Duplex-tetraplex equilibrium between a hairpin and two interacting hairpins of d(A-G)10 at neutral pH

Mixed guanine, adenine base quartets: possible roles of protons and metal ions in their stabilization

Structural variations of the well-known guanine quartet (G₄) motif in nucleic acid structures, namely substitution of two guanine bases (G) by two adenine (A) nucleobases in mutual trans positions, are discussed and studied by density functional theory (DFT) methods. This work was initiated by three findings, namely (1) that GA mismatches are compatible with complementary pairing patterns in duplex-DNA structures and can, in principle, be extended to quartet structures, (2) that GA pairs can come in several variations, including with a N1 protonated adeninium moiety (AH), and (3) that cross-linking of the major donor sites of purine nucleobases (N1 and N7) by transition metal ions of linear coordination geometries produces planar purine quartets, as demonstrated by some of us in the past. Here, possible structures of mixed AGAG quartets both in the presence of protons and alkali metal ions are discussed, and in particular, the existence of a putative four-purine, two-metal motif.

Double zipper helical assembly of deoxyoligonucleotides: mutual templating and chiral imprinting to form hybrid DNA ensembles

Herein, the conventional and unconventional hydrogen bonding potential of adenine in APA for double zipper helical assembly of deoxyoligonucleotides is demonstrated under ambient conditions. The quantum mechanical calculations supported the formation of hybrid DNA ensembles.

Fragile X mental retardation protein interactions with the microtubule associated protein 1B RNA

Fragile X mental retardation syndrome, the most common form of inherited mental retardation, is caused by the absence of the fragile X mental retardation protein (FMRP). FMRP has been shown to use its arginine-glycine-glycine (RGG) box to bind to a subset of RNA targets that form a G quadruplex structure. We performed a detailed analysis of the interactions between the FMRP RGG box and the microtubule associated protein 1B (MAP1B) mRNA, a relevant in vivo FMRP target. We show that MAP1B RNA forms an intramolecular G quadruplex structure, which is bound with high affinity and specificity by the FMRP RGG box. We determined that hydrophobic interactions are important in the FMRP RGG box-MAP1B RNA association, with minor contributions from electrostatic interactions. Our findings that at low protein:RNA ratios the RNA G quadruplex structure is slightly stabilized, whereas at high ratios is unfolded, suggest a mechanism by which the FMRP concentration variation in response to a neurotransmitter stimulation event could act as a regulatory switch for the protein function, from translation repressor to translation activator.

Interactions of the G quartet forming semaphorin 3F RNA with the RGG box domain of the fragile X protein family

Fragile X syndrome, the most common cause of inherited mental retardation, is caused by the transcriptional silencing of the fmr1 gene due to an unstable expansion of a CGG trinucleotide repeat and its subsequent hypermethylation in its 5′ UTR. This gene encodes for the fragile X mental retardation protein (FMRP), an RNA-binding protein that has been shown to use its RGG box domain to bind to G quartet-forming RNA. In this study, we performed a detailed analysis of the interactions between the FMRP RGG box domain and one of its proposed RNA targets, human semaphorin 3F (S3F) RNA by using biophysical methods such as fluorescence, UV and circular dichroism spectroscopy. We show that this RNA forms a G quartet-containing structure, which is recognized with high affinity and specificity by the FMRP RGG box. In addition, we analyzed the interactions of human S3F RNA with the RGG box and RG cluster of the two FMRP autosomal paralogs, the FXR1P and FXR2P. We found that this RNA is bound with high affinity and specificity only by the FXR1P RGG box, but not by the FXR2P RG cluster. Both FMRP and FXR1P RGG box are able to unwind the G quartet structure of S3F RNA, however, the peptide concentrations required in this process are very different: a ratio of 1:6 RNA:FMRP RGG box versus 1:2 RNA:FXR1P RGG box.

