Some insights into protein structural class prediction

It has been quite clear that the success rate for predicting protein structural class can be improved significantly by using the algorithms that incorporate the coupling effect among different amino acid components of a protein. However, there is still a lot of confusion in understanding the relationship of these advanced algorithms, such as the least Mahalanobis distance algorithm, the component-coupled algorithm, and the Bayes decision rule. In this communication, a simple, rigorous derivation is provided to prove that the Bayes decision rule introduced recently for protein structural class prediction is completely the same as the earlier component-coupled algorithm. Meanwhile, it is also very clear from the derivative equations that the least Mahalanobis distance algorithm is an approximation of the component-coupled algorithm, also named as the covariant-discriminant algorithm introduced by Chou and Elrod in protein subcellular location prediction (Protein Engineering, 1999; 12:107-118). Clarification of the confusion will help use these powerful algorithms effectively and correctly interpret the results obtained by them, so as to conduce to the further development not only in the structural prediction area, but in some other relevant areas in protein science as well.     

Monte Carlo simulation studies on the prediction of protein folding types from amino acid composition

A number of methods to predicting the folding type of a protein based on its amino acid composition have been developed during the past few years. In order to perform an objective and fair comparison of different prediction methods, a Monte Carlo simulation method was proposed to calculate the asymptotic limit of the prediction accuracy [Zhang and Chou (1992),Biophys. J. 63, 1523–1529, referred to as simulation method I]. However, simulation method I was based on an oversimplified assumption, i.e., there are no correlations between the compositions of different amino acids. By taking into account such correlations, a new method, referred to as simulation method II, has been proposed to recalculate the objective accuracy of prediction for the least Euclidean distance method [Nakashimaet al. (1986),J. Biochem. 99, 152–162] and the least Minkowski distance method [Chou (1989),Prediction in Protein Structure and the Principles of Protein Conformation, Plenum Press, New York, pp. 549–586], respectively. The results show that the prediction accuracy of the former is still better than that of the latter, as found by simulation method I; however, after incorporating the correlative effect, the objective prediction accuracies become lower for both methods. The reason for this phenomenon is discussed in detail. The simulation method and the idea developed in this paper can be applied to examine any other statistical prediction method, including the computersimulated neural network method.     

Woodward''s reagent K reacts with histidine and cysteine residues inEscherichia coli andSaccharomyces cerevisiae phosphoenolpyruvate carboxykinases

The reaction of Woordward's reagent K (WRK) with model amino acids and proteins has been analyzed. Our results indicate that WRK forms 340-nm-absorbing adducts with sulfhydryl- and imidazol-containing compounds, but not with carboxylic acid derivatives, in agreement with Llamaset al. [(1986),J. Am. Chem. Soc. 108, 5543–5548], but not with Sinha and Brewer [(1985),Anal. Biochem. 151, 327–333]. The chemical modification ofEscherichia coli andSaccharomyces cerevisiae phosphoenolpyruvate carboxykinases with WRK leads to an increase in the absorption at 340 nm, and we have demonstrated its reaction with His and Cys residues in these proteins. These results caution against claims of glutamic or aspartic acid modification by WRK based on the absorption at 340 nm of protein-WRK adducts.Abbreviations HPLC high-performance liquid chromatography - MES 2-(N-morpholino)ethanesulfonic acid - PEP phosphoenolpyruvate - PEPCK phosphoenolpyruvate carboxykinase - PTH phenylthiohydantoin - WRK Woordward's reagent K (2-ethyl-5-phenylisoxazolium-3-sulfonate)     

Is it a paradox or misinterpretation?

The paradox recently raised by Wang and Yuan (Proteins 2000;38:165-175) in protein structural class prediction is actually a misinterpretation of the data reported in the literature. The Bayes decision rule, which was deemed by Wang and Yuan to be the most powerful method for predicting protein structural classes based on the amino acid composition, and applied by these investigators to derive the upper limit of prediction rate for structural classes, is actually completely the same as the component-coupled algorithm proposed by previous investigators (Chou et al., Proteins 1998;31:97-103). Owing to lack of a complete or near-complete training data set, the upper limit rate thus derived by these investigators might be both invalid and misleading. Clarification of these points will further stimulate investigation of this interesting area.     

