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.
Interaction of DNA Bases and Base Pairs with Various Metal Cations (Li+ Na+, K+, Rb+, Cs+, Mg2+, Ca2+, Sr2+, Ba2+, Cu+, Ag+, Au+, Zn2+, Cd2+, and Hg2+): ab initio MP2 Calculations and Nonadditivity of the Interaction
Jaroslav V. Burda(1), Jirí Sponer(1), Jerzy Leszczynski(2)
and Pavel Hobza(1)
(1)J. Heyrovsky Institute of Physical Chemistry,
Academy of Sciences of the Czech Republic,
Dolejskova 3, 182 23 Prague 8, Czech Republic
(2)Department of Chemistry,
Jackson State University,
1325 J.R. Lynch Street,
Jackson, Mississippi 39217-0510 USA
Interactions of single nucleobases (adenine and guanine) and Watson-Crick (adenine-thymine and guanine-cytosine) base pairs with alkali metals, alkaline earth metals, coinage metals, and metals of zinc group (Li+ Na+, K+, Rb+, Cs+, Mg2+, Ca2+, Sr2+, Ba2+, Cu+, Ag+, Au+, Zn2+, Cd2+, and Hg2+) were optimized by the Hartree-Fock method. Energetics with inclusion of MP2 correlation energy and basis set superposition errors were performed for final structures. The cations were described by relativistic effective potentials using a modified DZ+P valence basis set, and the 6-31G** basis set was used for other elements. The metal cations were allowed to interact with nitrogen N7 of adenine and the N7 and O6 atoms of guanine. Opening of the adenine-thymine and the guanine-cytosine pairs in the presence of all the cations studied requires more energy than in the case of isolated base pairs; in the case of dication metal complexes, it requires about 100% more energy. Three-body non-additivity is considerably larger for MGC complex than for the MAT complex; the intercomplex charge-transfer is also much larger for the former complex.
CD and NMR Studies on a La3+ Binding Nucleated Synthetic EF Peptide
Youxiang Gong, Jianxin Yang, Andrew J. Bierzynski* and
Neville R. Kallenbach
Department of Chemistry,
New York University,
New York, NY 10003
*Institute of Biophysical Chemistry,
Polish Academy of Science, Poland
A synthetic peptide with the sequence acetyl-Asp-Lys-Asp-Gly-Asp-Gly-Tyr-Ile-Ser-Ala-Ala-Glu-Ala-Ala-Ala-Gln-amide where Asp1, Asp3, Asp5, Tyr7 and Glu12 form an EF hand binding loop in native proteins was designed. Lanthanum was used in this study due to the small difference between its radius and that of divalent calcium ion (Ca2+ 0.99, La3+ 1.03). La3+ has a stronger affinity than calcium for binding to small synthetic peptides which interact less strongly with ions than the intact calmodulin. In the absence of La3+ the peptide has a weak helix spectrum. CD studies show strong helix formation when 1.3:1 (La3+:peptide) La3+ added at pH 6.9, 4C. NMR analysis of the complex reveals a tightly folded structure with a 12 residue binding loop, 3 residues in b-sheet conformation (Tyr7, Ile8, Ser9) and a 4 residue helix at the C terminus. The NOE distance constraints were used to generate solution structures using the simulated annealing strategy. The resulting structures assume the shape of a classical EF helix in native calmodulin. This loop and helix structure model opens a way to study metal-peptide interaction, the intrinsic propensity for helix formation by each amino acid without side-side interactions and the quantitative evaluation of specific helix stabilizing interactions such as salt bridges and side chain-side chain hydrophobic interactions.
This research was supported by grants from the NIH (GM 40746) and the Human Frontiers of Science Program.
Theoretical Study on Structure, Dynamics and Properties of DNA Bases and Base Pairs
Pavel Hobza, Jirí Sponer and Vladimír
Spirko
J. Heyrovsky Institute of Physical Chemistry,
Academy of Sciences of the Czech Republic,
182 23 Prague, Czech Republic
Various approaches of theoretical chemistry (quantum chemistry, statistical mechanics, computer simulations) were utilized to study the structure, dynamics and properties of DNA bases and base pairs.
These calculations revealed nonplanarity of amino group of isolated bases which contributes to the nonplanarity of most H-bonded pairs. In addition consideration of effective geometries, which cover the shift of nuclear positions by molecular vibrations shows that all the base pairs are nonplanar; this effective nonplanarity increases with temperature.
Accurate quantum chemical calculations, covering all the energy contributions were for the first time used for evaluation of stabilization energy of more than 50 H-bonded pairs and 10 stacked pairs. Stabilization energy of H-bonded pairs is larger than that of stacked pairs; while the former pairs are stabilized by electrostatic interactions the stabilization of the latter ones originates in the dispersion forces. For all H-bonded pairs the buckle and propeller vibrations are the lowest ones and are located bellow 50 cm-1 .Quantum-chemical results were utilized for testing the performance of various empirical force fields (e.g., AMBER, OPLS, CHARMM) frequently used for modeling the DNA and best results were found for the AMBER potential utilizing Cornell et al force field (J. Am. Chem. Soc. 117, 5179 (1995)).
Thermodynamic characteristics for the formation of base pairs, evaluated with statistical-mechanics and computer simulations confirmed the mass spectrometry results (Biopolymers 18, 1149 (1979)) and ruled out results obtained from supersonic beam experiments (J. Am. Chem. Soc. 116, 9211 (1994)). While the stabilization energies for various pairs differ considerably, the respective entropy terms are more uniform.
DNA Supercoiling Alters Calicheamicin
g1I Activation by Glutathione
William A. LaMarr and Peter C. Dedon
Division of Toxicology,
MIT,
Cambridge, MA 02139
As part of our effort to el
ucidate
genotoxin target selection mechanisms, we are studying the effect of supercoiling
on calicheamicin g1I (CAL g)-induced DNA damage. This is a biologically
important problem since transcription generates high levels of positive
and negative DNA superhelical tension that may affect drug-induced DNA damage
by altering DNA structure and dynamics.
CAL g is a member of the enediyne class of potent antitumor antibiotics (LD50 ~ fM-pM). This molecule binds site-selectively in the DNA minor groove and abstracts 4' and 5' deoxyribose hydrogen atoms from opposing strands following a thiol-induced rearrangement to form a benzenoid diradical (see figure at left).
A role for superhelical tension in CAL g target selection
was suggested by recent studies in which we observed that CAL g induces
negative writhing of plasmid DNA (see below). This suggests that CAL g may
bind preferentially to positively supe
rcoiled
DNA, by analogy to intercalators that bind with higher affinity to negatively
supercoiled DNA.
To test this hypothesis and other effects of superhelical tension on CAL g target selection, we have developed a technique that uses the archaeal histone HMf to prepare plasmid DNA substrates with physiologically relevant levels of positive (s~+0.04) superhelicity. We then compared the interaction of CAL g with this substrate to that in negatively supercoiled plasmid DNA (s~-0.05).
In the first set of studies, we found that CAL g, activated
by glutathione (GSH), induces ~50% more strand breaks in negatively supercoiled
DNA t
han in DNA with positive superhelical tension.
Furthermore, there was no change in the sequence selectivity of the DNA
damage or in the ratio of single to double-strand breaks. This agrees with
the hypothesis in which a tighter binding of CAL g to ively supercoiled
DNA leads to a less efficient glutathione activation. The large size or
negative charge of the tripeptide glutathione may limit its accessibility
to CAL g bound deeply oove or in some other supercoiling-dependent conformation.
To investigate the role of GSH activation we have undertaken two sets of experiments: (i) a comparison of CAL g activation by GSH and methylthioglycolate (MTG), a small, uncharged thiol; and (ii) a comparison of DNA damage induced by CAL g and calicheamicin q1I (CAL logue activated by hydrolysis of a thioacetate trigger (see below).
In the presence of MTG, CAL g produces similar levels of
strand breaks in both plasmid substrates. This is in contras
t to the results seen with GSH and suggests that MTG has greater
access than GSH to CAL g bound to positively supercoiled DNA. This may be
due to its smaller size, its lack of charge, or its hydrophobicity; the
latter may allow MTG to concentrate in the minor groove of DNA. Studies
are underway to investigate CAL g activation in the presence of thiols of
differing size and charge.
Further evidence for supercoiling-dependent effects on
drug activation was observed in CAL q studies (structure below left). This
drug produced roughly equal numbers of strand breaks in both positively
and negatively supercoiled plasmid substrates. Since CAL q is activated
by hydrolysis in the absence of thiols, this observation is again consistent
with a role of supercoiling in the thiol activation of CAL g.
The results of these experiments indicate that superhelical tension alters DNA structure and dynamics in such a way as to affect CAL g binding or accessibility to thiol activation. The observations also suggest a role for superhelical tension in the in vivo selection of genotoxin targets.
Supported by Grants from the NIH/NCI (CA64524, CA72936) and the Samuel A. Goldblith Career Development Professorship.
The Iso-Competition Point For Multivalent: Univalent Cation Binding To DNA
Anzhi Z. Li, Haiyan Huang and Kenneth A. Marx
Department of Chemistry,
University of Massachusetts Lowell,
Lowell, MA 01854
The interactions of divalent cations (Mg2+ and Ca2+), and trivalent cations (hexammine cobalt(III) and spermidine3+) with l-DNA-Hind III fragments ranging from 2,027 - 23,130 bp were investigated by pulse gel electrophoresis [1]. These data were interpreted by electrophoresis models (Henry model, and reptation model), Manning's counterion condensation (CC) theory and the Iso-Competition Point (defined by the authors). Data for the normalized mobility reduction (µ/µ0) of DNA due to Mg2+ (0-200 µM), Ca2+(0-40 µM), hexammine cobalt(III) (0-200 µM), and spermidine3+ (0-50 µM) counterion binding in tris-borate buffer were well fit by Manning's CC theory. The ionic strength effect, valence effect, and DNA size effect on the competition binding can all be connected and interpreted by an important parameter we call the iso-competition point. This is where the monovalent and multivalent cations possess equivalent charge neutralization fractions on DNA. For example, we observed the normalized mobility reduction to be shifted by a small amount D(µ/µ0) relative to the CC prediction value, or to a reference value measured from a specific fragment length. The normalized shift D(µ/µ0)/ (µ/µ0) is a function of DNA length, and the ion environment ( multivalent cation concentration CN and ionic strength). The effect was enhanced by an increase of DNA length and the ion concentration CN, and a decrease of the ionic strength. The 'shift' phenomena was observed to occur only close to where CN begins dominating the counterion binding, marked by the iso-competition point [2-3]. In fact, the iso-competition point presents a clear picture of the competition binding under a certain ionic environment condition. The nature of the iso-competition point and iso-competition line will be discussed.
References
1. Li, A. Z., Qi, L. J., Shih, H. H. and Marx, K. A. Biopolymers,
Vol. 38, 367-376 (1996)
2. Li, A. Z., Huang, H. and Marx, K. A. ABSTRACTS OF MRS 1996 Fall Meeting,
EE 2.9, Boston.