Structural polymorphism exhibited by a quasipalindrome present in the locus control region (LCR) of the human beta-globin gene cluster

Structural polymorphism of DNA is a widely accepted property. A simple addition to this perception has been our recent finding, where a single nucleotide polymorphism (SNP) site present in a quasipalindromic sequence of beta-globin LCR exhibited a hairpin-duplex equilibrium. Our current studies explore that secondary structures adopted by individual complementary strands compete with formation of a perfect duplex. Using gel-electrophoresis, ultraviolet (UV)-thermal denaturation, circular dichroism (CD) techniques, we have demonstrated the structural transitions within a perfect duplex containing 11 bp quasipalindromic stretch (TGGGG(G/C)CCCCA), to hairpins and bulge duplex forms. The extended version of the 11 bp duplex, flanked by 5 bp on both sides also demonstrated conformational equilibrium between duplex and hairpin species. Gel-electrophoresis confirms that the duplex coexists with hairpin and bulge duplex/cruciform species. Further, in CD spectra of duplexes, presence of two overlapping positive peaks at 265 and 285 nm suggest the features of A- as well as B-type DNA conformation and show oligomer concentration dependence, manifested in A --> B transition. This indicates the possibility of an architectural switching at quasipalindromic region between linear duplex to a cruciform structure. Such DNA structural variations are likely to be found in the mechanics of molecular recognition and manipulation by proteins.

Towards a better understanding of the unusual conformations of the alternating guanine–adenine repeat strands of DNA

Alternating guanine-adenine strands of DNA are known to self-associate into a parallel-stranded homoduplex at neutral pH, fold into an ordered single-stranded structure at acid pH, and adopt yet another ordered single-stranded conformer in aqueous ethanol. The unusual conformers melt cooperatively and exhibit distinct circular dichroism spectra suggestive of a substantial conformational order, but their molecular structures are not known yet. Here, we have probed the molecular structures using guanine and adenine analogs lacking the N7 atom, and thus unable of Hoogsteen pairing, or those restrained in the less-frequent syn glycosidic orientation. The studies showed that the syn glycosidic orientation of dA residues promoted the neutral homoduplex, whereas the syn orientation of dG was incompatible with the homoduplex. In addition, Hoogsteen pairing of dA seemed to be a crucial property of the homoduplex whereas dG did not pair in this way. The situation was the same in both single-stranded conformers with the dG residues. On the other hand, the presence of N7 was important with dA but its syn geometry was not favorable. The present data can be used as restraints to model the unusual molecular structures of the alternating guanine-adenine strands of DNA.

Feasibility of a two-metal, four-purine nucleobase quartet motif

A mixed-purine nucleobase complex of composition trans-[(NH(3))(2)Pt(9-EtA-N1)(9-MeHx-N7)](NO(3))(2).2H(2)O (1) (9-EtA = 9-ethyladenine; 9-MeHx = 9-methylhypoxanthine) has been prepared and characterized by X-ray crystallography. Cations of 1 are self-complementary as far as hydrogen bonding properties are concerned and form H bonded dimers, containing four intermolecular hydrogen bonds in addition to two intramolecular ones. The resulting mixed-purine square is considered a model compound for a putative mixed-purine tetrad consisting of two adenines and two guanines. In this model, the one-metal, four-nucleobase quartet motif, as seen in guanine or uracil quartets of nucleic acids, with the metal located in the center of the base tetrad, has been altered to a two-metal, four-nucleobase motif, with the two metal ions localized at the periphery.

Circular dichroism spectroscopy of conformers of (guanine + adenine) repeat strands of DNA

(Guanine+adenine) strands of DNA are known to associate into guanine tetraplexes, homodimerize into parallel or antiparallel duplexes, and fold into a cooperatively melting single strand resembling the protein alpha helix. Using CD spectroscopy and other methods, we studied how this conformational polymorphism depended on the primary structure of DNA. The study showed that d(GGGA)(5) and d(GGA)(7) associated into homoduplexes at low salt or in the presence of LiCl but were prone to guanine tetraplex formation, especially in the presence of KCl. In addition, they yielded essentially the same CD spectrum in the presence of ethanol as observed with the ordered single strand of d(GA)(10). Strands of d(GA)(10), d(GGAA)(5), d(GAA)(7), and d(GAAA)(5) associated into homoduplexes in both LiCl and KCl solutions, but not into guanine tetraplexes. d(GAAA)(5) and d(GAA)(7) further failed to form the single-stranded conformer in aqueous ethanol. Adenine protonation, however, stabilized the single-stranded conformer even in these adenine-rich fragments. The ordered single strands, homoduplexes as well as the guanine tetraplexes, all provided strikingly similar CD spectra, indicating that all of the conformers shared similar base stacking geometries. The increasing adenine content only decreased the conformer thermostability.

