Progress in fold recognition

The prediction experiment reveals that fold recognition has become a powerful tool in structural biology. We applied our fold recognition technique to 13 target sequences. In two cases, replication terminating protein and prosequence of subtilisin, the predicted structures are very similar to the experimentally determined folds. For the first time, in a public blind test, the unknown structures of proteins have been predicted ahead of experiment to an accuracy approaching molecular detail. In two other cases the approximate folds have been predicted correctly. According to the assessors there were 12 recognizable folds among the target proteins. In our postprediction analysis we find that in 7 cases our fold recognition technique is successful. In several of the remaining cases the predicted folds have interesting features in common with the experimental results. We present our procedure, discuss the results, and comment on several fundamental and technical problems encountered in fold recognition. © 1995 Wiley-Liss, Inc.     

Homology modeling by the ICM method

Five models have been built by the ICM method for the Comparative Modeling section of the Meeting on the Critical Assessment of Techniques for Protein Structure Prediction. The targets have homologous proteins with known three-dimensional structure with sequence identity ranging from 25 to 77%. After alignment of the target sequence with the related three-dimensional structure, the modeling procedure consists of two subproblems: side-chain prediction and loop prediction. The ICM method approaches these problems with the following steps: (1) a starting model is created based on the homologous structure with the conserved portion fixed and the noncon-served portion having standard covalent geometry and free torsion angles; (2) the Biased Probability Monte Carlo (BPMC) procedure is applied to search the subspaces of either all the nonconservative side-chain torsion angles or torsion angles in a loop backbone and surrounding side chains. A special algorithm was designed to generate low-energy loop deformations. The BPMC procedure globally optimizes the energy function consisting of ECEPP/3 and solvation energy terms. Comparison of the predictions with the NMR or crystallographic solutions reveals a high proportion of correctly predicted side chains. The loops were not correctly predicted because imprinted distortions of the backbone increased the energy of the near-native conformation and thus made the solution unrecognizable. Interestingly, the energy terms were found to be reliable and the sampling of conformational space sufficient. The implications of this finding for the strategies of future comparative modeling are discussed. © 1995 Wiley-Liss, Inc.     

Dynamic properties of some β-chain mutant hemoglobins

The thermal behavior of the Soret band relative to the carbonmonoxy derivatives of some β-chain mutant hemoglobins is studied in the temperature range 300–10 K and compared to that of wild-type carbonmonoxy hemoglobin. The band profile at various temperatures is modeled as a Voigt function that accounts for homogeneous broadening and for the coupling with high- and low-frequency vibrational modes, while inhomogeneous broadening is taken into account with a gaussian distribution of purely electronic transition frequencies. The various contributions to the overall bandwidth are singled out With this analysis and their temperature dependence, in turn, gives information on structural and dynamic properties of the system studied. In the wildtype and mutant hemoglobins, the values of homogeneous bandwidth and of the coupling constants to high-frequency vibrational modes are not modified with respect to natural human hemoglobin, thus indicating that the local electronic and vibrational properties of the heme–CO complex are not altered by the recombinant procedures. On the contrary, differences in the protein dynamic behavior are observed. The most relevant are those relative to the “polar isosteric” βVal-67(Ell) →Thr substitution, localized in the heme pocket, which results in decreased coupling with low-frequency modes and increased anharmonic motions. Mutations involving residue βLys-144(HC1) at the C-terminal and residue βCys-112(G14) at the α₁β₁ interface have a smaller effect consisting in an increased coupling with low-frequency modes. Mutations at the β-N-terminal and at the α₁β₂ interface have no effect on the dynamic properties of the heme pocket. © 1995 Wiley-Liss, Inc.     

Investigation of the folding pathway of the TEM-1 β-lactamase

The TEM-1 β-lactamase is a globular protein containing 12 proline residues. The folding mechanism of this enzyme was investigated by kinetic and equilibrium experiments with the help of fluorescence spectroscopy and circular dichroism. The equilibrium denaturation of the protein induced by guanidine hydrochloride occurs in two discrete steps, indicating the existence of a thermodynamically stable intermediate state. Thisstate is 5.2 ± 0.4 kcal/mol less stable than the native conformation and 5.7 ± 0.2 kcal/mol more stable than the fully denaturedprotein. This intermediate state exhibits a high content of native secondary structure elements but is devoid of specific tertiary organization; its relation to the “molten globule” is discussed. Refolding kinetic experimentsrevealed the existence of a transient intermediate conformation between thethermodynamically stable intermediate and the native protein. This transient intermediate appears rapidly during the folding reaction. It exhibits a secondary structure content very similar to that of the native protein and has also recovered a significant amount of tertiary organisation. The final refolding step of the TEM-1 β-lactamase, leading to the native enzyme, is dominated by two major slow kinetic phases which probablyreflect a very complex process kinetically limited by proline cis/transisomerization. © 1995 Wiley-Liss, Inc.     

