Proc. Natl. Acad. Sci. USA Vol. 84, pp. 7508-7512, November 1987 Biochemistry

Active site of tripeptidyl peptidase II from human erythrocytes is of the type (serine peptidase/cysteine peptidase/thennitase/thiol dependency) BIRGITTA ToMKINSON*, CHRISTER WERNSTEDTt, ULF HELLMANt, AND ORJAN ZETTERQVIST* *Department of Medical and Physiological Chemistry, University of Uppsala, Biomedical Center, Box 575, S-751 23 Uppsala, Sweden; and tLudwig Institute for Cancer Research, Uppsala Branch, Box 595, S-751 23 Uppsala, Sweden Communicated by Bruce Merrifield, July 20, 1987

ABSTRACT The present report presents evidence that the EXPERIMENTAL PROCEDURES site of amino acid sequence around the serine of the active Materials. [1,3-3H]iPr2P-F was purchased from Amersham human tripeptidyl peptidase II is of the subtilisin type. The International. Emulsifier scintillator 299 came from Packard. from human erythrocytes was covalently labeled at its Dithiothreitol and bovine were from Boehringer with [3H]diisopropyl fluorophosphate, and the Mannheim, and L-1-tosylamido-2-phenylethyl chloromethyl protein was subsequently reduced, alkylated, and digested with ketone-treated trypsin was from Worthington. Unlabeled trypsin. The labeled tryptic peptides were purified by gel iPr2P-F was obtained from Fluka, and Tween-20 was from filtration and repeated reversed-phase HPLC, and their amino- Atlas Chemie. Guanidinium chloride, acetonitrile (Lich- terminal sequences were determined. Residue 9 contained the rosolv), trifluoroacetic acid (Uvasol), and iodoacetic acid radioactive label and was, therefore, considered to be the active (recrystallized in petroleum ether before use) were bought serine residue. The primary structure of the part of the active from Merck. Sephadex G-50, fine, was purchased from site (residues 1-10) containing this residue was concluded to be Pharmacia. The Nucleosil C18 column (10 ,um; 4 x 250 mm) Xaa-Thr-Gln-Leu-Met-Asx-Gly-Thr-Ser-Met. This amino acid for HPLC was packed by Skandinaviska GeneTec AB sequence is homologous to the sequence surrounding the active (Kungsbacka, Sweden), and the Bondapak C18/Corasil pack- serine of the microbial peptidases subtilisin and thermitase. ing material was obtained from Millipore-Waters. These data demonstrate that human tripeptidyl peptidase II Preparation of Reduced and ALkylated, 3H-Labeled TPP-II. represents a potentially distinct class of human peptidases and TPP-II was prepared from human erythrocytes as described raise the question of an evolutionary relationship between the (2) with a few modifications. Outdated erythrocyte concen- active site of a mammalian peptidase and that of the subtilisin trate, obtained from the blood bank at the University Hos- family of serine peptidases. pital of Uppsala, was used. The erythrocytes were washed once with 4 vol of 0.9% (wt/vol) NaCl before lysis with 8 vol of 10 mM potassium phosphate, pH 7.5/2 mM EGTA/1 mM The discovery of an extralysosomal tripeptide-releasing ami- dithiothreitol. The centrifugation used for removing erythro- no-peptidase in rat liver has been reported (1). Its substrate cyte membranes was omitted, and the ion-exchanger was specificity, as studied