Hairpin formation in Friedreich's ataxia triplet repeat expansion

Triplet repeat tracts occur throughout the human genome. Expansions of a (GAA)(n)/(TTC)(n) repeat tract during its transmission from parent to child are tightly associated with the occurrence of Friedreich's ataxia. Evidence supports DNA slippage during DNA replication as the cause of the expansions. DNA slippage results in single-stranded expansion intermediates. Evidence has accumulated that predicts that hairpin structures protect from DNA repair the expansion intermediates of all of the disease-associated repeats except for those of Friedreich's ataxia. How the latter repeat expansions avoid repair remains a mystery because (GAA)(n) and (TTC)(n) repeats are reported not to self-anneal. To characterize the Friedreich's ataxia intermediates, we generated massive expansions of (GAA)(n) and (TTC)(n) during DNA replication in vitro using human polymerase beta and the Klenow fragment of Escherichia coli polymerase I. Electron microscopy, endonuclease cleavage, and DNA sequencing of the expansion products demonstrate, for the first time, the occurrence of large and growing (GAA)(n) and (TTC)(n) hairpins during DNA synthesis. The results provide unifying evidence that predicts that hairpin formation during DNA synthesis mediates all of the disease-associated, triplet repeat expansions.

Interaction in vitro of type III intermediate filament proteins with triplex DNA

As previously shown, type III intermediate filaments (IFs) select from a mixture of linear mouse genomic DNA fragments mobile and repetitive, recombinogenic sequences that have also been identified in SDS-stable crosslinkage products of vimentin and DNA isolated from intact fibroblasts. Because these sequences also included homopurine.homopyrimidine (Pu.Py) tracts known to adopt triple-helical conformation under superhelical tension, and because IF proteins are single-stranded (ss) and supercoiled DNA-binding proteins, it was of interest whether they have a particular affinity for triplex DNA. To substantiate this, IF-selected DNA fragments harboring a (Pu.Py) segment and synthetic d(GA)(n) microsatellites were inserted into a vector plasmid and the constructs analyzed for their capacity to interact with IF proteins. Band shift assays revealed a substantially higher affinity of the IF proteins for the insert-containing plasmids than for the empty vector, with an activity decreasing in the order of vimentin > glial fibrillary acidic protein > desmin. In addition, footprint analyses performed with S1 nuclease, KMnO(4), and OsO(4)/bipyridine showed that the (Pu.Py) inserts had adopted triplex conformation under the superhelical strain of the plasmids, and that the IF proteins protected the triple-helical insert sequences from nucleolytic cleavage and chemical modification. All these activities were largely reduced in extent when analyzed on linearized plasmid DNAs. Because intramolecular triplexes (H-DNA) expose single-stranded loops, and the prokaryotic ssDNA-binding proteins g5p and g32p also protected at least the Pu-strand of the (Pu.Py) inserts from nucleolytic degradation, it seemed likely that the IF proteins take advantage of their ssDNA-binding activity in interacting with H-DNA. However, in contrast to g5p and E. coli SSB, they produced no clear band shifts with single-stranded d(GA)(20) and d(TC)(20), so that the interactions rather appear to occur via the duplex-triplex and triplex-loop junctions of H-DNA. On the other hand, the IF proteins, and also g32p, promoted the formation of intermolecular triplexes from the duplex d[A(GA)(20).(TC)(20)T] and d(GA)(20) and d(TC)(20) single strands, with preference of the Py (Pu.Py) triplex motif, substantiating an affinity of the proteins for the triplex structure as such. This triplex-stabilizing effect of IF proteins also applies to the H-DNA of (Pu.Py) insert-containing plasmids, as demonstrated by the preservation of intramolecular triplex-vimentin complexes upon linearization of their constituent supercoiled DNAs, in contrast to poor complex formation from free, linearized plasmid DNA and vimentin. Considering that (Pu.Py) sequences are found near MAR/replication origins, in upstream enhancer and promoter regions of genes, and in recombination hot spots, these results might point to roles of IF proteins in DNA replication, transcription, recombination, and repair.

