Mechanism of intramolecular activation of pepsinogen. Evidence for an intermediate delta and the involvement of the active site of pepsin in the intramolecular activation of pepsinogen.

Intramolecular pepsinogen activation is inhibited either by pepstatin, a potent pepsin inhibitor, or by purified globin from hemoglobin, a good pepsin substrate. Also, pepsinogen at pH 2 can be bound to a pepstatin-Sepharose column and recovered as native zymogen upon elution in pH 8 buffer. Kinetic studies of the globin inhibition of pepsinogen activation show that globin binds to a pepsinogen intermediate. This interaction gives rise to competitive inhibition of intramolecular pepsinogen activation. The evidence presented in this paper suggests that pepsinogen is converted rapidly upon acidification to the pepsinogen intermediate delta. In the absence of an inhibitor, the intermediate undergoes conformational change to bind the activation peptide portion of this same pepsinogen molecule in the active center to form an intramolecular enzyme-substrate complex (intermediate theta). This is followed by the intramolecular hydrolysis of the peptide bond between residues 44 and 45 of the pepsinogen molecule and the dissociation of the activation peptide from the pepsin. Intermediate delta apparently does not activate another pepsinogen molecule via an intermolecular process. Neither does intermediate delta hydrolyze globin substrate.

After at least 20 min at alkaline pH, the entire volume of each solution was assayed for remaining pepsinogen activity as described below. The potential proteolytic activity was assayed at 37" with globin as substrate.
To the 0.28 ml of alkaline pepsinogen solution, 0.72 ml of a solution containing 0.55 M citric acid and 0.10 M HCI were added in order to adjust the pH to 2.0. The remaining pepsinogen was allowed to activate for 30 min. One milliliter of 2.5% globin solution, pH 2.0, was added to each tube; then, after 20 min, 1 ml of 10% trichloroacetic acid solution was added. The mixture was filtered on Whatman No. 50 filter paper and the optical density of the filtrate was read at 280 nm against a blank to which 1 ml of 10% trichloracetic acid solution was added prior to the addition of the 1 ml of globin substrate solution. Use of this blank allowed quantification of the pepsinogen remaining after the initial activation because optical density due to soluble peptides produced during the two activation periods was included appropriately in the blank. In all assay situations, the net optical density at 280 nm was 1.2 or less, and the blank did not exceed 0.75 optical density. Measurement of Appearance of Proteolytic Activity by pH-stat Assay-The method was the same as that described in a separate work (5) except that pepsinogen rather than pepsin was added to the globin solution. In a typical experiment, 0.01 ml of 1.82 x lo-' M pepsinogen was added to about 1.8 ml of 1.6 x lo-' M globin at pH 2 and 37". The

Activation of Pepsinogen in Presence of Pepstatin
Pepsinogen (3 mg) was dissolved in 60 ml of distilled water, pH about 5.3, which contained 0.24 mg of pepstatin (about 4 mol of pepstatin/mol of pepsinogen). When methanol was used, 0.35 mg of pepstatin was dissolved in 60 ml of 5% by volume methanol in distilled water. Pepsinogen (3 mg) was separately dissolved in about 0.5 ml of Hz0 and added to the alcoholic pepstatin solution. The zymogen solutions then were cooled in a water bath to 0". In a manner similar to that of Rajagopalan et al. (12), the activation was achieved by the addition of 6 ml of ice-cooled solution containing 9 parts 2 N monochloroacetic acid and 1 part 1 N HCl. This acidified solution, pH 1.9, was allowed to stand at 0" for 10 min and then was transferred to a constant temperature bath of 11". After various incubation times, aliquots of this solution were frozen in a dry ice bath and lyophilized.
The dry residue was extracted three times with acetone. The protein residue was redissolved in 0.1 M NH,OH and lyophilized. An alternative procedure was used to test the above procedure. The solution from the 11" bath was adjusted to pH 4.4 and the protein was separated from the peptides on SE-Sephadex C-25 column (12 Determination of pepsinogen and pepsin was carried out by two methods.
First, the ratio of the two proteins in a mixture was determined from the ratio of phenylthiohydantoin-valine and phenylthiohydantoin-glycine in the second cycle of an automated Edman degradation.
