Figure 1. Secondary structure of the complex
between the leadzyme and the substrate RNA used in this study. The arrow
indicates the cleavage site.Combination of Metal Ions Makes an Improved Ribozyme System
Naoki Sugimoto and Tatsuo Ohmichi
Department of Chemistry,
Faculty of Science, Konan University,
8-9-1 Okamoto, Higashinada-ku, Kobe 658, Japan
A ribozyme is an RNA enzyme that can catalyze a biochemical reaction. Because a ribozyme requires a metal ions for a catalytic function, the ribozyme is considered as one of the metalloenzymes. Ribozyme reactions are very sensitive to species and concentration of metal ions, suggesting that divalent metal ions may play important roles for the chemical reaction and structural stabilization of the ribozyme.
We have investigated the combined effects of metal ions on both the structural stabilization of the complex between a ribozyme (a leadzyme), CUGGGAGUCC, and a substrate, GGACCGAGCCAG, and the cleavage step in the active center of the complex has been also investigated kinetically.
Figure 1. Secondary structure of the complex
between the leadzyme and the substrate RNA used in this study. The arrow
indicates the cleavage site.
The results showed that the addition of Nd3+ in the presence of Pb2+ increased significantly the yield of the RNA cleavage reaction by the leadzyme, although other rare earth ions or divalent ions except Pb2+ did not promote the reaction. Further, kinetics for the leadzyme reaction was measured at various concentration ratios of Nd3+ and Pb2+. At low concentration ratios of Nd3+ under a constant total concentration of metal ions, Nd3+ increased the stability of the complex between the leadzyme and the substrate. In contrast, at high concentration ratios of Nd3+, the addition of Nd3+ decreased the stability of the complex. The rate constant of the cleavage step was maximized when the ratio of Nd3+ to Pb2+ was 1:1. These results suggest that the complex between the leadzyme and the substrate has binding sites for Nd3+ ion that influence complex stability and catalyze directly the cleavage reaction. On the basis of the results, we propose a two-metal-ion mechanism in which Pb2+ and Nd3+ play the roles of base and acid catalyst, respectively.
This work was supported in part by grants from the Ministry of Education, Science and Culture, Japan.
[1] Sugimoto, N. and Ohmichi, T., FEBS Lett. 393,
97-100, 1996.
[2] Ohmichi, T. and Sugimoto, N., Biochemistry 36, in press, 1997.
Influence of Electrostatic Mutations on Conformation and Thermal Denaturation Parameters of Apo Calmodulin
I. Protasevich(1), B. Ranjbar(1), V.Lobachov(1), A.
Makarov(1), R. Gilli(2), C. Briand(2), D. Lafitte(3) and J. Haiech(3)
(1)Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Moscow 117984, Russia,
(2)URA CNRS
1924 Faculte de Pharmacie,
Marseille 13385
(3)Laboratoire de Chimie bacterienne,
UPR 7221 CNRS,
Marseille 13009, France
Scanning microcalorimetry and circular dichroism were used to study conformational state and heat denaturation of Ca2+-free synthetic calmodulin (SynCaM) and its three electrostatic mutants. We evidenced the major role of the electrostatic potential in the stability and flexibility of SynCaM. The substitution of 118DEE120 cluster by 118KKK120 (SynCaM12A) does not influence the flexibility of the protein, the replacement of 82EEE84 by 82KKK84 (SynCaM8) decreases its level, while the combination of these two mutations in SynCaM18A significantly increases the flexibility. The heat denaturation of apoSynCaM and its mutants is well approximated by two two-state transitions with the lower temperature transition corresponding to C-terminal lobe melting and the higher temperature one to N-terminal lobe melting. The difference in transition temperatures for the two lobes decreases in SynCaM8 and increases in SynCaM18A which suggests a modification in the influence of one lobe on the other. The electrostatic mutations influence the parameters of thermal denaturation of SynCaM lobes in a similar way as pH conditions affect thermal transitions parameters of domain proteins, leading to a linear temperature dependence of transition enthalpy. One domain of the N-terminal lobe in apoSynCaM18A is unfolded in the native state. Near CD spectra point out the invariability of the local structure of aromatic residues upon mutations, although secondary structure undergoes striking transformations. Cacodylate ions strongly and specifically alter the helical content of SynCaM. Our data unambiguously demonstrate that the two lobes are not independent and interactions between the lobes are mediated by the electrostatic potential of the molecule.
This work was supported by INTAS grant 94-2068.
Analysis of Structural Environment of Amino Acids That Alter in vitro Protein Stability
Boojala V B Reddy and Sunando Datta
Centre for Cellular & Molecular Biology,
Hyderabad -- 500 007, India
Phone: +91-40-672241; Fax:+671195; E-mail: bvbreddy@ccmb.globemail.com
Fairly good number of protein structures with substituted point mutations that alter in vitro stability is available in the data bank. The structural studies revealed that mutant proteins retain overall structures that are very similar to those of their wild-type parents. However, the variation in structural environment that occur at the site of the substitution such as hydrogen bonding, hydrophobic interactions, accessibility to solvent, etc. do alter in vitro stability of the mutant proteins. The objective of this analysis is to calculate propensity of each of the twenty amino acids in structural environment of parent residue proteins, and to study their correlation with the increase/decrease of in vitro stability of the mutant protein with the substituted residue. Such an analysis would enable us to develop a method to suggest substitutes for point mutations to engineer in vitro stability for a given protein of known structure.
We collected about 200 substituted single residue mutants from the literature for which parent protein structure and altered in vitro (thermal, solvent or pH induced) stability information are available. The structural environment of each of the parent residue is characterized into combinations of following variables: (i) secondary structural class (alpha-helix, beta-strand, coil), (ii) hydrogen bonding (M-M, M-SN, M-SO, S-S), (iii) solvent accessibility (>7%, 7%-20%, >20%) - (a) all atom accessibility, (b) polar side-chain, (c) non-polar side-chain, (d) all atom side-chain, (e) main-chain, and (iv) packing density (in-terms of Ooi number in 8A(ngstrom) and 14A(ngstrom) radius sphere). A data set of 303 non-homologous (< 30% sequence identity), well resolved >= 2.0A(ngstrom) protein structures is used to compute natural propensities ( pi ) of all the 20 amino acids for structural environment of each of the parent residue. A preliminary analysis of the study gives about 72% positive correlation showing the decrease in stability corresponding to a residue of relatively low propensity and the increase in stability corresponding to a residue with relatively high propensity for their respective structural environments. We are working towards improving the characterization of structural environment of residues that accounts most of the stabilizing interactions to get a positive correlation between pi and in viro stability near to 100%. The data and analyses supporting all these observations would be discussed.
The Regioselective Effect of Zwitterionic DNA Substitutions on DNA Alkylation
P. Dange(1), G. Liang(1), F.-X. Chen(2) and B. Gold(1)
(1)Department of Pharmaceutical Sciences,
(2)Eppley Institute for Research in Cancer
University of Nebraska Medical Center,
Omaha, NE 68198-6805, USA
The incorporation of zwitterionic residues (5-substituted w-aminoalkyl-2* deoxypyrimidines) into DNA has been reported to bend DNA as measured by aberrant gel mobility [Strauss et al., (1996) Proc. Natl. Acad. Sci. USA 93, 9515-9520]. Herein, we report that DNA methylation by N-methyl-N-nitrosourea at N7-guanine is regioselectively inhibited by point substitutions of zwitterionic residues: 5-(6 aminohexyl)-2*-deoxycytidine, 5-(6-aminohexyl)-2*-deoxyuridine, or 5-(3 aminopropyl)-2*-deoxyuridine. No inhibition is observed for DNA methylation by dimethyl sulfate. Based on the inhibition patterns for methylation with the different zwitterionic substitutions and the different length tethers, the w-aminoalkyl sidechains prefer to adopt a conformation that points them toward the 3*-base. Molecular modeling grid searches coupled with energy minimizations, and simulated annealing molecular dynamics studies indicate that unfavorable steric interactions with the 5* base and backbone, as well as stabilizing electrostatic interactions with electronegative atoms on the 3* -side, are responsible for the observed conformational preference. No evidence for association of the cationic sidechain with the phosphate backbone is observed. The observed bending of DNA induced by the tethered amines may arise from their localization in the major groove, and their effect on the concentration of diffusible cations and water structure.
Breather Mechanism of DNA TATA-Box Recognition
Sergey N. Volkov
Bogolyubov Institute for Theoretical Physics,
National Academy of Sciences,
252143, Kiev, Ukraine
To clarify the molecular mechanism of DNA recognition by proteins the conformational breather formation in DNA macromolecule is studied. The experimental data of Burley with co-authers and Singler with co-authers (Nature, 1993) allow us to suggest that the breather excitation may be the predecessor of the structural perturbation observed in DNA conformation under TATA-box recognition. The breather occurrance allows the macromolecule to move out the nucleic bases usually hidden in the DNA helix for its recognition by proteins.
Using the two-component model of macromolecule conformational mobility it was shown that nonpropagated breathed excitation may occur in B-DNA under transition of double helix fragment in the metastable A-form. The breather has a view of localized conformational excitation where the DNA structural elements oscillate with large displacement amplitudes. The form of conformational breather was found for the parameters of B-A conformational transition dynamics. Under breather formation the DNA base pairs oscillate around some centre line displaced in the double helix small groove. The amplitudes of nucleic base motions in breather excitation are much larger than the same for thermal conformational vibrations. The conformational breather has the frequency less than 60cm-1 and is getting soft due to increasing of structural elements motion amplitudes. The dynamics of breather excitations in DNA provides exposition of the nucleic bases atomic groups for its recognition by proteins. Such a process supposes also the appearance of the backbone kinks at the boundaries of the TATA-box section that is completely in accordance with the experimental data.
The proposed breather mechanism of TATA-box recognition allows to predict some definite effects for experimental test. The structural elements of the macromolecule may be moved out more in the helix small groove under increasing of the free energy difference between the DNA B-form and the metastable A-conformation. Besides the breather excitation may occur also in the large groove if the DNA backbone is fixed.
The Study of the Stability of Watson-Crick Nucleic Acids Base Pairs in Water and Dimethyl Sulfoxide: Computer Simulation by Monte Carlo Method
V. I. Danilov(1), N. V. Zheltovsky(1), O. N. Slyusarchuk(1)
and J. L. Alderfer(2)
(1)Institute of Molecular Biology and Genetics,
Kiev, Ukraine
(2)Roswell Park Cancer Institute,
Buffalo, NY USA
Despite some theoretical and experimental efforts the role of solvents in the stabilization of base complexes (first of all, coplanar pairs) is not clear yet. In this report the results of the extensive computer simulation of Watson-Crick base pairs stability in water and DMSO by means of Monte Carlo method are presented. In order to study the stability of nucleic acid base pairs in DMSO and water we carried out a computer simulation of the solvation of uracil (U), thymine (T), adenine (A), guanine (G), cytosine (C), and A-U, A-T, and G-C base pairs. The calculations were performed using Monte Carlo - Metropolis method in the canonical (T, V, N) ensemble. In each of the systems studied (base/base pair + solvent) the number of DMSO molecules was equal to 200 and the number of water molecules - to 400. The temperature was equal to 298 K. Semiempirical atom-atom potential functions for the calculation of the intermolecular interaction energy were used.
It follows from the obtained data that the values of A-U, A-T, and G-C base pairs formation energy in DMSO are negative and, hence, the formation of those base pairs in that solvent is energetically favorable. The analysis of the data for DMSO shows that in all of the base pairs considered the interaction between bases and DMSO-DMSO interaction are stabilizing and DMSO-base interaction is destabilizing. It is shown that DMSO stabilizes A-U base pair (-4.8 Kcal/mol) and A-T base pair (-7.2 Kcal/mol). For A-U base pair the solvent contributes noticeably to its stabilization and for A-T base pair it is commensurable with the contribution due to the interaction between bases. At the same time, DMSO destabilizes G-C base pair (3.8 Kcal/mol). That testifies that the only factor which can stabilize G-C base pair is the base-base interaction. Such a difference in the bahavior of G-C base pair, on one hand, and A-U and A-T base pairs, on the other hand, is conditioned by the fact that due to the high polarity of G and C molecules the energetical loss resulting from DMSO-base interaction at the formation of G-C base pair in DMSO cannot be compensated by the energetical gain resulting from DMSO-DMSO interaction. It should be emphasized that the formation of A-T base pair in DMSO is energetically more favorable then the formation of A-U base pair. Thus, the introduction of the methyl group noticeably influences the energetical stability of base pairs in DMSO.
The situation is quite different when water is considered. Data show that the energy values for the formation of A-U, A-T, and G-C base pairs in water are positive and, therefore, their formation in water is energetically unfavorable. The reason for this is that unlike DMSO, base-base interaction is destabilizing in all base pairs. Water-base interaction destabilizes all of the studied base pairs, like in DMSO. Therefore, the stabilization of base pairs is determined only by the interaction between bases. It follows from the data obtained that it is not only hydration effect but also hydrophobic effect that are destabilizing. Ultimately, this leads to the unfavorability of the formation of base pairs in water. Thus, at the present time the conclusion we based on the crude estimate of the influence of the solvent that the observed absence of base pairs in water despite their high stability in vacuo is due to the gain resulting from hydration effect and hydrophobic effect, nonassociated bases are considered, can be taken as proved.
The results of the simulation of nucleic bases and base pair solvation are compared with available experimental data which were obtained in DMSO for base derivatives and their nucleosides by means of NMR method.
Simulation of Base and "Model" Amino Acids Complexes in DMSO by Monte Carlo Method
V. I. Danilov(1), O. V. Mikhaleva(1), O. N. Slyusarchuk(1),
J. L. Alderfer(2) and V. I. Poltev(3)
(1)Institute of Molecular Biology and Genetics,
Kiev, Ukraine
(2)Roswell Park Cancer Institute,
Buffalo, NY USA
(3)Institute of Theoretical and Experimental Biophysics,
Pushchino, Russia
To study the molecular acts of protein-nucleic recognition the complexes of guanine (G) and cytosine (C) with acetic acid in neutral (AcOH) and deprotoned (AcO-) forms which are the model of the side chains of aspartic and glutamic acids were researched. The formation of such specific complexes in nonpolar solvents is a competitive process in respect to the formation of G-C base pairs. We carried out computer simulation of G, C, AcOH, CH+, AcO-, G-C, C-AcOH, CH+ - AcO-, G-AcO- solvation in DMSO using Monte Carlo method. The calculations were performed in cluster approximation at temperature 298 K. The number of DMSO molecules in each of the systems studied was equal to 200. The results show that solvation energy values for all of the systems are negative (in kcal/mol): -32.8 for G, -22.6 for C, -11.8 for AcOH, -102.6 for CH+, -63.2 for AcO-, -74.0 for G-C, -35.6 for C-AcOH, -169.6 for CH+ - AcO-,-112.8 for G-AcO-. The obtained values show that G-C base pair formation energy is -18.6 kcal/mol and, hence, its formation in DMSO is energetically favorable. The analysis of the energy values of various interactions in G-C base pair shows that DMSO-DMSO and base-base interactions are stabilizing and DMSO-base interaction is destabilizing. G-AcO- complex formation is also favorable (-16.8 kcal/mol). This fact shows that G-AcO- complex formation is comparable with the formation of G-C base pair in energetical favorability. In this case acetate anion can replace in G-C base pairs.
The formation of C-AcOH complex is far less favorable (-1.2 kcal/mol) than the formation of G-C base pair. However proton transfer from AcOH to C leads to the formation of CH+ - AcO- base pair which is far more favorable than all of the complexes studied. In this case acetate anion can replace G in G-C base pairs. At the same time the mentioned amino acids which have the carboxyl group in a neutral form cannot replace or G in G-C base pair due to their lower energetical favorability.
Therefore, the deprotoned form of carboxyl group of aspartic and glutamic acid which are often located in binding sites of protein and nucleic acid can effectively form bonds with G and C by means of hydrogen bonds. There was made a theoretical conclusion that under protein-nucleic recognition acetate anion can replace C and G in Watson-Crick G-C base pairs. Thus, the formation of G-AcO- and CH+ - AcO- complexes may be considered to be the first step in the mechanism of protein-nucleic recognition.
DNA Kinking as Imaged with a New High-Resolution AFM
Mensur Dlakic(1), Wenhai Han(2), Rodney E. Harrington(1)
and Stuart M. Lindsay(2)
(1)Department of Biochemistry
University of Nevada Reno,
Reno, NV 89557-0014
Tel: 702-784-4619; Fax: 702-784-1419; E-mail: mensi@scs.unr.edu
(2)Department of Physics and Astronomy
Arizona State University
Tempe, AZ 85287-1504
A novel Atomic Force Microscope (AFM) with a magnetically-oscillated tip has provided unprecedented resolution of small DNA fragments spontaneously adsorbed to mica and imaged in situ in the presence of divalent ions. Kinks (localized bends of an average angle of 78o) were observed in axially strained minicircles consisting of tandemly repeated d(A)5 and d(GGGCC[C]) sequences, the only sequences presently known to impart substantial static bending to DNA. The frequency of kinks in identical minicircles increased fourfold in the presence of Zn2+ (Figure B) compared with Mg2+ (Figure A). We show that periodicity is absent in the kink spacing by employing the spectral analysis on distances between successive kinks which deflect DNA in the same direction. Since the sequences used in this study are repetitive, we therefore conclude that the kinks are not localized to within a single preferred location. The studies demonstrate the power of the new AFM technology and the results have important implications regarding the effects of axial strain and ionic conditions on local DNA structure.
