Sodium-23 and lithium-7 NMR spin-lattice relaxation measurements in the study intercalation in DNA.

Sodium-23 spin-lattice relaxation rate (the reciprocal relaxation time) measurements have been used to study the intercalation of 9-aminoacridine in calf thymus DNA. The results are analyzed by a two state model based on the counterion condensation theory and a theory for the quadrupolar relaxation of counterions in polyelectrolyte solutions. It is shown that change of the solvent from H2O to D2O has a negligible effect on the intercalation process. Furthermore, an attempt is made to analyze the dependence of the 7Li spin-lattice relation rate on intercalation of 9-aminoacridine in LiDNA. It is shown that both quadrupolar and dipolar mechanisms contribute to the bound 7Li relation rate, and that both these contributions are reduced upon intercalation of 9-aminoacridine.     

23Na NMR study of the effect of organic osmolytes on DNA counterion atmosphere.

The effect of different organic osmolytes on the DNA counterion condensation layer has been investigated by 23Na NMR relaxation measurements. The zwitterionic compounds glycine, beta-alanine, 4-aminobutyric acid, and 6-aminocaproic acid have shown an increasing capacity to decrease the amount of sodium ions in the vicinity of the macromolecule. The experimental data have been correlated with the dielectric constant increase in their corresponding solutions and have been compared with the prediction of counterion condensation theory. Polyols (sorbitol and mannitol) did not display the same effect. These compounds largely increase the relaxation rate of sodium ions in the proximity of DNA, unlike the zwitterionic compounds. This probably results from a perturbation of the water dynamic around the macromolecule, of the primary or secondary hydration shell of the sodium nuclei involved, or both.     

Sodium-23 NMR studies of cation-DNA interactions

Sodium-23 NMR has been used to study the extent to which monovalent cations associate with double stranded DNA in aqueous solution (28°C, pH = 7.5). On the basis of the two site model for rapid exchange the ²³Na linewidth can be related to the fraction of sodium ions associated with DNA. To test the applicability to this system of the condensation model for the association of small counterions with polyelectrolytes, the concentration dependence of the sodium linewidth has been determined by making additions of NaCl to solutions of tetraethyl or tetrabutylammonium DNA. ([P], the DNA phosphate concentration was about 0.02M). The resulting titration curves extend over a wide range of the ratio [Na]/[P] (0.3–30). When [Na]/[P] ? 3 only sodium is associated, and the extent to which it compensates the charges on DNA does not vary with the addition of salt, at least until [Na]/[P] ≈ 30, the highest concentration examined. When [Na]/[P] ? 3 the tetraalkylammonium species is also associated with DNA; an equation has been derived to account for the effect on the ²³Na linewidth of the competition between sodium and another monovalent cation. Based on the assumption that the fraction of uncompensated charge remaining on DNA after the condensation of both species is constant, this equation fits all the linewidth data if the charge fraction is in the range 0.25 ± 0.10. The value required by the condensation model for DNA in the presence of monovalent counterions is ξ^?1 = 0.24. The reasonable agreement between experimental and theoretical values of the charge fraction and its invariance with respect to large variations in the concentration of added salt indicate that even in moderately concentrated solutions of DNA, the association of sodium can usefully be described in terms of the condensation model. If the theoretical value of the charge fraction is assumed, it follows from fitting the titration curves that the approximate relative affinities for DNA of Na⁺, Et₄N⁺, and Bu₄N⁺ are in the ratio 20:5:1, and the transverse relaxation rate of condensed sodium is 180 ± 10 s^?1.     

Relative affinities of divalent polyamines and of their N-methylated analogues for helical DNA determined by 23Na NMR

