Note that web at present does not support special characters such as greek etc. and as such, be careful in reading certain abstracts in which the greek is replaced with english alphabet. For example, alpha = a, beta = b and so on.
Marked Discreteness in Coil-globule Transition of Large DNA Molecules Induced by PEG (Polyethylene Glycol)
Yukiko Matsuzawa(1) and Kenichi Yoshikawa(2)
(1)Department of Ecological Engineering,
Toyohashi University of Technology,
Toyohashi 441 Japan
Current address: Department of Biochemistry,
University of Minnesota,
St. Paul. MN, 55108, USA
(2)Graduate School of Human Informatics,
Nagoya University,
Nagoya 464-01, Japan
The condensation of DNA has fundamental biological consequences [1]. DNA is packed very compactly inside a cell, in spite of its elongated state in dilute solution. Using fluorescence microscopy, we have observed the conformational change in individual large DNA molecules (T4DNA, 166kbp) by the addition of PEG. The main results are as follows:
1) Individual DNA chains undergo discrete transition with
the addition of PEG; the change in the effective volume is more than 104
[2].
2) The transition appears continuous on the level of the ensemble of DNA
chains.
3) The transition from elongated coil to collapsed globule occurs through
a kinetic process of nucleation & growth [3].
4) The ends of the chain exhibit the highest probability of nucleation [4].
References and Footnotes
1. Reich, Z., Ghirlando, R. and Misnky, A., Biochemistry,
30, 7828-7836 (1991).
2. Matsuzawa, Y., Yoshikawa, K., Physica D, 84, 220-227 (1995).
3. Yoshikawa, K. and Matsuzawa, Y., J. Am. Chem. Soc., 118,
929-930 (1996).
4. Matsuzawa, Y., Yonezawa, Y. and Yoshikawa, K., Biochem. Biophys.
Res. Com., 225, 796-800 (1996).
Stereochemistry of Protein/DNA Interactions
Paul B. Sigler
Yale University/HHMI,
Dept. of Molecular Biophysics and Biochemistry,
New Haven, CT 06511
High resolution structural studies, hydrodynamics, mutational
analysis, chemogenetics, covalent and computational chemistry have combined
to present a far more elaborate picture than most expected for the formation
and function of proton/nucleic complexes. As the assemblies grow in complexity
along with the biology of regulation and enzymatic processes, we are confronting
serious questions as to what problems should be addressed and how. These
perspectives will be discussed primarily in the context of high resolution
crystallographic analysis.
Structural Investigation of a Binary-Addressed Oligonucleotide System Juxtaposing Pyrene and Perfluoroazide Units by Means of High-Field NMR Spectroscopy and Molecular Modeling
Elena V. Bichenkova(1), Debora S. Marks(1), Sergei G.
Lokhov(2), Michail I. Dobrikov(2), Valentin Vlassov(2) and Kenneth T. Douglas(1)
(1)School of Pharmacy and Pharmaceutical Sciences,
University of Manchester,
Manchester M13 9PL, UK
(2)The Institute of Bioorganic Chemistry,
8 Lavent'ev Ave.,
Novosibirsk, 630090 Russia
Sequence-specific modification of nucleic acids by reactive oligonucleotide derivatives bearing covalently attached chemical groups and complimentary to particular target sequences is a promising approach for the direct chemical study of genetic material. Recently, anew approach has been proposed to improve the site-specificity and efficiency of the modification of nucleic acid target sequences, the binary system of complementary-addressing nucleic acid sequences. The binary system comprises two oligonucleotides, one modified with a photosensitizing group and the other with a photoreactive group. The sites of chemical modification are arranged to bring the two chemical functions close enough together in space to allow efficient energy transfer from the photo-excited photosensitizer to the arylazide moiety which expels N2 to form a nitrene which covalently labels the target nucleic acid. In this contribution structural aspects of this system have been investigated using the model binary system 1:2:3, where 1 is the target 12-mer pdGTATCAGTTTCT, 2 is a photoactivatable fluoroazide derivative dAGAAACp-L-Az and 3 is the photosensitizer derivative Pyr-pdTGATAC (here: Az-p-azidotetrafluorobenzyl group, Pyr-pyrenyl-1-methylamino group, L-linker group). Comprehensive analysis of 1H NOESY spectra of 1:2:3 showed that terminal fragments of the complex [5'p-1T-2G-3A-4T-], [-21A-22T-23A-24C] [-8T-9T-10T-11C-12T] and [13A-14G-15A-15A-17A-18C-] gave a continuous set of intr- and inter-nucleotide interactions, typical of regular double-stranded B-DNA. In contrast, the central region of the complex composed by 5C, 6A, 7G, 19T and 20G nucleotide residues, nearest to the Pyr and Az groups, was found to be distorted. Thus some signals from aromatic and/or sugar-ring protons of the above nucleotide residues were extremely broadened or almost absent. Moreover, some intra- and/or inter-nucleotide interactions, typical of the tandem system and some cross peaks of low intensity between the H2 proton of the Pyr group and 7G(hl'), 7G(H2'), 7G(H3'), 4T(H2") and 22T(CH3) were observed. Additional NOE-interactions between methylene protons of the linker group L and 18C nucleotide residue were detected. Proton-proton distances obtained from 1H NOESY experiment at 300 ms mixing time were used as a constraints for the refinement of the structure. A series of molecular dynamics simulations was performed for a total of 106 femoseconds (SYBIL 6.3) followed by energy minimization methods using simplex algorithms to remove bad strike contacts and Power algorithms and convergence. The final structure which satisfied both the experimental data and the energy criteria displayed significant distortion of the double-stranded helix in the middle of the complex, with an RMS deviation of 7 Angstroms (standard deviation 4.2) of the [-5C-6A-7G.-18C-19T-20G-] atoms from B-DNA. The heterocyclic bases of 6C and 5A were found to be rotated away from the major groove, the torsion causing bulging and destroying the integrity of the helix. The Pyr group was found in what remains of the minor groove near 4T, 5T, 6A and 7G. The Az group nitrene generated by FRET-induced N2 expulsion has been shown in other studies to attack the target 12-mer chain from the major groove side. This model of the NMR structure sites the Az group to within 6 Angstroms of N7 of 7G from the side of major groove and the centroid of the azide ring less than 9 Angstroms from the centroid of the ring system of Pyr group.
How the TATA Box Selects its Protein Partner
Nina Pastor and Harel Weinstein
Dept. Physiology and Biophysics,
Mount Sinai School of Medicine,
New York, NY 10029
The TATA box-binding protein (TBP) binds in the minor groove of the TATA promoter sequence, drastically distorting the DNA helix: TBP inserts Phe residues at TA and at AG (or AA) steps, and unwinds the DNA, most notably at the center of the TATA box. To explore the role of the TATA sequence in determining the specificity of TBP binding to DNA, we studied the solution structure and dynamics of 7 DNA dodecamer sequences, 6 of which bind to TBP, in molecular dynamics (MD) simulations with the CHARMS23 potential, explicit waters, one Na+/phosphate, and periodic boundary conditions. A comparison of the structural parameters of these dodecamers with DNA in the crystal structures of TBP/DNA complexes shows that the probability of the various sequences to adopt conformations mimicking the distortions required for TBP binding correlates with TBP-binding ability. The results suggest how the propensity of various TATA sequences to prepare structurally for TBP binding, provides a key selectivity determinant.
A Backward Glance at the Nucleic Acids
Alexander Rich
Department of Biology, MIT
Cambridge, MA 02139
Forty-five years ago, only a little was known about the role of the nucleic acids in biological systems. With the discovery of the DNA double helix by Watson and Crick in 1953, a number of open questions remained. Does RNA form a double helix? If information travels from DNA to RNA, how is that done? What is the system for transferring information? If the synthesis of proteins are directed by RNA, how is that done? Today, very large textbooks are filled with an array of facts about biochemistry and molecular biology and a large number of powerful techniques are used. How was research done in the early days of molecular biology? This talk will present a personal account of research on the nucleic acids which started before the advent of modern molecular biology.
