- Volume 76, Issue 4, 1995

Intranuclear structure was studied in influenza virus-infected cells by immunofluorescence microscopy with antibodies specific for fibrillarin, the splicing factor SC35 and the autoantigen p80 coilin. In the course of the infection, an increase in the number of coiled bodies was observed, with a parallel decrease in their size. In addition, the normal speckled pattern of the SC35 factor was altered to generate a more punctate distribution. However, no alteration was observed in the fibrillarin staining pattern. Since an alteration in the splicing of both viral and cellular mRNAs upon expression of influenza virus NS1 protein has been reported previously, the possible effects of NS1 expression on intranuclear structure were assayed. The increase in the coiled body numbers was not specific for the expression of NS1 protein, but alterations in the nuclear location of small ribonucleoprotein particles, as determined by immunofluorescence with an anti-Sm serum or the SC35 splicing factor, were produced by the sole expression of NS1 protein. These results correlate with the previously reported inhibition of splicing induced by NS1 protein expression and suggest an interaction of this influenza virus protein with the cellular splicing machinery.

Influenza virus nucleoprotein (NP) was purified from the virus and found to be virtually free from RNA. The morphology of the negatively stained NP was studied using electron microscopy. The monomer protein was found to be a small rod with dimensions of 62 by 35 Å. However, most of the protein was found to exist in polymeric forms ranging from trimers to large structures that were morphologically indistinguishable from the intact viral RNP. Each monomer has two sites for NP–NP contacts at one extremity of the rod. The consequences of this finding for crystallization studies, for in vitro studies on NP–RNA interactions and the possible consequences for in vivo ribonucleoprotein particle assembly are discussed.

Many reports indicate that a long-term asymptomatic state following human immunodeficiency virus type 1 (HIV-1) infection is associated with a low amount of circulating virus. To evaluate the possible effect of stabilizing a low virus load by non-sterilizing preexposure vaccination, a quantitative virus isolation method was developed and evaluated in four chronically infected chimpanzees infected with a variety of HIV-1 related isolates. This assay was then used to monitor a group of chimpanzees (n = 6) challenged with HIV-1 following vaccination with gp120 or gp160. Data indicated that of the three vaccinated animals which became infected after challenge, the animal with the lowest neutralizing titre at the time of challenge acquired a virus load similar to the control animals, whereas the two other chimpanzees had reduced numbers of virus producing cells in their peripheral circulation. One animal became virus isolation negative, developed an indeterminant PCR signal on lymph node DNA and subsequently became negative for HIV-1 DNA as determined by PCR on PBMC (peripheral blood mononuclear cells) and bone marrow DNA. Recently, the second animal has also become PCR negative. To confirm observations from quantitative virus isolations, quantification of HIV-1 DNA in PBMC and virus RNA in serum was performed by PCR on serially diluted samples at two different time points. Comparison of virus load as determined by these three methods confirmed that there was an effect of vaccination in reducing virus load and demonstrated a correlation between decreased numbers of virus producing cells, HIV-1 DNA containing cells and virus RNA molecules in serum.

The envelope (env) gene of human T cell leukaemia virus type I (HTLV-I) was inserted into an expression vector, referred to as phMTenv, under the transcriptional control of the human metallothionein IIa gene promoter (hMT-IIa). When this vector was transiently transfected in HeLa cells treated with hMT-IIa inducers, formation of multinucleated cells was observed, indicating the expression of functional surface and transmembrane glycoproteins. Of several HeLa cell clones transfected with phMTenv together with a plasmid carrying the neomycin resistance gene and isolated after selection in G418-containing medium, env mRNA was detected in only two, in the presence of hMT-IIa inducers. Viral glycoproteins were found to be weakly expressed, as detected in immunoprecipitation assays of 125I-surface-labelled cells. These env-transfected HeLa cell clones, although unable to form syncytia when cocultivated with untransfected control HeLa cells, retained the capacity to fuse with HTLV-I-producing C91PL T cells. However, a significant decrease in their fusogenic ability was observed, after treatment with hMT-IIa inducers. Under identical experimental conditions, control HeLa cell clones stably transformed with the same plasmid, but lacking the env gene, were still able to fuse with C91PL cells. These observations suggest that a post-transcriptional step in HTLV-I env expression is impaired, probably leading to the establishment of superinfection interference.

