Altering the intermediate in the equilibrium folding of unmodified yeast tRNAPhe with monovalent and divalent cations

The isothermal equilibrium folding of the unmodified yeast tRNA(Phe) is studied as a function of Na(+), Mg(2+), and urea concentration with hydroxyl radical protection, circular dichroism, and diethyl pyrocarbonate (DEPC) modification. These assays indicate that this tRNA folds in Na(+) alone. Similar to folding in Mg(2+), folding in Na(+) can be described by two transitions, unfolded-to-intermediate-to-native. The I-to-N transition has a Na(+) midpoint of approximately 0.5 M and a Hill constant of approximately 4. Unexpectedly, the urea m-value, the dependence of free energy on urea concentration, for the I-to-N transition is significantly smaller in Na(+) than in Mg(2+), 0.4 versus 1.7 kcal mol(-1) M(-1), indicating that more structure is formed in the Mg(2+)-induced transition. DEPC modification indicates that the I state in Na(+)-induced folding contains all four helices of tRNA and the I-to-N transition primarily corresponds to the formation of the tertiary structure. In contrast, the intermediate in Mg(2+)-induced folding contains only three helices, and the I-to-N transition corresponds to the formation of the acceptor stem plus tertiary structure. The cation dependence of the intermediates arises from the differences in the stability of the acceptor stem and the tertiary structure. The acceptor stem is stable at a lower Na(+) concentration than required for the tertiary structure formation. The relative stability is reversed in Mg(2+) so that the acceptor stem and the tertiary structure form simultaneously in the I-to-N transition. These results demonstrate that formation of the RNA secondary structure can be independent or coupled to the formation of the tertiary structure depending on their relative stability in monovalent and divalent ions.     

Thermodynamics of RNA duplexes with tandem mismatches containing a uracil-uracil pair flanked by C.G/G.C or G.C/A.U closing base pairs

The thermodynamics governing the denaturation of RNA duplexes containing 8 bp and a central tandem mismatch or 10 bp were evaluated using UV absorbance melting curves. Each of the eight tandem mismatches that were examined had one U-U pair adjacent to another noncanonical base pair. They were examined in two different RNA duplex environments, one with the tandem mismatch closed by G.C base pairs and the other with G.C and A.U closing base pairs. The free energy increments (Delta Gdegrees(loop)) of the 2 x 2 loops were positive, and showed relatively small differences between the two closing base pair environments. Assuming temperature-independent enthalpy changes for the transitions, (Delta Gdegrees(loop)) for the 2 x 2 loops varied from 0.9 to 1.9 kcal/mol in 1 M Na(+) at 37 degrees C. Most values were within 0.8 kcal/mol of previously estimated values; however, a few sequences differed by 1.2-2.0 kcal/mol. Single strands employed to form the RNA duplexes exhibited small noncooperative absorbance increases with temperature or transitions indicative of partial self-complementary duplexes. One strand formed a partial self-complementary duplex that was more stable than the tandem mismatch duplexes it formed. Transitions of the RNA duplexes were analyzed using equations that included the coupled equilibrium of self-complementary duplex and non-self-complementary duplex denaturation. The average heat capacity change (DeltaC(p)) associated with the transitions of two RNA duplexes was estimated by plotting DeltaH degrees and DeltaS degrees evaluated at different strand concentrations as a function of T(m) and ln T(m), respectively. The average DeltaC(p) was 70 +/- 5 cal K(-)(1) (mol of base pairs)(-)(1). Consideration of this heat capacity change reduced the free energy of formation at 37 degrees C of the 10 bp control RNA duplexes by 0.3-0.6 kcal/mol, which may increase Delta Gdegrees(loop) values by similar amounts.     

The role of 5-methylcytidine in the anticodon arm of yeast tRNA(Phe): site-specific Mg2+ binding and coupled conformational transition in DNA analogs.

