Human Skin Chymotryptic Proteinase ISOLATION AND RELATION TO CATHEPSIN G AND RAT MAST CELL PROTEINASE I*

A chymotrypsin-like proteinase was purified 2400- fold from human skin. The procedure involves extraction of the proteinase from skin in 2 M KCl, precipitation with protamine chloride, fractionation by gel filtration chromatography, and fractionation by chromatogra- phy using a CH-Sepharose-D-tryptophan methyl ester affinity column. The properties of this proteinase were compared to the rat mast cell proteinase I and human cathepsin G. Differences were observed in the rates at which the proteinases were inhibited by diisopropyl fluorophosphate, the sensitivity of the proteinases to protein proteolytic inhibitors, the relative hydrolytic rates of the proteinases for a series of substrates, and the kinetic constants of the proteinases for synthetic substrates. The human skin proteinase did not react with antiserum to the rat skin proteinase and did not elute in the same position as the rat skin proteinase on gel filtration columns. These data demonstrate that the human skin proteinase is distinct from the other pro- teinases. Extracts of involved skin from patients with cutaneous mastocytosis had 15-fold higher levels of chymotryptic activity than extracts of uninvolved skin or skin from normal controls. The enzymatic properties of the material extracted from the biopsied skin were similar to those of the proteinase from normal skin, suggesting that the human skin chymotrypsin-like proteinase is a mast cell constituent. of at mM Bz-Tyr-OEt), proteinase Cathepsin G was purified by the method and The purified product units/mg) was about 50% pure and elastolytic activity was not detected in cathepsin G and Kinetic Analysis-The hydrolysis of Bz-Tyr- OEt measured method of 0.3 M Tris-HC1, pH 8.0, 1.5 M KC1, 15% and Ac-Tyr-OEt hydrol- ysis measured method of Schwert and Takenaka the same solution the ethanol concentration 3%. Absor- bance was measured on a Gilford 250 spectrophotometer equipped with data lister changes intervals. skin proteinase dialysis-precipitated material. Studies of Patients with Cutaneous Mast Cell Diseuse-Four patients with cutaneous mastocytosis (one child, three adults) were studied after informed consent was obtained. All patients were diag- nosed clinically (by G. S. L.), and after Xylocaine anesthetic 6-mm punch biopsies from involved and uninvolved skin were obtained. Sections of each biopsy were processed for histological examination including staining with toluidine blue. Involved skin demonstrated classic mast cell infiltration, whereas uninvolved skin was histologically normal. Biopsies from 30 normal patients served as controls. For proteinase analysis biopsies were blotted, weighed, and then extracted by the method of Fraki and (3) as the

ase may be common to the mast cells of numerous animals. Enzyme histochemical studies (11,12) and recent studies with human lung mast cell preparations (13) indicate that human mast cells contain trypsin-like and chymotrypsin-like proteinases.  have shown that the high salt extract of human skin contains chymotryptic and tryptic activity. The activities were separable by gel fitration chromatography of the extracts under high ionic strength conditions and fractions containing chymotrypsin-like activity were further purified by ion exchange chromatography. Most of the recoverable activity eluted as a single peak, suggesting the presence of one major chymotrypsin-like proteinase in this tissue.
The purpose of this study was to purify the human skin chymotrypsin-like proteinase and to determine whether it is similar or identical to other cellular chymotrypsin-like proteinases. The proteinases studied include the rat skin proteinase, rat mast cell proteinase I (15), and human cathepsin G (16-19), the chymotrypsin-like proteinase of human polymorphonuclear leukocytes. Chymotryptic activity in patients with cutaneous mastocytosis, a condition characterized by increased numbers of mast cells within the skin, was also examined to determine whether the skin chymotrypsin-like proteinase is a mast cell constituent.