by use ofa series ofpeptide substrates, added directly to the hemolysate after the addition ofglycerol is complex (1, 2). The peptidase releases tripeptides with little (2). The specific activity of the enzyme preparation used was apparent similarity, although the rate of cleavage varies 4000 units/mg, as measured by the standard assay (1). TPP-II considerably. The enzyme, currently named tripeptidyl pep- (1.0 mg, i.e., 7.7 nmol of subunit) was incubated with 0.2 mM tidase II (TPP-II) (2, 3), has a high native molecular weight [3H]iPr2P-F (1.0 Ci/mmol; 1 Ci = 37 GBq) in a total volume (>106) (1), a subunit molecular weight of 135,000, and a of 750 ,lI for 1.5 hr at 23°C to label the serine residue at the neutral pH optimum (2). The presence ofTPP-II in numerous active site of the enzyme. After this treatment, the enzyme tissues of the rat (2) and in erythrocytes and liver of several had <15% of the original activity. Guanidinium chloride (8 other species (4) has been demonstrated. The human eryth- M), Tween-20 [1.6% (vol/vol)], and dithiothreitol (64 mM) rocyte enzyme has been classified as a serine peptidase due were added to the incubation mixture to give the final to its complete inhibition by diisopropyl fluorophosphate concentrations of 6 M, 0.05%, and 2 mM, respectively. The (iPr2P-F) and phenylmethylsulfonyl fluoride (2). However, pH was adjusted to 8.0 with NaOH. Nitrogen gas was passed the enzyme was also inhibited by some of the thiol reagents through the solution, which was then incubated at 37°C. After tested, e.g., N-ethylmaleimide, p-chloromercuribenzoate, 1 hr, a solution of50 mM iodoacetate was added to give a final and Hg2+. TPP-II could be protected against iPr2P-F, concentration of4.4 mM. Nitrogen was again passed through phenylmethylsulfonyl fluoride, or N-ethylmaleimide by the the reaction mixture, and the alkylation was allowed to use of a competitive inhibitor (2), suggesting that not only a proceed in the dark at 37°C for 1 hr. The reaction was serine, but also a cysteine residue is of importance for the terminated by the addition of 2-mercaptoethanol to a final activity of the enzyme. To obtain an unambiguous classifi- concentration of 1%, and the sample was then dialyzed at cation of TPP-II as a serine peptidase, we considered it 23°C for 23 hr against three 2.5-liter changes of 50 mM necessary to isolate and determine the amino acid sequence ammonium bicarbonate, pH 8.3/0.05% Tween-20. The addi- of the part of the protein that contains the active serine tion of Tween-20 was necessary to prevent losses due to residue. Information on the primary structure of the active unspecific adsorption of the labeled protein. site would also be ofinterest for the investigation ofstructural To determine the apparent specific radioactivity of relationships between TPP-II and other peptidases. [3H]iPr2P-F, by a procedure that makes allowance for pos- sible elimination of3H-labeled isopropyl groups (5), 0.1 mg of The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviations: iPr2P-F, diisopropyl fluorophosphate; >PhNCS, in accordance with 18 U.S.C. §1734 solely to indicate this fact. phenylthiohydantoin; TPP-II, tripeptidyl peptidase II.

