Sequence Specific DNA Binding Of Ets-1 Transcription Factor .
doi:10.1016/S0022-2836(03)00726-5J. Mol. Biol. (2003) 331, 345–359Sequence Specific DNA Binding of Ets-1 TranscriptionFactor: Molecular Dynamics Study on the EtsDomain –DNA ComplexesSatoshi Obika, Swarnalatha Y. Reddy and Thomas C. Bruice*Department of Chemistry andBiochemistry, University ofCalifornia, Santa Barbara, CA93106, USAMolecular dynamics (MD) simulations for Ets-1 ETS domain– DNA complexes were performed to investigate the mechanism of sequence-specificrecognition of the GGAA DNA core by the ETS domain. Employing thecrystal structure of the Ets-1 ETS domain– DNA complex as a startingstructure we carried out MD simulations of: (i) the complex between Ets1 ETS domain and a 14 base-pair DNA containing GGAA core sequence(ETS –GGAA); (ii) the complex between the ETS domain and a DNA having single base-pair mutation, GGAG sequence (ETS – GGAG); and (iii) the14 base-pair DNA alone (GGAA). Comparative analyses of the MD structures of ETS – GGAA and ETS – GGAG reveal that the DNA bendingangles and the ETS domain– DNA phosphate interactions are similar inthese complexes. These results support that the GGAA core sequence isdistinguished from the mutated GGAG sequence by a direct readoutmechanism in the Ets-1 ETS domain– DNA complex. Further analyses ofthe direct contacts in the interface between the helix-3 region of Ets-1 andthe major groove of the core DNA sequence clearly show that the highlyconserved arginine residues, Arg391 and Arg394, play a critical role inbinding to the GGAA core sequence. These arginine residues make bidentate contacts with the nucleobases of GG dinucleotides in GGAA coresequence. In ETS – GGAA, the hydroxyl group of Tyr395 is hydrogenbonded to N7 nitrogen of A3 (the third adenosine in the GGAA core),while the hydroxyl group makes a contact with N4 nitrogen of C4 (thecomplementary nucleotide of the fourth guanosine G4 in the GGAGsequence) in the ETS –GGAG complex. We have found that this differencein behavior of Tyr395 results in the relatively large motion of helix-3 in theETS – GGAG complex, causing the collapse of bidentate contacts betweenArg391/Arg394 and the GG dinucleotides in the GGAG sequence.0q 2003 Elsevier Ltd. All rights reserved.*Corresponding authorKeywords: molecular dynamics; Ets-1; ETS domain; transcription factor;protein –DNA complexIntroductionThe ETS protein family contains more than 45eukaryotic transcription activators and inhibitors,such as Ets-1, PU.1, Fli-1, GABPa, SAP-1, TEL andElk-1.1 – 3 Members of this family play an importantrole in normal cell proliferation and differentiation.The DNA rearrangement and/or overexpression ofets gene have been known to lead totumorigenesis.4 In order to regulate geneexpression, the ETS family of proteins bind to aAbbreviation used: MD, molecular dynamics.E-mail address of the corresponding author:tcbruice@bioorganic.ucsb.educonsensus DNA sequence centered on the coresequence 50 -GGA(A/T)-30 through the highly conserved DNA-binding domain.3 The DNA-bindingdomain for ETS proteins, termed ETS domain, isabout 85 amino acid residues in length and formsa winged helix-turn-helix motif consisting of threea-helices and four b-strands. The recent X-ray5 – 10and NMR11 – 13 studies of the ETS domain –DNAcomplexes have shown that the helix-3 in thewinged helix-turn-helix motif binds in the majorgroove of the consensus DNA sequence.In the crystal structure of Ets-1 ETS domain –DNA [d(TAGTGCCGGAAATGT)2] complex (PDBcode: 1K79), two arginine residues, Arg391 andArg394, which are in the helix-3 region and0022-2836/ - see front matter q 2003 Elsevier Ltd. All rights reserved.
