The Proteolytic Enzymes of the K-1 Strain of Streptomyces griseus Obtained from a Commercial Preparation ( Pronase )

The commercial protease mixture, Pronase, which is obtained from the K-l strain of Streptomyces griseus, contains many exoand endopeptidases. Among the latter enzymes are several which are inhibited by diisopropyl fluorophosphate (DFP). Three of these DFP-reacting components, being of low molecular weight, were partially separated by filtration through Sephadex G-75. Chromatography through CM-cellulose of the filtration fractions with the smaller proteins yielded two well separated peaks (A and B) after the start of a sodium chloride gradient. Peak B consisted of a single enzyme which hydrolyzed N-cr-acetyl-L-tyrosine ethyl ester (Ac-Tyr-OEt). It was homogeneous by gel electrophoresis and its approximate molecular weight was 17,500 as determined by gel filtration. The material in Peak A was heterogeneous with activities against both AcTyr-OEt and N-a-benzoyl-L-arginine ethyl ester (Bz-ArgOEt). The component which hydrolyzes Bz-Arg-OEt was previously shown to be homologous with mammalian trypsin. Because of reports of the binding of alkylamines by trypsin, rechromatography of components in the earlier peak was carried out in the presence of n-butylamine. Slight retardation of the trypsin-like component was achieved. When a diamine, cadaverine, was substituted for the monoamine, further selective retardation was noted with almost complete resolution of the components in Peak A into two peaks (A1 and Az). These resolutions were thought to be due to the formation of an amine-enzyme complex at the active site; for, if the protein corresponding to Peak A was first reacted with DFP, no retardation was noted. Material from Peak


The Proteolytic
Enzymes of the K-

SUMMARY
A study was carried out on some of the properties of the two smallest serine endopeptidases from Pronase which had previously been purified to homogeneity. Each enzyme is homologous with bovine chymotrypsin and has isoleucine as the NHZ-terminal residue. The smaller enzyme is free of lysine, whereas the larger enzyme contains only 1 lysine residue. Reaction of the larger enzyme with acetic anhydride yielded a homogeneous, active, and stable derivative as indicated by ion exchange chromatography and acrylamide gel electrophoresis. Reaction of the smaller enzyme with acetic anhydride yielded two chromatographic components, of which only the larger demonstrated activity against N"acetyl-L-tyrosine ethyl ester. This active component autolyzes during acrylamide gel electrophoresis but appears as a single band by cellulose acetate electrophoresis. The excellent yields of these protein derivatives were only achieved by modifying the past standard techniques of acetylation. In each reaction mixture glycerol was included at a concentration of 20% by volume. As a result only small amounts of native proteins were required to prepare the derivatives. Analysis of each enzyme revealed complete acetylation of the NH&erminal isoleucine residue. Acetylation resulted in only modest changes in the Michaelis constant and maximal velocity of each enzyme with N"-acetyl-L-tyrosine ethyl ester as substrate. Despite the earlier observation that only the larger enzyme demonstrated marked stability in 6 M guanidinium chloride there was no difference in the heat stabilities of the two enzymes. Neither showed an effect of ethylenediaminetetraacetate (EDTA) on activity at 37" even after several hours; however, at temperatures above 4.5" each enzyme underwent marked loss of activity in the presence of EDTA, whereas activity was conserved up to 60" in the absence of chelating agent. Metal-free enzyme was prepared in each case by gel filtration at a low pH in the absence of metals. These metal-free proteins demonstrated the same temperature stabilities as the EDTA-treated native enzymes. Of the many cations tested, Cazf was specific in each case in restoring the stability at higher temperatures to the metal-free enzymes. 3233 * This investigation was supported by United States Public Health Service Grants NIH-AM-09001, NIH-AM-05472, and NIH-GM-02011, National Defense Educational Act Grant HEW OE 40-001536, and American Cancer Society Grant IN-51-l.
A preliminary report of a portion of this work has been presented (1).
1 Student in the combined M.D., Ph.D. Program of the Universitv of Miami School of Medicine.
This work was carried out in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biochemistry, University of Miami.
$ To whom all inquiries should be directed.
In an earlier report the purification of four serine endopeptidases in Pronase was described (2). These enzymes have also been isolated to varying degrees of purity in other laboratories (3)(4)(5)(6).
