Electrostatic Recognition between Enzyme and Inhibitor: Interaction between Papain and Leupeptin

Electrostatic forces are involved in a w ide variety of molecular interactions that are of biological interest, including, among others, DNA–Protein interactions, protein folding, and the interactions bet w een enzymes and their substrates and inhibitors. In this w ork, the interaction bet w een papain and an inhibitor, leupep- tin, is analyzed from the point of vie w of their electro- static interaction. The computations enable one to suggest that negatively charged amino acids located in the region of the active site are responsible for creating an environment that enables efficient bind- ing of the inhibitor. This binding occurs despite the fact that the net global charge of both molecules is positive; an explanation for this apparent contradiction is proposed.

Key Words: electrostatics; inhibitors– enzyme complexes.

tin (C-Leu-Leu-Arginal) (7), a relatively nonspecific blocker that inhibits other cysteinic proteases as well as such serine proteases as palmina and tripsine. To date, leupeptin has been studied primarily in the con- text of its actions in metastasis, protein synthesis, and muscular dystrophy (2).
At physiological p H, papain has a net positive charge of 10 electron units, while leupeptin carries a positive charge of 1 electron unit. Without taking the detailed electrostatic profiles, and other possible effect s, into account, the mutual charge of the enzyme and the inhibitor would be expected to make their interaction highly unfavourable, yielding a very low reaction rate. However, the very fact that the inhibitor gets into the active site and binds to the enzyme via hydrogen bonds (3) is strong evidence that, despite their overall net charges, papain and leupeptin are able to efficiently form a “tight” complex with a K d = 10 —9 mol/) (4).

Apparent electrostatic incompatibility between molecules forming complexes is not uncommon. For inPapain is a sulfhydric protease ( ~23 kDa) extracted from the latex of Carica Papaya. First isolated in crystalline form by Balls et al. (1), papain is a prime exam- ple of the family of cysteinic proteins. As a cysteinic protease with a thiol group at its active site, moreover, it is an excellent structural model for this group of enzymes known to participate in the degeneration of muscle proteins (2).

Over the years, a wide variety of papain-inhibitor complexes (5, 6) have been studied with the aim of better understanding the tertiary structure of the en- zyme and the specific activities of its inhibitors. Among these papain inhibitors is the aldehyde peptide leupepstance, the binding sites of phosphate-binding proteins from Escherichia coli and the sulfate-binding protein from Salmonella typhimurium all have negative sur- face potentials yet bind anionic ligands (8). Moreover, superoxide dismutase (SOD), one of the faster acting enzymes, has been carefully analyzed by Honig et al. who foun d that it too falls into this category (9, 10). By taking advantage of the computer’s computational power and a sophisticated approach to examining the electrostatic problem, it has been possible to carry out detailed analysis of the electrical potential profiles of various molecules and to propose sensible explanations of the electrical interactions taking place during the formation of complexes (9 –11). It is thus clear that important features of protein complex formation can be identified through careful study of electrostatic inter- actions (10, 12).

FIG. 1. Electrostatic potential profile of papain. The isopotential curves shown are located in a plane along the active site crossing the sulfur atom. Note the negative channel leading from the surface to the active site.

The purpose of this work was to analyze the electro- static profiles of papain and the papain–leupeptin com- plex in order to more precisely understand how this interaction is accomplished.

COMPUTATIONAL METHOD

The electrostatic potential was computed using of the MOLPOT program (13), which makes use of the boundary element method to solve the electrostatic problem. Briefly, the program considers in three dimensions a closed surface that contains the macromolecule in question, following its shape as closely as possible. Inside the surface (Region I) there are N charges q i located at points r i(i=1, 2 … N). Within Region I, the electrical potential is φ1 and the permittivity is ϵ1; outside region I (within Region II), the electrical potential is φ2 and the permittivity is ϵ2. In Region I, Poisson’s equation is satisfied and we have where n is the inverse Debye length.Using the appropriate boundary conditions (continuity of the elec- trical potential and continuity of the normal of the electrical dis- placement) and the additional condition of regularity of the electrical potential, ad infinitum, Eqs. [1] and [2] can be solved analytically, but only for simple geometry. General cases, however, can be solved numerically, which is what MOLPOT does.

In the present case, charges were assigned as atomic partial charges, according to the GROMOS force field (14) located at the center of each atom. The dielectric permittivity of the macromolecule (Region I) was considered equal to 2.0 and that of the solvent (Region II) equal to 78.54 —i.e., the water permittivity at 298.16 K. The Debye screening length was taken to be 0.96223 nm, which corre- sponds to a physiological solution. The protein coordinates for papain (15) and the papain–leupeptin complex (3) were taken from the Brookhaven Protein Data Bank (ID Codes 9PAP and 1POP, respec- tively).

