Circular dichroism spectroscopy of DNA: from duplexes to quadruplexes.

Nucleic acids bear the genetic information and participate in its expression and evolution during replication, repair, recombination, transcription, and translation. These phenomena are mostly based on recognition of nucleic acids by proteins. The major factor enabling the specific recognition is structure. Circular dichroism (CD) spectroscopy is very useful to study secondary structures of nucleic acids, in general, and DNA, in particular. CD sensitively reflects isomerizations among distinct conformational states. The isomerizations may operate as molecular switches regulating various physiological or pathological processes. Here, we review CD spectra of nucleic acids, beginning with early studies on natural DNA molecules through analyses of synthetic polynucleotides to study of selected genomic fragments.

Circular dichroism and guanine quadruplexes.

Circular dichroism (CD) is remarkably sensitive to the conformational states of nucleic acids; therefore, CD spectroscopy has been used to study most features of DNA and RNA structures. Quadruplexes are among the significant noncanonical nucleic acids architectures that have received special attentions recently. This article presents examples on the contribution of CD spectroscopy to our knowledge of quadruplex structures and their polymorphism. The examples were selected to demonstrate the potential of this simple method in the quadruplex field. As CD spectroscopy detects only the global feature of a macromolecule, it should preferably be used in combination with other techniques. On the other hand, CD spectroscopy, often as a pioneering approach, can reveal the formation of particular structural arrangements, to search for the conditions stabilizing the structures, to follow the transitions between various structural states, to explore kinetics of their appearance, to determine thermodynamic parameters and also detect formation of higher order structures. This article aims to show that CD spectroscopy is an important complementary technique to NMR spectroscopy and X-ray diffraction in quadruplex studies.

CGG repeats associated with fragile X chromosome form left-handed Z-DNA structure.

This work is a continuation of our effort to determine the structure responsible for expansion of the (CGG)(n) motif that results in fragile X chromosome syndrome. In our previous report, we demonstrated that the structure adopted by an oligonucleotide with this repeat sequence is not a quadruplex as was suggested by others. Here we demonstrate that (CGG) runs adopt another anomalous arrangement-a left-handed Z-DNA structure. The Z-DNA formation was induced by high salt and millimolar concentrations of Ni(2+) ions and likelihood of its formation increased with increasing number of repeats. In an oligonucleotide in which the CGG runs were interrupted by AGG triplets, as is observed in genomes of healthy individuals, the hairpin conformation was stabilized and Z-DNA formation was hindered. We show here that methylation of the (CGG) runs markedly stabilized Z-DNA formation. We hypothesize that rather than in the expansion process the Z-DNA may be formed by long, expanded (CGG) stretches that become hypermethylated; this would inhibit transcription resulting in disease.

Circular dichroism and conformational polymorphism of DNA.

Here we review studies that provided important information about conformational properties of DNA using circular dichroic (CD) spectroscopy. The conformational properties include the B-family of structures, A-form, Z-form, guanine quadruplexes, cytosine quadruplexes, triplexes and other less characterized structures. CD spectroscopy is extremely sensitive and relatively inexpensive. This fast and simple method can be used at low- as well as high-DNA concentrations and with short- as well as long-DNA molecules. The samples can easily be titrated with various agents to cause conformational isomerizations of DNA. The course of detected CD spectral changes makes possible to distinguish between gradual changes within a single DNA conformation and cooperative isomerizations between discrete structural states. It enables measuring kinetics of the appearance of particular conformers and determination of their thermodynamic parameters. In careful hands, CD spectroscopy is a valuable tool for mapping conformational properties of particular DNA molecules. Due to its numerous advantages, CD spectroscopy significantly participated in all basic conformational findings on DNA.

Role of loops in the guanine quadruplex formation by DNA/RNA hybrid analogs of G4T4G4.