The procedures and accuracy of this determination have been reported previously (4). Second, the pepsinogen activity was assayed independent of pepsin after alkaline inactivation of the latter. The details of this method also have been published (2

Pepsinogen Activation
Mechanism 7097 resulting pepsinogen concentration was 1 x lo-" M. The consumption of 0.04 M HCl as a function of time was recorded.
These experimental curves had a gradually increasing slope and ultimately attained a constant maximal velocity.
The slope of the tangent to these curves was deemed proportional to the pepsin concentration at that time. Maximum velocity was the same as that generated by a preactivated pepsinogen sample of the same concentration.
The ratio of the slope at a given time to the final maximal slope was multiplied by the initial pepsinogen concentration to yield the pepsin concentration.
Scheme 1 in Fig. 1 (4) where K,,, 01, 0, 6, G and 6. G are defined in Scheme 1 (Fig. l) ' The Greek letters chosen for the intermediates are based on their graphical similarity to the schematic characteristics of the intermediates. In these letters the circles are visualized as the active center of pepsin in pepsinogen and the extra line represents the activation peptide portion of pepsinogen. The 6 form then has the activation peptide outside the active center which can bind to either inhibitor or substrate.
Intermediate 0 has the activation peptide bound in the active center, an intramolecular enzyme'substrate complex. Intermediate 6 represents the hydrolyzed but yet undissociated activation peptide in pepsinogen.
In the (r form the active site is exposed but the conformation is different from that of 6. Therefore a plot of the reciprocal of the pepsinogen decay constant A uersus globin concentration should be linear and have an intercept whose reciprocal is equal to the pepsinogen decay constant in the absence of globin. Since in this treatment, the 6.G complex is a dead-end, the globin inhibition of the activation reaction is analogous to competitive enzyme inhibition. Scheme 2 of Fig. 1  to the hydrolysis of the peptide bond to form pepsin, then we can not distinguish with our measurements between the obligatory and the nonobligatory mechanisms (see "Discussion"). A more complicated system in which the pepsinogen.globin complex (6.G) can be transformed to pepsin (a) has been considered (see Scheme 3 in Fig. 1 shows that in this instance the plot of l/A' versus G is not linear but is a curve which approaches a maximum, which is llk'I,t, as G is increased. Here because the 6 .G complex is productive, albeit at a different rate, the inhibition is somewhat analogous to noncompetitive enzyme inhibition.

Znhibition of Pepsinogen Activation
by Pepstatin-Activation of 0.05% pepsinogen in the presence of about a 6-fold molar excess of pepstatin was almost completely inhibited (Table I). However, it was necessary to include 5% methanol in the zymogen solution to completely dissolve the pepstatin. Although the concentration of methanol used has been shown by Neuman and Shinitzky (13) to produce no inactivation of pepsinogen, a set of three experiments was carried out in the absence of methanol. In these experiments (Nos. 3, 4 and 5 in Table I), only about a 4-fold molar ratio of pepstatin to pepsinogen was used due to the low solubility of the inhibitor. As shown in Table I, a definite inhibition of pepsinogen activation by pepstatin also existed under these conditions.
Binding of Pepsinogen Intermediate to PepstatinlSepharose Column-Since pepsinogen activation is inhibited by pepstatin, we devised the following experiments to verify the binding of a pepsinogen intermediate to a pepstatin/Sepharose column. Fig. 2a shows that pepsinogen at pH 5.6 in 1 N NaCl was not retained by a pepstatin/Sepharose column and thus emerged in a breakthrough peak. Pepsin under the same conditions was retained and could be eluted from the column only as alkaline inactivated enzyme in pH 8.5 (Fig. 26). However, pepsinogen was absorbed on the pepstatin/ Sepharose immediately after being acidified to pH 2.2. (There was no significant breakthrough peak, Fig. 2c.) A peak was eluted in pH 8, 0.05 M sodium phosphate buffer (Fig. 2~). This peak, which represented about 30% of the original zymogen, had alkaline stability in pH 8 and had the same specific activity as the starting pepsinogen. The amino acid composition, NH*-terminal 2-residue sequence, and the mobility on polyacrylamide disc gel electrophoresis (Fig. 3) were identical to the native pepsinogen. Since the recovered pepsinogen was not eluted in pH 5.6 buffer with 1 N NaCl, the zymogen must have been absorbed on the affinity column as an intermediate which reverted to native pepsinogen at pH 8.
