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 examople, alpha = a, beta = b and so on.
Evolution of Human and Mouse Dinucleotide Repeat DNA
Pankaj Agarwal(1), George I. Bell(2), Gavin Huttley(3)
and Michael W. Smith(4)
(1)SmithKline Beecham Pharmaceuticals R&D,
UW2230, 709 Swedeland Road,
PO Box 1539,
King of Prussia, PA 19406-0939
E-mail: agarwal@mh.us.sbphrd.com
(2)Theoretical Biology and Biophysics Group (T-10),
MS K710, Los Alamos National Laboratory,
Los Alamos, NM 87545
E-mail: gib@t10.lanl.gov
(3)Laboratory of Viral Carcinogenesis --- NCI, Bldg. 560,
Frederick Cancer Research & Development Center,
Frederick, MD 21702-1201,
E-mail: huttley@ncifcrf.gov
(4)Biological Carcinogenesis and Development Program, SAIC Frederick,
National Cancer Institute,
Frederick Cancer Research & Development Center,
Frederick, MD 21702-1201,
E-mail: smithm@ncifcrf.gov
Microsatellites have become increasingly important in genetic mapping, disease gene identification, and the etiology of cancer (Dietrich et al. 1996; Mendel, 1994). An analytical evaluation of over 8000 loci from humans and mice is presented that will allow researchers to understand these loci and their properties and choose additional loci for characterization. Poly (CA) microsatellite loci were analyzed for both humans and mice. Sequence alignment and scoring techniques were modified to pinpoint the extent of the microsatellite. The classification scheme of Weber (1990) was extended to automatically classify these sequences based on their complexity. The heterozygosity of the repeat had been observed to correlate better with the longest uninterrupted run of CA's rather than with the total length of the microsatellite (Weber, 1990); this observation is extended to a much larger database including both simple and compound microsatellites in humans and mice. In addition, few uninterrupted runs of 30 CA's or more were found suggesting that large CA runs are broken up by substitutions; a prediction that we verified by sequencing additional alleles. Additional sequencing and a database search also revealed examples of alleles with significant non-microsatellite sequence duplication. This demonstrates that not only does the microsatellite length change at these loci, but some of the neighboring sequence get duplicated as well.
B-DNA Bending by Cationic Ligands
I.F.Rouzina and V.A. Bloomfield
University of Minnesota,
St.Paul MN 55108
E-mail: rouzina@biosci.cbs.umn.edu.
A large body of the recent experimental data suggests that binding of the cationic ligands to the double-helical DNA often results in its bending. The direction of this bend is such as if the negatively charged DNA were wrapping around the positively charged ligand. The magnitude of such bend correlates with the ligand's charge. It can reach significant values up to 900 without perturbing the B-DNA structure. The latter means that the bend occurs over several base pairs. All of these features make one seek for the electrostatic explanation of the effect. Despite of its seeming simplicity even the basic understanding is lacking.
We show that the physical reason for such bending is the local charge inhomogeniety at the DNA surface created by the multivalent cation. In contrast to the smooth uniform screening by monovalent sodium ions, the multivalent cation strongly repels other screening cations, creating around itself the correlation hole with the excessive negative charge of the background DNA phosphates. This results in the DNA tendency to warp around the multivalent cation which is opposed by its natural rigidity. We develop a simple analytical description of this effect which provides the reasonable semiquantitative account of the available experimental and computer simulated data. The main conclusions of this study are:
The bend is local; several bends associated with different multivalent cations are independent and additive. Bending involves DNA fragment of the size of the ligand's electrostatic correlation hole, which is ~15-20Å long, i.e. spans 5-10 bp. The equilibrium static bending angle and its stabilizing energy increase with the total charge of the ligand and its compactness. Electrostatic bending is stronger in the regions of the lower local DNA bending rigidity. Such bending has both: static as well as dynamic components, but the later is much less significant.
Specific DNA-ligand interactions are not required for this type of bending, since correlation hole exists around the nonspecifically electrostatically bound mobile cation as well. Sequence specificity of electrostatic bending resides not in the DNA-ligand interaction, but within the sequence dependent DNA rigidity. Bending contributes to the stronger binding of the cationic ligand to regions of the local higher flexibility.
From the dynamic point of view electrostatic bending is a cooperative motion of ~10 bp DNA fragment which requires relatively long time during which the counterion should not move. For the mobile non-site bound cation there is a possibility to get self localized by the DNA bend, forming the "bending polaron". The later requires strong enough electrostatic interaction and/or high local flexibility. This is achieved when the counterion with at least +2 charge self localizes within the DNA groove.
The large body of experimental data can be interpreted
within the electrostatic bending mechanism. For example it can account for
the strong reduction of the apparent DNA persistence length below its limiting
high salt value of 50nm down to 20-30 nm upon addition of the small amounts
of multivalent counterions to DNA solution. It also explains the sequence
specificity of the " nonspecific electrostatic" cation binding
associated with the local DNA bending. Our conclusions are relevant to DNA
bending by the basic region binding proteins, and by the neutralized phosphate
patches on one DNA face. Results similar to ours were obtained in the molecular
mechanics simulation of the spermine-DNA binding.
Self-Assembly of DNA Oligomers with XGG Trinucleotide Repeats
Fu-Ming Chen
Department of Chemistry,
Tennessee State University,
Nashville, Tennessee 37209-1561
Comparative studies of molar K+-induced aggregate formation of d(XGG)4, where X = A, T, or C, and some related oligomers were carried out to investigate various aspects of the self-assembly processes of DNA having such motifs. Turbidity measurements via absorbance monitoring at 320 nm were employed to obtain kineitc profiles which exhibit autocatalytic-like behaviors consisting of initial lag periods followed by accelerative and leveling phases. At pH 8, the relative propensity for aggregation is shown to be in the order: d(AGG)4 > d(TGG)4 > d(CGG)4. The oligomer with CGG repeats, however, is greatly facilitated by acidic conditions. Although both d(CGG)4 and d(TGG)4 exhibit sizable psi-type CD, their spectral features are strongly dependent on the ionic conditions. Interestingly, d(AGG)4 fails to exhibit significant y-CD formation despite its ease in the K+-induced aggregate formation. The presence of Mg++ greatly facilitates and dramatically reduces the requirement of K+ in the aggregation process. Effects of these two cations on aggregation appear to be synergistic in nature. Thermal stabilities of aggregates are strongly dependent on the concentrations of these two ions. Replacements of K+ by Na+ and/or Mg++ by other divalent cations fail to induce simialr phenomena. A mechanism for the self-assembly of these oligomers is speculated.
Two-domain Structure of the td Intron-encoded Endonuclease I-TevI Correlates with the two-Domain Configuration of the Homing Site
Victoria Derbyshire, Joseph C. Kowalski, John T. Dansereau, Susan Baxter, Charles R. Hauer and Marlene Belfort
Molecular Genetics and Structural Biology Programs,
Wadsworth Center, PO Box 22002,
New York State Department of Health,
Albany, NY 12201
I-TevI, the T4 td intron-encoded endonuclease, catalyzes the first step in intron homing by making a double-strand break in the intronless allele within a sequence designated the "homing site". The 28-kDa enzyme, which interacts with the homing site over a span of 37 bp, binds as a monomer, contacting two domains of the substrate. In this study, limited proteolysis experiments indicate that I-TevI consists of two domains that behave as discrete physical entities as judged by a number of functional and structural criteria. Overexpression clones for each domain were constructed and the proteins were purified. The carboxy-terminal domain has DNA-binding activity coincident with the primary binding region of the homing site and binds with the same affinity as the full-length enzyme. The isolated amino-terminal domain, which is toxic when expressed in Escherichia coli, contains the conserved GIY-YIG motif, consistent with its being the catalytic domain. Furthermore, site-directed mutagenesis of a conserved arginine residue within the motif eliminated the toxicity of the N-terminal domain and rendered the full-length protein catalytically inactive, although DNA binding was maintained. This is the first evidence that the GIY-YIG motif is important for catalytic activity. An enzyme with an N-terminal catalytic domain and a C-terminal DNA-binding domain connected by a flexible linker is in accord with the bipartite structure of the homing site. To better understand the structural and functional organization of the enzyme and its substrate, NMR studies on the individual domains and full-length enzyme are underway.
IS903 Insertion Specificity: A Distinct Preference for the F Transfer Operon
Wen-Yuan Hu and Keith M. Derbyshire
Department of Biomedical Sciences,
School of Public Health,
SUNY at Albany, Albany NY 12206
and
Molecular Genetics Program,
Wadsworth Center,
New York State Department of Health,
Albany NY 12201
IS903 is a small bacterial insertion sequence of 1057 bp. It is flanked by perfect 18 bp inverted repeats (IR) and encodes a 307 amino acid transposase protein which mediates the transposition process. IS903 generates 9 bp direct repeats on insertion into target DNA.
To test whether IS903 exhibits target preference, we have examined independent insertions into a 55-kbp F plasmid derivative, pOX38. It was chosen as a target since the majority of its sequence is known and transposition events can be selected by using a mating-out assay. The sites of insertion into pOX38 were determined by DNA sequencing analysis of the IR-target site junctions. Surprisingly, IS903 insertions were not randomly distributed around the plasmid, but instead, were clustered into a few small regions (1.2-3.8 kbp) in the transfer region (tra) and in the leading region adjacent to the origin of conjugal transfer, oriT. In addition, several sites were used more than once. Using a second IS903 derivative we showed that this regional preference is not influenced by the internal sequences of the transposon. Alignment of the 9-bp target sites and 10-bp flanking sequences were carried out to determine the significant frequency of A, T, G and C at each position. No consensus sequence within the 9 bp target was identified, however, in the flanking sequence a preference is seen for certain nucleotides that are symmetrically positioned around the insertion site. The observations indicate that IS903 transposition displays a distinct regional preference for target selection although the multiple use of some sites suggest they are composed of a preferred sequence.
To test whether the regional preference is specific to the plasmid F, we have determined the location of insertions in a second, unrelated conjugative plasmid pUB307, a deletion derivative of the IncP plasmid RP1. Independent transposition events in pUB307 have been identified and all insertions are broadly distributed around the plasmid, but in the same orientation. There is no insertion preference for the tra region and the leading region adjacent to oriT in general suggesting that the insertion preference observed with pOX38 is unique to the plasmid F.
To examine whether the insertion specificity observed in pOX38 can be detected in a different plasmid context, we have subcloned individual segments of the F plasmid into pUB307. We have found that the insertion distribution in pUB307 was changed dramatically in the presence of a 1.1 kbp traY-L fragment from the F tra region. Almost 90% of insertions were mapped to this traY-L fragment and all were inserted in the same orientation as seen with pUB307. This orientation bias for insertion is not related to the orientation of the inserted tra fragment in pUB307. Our results indicate that all signals necessary for insertion specificity are contained by the traY-L fragment, but the unique orientation of insertions is determined by the plasmid not the tra fragment per se. We have also determined that the traY promoter (PtraY), the major transfer operon promoter, contained within the traY-L fragment is important for insertion specificity since deletion of the -35 and -10 region of PtraY reduces the number of insertions targeted to this segment by 40%. Our data show that the DNA context is important for target recognition, but it is likely that transcription from PtraY is an additional factor in enhancing the use of traY-L fragment as a target.
Mutational Analysis of IS903 Transposase
Norma Tavakoli and Keith Derbyshire
Molecular Genetics Program,
Wadsworth Center,
New York State Department of Health,
Albany, NY 12201
IS903 is a small insertion element of 1057bp, which transposes predominantly by a conservative "cut and paste" mechanism. It encodes a 307 amino acid transposase protein and is flanked by perfect inverted repeats (IR) of 18bps. The IR is composed of two functional domains; bp 6-18 contain the essential contacts for transposase binding and the outer 5bp although not essential for transposase binding are required for subsequent steps in the transposition process.
We have isolated IS903 transposase mutants that suppress the deleterious effect of IR mutations on transposition in order to identify those amino acids that play a role in establishing and maintaining protein-DNA contacts. The mutations were isolated using a papillation assay to screen a library of transposase mutants that could suppress specific IR mutations that abolish transposition. 25 independent suppressors were isolated that map to two regions of the N-terminus of the protein; between amino acids 24-32 (class I) and 110-128 (class II). We have also used the papillation assay to examine the specificity of suppression by individual mutations on different mutant ends. These experiments show that although the transposase suppressors work most efficiently on the mutant end from which they were isolated, they do suppress other mutant ends. Therefore, these changes are not altering specific protein-DNA contacts but are acting at a more global level. This is in agreement with the fact that most of the mutations were isolated as suppressors of the 3A terminus mutation which is involved in a step subsequent to IR binding. We are currently using the more quantitative mating-out assay to monitor allele specificity. Our preliminary data appear to support our papillation results in that the most effective suppression is seen with the mutant end the transposase suppressor was isolated on.
We tentatively group the suppressor mutants into two classes according to the pattern of suppression observed, their in vitro binding behaviour and their clustering in two specific regions of the protein. The distinct phenotypes of each class suggest that they suppress the defective end in different ways. Class I mutants include 9 different amino acid changes in the region between amino acids 24 and 32. These mutants tend to be more specific and suppress most but not all of the terminus mutations tested. Furthermore, the Leu to Pro mutation at amino acid 24 (LP24) and TR25, although able to mediate transposition in vivo, bind extremely weakly to an inverted repeat in vitro in a band-shift assay. Class II suppressor mutants includes 5 different amino acid changes in the region between amino acids 110 and 128. Interestingly, in each case the wild-type amino acid is replaced by a less hydrophobic residue. These mutations tend to have a more general suppressive effect in that they suppress all the terminus mutations tested. Finally, in contrast to the first class of suppressors these mutants are active in DNA binding assays. We have purified F110S and F128S proteins as MBP fusions and shown that they exhibit the same binding specificity as wild-type transposase.
A number of possibilities may explain the suppression effect. The mutations may help stabilize the protein; allow the protein to adopt a conformation which is less stringent in its recognition specificity; improve the protein's ability to oligomerise or stabilize transposase-terminus interactions.
In order to determine the amino acids involved in catalysis, seven amino acid residues within the C-terminus of IS903 transposase that show a high degree of conservation when compared with the IS4 family of insertion elements and the retroviral integrases were mutated to alanine (D154, D172, D193, Y196, Y252, R255 and E259). These include residues that are expected to be part of the DDE and YREK motifs which are conserved among many transposases and have been shown for certain transposases to be involved in catalytic reactions. The frequency of transposition of the mutants, as determined by mating out assays, was reduced significantly. All the mutants were shown to bind specifically to an IR and are therefore not compromised in binding. Based on the conservation it is expected that these residues play an important role in steps subsequent to IR binding.
