Preferential action of rat brain cathepsin B as a peptidyl dipeptidase converting pro-opioid oligopeptides

ARCHIVES OF BIOCHEMISTRY Vol. 249, No. 2, September,

Preferential

AND BIOPHYSICS pp. 489-499,1986

Action of Rat Brain Cathepsin B as a Peptidyl Dipeptidase Converting Pro-Opioid Oligopeptides N. MARKS,

Center

for

Neurochemistry, Ward’s Received

February

M. J. BERG,

AND

M. BENUCK

The Nathan S. Kline Institute Island, New York, New York 18, 1986, and in revised

form

for Psychiatric 100.35 May

Research,

12, 1986

Purified rat brain cathepsin B (EC 3.4.22.1) converted prodynorphins or proenkephalins to shorter active forms by the preferential removal of C-terminal dipeptides. The substrate affinities for Met-enkephalin-Arg-Phe or -Arg-Gly-Leu were Km 46 and 11’7 PM, and k,,,/K, ratios were 67 and 115 PM-~, min-‘, respectively. Met-Enkephalin was inactivated by the same mechanism (Km-450 ,LLM; kcc,,,/Km = 0.12 PM-‘mind’). The comparison of cathepsin B hydrolysis for pro-opioids, a synthetic hexapeptide and its fragments, C-blocked peptides (pro-opioid amides, Met-enkephalin amide, substance P), and bovine myelin basic protein, provided information on the influence of the C-terminal residues on dipeptide release, the rates as correlated to peptide length, and the optimal arrangement of residues favoring scission at the PI-Pi sites. The brain enzyme was stereospecific and did not act on peptides with C-terminal D-amino acid substituents. Arg hindered and Pro blocked the release of C-terminal dipeptides when in the Ph positions. The suppression of dipeptide release by agents inhibiting endopeptidase actions such as E64 and leupeptin, and the endogenous brain factor (cerebrocystatin) point to similar catalytic mechanisms for the exopeptidase action. Q 1986 Academic Press, Inc. -

Cathepsin B (EC 3.4.22.1), a lysosomal cysteine proteinase acts either as a peptidyl dipeptidase (sometimes referred to as dipeptide carboxypeptidase) or as an endopeptidase (1). The factors accounting for this differential expression of activity are unclear as illustrated for a human liver enzyme, acting on the C-terminus of glucagon and aldolase, yet cleaving one-third of the available internal peptide bonds of insulin B chain (2-5). Previously, we showed that bovine brain or human pituitary cathepsin B cleaved internal bonds of /3-lipotropin or its fi-endorphin fragment (6-8). In contrast the rat brain enzyme was shown in preliminary studies to act preferentially as a peptidyl dipeptidase towards smaller pro-opioids yielding shorter active forms (9). These observations prompted the present study on structural

features of neuropeptides or other substrates that favor a C-terminal action by brain cathepsin B. For this purpose, the pro-enkephalins and prodynorphins provide convenient models to examine the influence on dipeptide release of peptide size, the effects of different N- and C-terminal substituents, and the composition of the domains flanking the scissable bonds (10, 11) and are compared to a synthetic hexapeptide and its fragments, amidated opioid peptides, and the undecapeptide amide substance P. To examine the catalytic mechanisms associated with dipeptide release, we have compared the effects of endopeptidase inhibitors (E-64, leupeptin), and an endogenous brain factor (cerebrocystatin) shown previously to suppress myelin basic protein (MBP)’ breakdown by the rat brain enzyme (12). 489

0003-9861/86 Copyright All rights

$3.00

0 1986 by Academic Press, Inc. of reproduction in any form reserved.

490

MARKS,

BERG,

The C-terminal modification of prohormones constitutes a pathway for their activation (11-13). This is especially relevant for pro-opioids containing an essential Nterminal opioid determinant (Tyr-Gly-GlyPhe-Met or Leu). Currently, there is interest in cathepsin B-like enzyme and its role in the processing of prohormones (PTH, insulin) or proproteins (albumin) in secretory tissues (see Refs. (14-17)). The mapping of cathepsin B specificity as a peptidyl dipeptidase, an area not previously investigated in a systematic manner, can provide insights into the general mechanisms available for neuropeptide activation. MATERIALS

AND

METHODS

Peptides were obtained from the following sources: Dynorphin A 1-6 and Phe-Met-Arg-PheNH, (FMRF. NHz, cardioactive neuropeptide) CCK-8, LH-RH, and angiotensin-I from Bachem (Torrance, Calif.), Leu and Met-enkephalin, or analogs with C-terminal Arg, Arg, Phe(NHz), and Dynorphin A l-8 Peninsula Labs (Belmont, Calif.), the model hexapeptide Leu-Trp-MetArg-Phe-Ala and its fragments from Research Plus (Bayonne, N.J.), CBZ-Arg-Arg-NHMec from Cambridge Biochemical Ltd. (Atlantic Beach, N.Y.), CBZPhe-Arg-NHMec and E-64 or N-[A’-(t-3-trans-carboxyoxiran-2-carbonyl)-L-leucyl] agamatine from Peptides International (Louisville, KY.), Met-enkephalin-Arg’-Glyr-Let? was synthesized and supplied by H.-L. Lahm (Hoffman-LaRoche, Nutley, N.J.), ArgPhe, Phe-Arg, Phe-Met, Phe-Gly, Met- and Leu-enkephalinamide, or the D-Met’-ProSNHz, D-Met’, DPhe5, and D-Ala’-D-Let? analogs of ME were from Sigma (St. Louis, MO.). Bovine MBP was supplied by George A. Hashim (St. Luke’s Hospital Center, N.Y.). Enzyme grade sucrose and ammonium sulfate were purchased from Sigma. Enzymes pu$ication. Cathepsin B was purified from rat brain Pz fraction by a modification of the procedure described by Suhar and Marks, and Suhar et aL (6,7). Brains were removed from 10 male Sprague-Dawley rats 6-8 weeks of age and homogenized in 10 vol of 0.32 M sucrose, and the debris and nuclear fraction

’ Abbreviations used: ME and LE are methionine or leucine enkephalin; MBP, myelin basic protein; ACE, angiotensin I-converting enzyme; CCK-8, cholecystokinin octapeptide; LH-RH, luteinizing hormone releasing hormone; -NHMec, 7-amino-(4methyl)coumarylamide; SDS, sodium dodecyl sulfate; RP-HPLC, reverse-phase high-performance liquid chromatography; FPLC, fast protein liquid chromatography.

