key: cord-321162-pgd34ewv
authors: Holmes, Kathryn V.; Doller, Elizabeth W.; Sturman, Lawrence S.
title: Tunicamycin resistant glycosylation of a coronavirus glycoprotein: Demonstration of a novel type of viral glycoprotein
date: 1981-12-31
journal: Virology
DOI: 10.1016/0042-6822(81)90115-x
sha: 
doc_id: 321162
cord_uid: pgd34ewv

Abstract Tunicamycin has different effects on the glycosylation of the two envelope glycoproteins of mouse hepatitis virus (MHV), a coronavirus. Unlike envelope glycoproteins of other viruses, the transmembrane glycoprotein El is glycosylated normally in the presence of tunicamycin. This suggests that glycosylation of El does not involve transfer of core oligosaccharides from dolichol pyrophosphate intermediates to asparagine residues, but may occur by 0-linked glycosylation of serine or threonine residues. Synthesis of the peplomeric glycoprotein E2 is not readily detectable in the presence of tunicamycin. Inhibition of N-linked glycosylation of E2 by tunicamycin either prevents synthesis or facilitates degradation of the protein moiety of E2. Radiolabeling with carbohydrate precursors and borate gel electrophoresis of glycopeptides show that different oligcsaccharide side chains are attached to El and E2. The two coronavirus envelope glycoproteins thus appear to be glycosylated by different mechanisms. In tunicamycin-treated cells, noninfectious virions lacking peplomers are formed at intracytoplasmic membranes and released from the cells. These virions contain normal amounts of nucleocapsid protein and glycosylated El, but lack E2. Thus the transmembrane glycoprotein El is the only viral glycoprotein required for the formation of the viral envelope or for virus maturation and release. The peplomeric glycoprotein E2 appears to be required for attachment to virus receptors on the plasma membrane. The coronavirus envelope envelope glycoprotein E1 appears to be a novel type of viral glycoprotein which is post-translationally glycosylated by a tunicamycin-resistant process that yields oligosaccharide side chains different from those of N-linked glycoproteins. These findings suggest that El may be particularly useful as a model for studying the biosynthesis, glycosylation, and intracellular transport of 0-linked glycoproteins.

of the synthesis, glycosylation, and intracellular transport of glycoproteins is essential to understanding the structure and function of cell membranes and the role of oligosaccharides in glycoprotein processing and secretion. Because glycosylation and transport of viral envelope glycoproteins depend upon cellular processes, the G glycoprotein of vesicular stomatitis virus has been used as an excellent model for glycosylation and transport of N-linked glycoproteins (Roth-1 To whom reprint requests should be addressed. man and Lodish, 1977; Rothman et d, 1978; Morrison et aL, 19'78; Gibson et al, 1979) . Oligosaccharides may also be Olinked to serine or threonine residues of the polypeptide chain by a process which is less well understood (Sharon and Lis, 1981) . O-Linked oligosaccharides are predominant in many cell surface glycoproteins such as glycophorin (Tomita and Marchesi, 1975) and in secreted glycoproteins such as submaxillary mucins (Slomiany and Slomiany, 1978) .

Although tunicamycin inhibits glycosylation of N-linked glycoproteins (Takatsuki et al, 1971 ; Lehle and Tanner, 1976; Schwarz et al, 1976; Schwarz et al, 1979; Elbein, 1979; Schwarz and Datema, 1980) , no drug to inhibit O-linked glycosylation has yet been identified (Schwarz et al., 1979; Sharon and Lis, 1981) . All viral envelope glycoproteins studied to date have been of the N-linked type of glycoproteins (Leavitt et ok, 19'7'7; Morrison et cd, 19'78; Schwarz et aL, 1979; Nakamura and Compans, 1978a; Cash et d, 1980; Pizer et aL, 1980; Klenk and Rott, 1980; Choppin and Scheid, 1980; Ghosh, 1980; Stallcup and Fields, 1981) . We now present evidence that a coronavirus glycoprotein may be glycosylated by a different mechanism. This glycoprotein may serve as a useful model for the study of O-linked glycoproteins.