Interactions of divalent polyamines with double-helical DNA in aqueous solution are investigated by monitoring the decrease in 23Na NMR relaxation rates as NaDNA is titrated with H3N(+)-(CH2)m-+NH3, where m = 3, 4, 5, or 6. Analogous measurements are made for the same homologous series of methylated polyamines (methonium ions). The dependence of the 23Na relaxation rates on the amount of added divalent cation (M2+) is analyzed quantitatively in terms of a two-state model. The sodium ions are assumed to be in rapid exchange between a "bound" state, where they are close enough to DNA so that it affects their relaxation rate, and a "free" state in bulk solution, where their relaxation rate is the same as in solutions containing no DNA. The distribution of Na+ and M2+ between these states is described quantitatively in terms of an ion-exchange parameter: DM = (pMB)(1-pNaB)n/(pNaB)n(1-pMB), where pNaB and pMB are the fractions of Na+ and M2+ that are close enough to DNA to be considered bound (by the NMR criterion), and n is the number of sodium ions displaced from DNA by the binding of one M2+ ion. For each of the polyamines and methonium ions investigated here, equations derived from this two-state model yield acceptable fittings of the titration curves if roNa, the number of sodium ions bound per DNA phosphate when no competing cations are present, is assigned a value between 0.6 and 1.00. Within this range, changing the value assigned to roNa does change the best-fitted values of DM determined for these polyamines (DH) and for the methonium ions (DMe) but does not alter the following conclusions about the trends in these parameters. (1) For polyamines and methonium ions of the same m, DH exceeds DMe by factors that are significantly larger for m = 3 and 4 than for m = 5 and 6. (2) DH for m = 3 and 4 is larger than DH for m = 5 and 6. (3) DMe for m = 3 and 4 is smaller than DMe for m = 5 and 6.     

Sodium and calcium binding to Panulirus interruptus hemocyanin as studied by 23Na nuclear magnetic resonance.

Addition of Panulirus hemocyanin to NaCl solutions produces marked changes in the 23Na relaxation parameters; they show that sodium ions interact with binding sites on the protein and exchange rapidly with the bulk. The observed non-lorentzian lineshapes and the non-exponential decay of the transverse magnetization indicate that non-extreme narrowing conditions apply and give information on the dynamics of the interaction. Panulirus hemocyanin has at least two classes of Na+ binding sites; the binding constant of the more strongly bound sodium ions is in the order of 1 X 10(2) M-1. Competition between Na+ and Ca2+ for protein binding sites is demonstrated by the effect of Ca2+ on the 23Na relaxation parameters. However, only the more strongly bound Na+ are displaced by Ca2+. The number of Ca2+ needed to displace these sodium ions is 3--5 per oxygen binding site. The 23Na relaxation parameters are influenced also by the state of oxygenation of the protein, indicating a linkage between Na+ and oxygen binding. The simplest interpretation of the data is that sodium ions bind more strongly to oxyhemocyanin in agreement with oxygen equilibrium experiments.     

23Na NMR study of DNA thermal transconformation in presence of cysteamine radioprotector

DNA thermal transconformation is studied in absence and in presence of the cysteamine radioprotector, by observing the delta nu 1/2 variation of 23Na NMR peaks. The sodium state (Free or Bound) is discussed with the help of a two states model with RF and RB relaxation rates. The delta nu 1/2 behaviour during the DNA transconformation shows clearly the electrostatic interaction with cysteamine which is accompanied by an Na+ ejection out of phosphate sites. The temperature dependence of delta nu 1/2 in all cases leads to the conclusion that RBc (the average relaxation rate of sodium nuclei that remain bound in the coil state of DNA) tends to zero.     

Displacement of sodium ions by surfactant ions from DNA. A 23Na-NMR investigation

23Na-NMR probes the ionic composition in the immediate vicinity of the DNA molecule, in the presence of a series of quaternary ammonium bromides, of varying hydrocarbon chain length. The 23Na-NMR line shows two Lorentzian components, in accordance with quadrupolar relaxation theory for S = 3/2 nuclei under slow modulation. Deconvolution of the observed lineshape provides, in a reliable manner, the relative fraction of sodium counterions neutralizing the phosphate sites on DNA. This quantity (p B chi 2) serves as an index of the relative affinities of various surfactant ions toward DNA, Na+ being the reference cation. The results are consistent with site binding of detergent ions to the nucleic acid, an interaction dominated by hydrophobic forces.     

The interaction of polyamines with DNA: a 23Na NMR study.

The interaction between a variety of polyamines, both naturally occurring and synthetic, and calf thymus DNA has been studied using 23Na NMR. The relaxation behaviour of 23Na reflects the extent of interaction of Na+ with DNA phosphate groups and therefore the extent of charge neutralisation of DNA phosphate groups (P) by polyamine amino and imino groups (N) in solutions of DNa, polyamine and Na+. The studies reveal that whereas spermine and spermidine are capable of expelling nearly all of the Na+ ions from DNA at N/P approximately 1, diamines such as putrescine and homologues of spermine and spermidine are capable of neutralising only roughly 50% of DNA phosphates. The results provide a challenge to current models of DNA-polyamine interactions.     