AND

BENUCK

were removed by centrifugation at 1000~ for 20 min. The Pa fraction obtained at 14,000~ was dispersed in 20 vol of 100 pM sodium acetate, pH 4.5, containing 1 mM EDTA, incubated for 1 h at 37°C and recentrifuged at 3O,OOOg, the pellet was discarded. The supernatant was fractionated first with 40% w/v ammonium sulfate, with the pellet discarded, and then with 80% w/ v; the pellet after centrifugation was dialyzed overnight in the cold against 20 mM sodium acetate, pH 5.0, containing 1 mM EDTA. Dialyzed material was applied to a 2.5 X 30-cm column of CM-cellex (BioRad, Richmond, Calif.) equilibrated with 50 mM sodium acetate buffer, pH 5.5, containing 1 mM EDTA. The protein was eluted with a one-column volume of buffer and then a O-O.4 M sodium chloride gradient (400 ml), and 4-ml fractions were collected. Active fractions were pooled and concentrated on a Centricon filter (Amicon, Lexington, Mass.). Concentrated enzyme was applied to a Mono-S (FPLC) column (Pharmacia, Piscataway, N.Y.), 5 X 50 mm, equilibrated in 50 mM sodium acetate, pH 5.5, containing 1 mM EDTA. A linear sodium chloride gradient of O-O.5 M was applied over a 40-min period with a flow rate of 1.2 ml per min, and 1.2-ml fractions were collected. Enzyme was stored at 4°C. Alternatively, enzyme was purified after the salt precipitation step by application to a phenyl-sepharose (Pharmacia) column, 25 X 2.5 cm, equilibrated in 100 mM sodium acetate buffer, pH 4.5, containing 1 mM EDTA and 0.1 mM HgClz. Enzyme was eluted in good yield but with lower specific activity than the FPLC step with a descending 90-O w/v% gradient of WW,SO,. Enzyme assays. The reaction mixture of 3 ml of 100 mM sodium phosphate, pH 6.0, containing 1 mM EDTA, and 2 mM cysteine was preincubated 5 min at 37°C with an aliquot of enzyme followed by addition of 3 nmol of Z-Phe-Arg-NHMec, and fluorescence was read in a Turner III fluorometer (excitation 310 nm and emission 415 nm). The quantity of protein was determined by titration with E-64 by addition of the inhibitor to the enzyme during the 15-min preincubation period followed by assay. In earlier steps of the purification, protein was determined by the method of Lowry et al. (18). Enzyme activity in some cases was assayed also by the method of Barrett (19) using ZPhe-Argor Z-Arg-Arg-NHMec as substrate with measurement of the released fluorophore by fluorometry. In typical assays on peptide metabolism, lo-20 ng of purified enzyme (based on the E-64 titration) was preincubated in a volume of 0.1 ml containing 5 pmol of sodium phosphate, pH 6.0,0.05 wmol of EDTA, 0.2 pmol of cysteine-HCI, and lo-100 nmol of peptide substrate, and incubated 5-120 min at 37°C. The reaction was terminated with 20% perchloric acid to give a final concentration of 5%, or with 20 ~1 of a mixture of methyl alcohol, and HCl (1:l). Aliquots

CATHEPSIN

B CONVERSION

were analyzed for breakdown products by RP-HPLC on a 5 fi Cis Adsorbosphere column (Alltech Associates, Waukegan, Ill.), eluted with 0.1 M potassium phosphate, pH 3.0, and acetonitrile O-60% ,and monitored at 214 nM. Studies on MBP. Digest mixtures were prepared by incubation of lo-20 mg of bovine MBP in 2 ml of 0.1 M potassium phosphate buffer pH 6.0, containing 5 mM cysteine and 1 mM EDT& and enzyme equivalent to 10 units (10 nmol of Z-Phe-Arg-NHMec hydrolyzed per min) in a ratio of enzyme to substrate of approximately 1:25-50. Incubations were from 1 to 6 h and aliquots were kept frozen at -20°C. The digest mixtures were separated on a calibrated Sephadex G-75 column, 30 X 2 cm, previously equilibrated with 0.2 M acetic acid and monitored at 235 nM on a Vydac C, column (Separations Group, Hesperia, Calif.), 250 X 4.5 mm, and eluted with a O-40% gradient formed between 95% acetonitrile:O.l% trifluoroacetic acid (TFA) and 0.1% TFA:water using the gradient system of Lewis (20). Eluants representing peaks absorbing at 235 nm (Sephadex) or 214 nm (RP-HPLC) were combined, lyophilized, and used for further analyses. Amino acid analysis. Samples were hydrolyzed in redistilled 6 N HCl containing 2.5% thioglycolic acid for ‘70 min at 155°C for 18 h at llO”C, dried by flash evaporation, and applied to an ion-exchange column for separation followed by o-phthalaldehyde detection. In other cases, amino acids were derivatized with ophthalaldehyde according to the procedure of FernStrom and Fernstrom (21) and separated by RP-HPLC on Cl8 columns. End-grmp analyses. The low-molecular-weight materials obtained from gel filtration were lyophilized and dansylated, and the products were separated on polyamide plates (22). Material purified by Vydac C1 HPLC was incubated at pH 6.8 with carboxypeptidase Y (Sigma Biochemicals) at room temperature, or with carboxypeptidases A and B (Sigma Biochemicals) (23), and the released amino acids were derivatized with o-phthalaldehyde and separated by RP-HPLC as described above. Gel electrophoresis. The fragmentation of bovine myelin protein (MBP) was monitored on a urea-SDS gel as described previously (12). Purified cathepsin B itself was analyzed on 15% acrylamide slab gels, 8 X 7 cm, according to the method of Laemmli (24) and silver-stained by a modification of the method of Wray et al. (25) (personal communication H. Levy, Dept. Cell Biol., N.Y.U. Medical Center). Gels were scanned at 580 nM on a Hoefer GS-300 densitometer (San Francisco, Calif.) attached to a Spectra-Physics integrator (San Jose, Calif.). RESULTS

