Search

The 5'(rGGCAAGCCU)(2) duplex contains tandem A.A pairs. The three-dimensional structure of the 5'(rGGCAAGCCU)(2) duplex was modeled by molecular dynamics and energy minimization with NMR-derived distance and dihedral angle restraints. Although the 5'(rCAAG)(2) loop is thermodynamically destabilizing by 1.1 kcal/mol, the tandem A.A pairs adopt a predominant conformation: a sheared anti-anti (A.A trans Hoogsteen/Sugar-edge) alignment similar to that observed in the crystal structure of the P4-P6 domain of the Tetrahymena thermophila intron [Cate, J. H., Gooding, A. R., Podell, E., Zhou, K., Golden, B. L., Kundrot, C. E., Cech, T. R., and Doudna, J. A. (1996) Science 273, 1678-1685]. The NMR-derived structure of the 5'(rGGCAAGCCU)(2) duplex exhibits cross-strand hydrogen bonds from N3 of A4 to an amino hydrogen of A5 and from the 2' oxygen of the A4 sugar to the other amino hydrogen of A5. An intrastrand hydrogen bond is formed from the 2' OH hydrogen of A4 to O5' of A5. The cross-strand A5 bases are stacked. The Watson-Crick G-C regions are essentially A-form. The sheared anti-anti (A.A trans Hoogsteen/Sugar-edge) alignment provides potential contact sites for tertiary interactions and, therefore, is a possible target site for therapeutics. Thus, thermodynamically destabilizing internal loops can be preorganized for tertiary interactions or ligand binding.

RNA is becoming an important therapeutic target. Many potential RNA targets require secondary or tertiary structure for function. Examples include ribosomal RNAs, RNase P RNAs, mRNAs with untranslated regions that regulate translation, and group I and group II introns. Here, a method is described to inhibit RNA function by exploiting the propensity of RNA to adopt multiple folded states that are of similar free energy. This method, called oligonucleotide directed misfolding of RNA (ODMiR), uses short oligonucleotides to stabilize inactive structures. The ODMiR method is demonstrated with the group I intron from Candida albicans, a human pathogen. The oligonucleotides, (L)(TACCTTTC) and T(L)CT(L)AC(L)GA(L)CG(L)GC(L)C, with L denoting a locked nucleic acid residue, inhibit 50% of group I intron splicing in a transcription mixture at about 150 and 30 nM oligonucleotide concentration, respectively. Both oligonucleotides induce misfolds as determined by native gel electrophoresis and diethyl pyrocarbonate modification. The ODMiR approach provides a potential therapeutic strategy applicable to RNAs with secondary or tertiary structures required for function.

This overview unit discusses fluorescence spectroscopy as a tool for studying RNA folding. Ribozymes and oligonucleotides can be labeled with a fluorescent probe and analyzed to give information about both slow and fast kinetic processes with real-time data acquisition. The unit discusses the advantages and disadvantages of various pendant probes and nucleotide analogs, the analytical methods that can be used, instrument setup, control experiments, and a variety of kinetic experiments that can be performed, such as determination of rate constants.

As discussed in this overview, chemical modification is sensitive to the accessibility of a nucleotide to the solvent, and many nucleotides become less accessible as an RNA folds into its structured form. Chemical modification reagents are therefore suitable for following RNA folding, and can be used to study the kinetics of structure formation on time scales ranging from minutes to hours.

A G.A pair at position -5 in the P1 helix of the Candida albicans ribozyme contributes to tertiary binding of the 5' exon substrate [Disney, M. D., Haidaris, C. G., and Turner, D. H. (2001) Biochemistry 40, 6507-6519]. Here, the G in the G.A pair is replaced with inosine (I) in both semisynthetic ribozymes and oligonucleotide mimics of the internal guide sequence. Comparisons of oligonucleotide binding affinity for these and other sequences indicate that the G.A pair is in an imino conformation where the exocyclic amine of G contributes approximately 1.4 kcal/mol to tertiary interactions that help dock the ribozyme's P1 helix. Furthermore, replacement of the G.A pair with a G-C pair produces less favorable interactions with the 2'-hydroxyl group at the -3 position and a less favorable K(M) for pG in a ribozyme-catalyzed transesterification reaction. These results are also consistent with the G.A pair promoting docking of the P1 helix into the catalytic core. Evidently, tertiary interactions with the exocyclic amino group of a G in a single G.A pair can increase the equilibrium constant for tertiary folding of RNA by roughly 10-fold at 37 degrees C. Results with a G.U or G.G pair replacing the G.A pair at the -5 position suggest similar tertiary interactions with these pairs.

With the rapid increase in the size of the genome sequence database, computational analysis of RNA will become increasingly important in revealing structure-function relationships and potential drug targets. RNA secondary structure prediction for a single sequence is 73 % accurate on average for a large database of known secondary structures. This level of accuracy provides a good starting point for determining a secondary structure either by comparative sequence analysis or by the interpretation of experimental studies. Dynalign is a new computer algorithm that improves the accuracy of structure prediction by combining free energy minimization and comparative sequence analysis to find a low free energy structure common to two sequences without requiring any sequence identity. It uses a dynamic programming construct suggested by Sankoff. Dynalign, however, restricts the maximum distance, M, allowed between aligned nucleotides in the two sequences. This makes the calculation tractable because the complexity is simplified to O(M(3)N(3)), where N is the length of the shorter sequence. The accuracy of Dynalign was tested with sets of 13 tRNAs, seven 5 S rRNAs, and two R2 3' UTR sequences. On average, Dynalign predicted 86.1 % of known base-pairs in the tRNAs, as compared to 59.7 % for free energy minimization alone. For the 5 S rRNAs, the average accuracy improves from 47.8 % to 86.4 %. The secondary structure of the R2 3' UTR from Drosophila takahashii is poorly predicted by standard free energy minimization. With Dynalign, however, the structure predicted in tandem with the sequence from Drosophila melanogaster nearly matches the structure determined by comparative sequence analysis., (Copyright 2002 Elsevier Science Ltd.)

Algorithms for predicting RNA secondary structure require approximations for the free energies of multibranch loops, also called junctions. The stabilities of 62 RNA duplexes with three- and four-way multibranch loops were determined by optical melting. To account for the observed sequence dependence, a revised loop free-energy approximation is proposed that accounts for the strain in three-way junctions with fewer than two unpaired nucleotides, penalizes asymmetry in the distribution of unpaired nucleotides, and gives a bonus for four-way loops relative to three-way loops. Parameters for this equation were determined by linear regression.