CD spectroscopy, gel electrophoresis and absorption-based thermal stability were used to analyze quadruplex formation of RNA and RNA/DNA hybrid analogs of the deoxyoligonucleotide G4T4G4, which forms a well-characterized basket-type quadruplex. All RNA-containing dodecamers, g4u4g4, G4u4G4 and g4T4g4 (RNA lower-case, DNA capital letters), formed parallel, namely tetramolecular quadruplexes in Na+-containing solutions. The u4 loop forced DNA tetrads into the same conformation as adopted by g4u4g4. In contrast, the T4 loop destabilized the RNA tetrads. Potassium ions markedly stabilized parallel quadruplexes of RNA-containing analogs as well as their bimolecular folding. In the presence of K+, g4T4g4 formed exclusively bimolecular quadruplexes of both parallel and antiparallel types as indicated by CD. Thus, the T4 loop permits RNA strands to adopt an antiparallel arrangement. These findings may be useful for engineering particular quadruplex foldings in different quadruplex-forming sequences.

Guanine quadruplex formation by RNA/DNA hybrid analogs of Oxytricha telomere G(4)T(4)G(4) fragment.

Using circular dichroism spectroscopy, gel electrophoresis, and ultraviolet absorption spectroscopy, we have studied quadruplex folding of RNA/DNA analogs of the Oxytricha telomere fragment, G(4)T(4)G(4), which forms the well-known basket-type, antiparallel quadruplex. We have substituted riboguanines (g) for deoxyriboguanines (G) in the positions G1, G9, G4, and G12; these positions form the terminal tetrads of the G(4)T(4)G(4) quadruplex and adopt syn, syn, anti, and anti glycosidic geometries, respectively. We show that substitution of a single sugar was able to change the quadruplex topology. With the exception of G(4)T(4)G(3)g, which adopted an antiparallel structure, all the RNA/DNA hybrid analogs formed parallel, bimolecular quadruplexes in concentrated solution at low salt. In dilute solutions ( approximately 0.1 mM nucleoside), the RNA/DNA hybrids substituted at positions 4 or 12 adopted antiparallel quadruplexes, which were especially stable in Na(+) solutions. The hybrids substituted at positions 1 and 9 preferably formed parallel quadruplexes, which were more stable than the nonmodified G(4)T(4)G(4) quadruplex in K(+) solutions. Substitutions near the 3'end of the molecule affected folding more than substitutions near the 5'end. The ability to control quadruplex folding will allow further studies of biophysical and biological properties of the various folding topologies.

The longest (A+T) and (G+C) blocks in the human and other genomes.

We analysed complete or almost complete nucleotide sequences of the human, chimp, mouse, rat, chicken, dog, and other genomes to find that they contain extremely long (A+T) a (G+C) blocks that do not occur at all in the corresponding randomized sequences. The longest is an (A+T) block containing 1040 consecutive AT pairs that occurs in the 16th human chromosome. The longest human (G+C) block has 261 bp in length. About a half of the longest blocks occur in introns. The (A+T) blocks are discrete units whereas the (G+C) blocks are diffuse. They are imbedded in the genome through connectors longer than 1 kilobase where the (G+C) content gradually decreases to the value of 50%. Remarkably, the (A+T) as well as (G+C) blocks are substantially shorter in the chimp genome. Chicken is characteristic by very long (G+C) blocks that are even longer than in the human genome. Though much shorter, long (G+C) and especially (A+T) blocks occur in lower organisms as well, which means that AT and GC pair clustering is an ancient property that has evolved into large scales in higher eukaryote genomes and the human genome in particular. Very long (A+T) and (G+C) blocks confer specific biophysical properties on DNA that are likely to influence genome folding in cell nuclei and its functional properties.

Conformations of DNA strands containing GAGT, GACA, or GAGC tetranucleotide repeats.

The (GA)(n) microsatellite has been known from previous studies to adopt unusual, ordered, cooperatively melting secondary structures in neutral aqueous solutions containing physiological concentrations of salts, at acid pH values or in aqueous ethanol solutions. To find more about the primary structure specificity of these structures, we performed parallel comparative studies of related tetranucleotide repeats (GAGC)(5), (GAGT)(5), and (GACA)(5). The general conclusion following from these comparative studies is that the primary structure specificity is fairly high, indicating that not only guanines but also adenines play a significant role in the stabilization of these unusual structures. (GAGC)(5) is a hairpin or a duplex depending on DNA concentration. Neither acid pH nor ionic strength or the presence of ethanol changed the secondary structure of (GAGC)(5) in a significant way. (GACA)(5) forms a weakly stable hairpin in neutral aqueous solutions but forms a duplex at acid pH where cytosine is protonated. (GAGT)(5) behaves most similar to (GAGA)(5). Salt induces its hairpin to duplex transition at neutral pH and an isomerization into another, probably parallel stranded, duplex takes place at acid pH. (GAGT)(5) is the only of the three present 20-mers that responds to ethanol like (GAGA)(5).