The enzyme purified from rat brain was identical in properties to the bovine enzyme used previously (5, 16). In some cases im-

OF

PRO-OPIOIDS

491

purities were observed when preparations were analyzed on an SDS gel according to the method of Laemmli (24), but these could be removed by ion-exchange on a Mono-S (FPLC), or by hydrophobic chromatography on a phenyl-Sepharose column as described. The specific activity of different preparations ranged between 2 and 3 pmol of Z-Arg-Arg-NHMec hydrolyzed per mg of purified enzyme per min (enzyme content in the final FPLC step was assayed by titration with E-64), equivalent to a 5000-fold purification as compared to the original (activated) Pa fraction. Exopeptidase activity. Incubation of opioid peptides with enzyme for 5-180 min at 37°C led to the appearance of smaller fragments and C-terminal dipeptides as detected by RP-HPLC (Table I). Some of the substrates were degraded rapidly, within 5-15 min, while others, notably the substrates 5-9 in the table, required up to 3 h in order to detect products by RPHPLC. On a relative basis, the best substrate was the pro-opioid ME-Arg-Gly-Leu yielding a rate of 1.6 pmol per mg of protein per min. The hydrolysis of this substrate led to the release of ME-Arg and Gly-Leu as products. Compared to the octapeptide, the rates observed for the other substrates were lower, 30-60% for substrates 2-4, and 2% or less for the remainder listed in Table I. The data indicated that arginine when it was C-terminal significantly lowered the rates of dipeptide release, as illustrated in the case of substrates 5-8. The hydrolysis rate shown for ME-Arg-Phe, for example, was 16 times higher than that of ME-ArgArg. Degradation of Dyn A l-6 (LE-Arg‘) and the two pentapeptide enkephalins led to the formation of tri- and tetrapeptide end-products. No enzyme activity was observed for the substrates Dyn l-10 (a prodynorphin with a C-terminal Pro), and the three substituted pentapeptide enkephalins having D-amino acids in positions 4 and 5 (substrates 11-13). These data show that rat brain enzyme only degraded peptides having C-terminal L-substituents other than proline. The bonds cleaved in this series included the following: Arg-Gly, LysLeu, Arg-Arg, Met-Arg, Arg-Ile, Leu-Arg, Phe-Leu, and Gly-Phe.

492

MARKS,

BERG,

AND

TABLE PEPTID~L

DIPEPTIDASE

I

ACTION OF PURIFIED RAT BRAIN CATHEPSIN SELECTED ENKEPHALIN PEPTIDES Cleavage point (---) and dipeptide released

Substrate 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

BENUCK

ME-Arg-Gly-Leu Dyn A 1-13 Dyn A 1-8 ME-Arg-Phe Dyn A 1-9 Dyn A 1-7 ME-Arg-Arg Dyn A l-6 ME or LE Dyn A l-10 (D-Met5)-ME (D-Phe4)-ME (D-Ala’, D-Leu’)-ME

Product

-Arg’---Gly’-Leu’ -Lys1’--Leu’2-Lys’S -Arg6---ArgrIle’ -Me@---ArgG-Phe’ -Arg’---I@-A$ -Let?---ArdArg’ -Mets---Arti-Arg’ -Phe”---Leu’-Ard -Gly3---Phd-Mef(Ld) ND” ND ND ND

B TOWARD

Relative rate*

measured

ME-Arg’ Dyn A l-11 Dyn A 1-6 ME, Arg-Phe Dyn A l-7 LE ME Tyri-Gly-Gly-Phe4 Tyr’-Gly-Gly3

100 63 42 32 2 2 2 0.5 0 0 0 0

Note. Incubation mixture of 100 ~1 or 0.1 M potassium phosphate buffer, pH 6.0, containing 1 mM EDTA, 2 mM cysteine HCI, 20 ng of enzyme (based on titration with E-64) was incubated at 37°C for 5-180 min depending on rate of product formed (see column 4). Reaction fixed with 5% perchloroacetic acid and aliquots used for RP-HPLC separations on 5 p Cl8 columns or in the case of Dyn A 1-13 on a 5 1 Cs column. Values are the means of 4-6 determinations agreeing within 12%. Abbreviations: ME, methionine enkephalin; LE, leucine enkephalin; Dyn A’-13, the prodynorphin fragment Tyr-G1y-Gly-Phe-Leu5-ArgG-Arg7-Ile8-Arg9-Pro’O-Lys-LeuLys’3. a Not detected. b Rate for substrate 1 (=lOO%) was 1.6 pmol per mg per min.

