Conformational Characteristics and Correlations in Crystal Structures of Nucleic Acid Oligonucleotides: Evidence for Sub-states

Abstract

Sugar phosphate backbone conformations are a structural element inextricably involved in a complete understanding of specific recognition nucleic acid ligand interactions, from early stage discrimination of the correct target to complexation per se, including any structural adaptation on binding. The collective results of high resolution DNA, RNA and protein/DNA crystal structures provide an opportunity for an improved and enhanced statistical analysis of standard and unusual sugar-phosphate backbone conformations together with corresponding dinucleotide sequence effects as a basis for further exploration of conformational effects on binding. In this study, we have analyzed the conformations of all relevant crystal structures in the nucleic acids data base, determined the frequency distribution of all possible ?, ζ, α, β and γ backbone angle arrangements within four nucleic acid categories (A-RNA and A-DNA, free and bound B-DNA) and explored the relationships between backbone angles, sugar puckers and selected helical parameters. The trends in the correlations are found to be similar regardless of the nucleic acid category. It is interesting that specific structural effects exhibited by the different unusual backbone sub-states are in some cases contravariant. Certain α/γ changes are accompanied by C3′ endo (north) sugars, small twist angles and positive values of base pair roll, and favor a displacement of nucleotide bases towards the minor groove compared to that of canonical B form structures. Unusual ?/ζ combinations occur with C2′ (south) sugars, high twist angles, negative values of base pair roll, and base displacements towards the major groove. Furthermore, any unusual backbone correlates with a reduced dispersion of equilibrium structural parameters of the whole double helix, as evidenced by the reduced standard deviations of almost all con- formational parameters. Finally, a strong sequence effect is displayed in the free oligomers, but reduced somewhat in the ligand bound forms. The most variable steps are GpA and CpA, and, to a lesser extent, their partners TpC and TpG. The results provide a basis for considering if the variable and non-variable steps within a biological active sequence precisely determine morphological structural features as the curvature direction, the groove depth, and the accessibility of base pair for non covalent associations.     

The Interaction Between the Novel Intercalator Diethidium Cation and B-Form DNA: a Theoretical Study

Abstract

This research is an effort to further understand the physicochemical interaction between the novel drug molecule diethidium (2,7-diamino 9-[2,7 diamino 10-nN- phenanthridium] 10- nN- phenanthridium) and its biological receptor DNA. The ultimate goal is the elucidation of this novel class of drugs as potential pharmaceutical agents. Understanding the physico- chemical properties of this drug as well as the mechanism by which it interacts with DNA should ultimately allow the rational design of novel anti-cancer or anti-viral drugs.

A novel binding structure for the diethidium cation to B-form DNA is herein described. Molecular modeling on the complex formed between diethidium and a dodecamer of double-stranded B-form DNA, CGCGAATTCGCG, has shown that this complex is indeed fully capable of participating in the formation of a stable intercalation site. It was expected that diethidium would have a mechanism of intercalation significantly different from other classical intercalators because a) Its structure, that of two perpendicular planes, each known to have excellent intercalation properties, is novel b) The linker region length is zero c) The tilt between the two planes of the drug matches the geometry of the space available to this drug in the major groove.

We have studied the complex formed when diethidium enters the central site of the B-DNA dodecamer through the major groove. The complex forms several classes of intercalation structures, which are all stable and vary from “partially intercalate” to “fully intercalated”. Multiple minimizations show the drug to be very mobile within the intercalation site. Further, some structures show organization and concomitant stiffening of the DNA above the intercalation site, with a disorganization and disruption of the regular B-DNA structure immediately below the intercalation site. This particular phenomena may be expected to lead to significantly different physicochemical properties for the diethidium complex with respect to other known intercalators, because this sort of vectorial difference in structure above and below the site of intercalation is unknown in existing intercalators, as far as the authors are aware. In addition, we expect the mechanism of interaction between drug and DNA to be described by “direct ligand transfer”, wherein the drug is transferred from duplex DNA to duplex DNA without re-entering the solvent.¹

This work is the first instance known to the authors of a novel drug entity that was deduced solely by mathematical reasoning ² and described subsequently by computational methods. Evidence that diethidium should interact with its target site DNA differently from other known intercalators is strong.     

