Genetic and crystallographic studies of the 3',5'-exonucleolytic site of DNA polymerase I.

Site-directed mutagenesis of the large fragment of DNA polymerase I (Klenow fragment) yielded two mutant proteins lacking 3',5'-exonuclease activity but having normal polymerase activity. Crystallographic analysis of the mutant proteins showed that neither had any alteration in protein structure other than the expected changes at the mutation sites. These results confirmed the presumed location of the exonuclease active site on the small domain of Klenow fragment and its physical separation from the polymerase active site. An anomalous scattering difference Fourier of a complex of the wild-type enzyme with divalent manganese ion and deoxythymidine monophosphate showed that the exonuclease active site has binding sites for two divalent metal ions. The properties of the mutant proteins suggest that one metal ion plays a role in substrate binding while the other is involved in catalysis of the exonuclease reaction.

9-beta-D-Arabinofuranosyl-8-n-butylaminoadenine, a C-8 substituted nucleoside in the anti conformation. Crystallographic and NMR studies.

The protein NMR spectrum of 9-beta-D-arabinofuranosyl-8-n-butylaminoadenine shows an unusually low-field 5'-hydroxyl proton resonance, which has been interpreted in terms of an anti glycosidic conformation together with an 05' ... N8 intramolecular hydrogen bond. Confirmatory evidence for this was obtained by an X-ray crystallographic study; in the crystal, the glycosidic angle chi is 52.7 degrees and the sugar pucker is C3' endo-C4' exo.

Crystal structure of a berenil-d(CGCAAATTTGCG) complex. An example of drug-DNA recognition based on sequence-dependent structural features.

The AT-selective drug berenil has been co-crystallized with the dodecanucleotide sequence d(CGCAAATTTGCG)2. The crystal structure has been solved to a resolution of 2.0 A and an R factor of 18.3%, with the location of 65 water molecules. The drug is symmetrically bound in the 5'-AATT region of the minor groove, with its amidinium groups hydrogen-bonding to O-2 atoms of the thymine base at each end of the binding site. This arrangement is distinct from that previously found for berenil with the sequence d(CGCGAATTCGCG)2, which has the drug bound to the sequencing 5'-ATT via hydrogen bonds to adenine N-3 atoms with the involvement of a bridging water molecule at one end of the binding site. The reasons for these differences are discussed in terms of changes in helical parameters; in particular propeller twist and base-pair roll are considered to be important. The conformational and base-pair geometry of the dodecanucleotide in the structure reported here, is closely similar to that for the native structure, suggesting that the 5'-AAATTT sequence does not significantly alter during drug binding, either because of its inflexibility or because its geometry is nearly ideal for berenil binding.

Redefinition of the cleavage sites of DNase I on the nucleosome core particle.

DNase I has been widely used for the footprinting of DNA-protein interactions including analyses of nucleosome core particle (NCP) structure. Our understanding of the relationship between the footprint and the structure of the nucleosome complex comes mainly from digestion studies of NCPs, since they have a well-defined quasi-symmetrical structure and have been widely investigated. However, several recent results suggest that the established consensus of opinion regarding the mode of digestion of NCPs by DNase I may be based on erroneous interpretation of results concerning the relationship between the NCP ends and the dyad axis. Here, we have used reconstituted NCPs with defined ends, bulk NCPs prepared with micrococcal nuclease and molecular modelling to reassess the mode of DNase I digestion. Our results indicate that DNase I cuts the two strands of the nucleosomal DNA independently with an average stagger of 4 nt with the 3'-ends protruding. The previously accepted value of 2 nt stagger is explained by the finding that micrococcal nuclease produces NCPs not with flush ends, but with approximately 1 nt 5'-recessed ends. Furthermore we explain why the DNA stagger is an even and not an odd number of nucleotides. These results are important for studies using DNase I to probe nucleosome structure in complex with other proteins or any DNA-protein complex containing B-form DNA. We also determine the origin of the 10n +/- 5 nt periodicity found in the internucleosomal ladder of DNase I digests of chromatin from various species. The explanation of the 10n +/- 5 nt ladder may have implications for the structure of the 30 nm fibre.

Purification and crystallization of thymidine kinase from herpes simplex virus type 1.

Thymidine kinase from herpes simplex virus type 1 (ATP:thymidine 5'-phosphotransferase; EC 2.7.1.21) has been purified from an overexpression system and crystallized against ammonium sulfate by using the hanging-drop technique. The tetragonal crystals are of space group P4122 or P4322, and have unit cell dimensions a = b = 84 A, c = 180 A.