Interactions of monovalent cations with sodium channels in squid axon. II. Modification of pharmacological inactivation gating

The time-, frequency-, and voltage-dependent blocking actions of several cationic drug molecules on open Na channels were investigated in voltage-clamped, internally perfused squid giant axons. The relative potencies and time courses of block by the agents (pancuronium [PC], octylguanidinium [C8G], QX-314, and 9-aminoacridine [9-AA]) were compared in different intracellular ionic solutions; specifically, the influences of internal Cs, tetramethylammonium (TMA), and Na ions on block were examined. TMA+ was found to inhibit the steady state block of open Na channels by all of the compounds. The time-dependent, inactivation-like decay of Na currents in pronase-treated axons perfused with either PC, 9-AA, or C8G was retarded by internal TMA+. The apparent dissociation constants (at zero voltage) for interaction between PC and 9-AA with their binding sites were increased when TMA+ was substituted for Cs+ in the internal solution. The steepness of the voltage dependence of 9-AA or PC block found with internal Cs+ solutions was greatly reduced by TMA+, resulting in estimates for the fractional electrical distance of the 9-AA binding site of 0.56 and 0.22 in Cs+ and TMA+, respectively. This change may reflect a shift from predominantly 9-AA block in the presence of Cs+ to predominantly TMA+ block. The depth, but not the rate, of frequency-dependent block by QX-314 and 9-AA is reduced by internal TMA+. In addition, recovery from frequency-dependent block is not altered. Elevation of internal Na produces effects on 9-AA block qualitatively similar to those seen with TMA+. The results are consistent with a scheme in which the open channel blocking drugs, TMA (and Na) ions, and the inactivation gate all compete for a site or for access to a site in the channel from the intracellular surface. In addition, TMA ions decrease the apparent blocking rates of other drugs in a manner analogous to their inhibition of the inactivation process. Multiple occupancy of Na channels and mutual exclusion of drug molecules may play a role in the complex gating behaviors seen under these conditions.     

Sodium ion and solvent nuclear relaxation results in aqueous solutions of DNA

The field-dependent ²³Na nuclear relaxation in aqueous DNA solutions has been obtained for a range of temperatures, including the DNA melting region. At least two correlation times are needed to characterize the spectral density function for the ²³Na relaxation. For the slow process (with the largest correlation time), the temperature dependence of the coupling constant and the correlation time were determined, and important premelting effects were observed. Possible origins of the slow process are discussed. The last process is shown to be correlated with the properties of the hydration water of DNA as reflected by the ¹⁷O relaxation rates in these solutions. The influence of the polyelectrolyte and NaCl concentrations on the ²³Na relaxation rate is compared with previous results from solutions of linear flexible polyelectrolytes.     

Interaction of an N-methylated polyamine analogue, hexamethonium(2+), with NaDNA: quantitative 14N and 23Na NMR relaxation rate studies of the cation-exchange process

The interactions of the divalent hexamethonium (Hex2+) cation with double-helical calf thymus DNA are investigated by means of 14N NMR and, indirectly, by means of 23Na NMR. During a titration of NaDNA with HexBr2, the displacement of Na+ from DNA by Hex2+ is monitored by concurrent measurements of the Lorentzian 14N signals and the bi-Lorentzian 23Na signals. The variations in the quadrupolar relaxation rates of 14N and 23Na are analyzed according to a simple two-state model for the competition between Hex2+ and Na+ associated with DNA. From this analysis parameters characterizing the exchange process are evaluated, and the following conclusions are drawn: (1) The association of one Hex2+ displaces 1.7-2.0 sodium ions from the vicinity of the DNA. (2) Cation accumulation near DNA neutralizes approximately half of the phosphate charge at all points in the titration. (3) The exchange coefficient characterizing the displacement of Na+ by Hex2+ is of the same order of magnitude as the exchange coefficients determined by NMR for other divalent cations such as Mg2+ and putrescine. These findings imply that the interaction of Hex2+ with DNA is primarily electrostatic in character. The transverse and longitudinal relaxation rates observed for 14N are analyzed under the assumption that the quadrupolar relaxation processes of 14N in Hex2+ associated with DNA can be characterized by a single-exponential correlation function with correlation time tau NB. The resulting value of tau NB, 7.8 +/- 0.8 ns, is 3 orders of magnitude greater than that estimated for Hex2+ in the absence of DNA and is only 3-4 times greater than correlation times reported for 23Na and other quadrupolar cations near DNA. These comparisons indicate that the observed enhancements in the relaxation rates of 14N are due mainly to slowing of the motions that modulate its quadrupolar interactions in Hex2+ near DNA. The magnitudes of tau NB and of the quadrupolar coupling constant of Hex2+ associated with DNA are consistent with the conclusion that this association is primarily electrostatic.     