The final step of tRNA splicing in Saccharomyces cerevisiae requires 2'-phosphotransferase (Tpt1) to transfer the 2'-phosphate from ligated tRNA to NAD, producing mature tRNA and ADP ribose-1' '-2' '-cyclic phosphate. To address how Tpt1 protein recognizes substrate RNAs, we measured the steady-state kinetic parameters of Tpt1 protein with 2'-phosphorylated ligated tRNA and a variety of related substrates. Tpt1 protein has a high apparent affinity for ligated tRNA (K(m,RNA), 0.35 nM) and a low turnover rate (k(cat), 0.3 min(-1)). Tpt1 protein recognizes both tRNA and the internal 2'-phosphate of RNAs. Steady-state kinetic analysis reveals that as RNAs lose structure and length, K(m,RNA) and k(cat) both increase commensurately. For a 2'-phosphorylated octadecamer derived from the anticodon stem-loop of ligated tRNA, K(m,RNA) and k(cat) are 5- and 8-fold higher, respectively, than for ligated tRNA, whereas for a simple substrate like pApA(p)pA, K(m,RNA) and k(cat) are 430- and 150-fold higher, respectively. Tpt1 is not detectably active on a trimer with a terminal 5'- or 3'-phosphate and is very inefficient at removal of a terminal 2'-phosphate unless there is an adjacent 3'-phosphate or phosphodiester. The K(m,NAD) for Tpt1 is substrate dependent: K(m,NAD) is 10 microM with ligated tRNA, 200 microM with pApA(p)pA, and 600 microM with pApApA(p). Preliminary analysis of KptA, a functional Tpt1 protein homologue from Escherichia coli, reveals that KptA protein is strikingly similar to yeast Tpt1 in its kinetic parameters, although E. coli is not known to have a 2'-phosphorylated RNA substrate.

UV melting experiments show that C5-(1-propynyl)ation of seven pyrimidines to give a fully propynylated oligodeoxynucleotide (PrODN) heptamer increases the thermodynamic stability of six Watson-Crick paired DNA:RNA duplexes by 8.2 kcal/mol, on average, at 37 degrees C. About 2.5 kcal/mol of this enhancement is due to long-range cooperativity between the propynylated pyrimidines, Y(p)'s. On average, penalties for dU(p):rG, dC(p):rA, dU(p):rC, and dC(p):rC mismatches are enhanced by 2.9 kcal/mol in PrODN:RNA duplexes over those in unmodified duplexes. This results in penalties as large as 10 kcal/mol for a single mismatch. Removing a single propyne two base pairs away from a mismatch in a PrODN:RNA duplex eliminates the enhancement in specificity. Evidently, enhanced specificity is directly linked to long-range cooperativity between Y(p)'s. In most cases, the enhanced specificity is larger for internal than for terminal mismatches. PrODN:RNA duplexes are destabilized by full phosphorothioate backbone substitution to give S-PrODN:RNA duplexes. The S-PrODN:RNA duplexes retain enhanced mismatch penalties, however. These results provide insight for utilizing long-range cooperativity and enhanced specificity to improve nucleic acid based probe and drug design.

Many internal loops that form tertiary contacts in natural RNAs have GU closing pairs; examples include the tetraloop receptor and P1 helix docking site in group I introns. Thus, thermodynamic parameters of internal loops with GU closing pairs can contribute to the prediction of both secondary and tertiary structure. Oligoribonucleotide duplexes containing small internal loops with GU closing pairs were studied by optical melting, one-dimensional imino proton NMR, and one-dimensional phosphorus NMR. The thermodynamic stabilities of asymmetric internal loops with GU closing pairs relative to those of loops with GC closing pairs may be explained by hydrogen bonds. In contrast, the free energy increments for symmetric internal loops of two noncanonical pairs with GU closing pairs relative to loops with GC closing pairs show much more sequence dependence. Imino proton and phosphorus NMR spectra suggest that some GA pairs adjacent to GU closing pairs may form an overall thermodynamically stable but non-A-form conformation.

RNA multibranch loops (junctions) are loops from which three or more helices exit. They are nearly ubiquitous in RNA secondary structures determined by comparative sequence analysis. In this study, systems in which two strands combine to form three-way junctions were used to measure the stabilities of RNA multibranch loops by UV optical melting and isothermal titration calorimetry (ITC). These data were used to calculate the free energy increment for initiation of a three-way junction on the basis of a nearest neighbor model for secondary structure stability. Imino proton NMR spectra were also measured for two systems and are consistent with the hypothesized helical structures. Incorporation of the experimental data into the mfold and RNA structure computer programs has contributed to an improvement in prediction of RNA secondary structure from sequence.

Candida albicans is one of many infectious pathogens that are evolving resistance to current treatments. RNAs provide a large class of targets for new therapeutics for fighting these organisms. One strategy for targeting RNAs uses short oligonucleotides that exhibit binding enhancement by tertiary interactions in addition to Watson-Crick pairing. A potential RNA target in C. albicans is the self-splicing group I intron in the LSU rRNA precursor. The recognition elements that align the 5' exon splice site for a ribozyme derived from this precursor are complex [Disney, M. D., Haidaris, C. G., and Turner, D. H. (2001) Biochemistry 40, 6507-6519]. These recognition elements have been used to guide design of hexanucleotide mimics of the 5' exon that have backbones modified for nuclease stability. These hexanucleotides bind as much as 100000-fold more tightly to a ribozyme derived from the intron than to a hexanucleotide mimic of the intron's internal guide sequence, r(GGAGGC). Several of these oligonucleotides inhibit precursor self-splicing via a suicide inhibition mechanism. The most promising suicide inhibitor is the ribophosphoramidate rn(GCCUC)rU, which forms more trans-spliced than cis-spliced product at oligonucleotide concentrations of >100 nM at 1 mM Mg(2+). The results indicate that short oligonucleotides modified for nuclease stability can target catalytic RNAs when the elements of tertiary interactions are complex.