346MD Simulations of Ets-1 ETS Domain–DNA Complexesconserved among the ETS family, make bidentateinteractions with G1 and G2, respectively (Figure1).5 However, the pattern of these hydrogen bondsis not maintained in the crystal structures of otherETS domain– DNA complexes.7,9,10 In addition, theinteraction between the arginine residues, Arg391and Arg394, and the consensus DNA sequence isnot observed in the NMR study on Ets-1 ETSdomain– DNA complex.12,13On the other hand, a few studies have proposeddirect contacts of the amino acid residues in ETSdomain with the AA region (þ 3 and þ 4 positions)in the GGAA core. For example, an X-ray study ofthe Ets-1 ETS domain –DNA complex indicatedwhat would be a vital role of hydrophobic interaction between the phenyl ring of Tyr395 and 5methyl group of T4 or T5 .5 However, this type ofinteraction was not observed in other ETSdomain– DNA complexes,6 nor is the tyrosine residue conserved in other ETS family proteins suchas PU.1 and TEL.9,10 Thus, the precise molecularmechanism that clearly explains the sequencespecific GGAA recognition by ETS domain is stilllacking.The phosphate groups of DNA adjacent to thecore sequence GGAA have contacts to the wingedsegment and the turn region between helix-2 andhelix-3 of the ETS domain. It was reported that theneutralization of anionic phosphate charges onone face of DNA resulted in the DNA bending,probably due to the electrostatic repulsions of theremaining anionic charges.14 – 16 In fact, DNA bending was observed in the crystal structures of theETS domain –DNA complexes.5 – 10 It was also supposed that the conformational change of DNAcaused by the DNA bending would serve to provide effective GGAA core recognition by the helix3 of the ETS domain. However, the bending anglesof DNA previously reported in the X-ray crystallographic analyses of ETS domain– DNA complexesvary from one system to another.5 – 10 Thus,additional investigation is required in order to clarify a common role of the DNA bending in thesequence-specific binding of the ETS domain.Significant developments have been made in thelast few years in the procedures of moleculardynamics (MD). Improvements in AMBER,17CHARMM18,19 and GROMOS20 force fields andeffective treatment of long-range electrostatic interactions by using particle mesh Ewald (PME)method,21 explicit inclusion of solvent and ionshave opened the possibilities for accurate determination of protein and DNA structures.22 Besidesavailability of supercomputers have enabled toundertake simulations on a nanosecond (ns) timescale which expanded conformational samplingand eventually to elucidate biomolecular interactions. Further understanding of the molecularmechanism of sequence-specific DNA binding ofthe ETS domain is likely to provide novel clues forthe design of drugs that bind to inhibit the interaction between the ETS domain and DNA. Wereport here 3.5– 3.9 ns MD simulations of two Ets1 ETS domain –DNA complex systems. The aminoacid sequence of Ets-1 ETS domain and the DNAsequences used in this study are shown in Figure 2.The binding activity of the ETS domain is knownto be higher than that of the whole ETS-1protein.23 – 27 Therefore, the ETS domain– DNAcomplex would be a good model system for MDsimulation. The first system has the ETS domain(103 amino acid residues) of Ets-1 protein (Figure2(a)) and the high affinity 14 base-pair DNA containing GGAA core (þ 1 to þ 4, Figure 2(b))sequence, while the second one has a low affinityDNA involving a mutation of a single base-pair,GGAG sequence. In the text, we refer to these complexes as ETS – GGAA and ETS – GGAG, respectively. In addition, results from the MD simulationof the 14 base-pair DNA having the GGAA coresequence (referred as GGAA) are also discussedfor comparison.00Figure 1. Crystal structure of Ets-1 ETS domain – DNA complex (PDB code: 1K79): (a) the whole structure, (b) theclose-up view of the ETS domain – DNA nucleobase contact site (helix-3 and GGAA core sequence).
