Do immunoglobulins have proteolytic activity?

Abstract

As more and more proteins are sequenced, the search for sequential and structural interrelationships among them is becoming a major activity, even though it is still in its infancy compared with investigations dealing with ancestral relationships^1,2. For our study of homologous regions found among ancestrally unrelated proteins we have modified McLachlan's double matching method³ and used it to find the statistically significant homologies. The method, the details of which are being published elsewhere⁴, enables a computer to perform a sliding match between two proteins, either when both proteins are aligned from N-terminal to C-terminal, which we call ‘normal matching’, or when one is aligned from N-terminal to C-terminal while the other is aligned from C-terminal to N-terminal, which we call ‘reverse matching’. At each step during this sliding match, some amino acids will be found to be identical between the two proteins being matched and some others will be different. Based on similarities found among ancestrally related proteins the probability of each amino acid being substituted by another is calculated—the M score of similarity³—and printed. The scores reach from 0, a non-match, all the way to 8 and 9 both of which stand for exact matches, 9 being assigned to phenylalanine, cysteine, tyro-sine and tryptophan. For ease of detection, the printout has a stop (˙) for a score of 8 and a prime (′) for a score of 9 (ref. 4). The same molecule can also be matched against itself to reveal the presence of repeating subsequences and symmetry patterns⁵.