Cell, Vol. 45, 761-770, June 6, 1966, Copyright 0 1966 by Cell Press A Second Virus-Encoded Pmteinase Involved in Pmteolytic Processing of Poliovirus Polyprotein Haruka Toyoda,’ Martin J. H. Nicklin,’ Michael G. Murray,’ Carl W. Anderson,t John J. Dunn,t F. William Studier,t and Eckard Wimmer’ * Department of Microbiology State University of New York at Stony Brook Stony Brook, New York 11794 t Department of Biology Brookhaven National Laboratory Upton, New York 11973 Summary The poliovirus polyprotein is cleaved at three different amino acid pairs. Viral polypeptide 3C is responsible for processing at the most common pair (glutamine- glycine). We have found that a cDNA fragment encod- ing parts of the capsid protein region (Pi) and the non- structural protein region (P2), and including the Pl-P2 processing slte (tyrosine-glycine), can be expressed in E. COIL The translation product was correctly processed. Disruption of the coding sequence of 2A, a nonstructural polypeptide mapping carboxy-termi- nal to the tyrosine-glyclne cleavage site, by linker mutagenesis or deietlon, prevented processing. De- letion of the adjacent polypeptide 28 had no such effect. Antibodies against 2A specifically inhibited processing at the 3cC3D’ processing site (tyrosine- glyclne) in vitro. We conclude that poliovirus encodes the second proteinase 2A, which processes the poly- protein at tyrosine-glycine cleavage sites. Introduction Proteolytic processing of virus-encoded translation prod- ucts is common in animal virus replication, but only the picornaviruses produce all of their gene products by pro- teolytic cleavage of a single polyprotein precursor (for re- cent reviews, see Toyoda et al., 1966; Nicklin et al., 1966). In the case of poliovirus, a member of a family of nonen- veloped plus-stranded RNA viruses (the Picornaviridae), only three types of cleavages occur during proteolysis: one at an asparagine-serine (N-S) amino acid pair, two at tyrosine-glycine (Y-G) pairs, and eight or nine at glutamine-glycine (Q-G) pairs (Kitamura et al., 1961; Sem- ler et al., 1961a, 1961b, 1963; Larsen et al., 1962; Emini et al., 1962). At least one virus-encoded proteinase of picor- naviruses has been known for some time (see Pelham, 1976), but its specificity was proved only recently by biochemical and genetic means (Hanecak et al., 1962, 1964). On the other hand, a wealth of arguments and some circumstantial evidence were used to conclude that cellular enzymes must also be involved in proteolytic pro- cessing of the polyprotein (Korant et al., 1960; Burroughs et al., 1964; for earlier work, see review by Adler et al., 1961). In poliovirus replication, the virus-encoded polypeptide 3C cuts the polyprotein at Q-G sites (Hanecak et al., 1962). Because 3C itself is flanked by Q-G sites within the poly- protein, its initial appearance is thought to occur through intremolecular cleavages (Palmenberg and Rueckert, 1962) a mechanism supported by expression of 3C in E. coli and in vitro mutagenesis of a gene segment specify- ing 3C (Hanecak et al., 1964). Polypeptide 3C, by analogy with the corresponding enzymes of other picornaviruses (Palmenberg et al., 1979; Palmenberg and Rueckert, 1962; Gorbalenya and Svitkin, 1963; Argos et al., 1964) is a thiol proteinase. Antibodies raised against 3C inhibit cleavage at Q-G sites, but not at Y-G sites (Hanecak et al., 1962). This observation led us to conclude that proteolytic pro- cessing at active Y-G amino acid pairs requires a separate proteinase. Proteolytic processing at one of the two active Y-G sites separates the capsid precursor Pl from the rest of the polyprotein (see Figure 1). This cleavage occurs while the protein is still growing, and shortly after the ribosome has traveled past the Y-G amino acid pair. This cleavage can be observed in vivo and in vitro, regardless of the system used for cell-free translation (Jackson, 1966, and refer- ences therein). Thus, under normal conditions, a Pl-P2 polypeptide is never observed. Because of the efficiency of the cleavage between Pl and P2 it has been suggested that the enzyme responsible for it may be a ribosome- bound, cellular proteinase (Korant et al., 1960). The other active Y-G cleavage site in the poliovirus poly- protein is within polypeptide 3CD (see Figure l), and cleavage yields 3C’ and 3D’ (Hanecak et al., 1962). The biological significance of this cleavage is unknown; in any event, it lowers the yield of the two Q-G cleavage products of 3CD, the 3C proteinase and the 3D RNA polymerase. Since some poliovirus strains produce very little 3C’ and 3D’(Nottay et al., 1961), processing at this second Y-G site may be fortuitous. Nevertheless, the existence of this site is important for the study of the Y-G-specific proteinase reported here. Expression of the amino-terminal half of the polyprotein in E. coli leads to the synthesis of unstable proteins that are sometimes shorter than the full-length translation products and are rapidly turned over (Enger-Valk et al., 1964; Ariga and Wimmer, unpublished data). We present evidence that precursor polypeptides of poliovirus that span the coding region of Pl and P2, and contain 2A, un- dergo rapid cleavage at the junction between Pl and P2 when expressed in suitable expression vectors in E. coli. Results Construction of pVS(3)2601 In a manner similar to that described by Hanecak et al. (1964), we have expressed sequences of poliovirus that were suspected of encoding proteinase, in bacterial ex- pression vectors. By this approach, we reduce the likeli- hood that any correctly processed proteins found are the products of a proteinase from the host cell. To study the activity responsible for the Y-G cleavages