Analysis of the selective advantage conferred by a C-E1 fusion protein synthesized by rubella virus DI RNAs

During serial passaging of rubella virus (RUB) in cell culture, the dominant species of defective-interfering RNA (DI) generated contains an in-frame deletion between the capsid protein (C) gene and E1 glycoprotein gene resulting in production of a C-E1 fusion protein that is necessary for the maintenance of the DI [Tzeng, W.P., Frey, T.K. (2006). C-E1 fusion protein synthesized by rubella virus DI RNAs maintained during serial passage. Virology 356 198-207.]. A BHK cell line stably expressing the RUB structural proteins was established which was used to package DIs into virus particles following transfection with in vitro transcripts from DI infectious cDNA constructs. Packaging of a DI encoding an in-frame C-GFP-E1 reporter fusion protein corresponding to the C-E1 fusion protein expressed in a native DI was only marginally more efficient than packaging of a DI encoding GFP, indicating that the C-E1 fusion protein did not function by enhancing packaging. However, infection with the DI encoding the C-GFP-E1 fusion protein (in the absence of wt RUB helper virus) resulted in formation of clusters of GFP-positive cells and the percentage of GFP-positive cells in the culture following infection remained relatively constant. In contrast, a DI encoding GFP did not form GFP-positive clusters and the percentage of GFP-positive cells declined by roughly half from 2 to 4 days post-infection. Cluster formation and sustaining the percentage of infected (GFP-positive) cells required the C part of the fusion protein, including the downstream but not the upstream of two arginine clusters (both of which are associated with RNA binding and association with mitochondrial p32 protein) and the E1 part through the transmembrane sequence, but not the C-terminal cytoplasmic tail. Among a collection of mutant DI constructs, cluster formation and sustaining infected cell percentage correlated with maintenance during serial passage with wt RUB. We hypothesize that cluster formation and sustaining infected cell percentage increase the likelihood of co-infection by a DI and wt RUB during serial passage thus enhancing maintenance of the DI. Cluster formation and sustaining infected cell percentage were found to be due to a combination of attenuated cytopathogenicity of DIs that express the C-E1 fusion protein and cell-to-cell movement of the DI. In infected cells, the C-GFP-E1 fusion protein was localized to potentially novel vesicular structures that appear to originate from ER-Golgi transport vacuoles. This species of DI expressing a C-E1 fusion protein that exhibits attenuated cytopathogenicity and the ability to increase the number of infected cells through cell-to-cell movement could be the basis for development of an attractive vaccine vector.

Superinfection of human immunodeficiency virus type 1 (HIV-1) to cell clone persistently infected with defective virus induces production of highly cytopathogenic HIV-1

Superinfection with human immunodeficiency virus type 1 (HIV-1) in human subjects, defined as reinfection with a heterologous strain of HIV-1, has become a topic of great interest. To illustrate the significance of this occurrence, we performed HIV-1 superinfection of L-2 cells, which were isolated from MT-4 cells persistently infected with subtype B HIV-1 as a cell clone continuously producing defective HIV-1 particles. L-2 cells carrying provirus with a one-base insertion in the pol protease were superinfected with HIV-1 derived from primary isolates of subtype B or CRF01_AE. The kinetics of the superinfection in L-2 were very slow compared with those of primary infections in MT-4. Interestingly, L-2 shifted after superinfection to become a producer of highly cytopathogenic HIV-1. Molecular characterization revealed that superinfection occurred in only about 10% of the CRF01_AE-superinfected L-2, which carried provirus of both subtypes and produced viral particles containing genomic RNA of both subtypes. Surprisingly, such cytopathogenic HIV-1 showed predominantly the original subtype B phenotype. Thus, the mechanism of the production of cytopathic HIV-1 seemed to be mediated by trans complementation with pol products of superinfected CRF01_AE. These findings suggest the significance of long-lived infected cells as recipients for superinfection that could result in the generation of new HIV-1 variants with high virulence in patients who are off therapy or do not adhere to treatment, and may indicate the need for precautions against such superinfection.

In vitro infection of human peripheral blood mononuclear cells by a defective hepatitis B virus with a deletion in the PreS1 region of the viral genome

Previously, we identified a defective hepatitis B virus (HBV) which contains a 183 nucleotide deletion in the PreS1 region of the viral genome affecting the S gene promoter in sera from hepatitis B surface antigen (HBsAg)-negative patients with serum HBV-DNA. The aim of this study was to analyse the infectivity of this mutant. Peripheral blood mononuclear cells (PBMC) from a healthy donor were incubated with serum samples from 2 HBsAg-negative patients with serum HBV-DNA (infected with wild-type and deletion mutant HBV), from an HBsAg carrier (infected with wild-type HBV) and from a healthy donor. After 1 week, HBV-DNA was detected by polymerase chain reaction (PCR) in all supernatants and cells incubated with the HBV-DNA-positive inocula. DNase and trypsin pretreatment confirmed intracellular localization of HBV-DNA in cells. HBV-RNA and covalently closed circular HBV-DNA were also detected in PBMC, indicating that the viral DNA infecting these cells was transcriptionally active. Deletion mutant and wild-type HBV were detected in the supernatants and cells infected with the two HBsAg-negative sera, while only wild-type HBV was detected in the supernatant and cells incubated with the serum from the HBsAg-carrier. In conclusion, this HBV deletion mutant can infect, replicate and release viral particles in in vitro infected PBMC.

Development of HSV-specific CD4+ Th1 responses and CD8+ cytotoxic T lymphocytes with antiviral activity by vaccination with the HSV-2 mutant ICP10DeltaPK

A growth compromised herpes simplex virus type 2 (HSV-2) mutant which is deleted in the PK domain of the large subunit of ribonucleotide reductase (ICP10DeltaPK) protects from HSV-2 challenge in the mouse and guinea pig cutaneous and vaginal models and reduces the incidence and frequency of recurrent disease (Vaccine (17) (1999) 1951; Vaccine (19) (2001) 1879). The present studies were designed to identify the immune responses induced by ICP10DeltaPK and define the component responsible for protective activity. We found that ICP10DeltaPK elicits a predominant HSV-specific T helper type 1 (Th1) response, as evidenced by: (1) higher levels of HSV-specific IgG2a (Th1) than IgG1 (Th2) isotypes and (2) higher numbers of CD4+ IFN-gamma than IL-10 secreting T cells in popliteal lymph nodes. This Th1 response pattern was associated with a significant increase in the levels of IL-12 produced by dendritic cells from ICP10DeltaPK than HSV-2 immunized animals. Lymph node cells (LNCs) from ICP10DeltaPK immunized mice had significantly higher levels of HSV-2 specific cytolytic activity than LNCs from mice immunized with HSV-2 and it was mediated by CD8+ T cells. CD8+ CTL were not seen in LNCs from HSV-2 immunized mice. In adoptive transfer experiments, CD8+ T cells and, to a lower extent, CD4+ T cells from ICP10DeltaPK immunized mice inhibited HSV-2 replication, suggesting that they are involved in the protective immunity induced by ICP10DeltaPK vaccination.

Evaluation of infection parameters in the production of replication-defective HSV-1 viral vectors

Herpes simplex virus type-1 (HSV-1) is a neurotrophic human pathogen that establishes life-long latency in the nervous system. Our laboratory has extensively engineered this virus to retain the ability to persist in neurons without expression of lytic genes or disease phenotype. Highly defective, replication-incompetent HSV mutants are thus potentially ideal for transfer of therapeutic transgenes to human nerves where long-term therapy of nervous system disease may be provided. A prerequisite for using recombinant HSV vectors for therapeutic gene delivery to humans is the development of methods for large-scale manufacture of HSV vectors. Here we report studies to identify infection parameters that result in high-yield production of immediate early gene deletion mutant HSV vectors in complementing cells that supply the deleted essential viral functions in trans. Virus yield was correlated with various culture media conditions that included pH, glucose metabolism, and serum levels. The results demonstrated that systematic media exchange to remove lactate derived from high-level glucose consumption, maintenance of tissue culture pH at 6.8, and the use of 5% fetal bovine serum gave the highest yield of infectious virus. The data indicate that these are important parameters to consider for high-yield, large-scale virus production.

