. Frolov, C. M. Rice, and S. Schlesinger, J. Virol. 67:6439-6446, 1993). Insertion of the 132-mer into the subgenomic RNA significantly increased the packaging of this RNA into viral particles. We have used this assay and defective helpers that contain the structural protein genes of Ross River virus (RRV) to investigate the location of the encapsidation signal in the RRV genome. Our results show that there are several fragments that could act as packaging signals. They are all located in a different region of the genome than the signal for the Sindbis virus genome. For RRV, the strongest packaging signal lies between nt 2761 and 3062 in the nsP2 gene. This is the same region that was proposed to contain the packaging signal for Semliki Forest virus genomic RNA.     

mRNA activity of a Sindbis virus defective-interfering RNA.

We obtained Sindbis defective-interfering particles by nine and undiluted passages of standard virus on chicken embryo fibroblasts. These particles contain a deleted 20S RNA molecule which has mRNA activity, as shown by translation in cell-free systems in vitro. In infected cells, this mRNA activity appeared to be totally inhibited except at very late times postinfection.     

Failure of defective interfering particles of Sindbis virus produced in BHK or chicken cells to affect viral replication in Aedes albopictus cells.

Whereas defective interfering particles of Sindbis virus are readily produced in BHK-21 cells or chicken embryo fibroblasts by the techniques of serial undiluted passage, similar methods failed to generate such particles in Aedes albopictus cell cultures. In addition, Sindbis virus stocks produced in BHK-21 cells or chicken embryo fibroblasts and which contained defective interfering particles, when tested in A. albopictus cells, failed (i) to interfere with the replication of standard Sindbis virus and (ii) to change the pattern of intracellular viral RNA synthesis from that produced by infection with standard Sindbis virus alone. We conclude that defective interfering particles of Sindbis virus generated in chicken or hamster cells are silent or inert in mosquito cells.     

Persistence of vesicular stomatitis virus in cloned interleukin-2-dependent natural killer cell lines.

We have investigated virus-lymphocyte interactions by using cloned subpopulations of interleukin-2-dependent effector lymphocytes maintained in vitro. Cloned lines of H-2-restricted hapten- or virus-specific cytotoxic T lymphocytes (CTL) and alloantigen-specific CTL were resistant to productive infection by vesicular stomatitis virus (VSV). In contrast, cloned lines of natural killer (NK) cells were readily and persistently infected by VSV, a virus which is normally highly cytolytic. VSV-infected NK cells continued to proliferate, express viral surface antigen, and produce infectious virus. Furthermore, persistently infected NK cells showed no marked alteration of normal cellular morphology and continued to lyse NK-sensitive target cells albeit at a slightly but significantly reduced level. The persistence of VSV in NK cells did not appear to be caused by the generation of temperature-sensitive viral mutants, defective interfering particles, or interferon. Consequently, studies comparing the intracellular synthesis and maturation of VSV proteins in infected NK and mouse L cells were conducted. In contrast to L cells, in which host cell protein synthesis was essentially totally inhibited by infection, the infection of NK cells caused no marked diminution in the synthesis of host cell proteins. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of immunoprecipitates of viral proteins from infected cells showed that the maturation rate and size of VSV surface G glycoprotein were comparable in L cells and NK cells. Nucleocapsid (N) protein synthesis also appeared to be unaffected in NK cells. In contrast, the viral proteins NS and M appeared to be selectively degraded in NK cell extracts. Mixing experiments suggested that a protease in NK cells was responsible for the selective breakdown of VSV NS protein. Finally, VSV-infected NK cells were resistant to lysis by virus-specific CTL, suggesting that persistently infected NK cells may harbor virus and avoid cell-mediated immune destruction in an immunocompetent host.     

Comparison of the effects of Sindbis virus and Sindbis virus replicons on host cell protein synthesis and cytopathogenicity in BHK cells.

Infection of BHK cells by Sindbis virus leads to rapid inhibition of host cell protein synthesis and cytopathic effects (CPE). We have been studying these events to determine whether the expression of a specific viral gene is required and, in the present study, have focused our attention on the role of the structural proteins--the capsid protein and the two membrane glycoproteins. We tested a variety of Sindbis viruses and Sindbis virus replicons (virus particles containing an RNA that is self-replicating but with some or all of the viral structural protein genes deleted) for their abilities to inhibit host cell protein synthesis and cause CPE in infected BHK cells. Our results show that shutoff of host cell protein synthesis occurred in infected BHK cells when no viral structural proteins were synthesized and also under conditions in which the level of the viral subgenomic RNA was too low to be detected. These results support the conclusion that the early steps in viral gene expression are the ones required for the inhibition of host cell protein synthesis in BHK cells. In contrast, the Sindbis viruses and Sindbis virus replicons were clearly distinguished by the time at which CPE became evident. Viruses that synthesized high levels of the two membrane glycoproteins on the surface of the infected cells caused a rapid (12 to 16 h postinfection) appearance of CPE, and those that did not synthesize the glycoprotein spikes showed delayed (30 to 40 h) CPE.     

