Structure determination of the Antp (C39----S) homeodomain from nuclear magnetic resonance data in solution using a novel strategy for the structure calculation with the programs DIANA, CALIBA, HABAS and GLOMSA.

The structure of a mutant Antennapedia homeodomain, Antp(C39----S), from Drosophila melanogaster was determined using a set of new programs introduced in the accompanying paper. An input dataset of 957 distance constraints and 171 dihedral angle constraints was collected using two-dimensional n.m.r. experiments with the 15N-labeled protein. The resulting high quality structure for Antp(C39----S), with an average root-mean-square deviation of 0.53 A between the backbone atoms of residues 7 to 59 in 20 energy-refined distance geometry structures and the mean structure, is nearly identical to the previously reported structure of the wild-type Antp homeodomain. The only significant difference is in the connection between helices III and IV, which was found to be less kinked than was indicated by the structure determination for Antp. The main emphasis of the presentation in this paper is on a detailed account of the practical use of a novel strategy for the computation of nuclear magnetic resonance structures of proteins with the combined use of the programs DIANA, CALIBA, HABAS and GLOMSA.     

Secondary structure determination for the Antennapedia homeodomain by nuclear magnetic resonance and evidence for a helix-turn-helix motif.

The homeodomain encoded by the Antennapedia (Antp) gene of Drosophila was studied in aqueous solution by nuclear magnetic resonance (NMR). Sequence-specific resonance assignments have been obtained for the complete polypeptide chain of 68 amino acid residues. The secondary structure determined from nuclear Overhauser effects (NOE) and information about slowly exchanging amide protons includes three helical segments consisting of the residues 10-21, 28-38 and 42-52, respectively. Combination of the presently available NMR data with computer modeling provided preliminary evidence for the presence of a helix-turn-helix motif in the homeodomain. Near the turn, this supersecondary structure appears to be very similar to the DNA binding site in the 434 and P22 c2 repressors, but both helices in the homeodomain include 2-3 additional residues when compared with these prokaryotic DNA-binding proteins.     

Quantitative measurement of water diffusion lifetimes at a protein/DNA interface by NMR

Hydration site lifetimes of slowly diffusing water molecules at the protein/DNA interface of the vnd/NK-2 homeodomain DNA complex were determined using novel three-dimensional NMR techniques. The lifetimes were calculated using the ratios of ROE and NOE cross-relaxation rates between the water and the protein backbone and side chain amides. This calculation of the lifetimes is based on a model of the spectral density function of the water-protein interaction consisting of three timescales of motion: fast vibrational/rotational motion, diffusion into/out of the hydration site, and overall macromolecular tumbling. The lifetimes measured ranged from approximately 400 ps to more than 5 ns, and nearly all the slowly diffusing water molecules detected lie at the protein/DNA interface. A quantitative analysis of relayed water cross-relaxation indicated that even at very short mixing times, 5 ms for ROESY and 12 ms for NOESY, relay of magnetization can make a small but detectable contribution to the measured rates. The temperature dependences of the NOE rates were measured to help discriminate direct dipolar cross-relaxation from chemical exchange. Comparison with several X-ray structures of homeodomain/DNA complexes reveals a strong correspondence between water molecules in conserved locations and the slowly diffusing water molecules detected by NMR. A homology model based on the X-ray structures was created to visualize the conserved water molecules detected at the vnd/NK-2 homeodomain DNA interface. Two chains of water molecules are seen at the right and left sides of the major groove, adjacent to the third helix of the homeodomain. Two water-mediated hydrogen bond bridges spanning the protein/DNA interface are present in the model, one between the backbone of Phe8 and a DNA phosphate, and one between the side chain of Asn51 and a DNA phosphate. The hydrogen bond bridge between Asn51 and the DNA might be especially important since the DNA contact made by the invariant Asn51 residue, seen in all known homeodomain/DNA structures, is critical for binding affinity and specificity.     

Isolation and sequence-specific DNA binding of the Antennapedia homeodomain.