Structural Studies on the Unstable Triplet Repeats
S.V. Santhana Mariappan(1), Paolo Catasti(1,3), Xian
Chen(1,3), Louis A. Silks III(2), Robert K. Moyzis(4), Larry L. Deavan(4),
E. Morton Bradbury(3,5) and Goutam Gupta(1)
(1)Theoretical Biology and Biophysics, T-10, MS K710,
(2)Chemical Science and Technology Division, CST-4,
(3)Life Sciences Division, LS-2, MS 880
(4) Center for Human Genome Studies,
Los Alamos National Laboratory,
Los Alamos, NM 87545
(5)Department of Biological Chemistry,
School of Medicine,
University of California at Davis,
Davis, CA 95616
Genetic instability due to expansions of triplet repeats is commonly associated with many neurological disorders such as fragile X syndrome (FRAX), myotonic dystrophy (DM), Huntington's disease (HD), and Friedreich's ataxia (FRDA). In each disease, the expansion of a specific triplet repeat causes abnormal expression of the gene associated with the disease.
Here, we summarize our structural studies on (a) the hairpins formed by the GC-rich triplet repeats associated with FRAX, DM, and HD and (b) the triple helix structures formed by the GAA triplet repeat associated with FRDA. More detailed studies are presented for the FRAX and FRDA triplets in order to emphasize the importance of the unusual DNA structures formed by the triplet repeats in the non-coding region of genes in the control of their expression.
In FRAX, the expansion of the upstream (CCG)n repeat is associated with hypermethylation of the CpG islands inside the repeat and the subsequent suppression of the FMR-1 gene. Structural studies by 2D NMR spectroscopy and gel electrophoresis reveal that the individual (CCG)n strands form singly folded hairpin structures. An in vitro replication assay demonstrates that the same hairpin structure of the (CCG)n strand can form even in presence of its complementary (CGG)n strand. These results suggest the possibility of slippage of the FRAX repeats during replication through the formation of a hairpin by the (CCG)n strand. This slippage structure, which is essentially a three-way junction with two Watson-Crick (CCG)n·(CGG)n arms and a (CGG)n hairpin, is also shown to be an excellent substrate for CpG methylation by the human methyltransferase. Hetero-nuclear NMR spectroscopy of the (CCG)n hairpins with 15N4-labeled C's reveals the presence of flexible C·C pairs at the CpG sites that are more susceptible to open-closure than the Watson-Crick C·G pairs. This flexibility allows for a facile transition of the C's at CpG to the "flipped out" or "activated" state for methylation. It is also shown that the larger the FRAX repeat length the higher is the probability of hairpin formation which facilitates CpG methylation upstream of the FMR-1 gene and the subsequent down regulation of transcription.
In FRDA, the expansion of the (GAA)n repeats inside an intron of the frataxin gene drastically reduces the level of the corresponding mRNA. A Combination of homo- and hetero-nuclear NMR spectroscopy and in vitro replication assay shows the presence of triple helix structures for this repeat. The probability of triple helix formation increases with an increase in the repeat length. It appears that the formation of such triple helices causes slippage during replication and repeat expansion. The formation of similar triple helices during transcription may also lead to the abrupt termination of the mRNA synthesis at the template site of the expanded intron resulting in a drastic reduction in the level of the frataxin mRNA.
Simulations of DNA Cyclization by Brownian Dynamics
Alexi Podtelzhnikov and Alexander Vologodskii
Department of Chemistry,
New York University,
4 Washington Place,
New York, New York 10003
The dynamic characteristics of long flexible polymer chains, such as the major relaxation time, tm, and the time of diffusion-controlled cyclization, ta, were studied using Brownian Dynamics simulations. Our major goal was to estimate the effect of impermeability of DNA backbone to itself (non-phantomity) on the dynamic properties. This property has never been taken into account in analytical treatments of the problem. We compared several chain models: phantom ones with excluded volume. The non-phantomity of a chain appeared to have small effect on the magnitude of tm. For sufficiently long chain tm exceeded the value for a phantom chain by factor of 1.5. The ratio ta/tm remained the same for the phantom and non-phantom models. From the other hand, the excluded volume effect changed the dependence of tm on the chain contour length, in accordance with theoretical conclusions. In addition, the presence of excluded volume caused two-fold increase of the scaling factor between ta and tm for sufficiently long chains. This scaling factor and, consequently, ta appeared to be reciprocally proportional to the radius of cyclization reaction, R. When the reaction radius is close to statistical segment, l, ta was estimated as ta=z l/R tm, where the coefficient z is equal to 0.8±0.1 and 1.6±0.2 for chains without and with excluded volume correspondingly.
DNA Replication in a Crystal: a High Resolution Snapshot of a Polymerase in Action
James Kiefer, Chen Mao and Lorena S. Beese
Department of Biochemistry,
Duke University Medical Center,
Durham, NC 27710
A thermostable DNA polymerase with optimal activity at 65°C has been isolated from a newly identified strain of Bacillus stearothermophilus. The amino acid sequence is 40% identical to E. coli DNA polymerase I and 43% identical to the thermostable DNA polymerase from Thermus aquaticus. The crystal structure of a 67.7 kD C-terminal fragment (Bacillus fragment; BF) was determined by the method of multiple isomorphous replacement including the anomalous scattering data from two heavy atom derivatives. The structure has been refined to an R-factor of 19.8% between 8Å and 17Å resolution (Rfree=26%) with excellent stereochemistry (0.01 Å rms deviation in bond lengths and 1.8° rms deviation in bond angles). The structure is folded into two domains and like Klenow fragment and HIV reverse transcriptase, the polymerase domain of this enzyme resembles a right hand. The very highly conserved "palm" subdomain where DNA synthesis occurs is almost identical to that of the Klenow fragment in both sequence and structure. This similarity permits the extrapolation of the results from years of site directed mutagenesis and biochemical studies on Klenow fragment to this polymerase.
BF polymerase has been crystallized in the presence of several duplex DNA primer templates, and the structures of these DNA complexes have been determined at high resolution. These polymerase-DNA co-crystals are highly ordered and diffract to better than 1.7Å resolution on a rotating anode X-ray source, revealing the most detailed view of any polymerase-DNA complex to date. The structure of the complex shows DNA bound to the polymerase active site. The conformation of the polymerase changes upon DNA binding. The most dramatic motions occur in the "fingers" and "thumb" regions as these subdomains adjust to accommodate the DNA in the polymerase cleft.
The BF polymerase retains catalytic activity in the crystals which has allowed us to successively incorporate several nucleotides. The polymerase both elongates and translocates the DNA in the crystals. This enables us to unequivocally determine the direction of DNA synthesis for Pol I polymerases by direct observation, resolving a recent controversy. In addition, the polymerase appears to retain the ability to discriminate between correct and incorrectly paired nucleotides in the crystal' we have been unable to find conditions where the polymerize crystal will incorporate mismatched bases. This high resolution structure of a catalytically competent polymerase active site with bound substrate, combined with site directed mutagenesis and kinetic studies, leads us to suggest a mechanism for the fidelity of DNA replication.
RNA-Protein Recognition in the HIV Rev-RRE Complex
J.L. Battiste(1), H. Mao(1), Thomas J. Tolbert(1), Kwaku
T. Dayie(1), R.Tan(2),
R. Sambasiva(3), L. E. Kay(3), A.D.Frankel(2) and J.R.Williamson(1)
(1)Dept. of Chemistry,
Massachusetts Institute of Technology,
Bldg 18-407, Cambridge, MA 02139
617-253-1837
617-258-7847 FAX
jrwill@mit.edu
http://volvox.mit.edu/Williamson/
http://web.mit.edu/afs/athena.mit.edu/org/c/chemistry/www/williamson.html
(2)Dept. of Biochemistry and Biophysics,
University of California, San Francisco CA 94143
(3)Depts. of Medical Genetics, Biochemistry, and Chemistry,
University of Toronto,
Toronto, Ontario, Canada M5S 1A8
We have characterized a 34 nucleotide RNA corresponding to the high affinity HIV Rev binding site in the RRE in complex with a 22 residue Rev-peptide, using 1H, 13C, and 15N NMR spectroscopy. The RRE adopts a helical structure containing G-G and G-A base-pairs, with an A bulge and a U bulge. The unusual geometry of these purine-purine pairs in the internal loop region significantly widens the major groove. The Rev-peptide forms an a-helix that binds in the opened major groove, and side chains make contacts with the bases and phosphates surrounding the internal loop. Some of the bases in the RRE are important for Rev binding because they are contacts to protein side chains, while some of the bases are mainly important for formation of the widened major groove structure.
