Hydrophobic cluster analysis: an efficient new way to compare and analyse amino acid sequences

A new method for comparing and aligning protein sequences is described. This method, hydrophobic cluster analysis (HCA), relies upon a two-dimensional (2D) representation of the sequences. Hydrophobic clusters are determined in this 2D pattern and then used for the sequence comparisons. The method does not require powerful computer resources and can deal with distantly related proteins, even if no 3D data are available. This is illustrated in the present report by a comparison of human haemoglobin with leghaemoglobin, a comparison of the two domains of liver rhodanese (thiosulphate sulphurtransferase) and a comparison of plastocyanin and azurin.     

A vector representation for amino acid sequences

A scheme for representing amino acids as vectors in the plane is presented and justified. The two dimensions of the plane are size and hydrophobicity. The vector representation is then applied to generate a consensus sequence for some sets of homologous proteins. A figure of merit for the degree of homology of a set of sequences results from the analysis. Some other applications of the scheme are considered also. This work grew from ideaseeds planted by Margaret Dayhoff's work in theAtlas of Protein Structure and Sequence. It is with gratitude that I dedicate this paper to her memory.     

Similar amino acid sequences revisited

The rapid accumulation of protein sequences, many bearing unexpected resemblances to each other, is providing a new perspective on evolution.     

A computer program to calculate and design oligonucleotide primers from amino acid sequences

A computer program to generate oligonucleotide sequences frompeptide data is presented. The program, PrimerGen, reverse-translatesa peptide string into its corresponding degenerate deoxyribonucleotidecounterpart, and generates sequences of oligonucleotide moleculeswhich can be synthesized and either used as probes for hybridizationsor as primers in polymerase chain reactions.     

The amino acid sensitive TOR pathway from yeast to mammals

The target of rapamycin (TOR) is an ancient effector of cell growth that integrates signals from growth factors and nutrients. Two downstream effectors of mammalian TOR, the translational components S6K1 and 4EBP1, are commonly used as reporters of mTOR activity. The conical signaling cascade initiated by growth factors is mediated by PI3K, PKB, TSC1/2 and Rheb. However, the process through which nutrients, i.e., amino acids, activate mTOR remains largely unknown. Evidence exists for both an intracellular and/or a membrane bound sensor for amino acid mediated mTOR activation. Research in eukaryotic models, has implicated amino acid transporters as nutrient sensors. This review describes recent advances in nutrient signaling that impinge on mTOR and its targets including hVps34, class III PI3K, a transducer of nutrient availability to mTOR.     

A rapid procedure for detection of bacterial amino acid decarboxylases

The detection of decarboxylases of arginine, glutamic acid, histidine, lysine, ornithine, phenylalanine, tryptophan, and tyrosine in bacteria by thin-layer chromatography on polyamide sheets is described. The bacteria were grown on agar medium plates supplemented with eight amino acids at pH 5.5 for induction of amino acid decarboxylases, then transferred to amine-production media. The decarboxylation products in the spent media (amines and/or γ-amino-n-butyric acid) were dansylated and the dansyl derivatives were separated by thin-layer chromatography on polyamide sheets. This method requires only two separate incubations of the decarboxylase-induced bacteria in amine-production media for 1 h at 37°C for simultaneous detection of eight bacterial amino acid decarboxylases using 0.4 μl of the spent media.     

Structual comparison of dermatopontin amino acid sequences

Dermatopontin is a tyrosine-rich acidic extracellular matrix protein of 22 kD with possible functions in cellmatrix interactions and matrix assembly. Database of GenBank+EMBL+DDBJ sequences from Nucleotide, Gene, and Expressed Sequence Tag (EST) Divisions was searched with a keyword “dermatopontin” or mouse dermatopontin amino acid sequence. In addition to five mammals previously described, five mammalian, two bird, one fish dermatopontin genes were detected in vertebrates. Additionally, a goat EST was also shown as goat dermatopontin missing 5′-end of the coding region. Moreover, a mRNA sequence of rhesus monkey dermatopontin was identified, but the deduced amino acid sequence was terminated abruptly due to a nonsense codon. For three 6-residue repeat regions (D-R-E/Q-W-X-F/Y) that may function as part of a glycosaminoglycan binding site, the first repeat sequence is D-R-Q-W-N-Y in all mammals while Glutamine is substituted for Leucine in birds. The second and the third repeats are conserved in all vertebrates. The N-Y-D sequence, the consensus in many amine oxidases, is conserved in mammals except rodents. Asparagine is substituted for Threonine in birds. The tetrapeptide R-G-A-T sequence possibly recognizing the integrin family is conserved in mammals and birds, but Alanine was substituted for Glutamine in zebrafish resulting in loss of activity. In conclusion, functionally significant amino acid sequences in vertebrate dermatopontins are conserved in mammals, while they are not necessarily identified in birds and fish. The original function of vertebrate dermatopontins may be glycosaminoglycan binding and functions as a ligand for integrin and an amine oxidase may be gained in the process of evolution.     

