Human Carboxypeptidase M - Semantic Scholar

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 264, No. 4, Issue of February 5,pp. 22362241,1989 Printed in U.S.A.

Human Carboxypeptidase M PURIFICATION AND CHARACTERIZATION OF A MEMBRANE-BOUND CARBOXYPEPTIDASE THAT CLEAVES PEPTIDE HORMONES* (Received for publication, September 21, 1987)

Randal A. Skidgel$$, Richard M. David, andFulong Tan$ From the Departments of $Anesthesiology and §Pharmacology, College of Medicine, Universityof Illinois, Chicago, Illinois 60680 and the IIDepartment of Neurology, University of Texas Health Science Center, Dallns, Texas 75235

A membrane-bound neutral carboxypeptidase B-like dase B (EC 3.4.17.2) was extensively characterized and is a enzyme was solubilized from human placental micro- prototype for this class of enzymes (1).Carboxypeptidase N villi with 3-[(3-cholamidopropyl)-dimethylammonio]- (arginine carboxypeptidase, kininase I, anaphylatoxin inacti1-propanesulfonate (CHAPS) and purified tohomoge- vator, creatine kinase conversion factor, EC 3.4.17.3), discovneity by ion-exchange chromatography and affinity ered in human plasma over 26 years ago,is an important chromatography on arginine-Sepharose. It gavea sin- plasma enzyme as it cleaves many endogenous peptides and gle band on sodium dodecyl sulfate-polyacrylamide gelproteins (1)(e.g. kinins (2,3), anaphylatoxins (4),fibrinopepelectrophoresis with an apparent M , of 62,000 with or tides (5), Arg6- or Lys‘-enkephalins (6), creatine kinase (7), without reduction. The enzyme is a glycoprotein as enolase (8), albumin propeptide (9), complement factor Ba shown by its high affinity for concanavalinA-Sepha(lo),etc.). A metallo-carboxypeptidase with an acid pH optirose and reduction in mass to 47,600 daltons after chemical deglycosylation.It hasa neutral pH optimum, mum, originally described in pancreatic islets (11), was puriis activated by CoC12, and inhibited by o-phenanthro- fied from adrenal medulla, brain, pituitary ( E ) , and panline, 2-mercaptomethyl-3-guanidinoethylthiopropa- creatic islets (13). Called “enkephalin convertase” and carnoic acid, or cadmium acetate, indicating itis a metal- boxypeptidase E, it was renamed carboxypeptidase H and is likely to be involved in the processing of peptide hormones in lopeptidase. The enzyme cleaves arginine or lysine from theCOOH terminus of synthetic peptides(e.g. Bz- the secretory granules (11-14).We recently detected and purified arginine/lysine carboxypeptidases from human urine Gly-Arg,Bz-Gly-Lys,Bz-Ala-Lys,dansyl-Ala-Arg, where Bz is benzoyl and dansyl is 5-dimethylamino- (15) and seminal plasma (16).These enzymes are clearly naphthalene- 1-sulfonyl)as well as from severalbiolog- distinct from carboxypeptidases A, B, N, and H although they icallyactivesubstrates:dynorphin A(1-13), Met6- do cleave many of the same substrates, for example Arg6-and Arg‘-enkephalin ( K , = 46 p ~ kcat , = 934 min”), braLys‘-enkephalins or bradykinin. dykinin (& = 16 p ~ kcat , = 147 min”), MetS-LyseWe discovered an arginine/lysine carboxypeptidase in the enkephalin ( K , = 375 p ~ kcat , = 663 min”), and Leu’plasma membrane fractions separated from various human Arg‘-enkephalin ( K , = 63 PM, kcat= 106 rnin”). Al- andanimal tissues and in cultured cells (6).In order to though the enzyme sharessome properties with other characterize the enzyme and its relationship to other carboxcarboxypeptidase B-like enzymes, it is structurally, ypeptidases, we purified it to homogeneity from the richest catalytically, and immunologically distinct from pan- source, human placental microvilli. Inaddition, we report creatic carboxypeptidase A or B, human plasma car- here that itdiffers structurally, catalytically, and immunologboxypeptidase N, and carboxypeptidaseH (“enkepha- ically from other, previously described carboxypeptidases. We lin convertase”). To denote that the enzyme is mem- propose the name “carboxypeptidase M” for this newly charbrane-bound, and to distinguish it from other known acterized enzyme. carboxypeptidases, we propose the name “carboxypeptidase M.” Because of its localization on the plasma MATERIALS AND METHODS’ membrane and optimal activity at neutralpH, carboxypeptidase Mcould inactivate or modulate the activity RESULTS of peptide hormones either before or after their interaction with plasma membrane receptors. Carboxypeptidase M Is a Glycoprotein-Purified carboxypeptidase M was tightly bound when applied to a column of concanavalin A-Sepharose preequilibrated with 0.05 M TrisMnC12, 1% HC1, pH 7.0, containing 1 mM CaCl’, 1 mM Carboxypeptidases that cleave COOH-terminal arginine or CHAPS’ and could not be eluted with the same buffer conlysine from peptides and proteinshave been studied for many taining 0.25 M NaCl or 0.25 M NaCl + 0.25 M a-methyl-Dyears (1, 2). The digestive enzyme pancreatic carboxypepti*These studies were supported by Grant HL 36473 from the National Institutes of Health and Grant 2-2-45420 from the Campus Research Board, University of Illinois at Chicago. Some of these results were reported in preliminary form at a meeting of the Federation of American Societies for Experimental Biology, St. Louis, MO, 1986. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Portions of this paper (including “Materials and Methods,” part of “Results,” Table I, and Figs. 1 and 2) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journalthat is available from Waverly Press. ’The abbreviations used are: CHAPS, 3-[(3-~holamidopropyl)dimethylammonio]-1-propanesulfonate; Bz, benzoyl; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl; HPLC, high performance liquid chromatography; PAGE, polyacrylamide gel electrophoresis; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

