Localization of viral structural proteins in the cytoplasm and nucleus of Rous-associated virus-2-infected chicken embryo fibroblasts

The cellular location of viral structural proteins was carried out by immunohistochemistry and by cell fractionation. Antibody against the structural protein p27 was used in immunohistochemical reactions to demonstrate the presence of viral proteins in the cytoplasm and nucleus of Rous-associated virus 2-infected chicken cells. Localization in the nucleus was found over heterochromatic regions; in the cytoplasm it was found in discrete particulate structures. These observations were extended in cell fractionation studies in which cytoplasmic and nuclear fractions were immunoprecipitated with antibody against the viral structural proteins.     

Nucleotide sequence of noncoding regions in Rous-associated virus-2: comparisons delineate conserved regions important in replication and oncogenesis.

Fujinami sarcoma virus (FSV) encodes a 140,000-dalton transforming protein, P140, which contains gag- and fps-specific sequences. The cellular localization of this protein was examined by fractionation of [35S]methionine-labeled, FSV-infected chicken embryo fibroblasts. In homogenates of cells infected by wild-type, temperature-resistant FSV prepared in either hypotonic or isotonic buffer, 60 to 80% of the P140 was particulate. Isopycnic separation on discontinuous sucrose gradients indicated that the majority of the particulate P140 was present in a light membrane fraction enriched for plasma membranes. Much of the particulate P140 could be solubilized by the addition of 0.6 M salt to a postnuclear supernatant, suggesting that P140 is not an integral membrane protein. Particulate P140 may be associated with membranes either directly as a peripheral membrane protein or indirectly via cytoskeletal elements. In cells infected by mutants of FSV temperature sensitive for cellular transformation, most of the P140 is particulate at the permissive temperature, whereas most is soluble at the nonpermissive temperature; this change in distribution is not a secondary consequence of the change in cellular phenotype, since it also occurs in nonconditionally transformed cells doubly infected with temperature-sensitive FSV and wild-type Rous sarcoma virus. The movement of P140 from the particulate to the soluble fraction occurs rapidly when cells infected by temperature-sensitive FSV are shifted from the permissive to the nonpermissive temperature. Furthermore, P140 moves from the soluble to the particulate fraction, although somewhat more slowly, when cells are shifted from the nonpermissive to the permissive temperature. These observations suggest that the association of P140 with plasma membranes or the cytoskeleton may play a role in transformation by FSV.     

Embryonic infection with the endogenous avian leukosis virus Rous-associated virus-0 alters responses to exogenous avian leukosis virus infection.

We inoculated susceptible chicken embryos with the endogenous avian leukosis virus Rous-associated virus-0 (RAV-0) on day 6 of incubation. At 1 week after hatching, RAV-0-infected and control chickens were inoculated with either RAV-1 or RAV-2, exogenous viruses belonging to subgroups A and B, respectively. The chickens injected with RAV-0 as embryos remained viremic with exogenous virus longer and either failed to develop type-specific humoral immunity to exogenous virus or developed it later than the control chickens not inoculated with RAV-0. The RAV-0-injected chickens also developed neoplasms at a much higher frequency than did the control chickens. We suggest that the lower immune responses of the RAV-0-injected chickens were due to an immunological tolerance to envelope group-specific glycoproteins shared among endogenous and exogenous viruses.     

Localization of two cellular forms of the vesicular stomatitis viral glycoprotein.

Two cell-associated forms of the glycoprotein (G) of vesicular stomatitis virus, termed G1 and G2, have been resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. G1 has the higher electrophoretic mobility, but both forms migrate more slowly than G protein synthesized in a wheat germ cell-free system (G0), which presumably is the unglycosylated form. G1 is a kinetic precursor of the G2 form, and the apparent cause of the electrophoretic difference between the two species is the presence of N-acetylneuraminic acid on the G2 form. Conversion of G1 to G2 occurs 10 to 20 min prior to the appearance of the G2 form of the protein on the cell surface. This suggests that the G protein may be completely glycosylated several minutes prior to its migration to the cell surface and that glycosylation is not the limiting step in its maturation. No glycoprotein comigrating with G0 can be detected in the infected cells, even after 5-min labeling periods; this suggests that partial clycosylation of G occurs concomitantly with or immediately after its synthesis.     

