CLPB

Abstract: Certain proteins and protein fragments in Escherichia coli are modified by carboxy-terminal addition of an 11-residue peptide tag (Tu et al. 1995). This tagging process requires functional SsrA RNA (10Sa RNA), which encodes the last 10 residues of the peptide (Tu et al. 1995) and results in rapid degradation of the tagged protein by carboxy-terminal-specific proteases (Keiler et al. 1996). SsrA-mediated tagging of proteins translated from defective messenger RNAs lacking termination codons has been demonstrated, and a model in which SsrA functions both as a tRNA and an mRNA has been proposed (Keiler et al. 1996). The 363-nucleotide SsrA RNA has sequences that form a tRNA-like structure and has been shown to be chargeable with alanine (Komine et al. 1994; Williams and Bartel 1996; Felden et al. 1997). In the model proposed by Keiler et al., when ribosomes stall at the 3′ end of the damaged message, SsrA charged with alanine binds to the ribosome like a tRNA and contributes the alanine to the idle nascent chain. Translation then switches from the mRNA to a small open reading frame (ORF) in SsrA that encodes the carboxy-terminal degradation peptide. This system provides both a method to avoid the accumulation of ribosomes stalled at the end of defective messages and a general quality-control mechanism that allows the cell to rid itself of incomplete protein fragments that might have inappropriate cellular activities. Cells devoid of SsrA RNA grow more slowly and show a certain degree of temperature sensitivity (Oh and Apirion 1991; Komine et al. 1994; Trempy et al. 1994). The involvement of carboxy-terminal amino-acid sequences in targeting proteins for rapid degradation was recognized before the discovery of the SsrA-tagging system (Bowie and Sauer 1989; Parsell et al. 1990), and a periplasmic protease (Tsp or Prc) that degrades protein substrates in a carboxy-terminal-specific manner was purified and characterized (Silber et al. 1992). The carboxy-terminal substrate sequences recognized by Tsp are similar to those of the SsrA tag (Keiler et al. 1995; Tu et al. 1995), and Tsp is responsible for degradation of SsrA-tagged proteins that are exported to the periplasm (Keiler et al. 1996). Cytoplasmic proteins with carboxy-terminal degradation sequences, however, are still proteolyzed rapidly in cells lacking Tsp (Silber and Sauer 1994; Keiler et al. 1996), indicating that other proteases must be responsible for carboxy-terminal-specific degradation of proteins in the bacterial cytoplasm. Essentially all cytoplasmic degradation in prokaryotes, archaea, and eukaryotes is energy-dependent. E. coli, for example, has at least five ATP-dependent proteases [Lon (La); HflB (FtsH); ClpAP; ClpXP; and ClpYQ (HslUV)] (for review, see Gottesman 1996). These enzymes appear to have distinct substrate preferences, as a mutation in a single protease gene is often sufficient to stabilize a specific unstable protein. For example, mutations in Lon lead to stabilization of the N protein of bacteriophage λ, the SulA and RcsA proteins of E. coli, and the CcdA protein of the episomal F factor. HflB appears to be responsible for degradation of the cII protein of λ and the heat-shock σ factor RpoH (Herman et al. 1993, 1995). The principal substrates for ClpYQ degradation have not yet been identified, although this two-component protease has been implicated in degradation of both Lon subtsrates and HflB substrates in vivo (Missiakas et al. 1996; Kanemori et al. 1997; Khattar 1997; W.-F. Wu and S. Gottesman, unpubl.). ClpAP and ClpXP are two-component proteases that share a common proteolytic subunit, ClpP, but have different ATPase regulatory subunits, ClpA or ClpX. Proteins stabilized by mutations in clpX but not in clpA include λ O, phage Mu repressor variants, and the stationary-phase σ factor, RpoS; clpA but not clpX mutants stabilize certain LacZ fusion proteins and the MazE protein. ClpB, an ATPase with extensive sequence similarity to ClpA, has not thus far been demonstrated to have a direct role in proteolysis but may act as a chaperone (Squires and Squires 1992). In the studies presented here, we show that intracellular degradation of variants of the amino-terminal domain of λ repressor containing the SsrA peptide tag is dramatically reduced in cells lacking ClpP or lacking both ClpX and ClpA, and is somewhat reduced in cells lacking ClpX or ClpA only. Purified ClpXP and purified ClpAP degrade SsrA-tagged protein substrates in vitro, suggesting that these ATP-dependent enzymes are directly responsible for degradation of SsrA-tagged proteins in the bacterial cytoplasm.

Abstract: We systematically analyzed the capability of the major cytosolic chaperones of Escherichia coli to cope with protein misfolding and aggregation during heat stress in vivo and in cell extracts. Under physiological heat stress conditions, only the DnaK system efficiently prevented the aggregation of thermolabile proteins, a surprisingly high number of 150-200 species, corresponding to 15-25% of detected proteins. Identification of thermolabile DnaK substrates by mass spectrometry revealed that they comprise 80% of the large (>/=90 kDa) but only 18% of the small (