Olga F. Borisova, Anna K. Shchyolkina, Irina A. Il'icheva and Michael
A. Livshits
Engelhardt Institute of Molecular Biology,
Russian Academy of Sciences,
Moscov, 117984, Vavilova 32, Russia
Telomeres are important multifunctional nucleoprotein structures found at the ends of eukaryotic chromosomes. Telomeric DNAs composed of GnTm tandem repeats are able to form compact four-stranded structures (G-quadruplexes) in vitro. In this work, the conformational flexibility of G-quadruplexes was examined with the help of intercalating fluorescence probes ethidium bromide (EtBr) and acridine orange (AO), which are capable to extend and untwist DNA multihelices. Linked oligonucleotides 5'-d(GT)5-pO(CH2CH2O)3p-d(GT)5-5' (parGT) and d(TTAGGG)4 were synthesized, which are capable to form G-quadruplexes (1-3).
The tertiary structure of the oligonucleotides parGT and d(TTAGGG)4 was revealed with the aid of several independent methods: UV-melting, CD-spectra, and two specially designed fluorescence approaches for determination of hydrodynamic volumes and the fraction of unpaired nucleotides. All experiments were run in 10 mM Na-phosphate buffer, pH 7, 0.1 M NaCl at 3°C. It was demonstrated that d(TTAGGG)4 forms rather stable intramolecular antiparallel G-quadruplex (Tm = 48°C). ParGT oligonucleotide was proved to form a weak bimolecular parallel four-stranded structure, containing five stacked G-quartets and bulged out thymines (GT4 quadruplex).
The ability of these two types of G-quadruplexes to bind EtBr and AO was examined and compared to that of DNA duplexes. A tenfold increase in the EtBr fluorescence yield and lifetime, as well as a threefold increase of these parameters for AO upon their binding to DNA, GT4 and d(TTAGGG)4, were indicative of intercalation of the ligands. The binding parameters for EtBr and AO molecules were estimated using the adsorption isotherms which were obtained by means of fluorimetry techniques. EtBr and AO intercalate into the GT4 and d(TTAGGG)4 with the affinity three times lower than into DNA duplexes. Eight EtBr molecules have been found to associate with GT4 by an intercalating cooperative mode leading to a twofold extension of the quadruplex length. In contrast, only two EtBr molecules can be non-cooperatively bound to d(TTAGGG)4.
The computer modelling of the GT4 architecture has demonstrated the possible formation of parallel GT4 quadruplexes with either intercalated or bulged out thymines. However, in GT4-EtBr complex with two EtBr molecules intercalated between each two adjacent guanine tetrads, the thymines are apparently bulged out.
Contrary to rather rigid d(TTAGGG)4 and GnTm quadruplexes, the GT4 quadruplexes are characterized by extraordinary conformational flexibility, which may be functionally important for specific nucleoprotein complexes.
Acknowledgment
This work was supported by Russian Foundation for Fundamental Research, grant N 96-04-50703 and in part by the State program for Support of the Russian Scientific Schools, N 98093.
References and Footnotes
1. O.F. Borisova, A K. Shchyolkina, E.N. Timofeev, V.L.
Florentiev, FEBS Letters. 1992, v. 306, p. 140-142.
2. Y. Wang and D.J. Patel, Biochemistry. 1992, v. 31, p. 8112-8119.
3. P. Balagurumorthy and S.K. Brahmachary, J. Biological Chemistry. 1994,
v.269, p. 21858-21869.
George C. Ruben1*, Jian-Zhi Wang2, Inge Grundke-Iqbal2 and
Khalid Iqbal2
1Dept. Biol. Sci,
Dartmouth College,
Hanover, NH 03755,
2NYS Inst. Basic Research in Develop. Disabilities,
Staten Island, NY 10314,
*Author to whom correspondence should be addressed. E-mail: GCR@Dartmouth.edu
Since dementia in Alzheimer's Disease (AD) correlates with prevalence of intraneuronal neurofibrillary tangles, understanding the chemistry of tangle disassembly could yield clues for prevention and treatment. Neurofibrillary tangles consist of Paired Helical Filaments, 10-15 nm cylindrical filaments and 2.1 nm tau filaments in varying proportions (Ruben et al., Brain Res. 590:164, 1992). Tangle filaments are composed of abnormally phosphorylated tau (AD P-tau) that are also glycosylated. Maximum disassembly of the PHF requires both deglycosylation and dephosphorylation, which restores tau's ability to stimulate microtubule formation. The PHF were deglycosylated with endoglycoidase F/N-glycosidase F and/or dephosphorylated with protein phosphatase-2A (PP-2A). The enzymatic treatments of PHF have been described (Wang et al., Nature Med. 2, 1996, 71; J. Biol. Chem. 270, 1995, 4854).
The effects on PHF preparations by deglycosylation and by dephosphorylation separately and in combination were studied with TEM. Treated and untreated PHF samples were prepared on thin (~10 nm) indirectly evaporated carbon films (+/-0.8 nm rough) treated with 25% aldehyde vapor (2 hr) so that PHF spread from a 10 $B5l drop would stick and 2% phosphotungstate acid (pH 6.8) would coat the PHF and the carbon film. Negatively stained PHF preparations were photographed with a JEM 100cx at 100 kilovolts at Schertzer focus at ~140 nm underfocus. Phase objects like the PHF were imaged at ~0.7 nm resolution.
The PHF repeat period, L, averaged from 77.6 nm to 85.0 nm. The PHF widest regions, W, averaged from 16.6 nm (without sonication) to 21.1 nm (with 1 min sonication) depending on prior treatmen. The PHF thin regions, T, averaged from 5.8 nm (w/o sonication) to 8.9 nm (w sonication). Cylindrical filaments with periodic thin regions (CF-PT) and without (CF) were present. The unmodulated filaments (CF) had diameters averaging 10.3 nm (w/o sonication) to 15.3 nm (w 1 min sonication) and the modulated filaments (CF-PT) had thin regions spaced at L = ~76.6 nm, wide regions W = ~13.8 nm and thin regions T = ~7.0 nm. After 1min sonication the width of the CF-PT was enlarged but most of the CF-PT also disappeared.
Deglycosylation (0.5 hr) and dephosphorylation (1 hr) disassembled nearly
all the PHF leaving residual filament-like traces on the carbon film. These
residual filament-like traces were also found in the dephosphorylated preparation.
Dephosphorylation (1 hr) of PHF (w sonication) reduced W from 18.7 nm to
12.7 nm and T from 7.2 nm to 4.6 nm but left the turn period, L, unchanged
. The cylindrical filaments, CF, were also reduced from 11.3 nm to 7.5 nm.
Deglycosylation (1 hr) only relaxed the PHF so the appearance of right-hand
helicity was lost. Dephosphorylation alone reduced PHF width but did not
reveal two core filaments shown in the PHF model in this poster. The cylindrical
filament modulation period (L = ~76.6 nm) suggests that it is a likely precursor
of the PHF. These deglycosylation and dephosphorylation studies have implications
for the removal of tangles in AD and indicate that PHF can be enzymatically
disassembled.
A. Ishchenko1, V. Koval1,
O. Fedorova1*, K. Douglas2
and G. Nevinsky1
1Novosibirsk Institute of Bioorganic Chemistry,
Siberian Division of the Russian Academy of Sciences,
Novosibirsk 630090, Russia
2School of Pharmacy and Pharmaceutical Sciences,
University of Manchester,
Manchester, M13 9PL, UK
*Author to whom correspondence should be addressed. Phone: +7(3832)34-42-74;
Fax: +7(3832)33-36-77; E-Mail: fedorova@niboch.nsc.ru
Formamidopyrimidine (Fpg protein or 8-oxoguanine DNA glycosylase) from E. coli catalyzes excision repair of several damaged purine bases, including 8-oxoguanine (oxoG) and 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine from DNA. The interaction of this enzyme with small ligands and oligonucleotides (ODNs) was analyzed. Orthophosphate (Ki = 10 mM), deoxyribosophosphate (Ki = 8.3 mM) and different deoxyribonucleotide monophosphates (Ki = 3.3-8.3 mM) were shown to be the minimal ligands of the enzyme. The lengthening of non-specific ODNs (d(pN)n, n = 1 to10) by one nucleotide unit increased their affinity by a factor of ~1.5. Weak, non-specific electrostatic contacts of Fpg with inter-nucleoside phosphate groups of ODNs can account for about 4 orders of magnitude in the ligand affinity, whereas the contribution of specific interactions between Fpg and d(pN)n is no more than 1 order of magnitude of an ODN's total affinity. Addition of the second complementary chain increased the affinity of the first chain by a factor of ~100. For a stretch of 10-13-links or base-pairs in long, double-stranded DNA the contribution of weak, non-specific contacts to the specific substrate affinity was estimated to be about 7 orders of magnitude and found to predominate over the specific ones (1-2 orders of magnitude) due to the additivity of free energies of recognition of individual nucleotide units. Thus, complex formation of Fpg with cognate DNAs cannot provide the explanation of the specificity of the enzyme's action.
A relative contribution of the adjustment of ODNs to the optimal conformation for the catalytic enzyme conformation and the catalytic step of the reaction providing the specificity of the enzyme action was also investigated. The high specificity of the enzyme functioning may be provided by the DNA adaptation processes and the kinetic constant of the reaction. Changing the substrate from non-specific to specific DNA leads to an increase of the catalytic constant kcat by 6-7 orders of magnitude. The enhancement of DNA substrate processing rate occurs if the DNA molecule can change its structure due to interaction with the enzyme to be optimal for catalysis.