Importance of coulombic end effects on cation accumulation near oligoelectrolyte B-DNA: a demonstration using 23Na NMR.

The local cation concentration at the surface of oligomeric or polymeric B-DNA is expected, on the basis of MC simulations (Olmsted, M. C., C. F. Anderson, and M. T. Record, Jr. 1989. Proc. Natl. Acad. Sci. USA. 86:7766-7770), to decrease sharply as either end of the molecule is approached. In this paper we report 23Na NMR measurements indicating the importance of this "coulombic" end effect on the average extent of association of Na+ with oligomeric duplex DNA. In solutions containing either 20-bp synthetic DNA or 160-bp mononucleosomal calf thymus DNA at phosphate monomer concentrations [P] of 4-10 mM, measurements were made over the range of ratios 1 < or = [Na]/[LP] < or = 20, corresponding to Na+ concentrations of 4-200 nM. The longitudinal 23Na NMR relaxation rates measured in these NaDNA solutions, Robs, are interpreted as population-weighted averages of contributions from "bound" (RB) and "free" (RF) 23Na relaxation rates. The observed enhancements of Robs indicate that RB significantly exceeds RF, which is approximately equal to the 23Na relaxation rate in an aqueous solution containing only NaCl. Under salt-fre-tconditions ([Na]/[P] = 1), where the enhancement in Robs is maximal, we find that Robs--RF in the solution containing 160-bp DNA is approximately 1.8 times that observed for the 20-bp DNA. For the 160-bp oligomer (which theoretical calculations predict to be effectively polyion-like), we find that a plot of Robs v. [P]/[Na] is linear, as observed previously for sonicated (approximately 700 bp) DNA samples. For the 20-bp oligonucleotide this plot exhibits a marked departure from linearity that can be fitted to a quadratic function of [P]/[Na]. Monte Carlo simulations based on a simplified model are capable of reproducing the qualitative trends in the 23Na NMR measurements analyzed here. In particular, the dependences of Robs--RF on DNA charge magnitude of Z(320 vs. 38 phosphates) and (for the 20-bp oligomer) on [Na]/[P] are well correlated with the calculated average surface concentration of Na+. Thus, effects of sodium concentration on RB appear to be of secondary importance. We conclude that 23Na NMR relaxation measurements are a sensitive probe of the effects of oligomer charge on the extent of ion accumulation near B-DNA oligonucleotides, as a function of [Na] and [P].     

Structure of a DNA intercalation complex as determined by NMR using a paramagnetic probe.

Longitudinal relaxation rates of the protons of the 3,8-dimethyl-N-methyl-phenanthrolinium (DMP) cation in solutions containing DNA are strongly affected by the addition of the paramagnetic manganese (II) ions due to the electron-nuclear dipolar interaction in the ternary Mn-DNA-DMP Complex. Two possible models for the DMP-DNA intercalation complex are examined and one of them is unequivocally discriminated on the basis of the proton relaxation data. It is concluded that in the intercalation complex the long axis of the DMP molecule is almost perpendicular to the hydrogen bonds of the DNA base-pairs.     

Structure of a DNA intercalation complex as determined by NMR using a paramagnetic probe

Longitudinal relaxation rates of the protons of the 3,8-dimethyl-N-methyl-phenanthrolinium (DMP) cation in solutions containing DNA are strongly affected by the addition of the paramagnetic manganese (II) ions due to the electron-nuclear dipolar interaction in the ternary Mn-DNA-DMP complex. Two possible models for the DMP-DNA intercalation complex are examined and one of them is unequivocally discriminated on the basis of the proton relaxation data. It is concluded that in the intercalation complex the long axis of the DMP molecule is almost perpendicular to the hydrogen bonds of the DNA base-pairs.     

