- Volume 79, Issue 5, 1998

Modified virus Ankara (MVA) is a vaccinia virus (VV) strain that was attenuated by serial passage through chick embryo fibroblasts (CEFs) and contains six large genomic deletions compared with parental virus. MVA replicates well in CEFs, but poorly in most mammalian cells. Recombinant MVA is a promising human vaccine candidate due to its restricted host range, immunogenicity and avirulence in animal models, and excellent safety record as a smallpox vaccine. Here we present a further characterization of MVA and demonstrate that: (i) MVA can replicate, albeit poorly, in transformed human cell lines, but not in primary human fibroblasts although there is limited cell-to-cell spread; (ii) MVA is a potent inducer of type I interferon (IFN) from primary human cells, which may restrict virus spread in vivo; and (iii) unlike other VV strains, MVA does not express soluble receptors for IFN-gamma, IFN-alpha/beta, tumour necrosis factor and CC chemokines, but does express a soluble interleukin-1beta receptor. This provides a plausible and testable explanation for the good immunogenicity of MVA despite its poor replication in mammals. The implications of these findings for the use of MVA as a safe and immunogenic human vaccine candidate are discussed.

The African swine fever virus (ASFV) strain Malawi LIL20/1 open reading frame (ORF) j13L was expressed in vaccinia virus (VV) from a strong synthetic late promoter as either a complete ORF (vSJ1) or lacking codons 1-31 (vSJ2). Each recombinant VV produced a small plaque which rapidly reverted to a normal size upon passage. The yield of infectious virus from a single cycle infection with vSJ1 or vSJ2 was reduced 50- to 100-fold compared to wild-type (wt) and a revertant virus (vSJ5) in which the j13L ORF was removed and the VV thymidine kinase gene restored. PCR analysis of nine spontaneous large plaque revertant viruses, recovered after passage of vSJ1 in BSC-40 cells, showed that six had lost the j13L ORF and the co-inserted beta-galactosidase gene. Three viruses retained the j13L and beta-galactosidase genes, but in each case the j13L protein was not expressed due to a different single base deletion near the 5' end of the j13L coding region which introduced a stop codon a short distance downstream. The formation of intracellular mature virus (IMV) and extracellular enveloped virus was reduced 50- to 75-fold in cells infected with vSJ1 compared to wt VV and revertant vSJ5. Electron microscopy showed aberrant IMV precursor structures in vSJ1-infected cells, and immunoelectron microscopy demonstrated that these structures contained j13L protein. These results indicate that expression of the j13L protein is toxic for VV replication due to interference with VV morphogenesis prior to IMV formation.

The j5R open reading frame (ORF) of the Malawi LIL 20/1 African swine fever virus (ASFV) isolate encodes a 111 amino acid protein with a putative transmembrane domain at the N terminus. Antisera raised against the predicted C-terminal peptide were used to identify the j5R protein by Western blotting in cells infected with ASFV or with recombinant vaccinia virus expressing the j5R ORF. This showed that the j5R protein migrates with an apparent molecular mass of 23-25 kDa, depending on the virus isolate, on SDS-PAGE and is expressed late during ASFV infection. The localization in infected cells and in virions of the j5R protein, and that of a previously characterized virion protein, j13L, which also contains a putative transmembrane domain, were studied by immunofluorescence and immuno-electron microscopy. Both proteins were expressed at 8-10 h post-infection (p.i.) as small multiple perinuclear foci which coalesced to a single area indicative of the virus factory at 18 h p.i. At the ultrastructural level j5R and j13L were detected mainly on membrane-like structures within the virus factory and on virus particles, suggesting that they may be involved in particle assembly. Negative contrast immuno-electron microscopy of mature extracellular virions confirmed that they are also integral structural proteins.