A group I intron precursor and ribozyme were cloned from the large subunit rRNA of the human pathogen Candida albicans. Both the precursor and ribozyme are functional as determined from in vitro assays. Comparisons of dissociation constants for oligonucleotide binding to the ribozyme and to a hexanucleotide mimic of its internal guide sequence lead to a model for recognition of the 5' exon substrate by this intron. In particular, tertiary contacts with the P1 helix that help align the splice site include three 2'-hydroxyl groups, a G.U pair that occurs at the intron's splice junction, and a G.A pair. The free energy contribution that each interaction contributes to tertiary binding is determined. When the G.A pair is replaced with a G-C pair, tertiary interactions to 5' exon mimic 2'-hydroxyl groups are significantly weakened. When the G.A pair is replaced with a G.U pair, tertiary interactions are retained and binding is 10-fold tighter. These results expand our knowledge of substrate recognition by group I introns, and also provide a basis for rational design of oligonucleotide-based therapeutics for targeting group I introns by binding enhancement by tertiary interactions and suicide inhibition strategies.

A heptamer composed of C5-(1-propynyl) pyrimidines (Y(p)'s) is a potent and specific antisense agent against the mRNA of SV40 large T antigen (Wagner, R. W.; Matteucci, M. D.; Grant, D.; Huang, T.; Froehler, B. C. Nat. Biotechnol. 1996, 14, 840-844). To characterize the role of the propynyl groups in molecular recognition, thermodynamic increments associated with substitutions in DNA:RNA duplexes, such as 5'-dCCUCCUU-3':3'-rGAGGAGGAAAU-5', have been measured by UV melting experiments. For nucleotides tested, an unpaired dangling end stabilizes unmodified and propynylated duplexes similarly, except that addition of a 5' unpaired rA is 1.4 kcal/mol more stabilizing on the propynylated, PODN:RNA, duplex than on the DNA:RNA duplex. Free energy increments for addition of single propynyl groups range from 0 to -4.0 kcal/mol, depending on the final number and locations of substitutions. A preliminary model for predicting the stabilities of Y(p)-containing hybrid duplexes is presented. Eliminating one amino group, and therefore a hydrogen bond, by substituting inosine (I) for guanosine (G), to give 5'-dC(p)C(p)U(p)C(p)C(p)U(p)U(p)-3':3'-rGAGIAGGAAAU-5', destabilizes the duplex by 3.9 kcal/mol, compared to 1.7 kcal/mol for the same change within the unpropynylated duplex. This 2.2 kcal/mol difference is eliminated by removing a single propynyl group three base pairs away. CD spectra suggest that single propynyl deletions within the PODN:RNA duplex have position-dependent effects on helix geometry. The results suggest long-range cooperativity between propynyl groups and provide insights for rationally programming oligonucleotides with enhanced binding and specificity. This can be exploited in developing technologies that are dependent upon nucleic acid-based molecular recognition.

Thermodynamic parameters measured by optical melting are reported for formation of RNA duplexes containing tandem noncanonical pairs with at least one guanosine-guanosine (GG) pair. For selected sequences, imino proton NMR provides evidence that the desired duplex forms and that the structure of a GG pair adjacent to a noncanonical pair depends on context. A GG pair next to a different noncanonical pair is more stable than expected from measurements of adjacent GG pairs. This is likely due to an unfavorable stacking interaction between adjacent GG pairs, where areas of high negative charge probably overlap. The results suggest a model where tandem noncanonical pairs closed by two GC pairs are assigned the following free energy increments at 37 degrees C: 0.8 kcal/mol for adjacent GG pairs, 1.0 kcal/mol for GG next to UU, and -0.3 kcal/mol for all others. These values are adjusted by 0.65 kcal/mol for each closing AU pair.

Isoguanosine (iG) and isocytidine (iC) differ from guanosine (G) and cytidine (C), respectively, in that the amino and carbonyl groups are transposed. The thermodynamic properties of a set of iG, iC containing RNA duplexes have been measured by UV optical melting. It is found that iG-iC replacements usually stabilize duplexes, and the stabilization per iG-iC pair is sequence-dependent. The sequence dependence can be fit to a nearest-neighbor model in which the stabilities of iG--iC pairs depend on the adjacent iG--iC or G--C pairs. For 5'-CG-3'/3'-GC-5' and 5'-GG-3'/3'-CC-5' nearest neighbors, the free energy differences upon iG-iC replacement are smaller than 0.2 kcal/mol at 37 degrees C, regardless of the number of replacements. For 5'-GC-3'/3'-CG-5', however, each iG--iC replacement adds 0.6 kcal/mol stabilizing free energy at 37 degrees C. Stacking propensities of iG and iC as unpaired nucleotides at the end of a duplex are similar to those of G and C. An NMR structure is reported for r(CiGCGiCG)(2) and found to belong to the A-form family. The structure has substantial deviations from standard A-form but is similar to published NMR and/or crystal structures for r(CGCGCG)(2) and 2'-O-methyl (CGCGCG)(2). These results provide benchmarks for theoretical calculations aimed at understanding the fundamental physical basis for the thermodynamic stabilities of nucleic acid duplexes.

Pneumocystis carinii is a mammalian pathogen that infects and kills immunocompromised hosts such as cancer and AIDS patients. The LSU rRNA precursor of P. carinii contains a conserved group I intron that is an attractive drug target because humans do not contain group I introns. The oligonucleotide r(AUGACU), whose sequence mimics the 3'-end of the 5'-exon, binds to a ribozyme derived from the intron with a K(d) of 5.2 nM, which is 61000-fold tighter than expected from base-pairing alone [Testa, S. M., Haidaris, G. C., Gigliotti, F., and Turner, D. H. (1997) Biochemistry 36, 9379-9385]. Thus, oligonucleotide binding is enhanced by tertiary interactions. To localize interactions that give rise to this tertiary stability, binding to the ribozyme has been measured as a function of oligonucleotide length and sequence. The results indicate that 4.3 kcal/mol of tertiary stability is due to a G.U pair that forms at the intron's splice junction. Eliminating nucleotides at the 5'-end of r(AUGACU) does not affect intron binding more than expected from differences in base-pairing until r((___)ACU), which binds much more tightly than expected. Adding a C at the 5'- or 3'-end that can potentially form a C-G pair with the target has little effect on binding affinity. Truncated oligonucleotides were tested for their ability to inhibit intron self-splicing via a suicide inhibition mechanism. The tetramer, r((__)GACU), retains similar binding affinity and reactivity as the hexamer, r(AUGACU). Thus oligonucleotides as short as tetramers might serve as therapeutics that can use a suicide inhibition mechanism to inhibit self-splicing. Results with a phosphoramidate tetramer and thiophosphoramidate hexamer indicate that oligonucleotides with backbones stable to nuclease digestion retain favorable binding and reactivity properties.

