Manju Bansal* and Shibasish Chowdhury
Molecular Biophysics Unit,
Indian Institute of Science,
Bangalore - 560012, India
*Author to whom correspondence should be addressed. E-mail: mb@mbu.iisc.ernet.in
Experimental and theoretical studies have clearly demonstrated that for both DNA and RNA, guanine rich sequences can form four stranded helical structures in the presence of cation which is believed to play a crucial role in quadruplex stability. In order to understand the effect of different size cations on the tetrad geometry and the stability of quadruplex, we have performed, nanosecond-scale molecular dynamics simulations on d(G)7 parallel quadruplex in presence of coordinated Na+ or K+ ions and also, without any coordinated ion. All simulations are carried out using the AMBER force field and the electrostatic interactions are calculated using the Particle Mesh Ewald (PME) method. The total energy for the quadruplex with coordinated Na+ ions and solvent is marginally better than the quadruplex with coordinated K+ ions, while the quadruplex without any coordinated ion has lowest stability. However, if we exclude the intrinsic solvent energy, then the quadruplex with coordinated K+ and Na+ ion have comparable energy. Base stacking energy is most favourable in the quadruplex, without any coordinated ion and this trend is also reflected in the intrinsic quadruplex energy. However, inclusion of DNA - coordinated ion interaction energies for K+ and Na+ ion quadruplexes, make these quadruplexes more stable than that without any coordinated ion. Root mean square deviation between K+ and Na+ ion coordinated quadruplex is less in comparison to that between either of these two quadruplexes and that without coordinated ion. Strong interactions between guanine O6 atom and coordinated ion make the quadruplex compact and rigid. Analysis of the M.D. structures reveals that there is a significant difference in base tetrad geometry and hydrogen bond patterns during the three simulations, with bifurcated three center hydrogen bonds being favoured over the initial eight hydrogen bonded G-tetrads. Phosphate cylinder radius in quadruplex follows an ascending order, with Na+ ion quadruplex being most compact, followed by K+ ion quadruplex and the quadruplex without any coordinated ion being largest. These MD results provide details about the structural features of parallel G-quadruplex and the effect of intercalated cation on the stability of the guanine rich quadruplex structures.
Anke Gelbin, Dennis Sprous and David L. Beveridge
Wesleyan University,
Department of Chemistry,
Middletown, CT 06459
All-atom MD simulations of about 2ns have been used to study the thermodynamics of ligand binding to DNA. Minor-groove binding drug-DNA complexes d(CGCGAATTCGCG)/Netropsin, d(CGCGAATTCGCG)/monolexitropsin and the intercalating Daunomycin/d(CGCGATCG) complex were chosen as case studies. The studies are built upon the free energy analysis of the EcoRI-DNA complex developed by Jayaram et al. (1) Here, a thermodynamic cycle of seven distinct steps with well-defined energetic components is considered to determine the standard free energy of binding. The model involves explicit accounts of the energetics of the structural adaptation of the DNA and the drug, electrostatic and nonelectrostatic components of solvation, counterion effects and entropic aspects upon complexation. Internal energetics estimating the bonded and non-bonded interaction terms as well as the electrostatic contribution to binding are obtained from the Cornell et al. force field of the AMBER 4.1 modeling package. Solvation energy components were estimated from the Generalized Born theory and solvent accessibility measures. Changes in rotational and translational entropy upon complexation are calculated using classical statistical mechanics. The overall results of the calculations illustrate that the hydrophobic and van der Waals interactions are favorable to complexation while entropic effects and structural adaptation are unfavorable. Electrostatics and small ion effects also favor complex formation, in contrast with the result obtained for the EcoRI-DNA binding case.
References and Footnotes
1. B. Jayaram, K.J. McConnell and D.L. Beveridge, "Free
Energy Analysis of Protein-DNA Binding: The EcoRI Endonuclease-DNA Complex"
J. Comp. Phys. In Press (1999).
Yate-Ching Yuan1, Robert H. Whitson2, Qin Liu1, Keiichi Itakura2 and Yuan Chen*1
1Divisions of Immunology and 2Biology,
Beckman Research Institute of the City of Hope,
1450 East Duarte Road,
Duarte, CA 91010
A novel class of homologous DNA-binding domains has been established from at least sixteen recently identified DNA-binding proteins. The three-dimensional structure of one of these domains, Mrf-2, has been solved using NMR methods. This structure is significantly different from that of known DNA-binding domains. The mechanism of DNA recognition by this motif has been suggested based on conserved residues, surface electrostatic potentials and chemical shift changes. This new DNA-binding motif shares structural homology with T4 RNase H, E. coli endonuclease III and Bacillus DNA polymerase I. The structural homology suggests a mechanism for substrate recognition by these DNA replication and repair enzymes.