2236

2237

Human Carboxypeptidase M

glucoside. The enzyme was finally eluted with buffer contain- cates that the enzyme has a high affinity for the metal ion which reassociated after dilution of the chelator. To determine ing 0.25 M NaCl 0.5 M a-methyl-D-glucoside (not shown). Chemical deglycosylation of carboxypeptidase M with tri- whether carboxypeptidase M could be reactivated after refluoromethanesulfonic acid decreased its apparent size by an moval of its active site metal ion, it was dialyzed overnight average of 14,400daltons on sodium dodecyl sulfate-PAGE to against 0.05 M Tris-HC1, pH 7.5, with 0.5% CHAPS and 1 mM o-phenanthroline, then preincubated for 2 h at 4 "C with 47,600 f 1,000 (+S.E., n = 5), indicating acarbohydrate 0.1 M HEPES, pH 7.7,containing eitherno addition, 1 X content of 23% by weight (Fig. 2 ) . Effect of pH-The peptidase activity of carboxypeptidase M zinc acetate or 1 mM cobalt chloride. After dialysis, the M was tested with Bz-Ala-Lys substrate at pH values ranging enzyme had only 16% of the initial activity with dansyl-Alafrom 5.0 to 9.0.The maximal activity was obtained at pH7.0, A g while zinc acetate increased the activity to 39% and cobalt but the enzyme still retained about 90% of its activity at pH chloride to 127%. Higher concentrations of zinc inhibited (e.g. 8.0 and 60% at pH 9.0.The activity dropped sharply at acid 1.6 X M zinc acetate inhibited the native enzyme by pH, with only 47% at pH 6.0,25% at pH5.5,and no activity 73%). Thesedata indicate that carboxypeptidase M is a at pH 5.0. metalloenzyme, probably containing zinc in theactive center To test its stability at acid pH, carboxypeptidase M was as do other mammalian carboxypeptidases (1,2,32, 33). incubated for 1 h at room temperature in buffers with pH Hydrolysis of Synthetic Substrates-Carboxypeptidase M values ranging from 4.0 to 5.0 and then tested for activity at hydrolyzed peptides containing COOH-terminal Arg or Lys pH 8.5 with Bz-Gly-argininic acid, Carboxypeptidase M was including Bz-Ala-Lys, Bz-Gly-Lys, and Bz-Gly-Arg (Table stable between pH 4.5 and 5.0 but lost 17% of its activity 11). However, the ester substrate(1mM Bz-Gly-argininic acid) after an hour at pH4.25 and 33% at pH 4.0. was cleaved the fastest, about 46 times as fast as the correInhibition-The hydrolysis of Bz-Gly-Lys by carboxypep- sponding peptide substrate, Bz-Gly-Arg (5mM). Carboxypeptidase M was completely inhibited by 1 mM o-phenanthroline tidase M cleaved Bz-Gly-Arg about 1.7-fold faster than Bzor 10 p~ DL-2-mercaptomethyl-3-guanidinoethylthiopropa- Gly-Lys, showing a preference for COOH-terminal Arg over noic acid. Aprotinin (100units/ml) and 1 mM phenylmethyl- Lys. The penultimate amino acid also had a significant effect sulfonyl fluoride (serine protease inhibitors) and 0.1 mM p- on the hydrolysis rate as 1 mM Bz-Ala-Lys was cleaved about chloromercuriphenylsulfonate (which inhibits sulfhydryl pro- 17 times faster than 5 mM Bz-Gly-Lys. Carboxypeptidase M teases and carboxypeptidase H) had no appreciable effect. As (3 pg) did not hydrolyze the peptidyl dipeptidase substrate measured by HPLC with bradykinin (0.1mM) as substrate, furylacryloyl-Phe-Gly-Gly, or the carboxypeptidase A sub1 p~ ~~-2-rnercaptomethyl-3-guanidinoethylthiopropanoic strates Bz-Gly-phenyllactic acid and furylacryloyl-Phe-Phe. acid completely inhibited bradykinin hydrolysis but 1 p~ Under the same conditions, the latter two substrates were captopril (an angiotensin I converting enzyme inhibitor) or 1 rapidly cleaved by 0.1 pg of bovine carboxypeptidase A. p~ phosphoramidon (an inhibitor of neutral endopeptidase Hydrolysis of Naturally Occurring Peptides-Carboxypep24.11)had no effect. tidase M readily hydrolyzed COOH-terminal arginine or lyEffect of Metal Ions-As with other mammalian basic car- sine of several active peptides (Table 111). Of the peptides boxypeptidases, preincubation for 2 h with 1 mMCoC12 in- tested, Met5-Arg'-enkephalin was cleavedmost favorably with creased the peptidase activity of carboxypeptidase M (&fold the highest kat(934min-') and a relatively low K, of 46 p~ with Bz-Gly-Lys, 1.5-fold with Bz-Gly-Arg (Table 11) and 1.4- to give the highest specificity constant (KCat/K,,, = 20.3 p ~ " fold with dansyl-Ala-Arg) and 0.1 mM cadmium acetate inhib- min"). Carboxypeptidase M again showed a preference for ited it (49% with Bz-Gly-Lys and 51% with dansyl-Ala-Arg). arginine because changing the COOH-terminal amino acid to However, preincubation for 2 h with 1 mM cadmium acetate lysine (Met6-Lys6-enkephalin)resulted in a large increase in or 1 mM CoClz had no effect onthe esterase activity of ) a decrease in the kc, (to 663 min"), the K,,, (to 375 p ~ and carboxypeptidase M with Bz-Gly-argininic acid. and aspecificity constant less than one-tenthof that of Met6To further investigate the natureof the essential metal ion, carboxypeptidase M was preincubated with 1 mM o-phenan- Arg6-enkephalin. The penultimate amino acid also signifithroline for 2 h on iceand thendiluted to a final concentration cantly affected the specificity constant as shown with Met6of either 0.01 or 0.6 mM o-phenanthroline in the assay with Arg'-enkephalin and Leu5-Arg6-enkephalin(Table 111).In this dansyl-Ala-Arg as substrate. Under these conditions, the en- case, the K,,, of Leu5-Arg6-enkephalinwas only slightly higher ) the kat was almost an order lower than that of zyme retained only 7% of its activity when assayed in the (63 p ~ but presence of 0.6 mM o-phenanthroline while dilution to 0.01 Met5-Arg'-enkephalin, resulting in a 12-fold lower specificity mM resulted in complete recovery of its activity. This indi- constant. Of the peptides tested, bradykinin had the lowest

+

TABLEI1 Hydrolysis of synthetic substrates by human carboxypeptidase M Substrate

Conc.