Comparative immunological and electrophoretic studies on ribosomal proteins of Bacillaceae

Summary The ribosomal proteins from several Bacillus species were compared by two-dimensional gel electrophoresis and immunological methods. The results revealed great heterogeneity among most Bacillus species. Comparison of ribosomal proteins from Bacilli with those of E. coli by two-dimensional gel electrophoresis showed little similarities, while structural homologies could be found by immunological methods. SDS two-dimensional gel electrophoresis revealed that the molecular weight of ribosomal proteins is conserved in all tested bacteria.Paper No. 68 on Ribosomal Proteins. Preceding paper is by Geisser et al., Molec. gen. Genet. 127, 111–128 (1973).     

Expression, Purification, Characterization, and X-Ray Analysis of Selenomethionine 215 Variant of Leukocyte Collagenase

Matrix metalloproteinases belong to the superfamily of metzincins containing, besides a similar topology and a strictly conserved zinc environment, a 1,4-tight turn with a strictly conserved methionine residue at position three (the so called Met-turn [Bode et al. (1993) FEBS 331, 134–140; Stöcker et al. (1995) Protein Sci. 4, 823–840]. The distal S–CH₃ moiety of this methionine residue forms the hydrophobic basement of the three His residues liganding the catalytic zinc ion. To assess the importance of this methionine, we have expressed the catalytic domain of neutrophil collagenase (rHNC, residues Met80–Gly242) in the methionine auxotrophic Escherichia coli strain B834[DE3](hsd metB), with the two methionine residues replaced by Selenomethionine. Complete replacement was confirmed by amino acid analysis and electrospray mass spectrometry. The folded and purified enzyme retained its catalytic activity, but showed modifications which are reflected in changed kinetic parameters. The Met215SeMet substitution caused a decrease in conformational stability upon urea denaturation. The X-ray crystal structure of this Selenomethionine rHNC was virtually identical to that of the wild-type catalytic domain except for a very faint local disturbance around the sulfur-seleno substitution site.     

Preparation of biologically active recombinant human progastrin(1-80)

The bacterial expression of human progastrin_6–80 has been reported previously [Baldwin, G.S. et al. (2001) J. Biol. Chem. 276: 7791-7796]. The aims of the present study were to prepare full-length recombinant human progastrin_1–80 and to compare its biological activity with that of progastrin_6–80 in vitro, to determine whether or not the N-terminal five amino acids contributed to activity. A fusion protein of glutathione-S-transferase and human progastrin_1–80 was expressed in Escherichia coli, collected on glutathione-agarose beads, and cleaved with enterokinase. Progastrin_1–80 was purified by reversed-phase and anion exchange HPLC and characterized by radioimmunoassay, amino acid sequencing, and mass spectrometry. No differences were detected in the extent of stimulation by progastrin_1–80 and progastrin_6–80 in proliferation and migration assays with the mouse gastric cell line IMGE-5. We conclude that residues 1–5 of progastrin_1–80 are not essential for biological activity.     

Accuracy of Secondary Structure and Solvent Accessibility Predictions for a Clostridial Neurotoxin C-Fragment

Earlier studies used Rost and Sander's artificial neural network [(1993a), J. Mol. Biol. 232, 584–599] to predict the secondary structures [Lebeda and Olson (1994), Proteins 20, 293–300] and residue solvent accessibilities [Lebeda and Olson (1997), J. Protein Chem. 16, 607–618] of the clostridial neurotoxins. Because the X-ray crystal structure of the 50-kDa C-terminal half of the heavy chain of tetanus toxin was recently determined, this report evaluates the accuracy of these network-derived predictions. For this predominantly -strand-containing fragment, predictions, on a per-residue basis, for both secondary structure and solvent accessibility were about 70% accurate. A more flexible and realistic analysis based on overlapping segments yielded accuracies of over 80% for the three-state secondary structure and for the two-state accessibility predictions. Because the accuracies of these predictions are comparable to those made by Rost and Sander using a dataset of 126 nonhomologous globular proteins, our predictions provide a quantitative foundation for gauging the results when building by homology the structures of related proteins.     