Coronaviruses are enveloped viruses containing -5.8 X lo6 daltons of positive sense, single-stranded polyadenylated RNA (Tyrrell et cd, 1978; Wege et cd, 1978; Lai and Stohlman, 1978; Macnaughton et d, 1978) . These viruses cause a variety of respiratory, enteric, or neurological diseases in animals and man (Andrewes et & 1978) . The virions of the A59 strain of mouse hepatitis virus contain three structural polypeptides: a phosphorylated nucleocapsid protein N, and two glycoproteins El and E2 which have several interesting properties (Sturman, 1977; Sturman and Holmes, 1977; Sturman et cd, 1980; Sturman, 1981) . The glycoprotein E2 forms the large petal-shaped peplomers characteristic of the coronavirus envelope. E2 is a 18OK-dalton glycoprotein which can be cleaved by trypsin to yield two 90K components. The glycoprotein El appears to be a transmembrane molecule with three domains: A glycosylated domain projects from the envelope, a second domain lies within the membrane, and a third domain appears to interact with the nucleocapsid inside the viral envelope. Unlike most proteins, when El is boiled in the presence of SDS and mercaptoethanol it aggregates into dimers, trimers, and tetramers.

The intracellular distribution of El is also unusual. Labeling with monospecific fluorescent antibody against isolated El or E2 (Sturman et cd, 1980) showed that El remains restricted to the perinuclear area of the cell while E2, like most other viral glycoproteins, migrates rapidly via intracellular membranes to the plasma membrane (Doller and Holmes, 1980) .

In the present study we have used the antibiotic tunicamycin to study the synthesis and glycosylation of the coronavirus MHV. Tunicamycin, an analog of UDP-Nacetylglucosamine, interferes with the formation of dolichol pyrophosphate-hracetylglucosamine which acts as a carrier for N-glycosidic linkage of core oligosaccharides to asparagine residues on glycoproteins. Tunicamycin interferes with the cotranslational glycosylation of glycoproteins (Takatsuki et al, 1971; Lehle and Tanner, 1976; Schwarz et al, 1979) . We have demonstrated that tunicamycin inhibits formation of the E2 glycoprotein of MHV, but does not prevent synthesis or glycosylation of the transmembrane glycoprotein El, formation of virions, or release of virions from cells. Our evidence indicates that the unique El glycoprotein of the coronavirus may be an O-linked glycoprotein, and thus could be a particularly useful model for studying the synthesis, glycosylation, and intracellular transport of O-linked glycoproteins in mammalian cells.

Virus propagation and putificatim The A59 strain of mouse hepatitis virus (MHV) was grown in the spontaneously transformed 17 Cl-l line of BALB/c 3T3 mouse fibroblasts as previously described (Sturman and Takemoto, 1972; Sturman, 1977) and assayed by plaque titration in 17 Cl 1 cells. To prepare radiolabled virus, cells in 150-mm2 plastic flasks were inoculated with 1.0 ml of MHV at a multiplicity of 0.3 to 3 PFU/cell in Dulbecco's modified Eagles' minimal essential medium, high glucose (DMEM;

