key: cord-0815930-5ugpul3v
authors: Sturman, Lawrence; Holmes, Kathyrn
title: The novel glycoproteins of coronaviruses
date: 1985-01-31
journal: Trends in Biochemical Sciences
DOI: 10.1016/0968-0004(85)90010-6
sha: e91f16906f632eccda51d69d8cc7242ead2827e3
doc_id: 815930
cord_uid: 5ugpul3v

Abstract The two glycoproteins of coronavirus envelopes have different functions in intracellular budding, cytopathic effects and virus infectivity, and have markedly different biochemical properties.

E2 may occur as intact virions are transported through intracellular compart-Glycoproteins of viral envelopes medi-Upon entry into the cell, the positive-ments. Pulse-labeling shows that most of ate many of the biological activities of strand RNA genome (Mr 5.4 x 106) the E2 is incorporated into virions and enveloped virions, such as attachment of apparently acts as messenger RNA released from the cell within 2 hours of virus peplomers (spikes) to receptors on (mRNA) for the synthesis of RNA-synthesis 19. A small proportion of E2 the cell surface, entry of virions into cells dependent RNA polymerase 7. This which is not incorporated into virions is by fusion of the viral envelope with cel-enzyme transcribes the genome, forming transported to the plasma membrane lular membranes, virus-induced cell a genome-sized, negative-strand tem-where it may participate in cell fusion fusion, and cellular and humoral plate s. From this template, new genomic and render cells susceptible to cell-mediresponses to viral infection. In this brief RNA, and six subgenomic mRNAs are ated cytotoxicity. This route for intra-, review we will summarize recent studies, transcribed 7,9. These capped and poly-cellular transport of E2 to the plasma on the envelope glycoproteins of coro-adenylated RNAs all share a common 3' membrane is the same as that of the naviruses, emphasizing the biochemical end and extend for different lengths in envelope glycoproteins of orthomyxo-, characteristics associated with their dif-the 5' direction, forming a set of seven paramyxo-, rhabdo-and alphaviruses. ferent roles in intracellular budding of overlapping transcripts 9. Recent studies Proteolytic cleavage of 180K E2 by virions, virus-induced cell fusion and by Lai et a/. ~°,n and Spaan et al. 12 show host-cell proteases to form products of immune responses to coronavirus that a leader RNA at least 70 M r 0.09 x 106 apparently occurs as a late infections, nucleotides in length, encoded by the 3' step in transport of virions and E2 to the A model for the structure of the end of the negative-strand template, is plasma membrane. Two such 90K specoronavirus virion is shown in Fig. 1 . added to the 5' end of each mRNA and cies, which we have called 90A and 90B, Coronaviruses are large, enveloped vir-genomic RNA. Although it is not yet can be separated by hydroxyapatite uses with helical nucleocapsids that con-certain how the leader RNA is added to chromatography 2°. These cleavage prodtain RNA genomes of positive or the noncontiguous sequences of the ucts have different amino acid composimessage-sense polarityL The viral envel-, mRNAs, the available evidence suggests tions, and only the 90A form is acylated. ope is a lipid bilayer with two viral--that the leader is synthesized separately On virions released from different cell encoded glycoproteins which we have and utilized as a primer for transcription types, the ratio of 180K E2 to 90K E2 called E1 and E2. The glycoprotein (Ref. 13 and M. Lai, personal communi-varies considerably, presumably reflectwhich forms the viral peplomers or cation). In vitro translation of the iso-ing differences in host-protease activity. spikes, E2, is similar to the spike glyco-, lated viral mRNAs suggests that only Cell fusion occurs in coronavirus proteins of other enveloped RNA vir-the gene at the 5' end of each mRNA infections in vivo and in vitro and is uses. In contrast, the membrane-species is translated, yielding a single mediated by E22. Host-dependent clearassociated glycoprotein, El, which polypeptide ~4- 16. age of E2 may be an important determiappears to function like the nonglycosyl-, nant of coronavirus virulence.