Purification of the Human Skin Chymotrypsin-like Proteinase-Chymotryptic and tryptic activities were quantitated by measuring the hydrolysis of 10 mM Ac-Tyr-OEt and 10 mM N-benzoyl-L-arginine methyl ester according to the method of Hestrin (20). Elastolytic activity was assayed by measuring the hydrolysis of 1.5 mM N-succinyl-L-alanyl-L-alanyl-L-analine-p-nitroanilide at 410 nm. Cathepsin D activity was measured by quantitating the hydrolysis of radioactive hemoglobin at pH 4.7, and neutral proteolytic activity was determined by measuring the hydrolysis of radioactive casein as described by Levine et al. (21). Protein concentration was measured by the method of McGuire et al. (22). Skin was obtained from breast reductions and leg amputations. Specimens were cleaned of subcutaneous tissue by scraping with a scalpel, washed with cold saline, blotted, weighed, and then stored at -20 "C. Preparations used about 700 g of skin which was minced with scissors, and then extracted by the two-step method of . Skin (1 g of tissue/lO ml of solution) was frozen and thawed five times in 0.01 M sodium phosphate, pH 7.0 (low salt buffer), collected by filtration through glass wool, and then extracted by stirring overnight in 2.0 M KCl, 0.01 M sodium phosphate, pH 7.0 (high salt buffer), at a concentration of 1 g of tissue/5 ml of solution. The 2 M KC1 extract was clarified by centrifugation at 10,000 rpm for 10 min in a Beckman JA-10 rotor. Further purification of the human skin chymotrypsin-like proteinase was achieved by the precipitation of the enzyme from the high salt extract. Dialysis of the extract against low salt buffer resulted in the precipitation of 30% of the activity and the addition of 0.025% protamine chloride resulted in the precipitation of the remainder of the activity. Precipitated material was collected by centrifugation (10,000 rpm for 10 min) and was solubilized in 0.01 of the original volume in high salt buffer. Solubilized proteinase from the dialysis and protamine-precipitated fractions was next chromatographed on a Sephacryl S-200 column (5 cm X 90 cm) equilibrated in high salt buffer. Both dialysis and protamine-precipitated material eluted as a single peak of activity at M , = 28,000 (the major trypsin-like proteinase of human skin (14) elutes at M, = 130,000). Fractions were concentrated to 10 ml in a pressure dialysis cell (Amicon, Y-5 filter) for use in affinity chromatography. The affinity column was prepared by linking D-tryptophan methyl ester (50 mM) to activated CH-Sepharose 4B as described by Pharmacia. The column (0.7 X 20 cm) was equilibrated with high salt buffer and the proteinase was eluted with 75 m~ D-tryptophan methyl ester.
Best results were obtained with freshly prepared resin.
Purification of the Rat Skin Proteinase and Human Polymorphonuclear Leukocyte Cathepsin G-The rat skin proteinase was purified as described by Seppa and Jarvinin (2) except that the hydroxylapatite step was omitted. Based on specific activity (6.2 units/mg measured at 0.5 mM Bz-Tyr-OEt), the proteinase was 8% pure. Cathepsin G was purified by the method of Schmidt and Havemann (16). The purified product (28 units/mg) was about 50% pure and elastolytic activity was not detected in the cathepsin G preparation.
Enzyme Assays and Kinetic Analysis-The hydrolysis of Bz-Tyr-OEt was measured by the method of Hummel (23) in a solution of 0.3 M Tris-HC1, pH 8.0, 1.5 M KC1, 15% ethanol, and Ac-Tyr-OEt hydrolysis was measured by the method of Schwert and Takenaka (24) in the same solution except the ethanol concentration was 3%. Absorbance was measured on a Gilford 250 spectrophotometer equipped with data lister that records absorbance changes at 10-s intervals. Velocities measured as absorbance change/min (1-cm path length) were converted to micromoles of substrate hydrolyzed/min using the following coefficients: the hydrolysis of 1 mM Bz-Tyr-OEt gives an absorbance change of 0.964 at 256 nm (23) and the hydrolysis of 1 mM Ac-Tyr-OEt an absorbance change of 0.25 at 237 nm. The latter value was determined experimentally.
Proteolytic activity was determined by monitoring the degradation of radioactive protein to trichloroacetic acid-soluble peptides. Casein was radioactively labeled using r3H]acetic anhydride as described by Levine et al. (21). Assays were performed in a total volume of 60 p1 of solution containing 1.5 M KCI, 0.10 M Tris-HCI, pH 8.0, and 0.3 mg of casein (900,000 cpm) at 37 "C the reactions were stopped after 30 min by the addition of 40 pl of unlabeled bovine serum albumin (2.5 mg/ml) and 100 p1 of 6% trichloroacetic acid. Samples were then centrifuged and the amount of radioactivity in the supernatant was quantitated by scintillation counting. The percentage of the total radioactivity solubilized was assumed to be proportional to the amount of protein hydrolyzed and plots of radioactivity solubilized as a function of proteinase concentration were linear up to the hydrolysis of 20% of the substrate. The average increase in protein concentration due to the introduction of proteinase samples was less than 5% of the casein concentration.