A second peak was eluted from the affinity column with pH 8 buffer and 1 N NaCl. This sample had an amino acid 7098 Pepsinogen Activation Mechanism center, material eluted in pH 8 buffer (first peak in Fig. 2~); right, material eluted in pH 8 buffer with 1 N NaCl  (second peak, Fig. 2~). The major band in this gel represents the denatured pepsinogen 0 (see text). A very light band slightly behind the major band is denatured pepsin.
mobility of about 85% of the native pepsinogen (Fig. 3). In addition, two minor bands with mobilities corresponding to those of alkaline denatured pepsin and native pepsinogen were present. When an unstained gel was sectioned, eluted, and assayed after electrophoresis, all the proteolytic activity was found in the native pepsinogen position. The alkalinedenatured pepsin must have been derived from a small amount of activation during mixing with the pepstatin/Sepharose. Native pepsinogen in the second peak was probably the result of "tailing" from the first peak. The major material, the heavy band in the third gel of Fig. 3, had no potential Pepsinogen Activation Mechanism 7099 proteolytic activity and must be an inactivated pepsinogen.
(We will call this species pepsinogen a). From the results described above, it can be concluded that, at pH 2, an intermediate form of pepsinogen was generated and bound to the pepstatin affinity column. In pH 8 buffer, only about 30% of the bound zymogen was recovered as native pepsinogen. About 70% of the bound zymogen was recovered as inactive pepsinogen 0. It is important to point out that the half-life of pepsinogen activation at pH 2, 28", is about 16 s (2). Even at a lower temperature, a large fraction of zymogen is activated within 2 h (see Table I). Therefore, inhibition of pepsinogen activation by Sepharose-bound pepstatin must have taken place.
It is interesting that pepsinogen 0 forms a single, well defined band in polyacrylamide electrophoresis. In contrast, alkaline or alcohol denatured pepsinogen each produced six or seven bands under the same electrophoretic conditions. This comparison suggests that inactive pepsinogen c assumed a rather uniform conformation which may resemble the conformation of the bound intermediate.
A dialyzed pepsinogen n fraction was completely absorbed again on the pepstatin/ Sepharose column and was recovered by elution with pH 8 buffer, 1 N NaCl. Attempts to regenerate potential proteolytic activity from pepsinogen 0 by dialysis in pH 8, 4", were not successful.
Effect of Globin on Pepsinogen Activation Rate-From the above experiments we deduced that a low pH form of pepsinogen is capable of binding pepstatin, a potent pepsin inhibitor. We then studied the activation of pepsinogen in the presence of high concentrations of globin, a good protein substrate for pepsin. Pepsinogen, 1 x lo-' M, was incubated at pH 2 in the presence of globin, whose concentration ranged from 1 x 10m5 to 28.41 x 10m5 M. Semilogarithmic plots of the per cent of pepsinogen remaining versus incubation time were linear (Fig.  4). All experiments except the slowest one were followed until at least one-half of the pepsinogen had been activated. The linearity of these plots gave credence to the assumption that the bond cleavage reaction described by k,,, was much slower than either of the conformational changes or the binding of globin to pepsinogen 6. Increasing globin concentration caused a dramatic decrease in the apparent first order activation rate constant (Table II). The reciprocal plot defined by Equation 6 is quite linear (Fig. 5). The linearity of the plot over a 20-fold change in globin concentration supports the reversible, nonproductive, competitive globin inhibition contained in Schemes 1 or 2 of Fig. 1. In the l/A uersus G plot, the intercept divided by the slope of this line is 1.71 x 10m5 M and should be equal to K&K, + 1). Unfortunately, we cannot calculate values of the constants Ki, and K, from this data. However, as described above, the inhibition of pepsinogen intramolecular activation by globin is "competitive," which implies that the pepsin active site is bound to globin and therefore cannot catalyze intramolecular pepsinogen activation. We would expect globin binding to pepsinogen to be similar to but no tighter than globin binding to pepsin. Consequently, K,, would not be smaller than K, for globin as a pepsin substrate. Our measured value of this K, is 1.41 x 10m5 M (5). This implies that K, is less than 0.1; i.e. the ratio of the concentrations of intermediates d and 0 is 1O:l or greater. The maiimum possible value of K,, is 1.7 x 10m5 M.
