Mg2+-induced triplex formation of an equimolar mixture of poly(rA) and poly(rU)

doi:10.1093/nar/gkg698

Journal List > Nucleic Acids Res > v.31(17); Sep 1, 2003

Nucleic Acids Res. 2003 September 1; 31(17): 5101–5107.

doi: 10.1093/nar/gkg698.

PMCID: PMC212798

Mg²⁺-induced triplex formation of an equimolar mixture of poly(rA) and poly(rU)

Besik I. Kankia^*

Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, 1479 Gortner Avenue, St Paul, MN 55108, USA

^*Tel: +1 612 624 7468; Fax: +1 612 625 5780; Email: bkankia/at/umn.edu

Received May 8, 2003; Revised July 3, 2003; Accepted July 3, 2003.

This article has been cited by other articles in PMC.

Abstract

Magnesium ions strongly influence the structure and biochemical activity of RNA. The interaction of Mg²⁺ with an equimolar mixture of poly(rA) and poly(rU) has been investigated by UV spectroscopy, isothermal titration calorimetry, ultrasound velocimetry and densimetry. Measurements in dilute aqueous solutions at 20°C revealed two differ ent processes: (i) Mg²⁺ binding to unfolded poly(rA)·poly(rU) up to [Mg²⁺]/[phosphate] = 0.25; and (ii) poly(rA)·2poly(rU) triplex formation at [Mg²⁺]/[phosphate] between 0.25 and 0.5. The enthalpies of these two different processes are favorable and similar to each other, ~–1.6 kcal mol^–1 of base pairs. Volume and compressibility effects of the first process are positive, 8 cm³ mol^–1 and 24 × 10^–4 cm³ mol^–1 bar^–1, respectively, and correspond to the release of water molecules from the hydration shells of Mg²⁺ and the polynucleotides. The triplex formation is also accompanied by a positive change in compressibility, 14 × 10^–4 cm³ mol^–1 bar^–1, but only a small change in volume, 1 cm³ mol^–1. A phase diagram has been constructed from the melting experiments of poly(rA)·poly(rU) at a constant K⁺ concentration, 140 mM, and various amounts of Mg²⁺. Three discrete regions were observed, corresponding to single-, double- and triple-stranded complexes. The phase boundary corresponding to the transition between double and triple helical conformations lies near physiological salt concentrations and temperature.

INTRODUCTION

RNA homopolynucleotide, poly(rA), and its complementary poly(rU) are characterized by a large degree of structural variability. For instance, at neutral pH and ambient temperature, poly(rA) adopts a single helical structure due to a stacking interaction between the adenine bases (1,2). Under acidic conditions, pH <4, the N1 atoms of adenines are protonated and the polynucleotide forms a parallel double helix. At ambient temperature, poly(rU) does not possess any specific structure: the lack of a stacking interaction between uridine bases leads to a random-coiled conformation (2). However, at low temperature, poly(rU) exhibits a secondary structure, which is defined as an antiparallel double helix (2). An equimolar (1:1) mixture of poly(rA) and poly(rU) usually forms the double helix at ambient temperature (3,4). At higher temperature and sodium concentration (>0.2 M) the duplex is converted to the triplex, poly(rA)·2poly(rU) and free poly(rA) (4,5). At a lower concentration of monovalent cations, an increase in temperature does not induce the disproportionation reaction (4). However, adding small amounts of divalent magnesium (1 mM) to the solution reveals the duplex to triplex transition at ~50°C (4).

The significance of magnesium ions in nucleic acid biochemistry has been well recognized and studied (6–9). Divalent cations are important in charge screening of negatively charged phosphate groups to counterbalance the high linear charge density, especially for triplexes (6,7,10). However, the role of magnesium ions in the stability of triplexes has not been fully assessed. Triple helical nucleic acids are of major interest because they can be used in the regulation of the genome and have potential for antisense and therapeutic applications (10–13).