Synthetic hexapeptide. The synthetic hexapeptide shown in Table II also served as a substrate for exopeptidase action with

TABLE PEPTIDYL

Substrate

cleavage at the Arg-Phe bond. In addition, fragments containing one residue less on the C-terminus also acted as substrates II

DIPEPTIDASE ACTION OF PURIFIED RAT BRAIN CATHEPSIN A SYNTHETIC PEPTIDE AND ITS FRAGMENTS

and cleavage

point

(T)

1. Leu’-Trp’-Mets-Arg4-Phe’-Ala’

Product

measured

B TOWARD

Relative

rate”

Leui---Arg4 Phe’-Ala’

100

2. Leu-Trp-Met-Arg-Phe 1

Leu’---Met3 Arg*-Phe5

56

3. Leu-Trp-Met-Arg t 4. Met-Arg-Phe-Ala T 5. Met-Arg-Phe

Leu’-Trp*

9

Phe’-Ala’

4.6

ND

0

T

Note. For conditions a Rate for substrate

of incubation and other details 1 was 0.72 pmol per mg per h.

see Table

I.

CATHEPSIN

B CONVERSION

yielding a C-terminal dipeptide as one of the products. No hydrolysis was observed, however, for the tripeptide fragment MetArg-Phe. The approximate ratio of hydrolysis rates for the hexapeptide and its penta- and tetrapeptide fragments was 1: 0.6:0.06; the hydrolysis rate of the synthetic hexapeptide itself was 0.45 compared to that of ME-Arg-Gly-Leu (Tables I, II). The two tetrapeptide fragments were cleaved to yield pairs of dipeptides. The sites cleaved in this series included Arg-Phe, Met-Arg, and Trp-Met. Endopeptidase action. Several amidated peptides served as substrates for brain cathepsin B yielding degradation products detected by RP-HPLC. On a relative basis, ME-Arg-Phe * NH2 was hydrolyzed at rates comparable to the “exopeptidase” cleavage observed for ME-Arg-Gly-Leu (Tables I, III). In contrast, ME. NH2 or LE * NH2 were cleaved at low rates to release Tyr-Gly-Gly and the C-terminal dipeptide amides. The presence of the amide group in the case of the heptapeptide, however, led to a shift in the point of cleavage from Met-Arg for the non-amidated form to Arg-Phe, with release of ME-Arg and Phe . NH2 as products (Table III; Fig. 1). The tetrapeptide fragment Phe-Met-Arg-Phe . NH2, a cardioactive material found in spinal cord (26), was cleaved similarly at the Arg-Phe site at a lower rate to release Phe-Met-Arg and Phe . NH2 as products (Fig. 1). Substance P, a undecapeptidyl amide, was cleaved by purified enzyme at the Gly-

TABLE ENDOPEPTIDASE

1. 2. 3. 4. 5. 6. Note. “Rate

OF

Leu bond to release substance P l-9 and Leu-Met . NH2 (Fig. 1, Table III). Increase in time of incubation led to the appearance of other products, one of which was identified as substance P l-6. MBP incubated with enzyme in the ratio of 25:l was cleaved within l-6 h to form three to four peptides as detected on SDSurea gels in agreement with previous observations (see (12)). To decide if these products resulted from either exo or endopeptidase actions, or a mixture of both, the 6-h digest mixture was fractionated on a Sephadex G-75 column into low- and high-molecular-weight materials. The lowmolecular-weight materials contained the dipeptide Arg-Arg as seen when aliquots were dansylated and products were separated on polyamide plates. Analysis of the high-molecular-weight fraction obtained by gel filtration indicated that it contained all the larger polypeptides observed in the original digest mixture. A partial separation of these polypeptides was achieved by RP-HPLC on a C4 column as illustrated in Fig. 2. Within 6 h at 37”C, the enzyme degraded 90% or more of MBP, yielding three large and several smaller products. The first and largest peak shown for the RP-HPLC profile (Fig. 2) was used for end-group analysis. The densitometry pattern of peak I indicated the presence of only one major polypeptide; the others shown were minor and appeared to be impurities associated with the MBP controls and were not altered significantly as a re-

III

ACTIVITY OF PURIFIED RAT BRAIN CATHEPSIN ENKEPHALIN AMIDES AND SUBSTANCE P

Substrate

Cleavage point (---) and product released

ME-Arg-Phe-NH, Substance P Phe-Met-Arg-Phe-NH, ME *NH2 LE.NH2 (D-Met?, Pro. NH&ME

-Arg6---Phe . NH, -Gly’---Led’Met * NH, -Ar$---Phe. NH -G1$---Phe-Met * NH2 -Gly3---Phe-Leu . NH* ND

For conditions for substrate

493

PRO-OPIOIDS

of incubation and other details 1 was 1.47 pmol per mg enzyme

see Table per min.

Product

B TOWARD

measured

Phe . NH2 Substance Phe . NH2 Tyr-Gly-Gly Tyr-Gly-Gly I.

P 1-9

Relative ratea 100 20 14 1.6 0.5 0

494

MARKS.