7508 Downloaded by guest on October 1, 2021 Biochemistry: Tomkinson et al. Proc. Natl. Acad. Sci. USA 84 (1987) 7509 bovine trypsin was labeled, reduced, alkylated, and dialyzed as described above for TPP-II. After the dialysis, the amount of protein was determined by the modified Bradford method (6, 7), using trypsin as the standard. The radioactivity was measured in a Philips PW4700 liquid scintillator counter after mixing the samples with 5 ml of emulsifier scintillator 299. The apparent specific radioactivity of [3H]iPr2P-F was cal- culated to be 26 x 104 cpm/nmol, and this value was used thereafter for determination of the amounts of labeled pep- tides. 40C Preparation of Tryptic Peptides. After dialysis, 6.5 nmol of 0 reduced and alkylated 3H-labeled TPP-II was incubated with 0co ._ 4._ L-1-tosylamido-2-phenylethyl chloromethyl ketone-treated 0 trypsin for 3 hr at 370C. The molar ratio of trypsin/TPP-II 04) subunit was 1:28. The digestion was stopped by the addition Q aO of 1 M HCl to give a final pH of2. The digest was then loaded onto a Sephadex G-50 column and chromatographed as 0 described in Fig. 1. As a control, the same amount of x L-1-tosylamido-2-phenylethyl chloromethyl ketone-treated E trypsin (0.2 nmol) was dissolved in 50 mM ammonium Q bicarbonate, pH 8.3/0.05% Tween-20 and chromatographed C._ on the same Sephadex G-50 column. Fractions from the two Sephadex G-50 chromatographies were pooled, as indicated 0 by the bar in Fig. 1, and lyophilized. Each lyophilized 0 Sephadex G-50 pool was dissolved in 470 ,u ofthe ammonium ._0 bicarbonate/Tween buffer. The inclusion of Tween-20 in the 0 buffer allowed a good recovery of the labeled peptides, generally 75-100%, as compared to only 10-30o when Tween-20 was excluded. The dissolved material was purified by reversed-phase HPLC on a Nucleosil C18 column as described in Fig. 2. Fractions from three identical prepara- tions were pooled as indicated. After lyophilization, the Elution time (min) pooled material was dissolved in 320 Al of the ammonium FIG. 2. Reversed-phase chromatography of tryptic peptides of bicarbonate/Tween buffer and rechromatographed as de- 3H-labeled TPP-II after Sephadex G-50 chromatography. An aliquot scribed in Fig. 3. Fractions from two identical preparations (150 pl) of the dissolved Sephadex G-50 pool was loaded on a were pooled, as indicated, and lyophilized. Pools 1 and 2 Nucleosil C18 column equilibrated with 0.1% trifluoroacetic acid and were to acid sequence analysis as described 6% (vol/vol) acetonitrile, pH 2.1, and eluted with a linear acetonitrile subjected amino gradient as indicated. The trifluoroacetic acid concentration was kept below. Pool 3 was dissolved in 160 Al of the ammonium at 0.1% throughout the gradient. The flow rate was 1.0 ml/min, and fractions of 0.5 ml were collected. Radioactivity was determined in 5 ,l ofeach fraction as described. The vertical bars correspond to an absorbance of0.02. The horizontal bar indicates the fractions pooled c for further purification. A, control; B, tryptic peptides of 3H-labeled 0 25 a b c TPP-II; C, enlargement from B, showing all radioactivity above 0~~~~~~ background level (histogram). Solid line, absorbance at 220 nm; dashed line, % acetonitrile. CU -33 bicarbonate/Tween buffer and rechromatographed as de- 0~f scribed in Fig. 4. Pools 3:1 and 3:2 were lyophilized, and their amino acid sequence was determined as described below. Cr The HPLC equipment used throughout the purification was from LDC (Riviera Beach, FL), with a Spectromonitor E (model 1204A, with a light path of 10 mm), a Dynamixer, and two pumps (Constametric III) controlled by a microproces- sor. The HPLC column was protected by a precolumn (Bondapak C18, 6 X 20 mm). 20 60 100 Determination of Amino Acid Sequences. Amino acid se- Elution volume (ml) quences were determined by use of an automated gas-phase sequencer (Applied Biosystems protein sequencer, model FIG. 1. Sephadex G-50 chromatography of trypsin-digested 3H- 470A). The sequencer was operated according to the recom- labeled TPP-II. The sample, containing 6.5 nmol of 3H-labeled mendations of the manufacturer, and all solvents used were peptides in 5.2 ml, was prepared, loaded on a Sephadex G-50 column purchased from Applied Biosystems (Foster City, CA). Forty (1.7 x 50 cm), equilibrated, and eluted with 50% (vol/vol) acetic acid. percent of the phenylthiohydantoin (>PhNCS)-amino acid The flow rate was 13 ml/hr and fractions of 2.2 ml were collected. derivatives produced in each degradation step were injected The elution position of cyanogen bromide fragments of equine on line to an Applied Biosystems 120A >PhNCS analyzer. cytochrome c, which were used for calibrating the column, are was indicated by arrows. These fragments correspond to 65 (arrow a), 24 The remaining 60% was collected, and the radioactivity (arrow b), and 15 (arrow c) amino acid residues, respectively. The determined as described above. void volume was 38 ml as determined with blue dextran. Radioac- The fraction between pool 3:1 and pool 3:2 (Fig. 4B), tivity was measured in 5 /,u from each fraction. The horizontal bar consisting of 0.09 nmol of a mixture of peptides 3:1 and 3:2, indicates fractions pooled for further purification. was loaded onto the gas-phase sequencer and analyzed on Downloaded by guest on October 1, 2021 7510 Biochemistry: Tomkinson et al. Proc. Natl. Acad. Sci. USA 84 (1987)