1H, 15N and 13C resonance assignments and monomeric structure of the amino-terminal extracellular domain of epithelial cadherin

Summary E-cadherin is a transmembrane protein that provides Ca²⁺-dependent cell adhesion to epithelial cells. The large majority of the ¹H, ¹⁵N, ¹³C and ¹³CO resonances of a 146-amino acid polypeptide from epithelial (E-) cadherin have been assigned using multidimensional NMR spectroscopy. The structure of the amino-terminal 100 amino acids, corresponding to the first extracellular repeat of E-cadherin [Overduin et al. (1995) Science, 267, 386–389], has been refined. The monomeric state of this isolated domain is demonstrated by light scattering and sedimentation analysis. Seven -strands and two short helices were identified by patterns of NOE cross-peaks, vicinal coupling constants and chemical shift indices. A novel structural motif termed a quasi--helix found in the crystal structure of a neural (N-) cadherin domain [Shapiro et al. (1995) Nature, 374, 327–337] is characterized in detail for the first time by NMR. Slowly exchanging amides were concentrated in the -sheet region and quasi--helix. The -barrel fold of the cadherin domain is topologically similar to the immunoglobulin fold. Comparison of this solution structure to the crystallized dimers of the N-terminal pair of E-cadherin domains [Nagar et al. (1996) Nature, 380, 360–364] and of the homologous single domain of N-cadherin reveals a conserved cadherin fold with minor structural differences, which can be accounted for by differences in metal ligation and oligomeric state.Abbreviations cad extracellular cadherin repeat - CAM cell adhesion molecule - CSI chemical shift index - DTT dithiothreitol - E-cadherin epithelial cadherin - N-cadherin neural cadherin - NOE nuclear Overhauser enhancement - PFG pulsed field gradient - rmsd root-mean-square deviation     

A general enhancement scheme in heteronuclear multidimensional NMR employing pulsed field gradients

Summary General pulse sequence elements that achieve sensitivity-enhanced coherence transfer from a heteronucleus to protons of arbitrary multiplicity are introduced. The building blocks are derived from the sensitivity-enhancement scheme introduced by Cavanagh et al. ((1991) J. Magn. Reson., 91, 429–436), which was used in conjunction with gradient coherence selection by Kay et al. ((1992) J. Am. Chem. Soc., 114, 10663–10665), as well as from a multiple-pulse sequence effecting a heteronuclear planar coupling Hamiltonian. The building blocks are incorporated into heteronuclear correlation experiments, in conjunction with coherence selection by the formation of a heteronuclear gradient echo. This allows for efficient water suppression without the need for water presaturation. The methods are demonstrated in HSQC-type experiments on a sample of a decapeptide in H₂O. The novel pulse sequence elements can be incorporated into multidimensional experiments.     

Purification, characterization, and relation to bikunin of rat urinary trypsin inhibitors

Two forms of urinary trypsin inhibitor (UTI-1 and UTI-2) were purified from pooled urine of normal male rats to apparent homogeneity by salting out, affinity chromatography, gel filtration, and reverse-phase HPLC. UTIs-1 and 2 were shown to be thermostable glycoproteins with the respective molecular weights of 22,000 and 18,000 estimated by SDS-PAGE. These inhibitors combined with bovine trypsin in a 1:1 molar ratio: the K _d values were 2.5 × 10^–10 and 2.3 × 10^–10 M, respectively. Amino acid composition and sequence analysis indicated that UTI-1 corresponded to rat bikunin of which the amino acid sequence was deduced from a rat liver cDNA clone encoding ₁-microglobulin [Lindqvist et al. (1992), Biochim. Biophys. Acta 1130, 63–67] except that the protein sequence seemed to lack C-terminal serine, and UTI-2 corresponded to UTI-1 lacking N-terminal 21 amino acid residues.     