In order to reveal different individual steps of FPG-DNA complex formation and to measure the rate constants of individual steps on the reaction pathway, as well as to detect transient intermediates, the interaction of the enzyme with ODNs is being studied using stopped-flow techniques with fluorescent detection. In order to trap various specific intermediate(s) of FPG with substrates, and to analyze the pre-steady-state kinetics of FPG complexes, photoreactive derivatives of ODNs and pulsed-laser irradiation are being used.
Acknowledgments
The research is supported by grants from The Wellcome Trust (UK) and from the Siberian Division of Russian Academy of Sciences.
References and Footnotes
1. A. A. Ishchenko, N. V. Bulychev, G. A. Maksakova, F.
Johnson and G. A. Nevinsky, Mol. Biol. (Moscow) 32, 454-462 (1998).
2. A. A. Ishchenko, N. V. Bulychev, G. A. Maksakova, F. Johnson and G. A.
Nevinsky, Biochem. Mol. Biol. Intern., accepted for publication.
O. Fedorova1*, V. Koval1,
A. Adeenah-Zadah1, N. Karnaukhova2
and S. Lokteva2
1Institute of Bioorganic Chemistry,
Lavrentyev Pr. 8,
Novosibirsk 630090, Russia
2Novosibirsk State University,
Pirogova St. 2,
Novosibirsk 630090, Russia
*Author to whom correspondence should be addressed. Phone:+7(3832)34-42-74;
Fax:+7(3832)33-36-77, E-Mail: fedorova@niboch.nsc.ru
Cooperative interactions occur between oligonucleotides bound at neighboring sites on single-stranded (1) and double-stranded nucleic acids (2). In this case the specific binding constant of each oligonucleotide is enchanced. The nature of cooperative interaction largely arises from stacking between bases at the duplex junction. Earlier we developed an original approach for evaluation of the parameters of cooperativity (3,4). This approach is based on the measurement of the equilibrium association constant of the oligonucleotide with single-stranded DNA in the absence and presence of a second bound oligonucleotide occupying an adjacent site. The first oligonucleotide is covalently linked through its 5'-phosphate with alkylating N-(2-chloroethyl)-N-methylaminophenyl group (RCl). The equilibrium association constant of the alkylating derivative of the oligonucleotide (Kx) is calculated using the dependence of extent single-strandedDNA alkylation ((() on the concentration of the alkylating derivative of oligonucleotide (x0) at time > infinity: vx=1-exp[-gefKxX0/(7+KxX0)] (where (gef is the efficiency of DNA modification within the duplex by the reagent).
This work is devoted to the study of the effects of single-base pair mismatch at the junction on the parameters of cooperativity. The base stacks 5'-PyPy-3', 5'-PuPy-3', 5'-PyPu-3' and 5'-PuPu-3' were studied. Eight deoxyribooligonucleotides, each 22 nucleotides in length were used as a target ssDNA. They differed only by nucleotides in the 11th and/or 12th positions, providing the site of mismatch (Table I).
Table I
The structures of targets, reagents and neighbor oligonucleotides.
'*Underlined nucleotide sequences form duplexes
with reagent and neighbor oligonucleotide.
**Underlined nucleotides at the junction.
The data obtained demonstrate that the cooperativity is practically absent in the stack 5'-PyPy-3' and that the parameters of cooperativity increased in the sequence 5'-PyPy-3' < (5'-PuPy-3', 5'-PyPu-3') < 5'-PuPu-3'.
Acknowledgements
This work was supported by grant from Russian Foundation for Basic Research (98-04-48312).
References and Footnotes
1. C. R. Cantor and P. R. Schimmel, Biophysical Chemistry.
Part III: The Behavior of Biological Macromolecules: W. H. Freeman and
Co.: New York, NY (1980).
2. N. Colocci and P. B. Dervan, J. Am. Chem. Soc. 117, 4781 -4787
(1995).
3. O. S. Fedorova, A. Adeenah-Zadah, E. V. Bichenkova and D. G. Knorre,
J. Biomol. Struct. Dyn. 13, 145-166 (1995).
4. A. Adeenah-Zadah, D. G. Knorre and O. S. Fedorova, J. Biomol. Struct.
Dyn. 15, 369-380 (1997).
Vladimir V. Koval1*, Elena V. Bichenkova2, Kenneth T. Douglas2, Michail
I. Dobrikov1 and Olga S. Fedorova1
1Institute of Bioorganic Chemistry,
Lavrentyev Pr. 8,
Novosibirsk 630090, Russia
2School of Pharmacy and Pharmaceutical Sciences,
University of Manchester,
Manchester, M13 9PL, UK
*Author to whom correspondence should be addressed. Phone:+7(3832) 39-62-74;
Fax:+7 (3832) 33-36-77, E-Mail: koval@niboch.nsc.ru
Photoaffinity labels are widely used to study the structure and dynamics of biopolymers. The advantage of such reagents in comparison with the chemically active reagents consists in the possibility to initiate modification reaction in any time. Increased specificity for nucleic acid sequences may be obtained using a pair of oligonucleotide derivatives, activatable due to an enforced proximity once arranged in complementary complexes. One example of this, so-called binary reagents, has recently been proposed (1). This system consists of two tandem oligonucleotides complementary to a target sequence of a nucleic acid. Each oligonucleotide is covalently linked through its terminal phosphate group with a photo-reactive or photo-sensitizing groups. For irradiation at the wavelength of sensitizer excitation, photoactivation occurs only due to energy transfer from the sensitizing group.
It has already been shown that in the case of p-azidotetrafluorobenzyl and pyrenyl-1-methylamino groups as a photo-reactive/sensitizing pair, the efficiency of target modification depends on the mutual arrangement of photo-reactive and -sensitizing groups at their interface (2). Preferably the photo-reactive and sensitizing groups were attached to 3'-end and 5'-end phosphate groups of oligonucleotides at their interface, respectively. Structural analysis was performed by high-resolution 2D NMR and molecular modelling (3).
In the present work we have studied the influence of nucleotide composition at the binary junction on the efficiency of modification of target single-stranded DNA. For this purpose the photo-modification of 22-mer deoxyribooligonucleotide with three different binary reactive systems was performed:
where Pyr is the pyrenyl-1-methylamino residue and ArN3 is a p-azidotetrafluorobenzyl residue attached to phosphate through an ethylenediamine spacer.
At 25°C in system (I) base G11 was modified with 25% efficiency. In system (II) the same base was modified with an efficiency of 40%. In system (III) bases G14 ·G15 were equally modified with a total efficiency 60%. Molecular modelling analysis, using the Kollman-All force field and the known 3D coordinates from NMR studies of the analogous binary pyrene-arylazide system (3), was carried out for systems I to III to provide a structural basis for the efficiency of energy transfer and labelling.
Supported by a grant from the Russian Foundation for Basic Research (98-04-48312).
References and Footnotes
1. I. Dobrikov, S. A. Gaidamakov, T. I. Gainutdinov, A.
A. Koshkin and V. V. Vlassov, Antisense Nucl. Acid Drug Dev. 7, 309-317
(1997).
2. I. Dobrikov, S. A. Gaidamakov, A. A. Koshkin, T. I. Gainutdinov, N.
P. Luk'yanchuk, G. V. Shishkin and V. V. Vlassov, Russsian J. Bioorg.
Chem. 23, 171-178 (1997).
3. V. Bichenkova, D. S. Marks, S. G. Lokhov, M. I. Dobrikov, V. V. Vlassov
and K. T. Douglas, J. Biomol. Struct. Dyn. 15, 307-320 (1997).
Xie, X., Geacintov, N. E. and Broyde, S.
Departments of Chemistry and Biology,
New York University,
New York NY 10003
Covalent binding of the two enantiomeric metabolites of benzo[a]pyrene, (+)- and (-)-anti-BPDE (7R,8S-dihydroxy-9S,10R-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene and the corresponding 7S,8R,9R,10S enantiomer, respectively) to deoxyguanosine residues in DNA occurs by cis or trans addition to yield four isomers. The 10S(+)-trans and 10R (-)-trans adducts assume oppositely oriented minor groove conformations in normal, fully base paired duplexes, while the 10R (+)-cis and 10S (-)-cis adducts adopt oppositely oriented base displaced-intercalated conformations. We have carried out an extensive computational study of these four adducts on the nucleoside level, creating 373,248 different conformers for each adduct and evaluating the energy of each, to pinpoint the origin of the opposite orientations and the trans/cis differences. Steric hindrance rooted in the mirror image relationship of the benzylic ring in a (+)/(-) pair causes the opposite orientations in both trans and cis isomers. The configurational difference between trans and cis adducts in the benzylic ring directly account for their distinct conformational preferences: the minor groove is disfavored by cis adducts because the benzylic ring O9H and O8H groups are directed inward in a crowded situation in the minor groove position, while these groups are directed outward from the minor groove in the trans adducts. Observed differential processing of these four adducts by replication, repair and transcription enzymes may well stem from their differing conformational preferences.
Supported by NIH and DOE.