NMR Studies of the Anticancer Drug Pepleomycin and its Complexes with DNA
Janet Caceres-Cortes(1), Hiroshi Sugiyama(2), Kenji
Ikudome(2), Isao Saito(2) and Andrew H.-J. Wang(1)
(1)Department of Cell & Structural Biology,
University of Illinois at Urbana-Champaign,
Urbana, Illinois 61801
(2)Department of Synthetic Chemistry and Biological Chemistry,
Faculty of Engineering,
Kyoto University, Kyoto 606-01, Japan
Pepleomycin (PEP) is a metalloglycopeptide that has stronger anticancer activity and less pulmonary toxicity than bleomycin (BLM). PEP, like BLM, exerts its action by binding to and degrading DNA in the presence of oxygen and certain metals. Two cobalt-PEP species, HO2-Co(III)-PEP and its deglycosylated HO2-Co(III)-dPEP, have been studied by 2D-NMR and NOE refined structures were obtained. Two chiral conformers (form A or B) can exist for these molecules with either the b-aminoalanine
primary amine (A,NH2) or the mannose carbamoyl nitrogen (M,NH2) as the axial ligand. Analysis of our NOESY data shows convincingly that form A is the most probable conformer with the mannose carbamoyl M,NH2 and the b-aminoalanine primary amine A,NH2 as the axial ligands in CoPEP and CodPEP, respectively. The NOE crosspeaks resulting from the interactions between the N-terminus (i.e., the metal binding domain) and the C-terminus of CoPEP and CodPEP have similar patterns, suggesting that they both adopt compact structures with the bithiazole group folded back over the N-terminus [1]. We have treated the self-complementary oligonucleotide d(CGTACG)2 with one equivalent of CoPEP or CodPEP and carefully examined these complexes _ denoted CGTACG-CoPEP and CGTACG-CodPEP, respectively _ by one- and two- dimensional 1H NMR. Refined solution structures of CGTACG-CoPEP and CGTACG-CodPEP have been obtained by NOE restrained refinement procedures. Addition of CoPEP or CodPEP to d(CGTACG)2 disrupts the two-fold symmetry of the DNA duplex, consistent with binding of the drugs to the DNA duplex. The binding mode of both CoPEP and CodPEP with respect to the DNA duplex is defined by approximately 60 intermolecular NOEs. The NOE data is consistent with a partial intercalative binding mode for both DNA-drug complexes with the drugs' metal-binding domain and peptide linker bound in the minor groove of the DNA close to residues G8 and T9, and the drug's bithiazole tail intercalated between the base pairs A4·T9 and T3·A10. A structural comparison between CGTACG-CoPEP and CGTACG-CodPEP is made. The solution structures of the free drugs and DNA-drug complexes and their relationship with the biological activities will be presented [2].
References
[1] Caceres-Coertes, J., Sugiyama, H., Ikudome, K., Saito,
I. and Wang, A. H.-J., European J. Biochem., in press, 1997.
[2] Caceres-Coertes, J., Sugiyama, H., Ikudome, K., Saito, I. and Wang,
A. H.-J., submitted, 1997.
Probing G.U Mismatches in RNA Fragments Using 15N NMR
Xiaohu Zhang, Barbara L. Gaffney and Roger A. Jones
Department of Chemistry,
Rutgers - the State University of New Jersey,
Piscataway, NJ 08855
The guanine-uracil ( GU ) base pair is a phylogenetically highly conserved motif in many forms of RNA, including tRNA, rRNA, ribozymes and spliceosomes. Crick first suggested that G and U could pair with only limited distortion of the RNA backbone. This GU "wobble" geometry was later observed in the X-ray structure of yeast tRNAphe and appeared to cause only slight disruption of the helical regularity of the acceptor stem.
Recently, the biological role of the GU pair has been defined in more detail. In the Tetrahymena Group I self-splicing intron, the fuctional consequence of a GU pair in the splice site helix ( P1 ) was shown to be more than local helical distortion created by the wobble geometry. A high resolution NMR structure has been determined for a model of this Group I splice site helix ( P1 ), showing an altered backbone structure and stacking pattern.
15N NMR chemical shift changes of selectively labeled DNA fragments have provided model-independent insight into localized hydrogen bonding, protonation, hydration and protection from hydration. In view of the importance of the GU pairing in RNA structure and fuction, we begun a program focused on this pair, in increasingly complex systems. We will present our results about the isolated and tandem GU pair in different sequences as well as the GU pair in the unusually stable UNCG tetraloop.
A New High-Resolution Atomic Force Microscope for Imaging Biomolecules in Fluids
S.M. Lindsay(1), Wenhai Han(1) and Tianwei Jing(2)
(1)Department of Physics and Astronomy,
Arizona State University,
Tempe, AZ 85287-1504
(2)Molecular Imaging Corporation,
1208 E. Broadway, Tempe, AZ 85282
Tapping mode AFM [1] is used to image soft materials, such as biopolymers, because it does not sweep the sample off the surface, even when imaging in aqueous solution [2]. However, it otherwise offers little improvement in resolution over contact mode unless specially sharpened tips are used [2]. We believe that this is because the limitations of the acoustic method used to oscillate the tip. The tip itself is heavily damped in fluids, and background noise is introduced by the required high-level of the driving acoustic excitation (which also vibrates other parts of the microscope).
Figure 1: Showing how the AFM cantilever is
oscillated by application of a magnetic force.
We have devised a method for driving the tip directly while it is submerged in fluid [3]. The force-sensing cantilever is coated with a magnetic film, and a solenoid beneath the sample stage is used to force the tip into oscillation. This arrangement is illustrated in Figure 1. The signal to noise ratio of the microscope is greatly improved and the energy deposited into the sample by the tip can be reduced about a thousand times [4] because the microscope works at very small vibration amplitudes.
The consequence of this reduction in tip-sample interaction is that sharp asperities on the tip are not damaged as the sample is scanned. Very high resolution can be obtained with nominally blunt tips. We show below (Figure 2) images of 168 base-pair DNA microcircles made by ligation of intrinsically-bent sequence motifs (see the abstract by Dlakic and Harrington). The DNA was spontaneously adsorbed onto a mica surface in the presence of Zinc ions and imaged in-situ with no further sample preparation. The average width of images obtained this way was 35Å, while many of the molecules in the image below are substantially narrower than that. The gentle interaction between the tip and the sample has allowed us to change the solution almost at will, and we have observed interesting effects on DNA conformation as the electrolyte is changed [5].
Figure 2: 168 base-pair DNA microcircles spontaneously adsorbed to mica and imaged in situ.
We are grateful to Rodney Harrington and Mensur Dlakic for allowing us to use an image of the microcircles prepared by them. This work was supported by the NSF and Molecular Imaging Corporation.
References
[1] Q. Zhong, D. Innis, K. Kjoller and V.B. Eilings, Surf.
Sci. Lett. 290, L688 ,1993.
[2] P.K. Hansma, J.P. Cleveland, M. Radmacher, D.A. Walters, P.E. Hilner,
M. Benzanilla, M. Fritz, D. Vie, H.J. Hansma, C.B. Prater, J. Massie, L.
Fukunga, J. Gurley and V. Eilings, Appl. Phys. Lett. 64, 1738, 1994.
[3] S.M. Lindsay, Y. Lyubchenko, N.J. Tao, Y. Li, P.I. Oden, J.A. DeRose
and J. Pan, J. Vac. Sci. Technol. 11, 808, 1993.
[4] W. Han, S.M. Lindsay and T.W. Jing, Appl. Phys. Lett. in press,
1996.
[5] W. Han, M. Dlakic, R.E. Harrington and S.M. Lindsay, submitted, 1996.
Studies on the Interaction of Nonintercalating Drugs with Parallel-Stranded DNA: Differential Effects of Netropsin and Related Lexitropsins
Ingo Förtsch(1), Thomas M. Jovin(2), J. William
Lown(3) and Christoph Zimmer(1)
(1)Institute of Molecular Biology,
Friedrich-Schiller-University Jena,
Winzerlaer Str. 10, D-07747 Jena, Germany
(2)Department of Molecular Biology,
Max-Planck-Institute for Biophysical Chemistry,
Postfach 2841, D-37018 Göttingen, Germany
(3)Department of Chemistry,
University of Alberta,
Edmonton, Alberta, T6G2G2, Canada
The secondary structure of parallel-stranded DNA (ps-DNA), now well established [1], is held together by reverse Watson-Crick base pairing forming two equivalent grooves. Using CD, fluorescence and melting temperature measurements we have investigated the ability of the minor groove binders netropsin (Nt) and lexitropsins containing N-methylimidazole rings instead of methylpyrrole for interacting with ps-DNA. CD spectra demonstrated the formation of the DNA-drug complexes consisting of oligonucleotide duplexes with dAdT base pairs. Distinct differences between ps-DNA and the antiparallel form of B-DNA (aps-DNA) were observed. To identify the relative orientation of the two strands after formation of DNA-drug complexes the excimer fluorescence of 5'-terminally pyrene labeled molecules of the ps-form was followed upon addition of the drugs. The fluorescence emission spectra and their temperature dependence between 4°C and 30°C led to two conclusions: (i) In the presence of the specific ligand Nt, ps-DNA is converted to aps-DNA with mismatched base pairs. This ps- to aps-conversion is much less pronounced for the imidazole-containing analogs, which have a lower affinity for d(AT)-sequences. (ii) The more weakly binding imidazole-bearing netropsin-analogs may bind to ps-DNA at lower temperatures, possibly in the grooves. Comparative results of studies with a ps-duplex DNA consisting of dAdU base pairs will be discussed.
References
[1] Rippe, K. & Jovin, T.M., Methods Enzymol. 211, 199-220, 1992.
Modulation of the Stability of a DNA Triple Helix by Interaction with Minor Groove-Binding Ligands
Ingo Förtsch(1), Holger Schütz(2) and Christoph Zimmer(1)
(1)Department of Molecular Biology
and
(2)Department of Biophysical Chemistry,
Institute of Molecular Biology,
Friedrich-Schiller-University Jena,
Winzerlaer Straße 10, D-07745 Jena, Germany
The interaction of DNA triplex structures with nucleic acid-binding ligands is an essential aspect in controlling the affinity of the third strand for its target duplex. The minor groove binders netropsin [1], DAPI or Hoechst 33258 [2] destabilize the triplex motif dTn·dAnxdTn, whereas berenil may also stabilize the triplex but at low ionic conditions only [3]. We have investigated the modulating effect on the triplex dTn·dAnxdTn of a series of minor groove binders belonging to the group of bisquaternary ammonium heterocyclic drugs. Using UV melting and CD measurements our data show that these ligands influence the triplex stability in a different way. Like netropsin, SN-6999 destabilizes the triplex structure, but stabilizes the DNA duplex. In constrast to that SN-18071, which is unable to form hydrogen bonds effectively stabilizes the triplex and has a much weaker effect on the duplex. The binding constant of SN-18071 was found to be higher for the triplex than for the duplex. Comparative UV melting experiments were performed with three other ligands of this class of compounds: SN-6132, SN-16814 and NSC-101327 (kindly donated by Dr. Bruce Baguley, Auckland, New Zealand). For SN-6132 and NSC-101327 a triplex-stabiling effect was also observed, but at rather low ionic concentration only, whereas SN-16814 showed no effect at all. Remarkably, SN-18071 promotes triplex formation up to 500 mM Na+, an effect not previously observed for minor groove binders. A possible triplex binding mode of SN-18071 is proposed. The ability to stabilize or destabilize the DNA-triplex structure depends on the chemical structure of minor groove binders under comparable ionic conditions.
References
[1] Park, Y.-W. and Breslauer, K.J., Proc. Natl. Acad.Sci
USA 89, 6653-6657, 1992; Durand, M., Thuong, N.T. and Maurizot, J.C.,
J. Biol. Chem. 267, 24394- 24399, 1992.
[2] Kim, H.-K., Kim, J.-M., Kim, S.K., Rodger, A. and Norden, B., Biochemistry
35, 1187-1194, 1996.
[3] Pilch, D.S., Kirolos, M.A. and Breslauer, K.J., Biochemistry 34,
16107-16124, 1995.
DNA-Intercalator Interaction Measurements Using Surface Plasmon Resonance
G. Bischoff(1), W.-V. Meister(1), E. Birch-Hirschfeld(2)
R. Bischoff(3), C. Bohley(1) and S. Hoffman(1)
(1)Martin-Luther University Halle-Wittenberg,
Institute for Biochemistry,
Kurt-Mothes-Str. 3,
D-06120 Halle, Germany
(2)Klinikum der Universität Jena,
Institute for Virology,
Winzerlaerstr. 10,
D-07740 Jena, Germany
(3)Klinik für Hals-,
Nasen-, Ohrenkrankheiten, Gesichts- und Halschirugie
Magdeburger Str. 12,
D-06110 Halle, Germany
Intercalation of various molecules in DNA double-strands are widely used for DNA-analysis, cancer therapy or antiviral systems for example. With real time monitoring Surface Plasmon Resonance (SPR) the intercalation reaction was studied for the intercalators tilorone, 2,7-bis[(dipropylamino)-acetamido] fluoren-9-one, Bis-benzimidazol Hoechst 33258 and haematoporphyrin to synthetic DNA (36mer). A monolayer of biotinylated DNA was immobilized to a streptavidin-dextran-gold triple-layer. Portions of small concentrations of intercalating molecules of approximately 0.01 pmol/µl were injected in continuous flow. The increasing mass corresponds to the amount of the intercalated molecules. Pulses of 50 mM sodium hydroxide remove the intercalated molecules completely reversible out of the DNA. Following treatments with the intercalators gives reproducible results. As a virtual help to understand the reaction mechanisms we used computural simulations and compare them with the experimental datas.
Mobility and Diffusion of a Polymer Chain in a Dynamic Matrix
Udayan Mohanty
Eugene F. Merkert Chemistry Center,
Boston College,
Chestnut Hill, MA 02167 USA
We develop novel theoretical techniques to describe the electrophoretic mobility and the diffusion of polymer chains through a dynamic matrix like agarose gel. Our approach takes into account the lifetime of the conformations of the gel in the presence of an external electric field as well as the "entropic" interactions between the matrix and the probe. The approach leads naturally to a generalization of the Zimm-Lumpkin model of electrophoresis.[1] The model elucidates a series of experiments by Stellwagen and coworkers[2,3 ] who have found that the mobility of DNA fragments as well as the direction it migrates is effected by the orientation of the matrix. Our approach can be generalized to pulsed-field gel electrophoresis.
References
[1] Zimm, B. H., Lumpkin, O., Macromolecules 26,
226, 1993.
[2] Holmes, D. L., Stellwagen, N. C., J. Biomolecular Structure &
Dynamics, 7, 311, 1989.
[3] Stellwagen, J., Stellwagen, N. C., Biopolymers, 34, 187, 1994.
DNA Bending is an Essential Component for the Site Specific Recognition by DNA Binding Domain of Tumor Suppressor Protein p53
Akhilesh K. Nagaich(1), Ettore Appella(2) and Rodney
E. Harrington(1)
(1)Department of Biochemistry/330,
University of Nevada Reno,
Reno, NV 89557-0014 USA
(2)Laboratory of Cell Biology,
NCI, NIDDK,
National Institute of Health,
Bethesda, MD 20892 USA
We have used cyclic permutation assay to analyze the extent and location of the DNA bend induced by the DNA binding domain of human wild type p53 (p53DBD) upon binding to different naturally occurring DNA response elements. We have found that the response elements having CATG sequence at the junction of the two consensus pentamers specially favour highly bent complex (bending angle is ~ 50°) whereas response elements having CTTG bases at the pentamer junction are less bent (bending angle 37° to 25°). Our observations are consistent with the earlier studies and the proposed model for the DNA bending by p53DBD [1,2]. The electrophoretic mobility shift assay of the different complexes show an interesting correlation between the bending angle and the bending affinity of the p53DBD with the response elements; that is the greater the stability of the complex, the more the DNA appeared to be bent by p53DBD. This study provides the first evidence that the structure and the stability of the p53-DNA complex may very with different response elements and may be correlated to the transcription activation. In addition, we propose that the energetics of the bending contribute significantly to the overall binding affinity of the p53DBD for different sequences.
References
[1] Balagurumoorthy, P., Sakamoto, H. M. S., Zambrano,
N., Clore, G. M., Gronenborn, A. M., Appella, E. and Harrington, R. E.,
Proc. Natl. Acad. Sci., 92, 8591-8585, 1995.
[2] Nagaich, A. K., Zhurkin, V. B., Sakamoto, H., Gorin, A. A., Clore, G.
M., Gronenborn, A., M., Appella, E. and Harrington, R. E., (data unpublished).
