Protonated polynucleotides. VII. Thermal transitions between different complexes of polyinosinic acid and polycytidylic acid in an acid medium

The interaction between poly (I) and poly (C) in acid medium has been studied by potentiometric titration, mixing curves and thermal denaturation. Phase diagramms as a function of ionic strength, pH, and temperature have been established. From these data it is shown that the acid titration of the complex poly (I) · poly (C) passes through a triple-stranded intermediate poly (I) · poly (C) · poly (C⁺) to yield finally the protonated double-helical complex poly (I) · poly (C⁺). The mixing curves indicate the sole presence of the three-stranded complex in the intermediate zone. On the basis of the pK's the coexistence between the three-stranded complex with the neighboring double-stranded structure is demonstrated in a narrow rang of pH and ionic strength. The geometry of the base arrangements, their conformation and the sense of the strands are discussed in the light of the data presented. A Hoogsteen-type pairing between the bases for poly (I) · poly (C⁺) is favored, although the reverse Hoogsteen pair cannot be excluded.     

Helix–coil transition for poly(L-glutamic acid) in water–ethanol solutions

Potentiometric titrations and some complementary optical rotation data are presented for solutions of poly(L - glutamic acid) (PGA) in several H₂O–ethanol mixtures. The data allow the determination of the intrinsic pK (pK₀), slope of the apparent. pK (pK_app), versus degree of ionization curves and of the enthalpy of ionization as a function of ethanol concentration. The variation of the degree of ionization at which the helix–coil transformation occurs with ethanol and temperature is also determined. Finally free energy, enthalpy, and intropy changes associated with the helix–coil transformation for the uncharged conformers are determined from the titration curves. The effect of the ethanol is to increase the stability of the helical conformation of PGA for both the charged and the uncharged forms of the polymer. The stabilization of the uncharged helix is essentially an entropic effect.     

Cooperative nonenzymic base recognition II. thermodynamics of the helix-coil transition of oligoadenylic + oligouridylic acids

The properties of oligonucleotide helices of adeuylic- and uridylic acid oligomers have been investigated by measurements of hypo-and hyperchromieity. High ionic strengths favor the formation of triple helices. Thus, the double helix-coil transition can be studied (without interference by triple helices) only at low ionic-strength. A “phase diagram” is given representing the T_m-values of the various transitions at different ionic strengths for the system A(pA)₁₇ + U(pU)₁₇. Oligonucleolides of chain lengths <8 always form both double and triple helices at the nucleotide concentrations required for base pairing. For this reason the double helix-coil transition without coupling of the triple helix equilibrium can only be measured for chain lengths higher than 7. Melting curves corresponding to this transition have been determined for chain lengths 8, 9, 10, 11, 14 and 18 at different concentrations. An increase in nucleotide concentration leads to an increase in melting temperature. The shorter the chain length the lower the T_m-value and the broader the helix-coil transition. The experimental transition curves have been analysed according to a staggering zipper model with consideration of the stacking of the adeuylic acid single strands and the electrostatic repulsion of tlip phosphate charges on opposite strands. The temperature dependence of the nucleation parameter has been accounted for by a slacking factor x. The stacking factor expresses the magnitude of the stacking enthalpy. By curve fitting xwas computed to be 0.7, corresponding to a stacking enthalpy of about S kcal/mole. The model described allows the reproduction of the experimental transition curves with relatively high accuracy. In an appendix the thermodynamic parameters of the stacking equilibrium of poly A and of the helix-coil equilibria of poly A + poly U at neutral pH are calculated (ΔH_A = ?7.9 kcal/mole for the poly A stacking and ΔH₁₂ = ?10.9 kcal/mole for the formation of the double helix from the randomly coiled single strands). A formula for the configurational entropy of polymers derived by Flory on the basis of a liquid lattice model is adapted to calculate the stacking entropies of adenylic oligomers.     

The structures and properties of poly-5-methylcytidylic acid

The optical rotatory dispersion properties of poly 5MeC, poly diMeC, and 5MeCMP-(5′) in 0.1M Na+ have been studied at various pH values and temperatures. Poly 5MeC and poly diMeC have optical properties which are similar to those for poly C; however, poly 5MeC has a biphasic melting profile in the pH range from 3.8 to 5.4 similar to that observed for poly 51C. Using titration, ionic strength, and pH dependence measuements, the data for poly 5MeC are interpreted in terms of the following scheme at pH 4.0 and 0.1 ionic strength: triple-strand helix ^37°C double-strand helix ^79°C single-strand coil. Support for this scheme is discussed. The effect of the methyl group is discussed in terms of similar structural possibilities for other polymers of cytidylic acid.     

