ClpA is the ATPase component of the ATP-dependent protease Ti (Clp) in Escherichia coli and contains two ATP-binding sites. A ClpA variant (referred to as ClpAT) carrying threonine in place of the 169th methionine has recently been shown to be highly soluble but indistinguishable from the wild-type, 84-kDa ClpA in its ability to hydrolyze ATP and to support the casein-degrading activity of ClpP. Therefore, site-directed mutagenesis was performed to generate mutations in either of the two ATP-binding sites of ClpAT (i.e. to replace the Lys220 or Lys501 with Thr). ClpAT/K220T hydrolyzed ATP and supported the ClpP-mediated proteolysis 10-50% as well as ClpAT depending on ATP concentration, while ClpAT/K501T was unable to cleave ATP or to support the proteolysis. Without ATP, ClpAT and both of its mutant forms behaved as trimeric molecules as analyzed by gel filtration on a Sephacryl S-300 column. With 0.5 mM ATP, ClpAT and ClpAT/K501T became hexamers, but ClpAT/K220T remained trimeric. With 2 mM ATP, however, ClpAT/K220T also behaved as a hexamer. These results suggest that the first ATP-binding site of ClpA is responsible for hexamer formation, while the second is essential for ATP hydrolysis. When trimeric ClpAT/K220T was incubated with the same amount of hexameric ClpAT/K501T (i.e. at 0.5 mM ATP) and then subjected to gel filtration as above, a majority of ClpAT/K220T ran together with ClpAT/K501T as hexameric molecules. Furthermore, ClpAT/K501T in the mixture strongly inhibited the ability of ClpAT/K220T to cleave ATP and to support the ClpP-mediated proteolysis. Similar results were obtained in the presence of 2 mM ATP and also with the mixture with ClpAT. On the other hand, the ATPase activity of the mixture of ClpAT and ClpAT/K220T was significantly higher than the sum of that of each protein, particularly in the presence of 2 mM ATP, although its ability to support the proteolysis by ClpP remained unchanged. These results suggest that a rapid exchange of the subunits, possibly as a trimeric unit, occurs between the ClpAT proteins in the presence of ATP and leads to the formation of mixed hexameric molecules.

A new method for assaying ubiquitin C-terminal hydrolases was developed using a 125I-labeled ubiquitin-alpha NH-MHISPPEPESEEEEEHYC was substrate. Since the peptide portion was almost exclusively radiolabeled, the enzymes could be assayed directly by simple measurement of the radioactivity released into acid-soluble products. Using this assay protocol, we identified at least 10 ubiquitin C-terminal hydrolase activities from the extract of chick skeletal muscle, which were tentatively named UCHs 1 through 10. Of these, UCH-6 was purified to apparent homogeneity. Purified UCH-6 behaved as a dimer of 27-kDa subunits. The apparent molecular masses of the other partially purified UCHs ranged from 35 to 810 kDa as determined under a non-denaturing condition. Muscle UCHs, except UCH-1, were activated dramatically by poly-L-Lys but with an unknown mechanism. All of the UCHs were sensitive to inhibition by sulfhydryl-blocking agents such as iodoacetamide. In addition, all of the UCHs were capable of releasing free ubiquitin from a ubiquitin-alpha NH-carboxyl extension protein of 80 amino acids and from ubiquitin-alpha NH-dihydrofolate reductase. Five of the enzymes, UCHs 1 through 5, were also capable of generating free ubiquitin from poly-His-tagged diubiquitin. In addition, UCH-1 and UCH-7 could remove ubiquitin that had been ligated covalently by an isopeptide linkage to a ubiquitin (RGA)-alpha NH-peptide, the peptide portion of which consists of the 20 amino acids of the calmodulin binding domain of myosin light chain kinase. These results suggest that the 10 UCH activities isolated from chick skeletal muscle appear to be distinct from each other at least in their chromatographic behavior, size, and substrate specificity.

No abstract

We have isolated, cloned, and characterized two cDNAs corresponding to the mRNAs for cationic peroxidases synthesized by cultured peanut cells. The first clone was obtained from a phage Agtll library screened with antibodies directed against the major secreted isozyme. Its predicted amino acid sequence, deduced from the 1228-base-pair (bp) cDNA, revealed a 22-amino acid signal peptide and a 294-amino acid mature protein (Mr, 31,228). The second clone was isolated from a AgtlO library screened with oligonucleotides corresponding to the regions for acid/base catalysis and the fifth ligand of heme. This cDNA (1344 bp) encodes a protein (330 amino acids) with a mature peptide of 307 residues (Mr, 32,954). The two peanut peroxidases are 46% homologous. The estimated gene copy numbers of these peroxidases might be close to 1 or 2 per haploid genome. A comparison of the amino acid sequence of these peanut peroxidases with other known isozymes shows two already known regions of homology (the region for acid/base catalysis and the fifth ligand of heme). Moreover, some new characteristics appeared such as a glycosylation site identical in five of the seven isozymes, a putative antigenic determinant common to all the isozymes, and a region of the highest homology. A secondary structure prediction showed that it corresponds to a 16-amino acid helix linked to the next one by a long stretch of f3 strands and coils and might represent a critical structural element.

No abstract