Inherent DNA Duplex Curvature Determined from 1.5 ps Molecular Dynamics Simulations for a Series of TATA-promoter Mutants and Comparison with Experimental Physical and Biological Observations
Osmar Norberto de Souza and Rick L. Ornstein
Environmental Molecular Sciences Laboratory,*
Pacific Northwest National Laboratory,
Richland, WA 99352
We present 1.5 ns molecular dynamics simulation on the GCTATAAAAGGG double helix dodecamer bearing the Adenovirus major late promoter TATA-element and three iso-composition mutants for which physical and biological data are available from the same laboratory. Three out of four such DNA sequences experimentally induce tight binding with the TATA box binding protein (TBP) and induce high transcription rates. The fourth DNA sequence induces much lower TBP binding and transcription. Do differences of inherent DNA sequence-dependent structure and flexibility play a significant role in establishing these observed preferences? X-ray crystal structure structures have previously shown that the duplex DNA is highly bent in a DNA-TBP complex. Experimental structures for the isolated DNA sequences, noted above, are not available. We therefore decided to use molecular dynamics simulations to investigate the extent to which inherent DNA sequence-dependent curvature is consistent with observed trends. The computations used the AMBER 4.1 program and the 1995 forcefield, with the particle-mesh Ewald method. The structural similarities and differences amongst all four sequences correlated well with their observed specificity for the TATA box binding protein (TBP) and transcription rates. Our simulations are fully consistent with the premise that inherent DNA duplex sequence-dependent curvature plays a significant role in TBP binding, recognition, and transcriptional activation.
*Pacific Northwest National Laboratory is operated for
the U. S. Department of Energy by Battelle Memorial Institute under contract
DE-AC06-76RLO 1830.
DNA Stress Induced by the TATA-box Binding Protein
Anne LeBrun(1), Zippora Shakked(2) and Richard Lavery(1)
(1)Laboratoire de Biochimie Théorique,
URA 77 CNRS,
Institut de Biologie Physico-Chimique,
13 rue Pierre et Marie Curie,
Paris 75005, France
(2)Department of Structural Biology,
Weizmann Institute of Sciences,
Rehovet 76100, Israel
The recent determination of the three dimensional stucture of complexes between the TATA-box binding protein and DNA shows a conformation unlike that of most proetin-DNA complexes [1,2]. This involves strong DNA bending away from the protein, coupled with helical unwinding. Such effects, also seen with other proteins such as SRY, LEF-1 and PurR, cannot be explained by the canonical model of Mirzabekov and Rich, which implies DNA bending towards the protein as a result of partial phosphate neutralization.
Proposed explanations for the observed deformations have included kinking in response to partial amino acid side chain intercalation [1-4] and increased phosphate repulsion due to the presence of a low dielectric protein volume.
We present an alternative explanation based on the fact that the distortions induced in DNA by these proteins locally resemble the DNA conformations linked with nanomanipulation experiments involving extreme DNA stretching [6,7]. Molecular modeling demonstrates that local stretching and unwinding are closely coupled to one another and that such deformations also lead to kinks at the ends of the deformed domain, similar to those observed experimentally. Usually a single geometrical contstraint, we are able to closely reproduce the deformation of DNA induced by their TATA-box binding protein [8]. This study thus leads to a simple model for the mechanism ansd the energetics of the deformation, which also appears to be applicable to a number of other protein complexes.
References
[1] Kim, Y., Gieger, J. H., Hahn, S. and Sigler, P. B.,
Nature 365, 512-520, 1993.
[2] Kim, J. L., Nikolov, D. B. and Burley, S. K., Nature 365, 520-527,
1995.
[3] Kim, J. L. and Burley, S. K., Nature Struct. Biol. 1, 638-653,
1994.
[4] Werner, M. H., Gronenborn, A. M. and Clore, G. M., Science 271,
778-784, 1996
[5] Elcolck, A. H. and McCammon, J. A., J. Am. Chem. Soc. 118, 3787-3788,
1996.
[6] Cluzel, P., Lebrun, A., Heller, C., Lavery, R., Viovy J.-L., Chatenay,
D. and Caron, F., Science 271, 792-794, 1996.
[7] Lebrun, A. and Lavery, R., Nucleic Acids Res. 24, 2260-2270,
1996.
[8] Lebrun, A., Shakked, Z. and Lavery, R., Proc. Natl. Acad. Sci.,
in press.
Structural Studies on Minor Groove Binder: DNA Interactions
J. William Lown
Department of Chemistry,
University of Alberta,
Edmonton, AB, Canada T6G 2G2
Molecular recognition of B-DNA from the minor groove underlies the mode of a class of structurally diverse naturally occurring antibiotics, inducing distamycin A, netropsin, chromomycins, calcheamycins, CC-1065, pyrrolobenzodiazepines and mytomycins. Structural studies have provided insights into the modes of binding and have directly assisted drug design. An x-ray cyrstal structure has been solved of the complex of a monocationic oligopeptidic lexitropsin with B-DNA complex of sequence: CATGGCCATG. The lexitropsin is identical to netropsin except for replacement of the methylpyrrole rings by methyllimidazole.
Crystal Structure of the Complex of Im-Im+ Lexitropsin with the B-DNA Double Helix C-A-T-G-G-C-C-A-T-G
An analog of distamycin A, P4+, was found to bind in a 2:1 cooperative manner into the narrow groove of d(CGTATATACG)2 by 1D-NMR titration. The resulting 2:1 ligand-DNA complex was characterized by 2D NMR experiments including NOESY, COSY and TOCSY. Two molecules are oriented side by side opposite to each other; each has the N terminus to C terminus direction parallel to the 5' to 3' direction of its adjacent DNA strand. This study corroborates CD and ethidium fluormetry results and provides new support for the generality of the antiparallel side by side binding motif first observed with distamycin A. Covalent linkage of two lexitropsin molecules, each a pyrrole-pyrrole-imidazole tripeptide, is shown to provide a further improvement in the binding affinity and specificity for a designated complex sequence containing G·C base pairs.
Binding Analysis
Three binding modes are available: the 1:1 stoiciometry modes including the monodentate WPW and the bidentate WPPW modes and the 2:1 stoiciometry WPPW mode, as depicted.
Evidence for 2:2 sequence- and strand-specific binding of the dimer and structural characterization of its complex with d(CGAACATGTTCG)2 using NMR and restrained molecular/dynamic methods are discussed. These results suggest that selectivity of lexitropsins for DNA sequences may arise primarily from an extended array of specific intermolecular hydrogen bonds at the floor of the minor groove in DNA.
Conformational Flexibility of Voltage Gated Dihydropyridine Sensitive Calcium Channel: Molecular Dynamics Simulation
V. Kothekar and D. Gupta
Department of Biophysics,
All India Institute of Medical Sciences,
Ansari Nagar,
New Delhi- 110029, India
1-4 dihydropyridine (DHP) derivatives modulate calcium ion flux through voltage gated slow L type calcium channel. The inner lining of the pore is formed by four a helices from S3 segment of four structural domains in a1 subunit [1]. F1013, E1014 and E1323 in extra cellular stem region between S5-S6 of domains III and IV-are important for its activity [2]. These are involved in allosteric coupling of Ca2+ and DHP, and interaction with Ca2+.
We built a model of Ca2+ channel using aminoacid sequence DPWNVFDFLIVIGSIIDVILSE from S3 helix of the fourth internal repeat conserved between skeletal, cardiac, brain and aorta. The synthetic peptide was found to mimic ion conductance, selectivity and pharmacological activity [3] The model was built on the basis of structural data on four helix bundle [4] but having same orientation of all the helices. We also created a 16 residue loopVSTFEGWPQLLYRAID (res 1010-1025) with hairpin bend between F4-W7 and incorporated DHP antagonist Nifedipine such that it could simultaneously interact with F4 by hydrophobic interaction and E5 by salt bridge through Ca2+. The N and C terminus of all five polypeptides were blocked by CH3.C=O and NH-CH3 to avoid art effects in the simulation. Since we did not know the exact location of the loop, it was positioned perpendicular to the pore axis at the carboxylic end so as to direct assembling of the four helix bundle. Several residues in the loop (S2, Q9, R13) could interact with it. We introduced two more Ca2+ ions inside the pore at the level of D7 and D17 to neutralise its negative charge. The peptides were treated in all atom model. The carbon atoms of the lipid molecule were treated in united atom model. The structure of the model was optimized in vacuum (model-I), aqueous environment with 1589 water molecules in TIP3P model (model-II), and in bilayer of 36 DMPC molecules with 1201 water (model-III) molecules using molecular mechanics (MM) and molecular dynamics (MD) approach, with AMBER 4.0 package in 471, 809 and 950 cycles with energy -1827, -4466 and -5667 Kcal/mole. For MD simulation we used residue based cut-off distance of 12.5 Å. 0.001 pico seconds (ps) time step, and updated non bonded pair list after every 20 cycles. Bond length and bond angles were constrained by SHAKE. The system was coupled to temperature bath at 300 K with coupling constant 0.4 ps. Vacuum simulation was done without any temperature or pressure monitoring. Heating in aqueous and membranous environment was carried out under constant volume conditions. Equilibration was done at constant pressure conditions. Model-II in aqeous environment showed drifting and bending of the helices due to increase in the pore water. Finally after 120ps the helix structure became unstable due to water penetration. Models I and III were fairly stable. These were simulated for 200 ps after equilibration. For these, various structure dependent parameters were calculated from subaveraged coordinates. Thus for the DMPC bilayer we calculated its thickness, overall surface area, torsional angles and water penetration profile. For peptides we calculated secondary structure, backbone and side chain flexibility, interhelix distances, pore geometry and water movement. The main observation is both in vacuum and in lipid bilayer the helix conformation is stable except for E22 in helix 2 and 4 where it was in coil. Coil conformation was also noted for L20 S21. The interhelix distances across the channel at N terminus and within the pore remained within 10 to 14 Å and 12 to 187Å. The Ca and side chain atoms in the loop region showed larger flexibility compared to pore. The water molecules were found to be transported through the pore. Implications of these results in calcium channel activity will be discussed.
References
[1] Tanabe, T., Takeshima, H., Mikami, A., Flockerzi, V.,
Takahashi, H., Kangawa, K., Kojima, M., Matsuo, H., Hirose, T. and Numa,
S., Nature 328, 313-318, 1987.
[2} Peterson, B. Z. and Catterall, W. A., J. Biol. Chem. 270, 18201-18204,
1995.
[3] Grove, A., Tomich, J. M. and Montal, M., Proc. Natl. Acad. Sci. 88,
6418-6422, 1991.
[4] Schafmeister, C. E., Miercke, L. J. W. and Stroud, R. M., Science
262, 734-738, 1993.
Intercalation of Ethidium Bromide into Short DNA Hairpins: Effect of Hairpin-Hairpin Dimerization
Olga F. Borisova and Anna K. Shchyolkina
Engelhardt Institute of Molecular Biology RASc,
Vavilova 32, 117984 Moscow, Russia
The four-stranded structures are known to be formed in the course of the first stage of the recombination of two homological DNA molecules. To study the hairpin-hairpin assotiation, the deoxyoligonucleotides 5'-oligo1-L-oligo2-3' were synthesized (L - the flexible linker -pO(CH2CH2O)3p-). Their primary structure determines parallel or antiparallel orientation of complementary strands [1]. Three fluorescence methods were used: 1) the measurement of the fluorescence polarization for calculation of rotational relaxation times (r) and hydrodynamic volumes of the structures formed; 2) the measurement of the heterogeneous fluorescence lifetime of acridine orange for determination of fraction of unpaired nucleotides; 3) the binding isotherms of EtBr.The oligonucleotides: 5'-d(T)10-L-d(A)10-3' (antiA/T), 5'-d(GC)5-L-d(GC)5 -3' (antiGC), 5'-d(AT)5-L-d(AT)5-3' (antiAT), 5'-d(AT)5-L-d(AT)5-5' (parAT) and 5'-CAGAATGTGGCAAAGG-L-CCTTTGCCACATTCTG-3' (Znf) were investigated. The experiments performed in 0.01 M Na-phosphate buffer, pH7, in the presence of Na+ or Li+ and 0.5 - 1 mM of oligonucleotides, revealed the formation of intramolecular hairpins with anti- and parallel oriented complementary strands: antiGC, antiAT and parAT. In the presence of Mg2+ ions the oligonucleotides antiA/T, antiGC and Znf, but not the alternating parAT and antiAT were shown to form dimers. This dimerization can be monitored by the increase of rotational relaxation time r of the oligonucleotides antiA/T and antiGC and Znf. The temperature dependences of r allow to conclude that four strands are rigidly bound in the dimers. A cooperative mode of EtBr intercalation into these dimers was observed. The stoichiometry of EtBr-oligonucleotide complexes was determined and the EtBr binding constants were calculated. Molecular models of the DNA dimers are discussed on the basis of the four-stranded structures, which were predicted earlier [2,3].
References
[1] Borisova O.F. et al., J. Biomol. Struct. & Dynam.
8, 1187-1210, 1991.
[2] Chernyi A.A. et al., J. Biomol. Struct. & Dynam. 8, 513-527,
1990.
[3] Lebrun A. & Lavery R., J. Biomol. Struct. & Dynam. 13,
459-464, 1995.
Supported by the Russian Foundation for Fundamental Research (Grant No. 96-04-70503).
A Double Stranded RNA Binding
Protein
A.G. Gabrielian, K.A. Bakuntz, R.A. Zakharian
Institute of Molecular Biology,
Nation. Acad. Sci., Yerevan, Armenia
The translocation of ds-RNA polyelectrolyte molecule through the hydrophobic lipid-protein layer of the cellular plasmatic membrane is highly specific. Earlier we have found out, that the binding factor of ds-RNA on the plasmatic membranes is represented by protein: ds-RNA-binding ability completely disappeared after treatment of the surface of the cells by tripsin. It is natural, as the peculiarity of membrane is conditioned just by its protein component. Neither DNA-ase , RNA-ase, nor DNA and oligonucleotides influenced on ds-RNA - plasmatic membrane specific binding process.
Both the known and the elaborated by us techniques (including chromatographic and electrophoretic approaches) were used for isolation, purification and solubilization of 60 kDa ds-RNA-binding protein from a number of ds-RNA-binding proteins of the plasmatic membrane of rat brain cortex cells. This protein was obtained suitable for the physical-chemical, including UV-spectral, investigations in a water environment. Some properties of this protein were studied. Combined ds-RNA preparation of virus-like particles of killer yeast S.cerevisiae (M437) was separated into the fractions L and M with 4 500 and 1 800 b.p. respectively.
Then the study of interaction of the 60 kDa-protein with Lds-RNA was carried out in details. The CD-spectrum nonadditive alteration indicates the complex formation. The UV-high-resolution melting experiments showed, that this protein preferably binds to ds-RNA double-helical regions compared with single-stranded ones. The ds-RNA melting curve in the presence of the protein was shifted to the right without a noticeable change of the melting interval.
There was a strong deformation of differential melting curve. It was also observed a considerable degree of reassociation of Lds-RNA structure after the recooling of denatured Lds-RNA-protein complexes solution. By the way the renaturation practically is absent in ds-RNA solutions without the protein. All the physical-chemical data of Lds-RNA conformational fluctuations under the pressure of 60 kDa-protein-ligand molecules were considered in terms of the modern theoretical ideas. Such treatment allowed to elucidate the questions of protein molecules distribution along ds-RNA macromolecules, the character of ds-RNA conformational reorganization at complex formation, the nature of interaction forces, as well as conclude about the biological role of this protein. The results showed that this 60 kDa protein has all the properties of ds-RNA receptor:
1. it has high selectivity and specificity for the recognized
ds-RNA molecules;
2. the affinity parameters are not high and are corresponding to physiological
concentrations of the bound molecules;
3. the reception protein binding to the recognized molecules is reversible
so that the physiological effect can be reversed after withdrawal of the
molecules.
Prediction of Bendability
and Curvature in Genomic DNA
Andrei Gabrielian*, Kristian Vlahovicek and Sandor Pongor
International Centre for Genetic Engineering and Biotechnology (ICGEB),
Area Science Park, Padriciano 99,
34012 Trieste, Italy
E-mail :pongor@icgeb.trieste.it, kristian@icgeb.trieste.it
*Present address:
National Center for Biotechnology Information,
National Library of Medicine, NIH.,
Bldg 38A, Rm 8N805,
Bethesda, MD, 20894. USA
E-mail:gabrieli@ncbi.nlm.nih.gov
The dynamic ability of DNA molecule to bend is thought to be a crucial factor in processes such as gene regulation, packaging and DNA replication. DNA bendability can be estimated by using a set of parameters derived from deoxy-ribonuclease I cleavage experiments [1]. Bendability parameters can be used to construct realistic anisotropic elastic models of DNA [2]. Distribution of bendable and rigid segments in DNA sequences apparently correlates with curvature and superhelicity of the molecule [3,4]. Bendability profiles and Fourier transform methods were used to study the distribution of bendable, rigid and curved segments in genomic sequences, including the H.influenzae, M.genitalium, jannaschii, S. cerevisiae genome, as well as a number of mitochondrial and viral genomes. Characteristic patterns of bendability distributions were observed in prokaryotic vs. eukaryotic as well as coding vs. non-coding regions of genomes. The possibility that DNA may contain localized conformational signals, in addition to the information stored in the base sequence, was first raised more than 15 years ago [5]. It appears that the dynamic bendability/static curvature characteristics of DNA meet the basic criteria of conservation and uniqueness expected from such conformational signals. For many functionally important sites we observed specific patterns of DNA bendability which may provide new possibilities for genome studies.
[1] Brukner, I., Sanchez, R., Suck, D. and Pongor, S.,
Sequence-Dependent Bending Propensity of DNA as Revealed by DNase I: Parameters
for Trinucleotides. EMBO J., 14, 1812-1818, 1995.
[2] Gromiha, M., Munteanu, M. G., Gabrielian, A.and Pongor, S., Anisotropic
Elastic Bending Models of DNA, J. Biol. Phys.22, 227-243 , 1996.
