Faculty Bibliography

Chaperonins are known to facilitate protein folding, but their mechanism of action is not well understood. The fact that target proteins are released from and rebind to different chaperonin molecules ('cycling') during a folding reaction suggests that chaperonins function by unfolding aberrantly folded molecules, allowing them multiple opportunities to reach the native state in bulk solution. Here we show that the cycling of alpha-tubulin by cytosolic chaperonin (c-cpn) can be uncoupled from the action of cofactors required to complete the folding reaction. This results in the accumulation of folding intermediates which are chaperonin-bound, stable, and quasi-native in that they bind GTP nonexchangeably. We present evidence that these intermediates can be generated without the target protein leaving c-cpn. These data show that, in contrast to prevailing models, target proteins can maintain, and possibly acquire, significant native-like structure while chaperonin-bound

Chaperonins are ubiquitous multisubunit toroidal complexes that aid protein folding in an ATP-dependent manner. Current models of folding by the bacterial chaperonin GroEL depict its role as unfolding and releasing molecules that have misfolded, so that they can return to a potentially productive folding pathway in solution. Accordingly, a given target polypeptide might require several cycles of binding and ATP-driven release from different chaperonin complexes before reaching the native state. Surprisingly, cycling of a target protein does not guarantee its folding, and we report here that unfolded beta-actin or alpha-tubulin both form tight complexes when presented to either GroEL or its mitochondrial homologue, and both undergo cycles of release and rebinding upon incubation with ATP, but no native protein is produced. We conclude that different chaperonins produce distinctive spectra of folding intermediates

Arginine biosynthesis in Escherichia coli is negatively regulated by a hexameric repressor protein, encoded by the gene argR and the corepressor arginine. By hydroxylamine mutagenesis two types of argR mutants were isolated and mapped. The first type is transdominant. In heterodiploids, these mutant polypeptides reduce the activity of the wild-type repressor, presumably by forming heteropolymers. Four mutant repressor proteins were purified. Two of these map in the N-terminal half of the protein. Gel retardation experiments showed that they bind poorly to DNA, but they could be precipitated by L-arginine at the same concentration as the wild-type repressor. The other two mutant repressors map in the C-terminal half of the protein. They are poorly precipitated by L-arginine and they bind poorly to DNA. In addition, one of these mutants appears to exist as a dimer. The second type of argR mutant repressor consists of super-repressors. Such mutants behave as arginine auxotrophs as a result of hyper-repression of arginine biosynthetic enzymes. They map at many locations throughout the argR gene. Three arginine super-repressor proteins were purified. In comparison with the wild-type repressor, two of them were shown to have a higher DNA-binding affinity in the absence of bound arginine, while the third was shown to have a higher DNA-binding affinity when bound to arginine

In Escherichia coli K12, formation of the enzymes of arginine biosynthesis are controlled by arginine, with complete repression during growth with added arginine, severe repression (about 95%) during growth without added arginine and complete derepression during arginine-limited growth. In E. coli B, the degree of repression is not correlated with arginine concentrations. Under all conditions of growth enzyme formation is repressed, with repression being somewhat less in a medium with arginine than in a medium without arginine. These differences in repressibility between the two strains have been shown previously to be due to the presence of different alleles of argR, the gene for the arginine repressor. Here we have compared the binding of the two repressors to the operator sites of argF (ARG boxes). In DNase I footprinting and gel retardation experiments with argF ARG boxes we have shown that the arginine repressor of E. coli K12 bound to arginine (ArgRK-arg) has a greater affinity than the arginine repressor of E. coli B bound to arginine (ArgRB-arg), whereas free ArgRB (ArgRBf) has a much stronger affinity than free ArgRK (ArgRKf). The stronger binding of ArgRBf can explain the repression seen in E. coli B during arginine-limited growth and indicates that ArgRBf, but not ArgRKf, is able to repress enzyme synthesis under physiological conditions. The weaker repression of E. coli B than of E. coli K12 seen in the presence of arginine can be explained by the lower affinity of ArgRB-arg for operator sites as compared to ArgRK-arg. Another contributing cause for the weaker repression is the reduction of ArgRBf concentration due to autoregulation of the gene for the repressor. Thus the combined effects of repression by ArgRBf, but not ArgRKf, with the weaker repression by ArgRB-arg as compared to ArgRK-arg, convert the arginine dependent regulation in E. coli K12 to arginine independent regulation in E. coli B

In the arginine regulon of Escherichia coli K12 each of the eight operator sites consists of two 18-base-pair-long palindromic sequences called ARG boxes. In the operator sites for the structural genes of the regulon the two ARG boxes are separated by three base-pairs, in the regulatory gene argR they are separated by two base-pairs. The hexameric arginine repressor, the product of argR, binds to the two ARG boxes in an operator in the presence of L-arginine. From the results of various kinds of in vitro footprinting experiments with the ARG boxes of argF and argR (DNase I protection, hydroxyl radical, ethylation and methylation interference, methylation protection) it can be concluded that: (1) the repressor binds simultaneously to two adjacent ARG boxes; (2) that it binds on one face of the double helix; and (3) that it forms contacts with the major and minor grooves of each ARG box, but not with the central three base-pairs. The repressor can bind also to a single ARG box, but its affinity is about 100-fold lower than for two ARG boxes. From gel retardation experiments with 3H-labeled repressor and 32P-labeled argF operator DNA, it is concluded that the retarded DNA-protein complex contains no more than one repressor molecule per operator site and that most likely one hexamer binds to two ARG boxes. The bound repressor was shown to induce bending of argF operator DNA. The bending angle calculated from the results of gel retardation experiments is about 70 degrees and the bending center was located within the region encompassing the ARG boxes. The main features that distinguish the arginine repressor from other repressors studied in E. coli are its hexameric nature and the simultaneous binding of one hexameric molecule to two palindromic ARG boxes that are close to each other