G-quadruplex DNA structures--variations on a theme

To be functional, nucleic acids need to adopt particular three-dimensional structures. For a long time DNA was regarded as a rigid and passive molecule with the sole purpose to store genetic information, but experimental data has now accumulated that indicates the full dynamic repertoire of this macromolecule. During the last decade, four-stranded DNA structures known as G-quadruplexes, or DNA tetraplexes, have emerged as a three-dimensional structure of special interest. Motifs for the formation of G-quadruplex DNA structures are widely dispersed in eukaryotic genomes, and are abundant in regions of biological significance, for example, at telomeres, in the promoters of many important genes, and at recombination hotspots, to name but a few in man. Here I explore the plethora of G-quadruplex DNA structures, and discuss their possible biological functions as well as the proteins that interact with them.

DNA tetraplex formation studied with fluorescence resonance energy transfer

It is emerging that DNA tetraplexes are pivotal for many major cellular processes, and techniques that assess their structure and nature to the point are under development. Here we show how the structural conversion of largely unstructured single-stranded DNA molecules into compact intrastrand DNA tetraplexes can be monitored by fluorescence resonance energy transfer. We recently reported that intrastrand tetraplex formation takes place in a nuclease hypersensitive element upstream of the human c-myc proto-oncogene. Despite the highly repetitive guanine-rich sequence of the hypersensitive element, fluorescence resonance energy transfer measurements indicate that only one well defined tetraplex structure forms therein. The proposed structure, which is specifically stabilized by potassium ions in vitro, has a core of three stacked guanine tetrads that is capped by two intrastrand A-T base pairs.

Structural properties of Friedreich's ataxia d(GAA) repeats

The expansion of trinucleotide repeat sequences is the underlying cause of a growing number of inherited human disorders. To provide correlations between DNA structure and mechanisms of trinucleotide repeat expansion, we investigated potential secondary structures formed from the complementary strands of d(GAA.TTC)n, a sequence whose expansion is associated with Friedreich's ataxia. In 50 mM NaCl, pH 7.5, d(GAA)15 exhibited a cooperative and reversible decrease in large circular dichroism bands at 248 and 272-274 nm over the temperature range of 5-50 degrees C, providing evidence for a base-paired structure at reduced temperatures. Ultraviolet absorbance melting profiles indicated that the melting temperature (Tm) of d(GAA)15 was 40 degrees C. At 5 degrees C, the central portion of d(GAA)15 was hypersensitive to single-strand-specific P1 nuclease degradation and diethyl pyrocarbonate modification, providing evidence for a hairpin conformation. At temperatures between 25 and 35 degrees C in 50 mM NaCl, the triplet repeat region of d(GAA)15 was uniformly resistant to degradation by P1 nuclease, including the central portion of the sequence. Our results indicate that the structure of d(GAA)15 is a hairpin at 5 degrees C, unknown but partially base-paired at 37 degrees C, and an approximately random coil above 65 degrees C.

Dimerization of the guanine-adenine repeat strands of DNA

Jovin and co-workers have demonstrated that DNA strands containing guanine-adenine repeats generate a parallel-stranded homoduplex. Here we propose that the homoduplex is a dimer of the ordered single strand discovered by Fresco and co-workers at acid pH. The Fresco single strand is shown here to be stabilized in aqueous ethanol where adenine is not protonated. Furthermore, we demonstrate that the strands dimerize at higher salt concentrations without significantly changing their conformation, so that the dimerization is non-cooperative. Hence, the Jovin homoduplex can form through a non-cooperative dimerization of two cooperatively melting single strands. The available data indicate that the guanines stabilize the Fresco single strand whereas the adenines cause dimerization owing to their known intercalation or clustering tendency. The guanine-adenine repeat dimer seems to be a DNA analog of the leucine zipper causing dimerization of proteins.