BERG,

0.015

0.010

0.005

0.015 6 2

0.010

0

5 ap

2 0.005

AND

BENUCK

values between 46 PM for the heptapeptide amide and 470 PM for ME (Table IV). The keatvalue for ME-Arg-Gly-Leu was 240-fold higher than that of ME, and 4- to lo-fold higher compared to ME-Phe-Arg or its amide, and substance P. The k,,JK, values were 67-113 PM-‘min-’ for the hepta/octapeptide enkephalins, or 50- to loo-fold higher compared to ME, and 3- to 6-fold higher compared to substance P. Relatively low rates of metabolism were observed for other octa- and decapeptide amides and were comparable to those of ME (CCK-8 and LH-RH) and for the decapeptide angiotensin-I. For purposes of simplicity the rates for the latter substrates are indicated as single points at 2 h in Fig. 3. Inhibitors. An assay system containing of ME-Arg-Phe and purified enzyme was used to compare potencies of added material on rates of C-terminal dipeptide release. The release of Arg-Phe was sup-

0.02

TIME(MIN)

FIG. 1. Comparison of RP-HPLC profiles on a 5 p Cl8 column for brain cathepsin B degradation of MEArg-Phe (top) and its amidated form (middle), and substance P (bottom). Incubations were for 15 min at 37°C under the conditions described under Materials and Methods. The composition of identified peaks was established by amino acid analyses. The first (unidentified) peak for each substrate represents the injection artifact. ME denotes methionine enkephalin.

sult of incubation (Fig. 2, inset). The treatment of the peak I polypeptide with carboxypeptidases as described led to the sequential release of Phe, His, and Val with time of incubation and this was consistent with cleavage of bovine MBP at the PhesgPhego bond (see Ref. (27) for MBP structure). Kinetics. The rates for substrate hydrolysis plotted against time were linear up to 20 min for substance P, the heptapeptide amide, and its non-amidated form (Fig. 3). All of these substrates obeyed MichaelisMenten kinetics over a lo- to 15-fold range of substrate concentration, yielding Km

0.01

? * 2 co : <

1

10

20

TIME

30

40

50

(MIN)

FIG. 2. RP-HPLC profile on a 5 w Ca column for cathepsin B digestion of bovine myelin brain protein at zero time (top) and after 6 h at 37°C (lower). For conditions of incubation and separation see Materials and Methods. The inset in the lower portion represents the SDS electrophoretic profile scanned by densitometry for the first and largest peak eluted from the Cd column (see text). The arrows on the densitogram from left to right represent protein standards trypsinogen (M, = 24,000) @-lactoglubulin (&fr = l&400), and cytochrome c (M, = 12,400).

CATHEPSIN 80

B CONVERSION

OF

PRO-OPIOIDS

[email protected]* /

1

TIME

(MINI

FIG. 3. Comparison of the rates of hydrolysis for representative opioids or amidated forms, and substance P. For purposes of clarity the amidated octapeptide CCK-8, and decapeptide LH-RH, and the nonamidated decapeptide angiotensin-I are included as single points after 2 h incubation. Rates were determined by the net decrease in peak heights for the individual substrates after separation by RP-HPLC on a 5 fi Cl8 columns by the procedures described. The rates observed for ME eNHa or LE. NH2 are not shown but were comparable to that indicated for non-amidated forms.

pressed 50% by 17-24 nM amounts of leupeptin and E-64, as compared to 6-fold higher amounts for cerebrocystatin, and 300- to 900-fold higher for other added materials (Table V). These included dipeptide end-products of opioid peptides formed as a result of breakdown by cathepsin B (ArgPhe, Phe-Arg, and Phe-Met), an inhibitor of angiotensin converting enzyme (MK421), an inhibitor of a membrane-bound metalloendopeptidase (phosphoramidon), and a noncompetitive inhibitor of cathepsin B (hoc-D-Phe-Pro-Arg-H) (6). DISCUSSION

Brain cathepsin dicarboxypeptidase

B acted exclusively as a toward several pro-

495

opioid peptides with 6-13 residues with free C-terminal carboxyl groups. The presence of cathepsin L was excluded on the basis of the equal hydrolysis of Z-Arg-Arg- and Z-Phe-Arg . NHMec, and the cleavage of the model hexapeptide at the Arg-Phe rather than the Met-Arg bond as reported by Kirschke and Barrett (1). Also the absence of activity toward Arg-NNap excluded the presence of cathepsin H. This may be of interest to activation of proopioids since C-terminal dipeptide cleavage was accompanied by the formation of shorter (active) forms, and in the specific case of Dyn A 1-8, a largely K-type agonist, into one having a preferential d-type action (see (10)). A consistent feature for the series investigated, pro-opioids and the synthetic hexapeptide, was the removal of the C-terminal dipeptide regardless of peptide length or composition unless otherwise noted (e.g., C-terminal Pro, D-amino acids, etc.). This contrasts with either C-terminal or endopeptidase action expressed toward larger substrates such as glucagon, oxidized insulin B chain, substance, P, /3-l;potropin, aldolase, and MBP (5,6,28). The present data for rat brain enzyme, together with that of other tissues using angiotensin-I, neurotensin, and the synthetic hexapeptide, show a preference for brain cathepsin B to act on the C-terminus of small (Mr) substrates (5, 29). The expression of either carboxypeptidase or endopeptidase cleavages is of cenTABLE KINETIC

VALUES

IV

FOR CATHEPSIN

OF ENKEPHALINS

AND

B HYDROLYSIS

SUBSTANCE

P

Substrate 1. 2. 3. 4. 5.

ME ME-Arg-Gly-Leu ME-Arg-Phe ME-Arg-Phe . NH2 Substance P

56.4 13,500 3.860 5,140 1,390

470 117 58 46 71

0.12 115 66.6 111 19.7

Note. Protein estimated by titration with E-64 and k.., values based on A& of 28,000 for brain cathepsin B. The product measured for substrate 1 was Tyr-Gly-Gly; for 2, ME-Arg; for 3, ME-Arg; for 4, ME-Arg; for 5, substance P l-9. K,,, values are the means of 2-3 determinations agreeing within 12%.