Proteolytic processing of human cytomegalovirus glycoprotein B is dispensable for viral growth in culture

Glycoprotein B (gB) of human cytomegalovirus (HCMV), which is considered essential for the viral life cycle, is proteolytically processed during maturation. Since gB homologues of several other herpesviruses remain uncleaved, the relevance of this property of HCMV gB for viral infectivity is unclear. Here we report on the construction of a viral mutant in which the recognition site of gB for the cellular endoprotease furin was destroyed. Because mutagenesis of essential proteins may result in a lethal phenotype, a replication-deficient HCMV gB-null genome encoding enhanced green fluorescent protein was constructed, and complementation by mutant gBs was initially evaluated in transient-cotransfection assays. Cotransfection of plasmids expressing authentic gB or gB with a mutated cleavage site (gB-DeltaFur) led to the formation of green fluorescent miniplaques which were considered to result from one cycle of phenotypic complementation of the gB-null genome. To verify these results, two recombinant HCMV genomes were constructed: HCMV-BAC-DeltaMhdI, with a deletion of hydrophobic domain 1 of gB that appeared to be essential for viral growth in the cotransfection experiments, and HCMV-BACDeltaFur, in which the gB cleavage site was mutated by amino acid substitution. Consistent with the results of the cotransfection assays, only the DeltaFur mutant replicated in human fibroblasts, showing growth kinetics comparable to that of wild-type virus. gB in mutant-infected cells was uncleaved, whereas glycosylation and transport to the cell surface were not impaired. Extracellular mutant virus contained exclusively uncleaved gB, indicating that proteolytic processing of gB is dispensable for viral replication in cell culture.

Short defective interfering RNAs of tombusviruses are not targeted but trigger post-transcriptional gene silencing against their helper virus

Post-transcriptional gene silencing (PTGS) is a sequence-specific degradation mechanism that operates in almost all eukaryotic cells. In plants, double-stranded RNA triggers PTGS, generating 21- to 25-nucleotide guide RNAs responsible for specific degradation of cognate mRNA. The double stranded RNA intermediates of replicating plant viruses often induce PTGS, leading to symptom attenuation. Here we demonstrate the role of PTGS in defective interfering (DI) RNA-mediated symptom attenuation in plants infected with Cymbidium ringspot tombusvirus (CymRSV). Analysis of 21- to 25-nucleotide RNAs in Nicotiana benthamiana infected with CymRSV indicated that PTGS was not spread homogeneously along the viral genome. The 21- to 25-nucleotide RNAs derived mainly from plus-stranded RNA and likely arose from local basepaired structures. In contrast to helper viral RNA, short DI RNAs were not accessible to helper virus-induced RNA degradation guided by the 21- to 25-nucleotide RNAs. Our results suggest a model in which PTGS plays an important role in the selective accumulation and symptom attenuation mediated by DI RNAs. Because PTGS operates in a wide variety of different organisms, this model is applicable to DI RNA generation and accumulation in both plant and animal cells.

A growth and latency compromised herpes simplex virus type 2 mutant (ICP10DeltaPK) has prophylactic and therapeutic protective activity in guinea pigs

A growth compromised herpes simplex virus type 2 (HSV-2) mutant which is deleted in the PK domain of the large subunit of ribonucleotide reductase (ICP10DeltaPK) protects from fatal HSV-2 challenge in the mouse model (Aurelian L, Kokuba H, Smith CC. Vaccine potential of a Herpes Simplex Virus type 2 mutant deleted in the PK domain of the large subunit of ribonucleotide reductase (ICP10). Vaccine 1999;17:1951-1963). Here we report the results of our studies with ICP10DeltaPK in the guinea pig model of recurrent HSV-2 disease. ICP10DeltaPK was also compromised for growth and disease causation in this model. It was not isolated from latently infected ganglia by explant co-cultivation. The proportions of latently infected ganglia were significantly lower for ICP10DeltaPK than HSV-2 [3/25 (12%) and 7/10 (70%), respectively]. Similar results were obtained for the levels of viral DNA (8 x 10(3) and 2 x 10(5) molecules/ganglion for ICP10DeltaPK and HSV-2, respectively]. ICP10DeltaPK immunization caused a significant (P< or = 0.001) decrease in the proportion of animals with primary [1/14 (6%) and 16/16 (100%) for ICP10DeltaPK and PBS, respectively) and recurrent [1/14 (6%) and 11/14 (79%) for ICP10DeltaPK and PBS, respectively) HSV-2 skin lesions. It also protected from genital HSV-2 disease [1/10 and 10/10 for ICP10DeltaPK and PBS, respectively] and decreased the severity of the lesions in both models. Quantitative PCR (Q-PCR) with primers that distinguish between HSV-2 and ICP10DeltaPK indicated that immunization reduced the proportion of ganglia positive for HSV-2 DNA [8/25 (32%) and 7/10 (70%) for ICP10DeltaPK and PBS, respectively) and its levels [3 x 10(3) and 2 x 10(5) molecules/ganglion for ICP10DeltaPK and PBS, respectively]. The proportion of HSV-2 infected animals with recurrent disease was also significantly (P < or = 0.001) decreased by immunization with ICP10DeltaPK [1/15 (7%) and 11/14 (79%) with recurrent disease for ICP10DeltaPK and PBS, respectively], suggesting that ICP10DeltaPK has prophylactic and therapeutic activity in the guinea pig.

Targeted foreign gene expression in spinal cord neurons using poliovirus replicons

A hallmark of poliovirus is the propensity to infect and replicate in spinal cord neurons of the central nervous system. Previously, we characterized a poliovirus self-replicating RNA genome (replicon), which encodes firefly luciferase in place of the capsid genes. This replicon is encapsidated into an authentic poliovirion by providing the poliovirus capsid protein in trans. The amount of enzymatically active luciferase in cells infected with this replicon correlated with the infectious dose. To begin to characterize the in vivo infectious potential of replicons, we have inoculated mice transgenic for the human receptor for poliovirus (PVR), either intracranially or intraspinally, with the replicon encoding luciferase. Wild-type poliovirus delivered to PVR mice via intracranial or intraspinal routes resulted in paralysis and death. Replicon preparations were shown by a sensitive biological assay to be free of infectious poliovirus. Neither intracranial nor intraspinal inoculation of the replicon encoding luciferase resulted in any obvious paralysis or disease symptoms. Following intraspinal inoculation with replicons encoding luciferase, luciferase enzyme activity was detected at 4 h post-inoculation, with peak activity at approximately 8 h post-inoculation; by 48 - 72 h, the luciferase activity had returned to background levels. Luciferase activity was detected in spinal cord predominantly near the site of inoculation, although activity was detected anterior and posterior to the site of inoculation, indicating that replicons undergo limited movement within the CNS presumably via the cerebrospinal fluid. In stark contrast to poliovirus though, inoculation of replicons into the spinal cords of PVR mice did not result in noticeable pathogenesis. Using immunofluorescence with antibodies to double-stain for replicons and neurons, we determined that replicons exclusively infect the neurons of the spinal cord, with the expression of the luciferase and replicon proteins confined to the cytoplasm of the infected cells. Replicons, then, possess the identical capacity for infection of spinal cord neurons in vivo as poliovirus. The lack of discernible neuronal destruction following replicon inoculation into the spinal cord suggests that some of the pathogenesis observed during a poliovirus infection might not be due entirely to primary infection of neurons. Finally, the results of this study point to future use of replicons as a means to target recombinant protein expression to neurons in the spinal cord.

Construction and characterization of an equine herpesvirus 1 glycoprotein C negative mutant

An equine herpesvirus 1 (EHV-1) strain RacL 11 mutant was constructed that carries the Escherichia coli LacZ gene instead of the open reading frame encoding glycoprotein C (gC). The engineered virus mutant (L11(delta)gC) lacked codons 46-440 of the 1404 bp gene. On rabbit kidney cell line Rk13 and equine dermal cell line Edmin337, the L11(delta)gC virus grew to titers which were reduced by approximately 5- to 10-fold compared with wild-type RacL11 virus or a repaired virus (R-L11(delta)gC). However, when L11(delta)gC growth properties were analyzed on primary equine cells a decrease of viral titers was observed such that extracellular L11(delta)gC titers were reduced by 48- to 210-fold compared with those of wild-type or repaired virus. Heparin sensitive and heparin resistant attachment was assessed by binding studies using radiolabeled virion preparations. These studies revealed that EHV-1 gC is important for heparin sensitive attachment to the target cell. Similar results were obtained when cellular glycosaminoglycan (GAG) synthesis was inhibited by chlorate treatment or when cells defective in GAG synthesis were used. L11(delta)gC also exhibited significantly delayed penetration kinetics on Rk13 and primary equine cells. Infection of mice with L11(delta)gC did not cause EHV-1-related disease, whereas mice infected with either RacL11 or R-L11(delta)gC exhibited massive bodyweight losses, high virus titers in the lungs, and viremia. Taken together, EHV-1 gC was shown to play important roles in the early steps of infection and in release of virions, especially in primary equine cells, and contributes to EHV-1 virulence.