Molecular-Dynamics Simulation of (dA·dT:dT)6 on Nanosecond Trajectories
S. Yu. Tsybenko, I. A. Il'icheva and V. L. Florentiev
V. A. Engelgardt's Institute of Molecular Biology
Russian Academy of Sciences,
Vavilov St. 32 Moscow B-334, 117984 Russia
A potential energy surface of three-stranded helices (dAdT:dT)6 both with antiparallel and parallel orientations of thymine chains has been investigated by means of minimization of atom-atom potential function and nanosecond molecular-dynamics simulation. All four possible types of H-bonding in triplets are investigated. The conformational prevalence of deoxynucleotide chains as well as the triplex stability is in grate extent dependent on mutual orientation of thymine chains.
A-form dominates in equilibrium population of antiparallel (dAdT:dT)6. The system spends about 92% of common time in this form. B-form exists only during 6% of the time. The possible variations of internal conformational and helix parameters and some correlations in intramolecular movements has been defined. Parameters of low energy structures are described for mono- and biregular helices. Characteristic interchain interactions, which are influenced by effective charges on phosphates, are capable to initiate local N to S transitions in different helical sites. It is very likely to be the cause of existing differences of NMR and FTIR conformational data.
Parallel-stranded (dAdT:dT)6 triplexes show conformational behavior, essentially different than antyparallel one. Repulsion of oligothymidyl chains provoke the displacement of the third chain thymines towards the X-axes and H-bonding with both bases of Watson-Crik duplex. In grate extent this kind of tripling is defined by the interactions between O2 thymine, belonging to the 3'-nucleotide of the Hoogsteen chain and O4 thymine, belonging to the 5'-nucleotide of the Crik chain. Every of the three chains are characterized by its own conformational behavior. Adenylic chain preferred A-type of conformation (95% of the registration time). Both tymine chains spent approximately the same time as in A- so and B-conformation, but sugars in Hoogsteen chain much more time spent in E-conformation region. Moreover, conformation of this chain is not regular in energy minima regions. None the less, helical parameters, namely helical pith, is rather conservative and does not influenced by dynamical conformational alterations. Data obtained allowed us to suppose that stability of parallel triplexes dA·dT:dT will diminish with the increasing of their length.
Estimation of enthalpy on nanosecond trajectories for parallel
and anthyparallel triplexes (dAdT:dT)6 correlates well with UV data on Tm,
earlier obtained for three-stranded clips.
Heterogen Complexes of Ethidium Bromide and Their Role in Stabilization of Helix DNA
Ara P. Antonian(1), Pogos O.Vardevanian(1), Genady A.
Terzikian(1),Armen T. Karapetian(2) and Olga F. Borisova(3)
(1)Biophysics Department,
Yerevan State University,
Yerevan 375049, Armenia
(2)Physics Department,
Yerevan Institute of Architecture and Construction,
Yerevan 375009, Armenia
(3)V.A. Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Moscow, 117984, Russia.
Existing experimental and theoretical data show that Ethidium Bromide (EtBr) forms several types of complexes with DNA. Nevertheless it is not quite clear what type of interaction plays the major role in stabilization of the helical structure of DNA. Here we present the results of spectrophotometric (absorption and flourometric methods) investigations which lead to the conclusion that thermostability of poly(dA)-poly(dT) - ligand complexes increases due to intercalated EtBr. The temperature of helix-coil transition (Tm) of the complexes increases by 14 degree C at the ligand : base ratio 1:5, where only intercalated EtBr exists in the complexes.
On the other hand the comparison of theory with experiment predicts that EB may form at least two types of complexes with coiled DNA. We are going to present the experimental data showing the existence of these types of interaction of the ligand with single stranded DNA. Thermodynamic parameters of the interaction are also calculated.
New Methods for the Free Energy Calculation with Implicit Solvent Model
Yury N. Vorobjev and Jan Hermans
Department of Biochemistry and Biophysics,
University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599
E-mail: vorobjev@femto.med.unc.edu
The implicit solvent HC-CD (Hydrophobic Cavity - Continuum Dielectric) model has been shown consistent and reliable. The model calculates the free energy of creation of a hydrophobic cavity in water solvent, and the free energy of electrostatic polarization of water solvent by a solute molecule. Computational reliability of the HC-CD method depends crucially on how the Molecular Surface is calculated and how the Poisson equation for the solute molecule is solved. New method (SIMS, Smooth Invariant Molecular Surface) for calculation of a smooth invariant molecular surface which envelops the solvent-excluded volume, and new method (FAMBE, Fast Adaptive Multigrid Boundary Element) method for solution of the Poisson equation are described. It is demonstrated that the CPU time for the SIMS and FAMBE methods is scaled linearly with the number of atoms of the solute molecule. SIMS is superior to Connolly's MS program: it is faster, more accurate and more stable, and smoothes singularities of the MS. The FAMBE method demonstrates good stability when applied to a calculation of the solvation free energy for different conformations of a solute molecule, e.g., along 200 ps molecular dynamic trajectory of protein eglin. We propose a new method of calculation of the free energy difference between two macroscopic conformations of a biomolecule solute, which takes into account conformational flexibility as a sum of differences in internal energy, solvation free energy and internal entropy.
The programs SIMS and FAMBE are available from athor Y.V. on request: (vorobjev@femto.med.unc.edu).
Beta-Sheet Assembles DNA's Homo- and Heterologous Quadruplexes
Sergei A. Streltsov, Marina V.Borodina and Tymur E.Semenov
V.A.Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Vavilov str. 32, Moscow B-334, 117984 Russia
We have first shown the formation heteroquadruplex structure containing two molecules calf thymus dsDNA complexed with trivaline tetramer (beta-sandwich) [1]. We used analog of trivaline - (L-Val)3-N2H2-DNS CF3COOH (DHTV) - which formed dimers in water methanol mixtures [2]. The formation of beta-sandwich-dsDNA complex is accompanied by the significant increase in the fluorescence intensity of the dansyl label covalently bound to the peptide molecule [3].
In the present work we studied formation of heterologous and homologous quadruplexes upon binding of the peptide to calf thymus and linearized plasmid pBR322 dsDNA, respectively, using flow linear (LD), circular dichroism (CD), fluorescent methods and electron microscopy [4]. It was shown that:
1. Formation a complex between DHTV beta-sandwich on homo-
and hetero dsDNA. It follows from the fact that there is S-shape dependence
of the DHTV fluorescence intensity versus DHTV concentration in the presence
of homo- and hetero- dsDNAs (titration curves). DHTV titration curves have
different midpoints and saturation levels for homo- and hetero- dsDNAs.
2. Consistent with the formation of quadruplex structure are the nonlinear
dependence of the LD amplitude for complex DHTV in the beta-sandwich form
with dsDNA on dsDNA concentration at saturated concentration of DHTV, and
the appearance of the homogeneous thickness compact rod-like particles on
electron microscope. The homo- and heteroquadruplex structures diameters
(were contrasted by rotary shadowing with Pt:Pd) were equal one another
and coincided with those found previously [1]. The diameter heteroquadruplex
structure contrasted with uranyl acetate staining was 6-8 nm [1].
3. The CD amplitudes in the spectral region of 300-420 nm, measurements
of fluorescence intensity - exitation at 400, emission at 500 nm as well
as LD amplitude at 265 nm for DHTV-DNA mixtures are reflected of dissociation
of DHTV-dsDNA complexes under 30 mM NaCl concentration is added.
4. Under these conditions homoquadruplexes retain their morfology, whereas
heteroquadruplexes disappeared in the presence of 20 mM NaCl as revealed
from electron microscopy studies.
We conclude:
1. The quadruplexes formed upon coalescence of linear homologous
plasmid dsDNA molecules complexes with DHTV were stable under the conditions
when practically all peptide molecules are removed from DNA.
2. The quadruplexes containing linearized plasmid pBR322 DNA and quadruplexes
containing calf thymus DNA exhibit differences in their structural organization.
