Description of Base Motion and Role of Excitations in the Melting of Poly(dG) · Poly(dC)

Abstract

We study the contribution of various vibrational modes to the melting of poly(dG) · poly(dC). We find that the principal contribution comes from the H-bond breathing modes that have been observed in Raman scattering and that we have associated with helix melting. We show the softening of these modes on approach to melting in agreement with the observed behavior. We also describe the contribution to melting from base rotation modes that others have suggested are important in melting.     

Description of base motion and role of excitations in the melting of poly(dG).poly(dC).

We study the contribution of various vibrational modes to the melting of poly(dG).poly(dC). We find that the principal contribution comes from the H-bond breathing modes that have been observed in Raman scattering and that we have associated with helix melting. We show the softening of these modes on approach to melting in agreement with the observed behavior. We also describe the contribution to melting from base rotation modes that others have suggested are important in melting.     

Sequence and temperature effect on hydrogen bond disruption in DNA determined by a statistical analysis

We use the modified self-consistent phonon approximation theory to calculate temperature dependent interbase hydrogen bond disruption profiles for a number of six base pair repeating sequence infinite B-DNA polymers with various guanine-cytosine/adenine-thymine ratios. For comparison we also include results we have obtained in our earlier work on several B-DNA homopolymers, copolymers and a four-base-pair repeating sequence polymer. Our theory gives a statistical estimate of thermal fluctuational disruption probability of individual hydrogen bonds in individual base pairs in DNA as a function of temperature. The calculated probabilities show no sequence dependence at premelting temperatures, in agreement with proton exchange measurements. These probabilities however become very sensitive to base sequence at temperatures close to the observed melting temperatures. Multi-phasic critical transitions are found in which a portion of base pairs are disrupted at temperatures below the final disruption temperature. These transitions include localized as well as non-localized base pair opening. The localized transitions involve disruption of a few base-pairs at every other location without large scale base unstacking, and they may not appear in the observed UV curves with current resolution. On the other hand the overall disruption behavior is consistent with observations. The midpoint transition temperatures are close to the observed melting temperatures and these temperatures show the observed linear dependence on guanine-cytosine content. Our calculations indicate that our theory can be used effectively to calculate H-bond disruption behavior of different DNA sequences. Received: 20 February 1996 / Accepted: 2 May 1996     

The vibrational normal modes of beta-barrels in an IgG antibody molecule.

Based on the quasi-continuity model, and using the method of group theory, we studied the normal vibrations of the VL- and the CHL-beta-barrels in an IgG molecule. We put emphasis on the Raman- and the infrared-active normal modes. The Raman modes we obtained include both the breathing motion mode (or the dominant low-frequency mode) which corresponds to the maximum peak in the Raman spectrum, and the normal modes that correspond to the lower peaks. Our calculated vibration frequencies are found to be in good agreement with the experimental results observed by Painter et al. (Biopolymers 20 (1981) 243). The method and work presented in this paper may improve Chou's quasi-continuity theory in calculating the vibrational modes of a beta-barrel protein.     

Melting and premelting phenomenon in DNA by laser Raman scattering.

Raman spectra of DNA from calf thymus DNA have been taken over a wide range of temperatures (25°–95°) in both D₂O and H₂O. A study of the temperature dependence of the Raman spectra shows that the temperature profiles of the intensities and frequencies of the various bands fall into four different categories: (1) base bands that show a reversible increase in intensity prior to the melting region, i.e., a definite premelting phenomenon; (2) base bands that show little or no temperature dependence; (3) deoxyribose-phosphate backbone vibrations that show no temperature dependence up to the melting region, at which point large decreases in intensity occur; and (4) slow frequency changes in certain in-plane vibrations of guanine and adenine due to deuteration of the C-8 hydrogen of these purines in D₂O. Certain Raman bands arising from each of the four bases, adenine, thymine, guanine, and cytosine have been found to undergo a gradual increase in intensity prior to the melting region at which point large, abrupt increases in intensity occur. The carbonyl stretching band of thymine, involved in the interbase hydrogen bonding actually undergoes both a gradual shift to a lower frequency as well as an increase in intensity. These changes provide evidence that some change in the geometry of the bases relative to each other begins to occur around 50°C, well below the melting region of 70°–85°C. From the spectra taken at various temperatures, the DNA appears to remain in the B conformation until the melting point is reached, at which time the DNA progresses into a disordered random-coil form. No A-form conformation is found either in the premelting or the melting region.     