Thomas L. James, Zhihua Du, Anton V. Filikov, Kenneth E. Lind and
Anwer Mujeeb
Department of Pharmaceutical Chemistry,
University of California,
San Francisco, CA 94143-0446
In recent years there has been considerable interest in academia, in public health agencies, and in the pharmaceutical industry in the development of oligonucleotide analogs recognizing mRNA or duplex DNA based on primary sequence. We have been using NMR to examine some of the structural attributes of DNA-based oligonucleotides, containing chiral phosphorothioates or methylphosphonates, paired with a target RNA oligonucleotide (1,2). There has also been much effort to design diagnostic and therapeutic agents which bind to protein receptors based on their three-dimensional structure. In contrast, there has been little effort to design drugs rationally on the basis of the three-dimensional structure of DNA or RNA. The reason for this is undoubtedly due to the paucity of three-dimensional structures of possible targets available. However, structure determination methods have advanced to the state that structures of potential DNA or RNA targets are increasingly available. Consequently, we can now entertain the concept of designing agents to bind to gene targets based on the detailed three-dimensional structure of the target. While our lab has been involved with determining the structure of a few different DNA and RNA molecules, we have also begun developing a strategy to develop ligands based on DNA and RNA structure. We have used a pseudobrownian Monte Carlo minimization in torsion angle space for efficient flexible docking of compounds from the Available Chemicals Directory (ACD) (3,4). To aid in ranking the ligandsaccording to binding avidity, an empirical binding free energy function was developed. Some experimental testing of compounds identified computationally has been promising.
References and Footnotes
1. A. Mujeeb, M. A. Reynolds and T. L. James, Biochemistry
36, 2371-2379 (1997).
2. P. Furrer, T. M. Billeci, A. Donati, C. Kojima, B. Karwowski, A. Sierzchala,
W. Stec and T. L. James, J. Mol. Biol. 285, 1609-1621 (1999).
3. R. A. Abagyan, in "Computer Simulations of Biomolecular Systems:
Theoretical and Experimental Applications" (W. van Gunsteren, ed.)
pp. 363-394, Kluwer Academic Publishers, Dordrecht, Vol. 3 (1997).
4. A. V. Filikov and T. L. James, J. Comp.-Aided Molec. Design 12,
1-12 (1998).
John G. Arnez*1,3, Rajan Sankaranarayanan*1, Anne-Catherine Dock-Bregeon*1,
Christopher S. Francklyn*2 and Dino Moras*1
1Laboratoire de Biologie Structurale du CNRS,
IGBMC, BP 163,
67404 Illkirch Cedex, France
2Department of Biochemistry,
University of Vermont,
Burlington, VT 05405, USA
3now at Department of Biochemistry,
Albert Einstein College of Medicine,
Bronx, NY 10461, USA.
*For author's correspondence. J.G. Arnez: 718-430-3233, 718-430-8565, arnez@aecom.yu.edu
R. Sankaranarayanan: +33 3-88.65.33.10, +33 3-88.65.32.76, sankar@igbmc.u-strasbg.fr
A.-C. Dock-Bregeon: +33 3-88.65.33.10, +33 3-88.65.32.76, dock@igbmc.u-strasbg.fr
C.S. Francklyn: 802-656-8450, 802-862-8229, franck@emba.uvm.edu
D. Moras: +33 3-88.65.33.51, +33 3-88.65.32.76, moras@igbmc.u-strasbg.fr
The crystal structures of histidyl- (HisRS) and threonyl-tRNA synthetase (ThrRS) from E. coli and glycyl-tRNA synthetase (GlyRS) from T. thermophilus, all homodimeric class IIa enzymes, were determined in enzyme-substrate and enzyme-product states corresponding to the two steps of aminoacylation. HisRS was complexed with the histidine analog histidinol plus ATP and with histidyl-adenylate, while GlyRS was complexed with ATP and with glycyl-adenylate; these complexes represent the enzyme-substrate and enzyme-product states of the first step of aminoacylation, i.e. the amino acid activation. In both enzymes the ligands occupy the substrate-binding pocket of the N-terminal active site domain, which contains the classical class II aaRS fold. HisRS interacts in the same fashion with the histidine, adenosine and a-phosphate moieties of the substrates and intermediate, and GlyRS interacts in the same way with the adenosine and a-phosphate moieties in both states. In addition to the amino acid recognition, there is one key mechanistic difference between the two enzymes: HisRS uses an arginine whereas GlyRS employs a magnesium ion to catalyze the activation of the amino acid. ThrRS was complexed with its cognate tRNA and ATP, which represents the enzyme-substrate state of the second step of aminoacylation, i.e. the transfer of the amino acid to the 3'-terminal ribose of the tRNA. All three enzymes utilize class II conserved residues to interact with the adenosine-phosphate. ThrRS binds tRNAThr so that the acceptor stem enters the active site pocket above the adenylate, with the 3'-terminal OH positioned to pick up the amino acid, and the anticodon loop interacts with the C-terminal domain whose fold is shared by all three enzymes. We can thus extend the principles of tRNA binding to the other two enzymes.
James M. Benevides*, Liviu Movileanu and George J. Thomas, Jr.
School of Biological Sciences,
Division of Cell Biology and Biophysics,
University of Missouri,
Kansas City, MO 64110
*Author to whom correspondence should be addressed. Phone: 816-235-2499;
Fax: 816-235-1503; E-mail: benevidesj@umkc.edu
Thermal denaturation studies of DNA provide a basis for evaluating relationships between base sequence and helix stability. Factors which affect such stability are: (i ) hydrogen bonding between complementary bases, (ii ) stacking of vertically neighboring bases, (iii ) electrostatic repulsions of charged phosphates and (iv ) entropy related to internal degrees of freedom of the duplex. Whereas factors (i ) and (ii ) promote helix formation, (iii ) and (iv ) oppose it. Although many techniques can measure DNA melting, Raman spectroscopy may prove one of the more informative. UV absorption spectroscopy, for example, measures primarily factor (ii) but is less informative of others. Raman spectroscopy, however, has the potential to resolve all four of these factors, as well as monitor changes in phosphodiester conformation (1). The Raman method is particularly advantageous for studying "premelting" phenomena, i.e. structural modifications preceding the onset of strand separation.
Although use of Raman spectroscopy to probe temperature transitions in DNA is not new, previous applications have been limited by instrumental constraints. Currently available spectrometers and detectors permit rapid accumulation of spectra of greatly improved signal-to-noise ratio. Here, we use a state-of-the-art Raman system to probe melting of the B-DNA sequence isomers, poly(dA-dT)·poly(dA-dT) and poly(dA)·poly(dT). These DNA structures are of interest because of possible differences between alternating and nonalternating A·T motifs in such phenomena as B-DNA bending and protein recognition. Bifurcated hydrogen bonds in poly(dA)·poly(dT) may also contribute to unusual melting properties of this B-DNA model (2,3). A consistent Raman assignment scheme has been reached for all prominent bands in the spectra of both isomers and several new assignments have been made for vibrations of base and backbone residues. Temperature dependence of the Raman spectra of poly(dA-dT)·poly(dA-dT) and poly(dA)·poly(dT), although qualitatively similar, are quantitatively different and indicate greater thermostability of poly(dA)·poly(dT), in accord with earlier studies (4). In the premelting domain (15-65°C) both DNA models exhibit significant changes in interbase hydrogen bonding and phosphodiester backbone conformation, without significant change in base stacking. The results are consistent with bifurcated hydrogen bonding of the bases in poly(dA)·poly(dT). [NIH Grant GM54378].
References and Footnotes
1. J. Duguid, V.A. Bloomfield, J.M. Benevides, G. J. Thomas,
Jr., Biophys. J., 71, 3350-3360, 1996.
2. S.S Chan, K.J. Breslauer, M.E. Hogan, D.J. Kessler, R.H. Austin, J. Ojemann,
J.M. Passner, N.C. Wiles, Biochemistry, 29, 6161-6171, 1990.
3. S.S. Chan, K.J. Breslauer, R.H. Austin, M.E. Hogan, Biochemistry,
32, 11776-11784, 1993.
4. I. Haq, B. Z Chowdhry, J. B. Chaires, Eur. Biophys. J. 26, 419-426,
1997.
Corie Ralston1, Bianca Sclavi1, Mike Sullivan1, Michael
Brenowitz2 and Mark Chance*1,2
1Department of Physiology & Biophysics,
2Department of Biochemistry,
Albert Einstein College of Medicine,
1300 Morris Park Ave.,
Bronx NY 10461
*Author to whom correspondence should be addressed. Phone: 718-430-4136;
E-mail: mrc@aecom.yu.edu
Background
The speed of chemical reactions carried out by RNA catalysts, or ribozymes, are often limited by conformational changes in the RNA (1). As a result, the process by which RNA molecules fold into their native conformation has received much attention. Early investigations into the folding of tRNA established approximate timescales for the formation of secondary (10-4 - 10-5 s) and tertiary interactions (10-2 - 10-1 s), with reorganization of incorrect secondary structure occurring more slowly (0.1 - 1 s)(1). Recent work has shown that folding of large RNAs is complex (1,2). Individual domains of an RNA may form at rates that differ by orders of magnitude, with some transitions requiring minutes to reach completion (3,4,5).
Establishing the paths by which large RNAs fold has been hampered by the lack of experimental methods capable of probing RNA conformation with nucleotide resolution at subsecond timescales (1). We have developed stopped-flow footprinting of RNA using a synchrotron x-ray beam to produce hydroxyl radicals. This approach provides the first direct visualization of the major steps of the folding pathway of the Tetrahymena ribozyme. The 385 nt. ribozyme derived from the Tetrahymena group I intron (Figure 1) folds into a well-defined tertiary structure in the presence of Mg2+, which is required for activity (1). A 160 nt domain containing paired regions P4-P6 (Figure 1) folds independently, and its structure determined by x-ray crystallography (4).
Figure 1: Summary of the rates of formation of tertiary interactions within the domains of the Tetrahymena ribozyme.
Results
The folding kinetics of the ribozyme were quantitated by determining the changes in solvent accessibility of individual sites as a function of time (Figure 2). Specific nucleotides within the P4-P6 domain became protected from hydroxyl radical cleavage within 100 ms after addition of Mg2+, and the extent of protection reached a plateau within several seconds (Figure 2). A subset of nucleotides in P5c (Figure 1) were protected about twice as rapidly as other regions in the P4-P6 domain. This region is a Mg2+-rich subdomain suggesting that formation of a "metal ion core" in P5a-P5c could serve as a nucleation site for additional tertiary structure formation.
