No abstract

The bases of yeast tRNA(Phe) which react with carbodiimide and methoxyamine have been determined and this information has been combined with chemical modification studies of other workers to produce a composite picture of base accessibility in this tRNA. The results are compared with the three-dimensional structure which we have recently determined. The bases which react chemically lie in exposed positions in the three-dimensional model and those which do not are either in the double helical stem regions or else are involved in maintaining the tertiary structure through pairing or stacking interactions.

No abstract

A comparison is made between the electron density maps of the monoclinic and orthorhombic crystal forms of yeast tRNAPhe which have been obtained respectively by ourselves and by another group. It is concluded that the molecular structures are essentially the same in both crystals, although the models derived from the maps are not the same. The relation between the two molecular packings is discussed, and it is suggested that the intermolecular contact in the orthorhombic form which is not present in the monoclinic form, may arise through base pairing of the anticodons of neighboring molecules.We have recently determined the crystal structure of a monoclinic form of yeast phenylalanine tRNA at 3-A resolution by x-ray crystallographic analysis (1), using the method of isomorphous replacement to obtain a map of the electron density in the crystal. The map is of sufficient quality to trace the ribose-phosphate chain over all the double helical regions of the familiar clover leaf formula and also over most of the loop sequences in between. As a result, we have been able to build a model which defines most of the various interactions involved in maintaining the tertiary structure. A model of the structure of the same molecule has also been proposed by a group at MIT who have obtained an electron density map of the orthorhombic crystal form to the same resolution (2). Although the crystal forms are different, there are common features in the unit-cell dimensions and symmetry and also in the distribution of x-ray intensities, so that the two structures are expected to be very similar (3). Our model has a similar overall shape to the MIT model, but our density maps differ in detail and our interpretations have major differences. Most of the nucleotide residues are differently placed in the two models and none of the tertiary interactions we have described are present in the MIT model.In a recent paper (4), the MIT workers have attempted to predict the density in the monoclinic crystal from the density map which they have obtained of the orthorhombic form. This procedure, though theoretically sound, can have many pitfalls in practice and may lead to questionable results at high resolution.Since a density map for the monoclinic crystal has now been obtained directly (1), it is possible to make a comparison of independent electron density maps for the two crystal structures. We do so in this paper. We conclude that the molecular configurations are virtually identical in the two crystal forms, but that there are errors in the MIT interpretation, not only in the loop regions, but also in the double * Present address: Department of Chemistry, The University of Texas, Austin, Texas. 3711 helical stem regions. The suggestion by the MIT group that the anticodon loop may have a slightly different conformation in the two structures is examined and seen to be wanting in evidence. We also discuss the relationship between the packing of the molecule in the two crystal forms. A suggestion is made on the nature of the new inter...