The defective interfering particles of equine herpesvirus 1 encode an ICP22/ICP27 hybrid protein that alters viral gene regulation

The genomes of equine herpesvirus 1 (EHV-1) defective interfering (DI) particles that mediate persistent infection were shown to encode a unique hybrid open reading frame composed of sequences that encode the 196 N-terminal amino acids of ICP22 linked in-frame to the C-terminal 68 amino acids of ICP27. Previous studies demonstrated that this hybrid gene, designated as ICP22/ICP27. was expressed abundantly at both the mRNA and the protein levels in DI particle-enriched infections, but not in standard EHV-1 infection (Chen et al., 1996 J. Virol. 70, 313-320). Since the ICP22/ICP27 hybrid protein contains portions of two EHV-1 early regulatory proteins, its effect on EHV-1 gene regulation was investigated. In EHV-1-infected cells, the ICP22/ICP27 hybrid protein expressed from plasmid vectors significantly reduced expression of a reporter gene under the control of the EHV-1 immediate-early (IE) gene promoter and early gene promoter, such as the viral ICP27 gene. In uninfected cells, the ICP22/ICP27 hybrid protein moderately down-regulated the IE and ICP22 promoters, up-regulated late gene promoters such as IR5, and altered the regulatory function of the IE and 1CP22 proteins in co-transfected cells. These results demonstrated that DI particles might alter viral gene regulation by expression of a unique hybrid gene encoded on the DI particle genome.

Deletion analysis of a defective interfering Semliki Forest virus RNA genome defines a region in the nsP2 sequence that is required for efficient packaging of the genome into virus particles

The 1,244-nucleotide genome of Semliki Forest virus (SFV) defective interfering (DI) RNA 19 (DI-19) is coterminal with the infectious genome and contains two major deletions. One deletion removes the end of the nsP1 gene and the beginning of the nsP2 gene, and the other removes the end of the nsP2 gene, the nsP3 and nsP4 genes, and all of the structural protein genes (M. Thomson and N. J. Dimmock, Virology 199:354-365, 1994). Like all DI SFV RNAs, DI-19 contains three regions that are conserved. Region a comprises the 5' terminus continuous with part of the nsP1 gene, region b comprises a central part of the nsP2 gene, and region c comprises the 3' terminus and the associated untranslated region. A deletion analysis of the 265-nucleotide b region (nucleotides 679 to 943, inclusive) was undertaken to determine its role in genome replication and packaging into DI virus particles. Deleted plasmids were constructed and transcribed, and the resulting DI RNAs were transfected into SFV-infected BHK cells. Putative progeny DI virus particles that had been released into the tissue culture fluid were then serially passaged in new monolayers together with added high-multiplicity SFV, and cells and tissue culture fluids were tested for the presence of DI RNA by reverse transcription-PCR. DI RNA that had all of the b region deleted was replicated well in BHK-21 cells, as shown by the presence of large amounts of negative-sense DI RNA and an increase in the amount of positive-sense RNA in the cytoplasm, but was packaged very inefficiently, as indicated by very low amounts of DI RNA in the tissue culture fluid. The genome of a deletion mutant that retained the 3' 224 nucleotides of region b was packaged successfully, but one that retained only the 5' 41 nucleotides was not detected in the tissue culture fluid. These and other data suggest that nucleotides 720 to 777 of region b are of particular importance in the packaging process. This finding agrees with data obtained with Ross River virus and contrasts with the well-studied Sindbis alphavirus major packaging signal that is located within the nsP1 gene.

Lysis genes of the Bacillus subtilis defective prophage PBSX

Four genes identified within the late operon of PBSX show characteristics expected of a host cell lysis system; they are xepA, encoding an exported protein; xhlA, encoding a putative membrane-associated protein; xhlB, encoding a putative holin; and xlyA, encoding a putative endolysin. In this work, we have assessed the contribution of each gene to host cell lysis by expressing the four genes in different combinations under the control of their natural promoter located on the chromosome of Bacillus subtilis 168. The results show that xepA is unlikely to be involved in host cell lysis. Expression of both xhlA and xhlB is necessary to effect host cell lysis of B. subtilis. Expression of xhlB (encoding the putative holin) together with xlyA (encoding the endolysin) cannot effect cell lysis, indicating that the PBSX lysis system differs from those identified in the phages of gram-negative bacteria. Since host cell lysis can be achieved when xlyA is inactivated, it is probable that PBSX encodes a second endolysin activity which also uses XhlA and XhlB for export from the cell. The chromosome-based expression system developed in this study to investigate the functions of the PBSX lysis genes should be a valuable tool for the analysis of other host cell lysis systems and for expression and functional analysis of other lethal gene products in gram-positive bacteria.

The null mutant of the U(L)31 gene of herpes simplex virus 1: construction and phenotype in infected cells

Earlier studies have shown that the U(L)31 protein is homogeneously distributed throughout the nucleus and cofractionates with nuclear matrix. We report the construction from an appropriate cosmid library a deletion mutant which replicates in rabbit skin cells carrying the U(L)31 gene under a late (gamma1) viral promoter. The mutant virus exhibits cytopathic effects and yields 0.01 to 0.1% of the yield of wild-type parent virus in noncomplementing cells but amounts of virus 10- to 1,000-fold higher than those recovered from the same cells 3 h after infection. Electron microscopic studies indicate the presence of small numbers of full capsids but a lack of enveloped virions. Viral DNA extracted from the cytoplasm of infected cells exhibits free termini indicating cleavage/packaging of viral DNA from concatemers for packaging into virions, but analyses of viral DNAs by pulsed-field electrophoresis indicate that at 16 h after infection, both the yields of viral DNA and cleavage of viral DNA for packaging are decreased. The repaired virus cannot be differentiated from the wild-type parent. These results suggest the possibility that U(L)31 protein forms a network to enable the anchorage of viral products for the synthesis and/or packaging of viral DNA into virions.

Assembly of hepatitis delta virus-like empty particles in yeast

Large delta antigen (L-HDAg) of hepatitis delta virus (HDV) and small-form hepatitis B surface antigen (HBsAg) of helper hepatitis B virus have previously been shown to be the minimum components for the assembly of HDV-like particles in mammalian cells. Extending from this finding, we coexpressed L-HDAg and small HBsAg in Saccharomyces cerevisiae to study their assembly in yeast cells. The assembly of virus particles from L-HDAg and HBsAg in yeast was demonstrated by their coexistence in the same isopycnic fractions and by the coimmunoprecipitation of L-HDAg with HBsAg using an antibody against HBsAg (anti-HBs). Furthermore, after purification by affinity chromatography with anti-HBs, HDV-like particles with size and morphology similar to those derived from mammalian cells could be visualized by electron microscopy. Mice immunized with yeast-derived HDV-like particles simultaneously acquired antibodies against HBsAg and HDAg, indicating that both viral proteins are antigenic. The results indicated that S. cerevisiae could serve as a host for the assembly of HDV-like empty particles. This system may be useful in investigating cellular processes involved in HDV assembly and in producing ample amount of HDV-like particles for structural and immunological studies.

Construction of a vaccinia virus deficient in the essential DNA repair enzyme uracil DNA glycosylase by a complementing cell line

The vaccinia virus D4R open reading frame, encoding the essential DNA repair enzyme uracil DNA glycosylase, was expressed in two permanent cell lines, the rabbit kidney cell line RK13 and the human fibroblast cell line 293. The temperature-sensitive vaccinia virus mutant ts4149, which maps within D4R, was able to grow under restrictive conditions in both of these transformed cell lines. Cell clones complemented D4R function to various degrees, demonstrating complementation of an essential vaccinia virus gene by a cell line constitutively expressing the essential function. Thus, the complementing host cells allowed the rescue of a virus defective in the D4R gene, demonstrating that this system may be used for the propagation of defective cytoplasmic DNA viruses. The defective virus grew to high yields only in the engineered cell lines. The data support the hypothesis that early gene products, such as uracil DNA glycosylase, supplied in trans can fully complement essential viral functions.