Another factor that plays a fundamental role in the stability of the secondary and tertiary structure of nucleic acids is hydration (1,2,14,15). The method of volume and compressibility measurements proved to be successful for studying the hydration of nucleic acids and their interaction with ligand molecules (16–18). These parameters are sensitive to the hydration of solute molecules, and can follow the transfer of water molecules from the bulk state to hydration shells and vice versa. The basis of the measurements is that the molar volume and molar compressibility of pure (bulk) water are significantly larger than the same parameters for water molecules in the hydration shells.

We employed a combination of UV spectroscopy, isothermal titration calorimetry densimetry and ultrasound velocimetry to study the optical, thermodynamic and hydration effects of Mg²⁺ binding to an equimolar mixture of poly(rA) and poly(rU). Besides the fact that poly(rA) and poly(rU) complexes are a good model system to study the structural polymorphism of nucleic acids, the knowledge obtained in such studies could be useful to understand the properties of messenger RNA, which contains poly(rA) tail and AU-rich elements (19,20). The present work shows that Mg²⁺ ions are able to induce the triplex formation from equilomar mixtures of poly(rA) and poly(rU) in dilute aqueous solutions at ambient temperature, as well as at physiological salt concentrations and temperature.

MATERIALS AND METHODS

Materials

Poly(rA)·poly(rU) (700 kDa), poly(rA) (450 kDa) and poly(rU) (1300 kDa) were obtained from Sigma. The polynucleotides were dissolved in 100 mM NaCl, 5 mM EDTA, pH 8 and dialyzed against 2 mM Na-HEPES, pH 7.5 for 3–4 days at 4°C. The concentration of the polynucleotides was determined optically in 100 mM NaCl, 10 mM Na-HEPES, pH 7.5 at 20°C using the molar extinction coefficients in M^–1 cm^–1 of nucleotide units (21,22): ε₂₅₇ = 7000 for poly(rA)·poly(rU), ε₂₅₇ = 10 700 for poly(rA) and ε₂₆₀ = 9100 for poly(rU). Analytical grade salts were purchased from Fisher and Merck. The concentration of Mg²⁺ was determined by weighing magnesium chloride hydrate crystals and the appropriate amount of buffer solution. The amount of water in the crystals was determined by measuring the ultrasonic velocities of magnesium chloride in the aqueous solutions at 25°C and then comparing these with data from the literature (23). No significant differences in the pH of polynucleotide solutions before and after binding were observed. All experiments, except for the melting curves on Figures 5 and 6, were conducted in 2 mM Na-HEPES, pH 7.5 at 20°C. Under these conditions, poly(rA)·poly(rU) duplex is almost completely unfolded: (i) adding sodium ions up to 100 mM concentration revealed a gradual hyporchromic effect of ~40% and an isosbestic point around 280 nm (data not shown), which are characteristic for this duplex formation; and (ii) UV melting at 260 nm of poly(rA)·poly(rU) in our buffer reveals uncooperative transition with 25% hyperchromicity [after subtracting the contribution of the poly(rU) random coil] typical of single-stranded poly(rA) (24).

Figure 5

Melting curves of poly(rA)·poly(rU) in 140 mM KCl without (closed circles) and with 10 mM MgCl₂ (open circles). (A) and (B) correspond to the melting at 260 and 280 nm, respectively.

Figure 6

Phase diagram of an equimolar mixture of poly(rA) and poly(rU) in 140 mM KCl, 10 mM Na-HEPES, pH 7.5. Melting temperatures are shown as a function of MgCl₂ concentration. In the case of higher Mg²⁺ concentrations, 14.9 and 19.6 mM, triplex melting (more ...)

Optical studies

All UV absorption experiments were conducted on a GBC 918 spectrophotometer equipped with a thermoelectrically controlled six-cell holder. Quartz cells with 0.05 and 1 cm path lengths were used in all studies. The optical titrations were performed by adding salt solutions to polynucleotide solutions in the cuvette. Stirring was carried out directly in the cuvette using a vibrating bar. Absorption spectra were recorded after full equilibration at the desired Mg²⁺ concentration.