Nucleotides in RNA that are not Watson-Crick-paired form unique structures for recognition or catalysis, but determinants of these structures and their stabilities are poorly understood. A single noncanonical pair of two guanosines (G) is more stable than other noncanonical pairs and can potentially form pairing structures with two hydrogen bonds in four different ways. Here, the energetics and structure of single GG pairs are investigated in several sequence contexts by optical melting and NMR. The data for r(5'GCAGGCGUGC3')(2), in which G4 and G7 are paired, are consistent with a model in which G4 and G7 alternate syn glycosidic conformations in a two-hydrogen-bond pair. The two distinct structures are derived from nuclear Overhauser effect spectroscopic distance restraints coupled with simulated annealing using the AMBER 95 force field. In each structure, the imino and amino protons of the anti G are hydrogen bonded to the O6 and N7 acceptors of the syn G, respectively. An additional hydrogen-bond connects the syn G amino group to the 5' nonbridging pro-R(p) phosphate oxygen. The GG pair fits well into a Watson-Crick helix. In r(5'GCAGGCGUGC3')(2), the G4(anti), G7(syn) structure is preferred over G4(syn), G7(anti). For single GG pairs in other contexts, exchange processes make interpretation of spectra more difficult but the pairs are also G(syn), G(anti). Thermodynamic data for a variety of duplexes containing pairs of G, inosine, and 7-deazaguanosine flanked by GC pairs are consistent with the structural and energetic interpretations for r(5'GCAGGCGUGC3')(2), suggesting similar GG conformations.

Optical melting experiments were used to determine the thermodynamic parameters for oligoribonucleotides containing small asymmetric internal loops. The results show a broad range of thermodynamic stabilities, which depend on loop size, asymmetry, sequence, closing base pairs, and length of helix stems. Imino proton NMR experiments provide evidence for possible hydrogen bonding in GA and UU mismatches in some asymmetric loops. The stabilizing effects of GA, GG, and UU mismatches on the thermodynamic stability of internal loops vary depending on the size and asymmetry of the loop. The dependence of loop stability on Watson-Crick closing base pairs may be explained by an account of hydrogen bonds. Models are presented for approximating the free energy increments of 2 x 3 and 1 x 3 internal loops.

G.U pairs occur frequently and have many important biological functions. The stability of symmetric tandem G.U motifs depends both on the adjacent Watson-Crick base pairs, e.g., 5'G > 5'C, and the sequence of the G.U pairs, i.e., 5'-UG-3' > 5'-GU-3', where an underline represents a nucleotide in a G.U pair [Wu, M., McDowell, J. A., and Turner, D. H. (1995) Biochemistry 34, 3204-3211]. In particular, at 37 degrees C, the motif 5'-CGUG-3' is less stable by approximately 3 kcal/mol compared with other symmetric tandem G.U motifs with G-C as adjacent pairs: 5'-GGUC-3', 5'-GUGC-3', and 5'-CUGG-3'. The solution structures of r(GAGUGCUC)(2) and r(GGCGUGCC)(2) duplexes have been determined by NMR and restrained simulated annealing. The global geometry of both duplexes is close to A-form, with some distortions localized in the tandem G.U pair region. The striking discovery is that in r(GGCGUGCC)(2) each G.U pair apparently has only one hydrogen bond instead of the two expected for a canonical wobble pair. In the one-hydrogen-bond model, the distance between GO6 and UH3 is too far to form a hydrogen bond. In addition, the temperature dependence of the imino proton resonances is also consistent with the different number of hydrogen bonds in the G.U pair. To test the NMR models, U or G in various G.U pairs were individually replaced by N3-methyluridine or isoguanosine, respectively, thus eliminating the possibility of hydrogen bonding between GO6 and UH3. The results of thermal melting studies on duplexes with these substitutions support the NMR models.

Pneumocystis carinii is a mammalian pathogen that contains a self-splicing group I intron in its large subunit rRNA precursor. We report the binding of methylphosphonate/DNA chimeras and neutral methylphosphonate oligonucleotides to a ribozyme that is a truncated form of the intron. At 15 mM Mg(2+), the nuclease-resistant all-methylphosphonate hexamer, d(AmTmGmAmCm)rU, with a sequence that mimics the 3' end of the precursor's 5' exon, binds with a dissociation constant of 272 nM. The hexamer's dissociation constant for binding by base-pairing alone to the ribozyme's binding site sequence is 8.3 mM. Thus there is a 30 000-fold binding enhancement by tertiary interactions (BETI), which is close to the 60 000-fold enhancement previously observed with the all-ribo hexamer, r(AUGACU). Evidently, backbone charge and 2' hydroxyl groups are not required for BETI. At 3-15 mM Mg(2+), the all-methylphosphonate and DNA oligonucleotides trans-splice to a truncated form of the rRNA precursor, but do not compete with cis-splicing when pG is present. These results suggest that uncharged or partially charged backbones may be used to design therapeutics to target RNAs through binding enhancement by tertiary interactions and suicide inhibition strategies.

A phosphoramidite, solid support method for the chemical synthesis of oligoribonucleotides containing 2'-O-phosphate at a selected position is presented. Synthesis of these oligoribonucleotides is based on uridine- and adenosine-(2'-O-phosphate)-3'-phosphoramidites, and a new condition for removal of 2'-O-phosphate protecting groups, which does not cleave internucleotide bonds. The structure of oligoribonucleotides with 2'-O-phosphate has been proven by enzymatic digestions and dephosphorylation by yeast 2'-phosphotransferase.

Myotonic dystrophy is caused by an expanded CTG repeat in the 3' untranslated region of the DM protein kinase (DMPK) gene. The expanded repeat triggers the nuclear retention of mutant DMPK transcripts, but the resulting underexpression of DMPK probably does not fully account for the severe phenotype. One proposed disease mechanism is that nuclear accumulation of expanded CUG repeats may interfere with nuclear function. Here we show by thermal melting and nuclease digestion studies that CUG repeats form highly stable hairpins. Furthermore, CUG repeats bind to the dsRNA-binding domain of PKR, the dsRNA-activated protein kinase. The threshold for binding to PKR is approximately 15 CUG repeats, and the affinity increases with longer repeat lengths. Finally, CUG repeats that are pathologically expanded can activate PKR in vitro. These results raise the possibility that the disease mechanism could be, in part, a gain of function by mutant DMPK transcripts that involves sequestration or activation of dsRNA binding proteins.