Gibco, Grand Island, N. Y.), and incubated for 1 hr at 37". The inoculum was removed and cells were refed with 30 ml EMEM + 10% dialyzed fetal bovine serum (dFBS) containing 20 &i/ml of L -rH]amino acid mixture (New England Nuclear) and incubated for 24 hr. Virus released into the supernatant medium was harvested and purified by a modification of the method described previously (Sturman et aL, 1980) , using discontinous and continuous sucrose density gradients in TMS buffer (containing 0.05 M Tris-maleate and 0.1 M NaCl, pH 6.0) and omitting the polyethylene glycol precipitation step. L4zbeling and electrophoresis of intracellular viral pol~peptides. To study synthesis, processing, and release of viral specific polypeptides, pulse-chase experiments were performed. Confluent monolayers of 1'7 Cl 1 cells in 60-mm petri dishes (Falcon, Inc.) were preincubated for 18 hr at 37" in L-leucine deficient EMEM (leu-def MEM) with 10% dFBS and either mock infected with 0.5 ml/plate of leu-def MEM with 10% dFBS or inoculated with 3 to 5 PFU of MHV/cell in 0.5 ml/plate of leudef MEM with 10% dFBS. After incubation for 1 hr at 37", the inocula were removed, the cells were refed with leu-def MEM with 10% dFBS, with or without 0.5 pg/ml of tunicamycin (Eli Lilly, Indianapolis, Ind.), and the cultures were held at 37". Four hours prior to pulse labeling, 5 pg/ml of actinomycin D was added to the medium. At 8 and 10 hr after virus inoculation infected and control cells were pulse labeled for 15 min with 20 or 40 &i/ ml of rHlL-3, 4, 5-leucine (New England Nuclear, Inc.). The labeled medium was removed, cells were washed and refed with DMEM containing a lo-fold excess of unlabeled L-leucine, 10% FBS, 0.5 pg/ml tunicamycin. At intervals after the pulse, labeled intracellular polypeptides were extracted. Cells were washed twice in PBS and solubilized with 1.0 ml/plate of 1% Nonidet P40 (NP40, Accurate Chemical Corp.) in PBS. Nuclei and debris were removed by centrifugation at 1800 Q for 10 min at 4". Radiolabeled polypeptides were analyzed directly by polyacrylamide gel electrophoresis (PAGE) or were immunoprecipitated with rabbit antiserum against purified, detergent-disrupted virions in the presence of staphylococcal protein A prior to analysis by PAGE (Sturman et aL, 1980) . Samples for SDS-PAGE slab gels were heated at 37" for 30 min with an equal volume of sample treatment mixture composed of 6M urea, 4% SDS, and 0.05% bromphenol blue in 0.0625M Tris-chloride, pH 6.7. SDS-PAGE in cylindrical gels was performed using a high pH discontinuous buffer system and fractionation of gels with a Gilson automatic linear gel fractionator as previously described (Sturman and Holmes, 1977) . Five to twenty percent polyacrylamide gradient slab gels were prepared and fluorographed as previously described (Sturman et al 1980) .

Anal@s of glycopeptides. The glycoproteins El and E2 were isolated from SDS-PAGE of gradient-purified MHV which had been grown for 24 hr in medium containing 3 &i/ml [3Hlglucosamine (New England Nuclear, Inc.). Isolated El and E2 were eluted from the gels, digested with 15 pg of self-digested Pronase (Sigma) per milliliter for 30 hr at 60", concentrated by lyophilization, and analyzed as borate esters by PAGE in Tris-borate buffer at pH 8.3 according to the method of Weitzman et al. (1979) .

Electron microscopy. Concentrated, gradient-purified virions were prepared for electron microscopy using 2% phosphotungstic acid (PTA) at pH 7.2 on carboncoated, Formvar-covered, 400-mesh copper grids. Electron microscopy of cells was done by fixation with 1% glutaraldehyde followed by postfixation with 1% osmium tetroxide, dehydration in a graded series of ethyl alcohol solutions and propylene oxide, and embedding in Epon 812 resin as previously described (Compans et al, 1966) . Sections were stained with lead citrate and uranyl acetate and examined in a JEOL 1OOCX electron microscope.

