Solution NMR studies of antiamoebin, a membrane channel-forming polypeptide

Antiamoebin I is a membrane-active peptaibol produced by fungi of the species Emericellopsis which is capable of forming ion channels in membranes. Previous structure determinations by x-ray crystallography have shown the molecule is mostly helical, with a deep bend in the center of the polypeptide, and that the backbone structure is independent of the solvent used for crystallization. In this study, the solution structure of antiamoebin was determined by NMR spectroscopy in methanol, a solvent from which one of the crystal structures was determined. The ensemble of structures produced exhibit a right-handed helical C terminus and a left-handed helical conformation toward the N-terminus, in contrast to the completely right-handed helices found in the crystal structures. The NMR results also suggest that a "hinge" region exists, which gives flexibility to the polypeptide in the central region, and which could have functional implications for the membrane insertion process. A model for the membrane insertion and assembly process is proposed based on the antiamoebin solution and crystal structures, and is contrasted with the assembly and insertion mechanism proposed for other ion channel-forming polypeptides.

Synthesis and function of 3-phosphorylated inositol lipids

The 3-phosphorylated inositol lipids fulfill roles as second messengers by interacting with the lipid binding domains of a variety of cellular proteins. Such interactions can affect the subcellular localization and aggregation of target proteins, and through allosteric effects, their activity. Generation of 3-phosphoinositides has been documented to influence diverse cellular pathways and hence alter a spectrum of fundamental cellular activities. This review is focused on the 3-phosphoinositide lipids, the synthesis of which is acutely triggered by extracellular stimuli, the enzymes responsible for their synthesis and metabolism, and their cell biological roles. Much knowledge has recently been gained through structural insights into the lipid kinases, their interaction with inhibitors, and the way their 3-phosphoinositide products interact with protein targets. This field is now moving toward a genetic dissection of 3-phosphoinositide action in a variety of model organisms. Such approaches will reveal the true role of the 3-phosphoinositides at the organismal level in health and disease.

Structural characterization of the N-terminal oligomerization domain of the bacterial chromatin-structuring protein, H-NS

The H-NS protein plays a key role in condensing DNA and modulating gene expression in bacterial nucleoids. The mechanism by which this is achieved is dependent, at least in part, on the oligomerization of the protein. H-NS consists of two distinct domains; the N-terminal domain responsible for protein oligomerization, and the C-terminal DNA binding domain, which are separated by a flexible linker region. We present a multidimensional NMR study of the amino-terminal 64 residues of H-NS (denoted H-NS1-64) from Salmonella typhimurium, which constitute the oligomerization domain. This domain exists as a homotrimer, which is predicted to be self-associated through a coiled-coil configuration. NMR spectra show an equivalent magnetic environment for each monomer indicating that the polypeptide chains are arranged in parallel with complete 3-fold symmetry. Despite the limited resonance dispersion, an almost complete backbone assignment for 1H(N), 1H(alpha), 15N, 13CO and 13C(alpha) NMR resonances was obtained using a suite of triple resonance experiments applied to uniformly 15N-, 13C/15N- and 2H/13C/15N-labelled H-NS1-64 samples. The secondary structure of H-NS1-64 has been identified on the basis of the analysis of 1H(alpha), 13C(alpha), 13Cbeta and 13CO chemical shifts, NH/solvent exchange rates, intra-chain H(N)-H(N) and medium-range nuclear Overhauser enhancements (NOEs). Within the context of the homotrimer, each H-NS1-64 protomer consists of three alpha-helices spanning residues 2-8, 12-20 and 22-53, respectively. A topological model is presented for the symmetric H-NS1-64 trimer based upon the combined analysis of the helical elements and the pattern of backbone amide group 15N nuclear relaxation rates within the context of axially asymmetric diffusion tensor. In this model, the longest of the three helices (helix 3, residues 22-53) forms a coiled-coil interface with the other chains in the homotrimer. The two shorter N-terminal helices fold back onto the outer surface of the coiled-coil core and potentially act to stabilise this configuration.

Solution structure and backbone dynamics of the DNA-binding domain of mouse Sox-5

The fold of the murine Sox-5 (mSox-5) HMG box in free solution has been determined by multidimensional NMR using (15)N-labeled protein and has been found to adopt the characteristic twisted L-shape made up of two wings: the major wing comprising helix 1 (F10--F25) and helix 2 (N32--A43), the minor wing comprising helix 3 (P51--Y67) in weak antiparallel association with the N-terminal extended segment. (15)N relaxation measurements show considerable mobility (reduced order parameter, S(2)) in the minor wing that increases toward the amino and carboxy termini of the chain. The mobility of residues C-terminal to Q62 is significantly greater than the equivalent residues of non-sequence-specific boxes, and these residues show a weaker association with the extended N-terminal segment than in non-sequence boxes. Comparison with previously determined structures of HMG boxes both in free solution and complexed with DNA shows close similarity in the packing of the hydrophobic cores and the relative disposition of the three helices. Only in hSRY/DNA does the arrangement of aromatic sidechains differ significantly from that of mSox-5, and only in rHMG1 box 1 bound to cisplatinated DNA does helix 1 have no kink. Helix 3 in mSox-5 is terminated by P68, a conserved residue in DNA sequence-specific HMG boxes, which results in the chain turning through approximately 90 degrees.

The three-dimensional solution structure and dynamic properties of the human FADD death domain

FADD (also known as MORT-1) is an essential adapter protein that couples the transmembrane receptors Fas (CD95) and tumor necrosis factor receptor-1 (TNF-R1) to intracellular cysteine proteases known as caspases, which propagate and execute the programmed cell death-inducing signal triggered by Fas ligand (FasL, CD95L) and TNF. FADD contains 208 amino acid residues, and comprises two functionally and structurally distinct domains: an N-terminal death effector domain (DED) that promotes activation of the downstream proteolytic cascade through binding of the DED domains of procaspase-8; and a C-terminal death domain (DD). FADD-DD provides the site of FADD recruitment to death receptor complexes at the plasma membrane by, for example, interaction with the Fas receptor cytoplasmic death domain (Fas-DD), or binding of the TNF-R1 adapter molecule TRADD. We have determined the three-dimensional solution structure and characterised the internal polypeptide dynamics of human FADD-DD using heteronuclear NMR spectroscopy of (15)N and (13)C,(15)N-labelled samples. The structure comprises six alpha-helices joined by short loops and displays overall similarity to the death domain of the Fas receptor. The analysis of the dynamic properties reveals no evidence of contiguous stretches of polypeptide chain with increased internal motion, except at the extreme chain termini. A pattern of increased rates of amide proton solvent exchange in the alpha3 helix correlates with a higher degree of solvent exposure for this secondary structure element. The properties of the FADD-DD structure are discussed with respect to previously reported mutagenesis data and emerging models for FasL-induced FADD recruitment to Fas and caspase-8 activation.