Antisense compounds are designed to optimize selective hybridization of an exogenous oligonucleotide to a cellular target. Typically, Watson-Crick base pairing between the antisense compound and target provides the key recognition element. Uridine (U), however, not only stably base pairs with adenosine (A) but also with guanosine (G), thus reducing specificity. Studies of duplex formation by oligonucleotides with either an internal or a terminal 2- or 4-thiouridine (s(2)U or s(4)U) show that s(2)U can increase the stability of base pairing with A more than with G, while s(4)U can increase the stability of base pairing with G more than with A. The latter may be useful when binding can be enhanced by tertiary interactions with a s(4)U-G pair. To test the effects of s(2)U and s(4)U substitutions on tertiary interactions, binding to a group I intron ribozyme from mouse-derived Pneumocystis carinii was measured for the hexamers, r(AUGACU), r(AUGACs(2)U), and r(AUGACs(4)U), which mimic the 3' end of the 5' exon. The results suggest that at least one of the carbonyl groups of the 3' terminal U of r(AUGACU) is involved in tertiary interactions with the catalytic core of the ribozyme and/or thio groups change the orientation of a terminal U-G base pair. Thus thio substitutions may affect tertiary interactions. Studies of trans-splicing of 5' exon mimics to a truncated rRNA precursor, however, indicate that thio substitutions have negligible effects on overall reactivity. Therefore, modified bases can enhance the specificity of base pairing while retaining other activities and, thus, increase the specificity of antisense compounds targeting cellular RNA.

A computer program, OligoWalk, is reported that predicts the equilibrium affinity of complementary DNA or RNA oligonucleotides to an RNA target. This program considers the predicted stability of the oligonucleotide-target helix and the competition with predicted secondary structure of both the target and the oligonucleotide. Both unimolecular and bimolecular oligonucleotide self structure are considered with a user-defined concentration. The application of OligoWalk is illustrated with three comparisons to experimental results drawn from the literature.

The thermodynamic properties and structures of single mismatches in short RNA duplexes were studied in optical melting and imino proton NMR experiments. The free energy increments at 37 degrees C measured for non-GU single mismatches range from -2.6 to 1.7 kcal/mol. These increments depend on the identity of the mismatch, adjacent base pairs, and the position in the helix. UU and AA mismatches are more stable close to a helix end, but GG mismatch stability is essentially unaffected by the position in the helix. Approximations are suggested for predicting stabilities of single mismatches in short RNA duplexes.

Free energies for stacking of unpaired nucleotides (dangling ends) at the termini of oligoribonucleotide Watson-Crick helixes (DeltaG(0)37,stack) depend on sequence for 3' ends but are always small for 5' ends. Here, these free energies are correlated with stacking at helix termini in a database of 34 RNA structures determined by X-ray crystallography and NMR spectroscopy. Stacking involving GA pairs is considered separately. A base is categorized as stacked by its distance from (

An improved dynamic programming algorithm is reported for RNA secondary structure prediction by free energy minimization. Thermodynamic parameters for the stabilities of secondary structure motifs are revised to include expanded sequence dependence as revealed by recent experiments. Additional algorithmic improvements include reduced search time and storage for multibranch loop free energies and improved imposition of folding constraints. An extended database of 151,503 nt in 955 structures? determined by comparative sequence analysis was assembled to allow optimization of parameters not based on experiments and to test the accuracy of the algorithm. On average, the predicted lowest free energy structure contains 73 % of known base-pairs when domains of fewer than 700 nt are folded; this compares with 64 % accuracy for previous versions of the algorithm and parameters. For a given sequence, a set of 750 generated structures contains one structure that, on average, has 86 % of known base-pairs. Experimental constraints, derived from enzymatic and flavin mononucleotide cleavage, improve the accuracy of structure predictions., (Copyright 1999 Academic Press.)

Binding enhancement by tertiary interactions is a strategy that takes advantage of the higher order folding of functionally important RNAs to bind short nucleic acid-based compounds tightly and more specifically than possible by simple base pairing. For example, tertiary interactions enhance binding of specific hexamers to a group I intron ribozyme from the opportunistic pathogen Pneumocystis carinii by 1,000- to 100,000-fold relative to binding by only base pairing. One such hexamer, d(AnTnGnAnCn)rU, contains an N3' --> P5' phosphoramidate deoxysugar-phosphate backbone (n) that is resistant to chemical and enzymatic decay. Here, it is shown that this hexamer is also a suicide inhibitor of the intron's self-splicing reaction in vitro. The hexamer is ligated in trans to the 3' exon of the precursor, producing dead-end products. At 4 mM Mg2+, the fraction of trans-spliced product is greater than normally spliced product at hexamer concentrations as low as 200 nM. This provides an additional level of specificity for compounds that can exploit the catalytic potential of complexes with RNA targets.

The last step of tRNA splicing in yeast is catalyzed by Tpt1 protein, which transfers the 2'-phosphate from ligated tRNA to NAD to produce ADP-ribose 1"-2"-cyclic phosphate (Appr>p). Structural and functional TPT1 homologs are found widely in eukaryotes and, surprisingly, also in Escherichia coli, which does not have this class of tRNA splicing. To understand the possible roles of the Tpt1 enzymes as well as the unusual use of NAD, the reaction mechanism of the E. coli homolog KptA was investigated. We show here that KptA protein removes the 2'-phosphate from RNA via an intermediate in which the phosphate is ADP-ribosylated followed by a presumed transesterification to release the RNA and generate Appr>p. The intermediate was characterized by analysis of its components and their linkages, using various labeled substrates and cofactors. Because the yeast and mouse Tpt1 proteins, like KptA protein, can catalyze the conversion of the KptA-generated intermediate to both product and the original substrate, these enzymes likely use the same reaction mechanism. Step 1 of this reaction is strikingly similar to the ADP-ribosylation of proteins catalyzed by a number of bacterial toxins.

The energetics of small internal loops are important for prediction of RNA secondary and tertiary structure, selection of drug target sites, and understanding RNA structure-function relationships. Hydrogen bonding, base stacking, electrostatic interactions, backbone distortion, and base-pair size compatibility all contribute to the energetics of small internal loops. Thus, the sequence dependence of these energetics are idiosyncratic. Current approximations for predicting the free energies of internal loops consider size, asymmetry, closing base pairs, and the potential to form GA, GG, or UU pairs. The database of known three-dimensional structures allows for comparison with the models used for predicting stability from sequence., (Copyright 2001 John Wiley & Sons, Inc.)