Isothermal titration calorimetry

A Microcal (Northampton, MA) MCS calorimeter was used to measure the heat evolved during Mg²⁺ binding to poly(rA)·poly(rU). The polynucleotide solution was placed in the reaction cell and titrated with Mg²⁺. Typically, 5 or 10 µl aliquots of MgCl₂ solution (~6 mM) were injected into a polynucleotide solution (~1 mM per nucleotide) by a syringe spinning at 400 rotations per minute. The data obtained were corrected by heats from a control experiment in which the cation solution was injected into the buffer.

Ultrasound velocity measurements

Relative ultrasound velocity was measured by the resonator technique (25–27). The molar increment of ultrasonic velocity, A, was calculated using the equation:

A = (U – U_o) / (U_oC) 1

where U and U_o are the ultrasound velocities in the solution and solvent, respectively, and C is the molar concentration of the polynucleotides. The change in molar increment of ultrasound velocity, ΔA, accompanying the interaction of Mg²⁺ with the polynucleotide was calculated using the equation:

ΔA = A – A_o 2

where A is the molar increment of the ultrasound velocity of the polynucleotide + MgCl₂ relative to the buffer + MgCl₂ and A_o is the molar increment of the ultrasound velocity of the polynucleotide solution relative to the buffer. As in the case of the optical studies, the ultrasound velocity was recorded after full equilibration.

Density measurements

The density of solutions was measured with DMA 602 and 5000 densimeters (Anton Paar). The molar apparent volume was calculated using the equation (28):

ΦV = M / ρ_o – (ρ – ρ_o) / (ρ_oC) 3

where ρ_o and ρ are the density of the solvent and solution, respectively, and M is the molecular mass of poly(rA)·poly(rU) per nucleotide unit. The volume changes, ΔV, accompanying cation interaction with the polynucleotide, were calculated using the equation:

ΔV = ΦV – ΦV_o 4

where ΦV is the apparent molar volume of the polynucleotide + MgCl₂ relative to the buffer + MgCl₂ and ΦV_o is the apparent molar volume of the polynucleotide solution relative to the buffer. The volume effects of binding were measured as batch experiments mixing certain amounts of polynucleotide and salt solutions on the balance after an equilibrium time of ~1 h.

Apparent molar adiabatic compressibility

The apparent molar adiabatic compressibility, Φκ_S, was determined as a function of the increment of ultrasonic velocity and apparent molar volume (29,30):

Φκ_S = 2β_o(ΦV – A – M / 2ρ_o) 5

where β_o is the adiabatic compressibility coefficient of the solvent. The value of β_o was calculated from our data on density, ρ_o, and the ultrasonic velocity, U_o, in the solvent using the equation β_o = (ρ _oU_o²)^–1. The change in the apparent molar adiabatic compressibility, Δκ_S, was calculated using the equation:

Δκ_S = 2β_o(ΔΦV – ΔA) 6

RESULTS AND DISCUSSION

Optical study

Poly(rA)·poly(rU) was titrated by MgCl₂ and its conformational states were assessed by UV spectroscopy. Figure 1 shows UV spectra at different Mg²⁺ concentrations, which reveals two successive processes. First, a moderate hypochromicity without any isosbestic point up to [Mg²⁺]/[phosphate] = 0.25 (Figs 1A and 2). Secondly, a strong hypochromic effect with a clear isosbestic point around 283 nm, which levels off at [Mg²⁺]/[phosphate] = ~0.5 (Figs 1B and 2).

Figure 1

UV absorbance spectra of an equimolar mixture of poly(rA) and poly(rU) in 2 mM Na-HEPES, pH 7.5 at 20°C at various concentrations of MgCl₂ or concentration ratios of [Mg²⁺]/[phosphate]. An increase (more ...)

Figure 2

Optical titration of an equimolar mixture of poly(rA) and poly(rU) with Mg²⁺ at 257 (A) and 280 nm (B) in 2 mM Na-HEPES, pH 7.5 at 20°C.