DNA-mediated transport of the intermediate filament protein vimentin into the nucleus of cultured cells

A number of characteristic properties of intermediate filament (IF) proteins, such as nucleic acid-binding activity, affinity for histones and structural relatedness to transcription factors and nuclear matrix proteins, in conjunction with the tight association of IFs with the nucleus, suggest that these proteins might also fulfill nuclear functions in addition to their structure-organizing and -stabilizing activities in the cytoplasm. Yet, cytoplasmic IF proteins do not possess nuclear localization signals. In a search for carriers capable of transporting the IF protein vimentin into the nucleus, complexes of FITC-vimentin with various DNAs were microinjected into the cytoplasm of cultured cells and the intracellular distribution of the protein was followed by confocal laser scanning microscopy. The single-stranded oligodeoxyribonucleotides oligo(dG)25, oligo[d(GT)12G] and oligo[d(G3T2A)4G] proved to be excellent nuclear carriers for vimentin. However, in fibroblasts, fluorescence-labeled vimentin taken up by the nuclei remained undetectable with affinity-purified, polyclonal anti-vimentin antibody, whereas it was readily identifiable in the nuclei of microinjected epithelial cells in this way. Moreover, when FITC-vimentin was preinjected into fibroblasts and allowed to assemble into the endogenous vimentin filament system, it was still transferred into the nucleus by post-injected oligo(dG)25, although to a lesser extent. Superhelical circular DNAs, like pBR322, SV40 and mitochondrial DNA, were also characterized by considerable capacities for nuclear vimentin transport; these transport potentials were totally destroyed by relaxation or linearization of the DNA molecules. Nevertheless, certain linear double-stranded DNA molecules with a high affinity for vimentin IFs, such as repetitive telomere and centromere or mobile long interspersed repeat (LINE) DNA, could carry FITC-vimentin into the nucleus. This was also true for a 375 bp extrachromosomal linear DNA fragment which occurs in the cytoplasm of mouse tumor cells and which is capable of immortalizing human lymphocytes. On the basis of these results, it appears very likely that cellular and viral products of reverse transcription as well as other extrachromosomal DNAs, which are circular, superhelical and apparently shuttling between the cytoplasm and the nucleus (eccDNA), are constantly loaded with vimentin in vimentin-positive cells. Since such DNAs are considered as markers of genomic instability, it is conceivable that vimentin directly participates as an architectural, chromatin-modifying protein in recombinatorial processes set off by these DNAs in the nucleus.

Antiparallel DNA duplex formation between alternating alpha d(GA)n and beta d(GA)n sequences

Alternating polypurine d(GA)n, sequences exhibit a considerable polymorphism. Here we report that alpha d(GA) x d(GA) sequences form an antiparallel stranded duplex DNA at neutral pH. The spectroscopic, electrophoretic and thermodynamic properties of the alpha/beta chimeric oligodeoxynucleotide, 5'-d(GA)4(T)4 alpha d(AG)4T-3', support the formation of a hairpin structure with antiparallel strands in the stem. The optical properties of this novel antiparallel structure are different from the parallel stranded homoduplex formed by d(GA)G7. This alpha/beta hairpin has a remarkably high Tm of 44.5 degrees C in 0.4 M NaCl with a van't Hoff enthalpy comparable to that of a parallel d(GA)n duplex. Base pairing was confirmed by T4 polynucleotide ligase catalyzed joining of the alpha/beta hairpin to an antiparallel bimolecular duplex and by non-denaturing gel electrophoresis using duplexes containing sequence constraints. Both support the presence of alphaG-G and alphaA-A base pairing in the antiparallel 5'-d(GA)4(T)4 alpha d(AG)4T-3' intramolecular duplex. This study adds to the polymorphic nature of alternating d(GA)n sequences as well as providing novel homopurine base pairing approaches for probing polypurine polypyrimidine sequences.

Conformational properties of DNA strands containing guanine-adenine and thymine-adenine repeats

CD spectroscopy and PAGE were used to cooperatively analyze melting conformers of DNA strands containing GA and TA dinucleotide repeats. The 20mer (GA)10 formed a homoduplex in neutral solutions containing physiological concentrations of salts and this homoduplex was not destabilized even in the terminal (GA)3 hexamers of (GA)3(TA)4(GA)3, although the central (TA)4 portion of this oligonucleotide preserved the conformation adopted by (TA)10. This observation demonstrates that homoduplexes of alternating GA and TA sequences can co-exist in a single DNA molecule. Another 20mer, (GATA)5, adopted as a whole either the AT duplex, like (TA)10, or the GA duplex, like (GA)10, and switched between them reversibly. The concentration of salt controlled the conformational switching. Hence, guanine and thymine share significant properties regarding complementarity to adenine, while the TA and GA sequences can stack in at least two mutually compatible ways within the DNA duplexes analyzed here. These properties extend our knowledge of non-canonical structures of DNA.