Synthesis of viral proteins by the avian myeloblastosis viral core component.

Avian Myeloblastosis Viral (AMV) core component was isolated and shown to synthesize AMV proteins in vitro. This reaction was linearly dependent on viral core concentration, proceeded linearly with time, and was inhibited by puromycin and aurintricarboxylic acid. The proteins synthesized in vitro co-electrophoresed and co-chromatographed with known proteins, and were immunoprecipitated by total and monospecific antibodies to known AMV proteins.     

Coronavirus glycoprotein E1, a new type of viral glycoprotein

The carbohydrate contents of coronavirus glycoproteins E1 and E2 have been analyzed. E2 has complex and mannose-rich-type oligosaccharide side-chains, which are attached by N-glycosidic linkages to the polypeptide. Glycosylation of E2 is initiated at the co-translational level, and it is inhibited by tunicamycin, 2-deoxy-glucose, and 2-deoxy-2-fluoro-glucose. Thus, E2 belongs to a glycoprotein type found in many other enveloped viruses. E1, in contrast, represents a different class of glycoprotein. The following observations indicate that its carbohydrate side-chains have 0-glycosidic linkage. (1) The constituent sugars of E1 are N-acetylglucosamine, N-acetylgalactosamine, galactose, and neuraminic acid; mannose and fucose are absent. (2) The side-chains can be removed by β-elimination. (3) Glycosylation of E1 is not sensitive to the compounds interfering with N-glycosylation. E1 is the first viral glycoprotein analyzed that contains only 0-glycosidic linkages. Coronaviruses are therefore a suitable model system to study biosynthesis and processing of this type of glycoprotein.     

A dramatic shift in the transmembrane topology of a viral envelope glycoprotein accompanies hepatitis B viral morphogenesis.

The envelope of hepatitis B virus contains three related glycoproteins (termed L, M and S) produced by alternative translation initiation in a single coding region. The smallest of these, the S protein, is a 24 kDa glycoprotein with multiple transmembrane domains. The M and L proteins contain the entire S domain at their C-termini, but harbor at their N-terminal additional (preS) domains of 55 or 174 amino acids, respectively. Most of these preS residues are displayed on the surface of mature virions and hence would be expected to be translocated into the endoplasmic reticulum (ER) lumen during biosynthesis. Using a coupled, in vitro translation/translocation system we now demonstrate that, contrary to expectation, virtually all preS residues of the L protein are cytoplasmically disposed in the initial translocation product. This includes some preS sequences which in the M protein are indeed translocated into the ER lumen. Since preS sequences are found on the external surface of the virion envelope, our results indicate that during or following budding a dramatic reorganization of either the envelope proteins or the lipid bilayer (or both components) must occur to allow surface display of these sequences. These findings imply that some membrane budding events can have remarkable and previously unsuspected topological consequences.     

Synthesis and processing of glycoprotein gG of herpes simplex virus type 2.

Monoclonal antibody 13 alpha C5-1-A11 immunoprecipitated two major polypeptides of molecular weights 108,000 and 120,000 from extracts of herpes simplex virus type 2-infected BHK-21 cells labeled with [35S]methionine or [3H]glucosamine. In pulse-chase experiments, both labels were chased from the 120,000-molecular-weight peptide (120K peptide) into the 108K molecule. Endoglycosidase H (endo H) reduced the 120K peptide to a 112K peptide but did not affect the 108K peptide. Similar profiles were obtained with monoclonal antibody AP-1 which reacts with a 92K glycoprotein, gG, which maps to the short unique region of the genome. Cross-absorption experiments indicated that both antibodies reacted with the same peptides, suggesting that the 120K peptide is a partially glycosylated high-mannose-type precursor of gG (pgG1). Immunoprecipitation from monensin-treated cells indicated that pgG1(120K) may undergo peptide cleavage to form a 74K high-mannose-type peptide (pgG2) and that this 74K peptide may be further processed into an endo H-resistant 110K to 116K peptide. In the presence of tunicamycin, gG(108K) was replaced by 110K and 105K peptides which were resistant to both endo H and endoglycosidase F. The 105K peptide was the only molecule labeled by [3H]galactose or [3H]glucosamine in the presence of tunicamycin, and none of the peptides were labeled with [3H]mannose, indicating the probable presence of O-linked sugars in the 105K peptide. Our results imply that cotranslational glycosylation of the unglycosylated precursor 110K peptide results in the high-mannose-type pgG1(120K), which probably undergoes peptide cleavage. This putative cleavage product may then mature into gG (108K) by the trimming of sugars and the addition of complex and probably O-linked sugars; the high-mannose-type pgG2(74K) is probably an intermediate peptide formed in this process.     