The homeodomain encoded by the Antennapedia (Antp) gene of Drosophila was overproduced in a T7 expression vector in Escherichia coli. The corresponding polypeptide of 68 amino acids was purified to homogeneity. The homeodomain was analysed by ultracentrifugation and assayed for DNA binding. The secondary structure of the isolated homeodomain was determined by nuclear magnetic resonance spectroscopy. DNA-binding studies indicate that the isolated homeodomain binds to DNA in vitro. It selectively binds to the same sites as a longer Antp polypeptide and a full-length fushi tarazu (ftz) protein. Therefore, the homeodomain represents the DNA-binding domain of the homeotic proteins.     

The three-dimensional structure of the vnd/NK-2 homeodomain-DNA complex by NMR spectroscopy.

The three-dimensional solution structure obtained by NMR of the complex formed between the uniformly singly15N and doubly13C/15N-labeled vnd/NK-2 homeodomain and its consensus 16 base-pair DNA binding sequence was determined. This work was carried out using the accepted repertoire of experiments augmented with a novel implementation of the water flipback technique to enhance signals from exchangeable amide protons. The results using this new technique confirm the existence of hydrogen bonding between the invariant Asn51 and the second adenine of the DNA binding sequence, as seen in crystal structures of other homeodomain-DNA complexes, but never before detected by NMR. Hydrogen bonding by Arg5 and Lys3 in the minor groove of the DNA appears to be responsible for two unusually upfield-shifted ribose H1' resonances. The DNA duplex is nearly straight and its structure is primarily that of B -DNA. A detailed comparison is presented for all available homeodomain-DNA structures including the vnd/NK-2 DNA complex, which demonstrates that homology is maintained in the protein structure, whereas for the orientation of the homeodomain relative to DNA, small but significant variations are observed. Interactions are described involving certain residues in specific positions of the homeodomain, namely Leu7, Thr41, and Gln50 of vnd/NK-2, where single amino acid residue mutations lead to dramatic developmental alterations. The availability of our previously determined three- dimensional structure of the vnd/NK-2 homeodomain in the absence of DNA allows us to assess structural changes in the homeodomain induced by DNA binding.     

Structural and dynamic aspects of non-intercalative (1:1) binding of a thiazole-lexitropsin to the decadeoxyribonucleotide d-[CGCAATTGCG]2: An 1H-NMR and molecular modeling study.

The location, orientation and dynamics of a thiazole-containing analogue of distamycin 1 bound to the decadeoxyribonucleotide d-[CGCAATTGCG]2 have been studied by non-exchangable and imino proton NMR resonances of the 1:1 complex. Using NOE difference, COSY and NOESY experiments, lexitropsin (1) was located in the minor groove of DNA at 5'-CAAT sequence. This was concluded by an intermolecular NOE between the ligand and a minor groove A4H2 proton. The NOE cross-correlations in the NOESY map confirmed that the DNA decamer duplex in the 1:1 complex remains in a right-handed B-conformation similar to that in the free decamer. Experiments on non-exchangeable and exchangeable proton NMR resonances placed the N-formylamino terminus of drug 1 on the 5'-C3 nucleotide, while the rest of the molecule extends onto the 5'-AAT sequence. The structural evidence for sequence preferential binding at 5'CAAT rather than 5'AATT suggests this reflects an attempt on the part of the sterically demanding inward directed sulfur of the thiazole to minimize compression by moving part of the molecule to the somewhat wider CG base site. The lack of evidence for a 2:1 drug:DNA complex, in contrast to distamycin, is in accord with this interpretation. The lexitropsin 1 was found to be in an exchange between the equivalent 5'-CAAT sites at a rate of approximately 35S-1 with a delta G degree of 65 +/- 5 kJ mol-1 at 303 K. The experimental data suggests a slide-swing mechanism for this exchange process.     

Hydration of DNA in aqueous solution: NMR evidence for a kinetic destabilization of the minor groove hydration of d-(TTAA)2 versus d-(AATT)2 segments.

