Molecular-dynamics simulation of phenylalanine transfer RNA. I. Methods and general results

A 24-ps molecular-dynamics simulation of motions in yeast tRNA^Phe has been completed. The overall structure of the molecule is well preserved, for the motions represent fluctuations about an average structure that is very much like the crystallographic structure. The four helical stems remain intact, the structures of the loop regions do not deteriorate, and even the base stacking in the single-stranded amino acid acceptor terminus is maintained. With two exceptions, none of the sugar puckers is significantly changed. The unconstrained floppy motions of base A76 are responsible for the repuckering of ribose 76. The other sugar that repuckers is ribose, 46, and this is the result of a very small structural change in the center of the molecule that is also responsible for the breakage of one tertiary hydrogen bond. This change in local structure does not seriously distort the base-stacking and intercalation patterns where the variable loop and the D-stem interact.     

Large-amplitude bending motions in phenylalanine transfer RNA

Conformational energy calculations on yeast tRNA^Phe reveal several likely modes of intramolecular bending, including both hingelike motions (rotations about a discrete point) and distributed flexibility (deformations that bend a double-helical segment along a smooth curve). By combining these modes of motion, the molecule can be bent from the L-shaped crystallographic structure to two extremes. It can be straightened into a nearly linear conformation at an energy cost of about 50 kcal/mol, and it can be doubled over to a conformation where the anticodon and the amino acid acceptor terminus are separated by about 40 Å at an energy cost of less than 100 kcal/mol. A bending range of over 100° can be covered for 50 kcal/mol, and we estimate that this value could be cut in half with a minimization algorithm that produced optimum stereochemistry. These energies are comparable to those that would be associated with changes in solvation due to changes in surface area as the molecule bends, indicating that there are no major steric barriers to tRNA flexibility and that variations in solvent conditions and interactions with other molecules may produce large changes in the overall conformation of tRNA.     

Idealized atomic coordinates of yeast phenylalanine transfer RNA.

The atomic coordinates are given for yeast phenylalanine transfer RNA in the orthorhombic crystal form. The structure has been refined by fitting to successively improved electron density maps at 2.7 Å resolution. The model fitting has been accomplished by using an interactive computer graphics system to minimize the errors inherent in manual model building and coordinate measurements, using an optical comparator. The atomic coordinates have then been “idealized” to make bond distances, bond angles, steric conformation and non-bonded contacts close to standard values, while constraining the model to fit the electron density maps.     

Yeast phenylalanine transfer RNA: atomic coordinates and torsion angles.

The atomic coordinates of yeast phenylalanine transfer RNA (tRNA) as well as the torsion angles of the polynucleotide chain are presented as derived from an x-ray diffraction analysis of orthorhombic crystals. A comparison is made between the coordinates obtained from analysis of monoclinic crystals of the same material. It is concluded that the molecule has substantially the same form in the orthorhombic and the monoclinic lattices, except for differences found between residues at the 3' end of the polynucleotides chain. A number of observations are made concerning hydrogen bonding interactions which may account for many of the residues conserved in all tRNA sequences.     

Crystal structure of yeast phenylalanine transfer RNA. II. Structural features and functional implications.

The structural features of yeast phenylalanine transfer RNA are analyzed and documented in detail, based on atomic co-ordinates obtained from an extensive crystallographic refinement of the crystal structure of the molecule at 2.7 Å resolution (see preceding paper). We describe here: the relative orientation and the helicity of the base-paired stems; more definitive assignments of tertiary hydrogen bonds involving bases, riboses and phosphates; binding sites for magnesium hydrates, spermine and water; iriter-molecular contacts and base-stacking; flexibility of the molecule; conformational analysis of nucleotides in the structure. Among the more noteworthy features are a considerable irregularity in the helicity of the base-paired stems, a greater flexibility in the anticodon and aminoacyl acceptor arms, and a “coupling” among several conformational angles. The functional implications of these structural features are also discussed.     

Molecular dynamics of phenylalanine transfer RNA

The atomic motions of yeast phenylalanine transfer RNA have been simulated using the molecular dynamics algorithm. Two simulations were carried out for a period of 12 picoseconds, one with a normal Van der Waals potential and the other with a modified Van der Waals potential intended to mimic the effect of solvent. An analysis of large scale motions, surface exposure, root mean square displacements, helical oscillations and relaxation mechanisms reveals the maintenance of stability in the simulated structures and the general similarity of the various dynamic features of the two simulations. The regions of conformational flexibility and rigidity for tRNA(Phe) have been shown in a quantitative measure through this approach.     

