- Volume 75, Issue 11, 1994

The putative attachment protein of the avian pneumovirus that causes turkey rhinotracheitis is, by analogy with mammalian pneumoviruses, expected to be the major antigenic determinant. We report the nucleotide sequence of the attachment (G) protein genes of five different continental European isolates and compare them with the previously published sequence of the G gene for the focal variant of a U.K. isolate. The nucleotide sequences and the predicted amino acid sequences indicate that there are at least two distinct subgroups, similar to the grouping described for human respiratory syncytial (RS) virus. The U.K. and French isolates form one group and the isolates from Spain, Italy and Hungary form a second. The two subgroups can be easily distinguished on the basis of restriction enzyme digestion of PCR-generated products representing the full-length gene. Within the subgroups the predicted G proteins were highly conserved (98·5 to 99·7% amino acid identity) compared to the levels of identity of RS virus G proteins in the same subgroup (80 to 95%). Between the avian pneumovirus subgroups described here there was an unexpected degree of divergence, the average amino acid identity between members of the two groups being only 38%. This compares with the 53% conservation seen between members of the RS virus subgroups A and B. Comparison of the predicted amino acid sequences showed that the G proteins of members of the two avian pneumovirus subgroups had similar structural features. All proteins had an amino-terminal membrane anchor and the positions of cysteine residues were highly conserved. The potential importance of the high level of variation between the two subgroups in terms of epidemiology of the disease is discussed.

A newly identified hantavirus, tentatively called Four Corners virus (FCV), was found to be the aetiological agent of a 1993 outbreak of hantavirus pulmonary syndrome (HPS) in the southwestern United States. Immunodominant epitopes of 43 and 31 amino acids were identified in the nucleocapsid protein and G1 glycoprotein, respectively. The G1 genes of different hantaviruses are highly divergent, suggesting that geographically diverse FCVs might fail to cross-react owing to antigenic drift. We now show that the immunodominant epitope of G1 is conserved among 18 FCVs from a broad geographical area, despite extensive nucleotide sequence heterogeneity. Antibodies from all 45 HPS patients, separated by more than 3000 km were shown to be reactive with the dominant G1 epitope. Evidence for limited cross-reactivity between the G1 antigen of a novel hantavirus of the cotton rat and that of FCV is presented.

The rabies virus phosphoprotein (P) and nucleoprotein (N) are involved in transcription and replication of the viral genome. Interaction between N and P was studied in vivo in transfected cells expressing both proteins. Co-immunoprecipitation assays revealed that the N-P complex is present in cells expressing both proteins as well as in infected cells. Furthermore, immunostaining showed that coexpression of N and P was sufficient to induce the formation of cytoplasmic inclusions similar to those found in infected cells. In addition, deletion mutant analysis of P was performed to identify the regions of P interacting with N. The results indicate that at least two independent N-binding sites exist on P protein: one is located in the carboxy-terminal part of the protein and another between amino acids 69 and 177. The formation of cytoplasmic inclusions seems to require the presence of both N-binding sites on P protein.

We have re-evaluated the relationships among the polypeptides of eastern equine encephalitis (EEE) viruses using SDS-PAGE and peptide mapping of individual virion proteins. Four to five distinct polypeptide bands were detected upon SDS-PAGE analysis of viruses: the E1, E2 and C proteins normally associated with alphavirus virions, as well as an additional more rapidly-migrating E2-associated protein and a high M r (HMW) protein. In contrast with previous findings by others, the electrophoretic profiles of the virion proteins of EEE viruses displayed a marked correlation with serotype. The protein profiles of the 33 North American (NA)-serotype viruses examined were remarkably homogeneous, with variation detected only in the E1 protein of two isolates. In contrast, considerable heterogeneity was observed in the migration profiles of both the E1 and E2 glycoproteins of the 13 South American (SA)-type viruses examined. Peptide mapping of individual virion proteins using limited proteolysis with Staphylococcus aureus V8 protease confirmed that, in addition to the homogeneity evident among NA-type viruses and relative heterogeneity among SA-type viruses, the E1 and E2 proteins of NA-and SA-serotype viruses exhibited serotype-specific structural variation. The C protein was highly conserved among isolates of both virus serotypes. Endoglycosidase analyses of intact virions did not reveal substantial glycosylation differences between the glycoproteins of NA- and SA-serotype viruses. Both the HMW protein and the E2 protein (doublet) of EEE virus appeared to contain, at least in part, high-mannose type N-linked oligosaccharides. No evidence of O-linked glycans was found on either the E1 or the E2 glycoprotein. Despite the observed structural differences between proteins of NA- and SA-type viruses, Western blot analyses utilizing polyclonal antibodies indicated that immunoreactive epitopes appeared to be conserved.