3. Li, A. Z., Huang, H. and Marx, K. A. in Statistical Mechanics in Physics
and Biology, Wirtz, D., Halsey, T. C. & Zanten, J. van Eds., MRS
Symposium Proceedings Series, Volume 463, in press (1997)
Comparison Of Theoretical And Experimental Melting Curves Of The Yeast Genome And Simulation Of Individual Yeast Chromosome Melting Using The Program Meltsim
J. W. Bizzaro, Kenneth A. Marx and R.D. Blake*
Department of Chemistry,
*Department of Biochemistry, Microbiology and Molecular Biology
University of Massachusetts,
Lowell, MA 01854
04469-5735
*present address: Williams College, Williamstown MA 01267
The first eukaryotic genome to be completely sequenced as part of the international Human Genome Project is the yeast- S. cerevisiae. Using the sequence database for this organism(12,067,277 bp), we have carried out a simulated melting of yeast DNA using the program MELTSIM and compared it to the experimental high resolution melting behavior of yeast DNA in the following buffer condition: 0.0716 M NaCl, 0.0038 M cacodylic acid, 0.00015 M EDTA, pH 6.7. There is good agreement between the calculated and experimental melts. This demonstration, as well as earlier comparisons for incomplete genomic databases such as D. discoideum [1], validates the utility of MELTSIM to accurately simulate the melting of complex eukaryotic DNAs. We have also melted the individual yeast chromosome sequences. In general, there is little variability in melting behavior for the 16 chromosomes. They all undergo largely symmetric melting, resembling the simulated melting of the overall yeast genome. MELTSIM can also be used to perform positional melting along the individual yeast chromosomes. We illustrate this with representative chromosome melting data which we compare to positional GC composition along the same yeast chromosome sequence. Comparison of theoretical melting of all coding and non-coding sequences from the yeast genome showed a significant difference in the melting behavior. The coding sequences melt 2-3 degrees higher in temperature, characteristic of a higher GC composition compared to the non-coding sequences. This behavior is similar to what we have observed in other eurkaryotic genomes. The authors acknowledge support from the Center for Intelligent Biomaterials at UML.
References
1. Marx, K. A., Bizzaro, J. W., Assil, I. and Blake, R. D. in Statistical Mechanics in Physics and Biology, Proceedings MRS, 463, in press(1997).
MELTSIM, A Windows Program for Simulating the Thermal Dissociation of Polymeric DNAs
R. D. Blake1,2, J. W. Bizzaro3, John Knowles1, Gary
Day1, Jonathan Blake1,4 and Kenneth A. Marx3
1Department of Biochemistry,
Microbiology & Molecular Biology,
University of Maine,
Orono ME 04469-5735
2Department of Chemistry,
Williams College,
Williamstown,
MA 01267 (present address)
3Department of Chemistry,
University of Massachusetts,
Lowell MA 01854
4Department of Biochemistry and Biophysics,
University of California,
San Francisco CA 94010 (present address)
Knowledge of the forces that govern the conformation and stability of DNA can be inferred from the equilibrium concentrations of individual base pairs in the variety of conformational sub-states they adopt during melting. Such knowledge is needed to develop plausible models for the mechanics and dynamic behavior of DNA during replication, repair and transcription. MELTSIM is a PC windows-based statistical mechanical program for calculating simulated derivative melting curves and denaturation maps of polymeric DNA sequences. The model for melting is the one-dimensional Ising lattice to which loop entropy has been appended5-7. The algorithm is that of Poland8, with the Fixman-Friere9 approximation of the loop function by a sum of exponentials. The program reads in a DNA sequence and, using Poland's recursion formulae, calculates conditional probabilities at each temperature that the (m-1)th residue is in the paired state, given that the mth residue is paired. Following this, the unconditional probabilities that the mth residue is in the paired state are calculated. The unconditional probabilities yield the fraction of residues in the paired state, qbp(T), and, thereby, dqbp(T)/dT. The latter is proportional to dA270nm/dT of experimental curves. Computation time is proportional to Nlen, the length of the sequence, where values for the coefficients of the exponential terms are determined in subroutine by quasi-Newton least-squares fit, such that the difference between the summation of exponentials and the exact loop weighting function is minimized over the range of Nlen>N>1. Five intermediate nonhelical sub-states are recognized, distinguishable by the physical states of neighboring domains. Values for the principal statistical weights for these states are provided from the recent literature10: cooperativity parameter, closed ring structures, and the unit equilibrium constants for unstacking-unpairing pair L on M at the end of a helical segment of length Nk. Agreement between Tm for observed and calculated curves is generally within ±0.22o. Good agreement is also observed for the amplitudes, breadths and areas of subtransitions.
References
1. Ising, E. (1925) Physik 31, 253, cited in: Hill,
T.L., Statistical Mechanics, McGraw-Hill, New York (1956).
2. Hill, T.L., Statistical Mechanics, McGraw-Hill, New York (1956).
3. Wartell, R.M. & Benight, A.S., Physics Rep. 126, 67-107 (1985).
4. Poland, D. & Scheraga, H.A., Theory of Helix-Coil Transitions
in Biopolymers, Academic Press, New York (1970); Poland, D., Biopolymers
13, 1859-1871 (1974).
5. Fixman, M. & Freire, J., Biopolymers 16, 2693-2704 (1977).
6. Delcourt, S.G. & Blake, R.D., J.Biol.Chem. 266, 15160-15169
(1991); Blake, R.D., in Encyclopedia of Molecular Biology and Molecular
Medicine Vol 2. (Meyers, R.A., ed.) VCH Publ., New York, 1-19 (1996).
Refinements of Thermodynamic Parameters Associated With Helix-Coil Transitions of Polymeric DNAs
D. Blake(1,2) and Scott G. Delcourt(1)
(1)Department of Biochemistry,
Microbiology & Molecular Biology,
University of Maine,
Orono, ME 04469-5735
(2)Department of Chemistry,
Williams College,
Williamstown MA 01267 (present address)
Questions persist regarding the accuracy of various parameters associated with helix-coil transitions of polymeric DNAs. We reevaluated some of these parameters through analyses of high resolution melting curves obtained by the difference-approximation method with a modified double-beam ratio-recording spectrophotometer. Precision in the evaluation of parameters is proportional to the precision in the maintenance of scheduled temperatures and DT, which here approaches ±0.1mdeg and ±10mdeg, respectively. The helix-coil dissociation of synthetic homopolymeric DNAs is very sharp, within 0.1o; indicative of a considerable degree of cooperativity and a small value for the helix initiation parameter. By contrast, DNAs of quasi-random sequence dissociate in piecemeal fashion with subtransitions from micro-domains of a few base pairs to macro-domains of >500, spread over a 15-20o range. Subtransitions for individual domains can be isolated from background contributions of other domains by subtracting the melting curves of plasmid DNAs without the domain from those that have it. Sizes and distributions of base compositions of domains were obtained from integrated areas and spectral decomposition of curves at selective wavelengths, respectively; thereby confirming assignments. Domains were identified that dissociate in cooperative all-or-none fashion, and therefore analyzable for unit transition enthalpies by equilibrium van't Hoff methods.
The principal specimens were plasmid DNAs constructed with
oligomeric repeats of biased base content inserted at sites with favorable
energetic characteristics. Repetitive sequences of the type: {AAGTTGAAC[A]M}NAAGTTG,
were generally used, where 6>M>0 and 60>N>10; inserted at the
unique SmaI locus of a multiple cloning sequence with recognition sites
for ten restriction enzymes at the unique NruI locus (972) of pBR322. This
locus is immediately adjacent to a unique KpnI locus, and 803bp away from
an EcoRV sequence. When these sites are used to linearize the plasmid in
preparation for melting, the repetitive domain dissociates from the ends
of the helix and as a closed loop, respectively, providing suitable model
systems for the evaluation of helix propagation parameters and of loop entropy.
Work supported by funds provided by NIH and MAES.
Studies of the Dynamics of Adenine Amino Protons in DNA by 15N-Labeling and Heteronuclear NMR Spectroscopy
Ryszard Michalczyk and Irina M. Russu
Department of Molecular Biology and Biochemistry,
Wesleyan University,
Middletown, CT 06459-0175
Conformational fluctuations of DNA are functionally significant and affect the recognition and reactivity of DNA molecules. The time scales of these fluctuations range from picoseconds to minutes. Processes on the slower time scale include the rotation of amino groups around the exocyclic C-N bond. NMR spectroscopy is uniquely suited to observe and characterize this process at the level of individual base pairs. However, in the case of Adenines, the use of 1H NMR methods is limited by the fact that the resonances of amino protons overlap with several non-exchangeable proton resonances in the aromatic region of the spectrum. To overcome this limitation, in the present work, we have used heteronuclear 15N -1H NMR to edit the resonances of Adenine amino protons and to characterize the dynamics of Adenine amino groups.
Adenine was labeled with 15N at N-6 nitrogens and then,
incorporated in the DNA dodecamer [d(CGCGAGCTCGCG)]2 which contains two
symmetrical AT base pairs. 15N -1H HSQC experiments indicated that, at temperatures
below 20°C, the rotation of the amino groups in these base pairs is
slow on the NMR chemical shift scale and thus, separate resonances are observed
for the hydrogen-bonded and non-hydrogen bonded amino protons. The rotation
rates of the Adenine amino groups were measured from selective population
transfers in 15N-edited proton exchange spectra, and from lineshape analysis
of the amino resonances in 15N -1H HSQC spectra. The temperature dependence
of the rotation rates was used to obtain the activation parameters for the
rotation process. The implications of these results for the mechanism of
amino group rotation in AT base pairs will be discussed.
Supported by a grant from the NIH.
Modeling DNA Replication Intermediates
Roy, D.(1), Hingerty, B.E.(2) and Broyde, S.(1)
(1)Biology Dept.,
New York University,
New York, NY 10003
(2)Health Sciences Division,
Oak Ridge National Lab.,
Oak Ridge, TN 37831
While there is now available a great deal of information on double stranded DNA from X-ray crystallography, high resolution NMR and computer modeling, very little is known about structures that are representative of replication intermediates. DNA replication occurs at a single strand/ double strand junction and bulged out intermediates near the junction can lead to frameshift mutations. The single stranded domains are particularly challenging. Our interest is focused on strategies for modeling these types of replication intermediates. Modeling such structures presents special problems in addressing the multiple minimum problem and in treating the electrostatic component of the force field. We are testing a number of search strategies for locating low energy structures of these types. We are also investigating several different distance dependent dielectric functions in the coulombic term of the force field. We are studying both unmodified DNA and DNA damaged by aromatic amines, carcinogens present in the environment in tobacco smoke, barbecued meats and automobile exhaust. The nature of the structure adopted by the carcinogen modified DNA at the replication fork plays a key role in determining whether the carcinogen will cause a mutation during replication that can initiate the carcinogenic process. Structures of some modeled replication intermediates will be presented.
Supported by NIH and DOE.
The submitted manuscript has been authored by a contractor of the U. S. Government under contract number DE-AC05-960R22464. Accordingly, the U. S. Government retains a non-exclusive royalty free license to publish or reproduce the published form of this contribution, or allow others to do so for U. S. Government purposes.
Proton Exchange and Internal Rotation of Cytosine Amino Groups in DNA Double Helices
Irina M. Russu and Ryszard Michalczyk
Department of Molecular Biology and Biochemistry,
Wesleyan University,
Middletown, CT 06459-0175
Proton exchange in nucleic acid molecules is a sensitive indicator of conformational fluctuations including transitions of the base pairs between closed and open forms. The exchange properties of the imino protons on nucleobases have been extensively characterized by this and other laboratories, and have been shown to provide a faithful representation of the kinetics and energetics of base-pair opening processes. In contrast, the exchange mechanisms for the amino protons are less understood. This is due, in part, to the fact that exchange of amino protons may occur from the closed state of the nucleobase and is also affected by the internal rotation of the amino group around the exocyclic C-N bond. Understanding the exchange properties of amino protons is of special interest for defining conformational fluctuations in alternate nucleic acid structures and in complexes of nucleic acids with proteins or drugs.
In the present work, we have investigated the solvent exchange of Cytosine amino protons and the rotation of the Cytosine amino groups around the exocyclic C-N bond in the DNA dodecamer [d(CGCGAGCTCGCG)]2. The solvent exchange rates were measured by magnetization transfer from water at temperatures between 5 and 45°C, in 10 mM phosphate buffer at pH 7.0. The exchange rates of the hydrogen-bonded amino protons were, within experimental errors, the same as those of the non-hydrogen bonded protons, and ranged from 0.1 to 15 s-1. The rotation rates of Cytosine amino groups were measured from magnetization transfers between the two protons in the amino group and from lineshape analysis of the amino proton resonances, at temperatures between 5 and 45°C. The rates were found to range from < 5 to 200 s-1, with an activation enthalpy for the rotation process of the order of 20 kcal/mol. The correlations between solvent exchange and internal rotation of Cytosine amino groups will be discussed and compared to that in isolated GC base pairs.
