Sequence of the M28 dsRNA: preprotoxin is processed to an alpha/beta heterodimeric protein toxin

The killer and immunity phenotypes of K28 killer strains of Saccharomyces cerevisiae are determined by the 1.75-kb M28 dsRNA virus. In the plus strand, M28p, the K28 preprotoxin gene, comprises bases 13-1047 and is followed, after an additional 85 bases, by a 63-bp poly(A) sequence and a 553-base 3\u27-sequence. This 3\u27-sequence contains two potential stem-loop structures predicted to bind the L-A encoded cap-pol protein, initiating encapsidation; high-level expression results in curing of M1 dsRNA. Expression of M28p confers the complete K28 killer and immunity phenotype on a cell lacking M28 dsRNA. K28 toxin is a disulfide-bonded heterodimer of alpha (10.5 kDa) and beta (11 kDa) components whose N-termini correspond to M28p residues 50-61 and 246-257, respectively. alpha is preceded by a potentially redundant pair of secretion signal peptides; deletion of the first reduces toxin secretion by 75%. While M28p bears no sequence similarity to M1p, the K1 preprotoxin, the predicted patterns of processing by glycosylation and cleavage are remarkably similar. The beta N- and C-termini are probably processed by Kex2p and Kex1p, respectively; the mechanism of cleavage at the less typical sites bounding the alpha component is under investigation. While a kex2 delta mutation prevents toxin secretion, secreted toxin retains 20% activity in a kex1 delta mutant. Neither mutation affects immunity