The right variable region of the African swine fever virus (ASFV) genome is known to contain genes with functions involving virus virulence and host range in swine. A novel open reading frame, ORF l11L, which was absent in the non-pathogenic, cell culture-adapted European isolate BA71V, was identified in the pathogenic African isolate Malawi Lil-20/1. The location of l11L in the right variable region, together with its absence in BA71V, suggested that l11L may have a function in virus virulence and/or host range. Here, we show that the l11L gene is highly conserved among pathogenic African, European and Caribbean ASFV field isolates and that it exists either in a short form, encoding a protein of 77-78 amino acids (9.1 kDa) or in a longer form of 93-94 amino acids (11.1 kDa). The presence of two predicted membrane-spanning segments suggests that l11L is an integral membrane protein. RT-PCR analysis demonstrated that l11L mRNA is expressed late in the virus replication cycle. A recombinant l11L gene deletion mutant, deltal11L, was constructed from the ASFV isolate Malawi Lil-20/1 to examine gene function. Deletion of l11L did not affect virus replication in swine macrophage cell cultures nor virulence in domestic pigs, indicating that l11L is non-essential for growth in vitro and for virus virulence in domestic swine.

The complete DNA sequence of equine herpesvirus-4 (EHV-4) strain NS80567 was determined. The genome is 145597 bp in size and consists of a long unique region (UL, 112398 bp) flanked by a short inverted repeat (TRL/IRL, 27 bp) linked to a short unique region (Us, 12789 bp) flanked by a substantial inverted repeat (TRs/IRs, 10178 bp). EHV-4 is predicted to contain 76 different genes; three of these are present twice in TRs/IRs, giving a total of 79 genes. The closely related virus equine herpesvirus-1 (EHV-1) also possesses 76 different genes corresponding to those of EHV-4, but has a total of 80 genes because four are present twice in TRs/IRs. Interpretations of the coding capacity of the EHV-4 and EHV-1 genomes were refined by comparing the complete DNA sequences.

Glycoprotein G (gG) homologues have been found in most alphaherpesviruses although little is known about their structure or function. In this study, three species of equine herpesvirus-4 (EHV-4) gG were identified: a full-length 68 kDa virion-associated species (gGVL), a 12 kDa virion-associated species (gGVS) and a 60 kDa secreted species (gGS), detected in the medium of infected cells. gGS and gGVS appear to be proteolytic cleavage products of gGVL and correspond to the N- and C-terminal regions, respectively. It was shown that gGS and gGVL are similarly glycosylated possessing mostly N-linked complex-type carbohydrate side chains. Western blots of proteins separated under nonreducing conditions established that gGS is secreted as a 120 kDa glycoprotein while the virion-associated species, gGVL and gGVS, are present in the virion as 140 and 20 kDa proteins, respectively. As gGS and gGVL do not appear to associate stably with other viral proteins, it is most likely that each species exists as a disulphide-linked homodimer. Pulse-chase experiments indicated that gGVL is rapidly assembled as a homodimer prior to both carbohydrate side-chain maturation in the Golgi and proteolytic cleavage. Proteolytic cleavage of full-length gG occurs during or immediately after passage through the Golgi. Secreted and virion-associated species of gG were identified in the closely related virus EHV-1 and were of similar molecular masses to the corresponding EHV-4 gG species.

Glycoprotein G is a major target for the humoral immune response against herpes simplex virus (HSV) and a prototype antigen for type-specific serodiagnosis discriminating HSV-1 and HSV-2 infections. The mature part of HSV-2 glycoprotein G-2 (gG-2) contains a unique stretch suspected to mediate type specificity, and in addition a region homologous to HSV-1 glycoprotein G-1 (gG-1). Antigenic determinants of the mature gG-2 were mapped by testing the reactivity of mouse anti-gG-2 monoclonal antibodies (MAbs) and purified human anti-gG-2 antibodies with synthetic peptides coupled to cellulose membranes. The anti-gG-2 MAbs bound to four epitopes localized in a narrow cluster within a gG-2 segment delimited by amino acids (aa) 552 and 611. This cluster was located between the predicted O-glycan-rich region and the transmembrane anchor sequence. The epitopes of the human anti-gG-2 antibodies were localized within three stretches of amino acids, two of which were overlapping with those recognized by anti-gG-2 MAbs. One of these stretches, delimited by aa 552 and 574, showed reactivity to all human HSV-2 sera tested, but not to HSV-1 sera or to purified anti-gG-1 antibodies. Neither the anti-gG-2 MAbs nor the purified human anti-gG-2 antibodies were cross-reactive to gG-1 peptides or HSV-1 antigen, although most of the epitopes were localized within the part of gG-2 which was homologous to gG-1. The findings concerning HSV-2 type-specific human antibody response to a defined stretch within gG-2 may be of importance for the further development of type-discriminating serodiagnosis.