Determination of pK(a) values of carboxyl groups in the N-terminal domain of rat CD2: anomalous pK(a) of a glutamate on the ligand-binding surface

The ligand-binding surface of the T-lymphocyte glycoprotein CD2 has an unusually high proportion of charged residues, and ionic interactions are thought to play a significant role in defining the ligand specificity and binding affinity of CD2 with the structurally homologous ligands CD48 (in rodents) and CD58 (in humans). The determination of the electrostatic properties of these proteins can therefore contribute to our understanding of structure-activity relationships for these adhesion complexes that underpin T-cell adhesion to antigen-presenting cells. In this study, we investigated the pH titration behavior of the carboxyl groups of the N-terminal domain of rat CD2 (CD2d1) using the chemical shifts of backbone amide nitrogen-15 ((15)N) and proton NMR resonances, and carboxyl carbon-13 ((13)C) signals. The analysis revealed the presence of a glutamate (Glu41) on the binding surface of rat CD2 with an unusually elevated acidity constant (pK(a) = 6.73) for CD2d1 samples at 1.2 mM concentration. pH titration of CD2d1 at low protein concentration (0.1 mM) resulted in a slight decrease of the measured pK(a) of Glu41 to 6.36. The ionization of Glu41 exhibited reciprocal interactions with a second glutamate (Glu29) in a neighboring location, with both residues demonstrating characteristic biphasic titration behavior of the carboxyl (13)C resonances. Measurements at pH 5.5 of the two-bond deuterium isotope shift for the (13)C carboxyl resonances for Glu41 and Glu29 [(2)DeltaC(delta)(O(epsilon)D) = 0.2 and 0.1 ppm, respectively] were consistent with the assignment of the anomalous pK(a) to Glu41, under the strong influence of Glu29. The characterization of single site mutations of CD2d1 residues Glu41 and Glu29 to glutamine confirmed the anomalous pK(a) for Glu41, and indicated that electrostatic interaction with the Glu29 side chain is a significant contributing influence for this behavior in the wild-type protein. The implications of these observations are discussed with respect to recent structural and functional analyses of the interaction of rat CD2 with CD48. In particular, CD2 Glu41 must be a candidate residue to explain the previously reported strong pH dependence of binding of these two proteins in vitro.

Backbone dynamics of the C-terminal SH2 domain of the p85alpha subunit of phosphoinositide 3-kinase: effect of phosphotyrosine-peptide binding and characterization of slow conformational exchange processes

The backbone dynamics of the C-terminal SH2 domain from the regulatory subunit p85alpha (p85alpha C-SH2) of phosphoinositide 3-kinase has been investigated in the absence of, and in complex with, a high-affinity phosphotyrosine-containing peptide ligand derived from the platelet-derived growth-factor receptor. (15)N R(1) and R(2) relaxation rates and steady-state [(1)H]-(15)N NOE values were measured by means of (1)H-(15)N correlated two-dimensional methods and were analyzed within the framework of the model-free formalism. Several residues in the BC loop and in the neighbouring secondary structural elements display fast local dynamics in the absence of phosphotyrosine peptide ligand as evidenced by below-average [(1)H]-(15)N NOE values. Furthermore, residue Gln41 (BC3) displays conformational exchange phenomena as indicated by an above-average R(2) relaxation rate. Upon binding of the phosphotyrosine peptide, the NOE values increase to values observed for regular secondary structure and the exchange contribution to the R(2) relaxation rate for Gln41 (BC3) vanishes. These observations indicate a loss of backbone flexibility upon ligand binding. Substantial exchange contributions for His56 (betaD4) and Cys57 (betaD5), which are known to make important interactions with the ligand, are attenuated upon ligand binding. Several residues in the betaD'-FB region and the BG loop, which contribute to the ligand binding surface of the protein, exhibit exchange terms which are reduced or vanish when the ligand is bound. Together, these observations suggest that ligand binding is accompanied by a loss of conformational flexibility on the ligand binding face of the protein. However, comparison with other SH2 domains reveals an apparent lack of consensus in the changes in dynamics induced by ligand binding. Exchange rates for individual residues were quantified in peptide-complexed p85alpha C-SH2 from the dependence of the exchange contributions on the CPMG delay in an R(2) series and show that peptide-complexed p85alpha C-SH2 is affected by multiple conformational exchange processes with exchange rate constants from 10(2) s(-1) to 7.10(3) s(-1). Mapping of the exchange-rate constants on the protein surface show a clustering of residues with similar exchange-rate constants and suggests that clustered residues are affected by a common predominant exchange process.

Oligomerization of the chromatin-structuring protein H-NS

H-NS is a major component of the bacterial nucleoid, involved in condensing and packaging DNA and modulating gene expression. The mechanism by which this is achieved remains unclear. Genetic data show that the biological properties of H-NS are influenced by its oligomerization properties. We have applied a variety of biophysical techniques to study the structural basis of oligomerization of the H-NS protein from Salmonella typhimurium. The N-terminal 89 amino acids are responsible for oligomerization. The first 64 residues form a trimer dominated by an alpha-helix, likely to be in coiled-coil conformation. Extending this polypeptide to 89 amino acids generated higher order, heterodisperse oligomers. Similarly, in the full-length protein no single, defined oligomeric state is adopted. The C-terminal 48 residues do not participate in oligomerization and form a monomeric, DNA-binding domain. These N- and C-terminal domains are joined via a flexible linker which enables them to function independently within the context of the full-length protein. This novel mode of oligomerization may account for the unusual binding properties of H-NS.

Alternative modes of binding of proteins with tandem SH2 domains

The issue of specificity in tyrosine kinase intracellular signaling mediated by src homology 2 (SH2) domains has great importance in the understanding how individual signals maintain their mutual exclusivity and affect downstream responses. Several proteins contain tandem SH2 domains that, on interacting with their ligand, provide a higher level of specificity than can be afforded by the interaction of a single SH2 domain. In this study, we focus on the comparison of two proteins ZAP70 and the p85 subunit of PI 3-kinase, which although distinctly different in function and general structure, possess tandem SH2 domains separated by a linker region and which bind to phosphorylated receptor molecules localized to the cell membrane. Binding studies using isothermal titration calorimetry show that these two proteins interact with peptides mimicking their physiological ligands in very different ways. In the case of the SH2 domains from ZAP70, they interact with a stoichiometry of unity, while p85 is able to make two distinct interactions, one with a stoichiometry of 1:1 and the other with two p85 molecules interacting with one receptor. The observation of two different modes of binding of p85 might be important in providing different cellular responses based on fluctuating intracellular concentration regimes of this protein. Thermodynamic data on both proteins suggest that a conformational change occurs on binding. On investigation of this structural change using a truncated form of p85 (including just the two SH2 domains and the inter-SH2 region), both NMR and circular dichroism spectroscopic studies failed to show significant changes in secondary structure. This suggests that any conformational change associated with binding is small and potentially limited to loop regions of the protein.