The Investigation of DNA-Protein Interactions by Molecular Mechanics and Protein-DNA Cross-Links Methods. Deliniation of Universal DNA, RNA-Binding and DNA-Binding Motifs
Natalya G.Esipova, Dmitry E.Kamachev, Dmitry A.Kuznetzov
and Andrey D.Mirzabekov
Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Vavilov str. 32, 117984 Moscow
By the method of DNA-protein cross-linking some amino acid sequences of strong protein-nucleic acid affinity have been obtained. As a result of data bank analysis different proteins including these sequences have been selected. Interestingly, the symmetry of disposition of charged amino acids were observed in these sequences. Principles of organization of the data base have been formulated which are able to estimate the energetic characteristics of regions of DNA-protein interactions. Electrostatic characteristics and their variations in the sequences under study in the course of DNA-protein interactions were discussed. The electrostatic potential functions have been computed as well. Conclusions on favorable orientation of the complexes components in respect of electrostatic have been drawn. Method of DNA-protein cross-linking was used for determination of contact sites between lac-repressor and operator DNA or nonspecific DNA. Only Lys-33 forms cross-link in specific lac-repressor-lac- operator complex. In any nonspecific complexes additionally N-terminal alpha-amino group and Lys-2 form cross-links with DNA. The method for location of cross-linked amino-acid residues on DNA have been developed. For triple complex inducer-lac-repressor-operator sites of contacts between repressor N-terminus and operator are adenines 15 and 18 of ideal lac-operator. We suggest that the interaction with the operator leads to orientation of DNA-binding domains of repressor as in NMR experiment. Thus, mechanism of lac-repressor switch-off is discovered. Orientation of the lac-repressor DNA-binding domain is reversed upon inducer binding. For Tramtrack protein cross-linked amino-acid residues in complex of Tramtrack protein with specific and non-specific DNA were identified. N-terminal alpha-amino group is cross-linked to DNA. Location of N-terminus on the DNA-binding site have been determined through identification of nucleotides which were cross-linked to this alpha-amino group. Conformational changes of N-terminal residues as the protein is bound to DNA have been revealed. Adequacy of cross-linking method for study of dynamics of DNA-protein interaction have been demonstrated. The structure of DNA-binding domains in a complex Lac-repressor DNA was build on the basis of only crystallographically available C-alpha coordinates (the polupeptide chain from 1st to 61st residue). The determination of side-chain positions for the above-stated region in a field of atoms of both protein and DNA was undertaken. The further procedure of prediction of side-chain's conformation for the residues 1-61 of each subunit consisted of search by Monte-Carlo method with the subsequent minimization in a found local minimum. The resultant molecular mechanics findings are in close agreement with cross-links observation and posses to understand concrete interactions to be responsible for molecular recognition.
Structural Basis For Heterogeneous Kinetics: Reengineering the Hairpin Ribozyme
J. Esteban, N. Walter, G. Kotzorek*, J. Heckman, J.
Bond and J. Burke
Markey Center for Molecular Genetics,
Department of Microbiology & Molecular Genetics,
The University of Vermont,
Burlington, Vermont 05405, USA
*NAPS Göttingen GmbH,
Rudolf-Wisell Str. 28,
37079 Göttingen, Germany
We have recently completed a detailed steady-state and pre-steady state kinetic analysis of the cleavage and ligation reactions catalyzed by the hairpin ribozyme. In the course of these studies, we developed fluorescence-based assays capable of real-time monitoring of binding, cleavage and dissociation steps of the reaction pathway. Results define the kinetic mechanism of the ribozyme, and show that the cleavage kinetics are inherently biphasic. That is, single-turnover reactions using a large number of ribozyme variants show both a rapid cleavage phase and a slow phase. By fitting kinetic results to double-exponential rate equations, we can measure the rates associated with both phases, as well as the proportion of substrate that reacts in each phase. Chase experiments establish the presence of an inactive conformer that reversibly binds substrate and is responsible for the slow cleavage phase. The active and inactive conformers do not exchange over typical reaction time courses. Using RNA ligase to explore the proximity of 5' and 3' ends, we have found that the active form contains a sharp bend between helices 2 and 3, while these helices are stacked in the inactive conformer, preventing interaction of the two domains of the ribozyme substrate complex. Two different strategies were used to reengineer the ribozyme so as to prevent formation of the inactive conformer: (1) introduction of a new helix between H2 and H3, and (2) introduction of an ortho-benzene linkage. Each of these modifications is effective in preventing formation of the inactive, stacked conformation, and results in homogeneous kinetic behavior. The new hairpin ribozyme constructs are being used for structural analysis, kinetic and mechanistic studies, and possibly for intracellular RNA targeting studies.
Effects of Different Chemical Groups Covalently Attached to Oligodeoxyribonucleotides as well as the Presence of the Mismatch at these Junctions on the Co-operative Interactions
Abdussalam Adeenah-Zadah and Olga Fedorova
Institute of Bioorganic Chemistry,
Lavrentyev Pr. 8,
Novosibirsk 630090, Russia
Earlier we have developed an original approach to evaluation of the co-operativity parameters for the interaction of the oligonucleotides aligned on the complementary template1. The approach was developed based on chemical modification data. This work is the extension of the previous one1 and is devoted to the study the effects of different chemical groups covalently attached and located at the junctions as well as the presence of mismatch base pair at these junctions on co-operative interactions. For this purpose a double-stranded tandem system consisting of four oligodeoxyribonucleotides have been chosen. 26-meric oligonucleotide d(TTGCCTTGAATGGGAAGAGGGTCATT) (P) modeled a single-stranded nucleic acid target and had three binding sites for the oligonucleotides forming the complementary tandem sequence N1-X-N2 at the target (X= ClRCH2-pd(TTCCCA), where ClRCH2- is 4-[N-(2-chloroethyl)-N-methylamino]benzyl group, N1 was dp(TTCAAGGC) or Phn-dp(TTCAAGGC)-Phn and N2 was dp(TGACCCTC) or Phn-dp(TGACCCTC)-Phn (Phn is N-(2-hydroxyethyl)phenazinium group)). The effect of a mismatch on cooperative interactions was studied using the reactive oligonucleotide derivative ClRCH2NHpd(TCTTCCCT) (Y8) forming mismatch base pair TT at complex formation with the target P. The modification of the target P with the reagent X in tandem complexes with N1 and N2 was performed. The quantitative treatment of the modification data demonstrated the following results. 1) The efficiency of co-operative interaction of oligonucleotides increases by factor 3 in the presence of Phn-group covalently attached to oligonucleotide-effectors and located at the junctions with respect to its absence. 2) In the presence of alkylating group -CH2RCl at the junctions cooperative interactions of oligonucleotides is eliminated, apparently, due to the preventing of base stacking. 3) Moderate positive cooperative interaction takes place in the simultaneous presence of both Phn- and RCl- groups at the junction. 4) The presence of the TT-mismatch at the junction prevents cooperative interaction, apparently, due to the absence of base stacking originated from the absence of hydrogen bonds in the TT-mismatch base pair. 5) Co-operative interaction does occur in this region in the case of the simultaneous presence of Phn- group covalently attached and the TT-mismatch base pair formed by the reagent Y8 with the target P.
It was concluded that in our case the cooperativity originated not only from the stacking interaction of neighboring oligonucleotides but from the influence of the conformational transitions of target. The system under investigation represented an example of cooperative interaction of ligands with the biopolymers, similar to interaction of substrates with allosteric enzymes.
1) Fedorova O.S., Adeenah-Zadah A., Bichenkova E.V., Knorre D.G., Thermodynamic and Structural Features of Cooperative Interactions in Tandem Oligonucleotide Derivatives Arranged at the Complementary Template. Chemical Modification Data. Journal of Biomol. Struct. Dyn. 13, 145-166 (1995).
A Theoretical Analysis of Specificity of Nucleic Acid Interactions with Oligonucleotides and Peptide Nucleic Acids (PNAs)
Aleksey Lomakin(1) and Maxim Frank-Kamenetskii(2)
(1)Physics Department,
Massachusetts Institute of Technology,
Cambridge, MA 02139
(2)Center for Advanced Biotechnology and Department of Biomedical Engineering,
Boston University,
Boston, MA 02215
We treat theoretically a problem of specificity of interaction between a nucleic acid and an oligonucleotide or its analog. Simple models are considered, which are rid of all unnecessary details. The calculations are performed by the kinetic Monte Carlo method. Using our Model #1, anti-correlation is demonstrated between specificity and affinity for nucleic acid/oligonucleotide interaction. We conclude that oligonucleotides and their modifications are not efficient in recognizing specific sites on nucleic acids. Oligonucleotides are widely used in numerous techniques not because they are very efficient in targeting specific sites but because these technique do not require high specificity of recognition.
A notable exception from the general situation is studied in detail. Homopyrimidine PNAs form DNA/(PNA)2 triplexes with complementary DNA strand, which are exceptionally stable so the affinity is very high. At the same time, such PNAs show remarkable sequence specificity of binding to duplex DNA. A theoretical model of the process of homopyrimidine PNA interaction with nucleic acids is formulated (Model #2). The calculations demonstrate that the two-step process of PNA interaction with DNA, first postulated by Demidov et al. (1995, 1996), secures both, high affinity and very high specificity of PNA interaction with DNA. Our computer simulations made it possible to define the range of parameter values, in which high specificity is achieved. These findings are of great importance for numerous applications of PNA and for design of future drugs specifically interacting with DNA.
1. Demidov, V.V., Yavnilovich, M.V., Belotserkovskii, B.P.,
Frank-Kamenetskii, M.D. and Nielsen, P.E. (1995) Proc. Natl.Acad.Sci.
USA 92, 2637-2641 (1995).
2. Demidov,V.V., Frank-Kamenetskii, M.D. and Nielsen, P.E. (1996) in: Biomolecular
Structure and Dynamics, v. 2, pp. 129-134, eds.R.H.Sarma, M.H.Sarma, Adenine
Press, NY.
Supported by NIH grant GM52201
Kinetic Analysis of Specificity of Duplex DNA Targeting by Homopyrimidine PNAs
Vadim V. Demidov(1), Michael V. Yavnilovich(2) and Maxim
D. Frank-Kamenetskii(1)
(1)Center for Advanced Biotechnology,
Department of Biomedical Engineering,
Boston University,
36 Cummington Street,
Boston, MA 02215, USA,
(2)Department of Structural Biology,
Weizmann Institute of Science,
76100 Rehovot, Israel
A simple theoretical analysis shows that specificity of double-stranded DNA (dsDNA) targeting by homopyrimidine peptide nucleic acids (hpyPNAs) is a kinetically controlled phenomenon. Our computations give the optimum conditions for sequence-specific targeting of dsDNA by hpyPNAs. The analysis shows that, in agreement with the available experimental data, kinetic factors play a crucial in the selective targeting of dsDNA by hpyPNAs. The selectivity may be completely lost if PNA concentration is too high and/or during prolonged incubation of dsDNA with PNA. However, quantitative estimations show that the experimentally observed differences in the kinetic constants for hpyPNA binding with the correct and mismatched DNA sites are sufficient for sequence-specific targeting of long genomic DNA by hpyPNAs with a high yield under appropriate experimental conditions. Differential dissociation of hpyPNA/dsDNA complexes is shown to additionally enhance the selectivity of DNA targeting by PNA.
Supported by NIH grant GM52201
Salt Effects on DNA Structure and Dynamics Studied by H1-NMR Spectroscopy
Patrick Furrer, Nick Ulyanov and Tom James
Dept. of Pharm. Chem.,
University of California,
San Francisco 94143-0446, USA
In solution, DNA persistence length has been shown to be ionic strength dependent, as a stiffening of the double-helix can be achieved by reducing salt concentration (1-3). We're using high-resolution H1-NMR spectroscopy to study this salt dependence on a model duplex of 14bp (5'-GCCTTGAAGACAGC-3'), both in low (20mM phosphate buffer) and high (20mM phosphate buffer + 1M NaCl) salt conditions. In a first stage, we're comparing average interproton distances obtained from NOE intensities by a complete relaxation matrix approach, via the program RandMardi (an extension of Mardigras). The bounds for these interproton distances are further used as constraints in a refined monte-carlo simulation combined with Parse, a new probability assessment method (4). Correlations between higher flexibility and the conformations digged out from the conformational pool by the Parse algorithm will be discussed.
References
1. Porschke, D. (1991), "Persistence length and bending
dynamics of DNA from electrooptical measurements at high salt concentrations",
Biophysical Chemistry 40(2): 167-169.
2. Fenley, M. O., Manning G. S. and Olson W. K. (1992), "Electrostatic
persistence length of a smoothly bending polyion computed by numerical counterion
condensation theory", Journal Of Physical Chemistry 96(10): 3963-3969.
3. Furrer, P. et al. (1997), "Opposite effect of counterions on the
persistence length of nicked and non-nicked DNA", J. Mol. Biol. March
7.
4. Ulyanov, N. et al (1995), "Probability assessment of conformational
ensembles: sugar repuckering in a DNA duplex in solution", Biophysical
Journal 68(1): 13-24.
Studies on Direct and Indirect Hydrogen Bonds and the Operator Conformation in the Interaction of the trp Repressor with its Binding Sites
Avital Bareket-Samish, Ilana Cohen and Tali. E. Haran
Department of Biology,
Technion, Technion city,
Haifa 32000, Israel.
The interaction of the trp repressor with its binding sites, as revealed from the crystal structure of the repressor/operator complex, is the most striking example to date of indirect readout and structural recognition of DNA by sequence-specific DNA binding proteins (1,2). In the crystalline complex there are only two direct hydrogen bonds from the repressor to DNA bases. The protein contacts the DNA mainly through direct hydrogen bonds to the non-bridging oxygens of the phosphates on the DNA backbone and by indirect hydrogen bonds, mediated by water molecules, to both the DNA bases and the phosphate backbone. Earlier biochemical studies have found that mutating the base pairs at position +1, +8, and +9 had little effect on the binding of the trp repressor to its operator (3). Hence, the bidentate direct hydrogen bonds from Arg69 on the repressor to two G-9 bases on the operator were not considered crucial for specific binding.
We have studied the interaction of the trp repressor with several trp binding-site mutants (4) and have addressed the question of the role played by the base pairs at position +9 and +1 in forming the specific complexes of the trp repressor with its binding sites. We show that the identity of the dinucleotide at position -1/1 does not affect the interaction of the first trp repressor molecule with the primary DNA target site, however, it influences the assembly of additional repressor molecules at adjacent sites. We argue that direct hydrogen bonds may exist whenever there is a guanine base at position -1, because the DNA base at position -1 is equivalently positioned in 2:1 and 3:1 complexes, with respect to Arg69, as G-9 is in the 1:1 complex. These hydrogen bonds may stabilize tandem 2:1 complexes of these targets with the trp repressor.
Our analysis of the interaction of the repressor with its binding sites supports the binding hierarchy model of gene regulation in the trp system (5). We ascribe this ability to two sequence-dependent factors which act together: the identity and number of half-site sequences, recognized by water-mediated hydrogen bonds, and the number of guanine bases, capable of direct hydrogen bonding to the repressor, at position +9 and +1.
In addition we have measured the extent of bending of the trp operator targets in both their free state and in the complex with the repressor protein. In doing so we further address the relevance of the crystal structure of the complex (1), as well as that of the operator sequence alone (6), to solution structures.