A new double-labelling procedure for determination of amino acid composition: application to bacteriorhodopsin

A new double-labelling procedure for amino acid analysis which requires only routine chromatographic equipment is described. When 1-fluoro-2,4-dinitro[3H]benzene is reacted with a mixture of 14C-labelled amino acids followed by reaction with the same 14C-labelled amino acid mixture diluted with an unlabelled sample of amino acids, the 3H:14C ratio in the resulting 2,4-dinitrophenyl (DNP) amino acid derivatives of the diluted sample will be increased in proportion to the quantity of unlabelled amino acid in the diluted sample. This procedure gave reliable results when applied to the known proteins insulin and lysozyme. The procedure is most advantageous when applied to amino acids which are unstable during acid hydrolysis or present in low molar fractions. When applied to the analysis of the bacteriorhodopsin in Halobacterium cutirubrum, this procedure showed the presence of one histidine residue and four tryptophan residues per mole protein but no cystine or cysteine; in general, the analyses obtained were consistent with those originally reported by Oesterhelt, D. and Stoeckenius, W. (1971) (Nature (London) New Biol. 233, 149-152) for bacteriorhodopsin of H. halobium.     

A procedure for amino acid sequencing in internal regions of proteins

We describe a novel procedure for determining the amino acid (aa) sequence of the internal regions of proteins. This procedure has been implemented by directly determining the sequence of aa 65-75 of the product of the trpR gene of Escherichia coli, the trp repressor. This method is based on the insertion of the cleavage site of a specific protease (factor Xa) into the protein immediately before the region to be sequenced by Edman degradation. The simplicity of the procedure makes it appealing for studies of protein structure-function relationships, and of the expression of genetic information. The method is particularly useful when there is ambiguity concerning the co-linearity of the aa and nucleotide sequences.     

Improving protein pharmacokinetics by genetic fusion to simple amino acid sequences

The role of primary amino acid sequences in protein pharmacokinetics, an issue of relevance in both basic knowledge and biotechnology, was addressed here using as a starting point two repetitive antigens from the hemoflagellate Trypanosoma cruzi that are known to stabilize their associated proteins in the bloodstream. A major drawback to their pharmacological application is that these repetitive sequences are highly immunogenic, being therefore the deletion of this characteristic desirable. Based on sequence homology and epitope mapping analyses, an artificial repetitive sequence (PSTAD) was engineered. This motif was tested by genetic fusion to the C terminus of both the trypanosomal trans-sialidase and the rat tyrosine aminotransferase and found to produce a 4.5-6-fold increase in the half-life of the associated proteins in blood while displaying significantly lower immunogenicity. Residues involved in the stabilizing properties of the novel peptide were mapped by a site-directed mutagenesis approach, allowing us to successfully identify another two motifs. Searching databases for sequences displaying some homology, embedded in proline frameworks and associated to shed virulence factors from unrelated microorganisms, resulted in the identification of four other protein extensions. Remarkably, three of them (from Streptococcus pneumoniae, Actinomyces viscosus, and Escherichia coli) revealed similar pharmacokinetic features, suggesting therefore an analogous evolutionarily acquired mechanism to ensure the biodistribution of their corresponding proteins. Our findings indicate that the insertion of defined motifs into a proline-rich framework constitutes a suitable alternative to construct a chimeric protein with extended half-life in blood.     