[3] Gabrielian, A, Simoncsits, A. and Pongor, S., Distribution of Bending
Propensity in DNA Sequences, FEBS Letters,393, 124-130, 1996.
[4] Gabrielian, A. and Pongor, S., Correlation of Intrinsic DNA Curvature
with DNA Property Periodicity, FEBS Letters, 393, 65-68, 1996.
[5] Trifonov, E. N. and Sussman, J. L., The Pitch of Chromatin DNA is Reflected
in its Nucleotide Sequence, Proc. Natnl. Acad. Sci. USA ,77, 3816-3820,
1980.
Conformational NMR-Analysis of the Principal Neutralizing Determinant of the HIV-1 Envelope Protein gp120
Alexander M. Andrianov and Afanasii A. Akhrem
Institute of Bioorganic Chemistry,
Belarus Academy of Sciences,
Zodinskaya 5/2 Minsk-141,
220141 Republic of Belarus
A number of HIV-1 envelope reagents have been tested so far as immunogens to elicit neutralizing antibodies. Several lines of evidence suggest that the principal neutralizing determinant (PND) is located within a disulfide-linked loop in the V3 region of the envelope protein gp120.
The model of spatial structure for the HIV-1 PND is proposed in the present study in terms of two-dimensional nuclear Overhauser effect spectroscopy data [1]. To build the model, theoretical conformational NMR-analysis of synthetic peptides rp70, rp142 and rp342, representing the peptide copies of the PND fragments, is executed. Primary structures of these peptides, including 40 (peptide rp70), 24 (rp142) and 12 (rp342) amino acid residues, contain the hexapeptide Gly-Pro-Gly-Arg-Ala-Phe that lies in the central region of PND and forms the most probable immunodominant epitope of the HIV-1 protein gp120 [2]. The modeling of the molecular spatial structures is carried out by a new approach to research of conformationally flexible peptides based on the algorithms of the restrained molecular mechanics method developed earlier [3]. As a result of the analysis of the conformations obtained, the elements of secondary structure and probable systems of hydrogen bonds of studied peptides have been determined, the conformationally rigid and flexible segments have been identified and spatial backbone forms that are characteristic for these molecules have been described. Comparison of spatial structures of the fragment Gly-Pro-Gly-Arg-Ala-Phe, which are realized in the preferable conformations of peptides rp70, rp142 and rp342, shows that the constancy of its amino acid sequence is preserved on the level of spatial segment structure.
The following major conclusions have been made based on the analysis of the simulated peptide conformations: i) there is not unique PND structure in solution, ii) there are seven different PND structures each of which agrees with the experimental data and stereochemical criteria used in computing its spatial model, iii) the PND is characterized by irregular conformation containing a number of reverse turns, iv) all of the selected conformations have similar spatial structure of the conservative hexapeptide Gly- Pro-Gly-Arg-Ala-Phe. These data allow to suppose that binding properties of this site are determined by the structural motif which appears common for all hexapeptide conformations.
The proposed PND structural model is consistent with the reported NMR spectroscopy data [1] that were utilized in our calculations.
References
[1] K. Chandrasekhar, A.T. Profy and H.J. Dyson, Biochemistry
30, 9187-9194, 1991.
[2] K Javaherian, A.J. Langlois, G.J. LaRosa, A.T. Profy, D.P. Bolognesi,
W.C. Herlihy, S.D. Putney and T.J. Matthews, Science 250, 1590-1593,
1990.
[3] S.A. Sherman, A.M. Andrianov and A.A. Akhrem, J. Biomol. Struct.
Dynam. 5, 785-801, 1988.
Spatial Structure Model of the Fragment Ala-Ser-Thr-Thr-Thr-Asn-Tyr-Thr Responsible for Binding the HIV Envelop Prrotein gp120 to T4 Receptor of T4 Cells
Alexander M. Andrianov and Afanasii A. Akhrem
Institute of Bioorganic Chemistry,
Belarus Academy of Sciences,
Zodinskaya 5/2 Minsk-141
220141 Republic of Belarus
Of late, diverse inhibitors of duplication of human immunodeficiency virus (HIV), on the basis of which creating selective preparations for preventive maintenance and treatment of AIDS may be possible, are extensively studied. One of approaches to developing effective HIV inhibitors is based on oppression of virus adsorption on the surface of T4 cells and assumes the use, as antiviral preparations, of synthetic peptides with amino acid sequences corresponding to the conservative regions of HIV proteins. At present, peptide T [1], a synthetic molecule, amino acid sequence of which is identical to the primary structure of the site Ala-Ser-Thr-Thr-Thr-Asn-Tyr-Thr of protein gp120, is widest known among the HIV inhibitors participating in blocking viral attachment to the T4 receptor.
In this study we have undertaken attempt to predict three-dimensional structure of this site of protein gp120, ensuring the interaction of virus with T4 lymphocytes, based on the analysis of low-energy peptide T conformers, executed by making allowance for data on the spatial structures of similar octapeptide segments of unrelated proteins. To do this, the following researches have been carried out: i) the spatial structure models of peptide T and similar fragment 4-11 of an analogues of vasoactive intestinal peptide (VIP (4-11)) have been constructed by the restrained molecular mechanics method developed earlier [2] (to build the models, two-dimensional NMR spectroscopy data were taken from the literature); ii) conformational parameters of these models have been compared to geometrical characteristics of similar segments of unrelated proteins with known spatial structures. The result of the presented study for peptide T was a set consisting of six types of low-energy structures with different spatial packing of the peptide main chain. All structural types have been shown to be characterized by the lack of strict determination of the side chain conformations of the amino acid residues that can be realized in a few states providing approximately equal (within the given type) stabilization of one main chain form. At the same time, despite the definite differences, all of the selected structures were characterized by the presence of two consecutive reverse polypeptide chain turns at the C-terminal pentapeptide fragment. >From the analysis of the preferable VIP (4-11) conformations it was established that backbone of this fragment, that competes in isolated state with HIV and peptide T for the binding to the T4 receptor [3], forms two helical sites at residues 4-8 (distorted helix) and 8-11 (well defined regular helix). Comparison of three-dimensional structures of peptide T and VIP (4-11) among themselves and with X-ray crystal conformations of similar fragments of Rnase A (residues 19-26) and penicillopepsin (residues 169-176) has been performed by means of superimposition of nonhydrogen atoms of their main chains with subsequent calculation of root-mean-square deviations between atomic coordinates. The following major conclusions have been made based on the comparative analysis of the simulated and X-ray structures: i) the conformation of C- terminal fragment Thr-Thr-Asn-Tyr-Thr of peptide T, responsible for the biological activity of the molecule (3), does not undergo the essential distortions while embedding into the peptide chains of unrelated proteins; ii) this conformation, that is realized in isolated molecule and includes two consecutive reverse turns of the polypeptide chain, adequately describes the main conformational features of an appropriate site of the HIV protein gp120; iii) the fragment Ala-Ser-Thr-Thr-Thr-Asn-Tyr-Thr of protein gp120 accepts one of six spatial forms which are characteristic for peptide T.
References
[1] C.B. Pert, J.M. Hill, M.R. Ruff, R.M. Berman, W.G.
Robey, L.O. Arthur, F.M. Ruscetti and W.L. Farrar, Proc. Natl. Acad.
Sci. USA 83, 9254-9258, 1986.
[2] S.A. Sherman, A.M. Andrianov and A.A. Akhrem, J. Biomol. Struct.
Dynam. 5, 785-801, 1988.
[3] M.R. Ruff, B.M. Martin, E.I. Ginns, W.L. Farrar and C.B. Pert, FEBS
Lett. 211, 17-22, 1987.
NMR Structure of
a Tetrad Forming Oligonucleotide HIV Inhibitor
Naijie Jing, Xiaolian Gao(2) and Michael E. Hogan(1)
(1)Department of Molecular Physiology and Biophysics,
Baylor College of Medicine,
Houston, TX 77030
(2)Department of Chemistry,
University of Houston,
Houston, TX 77204-5641
Previously, we have developed therapeutic drug candidates for AIDS and AIDS related disease, based upon the use of G-quartet forming oligonucleotides, 5'-d(GGGTGGGTGGGTGGGT) (T30695) and 5'-d(GTGGTGGGTGGGTGGGT) (T30177), which are potent inhibitors of HIV-1 integrase in vitro (Rando et al., (1995) J. Biol. Chem. 270:1754; Jing et al., (1996) Biochemistry (submitted)). These tetrad folds are requirement for activity, leading to the most potent inhibitor of HIV-1 integrase to have been identified thus for (Mazumder et al., (1996) Biochemistry 35:13762). Spectroscopic, thermal denaturation and kinetic studies have revealed that DNA oligonucleotides, T30695 and T30177, form extremely stable intramolecular G-tetrads via a two-step process that involves the binding of one K+ ion to a central pair of G-quartets and two additional K+ ions, presumably, to the loops (Jing et al., (1996) Biochemistry (submitted)). In that these oligonucleotides are potent HIV-1 inhibitors and among the most active HIV-1 integrase inhibitors yet identified, we have sought to further characterize the K+-induced folding process for the purpose of rational chemical modification of these anti-HIV agents. Our NMR investigation demonstrates that in the presence of Li+ ions, T30695 forms a unimolecular tetrad fold structure stabilized by a pair of syn-anti-syn-anti G-quartets comprising a central core. The NMR spectrum of T30695 as a function of K+ titration reveals a well-defined transition that saturates upon addition of three K+ ions per oligomer. During this process, the initial Li+-dependent G-quartet structure converts into a highly symmetrical, stable form (the NMR detected melting transition temperature is increased by ~ 20 C). The conformation of the G-quartet core remains unchanged, while the loosely structured loop residues become organized in a fashion which is stabilized by K+ ion binding and by interactions with the core. Based upon our NMR data, the molecular structure of T30695 in the absence and presence of K+ ions were obtained and refined by introducing the restraints, including chiral (48), distance (70), and NOE distance (326) restraints. Wherein K+ binding to the loops, the loops of the structural of T30695 are rearranged to yield a planar array of loop bases, in proximity to the underlying G-quartets. By reference to closely related homologues, which lack activity as an HIV-1 or integrase inhibitor, the possibility is discussed that this flat, symmetrical, rigid, ion-coordinated loop structure is crucial to the biological activity of T30695.
Ab Initio Quantum Mechanics Analysis of C-H...O/N Hydrogen Bonding and Implications for Modeling Enzymes and Nucleic Acids
Rick L. Ornstein
Environmental Molecular Sciences Laboratory*,
Pacific Northwest National Laboratory,
Richland, WA 99352
While it is well established that classical hydrogen bonds play an important role in enzyme and nucleic acid structure, function and dynamics, the role of weaker, but 'activated' C-H donor hydrogen bonds is poorly understood. The most important such case for enzymes involves histidine which often plays a direct role in catalysis and possesses the most acidic C-H donor group of the standard amino acids. The histidine imidazole side chain moiety is also a component fragment of purine nucleic acid bases. In the present study, we obtained optimized geometries and hydrogen bond interaction energies involving the imidazole N-H and C-H donors with various acceptors at the MP2-FC/6-31++G(2d,2p) and MP2-FC/aug-cc-pVDZ//MP2-FC/6-31++G(2d,2p) levels of theory. We compare these hydrogen bonded interactions with those involving numerous other small molecules containing classical and 'weak' donors. A strong linear relationship is obtained between the stability of the various hydrogen bonded complexes and both separation distances for H...acceptor and donor----acceptor. In general, these calculations indicate that C-H donor containing interactions can be classified as hydrogen bonding interactions, albeit significantly weaker than the classical hydrogen bonds, but significantly stronger than just van der Waals interactions. For instance, while the electronic energy of stabilization at the MP2-FC/aug-cc-pVDZ//MP2-FC/6-31++G(2d,2p) level of theory of a water O-H...O water hydrogen bond is 4.36 kcal/mol more stable than the methane C-H...O water interaction, the water-water hydrogen bond is only 2.06 kcal/mol more stable than the imidazole Ce-H...O water hydrogen bond. Neglecting this latter hydrogen bonding interaction is obviously unacceptable. We also compare our ab initio results with trends from some popular molecular mechanics forcefields.
*Pacific Northwest National Laboratory is operated for the U.S. Department of Energy by Battelle Memorial Institute under contract DE-AC06-76RLO 1830.
DNA Separation by Polymer Micelle Capillary Electrophoresis
Chunhung Wu and Benjamin Chu*
Department of Chemistry,
State University of New York at Stony Brook,
Stony Brook, NY 11794-3408
Amphiphilic block copolymers of ethylene oxide, propylene oxide, and butylene oxide can self-assemble to form micelles with different associated structures. At appropriate high concentrations, polymer networks with DNA sieving ability are formed in various micellar associated structures including closed-packed core-shell micelles, flower-like micelles, and bridged-micelle clusters. The individual solutions or the mixtures of block copolymers EmPnEm, EmBnEm, BnEmBn and EmBn, where E, P, and B represent poly(ethylene oxide), poly(propylene oxide), and poly(butylene oxide) respectively, are used as DNA separation media by capillary electrophoresis(CE). Good ds and ss DNA electrophoretic separations can be achieved in the polymer networks with closed or open micellar associated structures. The application of these block copolymers in CE can also provide the advantages of easy injection and replacement of the block copolymer solutions as separation media, and convenient capillary column treatment, because the polymer itself can serve as a dynamic coating material and effectively suppress the electroosmotic flow. Filled with the block copolymer, short capillary with detection length of only 4 cm, which can provide the potential for miniaturization in a CE setup, can achieve good DNA electrophoretic separation in the DNA size range of 72 bp to 1353 bp.
Periodical Recurrence of Translation Pause Sites in mRNA and Standard Sizes of Protein Sequence Segments and Independently Folding Domains
C. Makhoul and E. N. Trifonov
Department of Structural Biology,
The Weizmann Institute of Science,
Rehovot 76100, Israel
It is believed that one of the main purposes of the pausing during mRNA translation is to ensure proper folding of the newly synthesized sections of the protein chain. The pausing occurs when rare triplets are encountered in the mRNA, so that it takes some time before the corresponding rare variant of tRNA is delivered. In an earlier work [1, 2] it has been shown that protein sequences are organized as consequtively joined segments of a standard size, about 120 residues in eukaryotes and 150 residues in prokaryotes. More detailed analysis indicates that the prokaryotic sequences may also consist of halfunits of about 75 aa residues [3]. The same segmentation in halfunits of about 210 bp is also typical of prokaryotic mobile DNA elements [4]. The size (210 bp, 70 triplets) is also consistent with the typical sizes of independently folding protein domains, 50 to 90 residues [5]. If, indeed, the translation pausing serves the proper folding of the individual domains then one would expect that the translation pause sites in mRNA would be typically located at the about 70 triplet distance from one another, or at the about 150 triplet distance to reflect the segmented organization of the protein sequences. To generate positional distributions of the pause sites along prokaryotic mRNA (cDNA) we have randomly chosen 500 prokaryotic cDNA sequences, with no detectable similarity to one another, from 5 different bacteria. From the individual profiles of local codon frequencies calculated with various windows, the coordinates of the clusters of the rarest codons were taken for generation of the overall histograms of positional preferences of the pause sites. The final histograms show, as expected, that the pause sites are located preferentially at the positions about 70 to 75 codons from one another, making a periodical pattern with the period 72 codons (216 bases), which closely corresponds to a half of a unit size, 420 - 450 bp, as estimated earlier. The profiles of local codon frequencies in mRNA can be used for detection (prediction) of the boundaries between the independently folding domains of the translated protein, when the crystal structure of the protein is not available. The pause sites revealed by the profiles also indicate locations of potential point mutations influencing protein folding rather than protein function per se.
References
[1] Berman, A. L., Kolker, E. and Trifonov, E. N., Proc.
Natl. Acad. Sci. USA 91, 4044-4047, 1994.
[2] Kolker, E. and Trifonov, E. N., In "Biological Structure and Dynamics",
pp. 257-266, Eds. R. H. Sarma and M. H. Sarma, Adenine Press, 1996.
[3] Kolker, E. Ph. D. Thesis, The Feinberg Graduate School, The Weizmann
Institute of Science, Rehovot, 1997.
[4] Trifonov, E. N., J. Biomolec. Str. Dyn., in press.
[5] Hubbard, S. J. and Argos, P., J. Molec. Biol. 261, 289-300, 1996.
Structural Variability of the A-DNA in Crystals of CCCGCGGG
Lucy Malinina(1), Luzimar Gonzaga-Fernandes(2), Dmitry
Pyshnyi(3), Nuria Verdaguer(2), Lourdes Campos(2) and Juan A.Subirana(2)
(1)Engelhardt Institute of Molecular Biology RAS,
Vavilov Str. 32,
117984 Moscow, Russia
(2)Departament d'Enginyeria Quimica,
Universitat Politecnica de Catalunya,
Avenida Diagonal 647,
E-08028 Barcelona, Spain
(3)Novosibirsk Institut of Biorganic Chemistry SB RAS,
Lavrentiev Prospect 8,
630090 Novosibirsk, Russia
We have determined the single crystal x-ray structure of the pCpCpCpGpCpGpG deoxyoligonucleotide in five different crystal forms with resolutions of 2.5A, 1.9A, 1.6A, 1.9A, 1.6A. In all these crystals, the octamer forms duplexes of A-DNA and all crystals display a similar packing mode, typical for the A-DNA. The structure of the duplex varies from loose to very compact when going from one crystal form to another. The most compact form exhibits the Vol./bp=995 A**3 that has never been found in the A-DNA being more characteristic of Z-DNA crystals.