Peripheral membrane proteins: FYVE sticky fingers

The recently determined structure of the lipid-binding 'FYVE' domain provides several clues to the mode of interaction for this class of peripheral membrane proteins. However, the application of traditional modes of structural analysis to diffusible membrane-binding proteins exposes some limitations of these techniques.

Crystal structure of the C-terminal SH2 domain of the p85alpha regulatory subunit of phosphoinositide 3-kinase: an SH2 domain mimicking its own substrate

The binding properties of Src homology-2 (SH2) domains to phosphotyrosine (pY)-containing peptides have been studied in recent years with the elucidation of a large number of crystal and solution structures. Taken together, these structures suggest a general mode of binding of pY-containing peptides, explain the specificities of different SH2 domains, and may be used to design inhibitors of pY binding by SH2 domain-containing proteins. We now report the crystal structure to 1.8 A resolution of the C-terminal SH2 domain (C-SH2) of the P85alpha regulatory subunit of phosphoinositide 3-kinase (PI3 K). Surprisingly, the carboxylate group of Asp2 from a neighbouring molecule occupies the phosphotyrosine binding site and interacts with Arg18 (alphaA2) and Arg36 (betaB5), in a similar manner to the phosphotyrosine-protein interactions seen in structures of other SH2 domains complexed with pY peptides. It is the first example of a non-phosphate-containing, non-aromatic mimetic of phosphotyrosine binding to SH2 domains, and this could have implications for the design of substrate analogues and inhibitors. Overall, the crystal structure closely resembles the solution structure, but a number of loops which demonstrate mobility in solution are well defined by the crystal packing. C-SH2 has adopted a binding conformation reminiscent of the ligand bound N-terminal SH2 domain of PI3K, apparently induced by the substrate mimicking of a neighbouring molecule in the crystal.

NMR exchange broadening arising from specific low affinity protein self-association: analysis of nitrogen-15 nuclear relaxation for rat CD2 domain 1

Nuclear spin relaxation monitored by heteronuclear NMR provides a useful method to probe the overall and internal molecular motion for biological macromolecules over a variety of time scales. Nitrogen-15 NMR relaxation parameters have been recorded for the N-terminal domain of the rat T-cell antigen CD2 (CD2d1) in a dilution series from 1.20 mM to 40 microM (pH 6.0, 25 degrees C). The data have been analysed within the framework of the model-free formalism of Lipari and Szabo to understand the molecular origin of severely enhanced transverse relaxation rates found for certain residues. These data revealed a strong dependence of the derived molecular correlation time tau c upon the CD2d1 protein concentration. Moreover, a number of amide NH resonances exhibited exchange broadening and chemical shifts both strongly dependent on protein concentration. These amide groups cluster on the major beta-sheet surface of CD2d1 that coincides with a major lattice contact in the X-ray structure of the intact ectodomain of rat CD2. The complete set of relaxation data fit well to an equilibrium monomer-dimer exchange model, yielding estimates of exchange rate constants (kON = 5000 M-1 s-1; kOFF = 7 s-1) and a dissociation constant (KD approximately 3-6 mM) that is consistent with the difficulty in detecting the weak interactions for this molecule by alternative biophysical methods. The self-association of CD2d1 is essentially invariant to changes in buffer composition and ionic strength and the associated relaxation phenomena cannot be explained as a result of neglecting anisotropic rotational diffusion in the analysis. These observations highlight the necessity to consider low affinity protein self-association interactions as a source of residue specific exchange phenomena in NMR spectra of macromolecular biomolecules, before the assignment of more elaborate intramolecular conformational mechanisms.

Structural and biochemical evaluation of the interaction of the phosphatidylinositol 3-kinase p85alpha Src homology 2 domains with phosphoinositides and inositol polyphosphates

Src homology 2 (SH2) domains exist in many intracellular proteins and have well characterized roles in signal transduction. SH2 domains bind to phosphotyrosine (Tyr(P))-containing proteins. Although tyrosine phosphorylation is essential for protein-SH2 domain interactions, the binding specificity also derives from sequences C-terminal to the Tyr(P) residue. The high affinity and specificity of this interaction is critical for precluding aberrant cross-talk between signaling pathways. The p85alpha subunit of phosphoinositide 3-kinase (PI 3-kinase) contains two SH2 domains, and it has been proposed that in competition with Tyr(P) binding they may also mediate membrane attachment via interactions with phosphoinositide products of PI 3-kinase. We used nuclear magnetic resonance spectroscopy and biosensor experiments to investigate interactions between the p85alpha SH2 domains and phosphoinositides or inositol polyphosphates. We reported previously a similar approach when demonstrating that some pleckstrin homology domains show binding specificity for distinct phosphoinositides (Salim, K., Bottomley, M. J., Querfurth, E., Zvelebil, M. J., Gout, I., Scaife, R., Margolis, R. L., Gigg, R., Smith, C. I., Driscoll, P. C., Waterfield, M. D., and Panayotou, G. (1996) EMBO J. 15, 6241-6250). However, neither SH2 domain exhibited binding specificity for phosphoinositides in phospholipid bilayers. We show that the p85alpha SH2 domain Tyr(P) binding pockets indiscriminately accommodate phosphoinositides and inositol polyphosphates. Binding of the SH2 domains to Tyr(P) peptides was only poorly competed for by phosphoinositides or inositol polyphosphates. We conclude that these ligands do not bind p85alpha SH2 domains with high affinity or specificity. Moreover, we observed that although wortmannin blocks PI 3-kinase activity in vivo, it does not affect the ability of tyrosine-phosphorylated proteins to bind to p85alpha. Consequently phosphoinositide products of PI 3-kinase are unlikely to regulate signaling through p85alpha SH2 domains.

Cross-restriction of a T cell clone to HLA-DR alleles associated with rheumatoid arthritis: clues to arthritogenic peptide motifs

OBJECTIVE

To identify distinctive sequence motifs required for productive peptide presentation by those HLA-DR alleles/DR4 subtypes that predispose to rheumatoid arthritis (RA).

METHODS

We tested 10 different HLA-DR4 subtypes for presentation of acetylcholine receptor peptides to 8 different DR4-restricted T cell lines/clones in proliferation assays.