347MD Simulations of Ets-1 ETS Domain–DNA ComplexesFigure 2. Sequences of Ets-1 ETSdomain and 14 base-pair DNA: (a)amino acid sequence with the residue numbers and secondary structure indicated above the sequence,(b) DNA base sequence with thenumbering provided above orbelow the sequence. Residues inthe core GGAA are shown in red.In the GGAG sequence, themutated GC base-pair is shown inblue.ResultsThe root-mean-square deviations (RMSD) of theprotein backbone and DNA heavy-atoms withrespect to the minimized structures of ETSdomain– DNA complexes and DNA (GGAA) aregiven in Figure 3. During the simulation, theRMSD values of the protein fluctuated around1.16 –2.12 Å in ETS –GGAA and 0.92– 1.55 Å inETS –GGAG (Figure 3(a)), while those of DNA inETS –GGAA and ETS – GGAG are from 1.06 Å to2.18 Å and from 1.12 Å to 1.96 Å, respectively(Figure 3(b)). The plots indicate the stability atabout 900 ps, except in the ETS –GGAA proteinstructure. The DNA structure of GGAA exhibitsrelatively large RMSD values, compared to theETS –GGAA and ETS –GGAG complexes (greenline in Figure 3(b)). This indicates the absence ofETS domain would cause conformational changesin the DNA structure.The positional fluctuations of Ca atoms (CA) ofEts-1 ETS domain evaluated from MD trajectoriesare shown in Figure 4 along with that from thecrystal structure. Although the magnitude of thefluctuations from X-ray and MD structures aredifferent, the fluctuation pattern is similar. Thehelix-3 region (residues 386 – 396) of the ETSdomain that recognizes the core DNA sequencehas smaller fluctuations, compared to the otherpart of the protein. On the other hand, the turnregion (379 – 384) between helix-2 and helix-3 andthe winged region (405 –410) show larger fluctu-ations. The steep peak observed in the C-terminalof helix-1 (348 – 353) of ETS – GGAA MD structureis likely due to the contact between the residues348– 353 and helix-5 region (430 – 436). However, itdoes not seem to affect any other structural features of the ETS domain.DNA structure of Ets-1 ETSdomain –DNA complexDNA bendingThe time-variation plots of DNA bending anglefor the ETS – GGAA and ETS – GGAG MD structures are given in Figure 5(a). According to theliterature,14,28 the DNA bending angle is defined asshown in Figure 5(b). Large fluctuations arenoticed in the bending angle with average valueof 168 and 228 for ETS – GGAA (900 – 3480 ps) andETS – GGAG (900 – 3930 ps), respectively. In GGAAhelix, a relatively extended DNA structure isobtained compared to the ETS –DNA complexes.The average (900 –3255 ps) bending angle ofGGAA is found to be 118. A stereo plot of averageDNA structures of ETS –GGAA (900 – 3480 ps),ETS – GGAG (900 – 3930 ps) and GGAA (900 –3255 ps) are given in Figure 5(c) – (e), superimposedon the canonical B-DNA structure.29 The plots indicate that the presence of ETS domain influences theDNA bending. As can be seen in Figure 5(a), (c),and (d), no significant difference in DNA bending
348MD Simulations of Ets-1 ETS Domain–DNA Complexesis observed in the ETS –GGAA and ETS –GGAGcomplexes.Major and minor groove widthsFigure 3. Time evolution of RMSD of the MD complexes of ETS– GGAA (black) and ETS– GGAG (red)with respect to the corresponding minimized structures:(a) the backbone heavy-atoms of the ETS domain and(b) all the heavy-atoms of DNA except terminal basepairs. The RMSD value of the heavy-atoms of GGAAhelix (green) with respect to minimized structure isshown in (b).According to the literature,14 groove widths aredefined as distances between two appropriatephosphate atoms (see Method). The average majorand minor groove widths of 14 base-pair DNA inETS –GGAA, ETS – GGAG and GGAA are summarized in Table 1. The major groove width around thecore GGAA region is found to be comparativelylarge in the crystal structure (PDB code: 1K79),although the DNA bends into the major groove. Inthe MD averaged structure, an expansion of majorgroove width around the core region is alsoobserved. For example, the averaged major groovewidths at base-pair 1 in ETS –GGAA and GGAAare 20.43 Å and 17.11 Å, respectively. Thus, theMD simulation of the ETS domain– DNA complexreproduced this