Hideyasu Okamura

Solution structure of a telomeric DNA complex of human TRF1

BACKGROUND Mammalian telomeres consist of long tandem arrays of double-stranded TTAGGG sequence motif packaged by TRF1 and TRF2. In contrast to the DNA binding domain of c-Myb, which consists of three imperfect tandem repeats, DNA binding domains of both TRF1 and TRF2 contain only a single Myb repeat. In a DNA complex of c-Myb, both the second and third repeats are closely packed in the major groove of DNA and recognize a specific base sequence cooperatively. RESULTS The structure of the DNA binding domain of human TRF1 bound to telomeric DNA has been determined by NMR. It consists of three helices, whose architecture is very close to that of three repeats of the c-Myb DNA binding domain. Only the single Myb domain of TRF1 is sufficient for the sequence-specific recognition. The third helix of TRF1 recognizes the TAGGG part in the major groove, and the N-terminal arm interacts with the TT part in the minor groove. CONCLUSIONS The DNA binding domain of TRF1 can specifically and fully recognize the AGGGTT sequence. It is likely that, in the dimer of TRF1, two DNA binding domains can bind independently in tandem arrays to two binding sites of telomeric DNA that is composed of the repeated AGGGTT motif. Although TRF2 plays an important role in the t loop formation that protects the ends of telomeres, it is likely that the binding mode of TRF2 to double-stranded telomeric DNA is almost identical to that of TRF1.

Structure of the central core domain of TFIIEβ with a novel double-stranded DNA-binding surface

Human general transcription factor TFIIE consists of two subunits, TFIIEα and TFIIEβ. Recently, TFIIEβ has been found to bind to the region where the promoter starts to open to be single‐stranded upon transcription initiation by RNA polymerase II. Here, the central core domain of human TFIIEβ (TFIIEβc) has been identified by a limited proteolysis. This solution structure has been determined by NMR. It consists of three helices with a β hairpin at the C–terminus, resembling the winged helix proteins. However, TFIIEβc shows a novel double‐stranded DNA‐binding activity where the DNA‐binding surface locates on the opposite side to the previously reported winged helix motif by forming a positively charged furrow. A model will be proposed that TFIIE stabilizes the preinitiation complex by binding not only to the general transcription factors together with RNA polymerase II but also to the promoter DNA, where double‐stranded DNA starts to open to be single‐stranded upon activation of the preinitiation complex.

A novel zinc finger structure in the large subunit of human general transcription factor TFIIE.

The zinc finger domain in the large subunit of TFIIE (TFIIEα) is phylogenetically conserved and is essential for transcription. Here, we determined the solution structure of this domain by using NMR. It consisted of one α-helix and five β-strands, showing novel features distinct from previously determined zinc-binding structures. We created point mutants of TFIIEα in this domain and examined their binding abilities to other general transcription factors as well as their transcription activities. Four Zn2+-ligand mutants, in which each of cysteine residues at positions 129, 132, 154, and 157 was replaced by alanine, possessed no transcription activities on a linearized template, whereas, on a supercoiled template, interesting functional asymmetry was observed: although the C-terminal two mutants abolished transcription activity (<5%), the N-terminal two mutants retained about 20% activities. The N-terminal two mutants bound stronger to the small subunit of TFIIF than the wild type and the C-terminal two mutants were impaired in their binding abilities to the XPB subunits of TFIIH. These suggest that the structural integrity of the zinc finger domain is essential for the TFIIE function, particularly in the transition from the transcription initiation to elongation and the conformational tuning of this domain for appropriate positioning of TFIIF, TFIIH, and polymerase II would be needed depending on the situation and timing.

Water-mediated interactions between DNA and PhoB DNA-binding/transactivation domain: NMR-restrained molecular dynamics in explicit water environment

The solution structure of the complex between the transcription factor PhoB DNA‐binding/transactivation domain and DNA was determined by NMR spectroscopy and simulated annealing in a periodic boundary box of explicit water with the particle mesh Ewald method. The refined structures provided better convergence and better local geometry compared with the structures determined in vacuum. The hydrogen bond interactions between the PhoB domain and DNA in the aqueous environment were fully formed. The complex structure was found to be very similar to the crystal structure, particularly at the PhoB‐DNA interface, much more so than expected from the vacuum structure. These results indicate the importance of the proper treatment of electrostatic and hydration influences in describing protein‐DNA interactions. The hydration structures observed for the refined structures contained most of the crystal waters as a subset. We observed that various water‐mediated PhoB‐DNA interactions contributed to the molecular recognition between PhoB and DNA. Proteins 2008.

Side-chain conformational changes of transcription factor PhoB upon DNA binding: a population-shift mechanism.

Using molecular dynamics (MD) simulations and analyses of NMR relaxation order parameters, we investigated conformational changes of side chains in hydrophobic cores upon DNA binding for the DNA binding/transactivation domain of the transcription factor PhoB, in which backbone conformational changes upon DNA binding are small. The simulation results correlated well with experimental order parameters for the backbone and side-chain methyl groups, showing that the order parameters generally represent positional fluctuations of the backbone and side-chain methyl groups. However, topological effects of the side chains on the order parameters were also found and could be eliminated using normalized order parameters for each amino acid type. Consistent with the NMR experiments, the normalized order parameters from the MD simulations showed that the side chains in one of the two hydrophobic cores (the soft core) were highly flexible in comparison with those in the other hydrophobic core (the hard core) before DNA binding and that the flexibility of the hydrophobic cores, particularly of the soft core, was reduced upon DNA binding. Principal component analysis of methyl group configurations revealed strikingly different side-chain dynamics for the soft and hard cores. In the hard core, side-chain configurations were simply distributed around one or two average configurations. In contrast, the side chains in the soft core dynamically varied their configurations in an equilibrium ensemble that included binding configurations as minor components before DNA binding. DNA binding led to a restriction of the side-chain dynamics and a shift in the equilibrium toward binding configurations, in clear correspondence with a population-shift model.

Crystallization and preliminary X-ray diffraction studies on the DNA-binding domain of the transcriptional activator protein PhoB from Escherichia coli

PhoB is a transcriptional factor that activates more than 30 genes of the pho regulon in response to phosphate starvation. Crystals of its C-terminal domain (PhoBC) were obtained in two forms. The first crystal form, obtained from phosphate solution, belongs to space group P2(1), with unit-cell parameters a = 30.7, b = 105.9, c = 30.9 A, beta = 110.3 degrees. The second form, crystallized from PEG solution, belongs to the same space group, but has a smaller unit cell (a = 30.6, b = 37.5, c = 44.4 A, beta = 109.4 degrees ). Crystals of selenomethionyl-derivatized PhoBC were obtained using the conditions for the second crystal form. Diffraction data from wild-type PhoBC (2.0 A resolution) and MAD data sets from selenomethionyl-derivative PhoBC (3.0 A resolution) have been collected at 100 K with a synchrotron-radiation source. MAD data analysis is in progress.