References and Footnotes
1. Yuan, Y.-C., Whitson, R., Liu, Q., Itakura, K., Chen
Y., Nature Structural Biology 5, 959-964.
2. Yuan, Y.-C., Whitson, R., Liu, Q., Itakura, K., Chen Y., J. Biomol.
NMR, 11, 459-460.
Thang Kien Chiu and Richard E. Dickerson
Dept. of Chemistry and Biochemistry,
University of California at Los Angeles,
405 Hilgard Avenue, Los Angeles,
CA 90095
We have solved the 1A crystal structures of the B-DNA decamers CCAACGTTGG and CCAGCGCTGG crystallized with either magnesium or calcium cations, and the 1.43A crystal structure of the A6/A16 dithiobispropane-crosslinked Dickerson- Drew dodecamer CGCGAATTCGCG crystallized with magnesium cation (X-link). From these crystal structures, we will examine the similarities and differences of magnesium versus calcium cation binding to B-DNA on (1) groove hydration, (2) groove width, (3) crystal packing and (4) divalent cation-induced DNA bending. To examine the issue of monovalent cation binding to B-DNA, we compare the crystal structures of our sodium-free and potassium-free X-link with the previously published crystal structures of the Dickerson-Drew dodecamer crystallized with either sodium or potassium cations added (Xiuqi et al., Biochemistry, 1998). We do this by examining (1) the chemical environment of the groove, (2) the coordination geometry of each solvent atom, (3) difference maps, and (4) the behavior of atomic temperature factors and valence values of solvent atoms of the minor groove versus all solvent atoms present.
Ho Leung Ng, Mary L. Kopka and Richard E. Dickerson
DOE-Molecular Biology Institute and Dept. of Chemistry and Biochemistry,
University of California, Los Angeles
Los Angeles, CA 90095
We report the crystal structure of the dodecamer CATGGGCCCATG. Xray diffraction data were collected to 1.3 angstroms, and phases were provided by isomorphous replacement and anomalous scattering data with an iodinated derivative. In contrast to the decamer CATGGCCATG which adopts the B conformation, CATGGGCCCATG, which differs by only an additional G-C step, adopts a conformation intermediate between the canonical A-DNA and B-DNA geometries. Under a wide range of conditions, CATGGGCCCATG crystallizes in this form, and we have been unable to find conditions which yield B form crystals.
The dodecamer has a normal vector plot typical of B-DNA. The groove width, helical rise, inclination, and X-displacement have values intermediate between those found in A and B DNA. The helical twist is typical of A-DNA as are the pucker of most of the sugars, being C3'-endo. The packing of the duplexes in the crystal, with the end of one duplex filling the minor groove of another duplex, is also typical of A-DNA crystals.
The capture of a stable A/B-DNA intermediate in the crystal form suggests that the A and B states of DNA are not as discrete as believed. Interconversion between A-DNA and B-DNA by G-tracts may be a biologically significant phenomenon. Like A-tracts, G-tracts also demonstrate unique structural properties, but whereas A-tracts are rigid, G-tracts are surprisingly deformable.
Wojciech Bal1*, Jacek Wojcik2, Maciej Maciejczyk2 and
Kazimierz S. Kasprzak3
1Faculty of Chemistry,
University of Warsaw
F. Joliot-Curie 14,
50-383 Warsaw, Poland
2Insitute of Biochemistry and Biophysics,
Polish Academy of Sciences
Pawinskiego 5A,
02-106 Warsaw, Poland
3Laboratory of Comparative Carcinogenesis,
Bldg. 538, Rmm 205,
NCI-FCRDC,
Frederic, MD 21702
Protamines are arginine-rich peptides that bind DNA in the sperm head. Protamine HP2, constituting ca. 50-70% of total human protamines is a 57-peptide, containing apart from 27 arginines, also 9 histidine and 5 cysteine residues. This feature makes HP2 a target for metal ions, like essential Zn(II) and toxic Pb(II). Another striking feature of HP2 is the presence of the N-terminal domain Arg-Thr-His. This sequence is a high-affinity binding site for Cu(II) and Ni(II) ions. In a previous study, the co-ordination of these metal ions to the N-terminal 15-peptide of HP2, RTHGQSHYRRRHCSR-amide was studied by potentiometry and a range of spectroscopic techniques (1). The oxidative reactivity of respective complexes was also investigated (2,3). These studies allowed to suggest that the N-terminal motif of HP2 serves as a protective site against copper genotoxicity, but may be an activator of nickel-inflicted damage. However, some spectroscopic effects, as well as details of oxidative reactions, indicated at the structuring of the peptide molecule resulting from metal ion coordination. In order to study this phenomenon a series of NMR experiments was performed on the Ni(II) complex of the 15-peptide. NMR NOESY experiment allowed us determination of interproton distance restraints. Standard X-PLOR (4) distance geometry and simulating annealing protocol was used for structure determination. A total of 300 structures were computed, from which 20 best were taken and refined. The average root mean square difference (rmsd) between final 20 structures and the mean structure of backbone atoms of the first eight aminoacids is below 1 A, showing secondary structure formation near Ni(II) complex.