Sequence preference for BI/BII conformations in DNA: MD and crystal structure data analysis

Deciphering sequence information from sugar-phosphate backbone is finely tuned through the conformational substates of DNA. BII conformation, one of the conformational substates of B-DNA, is known to play a key role in DNA-protein recognition. BI and BII are identified by the epsilon-zeta difference, which is negative in BI and positive in BII. Our analysis of MD and crystal structures shows that BII conformation is sequence specific and dinucleotides GC, CG, CA, TG, TA show high preference to take up BII conformation, while TT, TC, CT, CC dinucleotides rarely take up this conformation. Significant changes were observed in the dinucleotide parameters viz. twist, roll, and slide for the steps having BII conformation. Interestingly, the magnitude of variation in the dinucleotide parameters is seen to depend mainly on two factors, the magnitude of epsilon-zeta difference and the presence or absence of BII conformation in the second strand, across the WC base-paired dinucleotide step. Based on these two factors, the conformational substate of a dinucleotide step can be further classified as BI.BI (BI conformation in both strands), BI.BII (BI conformation in one strand and BII conformation in the other), and BII.BII (BII conformation in both strands). The occurrence of BII in both strands was found to be quite rare and thus, it can be concluded that BI.BI and BI.BII hybrid steps are more favorable than a BII.BII step. In conformity with the sequence preference seen for dinucleotides in each strand, BII.BII combination of backbone conformation was observed only for GC, CG, CA, and TG containing dinucleotide steps. We further classified BII.BII step as strong BII and weak BII depending on the magnitude of the average epsilon-zeta difference. The dinucleotide steps which belong to the category of strong BII, have large twist, high positive slide and negative roll values, while those in the weak BII group have roll, twist, and slide values similar to that of hybrid BI.BII steps. This conformational property could be contributing to the groove opening/closing and thus can modulate protein-DNA interaction.     

Assessment of the molecular dynamics structure of DNA in solution based on calculated and observed NMR NOESY volumes and dihedral angles from scalar coupling constants

To assess the accuracy of the molecular dynamics (MD) models of nucleic acids, a detailed comparison between MD-calculated and NMR-observed indices of the dynamical structure of DNA in solution has been carried out. The specific focus of our comparison is the oligonucleotide duplex, d(CGCGAATTCGCG)(2), for which considerable structural data have been obtained from crystallography and NMR spectroscopy. An MD model for the structure of d(CGCGAATTCGCG)(2) in solution, based on the AMBER force field, has been extended with a 14 ns trajectory. New NMR data for this sequence have been obtained in order to allow a detailed and critical comparison between the calculated and observed parameters. Observable two-dimensional (2D) nuclear Overhauser effect spectroscopy (NOESY) volumes and scalar coupling constants were back-calculated from the MD trajectory and compared with the corresponding NMR data. The comparison of these results indicate that the MD model is in generally good agreement with the NMR data, and shows closer accord with experiment than back-calculations based on the crystal structure of d(CGCGAATTCGCG)(2) or the canonical A or B forms of the sequence. The NMR parameters are not particularly sensitive to the known deficiency in the AMBER MD model, which is a tendency toward undertwisting of the double helix when the parm.94 force field is used. The MD results are also compared with a new determination of the solution structure of d(CGCGAATTCGCG)(2) using NMR dipolar coupling data.     

Solution structure and backbone dynamics of the AF-6 PDZ domain/Bcr peptide complex