3. DHTV-DNA complexes may serve as a model of RecA-DNA interaction.
References and Footnotes
1. Makarov, V.L., Streltsov, S.A., Vengerov, Yu.Yu., Khorlin,
A.A., Gursky, G.V., Molekularnaya Biologiya (Rus.) 17, 1089-1102.
1983.
2. Streltsov, S.A. and Gursky, G.V., J. Biomol. Struc. Dynam. 13,
N6, a231 1995.
3. Gursky, G.V., Zasedatelev, A.S., Zhuze, A.L, Grokhovsky, S.L.,Streltsov,
S.A., Surovaya, A.N., Nikitin, S.M., Retchinsky, V.O., Michailov, M.V.,
Beabealashvili, R.S. and Gottikh, B.P., Cold Spring Harbor Symp.
Quant. Biol. XLVII, 367-378 1983.
4. Streltsov, S.A., Borodina, M.V. and Semenov, T.E., J. Biomol. Struc.
Dynam. 14, 357-363 1996.
Spectroscopic Studies of
the Stability Enhancement of Selective
5-Substituted Cytidine and Hydration on the Conformations of DNA Triple Helices Containing a TTTT Loop
Lou-sing Kan(1), Ye Fang(2), Chunli Bai(2) and S. B.
Lin(3)
(1)Institute of Chemistry,
Academia Sinica, Taipei, Taiwan
(2)Institute of Chemistry,
Academia Sinica, Beijing, China
(3)The Graduate Institute of Medical Technology,
National Taiwan University, Taipei, Taiwan
5-Methyl- and 5-bromocytidines have been introduced into triplex-forming-oligonucleotides, 5'-d(TC)3T4(CT)3 ([CC]) that was designed to form a hairpin triplex with a 5'-dA(GA)2G ([AG6]) purine strand at acidic pH. The formation and conformation of the hairpin triplex as a function of the selective substitution of cytidines by 5-methyl- and 5-bromocytidine (M and B) in [CC], respectively, were studied by UV and FT-IR. Namely, cytidines are replaced in either the 3'-pyrimidine portion ([CM] and [CB]) or the 5'-pyrimidine portion ([MC] and [BC]) or in both ([MM] and [BB]) of the [CC]. The acidic-induced transitions of the equimolar mixtures of [AG6] with [MM], [MC], or [CM] gives rise to different apparent pK values, i.e., [MM].[AG6] (6.2) > [MC].[AG6] (6.0) > [CM].[AG6] (5.7) > [CC].[AG6] (5.2) > single-stranded oligopyrimidines (4.6 ± 0.2), indicating that cytosine methylation expands the pH range compatible with the hairpin triplex formation regardless of whether the substitution is in the 5'-pyrimidine (Hoogsteen) portion or in the 3'-pyrimidine (Watson-Crick) portion. Thermal denaturation profiles of 5-bromocytidine substituted [CC] indicated that [BB].[AG6] > [BC].[AG6] > [CB].[AG6] > [CC].[AG6] at slightly acidic conditions. Thus, substitutions of 5-methyl- and 5-bromocytidine in any position of [CC] increase the stability of the hairpin triplex DNA. IR spectra revealed the conformation of sugar conformation of triplexes are predominately in a S-type sugar pucker but with a relative proportion of two S-type sugars for one N-type at low humidities.
Identification and Characterization of Genomic Nucleosome-Positioning Sequences
Hans R. Widlund(1), Hui Cao(1), Stina Simonsson(2),
Elisabet Magnusson(3), Tomas Simonsson(1), Peter E. Nielsen(4), Jason D.
Kahn(5), Donald M. Crothers(6) and Mikael Kubista(1)
(1)Department of Biochemistry and Biophysics, The Lundberg Institute,
Chalmers University of Technology and Göteborg University, S-413 90
Göteborg, Sweden
(2)Department of Medical Biochemistry and Microbiology,
Göteborg University, S-413 90 Göteborg, Sweden
(3)Department of Genetics, The Lundberg Institute,
Göteborg University, S-413 90 Göteborg, Sweden
(4)Center for Biomolecular Recognition,
Department of Medical Biochemistry and Genetics, Panum Institute, DK-2200
Copenhagen N, Denmark.
(5)Department of Chemistry and Biochemistry,
University of Maryland, College Park, MD 20742-2021
(6)Department of Chemistry,
Yale University, New Haven, CT 06511
Positioned nucleosomes are believed to play important roles in transcriptional regulation and for the organisation of chromatin in cell nuclei. In this work we have isolated the DNA segments in the mouse genome that form the most stable nucleosomes yet characterised. In separate molecules we find phased runs of 3-4 adenine nucleotides, extensive CA repeats, and in a few cases phased TATA tetranucleotides. The latter forms the most stable nucleosome yet characterised. One sequence with CAG repeats was also found. By fluorescence in situ hybridisation the selected sequences are shown to be localised at the centromeric regions of mouse metaphase chromosomes.
GAT(G)-Repeats Form Unstable Nucleosomes
Hui Cao, Hans R. Widlund and Mikael Kubista
Department of Biochemistry and Biophysics, The Lundberg Institute,
Chalmers University of Technology and Göteborg University, S-413 90
Göteborg, Sweden.
Nucleosomes, the building blocks of chromatin, are responsible for the packaging of DNA in the cell nucleus. They play a structural role for condensation of the genome and act also as regulators of transcription and replication, by direct and indirect mechanisms. The properties of the DNA sequence, such as flexibility and static bending, direct the location of nucleosomes both in vitro and in vivo. Many DNA sequences that position nucleosomes have been identified and rules that govern the properties have been formulated. However, there have been no studies of nucleosome defiant sequences. Such sequences could loosen chromatin structure and enhance the transcriptional activity. Here we identify sequences that refrain from forming nucleosomes by selecting them from a large pool of synthetic DNA with a central region of 146 random base-pairs enclosed by adapters. These were used in an in vitro salt-induced reconstitution of nucleosomes under thermodynamic equilibrium conditions. The sequences that were not incorporated into nucleosomes were purified, amplified by polymerase chain reaction (PCR), and then the procedure was repeated. After 17 rounds of selection the material was enriched in sequences reluctant to form nucleosomes. Cloning and sequencing revealed that 35% of the molecules have long repeats of GATG and in some cases GAT nucleotides. Their affinity for histone octamers is about half of that of average nucleosome DNA.
Mouse Major Satellite DNA in a Positioned Nucleosome
Hans R. Widlund(1), Prasad N. Kuduvalli(2), Thomas D.
Tullius(2) and Mikael Kubista(1)
(1) Department of Biochemistry and Biophysics, The Lundberg Institute,
Chalmers University of Technology and Göteborg University, S-413 90
Göteborg, Sweden
(2) Department of Chemistry,
The Johns Hopkins University, Baltimore, Maryland 21218
Nucleosomes and their relation to the incorporated DNA sequence is of immense importance for the understanding of gene regulation, replication and higher order chromatin organisation. In a recent study we identified mouse genomic nucleosome sequences with a high affinity for forming nucleosomes in vitro. In that study one group of sequences was identified as mouse major satellite sequences, which are an important constituent of the DNA within the mouse chromosome centromers. This specific satellite DNA is characterised by having an abundance of AAA/TTT regions in a phase, near the helical repeat of the DNA. Here, we further analysed this mouse major satellite sequence rotational and translational positioning within a nucleosome. In vitro reconstituted nucleosomes on a molecule containing a 122 base pair (bp) mouse major satellite was found to occupy one main and one minor translational setting as measured by high resolution electrophoretic mobility shift assay (EMSA). By using the hydroxyl radical to probe the secondary structure of the DNA in the nucleosome we found that the AAA/TTT regions were either going from the outside to inside or the opposite in-to-out in the 5' to 3' direction. This is not in agreement with earlier observations of locations of A/T rich stretches found were the minor groove was facing inwards towards the histone octamer.
Genome Instability: Structural Basis of Triplet Repeat Expansion and Genetic Disorders
Samir K. Brahmachari, P.S. Sarkar, U. Shaligram, M.