Oligo(dT) is not a correct native PAGE marker for single-stranded DNA.

Polyacrylamide gel electrophoresis is a widely used method to study short DNA fragments in solution. It is, however, a relative method requiring length markers to assess mobility, shape, flexibility, and molecularity of the DNA structures of interest. In recent literature we have encountered the use of oligo(dT) fragments as the native PAGE length markers. We show here that this practice is inadequate because oligo(dT) migration is strongly retarded in native polyacrylamide gels. This conclusion is qualitatively true irrespective of the conditions of electrophoresis, oligo(dT) length, and gel concentration. Depending on their length, oligo(dT) fragments migrate 2--4 times slower than that would correspond to their nucleotide number. This leads to erroneous conclusions, e.g., determination of the number of associated molecules in guanine quadruplexes or other DNA complexes.

Intramolecular and intermolecular guanine quadruplexes of DNA in aqueous salt and ethanol solutions.

DNA guanine quadruplexes are all based on stacks of guanine tetrads, but they can be of many types differing by mutual strand orientation, topology, position and structure of loops, and the number of DNA molecules constituting their structure. Here we have studied a series of nine DNA fragments (G(3)Xn)(3)G(3), where X = A, C or T, and n = 1, 2 or 3, to find how the particular bases and their numbers enable folding of the molecule into quadruplex and what type of quadruplex is formed. We show that any single base between G(3) blocks gives rise to only four-molecular parallel-stranded quadruplexes in water solutions. In contrast to previous models, even two Ts in potential loops lead to tetramolecular parallel quadruplexes and only three consecutive Ts lead to an intramolecular quadruplex, which is antiparallel. Adenines make the DNA less prone to quadruplex formation. (G(3)A(2))(3)G(3) folds into an intramolecular antiparallel quadruplex. The same is true with (G(3)A(3))(3)G(3) but only in KCl. In NaCl or LiCl, (G(3)A(3))(3)G(3) prefers to generate homoduplexes. Cytosine still more interferes with the quadruplex, which only is generated by (G(3)C)(3)G(3), whereas (G(3)C(2))(3)G(3) and (G(3)C(3))(3)G(3) generate hairpins and/or homoduplexes. Ethanol is a more potent DNA guanine quadruplex inducer than are ions in water solutions. It promotes intramolecular folding and parallel orientation of quadruplex strands, which rather corresponds to quadruplex structures observed in crystals.

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.

Photochemical probing of the B--a conformational transition in a linearized pUC19 DNA and its polylinker region.

We induced the B-to-A conformational transition by ethanol in a linearized pUC19 DNA. A primer extension method was used in combination with UV light irradiation to follow the transition, based on pausing of DNA synthesis due to the presence of damaged bases in the template. Primer extension data highly correlated with the results of another method monitoring the B-A transition, i.e. inhibition of restriction endonuclease cleavage of UV light-irradiated DNA. Primer extension enabled us to locate damaged nucleotides within the region of interest. Most damaged nucleotides were located in B-form trimers, exclusively containing both pyrimidine bases (TTC, TCT, CTC, and CTT), and in a cytosine tetramer. The amount of damaged bases decreased in the course of B-A transition. Some of the damage even disappeared in the A-form, which mainly concerns the C(4) and C(3) blocks. The cleavage was nearly restored in the A-form within this region (Eco88I). On the contrary the decrease of damage was less significant with thymine dimers, only dropping to 50-60% of the B-form level. Consequently, the cleavage with EcoRI and HindIII remained mostly as before the transition (75% and 60% of uncleaved DNA preserved). We found significant differences in the B- and A-form pattern of UV light-damaged bases within the same region (polylinker) of DNA embedded within long (plasmid) or short (127 bp fragment) DNA molecules. The B-A transition of the fragment was found less cooperative than with linearized plasmid, which was confirmed by both CD spectroscopy and restriction cleavage inhibition.

Factors influencing DNA expansion in the course of polymerase chain reaction.