Although in all these crystals the octamer forms A-DNA duplexes, the structures of these duplexes are different. A comparison of the most compact form with the least compact one gives the RMS value of 1.7 A, with the distance between phosphates through the major groove being almost twice shorter in the compact form. The local helical parameters within a step also vary significantly with a crystal form. For instance, differences in the helical twist values reach 12 degrees when we consider one and the same step of the octamer in different crystal forms. We will demonstrate the details of the behavior of all structural parameters in all crystal forms. The results prove that the A-DNA is structurally very variable and demonstrate that the fine DNA structure of one and the same fragment can strongly depend on the environment (on the crystal form).
Conformational Behavior
Of CGG Triplet Repeats That Are Associated With Fragile X Syndrome
Xuening Huang and Xiaolian Gao
Department of Chemistry,
University of Houston,
Houston, TX 77204
Anomalous expansion of CGG triplet repeats (TRs) is associated with one of the most common cause of mental retardation, fragile X syndrome. However, there is only limited information in literature concerning the molecular mechanism of CGG expansion. In an effort to understand the role of this tandem TR in gene mutation, we systematically studied the conformational behavior of single stranded (CGG)n (n=2 to 20) TR sequences using a combination of NMR, UV and other biophysical methods. The 1H and 31P spectra of each CGG TRs of different lengths display major conformational transitions upon changing temperatures, while the patterns of the spectral variations are similar. NMR has been used to probe the possible presence of G-quartet, G-syn and other unusual conformations, but the investigations have led to the exclusion of these possibilities at close to physiological conditions. The UV melting measurements reveal an interesting periodic pattern, which may provide clues to the folding of longer CGG TRs. These results and their biological implications will be presented.
Insight Into the Stabilization of A-DNA From Molecular Dynamics Simulations Performed in Aqueous Solvent in the Presence of Cobalt Hexaamine or Ethanol
Thomas E. Cheatham III*(1) and Peter A. Kollman(2)
(1)LSB, DCRT, 12A-2041,
National Institutes of Health,
Bethesda, MD 20892-5626
cheatham@helix.nih.gov
(2)Department of Pharmaceutical Chemistry,
University of California,
San Francisco, CA 94143-0446
pak@cgl.ucsf.edu
Advances in computer power, empirical force field representations, and the development of more reasonable means to handle the long-ranged electrostatic interactions now allow routine "stable" nanosecond-length unrestrained molecular dynamics simulations in aqueous solution (for a review see Louise-May et al., 1996 [1]). Simulations of this type have given insight into sequence specific nucleic acid structure, including hydration and counter-ion association, and also insight into nucleic acid dynamics and flexibility [2-6]. From these studies, it has become clear that a well balanced force field is necessary to properly represent the expected structural preferences and dynamics. An example is the subtle dependence of the relative stability of A-DNA and B-DNA depending on the empirical force field employed. In solution, the Cornell et al. force field [7] appears to favor B-DNA structures [2] in contrast to the CHARMM all-hydrogen parameters for nucleic acids [8] which appear to preferentially stabilize A-DNA structures [4,5].
Although our previous simulations [2] suggest that B-DNA is favored over A-DNA in aqueous solution (as is expected), B-DNA could be artificially stabilized by the force field under all conditions. Therefore to move to the next level and validate the force field and methods, it is necessary to demonstrate stability of A-DNA (with the same force field and simulation protocol) under conditions which are expected to stabilize A-DNA. To this end, a series of simulations on the d[CCAACGTTGG]2 sequence in a canonical A-DNA geometry with the Cornell et al. [8] force field applying the particle mesh Ewald [11] method were performed either in the presence of ~85% ethanol or four cobalt hexaamine ions. Consistent with experiment [9,10], A-DNA appears to be stabilized on at least a two nanosecond time scale. These simulations suggest that ionic interactions and hydration in the major groove stabilize A-DNA. The results from these series of simulations (starting from both canonical A-DNA and B-DNA geometries) is discussed along with a perspective on the "sensitivity" of the force field to various simulation protocols, conditions and small modifications to the force field parameters.
*Presenter and to whom correspondence should be addressed.
References
[1] S. Louise-May, P. Auffinger and E. Westhof, Cur.
Op. Struct. Biol. 6, 289-298, 1996.
[2] T. E. Cheatham, III and P. A. Kollman, J. Mol. Biol. 259, 434-444,
1996.
[3] Y. Duan, P. Wilkosz and J. Rosenberg, J. Mol. Biol. 264, 546-555,
1996.
[4] A. D. MacKerell, Jr, J. Phys. Chem. 101, 646-650, 1997.
[5] L. Q. Yang and B. M. Pettitt, J. Phys. Chem. 100, 2564-2566,
1996.
[6] M. A. Young, B. Jayaram and D. L. Beveridge, J. Amer. Chem. Soc.
119, 59-69, 1997.
[7] W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M. Merz, D.
M. Ferguson, D. C. Spellmeyer, T. Fox, J. W. Caldwell and P. A. Kollman,
J. Amer. Chem. Soc. 117, 5179-5197, 1995.
[8] A. D. MacKerell, Jr., J. Wiorkiewicz-Kuczera and M. Karplus, J. Amer.
Chem. Soc. 117, 11946-11975, 1995.
[9] J. Piskur and A. Rupprecht, FEBS Lettt. 375, 174-178, 1995.
[10] H. Robinson and A. H.-J. Wang, Nuc. Acids Res. 24, 676-682,
1996.
[11] U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee and L. G.
Pedersen, J. Chem. Phys 103, 8577-8593, 1995.
How Does Metalloprotein Oxidation State Affect Structure, Dynamics and Function?
Thomas C. Pochapsky
Department of Chemistry,
Brandeis University,
Waltham, Massachusetts 02254
Putidaredoxin (Pdx), a 106-residue 2Fe-2S ferredoxin, is the physiological reductant of cytochrome P450cam in Pseudomonas putida. Pdx has two accessible oxidation states, Fe+3-Fe+3 (Pdxo) and Fe+3-Fe+2 (Pdxr). A model for solution structure of Pdxo has been determined [1], and redox-dependent structural features of Pdx have been characterized [2]. Non-native metal reconstitution of Pdx has been accomplished using Ga+3, resulting in a protein with a similar global fold to the native Fe-containing ferredoxin [3 ].1H, 13C and 15N assignments have been made for most resonances in the metal binding loop in Ga-Pdx, providing the first high-resolution structural data concerning the metal binding site in this class of 2Fe-2S ferredoxins. Recently, the structures of Pdx and cytochrome P450cam have been incorporated into a model for the diprotein complex which is based on experimental data and the results of molecular dynamics simulations [4].
The interaction between Pdx and cytochrome P450cam exhibits considerable redox dependence. The Kd for the product complex after electron transfer (Pdxo/P450r) is 55 times greater than that of the reactant complex [5] Pdxr shows slower amide exchange kinetics than Pdxo, particularly in regions adjacent to the metal cluster binding site [6]. Similar dynamic changes have been observed for other metalloproteins, e.g., cytochrome c and cytochrome b5 [7,8]. In all cases so far reported, the reduced protein shows slower amide exchange and lower amplitude motions than the oxidized form, indicating that there may be a common principle underlying these observations. It has been recently observed that the DNA-binding activity of SoxR, a Fe2S2-containing protein involved in oxidative shock response, is dependent upon the oxidation state of the metal center, providing another example of redox-dependent binding [9]. Possible linkage between redox-dependent functional and dynamic changes in redox-active proteins will be discussed.
References
[1] Pochapsky et al., Biochemistry 33, 6424,
1994.
[2] Pochapsky et al., Biochemistry 33, 1994.
[3] Kazanis et al., J. Am. Chem. Soc. 117, 6625, 1995.
[4] Pochapsky et al., Biochimie 78, 723, 1996.
[5] Davies & Sligar, Biochemistry 31, 11383, 1992.
[6] Lyons et al., Protein Sci. 5, 627, 1996.
[7] Marmorino et al., Protein Sci. 2, 627, 1993.
[8] Pochapsky et al., J. Am. Chem. Soc. 112, 5258, 1990.
[9] Ding et al., J. Biol. Chem. 271, 33173, 1996.
Satellite Tobacco
Ringspot Virus RNA in Sympton Reduction in Tobacco Ringspot Virus-Infected
Plants
N. I. Imthurn, K. L. Tessitore and G. A. Prody
Department of Chemistry,
Western Washington University,
Bellingham, WA 98225-9150
Tobacco ringspot virus (TobRV) contains two genomic RNAs, RNA 1 and RNA 2. During the course of serial inoculation of TobRV into plants in the laboratory, or following inoculation with a heavy load of virus, plant virus symptoms dramatically decrease due to the presence of the satellite of tobacco ringspot virus (sTobRV), a ca. 360 nucleotide RNA compete with genomic RNAs for TobRV replicase activity or coat protein availibility. sTobRV replicates by rolling circle replication followed by autocatalytic cleavage of the RNA multimers. To test the hypothesis that ribozyme activity may facilitate sympton reduction, we produced 32P-labeled in vitro transcripts from a partial clone of RNA 1 and mixed these with in vitro produced from sTobRV RNA. Results of these experiments were analyzed by gel electrophoresis followed by auto radiography and/or scintillation counting.
Environmental Influences on DNA Curvature
David W. Ussery(1) and Alex Bolshoy(2)
(1)Department of Pharmacology,
Microbiology and Food Hygiene,
Norwegian College of Veterinary Medicine, Oslo, Norway
(2)Center for Biological Sequence Analysis
The Technical University of Denmark,
Lyngby, Denmark
DNA curvature plays an important role in many biological processes. To study environmental influences on DNA curvature we compared the anomalous migration on polyacrylamide gels of ligation ladders of 15 specifically-designed oligonucleotides, which ranged from 0% to 70% GC content. At low temperatures (25°C and below) 14 of the 15 sequences exhibited some degree of anomalous migration. Increased temperature had a significant effect on the anomalous migration (curvature) of some sequences but limited effects on others, although at 50°C (in the presence of EDTA) only 1 sequence migrated anomalously. Mg2+ had a strong influence on the migration of certain sequences, whilst spermine enhanced the anomalous migration of a different set of sequences. Oligomers with a GGC motif exhibited greater curvature (even in the presence of EDTA) than predicted by the "standard" angles used for the nearest-neighbour wedge model and are especially sensitive to divalent cations. The data are have implications for models of DNA curvature and for environmentally-sensitive DNA conformations in the regulation of gene expression.
Extraction of Nucleosome DNA Sequence Pattern by Match Resonance Analysis
Ilya Ioshikhes, Lev Lvovsky and Edward N. Trifonov
Department of Structural Biology,
The Weizmann Institute of Science, Rehovot 76100, Israel
Tel : (972)-8-9343367; Fax: (972)89344136; E-mail: csiosch@dapsas1.weizmann.ac.il
As has been demonstrated by multiple sequence alignment [1], the main feature of the nucleosome DNA sequence pattern is periodic distribution of AA (TT) dinucleotides along the sequence. The periodic distribution of CC (GG) dinucleotides along nucleosome sites has been recently demonstrated as well [2]. Since the periods of these distributions are the same, one can assume that the periodicity would be the main feature of some other dinucleotides' patterns as well. For evaluation of the contributions of other dinucleotides to the nucleosome pattern an original match resonance analysis was applied. In this procedure the sequences were aligned to harmonic distributions of the dinucleotides of interest with the period of 10.3 bp, and the summed output patterns compared to similar outputs generated for random sequences. By varying period and phase shifts between the periodic distributions we maximized the sum of the amplitudes of the oscillating components in the output dinucleotide distributions, thus, calculating the resonance parameters. In this multidimensional optimization one could use many different ascent strategies to reach global maximum. In one procedure the dinucleotides were introduced one after another, according to their effect on the output expressed as standard deviates of the output amplitudes of the dinucleotides from the corresponding reshuffled controls. The AA (TT) dinucleotides were confirmed as substantial contributors. Their relative phase shift and period obtained by this technique are, within the error bars, identical to ones obtained by multiple alignment. Four other dinucleotide complementary pairs were detected as substantial contributors to the nucleosome signal: AC/GT, AG/CT, CC/GG and GA/TC. In second procedure the dinucleotides are introduced one by one according to the difference in the overall outputs caused by a given dinucleotide. The AG/CT, AA/TT, AC/GT, CC/GG and GA/TC have been confirmed as the main contributors to the nucleosome signal. GC appears as a contributor as well. The set of contributing dinucleotides with their respective phases and amplitudes will be presented as the nucleosome dinucleotide matrix suitable for prediction of nucleosome positions in the sequences.
References
[1] Ioshikhes, I., Bolshoy, A., Derenshteyn, K., Borodovsky,
M. & Trifonov, E.N., Nucleosome DNA Sequence Pattern Revealed by Multiple
Alignment of Experimentally Mapped Sequences, J. Mol. Biol. 262,
129-139, 1996.
[2] Bolshoy, A., CC Dinucleotides Contribute to the Bending of DNA in Chromatin,
Nature Struct. Biol. 2, 446-448, 1995.
At Physiological pH, d(CCG)15 Forms a Hairpin Containing Protonated Cytosines and a Distorted Helix
Adong Yu(1), Maria D. Barron(2
), Rebecca M. Romero(3), Mellisa Christy(1), Barry Gold(4),
Jianli Dai(1), Donald M. Gray(2), Ian S. Haworth(3) and Michael Mitas(1)
(1)Department of Biochemistry & Molecular Biology,
Oklahoma State University,
Stillwater, OK, 74078-3035
(2)Program in Molecular and Cell Biology,
The University of Texas at Dallas,
Richardson, TX 75083-0688
(3)Department of Pharmaceutical Sciences,
University of Southern California,
Los Angeles, CA 90033
(4)Eppley Institute for Research in Cancer,
University of Nebraska Medical Center,
Omaha, NE 68198-6805
Previous 1H NMR studies have revealed that the mismatched
cytosines in a [(CCG)2]2 duplex structure are extrahelical and stacked into
the minor groove of the helix (Gao, et al., 1995, J. Am. Chem. Soc.,
114:8883-8884). To investigate the interaction of the mismatched cytosines
in a DNA sequence containing larger numbers of CCG triplets, the structure
of a single-stranded (ss) oligonucleotide containing d(CCG)15 [ss(CCG)15]
was examined by studies of the pH and temperature dependence of electrophoretic
mobility, UV absorbance, circular dichroism, chemical modification, and
P1 nuclease digestion. ss(CCG)15 had an unusually high pKa (7.7±0.2).
At pH 8.5, ss(CCG)15 formed a relatively unstable (Tm = 30°C
in 1 mM Na+) hairpin containing CpG base-pairs steps. At pH 7.5, the hairpin
contained protonated cytosines but no detectable C·+C base-pairs,
increased thermal stability (Tm = 37°C), increased stacking of
the CpG base-pair steps, and a single cytosine that was flipped away from
the central portion of the helix. In 50 mM Na+ at pH 6.5-8.5, the C·G
base pairs melted between 45 and 55°C. Examination of ss(CCG)18 and
ss(CCG)20, which were designed to adopt hairpins containing alternative
GpC base-pair steps (by virtue of four Watson-Crick base pairs flanking
the repeats), revealed hairpins containing CpG base-pair steps, pKas of
~8.2 and ~8.4, respectively, and distorted helices. The results suggest
that DNA sequences containing (CCG)n>15 adopt hairpin conformations
that contain CpG rather than GpC base-pairs steps; the mismatched cytosines
are protonated at physiological pH but are not H-bonded. We propose that
protonation arises from the stacking of two cytosines in the minor groove
of a distorted helix.
The Adenylates of Friedreich's
Ataxia d(GAA) Repeats Are Nuclease-Resistant and Predominantly Unstacked:
Evidence For a New Helical Single-Stranded DNA Structure
Iang-Shan Suen, Mellisa Christy, Brian McEwen, and Michael
Mitas
Department of Biochemistry & Molecular Biology,
Oklahoma State University,
246 Noble Research Center,
Stillwater, OK, 74078-3035
Friedreich's ataxia is caused by expansion of an intronic d(GAA)·d(TTC) triplet repeat located within the frataxin gene. To investigate the possibility that the complementary strands of this sequence contained secondary structure, the electrophoretic mobilities of single-stranded (ss) oligonucleotides containing d(TTC)15 [ss(TTC)15] or d(GAA)15 [ss(GAA)15] were determined relative to their Watson-Crick forms at 28°C, pH 8.5. The relative electrophoretic mobility (Mrel) of ss(TTC)15 was 0.88, a value corresponding to DNA that contains no preferred secondary structure. The Mrel of ss(GAA)15 was 0.94 [for comparison, the Mrel of the d(CTG)15 hairpin at 28°C was 1.12], indicating that this DNA sequence formed a stable (Tm > 50°C) structure. Although nucleotides immediately flanking the 3' end of the d(GAA) repeats were hydrolyzed by P1 nuclease, the triplet repeat region was completely refractory to the action of this enzyme at temperatures < 65 oC in 50 mM Na+, pH 7.5. The 5'-, but not the 3'-adenines within 5'-GAA-3' triplets were readily modified by diethyl pyrocarbonate in 65-600 mM NaCl (pH 7.5, 37°C), indicating that the 5'-adenines lacked stacking/pairing interactions. The guanine residues were modified by dimethyl sulfate, indicating a lack of H-bonds at N7. We suggest that d(GAA) repeats adopt a stable single-stranded helical structure that contains stacked and unstacked bases.