Supported by a grant from the NIH.
(CTG)n·(CAG)n and (CGG)n·(CCG)n Triplet Repeats: Unusual Helical Properties and Alternative DNA Secondary Structures
Richard R. Sinden(1), Paul D. Chastain II(1), Evan E.
Eichler(2), Stephen D. Levene(3), David L. Nelson(2) and Christopher E.
Pearson(1)
(1)Institute of Biosciences and Technology,
Center for Genome Research,
Department of Biochemistry and Biophysics,
Texas A&M University,
2121 W. Holcombe Blvd.,
Houston, TX 77030-3303
(2)Department of Molecular and Human Genetics,
Baylor College of Medicine,
One Baylor Plaza,
Houston, Texas 77030
(3)Program in Molecular and Cell Biology,
The University of Texas at Dallas,
P.O. Box 830668,
Richardson, Texas 75083
About twelve human genetic diseases have been associated with the expansion of CTG or CGG triplet repeats. While the molecular etiology is unknown, expansion may involve the participation of an unusual helix structure or alternative DNA structure, specifically slipped strand structures, in replication, repair, or recombination. We have shown that DNA fragments containing (CTG)n·(CAG)n (where n = 17 to 255) derived from the human myotonic dystrophy gene have up to a 20% faster-than-expected mobility in nondenaturing polyacrylamide gels. The anomalous mobility is dependent upon the number of triplet repeats, the length of the flanking DNA, and the percentage and temperature of the polyacrylamide. The addition of the intercalating drug actinomycin D abrogates the rapid mobility. Applying a reptation model for electrophoresis, these results are consistent with a 20% increase in persistence length of the DNA, implying perhaps a DNA helix with a straight trajectory. PCR products containing CTG or CGG repeats from the spinocerebellar ataxia type I gene (SCA1) or the fragile X FMR1 gene, respectively, also showed faster electrophoretic mobility.
We have also investigated the mobility of DNA containing phased regions of curvature interspersed with (CTG)n tracts to further elucidate the unusual physical properties associated with triplet repeats. Results demonstrate that (CTG)n sequences do not possess static curvature but they reduced curvature resulting from A-tract or non-A-tract bends in an orientation independent manner. This decrease in the apparent overall curvature of A-tract bends was comparable to the effect of placing regions of increased bendability adjacent to the A-tract bends. Initial cyclization kinetic measurements are consistent with a flexible DNA helix. These results suggest that CTG triplet repeats may be inherently flexible.
Using cloned fragments from the DM and FRAXA loci containing normal, premutation, and full mutation lengths of repeats, we have reported the formation of novel alternative DNA secondary structures, presumably slipped strand DNA structures (S-DNA), that map within the repeat tracts during reannealing of complementary strands, containing equal lengths of repeats, into duplex DNA molecules. Duplex DNA molecules containing these alternative DNA secondary structures are characterized by reduced electrophoretic mobilities in polyacrylamide gels. S-DNA structures are stable at physiological ionic strengths and up to temperatures of at least 55 °C. Following reduplexing, the CAG strand of the (CTG)n·(CAG)n repeats is preferentially sensitive to mung bean nuclease suggesting the presence of single-stranded DNA in S-DNA. For (CTG)17, which is a repeat length found in normal individuals, less than 3% of the DNA molecules formed alternative DNA structures upon reduplexing. DNA molecules containing (CTG)50 or (CTG)255, which represent premutation and full mutation lengths of triplet repeats, respectively, formed a heterogeneous population of alternative DNA structures in >50% of DNA molecules. The propensity to form S-DNA and the heterogeneity of S-DNA structures increased with increasing repeat length. The presence of AGG interruptions in the FRAXA (CGG)n tract reduced the heterogeneity and extent of S-DNA formation. The effect of both length and the purity of the repeat tract on the extent and heterogeneity of S-DNA formation correlates with their effect on the genetic stability in humans. These are the first results consistent with the existence of slipped strand DNA structures.
Nonadditivity of Interactions in H-bonded and Stacked Complexes of Nucleic Acid Bases. An Important Contribution, Not Included in the Empirical Potentials
J. Sponer(1,2), J.V. Burda(1), J. Leszczynski(3) and
P. Hobza(1)
(1)J. Heyrovsky Institute of Physical Chemistry,
Academy of Sciences of the Czech Republic,
Dolejskova 3,
182 23 Prague, Czech Republic
(2)Institute of Biophysics,
Academy of Sciences of the Czech Republic,
Královopolská 135,
612 65 Brno, Czech Republic
(3)Department of Chemistry,
Jackson State University,
Jackson 39217 MS, USA
Current molecular modeling of nucleic acids is based on the use of pairwise-additive empirical potentials which do not cover the many-body contributions. Nonadditivity of interactions can be studied only by quantum-chemical calculations and here we present a preliminary characterization of nonadditivity of interactions in selected complexes of nucleobases.
i) Stacking of base pairs [1]. The nonadditivity of base
stacking (a difference between base stacking energy evaluated as the interaction
of two base pairs, and as a sum of the four separate base-base contributions)
is usually small, 0-3 kcal/mol in absolute value. Neverthelles, it can influence
both sequence-dependence and conformational dependence of stacking energy
in base pair steps consisting of two GC base pairs. The nonadditivity is
of induction origin and is enhanced when replacing standard bases by more
polarizable thioanalogues.
ii) Three-body term in neutral H-bonded trimers is mostly negligible, though
it is about 3 kcal/mol of additional stabilization in the G...G...C Hoogsteen
trimer [2].
iii) Protonated H-bonded and stacked pairs of nucleic acid bases are characterized
by a rather significant stabilizing induction contribution (3-4 kcal/mol);
further, there is a stabilizing three-body term in protonated H-bonded trimers
[2,3].
iv) Complexes of base pairs with metal cations are characterized by huge
polarization and charge-transfer effects. Empirical potentials are not able
to describe these complexes [3,4].
Our calculations could be made at best using the second-order perturbational treatment (MP2). Thus, they cover only the first-order and second-order exchange and induction nonadditivites, while the (third-order) dispersion nonadditivity is still absent. It seems to be possible to include induction energy (the three-body induction nonadditivity is the most important) into force fields in a near future. However, the charge transfer will not be included, and ab initio techniques are the only methods to study the energetics of complexation of base pairs with metal cations.
References
1. J. Sponer, H.A. Gabb, J. Leszczynski, P. Hobza Biophys.
J. submitted.
2. J. Sponer, J.V. Burda, P. Mejzlík, J. Leszczynski, P. Hobza. J.
Biomol. Struct. Dyn. in press.
3. J. Sponer, J. Leszczynski, V. Vetterl, P. Hobza J. Biomol. Struct.
Dyn. 13, 695-707 (1996).
4. J. Sponer, J.V. Burda, J. Leszczynski, P. Hobza J. Phys. Chem. in
press.
Molecular Interactions in i-DNA and Z-DNA. A Crystallographic, Quantum-Chemical and Preliminary Empirical Potential Study
J. Sponer(1,2), I. Berger(3), M. Egli(4), H.A. Gabb(5),
N. Spacková(2), P. Hobza(1) and A. Rich(3)
(1)J. Heyrovsky Institute of Physical Chemistry,
AS CR, Dolejskova 3,
182 23 Prague, Czech Republic
(2)Institute of Biophysics,
Královopolská 135,
612 65 Brno, Czech Republic
(3)Department of Biology,
MIT, Bdg. 68-233,
Ames Street 31, Cambridge, MA 02141, USA
(4)Department of Molecular Pharmacology and Biological Chemistry,
Northwestern University Medical School,
303 East Chicago Avenue,
Chicago, IL 60611-3008
(5)Imperial Cancer Research Fund,
44 Lincoln's Inn Fields,
London WC2A 3PX, UK
Crystallographic analyses of nucleic acid fragments adopting unusual conformations such as left-handed Z-DNA or four-stranded cytosine-rich DNA (i-DNA) reveal stabilization motifs unlike the familiar base pairing and base stacking schemes [1]. Various sugar-base stacking interactions were found in crystals of Z-DNA and i-DNA fragments, and systematic close contacts between adjacent ribose sugar moieties observed in i-DNA crystals were interpreted in the context of C-H...O hydrogen bond zipper [2].
Here, we report high-level quantum chemical and preliminary molecular mechanical analyses of the sugar-base stacking and sugar-sugar zipper motifs. The calculations show that these interactions are dispersion controlled. The molecular orbital analysis supports the previously postulated n-p* charge transfer interaction for sugar base stacking, though this contribution appears to be weak, or counterbalanced by another repulsive contribution. Both sugar-base and sugar-sugar interactions are found to be attractive, contributing an energy of -3.5 to -5 kcal/mol. This stabilization is comparable to base-base stacking in regular DNA forms. Crystallographic studies show that water molecules are excluded from the contact areas, and thus all interactions studied can be considered as hydrophobic. Our calculations indicate that the interaction observed in the sugar-sugar zipper in i-DNA is not directly stabilized through the H-O interaction (in contrast to usual hydrogen bonds). This somewhat contradicts the current view on C-H...O interactions, but is fully consistent with recent studies carried out on related C-H...O systems.
In i-DNA, the charged phosphate groups approach each other more closely than in regular DNA. Also, i-DNA requires hemi-protonation of the cytosines resulting in closely stacked (3.1 Å) consecutive protonated cytosine base pairs in the center of the molecule. To both features one is inclined to attribute a destabilizing effect. Force field calculations show that inclusion of counterions and the use of distance-dependent dielectric constant would be sufficient to eliminate the phosphate-phosphate repulsion, and to sharply reduce the repulsion between the protonated base pairs.
All interactions studied are reproduced by a standard (AMBER-based) empirical force field.
References
1. Wang et al., Nature 282, 680 (1979); Kang et
al., Proc. Natl. Acad. Sci. USA, 91, 11636 (1994)
2. Berger et al., Proc. Natl. Acad. Sci USA 93, 12116 (1996).
3. Chaney et al., J. Am. Chem. Soc. 118, 9432 (1996); Mourik
& Duijneveldt, J.Mol.Struct. (Theochem) 341, 63 (1995).
Crystal Structure of r(GUAUACA)dC: 3' Double Base Overhangs With Hemiprotonated Trans (dC·dC)+ Base Pairs and Minor Groove A*(G·C) Triplets
Ke Shi, Roopa Biswas, Shome Nath Mitra and Muttaiya
Sundaralingam
Biological Macromolecular Crystallography Laboratory,
Departments of Chemistry & Biochemistry,
The Ohio State University,
Columbus, OH 43201
RNA molecules, with 2'-hydroxyl groups which endows them significant differences from DNA molecules, can fold into numerous tertiary structures to perform various biological functions. RNA folding and related functions are of great interest and recently some breakthroughs have been made. Here we report the crystal structure of an alternating RNA octamer with a 3'-deoxycytosine, r(GUAUACA)dC at 2.2Å. The space group is P212121 with the unit cell dimensions a=24.19Å, b=45.33Å, c=74.20Å. There are two independent duplexes (I and II) in the asymmetric unit. The structure was solved by molecular replacement using the program AMoRe and refined by X-PLOR to an R of 19.6%. Instead of forming a blunt end duplex with eight base pairs including two non-adjacent A+·C mispairs, the RNA strand slides two bases forming a six base pair duplex and a double base (A7, C8) overhang. The duplexes (I and II) stack on each other in an unusual head-to-head and tail-to-tail fashion. The terminal deoxycytosines loop out and form hemiprotonated trans base pairs (dC·dC)+ with symmetry related molecules, while the penultimate adenines (A7, A15) of molecule I form novel minor groove A*(G·C) base triplets with molecule II. Various multiple base interactions enable RNA molecules to form complex tertiary structures to perform their biological functions.