A critical review of the nucleosidic vibration modes appearing in the 800-500-cm-1 spectral region, based on new harmonic dynamics calculations

On the basis of a harmonic dynamics calculation, it is shown that in the 800–500-cm^?1 spectral region of DNA vibrational spectra, the characteristic Raman peaks and ir bands do not arise from the same nucleosidic motions. The Raman spectra involve mainly the ring-breathing modes of nucleic bases while the ir spectra reveal essentially their out-of-plane vibrations. Moreover, the calculated results show the splitting of the guanine- and adenine-residue breathing modes upon their coupling with the sugar-pucker motions. This fact is in agreement with the poly[d(G-C)] and poly[d(A-T)] Raman spectra.     

A Raman spectroscopic study on 5′-rGMP gel and self-aggregate

In order to analyze the melting behavior of 5′-rGMP gel at acidic pH and self-aggregate near neutral pH we have obtained Raman spectra of aqueous solutions of 5′-rGMP at various temperatures. At low temperature the intensities of Raman peaks at 502, 585, 1083, 1179, 1322, 1366, 1487, and 1578 cm^?1 decrease due to the formation of ordered structure (Raman hypochromism). In contrast, the peaks at 671, 725, 813, and 1338 cm^?1 become stronger at low temperature (Raman hyperchromism). The Raman hyperchromism of the 671- and 813-cm^?1 peaks have been explained in terms of detailed structural models. Recently, the 668- and 682-cm^?1 peaks in the Raman spectrum of aqueous 5′-rGMP solution have been attributed to the guanine ring breathing vibrations in C3′- and C2′-endo conformers [Benevides, J. B., Lemur, D. & Thomas, G. J., Jr. (1984) Biopolymers 23 , 1011–1024]. On the basis of this information our Raman data can be interpreted to suggest that the continuous helix model of 5′-rGMP gel is right-handed. The 1487-cm^?1 peak intensity has been used to monitor the melting profies at several pHs. Near neutral pH the melting profile shows a single transition, whereas at acidic pH it shows two transitions. From these observations we propose possible pathways for the melting of 5′-rGMP gel formed at acidic pH and self-aggregate formed near neutral pH.     

Vibrational modes of A-DNA,B-DNA,and A-RNA backbones: An application of a green-function refinement procedure

Normal vibrational analysis was carried out for DNA molecules in both A and B conformations as well as for A-RNA. A simplified backbone model was examined and expanded to include the backbone phosphate-group and the ribose ring. We applied the new force-constant refinement procedure discussed in the preceeding paper [Van Zandt, L. L., Lu, K.-C. & Prohofsky, E. W. (1977) Biopolymers, 16 , 2481–90] to fit some observed frequencies in the Raman spectra for all three nucleic acids with the same set of force constants. The results indicate that the observed frequency shift can be attributed to the conformational change solely. We ignored the second-order differences in force constants for the different geometries. The agreement between the observed and calculated frequencies derived from the final refined set of force constants is good and apparently justifies this assumption. Two modes previously assigned to the symmetric diester O-P-O stretch and the symmetric dioxy O‥P‥O stretch are actually fitted. They are mainly backbone phosphate-group modes. The refined ribose-ring force-constants were transferred to the calculation of the vibrational spectrum of tetrahydrofuran. The overall agreement is again good. We discuss these calculations and the resulting normal modes. We also discuss the application of the Green-function refinement scheme and several strategies adopted to bias the convergence of the procedure.     