The three smaller enzymes are homologous with mammalian chymotrypsin (2,7,8), whereas, in contrast, the largest enzyme demonstrated at least partial homology with the subtilisins (2). The two smallest enzymes have activity against N"-acetyl-L-tyrosine ethyl ester. The third enzyme of the chymotrypsin family of proteins hydrolyzes N"-benzoyl-L-arginine ethyl ester and, therefore, not unexpectedly demonstrated extensive homology with bovine trypsin (8). To date, all of the enzymes of the chymotrypsin family of serine proteases have either an isoleucine or valine residue which is implicated in a functional role at the active site (9). With one exception (8) these residues are situated at the NHz-terminal position.
Because of the findings of an NHz-terminal isoleucine in each of the two smallest serine enzymes in Pronase, the effect of estensire acetylation on the function of each protein was examined. A modification of the usual technique of the reaction of proteins with acetic anhydride was developed in order to obtain high yields of homogeneous derivatives from modest amounts of native precursors.
The larger of the two enzymes was demorlstrated earlier to conserve'activity in the presence of either 6 M guanidinium chloride or 8 M urea (9). This enzyme c'emonstrated significant autolysis in 6 M guanidinium chloride if EDTA 3234 was present.
In the absence of denaturant no effect of the chelating agent was found.
Therefbre, an investigation was carried out to discover the specific cation responsible for stabilizing the euzyme.
As will be demonstrated, the smaller of the two enzymes contains no lysine.
Since both enzymes demonstrate elastase activity (10) and also, like chymotrypsin, hydrolyze Ac-Tyr-OEtl we propose the trivial names lysine-free chymoelastase and guanidine-stable chymoelastase for the smaller and larger proteins, respectively.

MATERIALS AND METHODS
Pronase (grade B) was obtained from Calbiochem; several lots (numbers 900053, 000130, 000133, and 001828) were used for these studies. The serine proteases were purified as previously described (1). Only the two smallest enzymes with activity against Ac-Tyr-OEt were utilized in these studies. They were determined to be homogeneous by chromatography through CM-cellulose, by gel filtration, and by acrylamide gel electrophoresis.
Ac-Tyr-OEt and Na-acetyl-L-phenylalanine ethyl ester were obtained from Mann Research Laboratories.
Casein was obtained from Difco and purified (1). Urea, reagent grade, was obtained from Mallinckrodt, and guanidinium HCl of highest purity was obtained from Eastman Kodak.
Sephadex G-25 was obtained from Pharmacia Fine Chemicals.
Polyamide layers, a product of the Cheng Chin Trading Co., Ltd., were purchased from Gallard-Schlesinger. The sheets (15 x 15 cm) were quartered before utilization for analysis of dansyl amino acid derivatives.
All other chemicals were of reagent grade.
Amino Acid Analyses-Analyses were performed according to standard techniques (11) on a Beckman amino acid analyzer. Scid hydrolyses (24-, 4%, 72-, and 96.hour) with about 2 mg of protein per tube were carried out in vucuo with 6 N HCl at 108". Cysteine and methionine were analyzed as their oxidized products following reaction with performic acid (12). Reported serine and threonine values are those obtained after extrapolation to zero time of hydrolysis (13). Tryptophan content was determined by a spectrophotometric assay in base (14). The quantitation of NHz-terminal residues was by the cyanate procedure (15) after the proteins had been oxidized by performic acid (12). Chromatography through a Dowex l-X8 column was carried out in order to remove peptides containing cysteic acid from the hydantoins (15). The recoveries for the NHz-terminal residues were calculated using the data of Stark (15) without accounting for losses after passage through the Dowex 1 column. Therefore, the values reported here are probably low.
Measurement of Catalytic Activity-The extent of casein digestion was determined as described earlier (9). The activities against Ac-Tyr-OEt and Ac-Phe-OEt were measured by previously described techniques using the pH-stat (16). The kinetic constants toward Ac-Tyr-OEt and Ac-Phe-OEt were determined at substrate concentrations from 5 to 30 mM. The enzyme concentrations for each assay are listed in Table III. The assay medium in the kinetic studies contained 3% dioxan to permit the complete solubility of the ester substrates at the higher concentrations.
Titrations were performed at pH 8.0 at 25" with 0.049 M NaOH.
dcetylation of Proteins-The technique of Oppenheimer et al. was followed in reacting each enzyme with acetic anhydride (18). However, because of the paucity of material the procedure was scaled down considerably.
Five or 10 mg of enzyme were added to 3 ml of 5 mM sodium acetate-5 mM calcium acetate (pH 5.0). The solution was cooled to 2" in a water-jacketed vessel. The pH was raised to 6.7 and maintained at this level by the addition of 2 M NaOH with the use of a pa-stat (Radiometer TTTlc).