The tDNA(Phe)AC, d(CCAGACTGAAGAU13m5C14U15GG), with a DNA sequence similar to that of the anticodon stem and loop of yeast tRNA(Phe), forms a stem and loop structure and has an Mg(2+)-induced structural transition that was not exhibited by an unmodified tDNA(Phe)AC d(T13C14T15) [Guenther, R. H., Hardin, C. C., Sierzputowska-Gracz, H., Dao, V., & Agris, P. F. (1992) Biochemistry (preceding paper in this issue)]. Three tDNA(Phe)AC molecules having m5C14, tDNA(Phe)AC d(U13m5C14U15), d(U13m5C14T15), and d(T13,5C14U15), also exhibited Mg(2+)-induced structural transitions and biphasic thermal transitions (Tm approximately 23.5 and 52 degrees C), as monitored by CD and UV spectroscopy. Three other tDNA(Phe)AC, d(T13C14T15), d(U13C14U15), and d(A7;U13m5C14U15) in which T7 was replaced with an A, thereby negating the T7.A10 base pair across the anticodon loop, had no Mg(2+)-induced structural transitions and only monophasic thermal transitions (Tm of approximately 52 degrees C). The tDNA(Phe)AC d(U13m5C14U15) had a single, strong Mg2+ binding site with a Kd of 1.09 x 10(-6) M and a delta G of -7.75 kcal/mol associated with the Mg(2+)-induced structural transition. In thermal denaturation of tDNA(Phe)AC d(U13m5C14U15), the 1H NMR signal assigned to the imino proton of the A5.dU13 base pair at the bottom of the anticodon stem could no longer be detected at a temperature corresponding to that of the loss of the Mg(2+)-induced conformation from the CD spectrum. Therefore, we place the magnesium in the upper part of the tDNA hairpin loop near the A5.dU13 base pair, a location similar to that in the X-ray crystal structure of native, yeast tRNA(Phe).(ABSTRACT TRUNCATED AT 250 WORDS)     

Thermodynamic, spectroscopic, and equilibrium binding studies of DNA sequence context effects in four 40 base pair deoxyoligonucleotides

Effects of different end sequences on melting, circular dichroism spectra (CD), and enzyme binding properties were investigated for four 40 base pair, non-self-complementary duplex DNA oligomers. The center sequences of these oligoduplexes have either of two 22 base pair modules flanked on both sides by sequences differing in AT content. Temperature-induced melting transitions monitored by differential scanning calorimetry (DSC) and ultraviolet absorbance were measured for the six duplexes in buffered 115 mM Na(+) solutions. Values of the melting transition enthalpy, DeltaH(cal), and entropy, DeltaS(cal), were obtained directly from DSC experiments. Melting transition parameters, DeltaH(vH) and DeltaS(vH), were also estimated from a van't Hoff analysis of optical melting curves collected as a function of DNA concentration, assuming that the melting transition is two-state. Melting free energies (20 degrees C) evaluated from DSC melting experiments on the four duplex DNAs ranged from -52.2 to -77.5 kcal/mol. Free energies based on the van't Hoff analysis were -37.9 to -58.8 kcal/mol. Although the values are different, trends in the melting free energies of the four duplex DNAs as a function of sequence were identical in both DSC and optical analyses. Subject to several assumptions, values for the initiation free energy were estimated for each duplex, defined as DeltaG(int) = DeltaG(cal) - DeltaG(pred), where DeltaG(cal) is the experimental free energy at 20 degrees C determined from the experimentially measured values of the transition enthalpy, DeltaH(cal), and entropy, DeltaS(cal). The predicted free energy of the sequence, DeltaG(pred)(20 degrees C), is obtained using published nearest-neighbor sequence stability values. For three of the four duplexes, values of DeltaG(int) are essentially nil. In contrast, the duplex with 81.8% GC has a considerably higher estimate of DeltaG(int) = 7.1 kcal/mol. The CD spectra for the six duplexes collected over the wavelength range from 200 to 320 nm are also sequence-dependent. Factor analysis of the CD spectra by singular value decomposition revealed that the experimental CD spectra could be reconstructed from linear combinations of two minor and one major subspectra. Changes in the coefficients of the major subspectrum for different sequences reflect incremental sequence-dependent variations of the CD spectra. Equilibrium binding by BamHI restriction endonuclease to the 40 base pair DNAs whose central eight base pairs contain the recognition sequence for BamHI restriction enzyme bounded by A.T base pairs, 5'-A-GGATCC-A-3' was investigated. Binding assays were performed by titering BamHI against a constant concentration of each of the duplex DNA substrates, in the absence of Mg(2+), followed by analysis by gel retardation. Under the conditions employed, the enzyme binds but does not cleave the DNAs. Results of the assays revealed two binding modes with retarded gel mobilities. Binding isotherms for the fraction of bound DNA species versus enzyme concentration for each binding mode were constructed and analyzed with a simple two-step equilibrium binding model. This analysis provided semiquantitative estimates on the equilibrium binding constants for BamHI to the four DNAs. Values obtained for the binding constants varied only 7-fold and ranged from 6 x 10(-)(8) to 42 x 10(-)(8) M, with binding free energies from -8.6 to -9.7 (+/- 0.2) kcal/mol depending on the sequence that flanks the enzyme binding site. Unlike what was found earlier in binding studies of the 22 base pair duplexes that constitute the core modules of the present 40-mers [Riccelli, P. V., Vallone, P. M., Kashin, I., Faldasz, B. D., Lane, M. J., and Benight, A. S. (1999) Biochemistry 38, 11197-11208], no obvious relationship between binding and stability was found for these longer DNAs. Apparently, effects of flanking sequence stability on restriction enzyme binding may only be measurable in very short duplex deoxyoligonucl     