References
1. Otwinowski, Z., Schevitz, R. W., Zhang, R. G., Lawson,
C. L., Joachimiak, A., Marmorstein, R. Q., Luisi, B. F., and Sigler, P.
B. (1988) Nature 335, 321-329.
2. Sigler, P. B. (1992) in Transcriptional Regulation (McKnight,
S. L., and Yamamoto, K. R., eds), pp. 475-499, Cold Spring Harbor Laboratory
Press, New York.
3. Bass, S., Sugiono, P., Arvidson, D. N., Gunsalus, R. P., and Youderian,
P. (1987) Genes Dev. 1, 565-572.
4. Bareket-Samish, A., Cohen, I., and Haran, T. E. (1996) J. Mol. Biol.
267, 103-117.
5. Kumamoto, A. A., Miller, W. G., and Gunsalus, R. P. (1987) Genes Dev.
1, 556-564.
6. Shakked, Z., Guzikevich-Guerstein, G., Frolow, F., Rabinovich, D., Joachimiak,
A., and Sigler, P. B. (1994) Nature 368, 469-473.
Computer Modeling of
the Ternary Complex of the Lac Repressor, CAP and DNA Loop
Steve Harvey(1), David Case(2) and Robert K.-Z. Tan(1)
(1)Department of Chemistry,
University of Alabama at Birmingham,
Birmingham, AL 35294-0005
(2)Department of Molecular Biology, TPC15
The Scripps Research Institute,
10550 N. Torrey Pines Rd.,
La Jolla, CA 92037, USA
Computer models of a ternary complex of the lac repressor (lacR), the Catabolite gene activator protein (CAP) and a tract of the lac promoter of Escherichia coli are built using a reduced representation of the DNA. The base pairs of double stranded DNA are each modeled by three atoms placed in a plane. The reduction in detail allows a large number of models to be examined at different threading, linking number and loop length. Implausible models are eliminated and the models that remain are converted to full-atom representation. The model of the lac repressor complexed with CAP and the DNA loop between the O1 and O3 operator sites (M. Lewis, G. Chang, N.C. Horton, M.A. Kercher, H.C. Pace, M.A. Schumacher, R.G. Brennan, P. Lu (1996) Science 271, 1247-1254) suffers from close contacts between the lac repressor and CAP according to our model. However the steric hindrance can be relieved by moving the repressor contact site at O3 upstream. The optimal displacement according to our calculations is the length of about half a turn of the helix.
The Role of Water in Complex Formation between Peptides and the MHC Molecule HLA-A2
Wilson S. Meng, Hermann von Grafenstein and Ian S. Haworth
Department of Pharmaceutical Sciences,
University of Southern California,
Los Angeles, CA 90033, USA
Class I major histocompatibility complex (MHC) molecules are proteins that bind short peptide fragments (derived from intracellularly processed proteins) and present them at the surface of infected cells to cytotoxic T cells. Since an individual can only express six different class I MHC molecules, and these are used to bind a large variety of peptide sequences, flexibility is required in the chemical environment of the MHC peptide-binding groove. However, in addition to binding peptides of diverse sequences with high affinity, the MHC molecule is also able to discriminate between two peptide sequences that may only differ by one amino acid.
It has been suggested that these requirements are facilitated
by concerted side chain motions of the protein, and by the presence of bound
water molecules in the binding groove. Here we extend this latter suggestion
and show that a structured network of eleven water molecules may contribute
to the formation of the complex between the peptide GILGFVFTL and the MHC
molecule HLA-A2. The water molecules are located in the binding groove as
shown in the figure, but, based on a nanosecond molecular dynamics simulation,
are not bound, but exhibit mobility and exchange b
etween
sites within the binding groove. We will show that the water network may
play a role in determining the appropriate conformations and orientations
of key protein side chains, which can then provide a complementary surface
for the binding of GILGFVFTL.
A complete analysis of the water motion in the simulation will be presented, and contrasted with simulations in which fewer than eleven water molecules were included in the binding groove. This comparison suggests that the water network fulfills two roles in contributing to the formation of the peptide-MHC complex. The first of these is a mediation of hydrogen bonding between peptide and protein, and several examples of this can be discerned. However, the water molecules in these locations are not tightly bound and secondary water to water interactions are required to maintain the peptide conformation in the complex. As shown in the figure above, we view the water network as a malleable interface which can adopt a particular 'shape', depending on the sequence of the peptide and the orientation of the HLA-A2 side chains.
6-Amino-6H-4,5,7,8-Tetrahydroimidazo[4,5-e][1,4]Diazepine-5,8-Dione: A Synthetic Analogue of Azepinomycin as a Transition State-Mimicking Inhibitor of Guanese
Vasanthakumar Rajappan and Ramachandra S. Hosmane
Laboratory for Drug Design and Synthesis,
Department of Chemistry and Biochemistry,
University of Maryland Baltimore County,
1000 Hilltop Circle, Baltimore, MD 21250
Guanase (guanine deaminase or guanine aminohydrolase, EC 3.5.4.3) is an important enzyme in the purine salvage pathway, and catalyzes the hydrolysis of guanine to xanthine via the tetrahedral intermediate (I). Since xanthine is an important biosynthetic precursor to other purine-based nucleotides, the inhibition of guanase has beneficial implications in cancer chemotherapy, especially in view of recent reports of significantly increased guanase activity in lung and gastric cancer tissues. Furthermore, abnormally high levels of serum guanase activity have been reported in patients with liver diseases like hepatitis. Such a high activity is also believed to be a biochemical indicator of rejection in liver transplant recipients. There are also reports of cases in which patients developed non-A/non-B hepatitis upon transfusion with blood containing high levels of serum guanase activity, with a linear correlation between the number of incidences of posttransfusional hepatitis and the extent of guanase activity in the donor blood. It has been further shown that patients with multiple sclerosis (MS) exhibit significantly elevated levels of guanase activity in their cerebral spinal fluids (CSF), and that a close correlation existed between the extent of disability and the level of CSF guanase activity. In light of these observations, a potent guanase inhibitor is likely to be useful (a) in exploring biochemical mechanisms of the above metabolic disorders, (b) in understanding the specific physiological role played by guanase, and (c) as a potential chemotherapeutic agent that could selectively inhibit the growth of rapidly proliferating cancerous cells via depletion of the purine nucleotide pool. We report herein the synthesis and biochemical screening of the title compound (III) as an analogue of azepinomycin (II), a naturally occurring moderate guanase inhibitor (IC50 = ~ 10-5 M) that is known to mimic the transition state structure leading to the intermediate (I) during the guanase-catalyzed hydrolysis reaction.
Affinity Reagents For Cross-Linking Hemoglobin: Bis(Phenoxycarbonylethyl)Phosphinic Acid (Bpcep) and Bis(3-Nitrophenoxycarbonylethyl)Phosphinic Acid (Bncep)
S. Prasad Peri(1), Vishweshwar S. Bhadti(1), Rong Liang(1),Ramachandra
S. Hosmane(1) and Victor Macdonald(2)
(1)Laboratory for Drug Design and Synthesis
Department of Chemistry and Biochemistry
University of Maryland Baltimore County
1000 Hilltop Circle, Baltimore, Maryland 21250
(2)U.S. Army Blood Research Detachment
Walter Reed Army Institute of Research
Washington, D. C. 20307
A suitably modified human hemoglobin currently holds the best promise as an acellular oxygen carrier, with compelling advantages including the elimination of need for typing and cross-matching before transfusion, limitations of storage of intact blood, and problems of inadequate supply for emergency transfusions, and not to mention the well tolerance and facile metabolism of hemoglobin by the human body. A crucial modification of hemoglobin concerns interdimer cross-linking that is necessary to overcome the two major problems inherent with cell-free hemoglobin: (a) too short intravascular retention time, and (b) too high oxygen affinity to deliver adequate amounts of oxygen to tissues. We report herein the synthesis and properties of two new hemoglobin cross-linking reagents, bis(phenoxycarbonylethyl)phosphinic acid (BPCEP; I) and bis(3- nitrophenoxycarbonylethyl)phosphinic acid (BNCEP; II). The reagents have been successfully employed to cross-link human hemoglobin under oxygenated conditions. The SDS-PAGE, HPLC, and FPLC analyses of the reaction products indicated that the cross-link was intramolecular in nature, and that it was between the two b subunits of hemoglobin. The products were purified by DEAE- cellulose chromatography, and the purified material was employed for oxygen-binding assessments. The oxygen equilibrium curve of the cross-linked material, in each case, was right-shifted toward lower oxygen affinity as desired. The computed value for the Hill coefficient, in each case, was lower than that of the stroma-free hemoglobin, suggesting somewhat reduced cooperativity of the cross-linked material.
Inhibitors of Glycogen Phosphorylase B: Synthesis, Molecular Modeling, and Biochemical Screening of Analogues of Hydantocodin
Mary W. Mumper, Yankanagouda S. Agasimundin and Ramachandra
S. Hosmane
Laboratory for Drug Design and Synthesis,
Department of Chemistry and Biochemistry,
University of Maryland Baltimore County,
Baltimore, Maryland 21228 USA
Glycogen phosphorylase is a key enzyme in the regulation of muscle and hepatic glycogen metabolism and, in turn, in the control of diabetes mellitus. Glycogen phosphorylase b (GPb) catalyzes the first step in the intracellular degradation of glycogen, and is regulated both by allosteric interactions and reversible phosphorylation. Glycogen is the carbohydrate reserve of most metabolically active cells in mammals. As warranted by cellular demands for energy, (_-1,4-glucoside)n (glycogen) + Pi _ (_-1,4-glucoside)n-1 + _-glucose-1-phosphate glycogen is broken down by GPb to glucose-1-phosphate which, in turn, either undergoes glycolysis to provide energy for muscle contraction or is converted to glucose in the liver to supply food to other organ tissues. The reverse reaction is also equally important when there is comparatively less cellular demand for energy, for example in the resting state or after a meal, and the energy needs to be stored as glycogen. The inhibition of GPb is anticipated to assist in shifting the equilibrium between glycogen degradation and glycogen synthesis in favor of glycogen synthesis in both muscle and liver, and therefore, GPb inhibitors may be clinically useful for the treatment of diabetes mellitus, especially the non-insulin-dependent diabetes mellitus (NIDDM or Type II diabetes). We report here the synthesis, molecular modeling, and biochemical screening studies of four spiro nucleosides, 1-4, that are analogues of the recently discovered powerful and non-toxic herbicide hydantocidin.
NMR Structure Studies of the Prion Protein
He Liu(1), Nicolai B. Ulyanov(1), Shauna Farr-Jones(1),
Hong Zhang(1), David Donne(3), Peter Wright(3), Fred Cohen(1), Darline Groth(2),
Stanley Prusiner(2) and Thomas L. James(1)
(1)Department of Pharmaceutical Chemistry,
(2)Neurology, University of California,
San Francisco, CA 94143-0446
(3)Department of Molecular Biology
The Scripps Research Institute
La Jolla, CA 92037
Four neurologic disorders in humans have been attributed to prions: kuru, Creutzfeldt- Jakob disease (CJD), Gerstmann-Stra|ssler-Scheinker disease (GSS), and fatal familial insomnia (FFI). Interestingly, familial CJD, GSS and FFI are not only inherited illnesses due to point mutations in the prion protein (PrP gene), they are also transmissible to experimental animals. In animals, bovine spongiform encephalophathy (BSE) or "mad cow" disease has caused more than 160,000 cattle deaths in Great Britain and is thought to be caused by prion-contaminated meat and bone meal dietary supplement. Recent reports have raised the possibility that bovine prions from sick cows have been transmitted to humans, manifesting a new variant vCJD.
There is now much evidence that prion diseases arise through the prion protein PrP. This protein is present in all animals in a normal cellular form designated PrPC. When exposed to the abnormal disease (scrapie) form of this same protein, designated PrPSc, normal animals develop the degenerative disease. Several forms of evidence have indicated that PrPC and PrPSc differ only in conformation, not in primary sequence or posttranslational modification. The working hypothesis is that PrPSc catalyzes the conversion of PrPC to PrP Sc. This is supported by the fact that transgenic mice missing the PrP gene do not develop scrapie diseases at all, regardless of exposure to PrPSc. Critical for understanding the molecular basis for these diseases is determination of the conformations of both PrPC and PrPSc and factors influencing them.
Attempts to produce PrPC molecules in large quantities for structural studies have been fraught with difficulties but, recently, expression of a 142-residue polypeptide in E. coli, corresponding to residues 90-231, from the sequence of hamster PrP has been successful. Treatment of PrPsC isolated from infected hamster, with proteinase K yields a protein corresponding to residues 90-231 which is infectious when injected into normal healthy hamsters.
The structure of PrP90-231 has been studied by multi-dimensional NMR with 15N- and 15N/13C-labeling. The secondary structure and the primary folding of the protein are obtained with complete backbone assignment and 95% sidechain assignment. The structure refinement is in progress.
Phosphorothioate and Phosphorodithioate Modifications of Antisense Oligomers: A Structural Comparison by NMR
Todd M. Billeci, Patrick Furrer, Alessandro Donati,
Chojiro Kojima, Wojcieh Stec and Thomas L. James
Department of Pharmaceutical Chemistry
UCSF,
San Francisco, CA 94143-0446
The promise of antisense technology is the surgical inactivation of a specific mRNA in vivo. With such an approach, the rational design of oligonucleotide-based pharmaceuticals, diagnostics, and research tools is permitted. Following initial successes with unmodified oligonucleotides, the introduction of chemically modified sequences has cleared some major technological hurdles at the cellular level. The most notable challenges have included the requirements for effective membrane permeability, target affinity, and stability to nucleases. Derivitization of the phosphate backbone has been the most successful solution to the above problems. At present, over 40 oligonucleotide-based therapeutics are currently under development in pharmaceutical companies, and 15 of these molecules are in clinical trials.
Despite the importance of chemical modifications to the antisense field, little is known about their impact on structure. We are engaged in the study of three hybrid duplexes with the sequence d(CCTATAATCC)* r(GGATTATAGG), differing only in the structure of the phosphate backbone of the DNA strand. One species is fully unmodified, the second contains a chirally pure R-phosphorothioate modification at each deoxynucleotide, while the third contains a phosphorodithioate moiety at each deoxynucleotide. Biophysical and NMR analyses are being performed, and our results may be summarized as follows:
1) UV and NMR melting profiles demonstrate that increasing
sulfur substitution destabilizes the duplex.
2) Sulfur substitution significantly alters H3' chemical shift.
3) Interproton distances obtained quantitatively from NOE intensities via
complete relaxation matrix analysis are compared and structural implications
are discussed.