Fluorescence microscopic comparison of the binding of phosphodiester and phosphorothioate (antisense) oligodeoxyribonucleotides to subcellular structures, including intermediate filaments, the endoplasmic reticulum, and the nuclear interior

To detect potential intracellular binding sites for antisense oligodeoxyribonucleotides (ODN), 3'-fluorescence-tagged phosphodiester (P) and phosphorothioate (S) analogs of a series of model and vimentin and actin antisense ODN were applied to digitonin-permeabilized fibroblast and epithelial PtK2 cells. Fluorescence microscopy revealed binding of the ODN to intermediate filaments (IFs) with a preference for cytokeratin IFs, cytoplasmic membranes (endoplasmic reticulum), and, above all, the nuclear interior. The affinity of the ODN for these cellular substructures was dependent on their base composition, and the S-ODN were by far superior to the corresponding P-ODN in binding activity. Fluorescence polarization measurements of the interaction of ODN with purified IF proteins in vitro confirmed the differential, high-affinity binding of S-ODN to IFs. In permeabilized cells, the ODN readily migrated into the nucleus where, at ambient temperature, preferentially the S-ODN gave rise to a multitude of large, irregular aggregates. Nuclear uptake of the ODN was considerably and differentially inhibited by wheat germ agglutinin. High-affinity S-ODN, but not P-ODN, additionally reacted with a structure presumably identical with the nuclear lamina. Simultaneously, they cause decompaction of chromatin, whereby the S-ODN aggregates appeared as compact inclusions in homogeneously dispersed chromatin. After microinjection of S-ODN into intact cells, these effects were not observed, although the nucleic acids rapidly moved into the nucleus and condensed into a large number of well-defined, spherical speckles or longitudinal rodlets. The methylphosphonate analogs of some of the ODN used exhibited only extremely low affinities for intracellular constituents. These results show that excess amounts of S-ODN saturate a host of both low-affinity and high-affinity binding sites on cellular substructures, whereas limited quantities as used for microinjection recognize only the high-affinity binding sites. The results support the notion that the nonsequence-specific, often toxic effects of antisense S-ODN result from their strong binding to cellular components and substructures involved in replicational, transcriptional, and translational processes. On the other hand, the association of the ODN with membranes and cytoskeletal and karyoskeletal elements may serve to optimize their sequence-specific interaction with their intended target sites and also increase their cellular retention potential. These cellular structures would thus fulfill a depot function.

A UV resonance Raman study of hairpin dimer helices of d(A-G)10 at neutral pH containing intercalated dA residues and alternating dG tetrads

The structure of the oligonucleotide d(A-G)10 in 0.6 M Na+, pH 7.0 has been investigated with UV resonance Raman (UVRR) spectroscopy. Variable wavelength excitation was used to distinguish the spectral contributions of dG and dA residues. Both classes of residues show UVRR hyperchromism with increasing temperature, reflecting unstacking of the bases. The dG residues melt relatively cooperatively with a Tm of approximately 42 degrees C. Unstacking is non-cooperative for the dA residues, increasing linearly between 4 and 80 degrees C. G-tetrads at low temperature are indicated by UVRR frequency shifts of modes associated with C6=O and C2-NH2 of the dG residues, and of vibrations involving N7, all sites of H-bonding. However, there are no indications of interbase H-bonds for the dA residues, showing they do not form H-bonded tetrads. Most of the bases are oriented anti about the glycosyl bond, but at 4 degrees C a fraction of the residues are syn. These results, together with the findings by Shiber et al. [Shiber,M.C., Braswell,E.H., Klump,H. and Fresco,J.R. (1996) Nucleic Acids Res. 24, 5004-5012] that d(A-G)10 under comparable conditions has the molecular weight of a dimer, support a model in which two hairpins interact to form a helical structure with G-tetrads and intercalated dA residues.