SmtB-DNA and protein-protein interactions in the formation of the cyanobacterial metallothionein repression complex: Zn2+ does not dissociate the protein-DNA complex in vitro.

The synechococcal metallothionein locus smt consists of two divergent genes: smtA coding for the metallothionein SmtA, and smtB coding for the trans-acting regulator SmtB. The latter binds at two inverted repeats, designated S1/S2 and S3/S4, in the overlapping promoter/operator sites between the two genes. We have determined the binding stoichiometries to the entire operator/promoter DNA and to the separate S1/S2 and S3/S4 half-operator oligonucleotides using sedimentation equilibrium and sedimentation velocity measurements. The full promoter/operator DNA binds two SmtB dimers. The hydrodynamic behavior of this complex supports a compact nucleoprotein structure. Each separate S1/S2 and S3/S4 operator sequence also binds two dimers. An equal molar mixture of separate S1/S2 and S3/S4 operator sequences, in excess SmtB, forms a S1/S2-SmtB:SmtB-S3/S4 bridge complex. Combining these results with previously published binding interference data, which showed consecutive S1/S2 and S3/S4 SmtB occupancy on the operator/promoter DNA, we have developed a model for the establishment of the repression complex that appears to involve significant DNA compaction, presumably DNA bending, stabilized by SmtB-SmtB bridge interactions. DNase I footprinting titrations also showed consecutive S1/S2 and S3/S4 SmtB occupancy. The footprints expand considerably in the presence of Zn2+. Hence, SmtB remains bound to the operator sites when Zn2+ ions are present. This result is further supported by gel retardation assay. Failure of the metal ions to dissociate SmtB from the DNA points to a hitherto unknown function of SmtB in the regulation of the smt locus.

Analysis of transport experiments using pseudo-absorbance data.

The measurement of the concentration distribution of a macromolecule across a solution column by absorption optics usually requires optical transmission profiles of both the sample solution and the buffer, measured under identical conditions, to calculate the absorbance as the logarithm of the ratio of reference to sample intensity. For transport experiments, however, where the changes in the local macromolecule concentration with time are measured, a reference buffer intensity is not necessarily required. We demonstrate that the logarithm of the light transmitted through the sample solution, referred to as pseudo-absorbance, can suffice to determine macromolecular transport parameters of interest, with little loss of precision. Local changes in illumination of the sample column or in the detection efficiency of the transmitted light, as well as temporal fluctuations of the light source intensity can be well-described by consideration of time-invariant and radial-invariant signal components in the pseudo-absorbance data, using the systematic noise decomposition techniques developed recently (Schuck, P., and Demeler, B. (1999) Biophys. J. 76, 2288-2296). The practical use of the method is demonstrated with double-sector and single-sector sedimentation velocity experiments, and with analytical electrophoresis experiments. It is shown that pseudo-absorbance analysis can increase the capacity of a sedimentation velocity experiment in ultracentrifugation, and, in general, can considerably simplify the requirements of optical design.

Reexamination of the high mobility group-1 protein for self-association and characterization of hydrodynamic properties.

Previous studies of the 25 kDa high mobility group-1 (HMG-1) protein have generated conflicting results regarding whether HMG-1 exists as a monomer or is capable of oligomerizing to (functional) tetramers. To resolve this question, sedimentation velocity analysis yielded a s20,w value of 2.59S, which is consistent with a monomeric protein. Equilibrium sedimentation data were obtained for three HMG-1 concentrations at two rotor speeds. The six sets of data were fit to both an ideal single component and monomer-dimer equilibrium model, with essentially identical fits produced for both models, with the latter indicating a low extent (7%) of dimerization. Reaction of HMG-1 with glutaraldehyde produced a small population of oligomers consistent with a low level of dimers. This supported the monomer-dimer equilibrium model. Surprisingly, gel permeation chromatography yielded an apparent molecular mass of approx. 55 kDa for both HMG-1 and HMG-2. This finding is considered anomalous and presumably due to the high negative charge density in the C terminus of HMG-1. The sedimentation data also permit one to model HMG-1 as a hydrated prolate ellipsoid with a major axis/minor axis ratio of 2. 79. The collective evidence from the sedimentation and chemical cross-linking studies strongly supports a moderately asymmetric monomer in solution and unequivocally eliminates the possibility of a highly extended shape for HMG-1 or the existence of any extensive oligomerization.