The residence times of the hydration water molecules near the base protons of d-(GTGGAATTCCAC)2 and d-(GTGGTTAACCAC)2 were investigated by nuclear magnetic resonance (NMR) spectroscopy. Nuclear Overhauser effects (NOE) were observed between base protons of the DNA and hydration water in NOESY and ROESY experiments. Large positive NOESY cross peaks observed between the resonances of the water and the adenine 2H protons of the central d-(AATT)2 segment in the duplex d-(GTGGAATTCCAC)2 indicate the presence of a 'spine of hydration' with water molecules exhibiting residence times on the DNA longer than 1 nanosecond. In contrast, no positive intermolecular NOESY cross peaks were detected in the d-(TTAA)2 segment of the duplex d-(GTGGTTAACCAC)2, indicating that no water molecules bound with similarly long residence times occur in the minor groove of this fragment. These results can be correlated with the larger width of the minor groove in d-(TTAA)2 segments as compared to that in d-(AATT)2 segments, as observed previously in single crystal structures of related oligonucleotide duplexes in B type conformation. The present experiments confirm earlier experimental results from single crystal studies and theoretical predictions that a 5'-dTA-3' step in the nucleotide sequence interrupts the spine of hydration in the minor groove.     

Conformational changes of the BS2 operator DNA upon complex formation with the Antennapedia homeodomain studied by NMR with 13C/15N-labeled DNA.

The NMR structures have been determined for a 13C/15N doubly labeled 14 base-pair DNA duplex comprising the BS2 operator sequence both free in solution and in the complex with the Antennapedia homeodomain. The impact of the DNA labeling is assessed from comparison with a previous structure of the same complex that was determined using isotope labeling only for the protein. Differences between the two structure determinations are nearly completely limited to the DNA, which retains the global B -conformation of the free DNA also in the complex. Local protein-induced conformational changes are a narrowing of the minor groove due to the interaction with the N-terminal arm of the homeodomain, and changes of the sugar puckers of the deoxyriboses G5 and C6, which are apparently induced by van der Waals interactions with Tyr25, and with Gln50 and Arg53, respectively. The high conservation of these amino acid residues in homeodomains suggests that protein-induced shifts in some sugar puckers contribute to the affinity of homeodomains to their cognate DNA. The data obtained here with the Antennapedia homeodomain-DNA complex clearly show that nucleic acid isotope-labeling can support detailed conformational characterization of DNA in complexes with proteins, which will be indispensable for structure determinations of complexes containing globally distorted DNA conformations.     

1H NMR studies on the interaction between distamycin A and a symmetrical DNA dodecamer

High-resolution NMR techniques have been used to examine the structural and dynamical features of the interaction between distamycin A and the self-complementary DNA dodecamer duplex d-(CGCGAATTCGCG)2. The proton resonances of d(CGCGAATTCGCG)2 have been completely assigned by previous two-dimensional NMR studies [Hare, D. R., Wemmer, D. E., Chou, S. H., Drobny, G., & Reid, B. R. (1983) J. Mol. Biol. 171, 319-336]. Addition of the asymmetric drug molecule to the symmetric dodecamer leads to the formation of an asymmetric complex as evidenced by a doubling of DNA resonances over much of the spectrum. In two-dimensional exchange experiments, strong cross-peaks were observed between uncomplexed DNA and drug-bound DNA resonances, permitting direct assignment of many drug-bound DNA resonances from previously assigned free DNA resonances. Weaker exchange cross-peaks between formerly symmetry related DNA resonances indicate that the drug molecule flips head-to-tail on one duplex with half the frequency at which it leaves the DNA molecule completely. In experiments performed in H2O, nuclear Overhauser effects (NOEs) were observed from each drug amide proton to an adenine C2H and a pyrrole H3 ring proton. In two-dimensional nuclear Overhauser experiments performed on D2O solutions, strong intermolecular NOEs were observed between each of the three pyrrole H3 resonances of the drug and an adenine C2H resonance, with weaker NOEs observed between the drug H3 resonances and C1'H resonances. The combined NOE data allow us to position the distamycin A unambiguously on the DNA dodecamer, with the drug spanning the central AATT segment in the minor groove.     