Potentiometric titration of poly-S-carboxyethyl-L-cysteine

Potentiometric titration curves have been determined for aqueous solutions of poly-S-carboxyethyl-L -cysteine, which is subject to the β-coil transition by a change in pH. Reversibility and time dependence of the titration curves are examined by different methods in order to establish the conditions for obtaining equilibrium curves. The β-coil transition is manifest, at some region on the equilibrium titration curve, if pH – log (α/1 – α) is plotted against α. Assuming a value, 4.00, for pK_int, the free-energy change for the β-coil transition of uncharged polymer has been evaluated from the extrapolation of the observed titration curves and is found to depend on the ionic strength and polymer concentration. The Henderson-Hasselbach plot of the titration curve yields clearer distinction between the β-form and random coil, and it permits estimation of the content of β-form at, a given pH. Comparison of the conformational titration curve with the circular dichroic measurements leads to a value of ?10,000° for [θ]₂₂₃ for the pure β-structure. Precipitation which occurs at low degrees of ionization and, especially, at high ionic strength does not reveal any discontinuous change of the titration curve, which suggests that, the degree of ionization of the precipitated β-form is not very different from that in solution.     

On the kinetics of helix formation between complementary ribohomopolymes and deoxyribohomopolymers

The rate of double helix formation by single-stranded poly A plus poly dT, poly dA plus poly U, poly dA plus dT, poly G plus poly dC, poly dG plus poly C, and poly dG plus poly dC have been investigated and compared to rates of ribohomopolymer helix formation rates. After correction for molecular weight, comparisons of rate data at 30°C below the melting temperature of the double helix show that:

• 1 Rates of helix formation by all combinations of guanine plus cytosine homopolymers are the same.
• 1 The rate of helix formation for poly dA plus poly dT is three times faster than the rate for poly A plus poly U. Rates of formation of DNA-RNA hybrid molecules are intermediate between these two rates, but closer to the poly dA plus poly dT rate.

The effect of temperature on the rate of helix formation is interpreted in terms of a steady-state model for helix propagation. The results are consistent with a mechanism in which the formation of the second base pair is the rate-determining step.     

Comparison of helix stabilities of poly-L-lysine, poly-L-ornithine, and poly (L-diaminobutyric acid)

The helix–coil transitions of aqueous solutions of poly-α-L -lysine (PLL), poly-α-L -ornithine (PLO), and poly(α,γ-L -diaminobutyric acid) (PLDBA) have been investigated as functions of pH at 25°C and of temperature at pH 11.75, where these polymers are uncharged; in the cases of the latter two polyamino acids, the transitions have also been studied as functions of apparent pH in methanol-water solution (50/50 by volume). The helix stability of the polypeptides is shown to be a direct function of the number of methylene groups on the side chain. From an analysis of potentiometric titration data, we find that the difference between the helix stability of PLL and that of PLO is due to a difference of about 1 e.u. in the ΔS° of the transition. Combining the “melting curves” obtained from optical rotatory dispersion studies with the potentiometric titration data permits evaluation of the initiation parameter Z (or 1/σ^½) of the statistical mechanical theories for these transitions. The value obtained for Z in the case of uncharged aqueous PLO is ca. 35.     

Electrostatic interactions in ionic homopolypeptides in solutions of moderate ionic strength

At sufficiently high ionic strength, long-range electrostatic interactions in a polyelectrolyte such as poly(L -glutamic acid) might be adequately approximated in matrix calculations by use of statistical weights representing second-order interactions. The validity of this assumption has been investigated making use of experimental observations (CD spectra and titration curves) for poly(L -glutamic acid) as a function of temperature in 0.1–0.5M sodium chloride. Theoretical analysis, using a statistical weight matrix proposed by Warashina and Ikegami, is based on the Zimm-Rice theory. Implementation differs from that of Warashina and Ikegami in one respect. Refinement of the initial estimates is achieved using a form of the configuration partition function which does not assume diagonalization of the statistical weight matrix. This difference is of no consequence for the values of σ and s, but it does produce somewhat different values for the statistical weights used to represent the electrostatic interactions. The method used to treat electrostatic interactions in poly(L -glutamic acid) in 0.1M sodium chloride can be viewed as successful in that it properly reproduces the helix–coil transition and titration curves in this solvent and the molecular-weight dependence of the titration curves yields values for s in harmony with those obtained using a treatment which is independent of model, and gives a reasonable ionic-strength dependence for the electrostatic parameters. Furthermore, the model can account for measured helix–coil transitions and titration curves in homopolypeptides in which the side chain is —(CH₂)_xNHCO(CH₂)_yCOOH. The model, however, is not exact. It does not properly account for the molecular-weight dependence of the helical content for polymers of low degree of polymerization.     

