- Volume 76, Issue 7, 1995

The human papillomavirus type 16 (HPV-16) E7 and adenovirus (Ad) E1A oncoproteins share a common pathway of transformation. They disrupt the cell cycle G1 phase-specific protein complex containing the E2F transcription factor and the regulatory protein Rb1, the retinoblastoma tumour suppressor gene product. In the G1 and S phases of the cell cycle, E7 and E1A bind two other cellular complexes containing the Rb1-related protein p107 and E2F. Ad E1A disrupts both complexes and releases active E2F. In contrast, HPV-16 E7, although it efficiently binds both E2F-p107 complexes, causes dissociation of the G1 phase complex only. Using chimeric proteins of HPV-16 E7 and Ad E1A we were able to demonstrate that the ability of E1A to disrupt both G1 and S phase E2F-p107 complexes is not due to the higher concentration of Ad E1A in the cell, but is an intrinsic property of the Ad E1A transforming region. These data suggest that E1A and E7 may function in cellular transformation in similar, but not identical ways.

This study was undertaken to determine the prevalence of core-gene-defective hepatitis B virus (HBV) in patients with chronic HBV infection, and the nature and significance of the deletions. PCR was performed on sera from 263 patients with chronic HBV infection. Seventeen (6.5%) patients had smaller band(s) in addition to a band of the expected size. Additional bands were also detected in the follow-up samples from 80% and 3% patients who had respectively, multiple and single bands initially. Six patients were further studied by direct sequencing. Four patients had in-frame deletions leading to loss of codons 79–122 of the core gene. Two patients had identical frameshift deletions of nucleotides 2204–2333 resulting in the loss of the first nine codons of the overlapping P gene. Follow-up samples from three of four patients studied showed deletions identical to those in the initial samples. The persistence of these deletions suggests that they were stable and that they may contribute to the chronicity of infection.

Although the guinea-pig cytomegalovirus (GPCMV) displays a similar pathogenesis to human cytomegalovirus (HCMV), there have unfortunately been few molecular analyses of the GPCMV genome to date. The guinea-pig has proved useful for the testing of drugs active against CMV infection, and insights derived from characterization of the specific virally encoded molecular targets of antiviral therapies would allow this model system to be more fully developed. Because the DNA polymerase serves as an important target for nucleoside antiviral agents active against herpesviruses, experiments were undertaken to identify, clone and sequence the GPCMV DNA polymerase gene (pol). A 3285 bp ORF capable of encoding a 1094 amino acid protein was identified spanning portions of the HindIII Q and P fragments of the genome. This ORF contained extensive homology to other herpesvirus DNA pol genes. Northern blot analyses identified two 3′ coterminal pol-specific mRNAs of 3.9 and 1.9 kb at early times post-infection. Primer extension and nuclease protection analyses mapped the 5′ end of the 3.9 kb transcript to a site 275 bases upstream of the pol initiation codon. Comparison of the GPCMV pol-encoded sequence to those of other herpesvirus polymerases identified non-conservative amino acid substitutions in a domain involved in substrate recognition.

Bovine herpesvirus 4 (BHV-4) DNA sequences located outside the gene blocks conserved among the gamma-herpesviruses BHV-4, herpesvirus saimiri (HVS) and Epstein—Barr virus (EBV) were analysed. Twelve potential open reading frames (ORFs) were found. Protein database comparisons showed that no ORF translation products were similar to proteins encoded by alpha- or betaherpesviruses. Nevertheless, six of the ORFs were homologous in amino acid sequences to proteins encoded by HVS but apparently not to those encoded by EBV. Furthermore, the location and orientation of these six ORFs in the BHV-4 genome were similar to the corresponding ORFs in the HVS genome. No genes homologous to known cellular genes were found in the BHV-4 genome; this feature is the major difference between the BHV-4 and HVS genomes with regards to the overall gene content.

Syncytial mutations of herpes simplex virus type 1 (HSV-1) strains ANG, ANG path, HFEM, tsB5 and HSZP cause extensive cell fusion and were mapped to the cytoplasmic domain of glycoprotein B (gB), within the syn 3 locus. These strains are so far the only ones which show the phenotype ‘fusion from without’ (FFWO): 60 min after infection with high m.o.i., cells in a tissue culture are fused without transcription and translation of the viral genome. In this report we detected, using the recombinants 27/III and K-7, that an amino acid exchange from Ala to Val at aa position 854 of gB is the main determinant for FFWO activity of strains ANG, ANG path and recombinant K-7. The transfer of this mutation to wild-type strains KOS and 17 syn+ by co-transfection results in recombinants KOS-854Q, 17-syn3, 17-syn3a and 17-syn3b. As a selection marker we used the cyclosporin A resistance of fusion which was shown to be a unique characteristic of syn 3 locus mutants. The recombinants show the FFWO phenotype in BHK cells but not in Vero cells. FFWO was shown to be cell-type dependent by comparing the number of p.f.u. needed to induce FFWO in various cell types.

