The kinetics of the reaction of alpha 2-macroglobulin (alpha 2M) with human thrombin were studied by recording the appearance of thiol groups spectrophotometrically and by measuring the distribution of protein species by denaturing non-reducing gel electrophoresis. The goals were to study the relation between the formation of various covalent enzyme-inhibitor complex species and the appearance of free thiol, and from the kinetic analysis, to try to characterize the chemical nature of the protein complexes. The kinetics of thiol-group release were observed to be biphasic, the early phase showing second-order behaviour, results consistent with previous reports in the literature. The observed second-order rate constant for thiol-group release was found to be faster than the second-order rate constant for the disappearance of the band corresponding to native alpha 2M on gel electrophoresis. This may be a reflection of the multiple products formed from the thioester. Alternatively, it is possible that covalent-bond formation is slower than some enzyme-induced change in the thioester centre, and this may be suggestive evidence for a reactive alpha 2M centre that does not contain an intact thioester. The kinetics of covalent-bond formation were found to be consistent with the internal cross-link of several alpha 2M chains by the bound proteinase, providing further evidence that the very-high-Mr species seen on gels may arise from dimers of the alpha 2M molecule held together by covalent bonds to the enzyme.

No abstract

The primary observation, from our laboratory and others, of the effect of blocking the lysyl amino groups of enzymes is the reduction in the fraction of complexes that are resistant to SDS. The blocked enzyme derivatives do cause the specific proteolysis of the alpha 2M subunit to the 85K/100K fragments, and do cause the appearance of new thiol groups. With respect to the sequence of reaction, we may summarize the results by saying that if the reversible DMM-trypsin is, in fact, a model for the native enzyme, proteolysis can precede formation of the presumed covalent bond between bound enzyme and inhibitor. If our preliminary observations are borne out by later experiments, thiol release may precede covalent bond formation or loss of reactivity with amines, suggesting that an intact thiolester need not be the immediate target for amines; another intermediate, possibly the internal pyroglutamate originally proposed by Howard et al. and seen in model studies, may be an additional, or even the primary, target for covalent bonding with native enzymes. With regard to the "trap" hypothesis, the limited release of thiols in a slow phase is suggestive of enzyme activity within the alpha 2M-protease complex, consistent with the theory. Noncovalent irreversible complexes, however, are not a necessary part of associations seen with lysyl-blocked enzymes (which do cause proteolysis and do release thiols); this result is supported by limited data with noncovalently bound native enzymes. Some fraction of irreversible noncovalently bound enzymes may occur, but our results suggest that although alpha 2M-bound enzymes are unusually sterically hindered, the transformation to the presumed covalent state that appears to depend on intact amino groups, may be sufficient to explain the low dissociation of native enzymes. We feel that more experimental evidence is needed to resolve some of the ambiguities on this question but, we feel the existence of a "trapping" reaction has not been proved. In fact, given the possible existence of equilibria between covalent and noncovalent complexes observed, for example, in soybean trypsin inhibitor, and the very low dissociation constants observed with traditional protein-protein complexes, the question of physically encapsulated structures in alpha 2M may not be resolvable without direct evidence from crystal structures.