The human AF-6, a scaffold protein between cell membrane-associated proteins and the actin cytoskeleton, plays an important role in special cell-cell junctions and signal transduction. It can be phosphorylated by the protein kinase Bcr, which allows efficient binding of the C terminus of Bcr to the PDZ domain of AF-6 and consequently enhances the binding affinity of AF-6 to Ras. Formation of the AF-6, Bcr, and Ras ternary complex results in down-regulation of the Ras-mediated signal transduction pathway. To better understand the molecular basis for the recognition of the AF-6 PDZ domain and Bcr, we solve the solution structure of the AF-6 PDZ domain complexed with the C-terminal peptide of Bcr and explore the interactions between them in detail. Compared with previously reported structures, the complex exhibits a noncanonical binding mode of PDZ/peptide. Owing to the distinct residues involved in the AF-6 PDZ domain and Bcr peptide interaction, the interaction mode does not adapt to the existing classification rules that have been put forward, based on the ligand or the PDZ domain specificity. Furthermore, the PDZ domain of AF-6 can bind to the C terminus of Bcr efficiently after phosphorylation of AF-6 by the Bcr kinase. The phosphorylation may induce a conformational change of AF-6, which makes the binding surface on the PDZ domain accessible to Bcr for efficient binding. This study not only characterizes the structural details of the AF-6 PDZ/Bcr peptide complex, but also provides a potential target for future drug design and disease therapy.     

Computation of nitroxide-nitroxide distances in spin-labeled DNA duplexes

Nanometer distances in nucleic acids can be measured by EPR using two 1-oxyl-2,2,5,5-tetramethylpyrroline radicals, with each label attached via a methylene group to a phosphorothioate-substituted backbone position as one of two phosphorothioate diastereomers (R(P) and S(P)). Correlating the internitroxide distance to the geometry of the parent molecule requires computational analysis of the label conformers. Here, we report sixteen 4-ns MD simulations on a DNA duplex d(CTACTGCTTTAG) .d(CTAAAGCAGTAG) with label pairs at C7/C19, T5/A17, and T2/T14, respectively. For each labeled duplex, four simulations were performed with S(P)/S(P), R(P)/R(P), S(P)/R(P), and R(P)/S(P) labels, with initial all trans label conformations. Another set of four simulations was performed for the 7/19-labeled duplex using a different label starting conformation. The average internitroxide distance r(MD) was within 0.2 A for the two sets of simulations for the 7/19-labeled duplex, indicating sufficient sampling of conformational space. For all three labeled duplexes studied, r(MD) agreed with experimental values, as well as with average distances obtained from an efficient conformer search algorithm (NASNOX). The simulations also showed that the labels have conformational preferences determined by the linker chemistry and label-DNA interactions. These results establish computational algorithms that allow use of the 1-oxyl-2,2,5,5-tetramethylpyrroline label for mapping global structures of nucleic acids.     

Molecular dynamics simulations show how the FMRP Ile304Asn mutation destabilizes the KH2 domain structure and affects its function

Mutations or deletions of FMRP, involved in the regulation of mRNA metabolism in brain, lead to the Fragile X syndrome (FXS), the most frequent form of inherited intellectual disability. A severe manifestation of the disease has been associated with the Ile304Asn mutation, located on the KH2 domain of the protein. Several hypotheses have been proposed to explain the possible molecular mechanism responsible for the drastic effect of this mutation in humans. Here, we performed a molecular dynamics simulation and show that the Ile304Asn mutation destabilizes the hydrophobic core producing a partial unfolding of two α-helices and a displacement of a third one. The affected regions show increased residue flexibility and motion. Molecular docking analysis revealed strongly reduced binding to a model single-stranded nucleic acid in agreement with known data that the two partially unfolded helices form the RNA-binding surface. The third helix, which we show here to be also affected, is involved in the PAK1 protein interaction. These two functional binding sites on the KH2 domain do not overlap spatially, and therefore, they can simultaneously bind their targets. Since the Ile304Asn mutation affects both binding sites, this may justify the severe clinical manifestation observed in the patient in which both mRNA metabolism activity and cytoskeleton remodeling would be affected.     

Solution structure and backbone dynamics of the KH-QUA2 region of the Xenopus STAR/GSG quaking protein