Figure 2: A kinetics progress curve obtained by stopped-flow synchrotron x-ray footprinting for tertiary structure formation of the Tetrahymena ribozyme. The insert is an expansion of the first three seconds of the reaction.
The ability to probe solvent accessible regions of an RNA backbone within the first tens of milliseconds of a reaction provides the first direct visualization of all major steps in the folding pathway of the Tetrahymena ribozyme. After addition of Mg2+, the earliest evidence of tertiary structure appears within the P5a-P5c subdomain. Subsequent collapse of the P4-P6 domain is concerted and occurs in a few seconds. This is followed by establishment of interdomain contacts with P2 & P9 within 10 s. Organization of the catalytic core, including formation of P3 and P7, requires 1 - 10 minutes (5).
Conclusions and Outlook
The results presented in this report demonstrate that most of the RNA tertiary structure can form in a physiologically relevant time in vitro. Under these conditions, we find that formation of tertiary interactions within the P4-P6 domain of the Tetrahymena ribozyme is only 10-fold slower (1 s-1) than formation of tertiary interactions in tRNA (100 ms); (2). Additional experiments using urea have examined the stability of the individual domains, further suggesting that folding of this domain is independent of other regions of the ribozyme. As formation of the native structure requires compaction of the RNA and coordination of multiple Mg2+ ions we anticipate that these fast folding events will be sensitive to environmental conditions such as divalent ion concentration and solvent viscosity. Stopped-flow synchrotron x-ray footprinting provides a method for rapidly probing the structure of RNA folding intermediates and will enable these questions to be answered in the near future for the Tetrahymena ribozyme. These studies provide the "proof of principle and practice" for stopped-flow synchrotron x-ray footprinting that is the foundation for the core and collaborative projects in both nucleic acid and protein footprinting of the Resource.
References and Footnotes
1. T. R. Cech and D. Herschlag, Nucleic Acids and Molecular
Biology 10, 1 (1996); L. Jaeger, F. Michel, E. Westhof, Nucleic Acids
and Molecular Biology 10, 34 (1996).
2. D. E. Draper, Nature Struct. Biol. 3, 397-400 (1996).
3. J. Pan, D. Thirumalai and S. A. Woodson, J. Mol. Biol., in press.
4. V. L. Emerick and S. A. Woodson, Proc. Natl. Acad. Sci. USA 91,
9675 (1994); P. P. Zarrinkar, J. Wang, J. R. Williamson, RNA 2, 564
(1996).
5. B. Sclavi, S. A. Woodson, M. Sullivan, M. R. Chance and M. Brenowitz,
J. Mol. Biol. 266, 144 (1997); M. R. Chance, B. Sclavi, S. A. Woodson,
and M. Brenowitz, Structure 5, 865 (1997).
6. J. H. Cate, A. R. Gooding, E. Podell, K. Zhou, B. L. Golden, C. E. Kundrot,
T. R. Cech and J. A. Doudna, Science 273, 1678 (1996).
7. V. Lehnert, L. Jaeger, F. Michel, E. Westhof, Chemistry & Biol.
3, 993 (1996).
Anjna Dixit and Uma Roy
Department of Biochemistry,
Central Drug Research Institute,
Lucknow 226001, India
For author correspondence. E-Mail: root@cscdri.ren.nic.in; Fax: 91-0522-223405
A formidable worldwide resurgence of tuberculosis is being witnessed in the wake of AIDS (1), and leprosy although on the decline, persists as a major public health problem in developing countries. Dismissal performances of BCG in some populations, as well as of other integral candidate vaccines undergoing human trials (2,3), have underlined the need for molecular characterization of immuno-pathologically important mycobacterial constituents in the quest for better drugs, diagnostic techniques and vaccines. Attention has primarily been focussed on proteins, with a growing realization that the T-cell mediated immune response generated by them (4) could be of a diagnostic or prophylactic value.
In the course of investigations, Mycobacterium habana, a candidate vaccine for mycobacterial infections has been dissected for analyzing its antigenic myriad (5,6,7). A 65 kDa protein of this mycobacterium has been isolated and characterized for its protective and cell-mediated immune responses (8).
In the present study, a molecular biology approach was used to define the species identity of M. habana in a manner that allows reliable comparison to be made with over 30 known mycobacterial species. A 441 bp region present at the N-terminus of 65 kDa gene of M. habana has been PCR amplified and the DNA sequence was analyzed (9). A comparison of the M. habana DNA sequence with those of M. tuberculosis, M. avium, M. paratuberculosis, M. leprae, M. bovis BCG and M. fortuitum revealed a species specific polymorphism, i.e. the presence of nucleotide substitution specific to M. habana. In an alternate approach the restriction pattern of 441 bp product was generated using Hae III and BstE II. (10). It was found to be unique when compared with that of the other mycobacteria for which the restriction patterns for same have been documented in the literature. These results established the species identity of M. habana at the nucleotide level.
References and Footnotes
1. Weiss, R., Science 235, 148-150 (1992).
2. Fine, P.E.M. & Rodriguer, L.C., Lancet 335, 1016-1020 (1990).
3. Convit, J., Sampson, C., Zuniga, N., Smith, P.G., Plata, J., Silva, J.,
Molina, J., Pinardi, M.E., Bloom, B.R. & Salgado, A., Lancet 339,
446-450 (1992).
4. Germain, R.N., Cell, 76, 287-299 (1994).
5. Singh, N.B., Srivastava, A., Gupta, H.P., Sreevatsa, Desikan, K.V., Ind.
J. Lepr., 57, 277-281 (1985).
6. Gupta, H.P., Singh, N.B., Mathur, I.S., & Gupta, S.K., Ind. J.
Ext. Biol. 17, 1190-1193 (1979).
7. Singh, N.B., Mathur, I.S., Gupta, H.P., Srivastava, A. Cur. Sci.,
50, 994-996 (1981).
8. Singh, N.B., Srivastava, K., Malviya, B., Kandpal, H., Srivastava, A.
and Gupta, H.P., Imm. and Cell. Biol. 73, 372-376 (1995).
9. Hance, A., Grand Champ, B., Levy Frebault, V., Lecossier, D., Rauzier,
J., Bocart, D. and Gicquel, B., Mol. Microbiol 3, 843-849 (1989).
10. Telenti, A.F., Marchesi, M., Balz, F., Bally, E.C., Bottger and T. Bodmer,
J. Clin. Microbiol. 31, 175-178 (1993).
D. Dubins*, T. Chalikian and R.B. Macgregor, Jr.
Dept. of Pharmaceutical Sciences,
University of Toronto,
M5S 2S2, Canada
*Author to whom correspondence should be addressed. E-mail: david.dubins@utoronto.ca
Bovine pancreatic ribonuclease A (RNase A) is an excellent model system for examining the volume changes arising from denaturation, as it is a relatively stable and well-studied enzyme (1). We are studying the influence of hydrostatic pressure on the temperature induced conformational transition of RNase A to elucidate the role of water in stabilizing its higher order structure. The comparison of the transition volume of RNase A involved in denaturation with that of RNase A - cytidine monophosphate (CMP) complexes will yield volume changes associated with binding, and give insights about the efficacy of using a stability / pressure dependence relationship to ascertain transition volume.
The transition volume of RNase A was derived from the pressure dependence of the melting temperature (T M), using the Clapeyron equation. Melting curves were collected at constant hydrostatic pressures, ranging from 5 to 200 MPa. We measured an atmospheric transition volume of DeltaV T = -17 ± 3 mL/mol, corresponding well with literature values (2). In comparison, 2 ' cytidine monophosphate (2 ' CMP) bound to RNase A exhibited an increase in overall stability (~ 4°C increase in melting temperature) but resulted in only a modest change of transition volume (DeltaV T = -12 ± 3 mL/mol). This would suggest that 2 ' CMP does not play a large role in the change of hydration of RNase A during the transition. The effect of binding 3 ' CMP were also investigated. The pressure dependence of the T M of RNase A when bound to 3 ' CMP is non-linear. At pressures below 60 MPa, the T M increases with pressure; above 60 MPa, the T M decreases. From the non-linearity of the T M with pressure, we calculated the compressibility of the RNase A-3 ' CMP complex to be DeltaK T ~ 0.03 mL/mol·bar. Additionally, the overall stability of the RNase A-3 ' CMP complex was ~1°C lower than that of the RNase A-2 ' CMP complex, indicating that the 3 ' isomer interacts less strongly than the 2 ' isomer.
Volume differences of binding were calculated, assuming the transition volumes of denaturation and ligand binding were additive. Transition volumes of binding for the RNase A-2 ' CMP complex were negative (DeltaVB = -5 ± 3 mL/mol), indicating that the stability of the complex was decreased with pressure. The RNase A-3 ' CMP complex exhibited a pressure dependent change in transition volume, starting from DeltaVB = -38 ± 5 mL/mol at 5 MPa, and increasing linearly to DeltaVB = 27 ± 5 mL/mol at 200 MPa. The differences in volume binding between the 2 ' and 3 ' CMP species suggests a profound difference in the way the complex is destabilised by pressure. It also demonstrates that our pressure-derived transition volume acquisition method is sensitive to subtle changes in ligand binding to RNase A.
Acknowledgements
This work was supported by a grant from NSERC.
References and Footnotes
1. Raines, R.T., Chem. Rev. 98, 1045-1065 (1998).
2. Prehoda, K.E. Mooberry, E.S. Markley, J.L., Biochemistry 37, 5785-5790
(1998).
