Interpretation of the 10 Å rotation function of the satellite tobacco necrosis virus

Abstract

The rotation function calculated with 10 A three-dimensional data from monoclinic crystals of the satellite tobacco necrosis virus was fitted numerically to an icosahedral axis set. The r.m.s, angular deviation of the observed peak maxima from the calculated model axis set was 0.67 ° and the largest deviation was 1.4 °. Thus, there is no significant deviation from icosahedral symmetry at 10 A resolution. An investigation of the effects of the data inclusion limits and the radius of integration on the resolution of neighboring peaks in the rotation function showed that the best resolution was obtained by using only a thin shell of the highest-resolution data available and a radius of integration no larger than the estimated diameter of the virus protein subunit.

The molecular replacement method. I. The rotation function problem, application to bovine liver catalase and STNV

“…This fact has been confirmed by Akervall et al (1971b). (However, we note that in a recent rotation function study of crystalline STNV, Lentz & Strandberg (1974) found 'no obvious correlation of peak height with the symmetry type of axis'.) Klug (1971) and Akervall et al (1971b) reinterpreted the Akervall et al (1971a) data to indicate a icosahedral symmetry point group of the STNV molecules, and explained the strong rotation function peaks corresponding to the point group O (432) as being due to rotations which relate the orientations of the two icosahedral STNV molecules in the primitive unit cell of the molecular crystal.…”

“…Moreover, it was concluded that the analysis could have been carried out in a 'logical' order by starting from the point that the rotation function data indicated the 'approximate' crystal class O (432) of the STNV molecular crystal. * It has been pointed out by one of the referees of this paper that with the present knowledge of data collection and computational procedures (data cut-off limits) (see Lentz & Strandberg, 1974) this misinterpretation of the rotation function data of STNV could have been avoided.…”

The refinement of the heavy-atom parameters in the presence of non-crystallographic symmetry

“…The other two problems were tests on model data for viruses. The first test was a simulation of satellite tobacco necrosis virus (STNV) where one whole particle is in the crystallographic asymmetric unit of the C2 cell (Lentz & Strandberg, 1974). The CPU time per independent parameter per reflection was 1.4× 10 -2 seconds.…”

Rotation function studies of southern bean mosaic virus at 22 Å resolution

“…The radius of integration was then chosen to be 100 A, rather less than the particle diameter of 280 A. The resolution limits were set between 40 and 22 fit, corresponding to the experiences of Lentz & Strandberg (1974) with a rotation function for satellite tobacco necrosis virus (STNV). The rotated Patterson synthesis was represented by the 46 largest terms of the observed 376 coefficients.…”

“…Peaks can arise in the rotation function from three sources (Lentz & Strandberg, 1974): (1) symmetry elements within the virus particle, to be termed here 'particle peaks'; (2) pseudo-symmetry elements relating one particle to another in the real crystal lattice when the particles are considered to be spherically averaged, to be termed here 'packing peaks'; (3) symmetry elements that are generated by superimposing the selfPattersons of crystallographically related particles at a common origin. The latter have been termed 'packing peaks' (Klug, 1971) but will be referred to here as 'Klug peaks'.…”