structural feature well. Noteworthy are the observations that fluctuations inthe major groove width around the core region(base-pairs 1– 4) are quite small in ETS – GGAA,while large fluctuations are observed in the majorgroove width at the base-pairs 3 and 4 in ETS –GGAG complex. These small fluctuations in themajor groove width of ETS – GGAA may reflectthe stability of the interaction between the helix-3of ETS domain and the GGAA core sequence.In comparison to the dynamics of the majorgroove, the minor groove widths obtained fromMD simulations of ETS –GGAA, ETS – GGAG andGGAA show little difference between each other.The values of minor groove width in the crystalTable 1. Average major and minor groove widths (Å)and their standard deviations (in parenthesis) of the 14base-pair DNA duplexes of ETS– GGAA, ETS– GGAGand GGAA MD structures. The values of X-ray structureof ETS– GGAA (PDB code: 1K79) are givenETS–GGAA(900–3480 ps)ETS–GGAG(900–3930 ps)GGAA(900–3255 ps)Major 6518.82618.9017.91( 1.50)17.91( 1.64)18.56( 1.17)20.43( 0.84)20.82( 0.43)20.40( 0.85)18.22( 0.75)17.94( 1.00)18.61( 1.54)17.55( 1.64)18.29( 1.67)18.13( 1.15)19.33( 0.80)20.52( 0.40)19.10( 1.63)18.97( 1.79)19.96( 1.90)19.92( 1.85)17.00( 1.49)17.54( 1.71)17.32( 1.55)17.11( 1.93)17.41( 1.78)16.80( 1.50)16.82( 1.24)15.90( 1.46)16.80( 1.40)Minor 59.0614.20( 0.88)13.77( 1.27)15.48( 1.13)14.96( 1.05)14.60( 1.23)13.80( 1.24)12.67( 1.34)12.02( 1.21)13.86( 1.09)14.74( 1.21)15.81( 1.57)15.60( 1.28)14.68( 1.70)14.83( 1.33)13.54( 1.05)11.11( 1.17)13.70( 1.16)14.11( 1.39)14.57( 1.17)14.52( 1.14)13.84( 1.12)13.22( 0.88)13.11( 1.02)13.40( 1.02)BasepairFigure 4. Atomic fluctuations of Ca atoms of the Ets-1ETS domain in MD structures of ETS – GGAA (averagedfor 900– 3480 ps) in black, and ETS – GGAG (averagedfor 900–3930 ps) in red. Fluctuations of Ca atom of thecrystal structure are shown in blue.X-raystructure(PDB:1K79)
349MD Simulations of Ets-1 ETS Domain–DNA ComplexesFigure 5. (a) Time evolution ofthe DNA bending angle u (deg) inthe MD structures of ETS – GGAA(black) and ETS– GGAG (red). Thebending angle of the crystal structure is shown in blue. (b) Schematicrepresentation of definition of DNAbending angle u. Stereo view of theaveraged DNA structures obtainedfrom the MD simulations of (c)ETS– GGAA(900 – 3480 ps)inblack, (d) ETS –GGAG (900 –3930 ps) in red and (e) GGAA(900 – 3255 ps) in green. The canonical B-form DNA structure is alsoshown (blue) in (c) – (e) for comparison. Superimpositioning is performedaccordingtotheorientations of base-pairs 24, 25and 26. All hydrogen atoms areomitted for clarification.structure are consistent with those in the MD structure of ETS – GGAA, except for the values at basepairs 4 and 5.Sugar puckeringThe time evolution of pseudorotational phaseangles ðPÞ of selected nucleosides in both ETSdomain– DNA complexes are represented inFigure 6. The switching of sugar puckering fromC2 -endo (S-form) to C3 -endo (N-form) or from Nform to S-form is observed in some nucleoside residues (Figure 6(b), (c), and (e)). On the other hand,the flexibility in sugar puckering is restricted insome DNA regions where the phosphates ornucleobases have contacts with the ETS domain.00Especially, the sugar conformations of G23 and A6are found to be almost locked in S-form puckering(Figure 6(a) and (f)) due to their 30 -phosphategroup being rigidly held in an interaction withspecific amino acid residues in the ETS domain.The 30 -phosphate oxygen atoms, O1P and O2P ofG23 (by convention listed in Table 2 as the 50 -phosphate oxygen atoms of the neighboring C22,O1P(C22) and O2P(C22), respectively) contact withthe hydroxyl oxygen OH of Tyr386, side-chainamino nitrogen NZ of Lys404 and main-chainamino nitrogen N of Tyr410, while those of A6(listed as O1P(T5 ) and O2P(T5 ) form a salt bridgewith N(Leu337) and OH(Tyr396). A considerabledifference in the puckering between the ETS –GGAA and ETS – GGAG structures is observed at0000