CoCL"

Bz-Gly-argininic acid Bz-Ala-Lys Bz-Gly-Lys Bz-Gly-Lys Bz-Gly-Arg Bz-Gly-Arg Bz-Gly-phenyllactic acid FA-Phe-Phe FA-Phe-Gly-Gly

1 1

Activityb

pmollminlmg

rnM

-

102 2 14 22 f 7 1.3 f 0.2 6.6 f 1.3 2.2 2 0.3 3.4 f 0.2 Not cleaved Not cleaved Not cleaved

5 5 + 5 5 + 1 1 0.5 Enzyme was preincubated (+) for 2 h on ice in the presence of 1 mM CoC12. 'Results are the average values f standard errorof the mean obtained from three different preparations of carboxypeptidase M.

TABLE 111 Hydrolysis of naturally occurringpeptides by human carboxypeptidase M Reactions were conducted in duplicate at each of 8-12 substrate concentrations ranging from well below the K,,, (0.03 K,,, to 0.11 K,) to well above the K, (3-15 times K,,,). Initial rates of the appearance of product (des-Arg or Lys peptides) were calculated and the data fit to the best straight line (plotting [SIuersus [S]/v) by linear regression. Results shown are the average values from two separate determinations. Turnover numbers (kcat)were calculated assuming M, = 62,000 for carboxypeptidase M,the value obtained in SDS-PAGE. Peptide

K, p~

Enkephalin-Met'-Arg' Enkephalin-Met'-Lys' Enkephalin-Leu5-Arg6 Bradykinin

46 375 63 16

t * t

rnin"

934 663

CdK, p ~ "min"