Induction of Self-ligation of Hairpin Ribozyme RNAs by Freezing and Dehydration
Brian H. Johnston(1,2), Svetlana V. Balatskaya(1,3)
and S. A. Kazakov(1)
(1)Somagenics,
239 Cypress Point Drive,
Mountain View, CA 94043
(2)Department of Pediatrics,
Stanford University School of Medicine,
Stanford, CA 94305
(3)Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry,
Russian Academy of Sciences,
Moscow 123182, Russia
Observations: RNAs containing the hairpin ribozyme (HPR) domain spontaneously interconvert between linear and covalently closed circular forms in dilute aqueous solutions if magnesium or certain other multivalent cations are present. We have discovered that linear forms of HPR derivatives can rapidly convert to the circular form even in the absence of multivalent cations if their aqueous solutions are frozen, evaporated under vacuum, or incubated with > 40% ethanol. In the case of solution evaporation, the presence of poly(ethylene glycol) (PEG) resulted in multimerization of the ribozyme in addition to circularization, indicating reaction in trans .
Possible Mechanisms: The conditions that promote this apparently Mg+2-independent reaction have as common features the reduction in water activity and the stimulation of intermolecular association or aggregation, suggesting that the catalytically active conformation of the ribozyme can be stabilized in partially dehydrated specific RNA aggregates. Also supporting an aggregation-mediated mechanism is the observation that addition of an excess of carrier tRNA to solutions before either evaporation or freezing inhibits self-ligation, presumably by competing for binding sites in the complex. In contrast, addition of unlabeled HPR1 itself to the same high initial concentrations has no inhibitory effect. The self-ligation reactions result in the formation of a normal 3'-5' phosphodiester bond from the 2',3'-cyclic phosphate and 5'-hydroxyl RNA termini, as with the ordinary Mg+2-dependent reaction. However, the two reactions differ significantly in the dependence of their rates and product yields on several biological inorganic and organic solutes. Dependence on pH upon freezing is shown in.
Implications: The 2'-OH group renders RNA highly susceptible to cleavage through transesterification, generating the same termini that are substrates for the self-ligation reaction. Based on closely related RNA reactions, we speculate that the dehydration-induced self-ligation could be catalyzed by a variety of RNA structures. Since drying and (perhaps) freezing conditions are likely to have been available on the primitive earth, the ability of those conditions to stimulate the ligation of certain cleaved RNAs might have helped maintain the integrity and diversity of RNA during chemical evolution. Moreover, the importance of aggregation in the observed reactions may lend support to the concept that life had its origins in quasi-solid-state processes rather than in aqueous solution. Because macromoleculer crowding and low water activity are also features of intracellular conditions, we suggest that in vitro studies of any ribozyme under such conditions could provide insights into their activities in vivo.
Antiparallel-Parallel Combination (APC) in DNA: The Formation is Facilitated by Groove Binding Ligand Distamicin A
Elena B. Khomyakova(1), Anna K. Shchyolkina(1,2), Lucy
E. Minchenkova(1), Elvira E. Minyat(1), Valery I. Ivanov(1) and Tomas M.
Jovin(2)
(1)The Engelhardt Institute of Molecular Biology,
Russian Academy of Science,
117984 Moscow, Russia
(2) Department of Molecular Biology,
Max Plank Institute for Biophysical Chemistry,
P.O.Box 2841, W-37018 Gottingen, FRG
Formation of Antiparallel-Parallel Combination (APC), a conformation with the segment of parallel-stranded helix 5'-AAATT/5'-TTTAA imbedded into usual B-DNA, was studied by a number of designed and synthesised oligonucleotides.
To reveal the parallel moiety a groove binding ligand, distamicin A, was used. It was proposed that due to equivalence of the grooves in parallel DNA unlike the antiparallel one, the stoichiometry and other properties could differ from those in the B-DNA.
The DstA binding phenomenon was characterized by changing of the circular dichroism (CD) spectra upon addition of the ligand. The titration with DstA revealed the differences in the binding stoichiometry and the form of CD spectra of the complexes DNA/ligand for APC and the control antiparallel duplex with the same sequence. These results are in agreement with the parallel duplex formation as a component of Antiparallel-Parallel Combination.
To study the involvement of the nucleotides in the base pairing the chemical modification study was performed. As it follows from the the KMnO4 and DEPC probing the nucleotides which are expected to be involved in the parallel helix formation are protected from modification both in presence of DstA and without ligand.
The obtained results strongly support the existence of APC folding in solution.
Structural Recognition of DNA by HMG Proteins
Hong Xin, Wenjie Cheng and Neville R. Kallenbach
Department of Chemistry,
New York University,
31 Washington Place,
New York, NY 10003 USA
The HMG1/2 proteins are a widespread family of chromosomal proteins that are active in DNA binding and looping, transcriptional control and DNA repair. Members of this class include sequence specific DNA binding proteins as well as structure responsive proteins. They share a conserved sequence of about 80 amino acids rich in basic, aromatic and proline side chains, the HMG box. We report here studies on the preferences for three and four way DNA junction substrates of four HMG proteins. The four proteins are: a structurally responsive single HMG box from rat (HMGb), two adjacent box sequences (HMGab), the intact rHMG1 containing a strongly acidic tail, as well as two HMG boxes cloned from the Xenopus transcriptional enhancer xUBF (xUBF'). We show here that all these HMG's bind four arm DNA more tightly than three arm structures. Intact rHMG1 protein binds about 10 fold more weakly than either single box or two-box protein, while the two box HMGab protein has a much higher affinity for junction molecules than the two boxes in xUBF'. The substrate selectivity is altered by the presence of mismatches and bulged bases at the branch. Mispairing at the branch increases the affinity of HMGb, HMGab and xUBF', but weakens binding of HMG1. None of these proteins binds tightly to DNA "necks", three arm DNA junction structures that contain a truncated third arm, and behave as bent structures in electrophoretic assays.
This research was supported by grants from the US NIH (CA 24101) and the International HFSP (RG 331/93).
The Effects of a Mutation in the B Binding Domain of rHMG-1
Susann Taudte, Amy Tam, Wenjie Cheng, Wendy P. Patton
and Neville R. Kallenbach
Department of Chemistry,
New York University,
31 Washington Place,
New York, NY 10003 USA
HMG-1, a member of the high mobility group (HMG) protein family, is composed of tandem HMG boxes (domains A and B) and an acidic C-terminal tail (Weir et al., 1993). HMG-1 is thought to be active in DNA repair, recombination, and assembly of DNA into chromatin fibers (Grosschedl et al., 1994). We have characterized the binding of the B-domain of HMG-1 to four-way branched DNA and to control duplexes using the wild type HMG-1B and a mutant protein, F19A. While the mutant binds slightly more weakly to four-way junction DNA than the wild type HMG-1B, it retains its selectivity for four-way junction DNA over duplex DNA. The stability of this mutant has been investigated: a folding study has been performed using circular dichroism (CD). The results suggest that a significant fraction of the normal HMG-1B secondary structure is lost as a consequence of the single core substitution.
This research was supported by grants from the US NIH (CA 24101) and the International HFSP (RG 331/93).
Thermal Unfolding of Cytochrome
C. Leyu Wang and Neville R. Kallenbach
Department of Chemistry,
New York University,
New York, NY10003
Recent hydrogen exchange experiments on native cytochrome c implicate a sequential and progressive unfolding pathway in contrast to a simple two state process. We have studied the heat induced unfolding of this protein using a combination of spectroscopic measurements to determine overall conformation and proteolytic enzyme digestion to identify regions of the protein that are exposed upon heating. Several spectroscopic profiles were monitored: CD at 222 nm, a measurement of secondary structure change in the protein, the absorbance at 280 nm, the region Trp 59 absorption, and absorbance at 420 nm, the Soret band of the heme. We detect a significant difference between apparent Tm values for these probes, consistent with unfolding via a pathway that contains intermediates. The limited digestion with proteinase K is consistent with population of an intermediate state in unfolding. We find a single strong region of cleavage at low temperature with retention of structure in each fragment.
This research was supported by grant GM40746 from NIH, LW is a recipient of Henry Mitchell MacCracken Fellowship at NYU.
Quantitative Footprinting Studies On HIV-1 RNA
Christopher Kirk, Jerry Goodisman, Lucia Pappalardo,
Philip Borer and James Dabrowiak
Department of Chemistry,
CST 1-014, Syracuse University,
Syracuse, NY 13244-4100.
The HIV-1 genome is a 9.6 kb RNA molecule which is enzymatically copied into double stranded DNA and inserted into the genomic DNA of the host cell. One of the hairpins in the packaging signal region of the HIV-1 genome, having the sequence, 5'-GGACUAGCGGAGGCUAGUCC, designated H3, is implicated in binding to a 7 kD nucleocapsid protein, NCp7. The NCp7 protein, a two zinc finger protein, is a critical component in the viral capsid, which is integral to viral packaging and infectivity. We are performing quantitative footprinting studies on the H3-NCp7 complex to determine the binding constant for the protein to the target RNA. The RNA cleavage agent utilized for the quantitative assays is ribonuclease T1, an endoribonuclease which cleaves on the 3' side of single stranded guanosines. Sequence assignment of the H3 RNA was performed using RNases T1 and A and nuclease S1. The primary sites for cleavage were G9 and G10 in the loop of the hairpin, with G2 in the stem of the hairpin, being cut to a lesser degree. Kinetic studies of the T1/H3 cleavage reaction in the absence of NCp7 have been completed to ascertain the proper concentration of T1 to use in the footprinting studies. Considering the importance of NCp7 to the survival of the HIV-1 retrovirus, any new information about the structure of the H3/NCp7 complex will be instrumental for future development of drugs for combating AIDS.
Diastereoselectivity Of Nucleolytic Enzymes Towards Phosphorothioate Analogues Of Oligonucleotides
Maria Koziolkiewicz, Marzena Wojcik, Edyta Gendaszewska,
Maria Maszewska and Wojciech J. Stec
Department of Bioorganic Chemistry,
Centre of Molecular and Macromolecular Studies,
Polish Academy of Sciences,
Sienkiewicza 112, 90-363 Lodz, Poland
In the course of our studies on the stereocontrolled method of synthesis of oligo(nucleoside phosphorothioate)s (PS-ODN), valuable tools in molecular biology and promising class of antiviral, anticancer and anti-restenotic compounds [1], it was necessary to establish a method for assessment of their diastereomeric purity (d.p.). From seminal work of Eckstein [2] it was known that Nuclease P1, possessing exo- and, what is in this case more important, endonucleolytic activity, can be used for d.p. determination of [all-RP]-PS-ODN, because this enzyme cleaves each internucleotide phosphorothioate bond of [SP]-absolute configuration. Because all so far studied endonucleases appeared to be [SP]-specific, the assignment of diastereomeric purity of [all-SP]-[PS-ODN] was not a trivial task. All known [RP]-specific enzymes are exonucleases and cannot hydrolyze internucleotide phosphorothioate linkages of [RP]-configuration located inside oligonucleotide of chain. We have found that endonuclease from Serratia marcescens (benzonase) is [RP]-specific. This enzyme does not accept as substrates short oligonucleotides (2-6 mers). Therefore, for the studies of its diastereoselectivity we have used stereoregular PS-ODNs (longer than heptamers) synthesized via the oxathiaphospholane method [3]. It has been found that digestion of PS-ODNs prepared by non-stereocontrolled methods ([Mix]-PS-ODNs) requires 200-400 times higher concentration of the enzyme than the hydrolysis of isosequential
PO-ODNs. Under similar conditions the [all-SP]-isomers are resistant to this enzyme, while the [all-RP]-isomers are completely hydrolyzed by Serratia endonuclease even at the enzyme concentration only 10 times higher than that necessary for hydrolysis of the PO-oligonucleotides. Therefore, Serratia marcescens endonuclease has been used, together with snake venom phosphodiesterase and Nuclease P1, for assignment of diastereomeric purity of P-stereoregular PS-ODNs [4].
Independent studies on the stability of stereoregular PS-ODNs in the cellular media led to the conclusion that human plasma and certain types of normal and tumor human cells do contain 3'-exonucleases which are also [RP]-specific [5].
Observations collected so far allow to suggest that the [all-SP]-PS-ODNs, although being weaker activators of RNase H and possessing poorer avidity towards complementary mRNA than their [all-RP]-PS-counterparts [6], may be superior antisense effectors due to their higher stability in the body fluids.
Acknowledgment
Results presented in this communication were financially assisted by the State Committee for Scientific Research grants no. 6 P203 007 05 and 4 PO5F 023 10
References
1. Antisense Therapeutics, 1996, Ed. Agrawal, S.,
Humana Press, Totowa, NJ.
2. Potter, B.V.L., Connolly, B.A., Eckstein, F. (1983), Biochemistry,
22, 1369-1377.
3. Stec, W.J., Grajkowski, A., Kobylanska, A., Karwowski, B., Koziolkiewicz,
M., Misiura, K., Okruszek, A., Wilk, A., Guga, P., and Boczkowska, M. (1995),
J.Am.Chem.Soc., 117, 12019-12029.
4. Koziolkiewicz, M.- manuscript in preparation
5. Koziolkiewicz, M., Wojcik, M., Kobylanska, A., Karwowski, B., Rebowska,
B., Guga, P., Stec, W.J., (1997), Antisense Res. Dev., - in press.
6. Koziolkiewicz, M., Krakowiak, A., Kwinkowski, M., Boczkowska, M., and
Stec, W.J. (1995), Nucleic Acids Res., 23, 5000-5005.
3D Domain Swapping in an RNA Pseudoknot
Craig E. Kundrot
Department of Chemistry and Biochemistry,
University of Colorado,
Boulder, CO 80309-0215
Phone: 303/492-0855; Fax: 303/492-5894; E-mail: kundrot@colorado.edu
RNA pseudoknots are required structures in several biological processes. Pseudoknots also arise frequently in in vitro selection experiments directed against protein targets. We have determined the 2.9 A resolution crystal structure of an RNA pseudoknot, PK26, that specifically inhibits HIV reverse transcriptase. The structure was determined using a combination of multiple isomorphous replacement and molecular replacement methods; the model contains all 26 nucleotides of the pseudoknot and has R = 20 % and Rfree = 24 %. The molecule forms a dimer with extensive base pairing between and within strands. The two residues comprising loop 1 adopt a "switchback" conformation that traverses the major groove of stem 2 and reverses the direction of the backbone. The residues in loop 1 participate in a junction region between stems 1 and 2; this junction maintains stacking interactions and contains one base triple and one non-canonical base pair. The formation of the dimer requires a three nucleotide "skip and detour" motif that allows bases within stem 1 to skip one base in the stacking register. Loop 2 forms base triples in the minor groove of stem 1.
Chemical modification data suggest how the dimer might bind to and inhibit HIV reverse transcriptase. Native gel electrophoresis and UV melting curves suggest that the dimer formation is facilitated by high salt conditions.
Despite the extensive intermolecular contacts in the PK26
dimer, this structure shows that a monomeric form of PK26 could be formed
from the dimeric structure by changing the conformation of the three residues
comprising the "skip and detour" region. Thus, the dimeric structure
of PK26 is formally identical to 3D domain swapping observed in proteins.
The structure shows how RNA can build large scale oligomers or conformational
switches through 3D domain swapping.