S-DNA and SI-DNA: Genetic and Biophysical Evidence for Slipped DNA Structures in Disease-Associated Triplet Repeat Instability
Pearson, C. E.(1), Eichler, E. E.(2), Lorenzetti, D.(2), Acharya, S.(3), Kramer, P. R.(1), Kramer, S. F.(1), Nelson, D. L.(2), Zoghbi, H. Y.(2), Fishel, R. A.(3) and Sinden, R. R.(1)
(1) Center for Genome Research,
Texas A&M University,
Houston, TX
(2) Molecular & Human Genetics,
Baylor College of Medicine,
Houston, TX
(3) DNA Repair and Molecular Carcinogenesis Program,
Thomas Jefferson University,
Philadelphia, PA
The mechanism(s) of neurodegenerative disease-associated expansions of the (CAG)n·(CTG)n and (CGG)n·(CCG)n trinucleotide repeats is unknown but is thought to involve DNA slippage at the repeats during replication or recombination [1]. In order to gain insights as to the possible mechanism(s) of triplet repeat instability we investigated alternative triplet DNA structures, their formation, and interaction with human proteins [1,2,3].
In normal individuals the (CAG)n tract of spinocerebellar ataxia type 1 (SCA-1) is interrupted with 1-3 centrally located CAT interruptions. While in normal individuals the (CGG)n tract of fragile X A (FRAXA) is interrupted with 1-3 AGG interruptions located in the 5' end. In both cases the non-repeat interruptions confer increased genetic stability to the repeat tract, such that interrupted tracts are less likely to undergo expansions. Loss of non-repeat interruptions, and hence the length of the pure tract are correlated with instability and disease. The stability threshold lengths are 43 and 34 for SCA-1 and FRAXA, respectively. Thus, the presence of these non-repeat interruptions provides a protective effect to both the genetic stability of the repeat tract and to the well being of the individual. The stabilizing affect of the interruptions is not well understood. We investigated the formation of slipped homoduplex DNA structures (S-DNA), where both strands have equal lengths of repeats. Using SCA-1 and FRAXA clones having triplet repeat lengths above and below the stability thresholds, with various genomic patterns of interspersed non-repeat interrupts, we describe slipped DNA structures formed in otherwise duplex DNA between complementary strands [1,2]. The propensity to form S-DNA increased with increased repeat length. The presence of CAT interruptions in the SCA-1 (CAG)n tract hinders the ability to form S-DNA. Similarly the presence of AGG interruptions in the FRAXA (CGG)n tract hinders the ability to form slipped structures. The effect of both the length and the purity of the repeat tract on the propensity of S-DNA formation correlates with their effect on genetic stability in human diseases. This is the first presentation of an alternative trinucleotide repeat structure whose formation correlates with the genetics of the neurodegenerative disease genes, suggesting that these structures may participate in triplet repeat instability.
Slipped intermediate heteroduplexes (SI-DNA) are expected to arise during the replication-mediated expansion from normal (n=30) to diseased (n=50) lengths of repeats [3]. We made SI-DNAs using two genomic clones of the myotonic dystrophy (DM) loci having 30 or 50 repeats [1,3]. Both (CTG)30·(CAG)50 and (CTG)50·(CAG)30 SI-DNAs were structurally distinct in that they could be resolved electrophoretically. We used gel-purified linear non-treated DNAs, S-DNAs or SI-DNAs in a band-shift assay to investigate the binding of the human mismatch repair protein hMSH2. hMSH2 did not bind to the linear DNAs. hMSH2 bound to the S-DNAs in a length dependent manner. hMSH2 also bound to both of the SI-DNAs. hMSH2 bound preferentially to structures having an excess of CAG repeats, implicating strand asymmetry in the mechanism of instability. Our results suggest that mismatch repair may participate in triplet tract length polymorphisms.
References
[1] Pearson, C. E. & Sinden, R. R., Biochmistry,
35:5041-5053, 1996.
[2] Pearson, C. E., Eichler, E. E., Lorenzetti, D., Acharya, S., Kramer,
P. R., Kramer, S. F., Nelson, D. L., Zoghbi, H. Y., Fishel, R. A., &
Sinden, R. R., Am. J. Hum. Genet., 59(4 Suppl.):A48, 244, 1996.
[3] Pearson, C. E., Ewel, A., Acharya, S., Fishel, R. A., & R. R. Sinden,
submitted, 1997.
Design And Structural Analysis Of Novel Nucleic Acids Duplexes With Mixed Strand Disposition
Markus W. Germann(1), James M. Aramini(1), Bernd W.
Kalisch(2) and Johan H. van de Sande(2)
(1) Kimmel Cancer Institute,
Thomas Jefferson University,
Philadelphia, PA, 19107.
(2)Department of Medical Biochemistry,
University of Calgary,
Calgary, Alberta, Canada T2N 1N4
Oligonucleotides containing alpha and beta anomeric nucleotides as well as polarity reversals have potential for application in the area of antisene therapy. Such DNA constructs exhibit a number of desirable properties including nuclease resistance and can permit RNAse H cleavage of a DNA·RNA duplex.
The use of a anomeric nucleotides in conjunction with 5'5' and 3'3' phosphodiester linkages allows the formation of stable DNA and DNA·RNA duplexes with local parallel stranded components embedded in an overall antiparallel helix. We present a thermodynamic and NMR spectroscopic study of 4 DNA decamers which are identical in sequence but contain one a anomeric nucleotide per strand flanked by 5'5' and 3'3' linkages as well as a control decamer. Using this design we systematically determined the consequences of an aT, aC, aA and aG nucleotide as well as the requirement for polarity reversals on structure and stability.
UV melting studies established that the a anomer containing duplexes, with the exception of the alphaC decamer, are only slightly less stable than the control duplex; all are significantly more stable than a decamer containing an aT nucleotide without linkage reversal. Several lines of evidence, in particular imino proton, phosphorus, nuclear Overhauser enhancement, and deoxyribose ring proton spin-spin coupling data, convincingly demonstrate that the overall structural integrity of the alpha and control duplexes are quite comparable, with any perturbations in the former localized to the regions of the construct encompassing the a-nucleotide and the unique backbone linkages. Specifically, Watson-Crick base pairing and anti conformation are observed for all nucleotides. However, the precise extent and type of perturbations are dependent on the a nucleotide. For instance only in the case of the alphaC decamer, which is the least stable duplex, do we observe a perturbation of the phosphodiester backbone across from the 3'3' phosphodiester linkage.
Model of a region in the alphaT decamer duplex d(GCGAAT-3'-3'-(aT)-5'-5'-CGC)2
encompassing the alpha anomeric nucleotide as well as the 3'-3' and 5'-5'
phosphodiester linkages. Arrows indicate sequence polarity.
The sugar puckering in these decamers is highly skewed towards the S-type conformation. However, the b anomeric nucleotide following the 5'5' linkage (C in Figure), is shifted towards the N-type conformation in all decamer duplexes with exception of the alphaG decamer.
The Helical Structure of d(CCG)n Trinucleotide Repeat DNA Probed by Mechlorethamine Crosslinking and Molecular Dynamics Simulations
Rebecca M. Romero(1), Huai Yong Cheng(1), Michael Mitas(2) and Ian S. Haworth(1)
(1)Department of Pharmaceutical Sciences,
Univ. of Southern California,
Los Angeles, CA 90033
(2)Department of Biochemistry and Molecular Biology,
246 Noble Research Center,
Oklahoma State Univ.,
Stillwater, OK 74078
The neurodegenerative diseases Fragile XA and Fragile XE are characterized by the expansion of the (CGG)n and (CCG)n trinucleotide repeat sequence, respectively. The expansion mechanism may be dependent on the single-stranded conformation of these sequences. Previous work has shown that both sequences can form stable hairpins and, for (CCG)15, that the hairpin adopts the alignment described in Yu et al.[2].
The CCG repeat hairpin contains the duplex repeat d(GCC).d(GCC), which in turn contains a central C-C mismatch pair. The cytosine bases of this C-C mismatch have been shown to adopt an extrahelical, minor groove location in the so-called e-motif DNA characterized by Gao et al. [3]. The DNA distortion required to permit the extrahelical cytosines is compensated for by stacking in the pseudo-5'-GC step created by the cytosine motion. Here we describe further evidence of unusual structures that d(GCC).d(GCC) can adopt, both in 'isolated' trinucleotide fragments within random sequence duplexes, and in extended runs of CCG repeat duplexes.
Mechlorethamine ( CH3-N-(CH2CH2Cl)2 ) can form a covalent crosslink between the N7 atoms of the two guanine bases in a d(GXC).d(GYC) fragment, a so-called 1,3-crosslink. We have previously shown for X-Y = G-C that such crosslink formation requires some distortion of the DNA and a limited displacement of the central (X-Y) base pair into the minor groove, induced by the crosslinking moiety [4]. Given this observation, we believed that a C-C mismatch in the X-Y position, in which the cytosine bases might already be displaced into the minor groove, should provide a favorable crosslinking site for mechlorethamine. Data for crosslinking of a d(GXC).d(GYC) fragment centrally placed in a 19mer duplex, where X-Y = C-G, C-A, C-C, T-C, T-A or T-G, show that the d(GCC).d(GCC) site is much more efficiently crosslinked than any other sequence. A characterization of the nature of this crosslink will be presented.
We have performed similar experiments on duplexes containing multiple d(GCC).d(GCC) fragments. These data will be correlated with data from molecular dynamics simulations. In simulations of CCG repeat hairpins we find that the extrahelical bases can stack within the minor groove, such that a given cytosine is stacked with a second cytosine from the opposite strand and originating from the mismatch pair in a 1,4 position, relative to the first cytosine.
References
[1] A.Yu, R.M. Romero, M.D. Barron, J. Dill, M. Christy,
B. Gold, D.M. Gray, I.S.Haworth and M. Mitas. Biochemistry, in press,
1997.
[2] A.Yu et al.,this web page.
[3] M.G. Remias, C-S. Lee and I.S. Haworth, J. Biomol. Struct. Dyn.,
12, 911, 1995.
[4] X. Gao, X. Huang, K.G. Smith, M. Zheng and H. Liu. J. Am. Chem. Soc.,
117, 8883, 1995.
Metals, Motifs and Recognition: 5S rRNA Domain Crystal Structure
C.C. Correll(1,3), B. Freeborn(2), P.B. Moore(1,2),
and T.A. Steitz(1,2,3)
(1)Departments of Molecular Biophysics & Biochemistry and (2)Chemistry,
(3)Howard Hughes Medical Institute,
Yale University, New Haven, CT 06520 USA
An integral part of all cytosolic ribosomes is a ~120 nt RNA called 5S rRNA. We have determined two crystal structures that contain all or part of a 62-nt nuclease resistant domain of E. coli 5S rRNA. This domain binds the ribosomal protein L25 and contains an internal loop, loop E, that requires divalent metal for stability. The first structure, determined at 3.0 Å resolution, contains this domain (1-11, 69-87, and 89-120) that includes helix I, loop E and helix IV. The second structure, determined at 1.5 Å resolution, contains loop E (70-81 and 96-107). This dodecamer contains the minimum 11 pairs required for L25 binding and loop E duplex formation.
Four metals bound in the major groove stabilize the loop E duplex by forming a 'metal zipper'. These metals anchor to one strand with both inner- and outer-sphere coordination and bridge the other strand through outer-sphere coordination.The irregular loop E geometry is defined by three additive RNA motifs termed 'cross strand purine stacks'. In these motifs purines in one strand stack with purines from the other. There are two A-stacks and one G-stack in loop E. Each A-stack contains a single residue kink, creates a major groove pocket, and is directional. Upstream of the juxtaposed A- and G-stack the major groove is wide enough to permit protein recognition. Downstream of the two A-stacks the minor groove is wide enough to permit formation of three novel base pairs that are water-mediated and metal stabilized.
Loop E and helix IV present a surface containing a large variety of H-bond donors and acceptors. This surface is comprised of the three novel pairs in the loop E minor groove and the adjacent widened major groove at the junction with helix IV. Published ethylation and nuclease protection data suggest that L25 binds to this surface.
Conformation Changes in DNA-peptide Complexes
Hana Votavová and Jaroslav Sponar
Institute of Organic Chemistry and Biochemistry,
Academy of Sciences of the Czech Republic,
Prague, Czech Republic
Phone: +42 2 20183 445, Fax: +42 2 243 105 03: E-mail: hanka@marilyn.uochb.cas.cz
A series of hexadecanucleotides containing the ATF/CRE binding site specific for one class of the bZIP proteins was synthesized and their interaction with a simple dimeric peptide modelling the binding domain of bZIP proteins was studied. The model peptide BP1 of sequence RAALKRARNREAARKSRARRAQRLKAGGC (forming a disulphide dimer) was derived from the binding sequence of the GCN4 protein, truncated to the minimum length shown to bind ATF/CRE containing DNA (Talanian et al., Science 249, 769, 1992), by introducing amino-acids changes at six positions. The oligonucleotides differed mainly in the sequences flanking the specific binding site and were compared with some oligonucleotides containing changes within the target site.
The binding of peptide BP1 to oligonucleotides is accompanied by an increase in the a-helix content, which depends strongly on the oligonucleotide sequence. Our experiments with oligonucleotides containing identical ATF/CRE specific target site have shown the importance of sequences flanking the binding site. The best ability for a-helix formation in BP1 was found in complexes with oligonucleotides that contained flanking sequences of homopurine.homopyrimidine type. On the other hand, lowest a-helix formation was found in complexes with the oligonucleotide where the complete target sequence ATF/CRE was flanked by the alternating CA/TG sequence, the conformation of which is known to differ in several respects from the average B-form of DNA. In this case the a-helix content in the complex was almost as low as in the complex with an oligonucleotide containing changes at two base pairs in the target half-site.
It was reported recently that the formation of the complex of bZIP proteins with the ATF/CRE site is accompanied by an increase of the intensity in the high wavelength CD band of the DNA (John et al., Nucleic Acids Res. 24, 4487, 1996). This effect was interpreted as a change in DNA conformation towards the A-DNA form. We have found similar type of CD changes for complexes with two oligonucleotides that exhibit strongest binding. Three other oligonucleotides show this effect only at a temperature of about 10°C. The changes of the DNA CD bands on peptide binding in our complexes are smaller than those reported for the binding of longer peptides. Complexes with some other oligonucleotides containing the ATF/CRE site show either no changes in this CD band or a decrease of the CD band intensity. The observed changes of DNA conformation can be roughly correlated with the ability of the oligonucleotide to induce a-helix formation in the peptide.
The conformation of peptides or nucleotides may be affected by the character of salt present in the solution. In order to find out how salts affect the conformation of both interacting components in the complex , we have studied the DNA-peptide interaction in the presence of NaClO4 , which is known to increase the ability of peptides to form an a-helix and in the presence of Cs+ ions affecting the conformation of DNA. Cs+ ions have no effect, but NaClO4 can change the a-helix content in some of the complexes.
The results of our study of the peptide-DNA interaction
using model compound exhibiting the binding specificity were compared with
results obtained with more simple peptide models which are supposed to bind
unspecifically or only selectively.
Hydration and the Stability of Protein-Protein Complexes
M. C. Strahm and D. M. Soumpasis
Biocomputation Group,
Max-Planck-Institute for Biophysical Chemistry,
37070 Goettingen, FRG
The Potential of Mean Force (PMF) strategy [1], reviewed in [2,3,4] provides the most ecient route for the treatment of solvent efects on structurally complex biomolecular systems like proteins. Here we show the application of the PMF strategy to the discrimination between correctly and incorrectly docked protein-protein complexes. These calculations are part of a project to create a new tool for protein docking. For effciency reasons most of the initial steps for selecting potential docking partners are based on geometrical arguments only. Here we discuss the last ltering which involves the discrimination between stable and non- stable docking complex candidates thus created. We use the PMF strategy to evaluate the charge-charge, hydrophilic, and hydrophobic contributions to the total free binding energy.
References
[1] D. M. Soumpasis, Statistical Mechanics of the B-Z Transition
of DNA: Contribution of diffuse ionic interactions, Proc. Natl. Acad.
Sci. USA, 81:5116-5120, 1984.
[2] D. M. Soumpasis, A. E. Garcia, R. Klement, and T. M. Jovin, The Potentials
of Mean Force (pmf) Approach for Treating Ionic Effects on Biomolecular
Structures in Solution, in D. Beveridge & R. Lavery, editor, Theoretical
Biochemistry and Molecular Biophysics, pages 343-360, Adenine Press NY,
1990.
[3] D. M. Soumpasis, Formal Aspects of the Potential of Mean Force Approach,
in D. M. Soumpasis and T. M. Jovin, editors, Computation of Biomolecular
Structures. Achievements, Problems and Perspectives, Springer Verlag, Berlin
Heidelberg, 1993.
[4] G. Hummer, D. M. Soumpasis, and A. E. Garcia, Potential-of-Mean-Force
Description of Ionic Interactions and Structural Hydration in Biomolecular
Systems, in M. Peyrard, editor, Nonlinear Excitations in Biomolecules,
pages 83-99. Springer-Verlag Berlin, 1995.