Detection of adeno-associated virus type 2 in sorted human bone marrow progenitor cells

Wild-type adeno-associated virus (wtAAV) is a helper-dependent human parvovirus which has the ability to integrate into the genome of a wide variety of human cells, including those of the hemopoietic lineages. Recombinant adeno-associated virus (rAAV) is becoming a good candidate for virally mediated gene therapy. rAAV is likely to be a safe vector in clinical gene transfer, as it has never been associated with any disease despite previous studies showing that up to 70% of adults are seropositive for wtAAV. Seroconversion appears to occur early in life. wtAAV is an upper respiratory tract virus that is gut secreted, but little is known about the integration of latent wtAAV in hemopoietic lineages. Unlike retroviruses, which have been the most common vehicles for gene transfer to date, wtAAV appears to have a preferred integration site in the target cell which has been termed AAVS1. Several studies have shown that wtAAV can only integrate into only one of the pair of chromosome 19 in a cell. This may have implications for the use of rAAV in gene transfer because patients with latent virus would be refractory to further infection with rAAV. We used a polymerase chain reaction (PCR) assay to detect the presence of wtAAV in the bone marrow samples from 106 patients who presented at our institution. We were able to detect the presence of integrated virus in 18 whole marrow samples. Subsequently CD34+ and CD3+ cell subsets were sorted from the cryopreserved marrow of three PCR-positive individuals to assess integration of virus in these cell lineages. In all three samples tested, we were unable to detect wtAAV virus in the CD34+ hematological precursor cells, but a detectable level of integrated viral DNA was demonstrated in the CD3+ cell fraction. Our findings therefore suggest that CD34+ cells might remain a good target for rAAV-mediated gene transfer despite previous wtAAV infection.

Improved adenovirus packaging cell lines to support the growth of replication-defective gene-delivery vectors

Adenovirus (Ad) vectors have been extensively used to deliver recombinant genes to a great variety of cell types in vitro and in vivo. Ad-based vectors are available that replace the Ad early region 1 (E1) with recombinant foreign genes. The resultant E1-deleted vectors can then be propagated on 293 cells, a human embryonal kidney cell line that constitutively expresses the E1 genes. Unfortunately, infection of cells and tissues in vivo results in low-level expression of Ad early and late proteins (despite the absence of E1 activity) resulting in immune recognition of virally infected cells. The infected cells are subsequently eliminated, resulting in only a transient expression of foreign genes in vivo. We hypothesize that a second-generation Ad vector with a deletion of viral genes necessary for Ad genome replication should block viral DNA replication and decrease viral protein production, resulting in a diminished immune response and extended duration of foreign gene expression in vivo. As a first step toward the generation of such a modified vector, we report the construction of cell lines that not only express the E1 genes but also constitutively express the Ad serotype 2 140-kDa DNA polymerase protein, one of three virally encoded proteins essential for Ad genome replication. The Ad polymerase-expressing cell lines support the replication and growth of H5ts36, an Ad with a temperature-sensitive mutation of the Ad polymerase protein. These packaging cell lines can be used to prepare Ad vectors deleted for the E1 and polymerase functions, which should facilitate development of viral vectors for gene therapy of human diseases.

Production and characterization of high-titer stocks of rev-defective HIV-1

High-titer stocks of rev-defective HIV-1 virions have been produced and characterized. To produce these stocks, CEM cells transduced with a murine retroviral vector that expresses Rev were used to support the growth of an HIV-1 genome containing four inactivating mutations in the rev gene. The resulting viral stocks were of high-titer when used to infect CEM cells that expressed Rev, but had no measurable titer when used to infect parental CEM cells; replication-competent virus was not detected. As expected, these rev-defective viruses established early gene expression in parental CEM cells as demonstrated by the accumulation of viral mRNAs and Nef protein. The expression of late genes was impaired as demonstrated by the inability of these viruses to direct the accumulation of singly spliced mRNA or p24 antigen. These defective viruses did not interfere with replication of the wild-type virus, although they were produced at low levels during coinfection experiments. This experimental system may be useful for the study of early events during the HIV-1 life cycle. In addition, this system represents a simple method of gene transfer into CD4-positive cells using replication-defective but transcriptionally active HIV-1.

Production of defective virus by terminally differentiated myotubes infected with Rous sarcoma virus

The generally accepted concept that the replication of Rous sarcoma virus (RSV) is dependent on host cell DNA synthesis was reexamined. As the host we used terminally differentiated myotubes (MT), in which no cellular DNA synthesis is observed. As an extension of our previous study which indicated that RSV-infected MT produce various virus components, we examined viral particles produced by infected MT. Electron microscopy showed presence of viral particles released from infected MT. Immunoprecipitation analysis revealed that these particles contained an equal amount of the gag but a decreased amount of the env proteins as compared with the particles from infected chicken embryo fibroblasts (CEF). Consequently, viral particles from infected MT had an infectivity only 6% of that of particles from infected CEF cells. In a parallel experiment, we microinjected molecularly cloned RSV DNA into MT. In contrast to the infection mediated by viral particles, both MT and CEF cells produced the same amount of infectious particles when microinjected with viral DNA. We conclude that RSV replicates in the complete absence of host DNA synthesis, though infectivity of the progeny virus depends on the initial condition of the infection.

Assembly functions of vesicular stomatitis virus matrix protein are not disrupted by mutations at major sites of phosphorylation

The matrix (M) protein of vesicular stomatitis virus (VSV) plays a central role in virus assembly by binding the nucleocapsid core to the viral envelope during the budding process. A small percentage of M protein molecules are phosphorylated in vivo, but the role of phosphorylation in M protein function is unknown. Using limited proteolysis, we previously determined the sites of in vivo phosphorylation for VSV M protein to be Thr 31 (and possibly Ser 32) and a site N-terminal to position 19 (Ser 2, Ser 3, or Ser 17) (P. E. Kaptur et al., J. Virol. 66, 5384-5392, 1992). M protein mutants were constructed using site-directed mutagenesis by substituting Ala for Ser or Thr at these sites in the M gene of the San Juan strain of VSV. One mutant had substitutions at the major in vivo phosphorylation site(s) at positions 31 and 32 (M31.32) while two others had additional substitutions at positions 2 and 3 (M2.3.31.32) or at position 17 (M17.31.32). Mutant M proteins were expressed in BHK cells using the vaccinia/T7 system, radiolabeled with 32Pi, and then analyzed for 32P content by PAGE and autoradiography. The data show that the site of phosphorylation near the N-terminus is at Ser 2 or 3 and not Ser 17. Further, Ser 38 was not phosphorylated. Mutation of the major phosphorylation site enhanced phosphorylation at alternative sites in the M protein C-terminal to amino acid 43 and at Ser residues 2 and 3. Mutant M proteins were tested for their ability to complement growth of the temperature-sensitive M protein mutant virus tsO23 at the nonpermissive temperature. Mutant M2.3.31.32 was further tested for its ability to assemble into VSV-defective interfering (DI) particles, using a replication system in which the DI genome and all five VSV proteins were expressed from plasmid DNA. Assembly of tsO23 virions or DI particles in the presence of mutant M proteins was similar to that observed for wild-type M proteins. These data indicate that phosphorylation of M protein at the major in vivo sites is not essential for virus assembly.

Tat-independent replication of human immunodeficiency viruses

The replication of human immunodeficiency retroviruses involves a complex series of events that is regulated at both transcriptional and posttranscriptional levels. The tat gene product is a potent trans-activator of viral transcription and therefore an attractive target for the development of antiviral drugs. Tat-defective HIV-1 proviral DNA clones have been shown previously to be replication defective. In this study, we report that tat-defective HIV-1 and HIV-2 viral DNA transfected into U937 cells can direct efficient viral replication in the presence of transcriptional stimulators such as TNF-alpha and PMA. In MT-4 cells, tat-defective HIV-1 can replicate without any stimulation. The viruses recovered from MT-4 cells remained tat defective defined by their inability to infect T cell lines (e.g., Molt 4/8) although replication could be rescued with cytokines. Limited replication was observed in primary mononuclear cells. Furthermore, we showed that Ro 24-7429, a potent tat antagonist and antiviral compound, failed to suppress HIV-1 replication in TNF-alpha-stimulated T cells. These results have important implications for targeting tat as a therapeutic strategy for AIDS.