20.3 1.8

106

1.7

147

9.2

2238


K, (16 p M ) and arelatively high specificity constant (9.2 p

~ " In addition to structural differences, the kinetic constants min"). of some biologically active peptides obtained with carboxyCarboxypeptidase M readily converted dynorphin A( 1-13) peptidases M and N differed (6). For example, the k&K, of (Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Ile-Arg-Pro-Lys-Leu-Lys) to Met5-Lyss-enkephalin is much higher with carboxypeptidase dynorphin A(1-E!). A t a substrate concentration of 100 p ~ , N (28.7 p ~ "rnin") than with carboxypeptidase M (1.8 p ~ - l the rate was 2.0 pmol/min/mg, similar to therates of hydrol- rnin"). The arginine derivative (Met6-Arg6-enkephalin)is the ysis of bradykinin (2.0 pmol/min/mg) or Met5-Lys6-enkeph- most favorable substrate for carboxypeptidase M (kcat/Km= alin (2.2 pmol/min/mg) calculated at 100 p~ substrate con- 20.3 pM" min"). The $,/K, for the hydrolysis of bradykinin centration. However, when assayed under the same condi- by carboxypeptidase M is about 3-fold higher than with tions, carboxypeptidase M did not hydrolyze dynorphin A(1- carboxypeptidase N (6). In addition, the hydrolysis rates of 13) amide, even after prolonged (23 h at 37 "C) incubation. smaller synthetic substrates by the two enzymes were differThis indicates that the enzyme is indeed a carboxypeptidase ent. These experiments show that carboxypeptidase M prefrequiring a free COOH at theCOOH terminus of a substrate erentially cleaves peptides containing COOH-terminal argiand that it does not act as an endopeptidase. The kinetic nine over those containing COOH-terminal lysine, just the studies with the other peptides described above also show that opposite of carboxypeptidase N which prefers COOH-termicarboxypeptidase M is specific for COOH-terminal basic nal lysine (1, 2, 6). Carboxypeptidase M differs from known B-type carboxyamino acids and does not have endopeptidase activity because the HPLC analyses showed that the peptide substrates are peptidases in many other ways. Carboxypeptidase M is less not cleaved further after theremoval of the COOH-terminal stable at acid pH than carboxypeptidase N (17). It does not cross-react with antiserum to carboxypeptidase N in Western ) basic amino acid. In addition, Leu5-enkephalin (100 p ~ was not hydrolyzedwhen incubated with carboxypeptidase M blotting and antiserumto carboxypeptidase M does not crossunder the same conditions as the peptides with COOH-ter- react with carboxypeptidase N or human pancreatic carboxypeptidase B. Cobalt and cadmium have no effect on the minal arginine or lysine. Lack of ~ m m u n o ~ g i cCrossreactivity al with Other Carboxy- esterase activity of carboxypeptidase M. In contrast, cobalt peptidases-In a Western blot with polyclonal antiserum to inhibits and cadmium activates the esterase activity of carcarboxypeptidase N (1:2,000 dilution), darkly staining bands boxypeptidase B (17, 18, 33), while cadmium inhibits the of MI = 83,000 and MI = 48,000 and a weaker band of M, = esterase activity of carboxypeptidase N (2, 17). Carboxypep55,000, corresponding to the high and low molecular weight tidase M is aglycoprotein (as shown by its binding to concansubunits, were seen in the lane containing 260 ng of carbox- avalin A-Sepharose and a decrease in molecular weight after ypeptidase N, while no bands were evident in the lane con- chemical deglycosylation) while carboxypeptidase B and the taining 460 ng of carboxypeptidase M(not shown). Con- active subunit of carboxypeptidase N are not glycosylated (1, versely, polyclonal antiserum to carboxypeptidase M (1:1,000 2, 17). Carboxypeptidase M is also