Improved thermodynamic parameters for prediction of RNA duplex formation are derived from optical melting studies of 90 oligoribonucleotide duplexes containing only Watson-Crick base pairs. To test end or base composition effects, new sets of duplexes are included that have identical nearest neighbors, but different base compositions and therefore different ends. Duplexes with terminal GC pairs are more stable than duplexes with the same nearest neighbors but terminal AU pairs. Penalizing terminal AU base pairs by 0.45 kcal/mol relative to terminal GC base pairs significantly improves predictions of DeltaG degrees37 from a nearest-neighbor model. A physical model is suggested in which the differential treatment of AU and GC ends accounts for the dependence of the total number of Watson-Crick hydrogen bonds on the base composition of a duplex. On average, the new parameters predict DeltaG degrees37, DeltaH degrees, DeltaS degrees, and TM within 3.2%, 6.0%, 6.8%, and 1.3 degreesC, respectively. These predictions are within the limit of the model, based on experimental results for duplexes predicted to have identical thermodynamic parameters.

Pneumocystis carinii is the most common lethal opportunistic pathogen infecting Acquired Immune Deficiency Syndrome (AIDS) patients, and more effective therapeutics for it are needed. P. carinii, but not humans, contain RNA self-splicing group I introns, so these functionally important RNAs are potential anti-fungal targets. In vitro, d(ATGACT), which mimics the 3' end of the 5' exon of a conserved ribosomal RNA group I intron from mouse-derived Pneumocystis carinii binds to a ribozyme that is a truncated form of this intron. The binding is about 30,000 times tighter than expected for simple base-pairing because binding is enhanced by tertiary interactions. Here we report the effects of modifying the phosphodiester backbone of d(ATGACT) with phosphorothioate and of d(ATGAC)rU with N3'-->P5' phosphoramidate linkages. The enhancement of binding by tertiary interactions is not substantially decreased, and in some cases is increased when single Rp and Sp phosphorothioate substitutions are made, although overall binding is weaker by up to 6-fold. A mixture of 5' exon mimic isomers that each contain five phosphorothioate linkages binds to the ribozyme at least 14-fold less tightly than the corresponding phosphodiester mimic. In contrast, the 5' exon mimic with five internal N3'-->P5' phosphoramidate linkages binds 4-fold more tightly than d(ATGAC)rU. This increased binding is largely due to more favorable base-pairing, but tertiary interactions still enhance binding by more than 2, 000-fold. These results indicate that chemically modified, nuclease stable 5' exon mimics can act as antisense agents with binding enhanced by tertiary interactions (BETI). This strategy permits design of short antisense agents with high specificity.

The recent increase in the population of immunocompromised patients has led to an insurgence of opportunistic human fungal infections. The lack of effective treatments against some of these pathogens makes it important to develop new therapeutic strategies. One such strategy is to target key RNAs with antisense compounds. We report the development of a model system for studying the potential for antisense targeting of group I self-splicing introns in fungal pathogens. The group I intron from the large ribosomal subunit RNA of mouse-derived Pneumocystis carinii has been isolated and characterized. This intron self-splices in vitro. A catalytically active ribozyme, P-8/4x, has been constructed from this intron to allow measurement of dissociation constants for potential antisense agents. At 37 degrees C, in 50 mM Hepes (25 mM Na+), 15 mM MgCl2, and 135 mM KCl at pH 7.5, the exogenous 5' exon mimic r(AUGACU) binds about 60 000 times more tightly to this ribozyme than to r(GGUCAU), a mimic of its complementary binding site on the ribozyme. This enhanced binding is due to tertiary interactions. This tertiary stabilization is increased by single deoxynucleotide substitutions in the exon mimic at every position except for the internal A, which is essentially unchanged. Thus 2' OH groups of the 5' exon mimic do not form stabilizing tertiary interactions with the P-8/4x ribozyme, in contrast to the Tetrahymena L-21 ScaI ribozyme. Furthermore, at 37 degrees C, the exogenous 5' exon mimic d(ATGACT) binds nearly 32 000 times more tightly to the P-8/4x ribozyme than to r(GGUCAU). Therefore, oligonucleotides without 2' OH groups can exploit tertiary stabilization to bind dramatically more tightly and with more specificity than possible from base pairing. These results suggest a new paradigm for antisense targeting: targeting the tertiary interactions of structural RNAs with short antisense oligonucleotides.

The X-ray crystallographic structure of the RNA duplex [r(CGCAIGCG)]2 has been refined to 2.5 A. It shows a symmetric internal loop of two non-Watson-Crick base pairs which form in the middle of the duplex. The tandem A-I/I-A pairs are related by a crystallographic two-fold axis. Both A(anti)-I(anti) mismatches are in a head-to-head conformation forming hydrogen bonds using the Watson-Crick positions. The octamer duplexes stack above one another in the cell forming a pseudo-infinite helix throughout the crystal. A hydrated calcium ion bridges between the 3'-terminal of one molecule and the backbone of another. The tandem A-I mismatches are incorporated with only minor distortion to the backbone. This is in contrast to the large helical perturbations often produced by sheared G-A pairs in RNA oligonucleotides.

The thermodynamic stabilities and structures of a series of RNA duplexes containing nonsymmetric tandem mismatches in the context of , where are tandem mismatches, were studied by UV melting and imino proton NMR. The contribution of one mismatch to the free energy increment for tandem mismatch formation depends on the identity of the other mismatch. Imino proton NMR indicates that this is partly because the structure of a mismatch is dependent on the adjacent mismatch. The results suggest that differences in size, shape, and hydrogen bonding of the adjacent mismatches play important roles in determining loop stability. A model for predicting stabilities of all possible tandem mismatches is proposed based on these and previous results.