Infection with herpes simplex virus or hepatitis viruses can lead to fulminant hepatitis, but there is controversy about the underlying conditions needed for such disease. To investigate how the impairment of host defences might be involved, macrophages were depleted by administration of silica to mice before intravenous injection with herpes simplex virus type 1 (HSV-1). Such mice died rapidly and their livers were yellowish and shrunken (acute yellow atrophy), and occasionally grossly haemorrhagic. Small foci of apoptotic cells developed in the liver lobules; these rapidly became confluent and zonal over time. The overall lesion pattern was similar to massive hepatic necrosis, and there was extensive HSV replication in the liver lesions. In the liver, DNA fragmentation characteristic of apoptosis followed the time course of HSV-1 propagation. These findings suggest that one of the underlying conditions for fulminant viral hepatitis may be inadequate macrophage response, and that the massive hepatic damage, often defined as cell necrosis, may actually be apoptosis of liver cells subsequent to virus infection.

The PML gene product is associated with a defined nuclear structure (10-20 per cell) known variously as PML-bodies, ND10, PODs or Kr bodies. Certain conditions are known to compromise the integrity of PML-bodies; these include environmental stress (e.g. heat shock), a chromosomal translocation-associated acute promyelocytic leukaemia, and infection with certain viruses [including human cytomegalovirus (HCMV), herpes simplex virus type 1 and adenovirus]. Expression of the HCMV major immediate early (IE) protein (IE1(491aa)) is by itself sufficient to cause disruption of PML-bodies, resulting in the dispersal of the PML antigen uniformly throughout the nucleus. In uninfected cells undergoing mitosis PML is excluded from chromatin. However, both IE1(491aa) and PML were observed to associate with mitotic chromosomes in cells infected with HCMV or transfected with the IE1 gene. A series of in-frame IE1 deletion mutants was used in DNA transfection experiments to identify two large sequence elements (aa 132-274 and the C-terminal aa 347-491) not required for dispersal of the PML antigen. However, a putative leucine-zipper domain (aa 105-139), a putative zinc-finger domain (aa 267-286) and exon 2 and 3 coding sequences (aa 6-85) were required. The association of the IE1 gene product with chromatin required an acidic domain near the C terminus (aa 421-486). The interaction of IE1(491aa) with chromatin was therefore not required for the disruption of PML-bodies. Exon 2 (aa 1-24) was shown to encode a nuclear localization signal.

Human herpesvirus-6 (HHV-6) exhibits a predominant tropism for CD4+ T-lymphocytes, but can infect other components of the blood as well as surrounding tissue and organs. To understand the role of the endothelium in the transmission and haematogenous spread of this virus, human umbilical vein endothelial cells (HUVEC) were infected with HHV-6 and monitored for viral gene expression. The presence of both early and late viral antigens was demonstrated by indirect immunofluorescence in 37.6 and 6.5%, respectively, of HUVEC. However, attempts to detect the release of infectious virus were not successful, indicating infection is semipermissive in nature. Upon continued passage of infected HUVEC monolayers, HHV-6 antigen-positive cells persisted up to 27 days post-infection. Furthermore, the virus could be recovered from HUVEC monolayers that contained fewer than 1% antigen-positive cells by co-cultivation with peripheral blood mononuclear cells. Together, these findings suggest that endothelial cells may serve as a reservoir for harbouring HHV-6.

The Epstein-Barr virus (EBV) ssDNA-binding protein (SSB) encoded by the BALF2 gene is one of the essential replication proteins in the lytic phase of EBV DNA replication. In order to obtain the amount of EBV SSB required for characterization, a recombinant baculovirus containing the complete sequence of the BALF2 open reading frame under the control of the baculovirus polyhedrin promoter was constructed. Insect cells infected with the recombinant virus produced a protein of 130 kDa, recognized by anti-BALF2 protein-specific polyclonal antibody. The overexpressed EBV SSB was purified homogeneously from the cytosolic fraction of the recombinant virus-infected cells. The purified protein displaced short DNA strands from their complementary sequences in the single-stranded form of M13. The helix-destabilizing activity was neutralized by the anti-BALF2 protein-specific antibody. Maximum unwinding occurred at EBV SSB concentrations exceeding saturation level of the DNA substrate. The DNA unwinding reaction mediated by the EBV SSB was highly cooperative and extremely rapid. The reaction displayed no directionality and required neither ATP nor MgCl2, two essential cofactors for DNA helicase activity. The helix-destabilizing property of the EBV SSB may function to melt out secondary structures on the ssDNA template, thereby facilitating the movement of the EBV DNA polymerase.