clearly distinct from carboxydilution) gave a strong band with carboxypeptidase M (400 peptidase H. Carboxypeptidase M has a neutral pH optimum, ng), butno bands were detected with carboxypeptidase N (400 ng) or either form ( M , = 34,000 or M, = 25,000) of human while pH 5.5 is optimal for carboxypeptidase H (11-14), and pancreatic carboxypeptidase B (1 pg) (not shown). These carboxypeptidase M is not inhibitedby p-chloromercuriphenresults were confirmed in double immunodiffusion (1 pg of ylsulfonate at a concentration which completely inhibits carantigen and antiserum at 1:4 and 1:16 dilutions) where anti- boxypeptidase H (6, 12, 14). In addition, carboxypeptidase H serum to carboxypeptidase N did not cross-react with carbox- is an intragranular enzyme (12-14) while carboxypeptidase M ypeptidase M and antiserum to carboxypeptidase M did not is on plasma membranes (6, 22). We have recently isolated and sequenced the cDNA for cross-react with carboxypeptidase N or either form of human human carboxypeptidase M.3 The deduced amino acid sepancreatic carboxypeptidase B (not shown). quence is fully consistent with the data presented in this paper, including the presence of potential glycosylation sites, DISCUSSION the molecular weight of the deglycosylated enzyme, and its This report describes the purification and characterization uniqueness when compared with known carboxypeptidases. of a human membrane-bound carboxypeptidase. We propose Although carboxypeptidase M was purified from human the name carboxypeptidase M to denote with the letter " M placenta, the enzyme is present in other tissues as well. We that it is membrane-bound and to distinguish it from other previously detected significant carboxypeptidase activity in known mammalian carboxypeptidases which are primarily membrane fractions of various human and animal tissues as soluble enzymes (1). well as in several lines of cultured cells (6). Besides the Carboxypeptidase M was purified seven times using some- placenta, these included human kidney, lung, pulmonary arwhat different procedures (see "Materials and Methods"). We terial endothelial cells, fibroblasts, and amniotic fluid, and found the most reproducible procedure to be sequential chro- bovine lung and pulmonary artery (6). More recently, others matography on thefollowing four columns in the order given: also reported the presence of a carboxypeptidase cleaving the 1) DEAE-Trisacryl, 2) Q-Sepharose, 3) arginine-Sepharose, COOH-terminal arginine of bradykinin or Leu5-Arg6-enkephand 4) HPLC on Mono-Q HR. alin in membrane fractions from porcine blood vessels and Although carboxypeptidase M shares some properties with hog aortic endothelial cells (34). other carboxypeptidase B-like enzymes, it is a distinctly difThe potential importance of cleaving COOH-terminal arferent enzyme. Thus, while carboxypeptidase M is a single ginine or lysine in the activation or regulation of peptide chain protein with an M, = 62,000, human plasma carboxy- hormone activity has generated a great deal of interest in this peptidase N is a tetramer (containing two high and two low type of carboxypeptidase (1). For instance, it is now known molecular weight subunits) of M , = 280,000 (1,2, 17),human that most peptide hormones are synthesized as partof a larger pancreatic carboxypeptidase B has an M , of 34,250 (18),and precursor protein which must be enzymatically processed in carboxypeptidase H has an M, of 50,000 to 52,000 (12-14). The NHz-terminal amino acid of carboxypeptidase M also F. Tan, S. J. Chan, R. A. Skidgel, andJ. W. Schilling,manuscript in preparation. differs from those of the other B-type carboxypeptidases.