Genetic control of bacterial suicide: regulation of the induction of PBSX in Bacillus subtilis

PBSX is a phage-like bacteriocin (phibacin) of Bacillus subtilis 168. Bacteria carrying the PBSX genome are induced by DNA-damaging agents to lyse and produce PBSX particles. The particles cannot propagate the PBSX genome. The particles produced by this suicidal response kill strains nonlysogenic for PBSX. A 5.2-kb region which controls the induction of PBSX has been sequenced. The genes identified include the previously identified repressor gene xre and a positive control factor gene, pcf. Pcf is similar to known sigma factors and acts at the late promoter PL, which has been located distal to pcf. The first two genes expressed from the late promoter show homology to genes encoding the subunits of phage terminases.

Cycles, chaos, and evolution in virus cultures: a model of defective interfering particles

Defective interfering particles (DIP) are spontaneous deletion mutants of viruses that replicate at the expense of the parent virus. DIP have complex effects on the growth of viruses in vitro, including the establishment of persistent infection, cyclical variation in virus titer, eradication of replicating virus, and rapid evolution of the virus. We show here that a simple mathematical model, based only on experimental observations, can explain all of the major effects of DIP on the population dynamics of virus growth. The variation in virus titer caused by DIP has many features that are characteristic of deterministic chaos: it follows that the quantitative effects of DIP are intrinsically unpredictable beyond a short time. We conclude (i) that other factors, such as temperature-sensitive virus mutants or interferons, need not be invoked to explain the complex effects of DIP; and (ii) that dominantly interfering viruses should only be used with great caution for therapeutic purposes, since their effects are, in principle, unpredictable.

Tacaribe virus gene expression in cytopathic and non-cytopathic infections

Tacaribe virus (TV) replication was compared in Vero cells infected under conditions leading either to cell death (c.p.e.(+) infection) or to the establishment of persistence (c.p.e.(-) infection). To this end, two virus preparations were employed: one containing a ratio of standard (plaque-forming) viruses to interfering particles (IP) that would induce a distinct lytic response in Vero cells infected at multiplicities giving synchronous infection and another virus stock enriched in IP that would block the cell-killing potential of the cytolytic virus stock. The following results were obtained: (1) No qualitative differences were observed in the species of intracellular viral RNAs in the lytic infection in comparison with infections leading to persistence or during the early stages of persistence. (2) Levels of viral RNAs were severely reduced when the cells were infected with IP in addition to standard viruses, the RNA accumulation being inversely proportional to the ratio of IP to standard viruses used in the infections. (3) Accumulation of the three measurable mRNAs (those corresponding to the glycoprotein precursor [GPC], to the nucleoprotein [N], and to the p11Z protein) ended earlier in the c.p.e.(-) infections (around 18 hr p.i.) than in the c.p.e.(+) infection (45-68 hr p.i.). (4) The rates of synthesis of the GPC, N, and p11Z proteins were largely determined in both the c.p.e.(+) and c.p.e.(-) infections by the amounts of their corresponding mRNAs. (5) The kinetics of accumulation of the S genomes and also the ratios of the S genome to S antigenome were similar in the different infections (accumulation ending at 45-68 hr p.i.). (6) L genome accumulation proceeded for longer time (until 92 hr p.i.) in the c.p.e.(+) infection than in the c.p.e.(-) infections. In the latter accumulation ended at around 45 hr p.i. Until this time ratios of L genome to L antigenome were similar in the different infections. It is concluded that IP affect virus mRNA synthesis early after infection reducing in this way the rate of viral protein synthesis. Low levels of viral proteins might then limit virus replication. In addition, the results support the idea that in TV infections transcription and replication are independently regulated. The implications of these results with regards to the nature and mode of action of TV IP are discussed.

The Sendai virus matrix protein appears to be recruited in the cytoplasm by the viral nucleocapsid to function in viral assembly and budding

The matrix (M) protein is viewed as the regulator of paramyxovirus particle assembly and budding. Accordingly it was observed to be mutated, and/or decreased in amount, in cases where virus particle production was significantly reduced. Here, a non-productive [non-defective and defective interfering (DI)] Sendai virus infection of COS cells is presented where virus particle production is abolished in the presence of a normal amount of intracellular M protein. In this infection the haemagglutinin-neuraminidase envelope glycoprotein is shown to be dispensable for virion production, and the fusion (F0) envelope glycoprotein behaves as in a productive infection. The M protein is shown to accumulate in perinuclear patches within the cytoplasm. In contrast, localization in the plasma membrane is observed in productive infections. However in both productive and non-productive infections a significant fraction of M protein is found in association with cellular membranes. The M protein-membrane association is shown to take place in the absence of any other viral component, and the M protein-membrane complex exhibits properties similar to those observed for the integral membrane protein F0. However these properties are distinct from those of the phosphoprotein, which is thought to associate with membranes in a non-specific manner. Concomitant with the cytoplasmic accumulation of M protein and the reduction of virus particle production in this non-productive infection, DI nucleocapsids are shown not to associate with cellular membrane fractions. This is a property which coincides with their poor envelopment in virus particles. Taken together, these data indicate the need for M protein to be recruited at the perinuclear membranes by the nucleocapsids to participate in viral assembly and budding. This view is consistent with a process of viral assembly taking place on internal cytoplasmic membranes rather than at the plasma membrane.

Defective endogenous proviruses are expressed in feline lymphoid cells: evidence for a role in natural resistance to subgroup B feline leukemia viruses

Endogenous feline leukemia virus (FeLV)-related sequences (enFeLV) are a family of proviral elements found in domestic cats and their close relatives. These elements can recombine with exogenous, infectious FeLVs of subgroup A (FeLV-A), giving rise to host range variants of FeLV-B. We found that a subset of defective enFeLV proviruses is highly expressed in lymphoma cell lines and in a variety of primary tissues, including lymphoid tissues from healthy specific-pathogen-free cats. At least two RNA species were detected, a 4.5-kb RNA containing gag, env, and long terminal repeat sequences and a 2-kb RNA containing env and long terminal repeat sequences. Cloning of enFeLV cDNA from two FeLV-free lymphoma cell lines (3201 and MCC) revealed a long open reading frame (ORF) encoding a truncated env gene product corresponding to the N-terminal portion of gp70env. Interestingly, all of three natural FeLV-B isolates include 3' env sequences which are missing from the highly transcribed subset and hence must be derived from other enFeLV elements. The enFeLV env ORF cDNA clones were closely similar to a previously characterized enFeLV provirus, CFE-16, but were polymorphic at a site corresponding to an exogenous FeLV neutralization epitope. Site-specific antiserum raised to a C-terminal 30-amino-acid peptide of the enFeLV env ORF detected an intracellular product of 35 kDa which was also shed from cells in stable form. Expression of the 35-kDa protein correlated with enFeLV RNA levels and was negatively correlated with susceptibility to infection with FeLV-B. Cell culture supernatant containing the 35-kDa protein specifically blocked infection of permissive fibroblast cells with FeLV-B isolates. We suggest that the truncated env protein mediates resistance by receptor blockade and that this form of enFeLV expression mediates the natural resistance of cats to infection with FeLV-B in the absence of FeLV-A.