Constant stirring was maintained with a magnetic stirrer.
Acetic anhydride was added (5 ~1 every 7 min) for six or seven additions.
Following the final addition the reaction was allowed to proceed for a further 30 min. Despite the great care taken in the slow addition of the anhydride, substantial amounts of protein precipitated irreversibly from solution.
The remaining protein in solution was polydisperse when analyzed by CM-cellulose chromatography and, furthermore, was inactive when tested against Ac-Tyr-OEt. Attempts were made without success to circumvent this problem by running the reaction in half-saturated sodium acetate in order to reduce the number of acetylated tyrosine residues (19). Halting the reaction with anhydride at the incipience of precipitate formation or running the reaction in 4 M urea was unsatisfactory. These latter methods yielded heterogeneous products with incomplete reaction and little activity.
The problem of gross denaturation was eliminated when the standard reaction was carried out with the addition of glycerol (20% by volume). With this modification no precipitate was noted. An entirely satisfactory product was obtained in the case of the guanidinestable chymoelastase, but in the case of the lysine-free chymoelastase two chromatographic components were noted (see below).
Acetylation of the latter enzyme in the presence of 13 or 30% glycerol (by volume) yielded results less satisfactory than those achieved at 20% concentration.
The native proteins and acetylated derivatives were separately analyzed by chromatography through a CM-cellulose column (1.1 x 7.5 cm) in the sodium and calcium acetate buffer as previously described (2). Because of the small amount of protein applied, the effluent fractions were monitored for absorbance at 230 nm. The glycerol in the solutions containing the acetylated proteins was removed by passage through a Sephadex G-25 column equilibrated with the above buffer before the proteins were passed through the ion exchange column.
The homogeneity of native and acetylated guanidine-stable chymoelastase was analyzed by electrophoresis in polyacrylamide gel at pH 4.3 as described earlier (2). Twenty microliters of protein solution were added for each run. The amount of protein applied was determined according to the method of Lowry et al. (20) with bovine serum albumin used as the standard. The gels were stained with Amido Schwarz. Acetylated lysine-free chymoelastase was not stable in this system. Therefore, this derivative and its precursor were analyzed by electrophoresis on cellulose acetate membranes with a Beckman Microzone system. The buffer composition (sodium diethylbarbiturate, pH 8.6) and staining technique with Ponceau S were those described by the Beckman Jlethods Manual (RM-TJ3-010).
The NHQ-terminal residues of the native and acetylated proteins were examined qualitatively after performic acid oxidation by reaction with dansyi chloride (21). The results were analyzed by two-dimensional chromatography on quartered polyamide sheets (22) (14).
Spectra of native and acctylated proteins in 5 mM sodium acetate-5 mM calcium acetate (pH 5.0) were measured with a recording spectrophotometer (Zeiss DMR-21).

Heat Stability and Calcium
Dependence--The heat stability of each enzyme was studied in thr presence and absence of EDTA. One milligram of each enzyme was dissolved separately in 0.5 ml of 10 mM Tris (pH 8.0). Thereafter 25.~1 aliquots were added to 75 ~1 of 10 mM Tris (pH 8.0) containing either 10 mM EDTA or 10 mM CaC12. The solutions were iucubated at the desired temperature for 15 min and then plunged into ice. After 1 hour at O", a 50.~1 aliquot was used to assay for activity against Ac-Tyr-OEt.
All results have been espressed as percentages of the masimal activity noted.
The following procedure was carried out to determine the specific cation which stabilized the two enzymes against heat denaturation.
Five milligrams of each enzyme were dissolved separately in 1.0 ml of 0.1 M glycine (pH 3.4) containing 10 mM EDTA.
It was later appreciated that at this pH EDTA has a very low affinity for divalent cations and therefore probably contributed nothing to these studies (23). The solution was passed through a Sephades G-25 column equilibrated with the above glycine solution.
The fractions containing the eluted protein were combined (total volume was 6 ml). Fifty-microliter aliquots were added to separate tubes containing 50 ~1 of different chloride salts at 0.1 M concentration.
The aliquots were incubated for 10 min at the desired temperature. After cooling for 1 hour at O", 100.~1 aliquots were removed to assay for activity against Xc-Tyr-OEt.