The first process corresponds to Mg²⁺ binding to poly(rA) and poly(rU) single strands, without significant rearrangement of the secondary structure. We recall that under the present experimental conditions, poly(rA)·poly(rU) is unfolded and exists as a mixture of dissociated poly(rA) and poly(rU) single strands. Poly(rU) adopts a random-coiled conformation while poly(rA) occurs as a right-handed single helix with an ordered secondary structure (2). Binding of Mg²⁺ to poly(rU) has no effect on spectroscopic properties (UV and CD) of poly(rU) (24), while it stabilizes the single helical structure of poly(rA) (24,31). Thus, the slight hypochromicity observed in the first stage of binding must correspond to the stabilization of the single helical structure of poly(rA) by Mg²⁺.

Taking into consideration that changes in the optical density at 280 nm and the isosbestic point around 283 nm are characteristic of triplex formation by mixing of the complementary strands in the range 0–33% poly(rA) (4), one can conclude that the second process, [Mg²⁺]/[phosphate] = 0.25–0.5, corresponds to the triplex formation:

poly(rA), poly(rU) → 1/2 poly(rA)·2poly(rU) + 1/2 poly(rA) 7

We would like to emphasize that in our Mg²⁺ titration experiments the formation of poly(rA)·poly(rU) duplex can be ruled out. As mentioned in Materials and Methods, titration of the same samples with NaCl, accompanied by the duplex formation, showed an entirely different result: a hyperchromic effect at 260 nm and an isosbestic point around 280 nm.

Poly(rA) and poly(rU) mixtures usually form a poly(rA)·2poly(rU) triplex. However, under special circumstances a triplex with one homo-pyrimidine and two homo-purine strands, poly(rU)·2poly(rA), can be formed (32). The latter complex requires a poly(rA) strand between 28 and 150 nt in length, while the size of the poly(rU) strand has no effect on the type of complex (32). The length of the poly(rA) strand in our experiment is significantly larger (~1000 nt). However, to be sure about the type of triplex, we performed a continuous variation experiment in 100 mM NaCl, 5 mM MgCl₂, 5 mM HEPES, pH 7.5 at 20°C. The plot showed that under our experimental conditions only poly(rA)·2poly(rU) could be formed (data not shown).

Isothermal titration calorimetry

In the calorimetric experiments, the equimolar mixture of poly(rA) and poly(rU) was titrated by magnesium chloride (Fig. 3). As in the case of optical titration, two independent processes were observed: Mg²⁺ binding to unfolded poly(rA)·poly(rU) up to [Mg²⁺]/[phosphate] ≈ 0.25, and subsequent triplex formation. The raw experimental data with a 3 min equilibration time between the injections (Fig. 3A) is affected by the kinetics of the triplex formation: the calorimetric base line deviates from the horizontal between the 10th and 27th points. Additional experiments with 10 min equilibration times (Fig. 3B) showed a significantly better base line. Figure 3C shows the enthalpy of interaction for both experiments, calculated per mole of nucleotide as a function of the concentration ratio [Mg²⁺]/[phosphate].

Figure 3

Typical ITC experiments. (A) The titration of an equimolar mixture of poly(rA) and poly(rU) with Mg²⁺. The injections of Mg²⁺ solutions (5 µl volume) were delivered into the measurement cell with a 3 min waiting time between injections. (more ...)

The enthalpies of these two different processes, Mg²⁺ binding to unfolded poly(rA)·poly(rU) and triplex formation, are favorable and similar to each other, at ~–1.6 kcal mol^–1 of base pairs. Thus, the overall heat of the interaction is exothermic, –3.2 kcal/mol, revealing the enthalpic nature of the process. The heats of conformational transitions of poly(rA)·poly(rU) and poly(rA)·2poly(rU) have been extensively studied (5). These heats are mutually consistent and using Hess’ law one can estimate the enthalpy of reaction 7. For instance, the heat of the similar reaction:

poly(rA)·poly(rU) → 1/2 poly(rA)·2poly(rU) + 1/2 poly(rA) 8

from DSC measurements is endothermic (1.9 kcal/mol) (5). The opposite sign of the heats involved in reactions 7 and 8 is due to the initial conformation of the polynucleotide: folded poly(rA)·poly(rU) duplex for the latter reaction, and unfolded for the former. The enthalpy of poly(rA)·poly(rU) duplex formation at 20°C was directly measured by mixing poly(rA) with poly(rU) in ITC experiments giving a value of –6.0 kcal/mol (33). Thus, the estimated value for reaction 7, –4.1 kcal/mol, is somewhat higher than our experimental value, –3.2 kcal/mol. However, agreement between the calculated and measured values is rather good, keeping in mind the experimental differences between the present and literature data: in our experiments triplex formation is induced by Mg²⁺ at constant temperature, while the calculations used data from duplex and triplex formation in Na ions by mixing complementary strands (33) and increase of temperature (5).