496

MARKS, TABLE

EFFECTS OF ADDED ENKEPHALIN-ARG-PHE BY PURIFIED

BERG,

V

PEPTIDES ON CLEAVAGE AT THE MET-ARG

OF METBOND

RAT

B

BRAIN

CATHEPSIN

IGO

Addition 1. Leupeptin (AC-Leu-Leu-Arg-H) 2. E-64 inhibitor 3. Brain cytosolic (cerebrocystatin) 4. Arg-Phe 5. MK-421 6. Phosphoramidon 7. Phe-Arg 8. hoc-D-Phe-Pro-Arg-H 9. Phe-Met

(W) 0.017 0.024 0.126 5 5 8.5 10.6 10.6 15.5

Note. Cerebrocystatin purified by the method of Kopitar et al (12). The ICso values are based on the quantity of inhibitor required to suppress substrate hydrolysis by 50%. Incubation mixture of 50 pl contained 5 nmol of peptide; 2 ng of cathepsin B, incubated for 20 min.

tral importance to the processing of many neuropeptides. It is known that most secretory products are transported in vesicular form along the rough endoplasmic reticulum, and that during transport they are processed by peptidases in a coordinated manner. Of particular significance to cathepsin B-like enzymes is the sorting or guiding of product by the Golgi apparatus towards a secondary catabolic pathway followed by fusion of vesicles with primary or secondary lysosomes (see (14, 1’7)). Although the C-terminal action of cathepsin B was described in 1978 by Aronson and Barrett (2) and Nakai et al. (3), insufficient attention has been paid to the role of cathepsin B as a processing enzyme capable of activating prohormones by C-terminal modification. There doesnot appear to have been any systematic studies conducted on dipeptidyl carboxypeptidase action similar to those reported here. The finding for a preferential action by brain cathepsin B toward the proenkephalins and prodynorphins was unsuspected and may be relevant to their turnover. These proopioid peptides are distributed widely in brain and in peripheral tissues and require shortening to

AND

BENUCK

release the active N-terminal (opioid) fragments (see (10, 11)). The alignment of the susceptible peptide domains of representative substrates used in this study with the putative catalytic center of the enzyme, numbered according to the method of Schechter and Berger (30) provide clues concerning the optimal juxtaposition of residues for hydrolysis (Fig. 4). In the series shown the vulnerable substrates contained a hydrophobic residue (Phe, Trp, or Leu) at the Pa/P3 positions. Interestingly, when a hydrophobic residue was absent at P2 it was to be found at P3 (except for substance P and Dyn l-7 where hydrophobic groups occupied both positions). These observations modify earlier conclusions of the essentiality of a hydrophobic residue at only the PZ position (see Ref. (38)). The marked decrease in rates of hydrolysis noted for the smallest substrates (ME or LE, the synthetic penta, or tetrapeptides) may be related to absence of residues at P3/P4 for interaction with putative S3/S4 subsites. Another conclusion that can be drawn from the alignments shown in Fig. 4 is that Arg when C-terminal was highly detrimental to rates observed for PI-P1 scission. This most likely can be attributed to the presence of the guanidium grouping rather than simply the basic amino acid at Ph, since Lys when C-terminal (Dyn A 1-13) constituted one of the better substrates (see Table I). It is of interest to processing of prohormones that C-terminal containing Lys at PZ, or Arg at P1 or Pi, were hydrolyzed comparatively well, e.g., Lys-X-Lys, Arg-Arg-X, Arg-XX, X-Arg-X. C-Termini such as these frequently occur in prohormonal fragments and, as a consequence they are potential candidates for cathepsin B. Moreover, a role for cathepsin B in prohormone processing is supported by the observation that secretory vesicles or granules of the anterior pituitary or adrenal medulla have an acid milieu favoring enzymes active at pH 6.0 (see (31)). The data also show that the Sr/S, subsites did not tolerate well Pro or D-amino acids; alternatively, the binding at these sites suppressed the subsequent enzymatic scission at the PI-P; bond. Finally, the absence of activity toward the

CATHEPSIN

B CONVERSION

OF

Sl’

S2 ’

Pl

Pl ’

P2’

Met

!I?5

GUY

LW

-Phe

-LeU

A=8

&is

Il.2

GUY

GUY

-Phe

Met

Aw

Phe

LW

Lx

Met

F&l

Phe

Ala

s4

S3

S2

Sl

P4

P3

P2

ME-Arg-Gly-Leu

GUY

-Phe

Dyn A l-s

‘=Y

ME-Arg-Phe Hexapeptide

1nl--lr---lr---lr-l Substrate

ME-Arg-Phe.NH2 Substance

P

FIG.

4. The

scission

putative

according for

substrates

I

GUY

-Phe

Met

!!=x

Phe . NH2

Gln

-Phe

-Phe

GUY

Ll?U

Het.NFl2

TYS

Gly

GUY

Phe

Met(NH2)

ME(NH2)

numbered

active

center

to the method with

free

for

cathepsin

of Schechter C-terminal

In the case of ME the point of cleavage Hydrophobic groups at Pz/Ps are underlined groups.

line

for

497

PRO-OPIOIDS

B with subsites (S) and and Berger (30). The arrow

carboxyl was

groups identical

and basic groups

or for

for

amidated

at P1/P;

those

residue with

and

positions

indicates

C-blocked

non-amidated

are indicated

the point

(P) of

amide forms.