Research supported by NIH Grant GM17378 and an Ohio Eminent Scholar Endowment.
Side-by-Side Binding of Distamycin A to B-DNA: Crystal Structure of d(GTATATAC)2 - Distamycin Complex With Two Independent Molecules
Shome Nath Mitra, Markus C. Wahl and Muttaiya Sundaralingam
Biological Macromolecular Crystallography Laboratory,
Departments of Chemistry and Biochemistry,
The Ohio State University,
120 West 18th Ave.,
Columbus, OH 43210.
Complexes of the groove binding drug distamycin A with DNA oligomers have been studied extensively. Knowledge of these structures and thermodynamics of complex formation can throw light on the initial recognition of DNA by the drugs and their binding specificity. Crystal structures of drug-DNA oligomers have indicated that the drug binds to a stretch of 4 to 5 A·T/I·C base pairs but not to G·C base pair. Most drug- binding involve 1:1 complexes, both in solution and in crystals. Here we report the crystal structure of a 2:1 side-by-side binding of distamycin to a DNA d(GTATATAC)2, containing natural bases at 2.35Å. The drug-DNA complex crystallized with cell parameters a=29.55Å, b=42.15, c=43.38, b=96.6o and space group P21 with two independent molecules in the asymmetric unit. The structure was solved by molecular replacement method using the program AMoRe and refined by X-PLOR to an R of 19.8%. In the complex the two dyad related distamycin molecules are in the expanded minor groove and in the antiparallel orientation as the DNA duplex itself. The amide NH's of the drug molecules are hydrogen bonded to the minor groove base atoms of the DNA strand closest to it. The alternating octamer structure exhibits low-high alternations in the helical twist, sugar puckering and phosphate conformations.
Research supported by NIH grant GM 17378 and an Ohio Eminent Scholar Endowment.
The A-DNA Duplex - How it Differs From B-DNA
Muttaiya Sundaralingam
Biological Macromolecular Crystallography Laboratory,
Departments of Chemistry and Biochemistry,
The Ohio State University,
120 West 18th Ave.,
Columbus, OH 43210
Fibers of A-DNA often form at low relative humidity compared to B-DNA, therefore, can they be considered to be more hydrophobic? A-DNAs exhibit a shallow narrow groove bringing the bases close to the sugar to form a hydrophobic sugar-base 'backbone' with intervening phosphates. The contraction of the backbone due to the C3'-endo sugar puckering leads to a continuous hydrophobic ribbon formed by the sugar protons (H1', H2'1, H4', H5'1, H5'2) which promotes stacking interactions on the shallow groove site. In oligonucleotide A-DNA structures the terminal base pairs of symmetry related duplexes abut into the shallow minor groove, unlike the B- and Z-DNA duplexes which stack on top of each other to form pseudocontinuous helices. This abutting interaction leads to the formation of shallow groove base triplets and quadruplets. It is of interest to know whether the abutting interaction also occurs in longer helices. The economy of hydration of the low humidity A-DNAs is probably due to their hydrophobicity. The above and related features will be discussed with the known A-DNA oligonucleotide crystal structures.
Research supported by NIH grant GM 17378 and an Ohio Eminent Scholar Endowment.
Crystal Structure of r(CCCIUGGG) with Tandem I·U Wobble Base Pairs-Three Independent Octamer Duplexes
Baocheng Pan, Shome Nath Mitra and Muttaiya Sundaralingam
Biological Macromolecular Crystallography Laboratory,
Departments of Chemistry and Biochemistry,
The Ohio State University,
120 West 18th Ave.
Columbus, OH 43210
Inosine (I) has been found in the first position of anti-codons in tRNA's and can translate codons ending in U, A and C. So far base pairings with inosine in RNA crystal structures have not been reported. To study the mode of base pairing involving inosine, the RNA octamer r(CCCIUGGG)2 with two central I·U base pairs has been synthesized and its crystal structure determined at 2.5 Å resolution by molecular replacement method using AMoRe and refined by X-PLOR to an R-factor of 0.192 for 3765 reflections. The crystal belongs to the monoclinic space group P21 with cell constants a=33.44 Å , b=43.41 Å, and c=49.39 Å and b=104.07°, with three independent duplexes in the asymmetric unit. The volume per base pair of 1,400 Å3. The three independent duplexes are stacked in the usual head-to-tail fashion.The adjacent I·U base pairs stack with good overlap of the bases and the RNA helices are not distorted. The I·U base pairs in the structure are in the wobble conformation, with two hydrogen bonds between I and U: O6(I) ······ N3(U) (average 2.82 Å, range: 2.66 Å to 3.02 Å) and N1(I) ······ O2(U) (average 2.83 Å, range: 2.69 Å to 2.96 Å). The average twist angle between the tandem I·U base pairs is 28.2°, compared with the average value of 32.9° for all the base pairs in the structure.
Research supported by NIH grant GM17378 and an Ohio Eminent Scholar endowment.
Effect of the Substitution of Inosine on the Structure of A-tract DNA
Ruth M. Ganunis and Thomas D. Tullius
Department of Chemistry,
The Johns Hopkins University
3400 N. Charles Street,
Baltimore, MD 21218
The frequency of cleavage by hydroxyl radical of mixed sequence DNA is relatively even. However, the frequency of cleavage for A-tract DNA resembles a sinusoidal pattern, with a decrease in cleavage from the 5' end to the 3' end of the A-tract. The cleavage then increases back to that of mixed sequence DNA over the adjacent several base pairs. This sinusoidal pattern is thought to reflect the narrowing of the minor groove through the A-tract, and the subsequent return to expected minor groove width through the surrounding mixed sequence DNA.[1]
Studies involving the incorporation of modified nucleotides into A-tract DNA provide a better understanding of the influences of the chemical nature of the base on the structure of bent DNA.2-5 Using gel mobility assays, Diekmann and coworkers have studied the effect of the substitution of inosine nucleotides into A-tract DNA. The results indicate that RL values correspond well to the number of inosine substitutions, and that the position of the inosine substitution is also important.[2] To better understand the effect of inosine (I) substitution on A-tract structure, we have performed hydroxyl radical cleavage experiments on an A5 sequence, and compared it to the modified sequences AAIAA, AIAIA, AIIIA, and IIAII. Almost no change in cleavage pattern for the substituted sequences is seen at the 5' end of the modified A-tracts. An increase in the amount of cleavage is seen at the 3' end of the modified A-tract, as well as at the nucleotide immediately 3' to the A-tract. The differences in the hydroxyl radical cleavage patterns suggest that the substitution of inosines into A-tract DNA changes the overall structure of the A-tract, and that these changes are focused at the 3' end of the A-tract.
References
1. Burkhoff, A. M., & Tullius, T. D. (1987) Cell
48, 935-943.
2. Diekmann, S., von Kitzing, E., McLaughlin, L. Ott, J., & Eckstein,
F. (1987) Proc. Natl. Acad. Sci. USA 84, 8257-8262.
3. Koo, H.-S., & Crothers, D. M. (1987) Biochemistry 26,
3745-3748.
4. Hagerman, P. J. (1990) Biochemistry 29, 1980-1984.
5. Diekmann, S., Mazzrelli, J. M., McLaughlin, L. W., von Kitzing, E., &
Travers, A. A. (1992) J. Mol. Biol. 225, 729-738.
External Nucleophiles That Induce DNA Cleavage in Conjunction With an Organometallic Complex
Celeste P. Jamison McDaniels and Thomas D. Tullius
Department of Chemistry,
The Johns Hopkins University
3400 N. Charles Street,
Baltimore, MD 21218
There exist few cases of synthetic metal catalysts that are capable of hydrolysis of unactivated phosphate diesters like DNA. Previous research in our lab led to development of a system whereby an external nucleophile is able to induce hydrolytic cleavage of a specific DNA phosphodiester bond.[1] These experiments showed that an external nucleophile can effect the second step of phosphodiester hydrolysis, as also was found for a lanthanum(III) complex.[2] Here we report new studies of DNA cleavage mediated by the reactive nucleophiles hydrogen peroxide (H2O2) and hydroxylamine (H2NOH), in conjunction with a synthetic metal complex. We have chosen to analyze a metallocene-based organometallic carcinostatic agent, bis(h5-cyclopentadienyl)molybdenum(IV) dichloride (Cp2MoCl2). In recent work, Cp2MoCl2 was shown to hydrolyze activated phosphodiesters.[3] This prompted us to see if Cp2MoCl2 could activate the hydrolysis of unactivated phosphodiesters in DNA, with H2O2 or H2NOH as external nucleophiles. Cp2MoCl2 has biological parallels, but also fundamental differences, to the well characterized antitumor drug cis-diamminedichloroplatinum(II), cisplatin; however, its interactions with DNA and mechanism of action warrant further characterization.[4]
The first set of experiments involved a 3' radiolabeled 27 base pair DNA duplex sequence. The cleavage products were run on 20% and 25% denaturing polyacrylamide gels so that the ends of the cleavage products could be determined. The 3' end labeled DNA showed cleavage at guanine, leaving a 5'-phosphate end. These results were very interesting since NMR studies had shown that Cp2MoCl2 had no specificity for a particular nucleotide, unlike cisplatin, which has been shown to have specificity for binding to guanine.[4] The cleavage pattern of a 5' radiolabeled strand showed an even more interesting cleavage pattern. In addition to bands resulting from cleavage at the 3' phosphate of guanine nucleotides, an associated band was observed which migrated more slowly. The slower migrating band ran between DNA fragments terminated by phosphates. These bands could very well be the result of hydrolysis specifically at G-C base pairs. These findings are interesting in that they parallel the guanine-specificity of cisplatin. These studies will help give a clearer picture of the mechanism by which Cp2MoCl2 combats tumors.
References
1 Kimball, A. S., Lee, J., Jayram M., & Tullius, T.
D. (1993) Biochemistry 32, 4698.
2 Takasaki, B. K., & Chin, J. (1993) J. Am. Chem. Soc. 115,
9337.
3 Kuo, L. Y., Kuhn, S., & Ly D. (1995) Inorg. Chem. 34,
5341.
4 Kuo, L. Y., Kanatzidis, M. G., Sabat, M., Tipton, A. L., & Marks,
T. J. (1991) J. Am. Chem. Soc. 113, 9027.
Chimeric Homeodomains Reveal the Importance of the N-terminal Arm in Determining the Structural DNA Binding Characteristics of Dfd and Ubx Homeodomains
Richard W. Frazee and Thomas
D. Tullius
Department of Chemistry,
The Johns Hopkins University,
3400 N. Charles Street, Baltimore,
Maryland, 21218 USA
Developmental gene expression in higher organisms is regulated in part by a family of transcription factors which contain a highly conserved 61 amino acid segment termed the homeodomain (HD) [1]. Acting with multiple and mixed target specificities, HD containing factors are involved in autoregulation and cross-regulation as well as regulation of effector genes. HD swapping experiments suggest that target specificity is associated with the identity of the HD rather than other domains of the transcription factor in which it is placed [2]. Possessing three a-helices in which the latter two approximate a helix-turn-helix motif, the HD constitutes the DNA-binding domain. X-ray [3] and NMR [4] structural data suggest that DNA binding is achieved via helix 3 contacts in the major groove together with adjacent minor groove interactions mediated by the HD N-terminus extending from helix 1. Hydroxyl radical footprinting, missing nucleoside, and methylation interference data for ultrabithorax (Ubx) and deformed (Dfd) HDs are in agreement with the three dimensional structural data [5]. Several studies have demonstrated the importance of the N-terminal arm in both in vivo [2] and in vitro [6] DNA binding site selectivity. However, few structural studies directly address the function of the N-terminal arm in binding site stability and selectivity. To determine N-terminal arm contributions to DNA binding specificity and stability at the nucleotide level, hydroxyl radical footprinting and missing nucleoside experiments were performed on a Ubx/Dfd chimera in which the Ubx N-terminal arm replaced the corresponding Dfd N-terminal arm. Both Dfd and Ubx optimal DNA binding sites in complex with the chimera were probed. Since both the Dfd and Ubx HDs bind the Ubx optimal site in a similar fashion, it was not surprising that the chimera's footprint on Ubx optimal DNA was similar to that observed for either Ubx or Dfd HDs. The missing nucleoside patterns were also similar to those observed for both wild type HDs. In contrast to the Ubx optimal sequence, both Dfd and Ubx HDs show marked differences in their footprints and missing nucleoside patterns when complexed with the Dfd optimal sequence. Interestingly, the chimera's footprint on Dfd optimal DNA more closely resembles that observed for the Ubx HD. Moreover, the chimera's missing nucleoside pattern is identical to that of the Ubx HD at key nucleoside positions which clearly distinguish Ubx HD complexes from those of the Dfd HD. Taken together, these data suggest that the HD N-terminal arm plays a major role in determining both the HD binding position and nucleoside contacts crucial for selective, stable HD/DNA complex formation.