Editing corrects mispairing in the acceptor stem of bean and potato mitochondrial phenylalanine transfer RNAs.

Editing is a general event in plant mitochondrial messenger RNAs, but has never been detected in a plant mitochondrial transfer RNA (tRNA). We demonstrate here the occurrence of a tRNA editing event in higher plant mitochondria: in both bean and potato, the C encoded at position 4 in the mitochondrial tRNA(Phe)(GAA) gene is converted into a U in the mature tRNA. This nucleotide change corrects the mismatched C4-A69 base-pair which appears when folding the gene sequence into the cloverleaf structure and it is consistent with the fact that C to U transitions constitute the common editing events affecting plant mitochondrial messenger RNAs. The tRNA(Phe)(GAA) gene is located upstream of the single copy tRNA(Pro)(UGG) gene in both the potato and the bean mitochondrial DNAs. The sequences of potato and bean tRNA(Pro)(UGG) genes are colinear with the sequence of the mature bean mitochondrial tRNA(Pro)(UGG), demonstrating that this tRNA is not edited. A single copy tRNA(Ser)(GCU) gene was found upstream of the tRNA(Phe) gene in the potato mitochondrial DNA. A U6-U67 mismatched base-pair appears in the cloverleaf folding of this gene and is maintained in the mature potato mitochondrial tRNA(Ser)(GCU), which argues in favour of the hypothesis that the editing system of plant mitochondria can only perform C to U or occasionally U to C changes.     

Mg2+-dependent compaction and folding of yeast tRNAPhe and the catalytic domain of the B. subtilis RNase P RNA determined by small-angle X-ray scattering

We apply synchrotron-based small-angle X-ray scattering to investigate the relationship between compaction, metal binding, and structure formation of two RNAs at 37 degrees C: the 76 nucleotide yeast tRNA(Phe) and the 255 nucleotide catalytic domain of the Bacillus subtilis RNase P RNA. For both RNAs, this method provides direct evidence for the population of a distinct folding intermediate. The relative compaction between the intermediate and the native state does not correlate with the size of the RNA but does correlate well with the amount of surface burial as quantified previously by the urea-dependent m-value. The total compaction process can be described in two major stages. Starting from a completely unfolded state (4-8 M urea, no Mg(2+)), the major amount of compaction occurs upon the dilution of the denaturant and the addition of micromolar amounts of Mg(2+) to form the intermediate. The native state forms in a single transition from the intermediate state upon cooperative binding of three to four Mg(2+) ions. The characterization of this intermediate by small-angle X-ray scattering lends strong support for the cooperative Mg(2+)-binding model to describe the stability of a tertiary RNA.     