2239

Human Carboxypeptidase M order to release the active hormone (35). In many cases, this involves initial enzymatic cleavage at paired basic residues in the precursor molecule to release a peptide containing additional COOH-terminal arginine or lysine residues which must be removed by a carboxypeptidase to generate a fully active peptide. These carboxypeptidases can also inactivate peptide hormones (1).For example, anaphylatoxin C3a is inactivated by carboxypeptidase N (4,36). Inaddition, removal of COOHterminal basic residues could modulate the activity of peptide hormones (1). For instance, bradykinin exerts many of its biological effects through the so-called B2 receptor (37). Conversion of bradykinin to des-Arg-bradykinin renders the peptide inactive at theBzreceptor and converts it toa specific agonist for the B1receptor (37). Similarly, anaphylatoxin C5a has both spasmogenic and chemotactic properties (36). Removal of the COOH-terminal arginine by a carboxypeptidase abolishes the spasmogenic and histamine-releasing activity of C5a while the chemotactic effectiveness is retained (36). The actual functional role that the various arginine/lysine carboxypeptidases play in vivo is probably related to their localization as well as their physical properties. Carboxypeptidase M, because of its presence on plasma membranes and optimal activity at neutral pH, is ideally situated to act on peptide hormones at local tissue sites (1).It could inactivate or modulate their activity either before or after their interaction with specific plasma membrane receptors. Placental microvilli are the sites at which the materno-fetal exchange takes place (38) and arerich in peptidases such as angiotensin I converting enzyme, neutral endopeptidase 24.11, and aminopeptidase A (22). Carboxypeptidase M at thislocation could play a protective role, contributing to the inactivation of potentially deleterious peptides before they cross this important barrier. Acknowledgments-We thank Elizabeth Obiku and Norman Albertson for their expert assistance, Dr. James Schilling of California Biotechnology, Inc. for performing the NHZ-terminalsequencing, and Dr. Ervin G. Erdos for his encouragement and helpful discussions. REFERENCES 1. Skidgel, R. A. (1988) Trends Pharmacol. Sci. 9, 299-304 2. Erdos, E. G. (1979) in Handbook of Experimental Pharmacology (Erdos, E. G . , ed) vol. 25, pp. 428-487, Springer, Heidelberg 3. Erdos, E. G., and Sloane, E. M. (1962) Biochem. Pharmacol. 11, 585-592 4. Bokisch, V. A., and Muller-Eberhard, H. J . (1970) J. Clin. Invest. 49,2427-2436 5. Teger-Nilsson, A. C. (1968) Acta Chem. Scand. 22, 3171-3182 6. Skidgel, R. A., Johnson, A. R., and Erdos, E. G. (1984) Biochem. Pharmacol. 33,3471-3478 7. Michelutti, L.,Falter, H., Certossi, S., Marcotte, B., and Maz-