An estimate of spin diffusion in a spin subset: Application to iterative distance calculation from 3D 15N NOESY-HMQC

Summary A method for quantification of distances between amide hydrogens using only the 3D NOESY-HMQC experiment recorded on a ¹⁵N-labelled protein is presented. This method is based on an approximate expression of the NOE intensities between amide hydrogens obtained from continuum modelling of the non-amide spins; this expression is used in a distance calculation algorithm. The algorithm has been named CROWD, standing for Continuum approximation of Relaxati On path Ways between Dilute spins. This approximation as well as the CROWD algorithm are tested on a simulated case; the CROWD algorithm is then applied to experimental data, measured on a fragment of bacteriorhodopsin.     

Three-dimensional structure of ectatomin from Ectatomma tuberculatum ant venom

Summary Two-dimensional ¹H NMR techniques were used to determine the spatial structure of ectatomin, a toxin from the venom of the ant Ectatomma tuberculatum. Nearly complete proton resonance assignments for two chains of ectatomin (37 and 34 amino acid residues, respectively) were obtained using 2D TOCSY, DQF-COSY and NOESY experiments. The cross-peak volumes in NOESY spectra were used to define the local structure of the protein and generate accurate proton-proton distance constraints employing the MARDIGRAS program. Disulfide bonds were located by analyzing the global fold of ectatomin, calculated with the distance geometry program DIANA. These data, combined with data on the rate of exchange of amide protons with deuterium, were used to obtain a final set of 20 structures by DIANA. These structures were refined by unrestrained energy minimization using the CHARMm program. The resulting rms deviations over 20 structures (excluding the mobile N- and C-termini of each chain) are 0.75 ? for backbone heavy atoms, and 1.25 ? for all heavy atoms. The conformations of the two chains are similar. Each chain consists of two α-helices and a hinge region of four residues; this forms a hairpin structure which is stabilized by disulfide bridges. The hinge regions of the two chains are connected together by a third disulfide bridge. Thus, ectatomin forms a four-α-helical bundle structure.     

Model experiments on squid axons and NG108-15 mouse neuroblastoma × rat glioma hybrid cells

Three types of ionic current essentially determine the firing pattern of nerve cells: the persistent Na⁺ current, the M current and the low-voltage-activated Ca²⁺ current. The present article summarizes recent experiments concerned with the basic properties of these currents. Keynes and Meves (Proc R Soc Lond B (1993) 253, 61–68) studied the persistent or steady-state Na⁺ current on dialysed squid axons and measured the probability of channel opening both for the peak and the steady-state Na⁺ current (PF_peak and PF_ss) as a function of voltage. Whereas PF_peak starts to rise at −50 mV and reaches a maximum at +40 to +50 mV, PF_ss only begins to rise appreciably at around 0 mV and is still increasing at +100 mV. This differs from observations on vertebrate excitable tissues where the persistent Na⁺ current turns on in the threshold region and saturates at around 0 mV. Schmitt and Meves (Pflügers Arch (1993) 425, 134–139) recorded M current, a non-inactivating K⁺ current, from NG108-15 neuroblastoma × glioma hybrid cells, voltage-clamped in the whole-cell mode, and studied the effects of phorbol 12,13-dibutyrate (PDB), an activator of protein kinase C (PKC), and arachidonic acid (AA). PDB and AA both decreased I_M, the effective concentrations being 0.1–1 μM and 5–25 μM, respectively; while the PDB effect was regularly observed, the M current depression by AA was highly variable from cell to cell. The PKC 19–31 peptide, an effective inhibitor of PKC, in a concentration of 1 μM almost totally prevented the effects of PDB and AA on M current, suggesting that both are mediated by PKC. Schmitt and Meves (Pflügers Arch (1994a) 426, Suppl R 59) measured low-voltage-activated (l-v-a) and high-voltage-activated (h-v-a) Ca²⁺ currents on NG108-15 cells and investigated the effect of AA and PDB on both types of current. At pulse potentials > −20 mV, AA (25–100 μM) decreased l-v-a and h-v-a I_Ca. The decrease was accompanied by a small negative shift and a slight flattening of the activation and inactivation curves of the l-v-a I_Ca. The AA effect was not prevented by 50 μM eicosa-5,8,11,14-tetraynoic acid (ETYA), an inhibitor of AA metabolism, or PKC 19–31 peptide and not mimicked by 0.1–1 μM PDB. Probably, AA acts directly on the channel protein or its lipid environment. The physiological relevance of these three sets of observations is briefly discussed.     