The synthesis and processing of coronavirus-specific polypeptides were analyzed by pulse-labeling techniques in cells treated for 4 hr with actinomycin D to reduce cellular protein synthesis. Without tunicamycin, a 15-min pulse label with rH]leucine 8 hr after virus inoculation showed synthesis of the three structural polypeptides El, N, and E2 ( Fig. 1 , channels l-8). No high-molecular-weight polyproteins were detected. Thus each structural polypeptide appears to be translated inedpendently, as also shown by in vitro translation studies with isolated MHV mRNAs (Siddell et c& 1980; Rottier et cd, 1981) . During successive chase periods, no shift in the molecular weight of the nucleocapsid protein N or the peplomeric glycoprotein E2 was observed. Since the E2 glycoprotein is known to be extensively glycosylated (Sturman, 1977; Sturman and Holmes, 197'7) , this suggests either that glycosylation of E2 is a cotranslational event and/or that the glycosylated and nonglycosylated E2 are not resolved in this region of the gradient slab gel. Pulsechase experiments with labeling at 6,8, or 10 hr after infection show that El is synthesized as a 20K species and then chased up to a broad band of up to 23K (not apparent in Fig. 1 , channels 7 and 8 due to overexposure). Double labeling studies of intracellular viral polypeptides demonstrated that the 20K form of El is not glycosylated whereas the 23K species can be labeled with PHlglucosamine (data not shown). Antibody against El purified from NP40-disrupted virions by sucrose density gradient sedimentation immunoprecipitates the broad band of El from 20K to 23K (Sturman et cd, 1980) . Similar pulse-chase radiolabeling studies were performed on cells infected with MHV, treated with 0.5 pg/ml of tunicamycin, and labeled with [8H]leucine ( Fig.  1 , channels 9-16). Synthesis of the nonglycosylated nucleocapsid protein N was not affected by tunicamycin.

Synthesis of the 180K E2 glycoprotein was not detectable in the presence of tunicamycin.

Thus, tunicamycin either inhibits synthesis of E2 or facilitates rapid degradation of newly synthesized E2. Tunicamycin has been shown to interfere with synthesis or detection of several other glycoproteins which are cotranslationally glycosylated via N-linked glycosidic bonds (Lehle and Tanner, 1976; Leavitt et al. 1977; Olden et al. 1978; Gibson et aL, 1979; Stallcup and Fields, 1981) . The three polypeptides of molecular weight 70 to 90K seen in channels 9-16 are cellular polypeptides since they also appear with equal intensity and kinetics in uninfected cells treated with tunicamycin.

In the presence of tunicamycin the rate of synthesis of the membrane glycoprotein El is reduced (Fig. 1, channels 9 and 10) in comparison to the control (Fig. 1, channels 1 and 2) . However, it is clear that El is synthesized as a 20K polypeptide (Fig.  1, channels 9 and 10) which is chased into the glycosylated 23K form (Fig. 1, virions have not been observed budding from the plasma membrane. Formation of virions was not inhibited by 0.5 pg/ml of tunicamycin (Fig. 2b) . Numerous virions were observed in dilated cisternae of the RER and in smooth-walled vesicles of tunicamycin-treated cells. Although no virions were adsorbed to the plasma membrane (Fig. 2b) , virions were released from tunicamycin-treated cells. Although 0.5 rg/ml of tunicamycin reduced the 24-hr yield of infectious virus lOOO-fold, large quantities of virions could be purified from the medium by sucrose density gradient ultracentrifugation.

The number of virions released from tunicamycin-treated cultures was only about &fold less than from untreated cultures, as estimated by electron microscopy. The virions from tunicamycin-treated cells lacked the characteristic large peplomers of coronaviruses (Fig. 3) . The absence of peplomers correlated with the inability of the virions to attach to receptors on the cell surface (Fig. 2b) or to initiate infection.

The structural proteins of virions purified from tunicamycin-treated or control cultures were compared (Fig. 4) . The normal virus contained El, N, and three forms of the E2 glycoprotein: the native 180K form, the 90K cleavage products, and a high-molecular-weight aggregated form (Fig. 4A ). In contrast, virus grown with tunicamycin contained El and N in normal amounts but completely lacked the peplomeric glycoprotein E2 (Fig. 4B) . These biochemical data thus confirm the ultrastructural observations on the absence of peplomers on virions from tunicamycintreated cells.