The main features of the complex structure are in full agreement with the properties observed in previous studies. The ability of Ni(II) to induce secondary structure in a peptide, running as far as five aminoacid residues beyond the coordination site, is a novel and unexpected result, extending our understanding of the roles of metal ions in biological systems.
Acknowledgements
This work was sponsored by The Polish Committee for Scientific Research (KBN) Grant np. 6 PO4A 024 13
References and Footnotes
1. Bal, W., Je_owska-Bojczuk, M., and Kasprzak, K.S., Chem.
Res. Toxicol. 10, 906-914 (1997).
2. Bal, W., Luszko, J., and Kasprzak, K.S., Chem. Res. Toxicol. 10,
915-921 (1997).
3. Liang, R., Senturker, S., Shi, X., Bal W., Dizdaroglu, M., and Kasprzak,
K.S., Carcinogenesis, (1997), in press
4. Brunger, A.T., X-PLOR Version 3.1, (1993).
K. J. McConnel2*, H. Arthanari1, M. A. Young2, R. Beger3, P. H. Bolton1 and D. L.
Beveridge1
1Department of Chemistry and Program in Molecular Biophysics,
Wesleyan University,
Middletown, Connecticut 06459
2Laboratories of Molecular Biophysics,
3Howard Hughes Medical Institute,
1230 York Avenue, New York, NY 10021
*Author to whom correspondence should be addressed. Phone: (860) 685-3214
; Fax: (860) 685-2211; E-mail: kmcconnell@wesleyan.edu
We describe a critical comparison of NMR structures, NOE and J-coupling data of the DNA oligonucleotides d(CGCGAATTCGCG) and d(CGCAAAAATGCG) with back calculated NOE and J-coupling data from crystal structures and molecular dynamics simulations in solution and crystal conditions. Based on previous study we have established that using back calculated NMR data of a DNA structure in a NMR refinement protocol generates an ensemble of structures which agrees with the original target structure to less than 1Å RMSD. In this study we describe the time scale for convergence of back calculated NMR observables from the molecular dynamics(MD) simulation and how the ensemble of structures relates to the MD average structure. In addition we compare different measures of fit including R-factor, Q-factor and RMS of two data sets and describe the degree of sensitivity of these to conformational changes.
The sequence d(CGCGAATTCGCG) is chosen as a prototypical system, which has been extensively studied in the field of molecular biophysics. Here we compare a 14ns solution phase MD simulation, 12ns crystalline phase MD simulation, 1.4 A crystal structure and NMR structure to the observed NMR data set. We analyze the differences between different classes of NOEs and their structural implications.
To further test the sensitivity of NMR data to conformational variability
we have chosen the d(CGCAAAAATGCG) sequence. The crystal structure of this
sequence has been previously solved by DiGabriele et al. (PNAS 1989) in
which the electron-density map fits two different orientations, "Up"
and "Down" of the sequence confirmed by partial occupancy of a
Bromine tag on one strand. The differences of these two orientations will
be analyzed through the NMR back calculation protocol and compared to MD
simulations and NMR data. The model of the solution structure is discussed
in context of these results.
R. L. Ornstein, C.-G. Zhan, O. Norberto de Souza and R. Rittenhouse
Battelle-Northwest,
Mailstop K2-21,
Richland, WA 99352.