Purification and characterization of M3 protein expressed on the surface of group A streptococcal type 3 strain C203

Abstract Monoclonal antibodies (mAbs) have been produced by immunizing BALB/C mice with whole M⁺ bacteria in incomplete Freund adjuvant and the resulting mAbs for M3 protein have been selected by an indirect immuno-fluorescent technique using formaldehyde-fixed M⁺ and M⁻ bacteria. Four mAbs reacted with a 65 kDa protein in an extract obtained from the cell wall of M⁺ bacteria after treatment with N -acetyl muramidase and lysozyme. The purified 65 kDa protein neutralized the phagocytic activity of rabbit anti-M3 antibody. The N-terminal amino acid sequence of the 65 kDa protein was identical with that of protein generated by the M3 gene which has been previously cloned and sequenced. The evidence indicates that the 65 kDa protein is M3 protein. The M3 protein bound not only human fibrinogen but also human serum albumin (HSA). When the M3 protein was purified by gel-filtration and ion-exchange chromatography in the absence of phenylmethyl sulfonyl fluoride (PMSF), four fragments (35 kDa, 32 kDa, 30 kDa, and 25 kDa) in addition to the intact molecule appeared. N-terminal amino acid sequence analysis showed that 35 kDa and 25 kDa fragments were ANAAD and DARSV, respectively, being identical at positions 1–5 and 198–202 to the M3 gene derived protein. Therefore, the 35 kDa and 25 kDa fragments, which were presumed to be cleavage products, may be derived from the C-terminal part and N-terminal part of the intact molecule, respectively. When the effect of purified M3 protein in the bactericidal activity of normal human blood in the presence of M⁻ bacteria was investigated, the M3 protein was responsible for the organism's resistance to attack by phagocytic cells.     

Multiple molecular recognition properties of the lipocalin protein family

The lipocalins, a diverse family of small extracellular ligand proteins, display a remarkable range of different molecular properties. While their binding of small hydrophobic molecules, and to a lesser extent their binding to cell surface receptors, is well known, it is shown here that formation of macromolecular complexes is also a common feature of this family. Analysis of known crystallographic structures reveals that the lipocalins process a conserved common structure: an antiparallel β-barrel with a repeated +1 topology. Comparisons show that within this overall similarity the structure of individual proteins is specifically adapted to bind their particular ligands, forming a binding site from an internal cavity (within the barrel) and/or an external loop scaffold, which gives rise to different binding modes that reflects the need to accommodate ligands of different shape, size, and chemical structure. The architecture of the lipocalin fold suggests that the both the ends and sides of this barrel are topologically distinct, differences also apparent in analyses of structural and sequence variation within the family. These different can be linked to experimental evidence suggesting a possible functional dichotomy between the two ends of the lipocalin fold. The structurally invariant end of the molecule may be implicated in general binding small ligands and forming macromolecular complexes via an exposed binding surface.     

Application of protein-A indirect ELISA (PAI-ELISA) for the detection of anti-Smith antibodies in systemic lupus erythematosus patients

An indirect form of protein-A ELISA (PAI-ELISA) was optimized and, when used to detect anti-Smith antibodies in sera of 31 systemic lupus erythematosus (SLE) patients, gave results comparable with those using a commercial immunodiffusion kit. The number of sera found to be positive for anti-Smith antibodies by ELISA was seven, four of which were also found positive by immunodiffusion.B.O. Siti-Rohana is with the Universiti Kebangsaan Malaysia (UKM), Faculty of Medicine, Kuala Lumpur, Malaysia. I.B. Ahmad is with the Department of Microbiology, Faculty of Life Sciences, Universiti Kebangsaan Malaysia, Bangi, 43600 UKM, Malaysia; B.A. Nasaruddin is with the Institute for Medical Research, Malaysia.