350MD Simulations of Ets-1 ETS Domain–DNA ComplexesFigure 6. Time evolution of the pseudorotation phase angle P of sugar ring of nucleosides (a) G23, (b) C23 , (c) C21, (d)G2, (e) C2 and (f) A6 in the MD structures of ETS– GGAA (black) and ETS– GGAG (red).000C21 (Figure 6(c)). This nucleoside residue in ETS –GGAA complex likely prefers the S-type sugar conformation, while the sugar puckering of C21 inETS –GGAG remains N-type after 2500 ps.Interaction between Ets-1 ETS domain and DNAContacts of Arg391 and Arg394 with nucleobasesin the core sequenceThe distances between non-bonded entities inthe contact region of the ETS domain– DNA complexes determined by MD simulation and crystalstructures are summarized in Table 3. The hydrogen bonding structures involving Arg391 andArg394 with nucleobases are shown in Figure 7.The time variations of selected heavy-atom nonbonded distances are shown in Figure 8. In theETS –GGAA MD structure, Arg391 of ETS is inbidentate contacts with G2 nucleobase by hydrogenbonds between secondary amino nitrogen NE ofArg391 and O6 oxygen of G2, and the otherbetween guanido nitrogen NH2 of Arg391 and N7nitrogen of G2 (Figure 7(a)). These two hydrogenbonds are maintained during the entire simulation(black line of Figure 8(a) and (c)). Such hydrogenbond interactions are observed between Arg394and G1 nucleobase (Figure 7(b)) for the durationafter 1700 ps in ETS – GGAA (black line of Figure8(d) and (f)). These separations are large for theTable 2. Average non-bonded distances (Å) and their standard deviations (in parenthesis) of the contact sites of Ets-1 –DNA phosphate backbone of ETS – GGAA and ETS– GGAG MD structures. The values of X-ray structure of ETS –GGAA (PDB code: 1K79) are givenETS–GGAAAtomsN(Leu337)· · ·O1P(T5 )NE1(Trp375)· · ·O2P(T4 )NZ(Lys379)· · ·O1P(T4 )OH(Tyr386)· · ·O2P(C22)NZ(Lys388)· · ·O2P(T3 )OH(Tyr396)· · ·O2P(T5 )OH(Tyr397)· · ·O2P(C21)NZ(Lys399)· · ·O2P(A6 )NZ(Lys404)· · ·O1P(C22)N(Tyr410)· · ·O1P(C22)OH(Tyr410)· · ·O1P(G23)000000ETS–GGAGX-ray structure (PDB: 1K79)(900–1650 ps)(1800–3480 ps)(900– 2100 ps)(2700–3930 4( 0.20)*3.53( 0.50)4.87( 1.18)2.65( 0.10)*4.71( 0.32)2.64( 0.10)*2.74( 0.14)*3.79( 1.21)2.72( 0.11)**2.83( 0.13)*2.86( 0.48)**2.96( 0.27)*3.22( 0.51)**5.35( 0.67)2.65( 0.10)*2.79( 0.33)*2.81( 0.37)**2.72( 0.13)*3.72( 0.93)2.74( 0.12)**2.87( 0.15)*3.03( 0.69)**3.00( 0.29)*2.96( 0.31)*2.77( 0.18)*2.65( 0.10)*2.75( 0.13)*2.94( 0.49)**2.71( 0.12)*3.98( 0.94)2.87( 0.33)**3.01( 0.28)*2.66( 0.15)*2.96( 0.18)*3.02( 0.40)*2.73( 0.14)*2.65( 0.10)*2.70( 0.10)*2.64( 0.10)*2.69( 0.12)*3.38( 0.83)2.73( 0.11)**3.00( 0.21)*2.64( 0.10)*Values within *0–0.25 Å and **0.25–0.50 Å different from the X-ray structure.
351MD Simulations of Ets-1 ETS Domain–DNA ComplexesTable 3. Average non-bonded distances (Å) and their standard deviations (in parenthesis) of the contact sites of Ets-1 –DNA nucleobase of ETS– GGAA and ETS – GGAG MD structures. The values of X-ray structure of ETS – GGAA (PDBcode: 1K79) are givenETS –GGAAAtomsNE(Arg391)· · ·O6(G2)NH2(Arg391)· · ·O6(G2)NH2(Arg391)· · ·N7(G2)NE(Arg394)· · ·N7(G1)NH2(Arg394)· · ·N7(G1)NH2(Arg394)· · ·O6(G1)OH(Tyr395)· · ·N6(A3)OH(Tyr395)· · ·N6/O6(A4/G4)OH(Tyr395)· · ·O4/N4(T4 /C4 )CD2(Tyr395)· · ·C5M(T5 )CE2(Tyr395)· · ·C5M(T5 )0000ETS –GGAGX-ray structure (PDB: 1K79)(900 –1650 ps)(1800–3480 ps)(900–2100 ps)(2700– 3930 0( 0.15)*3.94( 0.29)**2.90( 0.10)*3.96( 0.53)3.04( 0.21)3.97( 0.62)3.42( 0.38)**3.98( 0.60)**3.52( 0.55)*3.96( 0.28)3.82( 0.30)**2.96( 0.19)**4.20( 0.35)2.90( 0.13)*2.97( 0.12)*3.58( 0.28)*2.91( 0.17)*3.09( 0.26)*3.63( 0.40)*3.58( 0.51)*3.92( 0.38)4.05( 0.43)2.97( 0.17)**4.19( 0.31)2.89( 0.10)*2.94( 0.13)*3.67( 0.30)*2.91( 0.17)*3.29( 0.27)*4.36( 0.38)3.08( 0.25)3.97( 0.37)3.87( 0.31)**3.91( 0.46)2.97( 0.33)3.00( 0.15)*5.16( 0.17)3.00( 0.13)5.13( 0.20)4.27( 0.63)4.64( 0.49)3.00( 0.18)3.93( 0.34)3.75( 0.27)***Values within 0–0.25 Å and **0.25–0.50 Å different from the X-ray structure.time-period 250– 1700 ps, during which a hydrogen bond interaction is observed between the guanido nitrogen NH2 of Arg394 and N7 nitrogen ofG1 nucleobase (black line of Figure 8(e)).In the case of ETS – GGAG, the contact betweenNE of Arg391 and O6 of G2 (red line of Figure8(a)) give way to a hydrogen bond between NH2of Arg391 and O6 of G2 during 2850– 2900 ps(Figure 7(c) and red line of Figure 8(b)). AlthoughNH2 of Arg391 is within 3.0 Å distance