Search

The solution structure of a 16-nucleotide RNA hairpin, 5'-GCCUAG[CAAC]CUGGGC (loop bases in square brackets), has been determined by proton, phosphorus, and carbon (natural abundance) nuclear magnetic resonance (NMR) spectroscopy. This RNA tetraloop hairpin varies in four loop nucleotides from the wild-type T4 RNA hairpin (with eight loop nucleotides) involved in the translational repression of bacteriophage T4 DNA polymerase. Despite the differences in their sequence and proposed secondary structures, these two hairpins bind T4 DNA polymerase with equal affinity. The NMR spectra of the mutant hairpin indicate that its stem is extended in comparison to that of the wild-type hairpin by the formation of two additional Watson-Crick base pairs. The NMR data provide a precisely defined structure for the mutant hairpin with an average root mean square deviation of approximately 0.7 A for all 16 residues in the molecule. The structure of the mutant loop is very similar to that determined previously for the wild-type hairpin. The three loop bases that are conserved between the mutant and wild-type hairpins point out in solution with the groups capable of hydrogen bond formation exposed to the solution. This is exactly what was seen for the wild-type hairpin. Also, unusual, long-range NOEs, loop hydrogen bonds, and even the position at which the loop bends are common features between the two loops. This explains how two different hairpins, by adopting similar three-dimensional structures, have the same affinity for the DNA polymerase.

A high-resolution structure of a 16-nucleotide bacteriophage T4 RNA hairpin, 5'-GCCU[AAUAACUC]GGGC (loop bases in square brackets), has been determined in solution by proton, phosphorus, and carbon (natural abundance) NMR spectroscopy. This RNA hairpin is known to play a crucial role in the translational repression of bacteriophage T4 DNA polymerase. Ultraviolet absorbance melting curves indicate that the structure formed is unimolecular. The NMR spectra indicate that a single conformation consistent with a hairpin structure is formed. Strong imino-imino NOEs confirm the formation of the G.U base pair at the stem-loop junction. There is no evidence that A5 is protonated (at pH 6.0) and involved in an A+.C pair. However, the NMR data indicate that the stem is extended beyond the G.U pair and that A-form stacking continues for three nucleotides on the 5' side and one nucleotide on the 3' side. Structure calculations using restraints obtained from NMR data give a precisely defined structure with an average root mean square deviation (RMSD) of approximately 1.2 A for the entire molecule. The assignment of all the protons and most of the 31P resonances in the loop yielded a large number of distance and torsion angle restraints for these nucleotides. These helped obtain a well-defined loop with an average RMSD of 1.1 A for the loop nucleotides of 11 converged structures.

We determined the melting temperatures (Tm) and thermodynamic parameters of 15 RNA and 19 DNA hairpins at 1 M NaCl, 0.01 M sodium phosphate, 0.1 mM EDTA, at pH 7. All these hairpins have loops of four bases, the most common loop size in 16S and 23S ribosomal RNAs. The RNA hairpins varied in loop sequence, loop-closing base pair (A.U, C.G, or G.C), base sequence of the stem, and stem size (four or five base pairs). The DNA hairpins varied in loop sequence, loop-closing base pair (C.G, or G.C), and base sequence of the four base-pair stem. Thermodynamic properties of a hairpin may be represented by nearest-neighbor interactions of the stem plus contributions from the loop. Thus, we obtained thermodynamic parameters for the formation of RNA and DNA tetraloops. For the tetraloops we studied, a free energy of loop formation (at 37 degrees C) of about +3 kcal/mol is most common for either RNA or DNA. There are extra stable loops with delta G degrees 37 near +1 kcal/mol, but the sequences are not necessarily the same for RNA and DNA. The closing base pair is also important; changing from C.G to G.C lowered the stability of several tetraloops in both RNA and DNA. These values will be useful in predicting RNA and DNA secondary structures.

The structure of a very common RNA hairpin, 5'GGAC(UUCG)GUCC, has been determined in solution by NMR spectroscopy. The loop sequence, UUCG, occurs exceptionally often in ribosomal and other RNAs, and may serve as a nucleation site for RNA folding and as a protein recognition site. Reverse transcriptase cannot read through this loop, although it normally transcribes RNA secondary structure motifs. A hairpin with that loop displays unusually high thermodynamic stability; its stability decreases when conserved nucleotides are mutated. The three-dimensional structure for the hairpin was derived from interproton distances and scalar coupling constants determined by NMR using distance geometry, followed by restrained energy minimization. The structure was well-defined despite the conservative use of interproton distances, by constraining the backbone conformation by means of scalar coupling measurements. A mismatch G.U base pair, with syn-guanosine, closes the stem. This hairpin has a loop of only two nucleotides; both adopt C2'-endo sugar pucker. A sharp turn in the phosphodiester backbone is stabilized by a specific cytosine-phosphate contact, probably a hydrogen bond, and by stacking of the cytosine nucleotide on the G.U base pair. The structural features of the loop can explain the unusual thermodynamic stability of this hairpin and its sensitivity to mutations of loop nucleotides.

NMR and circular dichroism studies show that the RNA tetranucleotide rCpGpCpG can form a Z-RNA left-handed double-helix. In 1.0 M NaClO4, circular dichroism measurements indicate that the tetramer is in the A-form. In 6.0 M NaClO4, there is a characteristic change in the circular dichroism, indicating that the tetramer adopts a left-handed Z-form. This conformation is verified by phosphorus and proton NMR studies. The 31P spectrum shows a large downfield shift in one of the resonances upon an increase in salt concentration. Proton nuclear Overhauser effect (NOE) experiments indicate that the guanosines are in the syn conformation. These results are consistent with the formation of a Z-form double-helix.