The Quaking protein belongs to the family of STAR/GSG domain RNA-binding proteins and is involved in multiple cell signalling and developmental processes in vertebrates, including the formation of myelin. Heteronuclear NMR methods were used to determine the solution structure of a 134 residue fragment spanning the KH and QUA2 homology regions of the Quaking protein from Xenopus laevis (pXqua) in the absence of RNA. The protein is shown to adopt an extended type I KH domain fold that is connected to a structured alpha-helix in the C-terminal QUA2 region by means of a highly flexible linker. A comparison with the solution structure of the related protein splicing factor 1 (SF1) indicates that most aspects of the RNA-binding interface are conserved in pXqua, although the "variable loop" region that follows the second beta-strand possesses two additional alpha-helices. The structure of pXqua provides an appropriate template for building models of important homologues, such as GLD-1 and Sam68. Measurements of the (15)N relaxation parameters of pXqua confirm that the polypeptide backbone of the QUA2 region is more dynamic than that of the KH portion, and that the C-terminal helix is partially structured in the absence of RNA. By comparison with a random coil reference state, the nascent structure in the QUA2 region is estimated to contribute 15.5kJmol(-1) to the change in conformational free energy that occurs on forming a complex with RNA. Since STAR/GSG proteins may regulate alternative splicing by competing with SF1 in the nucleus for specific branch-point sequences that signal intronic RNA, the formation of secondary structure in the QUA2 region in the unbound state of pXqua has important functional consequences.     

NMR studies of the backbone flexibility and structure of human growth hormone: a comparison of high and low pH conformations

(15)N NMR relaxation parameters and amide (1)H/(2)H-exchange rates have been used to characterize the structural flexibility of human growth hormone (rhGH) at neutral and acidic pH. Our results show that the rigidity of the molecule is strongly affected by the solution conditions. At pH 7.0 the backbone dynamics parameters of rhGH are uniform along the polypeptide chain and their values are similar to those of other folded proteins. In contrast, at pH 2.7 the overall backbone flexibility increases substantially compared to neutral pH and the average order parameter approaches the lower limit expected for a folded protein. However, a significant variation of the backbone dynamics through the molecule indicates that under acidic conditions the mobility of the residues becomes more dependent on their location within the secondary structure units. In particular, the order parameters of certain loop regions decrease dramatically and become comparable to those found in unfolded proteins. Furthermore, the HN-exchange rates at low pH reveal that the residues most protected from exchange are clustered at one end of the helical bundle, forming a stable nucleus. We suggest that this nucleus maintains the overall fold of the protein under destabilizing conditions. We therefore conclude that the acid state of rhGH consists of a structurally conserved, but dynamically more flexible helical core surrounded by an aura of highly mobile, unstructured loops. However, in spite of its prominent flexibility the acid state of rhGH cannot be considered a "molten globule" state because of its high stability. It appears from our work that under certain conditions, a protein can tolerate a considerable increase in flexibility of its backbone, along with an increased penetration of water into its core, while still maintaining a stable folded conformation.     

Stabilization of duplex DNA and RNA by dangling ends studied by free energy simulations

Single unpaired nucleotides at the end of double‐stranded nucleic acids, termed dangling ends, can contribute to duplex stability. Umbrella sampling free energy simulations of dangling cytosine and guanine nucleotides at the end of duplex and single stranded RNA and DNA molecules have been used to investigate the molecular origin of dangling end effects. In unrestraint simulations, the dangling end nucleotides stayed close to placements observed in experimental structures. Calculated free energy contributions associated with the presence of dangling nucleotides were in reasonable agreement with experiment predicting the general trend of a more stabilizing effect of purine vs. pyrimidine dangling ends. In addition, the calculations indicate a more significant stabilizing effect of dangling ends at the 5′‐end vs. 3′‐end in case of DNA and the opposite trend in case of RNA. Both electrostatic and van der Waals interactions contribute to the duplex stabilizing effect of dangling end nucleotides. The free energy simulation scheme could also be used to design dangling end nucleotides that result in enhanced duplex stabilization. © 2013 Wiley Periodicals, Inc. Biopolymers 101: 418–427, 2014.     

NMR solution structure and backbone dynamics of domain III of the E protein of tick-borne Langat flavivirus suggests a potential site for molecular recognition

Flaviviruses cause many human diseases, including dengue fever, yellow fever, West Nile viral encephalitis, and hemorrhagic fevers, and are transmitted to their vertebrate hosts by infected mosquitoes and ticks. Domain III of the envelope protein (E-D3) is considered to be the primary viral determinant involved in the virus-host-cell receptor interaction, and thus represents an excellent target for antiviral drug development. Langat (LGT) virus is a naturally attenuated BSL-2 TBE virus and is a model for the pathogenic BSL-3 and BSL-4 viruses in the serogroup. We have determined the solution structure of LGT-E-D3 using heteronuclear NMR spectroscopy. The backbone dynamics of LGT-E-D3 have been investigated using 15N relaxation measurements. A detailed analysis of the solution structure and dynamics of LGT-E-D3 suggests potential residues that could form a surface for molecular recognition, and thereby represent a target site for antiviral therapeutics design.     