In the presence of the methyltransferase inhibitor 3-deazaadenosine (3DA-Ado) the production of infectious Autographa californica nuclear polyhedrosis virus (AcMNPV) in tissue culture was only slightly affected, while the synthesis of very late proteins (polyhedrin and p10) was abolished. The synthesis of the influenza virus proteins NS1 and HA, expressed under the polyhedrin promoter, was also abolished by 3DA-Ado. Furthermore, 3DA-Ado interfered with the shut-off of early and late AcMNPV proteins. Most of these results were also obtained with 5-azadeoxycytidine (5A-dCyt). In cells in which NS1 was produced abundantly, at least one specific AcMNPV protein was not synthesized. However, if the production of NS1 was inhibited by 3DA-Ado, or if HA was synthesized instead, this AcMNPV protein showed up normally.

Transient transfections of tissue culture cells with plasmids encoding the transactivating factors MHBst and HBx of hepatitis B virus result in transcriptional stimulation of multiple target genes. Our experiments show that the NF-κB-binding enhancer element of simian virus 40 (SV40) and the AP-1-binding enhancer element of the human metallothionein IIA gene mediate the transactivating function of MHBst and HBx. In contrast, the elements GT(IIC+I) and Sph(II+I) of the SV40 enhancer, that, as a common feature, require binding of transcription factor TEF-1 for activity, efficiently mediate transactivation only by HBx but not by MHBst. This finding suggests that at least one regulatory pathway exists that can only be activated by HBx but not by MHBst.

The hepatitis B virus (HBV) e antigen (HBeAg) is a 15 kDa soluble antigen derived from a precursor protein (precore protein) by two processing events, cleavage of the N-terminal signal peptide and cleavage of the C-terminal 34 amino acids. So far, the role of the C-terminal sequences in secretion has not been analysed in full. In this study deletion of the last 60 amino acids was found to abrogate HBeAg secretion whereas deletions of the last 10, 25 or 39 amino acids decreased its secretion rate. These data demonstrate that C-terminal precore protein sequences are crucial for HBe secretion and determine its secretion rate.

Previously, we identified human liver endonexin II (EII) present on human hepatocyte plasma membrane as a specific hepatitis B surface antigen (HBsAg) binding protein. We also showed the spontaneous development of anti-idiotypic (anti-HBsAg) antibodies in rabbits immunized with EII and in chicken immunized with the F(ab′)2 fragment of rabbit anti-EII IgG. These findings suggest the existence of a receptor-ligand relationship between EII and HBsAg. In the present study, we demonstrate that small HBsAg conjugated to 10 nm colloidal gold also binds specifically to human hepatocytes. Invagination of the coated pit region at the HBsAg binding sites on the human hepatocyte plasma membrane results in the internalization of the HBsAg-gold particles. The binding and consequently the internalization of HBsAg is inhibited by anti-EII or anti-idiotypic (anti-HBsAg) antibodies. These findings indicate that EII is directly involved in the binding and uptake of hepatitis B envelope proteins.

The human adenovirus type 5 E4 transcription unit has the potential to encode at least seven distinct polypeptides from reading frames accessed by differential splicing of a single primary transcript. Only some of these polypeptides have yet been detected during viral infection of cultured cells. Mutational inactivation of the reading frames whose products have not been described has no apparent effect on the growth of virus in standard cultured human cell lines, indicating that these proteins, if they exist, have only a subtle, non-essential role in the replication cycle. We have raised an antiserum to one of these undefined products, E4 Orf2, expressed in bacteria. Using this reagent, it was possible to show that Orf2 was expressed during the lytic cycle in HeLa cells, being a soluble cytoplasmic component appearing with early kinetics. No association of Orf2 protein with other infected cell components was detected.