Hydration parameters

As in the case of optical and calorimetric experiments, the acoustic titration curve (Fig. 4, open circles) reveals two different processes: (i) an initial decrease of ΔA up to [Mg²⁺]/[phosphate] ≈ 0.25 due to Mg²⁺ binding to single-stranded polynucleotides; and (ii) a steeper decrease of the ΔA value at [Mg²⁺]/[phosphate] > 0.25, that levels off at [Mg²⁺]/[phosphate] = 0.5, corresponding to reaction 7. In the case of volumetric titration (Fig. 4, closed circles) we observe an effect only for the first process. The triplex formation is not accompanied by any significant volume change. The changes in compressibility (Fig. 4, squares) are calculated from acoustic and densimetric experiments according to equation 6. The overall volume and compressibility effects for both processes are given in Table 1.

Figure 4

Ultrasound (open circles), volume (closed circles) and compressibility (squares) effects of Mg²⁺ binding to an equimolar mixture of poly(rA) and poly(rU) in 2 mM Na-HEPES, pH 7.5 at 20°C.

Table 1.

Hydration effects of Mg²⁺ binding to poly(rA) and poly(rU) and the triplex formation calculated per mole of base pairs

In the present work, acoustic and density measurements of Mg²⁺ interaction with poly(rA)·poly(rU) are accompanied by 10–15% dilution of the polynucleotide solution. However, the dilution process cannot alter the hydration parameters because the concentration dependence of the increment of ultrasound velocity and the apparent molar volume of single- and double-stranded polynucleotides did not reveal any measurable effects (24,34). It should be emphasized that we discuss here the concentration dependence of the hydration parameters and not the actual drop in the polynucleotide concentration upon addition of Mg²⁺ solution. The latter effect is taken into account in equations 1 and 3.

Before analyzing the volume and compressibility effects one should discuss the physical meaning of these parameters. The molecular interpretations of the apparent molar volume, ΦV, and the apparent molar adiabatic compressibility, Φκ_S, in a dilute solution are based on the following relationships (35):

ΦV = V_m + ΔV_h 9

Φκ_S = κ_m + Δκ_h 10

where V_m and κ_m are the intrinsic molar volumes of a solute molecule that is inaccessible to the surrounding solvent, and the intrinsic molar compressibility of this volume, respectively. ΔV_h represents the hydration contribution and consists of the volume change of water around the solute molecule as a result of the solute–water interactions and the void volume between the solute molecule and the surrounding water. Δκ_h is the hydration contribution to the apparent molar adiabatic compressibility, consisting of the changes in the compressibility of water around the solute molecule and the compressibility of the voids between the solute molecules and the surrounding water. The changes in volume and compressibility due to Mg²⁺ interaction with unfolded poly(rA)·poly(rU) or the triplex formation can be expressed as:

ΔV = ΔV_m + ΔΔV_h,b 11

Δκ_S = Δκ_m + ΔΔκ_h,b 12

where ΔV_m and Δκ_m are the effects on intrinsic molar volume and intrinsic molar adiabatic compressibility, respectively, due to structural changes of the polynucleotides, and ΔΔV_h,b and ΔΔκ_h,b are the hydration effects.