These results were expressed as percentages of the activity noted after incubation in 12.5 mM CaClz and the above Tris buffer at 24". To l.O-ml aliquots of each of these solutions was added 0.5 mg of enzyme in 50 ~1 of distilled HzO. after incubation at room temperature for various time periods lOO+l aliquots were removed for assay against iic-Tyr-OEt (1,2,16). Table I gives the amino acid composition of the two proteins. Since preliminary csperiments have revealed no free sulfhydryl groups and since these enzymes are at least in part homologous with chymotrypsin, the half-cystine residues are probably fully incorporated into disulfide bonds. There are probably 4 halfcystine residues in each protein (see below); the present low recovery is not readily esplicable.

RESULTS
The smaller protein is free of lysine.
The larger enzyme has a relatively high arginine to lysine ratio.
The other feature of interest is the high content of glycine, alanine, threonine, and serine in each protein.
' Table  II gives the results of the NHz-terminal analyses.
Guanidinestable chymoelastasc has 011l y isoleucine as the NHL-terminal residue; lysine-free chymoelastase demonstrates primarily isoleucine as the NHz-terminal residue with also a significant amount of NHz-terminal glycine.
Though these results indicate the possibility that the smaller protein may contain two peptide chains, the reaction with dansyl chloride (see below) suggests that only isoleucine is the significant NHp-terminal residue. Fig. 1 shows the results of chromatography through CM-cellulose of native and acetylated guanidine-stable chymoelastase. The derivative demonstrates a single peak which is eluted at a lower NaCl concentration than the parent protein. This is expected if acetylation is successful in blocking the o(-and e-amino groups. In contrast to chymotrypsin (18), a protein with 14 lysine residues, the bacterial protein with only 1 lysine residue demonstrates only a modest change in the affinity for CM-cellulose after acetylation.
This esplains also the high solubility of the acetylated bacterial enzyme in an acidic medium.
Acetylated chymotrypsin would have only 4 arginine and 2 histidine residues as the charged groups at a low pH. This could be the reason for the low solubility below pH 5 of the acetylated bovine enzyme (18). Fig. 2 shows the results of chromatography through CM-cellulose of native and acetylated lysine-free chymoelastase. Consistently, two peaks of material absorbing at 230 nm were obtained after reaction with acetic anhydride.
Eluting without the application of a NaCl gradient, the derivatized proteins were evidently different from the native proteins.
Application of a NaCl gradient after elution of these two components did not yield further material.
The larger peak contained most of the activity against Ac-Tyr-OEt.
It is surprising that this derivative is eluted so much earlier than the parent protein.  The activity against Ac-Tyr-OEt is demonstrated; loo-p1 aliquots were used for the assays. See text for details.
protein, the NHz-terminal isoleucine. Acetylation of this amino group and of tyrosine residues may be hindering the interaction of arginine residues with the CM-cellulose (see below).  (2) acetylated lysine-free chymoelastase after reaction with diisopropylphosphorofluoridate.
About 40 pg were applied for each protein.
as demonstrated by acrylamide gel electrophoresis is depicted in Fig. 3. The derivative, as expected, migrates less rapidly toward the cathode than the parent protein.
The acetylated protein demonstrated the poor retention of Amido Schwarz as noted earlier with the native enzyme (2,9). The derivative, like the native enzyme, showed stability in 6 M guanidinium HCl.
Therefore, the loss of dye was attributed to diffusion of the stable enzyme-dye complex from the gel (9). Acetylated lysine-free chymoelastase was not stable in the conditions of gel electrophoresis; no stainable material was discerned after the run. This derivative was reacted with diisopropylphosphorofluoridate with apparent complete inhibition of activity. Electrophoresis of the double derivative through acrylamide gel revealed the presence of several bands.
The possibility was considered that the heterogeneity arose from denaturation during acrylamide gel electrophoresis; therefore, acetylated lysine-free chymoelastase was examined by cellulose acetate membrane electrophoresis. Fig. 4 gives the results of studies on native and acetylated enzyme after reaction with diisopropylphosphorofluoridate; it appears by this technique that the proteins are homogeneous.
Furthermore, as expected from the absence of lysine in this protein, there is little difference in the migration of acetylated and native proteins. Fig. 5 demonstrates the results of studies on the NHn-terminal residues of the native and derivative proteins.
Guanidine-stable chymoelastase has only isoleucine as the NH&terminal residue. Lysine-free chymoelastase has primarily isoleucine, as noted by the very large fluorescent spot, of dansyl-isoleucine; however, this enzyme also demonstrates several other dansyl residues in very low amounts.