The High Resolution Crystal Structure of a Guanine Tetraplex
Kathryn Phillips(1), Zbyzek Dauter(2), Alaistair Murchie(3),
David Lilley(3) and Ben Luisi(1)
(1)Department of Biochemistry,
Cambridge University,
Tennis Court Road,
Cambridge, CB2 1QW, U.K.
(2)Department of Chemistry,
York University,
Heslington, York YO1 5DD, U.K.
(3)Department of Biochemistry,
Dundee University,
Dundee, DD1 4HN, U.K.
We report the crystal structure, at 0.95 angstrom resolution, of a four-stranded DNA helix formed by the association of hydrogen-bonding guanine tetrads. The unit of the tetraplex is the simple hexanucleotide d(TGGGGT). All of the DNA strands are parallel and the helix is right handed. In addition to the hydrogen-bonding guanine tetrads, the structure is stabilized by coordination of sodium ions which lie between successive tetrad planes. Four complete tetraplexes occupy the asymmetric unit. The thymine bases do not form part of the tetraplex unit, but instead are mostly stacked to link neighboring tetraplexes in a compact lattice. The high resolution of the structure permits a detailed analysis of the hydration patterns in and around the helical grooves and of the interaction of ions with the phosphate groups. Well defined networks of water molecules are observed in the grooves, and distributions of water molecules around the guanine and thymine bases can be defined. The sugars have a well defined puckering geometry. Two distinctive types of base stacking arrangements are observed. Four-stranded DNA structures have been implicated (but not yet proven) to play a number of biological roles, and these will be discussed.
High-Resolution Solution NMR Structure of the Anticancer Drug Cisplatin Bound to its Major Site in Duplex DNA
Andrew Gelasco and Stephen J. Lippard
Department of Chemistry,
Massachusetts Institute of Technology,
Cambridge, MA 02139
The solution structure of the duplex dodecamer d(CCTCTG*G*TCTCC)-(GGAGACCAGAGG), where G*G* represents the d(GpG) 1,2-intrastrand cross-link, was determined by 2D NMR spectroscopy at 750 MHz. The x-ray crystal structure of this same platinated DNA has recently been solved at a resolution of 2.6 Å.1,2 Interesting features in the crystal include a 26° roll of the platinated guanosines toward the major groove, a widening and flattening of the minor groove, an A form-B form junction, and a nearly 1 Å displacement of the platinum atom from the planes of the two adjacent guanosine residues. The solution structure was determined in order to compare the geometric distortions associated with the cis-[Pt(NH3)2{d(GpG)-N7(G1),-N7(G2)}] adduct on duplex DNA with those obtained in the solid state on the same sequence and with previous studies of the same adduct in duplex DNAs of different sequence.3,4 Recent work reveals a profound dependence of HMG-domain protein recognition of cisplatin-DNA adducts on local sequence.5
The use of the high field spectrometer has allowed us to measure and assign unambiguously a large number (510) of 1H-1H nOe's in the duplex dodecamer. A series of 2D NOESY spectra were collected with mixing times ranging from 70 - 300 ms in D2O at 15°C and used to assign non-exchangeable protons. Two 2D NOESY spectra in H2O at 110 and 200 ms, also recorded at 15°C, were used to assign exchangeable protons, including those of the NH3 groups on platinum, and to determine hydrogen bonding interactions.
The platinated guanosine residues have an approximately 30° roll, and the duplex bends by ~38° toward the major groove. These values are consistent with those obtained from gel electrophoresis studies,6,7 but smaller than the bend angle of 58° reported in a NMR structure of an octamer d(CCTG*G*TCC-GGACCAGG) containing the Pt-G*G* lesion.3 The duplex is unwound by ~20° near the site of platination. This feature and the bend toward the major groove cause the minor groove opposite the Pt lesion to be widened and flattened. The deoxyribose rings for G6* and G7* have C3'-endo and C2'-endo conformations, respectively, but for the entire duplex the sugar puckers have generally a C2'-endo conformation. This result is distinctly different from that observed in the crystal structure, where the long stretch of A-DNA and the A-B junction are most likely due to crystal packing forces.2
The bend of the duplex at the platinum site together with the widening of the minor groove are similar to the structural deformations observed in DNA complexes with the HMG-domain proteins SRY and LEF-1.8,9 These modulations of the DNA structure suggest why the binding of cisplatin to DNA leads to recognition by HMG-domain proteins.
References
1. Takahara, P. M.; Rosenzweig, A. C.; Frederick, C. A.;
Lippard, S. J. Nature 1995, 377, 649-652.
2. Takahara, P. M.; Frederick, C. A.; Lippard, S. J. J. Am. Chem. Soc
1996, 117, 12309-12321.
3. Yang, D.; van Boom, S. S. G. E.; Reedijk, J.; van Boom, J. H.; Wang,
A. H.-J. Biochemistry 1995, 34, 12912.
4. van Boom, S. S. G. E.; Yang, D.; Reedijk, J.; van der Marel, G. A.; Wang,
A. H.-J. J. Biomol. Struct. Dyn. 1996, 13, 989-998.
5. Dunham, S. U.; Morse, J. M.; Lippard, S. J. submitted for publication
1997.
6. Bellon, S. F.; Lippard, S. J. J. Biophys. Chem. 1990, 35,
179.
7. Bellon, S. F.; Coleman, J. H.; Lippard, S. J. Biochemistry 1991,
30, 8026.
8. Werner, M. H.; Huth, J. R.; Gronenborn, A. M.; Clore, G. M. Cell
1995, 81, 705.
9. Love, J. J.; Li, X.; Case, D. A.; Giese, K.; Grosschedl, R.; Wright,
P. E. Nature 1995, 365, 791-795.
Analysis of the Interactions of the DNA binding domain of PU.1 by Hydroxyl Radical Footprinting
Petra Groß(1), Cheryl H. Arrowsmith(2) and Robert
B. Macgregor, Jr.(1)
(1)Department of Pharmaceutical Sciences,
(2)Department of Medical Biophysics,
University of Toronto and Ontario Cancer Institute,
Toronto, Ontario, Canada
The PU.1 transcription factor is a member of the ets gene family. It plays an important role in normal development by regulating the differentiation of mast cells and macrophages and in the maturation of B cells (1). The ets proteins share a conserved 85-amino acid DNA binding domain, the so-called ets domain, that recognizes the sequence 5'-GGAA-3' flanked by more variable but not random 5' and 3' sequences. These proteins bind the DNA as monomers. PU.1 is the least conserved member of the ets family and shares a 35% homology with ETS1, the first ets protein that was characterized (1,2).
Hydroxyl radical footprinting has been used to probe interactions in a complex between the ets domain of the murine transcription factor PU.1 and a DNA restriction fragment containing one copy of the recognition sequence 5'-GGAA-3'. The extent of contact with the protein was monitored for every base in the complex. A concentration dependent cleavage site on the complementary TTCC strand is evidence of a specific interaction between PU.1 and the DNA. The association constant for the PU.1 interaction with the binding site, Ka = 9.2x106 M-1, was calculated from the concentration dependence of the protection. Two more protection sites and a hypersensitive cleavage site on the GGAA strand were observed as shown in Figure 1. This is in agreement with recently reported crystal structure of the PU.1-DNA complex (3). The footprinting data reveal small differences in the flanking region of the PU.1-DNA complex in solution and in the crystal state; however, these data confirm the global structure of the complex as suggested by crystallography.
FIGURE 1: a. The sequence of the PU.1 binding region of the restriction fragment used in these experiments. The consensus sequence is shown in bold letters, the bases protected from hydroxyl radical cleavage are indicated with solid arrowheads. Weak protection sites are marked with open symbols. DNase I protection sites are marked with arrows. b. Contacts between the ets domain of PU.1 and the DNA backbone as observed in the crystal structure (3).
References
1. Wasylyk, B., et al. (1993) Eur. J. Biochem. 211,
7-18.
2. Nye, J. A., et al. (1992) Genes Dev. 6, 975-990.
3. Kodandapani, R, et al. (1996) Nature 380, 456-460.
Infrared Spectroscopy of (dA)·(dT) Polymers at High Pressure
M.-C. Lin(1), P. Eid(2), P. T. T. Wong(2) and R. B.
Macgregor, Jr.(1)
(1)Department of Pharmaceutical Sciences,
University of Toronto,
Toronto, Ontario, Canada
(2)Department of Biochemistry,
University of Ottawa,
Ottawa, Ontario, Canada
Several studies have probed the influence of hydrostatic pressure on the stability and rate of formation of DNA helices. The effect of hydrostatic pressure on the DNA duplex, poly(dA)·poly(dT), and its component single strands, poly(dA) and poly(dT) in aqueous solution containing 20 mM Tris HCI, pH 8.8, 20 mM NaCl and 0.1 mM EDTA have been studied by infrared spectroscopy. Spectra were obtained at pressures up to 12 kbar (1 kbar = 100 Mpa = 978 atm) throughout the infrared absorption spectrum. However, particular attention was focused on the IR bands that are sensitive to the conformational state of the polymers.
Deoxyribose absorption bands were observed at 840 cm-1 in all three polymers throughout the pressure range. The position of this band is consistent with the C2' endo/anti geometry of sugar in helices with the B conformation. The deoxyribose intra-ring vibration band at 967 cm-1, seen in the spectra of the three polymers, is also indicative of the B-conformation. This band shifts to higher frequencies with pressure, indicating increasing hydration of the moiety. In addition, the band at 1188 cm-1 attributable to motion of sugars in A-conformation helixes was not observed in any spectrum.
The antisymmetric phosphate stretching vibration mode was observed at 1230 bcm-1, and its position was unaffected by pressure. This is consistent with the B conformation being maintained at elevated pressure.
Pressure-induced changes in the conformation of the bases were assessed by monitoring bands at 730 cm-1 and 1301 cm-1 for adenine. A hypochromicity was observed in the 730 cm-1 band of poly(dA) and in the 1301 cm-1 band of poly(dA)·Poly(dT). Hypochromic effects imply that base stacking increases with pressure.
The conformation-sensitive mode that arises from the N1-C6-H and C5-C6-H deformations of thymine are found at 1281 cm-1 at ambient pressure and at high pressure. This band appears at 1273 cm-1 for DNA in the A conformation.
Thus, all of the data indicate that up to pressures of 12 kbar poly(dA), poly)dT) and poly(dA)·poly(dT) remain in the B conformation. It appears that the extent of base stacking increases with pressure and that the hydration of the deoxyribose sugars is also enhanced. They provide useful data for interpreting previous results on the influence of pressure on the physical properties of DNA. For example, these findings are consistent with the previous proposal that a pressure-induced increased base stacking is one of the factors involved in the increased thermal stability of double stranded DNA at high pressure.
This work was supported, in part, by a grant from the National Science and Engineering Research Council of Canada.
Working Towards an Optimized Library of Routines for the Analysis and Reconstruction of Nucleic Acid Structure
Xiangjun Lu and Wilma K. Olson
Department of Chemistry,
Rutgers, The State University of New jersey,
Piscataway, NJ 08855-0939
Nearly ten years have passed since the "Cambridge Conventions" for definitions and nomenclature of nucleic acid structure parameters were formulated [1]. However, the stated goal of creating a new and optimized library of routines for the analysis and reconstruction of nucleic acid structure has yet to be fulfilled. On the contrary, many programs have appeared, claiming to be fully consistent with the conventions [2]. These various approaches differ subtly in ways which have caused confusion in this field.
The arguments with regard to local vs. global parameters, a single axis rotation vs. three Euler rotations etc, come mainly from different interpretations of the "Cambridge Conventions". As far as mathematics is concerned, the essential differences between the various approaches lie in three parts: (1) the definition of the reference frame used to fix a base or a base-pair in three-dimensional space; (2) the definition of a "middle-triad" to ensure the same numerical values of parameters are achieved independent of the direction in which a base-pair step is counted; (3) the algorithm used to calculate the three rotational and three displacement parameters.
Due to the complexities of nucleic acid structures, it is very difficult, if not impossible, to decide which approach is better by simply applying different programs against a few test cases. Key to this issue is a thorough understanding from both a mathematical and programmatic viewpoint of these analyses. We are testing rigorously several programs and our findings show that there are great overlaps of functionalities between them. Our goal is thus to develop a package with these redundant codes deleted and each distinct feature preserved.
References
1. Dickerson, R. E., et al, J. Biomol. Struct.
& Dynam. 6, 627--634, 1989.
2. Olson, W. K., Curr. Opin. Struct. Biol. 6, 242--256, 1996.
Influence of Stretching Deformations on DNA Structure
Konstantin Kosikov(1), Andrey Gorin(2), Victor Zhurkin(3)
and Wilma Olson(1)
(1)Department of Chemistry,
Rutgers University,
Piscataway, NJ 08855
(2)Memorial Sloan-Kettering Cancer Center,
1275 York Avenue, NY, NY 10021
(3)Laboratory of Mathematical Biology,
National Cancer Institute,
National, Institutes of Health,
Bethesda, MD 20892
The enzymatically "activated" unwound and stretched forms of DNA involved in transcription and recombination are high energy states in the absence of proteins.
Here the distortions of double helix are thought to be much more severe than those observed in regulatory protein-DNA crystal complexes; the Watson-Crick base pairs are expected to be either completely or partially disrupted during biochemical events.
As the first step towards visualizing these high energy conformations, we have undertaken systematic all-atom potential energy studies of stretching and compression of the A- and B-DNA double helices. The simple model that we used not only reproduces the compressed double helical form with highly inclined base pairs observed in protein-DNA complexes but also suggests the existence of different DNA families that become energetically favorable in highly stretched duplexes including an "activated", highly stretched unwound form.
Gaussian-Chain Model For Interphase Chromosomes
Joseph Y. Ostashevsky
Radiation Biology Division,
Department of Radiation Oncology,
State University of New York,
Health Science Center at Brooklyn,
450 Clarkson Avenue, Box 1212, Brooklyn, NY 11203
Recently Trask, Sachs and coworkers (Sachs et al. (1995), Yokomoto et al. (1995)) used the FISH technique to study the dependence of the average square of geometrical distance (<hx2>, Êm2) between two probes separated by a known genomic distance (Mx, Mbp) for whole chromosomes # 4, 5 and 19 in G1 phase human fibroblasts. They showed that G1 phase chromosomes behave as Gaussian chains which obey random-walk statistics : <hx2> = BMx, where the coefficent B = 0.05 - 0.08 Êm2/Mbp.
In this report, a model of the interphase nucleus was developed by approximating interphase chromosomes as Gaussian coils, each of which occupies a spherical volume of radius equal to the chromosomal radius of gyration. These spheres tightly fill all available space in the nucleus. The following results support this model:
1) Estimates of B for seven different types of cells (Drosophila,
maize, human fibroblasts and lymphocytes, Chinese hamster V79 and CHO, and
mouse C3H10T1/2 cells) were calculated to be of the same order of magnitude
(10-2 - 10-1 Êm2/Mbp) as are measured B-values.