from N7nitrogen of G2 even after 2850 ps, the non-bondedangle for NH2(Arg391)– HH21(Arg391)· · ·N7(G2)is 120.7( 27.5)8. This does not allow hydrogenbond formation (Figure 7(c)).30 Arg394 has a contact with G1 forming two hydrogen bonds onebetween NE of Arg394 and N7 nitrogen of G1 andthe other between NH2 of Arg394 and O6 oxygenof G1. These interactions exist until 2100 ps ofdynamics (red line of Figure 8(d) and (f)). The contact between NH1 of Arg394 and OG of Ser390,which would stabilize the helix-3 structure, is alsobroken during the same time (data not shown). Inplace, NH2 of Arg394 forms a hydrogen bondwith N7 nitrogen of G1 (Figures 7(d) and 8(e)). SoFigure 7. Molecular plot showing the contact interactions of MD structures. (a) Arg391 with G2C2 DNA base-pairand (b) Arg394 with G1C1 base-pair in ETS – GGAA, averaged for 1800 –3480 ps; (c) Arg391 with G2C2 base-pair and(d) Arg394 with G1C1 base-pair in ETS – GGAG, averaged for 2700– 3930 ps. See the stable bidentate hydrogen bondsin (a) and (b), while they are absent in (c) and (d).0000
352MD Simulations of Ets-1 ETS Domain–DNA ComplexesFigure 8. Time-dependent variation of separations of (a)NE(Arg391)· · ·O6(G2),(b) NH2(Arg391)· · ·O6(G2),(c) NH2(Arg391)· · ·N7(G2),(d) NE(Arg394)· · ·N7(G1),(e) NH2(Arg394)· · ·N7(G1) and(f) NH2(Arg394)· · ·O6(G1) of ETS –GGAA (black) and ETS– GGAG(red) MD structures.during the course of dynamics certain structuralalignments prevail favoring some interactions atthe cost of others.In the crystal structure of Ets-1 ETS domain–DNA complex, Tyr395 is proximal to A4 and T4nucleobases in the major groove of the core region(Figure 1).5 As shown in Table 3, the hydroxylgroup of Tyr395 is at 3.09 and 3.63 Å from the exocyclic amino nitrogen atoms N6 of A3 and A4,respectively in the MD structure of ETS – GGAA.This indicates that the hydroxyl group forms ahydrogen bond with N6 of A3, while it makes aweak contact at A4. The interaction between thehydroxyl group of Tyr395 and the carbonyl oxygenO4 of T4 is also observed. However, the close contact, in which the distance between the hydroxylgroup of Tyr395 and O4 of T4 is less than 3.2 Å, isidentified only for short time-periods (1000 –1200 ps and 2250 – 2600 ps). The steric hindrancecaused by the 5-methyl group of T4 likely preventshydrogen bonding between the O4 carbonyl oxygen of T4 and the hydroxyl group of Tyr395.During dynamics the delta carbon CD2 of Tyr395is at 3.92 Å from the 5-methyl carbon C5M of T5 ,indicating hydrophobic interaction between thephenyl ring of Tyr395 and the 5-methyl group ofT5 (Table 3).In the MD structure of ETS – GGAG, a contactbetween the hydroxyl group of Tyr395 and N6atom of A3 is observed until about 2800 ps(Table 3). The Tyr395 hydroxyl group forms ahydrogen bond with the 4-amino nitrogen N4 ofC4 nucleobase. Unlike the ETS – GGAA complexwhere the 5-methyl group of T4 prevents hydrogenbonding, the C4 in ETS – GGAG is in contact withTyr395. The hydrophobic interaction between C5M000000000Motion of the helix-3 on the interface betweenETS domain and DNAContact of Tyr395 with the DNA0of T5 and the phenyl ring of Tyr395 is also seen inthe ETS – GGAG complex.In order to investigate the motion of the helix-3in the major groove of the DNA, the structureswere averaged at the intervals of 300 ps and analyzed. Some structures are superimposed according to the orientations of G2, A3 and A4/G4 (Figure9). The stereo pictures indicate that the positionand motion of the helix-3 in each complex arequite different. In the ETS –GGAA complex, thehelix-3 is settled in the major groove of the consensus DNA sequence without significant positionalfluctuations. The main-chain of the helix-3 (CA, Cand N atoms) is almost at the same position duringthe entire duration of MD simulation (Figure 9(a)),and the contact residues, Arg391, Arg394 andTyr395 show no change in conformation (Figure9(b)). In addition, overall the helix-3 region is similar in both X-ray and MD averaged structures (datanot shown).On the contrary, a distinct motion of the helix-3region can be seen in the ETS – GGAG complex(Figure 9(c) and (d)). The main-chain of the helix-3gradually moves as the simulation proceeds. Forexample, the Ca atom of Tyr395 in the MD structure averaged for 1200 –1500 ps (yellow structurein Figure 9(d)) shows a movement of 4.11 Å withrespect to the MD structure averaged for 3000 –3300 ps (blue structure in Figure 9(d)).Interaction between the ETS domain andphosphate backbone in the DNAEleven direct contacts between hydrogen donorsin the ETS domain and phosphate backbone in the