RESULTS

Seven of the 8 T cells depended absolutely on either the autologous Lys71 (in Dw4) or Arg71 (e.g., Dw14), despite these alleles' similar charge and RA associations. In contrast, the PM-A T cell was only mildly affected by this interchange. Moreover, after minor modifications, peptides were presented to this unusual T cell preferentially by all the RA-associated subtypes of DR4 as well as by 2 other DR alleles (DR1 and DR1402) that predispose to RA.

CONCLUSION

This coincident cross-restriction to all the RA-associated HLA-DR alleles except DR10 shows that there could even be a single arthritogenic peptide; we now suggest a possible consensus motif.

Intermolecular interactions of the p85alpha regulatory subunit of phosphatidylinositol 3-kinase

The regulatory subunit of phosphatidylinositol 3-kinase, p85, contains a number of well defined domains involved in protein-protein interactions, including an SH3 domain and two SH2 domains. In order to investigate in detail the nature of the interactions of these domains with each other and with other binding partners, a series of deletion and point mutants was constructed, and their binding characteristics and apparent molecular masses under native conditions were analyzed. The SH3 domain and the first proline-rich motif bound each other, and variants of p85 containing the SH3 and BH domains and the first proline-rich motif were dimeric. Analysis of the apparent molecular mass of the deletion mutants indicated that each of these domains contributed residues to the dimerization interface, and competition experiments revealed that there were intermolecular SH3 domain-proline-rich motif interactions and BH-BH domain interactions mediating dimerization of p85alpha both in vitro and in vivo. Binding of SH2 domain ligands did not affect the dimeric state of p85alpha. Recently, roles for the p85 subunit have been postulated that do not involve the catalytic subunit, and if p85 exists on its own we propose that it would be dimeric.

Endocytosis: How dynamin sets vesicles PHree!

The dynamin GTPase is required for clathrin-dependent, receptor-mediated endocytosis. Exciting new studies have shown that dynamin's pleckstrin homology domain binds to phosphatidylinositol 4, 5-bisphosphate in vivo, thus localising dynamin directly at the plasma membrane and ultimately enabling vesiculation.

Scanning a DRB3*0101 (DR52a)-restricted epitope cross-presented by DR3: overlapping natural and artificial determinants in the human acetylcholine receptor

A recurring epitope in the human acetylcholine receptor (AChR) alpha subunit (alpha146-160) is presented to specific T cells from myasthenia gravis patients by HLA-DRB3*0101-"DR52a"-or by DR4. Here we first map residues critical for DR52a in this epitope by serial Ala substitution. For two somewhat similar T cells, this confirms the recently deduced importance of hydrophobic "anchor" residues at peptide p1 and p9; also of Asp at p4, which complements this allele's distinctive Arg74 in DRbeta. Surprisingly, despite the 9 sequence differences in DRbeta between DR52a and DR3, merely reducing the bulk of the peptide's p1 anchor residue (Trp149-->Phe) allowed maximal cross-presentation to both T cells by DR3 (which has Val86 instead of Gly). The shared K71G73R74N77 motif in the alpha helices of DR52a and DR3 thus outweighs the five differences in the floor of the peptide-binding groove. A second issue is that T cells selected in vitro with synthetic AChR peptides rarely respond to longer Ag preparations, whereas those raised with recombinant subunits consistently recognize epitopes processed naturally even from whole AChR. Here we compared one T cell of each kind, which both respond to many overlapping alpha140-160 region peptides (in proliferation assays). Even though both use Vbeta2 to recognize peptides bound to the same HLA-DR52a in the same register, the peptide-selected line nevertheless proved to depend on a recurring synthetic artifact-a widely underestimated problem. Unlike these contaminant-responsive T cells, those that are truly specific for natural AChR epitopes appear less heterogeneous and therefore more suitable targets for selective immunotherapy.

Structural analysis of the CD5 antigen--expression, disulphide bond analysis and physical characterisation of CD5 scavenger receptor superfamily domain 1

CD5 is a type-I transmembrane glycoprotein found on thymocytes, T-cells and a subset of B-cells. The extracellular region consists of three domains belonging to the scavenger receptor cysteine-rich (SRCR) superfamily, for which no three-dimensional structure has been obtained. Recombinant soluble CD5 domain 1 (CD5d1), the N-terminal SRCR domain, has been expressed in both chinese hamster ovary (CHO) cells and Pichia pastoris. CD5d1 was shown to be correctly folded by binding to the CD5 monoclonal antibody Leul. Circular dichroism and NMR analyses indicate that CD5d1 has a high beta-sheet content. CD5d1 from both CHO cells and P. pastoris have very similar properties. The disulphide bonding pattern was determined and is consistent with that found for the group-A SRCR domain of type-1 macrophage scavenger receptor and MARCO, the macrophage receptor with collagenous structure. Observations have been made of the role of glycosylation of CD5. P. pastoris expression provides large quantities of correctly folded recombinant CD5d1 for multidimensional NMR and for X-ray crystallographic studies. The whole extracellular region of CD5, expressed as a chimaera with rat CD4 domains 3 and 4 (cCD5d1-3-CD4d3+4), was studied by electron microscopy and carbohydrate analysis to gain an overview of the structure of the extracellular portion of intact CD5. Carbohydrate analysis identified N-linked glycans on CD5 domains 1 and 2, and sialylated O-linked glycans on the linker peptide between domains 1 and 2. Electron microscopy and carbohydrate analysis together suggest that the extracellular region of CD5 forms a rod-like structure with domain 1 distal from the cell surface and separated from domains 2 and 3 by an O-glycosylated peptide linker region.

NMR analysis of the N-terminal SRCR domain of human CD5: engineering of a glycoprotein for superior characteristics in NMR experiments

CD5 is a type-I transmembrane glycoprotein found on thymocytes, T-cells and a subset of B-cells. The extracellular region consists of three domains belonging to the scavenger receptor cysteine-rich (SRCR) superfamily for which the three-dimensional polypeptide fold is as yet unknown. Glycosylated CD5 domain 1 (CD5d1) has been obtained by expression by secretion from both Chinese hamster ovary (CHO) cells and Pichia pastoris. Recombinant CD5d1 expressed in this manner was shown to be correctly folded by binding to anti-CD5 L17F12/Leu1 monoclonal antibody. Preliminary nuclear magnetic resonance (NMR) spectra obtained for CD5d1 (residues 1-118) had spectral dispersion typical of a folded protein, but otherwise of such poor quality that NMR structural studies were not feasible. The analysis of glycoproteins by NMR is frustrated by sample heterogeneity and poor spectral quality associated with glycan resonance overlap and the potential for increased line-widths due to the large hydrodynamic volume. In order to pursue NMR structural studies of CD5d1 it was necessary to optimize the quality of NMR spectra of CD5d1. A range of constructs of varying length and carbohydrate content were expressed in CHO cells and in P. pastoris. In addition the P. pastoris CD5d1 proved susceptible to N-glycan cleavage with endoglycosidase H. The protein products were characterised using size exclusion chromatography, NMR measurement of translational self-diffusion coefficients and two-dimensional 1H nuclear Overhauser effect spectroscopy experiments. Removal of an eight residue O-glycosylated C-terminal peptide, in particular, resulted in significant improvements in the quality of the CD5d1 NMR data, while retaining native protein structure. Two-dimensional heteronuclear NMR spectroscopy of nitrogen-15 isotope labelled deglycosylated CD5d1 (residues 1-110) prepared from P. pastoris suggests that this protein product is now amenable to solution structure determination.