As was shown earlier (24), the ΔV_m and Δκ_m terms are negligibly small for Mg²⁺ binding to single-stranded polynucleotides. Thus, the volume and compressibility effects for the first process, Mg²⁺ binding before triplex formation, can be interpreted in terms of the hydration effects. The values are positive (Table 1) and correspond to the release of water molecules from hydration shells of Mg²⁺ and atomic groups of the polynucleotides involved in the binding (24,36,37). In a previous paper (24), we determined the volume and compressibility effects of Mg²⁺ binding to two separate solutes, poly(rA) and poly(rU). These values are also shown in Table 1 and their arithmetic mean is nearly equal to the dehydration effects of Mg²⁺ binding before triplex formation. As shown earlier, such dehydration effects correspond to inner-sphere complexes with several direct contacts between Mg²⁺ and the polynucleotides (24).

The compressibility effect of the triplex formation is significant, while the volume effect is negligibly small (Table 1). Unfortunately, straightforward analysis of the volume and compressibility effects without calculation of water inaccessible surfaces of the polynucleotides before and after the triplex formation is complicated. In addition, due to the increase in linear charge density upon transition from single-stranded polynucleotides to the triplexes, one cannot exclude transformation of Mg²⁺ binding modes. Thus, without structural data on Mg²⁺ complexes with poly(rA), poly(rU) and poly(rA)·2poly(rU), detailed interpretation of the present volume and compressibility data is impossible. However, to get a rough understanding of the magnitude of the present data one can compare the present values with the data on RNA duplex formation at constant temperature. Equimolar mixing of poly(rA) and poly(rU) solutions at 20°C in 116 mM NaCl shows an overall decrease in volume, –2.3 cm³ mol^–1 calculated per base pair (33). Similar measurements on RNA oligonucleotide CCAUCGCUACC, and its complement, GGUAGCGAUGG, reveals a negative effect in both volume and compressibility, ΔV_m = – 6.8 cm³ mol^–1 and Δκ_m= –9.1 × 10^–4 cm³ mol^–1 bar^–1 (38). These data were interpreted in terms of uptake of water molecules due to increased charge density and formation of grooves upon duplex formation (33,38). One can suggest that the positive effects of the triplex formation could be the result of removing water molecules from the major grove of poly(rA)·poly(rU) upon folding in the poly(rU) strand.

Melting curves at physiological salt concentrations

As shown in the previous sections, Mg²⁺ induces triplex formation from an equimolar mixture of poly(rA) and poly(rU) in dilute aqueous solutions (2 mM Na-HEPES, pH 7.5) at 20°C. These experimental conditions allowed us to compare the present hydration data with earlier measurements on Mg²⁺ binding to different nucleic acids (24,34). However, it is important to see whether Mg²⁺ is able to have an effect on reaction 8 at physiological salt concentrations and temperature. Therefore, we constructed optical melting curves of poly(rA)·poly(rU) solutions in 140 mM KCl, 5 mM HEPES, pH 7.5 at various concentrations of MgCl₂. Figure 5 demonstrates typical differential melting curves at 260 and 280 nm without and with 10 mM MgCl₂. As shown earlier (4), optical absorbance at 260 nm is sensitive to duplex formation, while that at 280 nm monitors triplex formation through a disproportionation of the duplexes. The melting experiment without Mg²⁺ shows a large hyperchromic effect at 260 nm and 59.5°C corresponding to duplex unfolding. At 280 nm and 52.6°C we see a small hypochromic effect which is immediately followed by a hyperchromic effect at 60.5°C. The hypochromic effect at 52.6°C corresponds to reaction 8 and hyperchromic effects near 60°C correspond to the melting of the triplex into single-stranded poly(rA) and poly(rU):

poly(rA)·2poly(rU) → poly(rA) + 2 poly(rU) 13

The melting temperatures for reactions 8 and 13 without Mg²⁺ are in good agreement with earlier reported data (4). In the presence of 10 mM Mg²⁺, reaction 8 occurs earlier, T_m = 34.3°C, while reaction 13 is shifted to higher temperatures, T_m = 78.5°C.

Figure 6 shows a phase diagram constructed from the melting experiments of poly(rA)·poly(rU) at different Mg²⁺ concentrations. Three discrete regions are seen, which are separated by phase boundaries corresponding to the reactions 8 and 13. Region I corresponds to the double helix. In region II the triplex helix is more stable, and region III corresponds to single-stranded polynucleotides. The most notable feature of the figure is that the phase boundary corresponding to reaction 8 is located close to physiological temperature and crosses it at ~8–9 mM Mg²⁺. Thus, changes in temperature or Mg²⁺ concentrations can have dramatic effects on the secondary structure of nucleic acids with homo-purine and homo-pyrimidine sequences in cells.