Comparison of a QM/MM force field and molecular mechanics force fields in simulations of alanine and glycine "dipeptides" (Ace-Ala-Nme and Ace-Gly-Nme) in water in relation to the problem of modeling the unfolded peptide backbone in solution

We compare the conformational distributions of Ace-Ala-Nme and Ace-Gly-Nme sampled in long simulations with several molecular mechanics (MM) force fields and with a fast combined quantum mechanics/molecular mechanics (QM/MM) force field, in which the solute's intramolecular energy and forces are calculated with the self-consistent charge density functional tight binding method (SCCDFTB), and the solvent is represented by either one of the well-known SPC and TIP3P models. All MM force fields give two main states for Ace-Ala-Nme, beta and alpha separated by free energy barriers, but the ratio in which these are sampled varies by a factor of 30, from a high in favor of beta of 6 to a low of 1/5. The frequency of transitions between states is particularly low with the amber and charmm force fields, for which the distributions are noticeably narrower, and the energy barriers between states higher. The lower of the two barriers lies between alpha and beta at values of psi near 0 for all MM simulations except for charmm22. The results of the QM/MM simulations vary less with the choice of MM force field; the ratio beta/alpha varies between 1.5 and 2.2, the easy pass lies at psi near 0, and transitions between states are more frequent than for amber and charmm, but less frequent than for cedar. For Ace-Gly-Nme, all force fields locate a diffuse stable region around phi = pi and psi = pi, whereas the amber force field gives two additional densely sampled states near phi = +/-100 degrees and psi = 0, which are also found with the QM/MM force field. For both solutes, the distribution from the QM/MM simulation shows greater similarity with the distribution in high-resolution protein structures than is the case for any of the MM simulations.     

Investigations into sequence and conformational dependence of backbone entropy,inter-basin dynamics and the Flory isolated-pair hypothesis for peptides

The populations and transitions between Ramachandran basins are studied for combinations of the standard 20 amino acids in monomers, dimers and trimers using an implicit solvent Langevin dynamics algorithm and employing seven commonly used force-fields. Both the basin populations and inter-conversion rates are influenced by the nearest neighbor's conformation and identity, contrary to the Flory isolated-pair hypothesis. This conclusion is robust to the choice of force-field, even though the use of different force-fields produces large variations in the populations and inter-conversion rates between the dominant helical, extended beta, and polyproline II basins. The computed variation of conformational and dynamical properties with different force-fields exceeds the difference between explicit and implicit solvent calculations using the same force-field. For all force-fields, the inter-basin transitions exhibit a directional dependence, with most transitions going through extended beta conformation, even when it is the least populated basin. The implications of these results are discussed in the context of estimates for the backbone entropy of single residues, and for the ability of all-atom simulations to reproduce experimental protein folding data.     

The structure and dynamics in solution of Cu(I) pseudoazurin from Paracoccus pantotrophus

The solution structure and backbone dynamics of Cu(I) pseudoazurin, a 123 amino acid electron transfer protein from Paracoccus pantotrophus, have been determined using NMR methods. The structure was calculated to high precision, with a backbone RMS deviation for secondary structure elements of 0.35+/-0.06 A, using 1,498 distance and 55 torsion angle constraints. The protein has a double-wound Greek-key fold with two alpha-helices toward its C-terminus, similar to that of its oxidized counterpart determined by X-ray crystallography. Comparison of the Cu(I) solution structure with the X-ray structure of the Cu(II) protein shows only small differences in the positions of some of the secondary structure elements. Order parameters S2, measured for amide nitrogens, indicate that the backbone of the protein is rigid on the picosecond to nanosecond timescale.     