We examined the influence of inactivation of various genes located in the unique short (Us) region of pseudorabies virus on virus replication and assembly in porcine nasal mucosa explant cultures. The following strains were used: the virulent wild-type strain NIA-3, and strains derived from NIA-3 containing a mutation inactivating the genes encoding either the US3-encoded protein kinase (PK), gG, gD, gI, gE, the 28 kDa (‘28K’) protein (single mutant), or the 28K and 11 kDa (‘11K’) proteins (double mutant). In addition a wild-type rescuant was used, which was generated by marker rescue from a PK− mutant. All virus strains infected nasal epithelium and had invaded the stroma after approximately 24 h. The morphogenesis in nasal epithelium cells of two PK− mutants showed the most striking differences compared to the parent NIA-3 strain and the other mutant strains. The changes could be ascribed to the US3-encoded PK because the rescue mutant showed a similar morphogenesis to wild-type NIA-3. The transmembrane transport of the PK− mutants was impaired at the outer nuclear membrane which resulted in an accumulation of virions in the perinuclear space. These results suggest that proteins, phosphorylated by the US3-encoded PK, are involved in debudding of virus particles at the outer nuclear membrane. This defect in the transport of the US3 mutant probably explains their reduced replication in vitro. The gG−, gD−, gI−, gE−, 28K− and 11K− mutant strains showed minor or no changes in viral assembly. Thus the reported decreased virulence of the gD−, gI− and gE− mutants was, in contrast to that of the PK− mutants, not associated with clear alterations in morphogenesis.

We constructred a hybrid type 1/type 3 poliovirus comprising the BC-loop of capsid protein VP1 of PV3/Finland/60212/84 and the rest derived from PV1/Mahoney, and cultured the virus in the presence of diluted rabbit antiserum to PV3/Finland/60212/84. Several strains isolated under this selection showed point mutations in the inserted type 3 poliovirus sequence but only in one case in the flanking PV1/Mahoney-derived RNA. These results indicate that, with the use of recombinant cDNA technology, it may be possible to study molecular interactions of defined regions of virus capsid proteins with neutralizing polyclonal antibodies.

Murine Mx1 protein is an interferon-inducible GTPase which localizes in nuclei and inhibits influenza virus infection. Wild-type Mx1 and two mutant Mx1 proteins, each carrying a single mutation either in the GTP-binding motif (S50I) or in the self-assembly motif (C71S), were expressed in MDCK cells. Wild-type Mx1 localized in nuclei, forming small granules with minute dots, and inhibited influenza virus growth. Mutant S50I, which had no GTP-binding or GTPase activities, formed linear structures in nuclei and lacked anti-viral activity, while C71S appeared diffuse in nuclei as minute dots without granules, but retained the inhibitory activity against influenza virus growth. A correlation existed between GTPase activity, intranuclear distribution and antiviral activity. We concluded that GTPase activity is essential for expression of the biological activity of Mx1 protein.

Hepatitis C virus (HCV) isolates from a cohort of 315 patients from the Benelux countries (Belgium, The Netherlands, Luxembourg) were genotyped by means of reverse hybridization Inno-LiPA (line probe assay). Genotypes 1a, 1b, 2a, 2b, 3a, 4a and 5a were detected. From the cohort, isolates representing all types and those showing an aberrant LiPA pattern were further analysed by sequencing parts of the 5′ UTR, core (nt 1 to 326; aa residues 1 to 108) and core/E1 (nt 477 to 924; aa residues 159 to 308) regions. Molecular evolutionary analysis of the core and core/E1 regions allowed discrimination between known and additional subtypes, especially within types 2 and 4. The core region is not suitable for classification of new subtypes because of the relatively high level of conservation. The core/E1 region displays a higher level of sequence variation and allows much more distinct discrimination between subtypes. Genotypes 2 and 4 are particularly heterogeneous, with at least 7 and 10 subtypes, respectively. In contrast to previous reports from Europe, HCV isolates from the cohort constituted a highly heterogeneous population of virus variants, especially within genotypes 2 and 4.