ACKNOWLEDGEMENTS

I am grateful to Professor Victor Bloomfield for stimulating discussions and for critically reading the manuscript. I thank Professor Luis Marky of the University of Nebraska Medical Center for use of the densimeter and for discussions. This investigation was supported by a NIH grant GM28093 to Professor Victor Bloomfield.

REFERENCES

Bloomfield V.A., Crothers,D.M. and Tinoco,I.,Jr (2000) Nucleic Acids: Structures, Properties and Functions. University Sciences Books, Sausalito, CA.

Saenger W. (1984) Principles of Nucleic Acid Structure. Springer-Verlag, New York.

Blake R.D., Massoulie,J. and Fresco,J.R. (1967) A spectral approach to the equilibria between polyriboadenylate and polyribouridylate and their complexes. J. Mol. Biol., 30:, 291–308. [PubMed].

Stevens C.L. and Fensenfeld,G. (1964) The conversion of two-stranded poly (A+U) to three-stranded poly (A+2U) and poly A by heat. Biopolymers, 2:, 293–314.

Krakauer H. and Sturtevant,J.M. (1968) The heats of helix-coil transitions of the poly A–poly U complexes. Biopolymers, 6:, 491–512. [PubMed].

Manning G.S. (1978) The molecular theory of polyelectrolyte solutions with applications to the electrostatic properties of polynucleotides. Q. Rev. Biophys., 11:, 179–246. [PubMed].

Anderson C.F. and Record,M.T.,Jr (1995) Salt-nucleic acid interactions. Annu. Rev. Phys. Chem., 46:, 657–700. [PubMed].

Misra V.K. and Draper,D.E. (1999) On the role of magnesium ions in RNA stability. Biopolymers, 48:, 113–135. [PubMed]

Misra V.K. and Draper,D.E. (2001) A thermodynamic framework for Mg²⁺ binding to RNA. Proc. Natl Acad. Sci. USA, 98:, 12456–12461. [PubMed].

10.

Cheng Y.-K. and Pettitt,B.M. (1992) Stabilities of double- and triple-strand helical nucleic acids. Prog. Biophys. Mol. Biol., 58:, 225–257. [PubMed].

11.

Sun J.S., Garestier,T. and Helene,C. (1996) Oligonucleotide directed triple helix formation. Curr. Opin. Struct. Biol., 6:, 327–333. [PubMed].

12.

Giovannangeli C. and Helene,C. (1997) Progress in developments of triplex-based strategies. Antisense Nucleic Acid Drug Dev., 7:, 413–421. [PubMed].

13.

Jenkins T.C. (2000) Targeting multi-stranded DNA structures. Curr. Med. Chem., 7:, 99–115. [PubMed].

14.

Westhof E. (1988) Water: an integral part of nucleic acid structure. Annu. Rev. Phys. Chem., 17:, 125–144.

15.

Berman H.M. (1991) Hydration of DNA. Curr. Opin. Struct. Biol., 1:, 423–427.

16.

Kankia B.I., Buckin,V.A. and Bloomfield,V.A. (2001) Hexamminecobalt(III)-induced condensation of calf thymus DNA: circular dichroism and hydration measurements. Nucleic Acids Res., 29:, 2795–2801. [PubMed].

17.

Kankia B.I. and Marky,L.A. (2001) Folding of thrombininto a G-quadruplex with Sr²⁺: stability, heat and hydration. J. Am. Chem. Soc., 123:, 10799–10804. [PubMed].

18.

Marky L.A., Kupke,D.W. and Kankia,B.I. (2001) Volume changes accompanying interaction of ligands with nucleic acids. Methods Enzymol., 340:, 149–165. [PubMed].

19.

Chen C.-Y. and Shyu,A.-B. (1995) AU-rich elements: characterization and importance in mRNA degradation. Trends Biochem. Sci., 20:, 465–470. [PubMed].