References
1. E. B. Lewis, Nature, 276, 565 (1978).
2. L. Lin and W. McGinnis, Genes Dev., 6, 1071 (1992).
3. C. Wollberger, A. K. Vershon, B. Liu, A. D. Johnson, and C. O. Pabo,
Cell, 67, 517 (1991).
4. Y. Q. Qian, K. Furukubo-Tokunaga, D. Resendez-Perez, M. Müller,
W. Gehring, and K. Wüthrich, J. Mol. Biol., 238, 333
(1994).
5. A. Draganescu, J. R. Levin, and T. D. Tullius, J. Mol. Biol.,
250, 595 (1995).
6. S. C. Ekker, K. E. Young, D. P. von Kessler, and P. A. Beachy, EMBO
J., 10, 1179 (1991).
Base-Triple Interactions in the Triple Helical Domain of Self-Splicing Group I Introns
Martina Lindqvist, Munna Sarkar and Astrid Gräslund
Department of Biophysics,
Stockholm University,
S-106 91 Stockholm, Sweden
Two synthetic oligonucleotides, corresponding to the P4-P6 and J3/4, J6/7 regions of the bacteriophage T4 nrdB pre-mRNA, were studied by UV and circular dichroism (CD) spectroscopy to probe the nature of the base-triple interaction between single-stranded joining region J6/7 and paired region P4 in the triple helical domain of the self-splicing group I intron. This two-strand system (23-mer RNA and 7-mer RNA) has been shown to serve as a good model for the triple helical domain. From selected variants of both the 23-mer RNA and the 7-mer RNA, four different triple-helical systems were formed to serve as a test for reverse Hoogsteen-type of base-pairing. J6/7 has been proposed to interact in the major groove of P4 forming either coplanar Hoogsteen- or reverse Hoogsteen-type of base-pairing. System1 (23-mer wild-type and 7-mer wild-type) and system2 (23-mer variant2 and 7-mer wild-type) had strong and weak phylogenetic existence respectively, but no reverse Hoogsteen-type of base-pairing. The base triple combination in system3 (23-mer variant1 and 7-mer variant4) and system4 (23-mer variant2 and 7-mer variant5) were, on the other hand, not seen in phylogeny but still had two reverse Hoogsteen-type of base-pairing. Possible triple-helix association between RNA hairpin (23-mer) and RNA single strand (7-mer) in a 1:1 stoichiometry was studied by the use of CD in presence of Mg2+ (5mM). The results show that interaction takes place in one of the phylogenetically existing systems (system1) and in the two systems with no phylogenetic existence (system3 and system4), the latter having reverse Hoogsteen-type of base-pairing. No base-triple interaction was seen in system2 which has a weak phylogenetic existence but no reverse Hoogsteen-type of base-pairing. This indicates that base-triple interactions of J6/7 in the major groove of P4 in the triple helical domain of self-splicing group I introns cannot primarily be described by coplanar reverse Hoogsteen-type of base-pairing. Instead the interaction is probably stabilized by additional parts of the full catalytic core.
Supercoiled DNA Polyelectrolyte
Marcia O. Fenley(1), A. H. Boschitsch(1), Todd R. Quackenbush(1)
and Wilma K. Olson(2)
(1)Continuum Dynamics, Inc.,
P. O. Box 3073,
Princeton, NJ 08543,
marcia@cdiprinceton.com
(2)Department of Chemistry,
Rutgers, the State University of New Jersey,
New Brunswick, NJ 08903
We have carried out Monte Carlo/simulated annealing and energy minimization studies in order to examine the minimum energy configurations of simplified models of closed circular DNA at various levels of supercoiling as a function of chain length and salt (both monovalent and divalent) concentration. Both elastic and electrostatic energy contributions are considered and a repulsive potential is employed to avoid distant points along the chain from coming into close contact. For long DNA chains we employ a fast adaptive multipole method for the computation of the electrostatic energy and its derivatives in both Monte Carlo/simulated annealing and energy minimization codes. We examine the effect of different electrostatic/excluded volume models on the minimum energy configurations of supercoiled DNA.
Slow Disproportionation of Duplexes Formed by Ssdna/RNA Targets and Mixed Purine-Pyrimidine Pnas With Lysines in the Backbone
E. Lesnik, F. Hassman, J. Barbeau, K. Teng, K. Weiler,
R. Griffey and S. Freier ISIS Pharmaceuticals,
2292 Faraday Avenue,
Carlsbad, CA 92008, USA
Peptide-nucleic acids (PNAs), nuclease resistant DNA analogs, were synthesized and described some years ago [1]. Mixed purine-pyrimidine PNAs with one lysine amide at the C terminus and neutral backbone bind ssDNA / RNA targets in duplex fashion obeying the Watson-Crick hydrogen bonding rules [2]. We have synthesized mixed purine-pyrimidine PNA analogs (14-mers) with four and eight positively charged lysines instead of glycines in the backbone and studied the effect of positive charges on a complex formation between PNA and ssDNA/RNA targets. UV melting and mixing curve experiments carried out in a buffer with 100mM Na+ suggested that replacement of glycines with lysines in PNA backbone resulted in slow, measured in hours, duplex disproportionation leading to triplex formation. High salt concentration (1 M NaCl) provided only duplex formation judging by melting and mixing curve experiments. Formation of putative Py/Pu/Py and Pu/Py/Pu triplets suggested earlier for oligonucleotides with neutral backbone [3]. Lysine PNAs may form similar triplexes facilitated by additional electrostatic interaction between DNA/RNA and PNA backbones.
References
1. Nielsen, P. E.; Egholm, M.; Berg, R. H.; Buchardt, O.
Science, 254, 1497-1500 (1991).
2. Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.; Freier, S. M.;
Driver, D. A.; Berg, R. H.; Kim, S. K.; Norden, B.; Nielsen, P. E. Nature,
365, 566-568 (1993).
3. Reynolds M. A.; Arnold L.J.; Altmazan M. T.; Beck T. A.; Hogrefe R. I.;
Metzler M. D.; Stoughton S. R.; Tseng B. Y.; Trapane T. L.; Ts'O P. o. p.;
Woolf T. M. Proc. Natl. Acad. Sci. USA, 91, 12433-12437 (1994).
Function and Mechanism of Type -2 Topoisomerases
Rybenkov V.(1), Vologodskii A.(2), Ullsperger C.(1),
Khodursky A.(1), Hildebrandt E. (3), Zechiedrich E.L.(1), Peter B.(1) and
Cozzarelli N.R.(1)
(1) Dept. of Molecular and Cell Biology,
University of California,
Berkeley, CA 94720-3204
(2) Dept. of Chemistry,
New York University,
New York, NY 10003
(3) Dept. of Biology,
Johns Hopkins University,
Baltimore, MD 21218-2685
Semiconservative DNA replication requires the complete
unlinking of chromosomal DNA during each round of replication. This process
must be fast (1000 links per second), faithful (Lk must be reduced to exactly
0), and operate against the load of huge concentrations of DNA, crowding
macromolecules, and condensing agents. We have discovered five factors in
bacteria that allow replication to achieve these ends:
1. The effective concentration of DNA is less than the overall concentration.
2. Supercoiling diffuses into replicated DNA from unreplicated regions.
3. Supercoiling greatly promotes decatenation.
4. Gyrase and topo IV are tailored to their distinct roles in unlinking
5. Type-2 topos remove links from DNA to below equilibrium values.
Interdomain Interactions and Multivalent Association in Signal Transducing Domains
David Cowburn, David Fushman, Jie Zheng, Quinghong Xu,
James McDonnell, Sean Cahill and Rong Xu
The Rockefeller University,
1230 York Avenue,
New York, NY 10021-6399 USA
Intial efforts on the structural characterization of domains involved in intracellular signal transduciton focussed on protein folds, and on the recognition of single ligand classes [Cowburn96]. Many pathways, however, involve complex clusters of multiple domains, either for enzymatic control, or in adaptors. We are probing these multidomain, multivalent interactions using NMR and synthetic multivalent ('consolidated') ligands [Cowburn95].
Consolidated bivalent ligands for the Abelson Src Homology domain (SH) dual domain, SH(32), have been designed and synthesized to investigate systematically the relative orientations of the two subdomains. Complexes have been studied by NMR, and the role of flexibility and the structral basis of the dual ligand/protein interaction characterized.
Pleckstrin homology (PH) domains have complex ligation properties, as a class, including both phosopholipids and proteins. The structural and dynamic properties of several PH domains, including dynamin[Fushman97], suggest that felxible loops and N and C terminal extensions may play important roles in PH domain function.
References
1. Cowburn, David; Zheng, Jie; Xu, Qinghong; Barany, George,
"Enhanced affinities and specificities of consolidated ligands for
the SH3 and SH2 domains of Abelson protein tyrosine kinase". J.
Biol. Chem. 270, 26758-26741 (1995).
2. Cowburn, David & Kuriyan, John, "SH2, SH3 and PH domains",
in "Signal Transduction", Modular Texts in Molecular and Cell
Biology, 1, Carl-Henrik Heldin and Mary Purton, Eds., Chapman & Hall,
London, pp.127-142 (1996).
3. Fushman, David; Cahill, Sean M.; Cowburn, David, "The main chain
dynamics of the dynamin pleckstrin homology (PH) domain in solution: Analysis
of 15N relaxation with monomer/dimer equilibration." J. Mol. Biol.,266,
173-194 (1997).
DNA Recognition by Peptide Nucleic Acid (PNA)
Pernilla Wittung-Stafshede*
Department of Physical Chemistry,
Chalmers University of Technology,
Gothenburg, Sweden
*present address: Beckman Institute,
California Institute of Technology,
Pasadena, California
Reagents that bind sequence-selectively to double-stranded DNA are of significant interest in molecular biology and medicinal chemistry as they may be developed into gene-targeted agents for diagnostic and therapeutic purposes, and to provide tools for sequence specific modification of DNA. Single-stranded oligonucleotides and their close analogs are the main candidates for such reagents. However, most of them are not optimal in terms of biological stability and so far only purine-rich stretches of DNA may be targeted by an oligonucleotide. PNA is a promising oligonucleotide mimic in which the entire deoxyribose phosphate backbone of DNA has been replaced by a peptide-like backbone.
Thymine-rich homopyrimidine PNA oligomers recognize double-stranded DNA targets by a novel mechanism in which two PNA molecules bind to the complementary DNA strand to form a PNA-DNA-PNA triplex and the identical DNA strand is displaced into a loop. This type of PNA binding is inhibited by salt, so for future applications of PNA in living cells it is essential to understand the mechanism of these PNA-DNA interactions and the factors that control them. By using spectroscopic techniques we have found that the rate-limiting step for thymine-PNA invasion is opening of a few base-pairs of DNA and involves both PNA molecules. Intercalating ligands pre-bound to the DNA enhance the rate of PNA binding. Covalent attachment of such ligands to PNA might thus enable PNA to perform DNA invasion also at higher salt concentrations.