Coxsackievirus B3 (CVB3)-induced myocarditis was studied in euthymic (nu/+) and athymic (nu/nu) C3H/ HeN (H-2k) mice. Mice were inoculated intra-peritoneally with 106 p.f.u. of CVB3 (Nancy strain) and sacrificed at intervals up to 92 days post-inoculation (p.i.). Viraemia peaked at day 2 to 3 p.i. and ceased at day 5 to 7 p.i. in a synchronized manner in both sets of mice. Very few infectious particles were detected in the blood of nu/nu mice after day 14 p.i. In nu/nu mice, CVB3 persisted in myocardial tissue with constant titres between 2·7 ± 1·9 × 104 and 7·6 ± 5·2 × 104 p.f.u./mg from day 3 to 92 p.i., which were comparable to those of nu/+ mice in the acute phase. In nu/+ mice, the virus was recovered from all animals examined by day 11 p.i. and from three out of 13 mice between days 14 and 21 p.i., yet no virus was recovered from nu/+ mice at day 42 p.i. In nu/nu mice, sense and antisense RNA for CVB3 was detected in the myocardial tissue up to day 42 p.i. by in situ hybridization and up to day 92 p.i. by reverse transcriptase-PCR. Neither sense nor antisense RNA was detected after day 21 p.i. in nu/+ mice with the same techniques. Myocardial tissue damage was analysed morphologically. At day 92 p.i., the area of myocardial injury peaked at 23% of the section in nu/nu mice. In contrast, less than 0·6% of tissue sections contained lesions in nu/+ mice. A neutralizing antibody response to CVB3 was observed in both nu/nu and nu/+ mice. The mean titre of neutralizing antibody was significantly higher at day 21 p.i. in nu/+ mice, but similar at day 42 p.i. with nu/nu and nu/+ mice. Perforin-producing natural killer-like cells, which are considered to play an important role in causing acute myocarditic lesions in immunocompetent mice, were found in the lesions of nu/nu mice persistently infected with CVB3. Prolonged tumour necrosis factor-α mRNA synthesis detected in nu/nu mice appears to reflect the continuous activation of macrophages, which extend phagocytic reactions to virus-infected myocytes. These immunological results suggested that the host immune response devoid of antigen-specific T cell function is not sufficient to terminate CVB3 infection in nu/nu mice. Also, it appears that competent cellular immunity, on the whole, plays a role in curing rather than in aggravating myocarditis in nu/+ mice. In conclusion, our results indicate that the presence of replicating CVB3 is an important factor in the development of myocarditic lesions in C3H/HeN mice.