Genetic variation between geographically and temporally distinct isolates of dengue-1 (DEN-1) and dengue-4 (DEN-4) viruses was investigated. The nucleotide sequences of a fragment of the envelope protein gene encoding amino acids 28 to 87 of 35 DEN-1 isolates and 28 DEN-4 isolates were determined. Maximum nucleotide sequence variation was 6.9% and 4.9% for DEN-1 and DEN-4 viruses, respectively. Taking a divergence of 6% between the nucleotide sequences as the cut-off value, three genotype groups were defined for DEN-1 viruses, whereas only one was observed for DEN-4 viruses. Molecular analysis of isolates from the South Pacific permits the classification of the recent strains of DEN-1 (1988–1989 epidemics) into a genotype distinct from the genotype which comprises earlier strains. This observation suggests that the recent epidemics were due to the introduction of a new genotype rather than to the re-emergence of the earlier strain.

Four ribozyme and antisense genes targeting citrus exocortis viroid (CEVd) positive- and negative-strand RNA molecules were constructed and used to transform the tomato Lycopersicon lycopersicum cv. UC82B. The tomato is a readily transformable plant and will support replication of CEVd following mechanical inoculation. The ribozyme genes contained three hammerhead catalytic motifs with long hybridizing arms and synthetic RNA transcripts were shown to cleave the target CEVd RNA molecule in vitro. Homozygous transgenic plants were produced from independent transformants expressing either ribozymes or antisense constructs. Inoculation of transgenic seedlings expressing antisense constructs targeting the negative-strand CEVd RNA molecule with CEVd resulted in a moderate reduction in the accumulation of CEVd RNA. In contrast, similarly inoculated transgenic plants expressing constructs targeting the positive-strand CEVd RNA molecule resulted in an increase in the rate of CEVd RNA accumulation. Addition of the ribozyme motifs to the antisense genes did not enhance their efficiency in the suppression of viroid replication and a moderation or elimination of the observed antisense effects was seen in plants expressing the corresponding catalytic RNA-encoding genes.

We examined the correlation of the amino acid at position 129 in the coat protein (CP) of cucumber mosaic virus (CMV) with the phenotype of the viral pathology in tobacco by using CP mutants in which several amino acid substitutions had been introduced. An exchange between Ser129 in CMV-Y, a chlorosis-inducing strain, and Pro129 of CMV-O, a green-mosaic-inducing strain, reciprocally altered the phenotypes of those virus strains on tobacco. Replacement of either Ser129 in CMV-Y or Pro129 in CMV-O with a Leu, as is found in a chlorosis-inducing strain, CMV-M, resulted in veinal necrosis. Furthermore, we created mutants that have a Phe or a Gly at position 129. Two Phe129 mutants induced necrotic lesions on the inoculated leaves, and a Gly129 mutant induced green mosaic symptoms. In inoculated protoplasts, the mutant viruses and the wild-type virus all replicated RNA well, and accumulated CP; however, infection with the Leu129 and Phe129 mutants yielded few virions. The Phe129 mutants lacked the capacity to move systemically in tobacco; by 2 weeks post-inoculation, the Phe129 mutants occasionally gave rise to revertants that elicited chlorosis, green mosaic or veinal necrosis. Sequence analysis revealed that one had reverted to the parental Y strain, and the others had additional single amino acid changes (positions 138, 144 or 147). We suggest that amino acids at specific sites affect the whole structure of the CP and affect virus assembly, virus transport and symptom expression.

The nucleotide sequence of lilac ring mottle ilarvirus (LRMV) RNA 3 consists of 2287 nucleotides and contains two open reading frames (ORF). The first encodes a putative translation product of 285 amino acids (M r 31308) and the second encodes a putative translation product of 206 amino acids (M r 22751). The 3′ terminal nucleotides can be folded into a loop structure similar to models proposed for other ilarviruses, although the last four nucleotides are UCGC not AUGC. The absence of the terminal AUGC motif in both LRMV and two isolates of apple mosaic ilarvirus (ApMV) provides circumstantial evidence which confirms the importance of AUGC motifs upstream of the terminal AUGC in the protein binding function associated with these models. Although the 3′ terminal structure of LRMV exhibits similarities to that of ApMV, comparison of the putative translation products of the two ORFs with similar products for other ilarviruses showed greatest identity with citrus leaf rugose (CiLRV) and citrus variegation (CVV) ilarviruses both of which are members of subgroup 2 of this genus. Thus it is proposed that LRMV be reassigned to subgroup 2 rather than remaining in its current subgroup, 7, or being reassigned to subgroup 3 which contains ApMV.

Processing of the 112 kDa (‘112K’) protein encoded by cowpea mosaic virus RNA 1 was examined in cowpea mesophyll protoplasts using a transient expression system. Cleavage of the 112K protein occurred via two alternative pathways either into VPg and 110K (24K + 87K) or into 26K (VPg + 24K) and 87K proteins. The 26K protein can be further cleaved into VPg and 24K proteins. The results support a model in which the 112K protein functions as the precursor of VPg during initiation of replication.