We have also investigated the possibility of targeting other DNA sequences with PNA, to extend the recognition code of double-stranded DNA. We find, using circular and linear dichroism, that cytosine-rich homopyrimidine PNA sequences do not invade double-stranded polynucleotide targets but instead add to the intact DNA as Hoogsteen strands forming PNA-DNA-DNA triplexes. Furthermore, PNA strands with homopurine, or alternating thymine-guanine sequences are shown to invade their respective DNA targets by displacing the identical DNA strands of the polynucleotides and forming new PNA-DNA duplexes. These results demonstrate distinct mechanistic variations as to how PNA interacts with a DNA target depending on choice of nucleobases, which could be of importance for future design of gene-specific agents.
Eukaryotic Recombinant Protein Rad51; its Thermal Stability and Complex Formation with Oligo(dT*)
Jong-Moon Kim, Masayuki Takahashi* and Seog K. Kim
Department of Chemistry,
College of Science,
Yeungnam University,
Kyung-san city,
Kyung-puk, 712-749 Republic of Korea
*UMR216,
Institut Curie and CNRS,
Centre Universite Paris-Sud,
F-91405 Orsay, France
Rad51 protein is an eukaryotic homologue of RecA protein and plays an essential role in homologous recombination like prokaryotic RecA protein. We have produced yeast Rad51 protein in E. coli and purified it to investigate the biochemical behaviors of this protein. The interaction of Rad51 with DNA was analyzed by fluorescence change of (dT)36 to which fluorescein is attached at the 5' end. We found that Rad51 requires high Mg2+ concentration (10 mM) for the interaction both in the presence and absence of ATP cofactor, which is in contrast with the fact that E. coli RecA requires less than 1 mM Mg2+. Mg2+ ion as well as ATP have been found to increase the thermal stability of Rad51 showing the presence of direct interaction of Mg2+ with the protein. The oligo(dT36) binding is stimulated by ATP and at lower pH (pH 6.4) as like the case of E. coli RecA/DNA interaction.
Binding Mode and Base Specificity of Norfloxacin with DNAs
Gwan-Su Son, Jeong-Ah Yeo, Mi-Sun Kim and Seog K. Kim*
*Department of Chemistry College of Sciences,
Yeungnam University,
Kyoungsan City, Kyoung-buk,
712-749, Republic of Korea
The binding properties of norfloxacin to calf thymus DNA and various synthetic polynucleotides were studied by optical spectroscopic methods including normal absorption spectroscopy, fluorescence techniques and linear dichroism. Norfloxacin forms complex with double stranded calf thymus DNA without ATP or Mg2+ mediation. Angles between the molecular plane of norfloxacin measured by linear dichroism are about 65 with respect to the helix axis. The equilibrium constant for the norfloxacin-DNA complex formation measured by Stern-Volmer method is about 3.0 ¥ 103 M-1 at 25C. The binding affinity to poly(dG)(dC) and poly[(dG-dC)2] is the highest among various polynucleotide, namely poly[(dG-dC)2], poly(dG)(dC), poly[(dA-dT)2], poly(dA)(dT), and DNA. That to denatured single stranded DNA is higher than to double stranded DNA. The sodium dependence for the norfloxacin complex formation with DNA indicates that binding of one norfloxacin molecule result in one Na+ ion release from DNA.
Structure-Based Design of Nucleic Acid Binding Ligands
Qi Chen, Irwin D. Kuntz and Richard H. Shafer
Department of Pharmaceutical Chemistry,
School of Pharmacy,
University of California,
San Francisco, CA 94143
We describe recent results based on implementation of the DOCK algorithm to design small molecules with affinity for specific nucleic acid structures. This approach matches the shape of the target nucleic acid structure to the shape of molecules contained in a database of small molecule structures. Our initial target was a groove in a dimeric hairpin quadruplex formed by the oligonucleotide d(GGGGTTTTGGGG). DOCK analysis led to a ligand with strong binding to this quadruplex with some degree of specificity. A high resolution NMR solution structure for this complex has recently been determined, revealing interactions with the quadruplex loop as well as groove regions. A second nucleic acid target was the deep groove in the RNA double helix. DOCK calculations led to several aminoglycosides which exhibit a strong preference for RNA duplexes in comparison to DNA duplexes, as determined by UV melting analysis. NMR experiments provided evidence for the binding of these compounds in the targeted deep groove of duplex RNA. These results demonstrate the effectiveness of the DOCK program in developing ligands capable of specific recognition of nucleic acid structures. It is a pleasure to acknowledge financial support from the National Institutes of Health.
Optical Imaging Approach to Study Protein-DNA Interaction and Nuclear Organization in Cultured Living Cells
Han Htun, Julia Brsony* and Gordon L. Hager
Laboratory of Receptor Biology and Gene Expression,
Building 41/Room B517,
National Cancer Institute,
National Institutes of Health,
9000 Rockville Pike,
Bethesda, MD 20892
*Laboratory of Cellular Biochemistry and Biology,
N.I.D.D.K., N.I.H.,
9000 Rockville Pike,
Bethesda, MD 20892
The recent availability of a naturally fluorescent protein, the green fluorescent protein (GFP) from Aequorea victoria jelly fish, provides a general method for labeling proteins and allows direct observation of intracellular trafficking and localization of the chimeric protein in cultured living cells with minimal perturbation. When fused to steroid receptors, which are ligand-dependent transcription factors, the effect of ligand on cytoplasm-to-nucleus translocation, nuclear distribution, and gene targeting can be observed in real time in living cells. In the case of the glucocorticoid receptor (GR), a previously undescribed pattern of GR distribution is seen in the interphase nucleus that is dependent on the type of ligand used to activate the receptor. For the dexamethasone-bound, transcriptionally competent form of the receptor, discrete nuclear foci are observed; in contrast, for the RU486-bound, transcriptionally weak form of the receptor, the receptor accumulates in a diffuse pattern with no evidence of discrete nuclear foci. The use of a unique cell line, 3134, which contains two hundred copies of a perfect head-to-tail tandem array of the MMTV LTR-ras-BPV DNA, reveals a new pattern of GR localization that demonstrates that sequence-specific protein-DNA interaction may be directly observed in living cells. An optical imaging approach is described based on GFP tagging of DNA binding proteins and the localization of the chimeric protein on an amplified DNA tandem array to study sequence-specific protein- DNA interaction, nuclear organization, chromatin decondensation, chromosome movement, transcriptional regulation, and alternate DNA structure formation in cultured living cells.
NMR Strategies for Structure Based Drug Design
Jonathan M. Moore
Vertex Pharmaceuticals Incorporated,
130 Waverly Street,
Cambridge MA 02139-4242
E-mail: moore@vpharm.com
Successful inhibitor design via a structure based approach requires the generation and incorporation of structural information from the diverse disci- plines of X-ray crystallography, NMR spectroscopy, and computational chem- istry. Although in practice NMR techniques are not able to generate high resolution structures of a protein target or complete target-inhibitor complex as rapidly as X-ray crystallographic methods, they are able to provide structural information for bound inhibitors rapidly enough to be useful in an iterative drug design approach. For example, by isotopically labeling either the ligand or protein target, the conformations of tightly bound inhibitors may be quickly determined using isotope editing or isotope filtering techniques, respectively. Alternatively, transferred NOE (tNOE) techniques are capable of characteriz- ing the conformations of weakly bound inhibitors or substrates. tNOE tech- niques are particularly useful in the early stages of a drug design program, where structural information is limited, and, due to weak binding, X-ray tech- niques may have limited success. Another area in which NMR may provide information not available via X-ray crystallography is in the characterization of protein and ligand dynamics via heteronuclear relaxation experiments. 13C and 15N relaxation experiments are capable of characterizing the motions of the protein target in both the apo- and ligand bound states, on several distinct time scales. Examples illustrating the utility of all of the above NMR methods in drug design programs will be given for two drug targets involved in immuno- suppression, the FK506 binding protein (FKBP-12) and inosine 5'-monophos- phate dehydrogenase (IMPDH).
Solution Structure of an Oligodeoxynucleotide Duplex Containing the Exocyclic Lesion 3, N4-Etheno-2'-deoxycytidine Opposite Deoxyguanosine: Determined by NMR Spectroscopy and Restrained Molecular Dynamics
David Cullinan, Moshe Eisenberg, Arthur P. Grollman
and Carlos de los Santos
Department of Pharmacological Sciences,
State University of New York at Stony Brook,
Stony Brook, NY 11794-8651
The solution structure of the d(C-G-T-A-C-eC-C-A-T-G-C).d(G-C-A-T-G-G-G-T-A-C-G) duplex containing the exocyclic 3,N4-etheno-2'-deoxycytidine (eC) adduct was determined by high-resolution NMR spectroscopy and molecular dynamics simulations. NMR data indicated that the duplex adopts a right-handed helical structure having all residues in anti orientation around the glycosidic torsion angle. The eC adduct has a sugar pucker on the C3'-endo/C4'-exo region while the rest of the residues are on the C2'-endo/C3'-exo range. NOE interactions established Watson-Crick alignments for canonical base pairs of the duplex. The G17(N1H) proton (the imino proton of the guanine opposite the adduct) resonated as a sharp signal, was resistant to water exchange and showed distance connectivities with protons on flanking residues, suggesting hydrogen bonding at the lesion site. Restrained molecular dynamics simulations generated three-dimensional models in excellent agreement with the spectroscopic data. The refined structures are slightly bent at the lesion site without major perturbations on the sugar-phosphate backbone. The adduct is displaced and shifted toward the major groove of the helix while its partner on the complementary strand remained stacked. The eC(anti).dG(anti) base pair alignment is highly sheared and stabilized by the formation of hydrogen bonds. The different conformations of the eC opposite dG, dA, and T (previously reported) are reflected in changes in chemical shifts.
1H - NMR and Molecular Dynamics Studies of a Model Hexapeptide GRGDTP
R. P. Ojha and A. K. Kulkarni
Biophysics Unit, Department of Physics,
University of Gorakhpur,
Gorakhpur - 273 009, India
The primary sequence of Arg-Gly-Asp has been found in a number of protein which binds to the cell surface receptors. Studies with synthetic peptides have shown that the presence of charged side chains is not sufficient to confer binding activity. Application of folding algorithms to proteins and peptides having similar sequences indicates that binding activity is strongly correlated with the presence of two or more closely spaced residues that each have a high probability of initiating a beta - bend. The proton NMR spectra of a hexapeptide GRGDTP, whose sequence is contained in fibronectin and shows binding activity, in aqueous solution ( 10 % D2O) show that the two amino protons of Asp and Thr exchange very slowly. Computer assisted modelling using restrained molecular dynamics and energy minimization results in conformation that include one bet- bend of type II fully consistent with constraints imosed by 1H-NMR data. Also one open - gama turn has been observed following molecular mechanics and molecular dynamics studies. It is suggested that this unusual secondary structure of beta-bend as well as gama-turn provides an additional specificity determinant.
Molecular Dynamics Simulations of the DNA Triplexe D(TG)5.D(CA)5.D(TG)5
R. P. Ojha and R. K. Tiwari
Biophysics Unit, Department of Physics, University of Gorakhpur
Gorakhpur - 273 009, India
The structure of the DNA triplex d-5'(TGTGTGTGTG).d-5'(CACACACACA).d-5' (TGTGTGTGTG) has been determined by computer simulation. the structure has been studied both in vacuum and in water. The simulation has been performed on a Power indigo2 using AMBER forcefield and includes counterions and explicit solvent under periodic boundary condition. The helical parameter of DNA sequences have been calculated using the programe Newhelix91. It has been observed that the third strand bind in the major groove of DNA duplex structure in such a way that the presence of third strand is recognised by both the strands of the double stranded DNA. It is suggested that such structures are possible because of the sequence specificity of nucleic acids.