Mutsuddi, S. Raghavan, R. Bhandari, S.S.Pataskar, M. Narayan and Quasar,
S.P
Molecular Biophysics Unit,
Indian Institute of Science,
Bangalore 560 012, India
Several years ago we initiated studies with the belief that chromosomal DNA will exhibit sequence dependent structural variability and majority of the repetitive sequences in the eukaryotic genome will manifest such a property[1,2]. Since then we have developed several biophysical, molecular biological and computational tools to establish that simple repetitive sequences depending on length, position and cellular context could manifest structural variability and regulate level of gene expression in vivo.[3,4 ]
The recent discovery of a novel type of mutation, where intragenic amplification of trinucleotide repeat leads to genetic disorders, has established the importance of repetitive sequences in genome organization and function. In this paper, we have addressed the two major questions raised by this phenomenon from a structural point of view :
1) How does the triplet repeat length affect expression and function of the genes and proteins? Why are all these diseases genetically dominant?
2) What is the mechanism of triplet repeat expansion?
Consequences of Triplet Repeat Expansion
To understand how the expansion of triplet repeat sequences influences gene expression we have synthetically constructed model systems for Myotonic Dystrophy where a step-wise increase of CTG repeat length is made in frame with lac Z gene in both E.coli and yeast vectors. We find that the in vivo expression of b-galactosidase decreases with increase in CTG repeat length due to the reduced level of full length transcript formation. Thus secondary structure of the DNA template or mRNA could be responsible for transcription elongation block.In several neurodegenerative disorders the polyglutamine tract (coded by CAG repeats) expansion has been implicated in the disruption of the normal gene function. To delink the effect of glutamine tracts from the probable cis regulation of the expanded CAG repeats, we designed vectors that encode polyserine, but have CAG repeats in the coding strand. These constructs also showed down regulation of transcription with increasing repeat length thus suggesting an additional cis acting role for these repeats in repression of transcription elongation. Autosomal dominant nature of these diseases have been so far explained on the basis of
a) a gain of function mediated by increase in polarity
leading to aberrant protein-protein interaction;
b) precipitation of the glutamine rich proteins due to self aggregation
We propose the following as equally probable explanations:
i) Repeat length dependent
cis regulation leading to reduced level of gene expression below a specific
threshold.
ii) Genome sequence analyses of CAG/CTG repeats suggest that (CAG)n-mRNA
can associate with other (CTG)n-mRNA in the cell thus leading to disruption
of gene function.
iii) Analysis of CAG repeat length variation at MJD locus for normal
and affected individuals show large differences in allele sizes of CAG repeats.
If dimers are the functional units, then presence of large difference in
glutamine stretches in monomers could manifest dominant disease status due
to formation of asymmetric inactive dimers.
Mechanism of Triplet Repeat Expansion
We have earlier proposed a mechanism for CNG repeat expansion via several intermediate metastable structures, such as slipped loop, hairpin and folded hairpins[5]. The process of triplet repeat expansion and stabilization of such a structure would then depend on the nature of the repeat as established by structural studies on (CTG)n and (CAG)n oligonucleotides by us and others. However in the presence of complementary strand, all oligonucleotides form stable duplex structures. The process of expansion of natural sequences in vivo could occur in two steps : i) nucleation - through bubble formation and initiation of hairpin formation ii) propagation - expansion of the hairpin through branch migration. Such hairpin structures could easily form when the double helix is unwound during replication or recombination.We observe that synthetic (CAG/CTG)12-48 shows faster anomalous mobility in PAGE while PCR products of various CAG/CTG length with about 50bp of flanking sequences from different genomic loci exhibit multiple slow moving electrophoretic species implying the importance of sequence context. Such multiple looped out structures are observed only in PCR products with more than 16 CAG repeats. Extent of loop formation increases with increase in CAG repeat length.
Genome sequence analysis for CAG/CTG repeats show that CAA triplet interruptions which could destabilize hairpin formation, are frequent at several loci thereby preventing expansion[6]. Long CAG/CTG repeats in E.coli exhibit spontaneous deletion in multiples of trinucleotides. The occurrence of point mutations within the triplet repeats and the stabilization of the triplet repeat length at a critical threshold appear to be correlated. The reversion of CAG repeat interrupting CAA or CAT sequences to CAG through one base change have been observed in several Spino-cerebellar ataxia patients. These suggest mismatch hairpin structure formation as a mechanism for expansion during replication. This is further supported by our observation that the frequency of spontaneous CAG expansion at two loci is enhanced by a thousand-fold in a Drosophila mutant that is deficient in post-replication DNA repair.
These observations support our model since stability of the hairpin formed during replication will depend on the extent of mismatch as well as the flanking sequences. The more stable the hairpin, higher the probability of expansion and lower the level of expression.
It is likely that unequal crossover during recombination could push the CAG repeat length to a threshold making it prone to replication slippage. Thus transient formation of unusual DNA structures is the first step in the mechanism of expansion of triplet repeats. This could also explain why similar repeats show different extent of expansion depending on their location in the genome and the frequency of CAA or CAT interruptions.
References and Footnotes
1. Samir K. Brahmachari, et al., Structures &
Methods, Vol 2, Eds. R.H.Sarma & M.H.Sarma, Adenine press (1990)
2. Jalaj Tripathi and Samir K.Brahmachari, J. Biomol. Str. Dyn. 9,
387 (1991).
3. P.S.Sarkar and Samir K.Brahmachari, Nucl. Acid Res., 20,5713 (1992)
4. Samir K.Brahmachari, et al., Gene, In press (1997)
5. Samir K.Brahmachari, et al., Electrophoresis 16, 1705,
(1995).
6. Rashna Bhandari and Samir K.Brahmachari, J.Biosci. 20, 613 (1995).
Analogues of Hoechst 33258 Bind to Transfer RNA with 1:1 Stoichiometry
S.E. Sadat Ebrahimi, Amanda N. Wilton and Kenneth T.
Douglas
School of Pharmacy and Pharmaceutical Sciences,
University of Manchester, Manchester, M13 9PL,U.K.
Early studies indicated netropsin and distamycin to bind strongly to the DNA minor groove but weakly to RNA (probably electrostatically to the phosphate backbone). However, classical DNA minor-groove ligands can interact strongly with RNA duplexes based on Tm and other studies1,2. However, these studies gave no indication of RNA-specifying interactions and the grooves of DNA and RNA differ in many respects. An X-ray diffraction study of tRNA with netropsin and distamycin showed specific hydrogen bonds and electrostatic interactions with phosphates but binding strengths were not measured3 . Bernard Pullman then detected this netropsin-binding site on tRNAphe by theoretical calculations 4. Hoechst 33258 binds strongly to B-DNA in AT-rich regions but the absolute sequence specificity is low, with multiple sites on calf thymus DNA. We report now that there is strong, specific binding to a unique site on tRNA of analogs of the Hoechst structure and of other benzimidazole-based species. In the UV-visible absorption spectra of mixtures of tRNA (from brewers' yeast) and trihydroxy-Hoechst the lambda max value of the ligand shifted from 326nm with increasing tRNA concentration, becoming 338nm for a 1:1 molar mixture. Spectra were obtained for a series of tRNA : trihydroxy-Hoechst mixtures, varying the molar ratio of the components but keeping a constant total concentration of the components. The absorbance at 360nm for each ratio was corrected for unbound ligand or tRNA assuming 1:1 tight binding. These corrected absorbances were plotted against the mole fraction of ligand as a Job plot which showed 1:1 complex formation and tight binding. Addition of tRNA to solutions of many of the analogs led to fluorescence quenching. This could be used to measyre binding constants, the ease of study depending on the strength of the native fluorescence of the free ligand. A sample of 3,4-dimethoxy-Hoechst in 0.1M sodium phosphate buffer, pH 7.80 was excited at 338nm and its fluorescence emission spectrum recorded between 300 and 600nm after each addition of a series of aliquots of tRNAPhe ( 20µl from 100µM stock). A plot of the change in the fluorescence intensity, corrected for dilution, versus [tRNA Phe] added, fit the equation for single-site binding (Fluorescence (F) = Fmax.[tRNAPhe]/(Kdiss+[tRNAPhe])), with Kdiss =14.8 + 0.9 µM. Other, much weaker binding site (s) were also detected. The molecular environment of these ligands on the RNA is differs different from that on DNA as the fluorescence of Hoechst 33258 is enhanced enormously on binding to the DNA minor groove but strongly quenched by tRNA. The molecular structures of this family of ligands and their specific recognition of a particular region of tRNA (they selectively occupy a unique run of 4-5 base-pairs out of approximately 75 bases) provides a new lead structure to the molecular basis of ligand:RNA recognition at an atomic level. The compounds also provide a rare example of stoichiometric ligand recognition by tRNA in a readily reversible equilibrium as opposed to covalent attack.