Amino acid sequence of squid troponin C

The complete amino acid sequence of squid Todarodes pacificus troponin C (TnC), which was shown to bind only 1 mol Ca²⁺/mol, was determined by both the Edman and cDNA methods. The squid TnC is composed of 147 amino acids including an unblocked Pro at the N-terminus and the calculated molecular weight is 17 003.9. Among the four potential Ca²⁺-binding sites, namely sites I–IV from the N-terminus, only site IV completely satisfied the consensus amino acid sequence for the active Ca²⁺-binding loop. This indicates that squid TnC possesses a single Ca²⁺-binding site at the site IV as scallop TnCs [Nishita et al., J. Biol. Chem. 269 (1994) 3464–3468; Ojima et al., Arch. Biochem. Biophys. 311 (1994) 272–276). The sequence homology of squid TnC to TnCs of scallop, arthropods, and rabbit was 61%, 31–38%, and 31%, respectively. In the sequence of the central D/E-helix region of squid and scallop TnCs, a deletion of three amino acids was required to maximize the homology with the other TnCs.     

Structural Features of Glycera dibranchiata Monomer Hemoglobins. Primary Sequences of Monomer Hemoglobin Components II and III

Primary sequences for the remaining two members (GMH2, GMH3) of the group of three major monomeric hemoglobins from the marine annelid Glycera dibranchiata have been obtained. Full sequences of each 147-amino acid globin were achieved with a high degree of confidence using standard Edman technology in combination with molecular mass determinations of the intact globins and of the cyanogen bromide cleavage fragments using electrospray ionization mass spectrometry. When minor assumptions concerning Q/E identities are made these new results indicate the likely correspondence of GMG2 with the protein represented by the first Glycera dibranchiata monomer hemoglobin complete sequence [Imamura et al., (1972), J. Biol. Chem. 247, 2785–2797]. When these new sequences are combined with the previously determined primary sequence for the third major monomer hemoglobin, GMH4 [Alam et al., J. Protein Chem. (1994), 13, 151–164], it becomes clear that these three (GMG2–4) are truly distinct proteins, contrary to previous suggestions. Surprisingly, our results show that none of these three primary sequences is identical to the published sequence of the refined monomer hemoglobin crystal structure protein; however, there is a strong correspondence to the GMG2 sequence. The present sequencing results, in combination with the published GMH4 sequence, confirm the presence of a distal Leu in place of the more commonly encountered distal His in all three of the major monomer hemoglobins isolated in this laboratory and indicate that the unusual B10 Phe occurs only in GMH4. Analysis of the sequences presented here, along with comparison of amino acid content for Glycera dibranchiata monomer hemoglobins isolated from three different laboratories, and comparison of NMR results from two laboratories suggest further correspondences which unify disparate published isolations.     

Purification and characterization of a 60-kDa protein from oat, formerly known as a TCP1-related chaperone

Recently, Mummertet al. [Nature 363, 644–648 (1993)] isolated a proposed TCP1-related chaperone. Here we report several findings concerning the protein which they sequenced. Two similar N-terminal sequences were obtained from this abundant 60-kDa protein. Internal sequences were also acquired by protease digestion. Initially it was believed the protein was able to completely inhibit citrate synthase aggregation, but later purifications demonstrated that the 60-kDa polypeptide lacked both chaperone activity and the previously reported kinase activity [Grimmet al., Planta 178, 199–206 (1989)]. It is now our belief that this protein is neither a chaperone nor a kinase.     