K. T. Douglas
School of Pharmacy and Pharmaceutical Sciences,
University of Manchester,
Manchester, M13 9PL, UK
In contrast to rational drug and ligand design based on molecular graphics for the field of proteins, the situation for nucleic acid ligand design is less advanced, but is now proceeding rapidly. For DNA the aspect most developed in this sense is the minor groove, but for RNA there is still rather little in the way of ligand design reported.
Insight into the DNA minor groove has reached the stage at which it is now possible to test the possibility of rational ligand design in a number of objective ways. The first stage of this is to rationalise observations already made, but more rigorous is to test the ability to predict structural and chemical properties. In this contribution the ability to predict binding interactions for ligands in the minor groove of B-DNA of a series of analogues of Hoechst 33258 will be analysed.
These compounds were designed using molecular graphic/dynamics based on high-field NMR structural determinations of Hoechst: duplex DNA complexes using a synthetic oligonucleotide sequence. Enthalpy and entropy as contributors to net binding strength will be considered.
The test of prediction powers is not merely to be able to achieve better net ligand binding strength, but also to propose specific bonding interactions. These predictions can be probed structurally using NMR analysis at high resolution of ligand: DNA complexes, again designed by molecular modelling. As well as using structural probes such as NMR spectroscopy, it is possible to test predictive ability by introducing novel chemical reactivity. In this context we shall describe a novel DNA strand-cleaving method, designed using molecular graphics of the above structures to locate a transition metal ion binding site very specifically and close to the phosphodiester backbone, allowing the generation of reactive free radicals to effect cleavage.
Relative to DNA, RNA-binding ligands are less widely studied at present and, in the final part of the contribution, the binding properties of some new ligands for tRNA, binding with 1:1 stoichiometry and low dissociation constants will be described. Their binding has been studied by UV-visible spectrophotometry and florescence titration.
The DNA Helical Phasing Between Promotes and Regulatory Sites of Escherichia Coli
Vladimir Yu. Shumilov
K.A.Timiryazev Institute of Plant Physiology,
Moscow, Russia
Initiation of transcription is the main control point of gene expression. In Procaryota, such regulation is provided by both activator or repressor proteins and by the RNA-polymerase sigma-subunit changes. Ten years ago I have found a discrete helical phasing between CRP-cAMP recognition sites (sCRP) and promoters [1]. Now four classes of phasing are observed for all known activator and repressor proteins of Escherichia coli (due to some correction of its DNA recognition sequences). For the Class I phasing, activatory sites are located upstream "-35" boxes of promoters, so that phase distances, Nf, (between 3'-end of site and 5'-end of "-35") are equal to integer numbers of DNA helix turns (Upstream Even positions: A0, A1, A2, A3, A4...An), whereas repressing operators are occupied Downstream Odd positions (R0,5 - R10,5) inside promoters, with Nf equal to half-integer DNA turn numbers. It includes global cellular regulons: CRP, Fnr, Lrp, LexA, OxyR, SoxS, NtrC, NifA, NarL, NarP, FadR, FruR, MarA, MarR, IHF, Fis, FurA and CpxR/AlgR1) as well as local pathway activators and repressors of E. coli (MalT; AraC, RhaR, RhaS, MelR; UhpA, FhlA; LacI, GalR, GalS, GlpR, NagC, MalI, DeoR, CytR, RbsR, RafR, ScrR, ExuR, UidR, UxuR; KdpE, NhaR, CadC; TreR, OstR; TetR, ArsR, MerR, MerD; PspA; SdiA; IclR, PdhR, LctP; GcvA, CynR, CysB, Cbl, MetR, LysR, XapR, DsdC, TdcA, TdcR; AsnC, PutA; TorR; TyrR; TrpR, ArgR, PurR, MetJ, BirA, BetI, SpeX; HipB, DicC, F+ PifC; FlhCD), and phages ( lambda , 434, phi80 cI and Cro; Mu and D108 C, c and Ner; P2 Cox and cI; P22 Mnt and Arc). A Class III phasing is an extreme case of the former, where Downstream Even positions (DA1 - DA6) are yet activatory, while Downstream Odd ones are still repressory. There are both global (PhoB, PhoP; SoxR) and local (MerR*; IlvY, DicA, lambda cII, P22 c2, HK022 cI, Mu Mor) regulons. For the Class II phasing Upstream Odd positions (A0,5-A5,5) are activatory and Downstream Evens (R0-R10) are repressory. It includes as global regulators: OmpR, Ada, ArcA, DnaA, as local ones: CarP and phage P2 Ogr - P4 delta. For a proposed Class IV phasing (putative cold shock sites, sCspA) Downstream Odd positions are activatory. In all cases the phasing is very precise for activatory sites but not for repressory ones. The internal helical phasing between promoter structural elements is found also. For the Even promoter family, "-10" boxes are phased in a Class III-fashion relative "-35" ones (at the DA3 position). It includes promoters for most sigma-factors of E. coli (70, 38, 32, E, F and 19). In contrast, the only Odd promoter family member is recognized by sigma-54 with the Class IV-phased "-12" boxes at DA2,5 position relative its "-24" ones. The model "sabre in scabbard" is proposed for mechanisms of both transcription activation/repression via spatially determined contacts between the RNA- polymerase and regulatory proteins and strong /weak promoters selection by the only RNA polymerase itself. The phasing enables us to predict regulation mode (activation or repression) for any procaryotic promoter from the only its DNA sequence data. For example, a novel bacteriophage phiX174 promoter, PCRP1 has multiple-factor regulation is specified by three sigma-factors: 70, 38 and 32. PCRP1 activation is found by CRP [1] and predicted for PhoB, PhoP, OxyR, ArcA, Lrp, SoxS and CspA. PCRP1 repression is found for LexA, NtrC [1,2] and predicted for IHF, NarL and NarP.
References
1. V.Yu. Shumilov, Mol.Biol.USSR, v.21, p.200-.223
(1987).
2. V.Yu. Shumilov, Proc. Acad. Sci. USSR, v.308, p.1008-1012 (1989).
RNA Splicing, Introns and Biology
Phillip A. Sharp
Center for Cancer Research and Department of Biology,
77 Massachusetts Avenue,
MIT,
Cambridge, MA 02139
The genes of mammals contain an average of 10 introns whose length can extend 100,000 nucleotides. Precise excision of these introns is essential for gene expression. The sequence specificity for intron excision can be considered in two classes, microspecificity directing the chemistry of the splicing process, and macrospecificity defining the exon sequences to be joined during splicing. The conserved consensus sequences at the boundary of introns and at the branch site are the microspecificity. Interestingly the sequences at the 5' splice site are recognized twice during the splicing process, once by U1 snRNP and subsequently by U6 snRNA and other components of the spliceosome. During reactions in vitro it is possible to bypass the U1 recognition during formation of a spliceosome. By inference the utilization of this bypass pathway could be important in regulation in vivo.
A subclass of proteins that regulate RNA splicing contains a diamino repetitive sequence, Ser-Arg (SR). This repetitive subdomain is partially phosphorylated in vivo and is thought to associate with other SR domains in mediating formation of splicing complexes. Recent experiments suggest that there is an extended family of proteins with SR subdomains. These proteins are concentrated in subregions of the nucleus where RNA splicing occurs. Some members of the SR family are best pictured as lattice proteins which associate with multiple SR proteins bound to RNA. It is proposed that some aspects of both nuclear structure and gene structure are related to the activities of SR proteins in RNA splicing.
X-Ray Diffraction Study of Histone Proteins HMfA and HMfB from the Hyperthermophilic Archaeon Methanothermus Fervidus
Klaas Decanniere(1), Kathleen Sandman(2), John N. Reeve(2)
and Udo Heinemann(1)
(1)Forschungsgruppe Kristallographie,
Max-Delbrueck-Centrum fuer Molekulare Medizin,
Robert-Roessle-Strasse 10,
D-13122 Berlin, Germany
(2)Department of Microbiology,
Ohio State University,
Columbus, OH 43210
The archaeon Methanothermus fervidus grows at 83 C at an ionic strength of 1 M salt. It produces the abundant, basic proteins HMfA and HMfB which share significant sequence homology with the eukarya l core histones and form nucleosome-like structures (NLS) with DNA [1]. The structure of the protein complex involved and the DNA topology in the M. fervidus NLS, however, appear to differ from those of the eukaryal histone octamer and the nucleosome [2]. A recent NMR structure of HMfB [3] has confirmed the basic core histone fold to be adopted by the protein. In spite of their high degree of se quence similarity, HMfA and HMfB bind to DNA with different stoichiometries and induce different degrees of supercoiling in topologically restrained molecules [4].
Recently, we have produced several crystal forms of HMfA and HMfB [5] and determined the protein structures by X-ray diffraction methods at resolutions between 1.48 and 2.1 A. The crystal structures reveal subtle differences between the two proteins and serve to explain the biochemical differences between them.
References
1. Sandman, K., Krzycki, J.A., Dobrinski, B., Lurz, R.
& Reeve, J.N., Proc. Natl. Acad. Sci. USA 87, 5788-5791 (1990).
2. Musgrave, D.R., Sandman, K.M. & Reeve, J.N., Proc. Natl. Acad.
Sci. USA 88, 10397-10401 (1991).
3. Starich, M.R., Sandman, K., Reeve, J.N. & Summers, M.F., J. Mol.
Biol. 255, 187-203 (1996).
4. Sandman, K., Grayling, R.A., Dobrinski, B., Lurz, R. & Reeve, J.N.,
Proc. Natl. Acad. Sci. USA 91, 12624-12628 (1994).
5. Decanniere, K., Sandman, K., Reeve, J.N. & Heinemann, U., Proteins:
Struct. Funct. Genet. 24, 269-271 (1996).
Studies of Sequence-Dependent Tertiary Structure in Supercoiled DNA by Site-specific Recombination
S. D. Levene and H. Tsen
Program in Molecular and Cell Biology,
University of Texas at Dallas,
PO Box 830688,
Richardson, TX 75083
Intrinsically bent DNA sequences have been implicated in the activation of transcription by promoting juxtaposition of DNA sequences near the terminal loop of a superhelical domain. We have developed a novel topological assay for DNA looping based on bacteriophage lambda integrative (Int) recombination to study the effects of intrinsically bent DNA sequences on the tertiary structure of negatively supercoiled plasmids. Remarkably, the localization of adenosine-tract (A-tract) sequences in the terminal loop of a supercoiled plasmid is independent of the extent of intrinsic bending. The results suggest that negative supercoiling induces an alternative DNA conformation in A-tract-containing sequences, thereby conferring other properties that organize the structure of superhelical domains. The fact that looping is independent of A-tract phasing may explain the lack of conservation of A-tract-dependent bending among DNA sequences located upstream of bacterial promoters.