References and Footnotes
1. Pilch, D.S., Kirolos, M.A., Liu, X., Plum, E. & Breslauer, K.J.
Biochemistry (1995) 34, 9962-9976
2. Wilson, W.D., Ratmeyer, L., Zhao, M., Srekowski, L. & Boykin, D.
Biochemistry (1993) 32, 4098-4104
3. Rubin, J. & Sundaralingam, M. J. Biomol. Struct. Dynamics 2,
165-174 (1984).
4. Zakrewski, K. & Pullman,B. J. Biomol. Struct. Dynamics 2,
737-743 (1984).
The Bi-loop, a New General Four-Stranded DNA Motif
Stephen A. Salisbury(1), Sarah E. Wilson(1), Harold
R. Powell(1), Olga Kennard(1), Paolo Lubini (2), George M. Sheldrick (2),
Nuria Escaja (3), ElMostafa Alazzouzi (3), Anna Grandas(3) and Enrique Pedroso
(3)
(1) Cambridge Crystallographic Data Centre,
12 Union Road, Cambridge CB2 1EZ, UK
(2) Institut für Anorganische Chemie der Universität Göttingen,
Tammannstrasse 4, D-37077 Göttingen, Germany
(3) Departament de Química Orgànica,
Universitat de Barcelona, Martí i Franquès 1-11, 08028 Barcelona,
Spain
X-ray crystallography of oligonucleotides is an unrivalled source of detailed information about nucleic acid structure. Intermolecular packing effects may contribute to the conformations observed, but results obtained nevertheless represent genuine points on the potential energy surface. Such structures may be relevant to biological processes and therefore provide a starting point when direct investigation at the molecular level is hampered by the complexity of these systems. The crystal structure of the cyclic octanucleotide d<pATTCATTC> contains a new motif which other evidence suggests could occur generally in DNA and which may be relevant in vivo.
The molecule crystallises as a nearly symmetrical dimer, stabilised by Watson-Crick A.T base pairs. Despite conformational restraints imposed by the the cyclic backbone, the A-T dinucleotides adopt geometries typical of B-DNA. Two small fragments of duplex, each with its component strands from different molecules, interlock via their minor grooves to form a compact unit, with the remaining thymines stacked above and below each section of double helix.
Schematic representation of the asymmetric unit. Watson-Crick
base pairs are formed between residues: A1, T12; A11, T2; A5, T16; and A15,
T6.
A closely similar structure was earlier observed for d(GCATGCT) [1] which, rather than crystallising as a duplex, assumed a dimeric, looped conformation that is almost superimposible with part of the cyclic octamer structure. The occurrence of this bi-loop arrangement in two quite dissimilar oligomers and with both G.C and A.T base pairs suggests that it may be a generally accessible motif. It could be induced in polymeric DNA by protein binding or applied torsional stress. Some sequence preference can be inferred since stacking overlap would be much less favourable when the base-pairs are reversed, i.e. the sequence is 5'-pyrimidine-purine-3'. Identical sequences in both portions of duplex make possible attractive interactions between functional groups of their minor grooves: when the unit is composed of all G.C pairs, four N2..H..O2 H-bonds form; while with A.T in the cyclic octamer, the four O2 atoms of thymine coordinate a metal cation. Sequences meeting these requirements are so short that statistically they are expected to be common in genomic DNA, but the actual occurrence in sequence databases is found to be much greater. While this is not evidence of a connection with bi-loop formation, it does indicate widesperead opportunity for formation of these structures.
Bi-loops can stack head-to-head, as is illustrated in the the crystal structure of d(GCATGCT), bringing the sugar-phosphate backbones of the two units into close proximity. Transesterification between the closest pairs of phosphodiester linkages in such an assembly could provide a molecular mechanism for recombination of double-stranded DNA in vivo.
Reference and Footnotes
1. G.A. Leonard, S. Zhang, M.P. Peterson, S.J. Horrop, J.R. Helliwell, W.B.T. Cruse, B. Langlois d'Estaintot, O. Kennard, T. Brown and W.N. Hunter. Structure 3, 335-340, 1995.
Role of the C-terminal Zinc Finger in the Recognition of Damaged DNA by Fpg, a DNA Repair Enzyme
Serge Smirnov(1), Carlos de los Santos(1), Teri Klein(2),
Arthur Grollman(1) and Moises Eisenberg(1)
(1)Dept. of Pharmacology,
SUNY at Stony Brook, Stony Brook, NY 11794
(2)Computer Graphics Laboratory,
University of California,
San Francisco, CA 94143
Fpg, a DNA glycosylase repair enzyme 269 residues long, is responsible for removing DNA lesions including those containing abasic sites [2] and oxidative damage as in 8-oxoguanine [3]. The enzyme has a (Cys)4-Zinc Finger motif 30 amino acid sequence at the C-terminus of the polypeptide chain. It has been shown that a mutation of any one of these four Cys residues causes the loss of its enzymatic activity[1]. These data is consistent with the generally accepted role of Zinc Fingers as DNA-bindig domains. In the present study we conducted a computer simulation of the interaction between the Fpg's Zinc-Finger with a 12-base pair DNA sequence. Our aim is to understand the role of each different structural component of the Zinc Finger (beta-sheet, loop, alpha-helix) in the binding to DNA. The results we show represent the end product of a 150 psec molecular dynamics of the complex DNA-Zinc Finger in vacuum followed by a 100 psec of a free molecular dynamics of the complex in a water box with positively charged counterions. The initial structure of the complex was built by analogy modeling from experimentally determined DNA-Zinc Finger complexes structures from NMR or X-ray crystallography data. The 3D structures of the DNA-Zinc Finger complex obtained from the simulation after the in vacuo and after in water are presented. The role of water molecules as well as conserved residues of the Zinc Finger sequence in the stabilization of the complex are described.
References and Footnotes
1. Tchou, J., Michaels, M.L., Miller, J.H. and Grollman,
A.P. Function of the Zinc-finger in Echerichia coli Fpg Protein. J. Biol.
Chem. 269, 26738-26744 (1993).
2. Baily, V., Verly, W.G., O'Connor, T.R and Laval, J. Mechanism of DNA
Strand Nicking at Apuric/apyrimidinic Sites by Escherichia coli [formamidopyrimidine]
DNA Glucosylase. Biochem. J. 262, 581-589 (1989).
3. Tchou, J., Kasai, H., Shibutani, S., Chung, M.H., Laval J., Grollman,
A.P. and Nishimura S. 8-oxoguanine (8-hydroxyguanine) DNA Glycosylase and
Its Substrate Specificity. Proc. Natl. Acad. Sci. USA 88, 4690-4694
(1991).
Structure of Mispairs Formed by 2'-Deoxy-6-Thioguanosine in DNA
Rahul Mitra and B. Montgomery Pettitt
Department of Chemistry,
University of Houston,
Houston, TX 77204-5641
2'-Deoxy-6-thioguanosine dGs is a therapeutic agent in the treatment of childhood leukemia. The antitumor properties of dGs are primarily due to the incorporation into DNA and disruption of the metabolic activities of the DNA. Mispairs formed by the dGs with the bases of the DNA determine the fate of the cell, in a dose dependent manner.
The high rate of misinsertion of dGs against Cytosine results in a canonical type of wobble pair, and thus escapes proofreading, to form a lesion. dGs mispairs with a Thymine in the next replicative cycle and invokes the post-replicative repair systems. Failure of the repair systems results in the formation of a single strand gap and a delayed cytotoxic effect. In order to explain the failure of the repair system the molecular structure of the mispair T.dGs has been studied in atomic detail, in different nearest neighbor contexts. The effect of a 5'Adenine on the extent of local perturbation of structure was found to be greater on a T.dGs than a (T.G) mispair. Thymine in the T.dGs with a 5'A was destacked and migrated to the major groove, invoking the repair enzymes. The high bias in the misinsertion of a T against a dGs, causes repeated cycles of excision and insertion of T against a dGs, resulting in the failure of the repair enzymes and single strand gaps. This allowed us to formulate conjectures on the sequence dependency of the stability of the mispairs and the dosage index of the therapeutic.