The El membrane glycoprotein in the virus grown with tunicamycin appeared to be glycosylated normally, as shown by the ratio of glucosamine to methionine labels in the El peaks in Fig. 4A and B. This confirms the observation made in the pulse-chase experiments (Fig. 1 ) that glycosylation of El is resistant to inhibition by tunicamycin.

In Fig. 4 it is also apparent that the ratio of glucosamine to methionine label was not constant across the broad peak of El. More extensive glycosylation corresponded with decreased electrophoretic mobility.

The oligosaccharide chains of N-and Olinked glycoproteins differ markedly in size, diversity, and carbohydrate composition (Sharon and Lis, 1981) . Useful information about the oligosaccharides can be obtained by analyzing the glycopeptides isolated from glycoproteins by PAGE in Tris-borate buffer (Weitzman et ak, 1979) . At alkaline pH, neutral sugars form negatively charged complexes with borate. The number of borate ions which react with glycopeptides is a function of the composition, sequence, and linkages of the carbohydrates.

Glycopeptides of dissimilar carbohydrate composition and length exhibit different electrophoretic mobilities. Although this separation of glycopeptides is not based on their molecular weights, in general, larger oligosaccha-rides bind more borate than smaller ones, and therefore migrate faster in the gel. Differences in the peptide components appear to have little, if any, effect on the electrophoretic mobilities of the glycopeptides. Thus glycopeptides from two different N-linked glycoproteins would be expected to migrate rather similarly whereas glycopeptides from N-and O-linked glycoproteins would differ markedly in electrophoretic mobility and distribution.

We have used this technique to compare the glycopeptides derived by Pronase digestion of isolated El and E2. Figure 5 shows that the borate complexes of El and E2 glycopeptides exhibited markedly different electrophoretic patterns. The glycopeptides of El exhibited significantly greater mobility than those of E2. Therefore, the borate-glycopeptide complexes of ~I-IJGlucosamine-labeled glycoproteins El and E2 of MHV were separated by SDS-PAGE, eluted from the gels, and digested with Pronase for 30 hr at 30". The glycopeptides were lyophilized and analyzed by electrophoresis on 10% polyacrylamide gels in Tris-borate buffer, pH 8.3. Profiles of [SHlglucosamine-labeled glycopeptides from E2 (A) and El (B) are shown. The anode is to the right.

El were significantly more negatively charged than those of E2. Figure 5 also shows that fewer glycopeptide components were resolved from El than from E2. This suggests that El may have less diversity of oligosaccharide side chains than E2. These data suggest that the oligosaccharides of El differ markedly from those of E2, and support the hypothesis that the oligosaccharides of El and E2 may be derived by different mechanisms of glycosylation. DISCUSSION 

inhibits glycosylation of the structural glycoproteins of alphaviruses (Leavitt et d, 1977) , bunyaviruses (Cash et al, 1980) , herpes viruses (Pizer et al, 1980) , myxoviruses (Nakamura and Compans, 1978a; Klenk and Rott, 19X) ), paramyxoviruses (Stallcup and Fields, 1981) , rhabdoviruses (Morrison et a& 19'78; Gibson et a& 1979; Klenk and Rott, 1980) , and retroviruses (Schwarz et al, 1976; Diggelman, 1979) , suggesting that all of these viruses contain N-linked glycoproteins which are glycosylated by the transfer of core oligosaccharides from a dolichol pyrophosphate carrier to asparagine residues on the polypeptide.

Use of tunicamycin has often permitted the identification of the nonglycosylated protein moiety of a viral glycoprotein (Morrison et al, 1978 , Gibson et a& 1979 Nakamura and Compans, 1978a; Diggelman, 1979) . In some virus strains, however, synthesis of the nonglycosylated polypeptide in the presence of tunicamycin is difficult to detect because complete translation of the glycoprotein mRNA may be dependent on cotranslational addition of N-linked oligosaccharide chains, or because the nonglycosylated polypeptide may be highly susceptible to degradation by host cell proteases, or because the nonglycosylated polypeptide may be insoluble (Schwarz et al, 1976; Leavitt et aL, 1977; Gibson et d, 1979; Diggelman, 1979; Pizer et al, 1980; Stallcup and Fields, 1981) .