Molecular dynamics simulations of HIV-1 protease monomer: Assembly of N-terminus and C-terminus into beta-sheet in water solution

HIV-1 protease (HIV-PR) consists of two identical subunits that are united together through a four-stranded antiparallel beta-sheet formed of the peptide termini of each monomer. Since the active site exists only in the dimer, a strategy that is attracting more and more attention in inhibitor design and which may overcome the serious drug resistance caused by competitive inhibitors is to block the peptide termini of the monomer, thereby interfering with formation of the active dimer. In the present work, we performed several extensive molecular dynamics (MD) simulations of the HIV-PR monomer in water to illustrate its solvated conformation and dynamics behavior. We found that the peptide termini usually assembled into beta-sheet after several nanoseconds' simulation, and became much less flexible. This beta-sheet is stabilized by intramolecular interactions and is not easily disaggregated under the present MD simulation conditions. This transformation may be an important transition during the relaxing and equilibrating of the HIV-PR monomer in aqueous solution, and the terminal beta-sheet may be one of the major conformations of the solvated HIV-PR monomer termini in water. This work may provide new insights into the dynamics behavior and dimerization mechanism of HIV-PR, and more significantly, offer a more rational receptor model for the design and discovery of novel dimerization inhibitors than crystalline structures.     

Role of structure and dynamics of DNA with cisplatin and oxaliplatin adducts in various sequence contexts on binding of HMGB1a

Anti-cancer drugs, such as cisplatin and oxaliplatin, covalently bind to adjacent guanine bases in DNA to form intra-strand adducts. Differential recognition of drug–DNA adducts by the protein HMGB1a has been related to the differences in efficacy of these drugs in tumours. Additionally, the bases flanking the adduct (the sequence context) also have a marked effect on HMGB1a binding affinity. We perform atomistic molecular dynamics simulations of DNA with cisplatin and oxaliplatin adducts in four sequence contexts (AGGC, CGGA, TGGA and TGGT) in the absence and presence of HMGB1a. The structure of HMGB1a-bound drug–DNA molecules is independent of sequence and drug identity, confirming that differential recognition cannot be explained by the protein-bound structure. The differences in the static and conformational dynamics of the drug–DNA structures in the absence of the protein explain some but not all trends in differential binding affinity of HMGB1a. Since the minor groove width and helical bend of all drug–DNA molecules in the unbound state are lower than the protein-bound state, HMGB1a must actively deform the DNA during binding. The thermodynamic pathway between the unbound and protein-bound states could be an additional factor in the binding affinity of HMGB1a for drug–DNA adducts in various sequence contexts.     

Differences in backbone dynamics of two homologous bacterial albumin-binding modules: implications for binding specificity and bacterial adaptation

Proteins G and PAB are bacterial albumin-binding proteins expressed at the surface of group C and G streptococci and Peptostreptococcus magnus, respectively. Repeated albumin-binding domains, known as GA modules, are found in both proteins. The third GA module of protein G from the group G streptococcal strain G148 (G148-GA3) and the second GA module of protein PAB from P.magnus strain ALB8 (ALB8-GA) exhibit 59% sequence identity and both fold to form three-helix bundle structures that are very stable against thermal denaturation. ALB8-GA binds human serum albumin with higher affinity than G148-GA3, but G148-GA3 shows substantially broader albumin-binding specificity than ALB8-GA. The (15)N nuclear magnetic resonance spin relaxation measurements reported here, show that the two GA modules exhibit mobility on the picosecond-nanosecond time scale in directly corresponding regions (loops and termini). Most residues in G148-GA3 were seen to be involved in conformational exchange processes on the microsecond-millisecond time scale, whereas for ALB8-GA such motions were only identified for the beginning of helix 2 and its preceding loop. Furthermore, and more importantly, hydrogen-deuterium exchange and saturation transfer experiments reveal large differences between the two GA modules with respect to motions on the second-hour time scale. The high degree of similarity between the two GA modules with respect to sequence, structure and stability, and the observed differences in dynamics, binding affinity and binding specificity to different albumins, suggest a distinct correlation between dynamics, binding affinity and binding specificity. Finally, it is noteworthy in this context that the module G148-GA3, which has broad albumin-binding specificity, is expressed by group C and G streptococci known to infect all mammalian species, whereas P.magnus with the ALB8-GA module has been isolated only from humans.