The energetics of HMG box interactions with DNA. Thermodynamic description of the box from mouse Sox-5

The structural energetics of the HMG box from the DNA-binding protein mouse Sox-5 were examined calorimetrically. It was found that this box, notwithstanding its small size (molecular mass about 10 kDa), does not behave as a single cooperative unit and, on heating, the box reversibly unfolds in two separate stages. The first transition (tt approximately 34 degrees C) involves about 40% of the total enthalpy and the second (tt approximately 46 degrees C) the remainder. Both transitions proceed with significant heat capacity increment, showing that they are associated with the unfolding of two sub-domains having non-polar cores. According to heat capacity, ellipticity, fluorescence and NMR criteria, this HMG box is in a fully compact native state only below 5 degrees C. HMG boxes consist of two approximately orthogonal wings: the minor wing comprises helix 3 and its associated antiparallel N-terminal strand, whilst the major wing is composed of helices I and II. Analysis of the fluorescence and NMR spectra for this box obtained at different temperatures shows that the lower melting transition can be assigned to the minor wing and the upper transition to the major wing. Under physiological conditions (37 degrees C), the minor wing is considerably unfolded, whilst the major wing is essentially fully folded. DNA binding in vivo therefore involves refolding of the minor wing.

Solution structure of the C-terminal SH2 domain of the p85 alpha regulatory subunit of phosphoinositide 3-kinase

Heterodimeric class IA phosphoinositide 3-kinase (PI 3-kinase) plays a crucial role in a variety of cellular signalling events downstream of a number of cell-surface receptor tyrosine kinases. Activation of the enzyme is effected in part by the binding of two Src homology-2 domains (SH2) of the 85 kDa regulatory subunit to specific phosphotyrosine-containing peptide motifs within activated cytoplasmic receptor domains. The solution structure of the uncomplexed C-terminal SH2 (C-SH2) domain of the p85 alpha subunit of PI 3-kinase has been determined by means of multinuclear, double and triple-resonance NMR experiments and restrained molecular-dynamics simulated-annealing calculations. The solution structure clearly indicates that the uncomplexed C-SH2 domain conforms to the consensus polypeptide fold exhibited by other SH2 domains, with an additional short helical element at the N terminus. In particular, the C-SH2 structure is very similar to both the p85 alpha N-terminal SH2 domain (N-SH2) and the Src SH2 domain with a root mean square difference (rmsd) for 44 C alpha atoms of 1.09 and 0.89 A, respectively. The canonical BC, EF and BG loops are less well-defined by the experimental restraints and show greater variability in the ensemble of C-SH2 conformers. The lower level of definition in these regions may reflect the presence of conformational disorder, an interpretation supported by the absence or broadening of backbone and side-chain NMR resonances for some of these residues. NMR experiments were performed, where C-SH2 was titrated with phosphotyrosine-containing peptides corresponding to p85 alpha recognition sites in the cytoplasmic domain of the platelet-derived growth-factor receptor. The ligand-induced chemical-shift perturbations indicate the amino-acid residues in C-SH2 involved in peptide recognition follow the pattern predicted from homologous complexes. A series of C-SH2 mutants was generated and tested for phosphotyrosine peptide binding by surface plasmon resonance. Mutation of the invariant Arg36 (beta B5) to Met completely abolishes phosphopeptide binding. Mutation of each of Ser38, Ser39 or Lys40 in the BC loop to Ala reduces the affinity of C-SH2 for a cognate phosphopeptide, as does mutation of His93 (BG5) to Asn. These effects are consistent with the involvement of the BC loop and BG loops regions in ligation of phosphopeptide ligands. Mutation of Cys57 (beta D5) in C-SH2 to Ile, the corresponding residue type in the p85 alpha N-SH2 domain, results in a change in peptide binding selectivity of C-SH2 towards that demonstrated by p85 alpha N-SH2. This pattern of p85 alpha phosphopeptide binding specificity is interpreted in terms of a model of the p85 alpha/PDGF-receptor interaction.

The adaptability of Escherichia coli thioredoxin to non-conservative amino acid substitutions

The adaptability of Escherichia coli thioredoxin to the substitution of a series of non-natural amino acids has been investigated. Different thiosulfonated alkyl groups were inserted into the hydrophobic core of the protein in position 78 via disulfide bonding with a buried cysteine residue as previously described (Wynn R, Richards FM. 1993. Unnatural amino acid packing mutants of Escherichia coli thioredoxin produced by combined mutagenesis/chemical modification techniques. Protein Sci 2:395-403). The side chains added to the cysteine included methyl, ethyl, n-propyl, n-butyl, n-pentyl, and cyclo-pentyl derivatives. The side chains appear to exploit the presence of the large cavities to incorporate these variant forms, enabling the protein to fold and have some activity. Solution structural and kinetic data suggested that these substitutions had little effect on the overall fold of the protein. Thermodynamic data revealed that the entropic effect of restricting the side chains in the folded protein has an effect on the stability. The variant forms of thioredoxin have different propensities to form dimers despite the limited structural perturbations. Molecular modeling studies allow the conformation of the side chains to be assessed.