Nicotiana benthamiana was transformed with three different constructs (pCRA1, pCRA2 and pJC1) containing the coat protein coding sequence of African cassava mosaic virus (ACMV). Transformed plants were inoculated with a coat protein deletion mutant of ACMV that induces mild systemic symptoms in control plants. Several inoculated plants of transgenic lines CRA1/3, CRA1/4, CRA2/1 and CRA2/2 developed severe systemic symptoms typical of ACMV. DNA analysis revealed that, in these plants, recombination had occurred between the mutant viral DNA and the integrated construct DNA, resulting in the production of recombinant virus progeny with 'wild-type' characteristics. No reversion of mutant to 'wild-type' virus was observed in pJC1-transformed plants. Recombinant virus from several transgenic plants was analysed by PCR and parts of DNA A of virus progeny were cloned. Sequence analysis revealed that only a few nucleotides were changed from the published sequence.

Brome mosaic bromovirus (BMV) 3a protein is required for cell-to-cell movement of the virus in host plants. The BMV 3a protein (B3a) was produced in Escherichia coli using an expression vector. Gel retardation analysis and UV cross-linking experiments demonstrated that B3a bound single-stranded RNA cooperatively without sequence specificity. Binding competition analysis showed that B3a bound to single-stranded nucleic acids more strongly than to double-stranded nucleic acids. Deletion mutagenesis located a nucleic acid-binding domain to amino acids 189-242. Western blot analysis of fractionated proteins of BMV-infected barley using monoclonal antibodies against B3a indicated that B3a may interact with membrane materials and form complexes in the cytoplasm. Immunogold labelling of thin sections of infected barley tissues revealed that B3a was associated with plasmodesmata and cytoplasmic inclusions.

Pea early browning virus (PEBV) is transmitted between plants by root-feeding trichodorid nematodes. Mutagenesis studies have implicated two non-structural viral proteins in the transmission process. These two proteins [the 29 kDa ('29K') protein and the 23K protein] were expressed in bacteria and used to raise antibodies. In Western blotting experiments, the antibodies detected both of these virus proteins in leaves and roots of infected Nicotiana bethamiana and N. clevelandii plants. Periodate treatment of proteins transferred to nitrocellulose membranes suggested that the PEBV 23K protein may be glycosylated. A PEBV mutant was constructed lacking the complete 23K coding sequence. The mutant was able systemically to infect Nicotiana spp. but caused striking chlorotic ringspot leaf symptoms and stunting of both leaves and roots. These symptoms were absent in plants doubly-infected with the mutant and wild-type PEBV. The 23K gene deletion mutant was transmitted by nematodes at a much reduced frequency compared to wild-type virus, indicating that the 23K protein is involved in but not essential for vector transmission. Western immuno-blot and ELISA experiments revealed that the reduction in the nematode-transmissibility of PEBV carrying mutations in the 23K gene did not result from interference in the expression of the 29K transmission protein or from gross changes in the titre of virus in the roots of infected plants.