Packaging and Dimer Linkage in Retroviral RNA: NMR of a GGAG-tetraloop Substructure of the HIV-1 Packaging Signal.
Lucia Pappalardo(1), Cui Liu(1), István Pelczer(2)
and Philip N. Borer(1)
(1)Chemistry Department,
Syracuse University,
Syracuse, NY 13244-4100
(2)Chemistry Department,
Princeton University,
Frick Lab, Washington Road,
Princeton, NJ 08544
Retroviruses contain two identical RNA strands of about 10,000 nt linked near their 5'-ends. Examples are HIV-1, and others that cause immunodeficiencies, leukemia, and solid tumors. About 200-800 nt near the 5'-end contains both the "packaging" signal (Psi) and a "Dimer Linkage Structure" (DLS). We developed a model for the retroviral Psi/DLS by comparing the RNA sequences in this region for twelve retroviruses and 36 strains of HIV-1, and estimating the free energies of plausible 2D-folds using Zuker's MFOLD program. The model satisfies most constraints from (1) electron microscope results, (2) chemical and enzymatic probing experiments, (3) in vitro stability and kinetic measurements, and (4) in vivo maturation of the DLS. The Psi-region of HIV-1 includes a GGAG-tetraloop hairpin represented in all acceptable MFOLD structures. We synthesized a 20-mer RNA model and our collaborators [B. Lee, R. de Guzman, & M. Summers, unpublished] showed that it forms a complex with the nucleocapsid protein NCp7. Using natural abundance experiments we have assigned nearly all of the RNA,s NMR signals, and find that NCp7 interacts with the loop and its neighboring stem base pairs, not affecting the stem-ends. In the absence of protein the GGAG-loop differs in structure from the GNRA-loop motif. Completion of the resonance assignments and NMR constraint-based modeling is in progress.
Comparision of the Dynamics
Of Angiogenin and RNase A
Madhusudhan, M.S. and Saraswathi Vishveshwara
Molecular Biophysics Unit,
Indian Institute Of Science,
Bangalore, 560 012 India
Angiogenin, a 14KD protein known to be a potent inducer of blood vessel formation is structurally and sequentially(35%) similar to RNase A. This similarity encompasses most of the binding and active sites as well.
Angiogenin is however atlest 4 orders of magnitude a weaker Ribonucleolytic enzyme. An insight into its catalytic activity can be obtained by studying the dynamics of the native and ligand bound enzymes.
As a first step towards this goal, we have carried out Molecular Dynamics simulations of Bovine Angiogenin in aqueous environment. The results are analysed in terms of the RMSD about the average MD structure, dynamical behaviour of secondary structure in general and the binding sites in particular ,intra-protein and protein-water hydrogen bonds. We have deliberated on the different conformations the protein takes during the course of dynamics and on their implications to substrate binding, keeping RNase A as a standard.
Environmentally Dependent
Molecular Dynamic Simulations of DNA Using the BMS Nucleic Acid Force Field
David R. Langley
Dept 502, CADD,
Bristol Myers-Squibb,
5 Research Parkway,
Wallingford, CT 06492
An nvironmentally dependent force field for DNA has been developed and validated using one nanosecond and longer molecular dynamic simulations. Simulations have been carried several hexanucleotides B-DNA, A-DNA and Z-DNA duplexes in low and high salt solutions. The conformational parameters have been analyzed for each simulation and compared with a statistical set generated from 48 B-DNA, 43 A-DNA and 26 Z-DNA high resolution x-ray structures obtained from the nucleic acid database. The average structure from each simulation was found to have a RMSD of 1A or less from its x-ray or canonical structure. Simulations of A-DNA in a low salt environment converts into B-DNA and converges into the RMS space sampled by a low salt simulation of the same sequence starting from B-DNA. High salt simulations of A-DNA either remain close to the x-ray structure or partially melt into single stranded DNA. High salt simulations of B-DNA either convert to the A-form or partially melt.
DNA Triplex Stabilization by
Cationic Polypeptides
Vladimir N. Potaman(1), David R. Corey(2) and Richard
R. Sinden(1)
(1)Institute of Biosciences and Technology,
Texas A&M University,
2121 W.Holcombe Blvd.,
Houston, TX 77030
(2)Howard Hughes Medical Institute,
Department of Pharmacology,
University of Texas Southwestern Medical Center at Dallas,
5323 Harry Hines Blvd.,
Dallas, Texas 75235
Triple-stranded DNA structures might be potentially stabilized by proteins enriched in positively charged amino acid residues that would electrostatically interact with triple helices with a high negative charge density. Surface-localized cationic domains of proteins might be approximated by basic polypeptides. We have recently shown that basic peptides may stabilize intermolecular triplexes, presumably by reducing interstrand phosphate repulsion (1). This phenomenon been used to develop two avenues of investigations.
We tested the interactions of basic peptides with supercoiled DNA capable of forming intramolecular triplex (H-DNA). The superhelical tension required for a local duplex unwinding and subsequent H-DNA formation may be reduced by a number of lysine- and arginine-rich peptides, as well as polyamine spermine. The optimum H-DNA stabilizing effect depends on both peptide sequence and concentration. Higher peptide concentrations may partially inhibit H-DNA formation, perhaps, due to a competitive stabilization of double-stranded structure. This suggestion is corroborated by experiments in simple buffers where the H-DNA formation requires increased superhelical tension as increasing Na+ concentration stabilizes the duplex. We suggest that polycations may shift the duplex-triplex equilibrium to the latter conformation by preferentially binding to a triple helix transiently formed upon fluctuations in local duplex unwinding.
We are developing a new class of triplex-forming molecules which consist of triplex-forming oligonucleotides (TFO) linked to triplex-stabilizing basic peptides. The chemical linkage is obtained by disulfide exchange between a terminal cysteine in the peptide and an S-thiopyridyl activated thiol in the TFO, or by other suitable procedure. Pyrimidine TFO-peptide conjugates efficiently and specifically hybridize to their target DNA at a pH close to physiologically relevant values. The hybridization rates of TFO-peptide conjugates are about three orders of magnitude faster than unmodified TFOs. Variation in the peptide length and sequence may provide freedom in optimizing selectivity and binding affinity of such conjugates.
The results available to date demonstrate that various oligopeptides enriched in lysine and arginine may stabilize inter- and intramolecular triplex DNA. Efficient triplex formation by TFO-peptide conjugates shows that binding a single basic protein fragment may provide sufficient stabilization of the DNA triple helix. These results are relevant to the potential therapeutic application of TFO interaction with DNA wrapped on a nucleosome core with a significant positive charge in histone proteins, and to interactions at regulatory regions of DNA that require protein recognition to mediate cellular processes such as transcription and replication.
References
[1] Potaman, V.N. and R. R. Sinden, Biochemistry 34, 14885-14892, 1995.
Molecular Modelling of
Interactions of Acridine Analogs With Double Helical DNA
V. K. Ganesh and D. Velmurugan
Department of Crystallography and Biophysics
University of Madras,
Guindy campus, Chennai-25, India
Acridines are well known potent antitumour and antibacterial agents and its biological activity is due to its ability to bind to DNA [1]. Acridines bind to DNA by intercalation and induce Frame shift mutations which inhibit Nucleic acid synthesis leading to cell death. Acridines with various substitutions have different binding characteristics and so have specific pharmocological significance. Molecular Modelling studies was initiated for better understanding of the structural and molecular aspects of DNA-Ligand interaction.
Molecular Mechanical calculations of the interactions of N-[ 2-(dimethylamino) ethyl]-9-amino] aminoacridine-4-carboxamide (AAC), N-[2-(dimethylamino)ethyl] acridine-4-carboxamide (DACA), 9-Methyl 3,4,6,9,10 -hexahydro 2,2,8,8 -tetramethyl 1,8 (2H,5H)- acridinedione (MAD) and 9-(4-Dimethylamino phenyl) 3,4,6,9,10 - hexahydro 2,2,8,8 - tetramethyl 1,8 (2H,5H) acridine dione (MAPAD) with double helical DNA were performed. AAC and DACA are clinically tested drugs and AAC was found to be active against leukemia cells [2]. Sequence preference of the binding of the ligands studied was probed by varying X and Y in the octanucleotide duplex d(GGCXYGCC).
In spite of the structural similarity between AAC and DACA, the electrostatic contribution is more in AAC and so AAC more strongly binds to DNA than DACA which agrees with the results from other technique [3].
The presence of dimethyl substitutions at the outer rings of MAD and MAPAD restricts the molecule to intercalate in conventional intercalative mode of binding, but stereochemically and energetically feasible model with MAD intercalating in a way quite similar to the mode of binding of Daunomycin to DNA has been proposed. A possible model for MAPAD with phenyl ring at C9 intercalating and the acridine chromophore interacting in the minor groove has been proposed. The chromophore of MAPAD is not planar which facilitate it to interact well in the minor groove.
Results from our calculations show the ability of acridine diones to bind to DNA. Analysis of the energy results show that the GC step is more preferred by all the ligands subjected in the study. Binding energies, drug helix energies, helix distortion energies and ligand induced structural distortions in the DNA helix will be discussed.
References
[1] Neidle, S., Prog. Med. Chem., 16, 151-221, 1979.
[2] Denny, W. A., Atwell, G. J., Rewcastle, G. W. and Baguley, B. C., J.
Med. Chem. 30, 658-663, 1987.
[3] James M. Crenshaw, David E. Graves and William A. Denny, Biochemistry,
34, 13682-13687, 1995.
1H NMR Investigations of Ion Binding by DNA G-quartets and A-tracts
Nicholas V. Hud and Juli Feigon
Department of Chemistry and Biochemistry and Molecular Biology Institute,
University of California,
Los Angeles, CA 90095
Cations play a major role in the stabilization of nucleic acid structures at all levels. While high resolution nucleic acid structures are now regularly reported, our understanding of the nature and location of cation binding sites remains limited. We have used 1H NMR spectroscopy to explore the ion binding properties of two DNA motifs, guanine quartets (G-quartets) and adenine tracts (A-tracts).
Guanine-rich polymers and oligonucleotides can form quadruplex structures which contain G-quartets[1]. It has been proposed that the preferential binding of K+ versus Na+, Li+, and Rb+ by quadruplexes is a result of the optimal fit of this cation between two G-quartets, coordinated by the guanine carbonyls[2,3]. The slow kinetics of quadruplex formation and disassociation have hindered the determination of the relative free energies of ion binding.1 We have studied the competition between Na+ and K+, under equilibrium conditions, for coordination by the G-quartets of [d(G3T4G3)]2 and [d(G4T4G4)]2. Proton chemical shifts, which are particular to the species of coordinated ion, have been used to monitor the conversion of these molecules from the sodium to the potassium form. We find that on average the replacement of a Na+ by a K+ from a quadruplex is associated with a net free energy change (G°) of approximately -0.85 kcal/mol. A consideration of the relatively large difference between the free energy of hydration for Na+ and K+ (~17 kcal/mol) indicates that the preferred coordination of K+ over Na+ is actually driven by the greater energetic cost of Na+ dehydration with respect to K+.
DNA molecules which contain four to eight consecutive adenine or thymine residues, without a 5'-TA-3' step, exhibit a considerable curvature of the helical axis[4]. The electrophoretic migration of curved DNA in polyacrylamide gels is anomalously slow with respect to straight DNA. The magnitude of this anomalous mobility for A-tract DNA, and presumably its curvature, depends upon the concentration and species of cation present in the running buffer of the gel[5]. These observations suggest that sequence-specific interactions between A-tract DNA and cations may be an important factor in the origin of sequence-directed curvature. To investigate this hypothesis we have monitored resonance line broadening in 1H NMR spectra of DNA oligonucleotide which contain A-tract sequences in the presence of the paramagnetic cation manganese (II). We report the identification of minor groove ion binding sites in A-tract DNA, the exact location of which is sequence dependence. Such ion binding sites may prove to be unique to AT-rich sequences, since all DNA-Mn+2 binding sites previously reported have been localized in the major groove and have involved at least one guanine residue.
Supported by NIH grant GM48123 to J.F. and by NIH postdoctoral fellowship GM17652 to N.V.H.
References
[1] Williamson, J. R., Ann. Rev. Biophys. Biomol. Struc.
23, 703 (1994).
[2] Sundquist, W. I., Klug, A., Nature 342, 825 (1989).
[3] Williamson, J. R., Raghuraman, M.K., Cech, T.R., Cell 59, 871
(1989).
[4] Hagerman, P. J., Annu. Rev. Biochem. 59, 755 (1990).
[5] Diekmann, S., Nucl. Acids Res. 15, 247 (1987).
E. coli DNA Topoisomerase III is an RNA Topoisomerase
Hui Wang(1)*, Russell J.
Di Gate(2) and Nadrian C. Seeman(1)
(1)Department of Chemistry,
New York University, New York, NY 10003
(2)Department of Pharmaceutical Science,
School of Pharmacy,
University of Maryland, Baltimore, MD 21201
The role of type I DNA topoisomerases is key in the cellular metabolism of DNA. These enzymes are intimately involved in replication, transcription, and in the maintenance of torsional stress in the genome. The importance of RNA within the cell is well recognized, as mRNA, as tRNA, in ribosomes, and in processing roles. The single-stranded character of cellular RNA has generally led to the assumption that functional RNA structures can be achieved without the need for strand passage activities to solve problems in RNA molecular topology. Hence, in contrast to DNA, the need for an RNA topoisomerase had not seemed compelling.
A synthetic strand of RNA has been designed so that it can adopt two different topological states when ligated into a cyclic molecule, a circle and a trefoil knot. The RNA knot and circle have been characterized by their behavior in gel electrophoresis and sedimentation experiments. This system allows one to assay for the existence of an RNA topoisomerase, because the two RNA molecules can be interconverted only by a strand passage event. We find that the interconversion of these two species can be catalyzed by E. coli DNA topoisomerase III, indicating that this enzyme can act as an RNA topoisomerase. The conversion of circles to knots is accompanied by a small amount of RNA catenane generation. These findings suggest that strand passage must be considered a potential component of the folding and modification of RNA molecules.
This research has been supported by grants from the NIH,
ONR and by a Margaret and Herman Sokol Fellowship.
*Address correspondence to this author at: Phone: 212-998-8464; Fax:212-260-7905;
E-mail: hqw1482@is2.nyu.edu.
Switching and Looping in i-Motif Structures
J-L Leroy, S. Nonin, X. Han, A. T. Phan, and M. Guéron
Groupe de Biophysique de l'Ecole Polytechnique,
91128 Palaiseau, France
C-rich DNA sequences bring systematic intercalation into the collection of nucleic acid structural motifs. They associate into a four-stranded structure built by the intercalation of two duplexes, each of which including parallel strands with hemi-protonated C.C+ pairs in a so called i-motif. The i-motif may also be formed by the inter-molecular assembly of two oligomers containing two repeats of cytidines, or by folding of a single strand containing four cytidine stretches. The oligomer stoichiometry is determined from the concentration dependence of the multimer-monomer equilibrium. The intercalation topology of the i-motif is read off the patterns of H1'-H1' and of t amino-H2'/H2" cross-peaks in NOESY spectra.
In many tetrameric i-motif structures, e.g. [d(TCC)]4, the four strands are identical on the NMR time scale. By contrast, symmetry is broken in the d(5mCCTCC) tetramer, the thymidines being paired in one duplex and unstacked in the other. Exchange cross-peaks show that the two duplexes switch configurations within the tetramer at a rate of 1.4 s-1 at 0 °C.
The oligonucleotide d(5mCCTTTACC) forms an i-motif dimer by intercalation of two symmetry related hairpins. The TTTA stretches are looped on opposite sides of the i-motif core and span the wide grooves. The core is extended by stacking of A6 on C2.C8+. The d(5mCCTTTTCC) oligonucleotide, whose position 6 is occupied by a thymidine instead of an adenosine, has a different folding topology: the two TTTT loops are on the same side of the core.
The oligonucleotide d(5mCCT3CCT3ACCT3CC) folds into an intramolecular i-motif up to pH 7.4. Four C.C+ pairs form the core of the structure. At 0 °C, the base-pair lifetimes range from 10 seconds (for pairs 5mC1.C12+ and C6.C17+ ) to 10 milliseconds for C7.C18+ and C2.C13+).
The two TTT fragments loop across the two narrows grooves and the TTTA loops across one of the wide grooves. The i-motif is extended in the direction of the TTT loops by stacking of the T5.T16 base-pair on the C2.C13+ pair. T8 H3 is H-bonded to the nitrogen N7 of A11 within the TTTA loop.
These observations illustrate the effect of nucleotides sequences on i-motif structures, and are part of an investigation of the structure of the C-rich strand of telomers and centromers.
DNA Bending by Small Molecules: Studies With Calicheamicin
Aaron A. Salzberg and Peter C. Dedon
Division of Toxicology,
MIT, Cambridge, MA 02139
We seek to understand those elements of drug and DNA structure and dynamics that determine the targeting of drugs to specific sites within DNA in vivo. As a part of this overall goal, we have investigated the role of DNA bending and flexibility in the selection of targets by calicheamicin g (CAL) in vitro. This potent enediyne antitumor antibiotic binds site-specifically and non-covalently in the minor groove of DNA and causes double strand breaks by way of a highly reactive diradical intermediate. Two observations suggest that DNA topology and flexibility play a role in its target selection:
(1) CAL targets the 3'-ends of purine tracts (poly-A and mixed A/G) and (2) bending of DNA in a model nucleosome increases CAL-induced DNA damage. We now present evidence that CAL bends DNA upon binding and therefore, may recognize sequences that are -- or can easily be -- bent. Our conclusions are based on gel mobility and DNA cyclization assays that provide a unique opportunity to investigate DNA structure and non-covalent drug binding.