Distinct specificity in the recognition of phosphoinositides by the pleckstrin homology domains of dynamin and Bruton's tyrosine kinase

Pleckstrin homology (PH) domains may act as membrane localization modules through specific interactions with phosphoinositide phospholipids. These interactions could represent responses to second messengers, with scope for regulation by soluble inositol polyphosphates. A biosensor-based assay was used here to probe interactions between PH domains and unilamellar liposomes containing different phospholipids and to demonstrate specificity for distinct phosphoinositides. The dynamin PH domain specifically interacted with liposomes containing phosphatidylinositol-4,5-bisphosphate [PI(4,5)P2] and, more weakly, with liposomes containing phosphatidylinositol-4-phosphate [PI(4)P]. This correlates with phosphoinositide activation of the dynamin GTPase. The functional GTPase of a dynamin mutant lacking the PH domain, however, cannot be activated by PI(4,5)P2. The phosphoinositide-PH domain interaction can be abolished selectively by point mutations in the putative binding pocket predicted by molecular modelling and NMR spectroscopy. In contrast, the Bruton's tyrosine kinase (Btk)PH domain specifically bound liposomes containing phosphatidylinositol-3,4,5-trisphosphate [PI(3,4,5)P3]: an interaction requiring Arg28, a residue found to be mutated in some X-linked agammaglobulinaemia patients. A rational explanation for these different specificities is proposed through modelling of candidate binding pockets and is supported by NMR spectroscopy.

NMR analysis of interacting soluble forms of the cell-cell recognition molecules CD2 and CD48

The T cell glycoprotein, CD2, is one of the best characterized molecules mediating recognition at the cell surface. The ligands of murine and human CD2 are CD48 and CD58, respectively, and interactions between these molecules have been shown to influence antigen recognition and T cell activation. The CD58 binding site of human CD2 has been characterized in mutational studies, and here we use heteronuclear NMR spectroscopy to identify the rat CD48 binding site of the N-terminal domain of rat CD2 (CD2d1). The NMR spectrum of bacterially expressed CD2d1, assigned initially at pH 4.3 in the course of determining the three-dimensional solution structure of this domain [Driscoll, P.C., et al. (1991) Nature 353, 762-765], has been reassigned as a two-dimensional 15N-1H heteronuclear single-quantum coherence (HSQC) spectrum at neutral pH. The CD48 binding surface was identified by monitoring perturbations in the line widths and chemical shifts of cross peaks in the HSQC spectrum of CD2d1 during titrations with a soluble form of CD48 expressed in Chinese hamster ovary cells. This first solution NMR analysis of interacting cell surface molecules shows that the ligand binding site extends across an area of ca. 700-800 A2 of the GFCC'C" face corresponding almost exactly to lattice contacts in crystals of soluble CD2 first proposed as a model of the interaction of CD2 with its ligands. The analysis finds no evidence for any large-scale structural changes in domain 1 of CD2 to accompany CD48 binding. Comparisons of the human and rat CD2 ligand binding sites suggest that species- and ligand-specific binding may be determined by as few as three amino acid residues, corresponding to Thr37, Leu38, and Glu41 in rat CD2 (Lys42, Lys43, and Gln46 in human CD2).

Structural similarity between the pleckstrin homology domain and verotoxin: the problem of measuring and evaluating structural similarity

An unexpected structural similarity is described between the pleckstrin homology (PH) domain and verotoxin. This similarity has escaped detection primarily due to the differences in topology that exist between the two proteins. By comparing this result with two previously reported similarities for the PH domain, one with the lipocalins and another with the FK506 binding protein, we discuss the problems of measuring and assessing structural similarities.

The solution structure and backbone dynamics of the fibronectin type I and epidermal growth factor-like pair of modules of tissue-type plasminogen activator

BACKGROUND

The thrombolytic serine protease tissue-type plasminogen activator (t-PA) is a classical modular protein consisting of three types of domain in addition to the serine protease domain: F1 (homologous to fibronectin type I); G (epidermal growth factor-like) and kringle. Biochemical data suggest that the F1 and G modules play a major role in the binding of t-PA to fibrin and to receptors on hepatocytes.

RESULTS

We have derived the solution structure of the F1 and G pair of modules from t-PA by two- and three-dimensional NMR techniques, in combination with dynamical simulated annealing calculations. We have also obtained information about the molecule's backbone dynamics through measurement of amide 15N relaxation parameters.

CONCLUSIONS

Although the F1 and G modules each adopt their expected tertiary structure, the modules interact intimately to bury a hydrophobic core, and the inter-module linker makes up the third strand of the G module's major beta-sheet. The new structural results allow the interpretation of earlier mutational data relevant to fibrin-binding and hepatocyte-receptor binding.

The solution structure of the F42A mutant of human interleukin 2

Interleukin 2 (IL-2) is one of the major cytokines produced by T lymphocytes in response to antigen. It is a potent growth and differentiation factor for several cell-types and is structurally related to the four-helix bundle family of cytokines. Mutation of residue Phe42 to Ala abolishes binding to the alpha chain of the tri-partite IL-2 receptor. The three-dimensional structure of the F42A mutant IL-2 has been calculated by two dimensional NMR methods and compared to a structure of wild-type IL-2 determined by X-ray crystallography. The overall topology of the two structures is the same. The main differences between the structures are within the ill-defined loops connecting the helices and the region of the protein that is believed to interact with the alpha-chain of the receptor. Thus, the mutation of Phe42 to Ala does not perturb the overall three-dimensional structure of IL-2, and does not appear to change the putative binding sites for the beta and gamma chains of the receptor. The structural differences observed in this mutant suggest that the replacement of Phe42 with Ala causes the re-orientation of neighbouring side-chains that are also involved in binding the alpha-chain of the receptor.

Crystal structure of an integrin-binding fragment of vascular cell adhesion molecule-1 at 1.8 A resolution

The cell-surface glycoprotein vascular cell adhesion molecule-1 (VCAM-1; ref. 1) mediates intercellular adhesion by specific binding to the integrin very-late antigen-4 (VLA-4, alpha 4 beta 1; ref. 3). VCAM-1, with the intercellular adhesion molecules ICAM-1, ICAM-2, ICAM-3 and the mucosal vascular addressin MAd-CAM-1, forms an integrin-binding subgroup of the immunoglobulin superfamily. In addition to their clinical relevance in inflammation, these molecules act as cellular receptors for viral and parasitic agents. The predominant form of VCAM-1 in vivo has an amino-terminal extracellular region comprising seven immunoglobulin-like domains. Functional studies have identified a conserved integrin-binding motif in domains 1 and 4, variants of which are present in the N-terminal domain of all members of the immunoglobulin superfamily subgroup. We report here the crystal structure of a VLA-4-binding fragment composed of the first two domains of VCAM-1. The integrin-binding motif (Q38IDSPL) is highly exposed and forms the N-terminal region of the loop between beta-strands C and D of domain 1. This motif exhibits a distinctive conformation which we predict will be common to all the integrin-binding IgSF molecules. These, and additional data, map VLA-4 binding to the face of the CFG beta-sheet, the surface previously identified as the site for intercellular adhesive interactions between members of the immunoglobulin superfamily.