Phosphotriesterase (PTE) from Pseudomonas diminuta catalyzes the hydrolysis of organophosphorus pesticides and related nerve agents, i.e. acetylcholinesterase inhibitors, with rate enhancements that approach 1012. There is much current interest both in chemistry and biochemistry in understanding how this remarkable enzyme and the related dinuclear metal complexes catalyze the hydrolysis so effectively. In order to elucidate the catalytic mechanism, it is first necessary to determine the structure of the active site. Recently, Vanhooke et al. reported a three-dimensional X-ray crystal structure of the zinc-substituted PTE complexed with the substrate analog, diethyl 4-methylbenzylphosphonate. The X-ray crystal structure indicates that the two zinc ions in the active site are separated by 3.3 Å. One of the two bridging ligands for the dinuclear metal center is a carbamylated Lys 169 residue. It is uncertain whether the other bridging ligand is a water molecule or a hydroxide ion, since hydrogen atoms cannot be determined by X-ray diffraction techniques. It is expected that this second bridging ligand, hydroxide or water, is directly involved in the catalytic hydrolysis process. Here we theoretically determine the identity of this critical bridging ligand by performing both molecular dynamics (MD) simulations on the solvated PTE-inhibitor complex, and quantum chemical calculations on simplified models of the active site. Both methods lead to exactly the same conclusion which will be described. We also now describe for the first time a minimum two-state reaction model. Finally, we will describe why this enzyme is an ideal candidate for rational enzyme redesign for enhanced degradation of nerve agents such as sarin and somain.
Paolo Catasti
LS-8, M888, Life Sciences Division,
Los Alamos National LAboratory,
Los Alamos, NM 87545
We summarize binding measurements by surface plasmon resonance technique
(a BIAcore 2000 instrument) on protein-protein and protein-DNA complexes.
As an example of protein-protein complex, we have chosen to study the interactions
of staphylococcus enterotoxin B (SEB) and chimeric protein that inhibits
SEB pathogenesis. The chimeric protein is bispecific, i.e., it contains
two domains (DRa and TCRVb) that binds to two non overlapping epitopes on
SEB. The two domains in the chimeric protein are linked by a flexible spacer
of 14 amino acids. We demonstrate that the chimeric protein, DRa-linker-TCRVb,
binds to SEB with 10 times higher affinity than the individual (DRa and
TCRVb) domains. As an example of protein-DNA complexes, we have chosen to
study the interaction of hairpin G-quartet with the Ku70/Ku80 heterodimer
that repair double-strand breaks. We show that under physiological conditions
Ku70/Ku80 bind with a high affinity (KD ~ 1 nM) to the hairpin G-quartet
structure formed by the human telomere repeat, (TTAGGG)n. Ku70/Ku80 does
not bind to the complementary strand (CCCTAA)n that is unstructured under
physiological conditions. Specific binding to Ku70/Ku80 to (TTAGGG)n may
be biologically important in rendering stability in length of the single-stranded
telomere overhang.
Prasad Kuduvalli and Nancy L. Craig*
Howard Hughes Medical Institute,
The Johns Hopkins University School of Medicine,
725, N. Wolffe St.,
Baltimore, MD 21205
*Phone: 410-955-3933; Fax: 410-955-0831; E-mail: ncraig@jhmi.edu
The bacterial transposon Tn7 inserts at a high frequency into a specific site on the E. coli chromosome called attTn7. This process requires four Tn7-encoded proteins: TnsA and TnsB, which constitute the transposase; TnsC, the ATP-dependent regulator protein; and TnsD, an attTn7-specific DNA binding protein. Our working model comprises the formation of a complex of TnsD and TnsC on attTn7 DNA, followed by the recruitment of the transposase and donor, yielding an insertion at attTn7.
We have characterized the interaction of TnsD and TnsC with attTn7
DNA by various solution methods. Our results indicate that a TnsD-induced
distortion on attTn7 DNA plays a critical role in the recruitment
of TnsC and transposase to this site. According to our hypothesis, target
DNA acts as an effector in Tn7 transposition, thus ensuring the high fidelity
and efficiency of the process. Moreover, a structural distortion induced
by triplex formation has been found to recruit TnsC and the rest of the
transposition machinery, and insert in a site-specific manner, analogous
to the TnsABCD pathway. The transposon exploits both sequence-encoded and
protein-induced DNA structure to insert into a specific target at a high
frequency. The specific role of intrinsic and protein-induced curvature
and flexibility will be discussed.
Jason E. Rao1, Paul S. Miller2
and Nancy L. Craig1*
1Howard Hughes Medical Institute,
Johns Hopkins University,
School of Medicine
2School of Hygeine and Public Health,
725, N. Wolffe St.,
Baltimore, MD 21205
*Phone: 410-955-3933; Fax: 410-955-0831; E-mail: ncraig@jhmi.edu
Found in virtually all organisms examined, transposons are mobile DNA segments implicated in playing a role in evolution and are central to a diverse range of biological processes which include genomic alteration, viral integration and propagation of antibiotic resistance genes (1). Transposons have been exploited for their use in molecular genetics as genetic mutagens, as 'reporters' of gene expression, for isolating genes and for manipulating chromosome structure. An underlying theme central to each application is target site selection. Each mobile element has a unique method of determining its insertion site which can vary from being highly selective to having almost no site preference.