ated M proteins or orthomyxo-, para.. E2, the spike glycoprotein Proteolytic cleavage of E2 to 90K myxo-, and rhabdoviruses, differs Translation in vitro indicates that syn-forms is required for rapid cell fusion by markedly from other viral glycoproteins thesis of E2 is directed by mRNA 3 (Mr exogenous virus 2°. Treatment of virions in its structure, processing, and intra-2.6 x 106). E2 is synthesized on ribo-with trypsin, elastase, thermolysin or cellular transport. The major differences somes bound to the rough endoplasmic chymotrypsin activates cell-fusing acbetween the E1 and E2 glycoproteins of reticulum (RER) and is co-transla-tivity so that cells can be fused in one the A59 strain of murine hepatitis virus tionally glycosylated by transfer of hour in the absence of protein synthesis (MHV-A59) are summarized in Table I N-linked oligosaccharides from dolichol (F. Baker, C. Ricard and L. Sturman, and will be discussed in detail after a phosphate carders to asparagine resi-unpublished results). As sequence data summary of the virus replication dues :-5. The apoprotein appears to be become available from molecular clonstrategy, about Mr 0.12 x 106 as shown by experi-ing and amino acid sequencing, it will be ments in which attachment of the of considerable interest to compare the L. Sturman is not inhibited by tunicamycin, and the virus. In these cases, cleavage yields a microsomal membranes were added oligosaccharides are O-linked to serine new amino terminus with a markedly after protein synthesis had begun, and threonine residues at the amino terhydrophobic sequence of amino acids, Rottier et al. found that about 65-70% minus which projects from the external which may participate in the cell-fusion of the E1 molecule was synthesized side of the viral envelope. Niemann and reaction, before the protein was inserted into tl'/e his colleagues have shown that the struc-The E2 glycoprotein also appears to lipid bilayer 22, suggesting that an inter-tures of these O-glycosidally linked be responsible for the unusual nal signal sequence was used. Although oligosaccharides are identical to those pH-dependentthermolabilityofcorona-E1 lacks an amino terminal signal of glycophorin A, a major sialoviruses. There is a rapid loss of corona-sequence, the cellular signal recognition glycoprotein of the human erythrocyte virus infectivity above pH 6.5 associated particle (SRP) is required for integration membrane 24. E1 and one type of with aggregation of the peplomeric into microsomal membranes, since addi-glycophorin A possess the same amino glycoprotein. This suggests that there tion of SRP to a cell-free translation sys-terminal tetrapeptide sequence Ser-Sermay be a pH-dependent conformational tem blocks elongation of E1 (P. Rottier, Thr-Thr. This region of the glycophorin change in the E2 glycoprotein 21. The J. Armstrong, and D. Meyer, in press), molecule with its associated oligosacamino acid sequence of E2 will soon be The block is released by addition of charities constitutes the blood group M known from cloning and sequencing studies in progress in several labora- glycoproteins in many ways (Table I) . Neutralization (complement-Neutralization (complement Messenger RNA 6 (M r 0.9 x 106) which independent) dependent) encodes E1 can be translated in vitro in Cell-mediated cytotoxicity the absence of membranes. When determinant, and purified E1 also pos-distribution of E1 in these cells has been presence of tunicamycin z. sesses blood group M activity, characterized by indirect immuno-No details of the interaction between Intracellular transport of E1 differs fluorescence microscopy. Preliminary the E1 and E2 glycoproteins have yet from that of E2. E1 is transported to the results indicate that at this level of resolu-been elucidated. We surmise, however, Golgi apparatus where it is glycosylated tion the distribution of E1 is identical to that at a domain in the membrane where and accumulates in thiamine pyrophos-that of E1 in virus-infected ceils. It is clear E1 molecules have inserted and aggrephatase-positive cisternae (Ref. 25, and that the intracellular transport signal for gated, the E2 glycoprotein on the E. Dollar and K. Holmes, unpubfished E1 is not determined by the O-linked lumenal side of the RER or Golgi memresults). Unlike E2, E1 is not transported glycosylation of El, since the E1 glyco-branes may assemble into spikes perpento the plasma membrane. The limited protein of the avian coronavirus IBV, dicular to the lipid bilayer. The helical intracellular migration of E1 appears to which contains N-glycosidically linked nucleocapsid of the coronavirus probably determine the intracellular budding site of oligomannosidic side-chains, also binds to the cytoplasmic domain of the E1 coronaviruses. Virions mature by bud-accumulates in the Golgi apparatus 26. molecule, anchoring the flexible helix to ding in RER and Golgi membranes, pre-Comparison of the predicted amino acid the forming viral envelope. This interacsumably because this is the location of the sequences of MHV-A59 E1 and the IBV tion may stabilize the helix in a tight coil E1 glycoprotein. Pulse-chase studies glycoprotein reveals that the sequences in which inhibits transcription or translation show that early in infection the synthesis the membrane-spanning region are highly of the plus-stranded genomic RNA. The of E1 is balanced with its release in vir-conserved. Although their cytoplasmic nucleocapsids seen in the cytoplasm of ions. E1 is the only viral glycoprotein domains are much less alike, both are rich infected cells appear to be much more required for virus budding, since virions in basic residues zT. It is likely that studies flexible and possibly more loosely coiled containing only nucleocapsid (N) and E1 with hybrid E1 proteins or site-directed than the nucleocapsids in the virion. Thus proteins can be assembled and released in mntagenesis of the cloned E1 gene will E1 might play a role in inhibition of trantunicamycin-treated cells zl.