K , and V,, were ascertained by Lineweaver-Burk graphic analysis. In nine experiments, the lines formed by the data had correlation coefficients of 0.99, and in three experiments the correlation coefficients were 0.98. The hydrolysis of Bz-Tyr-OEt by the rat proteinase was measured at proteinase concentrations of 17 and 27 nM and the hydrolysis of Ac-Tyr-OEt was measured at concentrations of 19, 37, and 46 nM. The hydrolysis of Bz-Tyr-OEt by the human skin proteinase was measured at 7, 5, and 3 n~ proteinase and the hydrolysis of Ac-Tyr-OEt was measured at 6 and 4 rn proteinase. The hydrolysis of Bz-Tyr-OEt by cathepsin G was measured at 18 and 22 nM proteinase. Initial velocities were measured at four to eight different substrate concentrations for each K, determination. The substrate concentration range was 0.25-0.50 m~ for Bz-Tyr-OEt and 0.65-2.2 m~ for Ac-Tyr-OEt.
Effect of Protein Proteolytic Inhibitors-Inhibition studies with protein inhibitors were performed in the high ionic strength conditions described for enzymatic assays and residual activity was measured after a 10-min incubation period by following the hydrolysis of Bz-Tyr-OEt. Inhibitor concentrations reported are those in the incubations prior to measurement of residual activity, and they represent the total of bound and unbound inhibitor. Dilution due to the assay procedure was 1.4-fold. Molarity of inhibitors were calculated from their molecular weights (25-28): soybean trypsin inhibitor (21,500), bovine pancreatic tyrpsin inhibitor (6,500), lima bean trypsin inhibitor (lO,OOO), and ovoinhibitor (46,500). Analytical Gel Filtration Chromatography-Apparent molecular weights of the proteinases were determined on Sephadex G-150 (1.5 cm X 90 cm) equilibrated in 1.2 M KCI, 0.01 M sodium phosphate, pH 7.0, at 4 "C. Elution of proteinases was monitored by following the hydrolysis of Ac-Tyr-OEt. Fractions containing 1 to 1.3 g of solution were collected and the flow rate was about 10 g of solution/h. Prior to the analyses the column was calibrated with the following proteins: bovine serum albumin (69,000), hemoglobin (64,OOO), ovalbumin (43,000), and cytochrome c (14,000). Calf thymus DNA was used to measure the void volume (VO) of the column and ,8-mercaptoethanol was used to measure the total volume (VJ. The distribution coefficient (Kav) of each protein and proteinase was determined from the relationship: where V, is the elution volume of each protein.
Inhibition of Proteinases by DFP-Inhibition rates were measured at 37 "C, in a solution containing DFP, 10% propylene glycol, 1.5 M KC1, and 0.01 m~ Tris-HCI, pH 8.0. The proteinase concentration in each inhibition study was less than 0.1 p~. After the addition of DFP, 0.1-ml aliquots were removed from incubations at the indicated times, diluted with 0.9 ml of substrate solution and assayed spectrophotometrically as described previously with Bz-Tyr-OEt (Figs. 3 and 4) or Ac-Tyr-OEt (Table IV) as substrate. Plots of the log fraction activity remaining as a function of time were linear as expected for a pseudof i s t order reaction. The slope of each line, which under the above conditions is equivalent to the pseudo-first order rate constant divided by 2.3, was used to calculate the length of time required to inactivate 50% (tlrz) of the proteinase. DFP concentrations were standardized by measuring the rate of chymotrypsin inactivation. A 0.02 mM DFP solution inhibited chymotrypsin 50% in 5 min at 37 "C.