Dissociation and reassociation of oligomeric viral glycoprotein subunits in the endoplasmic reticulum.

The vesicular stomatitis virus (VSV) glycoprotein (G) forms noncovalently linked trimers in the endoplasmic reticulum (ER) prior to transport to the cell surface. Here we examined the formation of heterotrimers between wild-type and mutant subunits that were retained in the ER by C-terminal retention signals. When G protein was coexpressed with mutant subunits that formed trimers at the wild-type rate and were transported from the ER at the wild-type rate, heterotrimers were readily detected. In contrast, when G protein was coexpressed with mutant subunits that formed trimers at the wild-type rate, but were retained in the ER, heterotrimers were not detected unless transport of the wild-type molecules from the ER was blocked. After removal of transport block, the heterotrimers then dissociated and reassorted to homotrimers of the mutant protein that were retained in the ER and wild-type trimers that were transported to the cell surface. These and other results presented here indicate that there is an equilibrium between G protein trimers and monomers in vivo, at least in the ER. This equilibrium may function to allow escape of wild-type subunits from aberrant retained subunits.     

Fatty acid binding to vesicular stomatitis virus glycoprotein: a new type of post-translational modification of the viral glycoprotein.

The glycoprotein (G) of vesicular stomatitis virus (VSV) binds 1–2 moles of fatty acid per mole of protein. The fatty acids cannot be released by repeated extractions of the protein with organic solvents, nor can they be released by denaturing the protein with ionic or nonionic detergents. Pronase digestion of G yields an organic extractable fragment that contains bound fatty acid. The fatty acid is quantitatively released from this fragment and from intact G by mild alkali treatment in methanol and is identified by gas-liquid and thin-layer chromatography as, predominantly, the methyl ester of palmitic acid. Insignificant amounts of phosphate are found in G, thus ruling out the presence of bound phospholipid. Chicken embryo fibroblast pre-labeled with ³H-palmitate and then infected with VSV for 4 hr show the presence of ³H label in G but not in other viral structural proteins. The ³H label is present only in the fatty acid moiety of the protein. Much smaller amounts of ³H fatty acid are bound to G protein formed by the VSV mutant ts045 grown at the nonpermissive temperature, and no ³H fatty acid is bound to G synthesized at 37°C in cells pretreated with tunicamycin, an inhibitor of glycosylation. However, infection with the VSV-Orsay strain at 30°C in the presence of tunicamycin allows for production of VSV particles with nonglycosylated G (Gibson, Schlesinger and Kornfeld, 1979), and this G has the same proportion of the fatty acid as does the normal glycosylated G. These data indicate that fatty acids become covalently attached to the G polypeptide chain during maturation of the protein—perhaps as the glycoprotein moves to the cell's plasma membrane.     

Differences between the endogenous and exogenous DNA sequences of Rous-associated virus-O.