Sequence-dependent recognition of DNA duplexes. Netropsin complexation to the AATT site of the d(G-G-A-A-T-T-C-C) duplex in aqueous solution

We have investigated intermolecular interactions and conformational features of the netropsin X d(G-G-A-A-T-T-C-C) complex by one- and two-dimensional NMR studies in aqueous solution. Netropsin removes the 2-fold symmetry of the d(G-G-A-A-T-T-C-C) duplex at the AATT binding site and to a lesser extent at adjacent dG X dC base pairs resulting in doubling of resonances for specific positions in the spectrum of the complex at 25 degrees C. We have assigned the amide, pyrrole, and CH2 protons of netropsin, and the base and sugar H1' protons of the nucleic acid from an analysis of the nuclear Overhauser effect (NOESY) and correlated (COSY) spectra of the complex at 25 degrees C. We observe intermolecular nuclear Overhauser effects (NOE) between all three amide and both pyrrole protons on the concave face of the antibiotic and the minor groove adenosine H2 proton of the two central A4 X T5 base pairs of the d(G1-G2-A3-A4-T5-T6-C7-C8) duplex. Weaker intermolecular NOEs are also observed between the pyrrole concave face protons and the sugar H1' protons of residues T5 and T6 in the AATT minor groove of the duplex. We also detect intermolecular NOEs between the guanidino CH2 protons at one end of netropsin and adenosine H2 proton of the two flanking A3 X T6 base pairs of the octanucleotide duplex. These studies establish a set of intermolecular contacts between the concave face of the antibiotic and the minor groove AATT segment of the d(G-G-A-A-T-T-C-C) duplex in solution. The magnitude of the NOEs require that there be no intervening water molecules sandwiched between the antibiotic and the DNA so that release of the minor groove spine of hydration is a prerequisite for netropsin complex formation.     

The interaction with DNA of wild-type and mutant fushi tarazu homeodomains.

The in vitro DNA binding properties of wild-type and mutant fushi tarazu homeodomains (ftz HD) have been analysed. The DNA binding properties of the ftz HD are very similar to those of the Antp HD. In interference experiments with mutant ftz HDs, close approaches between specific portions of the ftz HD peptide and specific regions of the binding site DNA were mapped. A methylation interference, G7 on the beta strand of BS2, is absent from the interference pattern with a mutant ftz HD [ftz (R43A) HD] in which the Arg43 at the second position of helix III (the recognition helix) is replaced by an Ala. This indicated that Arg43 of the ftz HD is in close proximity to the N7 of G7 of the beta strand of BS2 in the major groove. The methylation and ethylation interference patterns with the ftz (NTD) HD, in which the first six amino acids of the homeodomain were deleted, were extensively altered relative to the ftz HD patterns. Methylation of A11 and G12 of the alpha strand and ethylation of the phosphate of nucleotide A12 of the alpha strand no longer interfere with binding. This indicated that the first six amino acids of the homeodomain of ftz interact with A11 of the alpha strand in the minor groove, the phosphate of the nucleotide A13 on the alpha strand and G12 of the alpha strand in the adjacent major groove of BS2. In a binding study using a change of specificity mutation [ftz (Q50K) HD], in which the Gln50 at the ninth position of the third helix is exchanged for a Lys (as in the bicoid HD), and variant binding sites, we concluded that position 50 of the ftz HD and the ftz (Q50K) HD peptides interacts with base pairs at positions 6 and 7 of BS2. These three points of contact allowed us to propose a crude orientation of the ftz HD within the protein-DNA complex. We find that the ftz HD and the Antp HD peptides contact DNA in a similar way.     

Structure and dynamics of ligand-template interactions of topoisomerase inhibitory analogs of Hoechst 33258: high field 1H-NMR and restrained molecular mechanics studies.