The presence of tat protein or tumor necrosis factor alpha is critical for herpes simplex virus type 1-induced expression of human immunodeficiency virus type 1

Tat-independent transcription of human immunodeficiency virus type 1 (HIV-1) plays an important role in virus life cycle before biologically significant levels of Tat protein have been accumulated. Using a latently infected T-cell line containing an integrated Tat-defective HIV-1 provirus, we examined whether factors known to up-regulate the HIV-1 expression in vitro can replace the requirement for a functional Tat protein and induce the expression of the Tat-defective HIV-1 provirus. Both tumor necrosis factor alpha (TNF-alpha) and herpes simplex virus type 1 (HSV-1) infection stimulated transcription of the Tat-defective HIV-1 provirus to comparable levels, but in HSV-1-infected cells, the cytoplasmic HIV-1 transcripts were not efficiently translated in the absence of Tat protein and were excluded from the large polysomes. However, HSV-1 infection did not affect the distribution of cellular gamma-actin RNA or 28S RNA in the polysomal fractions. The translational block of HIV-1 RNA was not mediated by the virion-associated host cell shutoff protein (vhs); dissociation of HIV-1 transcripts from the polysomes and inefficient translation was also observed in cells infected with the vhs-defective mutant of HSV-1 (vhs-1). Overexpression of Rev protein did not rescue the synthesis of HIV-1 proteins in these cells; however, the observed inhibition of HIV-1 RNA translation was efficiently overcome in the presence of Tat protein or TNF-alpha. These findings suggest that, in contrast to TNF-alpha, HSV-1 infection is not able to induce a full cycle of HIV-1 replication and that cytokines and Tat have a critical role in the activation of HIV-1 provirus by HSV-1.

Preferential replication of defective turnip yellow mosaic virus RNAs that express the 150-kDa protein in cis

The turnip yellow mosaic virus genome encodes two proteins (the 150-kDa and 70-kDa proteins) that are proteolytically released from a single precursor and which are essential for RNA replication. Genomes with mutations in either of these coding regions were defective for independent replication in turnip protoplasts. The replication in trans of genomes with mutations in each region was studied by coinoculation with either a helper genome that carries a deletion in the coat protein gene, or with a second defective RNA that carries a mutation in the region encoding the other essential protein. Inefficient trans-replication of the defective RNAs was observed in most cases. In contrast, a defective RNA with a large deletion in the 70-kDa protein coding region could be replicated efficiently in trans, demonstrating that the cis-preference of replication can be overcome in some cases. Defective RNAs encoding wild type 150-kDa protein and defective 70-kDa protein were more efficiently replicated in trans than those encoding defective 150-kDa protein and wild type 70-kDa protein. The results suggest a model in which the 150-kDa and 70-kDa proteins form a relatively stable complex in cis on the viral RNA template.

Complementation of a poliovirus defective genome by a recombinant vaccinia virus which provides poliovirus P1 capsid precursor in trans

Defective interfering (DI) RNA genomes of poliovirus which contain in-frame deletions in the P1 capsid protein-encoding region have been described. DI genomes are capable of replication and can be encapsidated by capsid proteins provided in trans from wild-type poliovirus. In this report, we demonstrate that a previously described poliovirus DI genome (K. Hagino-Yamagishi and A. Nomoto, J. Virol. 63:5386-5392, 1989) can be complemented by a recombinant vaccinia virus, VVP1 (D. C. Ansardi, D. C. Porter, and C. D. Morrow, J. Virol. 65:2088-2092, 1991), which expresses the poliovirus capsid precursor polyprotein, P1. Stocks of defective polioviruses were generated by transfecting in vitro-transcribed defective genome RNA derived from plasmid pSM1(T7)1 into HeLa cells infected with VVP1 and were maintained by serial passage in the presence of VVP1. Encapsidation of the defective poliovirus genome was demonstrated by characterizing poliovirus-specific protein expression in cells infected with preparations of defective poliovirus and by Northern (RNA) blot analysis of poliovirus-specific RNA incorporated into defective poliovirus particles. Cells infected with preparations of defective poliovirus expressed poliovirus protein 3CD but did not express capsid proteins derived from a full-length P1 precursor. Poliovirus-specific RNA encapsidated in viral particles generated in cells coinfected with VVP1 and defective poliovirus migrated slightly faster on formaldehyde-agarose gels than wild-type poliovirus RNA, demonstrating maintenance of the genomic deletion. By metabolic radiolabeling with [35S]methionine-cysteine, the defective poliovirus particles were shown to contain appropriate mature-virion proteins. This is the first report of the generation of a pure population of defective polioviruses free of contaminating wild-type poliovirus. We demonstrate the use of this recombinant vaccinia virus-defective poliovirus genome complementation system for studying the effects of a defined mutation in the P1 capsid precursor on virus assembly. Following removal of residual VVP1 from defective poliovirus preparations, processing and assembly of poliovirus capsid proteins derived from a nonmyristylated P1 precursor expressed by a recombinant vaccinia virus, VVP1 myr- (D. C. Ansardi, D. C. Porter, and C. D. Morrow, J. Virol. 66:4556-4563, 1992), in cells coinfected with defective poliovirus were analyzed. Capsid proteins generated from nonmyristylated P1 did not assemble detectable levels of mature virions but did assemble, at low levels, into empty capsids.(ABSTRACT TRUNCATED AT 400 WORDS)

Encapsidation of a recombinant LuIII parvovirus genome by H1 virus and the fibrotropic or lymphotropic strains of minute virus of mice

We previously constructed a recombinant LuIII parvovirus genome lacking viral coding sequences and used it to generate luciferase-transducing virions, by cotransfection of cells with a helper plasmid expressing LuIII viral proteins. Here, we describe similar cotransfections using alternative, replication-defective helpers encoding the non-structural and capsid proteins of parvovirus H1, or of either the fibrotropic or lymphotropic parvovirus strain of minute virus of mice [MVM(p) or MVM(i)]. Each cotransfection generated transducing virus which directed luciferase expression after infection of HeLa cells. The transducing activity of virus produced using either LuIII or H1 helper plasmids could be specifically neutralized by antiserum raised against the corresponding infectious virus. When the recombinant LuIII parvovirus was pseudotyped with MVM(p) or MVM(i), the resulting virions efficiently expressed luciferase after infection in human or murine cells known to be permissive for both MVM strains. The MVM(p) pseudotyped virus also expressed this reporter efficiently when infected into the murine A9 fibroblast line. In contrast, the recombinant virus generated with an MVM(i) helper gave luciferase expression that was barely detectable after infection of A9 cells which are highly restrictive for MVM(i) productive infection. These results support the notion that the allotropic determinant of these MVM strains functions through their capsid proteins. Pseudotyping of recombinant parvovirus genomes should be useful in controlling their host range as vectors, and in studying mechanisms influencing the permissiveness of parvovirus infections.

Cell-dependent requirement of human immunodeficiency virus type 1 Vif protein for maturation of virus particles

A highly sensitive single-round infection assay using a bacterial chloramphenicol acetyltransferase was developed to analyze an early stage of human immunodeficiency virus type 1 replication. By a combination of transfection and single-round infection assay, a virus with a vif mutation, depending on host cells from which the virus was derived, was demonstrated to be defective at the early phase of infection cycle. Analysis of viral proteins synthesized in cells indicated that incorporation of the Env surface protein into virions of the vif mutant, again in a cell-dependent way, was greatly restricted. Taken together, it is concluded that the Vif protein acts through modulation of the Env protein in the virions, directly or indirectly, to enhance viral infectivity in a certain cell type.

Dissociation between lymphoproliferative responses and virus replication in mice with different sensitivities to retrovirus-induced immunodeficiency

Murine AIDS (MAIDS) is induced by a replication-defective virus (BM5d). In susceptible mice (C57BL/6J), inoculation with LP-BM5 murine leukemia virus, which consists of the BM5d virus and replication-competent B-tropic ecotropic (BM5e) and milk cell focus-inducing (BM5-MCF) helper viruses results in the polyclonal proliferation of T and B cells, immunodeficiency, and the expansion of B cells containing the BM5d provirus followed by the development of B-cell lymphomas. Several strains of mice that are resistant to LP-BM5-induced murine AIDS have been identified, and major histocompatibility complex genes as well as non-major histocompatibility complex genes were shown to play a role in this resistance. In the present study, we have examined and compared the replication of the BM5d and BM5e viruses after inoculation of LP-BM5 into sensitive (C57BL/6J) and resistant (C57BL/KSJ) mice. Using a specific polymerase chain reaction, we could detect the BM5d and BM5e proviruses as early as 1 week postinfection in the sensitive mice, and the levels of both viruses increased significantly with the progression of the disease. In contrast, in the resistant C57BL/KSJ mice, replication of BM5d and BM5e was restricted and no BM5d and only very low levels of the BM5e provirus could be detected either at early or late times postinoculation with the LP-BM5 virus mixture. Inoculation with LP-BM5 did not lead to the production of antibodies that could recognize the BM5d-encoded Pr60gag in either the sensitive or resistant mice; however, production of antibodies recognizing the env-related proteins of the helper virus was detected in the resistant but not in the sensitive mice at late times postinfection. Interestingly, inoculation with LP-BM5 increased polyclonal stimulation of spleen cells and decreased mitogen stimulation in both strains of mice. This stimulation of splenocytes persisted in the sensitive mice but decreased after a few weeks in the resistant mice. These results show an early block in BM5d and BM5e replication in the resistant C57BL/KSJ mice and indicate that resistance is a consequence of the inhibition of an onset of the BM5d virus infection and its expansion. However, initial responses to virus infection such as proliferation of spleen cells and response to mitogen are similar in both strains of mice and are therefore not necessarily related to the development of the disease.

