An Escherichia coli protease designated Tsp (tail-specific protease) has been purified, and its gene has been cloned and sequenced. Tsp specifically degrades a variant of the N-terminal domain of A repressor in which the five C-terminal residues, which are polar in wild type, have been replaced by nonpolar residues. This substrate specificity in vitro parallels the previously reported selective degradation in vivo of N-terminal-domain variants with nonpolar C-terminal residues. The gene sequence and N-terminal protein sequence of Tsp predict a protein of 660 amino acids. The deduced protein sequence of Tsp shows no significant homology to known protease sequences but does show sequence similarity to the human and bovine interphotoreceptor retinoid-binding proteins, which bind hydrophobic ligands.Within cells, different proteins are degraded with half-lives ranging from minutes to days (1). This heterogeneity implies that intracellular degradation is selective, but the mechanisms of this selectivity are incompletely understood. The C-terminal sequences of some proteins can have dramatic effects on their rates of degradation in Escherichia coli (2, 3). For example, the wild-type N-terminal domain of A repressor (residues 1-102) has the polar C-terminal amino acid sequence RSEYE102 and has a half-life in vivo of >10 hr. The rapidly degraded "#105" variant of this protein has the hydrophobic C-terminal sequence WVAAA102 and has a half-life of 15 min (3). To identify possible cellular component(s) responsible for the degradation of proteins with destabilizing C termini, we have purified an E. coli activity that specifically degrades the #105 variant but not the wild-type N-terminal domain. This activity resides in a single, purified polypeptide (Tsp, for tail-specific protease). We have cloned and sequenced the tsp gene.* Based on sequence comparisons, Tsp shows similarities to the human and bovine interphotoreceptor retinoid-binding proteins (IRBPs) but does not resemble any proteases in the protein sequence data base.
MATERIALS AND METHODSAssays for Tsp Activity. Tsp activity in crude lysates and column fractions was assayed by measuring the degradation of 35S-labeled substrates (either the #105 variant or the wild-type N-terminal domain of A repressor) to products soluble in 10% (wt/vol) trichloroacetic acid (3). Reaction mixtures contained _9000 cpm of 35S-labeled substrate, a sample of the E. coli fraction in lysis buffer or column buffer, and 0.02% Nonidet P-40 in a volume of 50 p1. Reaction mixtures were incubated at 370C for 1 or 2 hr and were processed as described (3).In more purified fractions, degradation was assayed by the appearance of substrate digestion products following SDS/ PAGE. Wild-type or #105 protein (2 ,ug), which was purified by the method of Lim and Sauer (4), was mixed with a sample in a 10-1LI reaction volume. Reaction mixtures were incubated at 370C for 3-6 hr and were stopped by boiling for 3 min in Laemmli sample buffer (5). Samples were electrophoresed in SDS/16.5% polyacrylam...