The cell wall of Chlamydomonas reinhardtii gametes: Composition,structure and autolysin-mediated shedding and dissolution

The cell walls of Chlamydomonas gametes are multilayered structures supported on frameworks of polypeptides extending from the plasma membrane. The wall-polypeptide catalogue reported by Monk et al. (1983, Planta 158, 517–533) and extended by U.W. Goodenough et al. (1986, J. Cell Biol. 103, 405–417) was re-evaluated by comparative analysis of mechanically isolated cell walls purified from several strains. The extracellular locus of wall polypeptides was verified by in vivo iodogen-catalysed iodination and by autolysin-mediated elimination of the bulk of these polypeptides from the cell surface. Three (w15, w16, w17) and possibly four (w14) polypeptides were located to the most exterior aspect of the wall because of their susceptibility to Enzymobeadcatalysed iodination and their retention by a cell-wall-less mutant. The composition of shed walls stabilised with ethylenediaminetetraacetic acid during natural mating and kinetic analysis of the dissolution of walls purified from a bald-2 mutant demonstrated the rapid and specific destruction of polypeptide w3. Differential solubilisation of wall polypeptides occurred after loss of w3. Wall dissolution, characterised by the generation of fishbone structures from the W2 layer, gave as many as four additional polypeptides. Charged detergents and sodium perchlorate extracted a comparable range of polypeptides at room temperature from mechanically isolated walls, i.e. components of the W4–W6 layers, hot sodium dodecyl sulphate solubilised framework polypeptides, while reducing agent was required to solubilise the W2 layer. A model of wall structure is presented.Abbreviations DTE dithioerythritol - EDTA ethylenediaminetetraacetic acid - M_r relative molecular mass - SDS-PAGE sodium dodecyl sulphate-polyacrylamide gel electrophoresis - Tris 2-amino-2-(hydroxymethyl)-1,3-propanediol     

An improved method for large-scale purification of recombinant human glucagon

Glucagon was expressed inEscherichia coli as a fusion protein including the glucagon sequence [Ishizakiet al. (1992),Appl. Microbiol. Biotechnol.36, 483–486]. The high-level expression of a protein inE. coli often results in an insoluble aggregate called an inclusion body containing a fusion protein. In our previous report [Yoshikawaet al. (1992),J. Protein Chem. 11, 517–525], we solubilized this inclusion body by using guanidinium chloride. However, the existence of denaturant caused problems such as a low proteolytic activity for transforming the fusion protein into glucagon and complicated purification methods. We tried to improve the method to enable large-scale purification. At alkaline pH, the inclusion body could be solubilized to a high concentration and cleaved by amino acid-specific endopeptidases. By utilizing isoelectric precipitations as a new economical purification method for glucagon from intermediates, the glucagon obtained was shown to be over 99.5% pure by analytical RP-HPLC. The yield was almost equal that of our previous method, and the glucagon produced was chemically and biochemically equivalent to natural glucagon.     

What is pea legumin — Is it glycosylated?

Since there is some question as to whether or not legumin is glycosylated, this storage protein was isolated by various procedures from developing cotyledons of Pisum sativum L. supplied with [¹⁴C]-labeled glucosamine and analyzed by sodium dodecylsulfate-polyacrylamide gel electrophoresis. Legumin isolated by the classical method of Danielsson [(1949) Biochem. J. 44, 387–400] a procedure in which globulins extracted with a buffered salt solution are precipitated with ammonium sulfate (70% saturation) and legumin separated from vicilin by isoelectric precipitation, was labeled. The glucosamine incorporated into legumin was associated with low-molecular-weight polypeptides. In contrast, legumin isolated by the method of Casey [(1979) Biochem. J. 177, 509–520], a procedure where legumin is prepared by zonal isoelectric precipitation from globulins precipitated with 40–70% ammonium sulfate, was not labeled. However, the globulin fraction precipitated with 40% ammonium sulfate was labeled and the radioactive glucosamine was associated with low-molecular-weight polypeptides. Legumin isolated from protein bodies [Thomson et al. (1978) Aust. J. Plant Physiol. 5, 263–279] was not extensively labeled. However, the saltinsoluble fraction of protein body extracts was labeled and the radioactivity was associated with low-molecular-weight polypeptides. These results indicate that protein bodies contain a glycoprotein of low-molecular-weight that co-purifies with legumin isolated by the method of Danielsson but that is discarded when isolation methods developed more recently are used.     