Poly(8-aminoguanylic acid): formation of ordered self-structures and interaction with poly(cytidylic acid).

Poly(8-aminoguanylic acid) has in neutral solution a novel ordered structure of high stability. The 8-amino group permits formation of three hydrogen bonds between two residues along the "top", or long axis, of the purines. The usual hydrogen bonding protons and Watson-Crick pairing sites are not involved in the association. The bonding scheme has a twofold rotation axis and is hemiprotonated at N(7). Poly(8NH2G) is converted by alkaline titration (pK = 9.7) to a quite different ordered structure, which is the favored form over the range approximately pH 10-11. The bonding scheme appears to be composed of a planar, tetrameric array of guanine residues, in which the 8-amino group does not participate in interbase hydrogen bonding. Poly (8NH2G) does not interact with poly(C) in neutral solution because of the high stability of the hemiprotonated G-G self-structure. Titration to the alkaline plateau, however, permits ready formation of a two-stranded Watson-Crick helix. In contrast to the monomer 8NH2GMP, poly(8NH2G) does not form a triple helix with poly(C) under any conditions. The properties of the ordered structures are interpreted in terms of a strong tendency of the 8-amino group to form a third interbase hydrogen bond, when this possibility is not prevented by high pH.     

Spectrophotometric titration studies on poly (uridylic acid)

In order to provide preliminary data for the interpretation of the spectrophotometric titration properties of RNA, the spectral changes accompanying the ionization of poly-(uridylic acid) have been determined, as have the ionization constants as a function of salt concentration and the enthalpies and entropies of ionization. The spectral propertics and ionization constants of poly (uridylic acid) have been compared with those of 2′(3′) Uridine monophosphate and of uridine; significant, differences have been established. The results obtained are consistent with the hypothesis that uracil residues in poly U are stacked only in concentrated salt solutions.     

Monte Carlo studies on potentiometric titration of poly(glutamic acid)

The potentiometric titration of poly(glutamic acid) with special attention to its helix-coil transition is investigated in terms of the previously developed Monte Carlo method. The simulations of the potentiometric titration are carried out for helical and coiled form of the peptide, separately. A cylindrical rod with spherical ionizable groups is adopted as each conformational model of poly(glutamic acid) molecule. A spherical charge with a hard core potential is assumed as a mobile hydrated ion. The helix-coil transition curves are analyzed by the Zimm-Bragg theory. A satisfactory agreement is achieved for the titration curves with the experimental data in most cases. The significance and the limitations of the simulation method are discussed.     

Ionic strength dependence of the hysteresis in the polyriboadenylate-polyribouridylate system.

The hysteresis observed in cyclic acid-base titrations of the three-standed polyribonucleotide helix poly (A)-2 POLY (U) strongly depends on ionic strength. For NaCl and at 25 degrees C, hysteresis occurs in the limited concentration range between 0.03 M and 1.0 M(NaCl). The transition points associated with the cyclic conversions between the triple helix and the poly (A)-poly (A) double helix and (free) poly (U) constitute a (pH ionic strength) phase diagram covering the ranges of stability and metastability of the hysteresis system. Variations with NaCl concentration of some hysteresis parameters can be quantitatively described in terms of polyelectrolyte theories based on the cylinder-cell model for rodlike polyions. The results of this analysis suggest that the metastability is predominantly due to dlectrostatic energy barriers preventing the equilibrium transition of the partially protonated triple helix above a critical pH value. Ultraviolet absorbance and potentiometric titration data of poly (A)in the acidic pH range can be analyzed in terms of two types of double-helical structures. Spectrophotometric titrations reveal isosbestic wavelengths for structural transitions of poly (A). "Time effects" commonly observed in poly (A) titrations are suggested to reflect helix transitions between the two acidic structures.     