Inhibition of the HIV-1 Integrase by DNA Quadruplexes is Dependent on Their Structure
Rahul Mitra(1,2), Sandy Lee(2), Michael E. Hogan(2)
and B. M. Pettitt(1)
(1)Department of Chemistry,
University of Houston,
4800 Calhoun, Houston, TX 77204-5641
(2)Department of Molecular Physiology & Biophysics,
Baylor College of Medicine,
Houston,TX 77030
A novel class of quadruplex forming oligonucleotides comprising of only deoxyguanosine and thymine were successfully used to inhibit the human immunodeficiency virus type-1 (HIV-1) integrase function both in vitro and in vivo. In this study we demonstrate that the inhibition of the HIV-1 integrase by the quadruplex inhibitors is dependent on their structure. Competitive binding with 3'-azido-3'-deoxythymidine (AZT) proves that these quadruplexes bind to the catalytic core of the integrase. Variation in the tertiary structure affects binding of the quadruplexes to the catalytic core of the HIV-1 integrase. Subtle variations in the structure result in at least a 100 fold difference in the binding affinity. Only one of the inhibitors tested (T30695) forms a 1:1 complex with HIV-1 integrase in the sub-micromolar range, which parallels the inhibitory concentration. We also propose the structure of the complex based on these results of binding of quadruplex inhibitors to the HIV-1 integrase.
Protein Displacement By Sequence-Directed DNA-Binding Molecules: The MERLINTM Assay System for Drug Discovery
Lisa M. Turin, Alka Chawla, Gary P. Schroth and Cynthia
A. Edwards
Genelabs Technologies,
505 Penobscot Drive,
Redwood City, CA 94063
MERLINTM is a novel assay system that profiles the sequence binding preferences of small organic molecules for their effectiveness in displacing proteins from DNA. This approach is unique because it does not target potential drug molecules towards specific proteins or DNA-binding sites for specific proteins. Instead we take the unique approach of targeting DNA-binding molecules to DNA sequences that are adjacent to a protein binding site. These flanking sequences are preferred drug targets since they have the potential to be gene-specific. We have shown that DNA-binding proteins can be displaced by different types of drugs bound to these flanking sequences. Displacement of bound protein occurs through either a steric or allosteric mechanism. This information can be applied to the design of DNA-binding molecules targeted to specific DNA sequences for the regulation of genes involved in disease processes. This approach offers the potential for creating an entirely new class of gene-specific therapeutics.
Tet Repressor Binding to Operator DNA and Tetracycline Studied by Infrared and Raman Spectroscopy
C. Krafft(1), W. Hinrichs(2), P. Orth(2), W. Saenger(2)
and H. Wefle(1)
(1)Max-Delbrück-Centrum für Molekulare Medizin,
D-13125 Berlin, Germany
(2)Institut für Kristallographie, Freie Universität Berlin,
D-14195 Berlin, Germany
Tet repressoe (TetR) is involved in the most abundant mechanism of tetrcyclie (Tc) resistance of Gram-negative bacteria(1). TetR regulates the expression of an intrinsic membrane protein TetA that mediates the transport of the antibiotic out of the cell before it can inhibit ribosomal activity. Two homodimeric TetR molecules bind to two tandemly oriented DNA operator regions of the resistance determinant named 01 and 02, thereby blocking the expression of two genes, one encoding for TetA and the other for TetR. The DNA binding properties of TetR were extensively analyzed(2). The crystal structure of the complex between TetR of class D (TetRD), Tc and a Mg2+ ion was determined by multiple isomorphous replacement at 2.5Å resolution 3,4 TetRD forms stable homodimers, the polypeptide chain of a nonomer being folded into ten alpha-helices. The N-terminal three-helix bundle of each monomer represents the DNA binding domain with typical alpha-helices-turn-alpha-helix motif. In the TetRD [Mg-Tc]1 complex, the two recognition helices are 39Å apart. Since the distance between two consecutive major grooves of B-DNA is only 34Å, the induced TetRD is unable to attach to the operators 01 and 02.
We measured Raman spectra of a recombinant TetRD protein, of 17mer and 18mer oligonucleotides with sequences corresponding to operator site 01, and of the TetRD oligonucleotide complex. Comparison of the spectra reveals evidence for specific contacts between TetRD residues and DNA bases: (i) intensity decrease connected with complex formation at thymine band 1376 cm-1 indicates the participation of thymine methyl groups, (ii) intensity increase at guanine bands 1467 and 1490 cm-1 is correlated with changes in the hydrogen bonding state of guanine N7 atoms, (iii) increased intensity at 728 cm-1 and 778 cm-1 shows involvement of adenine and cytosine, (iv) decreased intensity ratio I854/I823 signalizes changes of the tyrosine environment, (v) an increased peak at 1363 cm-1 is a market for a more hydrophobic tryptophan ring environment in the complex. Raman and infrared measurements indicate nearly identical conformations of TetRD with and without [Ni-Tc]1.
The results support a molecule model of the TetRD operator complex proposed from genetic experiments.
References and Footnotes
1. Hillen, W. and Berens, C., Ann. Rev. Microbiol. 48,
345-369, 1994.
2. Heble, V., Berens, C. and Hillen, W., J. Mol. Biol. 245, 538-548,
1995.
3. Hinrichs, W., Kisker, C., Düvel, M., Müller, A., Tovar, K.,
Hillen, W. and Saenger, W., Science 264, 418-420, 1994.
4. Kisker, C., Hinrichs, W., Tovar, K., Hillen, W. and Saenger, W., J.
Mol. Biol. 247, 260-280, 1995.
5. Baumeister, R., Hebl, V. and Hillen, W., J. Mol. Biol. 226, 1257-1270,
1992.
Mutational Analysis of Active Site of RNase Bacillus Intermedius (Binase)
N. Struminskaya(1), G. Yakovlev(1), G. Moiseev(1), L.
Kipenskaya(2), L. Znamenskaya(2), I. Leschinskaya(2) and R. Hartley(3)
(1)Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Moscow 117984
(2)Kazan State University,
Kazan, Russia
(3)National Institute of Diabetes and Digestive and Kidney Diseases,
NIH, Bethesda, MD 20892
To investigate the functional role of some residues in the active site of binase we made the mutations as follows: Arg86Ala, Arg82Ala, His101Glu, Asp53His, Lys26Ala and Lys26His. Mutant enzymes have been expressed in E. coli, purified to homogeneity and characterized by kinetic and spectroscopic methods. On cleavage of various substrates the catalytic activities of binase mutants Arg86Ala and His101Glu are 0.01-0.03% and 2.0-2.3% of that for native enzyme, respectively. The decreases in activity are determined mainly by the decreases in molecular rate constants kcat, with almost unchanged affinities of the enzyme for the substrates, characterized by Km. These are the expected results if Arg86 participates in electrostatic stabilization of the transition state of the binase-substrate complex and His101 acts as general acid, donating a proton to the leaving group on cleavage of a phosphodiester bond. The replacements of Lys26 by Ala or His cause comparatively low reductions in the enzyme activity. The substitution of His for Asp53 does not influence the kinetic parameters of poly(I) and poly(A) but one decreases kcat 100-fold for GpC. It shows that Asp53 belongs the subsite of the binase active site which binds a nucleoside at the 05'end of the scissle phosphodiester bond.
Computer Simulation of Codon - Codon Replacements
B. Borstnik
National Institute of Chemistry,
Hajdrihova 19, 1001 Ljubljana, Slovenia
An algorithm and the corresponding computer program were prepared which enable the simulation of point mutational component of the evolutionary processes. The basic units of mutational events were one - nucleotide interchanges. No special distinctions were made between transitions and transversions. Only coding regions of genes (exons) were processed. A two level filtering procedure was implemented: less stringent filtering to model the constraints imposed on the DNA and RNA level (control of silent mutations) and more stringent filtering (control of recognizable mutations) to model the constraints imposed on the amino acid sequence level. Both kinds of constraints were quantified in a simplified manner. DNA/RNA constraints were modeled as the requirements that the compositional parameters of short oligonucleotides reproduce the values which result from the massive analyses of DNA sequences. The constraints which were built in on the amino acid level were constructed by means of the requirements that the amino acid interchanges should exhibit proper pattern in the amino acid property space.