01£c 0 h-o 0 c 0

0._ 0

._- x._ CL._ 0 0 E 0 0 0 0 ) 0 Itx Irl0 E 0 0._ 0 T0 0 c: c) 0

Elution time(min) FIG. 3. HPLC purification oflabeled tryptic peptides. An aliquot (150 sul) ofthe samples, pooled as indicated by the bars in Fig. 2, was loaded onto a Nucleosil C18 column equilibrated with 22.5 mM ammonium acetate, pH 6.3, and 6% (vol/vol) acetonitrile, and eluted Elution time (min) with a gradient of acetonitrile as indicated. Ammonium acetate was not included in the acetonitrile solution, which gives a gradual FIG. 4. Rechromatography of labeled tryptic peptides. The decrease in ammonium acetate concentration (minimum 5 mM) and material ofpool 3 (see Fig. 3) was prepared and loaded on a Nucleosil background UV absorption during the procedure. The flow rate was C18 column equilibrated with 0.1% trifluoroacetic acid and 24% 1.0 ml/min and fractions of 0.5 ml were collected. Radioactivity in (vol/vol) acetonitrile, pH 2.1, and eluted with a linear acetonitrile 5 ALI of each fraction was determined. The vertical bars correspond gradient as indicated. The trifluoroacetic acid concentration was kept to an absorbance of 0.005. The horizontal bars indicate the fractions at 0.1% throughout the gradient. The flow rate was 1.0 ml/min, and pooled. A, control, pool from Fig. 2A; B, tryptic peptides, pool from fractions of 0.5 ml were collected. Radioactivity in 5 Al of each Fig. 2C; C, enlargement from B, showing all radioactivity above fraction was determined. The vertical bars correspond to an absorb- background level (histogram). Solid line, absorbance at 230 nm; ance of 0.005. The horizontal bars indicate fractions used for dashed line, % acetonitrile. sequence determination. A, reversed-phase chromatogram; B, en- largement from A showing all radioactivity above background level (histogram). Solid line, absorbance at 220 nm; dashed line, % line as described above. The remaining 60% ofthe >PhNCS- acetonitrile. amino acid derivatives from steps 1 to 4 were rechromato- graphed on the >PhNCS analyzer using a slightly modified gradient, which made it possible to discriminate between the labeled serine residue ofthe active site could not be identified >PhNCS derivatives of glutamine and carboxymethylated as a >PhNCS derivative, its position was determined by the cysteine. appearance of radioactivity. This residue was located at position 9 in all the peptides. RESULTS AND The background in the first cycle was too high to allow a DISCUSSION determination ofthe amino-terminal residue. This also lead to Purification of 3H-Labeled Peptides. The labeling ofTPP-II a difficulty in the unequivocal identification of the second with [3H]iPr2P-F has been shown to give an incorporation of residue of peptides 2 and 3:1 as a threonine. 0.7-0.9 mol per mol ofMr 135,000 subunit (2). After reduction As the >PhNCS derivatives of glutamine and carboxy- and alkylation, the 3H-labeled protein of the present exper- methylated cysteine cannot be separated in the HPLC system iment was dialyzed and digested with trypsin as described used in the gas-phase sequencer, a control experiment was above. Upon gel filtration on Sephadex G-50, the radioactive performed to discriminate between these two derivatives. material appeared in essentially one peak corresponding to From results of this experiment, it could be concluded that -90%o of the radioactivity loaded onto the column (Fig. 1). the amino acid residue in position 3 was glutamine. Fig. 2B shows a reversed-phase chromatogram of the An additional >PhNCS derivative could be detected in gel-filtered tryptic digest of 3H-labeled TPP-II. The main every cycle that showed a methionine residue. This material, labeled material (3.4 nmol) was pooled, lyophilized, and eluting earlier than the >PhNCS derivative of methionine, subjected to reversed-phase chromatography in a different was shown to elute with the same retention time as the solvent system (Fig. 3B), whereby the radioactive material >PhNCS derivative of methionine sulfone (data not shown). was separated into several peaks. Pool 1(0.40 nmol) and pool The amount of this derivative was 0.5, 1.5, 0.1, and 1.4 times 2 (0.23 nmol) were subjected to amino acid sequence deter- the amount of methionine at position 5 for peptide 1, 2, 3:1, mination. Pool 3, corresponding to 0.73 nmol, was and 3:2, respectively. This can explain the various yields of rechromatographed as described in Fig. 4. This purification methionine for the various peptides (Fig. 5). Oxidation of step gave two additional peptides for sequence determina- methionine residues during preparation of peptides has been tion, namely 3:1 (0.08 nmol) and 3:2 (0.13 nmol). reported (8). Amino Acid Sequence of Isolated 3H-Labeled Peptides. Fig. Peptide 1 contains an aspartic acid residue at position 6; 5 shows the results ofthe amino acid sequence analysis ofthe this position is occupied by asparagine in the other peptides. 3H-labeled tryptic fragments of TPP-II. Because the 3H- The small rise of asparagine in cycle 6 for peptide 1 might be Downloaded by guest on October 1, 2021 Biochemistry: Tomkinson et A Proc. Nati. Acad. Sci. USA 84 (1987) 7511

0 E 20 a ._ VD z 1

Cycle no. 1 2 3 4 5 6 7 8 9 1011 1 2 3 4 5 6 7 8 9 1011 Assigned X T O L M D G T S M X X X 0 L M N G T S M X amino acid TI aIT66 1 6 0 4 E 0.. 10 2 2 0