We performed more than 3,500 polymerase chain reactions (PCRs) under various conditions with more than 400 DNA fragments of 4-150 nucleotides in length. Some of the PCRs provided expanded DNA molecules of kilobase lengths whereas others led to no expansion. Repetitiveness of the primary structure was mostly found to be necessary but not sufficient for the expansion. (A+T)-rich fragments expand better than (G+C)-rich ones and pyrimidine-rich fragments expand better than purine-rich fragments. Terminal nucleotides and the fragment length also are important for the expansion. Examples are presented when relatively small alterations of the DNA primary structure caused a dramatic change in the expansion. For example, A8T8 expanded a lot whereas T8A8 did not expand at all. The present work has implications for pathological expansions of microsatellites in the human genome as well as regarding the genome evolution in general.

The thrombin binding aptamer GGTTGGTGTGGTTGG forms a bimolecular guanine tetraplex.

In the literature, the thrombin binding aptamer GGTTGGTGTGGTTGG is generally taken as a prototype of an intramolecular guanine tetraplex of DNA. Our results, however, show that this notion is not true in aqueous solutions. This conclusion is based on a dependence of the CD spectra on aptamer concentration, migration of the aptamer in polyacrylamide gels, and the Ferguson analysis of the gel migration data. The presented data document that the aptamer forms a bimolecular tetraplex. We furthermore show that only an extension of the aptamer by a sequence containing further guanines, or an elongation of loop regions, causes that its tetraplex folding is intramolecular.

Ethanol is a better inducer of DNA guanine tetraplexes than potassium cations.

Guanine tetraplexes are a biologically relevant alternative of the Watson and Crick duplex of DNA. It is thought that potassium or other cations present in the cavity between consecutive guanine tetrads are an integral part of the tetraplexes. Here we show using CD spectroscopy that ethanol induces the guanine tetraplexes like or even better than potassium cations. We present examples of ethanol stabilizing guanine tetraplexes even in cases when potassium cations fail to do so. Hence, besides the A-form or Z-form, ethanol stabilizes another conformation of DNA, i.e., the guanine tetraplexes. We discuss the mechanism of the stabilization. Use of ethanol will permit studies of guanine tetraplexes that cannot be induced by potassium cations or other tetraplex-promoting agents. This work demonstrates that a still broader spectrum of nucleotide sequences can fold into guanine tetraplexes than has previously been thought. Aqueous ethanol may better simulate conditions existing in vivo than the aqueous solutions.

Guanine tetraplex topology of human telomere DNA is governed by the number of (TTAGGG) repeats.

Secondary structures of the G-rich strand of human telomere DNA fragments G3(TTAG3)n, n = 1-16, have been studied by means of circular dichroism spectroscopy and PAGE, in solutions of physiological potassium cation concentrations. It has been found that folding of these fragments into tetraplexes as well as tetraplex thermostabilities and enthalpy values depend on the number of TTAG3 repeats. The suggested topologies include, e.g. antiparallel and parallel bimolecular tetraplexes, an intramolecular antiparallel tetraplex, a tetraplex consisting of three parallel chains and one antiparallel chain, a poorly stable parallel intramolecular tetraplex, and both parallel and antiparallel tetramolecular tetraplexes. G3(TTAG3)3 folds into a single, stable and very compact intramolecular antiparallel tetraplex. With an increasing repeat number, the fragment tetraplexes surprisingly are ever less thermostable and their migration and enthalpy decrease indicate increasing irregularities or domain splitting in their arrangements. Reduced stability and different topology of lengthy telomeric tails could contribute to the stepwise telomere shortening process.

Conformational properties of DNA containing (CCA)n and (TGG)n trinucleotide repeats.