Dynamics of the [4Fe-4S] cluster in Pyrococcus furiosus D14C ferredoxin via nuclear resonance vibrational and resonance Raman spectroscopies, force field simulations, and density functional theory calculations

We have used (57)Fe nuclear resonance vibrational spectroscopy (NRVS) to study oxidized and reduced forms of the [4Fe-4S] cluster in the D14C variant ferredoxin from Pyrococcus furiosus (Pf D14C Fd). To assist the normal-mode assignments, we conducted NRVS with D14C ferredoxin samples with (36)S substituted into the [4Fe-4S] cluster bridging sulfide positions, and a model compound without ligand side chains, (Ph(4)P)(2)[Fe(4)S(4)Cl(4)]. Several distinct regions of NRVS intensity are identified, ranging from "protein" and torsional modes below 100 cm(-1), through bending and breathing modes near 150 cm(-1), to strong bands from Fe-S stretching modes between 250 and ～400 cm(-1). The oxidized ferredoxin samples were also investigated by resonance Raman (RR) spectroscopy. We found good agreement between NRVS and RR frequencies, but because of different selection rules, the intensities vary dramatically between the two types of spectra. The (57)Fe partial vibrational densities of states for the oxidized samples were interpreted by normal-mode analysis with optimization of Urey-Bradley force fields for local models of the [4Fe-4S] clusters. Full protein model calculations were also conducted using a supplemented CHARMM force field, and these calculations revealed low-frequency modes that may be relevant to electron transfer with Pf Fd partners. Density functional theory (DFT) calculations complemented these empirical analyses, and DFT was used to estimate the reorganization energy associated with the [Fe(4)S(4)](2+/+) redox cycle. Overall, the NRVS technique demonstrates great promise for the observation and quantitative interpretation of the dynamical properties of Fe-S proteins.     

Far-Infrared Study of the Vibrational Modes of 5′-GMP Gels and Crystals of Na+ and K+

Abstract

Five Far-Infrared (50–600 cm^?1) spectra are presented: the sodium and potassium salts of 5′ Guanosine Monophosphate (GMP), each salt in both the gel and crystal conformations, and poly(rG). Measurements were performed at a sample temperature of 10 Kunder vacuum with a liquid He-cooled bolometer. The spectra were fit with Lorentzians and assignments are suggested. There are noteworthy differences in oscillator strengths and frequencies of the bands between all spectra. We report the tentative observation of a 100 cm^?1 mode which is in the neighborhood of a mode observed by Raman spectroscopy in solution (1) and dried gels (2).     

Tyrosine hydrogen-bonding and environmental effects in proteins probed by ultraviolet resonance Raman spectroscopy

Ultraviolet resonance Raman spectra with 229-nm excitation are reported for aqueous tyrosine and for ovomucoid third domain proteins from chicken [OMCHI3(-)] and from chachalaca [OMCHA(-)], as well as alpha 1-, alpha 2-, and beta-purothionin. At this excitation wavelength interference from phenylalanine is minimized, and it is possible to determine the frequencies of the Tyr ring modes nu 8a and nu 8b. The nu 8b frequency decreases with the degree of Tyr H-bond donation, reaching a limiting value for deprotonated tyrosine. This spectroscopic indicator of H-bond strength was calibrated by using the model compound p-cresol in H-bond acceptor solutions for which the enthalpy of H-bond formation can be obtained from the literature. With this calibration it is possible to estimate Tyr H-bond enthalpies in proteins for which Tyr is a H-bond donor; values of 13.7, 9.6, and 11.2 kcal/mol were found for OMCHA3(-) and for alpha 1- (or alpha 2-) and beta-purothionin, respectively. The intensity of the 1176-cm-1 nu 9a band of Tyr excited at 229 nm and also the intensity ratio of the Tyr 830/850-cm-1 Fermi doublet excited at 200 nm both correlate strongly with the estimated H-bond enthalpies, but large deviations are seen for the purothionins, reflecting a special environment for the Tyr residue of these proteins, which is believed to be constrained in a hydrophobic pocket. The molar intensity of the strong approximately 1000-cm-1 nu 12 band of phenylalanine in aqueous solution is about half the value observed in most proteins.(ABSTRACT TRUNCATED AT 250 WORDS)     

Vibrational fluctuations of hydrogen bonds in a DNA double helix with alternating TA type inserts.