DNA sequences related to the endogenous retrovirus of chickens, Rous-associated virus-O (RAV-O), have been examined using site-specific DNA endonuclease analysis of cellular DNA derived from line 15 and line 100 chickens. Individual embryos from both inbred lines were used as a source of embryonic fibroblasts from which cellular DNA was isolated. Analysis of DNA containing either endogenous RAV-O sequences alone or both endogenous and exogenous RAV-O sequences produced identical patterns of RAV-O-specific DNA fragments after digestion with the endonucleases Eco RI, Hind III, BgI II, Bam HI or Xho I. Similar analysis with endonucleases Hinc II or Hha I, however, produced several RAV-O-specific DNA fragments which were derived from cellular DNA containing both endogenous and exogenous RAV-O sequences but not from cellular DNA containing only endogenous sequences. Although some differences exist between the DNA fragments specific for the endogenous viral sequences of line 15 and line 100 cellular DNA, the DNA fragments specific for the exogenous viral sequences were identical between the two inbred lines. Cleavage of an unintegrated linear RAV-O DNA molecule with Hinc II or Hha I produced DNA fragments identical to those specific for the exogenously acquired RAV-O provirus. This suggests that these characteristic fragments contain no cellular DNA. The potential DNA junction fragments containing both viral and cellular DNA, identified after analysis of DNA that contains both endogenous and exogenous viral sequences, were identical to those observed after analysis of DNA containing only endogenous viral sequences. These results support the following conclusions. First, exogenous proviral sequences are integrated into chicken cell DNA following an interaction between viral and cellular DNA that is specific with respect to the virus and nonspecific with respect to the cell. Second, both the free linear RAV-O DNA intermediate and the newly integrated exogenous provirus contain specific endonuclease sites that are not found in endogenous RAV-O DNA sequences. These results suggest that the formation of the exogenous DNA provirus involves specific alteration of the endogenous viral DNA sequences before reinsertion of the sequences as the exogenous RAV-O DNA provirus. It is possible that newly integrated exogenous RAV-O sequences are characterized by specific differences in the pattern of base methylation and a limited sequence arrangement.     

Similarities and differences in the Fc-binding glycoprotein (gE) of herpes simplex virus types 1 and 2 and tentative mapping of the viral gene for this glycoprotein.

We performed affinity chromatography and immunoprecipitation experiments to determine whether cells infected with herpes simplex virus type 2 (HSV-2) expressed a glycoprotein that was functionally and antigenically related to the HSV-1 Fc-binding glycoprotein designated gE. We found that a protein from extracts of HSV-2-infected HEp-2 cells bound specifically to an Fc affinity column and that the electrophoretic mobility of this protein in sodium dodecyl sulfate-acrylamide gels was slightly less than the mobility of HSV-1 gE. Immunoprecipitation experiments performed with an antiserum prepared against HSV-1 gE revealed that (i) extracts from HSV-2-infected cells contained a glycoprotein that was antigenically related to HSV-1 gE; (ii) the electrophoretic mobility of the HSV-2 gE was indistinguishable from the mobility of the HSV-2 Fc-binding protein; (iii) the antiserum reacted with both newly synthesized transient forms and stable fully processed forms of both HSV-1 gE and HSV-2 gE; and (iv) the transient and stable forms of HSV-2 gE all had lower electrophoretic mobilities than their HSV-1 counterparts. Electrophoretic analyses of gE precipitated from extracts of HEp-2 cells infected with two sets of HSV-1 x HSV-2 intertypic recombinant viruses suggested that the gene for gE is located at the right end of the HSV genome (0.85 to 0.97 map units) in the unique portion of the S component.     

Fusion to C3d enhances the immunogenicity of the E2 glycoprotein of type 2 bovine viral diarrhea virus

The use of DNA and protein subunit vaccines in animals provides an opportunity to introduce vaccines that are arguably the safest that can be developed. For that reason, considerable effort is under way to devise methods of enhancing the immunogenicity of such vaccines. Seven years ago it was shown that fusing complement fragment C3d to hen egg lysozyme (HEL) enhanced the immunogenicity of HEL 10,000-fold. Based on this observation, we decided to evaluate the effect of C3d on the immunogenicity of the E2 protein of bovine viral diarrhea virus (BVDV). E2 is the major target of neutralizing antibody during BVDV infection. To test the effect of C3d on E2 immunogenicity, expression cassettes encoding a secreted form of E2 alone (E2s) or E2 fused to three copies of murine C3d (E2s-C3d) were constructed. The proteins were purified from the supernatants of transfected cells and used to immunize mice. The immune response was monitored by an enzyme-linked immunosorbent assay (ELISA) for E2s-specific antibody and by a virus neutralization test. The ELISA results indicated that the E2s-C3d protein is 10,000-fold more immunogenic than the E2s protein alone. The maximum primary immune response was elicited with <0.1 microg of E2s-C3d protein without an adjuvant. In addition, we have shown for the first time that high levels of anti-E2s and neutralizing antibodies can be elicited when this same low concentration of E2s-C3d is used to both prime and boost the immune response. We conclude that the E2s-C3d fusion protein has significant potential as a subunit vaccine against BVDV infection.     