Functional mutants of phenylalanine transfer RNA from Escherichia coli.

The gene pheV from Escherichia coli, coding for tRNAPhe and carried on a plasmid, has been mutagenised with hydroxylamine. Mutants in the structural gene have been identified using two criteria: (i) de-attenuation of beta-galactosidase expression, while under the control of the attenuator region of the pheS,T operon by means of an operon fusion; (ii) loss of ability to complement thermosensitivity of a mutant Phe-tRNA synthetase. Mutants showing de-attenuation were sequenced and two nucleotide changes identified: G44----A44 (found five times) and m7G46----A46 (found once). Sequencing of mutants that lost complementation identified two further tRNA mutants, C2---U2 and G15----A15; the mutant m7G46----A46 was also re-isolated by this criterion. Three of the mutants involve bases implicated in tertiary rather than secondary structure hydrogen bonding. One hypothesis for the mechanism of de-attenuation is that mutant tRNAPhe molecules compete with the wild-type tRNAPhe on the ribosome but are inefficient at some step in the elongation process.     

Strong magnesium binding sites in yeast phenylalanine transfer RNA.

To ascertain the sites that are available for strong binding between magnesium ions and phosphate groups in yeast phenylalanine transfer RNA, all distances below 5.5 A separating the phosphoryl oxygens (Op) of the 76 nucleotide residues have been computed from the latest atomic coordinates for the monoclinic form of the tRNA crystallized in the presence of magnesium chloride. The 5.5 A distance is chosen as the upper limit expected for Op....Op distances involved in strong magnesium-phosphate binding, on the basis of studies on a model magnesium phosphodiester hydrate, taking into account the quoted standard deviation in the tRNA atomic coordinates. It is concluded that there are four possible sites for strong magnesium binding in the tRNA molecule, in addition to the three sites previously reported. One of the hypothetical sites: m2G10-OL, U47-OR, could be involved in the first stage of melting of the tRNA molecule, and may be relevant to tertiary structure stabilization, since it links the dihydrouridine arm with the extra (V) loop.     

Raman spectra of ten aqueous transfer RNAs and 5S RNA. Conformational comparison with yeast phenylalanine transfer RNA.

Eleven native transfer RNAs have been prepared so as to maintain their Mg2+ content. Their aqueous Raman spectra show a high, relatively constant amount of order in the ribophosphate backbone, as indicated by the ratio 1.73 +/- 0.05 for I814/I1100 in all samples. Variation in the effectiveness of stacking of guanine and adenine bases is seen, though most of the transfer RNAs studied have a comparable degree of stacking to that found in phenylalanine transfer RNA from yeast, whose tertiary structure has been determined by X-ray crystallography. The spectrum of Escherichia coli 5S RNA indicates that the stacking efficiency of the guanine bases is much higher in 5S RNA than in yeast in phenylalanine transfer RNA, while that of the adenine bases is lower.     

Nucleotide sequence of phenylalanine transfer RNA from Schizosaccharomyces pombe: implications for transfer RNA recognition by yeast phenylalanyl-tRNA synthetase.

The nucleotide sequence of Schizosaccharomyces pombe tRNAPhe was determined to be pG-U-C-G-C-A-A-U-G**-G*-U-G-psi-A-G-D-D-G-G-G-A-G-C-A-psi-G*-A-C-A-G-A-Cm-U-Gm-A-A-Y-A-psi-m5C-U-G-U-U-G-m7G-U*-C-A-U-C-G-G-T-psi-C-G-A-U-C-C-C-G-G-U-U-U-G-U-G-A-C-A-C-C-AOH. This sequence differs from that of S. cerevisiae tRNAPhe in 27 nucleotides. Saccharomyces cerevisiae phenylalanyl-tRNA synthetase aminoacylates both the homologous tRNAPhe and S. pombe t-NAPhe; the reactions have similar Km and Vmax values. However, the nucleotide sequence in the D stem is different in the two tRNAs. This region was proposed by Roe, B., et al. [(1973) Biochemistry 12, 4146--4154] to be the major recognition site for yeast phenylalanyl-tRNA synthetase, but the present results cast doubt on the validity of this hypothesis.     