Cycle no. 1 2 3 4 5 6 7 8 9 1011 1 2 3 4 5 6 7 8 9 1011 Assigned X X Q L M N G T S M X X T O L M N G T S M X amino acid FIG. 5. Analysis of the amino acid sequence of tryptic peptides containing the active site serine residue of TPP-II. Amino acid sequence analysis was performed. The yield ofthe >PhNCS-amino acid derivatives released at each step of Edman degradation (peaks) is plotted against the cycle number. The amount ofserine (represented by the bars ofthe histogram) is calculated from the radioactivity measured in each fraction. The amino acids are denoted by the one-letter code. A, peptide 1; B, peptide 2; C, peptide 3:1, and D, peptide 3:2.

due to a small contamination by peptide 2. It remains to be a VAX 8200 computer. Although one mismatch was allowed elucidated whether the difference with respect to position 6 for, only six peptidase sequences were found. Five repre- is due to a deamidation during the preparation ofthe peptides sented various , or precursors of subtilisin, and one or whether the enzyme subunits are different in vivo. represented thermitase. We conclude from these data that the primary structure The serine peptidases are divided into two distinct families, around the serine residue of the active site of TPP-II is the trypsin type and the subtilisin type (15). The two families -Thr-Gln-Leu-Met-Asx-Gly-Thr-Ser-Met-. are considered to have evolved independently of each other, Since four 3H-labeled peptides were isolated, differences in as they have different primary and tertiary structure. Appar- addition to that at position 6, may exist. Considering their ently, no mammalian peptidase has so far been reported to elution position in the gel filtration step, these peptides were belong to the subtilisin family, whereas some bacterial expected to be of similar size-i.e., :20 amino acids long peptidases are shown to be like trypsin (16, 17). Subtilisin is (Fig. 1). As we have not yet been able to detect any residues an extracellular with Mr 27,500 (13), whereas after the second methionine residue, additional differences TPP-II is an intracellular exopeptidase with subunit Mr may reside in the carboxyl-terminal part of the peptides. For 135,000 (1, 2). It is thus obvious that the physicochemical instance, the possible existence of paired arginine and/or resemblance between these two is slight. Never- lysine residues at the cleavage sites would explain the theless, the similarity of the amino acid sequence containing appearance ofmore than one 3H-labeled, tryptic peptide even the active site residues ofthe two enzymes raises the question ifthe tryptic digestion was complete. In addition, a difference of an evolutionary relationship. in the amino-terminal residue of the peptides cannot be Thermitase, also included in the data baser is a thermo- excluded. stable endopeptidase of Thermoactinomycetes vulgaris (14, Homologies. One purpose ofthe determination ofthe amino 18). This enzyme, which belongs to the subtilisin family, has acid sequence around the serine residue of the active site of structural (Table 1) and catalytic features in common with TPP-II was to confirm that the enzyme belongs to the serine TPP-II. Both of the enzymes are inhibited by Hg2+ and peptidase class. It is clear from Table 1 that there are no p-chloromercuribenzoate as well as by iPr2P-F and distinct homologies between the active sites ofTPP-II and the phenylmethylsulfonyl fluoride (2, 18). Thermitase has been cysteine peptidases papain or ficin. Surprisingly enough, proposed to belong to a "new" subclass of subtilisin-like there was no homology with the mammalian serine peptidas- peptidases; namely, thiol-dependent serine peptidases es trypsin and either. Instead, TPP-II appar- (18-20). The enzymes of this class show, beside their ently must be classified as a serine peptidase of the subtilisin physicochemical characteristics, considerable homology to type, due to the striking homology ofthe amino acid sequence each other and are homologous to subtilisin mainly around around the active serine