2) Using the Mx-3/2 dependence of loop formation, as for ideal polymers,
we calculated the ratios (F) of centromere-containing interchromosomal to
intrachromosomal exchanges. For the case of equal-size equal-arms chromosomes,
F = 13.4 (experimental = 15) and 6.7 (experimental = 6), respectively, for
sparsely and densely ionizing radiations.
3) The Mx-3/2 dependence explains the experimental observation that small
intrachromosomal loops occur much more frequently than large ones.
4) The model predicts that probabilities for intrachromosomal and interchromosomal
exchanges are inversely proportional to nuclear volume. This agrees well
with data for chromosomal exchanges in human lymphocytes vs. fibroblasts,
and with data for chromosomal exchanges after a temporary shrinkage of the
nucleus caused by postirradiation treatment of cells with hypertonic salt
solutions.
5) The probability of interchromosomal exchanges is greater in spherical
nuclei (3D) than in thin flat nuclei (2D) because chromosomes in the former
have twice the number of adjacent neighbors. This agrees with the measured
difference in chromosome aberrations between monolayer-irradiated human
fibroblasts plated immediately (3D nucleus) vs. hours after irradiation
(2D nucleus).
DNA-Modified Electrodes: New Tools in Molecular Recognition, Detection of DNA Damage and Analysis of Environmental Pollutants
E. Palecek, M.Fojta, M. Tomschik, F. Jelen, V. Vetterl,
L. Havran and V. Stankova
Institute of Biophysics of the Academy of Sciences of the Czech Republic,
612 65 Brno, Czech Republic
Nucleic acids are electroactive species producing oxidation signals at carbon electrodes as well as reduction, oxidation and tensammetric signals at mercury electrodes (reviewed in 1,2). Signals obtained with the mercury electrodes sensitively reflect small changes in the structure the DNA double helix. Nucleic acid-modified electrodes can easily be prepared by immersing the electrode into a 5 ml drop of the DNA and/or RNA solution (2,3). Such nucleic acid-modified electrodes offer a number of advantages useful in biochemical and molecular-biological research (2).
The oxidation voltammetric signals obtained with carbon electrodes were little utilized in nucleic acid and protein research because high concentrations of the biomacromolecules were necessary for their analysis. Quite recently we have shown that with the constant current chronopotentiometric stripping analysis (CPSA) remarkably low levels of DNA, RNA (4,5), peptides and proteins (6,7) can be determined. For example the detection limit for tRNA is about 4x10-16 mol, if the tRNA is immobilized at the carbon paste electrode from a 5 ml drop of the RNA solution and the oxidation stripping peak of guanine residues is measured (4). Measuring the CPSA guanine anodic peak at the mercury electrode gives only a slightly higher detection limit. CPSA of proteins is based on the oxidation of tyrosine and tryptophan residues in the protein molecule. Bioactive peptides and proteins containing these amino acids can be determined at submicrogram concentrations (6,7).
Nucleic acid- and protein-modified electrodes can be utilized in molecular recognition studies including DNA hybridization and DNA-protein interactions. We have shown that surface-attached peptide nucleic acids (PNA) retain their unique and efficient hybridization properties, reported in solution studies (8,9). Adsorption behavior of the PNA with electrically neutral backbone differs greatly from that of the DNA (with negatively charged backbone) while the DNA-PNA hybrid shows intermediate behavior (10). At higher surface coverage PNA molecules associate at the mercury surface. Prolonged exposure of PNA to highly negative potentials does not result in PNA desorption while almost all DNA is removed from the surface at these potentials. The new capabilities and opportunities afforded by the use of PNA surface probes can be utilized in various electrochemical and other hybridization sensors .
Recently we have designed a sensor for the detection of DNA damage capable to detect a single interruption of the DNA sugar-phosphate backbone in small fractions of DNA molecules (11) and a sensor for the specific detection of damage of guanine residues in DNA (12). These sensors can be utilized for the detection of environmental pollutants (13). Application of the nucleic acid-modified and protein-modified mercury and carbon electrodes with CPSA open the door for a fast development of electrochemical tools in nucleic acid and protein research including studies of DNA-protein interactions.
References
1. E. Palecek: Modern polarographic (voltammetric) methods
in biochemistry and molecular biology. Part II. Analysis of macromolecules.
in G. Milazzo (Ed.), Topics in Bioelectrochem. Bioenerg. Vol. 5,
London, 1983, p. 65.
2. E. Palecek: From polarography of DNA to microanalysis with nucleic acid-modified
electrodes. Electroanalysis 8 (1996) 7.
3. E. Palecek and M. Fojta: Differential pulse voltammetric determination
of RNA at picomole level in the presence of DNA and nucleic acid components.
Anal. Chem. 66 (1994) 1566.
4. J. Wang, X. Cai, J. Wang, C. Jonsson and E. Palecek: Trace measurements
of RNA by potentiometric stripping analysis at carbon paste electrodes.
Anal. Chem. 67 (1995) 4065.
5. X. Cai, G. Rivas, P.A.M. Farias, H. Shiraishi, J. Wang, M. Fojta and
E. Palecek: Trace Measurements of Plasmid DNAs by Adsorptive Stripping Potentiometry
at Carbon Electrodes. Bioelectrochem. Bioenerg. 40 (1996) 41.
6. X. Cai, G. Rivas, P.A.M. Farias, H. Shiraishi, J. Wang and E. Palecek:
Trace measurements of insulin by potentiometric stripping analysis at carbon
paste electrodes. Electroanalysis, 8 (1996) 902.
7. X. Cai, G.Rivas, P.A.M. Farias, H. Shiraishi, J. Wang and E. Palecek:
Potentiometric Stripping Analysis of Bioactive Peptides at Carbon Electrodes
Down to Subnanomolar Concentrations Anal. Chim Acta 332 (1996) 49.
8. J. Wang, E. Palecek, P. Nielssen, G. Rivas, X. Cai, H. Shiraishi, H.
Dontha,
D. Luo and P.A.M. Farias: Peptide nucleic acid probes for sequence specific
DNA biosensors. J. Am. Chem. Soc. 118 (1996) 7667.
9. J. Wang, G. Rivas, X. Cai, M. Chicharro, N. Dontha and D. Luo, E. Palecek,
P. E. Nielsen: Adsorption and detection of peptide nucleic acids at carbon
paste electrodes. Electroanalysis , 9 (1997) 120.
10. M. Fojta, V. Vetterl, M. Tomschik, F. Jelen, P. Nielsen, J. Wang and
E. Palecek. Adsorption of peptide nucleic acid and DNA decamers at electrically
charged surfaces.
Biophys. J., in press
11. M. Fojta and E. Palecek: Supercoiled DNA-modified mercury electrode:
a highly sensitive tool for the detection of DNA damage Anal. Chim Acta,
in press
12. F Jelen, M Tomshcik and E Palecek: Adsorptive stripping square-wave
voltammetry of DNA. J. Electroanal. Chem., in press
13. M. Fojta, V. Stankova and E. Palecek : A supercoiled DNA-modified mercury
electrode based biosensor for the detection of DNA strand cleaving agents.
Anal. Chem., submitted
Modeling of DNA-polyamine Interactions by Cobalt Amine Complexes
Andrew Parkinson and Alison Rodger
Chemistry Department,
University of Warwick,
Coventry, England CV4 7AL
Email: msrkn@csv.warwick.ac.uk
DNA's structure in mammalian cells is in part regulated by natural polyamines including spermine and spermidine, possessing physiological charges of 4+ and 3+ respectively. In this study these natural polyamines are compared in their interactions with DNA to synthetic cobalt (111) amine transition metal complexes. The transition metal complexes used to model the polyamines include; [Co(NH3)6]3+, (+) - [Co(en)3]3+ , (-) - [Co(en)3]3+, trans - [Co(NH3)4 (H2O)2] and [Co(NH5)Cl]. These compounds are chosen primarily because of their similar functional groups to the natural polyamines, chemical inertness and well-defined geometry. Interactions of the amines have been studied with polynucleotides of varying guanosine-cytosine (GC) content including the synthetic poly[d(A-T)]2 and poly[d(G-C)]2 and the natural DNA's micrococcus lysodeikticus (ml-DNA, 72% GC), calf thymus (ct-DNA, 42% GC) and clostridium perfringens (cl-DNA, 26%GC).
The amine-DNA systems are studied by the spectroscopic techniques circular dichroism, linear dichroism and normal absorption. Thermal denaturation of the DNA has also been applied were possible.
The results have been interpreted non quantitatively in terms of the ligands ability to bend and condense the DNA. Quantitative analysis is performed on the circular dichroism data were applicable to obtain binding site size and equilibrium binding constants.
It is interesting to note which ligands are effective at inducing the B-Z transition in poly[d(G-C)]2 , this work will be the main focus of the poster and will hopefully give a greater understanding in the still poorly understood mechanism of the B-Z transition. The range of cobalt complexes give us some insight into the geometric arrangement of amine groups that are necessary to induce the B-Z transition.
Genome Sequence Analysis: A Pointer Towards Triplex-mediated cis Regulation of Gene Expression
Sowmya Raghavan, Pradeep K. Burma* and Samir K. Brahmachari
Molecular Biophysics Unit,
Indian Institute of Science,
Bangalore 560012. India.
*Present Address: Department of Genetics,
University of Delhi,
South Campus, New Delhi-110021. India
Several in vitro and in vivo studies show that polypurine/polypyrimidine motifs have the potential to adopt both inter- and intramolecular triple-helical structures. The presence of such motifs upstream of several genes have been found to be critical for the regulation of transcription and deletion of such sequences affects the transcription of these genes. Recently, we have shown that a designed polypurine/polypyrimidine sequence when present upstream of a gene can act in cis and modulate its expression in a chromosomal context. If this is a general regulatory mechanism, then the distribution of these sequences in genomes must follow a pattern. The availability of complete genomic sequences of S.cerevisiae (12.5Mb), M.genitalium (580kb) , M.jannaschii (1665kb), H.influenza (1830kb) and M.pneumoniae (826kb) provided an opportunity to test this hypothesis in a genomic context. We analysed the entire sequences of all these genomes for the presence of polypurine/polypyrimidine tracts of length 15 or more. The analysis reveals that S.cerevisiae shows a fifteen-fold overabundance of the polypurine/polypyrimidine motifs as does M.jannaschii (an archaebacterium). However, the eubacteria- M.genitalium, M.pneumoniae and H.influenza do not show such an overrepresentation. Due to the high density of coding regions in these bacterial genomes the motifs are present predominantly in coding regions while in S.cerevisiae they occur both in coding and non-coding regions. Moreover, we observe that when these motifs occur in non-coding regions, they are found predominantly at a unit nucleosomal distance from the ends of ORFs. Since this feature is unique to the yeast genome we postulate that these motifs may be involved in influencing nucleosome positioning by modulating histone affinity. Interestingly, in the yeast genome, when these motifs occur in coding regions we observe that polypurines preferentially occur in coding strands. This biased occurrence of polypurines persists even at distances ~50bp upstream of ORFs suggesting that this phenomenon may be transcription-specific. However, outside the coding regions such a strand bias is absent. This overrepresentation of polypurine tracts in coding strands appears to be a unifying feature even for the four bacterial genomes analysed here. In order to check if this bias in coding strands persists even after transcription, we scanned for the presence of these motifs in expressed sequence tags (ESTs). An analysis of 14,389 EST sequences from C.elegans reveals that even in these sequences polypurines are tenfold more overrepresented than polypyrimidines. We propose that polypyrimidines are universally underrepresented in coding strands (and consequently in mRNA sequences) because the extending mRNA chain with polypyrimidine tracts can transiently form a more stable intermolecular triplex with its progenitor template DNA than a polypurine-rich mRNA. Such a stable triplex formation could lead to severe inhibition of transcription elongation. Thus these motifs could act as cis regulatory elements through an intermolecular hybrid triplex formation of the mRNA with its progenitor template DNA and this mode of regulation could be a universal mechanism.
Crystal Structures of DNA Targets of the E2 Protein from Bovine Papillomavirus-1
H. Rozenberg(1), D. Rabinovich(1), F. Frolow(1), R.
H. Hegde(2) and Z. Shakked(1)
(1)Department of Structural Biology,
Weizmann Institute of Science,
Rehovot 76100, Israel
(2)Skirball Inst. of Biomolecular Medicine,
NYU Medical Center,
New York, NY 10016, USA
The E2 protein is the dominant transcriptional regulator of papillomaviruses. In bovine papillomavirus-1 (BPV-1), E2 binds to a consensus sequence d(ACCN6GGT) which is found 17 times in the BPV-1 genome. The crystal structure of the E2 DNA-binding domain complexed with a 17-mer target d(CCGACCGACGTCGGTCG) has been determined previously (Hegde et al., 1992). To understand the role of DNA structure and hydration in sequence-specific gene-regulatory recognition, it is necessary to compare the DNA target in its free state and when bound to its cognate protein. Here, we present crystal structures of the DNA dodecamer d(ACCGACGTCGGT) which is identical to the central 12 base-pair region of the sequence used in the protein-DNA complex. The DNA fragment was crystallized in two different crystal forms: R3 with one duplex per asymmetric unit, and P1 with three duplexes in the unit cell. This gives a total of four independent molecules for comparison with the bound DNA. The crystal structures were determined at a resolution of 1.6Å and 2.7Å respectively. The free targets adopt a B-DNA conformation with 10.4-10.6 bp/turn similar to the complexed structure. The central six base-pair region of the free molecules is bent by 10-17° towards the major groove, while this region in the bound DNA is bent by 20° towards the minor groove. This observation indicates that the central GACGTC region is inherently flexible. The present system provides an example where the flexibility of a DNA region not contacted by the protein is an important determinant of sequence-specific recognition.