zuchin, A. (1987) Clin. Biochem. 20, 21-29 8. Wevers, R.A., Boegheim, J. P. J., Hommes, 0. R., van Landeghem, A. A. J., Mul-Steinbusch, M. W.F. J., van der Stappen, J. W. J., and Soons, J. B. J. (1984) Clin. Chim. Acta139, 127135 9. Peters. T.. Jr.. and Davidson, L. K. (1986) J . Biol. Chem. 261, 7242-7246 10. Davrinche. C.. Charlionet. R.. Rivat. C.. Helal. A. N.. and Lefranc. G. (1984) Eur. J. Immunol 14,9571961 11. Zuhlke, H., Steiner, D. F., Lernmark, A., and Lipsey, C. (1975) Ciba Found. Symp. 41,183-195 12. Fricker, L. D., and Snyder, S. H. (1983) J. Biol. Chem. 258, 10950-10955 13. Davidson, H. W., and Hutton,J. C. (1987) Biochem. J. 245,575582 14. Fricker, L. D. (1985) Trends Neurosci. 8, 210-214 15. Skidgel, R.A., Davis, R.M., and Erdos, E. G. (1984) A d . Biochem. 140,520-531 16. Skidgel, R.A., Deddish, P. A., and Davis, R. M. (1988) Arch. '

'

Biochem. Biophys., in press Y.,Skidgel, R.A., and Erdos, E. G. (1982) Proc.Natl. Acad. Sci. U. S. A. 79,4618-4622 Marinkovic, D. V., Marinkovic, J. N., Erdos, E. G., and Robinson, C. J. G . (1977) Biochem. J. 163, 253-260 Porath, J., and Fornstedt, N. (1970) J. Chromatogr. 51,479-489 Skidgel, R. A., and Erdos, E. G . (1984) in Methods of Enzymatic Anulysis (Bergmeyer, H. U., ed) vol. 5, pp. 34-43, Verlag Chemie, Weinheim Dixon, M., and Webb, E. C. (1979) Enzymes, 3rd ed, pp. 60-62, Academic Press, New York Johnson, A. R., Skidgel, R. A., Gafford, J. T., and Erdos, E. G. (1984) Peptides 5, 789-796 Edge, A. S., Faltynek, C.R., Hof, L., Reichert, L. E., Jr., and Weber, P. (1981) Anal. Biochem. 118, 131-137 Laemmli, U. K.(1970) Nature 227,680-685 Morrissey, J. H. (1981) Anal. Biochem. 117, 307-310 Tarr, G. E. (1986) in Methods of Protein Microcharacterization (Shively, J. E., ed), pp. 155-194, Humana Press, Clifton, NJ Peterson, G. L. (1977) Anal. Biochem. 83, 346-356 Hjelmeland, L.M., and Chrambach, A. (1984) Methods Enzymol.

17. Levin, 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

104,305-328 29. Andrews, P. (1965) Biaehem. J. 96, 595-606 30. Skidgel, R. A., Bennett, C. D., Schilling, J. W., Tan, F., Weerasinghe, D. K., and Erdos, E. G. (1988) Biochem. Biophys. Res. Commun. 154,1323-1329 31. Fricker, L. D., Evans, C. J., Esch, F. S., and Herbert, E. (1986) Nature 323,461-464 32. Vallee, B. L., Rupley, J. A., Coombs, T. L., and Neurath, H. (1960) J. Biol. Chem. 235,64-69 33. Folk, J. E., and Gladner, J. A. (1961) Biochim. Biophys. Acta48, 139-147 34. Palmieri, F. E., Bausback, H. H., Churchill, L., and Ward, P. E. (1986) Biochem. Pharmacol. 35, 2749-2756 35. Steiner, D.F., Docherty, K., and Carroll, R. (1984) J. Cell. Biochern. 24,121-130 36. Hugli, T. E. (1980) Crit. Reu. Immunol. 1, 321-366 37. Regoli, D., and Barabe, J. (1980) Pharmacal. Rev. 32, 1-46 38. Kaufmann, P. (1985) Contr. Gynec. Obstet. 13,5-17

Continued on next page.

2240



Mlc.ro*illl

1~11(.*"1"l

51

95

CHAPS Swprn.lml

n w

T

~

I

~ I~

~

.

~ IP I

n w T ~ I . ~ II ~ ~ ~ I Ar~lnla-frphnror*

25 1.8

211 81.1

1.12

0.014

lW

I.o

5.bO

0.061

Ib9

h.9

1.98

h.59

2.12

118

Ibb

0.88

1.80

4.12

llh

1W

0.11

l.hO

h2

7lh

10.0

2241