Energetic basis of molecular recognition in a DNA aptamer

The thermal stability and ligand binding properties of the L-argininamide-binding DNA aptamer (5'-GATCGAAACGTAGCGCCTTCGATC-3') were studied by spectroscopic and calorimetric methods. Differential calorimetric studies showed that the uncomplexed aptamer melted in a two-state reaction with a melting temperature T(m)=50.2+/-0.2 degrees C and a folding enthalpy DeltaH(0)(fold)=-49.0+/-2.1 kcal mol(-1). These values agree with values of T(m)=49.6 degrees C and DeltaH(0)(fold)=-51.2 kcal mol(-1) predicted for a simple hairpin structure. Melting of the uncomplexed aptamer was dependent upon salt concentration, but independent of strand concentration. The T(m) of aptamer melting was found to increase as L-argininamide concentrations increased. Analysis of circular dichroism titration data using a single-site binding model resulted in the determination of a binding free energy DeltaG(0)(bind)=-5.1 kcal mol(-1). Isothermal titration calorimetry studies revealed an exothermic binding reaction with DeltaH(0)(bind)=-8.7 kcal mol(-1). Combination of enthalpy and free energy produce an unfavorable entropy of -TDeltaS(0)=+3.6 kcal mol(-1). A molar heat capacity change of -116 cal mol(-1) K(-1) was determined from calorimetric measurements at four temperatures over the range of 15-40 degrees C. Molecular dynamics simulations were used to explore the structures of the unligated and ligated aptamer structures. From the calculated changes in solvent accessible surface areas of these structures a molar heat capacity change of -125 cal mol(-1) K(-1) was calculated, a value in excellent agreement with the experimental value. The thermodynamic signature, along with the coupled CD spectral changes, suggest that the binding of L-argininamide to its DNA aptamer is an induced-fit process in which the binding of the ligand is thermodynamically coupled to a conformational ordering of the nucleic acid.     

Anticodon domain methylated nucleosides of yeast tRNA(Phe) are significant recognition determinants in the binding of a phage display selected peptide.

The contributions of the natural modified nucleosides to RNA identity in protein/RNA interactions are not understood. We had demonstrated that 15 amino acid long peptides could be selected from a random phage display library using the criterion of binding to a modified, rather than unmodified, anticodon domain of yeast tRNA(Phe) (ASL(Phe)). Affinity and specificity of the selected peptides for the modified ASL(Phe) have been characterized by fluorescence spectroscopy of the peptides' tryptophans. One of the peptides selected, peptide t(F)2, exhibited the highest specificity and most significant affinity for ASL(Phe) modified with 2'-O-methylated cytidine-32 and guanosine-34 (Cm(32) and Gm(34)) and 5-methylated cytidine-40 (m(5)C(40)) (K(d) = 1.3 +/- 0.4 microM) and a doubly modified ASL(Phe)-Gm(34),m(5)C(40) and native yeast tRNA(Phe) (K(d) congruent with 2.3 and 3.8 microM, respectively) in comparison to that for the unmodified ASL(Phe) (K(d) = 70.1 +/- 12.3 microM). Affinity was reduced when a modification altered the ASL loop structure, and binding was negated by modifications that disfavored hairpin formation. Peptide t(F)2's higher affinity for the ASL(Phe)-Cm(32),Gm(34),m(5)C(40) hairpin and fluorescence resonance energy transfer from its tryptophan to the hypermodified wybutosine-37 in the native tRNA(Phe) placed the peptide across the anticodon loop and onto the 3'-side of the stem. Inhibition of purified yeast phenylalanyl-tRNA synthetase (FRS) catalyzed aminoacylation of cognate yeast tRNA(Phe) corroborated the peptide's binding to the anticodon domain. The phage-selected peptide t(F)2 has three of the four amino acids crucial to G(34) recognition by the beta-structure of the anticodon-binding domain of Thermus thermophilus FRS and exhibited circular dichroism spectral properties characteristic of beta-structure. Thus, modifications as simple as methylations contribute identity elements that a selected peptide specifically recognizes in binding synthetic and native tRNA and in inhibiting tRNA aminoacylation.     