Marie L. Estock*1, Kenneth R. Brown2, Dennis R. Winge2 and Thomas
D. Tullius3
1Department of Chemistry,
The Johns Hopkins University,
3400 North Charles Street,
Baltimore, MD 21218
2University of Utah Health Sciences Center,
Salt Lake City, Utah 84132
3Department of Chemistry,
Boston University,
590 Commonwealth Avenue,
Boston, MA 02215
*Author to whom correspondencs should be addressed. Phone: (617) 353-8810;
Fax: (617)353-6466; Email: mestock@chem.bu.edu
Copper homeostasis is very important in all cells. Copper(I) sequestering proteins in yeasts Saccharomyces cerevisiae and Candida glabrata (CUP1 and MTI, MTIIalpha, MTIIbeta respectively) are regulated by Cu(I) induced transcription factors. Both CUP2, from S. cerevisiae, and AMT1, from C. glabrata, bind four Cu(I) ions cooperatively, undergo a conformational change and become active for DNA binding. Both proteins have the same structural features: amino acids 1-40 constitute a zinc(II) binding domain, amino acids 41-110 are the tetracopper domain, and the rest (residues 111-C-terminal) is the transactivation domain. The DNA binding domain contains both the zinc domain and the tetracopper domain. AMT1 can induce transcription of CUP1 and SOD in S. cerevisiae as well as its own gene (AMT1) and metallothioneins (MTI, MTIIalpha, MTIIbeta) in C. glabrata. However, CUP2 can effect the transcription of the S. cerevisiae genes CUP1 and SOD, not the genes in C. glabrata. In previous studies, the full DNA binding domains of the proteins (residues 1-110) have been footprinted and shown to have different footprints in spite of their many similarities. In this study, just the tetracopper domain has been footprinted on its proper binding site to determine if the tetracopper domains might account for the different transcription capabilities of the proteins.
References and Footnotes
1 Buchman, C., Skroch, P., Welch, J., Fogel, S., and Karin,
M., Mol. Cell. Biol. 9, 4091-4095 (1989).
2. Thorvaldsen, J. L., Sewell, A. K., McCowen, C. L., and Winge, D. R.,
J. Biol. Chem. 268, 12512-12518 (1993).
3. Thorvaldsen, J. L., Sewell, A. K., Tanner, A. M., Peltier, J. M., Pickering,
I. J., George, G. N., and Winge, D. R., Biochemistry 33, 9566-9577
(1994).
4. Dobi, A., Dameron, C. T., Hu, S., Hamer, D., and Winge, D. R., J.
Biol. Chem. 270, 10171-10178 (1995).
5. Dixon, W. J., Inouye, C., Karin, M., and Tullius, T. D., JBIC 1,
451-459 (1996).
6. Graden, J. A., Posewitz, M. C., Simon, J. R., George, G. N., Pickering,
I. J., and Winge, D. R., Biochemistry 35, 14583-14589 (1996).
7 Koch, K. A. and Thiele, D. J., Mol. Cell. Biol. 16, 724-734 (1996).
8. Johnson, J. A., Dixon, W. J., and Tullius, T. D., Inorganica Chimia
Acta 242, 233-238 (1996).
Bernd Giese*, Eric Meggers, Stephan Wessely and Martin Spormann
Department of Chemistry,
University of Basel,
St. Johanns Ring 19,
CH-4056 Basel, Switzerland
*Author to whom correspondence should be addressed. Phone: Int-41-61-267-1106;
Fax Int-41-61-267-1105; E-mail: giese@ubaclu.unibas.ch
During the last years the discussion about the mechanism of long range charge transfer through DNA has been centered around the beta-value in equation [1]. This equation describes the fast exponential rate decrease of charge transfer with increasing distance. Depending on the experiment, beta-values of about 1.0 Å-1 or 0.1 Å-1 were obtained (15). Such differences in beta-values have dramatic effects on the rate. Thus, for a charge transfer over 50 Å and a beta-value of 0.1 Å-1, the rate should be nearly twenty orders of magnitude faster than for a beta-value of 1.0 Å-1. It is unlikely that both b-values can be realized for the long range charge transport in DNA. Recently, we have proposed a hopping mechanism for the long range charge transfer (6) in which the charge tunnels between DNA bases of similar redox potentials. In the case that every single hopping step occurs over the same distance, the charge transfer is described by equation [2] where h is a proportionality factor and N is the number of hopping steps. In sharp contrast to equation [1], a weak distance dependence of the rate results from equation [2].
lnk µ beta · Deltar [1]
lnk µ n · ln
N [2]
We have now proven the validity of equation [2] experimentally through charge transfer reactions over 10 to 40 Å in DNA strands (7).
As a consequence, the base sequence plays a decisive role on the hole transfer. But the electron transfer should be less sequence dependent in intact DNA double strands because thymine and cytosine have similar redox potentials. Modifications change the redox potentials of the bases and therefore exert a strong influence on the efficiency of the charge transfer.
References and Footnotes
1. A. M. Brun, A. Harriman, J. Am. Chem. Soc. 116,
1038310393, 1994.
2. F. D. Lewis, T. Wu, Y. Zhang, R. L. Letsinger, S. R. Greenfield, M. R.
Wasielewski, Science 277, 673676, 1997.
3. K. Fukui, K. Tanaka, Angew. Chem. 37, 158161, 1998.
4. E. Meggers, D. Kusch, M. Spichty, U. Wille, B. Giese, Angew. Chem.
37, 460462, 1998.
5. S. O. Kelley, R. E. Holmlin, E. D. A. Stemp, J. K. Barton, J. Am.
Chem. Soc. 119, 98619870, 1997.
6. E. Meggers, M. E. Michel-Beyerle, B. Giese, J. Am. Chem. Soc. 120,
1295012955, 1998.
7. B. Giese, S. Wessely, M. Spormann, U. Lindemann, E. Meggers, M. E. Michel-Beyerle,
Angew. Chem. 38, in press, 1999.
Daniel Ventura1, David King1,
Xian Bin Yang1, Sheela Venkitachalam1, David E. Volk1, Susan
Fennewald2, Allan Brasier3,
Norbert Herzog2, Judy Aronson2,
Bruce Luxon1 and David Gorenstein1*
1Sealy Center for Structural Biology,
2Department of Pathology,
3Sealy Center for Molecular Sciences,
University of Texas Medical Branch,
Galveston, TX, 77555
*Author to whom correspondence should be addressed. Phone 409 747 6801;
Fax 409 747 6850; E-Mail: david@nmr.utmb.edu
Inhibition of NF-k B mediated inflammation could be achieved by selective binding of DNA aptamers to the p65 subunit of NF-k B, thereby reducing the number of inflammation-promoting dimers (p65/p65 and p65/p50) which are free to bind constitutive DNA, while leaving the inflammation-controlling p50/p50 homodimers unaffected. Likewise, inflammation could be promoted by selective binding to the p50/p50 homodimers. However, oligonucleotides are susceptible to nuclease digestion, making them unsuitable for oral or intravenous administration. DNA backbone modifications such as monothioation and dithioation inhibit nuclease digestion, thereby enhancing the therapeutic potential of DNA aptamers. Oligonucleotides possessing high thiophosphate backbone substitutions appear to be "sticker" towards proteins that normal phosphate esters. One needs, therefore, to optimize the total number of thioated phosphates in order to minimize non-specific protein binding while enhancing specific binding to the protein of interest.
Towards this goal, we have used a new thiophosphate combinatorial selection
methodology to isolate high affinity thiophosphate backbone modified duplex
aptamers targeted to NF-k B as well as another transcription factor
involved in the acute phase response, NF-IL6. We have also synthesized a
number of monothio- and dithio-substituted analogs of CK-14, a known binding
site for NF-k B, and have studied both their NMR solution structures
and their binding affinities to NF-k B. The fully monothio-substituted
analog binds equally to the p50 and p65 domains, while several of the selectively
monothioated analogs selectively bind to p65 and c-Rel only. The dithio
hybrid backbone aptamers bind more tightly to p65/p65 (5-15-fold) than to
p50 homodimer. Significantly, the hybrid backbone dithioate, XBY-6 aptamer
is also able to bind to an NF-k B complex present in cell extracts,
while the standard phosphodiester version of the oligonucleotide shows no
NF-k B specific binding in cell extracts. We have therefore succeeded
in synthesizing an oligonucleotide backbone modified agent which for the
first time increases the specific binding of the oligonucleotide to NF-k
B above that of other cellular proteins.
Feng Guo, Deshmukh N. Gopaul and Greg D. Van Duyne
Johnson Research Foundation and Department of Biochemistry and Biophysics,
University of Pennsylvania School of Medicine,
Philadelphia, PA 19104-6089
In Cre/loxP site specific recombination, two 34-base-pair loxP sites are recognized and synapsed by Cre recombinase and undergo two serial single strand exchanges to yield recombination products. Dramatic DNA distortion and formation of a Holliday junction intermediate are involved in this multi-step reaction. The structure of Cre covalently linked to a suicide loxP substrate showed a planar tetrameric architecture in which each Cre molecule is bound to a loxP half site. The structures of complexes corresponding to the other two major stages of the reaction, namely a Cre mutant R173K bound to a duplex DNA (loxS) and the Holliday junction DNA, are presented here.
The duplex complex structure reveals that the centers of two loxP sites are brought into close proximity. Each loxS duplex is bent severely as a result of a single kink between two neighboring base pairs at one end of the central crossover region. This kink contains a large tilt angle, requiring that the DNA backbone on one strand adopt a more extended conformation than the other. The energy stored in the bending and the close contact could serve as the driving force for the strand exchange after the cleavage of DNA backbone. The asymmetry and direction of the bend is linked to the activation of only one of the two DNA strands for cleavage at the start of the recombination reaction.