Radioactiue Labeling of Proteinases-Samples from gel filtration columns were concentrated using Aquacide I11 to obtain solutions of high activity. Either [3H]DFP (6.5 Ci/mmol) or [14C]DFP (100 mCi/ mmol) was incubated with proteinase samples at 37 "C in 0.6 ml of solution containing 2 M KC1, 0.01 M Tris-HC1, pH 8.0, and 30-50% propylene glycol. The concentration of DFP was between 0.6 and 0.3 mM, and incubations were between 1.5-2.5 h. Reactions were stopped by the addition of 0.05 ml of 0.2 M unlabeled DFP. Unbound DFP was removed by exhaustive dialysis against 2.0 M NaCI, 0.01 M sodium phosphate, pH 7.0, followed by precipitation of protein in 6% perchloric acid. To ensure quantitative precipitation, 0.01 ml of carrier proteins (0.25% bovine serum albumin, ovalbumin, and lysozyme) was added to the dialyzed proteinase solution prior to the addition of perchloric acid. Precipitated material was collected by low speed centrifugation and solubilized in 100-250 pl of a solution of 2.5% SDS, 5% dithiothreitol. The amount of bound DFP was then quantitated by scintillation counting. The precipitation step was also required to remove an apparent contaminant which affected the resolution of protein bands on SDS gels. We suspect contamination occurred during the concentration of samples in Aquacide 111.

Chymotrypsin-like
proteinase were performed in tubes by the method of Weber and Osborn (30). Molar Specific Actiuities-Specific activities based on the molarity of each proteinase were determined by correlating hydrolytic activity of a sample with the molarity of the proteinase. Standardizations for skin proteinases were performed on enzymes purified through gel filtration chromatography, and those for cathepsin G were performed on material purified by the method of Schmidt and Havemann (16). Hydrolytic activity was measured using 0.5 mM Bz-Tyr-OEt as substrate, and the concentration of proteinase was determined by radioactive DFP labeling as described above. The length of incubation time was such that 795% of the proteinase was inhibited. Moles of proteinase were calculated from bound radioactivity using the specific activities for each isotope. The efficiency for quantitating 3H by scintillation counting was 38%, and the efficiency for quantitating I4C was 90%. In a 1-ml assay performed under the conditions described in the previous section, 52 pmol of the rat skin proteinase, 13.4 pmol of the human skin proteinase, and 80 pmol of cathepsin G each produced a hydrolysis rate of 0.1 absorbance unit (256 nm)/min. Double Immunodiffusion Studies-Immunodiffusion plates were equilibrated with 2 M KCl, 0.01 M sodium phosphate, pH 7.0, prior to use. Phenylmethylsulfonyl fluoride-inhibited proteinase samples (6 4) in high salt buffer were placed in each well and allowed to diffuse against antiserum for 36-48 h at 25 "C. The gel was then washed with saline, washed with distilled water, dried onto a glass slide, and stained for 10 min with 0.1% Amido black dissolved in a solution of 5% methanol, 10% acetic acid. Unbound dye was removed by washing for 15 min in the above methanol/acetic acid solution. Antiserum to rat skin chymotrypsin-like proteinase was a generous gift from Dr. He& Seppa, University of Oulu, Finland. Lyophilized antiserum was dissolved to its original volume in 2 M KC1,O.Ol M sodium phosphate buffer, pH 7.0. Antiserum only reacted with the rat skin proteinase over the concentration range of 0.1-0.9 (for these studies rat skin proteinase was purified to homogeneity) and not with the human skin proteinase (0.6 PM) obtained from protamine-precipitated material or that obtained (0.4 PM) from dialysis-precipitated material.
Studies of Patients with Cutaneous Mast Cell Diseuse-Four patients with cutaneous mastocytosis (one child, three adults) were studied after informed consent was obtained. All patients were diagnosed clinically (by G . S. L.), and after Xylocaine anesthetic 6-mm punch biopsies from involved and uninvolved skin were obtained. Sections of each biopsy were processed for histological examination including staining with toluidine blue. Involved skin demonstrated classic mast cell infiltration, whereas uninvolved skin was histologically normal. Biopsies from 30 normal patients served as controls. For proteinase analysis biopsies were blotted, weighed, and then extracted by the method of Fraki and Hopsu-Havu (3) as described previously. Chymotryptic activity and protein measurements in the extracts were performed by previously described methods.