by the dotted

emphasis

tripeptide Met-Arg-Phe may have resulted from the proximity of the free -NH2 to the s&sable bond. Many chromogenic substrates with blocked N-termini have enhanced rates of hydrolysis (1, 6), and it is probable that for N-blocked di- and tripeptides the (bulky) substituent (N-benzoyl, N-carbobenzoxy) interact with the relevant subsites proximal to the scissable bond. The kinetic values obtained for selected pro-opioid peptides indicate that they are useful alternative (native) substrates for assay purposes. Their Km values were actually higher than those reported for fluorogenic substrates (Z-Arg-Arg-, Z-PheArg-NHMec) but with lower keat values (1, 32). Nevertheless, the k,,,/K, ratios compare favorably with values of 66-115 for the hepta/octapeptides as compared to 160 PM-1 min-’ for the specific cathepsin B substrate Z-Arg-Arg-NHMec (1). In contrast, the high Km and the low k&K, ratios found for the pentapeptide enkephalins appear to rule out the possibility that brain cathepsin B plays any significant role for their inactivation in situ. Rather, the data strongly support the notion that cathepsin B participates as a “convertase” by releasing the shorter (active) forms. These stud-

ies on neuropeptides emphasize the need to measure kinetic constants of potential processing enzymes as a guide to understanding their functional importance. In many respects, the specificity of cathepsin B resembles that of peptidyl dipeptidase A (ACE). The latter is a peptidyl dipeptidase converting several pro-opioid intermediates, and like cathepsin B, it can act also as an endopeptidase degrading substance P (albeit at different sites) or LH-RH (33, 34). Recently we have shown that human brain ACE also converts Dyn A 1-12 or l-10 to form l-8 by sequential removal of dipeptides (35). The presence in brain membranes of two distinct enzymes acting C-terminally to remove dipeptides is of interest, since one requires -SH and is active at low pH, and the other is inhibited by -SH but acts optimally at pH 8.0 in the presence of Cl- (36). Presumably they act within different compartments and complement each other in processes linked to the activation/inactivation of neuropeptides. The effect of inhibitors indicates that most probably the catalytic center(s) mediating peptidyl dipeptidase and endopeptidase cleavages are similar. It is of interest that E-64 contains a positively charged

498

MARKS,

BERG,

guanidino group and leupeptin a C-terminal arginyl aldehyde, and that these features may be related in some manner to the low hydrolysis rates observed for peptides with a C-terminal arginine (37). Barrett et al. (38) reported a loss of potency for E-64 epoxide analogs when the C-terminal guanidino was replaced by other groups. The suppression of dipeptide release observed for cerebrocystatin suggests that brain tissue contains a factor for regulating the C-terminal actions of cathepsin B. Currently the mechanisms accounting for cysteine proteinase catalysis and consequently the effects of inhibitors are unknown. Liver cathepsin B has been sequenced and shown to contain an essential cysteine2’ and histidinelg7 (39). The essential cysteine residue of several different cysteine proteinases of plant or animal origin are present in peptide domains that show marked homologies, but His is present in regions that are variable. It is possible that the domain of Gly-Gly-His-Ala contribute to the unique specificity of cathepsin B as a peptidyl dipeptidase as contrasted to Val-Asn-His-Ala for cathepsin H, a cysteine proteinase having aminopeptidase actions (40). Endopeptidase. Consistent with the dual specificity pattern of cathepsin B, the purified brain enzyme acted on the internal bonds of MBP, a myelin component with 169 amino acids, and on smaller C-blocked peptides. The data on MBP is in agreement with earlier observations on its fragmentation by rat brain enzyme (12) with the additional information that one of the sites cleaved was Phe*‘-Phego. This is one of the sites recognized also by cathepsin D (2’7), but it was not established in this study if the residual peptides so formed retained their (toxic) encephalitogenic properties. The number of products formed with cathepsin B was larger than with cathepsin D incubated for comparable periods under the same conditions, illustrating the broader specificity of the cysteine proteinase towards this substrate. The Arg-Arg found in the low-molecular-weight products separated by gel-filtration probably was formed secondarily, since Arg when C-terminal, as shown for pro-opioids, can

AND

BENUCK

be expected to yield low rates of C-terminal hydrolysis. It is also unlikely that MBP can be degraded extensively by the C-terminal action of cathepsin B since Pro165, as was the case for Dyn A l-10 (Pro”), can be expected to prevent further scission. The influence of a C-terminal -COO- on the point of cleavage is shown by comparison of ME-Arg-Phe (cleavage at Met-Arg) to its amidated form (Arg-Phe * NH2). The alignment of residues with enzyme subsites indicate interaction between S; and Phe . NH2 rather than with Arg in the nonamidated form (Fig. 4). Some role, therefore, is suggested for Arg in redirecting the point of cleavage, since substrates lacking Arg at or near the C-terminus such as ME * NHz, LE sNH2 and substance P were cleaved at their penultimate bonds to release dipeptide amides (Fig. 4). The kinetic data for ME-Arg-Phe * NH2 show that this is a suitable substrate to assay cathepsin B for “endopeptidase” actions with k,,,/K, values higher than the non-amidated form. The hydrolysis of peptide amides by a lysosomal enzyme may be related to inactivation and turnover of a number of neuropeptidyl amides such as CCK-8, LH-RH, and substance P. Present data, however, showed rather low rates of hydrolysis for CCK-8, LH-RH but higher ones for substance P. In the case of substance P, it is known that removal of the C-terminus containing the -NH2 is accompanied by loss of biological activity (13). ACKNOWLEDGMENTS Supported in part by Grants NS-12578 and the Kroc Foundation. We are indebted to George A. Hashim for providing MBP, and to Miriam Banay-Schwartz for assistance with the gel electrophoresis separation of MBP products. REFERENCES 1. BARRETT, A. J., AND KIRSCHKE, H. (1981) in Methods in Enzymology (Lorand, L., ed.), Vol. 80, pp. 535-561, Academic Press, New York. 2. ARONSON, N. N., AND BARRETT, A. J. (19’78) Biochem. J. 171,759-765. 3. NAKAI, N., WADA, K., KOBASHI, K., AND HASE, J. (1978) Bkxh~m. Biophgs. Res. Commun 83,881885. 4. MCKAY, M. J., OFFERMANN, M. K., BARRETT, A. J.,