Defective viral particles in caprine arthritis encephalitis virus infection

Attempts to isolate full-length unintegrated circular forms of the caprine arthritis encephalitis virus (CAEV) genome yielded only a large number of molecules with deletions. The 3' borders of most of these deletions were near the U3 region of the long terminal repeat whereas the 5' edges were found at various upstream sites within pol or env. With one exception, gag sequences were always present. Analysis of molecular clones derived from integrated proviral CAEV genomes from the same infected cells showed a similar spectrum of deletions. The presence of transcriptionally active elements within the U3 domain of the defective genomes, as well as cis-acting elements within the leader sequences known to be required for efficient encapsidation of viral RNA, suggested that the defective viral DNA genomes could be transcribed into defective RNA molecules which could then be packaged into virions. Isopycnic density gradient centrifugation of supernatants of infected cell cultures indicated the presence of particles with densities less than that expected for intact virions (1.16 g/cc). Northern analysis revealed the presence of smaller viral-specific RNAs that lacked env sequences. These data, along with the structures of the molecular clones, suggest that CAEV stocks contain particles with defective genomes. The role of these particles in influencing the course of virus infection remains to be determined.

Human immunodeficiency virus type 1 (HIV-1) superinfection of a cell clone converting it from production of defective to infectious HIV-1 is mediated predominantly by CD4 regions other than the major binding site for HIV-1 glycoproteins

A cell clone, L-2, which produces non-infectious doughnut-shaped human immunodeficiency virus type 1 (HIV-1) particles, was permissive for HIV-1 superinfection, which resulted in the production of infectious particles. The superinfection showed slow kinetics compared with primary HIV-1 infection of M10 cells, the parent of the L-2 cell clone. Inhibition studies on the superinfection of L-2 cells using several CD4-related reagents showed that the CD4 molecule was an essential component of the receptor for superinfection. Strong inhibitory effects were obtained using CD4 peptides such as CD4(68-130), which includes a portion homologous to the immunoglobulin third complementarity-determining region (CDR3), as well as recombinant soluble CD4. In contrast, a CD4(45-60) peptide, which includes most of the CDR2-related region, was not effective, although the Leu-3a monoclonal antibody (MAb), which recognizes a site near the CDR2-related region, did slightly, but significantly, delay the superinfection kinetics. Comparative flow cytometry of L-2 and M10 cells revealed that the cell surface of L-2 cells despite expressing HIV-1 env protein, reacted slightly with OKT4 or anti-CD4(68-130) MAb, but not with Leu-3a or OKT4A MAb. In contrast, no reaction was detected with any of these anti-CD4 MAbs on the surface of another HIV-1 superinfection-resistant cell clone, MOLT-#8IIIB-14, which expresses HIV-1 env proteins but does not produce infectious HIV-1 particles. These results strongly suggest that expression of the CD4 major receptor site for primary HIV-1 infection is preferentially decreased on the surface of L-2 cells, but that the OKT4 epitope and the nearby region corresponding to immunoglobulin CDR3 remain exposed on the cell surface. Consequently, the CD4 CDR3-related region could play a major role as the receptor for the superinfection reported here.

Comparison of two defective hepatitis A virus strains adapted to cell cultures

The replication of two defective hepatitis A virus strains in cell culture was examined. The w.t. HAS-15 strain growing in FRhK-4 cells produced infectious icosahedral virions 27 nm in size as well as round shaped particles with lipids attached to their surface. The morphogenesis of HAV was membrane-dependent and the detected particles were in various degree of maturation. The MBB 11/5 strain growing in PLC/PRF/5 cells produced mainly noninfectious empty procapsids without RNA genome. The translation of viral proteins was uninhibited in both strains. The reason for restricted replication competence of both strains seemed to be different. In HAS-15, highly efficient encapsidation of the progeny RNA positive-strand lowered the formation of replicative intermediate forms. In MBB 11/5, nearly exclusive empty procapsid production gave evidence for the failure of the VPg primer protein attachment to viral RNA. Changes in the efficacy of viral genome replication were a result of the adaptation of HAV to propagation in vitro.

Biological characterization of noninfectious HIV-1 particles lacking the envelope protein

To understand the role of the HIV-1 envelope protein in the assembly of virus, we constructed a proviral clone of HIV-1 where the methionine initiator codon of the env gene was substituted with a translational stop codon. Upon DNA transfection into permissive cells in culture, this clone produces virus-like particles similar in size to parental virus but are noninfectious in human T-cells, promonocytic cells, and primary macrophages. This mutant readily recombines with a deletion mutant provirus lacking the entire gag-pol region producing a recombinant virus that is infectious. Substitution of the same initiator methionine codon with valine results in a leaky missense mutant provirus capable of a low level of Env protein synthesis that leads to a productive infection. Thus, the prototype initiation codon AUG is dispensable for virus infectivity. Further, the expression of the envelope protein is not a prerequisite for the assembly of the virus particles in the HIV-1 system. These noninfectious envelope-less particles revert readily to wild-type phenotype upon cotransfection with Env-producing plasmid DNAs.

An L (polymerase)-deficient rabies virus defective interfering particle RNA is replicated and transcribed by heterologous helper virus L proteins

A rabies virus-derived defective interfering particle (DI) was isolated and characterized. The DI genome contained an internal deletion of 6.4 kb spanning the 3' moiety of the pseudogene region (psi) and most of the L gene. DI-specific monocistronic N, NS, and M mRNAs as well as a G/L fusion mRNA were transcribed in cells coinfected with DI and helper virus. In addition, polycistronic DI RNAs and standard virus RNAs with internal A stretches and intergenic regions were found. Superinfection experiments showed that heterologous rabies-related viruses (Lyssavirus serotypes 2, 3, and 4) can complement the L deficiency of the DI genome. The heterologous polymerase proteins recognize correctly the replicational and transcriptional signal sequences of the Lyssavirus serotype 1-derived DI.

Rescue of a Sendai virus DI genome by other parainfluenza viruses: implications for genome replication

Using a defective interfering Sendai virus stock (DIH4) freed of nondefective helper virus, we found that the closely related parainfluenza viruses 1 and 3 could substitute for the Sendai virus helper in replicating DIH4, creating chimeric nucleocapsids. The morbillivirus measles and the rhabdovirus VSV could not substitute. When DIH4 is incubated intracellularly for 5 days in the absence of help, the ability of PIV3 to rescue DIH4 at this time depended on fresh Sendai virus polymerase. The PIV3 polymerase apparently can only copy the chimeric template, but not that wrapped in the homologous Sendai NP protein. These results suggest that the cis-acting RNA sequences important for genome replication, e.g., the promoter and the encapsidation site, have been conserved among these viruses, but that the interactions between the polymerase and the template protein NP are unique for each virus.

Characterization and comparison of avian and murine helper cell lines for production of replication-defective retroviruses for avian transformation

Several approaches were taken to identify improved helper cell lines for the production of replication-defective avian retroviral vectors for avian transformation. Both QT6 and D17 cells were engineered to become helper cell lines for the production of reticuloendotheliosis virus vectors. The results showed that the majority of lines from the D17, QT6, and D17C3 cells produced titers in the 10(2) to 10(3) cfu/mL range, with one QT6 line producing 10(5) cfu/mL. This high producer line was relatively free of helper virus when restricted to low passage. An amphotropic murine cell line produced a 6- to 10-fold higher amount of virus and had a comparable higher titer on chicken cells, suggesting possible application to avian transformation.

