Differential binding of HMG1, HMG2, and a single HMG box to cisplatin-damaged DNA

The HMG box domain is a DNA binding domain present in the nonhistone chromosomal proteins HMG1 and HMG2 and in other proteins involved in the regulation of gene expression. Previous studies have demonstrated that HMG1 and HMG2 bind with high affinity to DNA modified with the cancer chemotherapeutic drug cisplatin (CDDP). In this report, we compare the binding of full-length HMG1 and HMG2 and the HMG boxes present in these proteins to that of CDDP-DNA. Complexes between HMG1, HMG2, or HMG Box A + B and CDDP-DNA were stable at > or = 500 mM salt, while complexes between a single HMG box and CDDP-DNA exhibited decreased stability. Analysis of a series of HMG1 Box A mutant constructs revealed different affinities for CDDP-DNA. Two constructs containing a Phe to Ala substitution at position 19 and a Tyr to Gly substitution at position 71, are noteworthy; these peptides exhibited reduced affinity for CDDP-DNA. We have generated a structure of HMG1 Box A and used it, along with the results of our binding studies, to model its interaction with CDDP-DNA. HMG1 Box A binds in the minor groove of CDDP-DNA, in agreement with earlier studies. Our model predicts that Tyr71 partially intercalates and forms an H bond with the sugar-phosphate backbone. The model also suggests that Phe 19 does not directly interact with DNA, and hence an Ala substitution at position 19 may alter protein structure. This model should provide a framework for future studies examining HMG Box-DNA interactions.     

Subdomain folding of the coiled coil leucine zipper from the bZIP transcriptional activator GCN4

One popular model for protein folding, the framework model, postulates initial formation of secondary structure elements, which then assemble into the native conformation. However, short peptides that correspond to secondary structure elements in proteins are often only marginally stable in isolation. A 33-residue peptide (GCN4-p1) corresponding to the GCN4 leucine zipper folds as a parallel, two-stranded coiled coil [O'Shea, E.K., Klemm, J.D., Kim, P.S., & Alber, T.A. (1991) Science 254, 539-544]. Deletion of the first residue (Arg 1) results in local, N-terminal unfolding of the coiled coil, suggesting that a stable subdomain of GCN4-p1 can form. N- and C-terminal deletion studies result in a 23-residue peptide, corresponding to residues 8-30 of GCN4-p1, that folds as a parallel, two-stranded coil with substantial stability (the melting temperature of a 1 mM solution is 43 degrees C at pH 7). In contrast, a closely related 23-residue peptide (residues 11-33 of GCN4-p1) is predominantly unfolded, even at 0 degrees C, as observed previously for many isolated peptides of similar length. Thus, specific tertiary packing interactions between two short units of secondary structure can be energetically more important in stabilizing folded structure than secondary structure propensities. These results provide strong support for the notion that stable, cooperatively folded subdomains are the important determinants of protein folding.     

DNA-binding properties of the yeast transcriptional activator, Gcr1p

In Saccharomyces cerevisiae the GCRI gene product is required for high-level expression of genes encoding glycolytic enzymes. In this communication, we extend our analysis of the DNA binding properties of Gcr1p. The DNA-binding domain of Gcr1p binds DNA with high affinity. The apparent dissociation constant of the Gcr1p DNA-binding domain for one of its specific binding sites (TTTCAGCTTCCTCTAT) is 2.9 x 10(-10) M. However, competition experiments showed that Gcr1p binds this site in vitro with a low degree of specificity. We measured a 33-fold difference between the ability of specific competitor and DNA of random sequence to inhibit the formation of nucleoprotein complexes between Gcr1p and a radiolabeled DNA probe containing its binding site. DNA band-shift experiments, utilizing probes of constant length in which the positions of Gcr1p-binding sites are varied relative to the ends, indicated that Gcr1p-DNA nucleoprotein complexes contain bent DNA. The implications of these findings in terms of the combinatorial interactions that occur at the upstream activating sequence elements of genes encoding glycolytic enzymes are discussed.     