353MD Simulations of Ets-1 ETS Domain–DNA ComplexesFigure 9. Stereo diagrams of the helix-3 in the major groove of core DNA sequence in (a),(b) ETS– GGAA and (c),(d)ETS– GGAG MD structure. The MD structures averaged for the period 1200– 1500 ps (yellow), 1800– 2100 ps (green),2400– 2700 ps (gray) and 3000– 3300 ps (blue) are superimposed according to the orientations of G2, A3 and A4/G4. In(a) and (c), the backbone atoms (CA, C and N) of binding site (residues 386– 396) corresponds to the major grooveview of DNA (base-pairs 1 –5). In (b) and (d), the close-view of (a) and (c) perpendicular to the helix axis are given,with only few important protein residues (Arg391, Arg394 and Tyr395) and nucleobases (G1, G2, A3, A4/G4, T4 andT5 ) shown.00
354MD Simulations of Ets-1 ETS Domain–DNA ComplexesDNA are observed in the crystal structure of theETS –GGAA complex (PDB code: 1K79).5 Theseare listed in Table 2, along with the correspondingdistances of ETS – DNA MD complexes. Only slightdifferences (, 0.5 Å) in the contact distances of saltbridge formations between the ETS domain andphosphate backbone of DNA are observedbetween the crystal and the averaged MD structures. In ETS –GGAA, the contacts between the primary amino nitrogen NZ of Lys379 and thephosphate oxygen O1P of T4 and between NZ ofLys399 and O2P of A6 are not seen.As seen in Figure 8, the drastic changes areobserved in the hydrogen bonding patternbetween DNA base and the helix-3 region of theETS –GGAG MD structure. However, no considerable change in the direct contact of the turn andwinged region of the ETS domain (two ends of thehelix-3 region) with the DNA phosphate backboneis noticed in the MD averaged structures (900 –2100 and 2700– 3930 ps) of ETS –GGAG. This resultsuggests the possibility that the helix-3 works independently of the flanked regions in the ETSdomain– DNA interaction.in the bending angle in the range 5– 398 and 5 –428, respectively (Figure 5(a)). These results indicate that the differences in the DNA bendingangle are likely to arise due to flexibility of DNAhelix in the complexes.Analysis of Ets-1 ETS domain– DNA complex inX-ray structures has led to the proposal that thereduction in binding affinity of ETS –GGAG complex is due to the absence of van der Waals contacts with C4 , and the reduction in the van derWaals overlap between the 5-methyl group of T5and the phenyl ring of Tyr395.5 The possibility ofhydrogen bond formation between the hydroxylgroup in Tyr395 and N4 of C4 was also reportedin the literature.5 However, it was concluded thatthe contact would not significantly affect the binding energy. These explanations from X-ray analysisare not consistent with the MD simulations. Thedistance between the 5-methyl group of T5 andthe delta carbon CD2 of Tyr395 observed in theMD structure of ETS –GGAA are comparable tothat of ETS – GGAG (Table 3), indicating that thevan der Waals interaction between the 5-methylgroup of T5 and the phenyl ring of Tyr395 doesnot play an important role in the recognition ofthe GGAA core sequence. In addition, the hydroxylgroup of Tyr395 is found to be hydrogen bonded toN4 of C4 in the MD simulation of ETS –GGAG andis maintained during the entire simulation. However, in ETS – GGAA the contact of the hydroxylgroup with O4 of T4 is only observed intermittently. These results indicate the significance of thecontact between the hydroxyl group of Tyr395 andN4 of C4 . It is noteworthy that the specific recognition of the nucleobase 40 -position by ETS domainis observed in ETS –GGAG, but not in ETS – GGAA.0Discussion000000Comparison of the MD structures to theexperimental data0MD simulations on ETS – GGAA and ETS –GGAG clearly show the presence of meta-stablestates of hydrogen bonding in which the