The coronavirus MHV contains two structural glycoproteins which have been isolated and partially characterized (Sturman et ak, 1980) . In 17 Cl 1 cells infected with the A59 strain of MHV, tunicamycin specifically interfered with the synthesis of the peplomeric glycoprotein E2 (Fig. 1) . Neither glycosylated nor nonglycosylated forms of E2 were detected either directly or by immunoprecipitation.

This suggests that E2, like other viral structural glycoproteins, may be an N-linked glycoprotein.

The transmembrane glycoprotein El or MHV is so far unique among viral structural glycoproteins in that it is glycosylated in the presence of tunicamycin.

Glycosylation of El appears to be a posttranslational event (Fig. 1 ) and the shift from the nonglycosylated 20K form to the glycosylated 23K form is not inhibited by tunicamycin.

This provides indirect evidence that El is not an N-linked glycoprotein but may be an O-linked glycoprotein. This hypothesis is also supported by direct evidence concerning the carbohydrate composition and the oligosaccharide side chains of El. Early studies on the incorporation of radiolabeled sugars into El and E2 showed that both El and E2 were labeled with [SHlglucosamine, but only E2 was labeled with ['Hlfucose (Sturman and Holmes, 1977) . Recent studies by Niemann and Klenk (1981) have identified additional differences between the carbohydrate composition of El and E2. El, like cellular O-linked glycoproteins (Thomas and Winzler, 1969; Spiro and Bhoyroo, 1974; Slomiany and Slomiany, 1978; Sharon and Lis, 1981) , contains little mannose, no fucose, and possesses a high proportion of N-acetyl galactosamine;

whereas E2, like many other N-linked glycoproteins (Nakamura and Compans, 197813; Prehm et al, 1979; Weitzman et uL, 1978; Burke and Keegstra, 1979) , contains both mannose and fucose but no N-acetyl galactosamine. Oligosaccharides of El but not E2 are removed by p-elimination (H. Niemann and H.-D. Klenk, personal communication) . Although the size, linkages, and sequences of sugars of the individual oligosaccharide chains of the two MHV glycoproteins have not yet been determined, analysis of the oligopeptides of El and E2 by Tris-borate PAGE has shown that the two glycoproteins have distinct oligosaccharide side chains (Fig. 5) . This technique affords resolution comparable to or better than Bio-Gel P6 gel filtration columns. The electrophoretic mobilities of the glycopeptides of E2 were similar to those of glycopeptides from N-linked glycoproteins of Sindbis virus or ovalbumin which have been previously described (Weitzman et al, 1979) .

In contrast, the glycopeptides of El were more negatively charged than those of E2. Thus, we have shown that El and E2 differ in carbohydrate composition, electrophoretie patterns of glycopeptides, and response to tunicamycin. These data suggest that E2 is an N-linked glycoprotein and El may be an O-linked glycoprotein.

The amino acid sequence and the locations and number of oligosaccharide side chains on the MHV glycoproteins

are not yet known. However, all of the oligosaccharide moieties of El appear to be located HOLMES, DOLLER, AND STURMAN near one end of the polypeptide chain. Treatment of MHV virions with bromelain or Pronase removed E2 and detached a 5K glycosylated polypeptide from El, leaving an 18K nonglycosylated polypeptide embedded within the viral envelope (Sturman and Holmes, 1977; Sturman, 1981) . It is not yet known whether the glycosylated portion of El which extends outside the viral membrane contains the amino terminal or carboxy terminal region of El. However, this region appears to contain multiple oligosaccharide side chains, since on PAGE El migrates either as a broad, diffuse band of 20K to 23K or, under other conditions, as three or more discrete bands.