Preparation of heavy-atom derivatives using site-directed mutagenesis. Introduction of cysteine residues into gamma delta resolvase.

The ability to determine protein structures by X-ray crystallography is often thwarted by the difficulty of finding isomorphous heavy-atom derivatives. The crystal structure of the site-specific recombinase, resolvase, has been difficult to determine for this reason. We have overcome this problem by introducing 13 single cysteine substitutions into the resolvase catalytic domain using oligonucleotide mutagenesis. The mutant proteins were screened for their ability to crystallize into the orthorhombic form and bind mercury ions isomorphously. Two mutant proteins provided excellent heavy-atom derivatives. This approach should be of general use and particularly helpful in cases where traditional methods have failed to produce a derivative.

Preliminary crystallographic data for NAD(P)H quinone reductase isolated from the Walker 256 rat carcinoma cell line.

An NAD(P)H quinone reductase isolated from Walker rat 256 carcinoma cells has been crystallized in a form suitable for high-resolution structural analysis. The crystals belong to orthorhombic space group P2(1)2(1)2(1) with cell parameters a = 168.15 A, b = 105.09 A and c = 67.38 A and contain four monomeric or two dimeric enzyme molecules per asymmetric unit. Diffraction extends beyond 2.3 A resolution.

NMR assignments and secondary structure of the UvrC binding domain of UvrB.

The 55 residue C-terminal domain of UvrB that interacts with UvrC during excision repair in Escherichia coli has been expressed and purified as a (His)6 fusion construct. The fragment forms a stable folded domain in solution. Heteronuclear NMR experiments were used to obtain extensive 15N, 13C and 1H NMR assignments. NOESY and chemical shift data showed that the protein comprises two helices from residues 630 to 648 and from 652 to 670. 15N relaxation data also show that the first 11 and last three residues are unstructured. The effective rotational correlation time within the structured region is not consistent with a monomer. This oligomerisation may be relevant to the mode of dimerisation of UvrB with the homologous domain of UvrC.

Structure to 1.9 A resolution of a complex with herpes simplex virus type-1 thymidine kinase of a novel, non-substrate inhibitor: X-ray crystallographic comparison with binding of aciclovir.

Treatment of herpes infections with nucleoside analogues requires as an initial step the activation of the compounds by thymidine kinase. As an aid to developing more effective chemotherapy, both for treatment of recurrent herpes infection and in gene therapy systems where thymidine kinase is expressed, two high-resolution X-ray structures of thymidine kinase have been compared: one with the relatively poor substrate aciclovir (Zovirax), the other with a synthetic inhibitor having an N2-substituted guanine. Both compounds have similar binding modes in spite of their size difference and apparently distinct ligand properties.

Crystal structure of Escherichia coli UvrB C-terminal domain, and a model for UvrB-uvrC interaction.

A crystal structure of the C-terminal domain of Escherichia coli UvrB (UvrB') has been solved to 3.0 A resolution. The domain adopts a helix-loop-helix fold which is stabilised by the packing of hydrophobic side-chains between helices. From the UvrB' fold, a model for a domain of UvrC (UvrC') that has high sequence homology with UvrB' has been made. In the crystal, a dimerisation of UvrB domains is seen involving specific hydrophobic and salt bridge interactions between residues in and close to the loop region of the domain. It is proposed that a homologous mode of interaction may occur between UvrB and UvrC. This interaction is likely to be flexible, potentially spanning > 50 A.

5-Substituted pyrimidines with a 1,5-anhydro-2, 3-dideoxy-D-arabino-hexitol moiety at N-1: synthesis, antiviral activity, conformational analysis, and interaction with viral thymidine kinase.

A new series of anhydrohexitol nucleosides are described. These compounds have a pyrimidine base moiety substituted in the 5-position with a chloro (1b), trifluoromethyl (1c), vinyl (1d), 2-thienyl (1e), ethynyl (1f) or propynyl (1g) substituent. The vinyl, propynyl, and, in particular, the 5-trifluoromethyl analogue showed potent activity against herpes simplex virus (HSV), 1c with a selectivity index of >16000 against HSV-1 and >1000 against HSV-2. Conformational analysis of anhydrohexitol nucleosides using computational methods indicates that these nucleosides occur in an equilibrium between the C1 and 1C form with a DeltaE of 5.9 kJ/mol. When the anhydrohexitol nucleoside is cocrystallized with the HSV-1 thymidine kinase it adopts a 1C conformation, which is opposite to the conformation found for the small molecule alone. The enzyme, apparently, induces a conformational change, and conformational flexibility of an anhydrohexitol nucleoside may be advantageous for recognition by viral enzymes.