Target site selection has been particularly well defined in bacterial transposon Tn7 (3). Unique amongst transposons, Tn7 is capable of high-frequency transposition into a single site in the E.coli chromosome known as attTn7. This ability comes from an elaborate array of transposition genes, which separate and distribute Tn7's transposition functions into several encoded proteins. TnsA+B make up the transposase, TnsC mediates the activity of the transposase to TnsD, the attTn7 selecting protein. In the absence of TnsD, the core transposition machinery shows no activity. However, we have found that TnsABC is able to carry out transposition to a specific site determined by the presence of triplex DNA. We have shown this activity is a result of TnsC's ability to act as a triplex binding protein. In addition, it has recently been shown that both triplex and TnsD induce conformational distortion on target DNA. This Distortion is believed to be the signal for recruitment of TnsABC transposase.
In recent years, oligonucleotide-directed triplex formation has been shown to be a versatile method for sequence-specific recognition of double helical DNA (4), with applications ranging from genomic mapping to gene therapy (5,6). An important piece of evidence to support the existence of triplex DNA in vivo, is the identification of proteins that specifically interact with triplexes. The recognition of triple helical DNA by TnsC and the applicatins of a triplex-directed transposition will be discussed.
References and Footnotes
1. D.E. Berg, M.M. Howe, eds., Mobile DNA. Washington
D.C.: American Society for Microbiology (1989).
2. N.L.Craig, Annu. Rev. Biochem., 66, 437-74 (1997).
3. N.L.Craig, Current topics in Microbiology and Immunology, 204,
27-48 (1996) .
4. M.D. Frank-Kaminetskii, S.M. Mirkin, Annu. Rev. Biochem., 64,
65-95 (1995).
5. C. Helene, N.T. Thuong, Angew. Chem. Int. Ed. Engl., 32, 666-690
(1993).
6. H.E. Moser, P.B. Dervan, Science, 238, 645-650 (1987)
.
A.P.Sokolov1* and H.Grimm2
1Department of Polymer Science,
University of Akron,
Akron, OH 44325-3909
2IFF,
Forschungszentrum Juelich, Germany
*Phone: 330-972-8409; Fax:330-972-5290, E-mail: sokolov@polymer.uakron.edu
Inelastic neutron scattering spectra of DNA-fibers are analyzed using ideas formulated recently in the field of the glass transition. The analysis [1] reveals two temperatures, namely T~180-200K and T~230K, at which the dynamics of DNA exhibits qualitative changes. The former is similar to the glass transition temperature, whereas the latter is similar to the crossover temperature recognized now as an important point for the dynamics of the glass transition. Exactly in this temperature range many other hydrated bio-polymers show some dynamic transition and strong slowing down of their functions. The crossover temperature appears to be close to the crossover temperature of bulk water. Analysis of DNA-fibers with low level of hydration clearly demonstrate a suppression of the transition at T~230K. A possible relation of the dynamic transition to functions of biomolecules and also to dynamic transition in hydration shell is discussed.
