MECHANISMS OF ENZYME RECYCLING AND HYDROGEN TRANSFER

Proton transfer is by far the most common process in reactions that enzymes catalyze. The chemical reaction itself, with few exceptions, leaves the enzyme in a different state of protonation than that required to initiate another reaction cycle. Proton recycling to the initial state may occur by internal transfer or by transfer(s) between the enzyme and solvent. Such transfers may come before or after free product is formed. Rates of proton transfer that occur between solvent and the acidic and basic residues of the enzyme are determined by different principles than those that operate in catalytic reactions. It can be expected that evolution will tend to maximize the rates of both components of the reaction cycle. Recently, we showed with fumarase that proton recycling is the rate-limiting process of a reaction of <10³ sec^-1, not a very fast rate. Our goal has been to characterize this process more fully.

FUMARASE: PROTON TRANSFER AND RECYCLING. ROSE, KUO

The fumarase reaction, which converts malate to fumarate, requires two proton transfers as shown:

Scheme 1

An acid supplies a proton to the C-2 hydroxyl of malate forming a readily dissociated water, and a basic group of the enzyme abstracts the C-3 proton which is more tightly bound than the product fumarate. Loss of this proton and recovery of the acid proton are required for recycling. The nature of the residues and the details of the reaction chemistry are still unknown. Previous studies using tritiated malate showed that almost all the label from the C-3 position could be captured by added fumarate indicating that tritium dissociated after fumarate was produced. From the amount of fumarate required for tritium capture, the rate of dissociation of this proton is expected to be a limiting factor in the recycling process. Imidazole buffer was shown to increase the malate to fumarate (M-->F) conversion and to increase the rate of dissociation of this proton.

Since the proton required for C₂-OH abstraction cannot be followed isotopically, other methods needed to be devised. The enzyme formed in the F-->M direction, E_M, was readily captured by labeled M (M*), which was converted to labeled F (F*) in the presence of high levels of F, a process called counterflow. The recycling rate of this form is slow enough to be rate-determining for the cycle and was shown to be activated by imidazole. E_M binds mesotartrate as a non-competitive inhibitor of the F-->M conversion, except when imidazole is present and inhibition becomes competitive. This shows that E_M does not bind F and is decreased greatly by buffer action. It seems reasonable to suppose that the malate specificity and slow recycling rate are due to the water-derived proton used to deliver -OH and form malate giving E^H (see Scheme 1). Conversion of E^H to E_H, the fumarate-specific form, may follow three pathways: E^H-->E--> E_H, E^H-->E_H (internal transfer), and E^H-->E^H _H-->E_H. Only the last is consistent with the nature of the buffer effects, general acid catalysis, which is found in both directions.

The two-step recycling pathway has an intermediate that may react with one or either of the substrates at high concentration to bypass the second recycling step. Thus, if high F reacted with E^H_H the new recycling path would result:

Scheme 2

in which the second proton would be lost from E^H_HF-->E_H F. This scheme might explain the cooperative kinetics of fumarase and the 2- to 4-fold activation above predicted V_max that is found in both directions. To test for this, product counterflow rates were measured at different concentrations of substrate in the activating range. There was no effect of [F] on M*-->F* either in the presence or absence of imidazole. However, high malate decreased counterflow F*-->M* greatly. As reported previously, high malate makes it more difficult for fumarate to capture the tritium from malate labeled at C-3. We have not been able to find an inhibitor that is non-competitive with respect to malate, which would indicate a fumarate specific intermediate. Thus, it seems that H recycling in the M-->F direction involves only one free enzyme intermediate, most likely the non-specific form E^H_H as in Scheme 3:

Scheme 3

Since H₂O readily exchanges from H_H^H₂^O F E_HF as shown by 2-¹⁸O-malate/ H₂O exchange studies, it is likely that it can easily be replaced by a proton before F dissociates. This leads to the unusual result in which the pathways followed in two directions are different, which may explain the puzzling finding that many simple anions, such as chloride, activate the conversion of F-->M but do not affect M-->F.

Enzymes that recycle slowly will have a number of unusual properties that derive from the fact that they are largely in the unliganded state at substrate saturation in one direction. To the extent that proton transfers are involved, one should find buffer catalysis at all concentrations of substrate, and buffer should not affect the slopes of double reciprocal plots, only the intercepts. Slow recycling can be used to dynamically couple a favorable to an unfavorable conversion, and where proton transfers are involved, one could apply a proton gradient to the same purpose. To identify the residues of enzymes that are unique to the recycling function, one should search for mutations that knock-out the catalytic rate, which includes recycling, but have no effect on equilibrium exchange or isotopic counterflow rates.

Paper in press at time of previous report:

ROSE, I.A. Isotopic strategies for the study of enzymes. Protein Sci. 4: 1430-1433, 1995.