The Biochemistry of Proteins Section conducts research on the function and control of protein degradation in bacterial and human cells and on the mechanism of action of the ATP-dependent proteases ClpAP and ClpXP. Clp proteases have three constituents: a substrate recognition domain (SspB, RssB, or ClpS), an ATP-driven protein unfoldase (ClpX or ClpA), and an associated self-compartmentalized protease, ClpP. In the past year we have extended our understanding of intracellular degradation carried out by ClpAP and the adaptor protein, ClpS, which is governed by a mechanism called the N-end rule. The N-end rule defines a mechanism by which proteins are targeted for degradation based on the identity of their N-terminal amino acids. In E. coli, N-end degrons are recognized by ClpS, which binds the N-terminal Leu, Phe, Tyr, and Trp. ClpS interacts with the N-domain of ClpA and hands off the N-end rule substrates to the ClpAP complex. In E. coli cells, proteins with N-terminal Lys and Arg are also targeted, because they acquire a Leu or Phe N-degron through the action of Aat, an aminoacyl tRNA protein transferase. We reported that a ClpS affinity column could capture more than 100 E. coli proteins with N-degrons. We have now shown that ClpS has general utility for capturing N-end rule proteins from other organisms. We have isolated scores proteins with N-degrons from extracts of bacterial cells (Vibrio cholerae and Bacillus subtilis), as well as from extracts of eukaryotic cells, including Saccharomyces cerevisiae and Homo sapiens. We have constructed a mutant of ClpS (M40A) that binds N-terminal amino acids but has lost the ability to discriminate. Using a peptide array we found that this mutant binds all N-terminal amino acids except aspartate and glutamate. Mammalian cells have several different classes of N-degrons but currently there is no mechanism for isolation of proteins bearing a specific N-degron. We will mutagenize ClpS and screen for the ability to bind specific classes of N-degrons and we will use them to pull out proteins from mammalian cells and test their ability to inhibit degradation of proteins with different N-degrons in vivo. Studies of N-end rule degradation in E. coli continue with attempts to identify the peptidase that expose N-degrons in proteins. We cloned YfbL, a putative protease that generates an N-degron in Dps, a DNA-protecting protein in bacteria. Dps is no longer pulled down from cells in which yfbL has been mutated. We also cloned putrescine aminotransferase (PATase), one of the most abundant N-end rule substrates. PATase is unique in that the N-terminal methionine is retained and is modified by addition of Leu and Phe to the N-terminus. We will reconstruct the modification reaction in vitro and identify factors that are responsible for regulating the modification. Studies with ClpP are focused on the mechanism of cell death that results from binding the acyldepsipeptide antibiotic ADEP and the structural changes needed for substrate entry into the degradation chamber. ADEP is an antibiotic made by Streptococcus hawaiiensis. When bound to ClpP ADEP opens the axial channel and activates indiscriminate protein degradation. The site of ADEP binding is also the docking site for ClpX and ClpA, which govern delivery of substrates to ClpP. ADEPs are being developed as novel antibiotics to target human pathogens. Current research is focused on the features of ClpP needed for ADEP binding and for the allosteric changes in ClpP that open the channel. We randomly mutagenized ClpP and identified mutants that are insensitive to ADEP but retain ClpP activity with its cognate ATPases. We found mutations in the axial channel that provides access to the ClpP active site and in sites that affect the shape of the docking site. We have purified several of the mutants and are studying their biochemical and enzymatic properties. We will purify larger quantities for crystallization in order to identify the structural changes that alter their response to ADEP binding. These mutants are rare and we expect to identify sites involved in allosteric communication between the docking site, the active site, and the subunit contact sites, all of which affect ClpP activity. Until recently, studies of Clp function have been hindered by the lack of compounds that can be added to cell cultures to inhibit ClpP. Divalent Zn inhibits ClpP, and we have obtained a crystal structure of ClpP and identified the sites at which Zn binds. Two critical residues that form the interface between subunits in the heptameric ring serve to chelate the Zn. Two catalytic residues, His122 and Asp171, also interact with the Zn. We have observed that Zn stabilizes a collapsed form of the handle region that forms the interface between the ClpP heptameric rings. We obtained a number of bis (benzimidazole) compounds from Prof. Holden Thorp at the University of North Carolina that can enhance Zn binding to proteases. Our preliminary screen of these compounds identified one compound that gave a slight enhancement of inhibition. We will ask our collaborators to prepare similar derivatized bis(benzimidazoles) and test them for their efficacy as co-inhibitors. We have made substantial progress in our collaboration with Alfred Goldberg at Harvard Medical School to obtain a crystal structure of the active form of ClpP from Mycobacterium tuberculosis. ClpP is essential for growth of M. tuberculosis and thus is a promising target for potential antimicrobials. We now have a 3.0 Angstrom crystal structure of the active form, which consists of a heptameric ring of ClpP1 complexed with a heptameric ring of ClpP2. Only this hetero-complex is active. The presence of two forms of ClpP in one complex will facilitate structural analysis of the ring interactions by allowing assembly of tetradecamers in which only one ring is mutated. We observe the activating peptide in the ClpP1 and ClpP2 active sites, but interestingly the peptide binds in opposite orientations in the two sites. The crystal structure should guide the design of small molecule inhibitors that will serve as leads for the development of compounds that can block the growth of M. tuberculosis and other pathogens. The goal of our studies of human ClpX and ClpP is to define their functions in mitochondria and to discover why they are needed for mitochondrial integrity and cell survival. We found that over expression of HClpP allows better survival of cells treated with the anti-cancer drug, cisplatin. Conversely, cells were more sensitive to cisplatin when HClpP was partially knocked down. Cisplatin accumulation increased when HClpP was knocked down, suggesting that HClpP activity might be needed to allow rapid uptake of cisplatin. Alternatively, HClpP activity could affect one or more enzymes that metabolize cisplatin or cisplatin adducts in the cell. We find that cisplatin is incorporated preferentially into mitochondrial DNA and that HClpP has a dramatic effect on the level of cisplatin adducts detected in mitochondrial DNA. HClpP has been implicated in a hereditary human disease called Perrault's syndrome. In addition, homozygous knockout of ClpP in mice leads to profound hearing loss and infertility. These results indicate that ClpP plays some important or even essential role in mammalian cells. We find that in human cell culture drastic depletion of hClpP or hClpX by treatment with siRNA leads to cell death. Because the conditions for transfection are stressful to cultured cells we propose that HClpP might be essential under conditions of stress, which would explain why mice with homozygous deletion of CLPP survive. Proteomics studies reveal that 30 proteins are increased within 16 hours of depletion of hClpP with siRNA and that many of the proteins are involved in stress responses.