Viral proteins and RNAs in BHK cells persistently infected by lymphocytic choriomeningitis virus

Some Syrian hamster cell lines persistently infected with lymphocytic choriomeningitis virus (LCMV) do not produce extracellular virus particles but do contain intracytoplasmic infectious material. The proteins of these cells were labeled with [35S]methionine or with [3H]glucosamine and [3H]mannose, and immunoprecipitates were prepared with anti-LCMV sera. A substantial amount of the LCMV nucleocapsid protein (molecular weight about 58,000) was detected, along with GP-C, the precursor of the virion glycoproteins GP-1 and GP-2. GP-1 and GP-2 themselves were not detected. A new method of transferring proteins electrophoretically from sodium dodecyl sulfate-polyacrylamide gels to diazotized paper in high yield revealed several additional LCMV proteins present specifically in the persistently infected cells, at apparent molecular weights (X10(3] of 112, 107, 103, 89, 71 (probably GP-C), 58 (nucleocapsid protein), 42 to 47 (probably GP-1), and 40 (possibly GP-2). By iodinating intact cells with I3, GP-1 but not GP-2 or GP-C was revealed on the surfaces of the persistently infected cells, whereas both GP-1 and GP-C were found on the surfaces of acutely infected cells. The absence of GP-C from the plasma membrane of the persistently infected cells might be related to defective maturation of the virus in these cells. Cytoplasmic viral nucleoprotein complexes were labeled with [3H]uridine in the presence or absence of actinomycin D, purified partially by sedimentation in D2O-sucrose gradients, and adsorbed to fixed Staphylococus aureus cells in the presence of anti-LCMV immunoglobulin G. Several discrete species of viral RNA were released from the immune complexes with sodium dodecyl sulfate. Some were appreciably smaller than the 31S and 23S species of standard LCMV virions, indicating that defective interfering viral RNAs are probably present in the persistently infected cells. Ribosomal 28S and 18S RNAs, labeled only in the absence of actinomycin D, were coprecipitated with anti-LCMV serum but not with control serum, indicating their association with LCMV nucleoproteins in the cells.     

Clonal analysis of mammalian cell cultures persistently infected with Japanese encephalitis virus.

More than 200 cells were cloned from populations of mammalian cells persistently infected with Japanese encephalitis virus. Only four cloned cultures contained cells that had viral antigen measurable by immunofluorescence and that released infectious virus, yet all clones harbored virus-specific RNA. Superinfection of cloned cells with wild-type Japanese encephalitis virus did not produce cytopathic effects, but resulted in production of viral antigen and infectious virus in formerly nonproducing clones. Cocultivation of nonproducer clone cells with normally permissive cells did not induce virus production, nor did treatment of nonproducer clones with various inhibitors of DNA, RNA, or protein synthesis. It is suggested that the cloning procedure may have selected for a particular subpopulation of cells and that defective virus is also involved in establishment and maintenance of persistent infection.     

Selection of functional 5' cis-acting elements promoting efficient sindbis virus genome replication

The 5' portion of the Sindbis virus (SIN) genome RNA is multifunctional. Besides initiating translation of the nonstructural polyprotein, RNA elements in the 5' 200 bases of the SIN genome RNA, or its complement at the 3' end of the negative-strand intermediate, play key roles in the synthesis of both negative- and positive-strand RNAs. We used here a combination of genetic and biochemical approaches to further dissect the functions of this sequence. Replacement of the SIN 5' end in defective-interfering (DI) and genome RNAs with sequences from a distantly related alphavirus, Semliki Forest virus (SFV), resulted in nonviable chimeras. The addition of five nucleotides from the 5' terminus of SIN restored negative-strand RNA synthesis in DI genomes but not their replication in vivo. Pseudorevertants of various SFV-SIN chimeras were isolated, and suppressor mutations were mapped to AU-rich sequences added to the 5' end of the original SFV 5' sequence or its "deleted" versions. Early pseudorevertants had heterogeneous 5' termini that were inefficient for replication relative to the parental SIN 5' sequence. In contrast, passaging of these pseudorevertant viral populations in BHK cells under competitive conditions yielded evolved, more homogeneous 5'-terminal sequences that were highly efficient for negative-strand synthesis and replication. These 5'-terminal sequences always began with 5'-AU, followed by one or more AU repeats or short stretches of oligo(A). Further analysis demonstrated a positive correlation between the number of repeat units and replication efficiency. Interestingly, some 5' modifications restored high-level viral replication in BHK-21 cells, but these viruses were impaired for replication in the cells of mosquito origin. These studies provide new information on sequence determinants required for SIN RNA replication and suggest new strategies for restricting cell tropism and optimizing the packaging of alphavirus vectors.     

Persistent infection of cultured mammalian cells by Japanese encephalitis virus.

Persistent infections were established by serial undiluted passage of flavivirus Japanese encephalitis virus in a line of rabbit kidney cells (MA-111). The persistently infected cells resembled uninfected cells in most respects. Low levels of infectious virions were released from a small percentage of cells, and a larger and more variable percentage was shown to possess viral antigen by fluorescent-antibody staining. Released viruses were shown to interfere with replication of wild-type Japanese encephalitis virus. Persistently infected MA-111 cells could not be superinfected with homologous wild-type Japanese encephalitis virus but could be superinfected with two heterologous viruses. Transfer of cell culture medium from persistently infected MA-111 cells to a line of African green monkey kidney cells (Vero) resulted in similar persistent infections in the latter cells. Temperature sensitivity and host-cell interferon production were not involved in establishment or maintenance of persistence. Determination of ratios of physical particles to infectious particles revealed that many defective, noninfectious viruses were present, suggesting that defective interfering particles may be responsible for persistency.