Crystallization and preliminary X-ray diffraction characterisation of both a native and selenomethionyl VLA-4 binding fragment of VCAM-1

Soluble fragments of the extracellular region of vascular cell adhesion molecule 1 (VCAM-1) expressed in Escherichia coli retain functional adhesive activity. An integrin (VLA-4) binding fragment consisting of the N-terminal two immunoglobulin-like domains (VCAM-d1,2) has been crystallized. The crystals belong to space group P2(1)2(1)2(1) with cell dimensions of a = 52.7 A, b = 66.5 A, c = 113.2 A and contain two molecules in the crystallographic asymmetric unit. A batch of protein produced in the standard E. coli strain (HW1110), but grown in the presence of selenomethionine enriched media, showed 85% incorporation of selenium in place of sulphur at methionine residues. The selenomethionyl VCAM-d1,2 was crystallized by microseeding techniques initially using the native crystals for nucleation. Both native and selenomethionyl crystals diffract X-rays to a minimum Bragg spacing of 1.8 A.

Expression and characterisation of a very-late antigen-4 (alpha 4 beta 1) integrin-binding fragment of vascular cell-adhesion molecule-1

We have used an Escherichia coli expression system to produce forms of vascular cell-adhesion molecule-1 (VCAM-1) containing the first two and three supposed immunoglobulin-like domains. A form consisting of the first two domains of VCAM-1 is shown to promote very-late antigen-4-dependent spreading of a melanoma cell line comparable to that found for the equivalent region in the full seven-domain form. Preliminary structural analysis by CD and NMR is consistent with an immunoglobulin fold which is predicted from sequence comparison studies.

Three-dimensional solution structure of the pleckstrin homology domain from dynamin

BACKGROUND

The pleckstrin homology (PH) domain is a region of approximately 100 amino acids, defined by sequence similarity, that has been found in about 60 proteins, many of which are involved in signal transduction downstream of cell surface receptors; the function of PH domains is unknown. The only clue to the function of PH domains is the circumstantial evidence that they may link beta gamma subunits of G proteins to second messenger systems. Knowledge of the three-dimensional structures of PH domains should help to elucidate the roles they play in the proteins that contain them.

RESULTS

Using homonuclear and heteronuclear magnetic resonance spectroscopy, we have determined the solution structure of the PH domain of the GTPase dynamin, one of a number of proteins that have PH domains and interact with GTP. The fold of the dynamin PH domain is composed of two antiparallel beta-sheets, which pack face-to-face at an angle of approximately 60 degrees. The first beta-sheet comprises four strands (residues 13-58) from the amino-terminal half of the protein sequence; the second beta-sheet contains three strands (residues 63-99). A single alpha-helix (residues 102-116) flanks one edge of the interface between the two sheets, parallel in orientation to the second sheet, in an alpha/beta roll motif similar to that of the B oligomer of verotoxin-1 from Escherichia coli.

CONCLUSIONS

The structure of the dynamin PH domain is very similar to the recently reported structures of the pleckstrin and spectrin PH domains. This shows that, despite the low level of sequence similarity between different PH domains, they do have a characteristic polypeptide fold. On the basis of our structure, the suggestion that PH domains engage in coiled-coil interactions with G protein beta gamma subunits seems unlikely and should be re-evaluated.

DNA binding and bending properties of the post-meiotically expressed Sry-related protein Sox-5

Sox-5 is one of a family of genes which show homology to the HMG box region of the testis determining gene SRY. We have used indirect immunofluorescence to show that Sox-5 protein is localized to the nucleus of post-meiotic round spermatids in the mouse testis. In vitro footprinting and gel retardation assays demonstrate that Sox-5 binds specifically to the sequence AACAAT with moderately high affinity (Kd of approximately 10(-9) M). Moreover, interaction of Sox-5 with its target DNA induces a significant bend in the DNA, characteristic of HMG box proteins. Circular dichroism spectroscopy of the Sox-5 HMG box and its specific complex with DNA shows an alteration in the DNA spectrum, perhaps as a consequence of DNA bending, but none in the protein spectrum on complex formation. The dependence of the change in the CD spectrum with protein to DNA ratio demonstrates the formation of a 1:1 complex. Analysis of the structure of the Sox-5 HMG box by 2D NMR suggests that both the location of helical secondary structure as well as the tertiary structure is similar to that of HMG1 box 2.

High-resolution solution structure of reduced parsley plastocyanin

A high-resolution three-dimensional solution structure of parsley plastocyanin has been determined using 1H-NMR-derived data. An ensemble of 30 conformers has been calculated, exhibiting an atomic root mean square distribution about the mean coordinate positions of 0.37 +/- 0.03 A for backbone atoms and 0.75 +/- 0.04 A for all heavy atoms. (These values exclude residues 8-10 which are disordered.) The global fold of parsley plastocyanin is closely similar to those of other plastocyanins which have been structurally characterized by X-ray diffraction and NMR methods. However, deletion of residues at positions 57 and 58 of the consensus plastocyanin sequence causes elimination of a turn found in most higher plant plastocyanins. This turn is located in an acidic patch binding site, which consists of two clusters of acidic residues at positions 42-45 and 59-61. These residues surround the side chain of Tyr 83, which has been shown to be involved in binding of and electron transfer from cytochrome f, one of plastocyanin's physiological partners. The acidic recognition site is further disrupted in parsley plastocyanin by nonconservative substitution of two charged residues at positions 59 and 60. The NMR-derived structures show that E53, E85, and E95 compensate for these substitutions and give parsley plastocyanin an acidic recognition site of similar extent to that of other higher plant plastocyanins.(ABSTRACT TRUNCATED AT 250 WORDS)

NMR and crystallography--complementary approaches to structure determination

A knowledge of the three-dimensional structure of a protein is essential to understand how a protein performs its functions. It is also a prime requirement for the rational design of novel sequences with specific structural, chemical or catalytic properties. Until recently, such information could only be obtained from X-ray diffraction studies on protein crystals. In the past few years, however, nuclear magnetic resonance (NMR) spectroscopy in solution has rapidly become established as an effective alternative method. This review briefly examines the two techniques and their relevance to protein engineering and design.

Three-dimensional solution structure of the extracellular region of the complement regulatory protein CD59, a new cell-surface protein domain related to snake venom neurotoxins

The cell surface antigen CD59 is an inhibitor of complement-mediated lysis and a member of the Ly6 superfamily (Ly6SF) of cysteine-rich cell-surface molecules whose sequences are related to those of snake venom neurotoxins. The three-dimensional solution structure of a recombinant form of the extracellular region of the molecule (residues 1-70 of the mature protein; sCD59) has been solved by 2D NMR methods. sCD59 is a relatively flat, disk-shaped molecule consisting of a two-standed beta-sheet finger loosely packed against a protein core formed by a three-stranded beta-sheet and a short helix. Structure calculations allowed an unambiguous assignment of the disulfide-bonded cysteine pairs as 3-26, 6-13, 19-39, 45-63, and 64-69. The topology of sCD59 is similar to that of the snake venom neurotoxins and consistent with an evolutionary relationship existing between the Ly6SF and the neurotoxins.