Functional anticodon architecture of human tRNALys3 includes disruption of intraloop hydrogen bonding by the naturally occurring amino acid modification, t6A

The structure of the human tRNA(Lys3) anticodon stem and loop domain (ASL(Lys3)) provides evidence of the physicochemical contributions of N6-threonylcarbamoyladenosine (t(6)A(37)) to tRNA(Lys3) functions. The t(6)A(37)-modified anticodon stem and loop domain of tRNA(Lys3)(UUU) (ASL(Lys3)(UUU)- t(6)A(37)) with a UUU anticodon is bound by the appropriately programmed ribosomes, but the unmodified ASL(Lys3)(UUU) is not [Yarian, C., Marszalek, M., Sochacka, E., Malkiewicz, A., Guenther, R., Miskiewicz, A., and Agris, P. F., Biochemistry 39, 13390-13395]. The structure, determined to an average rmsd of 1.57 +/- 0.33 A (relative to the mean structure) by NMR spectroscopy and restrained molecular dynamics, is the first reported of an RNA in which a naturally occurring hypermodified nucleoside was introduced by automated chemical synthesis. The ASL(Lys3)(UUU)-t(6)A(37) loop is significantly different than that of the unmodified ASL(Lys3)(UUU), although the five canonical base pairs of both ASL(Lys3)(UUU) stems are in the standard A-form of helical RNA. t(6)A(37), 3'-adjacent to the anticodon, adopts the form of a tricyclic nucleoside with an intraresidue H-bond and enhances base stacking on the 3'-side of the anticodon loop. Critically important to ribosome binding, incorporation of the modification negates formation of an intraloop U(33).A(37) base pair that is observed in the unmodified ASL(Lys3)(UUU). The anticodon wobble position U(34) nucleobase in ASL(Lys3)(UUU)-t(6)A(37) is significantly displaced from its position in the unmodified ASL and directed away from the codon-binding face of the loop resulting in only two anticodon bases for codon binding. This conformation is one explanation for ASL(Lys3)(UUU) tendency to prematurely terminate translation and -1 frame shift. At the pH 5.6 conditions of our structure determination, A(38) is protonated and positively charged in ASL(Lys3)(UUU)-t(6)A(37) and the unmodified ASL(Lys3)(UUU). The ionized carboxylic acid moiety of t(6)A(37) possibly neutralizes the positive charge of A(+)(38). The protonated A(+)(38) can base pair with C(32), but t(6)A(37) may weaken the interaction through steric interference. From these results, we conclude that ribosome binding cannot simply be an induced fit of the anticodon stem and loop, otherwise the unmodified ASL(Lys3)(UUU) would bind as well as ASL(Lys3)(UUU)-t(6)A(37). t(6)A(37) and other position 37 modifications produce the open, structured loop required for ribosomal binding.     

Energetics of DNA intercalation reactions

Isothermal titration calorimetry has been used to determine the binding enthalpy and heat capacity change (DeltaC(p)()) for a series of DNA intercalators, including ethidium, propidium, daunorubicin, and adriamycin. Temperature-dependent binding enthalpies were measured directly for the ligands, from which DeltaC(p)() values of -140 to -160 cal mol(-)(1) K(-)(1) were calculated. Published van't Hoff plots were reanalyzed to obtain DeltaC(p)() values of -337 to -423 cal mol(-)(1) K(-)(1) for the binding of actinomycin D to several DNA oligonucleotide duplexes with defined sequences. Heat capacity changes for DNA intercalation were found to correlate with the alterations in solvent-accessible surface area calculated from available high-resolution structural data. Multiple linear regression was used to derive the relationship DeltaC(p)() = 0. 382(+/-0.026)DeltaA(np) - 0.121(+/-0.077)DeltaA(p) cal mol(-)(1) K(-)(1), where DeltaA(np) and DeltaA(p) are the binding-induced changes in nonpolar and polar solvent-accessible surface areas (in square angstroms), respectively. The DeltaC(p)() terms were used to estimate the hydrophobic contribution to intercalative binding free energies, yielding values that ranged from -11.2 (ethidium) to -30 kcal mol(-)(1) (actinomycin D). An attempt was made to parse the observed binding free energies of ethidium and propidium into five underlying contributions. Such analysis showed that the DNA binding behavior of these simple intercalators is driven almost equally by hydrophobic effects and van der Waals contacts within the intercalation site.     