The analyses, which were performed by the above mentioned formalism, addressed in particular the question of existence or non - existence of detailed balance in the case that the mutational transitions are treated in the framework of master equation.
The results of the computer simulations were compared with the results of the analysis of the homology searches within the coding regions of natural gene sequences. It was found that the stochasticity component, which was the essential ingredient in the computer simulation model, is also present in great extent in the pattern of observed codon - codon interchanges.
Triple Helical Structures Involving Deoxyinosine: There is a Penalty for Promiscuity
Martin Mills, Jens Völker and Horst H. Klump
Department of Biochemistry,
University of Capetown,
Rondebosch 7700, RSA
According to the "Third-strand binding code" inosine has the ability to act as a "wild-card" binding nonspecifically to both A·T and G·C basepairs, and this has obvious implications for oligonucleotide site-directed probes.
In this poster a series of seven 35-mer oligonucleotides of 5'pu-py-py motif have been designed to fold up sequentially into intramolecular triple helices. The Watson-Crick structure is identical for all sequences. One or more inosines were incorporated into the Hoogsteen strands in place of T's and/or C's. Once folded into the triplex the inosines are incorporated into parallel orientation to the purine strand of the duplex. The influence of inosine on the melting temperature (Tm) and on the phase boundaries has been assessed by means of UV-melting and CD-spectroscopy.
There are two distinguishable influences, namely bulkiness (I for T) and/or loss of charge (I for C+). On introducing inosines in place of thymines the Tm of the triple helix to hairpin transition is lowered by 35.5 (± 1)°C at pH 6.0 and by 1.2 pH units at 25°C. On introducing inosines in place of a cytosine and a consecutive thymine the Tm of the same transition is lowered by 29.2 (± 1)°C at pH 6.0 and 0.9 pH units at 25°C. On replacing cytosines with inosines the Tm drops by 44.5 (± 1)°C at pH 6.0 and by 1.7 units at 25°C. All in all while inosine destabilizes triple helix formation considerably, it is a poor substitute and it's role as a wild-card is limited.
Influence of DNA Topology on the Homologous Sequence Search Mediated by RecA Protein of E. Coli
U. Marquardt and J. Chen
Department of Chemistry and Biochemistry,
University of Delaware,
Newark, DE 19716
The mechanism of search and recognition of homologous sequence mediated by RecA protein is one of the main obstacles to understand the homologous recombination; i. e., how are any two homologous DNA sequences brought together. We have used a restriction enzyme assay to determine if there is any significant difference in the rate of homologous search by a ssDNA-RecA filament that is mediated by paranemic joints or plectonemic joints, and to ask what are the structural effects of the duplex DNA (i.e. should it be relaxed or supercoiled) on the rate of homologous search.
Our results show that no matter paranemic joint or plectonemic joint was formed between the ssDNA-RecA filament and the duplex DNA, the rate of finding the target site on the linear duplex DNA is always faster than that of a supercoiled duplex DNA,. We also domonstrat that on the same duplex substrate, three different linear ssDNA-RecA filaments have very similar rates of finding the duplex target. Even though two of them should form 5' and 3' end plectonemic joints respectively, and one should form a paranemic joint with the duplex. However, when a circular ssDNA-RecA filament was used, the rate is slower but at appreciably level.
Finally, we found that adding heterologous duplex DNA to the search reaction or increase the size of the duplex DNA but keep the same target site profoundly slows down the rate of search. This final conclusion differs from recent published data (1) showing that the search for homology follows a reaction scheme that is overall second order and that the apparent second order rate constant is independent of the complexity of the duplex DNA.
References and Footnotes
1. J. E. Yancey-Wrona & R. D. Camerini- Otero, Current Biology, 5, 1149-1158 1995.
The Sep1 Strand Exchange Protein From Saccaromyces Cerevisiae Does Not Promote Strand Exchange Reaction
Z. Zhang and J. Chen
Department of Chemistry and Biochemistry
University of Delaware
Newark, DE 19716
The Saccharomyces cerevisiae strand exchange protein 1 (Sep 1; also referred to as Xrn1, kem1, Rar5, or Stp beta) promotes the transfer of one strand of a linear duplex DNA to a homologous single-stranded DNA circle. Sep1 is also an exonuclease active on DNA and RNA. Multiple roles for the in vivo function of Sep1 have been proposed; ranging from DNA recombination, cytoskeleton to RNA turnover and G4 DNA pairing among other cellular process.
It has been shown that the exonuclease activity of Sep1 is slow and nonprocessive, with a turn over number of 20 nucleotides per minute and a processivity of 45 nucleotides on duplex DNA (1-3). It was shown that for Sep1 to initiate strand exchange, the linear duplex must have a single-stranded tail greater than 20 nucleotides in length; this single-stranded tail can be produced by Sep1 or an exogeneous exonuclease. The conclusion was that the Sep1 nuclease activity is dispensable for the strand displacement and branch migration phase of the strand exchange reaction (1-3).
Here, by using smaller substrates and Exonuclease III generated linear duplex with various length of single-stranded tail, as wel as Ca+2 to inhibite the exonuclease activity of Sep1. We show that the processivity of exonuclese activity of Sep1 is at least a few hundred nucleotides which is much longer than previouly reported. We also demonstrate that the joint molecule between the linear duplex and single-stranded circle observed from the Sep1 promoted strand transfer reaction is just the pairing between the long single-stranded tail of the linear duplex (generated by the processive exonuclease activity of the Sep1 from both ends of the duplex) and the single-stranded circle. Finally, with a synthesized Holliday junction as substrat, we can not detect any branch migration facilitated by Sep1. Our results imply that Sep1 may not involed in recombination in vivo.
References and Footnotes
1. Johnson, A. W. and Kolodner, R. D., J. Biol. Chem.
266, 14046-14054 1991.
2. Johnson, A. W. and Kolodner, R. D., J. Biol. Chem. 269, 3664-3671
1994.
3. Johnson, A. W. and Kolodner, R. D., J. Biol. Chem. 269, 3672-3681
1994.
The Topological Linkage of the Kinetoplast Minicricle DNA From Leshimania Tarentolae
W. A. McLaughlin, R. Hanson and J. Chen
Department of Chemistry and Biochemistry,
University of Delaware,
Newark, DE 19716
Kinetoplast DNA (kDNA) of trypanosomatid parasites is a network containing several thousand catenated DNA minicircles and a few dozen maxicircles. We had developed the following strategy to deduce the topological linkage of the minicircles of the mature kDNA network of Crithidia fasciculata(1). We used graph theory to provide precise models of possible network structures. (2) Based on these models, we predicted the structure and relative frequencies of the minicircle oligomers expected from a random breakage of networks(3). We determined experimentally the fragmentation pattern of kDNA networks as a function of the extent of digestion(4). By comparison of the results with the predictions, we identified the model that best represents the network. We concluded that for Crithidia fasciculata each minicircle is linked on average to three other minicircles(1,2).
However, the minicircles in a kDNA network are heterogeneous in sequence, and the degree of heterogeneity varies between different species. Thus, due to the differences in the degree of the sequence heterogeneity and the numbers as well as arrangement of the conserved regions, it is important to know if the topological structures of kDNA networks from other species, such as Leishmania or Trypanosoma are similar to those of Crithidia.
Here, using the similar method we have determined the minicircle valence (how many minicircles are catenated to each minicircle) of Leishmania tarentolae. the kDNA network of L. tarentolae appears very different from that of C. fasciculata: the minicircles are much smaller and isolated networks appear by EM to be irregular and full of holes; one would expect the calence number of L. tarentolae should be smaller than 3. We found that the average minicircle valence in L. tarentolae networks is 2.5. This results confirm our prediction and indicate that the minicircle valence number correlates with the compaction of minicircles in the kDNA network.
References and Footnotes
1. J. Chen et al., Cell, 80, 61-69 1995.
2. J. Chen et al., EMBO J. 14, 6339-6347 1995.
Back to Table of Contents: Abstracts
Back to Table of Contents: Tenth
Conversation