To identify Mengo virus-specific T cell epitopes in mice (the natural host for the virus), lymph node cells were obtained from BALB/c (H-2d ) mice, previously immunized with u.v.-inactivated virus, and stimulated in vitro with each of 116 overlapping peptides (10 to 18 residues long) covering the entire capsid coding region (834 amino acids). T cell epitopes were defined on the basis of specific peptide-induced lymphocyte proliferation. Where proliferation occurred, immunological characterization showed that it was the CD4+ T helper (Th) cell subpopulation that was responsible for the Mengo virus-specific response. Surprisingly, no Mengo virus Th cell epitopes were found in capsid protein VP1 or VP4. Six peptides in VP2 (residues 1 to 15, 99 to 108, 118 to 132, 133 to 147, 227 to 236 and 247 to 256) identified the positions of separate Th cell epitopes, and two overlapping peptides (residues 173 to 182 and 178 to 192) defined an additional Th cell immunogenic sequence. Three individual peptides in VP3 (residues 46 to 58, 136 to 150 and 198 to 212) and two overlapping peptides (residues 1 to 15 and 11 to 20) also represent Th cell epitopes. Similar assays with C57BL/6 (H-2b ) and SJL/J (H-2s ) mice showed that the pattern of recognition of these peptides was H-2 restricted. Each of the previously identified sites of B cell antigenicity in VP2 and VP3 are associated with one Th epitope. Comparison of the experimentally determined Th epitopes with potential T cell epitopes identified by several predictive strategies revealed only a low correlation between authentic and predicted epitopes.

Cathepsin D and cathepsin B are endosomal/lysosomal proteases that are thought to play a role during in vivo antigen processing, releasing fragments for binding to major histocompatibility complex class II products and subsequent presentation to T cells. Here we treated purified foot-and-mouth disease virus (FMDV) strain A10Holland with both enzymes. Cathepsin D, but not cathepsin B, was shown to release fragments from reduced or non-reduced FMDV under mild conditions in vitro. Twenty-eight predominant cathepsin D-released fragments were purified by HPLC and identified by amino acid composition analysis and sequencing. The unseparated set of fragments produced (the digest) was able to stimulate T cells from eight vaccinated cattle. With respect to the response to intact virus the extent of the response to the digest differed between animals: four animals could be classified as good responders, three as intermediate responders and one as a low responder. Subsequently, we investigated the proliferative T cell response to a large set of synthetic peptides in detail for two animals, one belonging to the group of good responders, the other being the low responder. The peptides covered all 28 cathepsin D-released fragments analysed and also several sequences not recovered from the digest. In this way seven T cell sites could be identified, five of which coincided with cathepsin D-released fragments. The other two T cell sites were VP2[54–72], being a homologue of a T cell site identified for FMDV strain O1K and the N terminus of VP4. Whether the most dominantly recognized T cell site was recovered from the digest or not was shown to be related to the good or low response to the digest. These findings suggest a role for cathepsin D in the release of some but not all T cell-stimulatory fragments from FMDV.

Experimental infection of transmissible mink encephalopathy (TME) in two closely related mustelids, black ferret (Mustela putorius furo) and mink (Mustela visa), revealed differences in their susceptibility to the TME agent. When challenged with the Stetsonville TME agent, a longer incubation period was observed in ferrets (28 to 38 months) than mink (4 months). Western blot analysis of ferret and mink prion proteins (PrP) demonstrated no detectable differences between the proteins. Northern blot analysis of ferret brain RNA indicated that PrP mRNA abundance is similar in infected and uninfected individuals. We amplified the PrP coding region from ferret DNA using the polymerase chain reaction and compared the deduced amino acid sequence of the ferret PrP gene with the mink PrP gene. This comparison revealed six silent base changes and two amino acid changes between mink and ferret: Phe → Lys at codon 179 and Arg → Gin at codon 224, respectively. These changes may indicate the region of PrP that is responsible for the species barrier effect between mink and ferret.

Simian immunodeficiency virus (SIV) and human immunodeficiency virus type 2 (HIV-2) encode the accessory viral protein X (Vpx) known to be incorporated into virions in amounts comparable to those of the Gag proteins. The localization of Vpx within SIVmac-infected HUT-78 cells and SIVmac virions was studied by immunoelectron microscopy. Vpx appeared to be associated with extracellular virions as well as budding viral particles at the surface of infected cells. Immunolabelling of purified viral cores suggested that Vpx was a component of the amorphous material surrounding the core structure. Furthermore, a detergent insoluble fraction containing SIV core proteins was devoid of Vpx. To investigate the protein requirement for packaging of Vpx, BHK-21 cells were co-infected with vaccinia virus recombinants encoding Vpx and other SIV proteins able to assemble into virus-like particles. Analysis by immunoprecipitation of the extracellular particulate material as well as immunoelectron microscopy demonstrated that co-expression of Vpx with the Pr56 gag polyprotein was sufficient for the formation of pseudo-virions containing Vpx. Virus-like particles that appeared upon expression of p16 gag did not contain Vpx. The results suggest that Vpx is packaged into viral particles through its binding to the Gag polyprotein. The precise positioning of Vpx within the space separating the viral envelope from the core structure is postulated to result from the reorganization of viral proteins that occurs upon Gag polyprotein cleavage and budding.