Mg2+-induced tRNA folding

Mg(2+)-induced folding of yeast tRNA(Phe) was examined at low ionic strength in steady-state and kinetic experiments. By using fluorescent labels attached to tRNA, four conformational transitions were revealed when the Mg(2+) concentration was gradually increased. The last two transitions were not accompanied by changes in the number of base pairs. The observed transitions were attributed to Mg(2+) binding to four distinct types of sites. The first two types are strong sites with K(diss) of 4 and 16 microM. The sites of the third and fourth types are weak with a K(diss) of 2 and 20 mM. Accordingly, the Mg(2+)-binding sites previously classified as "strong" and "weak" can be further subdivided into two subtypes each. Fluorescent transition I is likely to correspond to Mg(2+) binding to a unique strong site selective for Mg(2+); binding to this site causes only minor A(260) change. The transition at 2 mM Mg(2+) is accompanied by substantial conformational changes revealed by probing with ribonucleases T1 and V1 and likely enhances stacking of the tRNA bases. Fast and slow kinetic phases of tRNA refolding were observed. Time-resolved monitoring of Mg(2+) binding to tRNA suggested that the slow kinetic phase was caused by a misfolded tRNA structure formed in the absence of Mg(2+). Our results suggest that, similarly to large RNAs, Mg(2+)-induced tRNA folding exhibits parallel folding pathways and the existence of kinetically trapped intermediates stabilized by Mg(2+). A multistep scheme for Mg(2+)-induced tRNA folding is discussed.     

Spectroscopic determination of the binding affinity of zinc to the DNA-binding domains of nuclear hormone receptors

Zinc binding to the two Cys(4) sites present in the DNA-binding domain (DBD) of nuclear hormone receptor proteins is required for proper folding of the domain and for protein activity. By utilizing Co(2+) as a spectroscopic probe, we have characterized the metal-binding properties of the two Cys(4) structural zinc-binding sites found in the DBD of human estrogen receptor alpha (hERalpha-DBD) and rat glucocorticoid receptor (GR-DBD). The binding affinity of Co(2+) to the two proteins was determined relative to the binding affinity of Co(2+) to the zinc finger consensus peptide, CP-1. Using the known dissociation constant of Co(2+) from CP-1, the dissociation constants of cobalt from hERalpha-DBD were calculated: K(d1)(Co) = 2.2 (+/- 1.0) x 10(-7) M and K(d2)(Co) = 6.1 (+/- 1.5) x 10(-7) M. Similarly, the dissociation constants of Co(2+) from GR-DBD were calculated: K(d1)(Co) = 4.1 (+/- 0.6) x 10(-7) M and K(d2)(Co) = 1.7 (+/- 0.3) x 10(-7) M. Metal-binding studies conducted in which Zn(2+) displaces Co(2+) from the metal-binding sites of hERalpha-DBD and GR-DBD indicate that Zn(2+) binds to each of the Cys(4) metal-binding sites approximately 3 orders of magnitude more tightly than Co(2+) does: the stoichiometric dissociation constants are K(d1)(Zn) = 1 (+/- 1) x 10(-10) M and K(d2)(Zn) = 5 (+/- 1) x 10(-10) M for hERalpha-DBD and K(d1)(Zn) = 2 (+/- 1) x 10(-10) M and K(d2)(Zn) = 3 (+/- 1) x 10(-10) M for GR-DBD. These affinities are comparable to those observed for most other naturally occurring structural zinc-binding sites. In contrast to the recent prediction by Low et. al. that zinc binding in these systems should be cooperative [Low, L. Y., Hernández, H., Robinson, C. V., O'Brien, R., Grossmann, J. G., Ladbury, J. E., and Luisi, B. (2002) J. Mol. Biol. 319, 87-106], these data suggest that the zincs that bind to the two sites in the DBDs of hERalpha-DBD and GR-DBD do not interact.