Purification of the Human Skin Chymotrypsin-like Proteinase
The purification method is described in Table I. Dialysis of the high salt skin extract against low ionic strength buffer results in the separation of chymotryptic activity into two fractions: a low salt insoluble fraction and low salt soluble   Chymotryptic activity (0) was measured by following the hydrolysis of 10 nm Ac-Tyr-OEt for 30 min. Protein (0) was measured at 280 nm. Arrow marks the elution with 75 m~ D-tryptophan methyl ester.
The volume of each fraction was 10 rnl and the column was eluted at 25 ml/h. fraction that precipitates in the presence of protamine. The data to be presented subsequently indicate that the proteinases in both fractions are the same catalytic entity. Gel fitration of the precipitated material in both fractions results in the elution of a single peak of hydrolytic activity at a position corresponding to M, = 28,000. Skin chymotrypsin-like proteinases bind strongly to a CH-Sepharose-D-tryptophan methyl ester affinity column (Fig. 1) and can be eluted with free D-tryptophan methyl ester. The proteinase at this stage of the purification is 25-60% pure based on the comparison of the specific activity of preparations with that determined by labeling with radioactive DFP (see Table 111). Removal of minor high and low molecular weight contaminants leading to a preparations routinely 6040% pure is accomplished by further fractionation on a Sephadex G-100 column. Human skin chymotrypsin-like proteinase preparations had no detectable tryptic, elastolytic, or cathepsin D-like activity. The hydrolysis of casein, a nonspecific proteinase substrate, was inhibited over 95% by DFP, indicating that the preparations were not contaminated with nonserine class proteinases capable of hydrolyzing casein.
Chymotrypsin-like proteinases obtained from Sephacryl S-200 chromatography were labeled with radioactive DFP and analyzed by SDS-slab gel electrophoresis as shown in Fig. 2a. One broad band with M , = 30,000 is the major radioactive component observed in proteinase preparations obtained from both the dialysis (DE')and protamine (PI')-precipitated fractions. Human polymorphonuclear leukocyte cathepsin G and the rat skin chymotrypsin-like proteinase purified as described under "Experimental Procedures" are shown in the tracks G and R, respectively. SDS gel analyses of the proteinases in the most highly purified preparations were performed by the method of Weber and Osborn (30) as shown in Fig. 2b. The Weber and Osborn method was employed as an alternative method of analysis to determine whether the diffuse banding pattern was the result of the discontinuous buffer system used in slab gel electrophoresis. Gels stained with Coomassie brilliant blue show broad bands at M, = 30,000, consistant with that observed for radioactively labeled proteinases. The high molecular weight bands are contaminants, since they were not labeled with DFP.
The diffuse banding pattern is a quality that has been observed in most SDS-polyacrylamide gel analyses of the human skin proteinases; the reason for this behavior has not been established. Kinetic and inhibition studies indicate there is only one catalytic entity in the preparations, however. Lineweaver-Burk plots to determine kinetic constants were a. DP P P G R b. SDS-polyacrylamide slab gel electrophoresis of proteinases labeled with radioactive DFP (a) and SDS-polyacrylamide gels of the most highly purified preparations of the human skin proteinases ( b ) . a, composite of several autoradiographs in which bands were aligned at their approximate positions relative to standards. Tracks DP and PP are of human skin chymotrypsin-like proteinases purified from dialysis-and protamine-precipitated material, respectively. Track G represents cathepsin G and track R represents rat skin proteinase. Carbonic anhydrase (M, = 30,000), and a-chymotrypsinogen (M, = 25,700) had the same migration rate, b, about 7 pg of protein were analyzed on each gel and the bands were stained with Coomassie brilliant blue. Gel S represents the migration of standard proteins. Gels DP and PP were from different analyses, but in both cases the major band migrated as a polypeptide chain with M , = 30,000.
linear, having correlation coefficients greater than 0.99 in most studies and the inhibition rate of the proteinase by DFP followed pseudo-first order kinetics (Fig. 3). Logarithmic plots of the inhibition were linear until greater than 75% of the activity was inhibited, and the inhibition rates ( t 1 / 2 of 18 and 33 min at 0.8 and 0.4 m~ DFP, respectively) determined at two concentrations of DFP and two concentrations of proteinase appeared to be only dependent upon the DFP concentration. Furthermore, similar inhibition rates were obtained with two substrates, Bz-Tyr-OEt (Fig. 3) and Ac-Tyr-OEt (Table  IV).