Pressure dependence of the helix-coil transition temperature for polynucleic acid helices

Thermal denaturation of DNA's and the corresponding helix–coil transformation of artificial polyribonucleic and polydeoxyribonucleic acids have been studied extensively both theoretically^1–13 and experimentally. ^14–30 Much less work has been carried out on the properties of these polynucleic acids at high pressure, and in particular, on the presure dependence of the helix–coil transition temperature.^31–33 Light-scattering techniques have been used in this study to measure the pressure dependence of the helix–coil transition temperature of the two- and three-stranded helices of polyriboadenylic and polyribouridilic acids and of calf thymus DNA. From the slopes of the transition temperature vs. pressure curves and heats of transition obtained from the literature,^20,34 the following volume changes from these helix–coil transitions have been obtained: (a) ?0.96 cc/mole of nucleotide base pairs for the poly (A + U) transition, (b) +0.35 cc/mole of nucleotide base trios for the poly (A + 2U) transition, and (c) +2.7 cc/mole of nucleotide base pairs for the DNA transition. The relative magnitudes and signs of these volume changes which show that poly (A + U) is destabilized by increased pressure, whereas poly (A + 2U) and calf thymus DNA are stabilized by increased pressure, indicates that further development of the helix–coil transition theory for polynucleotides is needed.     

A calorimetric study of the triple helix formation: interaction of poly(C) - guanosine - protonated poly(C).]

The triple helix formation of poly(C) - guanosine - poly(C+) was investigated by the help of an LKB scanning micro-calorimeter. The existence of the triple helix could also be shown by recording the melting curves. The ultraviolet absorption at different wave lengths namely 275 nm, 260 nm, and 245 nm was plotted as a function of the temperature. Furthermore formation of the triple helix was shown by plotting the ultraviolet absorption at 245 nm during the increasing addition of guanosine solution to a fixed amount of poly(C) in the solution. Finally the formation of the triple helix was demonstrated by plotting the ultraviolet absorption at 245 nm of a certain mixture of the components while the pH value of the solution was continuously lowered. All these methods show that the monomer interacts with the polymer double helix to form a triple helix. The calorimetric measurements show that the reaction enthalpy is concentration dependent. Above a threshold concentration a rapid increase of the reaction enthalpy is observed. This increase occurs in a very narrow concentration interval. Above this interval a final value of the reaction enthalpy is reached. The amount of the reaction enthalpy for the interaction of guanosine with poly(C) - poly(C+) double helix is 5.5 Kcal (mol base triplet)-1.     

Hydrogen-ion binding by tobacco-mosaic-virus protein polymers.

Hydrogen ion titration curves of tobacco mosaic virus protein have been measured in various conditions of protein concentration, temperature, ionic strength, and rate of pH change. The polymers present at each stage are deduced from turbidity and sedimentation data, plus published information. A simple semi-quantitative analysis of the curves is given, and the pK values of the two abnormal carboxylates in single helix are estimated as 6.4 and about 7.0. Disks, and some faster-forming unknown polymers in the same size range, have been abnormal carboxylate with pK 6.9. These results are most easily interpreted in terms of electrostatic interactions between carboxylates, probably at the axial ends of the protein subunits.     

Effect of adenosine protonation on its interaction with polyuridylic acid

The effect of adenosine protonation on complex formation between poly(U) and adenosine has been studied by UV spectroscopy, titration and equilibrium dialysis techniques. A method has been developed to estimate the "misincorporation" of ionized monomer molecules into a polynucleotide--monomer complex. The method is based on combining the titration and binding data. Using this method it is shown that protonated adenosine interacts to some extent with poly(U) in the course of A.2 poly(U) dissociation at acidic pH. Qualitative differences between the effects of ionization of the polymer and monomer components on polynucleotide--monomer interaction are discussed.     

Characterization and determination of titratable groups of proteins by linearization of titration curves. II. Application to lysozyme

The potentiometric acid-base titration curve of fully protonated lysozyme at ionic strengths of 0.10 and 1.0 m has been performed. The stoichiometry and the pK_a values of each titratable group have been determined through the linearization of titration curves. Two types of carboxylic groups with pK_a values of 3.76 and 5.02, the imidazole group with pK_a 7.37 and the amine group with pK_a 9.63, have been identified at an ionic strength of 0.10 m at 25.0°C. The number of titratable groups found per mole of protein has been 5.12 and 5.60 for the two types of carboxylic groups, 1.13 for the imidazole group, and 3.19 for the amino groups. The endpoint of the titration of the protein obtained by this method accords quite well with the endpoint obtained by the use of Gran function applied to the excess of strong base.     