References and Footnotes
1. Sokolov, A.P., Grimm, H., Kahn, R., J.Chem.Phys., 110, 7053-7057 (1999).
Yehuda Goldgur1, Alison B. Hickman1, Tamio Fujiwara2, Tomokazu
Yoshinaga2, Toshio Fujishita2,
Hirohiko Sugimoto2, Takeshi Endo2,
Hitoshi Murai2, Robert Craigie1
and David R. Davies1
1Laboratory of Molecular Biology,
National Institute of Diabetes and Digestive and Kidney Diseases,
National Institutes of Health,
Bethesda, MD 20892
2Shionogi Institute for Medical Science,
Shionogi and Co, Ltd.,
2-5-1, Mishma, Settsu-shi,
Osaka 566, Japan
HIV-1 integrase is responsible for catalyzing the insertion of the viral genome into the host cell chromosome; it provides an attractive target for anti-viral drug design. The previously reported crystal structure of the HIV-1 integrase core domain revealed that this domain belongs to the superfamily of polynucleotidyl transferases (1). However, the position of the conserved catalytic carboxylic acids differed from those observed in other enzymes of the class, and attempts to crystallize in the presence of the cofactor, Mg2+, were unsuccessful. We examined the available structures of the HIV-1 integrase core domain and identified W131 as possible mutation site. Since new crystal forms were obtained for both W131K and W131E mutants, this indicates that such a structure-based approach is fruitful and potentially generalizable. We report new crystal structures of the core domain of HIV-1 integrase mutants and of its complexes with Mg2+ and an inhibitor. The new crystal forms reveal substantial deviation from the original crystal structure with Calpha atoms of the catalytically essential D116 shifted apart by 3.5 A, while the distance between the corresponding carboxylate ends of the side chains is as long as 7 A. The new crystal forms exhibit a consistent position of the essential catalytic residue, E152, which was disordered in the original structure. The differences are attributed to the modificaton of two cysteine residues, C65 and C130, by cacodylate in the original structure. Our results resolve the controversy between the structures of polynucleotidyl transferases. The complex with metal cofactor contains one bound Mg2+ion per monomer. It has been proposed, by analogy to the 3'-5' exonuclease of E. coli polymerase I, that retroviral integrases bind two metal ion cofactors at each active site (2). If two metal ions are indeed required for catalysis, it is probable that the second ion binds only in presence of the DNA substrate. The complex with a potential inhibitor demostrates that binding occures at the active site of the enzyme with several catalytically important residues directly involved in the interactions with the inhibitor.
References and Footnotes
1. Dyda, F., Hickman, A. B., Jenkins, T. M., Engelman,
A., Craigie, R., Davies, D. R., Science 266, 1981-1986, 1994.
2. Kulkosky, J., Jones, K. S., Katz, R. A., Mack, J. P. G., Skalka, A. M.,
Mol. Cell. Biol. 12, 2331-2338, 1992.
Carolina M. Reyes and Peter A. Kollman*
Dept. of Pharmaceutical Chemistry,
Box 0446,
UCSF,
San Francisco, CA 94143-0446
*Phone:415-476-4637, Fax:415-502-1411; E-mail: pak@cgl.ucsf.edu
The spliceosomal protein U1A binds a hairpin RNA and an internal loop
RNA with subnanomolar affinities. To better understand the dynamics and
energetics of RNA-protein recognition, we studied both complexes using molecular
dynamics and free energy calculations. The absolute binding free energies
are estimated as the sum of gas phase energies and solvation free energies.
The solvation free energies are computed with finite-difference Poisson-Boltzmann
for the electrostatics contribution and a surface-area-dependent term for
the non-polar contribution. We present the calculations of free energies
of binding for both complexes and for different U1A mutants and compare
them with experimental results.
Manal A. Swairjo1, Arturo J. Morales1,2 and Paul Schimmel1
1Skaags Institute for Chemical Biology and Departments of
Molecular Biology and Chemistry,
The Scripps Research Institute,
10550 N. Torrey Pines Rd.,
La Jolla, CA 92037
2Department of Biology,
Massachusetts Institute of Technology,
Cambridge, MA 02139
Trbp111 is a newly identified tRNA-specific binding protein from the
extreme thermophile Aquifex aeolicus. The 111 amino-acid protein is functionally
unique among known tRNA binding proteins in its requiring the general L-shape
characteristic of full-length tRNA for recognition, without specificity
with respect to the type of tRNA. The presence of trbp111 homologues in
other prokarytes and eukaryotes has led to the proposal that trbp111 is
an ancient tRNA-binding element that has been incorporated into much larger
tRNA binding proteins in higher organisms. We are pursuing the crystal structure
of Aquifex aeolicus trbp111 and its homologue from E. coli (designated ectrbp111).
Predicted to exhibit a novel tRNA-binding fold, a high-resolution structure
of trbp111, ectrbp111 and their complexes with tRNA may provide unique insight
into the early evolution of the tRNA/synthetase recognition system. Crystals
of trbp111 and ectrbp111 were grown from PEG solutions. Ectrbp111 crystals
belongto space group P3121 and contain one molecule the asymmetric unit.
Diffraction data were collected from native and derivatized crystals to
1.5 Å and 2.8 Å, respectively. Difference-Patterson and Fourier
maps were used to determine heavy atom positions from three derivatives
and initial heavy-atom phases calculated at 3.5 Å. Initial solvent-flattened
electron density maps reveal a homodimer with half a dimer per asymmetric
unit and a helical dimer interface.