The Green's function technique is applied to a study of breathing modes in a DNA double helix which contains a region of different base pairs from the rest of the double helix. The calculation is performed on an alternating poly(dC-dG).poly(dC-dG) helix in the B conformation with four consecutive base pairs replaced by a model of a biological promoter region with four alternating T-A,A-T base pairs, henceforth referred to as (TATA)2. The average stretch of interbase hydrogen bonds is found to be amplified around the insert. This is likely related to the (TATA)2 insert having a lower stability against hydrogen bond melting than the two semi-infinite poly(dC-dG).poly(dC-dG) helices. The insert region may be considered to be a site of enhanced tendency to melt in such a helix. The results show that an alternating AT insert of four base pairs has a larger average hydrogen bond stretch inside and outside the insert region than the average hydrogen bond stretch inside and outside an insert of four consecutive A-T base pairs, henceforth referred to as (AAAA).(TTTT). Calculations are performed which show that the enhancement of the average hydrogen bond stretch around an alternating TA type insert is greatly dependent upon the local modes and not the inband modes. The amount of local mode enhanced average stretch is explored as a function of insert size.     

Vibrational normal modes and dynamical stability of DNA triplex poly(dA). 2poly(dT): S-type structure is more stable and in better agreement with observations in solution.

A normal-mode and statistical mechanical calculation was carried out to determine the vibrational normal modes, contribution of internal fluctuations to the free energy, and hydrogen bond disruption of DNA triplex poly(dA).2poly(dT). The calculation was performed on both the x-ray fiber diffraction model with a N-type sugar conformation, and a newly proposed model with a S-type sugar conformation. Our calculated normal modes for the S-type structure are in better agreement with observed IR spectra for samples in D2O solution. We also find that the contribution of internal fluctuations to free energy, premelting hydrogen bond disruption probability, and hydrogen bond melting temperatures for the Hoogsteen and Watson-Crick hydrogen bonds all show that the S-type structure is dynamically more stable than the N-type structure in a nominal solution environment. Therefore our calculation supports experimental findings that the triplex d(T)n.d(A)nd(T)n most likely adopts a S-type sugar conformation in solution or at high humidity. Our calculations, however, do not preclude the possibility of an N-type conformation at lower humidities.     

Melting profile and temperature dependent binding constant of an anticancer drug daunomycin-DNA complex

We calculate thermal fluctuational base pair opening probability and the drug binding constant of a daunomycin-bound Poly d(CGTA) · Poly d(TACG) at temperatures from room temperature to its melting temperature. For comparison we also carry out a calculation on a drug-free DNA with the same sequence. Our calculations are carried out by means of a statistical approach using microscopic structures and established force fields and with cooperative effects incorporated into the algorithm. Both hydrogen bond disruption probabilities and drug unstacking probability are determined self-consistently. These probabilities are then used to determine temperature dependent base pair opening probabilities and the drug binding constant. The calculated base pair opening probabilities and drug binding constant are found to be in fair agreement with experiments carried out at room temperature. Our calculation shows cooperative base pair disruption and drug dissociation at certain critical temperatures close to the observed melting temperatures for similar helices. We find that the temperature dependence of the drug binding constant fits well to the van't Hoff relation, in agreement with observations. Our calculation indicates the occurrence of a premelting transition in the drug-bound DNA helix. Some comments are made about this premelting transition.     

Resonance Raman enhancement of phenyl ring vibrational modes in phenyl iron complex of myoglobin.