We have used CD spectroscopy, polyacrylamide gel electrophoresis, and UV absorption spectroscopy to study conformational properties of DNA fragments containing (CCA)n and (TGG)n repeats, which are the most length-polymorphic microsatellite sequences of the human genome. The (CCA)n fragments are random single strands at neutral and alkaline pH but they fold into intramolecular intercalated cytosine tetraplexes at mildly acid pH values. More acid values stabilize intermolecular tetraplex formation. The behavior of (TGG)n repeats is more complex. They form hairpins or antiparallel homoduplexes in low salt solutions which, however, are transformed into parallel-stranded guanine tetraplexes at physiological KCl concentrations. Their molecularity depends on the repeat number: (TGG)4 associates into an octameric complex, (TGG)8 forms tetramolecular complexes. (TGG)n with odd repeat numbers (5, 7, and 9) generate bimolecular and tetramolecular tetraplexes. The only (TGG)7 folds into an intramolecular tetraplex at low KCl concentrations, which is antiparallel-stranded. Moreover, the (TGG)(n) fragments provide various mutually slipped conformers whose population increases with salt concentration and with the increasing repeat number. However, the self-structures of both strands disappear in the presence of the complementary strand because both (TGG)n and (CCA)n prefer to associate into the classical heteroduplex. We suppose that the extreme conformational variability of the DNA strands stands behind the length polymorphism which the (CCA)n/(TGG)n repeats exhibit in the human genome.

Mapping the B-A conformational transition along plasmid DNA.

A simple method is presented to monitor conformational isomerizations along genomic DNA. We illustrate properties of the method with the B-A conformational transition induced by ethanol in linearized pUC19 plasmid DNA. At various ethanol concentrations, the DNA was irradiated with ultraviolet light, transferred to a restriction endonuclease buffer and the irradiated DNA was cleaved by 17 restriction endonucleases. The irradiation damaged DNA and the damage blocked the restrictase cleavage. The amount of uncleaved, i.e. damaged, DNA depended on the concentration of ethanol in a characteristic S-shape way typical of the cooperative B-A transition. The transition beginning and midpoint were determined for each restriction endonuclease. These data map the B-A transition along the whole polylinker of pUC19 DNA and six evenly distributed recognition sequences within the rest of the plasmid. The transition midpoints fell within the B-A transition region of the plasmid simultaneously determined by CD spectroscopy. The present method complements the previous methods used to study the B-A transition. It can be employed to analyze multikilobase regions of genomic DNA whose restriction endonuclease cleavage fragments can be separated and quantified on agarose gels.

DNA in phosphorus atom representation: the heteronomous double helices of poly(dA).poly(dT) and poly(dG).poly(dC) and simulation of the yeast genome and of a human chromosome DNA.

We extracted phosphorus atom coordinates from the database of DNA crystal structures and calculated geometrical parameters needed to reproduce the crystal structures in the phosphorus atom representation. Using the geometrical parameters we wrote a piece of software assigning the phosphorus atom coordinates to the DNA of any nucleotide sequence. The software demonstrates non-negligible influence of the primary structure on DNA helicity, which may stand behind the heteromonous double helices of poly(dA).poly(dT) and poly(dG).poly(dC). In addition, the software is so simple that it makes possible to simulate the "crystal" structures of not only viral DNAs, but also the whole genome of Saccharomyces cerevisiae as well as the DNA human chromosome 22 having dozens of megabases in length.

Factors influencing resistance of UV-irradiated DNA to the restriction endonuclease cleavage.

DNA molecules of pUC19, pBR322 and PhiX174 were irradiated by various doses of UV light and the irradiated molecules were cleaved by about two dozen type II restrictases. The irradiation generally blocked the cleavage in a dose-dependent way. In accordance with previous studies, the (A + T)-richness and the (PyPy) dimer content of the restriction site belongs among the factors that on average, cause an increase in the resistance of UV damaged DNA to the restrictase cleavage. However, we observed strong effects of UV irradiation even with (G + C)-rich and (PyPy)-poor sites. In addition, sequences flanking the restriction site influenced the protection in some cases (e.g. HindIII), but not in others (e.g. SalI), whereas neoschizomer couples SmaI and AvaI, or SacI and Ecl136II, cleaved the UV-irradiated DNA similarly. Hence the intrastrand thymine dimers located in the recognition site are not the only photoproduct blocking the restrictases. UV irradiation of the A-form generally made the irradiated DNA less resistant to restrictase cleavage than irradiation in the B-form and in some cases, the A-form completely protected the UV-irradiated DNA against the damage recognized by the restrictases. The present results also demonstrate that the UV irradiation approach used to generate partial digests in genomic DNA studies, can be extended to the (G + C)-rich and (PyPy)-poor restriction sites. The present extensive and quantitative data can be used in genomic applications of UV damage probing by restrictases.