Crystal structure of the cyanobacterial metallothionein repressor SmtB: a model for metalloregulatory proteins.

SmtB from Synechococcus PCC7942 is a trans-acting dimeric repressor that is required for Zn(2+)-responsive expression of the metallothionein SmtA. The structure of SmtB was solved using multiple isomorphous replacement techniques and refined at 2.2 A resolution by simulated annealing to an R-factor of 0.218. SmtB displays the classical helix-turn-helix motif found in many DNA-binding proteins. It has an alpha + beta topology, and the arrangement of the three core helices and the beta hairpin is similar to the HNF-3/fork head, CAP and diphtheria toxin repressor proteins. Although there is no zinc in the crystal structure, analysis of a mercuric acetate derivative suggests a total of four Zn2+ binding sites in the dimer. Two of these putative sites are at the opposite ends of the dimer, while the other two are at the dimer interface and are formed by residues contributed from each monomer. The structure of the dimer is such that simultaneous binding for both recognition helices to DNA would require either a bend in the DNA helix or a conformational change in the dimer. The structure of Synechococcus SmtB is the first in this family of metal-binding DNA repressors.

The cyanobacterial repressor SmtB is predominantly a dimer and binds two Zn2+ ions per subunit.

The Synechococcus PCC7942 metallothionein repressor gene smtB has been cloned into a high expression vector and the protein purified to near homogeneity (>/=98%). Analytical ultracentrifugation studies demonstrate that the protein is predominantly dimeric in 0.1 M NaCl, pH 7.4, and 22 degrees C, exhibiting a monomer-dimer-tetramer equilibrium. The monomer-dimer (Ka(1,2)) and the dimer-tetramer (Ka(2,4)) association constants are 3.24 x 10(5) and 9.90 x 10(2) M-1, respectively. The repressor binds two Zn2+ ions per subunit with an overall Kd of 3.49 x 10(-6) M. In the presence of Zn2+, Ka(1, 2) increases by 2 orders of magnitude to 1.25 x 10(7) M-1 and the apparent weight-averaged sedimentation coefficient increases from 2. 00 to 2.22 S. The fact that the increase in sedimentation coefficient is greater than that predicted by increased dimerization is interpreted as caused by compaction of the structure in the presence of metal ions. At pH 6.0, 0.1 M NaCl, and 22 degrees C, the protein exhibits only a monomer-dimer equilibrium, with Ka(1,2) = 1.52 x 10(7) M-1 which is almost identical to that seen upon binding Zn2+ at pH 7.4. The compaction and conformational change in SmtB caused by Zn2+ is consistent with a role for this altered quaternary state in derepression of smtA in Synechococcus challenged with heavy metal ions.

Oligomerization of the hydrophobic heptad repeat of gp41.

The transmembrane protein of human immunodeficiency virus type 1 (HIV-1) contains a leucine zipper-like (hydrophobic heptad) repeat which has been predicted to form an amphipathic alpha helix. To evaluate the potential of the hydrophobic heptad repeat to induce protein oligomerization, this region of gp41 has been cloned into the bacterial expression vector pRIT2T. The resulting plasmid, pRIT3, expresses a fusion protein consisting of the Fc binding domain of monomeric protein A, a bacterial protein, and amino acids 538 to 593 of HIV-1 gp41. Gel filtration chromatography demonstrated the presence of oligomeric forms of the fusion protein, and analytical centrifugation studies confirmed that the chimeric protein formed a higher-order multimer that was greater than a dimer. Thus, we have identified a region of HIV-1 gp41 which is capable of directing the oligomerization of a fusion protein containing monomeric protein A. Point mutations, previously shown to inhibit the biological activity of the HIV-1 envelope glycoprotein, have been engineered into the segment of gp41 contained in the fusion protein, and expressed mutant proteins were purified and analyzed via fast protein liquid chromatography. A point mutation in the heptad repeat, which changed the central isoleucine to an alanine, caused a significant (> 60%) decrease in oligomerization, whereas changing the central isoleucine to aspartate or proline resulted in almost a complete loss of oligomerization. Deletions of one, two, or three amino acids following the first isoleucine also resulted in a profound decrease in oligomerization. The inhibitory effects of the mutations on oligomer formation correlated with the effects of the same mutations on envelope glycoprotein-mediated fusion. A possible role of the leucine zipper-like region in the fusion process and in an oligomerization event distinct from assembly of the envelope glycoprotein complex is discussed.