Influence of the sodium gradient on contractile activity in pregnant rat myometrium

The effects of varying the sodium gradient-either by lowering [Na+]o or by increasing [Na+]i on the electromechanical properties of pregnant rat uterine smooth muscle were studied. In normal tissues, complete removal of external sodium ions (choline, Tris or sucrose as substitutes) induced a strong and maintained contraction which was dependent on the presence of extracellular calcium ions, and was sensitive to Ca2+-antagonist drugs (Nifedipine; D 600, Mn2+). Electrical recordings showed that the membrane was transiently hyperpolarized (-10 +/- 2.4 mV, n = 20); after 1 minute depolarization accompanied by a spontaneous spike discharge occurred. Partial withdrawal of external sodium ions resulted in following changes in twitch contractions evoked by electrical stimulation: a linear relationship was found between the time constant of twitch relaxation and the external Na-concentration. In Na-rich tissues, where the Na/K pump was blocked, or in the presence of monensin, Na-free solutions (whatever the substitute, even K+ ions) again triggered strong contractions entirely dependent on external calcium but rather insensitive to Ca-antagonists. The Na-free (K+) induced contraction was larger than the Na-free (choline or Tris)-induced contraction. It was concluded that the sodium gradient was an important factor for the regulation of contractile activity of uterine smooth muscle. Na-Ca exchange appeared to mediate twitch relaxation in normal tissues and was responsible for Ca-influx in Na-rich tissues.     

Effect of ions on the dielectric relaxation of DNA

The dielectric relaxation of DNA solutions has been investigated with and without extraneous ions covering a wide frequency range. The effect of monovalent ions such as Na, K, and Li as well as divalent ions such as Mg, Ca, and Hg have been included in the study. These ions are found to have a profound effect on the dielectric increment and the relaxation time without affecting the molecular dimension drastically. This dielectric effect is interpreted as indicating the importance of counterion fluctuation on the low frequency dielectric constant of DNA in solution. The effect of an organic ion, tetra-methylammoniun bromide, has also been studied. This ion has no noticeable effect. A simple theory is derived on the basis of a microscopic model to account for the effect of external ions on the dielectric behavior of solutions of DNA.     

23Na NMR relaxation study of the effects of conformation and base composition on the interactions of counterions with double-helical DNA

NMR relaxation rates (T1(-1) and T2(-1)) have been determined for 23Na in aqueous salt solutions containing various types of helical double-stranded deoxyribonucleic acids. These measurements were performed on three synthetic polynucleotides having different overall conformations, poly-(dA-dT).poly(dA-dT) (alternating B-DNA), poly(dG-dC).poly(dG-dC) at low salt (B-DNA), and Br-poly(dG-dC).Br-poly(dG-dC) (left-handed Z-DNA), and on four types of natural DNA differing in base composition, Clostridium perfringens (26% GC), calf thymus (40% GC), Escherichia coli (50% GC), and Micrococcus lysodeikticus (72% GC). For all types of DNA investigated, except poly(dA-dT).poly(dA-dT), the 23Na NMR spectra measured at 21 degrees C and an applied field of 4.7 T are non-Lorentzian. These non-Lorentzian spectra were analyzed on the basis of the two-state model and the standard theory of nonexponential quadrupolar relaxation processes in order to obtain estimates of the correlation times (tau c) characteristic of the sodium nuclei associated with the various nucleic acids. All of the correlation times estimated in this way are in the range of nanoseconds. The magnitudes of these correlation times show a significant dependence on the overall conformation of the nucleic acid (B vs. Z) but not on its base composition. To investigate the concentration dependence of tau c, sodium or magnesium salts were added to solutions of Br-poly(dG-dC).Br-poly(dG-dC) (Z-DNA).(ABSTRACT TRUNCATED AT 250 WORDS)     

Inactivation of Rb+ and Na+ occlusion on (Na+,K+)-ATPase by modification of carboxyl groups

