Processing of the envelope glycoproteins of pestiviruses.

The genomic RNA of pestiviruses is translated into a large polyprotein that is cleaved into a number of proteins. The structural proteins are N terminal in this polyprotein and include three glycoproteins called E0, E1, and E2 on the basis of the order in which they appear in the polyprotein. Using pulse-chase experiments, we show that a pestiviral glycoprotein precursor, E012, is formed that is processed into E0, E1, and E2 in an ordered fashion. Processing is initiated by a nascent cleavage between the capsid and the translocated E012 followed by cleavage at the C terminus of E2. E012 is then rapidly cleaved to form E01 and E2. After E2 is released from the precursor, E01 is processed into E0 and E1. To identify the sites of cleavage, the N termini of the glycoproteins of the pestivirus classical swine fever virus (formerly termed hog cholera virus) were sequenced after expression in the vaccinia virus system. The N termini are Glu-268 for E0 (gp44/48), Leu-495 for E1 (gp33) and Arg-690 for E2 (gp55). The sequences around the cleavage sites capsid/E0 and E1/E2 conform to the rules known for cellular signal proteases, as does the sequence at the presumed C terminus of E2. The sequence upstream of the E0/E1 cleavage site also shows sequence characteristics of signalase processing sites but lacks the typical hydrophobic signal peptide; this cleavage site has characteristics in common with a site in flaviviruses that is also cleaved in a delayed fashion. The absence of any membrane-spanning region results in the shedding of E0 by infected cells, and E0 can be detected in the virus-free supernatant. Comparison of the sequences around the cleavage sites of pestiviruses suggests a general processing scheme for the structural glycoproteins. Comparison of the pesti- and flaviviral structural glycoproteins suggests analogies between E012 and prM-E.