Tertiary base pair interactions in slipped loop-DNA: an NMR and model building study

A triple helix within a pseudoknot is a conserved and essential element of telomerase RNA

Telomerase copies a short template within its integral telomerase RNA onto eukaryotic chromosome ends, compensating for incomplete replication and degradation. Telomerase action extends the proliferative potential of cells, and thus it is implicated in cancer and aging. Nontemplate regions of telomerase RNA are also crucial for telomerase function. However, they are highly divergent in sequence among species, and their roles are largely unclear. Using in silico three-dimensional modeling, constrained by mutational analysis, we propose a three-dimensional model for a pseudoknot in telomerase RNA of the budding yeast Kluyveromyces lactis. Interestingly, this structure includes a U-A.U major-groove triple helix. We confirmed the triple-helix formation in vitro using oligoribonucleotides and showed that it is essential for telomerase function in vivo. While triplex-disrupting mutations abolished telomerase function, triple compensatory mutations that formed pH-dependent G-C.C(+) triples restored the pseudoknot structure in a pH-dependent manner and partly restored telomerase function in vivo. In addition, we identified a novel type of triple helix that is formed by G-C.U triples, which also partly restored the pseudoknot structure and function. We propose that this unusual structure, so far found only in telomerase RNA, provides an essential and conserved telomerase-specific function.

Triplet repeat DNA structures and human genetic disease: dynamic mutations from dynamic DNA

Fourteen genetic neurodegenerative diseases and three fragile sites have been associated with the expansion of (CTG)n (CAG)n, (CGG)n (CCG)n, or (GAA)n (TTC)n repeat tracts. Different models have been proposed for the expansion of triplet repeats, most of which presume the formation of alternative DNA structures in repeat tracts. One of the most likely structures, slipped strand DNA, may stably and reproducibly form within triplet repeat sequences. The propensity to form slipped strand DNA is proportional to the length and homogeneity of the repeat tract. The remarkable stability of slipped strand DNA may, in part, be due to loop-loop interactions facilitated by the sequence complementarity of the loops and the dynamic structure of three-way junctions formed at the loop-outs.

Human NM23/nucleoside diphosphate kinase regulates gene expression through DNA binding to nuclease-hypersensitive transcriptional elements

NM23-H2/NDP kinase B has been identified as a sequence-specific DNA-binding protein with affinity for a nuclease-hypersensitive element of the c-MYC gene promoter (Postel et al., 1993). The ability of Nm23-H2 to activate c-MYC transcription in vitro and in vivo via the same element demonstrates the biological significance of this interaction. Mutational analyses have identified Arg34, Asn69 and Lys135 as critical for DNA binding, but not required for the NDP kinase reaction. However, the catalytically important His118 residue is dispensible for sequence-specific DNA binding, suggesting that sequence-specific DNA recognition and phosphoryl transfer are independent properties. Nm23-H2 also has an activity that cleaves DNA site-specifically, involving a covalent protein-DNA complex. In a DNA sequence-dependent manner, Nm23-H2 recognizes additional target genes for activation, including myeloperoxidase, CD11b, and CCR5, all involved in myeloid-specific differentiation. Moreover, both NM23-H1 and Nm23-H2 bind to nuclease hypersensitive elements in the platelet-derived growth factor PDGF-A gene promoter sequence-specifically, correlating with either positive or negative transcriptional regulation. These data support a model in which NM23/NDP kinase modulates gene expression through DNA binding and subsequent structural transactions.

Slipped loop structure of DNA: a specific nucleotide sequence forms only one unique conformer

Earlier with some DNA sequences we were able to prove the existence of a new polynucleotide chain folding named slipped loop structure, or SLS [1,2]. However, the possibility of the presence of two SLS isomers in equilibrium was not excluded in the experiments. Here we are dealing with a specially designed structure formed by two short oligonucleotides intended for avoiding such a situation. To minimize the possibility of alternative structure formation and stabilize the conformation under investigation, the oligonucleotide sequences were designed in such a way that the bimolecular structure SLS31 would have two binding sites for antibiotic distamycin A. The sample was exposed to chemical probing both in the presence of distamycin A and without the ligand and the accessible nucleotides were mapped. The results do not suggest the presence in the solution of two isomers with different types of loop slippage without interloop interactions and strongly support the formation of a unique slipped loop conformation stabilized by an additional interloop helix, or slipped loop structure.

Three-dimensional model for slipped loop RNA

Earlier a three-dimensional model for a new unusual DNA conformation referred to as Slipped Loop Structure (SLS) has been suggested by us (1). The same type of folding could occur with RNA as well which means that one must use the A-form of the double helix rather than the B-one. The present paper discusses the creation of an all-atom stereochemically sound model for SLS-RNA. This calculated model, while possessing the same folding topology as the SLS-DNA, differs dramatically from the SLS-DNA by an overall folding geometry. It also differs radically from the RNA-pseudoknot and can thus be regarded as a new type of an RNA folding.

RNA-RNA interactions between oligonucleotide substrates for aminoacylation

RNA stem-loop microhelices with helix sequences based on tRNA acceptor stems can be charged with specific amino acids. Experiments were designed to test the possibility that microhelices could laterally associate through complementary loop sequences and thereby bring their attached aminoacyl groups close enough together to form a peptide bond. Computer simulations suggested that formation of such complexes would be sensitive to the number of loop nucleotides needed to span the grooves of the quasi-continuous helix of the intermolecular pseudoknot so formed. These predictions were conformed experimentally by observation of complex formation sensitivity to loop size. Complexes with optimized loop sizes had apparent bimolecular dissociation constants of approximately 100 nM with only three complementary base pairs between the respective loops. Single nucleotide substitutions that disrupted the predicted intermolecular loop-loop base-pairing abolished detectable association. Similarly, placing a gap between the short helix formed by loop-loop pairing and the adjacent acceptor stems also diminished complex formation. These experiments establish an experimental basis for microhelix association for peptide synthesis.