Partial amino acid sequences of mouse transplantation antigens

The partial N-terminal amino acid sequences of the K and D gene products from theH-2 ^q andH-2 ^s haplotypes are presented. These data in conjunction with data already published demonstrate striking homology relationships among the transplantation antigens of mouse and other species. Moreover, these new data support the presence of certain sequence patterns noted in earlier sequence studies (e. g. no Kness or Dness, species-associated residues, and complex allotypes). These patterns place interesting constraints on the genetic organization and evolutionary history of the genes encoding the transplantation antigens which are discussed in this report.Abbreviations used in this paper HEPES N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid - HPLC High performance liquid chromatograph(y) - MEM Minimum essential medium - PTH Phenylthiohydantoin - SDS Sodium dodecyl sulfate     

Availability of short amino acid sequences in proteins

Much attention is being paid to protein databases as an important information source for proteome research. Although used extensively for similarity searches, protein databases themselves have not fully been characterized. In a systematic attempt to reveal protein-database characters that could contribute to revealing how protein chains are constructed, frequency distributions of all possible combinatorial sets of three, four, and five amino acids ("triplets," "quartets," and "pentats"; collectively called constituent sequences) have been examined in the nonredundant (nr) protein database, demonstrating the existence of nonrandom bias in their "availability" at the population level. Nonexistent short sequences of pentats were found that showed low availability in biological proteins against their expected probabilities of occurrence. Among them, six representative ones were successfully synthesized as peptides with reasonably high yields in a conventional Fmoc method, excluding the possibility that a putative physicochemical energy barrier in forming them could be a direct cause for the low availability. They were also expressed as soluble fusion proteins in a conventional Escherichia coli BL21Star(DE3) system with reasonably high yield, again excluding a possible difficulty in their biological synthesis. Together, these results suggest that information on three-dimensional structures and functions of proteins exists in the context of connections of short constituent sequences, and that proteins are composed of evolutionarily selected constituent sequences, which are reflected in their availability differences in the database. These results may have biological implications for protein structural studies.     

Prediction of gamma-turns from amino acid sequences.

We predicted gamma-turns from amino acid sequences using the first-order Markov chain theory and enlarged representative data sets corresponding to protein chains selected from the Protein Data Bank (PDB). The following data sets were used for training and deriving the probability values: (1) an initial data set containing 315 protein chains comprising 904 gamma-turns and (2) a later data set in order to include new entries in the PDB, containing 434 protein chains and comprising 1053 gamma-turns. By excluding 93 protein chains that were common to these two training data sets, we generated two mutually exclusive data sets containing 222 and 341 protein chains for testing our predictions. Applying amino acid probability values derived from training data sets on to testing data sets yielded overall prediction accuracies in the range 54-57%. We recommend the use of probability values derived from the data set comprising 315 protein chains that represents more gamma-turns and also provides better predictions.     

A single amino acid change makes a rat neuronal sodium channel highly sensitive to pyrethroid insecticides

Vitamin K-dependent protein S, which is a cofactor for activated protein C and thus important for down-regulation of the coagulation cascade, contains several Ca(2+)-binding sites with unusually high affinity. The 89 amino acid fragment constituting the third and fourth epidermal growth factor-like (EGF) modules of protein S is the smallest fragment that retains high-affinity Ca(2+) binding and is therefore useful for investigating the structural basis of this property. Heteronuclear multidimensional nuclear magnetic resonance experiments were used to obtain extensive assignments of the (1)H, 15N and (13)C resonances of the module pair with one Ca(2+) bound in EGF 4. In addition, nearly complete assignments of the (1)H resonances of the isolated Ca(2+)-free EGF 3 module were obtained. The assignment process was complicated by broadening of several resonances, spectral heterogeneity caused by cis-trans isomerisation of the peptide bond preceding Pro-168, and dimerisation. Analysis of weighted average secondary chemical shifts, (3)J(HNHalpha) coupling constants, and NOE connectivities suggest that both EGF modules in this fragment adhere to the classical secondary structure of EGF modules, consisting of one major and one minor anti-parallel beta-sheet.     

A neural network model for the prediction of membrane-spanning amino acid sequences

The architecture and weights of an artificial neural network model that predicts putative transmembrane sequences have been developed and optimized by the algorithm of structure evolution. The resulting filter is able to classify membrane/nonmembrane transition regions in sequences of integral human membrane proteins with high accuracy. Similar results have been obtained for both training and test set data, indicating that the network has focused on general features of transmembrane sequences rather than specializing on the training data. Seven physicochemical amino acid properties have been used for sequence encoding. The predictions are compared to hydrophobicity plots.