The Crystal Complex of Tetrameric PheRS and Cognate tRNA: The Pattern of Recognition
Mark Safro(1), Yehuda Goldgur(1), Lidia Mosyak(1), Ludmila
Reshetnikova(2), Valentina Ankilova(3), Olga Lavrik(3), Svetlana Khodyreva(3)
(1)Dept. of Structural Biology,
Weizmann Institute of Science,
76100 Rehovot, Israel
(2)Engelhard Institute of Molecular Biology,
RASc,Vavilova 32,
117984 Moscow, Russia
(3)Institute of Bioorganic Chemistry,
RASc,
690390 Novosibirsk, Russia
The crystal structure of Phenylalanyl-tRNA synthetase (PheRS) complexed with cognate tRNA has been determined at 3.28A resolution [1]. It reveals that one tRNA molecule binds across all four PheRS subunits. Multisubunit interaction of this type has not been observed in aaRS systems before and explaining the absolute necessity for the tetrameric architecture of the enzyme. The anticodon recognition upon tRNA binding is performed by the B8 domain, the structure of which is similar to that of the RNA-binding domain (RBD) of the small spliceosomal protein U1A. The tRNA increases the number of RNA types specifically recognized by RBD and, a new type of anticodon binding emerges in the aaRS family. The PheRS approaches the anticodon loop from the minor groove side. In contrast to other aaRS systems (GlnRS, AspRS, LysRS) where anticodon bases (approaching the anticodon loop from the major-groove side) have to protrude out to form base-specific contacts, in PheRS these bases (33-38) are maintained in almost undistorted conformation. The interactions of PheRS with tRNA stabilize the flexible N-terminal part of the small alpha-subunit, which appeared to form the enzyme's 11-th domain [2], comprising a coiled-coil structure (helical arm) built up of two long antiparallel alpha-helices. Judging from the structure of the tRNA(Ser)-SerRS complex, the coiled-coil is a major tRNA(Ser)-recognition element. The existence of the long variable loop in tRNA(Ser) was believed to be a prerequisite for the helical arm to perform its function. The phenylalanine system shows that, along with its specific recognition of the anticodon, the coiled-coil domain is also a part of the PheRS structure. This domain plays an important role in PheRS even in the absence of the long variable loop in tRNA(Phe). Observation of a coiled-coil in PheRS-tRNA(Phe) complex in addition to previously discovered in SerRS [3] raises the problem of occurence of coiled-coils among the aaRS family. Results of coiled-coil predictions [4] suggest that this characteristic domain is common in both classes of aaRS family: they expected to exist from prokaryotes to higher eukaryotes in PheRS, SerRS, ThrRS, AlaRS, dimeric GlyRS of class II and ValRS, CysRS of class I. The fact that the coiled-coil is the only domain shared by two classes of aaRSs allow us to hypothesize that this domain is even more ancient than the characteristic of class I ("Rossmann fold") and class II ("seven-stranded antiparallel beta-sheet") active site domains and have evolved before formation of two classes. The interactions of tRNA(Phe) with PheRS and particularly with N-terminal domain of the alpha-subunit result in conformational changes in TP(si)C and D loops are seen by comparison with uncomplexed yeast tRNA(Phe). The uniqueness of PheRS in charging 2'OH of tRNA is dictated by the size of its adenine-binding pocket and by the local conformation of the tRNA's CCA end.
References
1. Goldgur, Y., Mosyak, L., Reshetnikova, L., Ankilova,
V., Lavrik, O., Khodyreva,S. & Safro, M., Structure 5, 59-68,
1997
2. Mosyak, L., Reshetnikova, L., Goldgur, Y., Delarue, M. & Safro M.,
Nat. Struct. Biol. 2, 537-547,1995
3. Biou,V., Yaremchuk, A., Tukalo, M., & Cusack, S., Science 263,
1404-1410, 1994.
4. Lupas, A., Van Dyke, M. & Stock, J., Science 252, 1162-1164,
1991
Proline Is An Artificial Protein Folding Chaperone
D. Samuel and Chin Yu
Department of Chemistry,
National Tsing Hua University
Hsinchu, Taiwan, ROC
Among the twenty naturally occuring aminoacids, proline has the highest solubility in water. Aqueous solutions of proline at higher concentration(>2M) have been found to aggregate and form 'supramolecular' assemblies. Viscosity and 1-anilino napthalene sulphonic acid binding measurements clearly indicate that the 'supramolecular' assemblies of proline exhibit an amphipathic character. Owing to this amphipathic property, the higher order aggregaties of proline have been found to chaperone the oxidative refolding of hen egg white lysozyme. The results of the present study indicate that proline at higher concentration could behave as an artificial chaperone aiding the in vitro folding of proteins.
New Single Molecule Approaches to Genomic Analaysis: Optical Mapping
Schwartz, D.C.(1), Anantharaman, T.(2), Cai, W.(3), Clarke, V.(1), Delobette, S.(1), Dimalanta, E.(1), Edington, J.(1), Giacalone, J.(1), Hiort, C.(1), Hu, X.(4), Huff, E.(1), Jing, J.(1), Lai, Z.(1), Lee, E.(1), Mishra, B.(2), Murti, J. R.(1), Porter, B.(1), Qi, R.(1), Rabbah, R.(1), Ramanathan, A.(1), Reed, J.(1), Samad, A.(1), Shenker, A.(1), Skiadas, Y.(1), Tankhoveya, K.(1), Wang, W.(1) and Lin, J.(1)
(1) W. M.Keck Laboratory for Biomolecular Imaging,
Department of Chemistry,
New York University,
31 Washington Place, New York, NY 10003
(2) Courant Institute of Mathematical Sciences,
Department of Computer Science,
New York University,
251 Mercer Street,
New York, NY 100012
(3) presently located at: Baylor College of Medicine,
One Baylor Plaza,
Houston, TX 77030
(4) presently located at: CuraGen Corporation, 3
22 E. Main Street,
Branford, CT 06405
Current molecular biological approaches were developed primarily for characterization of single genes, not entire genomes, and, as such, are not ideally suited to analysis of complex traits and population-based molecular genetics. Despite rapid progress in the human genome project effort, there is little doubt that radically new conceptual approaches are needed before routine whole genome-based analyses can be undertaken by both basic research and clinical laboratories.
Physical mapping of genomes, using restriction endonucleases, has played a major role in the identification and characterizing various loci, for example, by aiding clone contig formation and by characterizing genetic lesions. Restriction maps provide precise genomic distances, unlike ordered sequence-based landmarks such as Sequence Tagged Sites (STSs), that are essential for optimizing the efficiency of sequencing efforts, and for determining the spatial relationships of specific loci. When compared to tedious hybridization-based fingerprinting approaches, ordered restriction maps offer relatively unambiguous clone characterization that is useful in contig formation, establishment of minimal tiling paths for sequencing, and preliminary characterization of sequence lesions. Despite the broad applications of restriction maps, the associated techniques for their generation have changed little over the last ten years, primarily because they still utilize electrophoresis. To help overcome these shortcomings, our laboratory developed the first practical non-electrophoretic genomic mapping approach, Optical Mapping, to meet this need.
Optical Mapping is a single molecule methodology for the rapid production of ordered restriction maps from single DNA molecules. Ordered restriction maps were constructed originally from yeast chromosomes by imaging restriction endonuclease cutting events on single, stained DNA molecules with fluorescence microscopy. Cut sites appeared as gaps that widened as the DNA fragments relaxed. Maps were then constructed by measuring fragment sizes via relative fluorescence intensity or apparent length measurements. In the original method, individual fluorescently labeled DNA molecules were elongated in a flow of molten agarose containing restriction endonucleases, generated between a coverslip and microscope slide, and the resulting cleavage events were recorded by fluorescence microscopy as time-lapse digitized images. Nevertheless, improvements were required if a wide range of cloning vectors (cosmid, bacteriophage, P1, YACs) were to be analyzed. Given the utility of such clones, we developed a second generation Optical Mapping approach which dispensed with agarose and time-lapse imaging, and involved fixing elongated DNA molecules onto polylysine-treated glass surfaces. For analysis of lambda clones, DNA samples were fixed onto derivatized glass surfaces by sandwiching between a treated coverslip and glass slide. In these experiments, a cooled CCD camera was used to image molecules from 28Kb down to 800bp ; more recent experiments have lowered the limit of resolution to ~300bp. Sizing accuracy is a function of number of molecules analyzed, and we have achieved rates as low as a few bp on Kb-range restriction digest fragments of lamda bacteriophage, representing much higher accuracy than is achievable by agarose gel electrophoresis.
Ordered restriction maps of Yeast Artificial Chromosomes (YACs) have also been generated in this laboratory by Optical Mapping, with overall relative sizing accuracies that are comparable to routine pulsed field gel electrophoretic analysis. Presently, a large fraction of the human genome is covered by YAC contigs, yet extensive high resolution restriction maps of YACs have not been widely generated. This fact is due primarily to the high frequency of rearrangements/chimerism among YACs, the low complexity of fingerprints generated by hybridization approaches, and the extensive labor required to overcome these problems. The approach taken by this laboratory to Optically Map YACs involved combining the fluid turbulence damping properties of molten agarose, with the stability and enzymatic accessibility of surface mounting, by fixing YACs (ranging up to 440Kb in size) in molten agarose onto derivatized glass surfaces.
Using the approaches discussed above, this laboratory has generated ordered restriction maps for the Beckwith-Wiedeman locus in humans (in collaboration with Dr. D. Housman's group at MIT), the Brca2 locus (in collaboration with Dr. S. Fisher's group at Columbia University), and the mouse olfactory locus (in collaboration with Dr. R. Axel's group at Columbia University). Optical Maps are currently being generated from phage, cosmid, YAC and Bacterial Artificial Chromosome (BAC) clones. In anticipation of the vastly increased throughput requirements for whole genome analysis, we have devised high-throughput approaches involving obviation of a glass sandwich layer. Prior to the development of Optical Mapping, restriction map construction was almost exclusively dependent on gel electrophoretic analysis. By this approach, maps were made for the genomes of E. coli, S. cerevisiae, and C. elegans. Despite the successes of electrophoretically generated maps, numerous limitations of this approach remain, formeost among them being the need for higher throughput with less labor intensiveness. Automation of Optical Mapping also holds enormous promise for miniaturization, with expected increases in throughput and reductions in cost. Thus, advantages of Optical Mapping include high throughput and resolution, safety, and low cost. We expect that the advantages of Optical Mapping will facilitate closure of the initial objecives of the human genome project, mapping of disease genes, and detection/mapping of genetic alterations afflicting the human genome.
Quantitative Gel Studies of the Dimer Linkage Site of HIV
Michael Shubsda, Jerry Goodisman, Philip Borer and James
Dabrowiak
Department of Chemistry,
CST 1-014, Syracuse University,
Syracuse, NY 13244-4100
Two copies of genomic RNA are stored in the HIV virion as a dimer linked near the 5' ends of both strands. Since the dimer-monomer equilibrium is important for the viability of the HIV virus and thus a potential target for drug action, we are studying the equilibrium using quantitative gel methods. A 41-mer RNA sequence occurring in the dimer linkage site of the Mal isolate of HIV was synthesized from the appropriate DNA template using T7 RNA polymerase. After 5' end labeling in the presence of 32-P-ATP and T4 polynucleotide kinase, the 41-mer was electrophoresed in a non-denaturing polyacrylamide gel at 4 C and the locations of RNA visualized using autoradiography. Electrophoresis yielded three major bands which were assigned to the hairpin monomer (fastest moving), duplex dimer (slowest moving) and "kissing dimer". The kissing dimer is a structure having two hairpin monomers joined through Watson-Crick base pairing involving the loops of the monomers. Quantitation of the gel data using scanning techniques followed by analysis revealed that the hairpin monomer-duplex dimer equilibrium constant at 37 C is 1.2 X 10+5 M-1 (100 mM NaCl) and that it varies linearly with NaCl concentration.
Genetic Control of Microsatellite Instability in Saccharomyces cerevisiae
E. A. Sia, R. J. Kokoska, M. Dominska, P. Greenwell,
M. Wierdl and T. D. Petes
Department of Biology and Curriculum in
Genetics and Molecular Biology,
University of North Carolina,
Chapel Hill, NC 27599-3280
Eukaryotic genomes contain many repetitive sequences, called microsatellites. Microsatellites are regions of DNA containing tandem repeats of a single base or a small number of bases. These regions of DNA alter at a significantly higher rate than non-repetitive regions. Consequently, such repetitive sequences are highly polymorphic among individuals within a population. Examination of the changes in these tracts reveal that most of the alterations are deletions or additions of an integral number of repeats (1, 2, 3) . Two models have been proposed to explain the high rates and types of alteration of simple repeated tracts, polymerase slippage and unequal recombination. Most of the available data in vitro and in vivo support the polymerase slippage In this model, transient dissociation of the two DNA strands occurs, followed by a reassociation in a misaligned configuration. This misalignment would result in a loop of unpaired bases consisting of one or more repeat units. Such unpaired loops are predicted to be substrates for DNA mismatch repair. Defects in mismatch repair have been shown to destabilize repetitive tracts in E. coli, yeast and humans (3, 5, 6) .
In Saccharomyces cerevisiae five genes have been demonstrated to be involved in DNA mismatch repair. Three of these, MSH2, MSH3, and MSH6 are homologous to the prokaryotic methyl-directed mismatch repair gene, mutS. Two others, PMS1 and MLH1 are homologous to another component of the prokaryotic mismatch repair system, mutL. Mutations in each of these genes has been demonstrated to destabilize poly GT dinucleotide tracts(5, 7, 8) . Instability of simple repeats has been observed for human hereditary nonpolyposis colorectal cancer (HNPCC) and many sporadic tumors. Subsequently, mutations in the human homologs of the yeast mismatch repair genes were found in many of these tumors. (reviewed in 6)
Using a plasmid-based assay system we have been able to study the rate of alteration of repetitive tracts in the yeast, Saccharomyces cerevisiae and to determine the types of alterations. We have been able to examine parameters affecting the stability of these tracts, such as the length of the tract and the transcriptional state. In addition, we have examined the stability of repetitive sequences with different repeat unit lengths in wild-type yeast strains and in yeast strains deficient in mismatch repair. We have found that even in repair-proficient strains repetitive tracts with repeat units up to 20 base pairs in length are unstable and that repetitive tracts with repeats from 1-13 base pairs are destabilized further by defects in DNA mismatch repair. We interpret this result as indicating that DNA loops up to 13 bases in size can be recognized and repaired by the DNA mismatch repair system, but larger DNA loops are immune to repair by this system.
References
1. S. T. Henderson, T. D. Petes, Mol. Cell. Biol.
12, 2749-2757 (1992).
2. A. J. Jeffreys, et al., Nature Genet. 6, 136-144 (1994).
3. G. Levinson, G. A. Gutman, Mol. Biol. Evol. 4, 203-221
(1987).
4. T. A. Kunkel, Biochemistry 29, 8003-8011 (1990).
5. M. Strand, T. A. Prolla, R. M. Liskay, T. D. Petes, Nature 365,
274-276 (1993).
6. P. Modrich, R. Lahue, Annu. Rev. Biochem. 65, 101-133 (1996).
7. R. E. Johnson, G. K. Kovvali, L. Prakash, S. Prakash, J. Biol. Chem.
271, 7285-7288 (1996).
8. M. Strand, M. C. Earley, G. F. Crouse, T. D. Petes, Proc. Natl. Acad.
Sci. USA 92, 10418-10421 (1995).
Conformational Changes in Fluoromethemoglobin S and Fluoromethemoglobin A Probed by Ultraviolet Resonance Raman Spectroscopy
L. Sokolov, K. Sheffield and I. Mukerji
Department of Molecular Biology and Biochemistry,
Molecular Biophysics Program,
Wesleyan University, Midletown, CT 06459
We are studying the mechanism of fiber formation in sickle cell hemoglobin (HbS), primarily using UV Resonance Raman Spectroscopy (UVRR). Initial investigations have focused on elucidating conformational differences between HbA and HbS tetramers. In these investigations, the a modified hemoglobin, fluoromethemoglobin (FmetHb) is used which allows us to chemically control the allosteric state of the protein.
