The Octameric Histone Core of the Nucleosome: Structural Issues Resolved

Abstract

The crystal structure of the histone octamer has now been determined at 3·1 Å resolution and refined to a crystallographic R value of 25·5%. The overall shape of the structure is significantly different from that originally reported by Burlingame et al. and its length is now in agreement with that observed by Klug et al. in their low-resolution studies. The experimental intensity data used in constructing the new electron map were the same as those used by Burlingame et al. for the original electron density map. In addition, the methods used in producing the new density map were also the same as those for the original map. The only difference between the two calculations was the selection of the heavy-atom location. The large change seen in the structural image (110 Å × 70 Å × 70 Å versus 55 Å × 70 Å × 70 Å) was due to a relatively small change (2·27 Å shift) of the heavy-atom site. The fact that the shape and size of the original structure were incorrect is surprising and unusual, since the electron density map that produced the original model was clear for most parts of the structure; one could easily see the well-formed right-handed helices of the H2A and H2B molecules, and the ordered parts of the H2A and H2B molecules could be easily traced from end to end. A comparison of the two maps shows that the original image was derived from two fused copies of the correct structure rotated by ±120° from its true location along a rotation axis parallel to the z-axis and the image seen was a partial (about 19·5%) overlap of two molecules. An explanation is given as to how such a small shift of the heavy-atom position could create such a double image in the unit cell, and how the original electron density map could be converted to the new map by a phase modification in the Fourier synthesis. This study resolves the differences between the analyses of the shape and size of the histone octamer structure.