Characterization of the receptor-binding domain of Ebola glycoprotein in viral entry

Ebola virus infection causes severe hemorrhagic fever in human and non-human primates with high mortality. Viral entry/infection is initiated by binding of glycoprotein GP protein on Ebola virion to host cells, followed by fusion of virus-cell membrane also mediated by GP. Using an human immunodeficiency virus (HIV)-based pseudotyping system, the roles of 41 Ebola GP1 residues in the receptor-binding domain in viral entry were studied by alanine scanning substitutions. We identified that four residues appear to be involved in protein folding/structure and four residues are important for viral entry. An improved entry interference assay was developed and used to study the role of these residues that are important for viral entry. It was found that R64 and K95 are involved in receptor binding. In contrast, some residues such as I170 are important for viral entry, but do not play a major role in receptor binding as indicated by entry interference assay and/or protein binding data, suggesting that these residues are involved in post-binding steps of viral entry. Furthermore, our results also suggested that Ebola and Marburg viruses share a common cellular molecule for entry.     

The cytoplasmic domain of alphavirus E2 glycoprotein contains a short linear recognition signal required for viral budding.

Intracellular alphavirus nucleocapsids express a binding site for the cytoplasmic domain of the viral E2 spike glycoprotein. This binding site is recognized by the anti-idiotype monoclonal antibody, F13. The monoclonal anti-anti-idiotype antibody, raised against F13 and designated 3G10, recognizes the carboxy-terminal eight residues of the E2 cytoplasmic domain in Semliki Forest virus (SFV), identifying this as the signal for nucleocapsid interaction. F13 binding to cells infected with SFV or a second alphavirus, Sindbis virus, is inhibited by a synthetic peptide corresponding to the entire 31 residue cytoplasmic domain (E2c), and also by a synthetic peptide corresponding to the eight residue epitope recognized by 3G10. Both E2c and the eight residue peptide inhibited viral budding in microinjection experiments and when conjugated to colloidal gold are bound specifically to nucleocapsids in infected cells. These results identify a short linear signal in the E2 cytoplasmic domain required for the interaction with nucleocapsids which leads to budding of at least two alphaviruses from infected cells.     

Selective isolation of mutants of vesicular stomatitis virus defective in production of the viral glycoprotein.

We describe a procedure that enriches for temperature-sensitive (ts) mutants of vesicular stomatitis virus (VSV), Indiana serotype, which are conditionally defective in the biosynthesis of the viral glycoprotein. The selection procedure depends on the rescue of pseudotypes of known ts VSV mutants in complementation group V (corresponding to the viral G protein) by growth at 39.5 degrees C in cells preinfected with the avian retrovirus Rous-associated virus 1 (RAV-1). Seventeen nonleaky ts mutants were isolated from mutagenized stocks of VSV. Eight induced no synthesis of VSV proteins at the nonpermissive temperature and hence were not studied further. Four mutants belonged to complementation group V and resembled other ts (V) mutations in their thermolability, production at 39.5 degrees C of noninfectious particles specifically deficient in VSV G protein, synthesis at 39.5 degrees C of normal levels of viral RNA and protein, and ability to be rescued at 39.5 degrees C by preinfection of cells by avian retroviruses. Five new ts mutants were, unexpectedly, in complementation group IV, the putative structural gene for the viral nucleocapsid (N) protein. At 39.5 degrees C these mutants also induced formation of noninfectious particles relatively deficient in G protein, and production of infectious virus at 39.5 degrees C was also enhanced by preinfection with RAV-1, although not to the same extent as in the case of the group V mutants. We believe that the primary effect of the ts mutation is a reduced synthesis of the nucleocapsid and thus an inhibition of synthesis of all viral proteins; apparently, the accumulation of G protein at the surface is not sufficient to envelope all the viral nucleocapsids, or the mutation in the nucleocapsid prevents proper assembly of G into virions. The selection procedure, based on pseudotype formation with glycoproteins encoded by an unrelated virus, has potential use for the isolation of new glycoprotein mutants of diverse groups of enveloped viruses.