Internal motions in yeast phenylalanine transfer RNA from 13C NMR relaxation rates of modified base methyl groups: a model-free approach

Internal motions at specific locations through yeast phenylalanine tRNA were measured by using nucleic acid biosynthetically enriched in 13C at modified base methyl groups. Carbon NMR spectra of isotopically enriched tRNA(Phe) reveal 12 individual peaks for 13 of the 14 methyl groups known to be present. The two methyls of N2,N2-dimethylguanosine (m22G-26) have indistinguishable resonances, whereas the fourteenth methyl bound to ring carbon-11 of the hypermodified nucleoside 3' adjacent to the anticodon, wyosine (Y-37), does not come from the [methyl-13C]methionine substrate. Assignments to individual nucleosides within the tRNA were made on the basis of chemical shifts of the mononucleosides [Agris, P. F., Kovacs, S. A. H., Smith, C., Kopper, R. A., & Schmidt, P. G. (1983) Biochemistry 22, 1402-1408; Smith, C., Schmidt, P. G., Petsch, J., & Agris, P. F. (1985) Biochemistry 24, 1434-1440] and correlation of 13C resonances with proton NMR chemical shifts via two-dimensional heteronuclear proton-carbon correlation spectroscopy [Agris, P. F., Sierzputowska-Gracz, H., & Smith, C. (1986) Biochemistry 25, 5126-5131]. Values of 13C longitudinal relaxation (T1) and the nuclear Overhauser enhancements (NOE) were determined at 22.5, 75.5, and 118 MHz for tRNA(Phe) in a physiological buffer solution with 10 mM MgCl2, at 22 degrees C. These data were used to extract two physical parameters that define the system with regard to fast internal motion: the generalized order parameters (S2) and effective correlation times (tau e) for internal motion of the C-H internuclear vectors.(ABSTRACT TRUNCATED AT 250 WORDS)     

X-ray crystallographic studies of polymorphic forms of yeast phenylalanine transfer RNA

Yeast phenylalanine transfer RNA has been found to crystallize in five different crystal systems involving eight different space groups. The X-ray diffraction characteristics of these forms are described. One of the orthorhombic forms yields a diffraction pattern with higher resolution than either the hexagonal, the cubic or the monoclinic forms. One region of this orthorhombic diffraction pattern is particularly sensitive to X-ray exposure and to changes in the concentration of various solutes. The diffraction pattern from the cubic crystal form extends to a resolution of 3 Å, and there are a number of strong reflections in the 3 to 4 Å region which suggest that double-helical segments of the tRNA molecules are oriented along the 4-fold axes. Some comments are made regarding the nature of the polymorphism in the transfer RNA crystals.     

Orientation of double-helical segments in crystals of yeast phenylalanine transfer RNA

A search of the X-ray intensities of the P2₁ crystal form of yeast transfer RNA^Phe has revealed the orientation of the double-helical segments in the crystal. Because of the ambiguity imposed by the crystal symmetry on choosing the helices belonging to the same molecule, and because of the difficulty of determining lengths and positions of helices, a unique model cannot be deduced, but only a small number of types are possible. Among the possibilities are a “boot”-shaped molecule, which may be derived from an earlier model proposed by the author, by bending out the anticodon arm, and also an L-shaped molecule. The latter is, however, not oriented in the way proposed by Kim et al. (1973) for the case of the closely related orthorhombic crystal form.     

Accessible and inaccessible bases in yeast phenylalanine transfer RNA as studied by chemical modification.

The selective modification of cytidine, uridine, guanosine and dihydrouridine residues in ³²P-labelled yeast phenylalanine transfer RNA has been studied by the use of specific reagents.The selective modification of cytidine residues with the reagent methoxyamine is described. Of the six cytidines in the single-stranded regions of the cloverleaf formula, only two are completely reactive, C74 and C75 at the 3′-terminus. Cm32 in the anticodon loop is reactive to only a small extent.The selective modifications of uridine and guanosine residues with 1-cyclohexyl 3-[2-morpholino(4)-ethyl] carbodiimide methotosylate, is described. The reagent is also shown to be reactive with dihydrouridine. In the single-stranded regions of the secondary structure of yeast phenylalanine transfer RNA there are 16 base residues which this reagent could be specific for. However, only G20, Gm34 and U47 are extensively modified, whilst U33 and D16 are partially modified. G18 is modified to a very small extent.The results obtained in this study are also in good agreement with previous chemical modification studied by other workers, carried out on unlabelled yeast phenylalanine transfer RNA using different reagents to the ones described here.The pattern of chemical modification is compared with the three-dimensional structure obtained by an X-ray crystallographic analysis of the same tRNA species. The correlation between exposed regions of the model and the regions of chemical reactivity are everywhere consistent.