Human immunodeficiency virus type 1 (HIV-1) chronically infected (Cl) cell lines were established from HIV- lIIIB/LAIinfected MT-4 cells that survived acute infection. The HIV env gene expressed in the two longterm cultured cell lines differed from that of the lines cultured for shorter periods, by coding for a glycoprotein gp160 that had the C terminus deleted. One long-term cultured cell line, Cl-17, was studied in detail. An insertion of a premature stop codon in the env gene caused about 90% of gp160 molecules to be truncated (gp160×), lacking both cytoplasmic and transmembrane domains; these species were secreted into the cell medium, and could form oligomers with other truncated gp160 molecules as well as with their normal counterparts. Cl-17 cells constantly yielded high levels of viral protein and relatively low quantities of infectious virus, without cytopathicity. However, acute infection of fresh MT-4 cells with CI-17-derived virus led to cytopathicity, the rate of which as well as the Env glycoprotein pattern depended on multiplicity: (i) using an infection dose of 10−4 ID50/cell, cells died 7 to 8 days post-infection with normal gp160 synthesis predominating; (ii) with 10−2 ID60, gp160× was produced as early as 48 h postinfection and cell death was delayed. Predominant gp160× formation occurred again when new Cl cell lines were obtained with CI-17-derived virus. Thus, two human immunodeficiency virus variants, a normal and a defective one, are persistently expressed in CI-17 cells. The other long-term cultured Cl cell line also expressed gp160 with a similar (albeit slightly longer) deletion of a C-terminal region in most molecules, but the cell lines that were cultured for shorter periods did not. These results suggest that the emergence of HIV variants with a C-terminal deletion in the Env glycoprotein, which coexist with normal virus, may play a role in maintaining the long-term growth capacity and viability of Cl cells.

During productive infection of human T lymphocytes in cell culture, the expression of human immunodeficiency virus type 1 is temporally regulated by virus-encoded regulatory proteins. Among these Nef, whose function has not been clearly elucidated, is thought to alter CD4+ T cells. We examined the possibility that the nef gene interferes with the translation process in a cell-free system. The results demonstrate that the nef gene product mediates an inhibitory effect on protein synthesis. Conversely, the use of antisense nef mRNA did not affect translation. Further observations suggest that this inhibitory effect is an inherent property of the nef gene product itself and not of its mRNA. The data show that the translational repression directed by Nef is a general phenomenon, acting on its own and on other messengers used as reporter mRNAs. We propose that, as a consequence, Nef can play an important role in the pathogenesis of AIDS.

Six mutants that differ in the extent of their carboxy- terminal sequences and two deletion mutants of the gag gene of HIV-1 have been characterized morphologically following their expression in Spodoptera Jrugiperda cells using recombinant baculoviruses. Electron microscopy has revealed distinct morphological forms of the Gag protein that can be classified as either (i) particulate, three-dimensional, spherical or tubular shells or (ii) nonparticulate, two-dimensional, flat, curved or convoluted sheets. Progressive truncation of the carboxy terminus of Gag was accompanied by changes in the morphology and formation of spherical particles from predominantly C-type assembly and budding at the plasma membrane, through B-type intracytoplasmic assembly, to A-type assembly with budding mainly into cytoplasmic vacuoles. Deletions within the Pr24 CA domain of Gag abolished particle formation but retained association of the protein with the plasma membrane. All of the observed morphologies of the mutant Gag proteins could be accommodated within an icosahedral model for the organization of spherical particles and a basic hexagonal arrangement of assembled Gag protein monomers.