This paper demonstrates and characterizes inactivation by N,N'-dicyclohexylcarbodiimide (DCCD) of Rb+ and Na+ occlusion in pig kidney (Na+,K+)-ATPase. Rb+ and Na+ occlusion dependent on oligomycin are measured with a manual assay. Parallel measurement of phosphorylation (by Pi plus ouabain) and Na+ or Rb+ occlusion lead to stoichiometries of 3 Na+ or 2 Rb+ per pump molecule. Inactivation of cation occlusion by DCCD shows the following features: (a) Rb+ and Na+ occlusion are inactivated with identical rates and (b) DCCD concentration dependence shows first-order kinetics and also proportionality to the ratio of DCCD to protein, (c) Rb+ and Na+ occlusion are equally protected from DCCD, by Rb+ ions with high affinity (or Na+ ions with lower affinity), (d) inactivation is only slightly pH-dependent between 6 and 8.5 but (e) is significantly accelerated by several hydrophobic amines while a water-soluble nucleophile, glycine ethyl ester has no effect, and (f) inactivation is exactly correlated with inactivation of (Na+,K+)-ATPase activity of ATP-dependent Na+/K+ exchange in reconstituted vesicles and with the magnitude of E1Na+----E2(Rb+) conformational transitions measured with fluorescence probes. The simplest hypothesis to explain the results is that DCCD modifies one (or a small number of) critical carboxyl residues in a non-aqueous cation binding domain and so blocks occlusion of 2 Rb+ or 3 Na+ ions. The results suggest further that Na+ and K+(Rb+) bind to the same sites and are transported sequentially on the same trans-membrane segments. A second effect of the DCCD treatment is a 4-8-fold shift of the conformational equilibrium E2(Rb+)----E1Rb+ toward E1Rb+. This is detected by (a) reduction in apparent Rb+ affinity for Rb+ occlusion or Rb+/Rb+ exchange in vesicles and (b) direct demonstration of an increased rate of E2(K+)----E1Na+ and decreased rate of E1Na+----E2(K+). This effect is not protected against by Rb+ ions and probably reflects modification of a second group of residues. Modification of (Na+,K+)-ATPase by carbodiimides is complex. Depending on the nature of the carbodiimide (water- or lipid-soluble), ratio of carbodiimide to protein, and perhaps source of the enzyme, inactivation might result either from modification of critical carboxyls, as suggested by this work, or from internal cross-linking as proposed by Pedemonte, C. H. and Kaplan, J. H. ((1986) J. Biol. Chem. 261, 3632-3639).     

Geometry of DNA–dye intercalation complexes from the study of linear dichroic spectra of stretched films

Films of DNA–dye complexes were combined with films of pure DNA deposited on poly(vinyl alcohol) support and stretched. Reproducible dichroic spectra were obtained after equilibration of the stretched films at 93% relative humidity. Dye diffusion into the supporting poly(vinl alcohol) matrix was eliminated. The long axis of intercalated acriflavine is perpendicular to the DNA helix; proflavine deviates slightly and 9-aminoacridine significantly from such an intercalation geometry. The dichroism of two mutually perpendicularly polarized transitions of 9-aminoacridine enabled us to determine both the angles of tilt and twist of the plane of the dye relative to the DNA helix in the complex.     

A study of the quadrupolar NMR splittings of 7Li+, 23Na+, and 133Cs+ counterions in macroscopically oriented DNA fibers.

The hydration and temperature dependencies of the 23Na+, 133Cs+, and 7Li+ quadrupolar splitting have been determined in hydrated, macroscopically oriented DNA fibers. At low water contents the quadrupolar splitting is found to decrease as the water content increases, regardless of counterion, while at high water contents the hydration dependence is reversed. The 23Na+ and 133Cs+ quadrupolar splittings decrease as the temperature increases, while the 7Li+ splitting shows the opposite behavior. At high water contents the 23Na+ and 133Cs+ splittings decrease, and then, after passing zero splitting, increase as the temperature increases. The interpretation of the temperature dependence is discussed in terms of a two-site model (free and bound ions) and a three-site model (free ions and specifically or nonspecifically bound ions). It is suggested that a three-site model is more consistent with the data for the present system. At high water contents, the temperature dependence of the 7Li+ splitting vanishes, indicating counterion condensation. The behavior of the 7Li+ splitting is confirmed by measurements on DNA fibers in equilibrium with a C2H5OD-D2O-LiCl solution. The salt dependence in this system is weak. The counterion quadrupolar splitting is seen to be very sensitive to structural transitions in double-helical DNA.