of TPP-II with that of the subtilisin the active site triad (aspartic acid, serine, and histidine) (19). family. A protein sequence data baser covering 963,031 So far, only microbial peptidases have been proposed to residues, was searched for the sequences Asn-Gly-Thr-Ser- belong to this class, but the existence of several mammalian Met and Asp-Gly-Thr-Ser-Met by use ofthe PSQ program on serine peptidases with an apparent sensitivity to thiol re- agents has been reported (21-23). However, none ofthe latter tProtein Identification Resource (1986) Protein Sequence Database peptidases appears to have been sequenced. Our findings of (Natl. Biomed. Res. Found., Washington, DC), Release 11.0. a mammalian serine peptidase that is thiol-dependent and Downloaded by guest on October 1, 2021 7512 Biochemistry: Tomkinson et al. Proc. Natl. Acad. Sci. USA 84 (1987) Table 1. Amino acid sequences surrounding the serine or cysteine residue of the active site of certain peptidases Peptidase Sequence Reference TPP-II (human) Thr-Gln-Leu-Met-Asx-Gly-Thr-Ser-Met This work Papain (papaya) Asn-Gln-Gly-Ser-Cys-Gly-Ser-Cys-Trp 9 Ficin (fig) Gln-Gln-Gly-Gln-Cys-Gly-Ser-Cys-Trp 10 Trypsin (porcine) Lys-Asp-Ser-Cys-Gln-Gly-Asp-Ser-Gly 11 Chymotrypsin A (bovine) Val-Ser-Ser-Cys-Met-Gly-Asp-Ser-Gly 12 Subtilisin (Bacillus subtilis) Tyr-Ala-Thr-Leu-Asn-Gly-Thr-Ser-Met 13 Thermitase (Thermoactinomycetes vulgaris) Tyr-Ala-Ser-Leu-Ser-Gly-Thr-Ser-Met 14 The active residue is underlined. subtilisin-like, indicates that this class of peptidases is more 12. Brown, J. R. & Hartley, B. S. (1966) Biochem. J. 101, widespread than was believed. 214-228. 13. Markland, F. S., Jr., & Smith, E. L. (1971) in The Enzymes, We thank Asa Flygh and Gunilla Pettersson for skillful technical ed. Boyer, P. D. (Academic, New York), 3rd Ed., Vol. 3, pp. assistance in preparing TPP-II. This work was supported by Project 561-608. 13X-04485 from the Swedish Medical Research Council and the 14. Meloun, B., Baudys, M., Kostka,V., Hausdorf, G., Frommel, Medical Faculty of Uppsala. C. & Hohne, W. E. (1985) FEBS Lett. 183, 195-200. 1. BAlow, R.-M., Ragnarsson, U. & Zetterqvist, 0. (1983) J. Biol. 15. Neurath, H. (1984) Science 224, 350-357. Chem. 258, 11622-11628. 16. Wahlby, S. & Engstrom, L. (1968) Biochim. Biophys. Acta 2. BAlow, R.-M., Tomkinson, B., Ragnarsson, U. & Zetterqvist, 151, 402-408. 0. (1986) J. Biol. Chem. 261, 2409-2417. 17. Whitaker, D. R., Jurdsek, L. & Roy, C. (1966) Biochem. 3. McDonald, J. K. & Barrett, A. J. (1986) in Mammalian Pro- Biophys. Res. Commun. 24, 173-178. teases: A Glossary and Bibliography (Academic, New York), 18. Stepanov, V. M., Chestukhina, G. G., Rudenskaya, G. N., Vol. 2, pp. 150-152. Epremyan, A. S., Osterman, A. L., Khodova, 0. M. & 4. BAilow, R.-M. & Eriksson, I. (1987) Biochem. J. 241, 75-80. Belyanova, L. P. (1981) Biochem. Biophys. Res. Commun. 5. van der Drift, A. C., Beck, H. C., Dekker, W. H., Hulst, 100, 1680-1687. A, G. & Wils, E. R. (1985) Biochemistry 24, 6894-6903. 19. Stepanov, V. M. (1985) in Environmental Regulation ofMicro- 6. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. bial Metabolism, eds. Kalaev, I. S., Dawes, E. A. & Tempest, D. W. (Academic, New York), pp. 295-300. 7. Read, S. M. & Northcote, D. H. (1981) Anal. Biochem. 116, 20. Jany, K. D. & Mayer, B. (1985) Biol. Chem. Hoppe-Seyler 53-64. 366, 485-492. 8. Savige, W. E. & Fontana, A. (1977) Methods Enzymol. 47, 21. Tsunasawa, S., Narita, K. & Ogata, K. (1975) J. Biochem. 77, 453-459. 89-102. 9. Light, A., Frater, R., Kimmel, J. R. & Smith, E. L. (1964) 22. McDonald, J. K. & Schwabe, C. (1977) in Proteinases in Proc. Natl. Acad. Sci. USA 52, 1276-1283. Mammalian Cells and Tissues, ed. Barrett, A. J. (North- 10. Husain, S. S. & Lowe, G. (1970) Biochem. J. 117, 333-340. Holland, Amsterdam), pp. 361-364. 11. Hermodson, M. A., Ericsson, L. H., Neurath, H. & Walsh, 23. Doebber, T. W., Divor, A. R. & Ellis, S. (1978) Endocrinology K. A. (1973) Biochemistry 12, 3146-3153. 103, 1794-1804. Downloaded by guest on October 1, 2021