The binding characteristics of Hoechst 33258 (1), a synthetic bis-benzimidazole, and its structural analog 2, with one of the benzimidazoles replaced by a pyridoimidazole, to the self-complementary decadeoxyribonucleotide sequences d(CGCAATTGCG)2 (A) and d-(CATGGCCATG)2 (B) respectively, were examined using high field 1H-NMR techniques. Selective complexation induced chemical shift changes, the presence of exchange signals and intermolecular NOE contacts between the ligands and the minor groove protons of the oligonucleotides suggest the preferred binding sites as the centrally located AATT segment for complex A1, and the CCAT segment for complex B2. The B-type conformations of the two DNA duplexes are preserved upon complexation, as confirmed by the 2D-NOESY based sequential connectivities involving DNA base and sugar protons. Close intermolecular NOE based contacts between the ligands and their respective DNA sequences were further refined to model the ligand-DNA complexes starting from the computer generated B-type structures for the oligonucleotides. Force field calculations of ligand-DNA interaction energies indicate a more favorable contribution from the van der Waals energy component in the case of complex A1 consistent with its stronger net binding compared with the complex B2. Overall, the incorporation of a pyridinic nitrogen in Hoechst 33258 structure alters its selectivity for base pair recognition from A.T to G.C, resulting largely from the formation of a hydrogen bond between the new basic center and the 2-NH2 group of a guanosine moiety. The rates for the exchange of ligands between the two equivalent binding sites (AATT for 1, and CCAT for 2) of the self-complementary DNA sequences, are estimated from analyses of coalescence of NMR signals to be 189s-1 at 301 K for A1 and 79s-1 at 297 K for B2; which correspond to delta G++ of 13.8 and 18.6 kcal.mol-1 respectively.     

Sequence specific molecular recognition and binding of a monocationic bis-imidazole lexitropsin to the decadeoxyribonucleotide d-[(GATCCGTATG).(CATACGGATC)]: structural and dynamic aspects of intermolecular exchange studied by 1H-NMR

The non-exchangeable and imino proton NMR resonances of the non self-complementary decadeoxyribonucleotide d-[(GATCCGTATG).(GATACGGATC)] as well as those of the 1:1 complex of the monocatonic bis-imidazole lexitropsin 1 to this sequence have been assigned by using a combination of NOE difference, COSY and NOESY techniques. Confirmation of complete annealing of the two non self-complementary decamer strands to give the duplex decadeoxyribonucleotide is obtained by the detection of ten imino protons. It is established that the sugar-base orientations of all the bases in the duplex decamer are anti. From NOE studies, it is concluded that the duplex oligomer is right-handed and adopts a conformation in solution that belongs to the B family. A population analysis reveals that the sugar moieties exist predominantly in the S-form (2'-endo-3'-exo). Addition of 1 to the DNA solution leads to doubling of the resonances for CH6(4,5), GH8(6), TH6(7) and T-CH3(7). The base, anomeric H1' and imino proton signals for the base sequence 5'-CCGT undergo the most marked drug-induced chemical shift changes. These results provide evidence that the lexitropsin is bound to the sequence 5'-CCGT in the minor groove of the DNA. NOE measurements between the amide protons (NH1 and NH4) and the imino proton (IV and V) signals confirmed the location and orientation of 1 in the 1:1 complex, with the imino terminus oriented to C(4). The specific binding of 1 to the sequence 5'-CCGT-3' deduced in this study is in agreement with the footprinting data obtained using the Hind III/Nci fragment from pBR322 DNA [Kissinger et al. 1987 (13)]. Intramolecular NOEs observed between H4 and H9 of the lexitropsin suggest that the molecule is not planar, but subjected to propeller twisting, in both the free and bound forms. Furthermore, NOE measurements permit assignment of the DNA duplex in the 1:1 complex to the B-form, which is similar to that of the free DNA. The [(T7A8T9).(A12T13A14)] segment of the DNA shows better stacking, by propeller twisting, compared to the rest of the molecule in the free as well as the complex forms. The intermolecular rate of exchange of 1 between the equivalent 5'-CCGT sites, at a concentration of 12 mM, is estimated to be approximately 88s-1 at 308 degrees K with delta G not equal to of 63 +/- 5 KJ mol-1.     