Figure 1 shows a map of the electrostatic potential of papain depicted as an isopotential contour in a plane.

FIG. 2. Electrostatic profile of the papain–leupeptin complex (the same plane as Fig. 1). Note that the negative channel seen in the isolated protein is now absent due to the interaction of leupeptin’s charges with those of the enzyme.

The existence of a channel having a negative electri- cal potential makes it possible for a substrate carrying a positive potential (e.g., leupeptin) to enter the active site, provided the site is located in the mouth of the channel. Similar situations occur with other enzymes (e.g., SOD), and it was thought (10) that charge screen- ing reduces the extent of the off-zero potential region without substantially affecting its shape. In our case, however, and most likely in the case of SOD as well, that would mean the efficiency of the enzyme could not be very high, as the channel cross section is small, and only substrate molecules approaching the enzyme from the right direction would be attracted along the chan- nel. It thus seems apparent that other effect s also need to be considered.

At physiological p H, papain has a net charge of +10, and its inhibitor, leupeptin, a net charge of +1. Since the two proteins do in fact interact, with the inhibitor entering the enzyme’s active site and binding to it via hydrogen bonds, a mechanism must exist by which the repulsive effect of the like charges is circumvented. It is clear that the local potential at the active site has a particular profile that the electrostatic potential shows the presence of a negative channel oriented toward the active site. This phenomenon is attributable in part to the distribution of negatively charged amino acids (e.g., ASP 158) and perhaps more importantly to the sulfur belonging to CYS 25, which is situated within the tertiary structure such that the electrostatic profile includes a region of negative potential within the active site. Figure 3 shows a detail of the active site in which the charges of the residues that form the site can be seen, as can the size of the channel.

FIG. 3. Stereo view of the binding site showing the charge of residues. The charges are represented by colors going from white, for the more negative charges to green, for the more positives.

FIG. 4. Detail of the active site of papain. The dipole of the helix facilitates proton transfer from cystein 25 to hystidine 159, enabling the sulfur atom to contribute to the electrostatic recognition.

Although the sulfur in the CYS is normally proton- ated, the presence of an α-helix spanning amino acids 24 to 43 facilitates proton transfer from CYS 25 to HIS 159, negatively ionizing the CYS and neutralizing the HIS. The presence of the “macro” dipole of the α-helix (17) changes the charge distribution, which allows the sulfide group to contribute to electrostatic recognition, thereby increasing the reactivity of the enzyme (see Fig. 4). The tiolate ion formed is thus the crucial point in the initial catalytic step, because it is the primary determinant of the particular electrostatic profile present at the active site.

To obtain a complete picture of the process, it is necessary to also take into account the dynamical ef- fects. Although the working details have not yet been fully described, it is known that there is an important mechanical component to enzyme activity. For in- stance, the effect of counter ions cannot be fully ex- plained by Debye–Hu¨ ckel’s theory if dynamical effect s are involved. For a conducting medium in the presence of oscillating fields, however, there is a critical fre- quency (Maxwell’s frequency ωM) above which the sub – stance behaves as a dielectric. The crossover from a conducting to a dielectric regime occurs in a physiolog- ical medium at about 150 MHz (16), which can be reached by some charged groups within proteins, but not by all. Those charges that exceed Maxwell’s fre- quency will, at least partially, escape the screening due to counter ions.

It is difficult to obtain dynamic information using diffraction techniques, but one can make a qualitative estimation from the so-called thermal B factor. With respect to papain, we know the negative atoms around the active site have a larger B factor than the rest of the atoms in the region, and that the B factor declines when the isolated enzyme form s a complex. In partic- ular, oxygen atom OD2 of ASP 158 goes from 27.5 Å 2 to
18.5 Å 2 in moving from one state to the other (15, 3). The movements of atoms are not the sole determinant of the value of the B factor, however. Consequently, additional information will be needed to precisely de- scribe the circumstances surrounding the binding of papain and leupeptin. Nevertheless, the present findings would seem to indicate that greater mobility may be a key characteristic of atoms involved in molecular recognition. Recently (18) this effect s have been shown for the SOD, which make this phenomenon as quite probable for the present case.

CONCLUSIONS

The present results show the importance of consid- ering the local potential rather than the global net charge when assessing the interaction between an en- zyme and its ligand. They also emphasize the impor- tance of identifying the amino acids that play key roles in the process of electrostatic focu sing of the ligand. It is important to note, however, that these arguments are based solely on the static distribution and that although we have mentioned the possible effect of charge movement on the screening of charges, this presumably important effect cannot be properly as- sessed with the available data.