1H and 15N nuclear magnetic resonance assignments, secondary structure in solution, and solvent exchange properties of azurin from Alcaligenes denitrificans

Complete sequential 1H and 15N resonance assignments for the reduced Cu(I) form of the blue copper protein azurin (M(r) = 14,000, 129 residues) from Alcaligenes denitrificans have been obtained at pH 5.5 and 32 degrees C using homo- and heteronuclear two-dimensional and heteronuclear three-dimensional NMR spectroscopy. Comparison of the resonance assignments for the backbone protons with those of Pseudomonas aeruginosa azurin, which is 68% homologous in its amino acid sequence and has a very similar three-dimensional structure, showed a high similarity in chemical shift positions. After adjustment for random coil contributions the mean difference in NH chemical shifts is 0.00 ppm (root mean square width = 0.30 ppm), whereas for C alpha protons the mean difference is 0.09 ppm (root mean square width = 0.23 ppm). Characteristic NOE connectivities and 3JHN alpha values were used to determine the secondary structure of azurin in solution. Two beta-sheets, one helix, and nine tight and four helical turns were identified, and some long-range NOE contacts were found that connect the helix with the beta-sheets. The secondary structure obtained is in agreement with the structure derived from X-ray diffraction data [Baker, E. N. (1988) J. Mol. Biol. 203, 1071-1095]. Studies of the hydration of the protein in the vicinity of the copper ligand residue His117 revealed that the solvent-exposed N epsilon 2 of His117 is in slow exchange with the bulk solvent. However, no evidence was obtained for the presence of a long-lived water molecule at the position corresponding to a well-defined water molecule observed in the crystal structures of A. denitrificans and Ps. aeruginosa azurin.

Secondary structure of fibronectin type 1 and epidermal growth factor modules from tissue-type plasminogen activator by nuclear magnetic resonance

A segment of human tissue-type plasminogen activator (t-PA) corresponding to the fibronectin type 1 (F1) and epidermal growth factor-like (G) pair of modules, residues 1-91, has been produced as a recombinant protein in Saccharomyces cerevisiae, with a single conservative Cys to Ser substitution. The sequence-specific assignment of the 1H and 15N nuclear magnetic resonances from the pair of modules has been completed using 2D 1H nuclear magnetic resonance (NMR) spectra in conjunction with 3D, 15N-edited, 1H and 2D 15N-1H NMR spectra. Slowly exchanging amide protons have been identified, and estimates of a number of backbone 3JNH-C alpha H coupling constants were obtained by line-shape-fitting. The secondary structure of the F1 module in the pair closely matches that previously determined for the isolated F1 module from t-PA, and that of the G module conforms to the "consensus" G module structure determined previously from several isolated G modules. In the module pair, the residues linking the two modules appear to form an extended beta-strand, the carboxy-terminal end of which makes up a third strand of the major beta-sheet of the G module. The intermodule interface is defined by NOEs between residues in the ranges 22-24 in the F1 module and 65-72 in the G module. The NMR data indicate that there is little or no reorientation of the two modules with respect to one another but rather that they combine with a fixed hydrophobic contact dominated by the side chain of leucine-22.

Mutational analysis of the CD2/CD58 interaction: the binding site for CD58 lies on one face of the first domain of human CD2

The adhesion interaction between the immunoglobulin superfamily molecules CD2 and CD58 (lymphocyte function-associated antigen 3) plays an important role in T cell and natural killer cell interaction with various antigen-presenting and target cells. Determination of the solution structure of rat CD2 domain 1 has allowed a model of human CD2 domain 1 to be generated, and a series of mutants based on this model have been made. Residues of domain 1 of human CD2 predicted to be solvent exposed were substituted with the equivalent residues present in the rat CD2 molecule. The ability of these mutants to mediate rosetting with human and sheep erythrocytes was studied. Results show that the binding site of CD2 for both human and sheep CD58 maps to the beta sheet containing beta strands CC'C"F and G. Residues K34 and E36 in beta strand C, R48 and K49 in beta strand C', and K91 and N92 in the loop connecting beta strands F and G are shown to be critical in the interaction. The data support the proposition that the interaction between CD2 and CD58 involves the major beta sheet face of CD2.

Secondary structure of a pair of fibronectin type 1 modules by two-dimensional nuclear magnetic resonance

The fourth and fifth type 1 module pair, corresponding to residues 151-244 from the amino terminus of human fibronectin, has been produced as a recombinant protein using a yeast expression system and studied by two-dimensional homonuclear 1H nuclear magnetic resonance (NMR) spectroscopy. The sequence-specific resonance assignment of the 1H NMR spectrum has been completed using a combination of 2D 1H nuclear Overhauser effect (NOE) spectroscopy, homonuclear Hartmann-Hahn, and correlated spectroscopy spectra recorded under a variety of pH and temperature conditions. Slow exchanging amide protons have been identified and estimates of many backbone 3JNH-C alpha H coupling constants were obtained by line shape fitting. The secondary structures of each module conform closely to the "consensus" fibronectin type 1 module structure determined previously for two other single type 1 modules. In the module pair described here, the two modules are linked by a short five-residue linker which appears to form a turn. The intermodule interface is defined by NOEs observed between a hydrophobic three-residue sequence from the fourth type 1 module and residues in the first double-stranded beta-sheet of the fifth type 1 module. The interaction is dominated by a tryptophan residue (unconserved in other type 1 sequences) within the fourth module, which causes large upfield ring current shifts for several proton resonances from the beta-sheet of the fifth module. The NMR data indicate that there is little or no relative reorientation of the two modules about the linker region but rather that the two modules combine with a fixed and intimate hydrophobic contact.

Solution structure of a DNA-binding domain from HMG1

We have determined the tertiary structure of box 2 from hamster HMG1 using bacterial expression and 3D NMR. The all alpha-helical fold is in the form of a V-shaped arrowhead with helices along two edges and one rather flat face. This architecture is not related to any of the known DNA binding motifs. Inspection of the fold shows that the majority of conserved residue positions in the HMG box family are those involved in maintaining the tertiary structure and thus all homologous HMG boxes probably have essentially the same fold. Knowledge of the tertiary structure permits an interpretation of the mutations in HMG boxes known to abrogate DNA binding and suggests a mode of interaction with bent and 4-way junction DNA.