An alternative reaction scheme (Scheme 3, Fig. 1) has been considered and has been rejected because its predicted kinetics are not in agreement with our observations. This scheme where   the l/A uersus G plot. If this were interpreted to be due to production of pepsin from the 6. G complex, the rate constant for the process (k',,,) would be less than %O of k,,,. Consequently the contribution of the alternate pathway would be insignificant. Another reasonable explanation for the curvature is slight insolubility of globin at the higher concentrations.
The rate of conversion of pepsinogen to pepsin in the presence of globin was also determined in a pH-stat assay. Fig.  6 shows the semilogarithmic plot derived from the pH-stat experiments for 1 x 1O-6 M pepsinogen and 16.14 x 10m5 M globin. The pepsinogen decay constant A was 0.23 min-'. This val,:e is in very good agreement with the proteolytic activity assay results shown in Table II for 16.16 x 10m5 M globin, which has a value of 0.25 min-' for A. This agreement not only substantiates the validity of the experiments described above but also indicates that the pepsinogen does not possess significant proteolytic capability toward globin substrate. (In Fig. 6, the zero time intercept of 90% rather than 100% pepsinogen may indicate some proteolytic capability for a pepsinogen species.) Total acid consumption in these assays corresponded to two to three cleavages per globin molecule. Therefore, the assumption in the kinetic treatment of constant globin inhibitor concentration during an experiment is justified. DISCUSSION Our current understanding of intramolecular pepsinogen activation is summarized in Fig. 7 and diagrammatically in Fig. 8. Native pepsinogen and active pepsin are referred to as n and cy, respectively. Intermediate d is the pepsinogen species which binds the inhibitors pepstatin and globin. Intermediate 0 has the activation peptide in the active site. The species @ is the pepsin .peptide complex after cleavage but before dissociation of the peptide. The inactive pepsinogen which is recovered from the pepstatin/Sepharose column is called B. Two lines of evidence for the existence of intermediate 6 have been obtained. First, pepstatin at pH 2, either in solution or attached to an affinity column, binds to pepsinogen and greatly retards pepsinogen activation (Table II and Fig. 2~). This interaction implies that at pH 2 pepsinogen assumes a conformation and a state of ionization capable of binding pepstatin and that this pepsinogen'pepstatin complex is Activation of pepsinogen. This scheme represents the minimum number of steps for the pepsinogen activation. The intermediates 0 and @ are known from the previous studies. Intermediate 6 is postulated based on the evidence presented in this paper; however, its formation may not be obligatory. G is globin; pst is pepstatin.

nonproductive,
i.e. cannot be converted to pepsin. The confirmation of the pepsinogen species which binds pepstatin, intermediate 6, must differ from that of the native zymogen. The affinities of the native pepsinogen and intermediate 6 toward the pepstatin/Sepharose column are strikingly different (Fig. 2). Second, the globin inhibition of pepsinogen activation indicates that a pepsinogen.globin complex forms. This pepsinogen/globin interaction and its inhibition of intramolecular pepsinogen activation is analogous to competitive inhibition of an enzyme/substrate reaction. These results are consistent with the formation of an intermediate b. Because our kinetic data fit the equations derived from Schemes 1 or 2 (Fig. l), we conclude that the pepsin substrate globin binds reversibly to a pepsinogen intermediate called 6. The kinetic data also indicate that the d.globin complex cannot be converted to pepsin. As shown diagrammatically in Fig. 8, the intermediate is thought to have a developed active site which can bind either a pepsin inhibitor, pepstatin; or a pepsin substrate, globin. The region near the NH,-terminal activation peptide does not cover the active site in the intermediate 6 but subsequently binds to this active site when