A new high-resolution structure of a Cre-bound wild type loxP Holliday
junction was determined at 2.0 Å resolution. The Holliday junction
is two-fold symmetric with the four half-site arms nearly coplanar. The
angles between the neighboring arms are 101° and 75° respectively.
There are two unique DNA strands in the junction, which we refer to as continuous
and crossing. The backbone conformation of the crossing strand is of particular
interest because the branch point phosphates are flipped pointing toward
the center of the complex. This reaction intermediate structure, together
with others mentioned, show in detail what the synaptic complex looks like
during each step of the reaction and suggest how isomerization of the Holliday
junction complex plays a role in regulating the cleavage events.
Avital Bareket-Samish, Ilana Cohen and Tali E. Haran
Department of Biology,
Technion, Technion City,
Haifa 32000, Israel
Phone: 972-4-8293767; Fax: 972-4-8225153; E-mail: bareket@tx.technion.ac.il;
bitali@tx.technion.ac.il
The TATA box binding protein (TBP) is an essential protein for transcription initiation by all three eukaryotic RNA polymerases. TBP binds the TATA box present at many RNA polymerase II promoters and thus initiates the assembly of the preinitiation complex (reviewed in 1). TBP recognizes its target sites, mostly via indirect readout (recognition of intrinsic DNA structure, or its ability to be deformed upon complex formation). To explore the particular indirect readout mechanism used by TBP to interact differentially with different TATA boxes, we studied the interaction of TBP with a set of TATA elements, derived from the Adenovirus major late promoter (AdMLP).
We find that the interaction of TBP with its binding sites is influenced by sequence signals both at the level of individual base-pair step and the entire binding site. In particular, we observe a correlation of the kinetic stability of TBP/TATA-box complexes to structural characteristics of base-pair steps 4/5 and 7/8, as well as to the overall mechanical rigidity of TATA boxes. However, TBP-induced DNA bending is not correlated well to the stability of the complexes. On the basis of these results we suggest a possible way for TBP to read the sequence of bases at the minor groove and thus to bind differentially to different TATA boxes.
The dissociation rate of TBP/TATA-box complexes is influenced by the sequences flanking the core TATA box although they do not contact TBP. We show that this effect is differential: the flanking sequences modulate the dissociation rate of a complex of TBP with a TATA box which contains an alternating (T-A)n tract, whereas they do not affect the dissociation rate of a complex of TBP with an A4-tract-containing TATA box. We suggest that the ability of the flanking sequences to differentially modulate complex stability may be a novel mechanism for fine regulation of transcription.
It was demonstrated that TBP remains stably bound on the TATA box after RNA polymerase II and the associated factors clear the promoter (2). Moreover, the in vitro studies of Yean and Gralla (3) showed that mutating the consensus TATA box decreases transcription reinitiation rate. Therefore, increased kinetic stability of TBP/TATA-box complexes may increase the transcriptional activity of the corresponding promoter, by increasing the number of transcription reinitiation events. Indeed, the transcriptional activity of two of our TATA elements, as determined in the study of Starr et al. (4), correlates with the kinetic stability of the corresponding complexes, observed in our study.
References and Footnotes
1. Burley, S. K., and Roeder, R. G., Annu. Rev. Biochem.
65, 769-799 (1996).
2. Zawel, L., Kumar, K. P., and Reinberg, D., Genes Dev. 9,1479-1490
(1995).
3. Yean, D., and Gralla, J. Moll. Cell. Biol. 17, 3809-3816 (1997).
4. Starr, D. B., Hoopes, B. C., and Hawly D. K., J. Mol. Biol. 250,
434-446 (1995).
Lou-sing Kan*1,2, Ming-hong Hou2, Wei-chen Lin1 and Shwu-Bin
Lin3
1Institute of Academia Sinica,
Institute of Biochemistry,
2National Chung Hsing University,
3The Graduate School of Clinical Medicine,
National Taiwan University,
Taipei, Taiwan
*Author to whom correspondence should be addressed. Phone: 886-22-788-4184;
Fax: 886-22-783-1237; E-mail: LSKAN@CHEM.SINICA.EDU.TW
The association and dissociation rate constants of oligonucleotide duplex formation were studied in the presence of the various cations with a BiaCore 2000 in tris buffer, pH 7. The cations were chosen from those found in cellular solutions (Na+, Mg++, spermidine and spermine). The synthetic oligonucleotide duplexes contain either (1) 14 AT base pairs with one GC in position 2, 8, and 14 (designated as P2, P8, and P14, respectively); or (2) 15 AT base pairs and one bulged G base at position 8 (B8) and 13 AT, two GC base pairs with one bulged G base also at position 8 (B8'); or (3) 14 AT base pairs with one AG mismatched base pair at position 8 (M8). Results showed that the association rate constants are fast, and reaching a ceiling of 106 M-1 sec-1 in all oligonucleotides and in the presence of any of the cations described above. This indicates that the formation of duplex of these oligonucleotides is instantaneous. On the other hand, the dissociating rate constants have the following order in the presence of various cations:
M8> B8, B8 ' > P2, P8, P14
However, the differences diminish gradually in the following order: tris
buffer > 0.1 M NaCl > 0.02 M MgCl > 0.005 M spermidine > 0.005
M spermine. We conclude that the presence of cations, especially the spermine,
can compensate the energy discrepancy caused by lesions (e.g. bulged loop
or mismatched base pairs) in the duplex formation. The same conclusion was
reached from the UV melting temperature study. Our study may provide a basis
in explaining the DNA polymorphism phenomenon, since the spermine is ubiquitous
in living cells. (Supported by National Science Council and Academia Sinica,
Taipei, Taiwan)
Maciej Maciejczyk1,2*, Witold R. Rudnicki2 and Bogdan Lesyng2,3
1Insitute of Biochemistry and Biophysics,
Polish Academy of Sciences,
Pawinskiego 5A,
02-106 Warsaw, Poland
2Interdyscyplinary Centre for Mathematical and
Computational Modeling,
Pawinskiego 5A,
02-106 Warsaw, Poland
3Department of Biophysics,
Warsaw University,
Zwirki i Wigury 93,
02-089 Warsaw, Poland
*Author to whom correspondence should be addressed. Phone: +48 22 658 46
83, Fax: +48 22 658 46 83, E-mail: maciej@icm.edu.pl
For past several years we have been developing a mezoscopic model of a double-helical DNA based on Quaternion and Lagrangian Dynamics (1-3). In the current study we present an effective Potential of Mean Force designed for this model (for PMF see (4,5)). The DNA model is built from pseudoatoms as well as rigid and pseudo-elastic bodies described by a limited number of selected cartesian and internal degrees of freedom. Phosphate groups, deoxyribose rings and bases are represented by pseudoparticles, some of them with internal degrees of freedom.
PMF has a functional form similar to potentials used routinely in atomic scale simulations. It is defined as a sum of effective bonded and long-range contributions. The bonded part is obtained from microscopic simulations and consists of bond stretching, angle bending and dihedral torsion potentials. The long-range part is approximated with effective non-bonded interactions. Each bonded potential has a harmonic form with an equilibrium position parameterized with the pseudorotation phase of the nearest deoxyribose ring. Harmonic potentials were fitted to the numerical free energy surfaces. Almost 100 free energy surfaces, each depending on a conformational variable (pseudobond length, angle or dihedral angle) and the pseudorotation phase of the nearest deoxribose ring, were computed. The numerical free energy surfaces were obtained from probability distributions derived from a 1.5 ns microscopic MD simulation of the d(GC)9 double helical DNA molecule. The microscopic simulation was performed in the canonical ensemble in the room temperature and with the distance dependent dielectric function, e = 4*r, using standard Amber parameterization as implemented in the Discover 2.97 program from MSI. An umbrella sampling method was used to simulate transitions between A and B forms.
References and Footnotes
1. Rudnicki, W.R., Lesyng, B. and Harvey, S.C., Biopolymers
34, 383 (1994).
2. Rudnicki, W.R, and Lesyng, B., Molecular Simulations 19, 247-266
(1997).
3. Rudnicki, W.R., PhD thesis, (1997).
4. Kuczera, K., J. Comp. Chem. 17, 1726-49 (1996).
5. McCammon, J.A., Curr. Opinion Struct. Biol. 1, 196-200 (1991).
Ishita Mukerji1 and Alison P. Williams2
1Department of Molecular Biology and Biochemistry,
2Department of Chemistry,
Molecular Biophysics Program,
Wesleyan University,
Middletown, CT 06459
Previous experimental work has demonstrated that the presence of AnTn and An tracts in DNA molecules has a significant impact on structural properties due to their intrinsic curvature. Double helical sequences containing these tracts exhibit an alternative to the standard B DNA conformation known as B'. This anomalous conformation has distinct structural features including a narrow minor groove. Crystallographic studies indicate that an unusual three-centered hydrogen bond can exist between successive AT base pairs in these tracts. Gel mobility of duplexes containing these tracts is significantly retarded indicative of a bent conformation. Premelting conformational transitions have been observed in molecules containing An tracts and are thought to indicate a conformational transition from B' to B form.
We report spectroscopic studies of three related DNA dodecamers, d(CGCAAATTTGCT)2 (abbreviated A3T3),
d(CGCTTTAAAGCG)2 (abbreviated T3A3) and d(CGCATATATGCG)2 (abbreviated
(AT)3). Previous NMR studies of the A3T3 sequence demonstrated that the central A residue exhibits
unusual thermodynamic properties. The above sequences are studied to determine
if 3-centered H-bonds exist in the premelting state and to gauge the impact
of these putative 3-centered H-bonds on the structural and thermodynamic
properties of the A3T3 duplex.