Defective interfering particles: effects in modulating virus growth and persistence

Defective interfering virus particles (DIP) frequently play an important part in viral persistence in vitro, and may in some instances modify a virus infection in vivo, causing attenuation or persistence of the infection. To explain certain aspects of the growth of these mutants in vitro, other factors have been invoked such as interferon, mutations in the wild-type virus or the infected cells, or other substances released by infected cells that attenuate the infection. We present here a simple model of the growth of DIP in vitro which shows that (a) the observed population dynamics of DIP can readily be explained without invoking such extrinsic factors; (b) the initial multiplicity of infection of DIP is the principal determinant of the outcome of infection in both single- and repeated-passage cultures; and (c) in a long-term culture in vitro, the criterion used to decide the time of virus passage directly determines how long the standard virus, DIP, and cells survive. This model may be used with minor modifications to predict the behavior in vitro of other mutant viruses with a dominantly interfering phenotype.

Release of pseudorabies virus from infected cells is controlled by several viral functions and is modulated by cellular components

The role of the nonessential glycoproteins gI, gp63, and gIII in the release of pseudorabies virus from different cell lines was investigated. We show that these glycoproteins may have a beneficial or deleterious effect on virus release depending on the type of cell in which the virus is grown. Inactivation of the genes encoding either gI, gp63, or gIII has no detectable effect on virus release from rabbit kidney cells. Inactivation of gI or gp63 strongly promotes virus release from chicken embryo fibroblasts, whereas inactivation of gIII reduces virus release from these cells. A defect in both gI and gIII or in both gp63 and gIII diminishes virus release from rabbit kidney cells but improves release from chicken embryo fibroblasts. We demonstrate that all three nonessential glycoproteins contribute to one specific aspect of viral growth, namely, virus release, and that they affect virus release in conjunction with each other. Furthermore, our results show that the manifestation of the role of each of these viral functions in virus growth may differ in different cell types, i.e., that release is affected by these viral functions in conjunction with some unknown cellular function.

Expression of v-src arrests murine glial cell differentiation

A replication-defective retroviral vector carrying the v-src oncogene and the gene for neomycin resistance was used to infect neurone- and fibroblast-depleted embryonic mouse brain cells in vitro. Cells resistant to the antibiotic G418 were obtained and continually passaged. Several cell lines were isolated which express high levels of v-src mRNA and v-src tyrosine kinase activity. The antigenic marker profile of either the pooled cells from an individual infection or sublines isolated from individual foci showed the cells to be immature glia: most cells expressed vimentin, A2B5 antigen and/or J1/tenascin glycoproteins, but not fibronectin. Sublines expressed different antigen profiles suggesting that the immortalised cells were derived from glial cells of different phenotypes. The cell lines expressed the 120 and 140 but not the 180 kd components of N-CAM as well as voltage-activated potassium channels, typical for glial cells. 01 antigen-positive oligodendrocytes were never observed in the lines or sublines after long term passage (over 1 year), but some cells expressed glial fibrillary acidic protein, a marker for mature astrocytes. Thus, expression of v-src in murine glial cells appears to arrest their development and prevent their differentiation.

Delayed formation of defective interfering particles in vesicular stomatitis virus-infected cells: kinetic studies of viral protein and RNA synthesis during autointerference

The time course of defective interfering (DI) particle and B particle release from vesicular stomatitis virus-infected BHK-21 cells was studied at different multiplicities of defective and infective particles. Particle release was progressively delayed in cells infected with an increasing DI-to-B particle ratio. The delayed particle release during interference was found to be connected with a reduced but prolonged synthesis of viral proteins, a slower accumulation of viral proteins, and a delayed shutoff of cellular protein synthesis. The relative synthesis of M and G proteins was reduced during interference, whereas the relative synthesis of N and NS proteins was increased. On the level of genomic RNA replication, we found that DI RNA was replicated more slowly during interference than the standard genomic RNA was during acute infection. The ratio of DI particles to B particles which were released increased throughout the infectious cycle. At a given time in the infectious cycle, this ratio was independent of the multiplicity of infecting DI and B particles. On the basis of the kinetic studies, we argue that cells infected with higher amounts of DI particles compared with B particles synthesize a higher DI-to-B particle ratio and release these progeny particles later than cells infected with a low DI-to-B particle ratio.

Defective interfering virus associated with A/Chicken/Pennsylvania/83 influenza virus

The A/Chicken/Pennsylvania/1/83 influenza virus, isolated from a respiratory infection of chickens, is an avirulent H5N2 virus containing subgenomic RNAs (W.J. Bean, Y. Kawaoka, J.M. Wood, J.E. Pearson, and R.G. Webster, J. Virol. 54:151-160, 1985). We show here that defective interfering particles are present in this virus population. The virus had a low ratio of plaque-forming to hemagglutinating units and produced interference with standard virus multiplication in infectious center reduction assays. Subgenomic RNAs were identified as internally deleted polymerase RNAs. We have confirmed that this virus protects chickens from lethal H5N2 influenza virus infection. This protective effect appeared to be due to the inhibition of virulent virus multiplication. Additionally, subgenomic RNAs derived from polymerase RNAs were detected in 5 of 18 RNA preparations from animal influenza virus isolates. Therefore, defective interfering particles are sometimes produced in natural influenza virus infections, not just under laboratory conditions. These particles may be capable of suppressing the pathogenic effect of virulent virus infections in nature.

Isolation of a novel type of interfering influenza B virus defective in the function of M gene

A novel type of interfering influenza B virus which is defective in the function of M gene has been reported. Clone 301, a B type virus clone obtained by successive back-crosses of A/Aichi/2/68 (H 3 N 2) with B/Yamagata/1/73, grew normally in MDCK cells when inoculated at a low multiplicity, but was easily converted to a hemagglutinating but non-infectious form by one cycle of high multiplicity infection. Within MDCK cells infected with infectious clone 301 at a high multiplicity, synthesis of M protein was greatly reduced. The virus particle produced by a high multiplicity infection was devoid of RNA segment 7 (M gene), contained less amount of M protein compared with the standard virus, and interfered with the replication of wild type B/Yamagata, again accompanied by a selective suppression of M protein synthesis within the co-infected cells.

Temperature-sensitive mutants of fowl plague virus defective in the intracellular transport of the hemagglutinin

Nine mutants of fowl plague virus with temperature-sensitive defects in the biosynthesis of the hemagglutinin have been characterized by analyzing the processing and the intracellular location of this glycoprotein in MDCK and chick embryo cells. It was found that with all of these mutants the transport of the hemagglutinin to the cell surface was impeded at the non-permissive temperature. There were differences, however, in the site of the block. With mutants tsl, ts227, ts478 and ts658 the precursor HA was not cleaved and the oligosaccharide side chains remained sensitive to endoglucosaminidase H. When the hemagglutinin was analyzed in permeabilized cells by immunofluorescence, usually only cytoplasmic labeling was seen. Immunofluorescence of non-permeabilized cells and hemadsorption revealed that the hemagglutinin did not reach the cell surface. In contrast, the hemagglutinin of mutants ts79, ts482, ts532, ts546 and ts651 was cleaved and oligosaccharides were processed to the endoglucosaminidase H-resistant form at non-permissive temperature. In permeabilized cells, the cytoplasm and juxtanuclear regions typical for the Golgi apparatus were labeled by immunofluorescence. Except for ts482, ts532 and ts546 which were leaky, hemagglutinin could not be detected at the cell surface. These observations indicate that, with the first group of mutants, hemagglutinin transport is usually arrested already in the rough endoplasmic reticulum, whereas with the second group it is inhibited at a late stage between the Golgi apparatus and the plasma membrane.

Very rapid generation/amplification of defective interfering particles by vesicular stomatitis virus variants isolated from persistent infection

Multiply cloned variants of vesicular stomatitis virus (VSV) were found to generate/amplify defective interfering (DI) particles at a rate greatly exceeding the rates normally observed for wild-type VSV (or for other mutants of VSV). A single undiluted passage of the first clonal pool of this variant virus produced concentrated visible bands of DI particles on sucrose gradients whereas wild-type and other mutant strains of VSV required from three to six or more serial undiluted passages. Since DI particle amplification by wild-type VSV at each undiluted passage can exceed 10,000-fold enrichment, these variant virus clones were generating/amplifying DI particles many millions of times more rapidly than were wild-type and other mutant strains of VSV. This rate of generation/amplification is so high that it was not feasible to obtain accurate estimates of the rates of generation (or amplification) of these DI particles.