The dynamics of binding of various guanosine, or G, substrates to the Tetrahymena thermophila L-21 ScaI ribozyme have been investigated by fluorescence-detected stopped-flow experiments. Upon rapid mixing of various G substrates with a preformed complex of the ribozyme and the fluorescent 5' splice site analogue CCUCUepsilonA, fluorescence transients that provide rates for binding of G substrates and the rate-limiting step for transesterification are observed. The measured apparent bimolecular rate constant for binding of pG is 10(3) M-1 s-1, much slower than expected for diffusion. pG appears to bind to the preformed complex of the ribozyme and CCUCUepsilonA in at least two steps, a bimolecular step followed by at least one conformational change. This two-step binding of pG, involving a rapid pre-equilibrium, leads to the slow apparent rate constant for binding of pG. Furthermore, the 2'-OH of pG and of the 3' terminal G of the G substrate GUCG and the nonbridging pro-Sp phosphoryl oxygen atom at the site of phosphoryl transfer on CCUCUepsilonA appear to mediate formation of a properly conformed docked ternary complex of the G substrate, 5' splice site, and ribozyme which may represent an intermediate required for initiation of transesterification. It is possible that the 2'-OH of pG and this nonbridging pro-Sp phosphoryl oxygen interact, directly or indirectly, with one another.

Transient kinetic studies with fluorescence detection were used to determine the effects of Mg2+ concentration and the 2' OH group of guanosine monophosphate, prG, substrate on various steps in the transesterification reaction of prG with 5' pyrene-labeled oligonucleotides as catalyzed by the L-21 ScaI ribozyme. The effect of increasing Mg2+ from 5 to 10 mM on the rate constants of association and dissociation of 5' pyrene-labeled CUCUA at 15 degrees C was measured. The rate constant of association increases about 3-fold to (8.7 +/- 0.7) x 10(6) M-1 s-1 at 10 mM Mg2+. The rate constant for dissociation is 25 +/- 4 s-1 at 10 mM Mg2+, within experimental error of the rate constant of 17 +/- 5 s-1 measured at 5 mM Mg2+. This Mg2+ dependence is attributed to nonspecific binding of Mg2+ to the duplex helix. In the absence of prG, no docking of substrate is observed. The effect of Mg2+ concentration on rates for docking of 5' pyrene-labeled substrate, pyrCCUCUA, were measured at [Mg2+] >/= 2 mM and at temperatures /= 4 mM. In the presence of Ca2+, such that [Ca2+] + [Mg2+] = 15 mM, the observed rate constants of both transients are constant when 4 mM

The symmetric, tandem GU mismatch motifs, and , which only differ in the mismatch order, have an average difference in thermodynamic stability of 2 kcal/mol at 37 degrees C. Thermodynamic studies of duplexes containing these motifs indicate the effect is largely localized to the mismatches and adjacent base pairs. The three-dimensional structures of two representative duplexes, (rGGAGUUCC)2 and (rGGAUGUCC)2, were determined by two-dimensional NMR and a simulated annealing protocol. Local deviations are similar to other intrahelical GU mismatches with little effect on backbone torsion angles and a slight overtwisting between the base pair 5' of the G of the mismatch and the mismatch itself. Comparisons of the resulting stacking patterns along with electrostatic potential maps suggest that interactions between highly negative electrostatic regions between base pairs may play a role in the observed thermodynamic differences.

The three-dimensional solution structure of the RNA self-complementary duplex [formula in text] was derived from two-dimensional NMR and the iterative relaxation matrix approach. Each GA mismatch forms two hydrogen bonds: A-NH6 to G-O6 and A-N1 to G-NH1 (imino). This GA structure differs from the sheared tandem GA structure in [formula in text] which also has two hydrogen bonds: A-N7 to G-NH2 and A-NH6 to G-N3 [SantaLucia, J., Jr & Turner, D. H. (1993) Biochemistry 32, 12612-12623], although the only difference between the two sequences is the order of the two GA mismatches. Inspection of three-dimensional structures indicates that substituting [formula in text] for [formula in text] makes GA mismatches unable to have stable sheared conformations. This may explain why the [formula in text] motif is rarely observed in nature, whereas [formula in text] common.

RNA transcripts corresponding to the 250-nt 3' untranslated region of the R2 non-LTR retrotransposable element are recognized by the R2 reverse transcriptase and are sufficient to serve as templates in the target DNA-primed reverse transcription (TPRT) reaction. The R2 protein encoded by the Bombyx mori R2 can recognize this region from both the B. mori and Drosophila melanogaster R2 elements even though these regions show little nucleotide sequence identity. A model for the RNA secondary structure of the 3' untranslated region of the D. melanogaster R2 retrotransposon was developed by sequence comparison of 10 species aided by free energy minimization. Chemical modification experiments are consistent with this prediction. A secondary structure model for the 3' untranslated region of R2 RNA from the R2 element from B. mori was obtained by a combination of chemical modification data and free energy minimization. These two secondary structure models, found independently, share several common sites. This study shows the utility of combining free energy minimization, sequence comparison, and chemical modification to model an RNA secondary structure.

The thermodynamics of RNA secondary structure formation in small model systems provides a database for predicting RNA structure from sequence. Methods for making these measurements are reviewed with emphasis on optical methods and treatment of experimental errors. Analysis of experimental results in terms of simple nearest-neighbor models is presented. Some measured sequence dependences of non-Watson-Crick motifs are discussed.

Optical melting experiments on oligoribonucleotides containing internal loops of three nucleotides (two nucleotides opposing one nucleotide) flanked by CG pairs show that the free energy of internal loop formation depends on the sequence of nucleotides within the loop. Internal loops with the potential for A.G and U.U mismatch pairing generally have more favorable free energies of formation than internal loops with potential AA, CC, AC, or UC pairings. In most cases, potential A.G and U.U pairings leaving a single unpaired nucleotide 5' to a flanking CG pair are most stable; those leaving a single nucleotide 3' to a flanking CG pair have intermediate stability. Internal loops of three nucleotides without potential A.G or U.U pairs are the least stable. The results provide an improved model for predicting the folding stabilities of internal loops of three nucleotides.

The duplex (rGAGGUCUC)2 contains the motif [sequence: see text] which is unusually stable compared with other symmetric tandem GU mismatches and occurs in the P5 helix of the group I intron of Tetrahymena thermophila. The three-dimensional solution structure of (rGAGGUCUC)2 was determined using two-dimensional NMR and a simulated annealing protocol. The structure is remarkably similar to the A-DNA crystal structure of (dGGGGTCCC)2 [Kneale, G., Brown, T., & Kennard, O. (1985) J. Mol. Biol. 186, 805-814] which contains the analogous motif [sequence: see text]. Incorporation of the [sequence: see text] motif has little effect on backbone torsion angles and helical parameters compared with standard A-form. The only significant departure from A-form is a slight overtwisting 5' of the G in the GU mismatch and a displacement of the mismatches toward the minor groove. Inspection of stacking patterns of this structure and comparison with symmetric tandem GT mismatches in A-DNA oligonucleotides from crystal structure data suggest that electrostatics are important in determining motif stability.