Crystal structure of human DT-diaphorase: a model for interaction with the cytotoxic prodrug 5-(aziridin-1-yl)-2,4-dinitrobenzamide (CB1954).

The crystal structure of human DT-diaphorase (NAD(P)H oxidoreductase (quinone); EC 1.6.99.2) has been determined to 2.3 A resolution. There are only minor differences in shape and volume between the active sites of the rat and human enzymes and in the hydrophobic environment in the vicinity of the substrate. The isoalloxazine ring of the bound FAD is more buried in the human structure. Molecular modeling was used to examine optimal positions for the antitumor prodrug CB1954 (5-(aziridin-1-yl)-2,4-dinitrobenzamide) in both the human and rat enzyme active sites. This suggests that the position of CB1954 in the active site of the human enzyme is very similar to that in the rat, although there are detailed differences in the predicted patterns of hydrogen bonding between side chains and the drug. Some of the differences are a consequence of the shift in position for the FAD molecule and may contribute to the observed differences in rate of the two-electron reduction of CB1954.

Solution structure, hydrodynamics and thermodynamics of the UvrB C-terminal domain.

The solution structure, thermodynamic stability and hydrodynamic properties of the 55-residue C-terminal domain of UvrB that interacts with UvrC during excision repair in E. coli have been determined using a combination of high resolution NMR, ultracentrifugation, 15N NMR relaxation, gel permeation, NMR diffusion, circular dichroism and differential scanning calorimetry. The subunit molecular weight is 7,438 kDa., compared with 14.5+/-1.0 kDa. determined by equilibrium sedimentation, indicating a dimeric structure. The structure determined from NMR showed a stable dimer of anti-parallel helical hairpins that associate in an unusual manner, with a small and hydrophobic interface. The Stokes radius of the protein decreases from a high plateau value (ca. 22 A) at protein concentrations greater than 4 microM to about 18 A at concentrations less than 0.1 microM. The concentration and temperature-dependence of the far UV circular dichroism show that the protein is thermally stable (Tm ca. 71.5 degrees C at 36 microM). The simplest model consistent with these data was a dimer dissociating into folded monomers that then unfolds co-operatively. The van't Hoff enthalpy and dissociation constant for both transition was derived by fitting, with deltaH1=23 kJ mol(-1). K1(298)=0.4 microM and deltaH2= 184 kJ mol(-1). This is in good agreement with direct calorimetric analysis of the thermal unfolding of the protein, which gave a calorimetric enthalpy change of 181 kJ mol(-1) and a van't Hoff enthalpy change of 354 kJ mol(-1), confirming the dimer to monomer unfolding. The thermodynamic data can be reconciled with the observed mode of dimerisation. 15N NMR relaxation measurements at 14.1 T and 11.75 T confirmed that the protein behaves as an asymmetric dimer at mM concentrations, with a flexible N-terminal linker for attachment to the remainder of the UvrB protein. The role of dimerisation of this domain in the excision repair mechanism is discussed.

Crystal lattice packing is important in determining the bend of a DNA dodecamer containing an adenine tract.

The crystal structure of a DNA duplex dodecamer d(CGCAAAAATGCG) and its complementary strand has been determined at 2.6-A resolution. Although our goal was to deduce the structural features of the static bending of the helical axis exhibited by adenine-tract structures in solution, we conclude that the overall bend of 20 degrees in the direction of the major groove observed here arises from the forces associated with crystal packing. An isomorphous dodecamer brominated on one strand provides experimental evidence that this asymmetric sequence is positioned in two orientations in the crystal lattice that are related by a 180 degrees rotation around the pseudodyad axis of the sequence. The bend in these two differently positioned DNA molecules depends on their orientation in the crystal, not on their sequence. As with previously determined structures containing adenine tracts, the adenine and thymine base pairs are highly propeller twisted. The N-6 of the adenine comes within hydrogen bonding distance of the O-4 of thymine one step down the helix, facilitating the formation of a series of bifurcated hydrogen bonds within the adenine tract. The adenine tract is relatively straight and the bending is localized outside this region.

Threonine phosphorylation of the beta 3 integrin cytoplasmic tail, at a site recognized by PDK1 and Akt/PKB in vitro, regulates Shc binding.