Open reading frame (ORF) 9 of the EcoBA I fragment of the baculovirus Autographa californica multinucleocapsid nuclear polyhedrosis virus (AcMNPV) genome encodes a protein, PK-1, which has strong similarity to serine-threonine protein kinases. The published sequence of the pk-1 gene contains an error that, when corrected, extends the ORF by 228 nucleotides. Transcription of pk-1 was first detectable by primer extension at 12 h post-infection, accumulating to high levels during the very late phase of infection. PK-1 produced in rabbit reticulocyte lysates was able to phosphorylate histone HI. Together, our results suggest that ORF 9 encodes a protein kinase that is expressed during the beginning of the late and throughout the very late phases of AcMNPV infection.

Microplitis demolitor is a polydnavirus-carrying wasp that parasitizes the larval stage of Pseudoplusia includens. M. demolitor eggs are never encapsulated by host hae- mocytes when coinfected with its associated polydnavirus (MdPDV) whereas eggs are encapsulated within 36 h when injected into hosts without virus. In this study, infection of specific classes of P. includens haemocytes by MdPDY was examined. Electron microscopic studies indicated that MdPDV entered all haemocyte morphotypes. Northern blot analysis revealed that similar size classes of viral mRNAs were produced in granular cells, plasmatocytes and spherule cells. Expression of a 1·6 kb MdPDV mRNA in haemocytes from parasitized hosts was detectable by in situ hybridization at 2 h post-parasitism (p.p.) and continued through until day 6 p.p. By 12 h p.p., viral expression was detected in greater than 80% of the haemocytes in circulation but thereafter the percentage of haemocytes exhibiting a hybridization signal declined. Similar patterns were observed in haemocytes from larvae injected with calyx fluid or MdPDV plus venom. Granular cells and plasmatocytes from unparasitized larvae were purified on Percoll cushions and maintained in vitro. Both morphotypes were successfully infected with MdPDV and exhibited changes in morphology and adhesiveness very similar to cells from parasitized hosts. Cell-free plasma from parasitized larvae had a variable effect on haemocyte adhesion. Haemocytes cultured in plasma from 1 or 4 day p.p. larvae rapidly spread whereas cells cultured in 7 day p.p. plasma did not. Reciprocally, adhesion of haemocytes from parasitized larvae could not be rescued by cell-free plasma from unparasitized larvae. Together, these data suggest that disruption of the host encapsulation response is mediated primarily by direct infection of granular cells and plasmatocytes by MdPDV.

As a putative mechanism of hepatitis B virus (HBV) uptake into hepatocytes the interaction between HBV and the hepatic, human-derived asialoglycoprotein receptor (ASGPR) was investigated. Sera from patients with different variations of hepatitis B surface antigen-(HBsAg) positive chronic hepatitis, HBV particles isolated from HBV carriers with high-titre viraemia and commercial HBsAg served as sources of HBV. ASGPR was affinity-purified from human liver. HBV that had bound to isolated ASGPR was either detected by radio-immunoassay using solid-phase bound ASGPR or enzyme immunoassay with biotin-ASGPR bound to immobilized HBV. Furthermore, binding and uptake of purified, 125I-labelled HBV particles into human hepatoma cell lines (HepG2 and HuH7), which constitutively express functional ASGPR molecules, were compared to that of ASGPR-negative COS cells. As a result HBV was found to bind to purified human ASGPR in two different assays. Circulating virus particles from sera with high titre viraemia showed the highest attachment activity to ASGPR. HBV binding to purified ASGPR was saturable and inhibitable by an excess of d-galactose-bearing ligands, by EDTA and anti-receptor immunoglobulin. Lysis of particles by adding detergent abolished immunodetectable HBV binding to purified ASGPR. Commercial HBsAg did not adhere to solid phase-immobilized ASGPR. Monoclonal anti-preS1 antibody (MA18/7) but not anti-preS2 antibody (Q19/10) inhibited virus attachment. Purified and radiolabelled HBV particles showed binding to HepG2 and HuH7 cells but to much lesser degree to COS cells. Cellular binding of HBV was significantly inhibited by blocking of ASGPR function. Both ASGPR ligands and rabbit anti-ASGPR immunoglobulin but not non-immune rabbit serum inhibited uptake of radiolabelled HBV particles into HepG2 cells or HuH7 cells, respectively. This study suggests that HBV virions may enter human hepatocytes via ASGPR molecules by attachment of viral preS1-related envelope binding sites to this receptor.