UV optical melting studies probe the effect of these H-bonds on global duplex
stability. Previous UV resonance Raman (UVRR) studies have detected the
presence of bifurcated H-bonds in the poly dA-poly dT homopolymer. These
two techniques are thus employed to monitor the structural and thermodynamic
changes in these dodecamers as a function of temperature. UVRR results indicate
that despite the sequence similarities, the molecules have dramatically
different thermal profiles. Moreover, the concerted shifts of adenine exocyclic
NH2 and thymine C4=O vibrational modes are suggestive that the A3T3 dodecamer forms 3-centered hydrogen
bonds at low temperature, while the (AT)3 and T3A3 dodecamers do not. From the frequency
shift of the thymine C4=O vibrational mode with temperature, the enthalpy
of the 3-centered H-bond is estimated to be approximately 1.8 kJ/mol or
0.43 kcal/mol. Thermodynamic analyses based upon UV melting profiles will
also be presented. The effect of cations on the stability of the complexes
and their conformation will also be addressed.
Maria C. Nagan, Stephanie S. Kerimo, Christopher J. Cramer and Karin
Musier-Forsyth*
Department of Chemistry and Supercomputer Institute,
University of Minnesota,
207 Pleasant Street SE,
Minneapolis, MN 55455
*Author to whom correspondence should be addressed. Phone: 612-624-6000;
E-mail: musier@chem.umn.edu
Accurate translation of the genetic code in protein synthesis depends on the capability of aminoacyl tRNA-synthetases to recognize their cognate tRNAs. In the alanine system, the primary recognition elements are contained in the acceptor stem. As a result, small RNA microhelices and duplexes derived from this domain are specific substrates for aminoacylation with alanine (1,2). In particular, the G3:U70 wobble base pair is the major determinant (3,4). Molecular dynamics studies of the microhelixAla and 3:70 variants exhibiting different aminoacylation activities were carried out for 2 ns per microhelix. We were interested in the propensity of water molecules to be tightly bound around the 3:70 base pair. In addition, the influence of base-pair substitution on helical structure and water binding was examined. Water density maps of snapshots obtained every 1 ps over the simulation predict that the G:U base pair in microhelixAla tightly binds a water molecule, as previously observed in other G:U-containing RNA helices (5-10). Our simulations find that both the minor groove 2-amino group of G3 and the 2 '-hydroxyl of U70 are crucial for water binding. Removal of either of these functional groups correlates with an observed reduction in alanine groove water may play a role as a specific recognition element or through its release, provide a free energy force for favorable synthetase binding.
References and Foonotes
1. Francklyn, C. and Schimmel, P., Nature 337, 478-481
(1989).
2. 2. Musier-Forsyth, K., Usman, N., Scarings, S., Doudna, J., Green, R.
and Schimmel, P., Science 253, 784-786 (1991).
3. McCain, W. H. and Foss, K., Science 240, 793-795 (1988).
4. Hou, Y.-M. and Schimmel, P., Nature 333, 140-145 (1988).
5. Auffinger, P. and Westhoff, E., J. Mol. Biol. 269, 326-341 (1997).
6. Biswas, R., Wahl, M. C., Ban, C. and Sunderalingam, J., J. Mol. Biol.
267, 1149 (1997).
7. Betzl, C., Lorenz, S., Furste, J. P., Bald, R., Zhang, M., Schnieder,
T. R., Wilson, K. S. and Erdmann, V. A., FEBS Lett. 351, 159 (1994).
8. Cruse, W. B. T., Saludjian, P., Biala, E., Strazweski, P., Prangé,
T. and Kennard, O., Biochemistry 91, 4160 (1994).
9. Holbrook, S. R., Cheong, C., Tinoco, I. and Kim, S.-H., Nature 353,
579-581 (1991).
10. Mueller, U., Schubel, H., Sprinzl, M. and Heinemann, U.
Sandeep Kumar1, Buyong Ma2,
Chung Jung Tsai1 and Ruth Nussinov1,3
1IRSP, SAIC Frederick,
LECB, NCI-FCRDC,
Frederick, MD 21702
2LECB, NCI - FCRDC,
Frederick, MD 21702
3Medical School,
Tel Aviv University, Israel
We have compared three dimensional structures of thermophilic and mesophilic proteins in 18 non-redundant families derived from Protein Data Bank (PDB). The thermophilic and the mesophilic proteins have similar hydrophobicities, compactness, oligomeric states, main chain hydrogen bonds, and main chain side chain hydrogen bonds. Frequencies of salt bridges increase in the thermophilic proteins and may be correlated with their melting temperatures. However, the salt bridges have been shown to destabilize the protein structure. Why then several thermophilic proteins contain a greater number of salt bridges and their networks as compared to their mesophilic homologues? We have compared the salt bridges and their networks in glutamate dehydrogenase (GDH) from hyperthermophilic archaeon Pyrococcus furiosus and mesophile Clostridium symbiosum. Our results show that several of the additional salt bridges within each monomer of the hyperthermophilic GDH are formed around the active site of the protein. Salt bridges in Pyrococcus furiosus GDH stabilize the protein. In contrast, the salt bridges in Clostridium symbiosum GDH are only marginally stabilizing. These results indicate that salt bridges and their networks may have important roles in maintaining protein structure, at least in critical regions like neighborhood of active site, at high temperatures.
Arthur Pardi, Emilia Mollova, Paul Hanson and Mark R. Hansen
Department of Chemistry and Biochemistry,
Campus Box 215,
University of Colorado,
Boulder, Boulder, CO 80309-0215
Filamentous phage represent a simple versatile method for generating partially ordered macromolecules in solution. The phage allow tunable degrees of alignment of the macromolecule over a large temperature range (5 60°C) and a wide variety of conditions. We have applied this technique to DNA and RNAs and a number of proteins. The negatively charged phage are ideal alignment tools for negatively charged molecules such as nucleic acids where these phage-macromolecule solutions are stable indefinitely. We will report on measurements of 1H-15N, 1H-13C dipolar couplings in tRNA and the Dickerson dodecamer for generating angle constraints for structure refinements. Examples will be also be presented where dipolar couplings are able to generate long range structural and dynamic information that is unavailable by standard NMR techniques. (Supported by NIH grant AI33098).
Ekaterina Protozanova* and Robert B. Macgregor, Jr.
Department of Pharmaceutical Sciences,
University of Toronto,
Toronto, Ontario M5S 2S2, Canada
*Author to whom correspondence should be addressed. E-mail: e.protozanova@utoronto.ca
Dynamic light scattering has been used to study the formation of DNA frayed wires multistranded complexes arising from the self-association of oligonucleotides possessing long terminal tracks of guanines, e.g. d(A15G15). DNA frayed wires originate from guanine-guanine interactions resulting in the formation of a guanine stem. Non-guanine portions of the parent strands are displaced from the stem and form single stranded arms. The two domains within the frayed wire are structurally and thermodynamically independent.
Since each sample contains species composed of various numbers of strands parent monomers and products of the association of the monomers the sample presents a polydisperse population of complexes differing in molecular weights and sizes. Underlying distributions of decay times deduced from the autocorrelation functions are bimodal. The slow mode, observed before for the solutions of short DNA and other polyelectrolytes, is attributed to gelation or clustering due to electrostatic or other long-range interactions. Its amplitude is favored by low ionic strength and high DNA concentration.
The fast mode corresponds to the collective diffusion of all types of
aggregates present in the sample. We use the mean of the fast mode to estimate
the apparent size associated with the whole population of frayed wires.
The calculated hydrodynamic diameter of frayed wires ranges from 7 to 28
nm depending on the history of the sample and salt conditions. There are
two factors affecting the overall amplitude and the mean of the fast mode
the presence of magnesium and heating. Magnesium containing samples
exposed to high-temperature incubation exhibit a notable increase in the
apparent size. The same trend is observed for the baseline intensity of
the autocorrelation functions the intensity increases with time and
temperature. As resolved by agarose gels, frayed wires tend to grow into
larger species when incubated at higher temperature. These experiments show
that both, the rate and the extent, of the aggregate formation increase
with temperature, consistent with the positive activation energy for the
forward reaction. This behavior, atypical for nucleic acid helix formation,
underscores the complexity of the mechanism of the frayed wire formation.
Evidently aggregation is hindered by an unspecified secondary structure
within the arms and/or the stem of the complex. These interactions are disrupted
upon heating and polymerization proceeds. Furthermore, the addition of another
strand to the frayed wire possibly requires the "opening" of the
existing structure to make it more accessible for the newly introduced strand.
This process is also favored by high temperature contributing to higher
rates of the frayed wire formation with heating.
A. R. Srinivasan1, Michael Weiss2 and Wilma K. Olson1
1Department of Chemistry,,
Wright-Rieman Laboratories,
Rutgers, the State University of New Jersey,
Piscataway, NJ
2Center for Molecular Oncology,
University of Chicago,
Chicago, IL
The sequence-dependent flexibility of short DNA double helices is probed
by a thorough examination of the spatial distributions of flexible alkane
chains covalently linked to the 5 '-phosphate groups. We have generated
a large number of alkane chains (106-107
configurations at each 5 '-end) by a Metropolis-Monte Carlo chain generation
scheme and then computed the radial distribution, R (r 12), of the separation r 12 between
alkane termini of the DNA. This method has been applied to naturally straight
and a variety of globally bent DNA structures and provides a criterion for
assessing fluorescence energy transfer measurements. Preliminary results
demonstrate the non-Gaussian nature of the radial distribution functions
and the need for a modified interpretation of experimental measurements.