Sequence analysis of human papillomavirus (HPV) general primer GP5/6 mediated PCR products revealed the presence of short highly conserved sequences adjacent to the 3′ ends of both primers. Part of these sequences was used to elongate GP5 and GP6 at their 3′ ends to generate the primers GP5+ and GP6+, respectively. Compared with the GP5/6 PCR, GP5+/6+ specific PCR on 22 cloned mucosotropic HPVs revealed an improved HPV detection, reflected by a 10- to 100-fold higher sensitivity and a markedly increased signal to background ratio, especially at the gel level. As determined on purified DNA, the sensitivity of this GP5+/6+ based assay was at the femtogram level for those HPV genotypes which match strongly with the primers (e.g. HPV-16) and at the picogram level for HPV types (e.g. HPV-39 and -51) having four or more mismatches with one or both primers. Application of both methods on 264 cervical scrapes of a cohort of women participating in a prospective follow-up study revealed an increase of total HPV positivity from 39% (GP5/6 PCR) to 43% (GP5+/6+PCR) of the scrapes. Additional HPV typing by PCR specific for the HPV-6, -11, -16, -18, -31 and -33 revealed that all GP5+/6+ PCR positive cases which were negative by GP5/6 PCR (n = 12) contained HPV types different from these six types. These data indicate that the GP5+/6+ PCR method provides an increased detection level mainly of uncommon, apparently poorly matched HPV types in cervical scrapes and most likely in the enlargement of the spectrum of HPVs detectable by this assay.

Alcelaphine herpesvirus type 1, a rhadinovirus causing malignant catarrhal fever of ruminants, has an 1959 nucleotide open reading frame with significant homologies to semaphorin genes. While truncated genes of similar structure have been found in poxviruses, this is the first known example of a semaphorin-like gene in a herpesvirus.

The complete nucleotide sequence of the gene encoding the protease and ICP35 capsid protein of herpes simplex virus (HSV) type 2 strain G has been determined. The gene consisits of a 1908 bp open reading frame capable of encoding 636 amino acids. Comparative nucleotide sequence analysis of the gene with published HSV-1 sequences revealed that the HSV-2 nucleic acid segment most likely encodes two distinct proteins. The first protein is the HSV-2 protease, which is required for the assembly of the herpes virus capsid. The second protein has been previously designated as ICP35 in HSV-1, and belongs to the family of herpesvirus capsid proteins.

The putative 38 kDa movement protein (P38) gene, located on the RNA2 of grapevine fanleaf nepovirus (GFLV), was cloned in Escherichia coli and expressed as a fusion protein fused to six histidines (6HisP38). This protein was purified and used to produce a specific polyclonal antiserum from which immunoglobulins were isolated by immunoaffinity against the recombinant 6HisP38. Western immunoblot analyses on GFLV RNA2 in vitro maturation products showed that the antibodies were specific for P38 protein. This protein was detected as early as 18 h post-inoculation in GFLV-infected Chenopodium quinoa protoplasts and accumulated to very high levels. Tissue-prints and time course experiments on infected C. quinoa plants confirmed that P38 is present at a high level late in infection and is a final maturation product of the GFLV RNA2 polyprotein in vivo. P94 and P66 intermediates of maturation and polyprotein P2 were also detected in vivo but in very low concentrations. No significant difference was observed in the relative amounts of P66 and P94 detected in vivo, contrary to what occurs in vitro. Subcellular fractionation studies showed that P38, although mainly cytosolic, is also found in association with cell wall and membranes. Thought to be the GFLV movement protein, P38 would thus behave in an ‘atypical’ manner.

We have cloned the region of tomato ringspot nepovirus (TomRSV) RNA-1 coding for the putative TomRSV 3C-related protease (amino acids 1213 to 1508) in a transcription vector and in a transient expression vector. Using cell-free transcription and translation systems and plant protoplasts, we have demonstrated that proteins produced from these clones possess a proteolytic activity in trans on the cleavage site between the TomRSV movement and coat proteins. By amino acid homology of the TomRSV 3C-related protease with other nepo- and comovirus proteases, His1283, Glu1331 (or Asp1354) and Cys1433 have been predicted to constitute the catalytic triad. Site-directed mutagenesis of His1283 to Asp abolished the TomRSV protease activity, in vitro and in vivo. The cleavage site between the TomRSV movement and coat proteins has been determined to be Q/G, by direct protein sequencing. Previously, His1451 located in the substrate binding pocket of the TomRSV 3C-related protease has been suggested to be involved in the cleavage site specificity. We show that an inactive TomRSV 3C-related protease is obtained after substitution of His1451 with Leu. These results are discussed in light of the possible relation of the TomRSV 3C-related protease to 3C-related proteases of nepo-, como- and potyviruses.