The role of glycosylation in the functions of El is not yet known. Glycosylation may affect the conformation of El or its orientation in the viral envelope. However, glycosylation of El may not be essential for virion formation since both nonglycosylated and glycosylated El are incorporated into A59 virions (data not shown) and the El-like membrane proteins found in some other coronaviruses are apparently not glycosylated (Garwes, 1979) . The novel O-linked post-translational glycosylation of El may be associated with its restricted intracellular migration.

El is restricted to the perinuclear area of infected cells, in contrast to E2 which migrates rapidly to the plasma membrane (Doller and Holmes, 1980) . Possibly the oligosaccharides of El act as signals for movement from the RER to other membranes. Oligosaccharides are important recognition signals for enzymes destined for lysosomes (Hasilik, 1980) and glycoproteins destined for endocytosis (Baenziger and Fiete, 1980; Stahl and Schlesinger, 1980) .

The differential effects of tunicamycin on the synthesis and glycosylation of the two coronavirus envelope glycoproteins has permitted tentative assignment of functions to these glycoproteins. In MHVinfected cells treated with tunicamycin, virions which lack the E2 glycoprotein, and hence the peplomers, bud from intracytoplasmic membranes and are released from the cell (Figs. 2, 3, and 4) . Thus the transmembrane glycoprotein El appears to be the only envelope glycoprotein required for virus budding. Indeed, the location of El on intracytoplasmic membranes may determine the budding site of the coronavirus. Budding may occur where the viral nucleocapsid recognizes intracellular membranes altered by the addition of El. The E2 glycoprotein appears-to be essential for virus attachment to receptors on the plasma membrane. Virus lacking peplomers showed a markedly decreased infectivity.

No virions were observed adsorbed to the plasma membrane of tunicamycin-treated cells although numerous virions lacking E2 were present in the medium (Figs. 2 and 3) .

Many cellular glycoproteins such as fetuin, cornea1 proteoglycan, glycophorin, thyroglobulin, and immunoglobulins contain both asparagine-linked and serine-or threonine-linked oligosaccharides (Sharon and Lis, 1981) . Further detailed study may reveal some O-linked oligosaccharides on viral glycoproteins now believed to contain only N-linked oligosaccharides. Indeed, the vaccinia virus hemagglutinin, a nonstructural glycoprotein, has both 0-and N-linked oligosaccharide chains (Shida and Dales, 1981) . To date, however, no structural viral glycoprotein except the El glycoprotein of MHV has been identified which might contain only O-linked carbohydrates.

It is unlikely that glycosylation of El is a virus-encoded process. Since MHV has a limited amount of genetic material and codes for only six species of mRNA (Siddell et aL, 1980; Rottier et a& 1981) , it is far more likely that glycosylation of El occurs via cellular processes. Production of the El glycoprotein may be a good model system for the study of tunicamytin-resistant glycosylation.

In MHV-infected cells treated with actinomycin D to inhibit synthesis of cellular glycoproteins and with tunicamycin to inhibit synthesis of the N-linked viral glycoprotein E2, El is the major glycoprotein synthesized. This glycoprotein can be purified in large amounts from virions released from tunicamycin-treated cells. Thus, the coronavirus glycoprotein El may be a useful model for analysis of the synthesis, gly-cosylation, and intracellular transport of O-linked glycoproteins.