The thermodynamics of RNA helix-helix interfaces with intervening single- and tandem-GA or single-CC mismatches were studied by UV melting experiments. The model system consists of a hairpin with a four- or five-nucleotide 5' overhang which is bound by a short oligomer, creating the helical interface. Single GA interfaces are found to have favorable free energy increments of about 2 kcal/mol. This is similar to those reported for coaxially stacked flush interfaces of AU base pairs [Walter A. E., & Turner, D. H. (1994) Biochemistry 33, 12715-12719]. The free energy increment of the GA mismatches depends little on the sequence of the closing base pairs of the helixes, whether the break in the phosphate backbone is 5' or 3' with respect to the mismatch or whether the chains are extended beyond the helix-helix interface. Surprisingly, interfaces with single-CC mismatches have favorable free energy increments similar to those of GA interfaces, even though CC mismatches in coaxial stacks occur much less frequently in known RNA secondary structures. The results provide experimental support for the assumption that a bonus free energy is required for coaxially stacked helixes with intervening GA mismatches when free energy minimization is used to predict RNA secondary structures.

The three-dimensional solution structure of the RNA self-complementary duplex [sequence: see text] was derived from two-dimensional NMR and the iterative relaxation matrix approach. Each GA mismatch forms two hydrogen bonds: A-NH6 to G-O6 and A-N1 to G-NH1 (imino). This is the first three-dimensional RNA structure with imino hydrogen-bonded tandem GA mismatches. This GA structure is totally different from the sheared tandem GA structure in [sequence: see text] which also has two hydrogen bonds: A-N7 to G-NH2 and A-NH6 to G-N3 [SantaLucia, J., Jr., & Turner, D. H. (1993) Biochemistry 32, 12612-12623]. In particular, the sheared and imino GA mismatches produce a narrowing and widening of the backbone, respectively. The results show that substitutions of Watson-Crick base pairs can have dramatic effects on the three-dimensional structures of adjacent non-Watson-Crick paired regions; i.e., the structure depends on sequence context. Thus compensating substitutions in site-directed mutagenesis experiments may not always restore biological activities.

The study of the thermodynamics of base pairing is a rapidly expanding field due to its importance for taking advantage of opportunities presented by the growing database of nucleic acid sequences. Both natural and unnatural nucleic acids are being studied. The determination of nearest-neighbor parameters from optical melting data for RNA-DNA and DNA-DNA duplexes with Watson-Crick base pairs is particularly noteworthy. The foundations are also being laid for a theoretical understanding of these thermodynamics.

The hairpin loop GUAAUA occurs frequently in ribosomal RNA. Optical melting studies show that r(GGCGUAAUAGCC) folds into a hairpin containing this loop. The structural features of the r(GGCGUAAUAGCC) hairpin have been determined by NMR and molecular modeling. NOEs from G4-H1' to A9-H2 and from A9-H2 to G10-H1' show that G4 and A9 form a sheared base pair with two hydrogen bonds: A-N7 to G-NH2 and A-NH6 to G-N3. One-dimensional NOE data show no NOEs between the imino protons of U5 and U8, but NOEs are observed between the U5-H1' and the U8-H6 and U8-H5, thus orienting the U8 imino proton away from U5. Thus U5 and U8 do not form an imino hydrogen-bonded U.U pair. The U5-H2' exhibits NOEs to both the A6-H8 and A7-H8, and the 3' phosphorus resonances of U5 and A6 are shifted downfield. This suggests that the helix turn is between the U5 and A6 nucleotides. The JH1'-H2 and JH3'-H4' coupling constants indicate that the loop is dynamic, particularly at 35 degrees C, well below the melting temperature of 63 degrees C. Structures were generated using 75 distance and 46 dihedral angle restraints. In these structures, the U5 base is stacked on the sheared base pair formed by G4 and A9 and can initiate a uridine turn similar to that observed in the anticodon loop of tRNA. The A6, A7, and U8 bases can stack on one another with their hydrogen-bonding surfaces exposed to the solvent, suggesting that they are available for tertiary interactions or protein recognition in rRNA. A range of loop structures are consistent with the data, however. The lack of formation of a U.U mismatch is consistent with a recent model that predicts the stability of hairpin loops of six nucleotides on the basis of the closing base pair and first mismatch in the loop [Serra, M. J., Axenson, T. J., & Turner, D. H. (1994) Biochemistry 33, 14289-14296].

A simple model system is described which mimics the second step of splicing and reverse cyclization reactions of the self-splicing intron from Tetrahymena thermophila. This model is based on the L-21 Sca I catalyzed ligation reaction between exogenously added oligomers: cucu + UCGa L-21 Sca I cucua + UCG. Steady-state kinetics for the forward and reverse direction were measured at 15 degrees C to find oligonucleotides that exhibit Michaelis-Menten behavior with acceptable KMS. CUCU and UCGA fit both criteria and were chosen for further studies. Steady-state kinetics reveal a lag that appears to be an RNA folding step that is eliminated by preincubation of the ribozyme with 2 mM and higher [Mg2+] and by UCGA. At constant ionic strength, the Mg2+ dependence of steady-state rates exhibits a sharp maximum near 5 mM Mg2+. Pre-steady-state and steady-state kinetics, along with active-site titrations, explain the Mg2+ profile: the rate of reaction up to and including chemistry increases with Mg2+ concentration, while the fraction of active ribozyme and the rate of postchemistry steps decrease with Mg2+ concentration. The rate-limiting step at 5 mM Mg2+ for the reaction mimicking the second step of splicing is either chemistry or a conformational change before chemistry involving ribozyme bound with substrates. The rate-limiting step at 50 mM Mg2+ appears to be a postchemistry conformational change of the ribozyme or product release. At 50 mM Mg2+, single-turnover experiments support ordered binding of substrates with 5'-exon mimic binding before 3'-splice site mimic. Moreover, the 3'-splice site mimic binds and reacts in the presence of 5'-exon mimics predocked into the catalytic core. Results also indicate that Mg2+ ions associate with the ribozyme upon docking.