Primary structures of aspartate aminotransferases.

three residues seems to be common to all the serine enzymes and has become known as the ‘charge-relay system’. Its function is to transfer a negative (or partial negative) charge from the aspartic acid residue to the oxygen atom of the hydroxy group of serine. This obviously increases the nucleophilicity of the serine, but the important point is that the charge is developed in the absence of bulk solvent. It has been known for a long time that nucleophiles of the type OR are much stronger in aprotic dipolar solvents than in solvents to which they can form hydrogen bonds (of which water is the prime example). The reason is that the charge is not dispersed by hydrogen-bonding in aprotic dipolar solvents (for a review see Parker, 1969). Hence the catalysis produced by chymotrypsin can be understood, at least in qualitative terms, by the generation ofcharge on the oxygen atom of the hydroxy group of the active serine in an essentially non-aqueous environment. A somewhat similar device has been revealed by crystallographic studies of the acid proteinase penicillinopepsin (James et al., 1977). It appears that there is a single water molecule in the active site in close proximity to the substrate. This water molecule can lose a proton to a neighbouring aspartic acid residue and this, of course, generates a powerful nucleophile in an essentially non-aqueous environment. To return to aspartate aminotransferase: a recent crystallographic study of the mitochondria1 enzyme from chicken appears to show that the prototropic rearrangements of the azomethine intermediates are controlled by the &-amino group of a single lysine residue and that this group acts as both proton acceptor and proton donor (Ford et al., 1980). The enzyme used in this study comes from a source different from that used in the kinetic studies described above, but the similarity between the members of this family of enzymes suggests that a common mechanism is involved. It is now necessary to enquire how it arises that this amino group can act as such a powerful base. In aqueous solution amines catalyse the prototropic rearrangement of azomethines, but the effect is relatively weak. Presumably the explanation is similar to that given above for the proteolytic enzymes; namely that in the absence of a hydrogenbonding solvent the basicity of the amino group is enormously increased. There are certainly precedents for this; halide ions in water are virtually non-basic, but in aprotic dipolar solvents they act as powerful bases (Parker, 1969). It would be interesting to study the catalysis of the rearrangements of the azomethine produced by amines in aprotic dipolar solvents, but as far as I know, such a study has not been made. It ought to be. The notion that the &-amino group of a lysine residue is both a proton acceptor and a proton donor might appear at first sight contradictory, since the first requires that the group is a very strong base and the second that the protonated form is a very strong acid. A way out of this difficulty might be to note that if a carbanion intermediate is formed this, because of the distribution of charge over the triad system, would itself be a very strong base. Further study, particularly using isotope techniques, is necessary. One may conclude that enzyme catalysis can be understood only if the microenvironment of the active site can be identified, together with its relationship to the reaction being catalysed (Doonan et al., 1970). The sequence of events in the aminotransferase reactions is clearly complex. It is clear that conformational changes in the azomethine derivatives occur during the complete cycle, but these do not, by themselves, explain catalysis. We need to know about the physical organic chemistry of these reactions and about the microenvironment produced by the active sites.