In the serum of hepatitis B virus (HBV)-infected patients, two different types of particles, a 42 nm virion and a 22 nm subviral particle, were identified. The envelope of both particles is composed of three proteins, the large (L), middle (M), and major/small (S) surface proteins but the ratio between these components varies in each. The M protein appears in a lesser amount than the S protein in both virion and subviral particles, although it is translated from the same subgenomic RNA, and this is due to its poor initiation context of translation. In addition, only the glycosylated form of M protein is secreted in contrast to both glycosylated and unglycosylated forms of L and S proteins that are secreted. To investigate the biogenesis of M protein, human hepatoma cells transfected with plasmids containing a mutated HBV DNA were used to produce a high amount of M protein. Electron microscopic observation revealed that despite a higher proportion of the M protein being found in the transfected cells, the secreted surface antigen particles possess similar size and density to 22 nm subviral particles. Detailed biochemical analyses showed the following. (1) The unglycosylated M protein was predominantly present in the microsomal fraction but not present in any other subcellular fractions. (2) The M protein formed 22-nm-like particles in the endoplasmic reticulum (ER) and was retained in the post-ER or pre- Golgi regions. (3) In addition to the complex glycosylated form of M protein, a high-mannose form of M protein could be secreted. (4) Normally, no unglycosylated M protein was secreted. However, glycosylation was not essential for M protein secretion since M protein deprived of glycosylation by tunicamycin treatment was detected in the medium. These findings suggest that (i) the M protein was probably translated and co-translocated into the ER and at least one site was glycosylated before leaving the ER resulting in no secretion of unglycosylated M protein, and (ii) the M protein had two secretion pathways, one through the conventional pathway and the other probably directly through the ER.

The unique region of murine hepatitis virus (MHV) mRNA 5 has two open reading frames, ORF 5a and ORF 5b, that encode small proteins of unknown function. In the experiments described here, we have used the in vitro translation of synthetic mRNAs to examine the expression of these ORFs. Our results show that a synthetic mRNA containing both ORFs is functionally bicistronic. More importantly, the expression of ORF 5b, but not ORF 5a, is maintained in a trieistronic mRNA containing an additional 5′-proximal ORF. Thus, in the context of the MHV mRNA 5 unique region, the initiation of protein synthesis on ORF 5b can occur independently of ribosomes that enter from the 5′ end of the mRNA. We conclude that the translation of ORF 5b is mediated by the internal entry of ribosomes.

Deletion of the NR2 element of the long control region (LCR) of bovine papillomavirus type 4 (BPV-4) was observed previously to lead to a fivefold increase in enhancer activity of a subfragment of the LCR. Further characterization of this element indicates that mutations in NR2 lead to increased enhancer activity in both mouse CT3 fibroblasts and in a transformed bovine epithelial cell line derived from an alimentary canal papilloma/m situ carcinoma, but not in primary bovine keratinocytes. Since similar oligonucleotide-nuclear factor complexes were obtained in electrophoretic mobility shift assays (EMSA) for all three cell types, the observed difference in negative activity may result from variation in the NR2-binding factor itself between primary and established/transformed cell lines, or from the involvement of other factors that vary between the lines. Characterization of the NR2-binding factor by heat stability and antibody supershifts in EMSA indicate that the factor is related to the CCAAT/enhancer- binding protein (C/EBP) family, and that one component of the complexes may be C/EBP/β.