CATHEPSIN

B CONVERSION

AND BOND, J. S. (1983) Biochem J. 213,467-471. 5. KATUNUMA, N., AND KOMINANI, E. (1983) Current Topics in Cellular Regulation, Vol. 22, pp. 71101, Academic Press, New York. 6. SUHAR, A., AND MARKS, N. (1979) Eur. .I B&hem. 101,23-30. 7. SUHAR, A., MARKS, N., TURK, V., AND BENUCK, M. (1981) in Proteins and Their Inhibitors (Turk, V., and Vitale, E. I., eds.), pp. 33-41, Pergamon, New York. 8. MARKS, N., SUHAR, A., AND BENUCK, M. (1981) Adv. Biochem. Psychopharmacol. 28,49-60. 9. MARKS, N., KOPITAR, M., STERN, F., AND BERG, A. J. (1985) in Intracellular Protein Catabolism (Khairallah, E. A., Bond, J. S., and Bird, J. W. C., eds.), Prog. Clin. Biol. Res., Vol. 180, pp. 247-249, Alan Liss Inc., New York. 10. GOLDSTEIN, A. (1984) in The Peptides (Meienhofer, J., and Udenfriend, S., eds.), Vol. 6, pp. 95-145, Academic Press, New York. 11. UDENFRIEND, S., AND KILPATRICK, D. L. (1984) in The Peptides (Meienhofer, J., and Udenfriend, S., eds.), Vol. 6, pp. 25-68, Academic Press, New York. 12. KOPITAR, M., STERN, F., AND MARKS, N. (1983) B&hem. Biophys. Res. Commun. 112, lOOO1006.

13. MARKS, N. (1977) in Peptides in Neurobiology (Gainer, H., ed.), pp. 221-258, Plenum, New York. 14. HABENER, J. F., ROSENBLATT, A., AND POTTS, J. T. (1984) PhysioL Rev. 64,985-1053. 15. DOCHERTY, K., CARROLL, R., AND STEINER, D. F. (1983) Proc. NatL Acad. Sci. USA 80,3245-3248. 16. QIJINN, P. S., AND JUDAH, J. D. (1978) Biochem. .I 172,301-309.

17. BIENKOWSKI, R. S. (1983) Biochem. J. 214, l-10. 18. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL, R. J. (1951) J. BioL Chem. 193, 265-275. 19. BARRETT, A. J. (1980) Biochem. J. 187,909-912. 20. LEWIS, R. V. (1979) Anal. B&hem. 98,142-145. 21. FERNSTROM, M. H., AND FERNSTROM, J. D. (1981) Life Sci. 29,2119-2130. 22. WOOD, K. R., AND WANG, K.-T. (1967) B&him. Biophys. Acta 133, 369-370.

OF

PRO-OPIOIDS

499

23. AMBLER, R. P. (1972) in Methods in Enzymology (Hire,, C. H. W., and Timasheff, S. N., eds.), Vol. 25, Part B, pp. 1433154, Academic Press, New York. 24. LAEMMLI, II. K. (1970) Nature &m&m) 227, 680685. 25. WRAY, W., BONLIKAS, T., WRAY, V. P., AND HANCOCK, R. (1981) Anal. Biochem. 118,197-203. 26. YANG, H.-Y. T., FRATTA, W., MAJANE, E. A., AND COSTA, E. (1985) Proc. Natl. Acad. Sci. USA 82, 7757-7761. 27. BENUCK, M., MARKS, N., AND HASHIM, G. A. (1975) Eur. J. Biochem. 52,615-621. 28. BOND, J. S., AND BARRETT, A. J. (1980) B&hem. J. 189,17-25.

29. AZARYAN, A., BARKHUDARYAN, N., GALOYAN, A., AND LAJTHA, A. (1985) Neurochem. Res. 10, 1525-1532. 30. BERGER, A., AND SCHECHTER, I. (1970) Philos. Trans. R. Sot. London Ser. B BioL Sci. 257,249264. 31. FRICKER, L. D. (1985) Trends Neurosci. 83, 210214.

32. KIRSCHKE, H., KEMBHANI, A. A., BOHLEY, P., AND BARRETT, A. J. (1982) Biochem. J. 201,367-372. 33. CASCIERI, M. A., BULL, H. G., MUMFORD, R. A., PATCHETT, A. A., THOMBERRY, N. A., AND LIANG, T. (1984) Mol. PharmacoL 25,287-293. 34. SKIDGEL, R. A., ENGELBRECHT, S., JOHNSON, A. R., AND ERDOS, E. G. (1984) Peptides 5,760-776. 35. MARKS, N., BENUCK, M., Lo, E.-S., NOVACENKO, H., SEYFRIED, C., AND WOLF, H. P. (1985) Alcohol Drug Res. 6, 123-124. 36. BENUCK, M., BERG, M. J., AND MARKS, N. (1984) Neurochem. Res. 9,733-749. 37. HANADA, K., TAWAI, M., YAMAGISHI, M., OHMURA, S., SAWADA, J., AND TANAKA, I. (1978) Agr. BioL Chem. 42,523-528. 38. BARRETT, A. J., KEMBHANI, A. A., BROWN, M. A., KIRSCHKE, H. K., KNIGHT, G. G., TAMAI, M., AND HANADA, K. (1982) Biochem. .I 201,189-198. 39. TAKIO, K., TOWATARI, T., KATUNUMA, N., TELLER, D. C., AND TITANI, K. (1983) Proc. NatL Acad. Sci. USA 80,3666-3670. 40. WILLENBROCK, F., AND BROCKLEHARST, K. (1985) Biochem. J. 227,521-528.