A distinct 14 residue site triggers coiled-coil formation in cortexillin I.

We have investigated the process of the assembly of the Dictyostelium discoideum cortexillin I oligomerization domain (Ir) into a tightly packed, two-stranded, parallel coiled-coil structure using a variety of recombinant polypeptide chain fragments. The structures of these Ir fragments were analyzed by circular dichroism spectroscopy, analytical ultracentrifugation and electron microscopy. Deletion mapping identified a distinct 14 residue site within the Ir coiled coil, Arg311-Asp324, which was absolutely necessary for dimer formation, indicating that heptad repeats alone are not sufficient for stable coiled-coil formation. Moreover, deletion of the six N-terminal heptad repeats of Ir led to the formation of a four- rather than a two-helix structure, suggesting that the full-length cortexillin I coiled-coil domain behaves as a cooperative folding unit. Most interestingly, a 16 residue peptide containing the distinct coiled-coil 'trigger' site Arg311-Asp324 yielded approximately 30% helix formation as monomer, in aqueous solution. pH titration and NaCl screening experiments revealed that the peptide's helicity depends strongly on pH and ionic strength, indicating that electrostatic interactions by charged side chains within the peptide are critical in stabilizing its monomer helix. Taken together, these findings demonstrate that Arg311-Asp324 behaves as an autonomous helical folding unit and that this distinct Ir segment controls the process of coiled-coil formation of cortexillin I.     

THE VALENCE OF CORPUSCULAR PROTEINS

By the use of two extreme models: a hydrated sphere and an unhydrated rod the valence (net charge) of corpuscular proteins can be successfully calculated from electric mobility data by the Debye-Hückel theory (modified to include the effect of the ions in the ion atmosphere) in conjunction with the electrophoretic theory of Henry. As pointed out by Abramson, this permits a comparison with values for the valence from titration data. Electrometric titration measurements of serum albumin B (Kekwick) have been determined at several ionic strengths. These results, together with the available data in the literature for serum albumin B, egg albumin, and β-lactoglobulin have been used to compare values for the valence calculated from measurements of titration, electrophoresis, and membrane potentials. The results indicate that the usual interpretation of titration curves is open to serious question. By extrapolation of the titration data to zero ionic strength and protein concentration, there results an "intrinsic" net charge curve describing the binding of H⁺ (OH^-) ion alone. This curve agrees closely, in each case, with values of the valence calculated from mobility data (which in turn are in close accord with those estimated from membrane potential measurements). The experimental titration curves in the presence of appreciable quantities of ions and protein deviate widely from the ideal curve. It is suggested that, under these conditions, binding of undissociated acid (base) leads to erroneous values for the net charge. This binding would not affect the electrophoretic mobility. Values of the net charge obtained by the two extreme models from electrophoretic data are in agreement within 15 to 20 per cent. The agreement between the cylindrical model and the titration data is somewhat better in each case than with the sphere; i.e., this comparison enables a choice to be made between asymmetry and hydration in the interpretation of results from sedimentation and diffusion measurements on proteins. It is concluded that the proteins discussed here are somewhat asymmetric and also hydrated.     

Disorder–order transitions induced in anionic homopolypeptides by cationic detergents

CD spectra have been obtained for poly(L -glutamic acid) and poly(L -aspartic acid) as functions of temperature and concentration of cationic detergents. Dodecylammonium chloride induces a coil–helix transition in fully ionized poly(L -glutamic acid). The interaction of the monomeric detergent with the polypeptide is responsible for the conformational transition. The detergent concentration required to produce the transition is independent of temperature. The CD of fully ionized poly(L -aspartic acid) is nearly unaffected by dodecylammonium chloride, in marked contrast to the situation found with poly(L -glutamic acid). However, these results do not imply that dodecylammonium chloride interacts differently with aspartyl and glutamyl residues. The observed results can be accounted for by the well-known fact that the glutamyl residue has a higher helix-forming tendency that the aspartyl residue. Cetyltrimethylammonium chloride destabilizes the helical form of poly(L -glutamic acid). This detergent presents an exception to the usual ability of ionic detergents to promote formation of ordered structures in oppositely charged homopolypeptides.