Comparison of Human Skin, Rat Skin Chymotrypsin-like
Proteinases and Cathepsin G Enzymatic Properties-The relative hydrolytic rates of the human skin proteinases for three substrates are compared to those found for cathepsin G and the rat skin proteinase in Table 11. Human skin proteinases purified from the dialysis (DP)-and protamine (PP)-precipitated fractions hydrolyzed these substrates at identical rates, providing evidence that they are similar catalytic entities. These ratios differed significantly from those obtained for the rat skin proteinase and cathepsin G. Caseinolytic activity as well as the esterolytic activity of all proteinases was over 95% inhibited by DFP, indicating that the above results are not influenced by contamination with nonserine class proteinases.
Further differences between the three proteinases are shown in Table 111, where kinetic constants and inhibition properties are compared. The human skin proteinase hydrolyzed the synthetic substrates Bz-Tyr-OEt and Ac-Tyr-OEt more efficiently than either the rat skin proteinase or cathep-sin G as evidenced by the higher k,,,/K, ratios and the higher specific activity for Bz-Tyr-OEt. The K,,, values obtained for the human, rat, and cathepsin G proteinases were 1.1, 1.0, and 3.4 mM, respectively, when Bz-Tyr-OEt was the substrate, and 1.8, 2.0, and not measured when Ac-Tyr-OEt was the substrate. The human skin proteinase is the least efficient in hydrolyzing casein.
Inhibition Properties-In Fig. 4, the inhibition rates of the three chymotrypsin-like proteinases by DFP are compared and the tIl2 values are reported in Table 111. The human skin proteinase is inhibited %fold more slowly by DFP than the rat skin proteinase and cathepsin G. Also, the human skin proteinases purified from the dialysis-and protamine-precipitated material have identical inhibition rates (Fig. 4). The inhibition of the proteinases by a series of protein proteolytic inhibitors is compared in Table 111. The patterns of inhibition obtained for the rat skin proteinase and the human skin proteinase were similar. Both differed markedly from cathepsin G, which was sensitive to all the inhibitors. Bovine pancreatic chymotrypsin was inhibited greater than 90% by all the protein proteolytic inhibitors.
Comparison of Physical Properties-The differences in the banding patterns of the human skin proteinase with the rat skin proteinase and cathepsin G on SDS gels suggests struc- and 0.8 mM DFP (0) were performed at 37 "C. Further conditions and the method for measuring residual enzymatic activity are described under "Experimental Procedures." Proteinase concentration was 35 nM in the former study and 50 nM in the latter study.

Relative hydrolytic rates
Esterase activity was determined as micromoles of substrate hydrolyzed/min-ml of sample, and caseinolytic activity was determined as microerams of orotein hvdrolvzed/min-ml. " PP, protamine-precipitated fraction; DP, dialysis-precipitated fraction.
'Parentheses represent the range of two determinations which were the average of measurements at two different substrate concentrations.
human and rat skin proteinases by gel fitration chromatography also suggests a marked difference in molecular weight and/or shape (Fig. 5). Schmidt and Havemann,(16) have shown that cathepsin G migrates as a broad band under similar chromatographic conditions indicating it may be undergoing a self-aggregation process. Cross-reaction was not observed between the rat and human skin proteinases by double immunodiffusion analysis using antiserum to the rat skin proteinase (data not presented). Cross-reaction was also    '' Values in parentheses are the S.D. of the ratios obtained for three patients and for three normal skin samples. Casein hydrolysis was measured at a substrate concentration of 2.2 ,ug/ml and Ac-Tyr-OEt hydrolysis was measured at a substrate concentration of 1.0 rnM. 'Data obtained from one mastocytoma patient and values in parentheses are the range of duplicates. Hydrolytic activity was measured using the substrate Ac-Tyr-OEt. not observed between the human skin proteinase and cathepsin G using antiserum to cathepsin G.'