Interactions of the Antiviral Quinoxaline Derivative 9-OH-B220 {2,3-dimethyl-6-(dimethylaminoethyl)-9-hydroxy-6H-indolo-[2,3-b]quinoxaline} with Duplex and Triplex Forms of Synthetic DNA and RNA
Ulrica Sehlstedt(1), Palok Aich(1),* Jan Bergman(2),
Hans Vallberg(2), Bengt Nordén(3) and Astrid Gräslund(1)
(1)Department of Biophysics,
Stockholm University,
S-106 91 Stockholm, Sweden
(2)Department of Organic Chemistry,
NOVUM, CNT,
S-141 52 Huddinge, Sweden
(3)Department of Physical Chemistry,
Chalmers University of Technology,
S-412 96 Gothenburg, Sweden
*Present address: Medicinal Physics,
Department of Medical Biochemistry and Biophysics,
Karolinska Institute,
S-171 77 Stockholm, Sweden
The binding of an antiviral quinoxaline derivative, 2,3-dimethyl-6-(dimethylaminoethyl)-9-hydroxy-6H-indolo-[2,3-b]quinoxaline (9-OH-B220), to synthetic double and triple helical DNA [poly(dA)·poly(dT) and poly(dA)·2poly(dT)] and RNA [poly(rA)·poly(rU) and poly(rA)·2poly(rU)] has been characterized using flow linear dichroism (LD), circular dichroism (CD), fluorescence spectroscopy, and thermal denaturation. The negative reduced LD displayed by all complexes is consistent with intercalation of the ring system between the base pairs/triplets of both DNA host structures and the RNA duplex, a geometry that is in agreement with the observed weak induced CD signals and strong increments of the fluorescence emission intensities upon binding of the drug to each of these nucleic acid polymers. Furthermore, 9-OH-B220 is found to effectively enhance the thermal stability of both the double and triple helical states of DNA as well as the RNA duplex. On the other hand, when the RNA triplex serves as host structure, the average orientation angle between the drug chromophore plane and the helix axis of the triple helical RNA is roughly 63. Moreover, the thermal stabilizing capability, as well as the fluorescence increment and perturbations of the absorption envelope, of 9-OH-B220 in complex with the RNA triplex is smaller than those observed for the other types of complexes. These spectroscopic features indicate that the drug most likely not is intercalated into the RNA triple helical structure. A binding geometry model for the 9-OH-B220-RNA triplex complex in consistency with the present data is suggested, and the implications of the results for the antigene/antisense strategies are discussed.
Chemical structure of 9-OH-B220.
Covalently Dimerized DNA-Binding Domains of the Phage 434 Repressor: A Novel Protein Framework to Study Protein-DNA Interactions
András Simoncsits and Sándor Pongor
International Centre for Genetic Engineering and Biotechnology (ICGEB),
Area Science Park, Padriciano 99, 3
4012 Trieste, Italy
E-mail : simoncs@icgeb.trieste.it, pongor@icgeb.trieste.it
A novel single-chain protein framework containing covalently dimerized N-terminal DNA-binding domains (DBD) of the phage 434 repressor has been designed and constructed. The prototype RR69 contains two wild-type DBDs, linked with a peptide linker in a head-to-tail arrangement. It was obtained by expression of a recombinant gene coding for tandem repeats of the N-terminal 1 to 69 amino acid residues and a natural linker segment of the intact 434 repressor. Compared to the natural 434 repressor cI, it has a substantially reduced molecular mass and a simpler architecture. These properties allowed us to study the conformational changes taking place upon binding to specific and non-specific DNA by CD spectroscopy and to show that a scanning recognition mechanism is plausible also for the helix-turn-helix DNA binding proteins [1]. A combination of in vivo and in vitro assays showed that the covalent linker can functionally replace the non-covalent dimerization domain of the natural repressor: the single-chain RR69 and the natural 434 repressor recognize isolated operator sites with the same specificity and affinity.
The single-chain architecture allows for the introduction of independent changes into the individual DBDs, and thereby to construct DNA binding proteins which recognize nonpalindromic DNA operators. A heterodimeric single-chain repressor (RR*69) has been constructed by replacing the DNA-contacting amino acid residues of the a3 helix of the 434 repressor with the corresponding residues of the related P22 repressor. This molecule recognized an asymmetric, 434-P22 hybrid operator both in vivo and in vitro. Further substitutions in the unchanged domain of RR*69 resulted in a new homodimeric single-chain repressor R*R*69, which recognized a symmetric P22 operator site [2].
We have used binding site selection, targeted operator mutagenesis and binding affinity techniques to define the optimum DNA binding site for RR69 and RR*69. These studies confirmed that the homodimeric single-chain repressor binds to symmetric, while the heterodimeric single-chain repressor binds to asymmetric operators. The contacted, outer operator boxes are separated by a noncontacted spacer region of constant length (6 bp). By altering this spacing, the binding affinity was dramatically decreased, indicating that the interdomain contacts observed in the 434 repressor - operator complexes are maintained in the DNA-bound single-chain repressors. Further, it was found that the high affinity binding is also dependent on the non-contacted base pairs. Thus, RR*69 represent an example for the combination of altered direct and unchanged indirect recognition mechanisms.
Combinatorial single-chain repressor libraries containing randomized amino acids at the place of the DNA-contacting residues in one of the domains were constructed. By using in vivo selection techniques, single-chain proteins showing high affinity binding for either symmetric or asymmetric DNA targets could be isolated. Such studies may provide useful reagents for the specific recognition of long DNA targets and reveal new recognition rules for amino acid side chain - base pair interactions in a given domain framework.
References
1. Percipalle, P., Simoncsits, A., Zakhariev, S., Guarnaccia,
C., Sanchez, R. and Pongor, S. EMBO J., 14, 3200-3205 (1995).
2. Simoncsits, A., Chen, J., Percipalle, P., Wang, S., Törö, I.
and Pongor, S. J. Mol. Biol., in press (1997).
DNA Structures Associated With Class I Expansion of GAA in Friedreich's Ataxia
Cynthia T. McMurray, A. Marquis Gacy, Geoff M. Goellner,
Craig Spiro, Roy Dyer, Marci Mikesell, Janet Z. Yao, Aaron J. Johnson, Nenad
Juranic, Slobodan Macura, Andrea Richter* and Serge B. Melancon*
Mayo Foundation,
Rochester, MN,
*Universite de Montreal,
Montreal, Quebec, Canada
Until the discovery of the GAA repeat instability in intron 1 of the frataxin gene corresponding with Friedreich's Ataxia (FRDA), all trinucleotide repeat diseases displayed an expanded CNG repeat pattern, where N=A,C,T,G. We find that the defect causing GAA expansion in frataxin is directed by the DNA itself, as with other trinucleotide repeat diseases, but through a different structural mechanism.
Intergenerational instability occurs frequently within long GAA/TTC repeats in frataxin whose copy number correlates with the onset of FRDA, while transmission of the normal length allele is stable. In contrast, multiple microsatellie instability at other loci as in colon cancer is not observed within FRDA families. Patterns of instability in genetic/family data indicate that unstable transmission occurs through an intra-allelic mechanism.
GAA/CTT repeats are capable of forming an intramolecular triplex structure in vitro. GAA repeats are capable of forming a parallel, intermolecular double helix in vitro. GAA/CTT repeats are not capable of in vitro hairpin formation as observed with CNG repeats.
The combination of genetic and structural evidence suggest a model by which DNA secondary structure may facilitate expansion in the GAA repeat of the frataxin gene. Further, examination of the frataxin locus in Friedreich Ataxia reveals that DNA instability occurs by an intra-allelic process.
Structural Ensemble of a Methylphosphonated-DNA·RNA Hybrid: Analysis of NMR Data
Anwer Mujeeb and Thomas L. James
Departments of Pharmaceutical Chemistry and Pathology,
University of California,
San Francisco, CA 94143-0446
We have been investigating the structure of an antisense DNA·RNA hybrid by NMR. This trisdecamer hybrid duplex contains alternating Rp-methylphosphonates in the DNA backbone. The analysis of NMR data indicates that the hybrid duplex has a DNA strand with significant conformational plasticity. A sixth-root-based residual factor analysis of experimental 2DNOE intensities was successfully used to generate the envelope of conformational preferences of the individual residues. In case of conformational flexibility, 2DNOE data embody time-average information and this may lead to internal inconsistencies in structural restraints. During NMR structure-refinement process, one can generate sets of three-dimensional structures, which would then fit NMR data better as an ensemble rather than a single conformer. We have used PARSE (Probability Assessment via Relaxation rates of a Structural Ensemble) and time-averaged restrained-Molecular Dynamics (r-MD) methods to generate and refine such structural ensembles for the antisense hybrid. We will present our results on the refinement procedure and the characteristics of the structural ensemble.
Impact of pi - sigma* Interactions on the Axial Preference of the Purine vs. Pyrimidine Nucleobase in 2',3'-unsaturated Pentopyranosyl Nucleosides
Janez Plavec(1), Matjaz Polak(1) and Piet Herdewijn(2)
(1)National Institute of Chemistry,
Hajdrihova 19,
SI-1115 Ljubljana, Slovenia
(2)Laboratory of Medical Research,
Rega Institute,
Katholieke Universiteit Leuven,
Minderbroedersstraat 10,
B-3000 Leuven, Belgium
The conformational properties of 2',3'-unsaturated pentopyranosyl nucleosides 1-101 have been studied by 1H-NMR spectroscopy at 300 MHz in aqueous solution in the temperature range from 278K to 358K.
The analysis2,3 of the experimental 3JHH coupling constants has shown that the constituent pyranosyl moiety is in 1 - 10 involved in a rapid two-state 6H5 (half-chair) 5E/5S4 (intermediate between envelope and screw-boat canonical forms) conformational equilibrium. The assumed two-state conformational equilibrium is in agreement with the available X-ray data and our HF/3-21G ab initio calculations on 1 - 3 in vacuo. At room temperature b nucleosides 1 - 5 adopt predominantly (95% in purine and 80% in pyrimidine nucleosides) the conformation which is intermediate between 5E and 5S4 canonical forms, where both the nucleobase and 4'-OH are in the axial orientation. The pyranosyl moiety of a-purine nucleosides 6 and 7 is involved in an unbiased two-state conformational equilibrium. The a pyrimidine nucleosides 8 - 10 show a conformational bias of ca. 75% towards 5E canonical form. The van't Hoff type analysis of the population changes in the conformational equilibrium with temperature afforded the thermodynamic data on 6H5 5E/5S4 conformational equilibria in 1 -10.
The conformational equilibria in 1 - 10 are enthalpy driven and tuned by the following individual contributions: O6'-C1'-N1/9 anomeric effect, gauche effect of [O6'-C5'-C4'-O4'] fragment, the interaction between p-system of the C2'-C3' bond and both the heterocyclic aglycone and 4'-OH in allylic positions and the intrinsic steric effect of the nucleobase. The dissection of the individual contributions to the drive of the conformational equilibrium in 1 - 10 has shown that the relative strength of pi - sigma*C1'-N9/1 interactions in purine vs. pyrimidine nucleobases increases in the following order: cytosine < uracil < thymine < adenine < guanine.
References
1 B. Doboszewski, H. De Winter, A. Van Aerschot and P.
Herdewijn Tetrahedron, 51, 12319 (1995).
2 C.A.G. Haasnoot J. Am. Chem. Soc, 115, 1460-1468 (1993).
ibid., 114, 882-887 (1992).
3 C. Altona, R. Francke, R. de Haan, J.H. Ippel, G.J. Daalmans, A.J.A.W.
Hoekzema and van J. Wijk Magn. Reson. Chem., 32, 670 (1994)
Alexander Rich
Biology Department
Massachusetts Institute of Technology
Cambridge MA 02139
I will touch upon a number of research topics covered in my laboratory during the last four decades or so working on nucleic acids. What becomes apparent is that in the early years very large and basic questions were asked, such as: Can RNA form a double helix? How is protein synthesis carried out? What does transfer RNA look like? In more recent years, however, increasingly specialized questions have been addressed, such as the biological implications of having DNA existing in alternate non-B conformations. Does nature utilize them? And if so, how?
The change in the type of question asked, of course, reflects the enormous maturation which the discipline has undergone. Today, it is no longer a question of whether we can explain living organisms at the molecular level. Rather, the question is how detailed our understanding can become and in how comprehensive a manner we can look at all the phenomena at the molecular level and integrate them into an uderstanding of the organism as a whole.
In this effort basic studies dealing with the nucleic acids will continue to play an important role, even though the main focus of activity increasingly has to do with the nature of the proteins, encoded by the nucleic acids and the manner in which these proteins interact, especially with the nucleic acids, to influence the behavior of biological systems.
Back to Table of Contents: Abstracts
Back to Table of Contents: Tenth Conversation