conservedresidues, Arg391 and Arg394, are participating(Figure 7). In the low affinity ETS – GGAG complex,both arginine residues change hydrogen bondingpartners after 2 ns MD simulation. In contrast,Arg394 in ETS –GGAA changes only its side-chainconformation preference at around 1.7 ns to formmore stable bidentate hydrogen bonds with thesame partner (Figure 8(d) and (f)). Thus, the arginine residues show a certain degree of conformational flexibility in the ETS domain– DNAcomplexes. These observations agree well with theresults from NMR experiments of Ets-112,13 andFli-1,11 in which the conserved arginine residues inthe ETS domain– DNA complexes were notassigned, and it was concluded that the Arg391and Arg394 did not have a single defined conformation in the complexes. Furthermore, the hydrogen bonding mode of these arginine residues inthe crystal structures of ETS domain –DNA complexes depends on the complex studied,5 – 10 whichalso supports the conformational flexibility of thearginine residues.In the crystal structures of ETS domain– DNAcomplexes, the DNA bending angle varies fromstructure to structure.5 – 10 The X-ray structure ofPU.1 –DNA complex shows a DNA bending angleof 88,9,10 while the value in the Ets-1 –DNA complexwas reported5 to be 2
served DNA-binding domain.3 The DNA-binding domain for ETS proteins, termed ETS domain, is about 85 amino acid residues in length and forms a winged helix-turn-helix motif consisting of three a-helices and four b-strands. The recent X-ray5-10 and NMR11-13 studies of the ETS domain-DNA complexes have shown that the helix-3 in the
To further quantify the SA2 binding specificity for DNA ends, we applied analysis the based on the fractional occupancies of SA2 at DNA ends (46). SA2 binding specificities for DNA ends (S DNA binding constant for specific sites/DNA binding constant for nonspecific sites K SP/K NSP) are 2945 ( 77), 2604 ( 68), and 2129 ( 76),
BCD Arg105 Within clamp-interacting helix 92.6 R105E Reduced clamp binding, eliminated DNA binding γ BCD Ser132 Before central helix 85.1 S132A Eliminated DNA binding γ BCD Arg133 Within central helix 48.0 R133A, R133E Reduced DNA binding γ BCD Lys161 Before SRC-containing helix 94.2 K161A, K161E Reduced DNA binding δ′
d Mutational disruption of DNA binding to XRCC1 impairs recruitment to DNA damage d Disruption of DNA binding by XRCC1 impairs repair of DNA single-strand breaks . observed perturbations upon DNA binding occurred in residues that were not strongly affected by PAR (Figure 2C), suggesting that the DNA and PAR molecules were binding to distinct .
The øX174 DNA binding protein contains two DNA binding domains, containing a series of DNA binding basic amino acids, separated by a proline-rich linker region. Within each DNA binding domain, there is a conserved glycine residue. Glycine and proline residues were mutated and the effects on virion structure were examined.
Discovering basic DNA-binding units A basic DNA-binding unit (DBU) is defined as a com-pact cluster of residues that is supposed to protrude into DNA grooves when a protein binds to DNA. The proposed method discovers DBUs by combining infor-mation of conservation, solvent accessibility, and DNA-binding propensity. Conserved residues are discovered
of DNA- and RNA-binding residues on the COMB_T dataset. 46 Figure 4.2. Comparison between the DNA and RNA machine learning (ML) consensus that targets combined prediction of DNA- and RNA-binding residues and the considered predictors of DNA- or RNA-binding residues on the COMB_T test
DNA, enabling us to envision how the binding of dimers is propagated along the length of the DNA. We have extended the ParA/Soj DNA binding work by identifying conserved positively charged residues in ParA from the plasmid pB171 that may be important for its DNA binding.
computational prediction of DNA- and RNA-binding residues from protein chains. These methods can be used to find the binding proteins in the vast sequence databases and to indicate sites of these interactions. Table 1 summarizes recent compara-tive reviews of the predictors of DNA-binding residues [13, 14] and RNA-binding residues [7, 15, 16].