For these studies, 21dp duplex oligonucleotides were designed with two CAL damage sites that were either in- or out-of phase with the 10.5 bp helical repeat of B-DNA. Gel mobility studies revealed that these target sequences do not possess a discernible degree of intrinsic curvature; ruling out fixed curvature as a requirement for CAL target selection.
To qualitatively assess the effects of CAL binding on DNA structure, the 21 bp constructs were polymerized, with and without CAL-e, and the amount of cyclized polymers individually measured by one-dimensional electrophoresis. This was made possible by first digesting the DNA with Bal31 exonuclease to remove linear polymers. CAL-e, the inactive form of the drug, was used since thiols present in the ligase preparation would activate the enediyne to produce DNA damage. (CAL-e binds to the same sequences as CAL with a ~10-fold lower affinity.) The mean circle size produced by the polymerization of the in-phase construct shifted two monomer lengths (from 210 bp to 168 bp) upon the addition of CAL-e (previous page). In contrast, the out-of-phase duplex was relatively non-responsive; suggesting that the drug increases the cyclization of the in-phase polymer by inducing a bend in the DNA rather than twisting the helix or causing some other "sequence independent" effect.
To distinguish between -- and quantify the magnitude of -- the bend and twist components, the probability of cyclization was measured in a 273 bp (13-mer) polymer of the in-phase binding construct at increasing CAL-e concentrations. J, the molar cyclization factor, was determined by curve-fitting the measured products of the time course experiments to the appropriate kinetic equations. Shown above, J increases smoothly with increasing CAL-e concentration which suggests that, in response to drug binding, the polymer undergoes bending with little twist.
Interestingly, the nonlinear behavior of J appears to result from a dramatic decrease in the rate of dimerization at high CAL-e concentrations. This may reflect additional affects of the drug or some form of inhibition resulting from the circular, but not cyclized, form of the DNA molecule and it suggests that reliable J data can only be obtained at low drug concentrations.
These results demonstrate that, in addition to base recognition and sequence-independent factors, DNA bending and flexibility are important determinants in the selection of DNA targets by CAL and potentially by many small molecules. Our findings with CAL also support the hypothesis that, in addition to the structural anomalies of A-tracts, there exist unique structural or dynamic characteristics that distinguish the 3'-ends of purine tracts from other sequences. Finally, the Bal31 modification of the cyclization assay allows it to be used as a facile screen for DNA bending by small molecules, while the J factor results suggest that the rate of dimerization may depend on the structure of the DNA molecule.
Sponsored by an MIT Sloan Basic Grant, the Samuel A. Goldblith
Professorship, NIH/NCI grant CA64524 and NIEHS grant ES07020.
Helix -Coil Transitions in DNA Modified by Novel Pt(II) Complexes: A pH Melting Study
R. Malathi(1)*, G. Natarajan(1), E. Holler(2)
(1)Department of Genetics,
Dr ALM Postgraduate Institute ofBasic Medical Sciences,
Taramani, Madras-600113, India.
(2)Institut fur Biophysik and Physikaische Biochemie,
Universitat Regensburg, D-93040, Regensburg, Germany
Phone: 00 - 91 - 44 - 492 5317; Fax : 00 - 91 - 44 - 492 6709
Recent reports have shown that the pH could also be used as a melting factor (1) to monitor the helix - coil transitions in DNA; the results being comparable to those obtained using T(m) studies; the rapidity with which this method can be performed to obtain similiar transition curves and the elimination of the evaporation factor (at high temperatures as seen in T(m) studies ) is one of the advantages offered by this technique. With regard to its suitability in studying DNA - Drug interactions, the addition of platinum(II) complexes changed the P(m) (pH of melting ) in a predictable manner thereby confirming the destabilisation of bases in DNA[1].
In the present study, melting properties of calf thymus DNA modified by certain chloro substituted platinum complexes have been generated using pH as a denaturing factor . Diamminediaquaplatinum (active form of the antitumor drug -cisplatin) was coupled with beta - poly -L- malate (a bioresorbable polymer), L-malate (it's monomer) and L-succinate (non-hydroxy group containing dicarboxylate). At a constant Pt : P ratio , (=5) the extent of the damage to DNA by these complexes in comparision to cisplatin was cispt> succ-Pt > Mal-Pt > PMA- Pt > carboplatin. Given the similiarity in the sidegroups of these Platinum complexes with that of carboplatin (a succesful second generation analogue of cisplatin), interesting variations have been obtained in the DNA melting profiles, the implications of which havebeen discussed in the present study .
*Author to whom the correspondence should be addressed.
References
[1] Natarajan , G. ; Malathi , R. ; Holler , E ., Analytical Biochemistry, 237, 152 - 155, 1996.
Molecular Recognition Between DNA And Chiral Intercalative Metalloprobes
K. A.Vickery(1), A. M. Bonin(2), P. A. Williams(3) and
R. S. Vagg(1)
(1)School of Chemistry,
Macquarie University,
Sydney, N.S.W. 2109, Australia
(2)Toxicology Unit,
N.I.O.H. & S.,
Sydney, N.S.W. 2050, Australia
(3)Department of Chemistry,
University of Western Sydney
Nepean, N.S.W. 2747, Australia
Polynucleic acids are capable of enantiopreferentially
recognising and binding to small molecules. Substitutionally inert chiral
ruthenium complexes have provided valuable use as spectroscopic probes in
the determination of the causal factors governing small-molecule recognition
by D
NA. In the past decade, considerable work has been focused
on the enantioselectivity observed by the D- and L-[Ru(phen)3]2+ ions towards
DNA,eg 1-3, where the right-handed propeller form (D) was proposed to preferentially
bind to DNA. Progressional studies, 1-4, have led to the use of the analogues
[Ru(L)2(L')]2+ , where L is 1,10-phenanthroline or
2,2'-bipyridine and L' is a diimine bidentate with an extended aromatic
surface area. These extended components function as the intercalating chromophore
and hence determine overall molecular binding affinity. We reason that enantioselectivity
does not require the remaining four coordination sites on the metal ion
to be occupied by the two secondary L bidentates. Hence we have developed
a series of novel octahedral complexes of the general type [Ru(picchxnMe2)(L')]2+
(shown), in which picchxnMe2 is a rigid chiral tetradentate that
controls both the stereo- (a vs b) and enantiomeric (D vs L) forms of the
cations. The diimine L' has been varied in order to optimise the
DNA binding abilities of these probe molecules through intercalation. The
enantioselectivity of this binding is determined by tetradentate design.
Structural information has been obtained on intermolecular hydrophobic interactions for the series of probes where L' is phen, dpq 5 (shown)and phdi. The chiral recognition functions of DNA towards the various isomeric complexes were investigated using natural DNA, synthetic oligonucleotides and bacterial DNA systems. The RR form of picchxnMe2 has been shown to provide cis-D diastereoisomers exclusively. Consistent with our structural hypothesis, the D-cis-a isomers demonstrate higher intercalation affinity than for the L or b forms ineach of the given systems. The interpretation of these interactionswill be presented.
References
[1] A. Yamagishi, J. Chem. Soc. Chem. Commun., 572,
1983.
[2] J. Barton, J. Am. Chem. Soc., 106, 2172, 1984.
[3] M. Erikkson, M. Leijon, C. Hiort, B. Norden and A. Grasland, Biochem.,
33, 5031, 1994.
[4] R. Morgan, S. Chatterjee, A. Baker and T. Strekas, Inorg. Chem.,
30, 2687, 1991.
[5] dpq is dipyrido[3,2-d: 2'3-f]quinoxaline and phdi is 9,10-phenanthrenequinone
diimine
Effect of Vicinal 2',5'
and 3',5'-Phosphodiester Linkages on the Formation of Hairpin DNA
Ravinderjit S. Braich and Masad J. Damha*
Department of Chemistry,
McGill University,
801 Sherbrooke St. W.,
Montreal, QC, CANADA H3A 2K6
We have been interested in the use of branched oligonucleotides for inducing the formation of novel triple-helical DNA and I-motif tetraplex DNA (see accompanying abstract). We have now prepared the first examples of DNA hairpins containing a branched X-5',2'-rA-3',5'-Y unit in their loop (1) and have studied their properties by UV-melting experiments. For example, a branched hairpin with a d(CC)-5',2'-rA-3',5'-d(CC) pentaloop exhibited better thermal stability to those with a natural loop, e.g., dCCACC pentaloop (DTm= 7°C), and dCCCC tetraloop (DTm= 4°C). Molecular modeling studies suggest that branched loops possess considerable self-structure (e.g., base-base stacking) that is reminiscent of exceptionally stable hairpin loops (2). Studies on various branched pentaloops containing various sequence combinations are being pursued to determine whether the gain of thermal stability is dependent on base sequence, or whether it is a general property of these branched pentaloops.
References
[1] R. Braich and M.J. Damha, submitted.
[2] C. Cheong, G. Varani and I. Tinoco Jr., Nature 346, 680, 1990.
Association of Branched Oligonucleotides into C-Tetraplexes: 2-Deoxyribose and Arabinose, but not Ribose, Stabilize the C-Tetrad Structure
S. Robidoux(1), M.J. Damha(1), K. Gehring(2) and R.
Klinck(2)
Contribution from the Departments of (1)Chemistry
and (2)Biochemistry,
McGill University,
and the (2)Biotechnology Research Institute,
Montreal, QC, Canada
The unique architecture of branched oligonucleotides mimicking lariat RNA introns (1) can be exploited to study novel DNA structures (2). Recently attention was focused on reverse-Hoogsteen triplexes formed by branched oligothymidylates, viz. 3'-(dT)105'p2'-rA-3'p5'(dT)10-3' complexed to dA10 (3). These experiments have been extended to branched oligomers associating as two parallel duplexes with intercalating C/C+ base pairs (I-motif DNA, or cytosine tetraplex DNA) (4,5)
.
The formation of branched C-tetraplexes was induced by linkage of two (deoxy-C)n strands (n= 3-10) through their 5'-ends via coupling to riboadenosine (rA) at the vicinal 2'- and 3'-oxygen atoms. This arrangement causes the orientation of the dC strands to be parallel, and forces the formation of a C/C+ duplex that self-associate into a C-tetrad. CD, NMR, thermal denaturation, and native gel electrophoresis experiments supporting this structure will be presented.
In order to test the hypothesis that favorable sugar-sugar interactions stabilize DNA C-tetrads (4), thermal denaturation and circular dichroism of sugar-modified branched oligomers, viz. 3'-(ara-C)55'p2'-rA-3'p5'(ara-C)5-3', and 3'-(ribo-C)55'p2'-rA-3'p5'(ribo-C)5-3' (in which the ribo-C5 "arms" are either 3',5'- or 2',5'-linked) were studied in 10 mM acetate buffer. The deoxy-C analogue served as model system. The stability of the arabinose tetraplex was found to be high and similar to the deoxyribose tetraplex, whereas the ribo analogues did not form a stable complex under the same conditions. These observations offer a basis as well as a need for understanding the factors that stabilize C-tetraplexes.
References
[1] J. C. Wallace and M. Edmons, Proc. Natl.
Acad. Sci. USA 80, 950-954, 1983.
[2] R. H. E. Hudson, K. Ganeshan and M.J. Damha, "Branched Oligonucleotides:
Synthesis and Biological Applications", In Carbohydrate Modifications
in Antisense Research, P. D. Cook and Y. S. Sanghvi, Eds.,
American Chemical Society Symposium Series 580, 133-152
,1994.
[3] R. H. Hudson, A. H. Uddin and M. J. Damha, J. Am. Chem.
Soc. 117, 12470-12477, 1995.
[4] K. Gehring, J. L. Leroy and M. Gueron, Nature 363, 561-565,
1993.
[5] L. Chen; L. Cai, X. Zhang and A. Rich, Biochemistry 33,
13540-13546, 1994.
Selective Association
and Antisense Properties of 2',5'-Linked Oligoribonucleotides
M. Wasner(1), A. Uddin(1), G. Borkow(2), D. Arion(2),
M. Parniak(2) and M.J. Damha(1)
(1)Department of Chemistry,
McGill University,
801 Sherbrooke St. W.,
Montreal, QC, Canada H3A 2K6
(2)McGill AIDS Centre and Faculty of Medicine,
McGill University,
3755 Cote Ste-Catherine Road,
Montreal, QC, Canada H3T 1E2
Our group discovered that 2',5'-oligoribonucleotides bind to normal single-stranded RNA but only weakly (if at all) to normal single-stranded DNA (1, 2). Such "RNA binding selectivity" has recently been observed with 2',5'-linked ssDNA (3-5), 2',5'-linked formacetal DNA (6), and mixed 2',5'/3',5' RNA/DNA oligonucleotides (7), suggesting that this may represent a general feature of oligomers constructed with 2',5'-internucleotide linkages.
We have now prepared 2',5'-oligoribonucleotides containing mixed sequences of the four nucleobases (A, G, C and U) and have studied the binding of these oligonucleotides to normal complementary RNA and DNA. We prepared two 2',5'-RNA sequences that were designed to be complementary to RNA strands that mimic the R and U5 regions of genomic HIV RNA. The corresponding DNA targets were also synthesized. The properties of the complexes formed between complementary 2',5'-RNA and 3',5'-RNA (and 3'5'-DNA) strands were assessed by UV-melting experiments and circular dichroism (CD) spectroscopy. RNase H activation, and biological properties of 2',5'-RNAs including in vitro inhibition of HIV reverse transcription were examined. The structural basis for the selective hybridization properties of 2',5'-linked nucleic acids will be discussed (8).
References
[1] P.A. Giannaris and M.J. Damha, Nucl. Acids Res.
Symp. Ser. 24, 290, 1991.
[2] P.A. Giannaris and M.J. Damha, Nucl. Acids Res. 21, 4742-4749,
1993.
[3] R. Alul and G.D. Hoke, Antisense Res. Dev. 5, 3-11, 1995.
[4] T.J. Sheppard and R.C. Breslow, J. Am.Chem. Soc. 118, 9810-9811,
1996.
[5] T.P Prakash, K.-E. Jung and C. Switzer, Chem Comm. 1793-1794,
1996.
[6] J.S. Pudlo et al. Tetrahedron Letters 50, 9315-9318, 1994.
[7] E.R. Kandimalla et al. Nucl. Acids Res. 25, 370-378, 1997.
[8] M.J. Damha et al., Nucl. Acids Res. 23, 3967-3973, 1995.
Hybridization Properties of Arabinonucleic Acids (ANA) Influence of Stereochemistry at 2' on the Stability of Double and Triple Helices
Anne Noronha and Masad J. Damha
Department of Chemistry,
McGill University,
801 Sherbrooke St. W.,
Montreal, QC, Canada H3A 2K6
We have been interested in the properties of an stereoisomer of normal RNA based on D-arabinose instead of the natural D-ribose (1). Related studies have been described by Pfleiderer and co-workers on the synthesis of a transfer ANA molecule (2), and Jacobsen has recently reported the properties of DNA duplexes incorporating a single 2'-O-methyl-b-D-araT unit (3). In our earlier study, we synthesized a series of ANA strands and demonstrated that octanucleotides of ara-C, ara-A and ara-U hybridize with their complementary strands resulting in complexes of high thermal stability. In fact, in complexes with poly-dT or poly-rU, ara(Ap)7A displayed greater hybridization affinity than ribo(Ap)7A (1).
We have now prepared oligoarabinonucleotides of mixed base composition, and the analogous DNA and RNA strands. The properties of the complexes formed between these oligomers and a 18-mer RNA sequence corresponding to the R-region of genomic HIV RNA were assessed by UV-melting experiments and circular dichroism (CD) spectroscopy. In addition, a polypyrimidine ANA strand of mixed composition was tested for its ability to form a stable triple helix with a DNA target duplex under physiological conditions.
References
[1] P. Giannaris and M.J. Damha, Can. J. Chem. 72,
909, 1994.
[2] M. Resmini and W. Pfleiderer, Helv. Chim. Acta 76, 158,
1993.
[3] C. Gotfredsen et al. Bioconjugate Chem. 7, 680, 1996.
Biophysical Studies and Synthesis of Oligonucleotides Containing a Peptide / Aromatic Backbone
C. J. Wilds(1), J. F. Lunetta(2), Y. S. Tsantrizos(2)
and M. J. Damha(1)
(1)Department of Chemistry,
McGill University,
801 Sherbrooke St. W.,
Montreal, QC, Canada H3A 2K6
(2)Department of Chemistry and Biochemistry,
Concordia University,
1455 deMaisonneuve Blvd. W.,
Montreal, QC, Canada H3G 1M8
The synthesis of novel oligonucleotide analogues containing a peptide / aromatic (APNA) backbone is described. The properties of the complexes formed between APNA-DNA chimera and ssDNA (and RNA) targets were assessed by UV-melting experiments and circular dichroism (CD) spectroscopy.
References
[1] P.E. Nielsen, M. Egholm, R.H. Berg and O. Buchardt,
Science 254, 1497, 1991.
[2] Y.S. Tsantrizos, J.F. Lunetta, M. Boyd and M.-C. Wilson, submitted.