Distamycin-induced inhibition of homeodomain-DNA complexes.

The mobility shift assay was used to study the competition of the minor groove binder distamycin A with either an Antennapedia homeodomain (Antp HD) peptide or derivatives of a fushi tarazu homeodomain (ftz HD) peptide for their AT-rich DNA binding site. The results show that distamycin and the homeodomain peptides compete under the conditions: (i) preincubation of DNA with distamycin and subsequent addition of HD peptide; (ii) simultaneous incubation of DNA with distamycin and HD peptide; and (iii) preincubation of DNA with HD peptide and subsequent addition of distamycin. There is also competition when using a peptide which lacks the N-terminal arm of ftz HD that is involved in contacts in the minor groove. It is proposed that the protein's binding affinity is diminished by distamycin-induced conformational changes of the DNA. The feasibility of the propagation of conformational changes upon binding in the minor groove is also shown for the inhibition of restriction endonucleases differing in the AT content of their recognition site and of their flanking DNA sequences. Thus, it is demonstrated that minor groove binders can compete with the binding of proteins in the major groove, providing an experimental indication for the influence of biological activities exerted by DNA ligands binding in the minor groove.     

Dynamics and binding mode of Hoechst 33258 to d(GTGGAATTCCAC)2 in the 1:1 solution complex as determined by two-dimensional 1H NMR.

We have investigated the interaction of the bisbenzimidazole derivative Hoechst 33258 with the self-complementary dodecadeoxynucleotide duplex d(GTGGAATTCCAC)2 using one-dimensional (1D) and two-dimensional (2D) proton nuclear magnetic resonance (1H NMR) spectroscopy. To monitor the extent of complex formation, we used the imino proton region of the 1D 1H NMR spectra acquired in H2O solution. These spectra show that the DNA duplex loses its inherent C2v symmetry upon addition of the drug, indicating that the two molecules form a kinetically stable complex on the NMR time scale (the lifetime of the complex has been measured to be around 450 ms). We obtained sequence-specific assignments for all protons of the ligand and most protons of each separate strand of the oligonucleotide duplex using a variety of homonuclear 2D 1H NMR experiments. The aromatic protons of the DNA strands, which are symmetrically related in the free duplex, exhibit exchange cross peaks in the complex. This indicates that the drug binds in two equivalent sites on the 12-mer, with an exchange rate constant of 2.2 +/- 0.2 s-1. Twenty-five intermolecular NOEs were identified, all involving adenine 2 and sugar 1' protons of the DNA and protons in all four residues of the ligand, indicating that Hoechst 33258 is located in the minor groove at the AATT site. Only protons along the same edge of the two benzimidazole moieties of the drug show NOEs to DNA protons at the bottom of the minor groove. Using molecular mechanics, we have generated a unique model of the complex using distance constraints derived from the intermolecular NOEs. We present, however, evidence that the piperazine group may adopt at least two locally different conformations when the drug is bound to this dodecanucleotide.     

Determination of the three-dimensional structure of the Mrf2-DNA complex using paramagnetic spin labeling

Understanding the mechanism of protein-DNA interactions at the molecular level is one of the main focuses in structural and molecular biological investigations. At present, NMR spectroscopy is the only approach that can provide atomic details of protein-DNA recognition in solution. However, determining the structures of protein-DNA complexes using NMR spectroscopy has been dependent on the observation of intermolecular nuclear Overhauser effects (NOE) and their assignments, which are difficult to obtain in many cases. In this study, we have shown that intermolecular distance constraints derived from a single spin-label in combination with docking calculations have defined many specific contacts of the complex between the AT-rich interaction domain (ARID) of Mrf2 and its target DNA. Mrf2 contacts DNA mainly using the two flexible loops, L1 and L2. While the L1 loop contacts the phosphate backbone, L2 and several residues in the adjacent helices interact with AT base pairs in the major groove of DNA. Despite the structural diversity in the ARID family of DNA-binding proteins, Mrf2 maintains contacts with DNA similar to those observed in the homologous Dri-DNA complex.