The diverse functions of RNA, which include enzymatic activities, regulatory roles in transcription and translation, are made possible by tertiary structure. Computer algorithms can predict the secondary structure of an RNA molecule using free-energy parameters for base pairing and stacking, loops and bulges. However, with the exception of transfer RNA, little is known about the structures and thermodynamics of interactions involved in the tertiary structure of RNA. Recently, it has been proposed that a novel form of RNA folding called pseudoknotting occurs at the 3' end of certain viral RNAs from plants. A pseudoknot involves intramolecular pairing of bases in a hairpin loop with a few bases outside the stem of the loop to form an additional stem and loop region (Fig. 1). If each stem contained a full helical turn, a true knot would be formed. We present evidence from single-strand specific (S1) and double-strand specific (V1) nuclease digestion, that a short RNA oligonucleotide (19 nucleotides long) adopts a stable pseudoknotted structure. The nuclease digestion and thermodynamic properties of this oligonucleotide were compared with those of oligonucleotides which form hairpin structures containing the two possible stem regions in the pseudoknot. These results show that appropriate sequences can form pseudoknots and indicate that pseudoknots are a significant type of local tertiary structure which must be considered in the folding of complex RNA molecules.

The conformations of dinucleoside phosphates have been reexamined by semiempirical potential energy calculations. Conformations I, II, and III, proposed by Lee & Tinoco [Lee, C. H., & Tinoco, I., Jr. (1977) Biochemistry 16, 5403], are possible species after refinement of their structures by potential energy minimization. These three conformers can represent three types of dinucleoside phosphate species in solution. Dhingra et al. [Dhingra, M. M., Sarma, R. H., Giessner-Prettre, C., & Pullman, B. (1978) Biochemistry 17, 5815] had concluded that conformations of type II and III were unlikely or impossible. They favored conformations g-g- (equivalent to I), g+g+,g+t, and tg+; the last three conformations have little stacking and are calculated to be energetically less favorable by more than 5 kcal/mol. Common structures of the types I, II, and III are found for dinucleoside phosphates with different purine-pyrimidine sequences. The sequence dependence of the potential energy of these three conformers has been calculated. The experimental nuclear magnetic resonance data of dinucleoside phosphates are consistent with these three conformations.

The kinetics of U-U-C-A binding to the dodecanucleotide (A-Cm-U-Gm-A-A-Y-A-psi-m5C-U-Gp) isolated from the anticodon region of yeast tRNA-Phe are similar to the kinetics of binding of U-U-C-A to intact tRNA-Phe. A large enhancement in binding constant over that predicted for U-U-C-A-U-G-A-A is observed for both the complexes of dodecanucleotide and tRNA-Phe with U-U-C-A. This strongly suggests that both the anticodon loop in tRNA-Phe and the dodecanucleotide can form four base pairs with U-U-C-A. Furthermore, the enhanced stability cannot be attributed to a special conformation of the anticodon loop, but instead the anticodon loop is probably flexible. A likely explanation for the increased binding is the effect of non-base-paired ends. This increased thermodynamic stability comes from a larger entropy gain rather than a larger enthalpy decrease.

The kinetics of double-strand formation were measured by using temperature-jump kinetic techniques for the DNA oligonucleotides dCA5G + dCT5G, the analogous RNA oligonucleotides rCA5G + rCU5G, and the hybrid rCA5G + dCT5G. The DNA oligonucleotides have a faster rate of recombination and a slower rate of dissociation at 12.0 degrees C than the RNA oligonucleotides; the hybrid has about the same recombination rate and a slightly faster dissociation rate than the RNA oligonucleotides. The activation energies for recombination for the DNA and RNA oligonucleotides are both near 0 kcal/mol. The difference in dissociation and recombination activation energies is consistent with the thermodynamic results obtained earlier. The relaxation process is composed of two exponential components for the RNA and hybrid oligonucleotides at temperatures of 12.0 degrees C and lower. One exponential component is observed for these oligonucleotides above 12.0 degrees C and for the DNA oligonucleotides at all temperatures.

The nonexchangeable base protons and the hydrogen-bonding NH--N imino protons were used to study the conformations and the helix--coil transitions in the following oligonucleotides: (I) dCT5G + dCA5G, (II) rCU5G + rCA5G, (III) dCT5 G + rCA5G, (IV) rCU5G + dCA5G. The first three mixtures all form stable double-helical structures at 5 degrees C, whereas IV forms a triple strand with an rCU5G:dCA5G 2:1 ratio. The chemical shifts of the imino protons in the double strands indicate that I, II, and III have different conformations in solution. For example, the hydrogen-bonded proton of one of the C.G base pairs is more deshielded (a 0.4-ppm downfield shift) in helix I than in helix II or III. This implies a significant change in helical parameters, such as the winding angle, the distance between base pairs, or overlap of the bases. The coupling constants of the H1' sugar protons show that helix I has 90% 2'-endo sugar conformation, whereas helix III has greater than 85% 3'-endo conformation for the observed sugar rings. The sugar pucker data are consistent with helix I having B-family geometry; III has A-family geometry. The chemical shifts of the nonexchangeable base protons in system I were followed with increasing temperature. The midpoints for the transitions, Tm's, for all the base protons were 28--30 degrees C; this indicates an all-or-none transition.