Improved frequency resolution in multidimensional constant-time experiments by multidimensional Bayesian analysis

Summary The resolution of spectral frequencies in NMR data obtained from discrete Fourier transformation (DFT) along D constant-time dimensions can be improved significantly through extrapolation of the D-dimensional free induction decay (FID) by multidimensional Bayesian analysis. Starting from Bayesian probability theory for parameter estimation and model detection of one-dimensional time-domain data [Bretthorst, (1990) J. Magn. Reson., 88, 533–551; 552–570; 571–595], a theory for the D-dimensional case has been developed and implemented in an algorithm called BAMBAM (BAyesian Model Building Algorithm in Multidimensions). BAMBAM finds the most probable sinusoidal model to account for the systematic portion of any D-dimensional stationary FID. According to the parameters estimated by the algorithm, the FID is extrapolated in D dimensions prior to apodization and Fourier transformation. Multidimensional Bayesian analysis allows for the detection of signals not resolved by the DFT alone or even by sequential one-dimensional extrapolation from mirror-image linear prediction prior to the DFT. The procedure has been tested with a theoretical two-dimensional dataset and with four-dimensional HN(CO)CAHA (Kay et al. (1992) J. Magn. Reson., 98, 443–450) data from a small protein (8 kDa) where BAMBAM was applied to the ¹³C and H constant-time dimensions.To whom correspondence should be addressed.     

1H-NMR study of GM2 ganglioside: evidence that an interresidue amide-carboxyl hydrogen bond contributes to stabilization of a preferred conformation

Several properties of the exchangeable amide protons of the ganglioside G_M2 were studied in detail by¹H-NMR spectroscopy in fully deuterated dimethylsulfoxide [²H₆]DMSO/2% H₂O, and compared with data obtained for the simpler constituent glycosphingolipids G_A2 and G_M3. In addition to chemical shifts,³ J _2,HN coupling constants, and temperature shift coefficients, the kinetics of NH/²H chemical exchange were examined by following the disappearance of the amide resonances in [²H₆]DMSO/2%²H₂O. The results included observation of an increase in half-life of theN-acetylgalactosamine acetamido HN by more than an order of magnitude in G_M2 compared to G_A2, attributable to the presence of the additionalN-acetylneuraminic acid residue. Additional one-dimensional dipolar cross relaxation experiments were also performed on nonexchangeable protons of G_M2. The results of all of these experiments support a three-dimensional model for the terminal trisaccharide in which a hydrogen bond is formed between theN-acetylgalactosamine acetamido NH and theN-acetylneuraminic acid carboxyl group. The interaction is proposed to be of the -acceptor type, a possibility which has not yet been explored in the literature on carbohydrates. The proposed model is discussed in comparison with that of Sabesanet al. (1984,Can J Chem 62: 1034–45), and the models of G_M1 proposed more recently by Acquottiet al. (1990,J Am Chem Soc 112:7772–8) and Scarsdaleet al. (1990,Biochemistry 29:9843–55).     

Doubly sensitivity-enhanced 3D TOCSY-HSQC

Summary Recently, strategies for double sensitivity enhancement in heteronuclear three-dimensional NMR experiments were introduced (Krishnamurthy, V.V. (1995) J. Magn. Reson., B106, 170–177; Sattler et al. (1995) J. Biomol. NMR, 6, 11–22; Sattler et al. (1995) J. Magn. Reson., B108, 235–242). Since a sensitivity enhancement of a factor 2^1/2 can be achieved for each indirect dimension, nD spectra can theoretically be enhanced up to a factor of 2^((n-1)/2). We propose and analyze a doubly enhanced three-dimensional TOCSY-HSQC sequence. The application of the doubly enhanced three-dimensional {¹⁵N, ¹H} TOCSY-HSQC sequence is shown for uniformly ¹³C-/¹⁵N- and ¹⁵N-labeled samples of the relatively large Azotobacter vinelandii flavodoxin II (179 amino acids). The main factors that contribute to the final signal-to-noise enhancement have been systematically investigated. The sensitivity enhancement obtained for the doubly enhanced TOCSY-HSQC pulse sequence as compared to the standard (unenhanced) version is close to the theoretically expected factor of two.