Polymerization has been observed for the fluoromet derivative of HbS upon addition of the allosteric effector inositol haexaphosphate (IHP). The minimal concentration required for fiber formation at 35 0C is 3.5 mM Fmet HbS with a ten fold excess of IHP. Fibers have also been characterized by electron microscopy.
The structural changes that occur during the transformation from the T to R state are probed using UVRR Spectroscopy. Fluoromethemoglobin A and S are investigated with and without IHP. The changes in environment of Trp, Tyr and Phe residues gives rise to a UVRR spectrum that is characteristic of T state formation.
The difference spectra between FmetHbS and FmetHbA show detectable features that arise from Trp residues. The most striking feature in these spectra is a Trp band occurring at 1555 cm-1 which can be attributed to the a14 Trp and b15 Trp located in the A helix. This feature, which has been previously observed for HbC, is suggestive of a tertiary structural difference between FmetHbA and FmetHbS.
AMBER4.1/PME simulations of the "Hagerman" sequences d[G5-{GA4T4C}2-C5] and d[G5-{GT4A4C}2-C5]
Dennis Sprous, Matthew A. Young and David L. Beveridge
Chemistry Department,
Wesleyan University,
Middletown, CT 06459
Hagerman [1986, Nature 321: 449-450] demonstrated by gel anomaly that DNA segments based on repeats of d[G-A4T4-C]n or d[C-A4T4-G]n are curved while those based on repeats of d[G-T4A4-C]n or d[C-T4A4-G]n are not. Berkoff & Tullius [1988, Nature 331: 455-451] showed that the d[A4T4] motif exhibits a sinusoidal hydroxyl radical cleavage pattern while the d[T4A4] motif does not. This cleavage pattern has been interpreted as evidence for a compressed minor groove; an interpretation consistent with the direction of curvature assigned to A-tracts by Zinkel & Crothers [Nature 328: 178-181] and Trifonov [Ulanovsky & Trifonov, 1987, Nature 326: 720-722; Bolshoy et al., 1991, PNAS 88: 2312-2316]. We have performed AMBER4.1/PME simulations on the 30mers d[G5-{GA4T4C}2-C5] and d[G5-{GT4A4C}2-C5]. At the writing of this abstract, the simulations have progressed to 800 ps for the d[A4T4] motif and to 1000 ps for the d[T4A4] motif. The d[G5-{GA4T4C}2-C5] trajectory shows a compressed minor groove and a stable curving towards the minor groove. The d[G5-{GT4A4C}2-C5] does exhibit these same traits over the 200 to 400 ps time window. However, for the remainder of the 1000 ps simulated to date, the d[G5-{GT4A4C}2-C5] is best described as being straight and having a wide minor groove. These simulations provide atomic resolution structures for A-tract curvature. In our opinion, these simulations strongly complement the low resolution solution structural studies which form the basis of understanding A-tract curvature.
The AMBER4.1 Force Field Favors B-DNA Pucker in 75% Methanol:TIP3P Water (v/v) and in Straight TIP3P Water
Dennis Sprous, Kevin J. McConnell and David L. Beveridge
Chemistry Department,
Wesleyan University
Middletown CT 06459
Ivanov et al. [1974, J. Mol. Biol. 87: 817-833] observed in solution that between 70% to 80% (v/v) alcohol, DNA will undergo a reversible transition from the B-form to the A-form. It is desirable that a potential function for DNA reproduce this phenomena, or at least favor the correct form under correct conditions. Cheatham & Kollman [1996, J. Mol. Biol. 259: 434-444] have demonstrated that B-form is well supported by AMBER4.1 under B-form conditions. Here, we present four simulations using the hexamer d[GCCGGC]: (1) starting from the A-form in 75% methanol, (2) starting form the B-form in 75% methanol, (3) starting from the A-form in straight water and (4) starting from the B-form in straight water. The simulations show that under conditions which in experiment would support A-DNA, the AMBER4.1 force field favors B-DNA pucker. Other parameters such as XDP and INC are in between the two canonical forms, possibly as far to A-DNA as possible with a backbone which favors B-DNA.
Kinetic Study of a (T,C)-motif Triple Helix Formation by Biomolecular Interaction Analysis Technology: a Comparative Solution and Biosensor-Based Study
Jian-sheng Sun, Thérèse Garestier and
Claude Hélène
Laboratoire de Biophysique,
Muséum National d'Histoire Naturelle,
INSERM U201, CNRS URA481
43, rue Cuvier
75231 Paris Cedex 05, France
Over the passed ten years, intermolecular triple helix formations have rised a great deal of intense research due to their potential applications in molecular biology and in therapeutics. A number of studies has been devoted at characterizing the thermal stability as well as the thermodynamics of triple helix formations. However, kinetic data on triple helix formation were limited and fragmentary due to the use of various techniques (biochemical assay, spectroscopy and calorimetry) and under different conditions.
Biosensor technology allows molecular interaction studies in real-time. It provides a direct access to kinetic data on triple helix formation. The present work compared the kinetic parameters of triple helix formation previously determined by analysing the hysteresis of melting curves by heating and cooling triplex solution using UV spectrophotometry (Rougée et al., Biochemistry 31, 9269-9278, 1992) and the same system using BIA technology. It was found that a number of factors affects kinetic parameters measured by biosensor:
1) the presence of dextran molecules on the sensor chip;
2) stable level of baseline after a regeneration pulse;
3) when the streptavidine-biotine capturing system was used for immobilizing
the target duplex, the length of the linker between the tagged biotine and
the duplex;
4) the ratio of biotine/streptavidine, and the density of surface-bound
streptavidine;
5) the time allowed to the dissociation of triplex;
6) the temperature range of measurement.
All these factors should be taken into account in order to determine quantitatively
kinetic data which are not severely biased by the use of surface based biosensor.
Kinetics of triple helix formation will be discussed.
Establishing in vitro Assays for IS903 Transposition
Valeri J. Thomson and Keith Derbyshire
Wadsworth Center
Albany, NY 12201
We are interested in establishing an in vitro system to gain insight into the mechanism of IS903 transposition. IS903 is a 1057-bp insertion sequence that was originally isolated as part of the composite transposon Tn903. IS903 encodes a single polypeptide, transposase, that binds to the 18-bp inverted repeats at the transposon ends, mediates cleavage at both ends and is required for integration into target DNA. We are establishing assays to detect the complete transposition reaction and each of the steps in the transposition process in vitro. The emphasis of my current work is developing assays for some of the individual steps. These assays provide ways of analyzing and characterizing intermediates in transposition and each successful assay can be then be optimized and this information applied to the complex reaction. The in vitro assays will provide a powerful tool for characterizing mutant derivatives of the transposase and for defining the roles of individual amino acids within the transposase.
IS903 transposase protein has been overexpressed and purified in a soluble form as a maltose-binding protein-transposase fusion and I am currently optimizing techniques to purify the native transposase with a carboxy-terminal histidine tag. Using a band shift assay we have determined that both the fusion protein and the histidine-tagged protein bind to the inverted repeat sequence. Two separate assays are being employed to look at the cleavage of DNA containing an inverted repeat sequence. We have designed a very straight forward assay to detect transposase mediated nicks in a supercoiled plasmid substrate. After incubating transposase with a transposon containing plasmid, the reaction products are separated in a tris-acetate agarose gel. Relaxation, specific for plasmids containing the IS903 inverted repeat sequence, will be indicative of transposase-mediated nicking at the transposon ends. We are also using the more sensitive primer extension assay to detect transposase mediated nicks or cleavage events at the transposon ends. This has the additional advantage that the precise site of transposase cleavage can be determined. The final step in transposition is the integration of the cleaved transposon into a target DNA. We have developed an integration assay to detect the insertion of a radiolabeled transposon fragment into a target plasmid by transposase. The reaction products will then be separated on an agarose gel. Insertion of the transposon into the target plasmid would result in the labeling of the plasmid DNA.
By establishing each of these reactions we hope to develop an in vitro transposition system that will allow us to analyze transposition at the molecular level. The role of individual amino acids in the transposase protein will be examined by including the mutant transposase enzymes we have generated in our laboratory.
Left-handed Kind Conformation of Polypeptide Chain in the DNA Minor Groove
Vladimir G.Tumanyan, Dmitry A.Kuznetsov and Natalya
G.Esipova
Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Vavilov str. 32, 117984 Moscow
Extended lefthanded conformation of polyproline type which originally have been found in fibrous protein collagen only at now is discovered in many globular proteins. This conformation being exposed to water provides evident advantages for complex formation. Indeed, recently this conformation was discovered in the influenza virus peptide bonded to the hystocompatibility protein and in the peptide bonded to chaperone Dnak. Naturally, it is tempting to search this conformation in protein-DNA minor groove complexes. A type of conformation of polyproline II kind one can see in the linker regions of Epstein-Barr Virus origin-binding protein in Hin recombinase and etc. Other example where also may realize a complex of extended structure-DNA minor groove complex is the binding of high mobility group chromatin protein with DNA. We undertaken analysis of X-ray data in this aspect. We undertaken also our own study to construct specific complex for some polypeptides under consideration on the basis of polyproline II comformation. Using molecular mechanic program ICM running om Silicon Graphics station we found that the polypeptide helix tightly fit minor groove closing hydrogen bonds between polypeptide and edges of bases. In this respect the role of glycine may be stressed. Complexes via alpha-helix and beta-structure and DNA are rather common. Thus, polyproline II lefthanded helix provides some peculiarities and preferences for DNA-protein complex formation.
Raman Spectroscopy as a Probe of DNA Minor-Groove Binding: The Complex of a High-Mobility-Group Domain (hSRY) and its Target DNA
James M. Benevides(1), Michael A. Weiss(2) and George
J. Thomas, Jr.(1)
(1)School of Biological Sciences,
University of Missouri,
Kansas City, MO 64110
(2)Center for Molecular Oncology,
The University of Chicago,
Chicago, IL 60637
Previous studies of DNA-protein complexes by Raman spectroscopy1 show that sequence-specific recognition can be distinguished from nonspecific binding. In the case of major-groove-binding proteins, such as transcriptional activators of the bacteriophages l and D108, the Raman signature of sequence-specific recognition includes markers of altered phosphodiester backbone conformation, altered nucleoside sugar conformation and base-specific interactions involving major groove sites of guanine (N7, O6) and thymidine (5CH3).2 In the case of nonspecific binding, these spectral markers are not observed. To extend this approach to proteins that bind specificity but utilize the minor groove, we have investigated binding between the hSRY-HMG box and a DNA target site. The hSRY-HMG box consists of an 87-residue fragment of hSRY, a 203-residue protein of the HMG domain to its DNA target occurs specifically via the minor groove and is accompanied by large conformational perturbances of DNA.3 The DNA target site in the present Raman study is the sequence d(5'GATTGTTA0·(3'CTAACAAG0. A second hSRY-HMG box containing the mutation of valine-60 to leucine (V60L) has also been investigated. NMR studies and assays of hSRY-HMG binding suggest that V60 is of critical importance for presenting the correct surface to the DNA target.3,4
The principle findings in this investigation are: (i) Large perturbations to the Raman spectrum of both DNA and protein occur with binding. This is consistent with sequence-specific recognition. (ii) Both the wild type and mutant (V60L) proteins produce a qualitatively similar Raman difference spectrum, indicating similar modes of interaction with DNA. (iii) The region of the Raman spectrum that is most sensitive to the conformation of the DNA phosphodiester backbone (i.e. 800-1000 cm-1) exhibits the greatest perturbations with hSRY-HMG binding, consistent with the DNA bending and helix unwinding proposed previously on the basis of NMR studies. (iv) Raman bands sensitive to base stacking exhibits enhanced intensities in the complex. This loss if Raman hypochromism implies increased rise between base pairs. A similar conclusion has been reached from the NMR results. (v) The Raman signature of the DNA/hSRY-HMG complex exhibits no markers of C3'-endo nucleoside conformation. (vi) Unlike previously investigated major-groove-binding proteins, hSRY-HMG does not interact with guanine N7 sites. This result is consistent with major-groove-binding.
We conclude that Raman spectroscopy provides an efficient and effective means of probing sequence specific recognition by closely related variants of the hSRY-HMG box. The present results also demonstrate that the Raman spectrum provides a unique signature of DNA minor-groove binding, which is distinct from previously reported Raman fingerprints of DNA major-groove and nonspecific binding. [NIH grant GM50776].
References
1. Benevides, J. M., Terwilliger, T. C., Vohník,
S. and Thomas, G. J., Jr., Biochemistry 35, 9603, 1996.
2. Benevides, J. M., Weiss, M. A. and Thomas, G. J. Jr., J. Biol. Chem.
269, 10869, 1994.
3. Werner, M. H., Huth, J. R., Gronenborn, A. M. and Clore, G. M., Cell
81, 705, 1995.
4. Harley, V. R., Jackson, D. I., Hextall, P. J., Hawkins, J. R., Berkovitz,
G. D., Socknathan, S., Lovell-Badge, R. and Goodfellow, P. N., Science
255, 453, 1992.
Oligodeoxyribonucleotide N3-P5' Phosphoramidate Duplexes Have A-Helix Conformations And Can Serve As Decoys In Binding The HIV-1 Rev Protein
W. David Wilson(1), Daoyuan Ding(1), C. Ted Rig(l),
Sergei M. Gryaznov(2) and David H. Lloyd(2)
(1)Department of Chemistry,
Georgia State University,
Atlanta, GA, 30303
(2)Lynx Therapeutics, Inc.,
3832 Bay Center Place,
Hayward, CA 94545
The solution conformation of the duplex d(CGCGAATTCGCG)2 modified with N3'-P5' phosphoramidate internucleoside linkages has been determined by 2D NMR methods, and compared to the isosequential phosphodiester RNA and DNA. NMR studies of model dimers show that the sugar ring conformation changes from predominantly C2'-endo to C3'-endo when the 3'-phosphoester is replaced by a phosphoramidate group. 2D NMR studies of the duplex indicate that it is closest to an A-DNA helix type, and confirm that all furanose rings of 3'-aminonucleotides adopt predominantly N-type sugar puckering.
The A-form conformation of the phosphoramidates indicates that they could mimic RNA in biological systems, and for example, could bind to control regions of viral genomes and act as "decoys" to sequester viral gene regulatory proteins. In agreement with this idea, we have found that oligonucleotide N3'-P5' phosphoramidate DNA analogs of HIV-1 RRE form stable duplexes that exist in A-form as judged by circular dichroism. Moreover, gel shift assays demonstrate that these phosphoramidates can specifically bind to peptides derived from HIV-1 Rev. Isosequential phosphodiester DNA duplexes do not bind to the respective peptides under the experimental conditions used. These results suggest the possibility that nuclease-resistant oligonucleotide N3'-P5' phosphoramidates might serve as RNA-like decoys and disrupt specific viral RNA/protein interactions such as RRE/Rev and TAR/Tat in HIV-1.
Supported by NIH and Lynx Therapeutics, Inc.
Back to Table of Contents: Abstracts
Back to Table of Contents: Tenth
Conversation