These other KHz-terminal residues are probably generated during the reaction with dansyl chloride, since the parent native enzyme was shown to be homogeneous by acrylamide gel electrophoresis (2). The acetylated proteins demonstrate the absence of patterns of any dansyl amino groups, indicating that the amino groups have been ccmpletely acetylated. 6. Ultraviolet spectra of native (--) and acetylated (---) proteins.
A, lysine-free protease, 0.80 mg per ml of native or about 0.7 mg per ml of acetylated enzyme. R, guanidine-stable protease, 0.72 mg per ml of native or about 0.6 m.g per ml of acetylated enzyme. Fig. 6 demonstrates the spectra of native and acetylated proteins. The spectral shifts of each protein after acetylation may be due in part to 0-acetylation of tyrosine residues. In contrast to the studies with chymotrypsin, the spectra of the derivatives were not converted back toward those of the native proteins after reaction with acetate at pE-I 5.0 for several days at 4" or by reaction with 0.1 M hydrosylamine for several hours at room temperature.
If the spectral shifts are attributable to O-acetylation of tyrosine, these liganded groups appear to be unusually resistant to cleavage in these proteins. are demonstrated. Fig. 7 depicts the activities of guanidine-stable and lysine-free chymoelastases against Ac-Phe-OEt as a function of pH. The pH range of maximal activity lies between 7.5 and 10; no differences are seen between each native enzyme and its acetylated derivative.
Recently we described the unique stability in denaturant of guanidine-stable chymoelastase (9,24). A study was carried out to examine the stability of lysine-free chymoelastase in urea and guanidinium HCI. Fig. 8 depicts the results.
In the presence of 10 IIIM CaC12 this enzyme appears to be entirely stable in 7.6 M urea but loses activity in high concentrations of guanidinium HCl.
In contrast, in 1 mM sodium EDTA a moderately rapid loss of activity is noted in 7.6 M urea and the rate of inactivation in guanidinium HCl is accelerated. As previously noted with guanidine-stable chymoelastase (9), no effect of EDTA on the stability of lysine-free chymoelastase could be demonstrated in the absence of denaturant. Table III demonstrates the kinetic constants of each enzyme under different conditions. The V,,l,, of each enzyme toward' Ac-Tyr-OEt is much less than that of bovine chymotrypsin (16). The K, for Ac-Tyr-OEt of each enzyme increases greatly in 8 M urea. The acetylated enzymes show only slight differences in kinetic constants from those seen with the native proteins.
The studies with EDTA suggested that in each enzyme there were tightly bound metal ions which were only accessible to the   of the proteases were relaxed. This relaxation can only be inferred, since circular dichroism studies of guanidine-stable chymoelastase revealed no spectral differences in the presence or absence of denaturant. It was considered that an elevation in temperature would similarly demonstrate a loosening of conformation.
Therefore, the heat stability of each enzyme was studied in the presence and absence of EDTA. Fig. 9 depicts the results. As noted earlier, no effect of chelating agent is demonstrable below 45"; however, at higher temperatures the protective effect of a metal ion is inferred by the destabilizing effect of EDTA.
There is no significant difference in the heat stabilities of the two enzymes. As described above, apparent metal-free enzyme was prepared in each case and an analysis was carried out on the protective effect of various cations on the stabilities at 55". As is demonstrated in Table IV, calcium was specific as the required metal for the two proteases.

DISCUSSION
Glycerol has been used in many enzymological studies as an empirical measure to stabilize proteins.
For instance, placental 17P-hydroxysteroid dehydrogenase when stored in a buffer with 50% glycerol was completely protected against loss of activity (25), in contrast to the rapid denaturation noted in glycerol-free solutions.
At concentrations less than 50%, glycerol gave intermediate degrees of protection.
Furthermore, glycerol protected the enzyme against heat denaturation.
In another study, buffers with 50% glycerol completely protected aminoacyl ribonucleic acid synthetases against loss of activity when stored at -15" (26). How glycerol stabilizes proteins remains obscure; to our knowledge no definitive studies have been carried out to determine the mechanism.
The enzymes in the present study, in contrast to the above examples, are among the most stable soluble proteins known.
Therefore, the addition of glycerol during our experiments requires some explanation.
We considered that there might be substantial flexibility of the proteins despite their stability.
This flexibility could permit the fleeting exposure of buried groups susceptible to acetylation.
