Purification and characterization of galanin from the phylogenetically ancient fish, the bowfin (Amia calva) and dogfish (Scyliorhinus canicula)

The isolation of a novel biologically active peptide, designated galanin, is described. The peptide was discovered by the detection of its C-terminal amide structure in porcine intestinal extract using a chemical method. It was found that galanin consists of 29 amino acids and the complete amino acid sequence is: Gly-Trp-Thr-Leu-Asn-Ser-Ala-Gly-Tyr-Leu-Leu-Gly-Pro-His-Ala-Ile-Asp-Asn-His -Arg-Ser -Phe-His-Asp-Lys-Tyr-Gly-Leu-Ala-NH2. Galanin was found to contract smooth muscle preparations from the rat and to cause a mild and sustained hyperglycemia in dog.

Many regulatory peptide precursors undergo post-translational processing at mono- and/or dibasic residues. Comparison of amino acids around the monobasic cleavage sites suggests that these cleavages follow certain sequence motifs and can be described as the rules that govern monobasic cleavages: (i) a basic amino acid is present at either 3, 5, or 7 amino acids N-terminal to the cleavage site, (ii) hydrophobic aliphatic amino acids (leucine, isoleucine, valine, or methionine) are never present in the position C-terminal to the monobasic amino acid at the cleavage site, (iii) a cysteine is never present in the vicinity of the cleavage site, and (iv) an aromatic amino acid is never present at the position N-terminal to the monobasic amino acid at the cleavage site. In addition to these rules, the monobasic cleavages follow certain tendencies: (i) the amino acid at the cleavage site tends to be predominantly arginine, (ii) the amino acid at the position C-terminal to the cleavage site tends to be serine, alanine or glycine in more than 60% of the cases, (iii) the amino acid at either 3, 5, or 7 position N-terminal to the cleavage site tends to be arginine, (iv) aromatic amino acids are rare at the position C-terminal to the monobasic amino acid at the cleavage site, and (v) aliphatic amino acids tend to be in the two positions N-terminal to and the two positions C-terminal to the cleavage site, except as noted above. When compared with a large number of sequences containing single basic amino acids, these rules and tendencies are capable of not only correctly predicting the processing sites, but also are capable of excluding most of the single basic sequences that are known to be uncleaved. Many of these rules can also be applied to correctly predict the dibasic and multibasic cleavage sites suggesting that the rules and tendencies could govern endoproteolytic processing at the monobasic, dibasic and multibasic sites.

From acid/ethanol extracts of surgical specimens of human large intestine we isolated two peptides, in approximately equal amounts, that reacted with an antiserum against porcine galanin. By amino acid analysis, sequence analysis and mass spectrometry, the larger of the two peptides was found to consist of 30 amino acid residues, the sequence of which was identical to that of porcine galanin except for the following substitutions: Val16, Asn17, Asn26, Thr29 and Ser30. Unlike porcine galanin, the carboxy-terminus was not amidated. The smaller peptide corresponded to the first 19 amino acid residues counted from the N-terminus of the 30 residue peptide (again without amidation). The structural analysis was repeated on another batch of tissue with identical results. By HPLC analysis of extracts of specimens from a further 4 patients, the same peptides were identified. Thus, human galanin includes two peptides of 19 and 30 amino acids that share the sequence of the N-terminal 15 residues with other mammalian galanins, but exhibit characteristic differences in the remaining part of the molecules.