The mechanism of outside-in signaling by integrins parallels that for growth factor receptors. In both pathways, phosphorylation of a cytoplasmic segment on tyrosine generates a docking site for proteins containing Src homology 2 (SH2) and phosphotyrosine binding domains. We recently observed that phosphorylation of a threonine (Thr-753), six amino acids proximal to tyrosine 759 in beta(3) of the platelet specific integrin alpha(IIb)beta(3), inhibits outside-in signaling through this receptor. We hypothesized that the presence of phosphothreonine 753 either renders beta(3) a poor substrate for tyrosine kinases or inhibits the docking capabilities of the tyrosyl-phosphorylated form of beta(3.) The first alternative was tested by comparing the phosphorylation of beta(3) model peptides by the tyrosine kinase pp60(c-src) and we found that the presence of a phosphate group on a residue corresponding to Thr-753 did not detectably alter the kinetics of tyrosine phosphorylation. However, the presence of phosphate on this threonine inhibited the binding of Shc to tyrosyl-phosphorylated beta(3) peptide. The inhibitory effect of the phosphate group could be mimicked by substituting an aspartic acid for Thr-753, suggesting that a negative charge at this position modulates the binding of Shc and possibly other phosphotyrosine binding domain- and SH2-containing proteins. A survey of several protein kinases revealed that Thr-753 was avidly phosphorylated by PDK1 and Akt/PKB in vitro. These observations suggest that activation of PDK1 and/or Akt/PKB in platelets may modulate the binding activity and/or specificity of beta(3) for signaling molecules.

Conformational analysis of the octamer d(G-G-C-C-G-G-C-C) in aqueous solution. A one-dimensional and two-dimensional proton NMR study at 300 MHz and 500 MHz.

Proton NMR studies in H2O and 2H2O were carried out on the self-complementary DNA octamer d(G-G-C-C-G-G-C-C) and compared with similar studies on the dimethylated analogue d(G-G-m5C-m5C-G-G-C-C) [Sanderson, M. R., Mellema, J.-R., van der Marel, G. A., Wille, G., van Boom, J. H. & Altona, C. (1983) Nucleic Acids Res. 11, 3333-3346]. Base, H1', H2' and H2" resonances were assigned by means of two-dimensional nuclear Overhauser spectroscopy (NOESY) and correlated spectroscopy (COSY) experiments. Chemical shift vs temperature profiles were used to analyze the temperature-dependent conformational behaviour and to extract thermodynamic parameters for the duplex-to-coil transition. Analysis of proton-proton couplings were used to discriminate between J1'2' and J1'2" and between the H2' and H2" resonances as well as to obtain conformational parameters of the sugar rings. From the NOESY and COSY experiments, the temperature profiles and the analysis of the coupling data it is concluded that: (a) the compound adopts a B-DNA-type helix in solution; (b) the sugar rings in the intact duplex display limited conformational freedom; (c) methylation of the cytosine bases at the C5 position has no measurable effect on the conformational behaviour of the octamer, nor on the conformation of the sugar rings; however, methylation does increase the stability of the duplex in aqueous solution under conditions of low salt concentration.

Crystal structure of a disulfide-linked "trefoil" motif found in a large family of putative growth factors.

Porcine pancreatic spasmolytic polypeptide (PSP) belongs to a large family of homologous growth factor-like polypeptides characterized by a disulfide-linked "trefoil motif," duplicated and conserved in various family members. PSP contains two trefoil motifs, has several pharmacological actions on the gut, and has growth factor properties on epithelial cells in vitro. The human PSP analogue, human spasmolytic polypeptide, appears to be involved in many regenerative situations and, especially, in healing gastrointestinal ulcers. One member of the trefoil family, pS2, is secreted in approximately 50% of estrogen-dependent human breast carcinomas, which has led to its use as a tumor prognostic marker. Both pS2 and human spasmolytic polypeptide are also widely expressed in chronic gastrointestinal ulcerative conditions such as Crohn disease. Here we report the three-dimensional structure at 2.6-A resolution of a trefoil-containing protein, namely PSP, purified from porcine pancreas. The structure shows two homologous domains that share a supersecondary structure and disulfide bond pattern. The two domains pack asymmetrically giving rise to a number of protruding loops, exposed clefts, and an unusual electrostatic surface potential. Knowledge of the structure of PSP should allow the design of mutants to investigate further the function of PSP and other trefoil-containing peptides.