Short length amino acid sequence homology among ancestrally unrelated proteins

A growth regulatory factor that can both inhibit and stimulate growth

A growth inhibitor that is produced by BSC-1 cells (African green monkey kidney epithelial cells) has been isolated from conditioned medium. It has been purified by gel chromatography and high performance liquid chromatography. It appears to be a protein with a relative molecular mass (Mr) of 24 000. It is extremely active as a growth inhibitor with some cells, but not with others. Approximately 50% inhibition of thymidine incorporation is observed with CCL64 cells at 0.05 ng/ml and with BSC-1 cells at 1 ng/ml. The growth inhibitor induces BSC-1 cells to synthesize and secrete a glycoprotein of approximately 48 000 Mr. It inhibits Na+ accumulation in BSC-1 cells. Recently, in collaboration with R. F. Tucker, G.D. Shipley and H. L. Moses (Mayo Foundation & Medical School), we have found that the growth inhibitor is very similar to and may be identical with transforming growth factor beta (TGF-beta). Our growth inhibitor stimulates colony formation in soft agar by AKR-2B cells, and it competes with TGF-beta in binding to cell surface receptors. TGF-beta, from human platelets, is extremely active as an inhibitor of thymidine incorporation by BSC-1 cells and CCL64 cells. The growth inhibitor/TGF-beta can, therefore, stimulate or inhibit growth, depending on the cells and the growth conditions.

Linear repetitions of amino acids and convergent evolution inside protein subregions of ordered secondary structures

51 polypeptides of known 3-dimensional structures have been submitted to a search for internal similarities. It is shown that the frequency of proteins displaying significant amounts of internal similarities is higher than predicted by chance. A non-negligible part of those similarities probably occurs in connection with the existence of ordered secondary structures. Indeed, similarity occurs at a much more important rate when analyses are restricted to protein subsequences corresponding to alpha helices or beta pleated sheets. Furthermore, the correlation existing between the rates at which linear and inverted repeats occur inside protein subregions of ordered secondary structures suggests that a significant part of short similarities are analogies rather than homologies. An hypothesis is put forward suggesting that the regular alternations of hydrophobicity which characterize most of alpha helices and beta strands could provoke the occurrence of significant amounts of similarities inside protein sequences.

The sharing of amino acid short spans by ancestrally unrelated proteins may be the result of ubiquitous alpha and beta secondary structures

Short homologies are often found when genetically unrelated proteins are compared but it is not known whether the rate at which they occur is or not above randomness. Comparing 190 pairs of unrelated proteins enable us to show that the frequency at which pairs of unrelated proteins share little spans of amino acids is compatible with chance. However, it appears that those short homologies are mainly located within protein subregions of identical secondary structure: the frequency at which pairs of unrelated proteins exhibit related spans of amino acids inside subregions of identical secondary structure is far above randomness. Those data suggest that the sharing of related spans of amino acids by genetically unrelated proteins could result from structural constraints imposed by the alpha or beta secondary structures.

Probable mechanism of enzyme evolution: how did EBG of E. coli originate?

A mechanism is proposed for the formation of Ebg-evolved beta galactosidase-of E. coli based on the following assumptions: 1. In the presence of lactose, certain proteins being translated bind to their m-RNA-ribosome complexes; 2. This binding interferes with the release of m-RNA from the bacterial chromosome, marking the gene; 3. Thereupon a cytosine specific methylase and methyl cytosine deaminase pair, modify - mutate - the marked gene; 4. The result, after five or so mutations, is a new gene capable of coding for a different protein which can split lactose; 5. I propose that this enzyme pair has evolved to produce mutations internally, when need arises, as is the case here; 6. This may be a general mechanism through which drug resistance and detoxification of a novel chemical, could be achieved in bacteria; 7. All of these ideas are experimentally testable.

Attempts at simulating evolution by a computer: I. evolution of proteins under prebiotic conditions

The effect of random mutations of five identical pentapeptides that have been inserted arbitrarily along a 100 amino acid-long protein chain has been studied by computer simulation. The method used was the application of mutation probability matrix for 2 PAMs of Dayhoff (1972) repeatedly to obtain the desired length of time. The results indicated that, given sufficient length of time, even the identical peptides could become drastically altered and in order to recognise them as stemming from the same origin one has to use reasonable statistical significance thresholds.

Structural analysis of the molecular evolution of some gastro-entero-pancreatic hormones

By means of a statistical analysis of the occurrence of amino-acid residues in the polypeptide chains of several gastro-entero-pancreatic (GEP) hormones an investigation was undertaken to determine whether any of these hormones might be related to each other--possibly from an evolutionary point of view. Particular interest was paid to the occurrence of small charged segments, i.e. those with acidic or basic amino acid residues, since such segments can be presumed to play a role in hormonal receptor binding mechanisms. By this method hormonal relationships were suggested by the observation that these small charged amino-acid sequences, contained in the hormonal structures, match as a result of non-randomness. It was found that hagfish and human insulin were related on a molecular level not only to the newly discovered (avian, bovine, human) pancreatic polypeptide (PP) but also to some other GEP hormones (VIP, GIP, glucagon) as well as to calcitonin and to the alpha-subunit of the glycoprotein hormones. Interpretation of the statistical data suggests that all these peptide hormones are related by a common hexapeptide sequence which contributed, at an evolutionary point, to their molecular architecture. A hexapeptide segment of APP is statistically related to a sequence of equal size in the carboxy terminal region of the A-chain of both hagfish and human insulin, providing the first instance of their structural similarity. Correlations between PP, insulin, glucagon, VIP, and calcitonin provide a tentative basis for predicting the production of one or more of these peptide hormones by immature or de-differentiated cells of neoplasms and non-neoplastic pathologic lesions of the GEP endocrine system.

Symmetry patterns in trypsinogen

When the primary structure of bovine trypsinogen is searched for the existence of regularities, according to Greller & Erhan (1974), one finds eight pairs of peptides, arranged in a symmetrical pattern along the molecule. These peptides cover 49% of the length of the molecule-112 of the 227 amino acids-and each pair folds in a similar way. This observation is in agreement with the observation that "Trypsin folds into two halves, each of which contains a pseudo-cylindrical arrangement of hydrogen bonds...", Stroud et al. (1971). Thus the above mentioned method is capable not only of detecting regulatities along the primary structure but also of predicting the folding of a protein.