Solution structure of the oxidized 2[4Fe-4S] ferredoxin from Clostridium pasteurianum

Following the recently developed approach to the solution structure of paramagnetic high-potential iron-sulfur proteins, the three-dimensional structure in solution of the oxidized Clostridium pasteurianum ferredoxin has been solved by 1H-NMR. The X-ray structure is not available. The protein contains 55 amino acids and two [4Fe-4S] clusters. In the oxidized state, the clusters have S = 0 ground states, but are paramagnetic because of thermal population of excited states. Due to the somewhat small size of the protein and to the presence of two clusters, approximately 55% of the residues have at least one proton with a non-selective T1 smaller than 25 ms. The protein has thus been used as a test system to challenge the present paramagnetic NMR methodology both in achieving an extended assignment and in obtaining a suitable number of constraints. 79% of protein protons have been assigned. Analogy with other ferredoxins of known structure has been of help to speed up the final stages of the assignment, although we have shown that this independent information is not necessary. In addition to dipolar connectivities, partially detected through tailored experiments, 3JHN-H alpha, H-bond constraints and dihedral angle constraints on the Cys chi 2 angles have been generated by using a recently derived Karplus-type relationship for the hyperfine shifts of cysteine beta CH2 protons. In total, 456 constraints have been used in distance geometry calculations. The final quality of the structures is satisfactory, with root-mean-square deviation values of 66 pm and 108 pm for backbone and heavy atoms, respectively. The resulting structure is compared with that of Clostridium acidi urici ferredoxin [Duée, E. D., Fanchon, E., Vicat, J., Sieker, L. C., Meyer, J. & Moulis, J.-M. (1994) J. Mol. Biol. 243, 683-695]. The two proteins are very similar in the overall folding, secondary structure elements and side-chain orientations. The C alpha root-mean-square deviation values between the X-ray-determined C. acidi urici ferredoxin structure and the conformer with lowest energy of the C. pasteurianum ferredoxin family is 78 pm (residues 3-53). Discrepancies in residues 26-28 may arise from the disorder observed in the X-ray structure in that region.     

Cleavage of transcriptional activator Oct-1 by poliovirus encoded protease 3Cpro

To understand the regulation of receptor-mediated endocytosis in hepatocytes, we have used two specific inhibitors of serine-threonine protein phosphatases (PP), microcystin (MCYST) and okadaic acid (OKA) as probes to alter protein phosphorylation in hepatocytes. We have then examined the impact of these changes on the specific binding and uptake of transferrin (Tf) in hepatocytes. The measurement of PP activity in hepatocyte lysates showed that OKA and MCYST shared a common inhibition of protein phosphatase 2A (PP2A). Our results showed that both OKA (250 nmol/L) and MCYST (500 nmol/L) significantly reduced Tf uptake at steady state (P < or = .05). The measurement of Tf internalization after 15 minutes in protein phosphatase inhibitor-pretreated cells revealed that the initial uptake was also significantly reduced. Binding studies showed that pretreatment with either of the phosphatase inhibitors did not result in significant changes in the K(d) for Tf binding to transferrin receptor (TfR). Additionally, no significant changes in the number of TfR in the plasma membrane were observed in phosphatase inhibitor-pretreated cells. The treatment of hepatocytes with nocodazole (NOC), which results in microtubule disassembly and inhibition of microtubule-based vesicle transport, caused comparable reductions in initial and steady state levels of transferrin accumulation. The changes in transferrin accumulation by both phosphatase inhibitors and nocodazole were accompanied by redistribution of the microtubule-anchored Golgi apparatus and lysosomal network from the perinuclear region to the cell periphery. Our data show that the regulation of Tf uptake by receptor-mediated endocytosis is mediated by PP2A and additionally may occur through regulation of microtubule-based vesicle transport.     

Crystal structure of a DExx box DNA helicase

There are a wide variety of helicases that unwind helical DNA and RNA substrates. The twelve helicases that have been identified in Escherichia coli play a role in almost all cellular processes involving nucleic acids. We have solved the crystal structure of a monomeric form of a DNA helicase from Bacillus stearothermophilus, alone and in a complex with ADP, at 2.5 and 2.9 A resolution, respectively. The enzyme comprises two domains with a deep cleft running between them. The ATP-binding site, which is situated at the bottom of this cleft, is formed by motifs that are conserved across the superfamily of related helicases. Unexpected structural homology with the DNA recombination protein, RecA, suggests how ATP binding and hydrolysis may drive conformational changes of the enzyme during catalysis, and implies that there is a common mechanism for all helicases.     

Solution structure of the tobramycin-RNA aptamer complex

We have solved the solution structure of the aminoglycoside antibiotic tobramycin complexed with a stem-loop RNA aptamer. The 14 base loop of the RNA aptamer 'zippers up' alongside the attached stem through alignment of four mismatches and one Watson-Crick pair on complex formation. The tobramycin inserts into the deep groove centered about the mismatch pairs and is partially encapsulated between its floor and a looped out guanine base that flaps over the bound antibiotic. Several potential intermolecular hydrogen bonds between the charged NH3 groups of tobramycin and acceptor atoms on base pair edges and backbone phosphates anchor the aminoglycoside antibiotic within its sequence/structure specific RNA binding pocket.