(Supported by USPHS research grant GM20861).
Eva de Alba1, Luc de Vries2,
Marilyn G. Farquhar2 and Nico Tjandra1
1Laboratory of Biophysical Chemistry,
NHLBI, NIH,
Bethesda, MD 20892
2Division of Cellular and Molecular Medicine,
The Center for Molecular Genetics and Department of Pathology,
UCSD,
La Jolla, CA 92093
Human Ga Interacting Protein (GAIP) is a member of a family of proteins which are involved in the regulation of G protein signaling (RGS). The RGS proteins bind the active conformation of the G protein a subunit, which is the GTP bound form. It is well known that a conserved switch surface of Galpha is the only interacting site for its effectors. The surface alone however does not explain the high specificity of the G protein a subunit and effector interaction. This study is aimed to look at specificity of the interaction by detailed structure and dynamics study of GAIP using solution NMR. GAIP is localized in the Golgi and bind specifically to the Galphai3. The structure of GAIP was determined with the aid of residual dipolar couplings. The dipolar couplings are reintroduced by a slight alignment of the protein in solution. This was achieved by addition of liquid crystal medium into the system. Two different liquid crystalline media were used in the case of GAIP, DMPC:DHPC lipid mixture and fd bacteriophage. The availability of the dipolar coupling data allow a detailed comparison between GAIP which represents the free RGS protein and the RGS4 bound to the transition state form of Galphai1. In the same manner the dipolar coupling data provide a confirmation on the aggregation state of GAIP. The combination of NMR relaxation data and dipolar coupling were used to calculate the rotational diffusion parameters of GAIP.
Tom Tullius*
Department of Chemistry,
Boston University,
590 Commonwealth Avenue,
Boston MA 02215
*For author correspondence. Phone: (617) 353-2482; Fax: (617) 353-3535;
E-mail: tullius@bu.edu
Low-valent metals like iron(II) react with hydrogen peroxide to form the hydroxyl radical, via the Fenton reaction. My group introduced the use of this reaction as a source of the hydroxyl radical for footprinting DNA. Recently we have become interested in (1) which deoxyribose hydrogen atoms are abstracted by the hydroxyl radical, and (2) what are the structural consequences to DNA of the lesion introduced by the radical. Our work also has implications for radiation chemistry, since the agent responsible for much of the damage induced in DNA by ionizing radiation is the hydroxyl radical.
We have used deuterium kinetic isotope effect experiments to show that the solvent-accessible surface area of a deoxyribose hydrogen atom governs its reactivity toward the hydroxyl radical. The 5' hydrogen atoms participate in more than half of the hydrogen abstraction events.
To study the struuctural consequences of the lesion introduced into DNA by the hydroxyl radical, we developed a new two-dimensional gel electrophoresis experiment to show that the single-nucleoside gap that is produced by the hydroxyl radical is a site of anisotropic bending or flexibility.
Feng Guo, Deshmukh N. Gopaul and Gregory D. Van Duyne
Department of Biochemistry and Biophysics,
University of Pennsylvania School of Medicine,
3700 Hamilton Walk,
Philadelphia, PA 19104-6089
Cre recombinase is a member of a large family of site-specific recombination enzymes that performs a cut-and-paste operation between two specific DNA sequences. Two Cre proteins bind specifically to a single 34-base-pair "loxP" site and mediate the association of two such Cre-DNA complexes through protein-protein interactions to form a recombination synapse. Two of the four DNA strands in this higher order complex are cleaved by the recombinases to form covalent 3 '-phosphotyrosine linkages between protein and DNA, while releasing free 5 '-hydroxyl groups. The 5 '-hydroxyl groups then attack the phosphotyrosine linkage on the opposite DNA substrate, resulting in an exchange of one pair of strands and formation of a Holliday junction intermediate. The Holliday junction is then a substrate for cleavage and exchange of the remaining pair of DNA strands to form recombinant products.
Our goal has been to understand the mechanism of this complex reaction
by trapping and characterizing the three-dimensional structures of each
of the reaction intermediates. This work has led to high resolution crystallographic
models of (i) the initial synaptic complex, (ii) the covalent Cre-DNA intermediate,
and (iii) the Holliday junction intermediate. The Cre-loxP system appears
to function by creating at the outset a protein-DNA architecture that resembles
the Holliday junction intermediate that is eventually formed. The "arms"
of the loxP sites are initially bent by about 90° in the synaptic complex,
forming a nearly planar arrangement that is held fixed, while cleavage and
strand exchange occur in the central region between the arms. The recombination
pathway contains two symmetrical halves, each of which uses this Holliday
junction-like architectural framework to mediate the cleavage and ligation
steps. The two halves are linked by a subtle isomerization of the Holliday
intermediate that switches the roles of the recombinase subunits and the
DNA strands in a manner that resembles crossover isomerization of Holliday
junctions in homologous recombination.
Kymberley Vickery1,2*, Robert Vagg1, Per Lincoln2, Bengt Nordén2 and Magdalena Eriksson2
1 Department of Chemistry,
Macquarie University,
NSW 2109, Australia
2 Department of Physical Chemistry,
Chalmers University of Technology,
S-412 96 Gothenburg, Swede
*Author to whom correspondence should be addressed. Phone: +612 9850 8296;
Fax: +612 9850 8313; E-mail: kvickery@alchemist.chem.mq.edu.au
The rigid and stereochemically well-defined framework of octahedral transition metal complexes makes them attractive as candidates for sequence selective DNA-binding agents and for molecular probes of local DNA structure (1). The capabilities of ternary cations of general form [Ru(tetradentate)(bidentate)]2+ to function as stereo- and enantiodiscriminatory intercalating probes of DNA structures have been investigated for some time (2). In this study, we have employed NMR techniques to investigate the interactions between the oligonucleotide duplex d(CGCGATCGCG)2 and the L and D diastereomers of cis-b-[Ru(RR-picchxn)(phen)]2+ (shown). This also allows assessment of 1,10-phenanthroline's capacity to intercalate DNA. Flow linear dichroism (LD) spectroscopic studies (3) also have been carried out to indicate the orientation of each probe with respect to the helical axis of calf thymus DNA.
L-cis-beta-[Ru(RR-picchxn)(phen)]2+ Delta-cis-beta-[Ru(RR-picchxn)(phen)]2+
Probe:oligo 1D NMR titrations, NOESY and WATERGATE experiments have been conducted on the diastereomeric probes with the oligonucleotide, and chemical shifts for proton resonances of the nucleic acid and the interacting probe were monitored. Each interaction occurs in the kinetically fast exchange regime. The results indicate that both chelate probes bind in the minor groove. The binding selectivities of the two diastereomers also are found to be significantly different. A detailed structural interpretation of both sets of spectroscopic studies will be presented.
References and Footnotes
1 P. Lincoln and B. Nordén, J. Phys. Chem. 102,
9583-9594, 1998.
2. K.A. Vickery, A.M. Bonin, P.A. Williams and R.S. Vagg, in Structure,
Motion, Interaction and Expression of Biological Macromolecules, H.S.
Sarma and M.H. Sarma (Editors), Adenine Press, New York, U.S.A. 1, 195-206,
1998.
3. P. Lincoln, B. Nordén, P. Williams, R. Vagg and K. Vickery, XXXIII
Intnl. Conf. Coord. Chem., Firenze, Italy, 325, 1998.
L. Zhu1, J. Wilken2,
G. Chan1, L. Milos1,
S. B. Kent2 and M. A. Weiss1
1The University of Chicago,
Center for Molecular Oncology,
924 East 57th Street,
Chicago, IL 60637
Phone: (773) 702-4652
2Gryphon Sciences,
250 East Grand Avenue,
South San Francisco, California, 94080
Phone: (650) 952-7714
A universal downstream pathway of sex determination is proposed to employ a metazoan DNA-binding motif shared by the Doublesex protein of D. melanogaster, the Mab 3 protein of C. elegans, and the mammalian homolog DMT1. This motif regulates sex-specific gene expression in D. melanogaster and C. elegans: its deletion in humans causes 46, XY gonadal dysgenesis with sex reversal (the 9p syndrome). Here, we present the solution structure of the Drosophila DNA-binding domain and associated biochemical studies of its DNA-binding properties. The structure reveals a novel minor-groove-binding Zn module unrelated to previous structures. Surprisingly, a DNA-recognition a-helix binds in the minor groove without significant DNA bending. Implications for sex reversal are discussed.
Alison P. Williams
Department of Chemistry,
Molecular Biophysics Program,
Wesleyan University,
Middletown, CT 06459
Many biological processes demonstrate a marked dependence on salt concentration. One example is the marked dependence of physical properties of nucleic acids on salt. Modified polyelectrolyte theories predict that the charge density of a short oligonucleotide decreases dramatically at the ends of the molecule relative to the internal regions of a double helix, resulting in reduced counterion association. This uneven distribution of charge also impacts on the interactions of shorter sequences with the surrounding environment. There are conflicting experimental results in the literature as to the extent of this effect. The Donnan coefficient and preferential interaction parameter quantify electrostatic nonideality in ternary solutions and can be used to ascertain the thermodynamic consequences of counterion association with nucleic acids. We have mapped the charge density of single and double stranded oligonucleotides by measuring the Donnan coefficient and the preferential interaction parameter. Equilibrium dialysis oligonucleotides of varying length was performed as a function of oliognucleotide concentration. Na+ and Cl- concentrations on both sides of the dialysis membrane were determined using capillary electrophoresis. The results are compared with previous measurements of ion release upone duplex melting. The thermodynamic behavior of these molecules is interpreted using current models of charged interactions of oligonucleotide double helices.