The X-ray crystal structure of the nucleosome core particle of chromatin shows in atomic detail how the histone protein octamer is assembled and how 146 base pairs of DNA are organized into a superhelix around it. Both histone/histone and histone/DNA interactions depend on the histone fold domains and additional, well ordered structure elements extending from this motif. Histone amino-terminal tails pass over and between the gyres of the DNA superhelix to contact neighbouring particles. The lack of uniformity between multiple histone/DNA-binding sites causes the DNA to deviate from ideal superhelix geometry.

No abstract

DNA in eukaryotic chromosomes is organized in arrays of nucleosomes compacted into chromatin fibres. This higher-order structure of nucleosomes is the substrate for DNA replication, recombination, transcription and repair. Although the structure of the nucleosome core is known at near-atomic resolution, even the most fundamental information about the organization of nucleosomes in the fibre is controversial. Here we report the crystal structure of an oligonucleosome (a compact tetranucleosome) at 9 A resolution, solved by molecular replacement using the nucleosome core structure. The structure shows that linker DNA zigzags back and forth between two stacks of nucleosome cores, which form a truncated two-start helix, and does not follow a path compatible with a one-start solenoidal helix. The length of linker DNA is most probably buffered by stretching of the DNA contained in the nucleosome cores. We have built continuous fibre models by successively stacking tetranucleosomes one on another. The resulting models are nearly fully compacted and most closely resemble the previously described crossed-linker model. They suggest that the interfaces between nucleosomes along a single helix start are polymorphic.

The X-ray crystal structure of the transcription factor IIA (TFIIA) in complex with the TATA-box-binding protein (TBP) and TATA-element DNA is presented at 2.5 A resolution. TFIIA is composed of a beta-barrel and a four-helix bundle motif that together have a boot-like appearance. The beta-barrel extends the TBP beta-sheet and bridges over the DNA major groove immediately upstream of the TATA box. The four-helix bundle contributes substantially to the surface of the complex available for interaction with additional transcription factors.

Catalysis of ligand-receptor interactions is proposed as an important function of the lipid phase of the cell membrane. The catalytic mechanism is deduced from observed specific interactions of amphiphilic peptides with artificial lipid bilayers. In our model a direct ligand-receptor reaction is replaced by multiple sequential steps including surface accumulation of charged ligands, ligand-membrane interactions, and ultimately binding to the receptor itself. By dividing the total free energy of binding among several steps, the energy per step, including the intrinsic receptor interaction energy, is kept to moderate values. The model thereby yields simple explanations for the large apparent association constants, the high association and dissociation rates, and the heterogeneity of binding sites so frequently found with pharmacological and biochemical ligand-receptor interactions. Furthermore, the measured apparent association constant is a function of the whole system rather than just the receptor. The same, fully functional receptor-may show different binding characteristics in different surroundings, such as in another tissue or in a reconstituted system. Although the receptor concept was introduced early in the 20th century and has been the subject of ever more intense research, the mechanisms by which polypeptide hormones bind to cells and trigger biological responses still presents many enigmas. In this paper we will address some fundamental questions about the mechanism of hormone-receptor interactions raised by the concentration dependence of the binding and response data.The form of receptor binding curves, which often show nonlinearities in the Scatchard plot, has led to various interpretations, such as high-and low-affinity binding sites or cooperativity between sites. While undoubtedly relevant in certain cases, such explanations do not supply a satisfying general rationale for the complex behavior that is often observed.Dose-response curves often indicate EC50 values in the nanomolar range or lower, in contrast to the Kms of enzyme-substrate reactions in solution, for which 1 ,tM (e.g., arginine-tRNA ligase/arginine) represents an extremely low value. Part ofthe difference could lie in different proportionation of the intrinsic binding energy between "productive" and "nonproductive" binding energy (1), but other basic factors may also be involved.The kinetic aspects of the hormone-receptor reaction must also be considered, especially if the low EC50 values are rationalized as reflecting a lower proportion of productive binding energy-i.e, less rate enhancement due to reactant destabilization. Reaction rate theory predicts rate constants proportional to exp(-AGb/RT), where AGb represents the height of the reaction barrier. If 1000 sec-1 can be taken as a reasonable rate constant for the forward reaction (1), then, given an overall AGb = -12 kcal/mol (EC50 1 nM), the off reaction would have a rate constant on the order of 10-7 sec-1.This corresponds to a time constant of about a year, which is ...

Site-specific recognition of DNA in eukaryotic organisms depends on the arrangement of nucleosomes in chromatin. In the yeast Saccharomyces cerevisiae, ISW1a and related chromatin remodelling factors are implicated in establishing the nucleosome repeat during replication and altering nucleosome position to affect gene activity. Here we have solved the crystal structures of S. cerevisiae ISW1a lacking its ATPase domain both alone and with DNA bound at resolutions of 3.25 Å and 3.60 Å, respectively, and we have visualized two different nucleosome-containing remodelling complexes using cryo-electron microscopy. The composite X-ray and electron microscopy structures combined with site-directed photocrosslinking analyses of these complexes suggest that ISW1a uses a dinucleosome substrate for chromatin remodelling. Results from a remodelling assay corroborate the dinucleosome model. We show how a chromatin remodelling factor could set the spacing between two adjacent nucleosomes acting as a 'protein ruler'.

Recognition of and discrimination between potential glyco-substrates is central to the function of galectins. Here we dissect the fundamental parameters responsible for such selectivity by the fungal representative, CGL2. The 2.1 A crystal structure of CGL2 and five substrate complexes reveal that this prototype galectin achieves increased substrate specificity by accommodating substituted oligosaccharides of the mammalian blood group A/B type in an extended binding cleft. Kinetic studies on wild-type and mutant CGL2 proteins demonstrate that the tetrameric organization is essential for functionality. The geometric constraints due to the orthogonal orientation of the four binding sites have important consequences on substrate binding and selectivity.

No abstract