lead to identification of the signals on the scription and in stabilization of the virion-As a consequence of intracellular bud-E1 protein which determine its routte of associated nucleocapsid. E1 may also ding, complete virions may be formed intracellular transport, function in uncoating. Following fusion of within the cell prior to insertion of large

The chemical and physical properties the viral envelope with cell membranes, amounts of E2 glycoprotein into the of E1 are unusual, as might be expected the binding between E1 and viral plasma membrane. This would mean that from such a unique molecular structure, nucleocapsid may be released so that the enveloped virions are completed before E1 is a highly hydrophobic glycoprotein nucleocapsid can be used in the the infected cell becomes susceptible to which aggregates on heating at 100°C in cytoplasm for translation, and then for attack by humoral or cell-mediated SDS. E1 aggregates and binds to isolated transcription. immune responses directed against E2 on nucleocapsids at 37°(7. Aggregation of E1 the cell surface. Clearly intracellular bud-molecules with each other in the plane of Conclusion ding may provide an important mecha-the lipid bilayer may be important in vir-

The glycoproteins of coronaviruses are nism for persistent infection. Indeed, ion formation at the budding site. This significant both for understanding of viral coronaviruses cause persistent infections lateral serf-aggregation may be similar to functions and as models for glycoproteins both in vitro and in vivo.

that of the nonglycosylated M proteins of with different patterns of membrane To analyse the signals responsible for paramyxoviruses which has recently been insertion, glycosylation, and intracellular the intracellular transport of E1 to the reported 2s. E1 may form long tubules, 50 transport. The E2 glycoprotein most Golgi apparatus, cells have been microin-nm in diameter, which are found in the closely resembles the spike glycoproteins jected with purified mRNA 6 or trans-lumen of the RER in cells late in infection of enveloped viruses with negativefected with the cloned E1 gene and the with MHV, or in cells infected in the stranded RNA genomes in structure, 

) attack. Consequently, genetic deficiency, or the oxidation of its reactive centre in Adv

131-163 family of serine proteinase inhibitors. The specificity of each of these inhibitors is

primarily dependent on a single amino acid at its reactive centre. Site-directed Spaan

Human plasma contains several inhibi-emphysema. Extensive studies on the

Virology 132, tors of proteolytic enzymes that together molecular pathology of this genetic 250-260 form some 10% of its protein content I

A major example is the broad-spectrum completion of the structure 3'4 of the and

ever, the best studied of the inhibitors, structural studies is that ct

and the one present in greatest concen-proved to be the archetype of a new Patton

Although this venience, the serpins since it is primarily Stohlman

it is really an anti-elastase and its It is becoming clear that the lessons and Stohlman

of elastase released by neutrophil applicable to the other members of the Armstrong

633-640 nective tissue

and ter in the lungs -a condition known as molecule

Christchurch Clinical School of Medicine. Christ-design the specificity of

Carrell was awarded Specific purpose is already being de

chemical Society. His presentation lecture covered simple recombinant-DNA modifications 18 Niemann