Yael Mandel-Gutreund1 and Victor B. Zhurkin2
1 Department of Molecular Genetics and Biotechnology,
The Hebrew University,
POB 12272,
Jerusalem 91120 Israel
2 Laboratory of Experimental and Computational
Biology,
National Cancer Institute,
NIH,
Bethesda, MD 20892-5677 USA
In addition to the donor-acceptor pattern, the protein binding sites are characterized by nonrandom positioning of easily deformable DNA elements facilitating sterical fit between a protein and double helix (the so-called "indirect" or "conformational" recognition). The DNA deformations usually occur in the flexible pyrimidine-purine (YR) "hinges" and involve bending, twisting and sliding motions that are closely correlated (1). To analyze the protein binding sites in this context, we applied the auto- and cross-correlation functions (2-4). These functions help revealing the hidden periodicity patterns, e.g., detecting imperfect direct and mirror repeats, as well as distinguishing between the coding and non-coding DNA fragments.
Three sets of the transcription factors binding sites (TF-sites) were studied, from E. coli, yeast and vertebrates, comprising ca. 500, 100 and 700 sequences respectively. Overall, the correlation functions for TF-sites reveal minima at 3 bp, as opposed to the coding sequences having maxima at 3 bp (2-4). Bacterial TF-sites are characterized by 10-11 bp and 20-21 bp periodicities, the strongest periodic patterns being demonstrated by the AA:TT and CA:TG dimers (and YR:YR in general). By contrast, eukaryotic TF-sites have periodicity profiles with the strongest maxima at 4-6 bp and 14-16 bp. The peaks at 10-11 bp and 20-21 bp are less pronounced than in E. coli.
We suggest this difference in the DNA periodicity can be related to the difference in architectural organization of the DNA-protein complexes. In prokaryots, in most of the known complexes the proteins are localized "internally" with respect to the DNA loop. Hence, the DNA bending appears to play a pronounced role, and the DNA recognition sequences are organized consistently with the DNA looping, in turn related to the 10-11 bp periodicity (2). In eukaryots, the DNA is tightly packed in nucleosomes, and the "internal localization" of the DNA-bound proteins would be hindered by histones. Thus, the transcription factors interact with the double helix predominantly on the "external" side of a DNA loop (e.g., TBP, p53). The recognition patterns in the DNA sequences are therefore separated by multiples of a half-pitch of the duplex, enabling the protein subunits to "grip" the double helix from the outside of a nucleosome.
References and Footnotes
1. Olson, W.K., Gorin, A.A., Lu, X.-J., Hock, L.M. and
Zhurkin, V.B,. Proc. Natl. Acad. Sci. USA 96, 1875-1880 (1999).
2. Trifonov, E.N. and Sussman, J.L. Proc. Natl. Acad. Sci. USA 77, 3816-3820
(1980).
3. Zhurkin, V.B., Nucleic Acids Res. 9, 1963-1971 (1981).
4. Hertzel, H., Weiss, O. and Trifonov, E.N., J. Biomol. Struct. Dynam.
16, 341-345 (1998).
Christophe Escudé*, Rula Zain, Christophe Marchand, Jian-Sheng
Sun, Chi-Hung Nguyen, Thérèse Garestier and Claude Hélène
Laboratoire de Biophysique,
Muséum National d'Histoire Naturelle,
INSERM U 201,
CNRS UMR 8646,
43, rue Cuvier, 75231 Paris Cedex 05, France
*E-mail: escude@mnhn.fr
Agents that bind double-stranded DNA with high affinity and sequence specificity could be used to control the expression of specific genes at the level of transcription. Triple helix forming oligonucleotides are among the most specific of these ligands. However, the formation of DNA triple helices might be limited under physiological conditions. Specific ligands can intercalate these structures and stabilize them. Based on molecular modeling studies and experimental results (1), a new pentacyclic ligand, named benzoquinoquinoxaline (BQQ), has been designed, which provides optimal stacking interactions with base triplets (2). Thermal denaturation experiments and inhibition of restriction enzyme cleavage show that this newly synthesized compound can indeed stabilize triple helices with great efficiency and/or induce triple helix formation under physiological conditions.
In order to design a triple helix-specific DNA cleaving reagent, BQQ has been covalently linked to ethylenediaminotetraacetic acid (EDTA). The intercalative binding of BQQ in triple helix should position EDTA in the minor (Watson-Crick) groove of triple helix. In the presence of Fe2+ and a reducing agent, the BQQ-EDTA conjugate cleaves an 80-bp DNA fragment selectively at the site where an oligonucleotide binds to form a local triple helix. The selectivity of the BQQ-EDTA conjugate for a triplex structure was high enough to induce an oligonucleotide-directed DNA cleavage at a single site on a 2718-bp plasmid DNA (3). This new class of structure-directed DNA cleaving reagent should be of interest to cleave DNA at specific sequences and to investigate triple-helical structures, such as H-DNA, which could play an important role in the control of gene expression in vivo.