Olive latent virus 2 (OLV2), a virus with particle shapes resembling those of alfalfa mosaic alfamovirus (AMV), has four major RNA species, two of which (RNA3 and RNA4) were completely sequenced. RNA3 was a bicistronic molecule containing two clear-cut ORFs, one of which (ORF1) coded for a 36.5 kDa polypeptide with conserved motifs of the ‘30K superfamily’ movement proteins and the other (ORF2) encoded a 20 kDa polypeptide identified as the viral coat protein. RNA4, which was a little smaller than RNA3 (2078 nt versus 2438 nt), also differed from RNA3 in a few positions, but its in vitro translation did not produce any protein. By contrast, an additional RNA, 1042 nt in size with strong sequence homology with RNA3 and RNA4, was identified in RNA extracts from infected plants. This RNA, which may be a subgenomic form of RNA3 carrying the coat protein cistron, is apparently encapsidated to a very small extent, or not at all. Comparative computer-assisted analysis of virus-coded protein sequences disclosed homologies suggesting that OLV2 may belong to the family Bromoviridae, but as an entity separated from the currently known genera.

Potato virus Y (PVY) the type member of the genus Potyvirus, occurs world-wide as isolates which differ in host range and the type of symptoms caused. The sequences of a 5′ segment of viral RNA overlapping the 5′ non-translated region (5′NTR) alone (ten isolates) or the 5′NTR and the adjacent P1 coding region (eight isolates) were established. These data were used to quantify the polymorphism in the 5′-terminal part of the PVY genome. Nucleotide sequence identity between isolates ranged from 66–100% in the 5′NTR and from 70–100% in the P1 coding region. The lowest amino acid sequence similarity between PVY P1 was 77%, illustrating the high variability of this protein in the PVY species. Phylogenetic trees based on either 5′NTR or P1 sequence analyses resulted in the same clustering of the studied isolates into three groups. Group I comprises potato isolates all inducing ‘tobacco veinal necrosis’ symptoms. Group II contains isolates inducing either ‘tobacco veinal necrosis’ or mosaic symptoms in tobacco. Group III contains mainly pepper or tomato isolates inducing mosaic symptoms in tobacco and shows a geographical clustering of the Tunisian isolates. This clustering into three groups is discussed in comparison with phylogenetic trees previously obtained from capsid gene or 3′NTR sequence analysis in the PVY species. Multiple sequence alignment indicated conserved motifs potentially involved in viral functions.

Proteolytic processing of the potyvirus polyprotein is mainly performed by the virus-encoded NIa protease, whose cleavage sites are characterized by conserved heptapeptide sequences. Partial processing at the cleavage site present between the P3 and 6K1 cistrons by the plum pox potyvirus (PPV) NIa protease has been previously shown to occur in vitro. We have now studied the role of polyprotein processing at the P3−6K1 junction in vivo, using a full-length PPV cDNA clone. PPV mutant transcripts containing a histidine for glutamine substitution in the cleavage site sequence (a change that abolishes in vitro processability) are able to infect Nicotiana clevelandii plants, indicating that normal processing at the P3−6K1 junction is not required for virus viability. However, disease symptoms were not detected and virus accumulation occurred after a second site mutation was introduced into the 6K1 cistron during replication. This additional change did not restore the in vitro processability of the mutant heptapeptide. Changes at other positions in the heptapeptide (that only slightly altered the in vitro processability of this NIa site) were also engineered and it was found that these mutations affected the time course and severity of the symptom induction process. A possible regulatory effect on the function of the potyvirus P3+6K1 protein by processing at the P3−6K1 junction is discussed in light of our present results with PPV.

Complete sequence data for the RNA 3 of both citrus leaf rugose (CiLRV) and citrus variegation (CVV) ilarviruses have been determined. The RNAs are 2289 nt (CiLRV) and 2309 nt (CVV) in length and both contain the typical Bromoviridae arrangement of two open reading frames (ORFs) which, when translated, code for proteins that correspond to the M r 32 000 (32K) putative movement proteins (ORF 1) and the coat proteins (ORF 2) of the respective viruses. The 3′ termini of both viruses can be folded to form a secondary structure similar to those reported for other ilarviruses. These are the first complete nucleotide sequences for RNA 3 of members of subgroup 2 of the ilarviruses. The two viruses share substantial homology in nucleic acid sequence, code for identically sized coat proteins and share high levels of identity in the translated products of both ORFs. Although related, these viruses differ sufficiently to be considered distinct. The RNA 3s of CiLRV and CVV appear to be distinct from those of other ilarviruses for which comparable sequence data are available and also from the closely related alfalfa mosaic virus.