Viruses of Vertebrates

Galactoee and N-acetylgalactosamine specific endocytosis of glyeopeptides by isolated rat hepatocytes

Carbohydrate structure of Sindbis virus glycoprotein E2 from virus grown in hamster and chicken cells

The effect of glycosylation inhibitors on the maturation and intracellular polypeptide synthesis induced by snowshoe hare

The role of viral glycoproteins in adsorption, penetration and pathogenicity of viruses

An electron microscopic study of moderate and virulent virus-cell interactions of the parainfluenza virus SV5. virdogy 30,411~422

Different Intracellular Transportation of the Envelope Glycoproteins El and E2 of the Coronavirus MHV

The role of lipid-linked saccharides in the biosynthesis of complex carbohydrates

Viral Enteritis in Humans and Animals

Synthesis and maturation of glycoproteins of enveloped animal viruses

The nonglycosylated glycoprotein of vesicular stomatitis virus is temperature-sensitive and undergoes intracellular aggregation at elevated temperatures

Cotranslational and posttranslational processing of viral glycoproteins

RNA of mouse hepatitis virus

Impaired intracellular migration and altered solubility of nonglycosylated glycoproteins of vesicular stomatitis virus and Sindbis virus

The specific site of tunicamycin inhibition in the formation of dolichol-bound N-acetylglucosamine derivatives

The genomes of three coronaviruses

Assembly of viral membranes: Maturation of the vesicular stomatitis glycoprotein in the presence of tunicamycin

Effects of glucosamine, 2-deoxy-D-glucose and tunicamytin on glycosylation, sulfation and assembly of influenza viral proteins

Glycopeptide components of influenza viral glycopeptides

The Biochemistry and Biology of Coronaviruses

Role of carbohydrates in protein secretion and turnover: Effects of tunicamycin on the major cell surface glycoprotein of chick embryo fibroblasts

Effect of tunicamycin on herpes simplex virus glycoproteins and infectious virus production

The carbohydrate structure of the glycoproteins of the paramyxovirus SV5 grown in bovine kidney cells

Glycosylation of a membrane protein is restricted to the growing polypeptide chain but is not necessary for insertion as a transmembrane protein

Synchronized transmembrane insertion and glycoaylation of a nascent membrane protein

Translation of three mouse hepatitis virus strain A59 subgenomic RNAs in Xenopus ti oocytes

Inhibitors of protein glycosylation

Suppression of glycoprotein formation of Semliki Forest, influenza and avian sarcoma virus by tunicamycin

Inhibition of glycosylation of viral glycoproteins

Glycoproteins; research booming on long-ignored, ubiquitous compounds

Biogenesis of vaccinia: Carbohydrate of the hemagglutinin molecule

Coronavirus JHM: Cell free synthesis of structural protein p60

Structures of the acidic oligosaccharides isolated from rat sublingual glycoprotein

Structure of the 0-glycosidically linked carbohydrate units of fetuin

Receptor-mediated pinocytosis of mannose/N-acetylglucosamine-terminated glycoproteins and lysosomal enzymes by macrophages

The replication of measles virus in the presence of tunicamycin

Characterization of a coronavirus. 1. Structural proteins: Effects of preparative conditions on the migration of protein in polyacrylamide gels

The Biochemistry and Biology of Coronaviruses

Characterization of a coronavirus. II. Glycoproteins of the viral envelope; tryptic peptide analysis

Isolation of coronavirus envelope glycoproteins and interaction with the viral nucleocapsid

Enhanced growth of a murine coronavirus in transformed mouse cells

Structural studies on human erythrocyte glycoproteins

Amino-acid sequence and oligosaccharide attachment sites of human erythrocyte glycophorin

Genomic RNA of the murine coronavirus

Analysis of glycopeptides as borate complexes by polyacrylamide gel electrophoresis

We thank Margaret Kerchief, Barbara O'Neill, Cynthia Ricard, Gale Schmidt, and Eugene Downing for excellent technical assistance. We are grateful to Dr. H.-D. Klenk and H. Niemann for making their data available to us in advance of its publication. This research was supported in part by Research Grant R67403 from the Uniformed Services University of the Health Sciences. The opinions expressed are the private views of the authors and should not be construed as official or necessarily reflecting the views of the Uniformed Services University School of Medicine or the Department of Defense. This work was presented in part at a conference on The Biochemistry and Biology of Coronaviruses in Wiirzburg, West Germany, in October, 1989.