Localization of the Chymotrypsin-like Proteinase in Skin
Cutaneous mastocytosis is an uncommon condition manifested by the presence of increased numbers of mast cells within the skin. Biopsies from four patients having this condition were examined. The hydrolysis rate of Ac-Tyr-OEt per gram of tissue in the high salt extracts from this skin were 15fold higher (patients = 4, mean = 480 activity units, range = 190-800) than that found in extracts of uninvolved skin or skin from normal patients (mean = 32, range 0-42). The kinetic properties of the "proteinase" solubilized from three patients having the highest hydrolytic activity were nearly identical to the chymotrypsin-like proteinase from normal human skin and different from those of the rat skin proteinase (Table IV). Chymotryptic activity in the mastocytoma extracts was not inhibitable by bovine pancreatic trypsin inhibitor. This observation coupled with the slow inhibition rate of C. Reilly and J. Travis, personal communication.
by guest on March 24, 2020 http://www.jbc.org/ Downloaded from from Human Skin proteinase by DFP demonstrates that the proteinase in the mastocytoma extracts is not polymorphonuclear leukocyte cathepsin G. These data suggest that the normal human skin chymotrypsin-like proteinase is a mast cell constituent. There appeared to be some correlation between the degree of mast cell infiltration established histologically and proteinase concentration in our specimens; however, such correlations are difficult to establish in a small number of biopsies.
Consistent with the above observation, most of the chymotryptic activity of skin was localized to the dermal layer of skin where mast cells are located. This was accomplished by separation of freshly obtained skin into dermal and epidermal components using 2 M KBr as described by Levine et al. (21), followed by the extraction of each layer as described under ''Experimental Procedures." About 90% of the hydrolytic activity for Ac-Tyr-OEt found in the whole skin control was present in the 2 M KC1 extract of the dermal component.

DISCUSSION
The 2 M KC1 extract of human skin contains one major chymotrypsin-like proteinase which was purified 2400-fold. The proteinase appears to be a constituent of dermal mast cells as evidenced by the elevated level of the proteinase in extracts of skin obtained from patients with cutaneous mastocytosis. The enzymatic and physical properties of the human skin proteinase were compared to the major chymotrypsinlike proteinase of rat skin (rat mast cell proteinase I) which has been shown to be a constituent of mast cell granules (4, 31-34). There was no detectable cross-reaction between the proteinases and the differences in elution of the proteinases from gel filtration columns suggest a difference in molecular weight and/or shape. Except for their similarity in interaction with a series of protein proteolytic inhibitors, both proteinases differed markedly in kinetic and inhibition properties, indicating important differences in their active site structures. The human skin proteinase was only 25% as active as the rat skin proteinase in hydrolyzing denatured casein, a nonspecific protein substrate. Efforts to map the extended substrate binding site of the human skin proteinase by the peptide analogues used to map rat mast cell proteinase I and cathepsin G (35) are in progress.
The concentration of the human skin chymotrypsin-like proteinase in skin is about 4 nmo1/100 g of tissue. There are 1-10 X lo6 mast cells/g of tissue (36). Using this value, there are 0.2-2 X IO7 molecules of the human skin proteinase/mast cell. This value is 5-50-fold less than reported for the rat mast cell proteinase I (37). Schwartz et al. (13,38) have found that isolated human lung mast cells contain a trypsin-like proteinase which is present at a concentration of approximately lo8 molecules (Mr = 35,00O)/mast cell. Chymotrypsin-like proteolytic activity was reported in their mast cell preparations, but the catalytic properties of this proteinase were not examined. Degranulation of lung mast cells by treatment with rabbit IgG anti-human IgE led to the solubilization of trypsinlike activity but not chymotrypsin-like activity, suggesting that the chymotrypsin-like proteinase of human mast cells may not be analogous to rat mast cell proteinase I.
The human skin chymotrypsin-like proteinase and cathepsin G are serine proteinases produced in different tissues of the same species. Cathepsin G is a constituent of human polymorphonuclear leukocytes and the human skin chymotrypsin-like proteinase appears to be a constituent of mast cells. The differences in catalytic and inhibition properties presented in this study demonstrate that both proteinases are different genetic products. The relationship between chymotrypsin-like proteinases of skin and polymorphonuclear leu-kocytes has not been reported for any other species.
The physical and kinetic dissimilarities between the rat and human skin proteinases may reflect the evolutionary divergence of these skin proteinases from a common ancestral gene product and may suggest that these enzymes have evolved different biological functions within the mast cell. Ultimately, the evolutionary relationship between these chymotrypsinlike proteinases and cathepsin G w i