References and Footnotes
1. Escudé, C., Nguyen, C.H., Mergny, J.L., Sun,
J.S., Bisagni, E., Garestier, T. & Hélène C. Journal
of the American Chemical Society , 117, 10212-10219 (1995).
2. Escudé, C., Nguyen, C.H., Kukreti, S., Janin, Y., Sun, J.S., Bisagni,
E., Garestier, T. & Hélène C. Proc. Natl. Acad. Sci.
USA. 95, 3591-3596 (1998).
3. Zain, R., Marchand, C., Sun, J.S., Nguyen, C.H., Bisagni, E., Garestier,
T. & Hélène C., Chem. Biol., in press (1999).
Angel E. Garcia1, Chang-Shung Tung1, Gina Arents2 and Evangelos
N. Moudrianakis2
1Theoretical Biology and Biophysics Group,
T10, MS K710,
Los Alamos National Laboratory,
Los Alamos, New Mexico 87545
2Department of Biology,
The Johns Hopkins University,
Baltimore, MD 21218
An atomic model for the nucleosome core particle has been constructed by utilizing, a) the crystallographic coordinates of the core histone octamer (Arents et al., 1991) and b) the geometric constraints of the octamer surface to "dock'' DNA on to it, as described by Arents et al. (1993). Specifically, the path of the DNA is defined by two sets of high-resolution crystallographic criteria:
1) The symmetries of the histone fold, the ``Handshake motif'', and the
periodic occurrence of the binary "Paired Element Motifs'' (PEM) occurring
12 times on the octamer surface.
2) The preponderance of positive charges appearing on the octamer surface
as a left-handed spiral.
Additional chemical constraints utilized in this modeling include: DNA-histone contacts and proximities derived from chemical crosslinking, nuclease digestion, and hydroxy-radical footprinting data.
We next employed molecular modeling methods to generate an atomic-level description of a superhelical, 149 base-pair DNA double helix. The fluctuations of the resulting system are simulated in aqueous solution, in the presence 150 mM sodium chloride (excess salt) at room temperature and under periodic boundary conditions, for a period of 1.2 nanoseconds.
We find that the nucleosome is a intrinsically dynamic structure where
fluctuations are dominated by motions in the DNA backbone. The DNA atomic
fluctuations are much larger than those of the protein atoms by a factor
of 10. Wedges among DNA base pairs form and disappear in the nanosecond
time scale of the simulation. The protein-DNA complex is stabilized by the
interactions of charged side-chains, hydrophobic contacts and ions from
the solution. The nucleosome is surrounded by a positive ion cloud with
an average local density exceeding by a factor of 5-10 the bulk phase ion
concentration. Particularly high ion concentrations are found within the
DNA grooves. We also see high water density at the protein-DNA boundaries,
at the DNA grooves, and specially between the two apposing gyres of the
nucleosomal DNA.
Chang-Shung Tung1*, Stephen Gallagher2 and Jill Trewhella2
1Theoretical Biology and Biophysics,
2Chemical Science and Technology,
Los Alamos National Laboratory,
Los Alamos, NM 87545
*Phone: 505-665-2597, E-mail: cst@t10.lanl.gov
Attempts to crystallize troponin C complexed with troponin I, or the ternary troponin complex, have to date failed to deliver a high resolution structure for this critical molecular switch responsible for calcium-dependent regulation of the contractile mechanism. Given this limitation, many groups have persued different experimental strategies to gain insights into the nature of the TnC/TnI interaction. The high resolution crystal structure of TnC has been available for more than a decade, and recently a structure of TnC complexed with the N-terminal sequence segment (residues 1-47) has been solved. Peptide binding and competition studies have defined a number of important regions in TnI that interact with TnC, while NMR studies of TnC with various TnI peptides have yielded information on sites of interaction as well as limited structural information on the peptides. Crosslinking studies and fluorescence energy transfer data provide additional constraints. Neutron contrast variation studies have provided general shape information on TnC and TnI in both the binary complex and in the ternary troponin complex. We have attempted to bring all of the experimental constraints available for the TnC/TnI interaction together to develop a model that best fits the currently available data. The model is surprisingly well packed and has been subjected to molecular dynamics simulation and energy minimization. We preset the model not as a final answer, but as a basis for discussion of the volumous structural data we have on this tantalizing system and as a basis for developing further experiments for testing and refinement.