The entire genome of grapevine leafroll-associated closterovirus-2 (GLRaV-2), except the exact 5' terminus, was cloned and sequenced. The sequence encompasses nine open reading frames (ORFs) which include, in the 5' to 3' direction, an incomplete ORF1a encoding a putative viral polyprotein and eight ORFs that encode proteins of 52 kDa (ORF1b), 6 kDa (ORF2), 65 kDa (ORF3), 63 kDa (ORF4), 25 kDa (ORF5), 22 kDa (ORF6), 19 kDa (ORF7) and 24 kDa (ORF8) respectively, and 216 nucleotides of the 3' untranslated region. An incomplete ORF1a potentially encoded a large polyprotein containing the conserved domains characteristic of a papain-like protease, methyltransferase and helicase. ORF1b potentially encoded a putative RNA-dependent RNA polymerase. The expression of ORF1b may be via a +1 ribosomal frameshift mechanism, similar to other closteroviruses. A unique gene array, which is conserved in other closteroviruses, was also identified in GLRaV-2; it includes genes encoding a 6 kDa small hydrophobic protein, 65 kDa heat shock protein 70, 63 kDa protein of function unknown, 25 kDa coat protein duplicate and 22 kDa coat protein. Identification of ORF6 (22 kDa) as the coat protein gene was further confirmed by in vivo expression in E. coli and immunoblotting. Phylogenetic analysis comparing different genes of GLRaV-2 with those of other closteroviruses demonstrated a close relationship with beet yellows virus (BYV), beet yellow stunt virus and citrus tristeza virus. GLRaV-2 is the only closterovirus, so far, that matches the genome organization of the type member of the group, BYV, and thus can be unambiguously classified as a definitive member of the genus Closterovirus.

The RNA genome of grapevine leafroll-associated closterovirus-3 (GLRaV-3) was cloned as a cDNA generated from GLRaV-3-specific dsRNA, and a partial genome sequence of 13154 nucleotides (nt) including the 3' terminus was determined. The sequenced portion contained 13 open reading frames (ORFs) potentially encoding, in the 5'-3' direction, proteins of > 77 kDa (ORF1a; helicase, HEL), 61 kDa (ORF1b; RNA-dependent RNA polymerase, RdRp), 6 kDa (ORF2), 5 kDa (ORF3, small transmembrane protein), 59 kDa (ORF4; heat shock protein 70, HSP70), 55 kDa (ORF5), 35 kDa (ORF6; coat protein, CP), 53 kDa (ORF7; diverged coat protein, CPd), 21 kDa (ORF8), 20 kDa (ORF9), 20 kDa (ORF10), 4 kDa (ORF11), 7 kDa (ORF12), and an untranslated region of 277 nt. ORF1b is probably expressed via a +1 ribosomal frameshift mechanism, most similar to that of lettuce infectious yellows virus (LIYV). Phylogenetic analysis using various gene sequences (HEL, RdRp, HSP70 and CP) clearly demonstrated that GLRaV-3, a mealybug-transmissible closterovirus, is positioned independently from aphid-transmissible monopartite closteroviruses (beet yellows, citrus tristeza and beet yellows stunt) and whitefly-transmissible bipartite closterovirus (lettuce infectious yellows, LIYV). However, another alleged mealybug-transmissible closterovirus, little cherry virus, was shown to be more closely related to the whitefly-transmissible LIYV than to GLRaV-3.

A putative DNA helicase gene from the granulovirus of Trichoplusia ni (TnGV) was cloned, sequenced, and compared with the corresponding gene of several multinucleocapsid nucleopolyhedroviruses (MNPVs) including those from Autographa californica (AcMNPV), Orgyia pseudotsugata (OpMNPV), Bombyx mori (BmNPV), and Spodoptera exigua (SeMNPV). The TnGV helicase gene (p137) encoded a helicase of 1158 amino acids with a predicted mass of 137 kDa. Comparison of p137 with AcMNPV p143 revealed 44.5% identity at the nucleotide level, and, respectively, 28.6% identity and 53.0% similarity at the amino acid level. Similar levels of identity and similarity were obtained when TnGV p137 was compared with the corresponding helicase genes of BmNPV, OpMNPV and SeMNPV. Using an antisense probe made from an internal 1.6 kb region of p137, a major transcript of approximately 3600 nt was detected by Northern blot analysis in fat body tissue from TnGV-infected larvae of T. ni. As both TnGV and AcMNPV replicate efficiently in larvae of T. ni, these results demonstrate that baculovirus putative DNA helicases which have diverged markedly can function efficiently in the same host. Three genes flanking TnGV p137, designated ORF68, ORF219 and ORF157, corresponded in order and orientation with AcMNPV ORFs 93, 94 and 96. However, the amino acid similarity between corresponding genes ranged from only 50.4 to 62.5%, providing further evidence that related baculovirus proteins which have diverged markedly can function efficiently in the same host.