When acetylated at these sites the proteins could undergo further conformational transitions which would make them susceptible to autolysis. If glycerol could restrict the degree of initial relaxation of confor- Lysine-free chymoelastase Guanidine-stable chymoelastase L a The chloride salt of Na+, K", Mg2+, Sr2+, Ba2+, Mn2+, Fe3+, Co2* Ni2+, Cuz+, Zn2+, or Cd2+ at 12.5 mM concentration. mation, the susceptible buried groups would not be acetylated. It is clear, in the case of guanidine-stable chymoelastase, that the major stabilizing effect of glycerol cannot be the restriction of conformational changes after acetylation. Purification of the acetylated derivative free of glycerol resulted in a completely stable protease.
However, in the case of lysine-free chymoelastnse the stabilization of the acetylated derivative by glycerol may also be important in view of the moderate rate of loss of activity following the removal of glycerol.
Another explanation for these results may be that glycerol, as a competitive nucleophile, significantly improves the selectivity of acetylation.
The reaction of glutamate dehydrogenase with acetic anhydride demonstrated significant improvement in selectivity when in the presence of 0.1 M Tris acetate, a nucleophilic buffer (27). Our very preliminary studies suggest the possibility that glycerol may generally stabilize protein substrates during other methods of *chemical modification.
Recently the carbosymethylation of sulfhydryl groups on aldehyde dehydrogenase was shown to be affected by concentrations of glycerol which stabilize that enzyme (28). Our procedure represents a significant advance over the methods previously used for acetylation of trypsin and chymotrypsin.
By minimizing denaturation and heterogeneity of products we were able to effect a loo-fold reduction in scale of the acetylation procedure as compared with the studies of chymotrypsin and trypsin (18, 29). The acetylation studies were carried out because of the identity of residues in the microbial enzymes Asp-194 and Be-16 in cY-chymotrypsin (2,7,9,16). In the bovine enzyme these 2 residues form an ion pair via the /3-carboxyl group of the aspartyl residue and the a-amino group of the terminal isoleucine (30). It has been demonstrated that specific substrates will bind to the enzyme only when this ion pair is formed (31). In or-chymotrypsin, acetylation of this a-amino group results in complete loss of activity towards specific substrates (18) but not towards p-nitrophenyl acetate (32). On the other hand, acetylation of the a-amino group of a homologous residue in porcine elastase results in no loss of activity (33). Finally, reaction of acetic anhydride with trypsin fails to acetylate the NHa-terminal isoleucine (29). These results have been interpreted as demonstrating variance in accessibility of the NHz-terminal residue to acetic anhydride and also the presence or absence of steric factors which would prevent the acetylated a-amino group from assuming the required conformation of the active enzyme. The initial consideration that formation of the ion pair in the pancreatic serine enzymes generated the catalytic center (30) has required further modification.
The catalytic center appears largely intact in the zymogen before formation of the above ion pair (34). X-ray diffraction studies of both chymotrypsinogen and oc-chymotrypsin suggest instead that the formation of the specific ion pair on zymogen activation generates the substratebinding site in chymotrypsin (30,34). In our studies the slight changes that are demonstrated in kinetics following acetylation appear to reflect primarily changes in substrate binding (See Table II), confirming the above hypothesis.
We believe that acetylation of the a-amino group on the isoleucine residue in each enzyme need not substantially change the postulated ion pair bond.
Acetylation of this residue may merely convert an electrostatic bond into a hydrogen bond (Fig. 10). This could also explain the retention of activity in acetylated porcine elastase.
The present study confirms again the lack of effect of EDTA on these enzymes in the absence of denaturant (9,35). As noted earlier with the guanidine-stable enzyme, a substantial effect by EDTA on the lysine-free enzyme is noted in denaturant. Because of the difficu1t.y of establishing in denaturant the specific cation requirements of each enzyme, the above-described heat stability studies were carried out. The surprising finding in these investigations was that, despite the marked difference in stabilities of the two enzymes in concentrated guanidinium HCl solutions, there were no difference in the heat stabilities.
However, like thermolysin (36-38) the stability of each of the Pronase enzymes is dependent specifically on calcium ions. In thermolysin 4 calcium ions are bound per enzyme molecule; the number of calcium ions bound to these serine proteases remains to be demonstrated. That calcium ion stabilizes these enzymes is not unexpected in view of the similar stabilizing effect of calcium ion on bovine chymotrypsin and trypsin (3943). Sequence studies on the lysine-free enzyme carried out elsewhere have confirmed our finding of only 1 NH*-terminal residue, isoleucine (44)(45)(46).
These studies have also demonstrated the presence of two disulfide bonds (and therefore 4 half-cystine residues) in this protein.
We have commenced an investigation of the primary sequence of guanidine-stable chymoelastase in an attempt to define further the remarkable stability of this enzyme.