Resonance Raman spectra are reported for the organometallic phenyl-FeIII complexes of horse heart myoglobin. We observed the resonance enhancement of the ring vibrational modes of the bound phenyl group. They were identified at 642, 996, 1,009, and 1,048 cm-1, which shift to 619, 961, 972, and 1,030 cm-1, respectively, upon phenyl 13C substitution. The lines at 642 and 996 cm-1 are assigned, respectively, as in-plane phenyl ring deformation mode (derived from benzene vibration No. 6a at 606 cm-1) and out-of-plane CH deformation (derived from benzene vibration No. 5 at 995 cm-1). The frequencies of the ring "breathing" modes at 1,009 and 1,048 cm-1 are higher than the corresponding ones in phenylalanine (at 1,004 and 1,033 cm-1) and benzene (at 992 and 1,010 cm-1), indicating that the ring C--C bonds are strengthened (or shortened) when coordinated to the heme iron. The excitation profiles of these phenyl ring modes and a porphyrin ring vibrational mode at 674 cm-1 exhibit peaks near its Soret absorption maximum at 431 nm. This appears to indicate that these phenyl ring modes may be enhanced via resonance with the Soret pi-pi transition. The FeIII--C bond stretching vibration has not been detected with excitation wavelengths in the 406.7-457.9-nm region.     

Temperature dependent vibrational modes of glycosidic bond in disaccharide sugars

We studied the temperature dependent vibrational modes of the glycosidic bond in trehalose, sucrose, and maltose at wavenumbers ranging from 1000 to 1200 cm(-1). We found that the slope of temperature dependent Raman shifts of the glycosidic bond in trehalose and sucrose changed at temperatures around 120 degrees C, indicating a bond length or a bond angle (dihedral and torsional angles) change. However, we did not observe any slope change in maltose because the melting temperature of maltose is very close to 120 degrees C. We also found, at temperatures below 120 degrees C, that Raman shifts of the vibrational modes of the glycosidic bond in trehalose showed the strongest temperature dependence among the three disaccharides.     

Calculation of the Raman frequencies as a function of pressure in the solid phases II and III (III’) of benzene

We calculate here the Raman frequencies of the lattice modes A(A _g ), B(B _2g ) and C(B _1g B _3g ) as a function of pressure at room temperature for the solid phases (II, III and III’) of benzene. This calculation is performed using volume data through the mode Grüneisen parameter. It is found that our calculated frequencies of those lattice modes increase with increasing pressure, as expected. Calculated frequencies are in good agreement with the measurements of the three lattice modes for the solid phases studied in benzene.     

The oscillator strengths of hemoproteins. Their relation to the coordination structure and magnetic susceptibility

The oscillator strengths of hemoproteins in the light frequency range of 1.11 X 10(4) to 3.23 X 10(4) cm-1 (wavelength range of 900 to 310 nm) were measured by means of computer-assisted spectrophotometry. The obtained values of oscillator strength per molar heme ranged from about 1.4 to 2.2. By comparing the oscillator strength values of the ferric and ferric cyanide-bound forms of hemoproteins and also the values of low molecular weight ferric heme complexes, it was found that the oscillator strength was lower for those hemoproteins whose heme was coordinated with strong field ligands. It was also found that the hemoproteins showing a smaller pH-dependent change in the carbon monoxide-difference spectrum had lower oscillator strengths. The following linear relation was observed, with various ligand complexes of bovine methemoglobin, horse metmyoglobin, and ferric horseradish peroxidase, between the oscillator strength (f) determined in the present study and the respective magnetic susceptibility (10(6) X chi 20 degrees M) values in the literature: f = A (10(6) X chi 20 degrees M) + B. The values of constants A and B in the equation were estimated for horseradish peroxidase, methemoglobin, and metmyoglobin. On varying the temperature in the range of 0 to 40 degrees C, the oscillator strength of the metmyoglobin-azide complex changed in parallel with the change in the spin state. Taking advantage of the fact that fluoride complexes of many hemoproteins show 10(6) X chi 20 degrees M values close to 14,500 and also that the values of intersection B are around 86.4% of the respective values of the fluoride complexes of ferric horseradish peroxidase, methemoglobin, and metmyoglobin, an empirical equation was evolved for the calculation of an approximate 10(6) X chi 20 degrees M value from the f value of a given complex (fobs) and that of the fluoride complex (fF) of a hemoprotein. The approximate magnetic susceptibilities of various ligand complexes of bovine lactoperoxidase could be thus calculated with the equation. The oscillator strengths of ferrous hemoproteins were also investigated and ligand-dependent regular changes were found.