The location of amino acid residues within the tobacco mosaic virus protein subunit is discussed. Sequence data, X-ray crystallographic measurements, and the availability of specific residues for enzymic, immunological or chemical reaction are amongst the information used to trace roughly how the tobacco mosaic virus polypeptide chain winds in and out from the virus axis. Published rules for predicting secondary structure are then applied to obtain a diagram of the course of the polypeptide chain. This map should be useful for the interpretation of X-ray diffraction data and already permits an outline of the main features of the inner third of subunit to be suggested.Tobacco mosaic virus (TMV) protein presents a difficult problem for X-ray crystallographic structure determination. It can be crystallised as disks [1,2] but the asymmetric unit is very large (about 600000 daltons) and oriented gels of the virus itself are not sufficiently ordered to yield information to atomic resolution. Therefore, as an aid to the interpretation of current X-ray results, we attempt here to make deductions about the structure of the protein subunit from other available evidence. Such deductions impose limitations on the range of interactions between amino acid residues that may be proposed to explain other experimental data and, of course, the work also represents an exploration of the current state of the art of protein structure prediction.TMV protein is an unusually favourable subject for structure prediction because so much is already known about it. The information summarised here was obtained by many methods, the most important of which is X-ray crystallography, but also including electron microscopy, amino acid sequencing, titration studies and studies of the chemical, enzymic and immunological reactivity of various amino acid residues. The secondary structure of the polypeptide chain is then predicted, using the rules of Chou and Fasman [3,4]. Together these results enable us to propose a diagram of the approximate path of the polypeptide chain. Earlier attempts at partial TMV protein structure prediction have been made by Schiffer and Edmundson [5], and Leberman [7]. We gladly acknowledge that our thinking is based upon the labours of many people, not always explicitly mentioned here. In particular we acknowledge our debt to the crystallographic studies of Dr K. C. Holmes and his colleagues, started in Cambridge and continued, more recently, in Heidelberg.
Low-Resolution Structural InformationThe TMV protein subunit contains 158 amino acid residues and has a molecular weight of 17493 [8,9]. Electron microscope and X-ray studies upon the virus (reviewed by Klug and Caspar [lo] and more recently by Barrett et al. [ll]) show that each subunit can be contained within a sector-shaped domain, like a slice of cake, within an angle of about 22". It extends radially from about 20 8, to about 90 8, from the virus helix axis, and is about 25 8, high. The RNA fits between the protein subunits in a groove at a radius of about 40 A,...