20.

Sachs A.B. and Buratowski,S. (1997) Common themes in transational and transcriptional regulation. Trends Biochem. Sci., 22:, 189–192. [PubMed].

21.

Riley M., Maling,B. and Chamberlin,M.J. (1966) Physical and chemical characterization of two- and three-stranded adenine–thymine and adenine–uracil homopolymer complexes. J. Mol. Biol., 20:, 359–389. [PubMed].

22.

Puglisi J.D. and Tinoco,I.,Jr (1989) Absorbance melting curves of DNA. Methods Enzymol., 180:, 304–325. [PubMed].

23.

Millero F.J., Ward,G.K. and Chetirkin,P.V. (1977) Relative sound velocities of sea salts at 25°C. J. Acoust. Soc. Am., 61:, 1492–1498.

24.

Kankia B.I. (2003) Binding of Mg²⁺ to single-stranded polynucleotides: hydration and optical studies. Biophys. Chem., 104:, 643–659. [PubMed].

25.

Eggers F. and Funck,T. (1973) Ultrasonic measurements with milliliter liquid samples in the 0.5–100 MHz range. Rev. Sci. Instrum., 44:, 969–977. [PubMed].

26.

Eggers F. and Kaatze,U. (1996) Broad-band ultrasonic measurement techniques for liquids. Meas. Sci. Technol., 7:, 1–19.

27.

Sarvazyan A.P. (1982) Development of methods of precise ultrasonic measurements in small volumes of liquids. Ultrasonics, 20:, 151–154.

28.

Millero F.J. (1972) The partial molar volumes of electrolytes in aqueous solutions. In Horne,R.A. (ed.), Water and Aqueous Solutions. Wiley-Interscience, New York, pp. 519–564.

29.

Barnatt S. (1952) The velocity of sound in electrolytic solutions. J. Chem. Phys., 20:, 278–279.

30.

Owen B.B. and Simons,H.L. (1957) Standard partial molal compressibilities by ultrasonics. I. Sodium chloride and potassium chloride at 25°C. J. Phys. Chem., 61:, 479–482.

31.

Porschke D. (1976) Thermodynamic and kinetic parameters of ion condensation to polynucleotides. Outer sphere complex formed by Mg⁺⁺ ions. Biophys. Chem., 4:, 383–394. [PubMed].

32.

Broitman S.L., Im,D.D. and Fresco,J.R. (1987) Formation of the triple-stranded polynucleotide helix, poly(AAU). Proc. Natl Acad. Sci. USA, 84:, 5120–5124. [PubMed].

33.

Rentzeperis D., Kupke,D.W. and Marky,L.A. (1993) Volume changes correlate with entropies and enthalpies in the formation of nucleic acid homoduplexes: different hydration of A and B conformations. Biopolymers, 33:, 117–125. [PubMed].

34.

Kankia B.I. (2000) Interaction of alkaline-earth metal ions with calf thymus DNA. Volume and compressibility effects in diluted aqueous solutions. Biophys. Chem., 84:, 227–237. [PubMed].

35.

Shio H., Ogawa,T. and Yoshihashi,H. (1955) Measurement of the amount of bound water by ultrasonic interferometer. J. Am. Chem. Soc., 77:, 4980–4982.

36.

Buckin V.A., Kankiya,B.I., Rentzeperis,D. and Marky,L.A. (1994) Mg²⁺ recognizes the sequence of DNA through its hydration shell. J. Am. Chem. Soc., 116:, 9423–9429.

37.

Buckin V.A., Tran,H., Morozov,V. and Marky,L.A. (1996) Hydration effects accompanying the substitution of counterions in the ionic atmosphere of poly(rA)·poly(rU) and poly(rA)·2poly(rU) helices. J. Am. Chem. Soc., 118:, 7033–7039.

38.

Kankia B.I. and Marky,L.A. (1999) DNA, RNA and DNA/RNA oligomer duplexes: a comparative study of their stability, heat, hydration and Mg²⁺ binding properties. J. Phys. Chem., 103:, 8759–8767.