Calcium Depletion Blocks Proteolytic Cleavages of Plasma Protein

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 24, Issue of August 25, pp. 17465-17471,1992 Printed in U. S . A

Calcium Depletion Blocks Proteolytic Cleavages of Plasma Protein Precursors Which Occur at the Golgi and/or Trans-Golgi Network POSSIBLE INVOLVEMENT OF Ca2+-DEPENDENT GOLGI ENDOPROTEASES*

(Received for publication, February 5,1992)

Kimimitsu OdaS From the Department of Biochemistry, Fukuoko University School of Medicine, Nanakuma, Jonan-ku, Fukuoka 814-01, Japan

The effects of calcium depletion on the proteolytic cleavage and secretion of plasma protein precursors were investigated in primary cultured rat hepatocytes and HepG2 cells. When the cells were incubated with A23187, the calcium-specific ionophore, in a medium lacking CaCl,, precursors of serum albumin and the third and fourth components of complement, C3 and C4, respectively, were found to be released into the medium. The addition of ionomycin or EGTA to the medium inhibited the processing of pro-C3 as well. Blocking the secretory pathway either at the mixed endoplasmic reticulum/Golgi in the presence of brefeldin A or at the endoplasmic reticulum/tubular-vesicular structure at a reduced temperature caused accu- mulation of pro-C3 within hepatocytes or HepG2 cells, indicating that the cleavage of the precursor occurs at a later stage of the secretory pathway. Once the blockade was released by incubating the cells either in the brefeldin A-free medium or at 37 O C , the secretion of plasma proteins resumed, irrespective of the presence of A23187. However, the processing of pro-C3 was almost completely inhibited in the presence of A23187, with only the precursor being released into the medium, implying that a decline in Ca” levels within the cell modulates the activity of a Golgi endoprotease responsible for the cleavage of pro-C3. When incubated with isolated Golgi membranes, pro-C3 secreted from Caz+-depleted cells was cleaved in vitro into their subunits in the presence of Ca2+ but not in its absence, pointing to the involvement of a Ca2+-dependent Golgi endoprotease in the processing of pro-C3. These results collectively suggest that calcium depletion blocks the proteolytic cleavages of plasma protein precursors presumably by exhausting a Ca2+ pool available to the Ca2+-dependent processing enzyme(s) located at the Golgi and/or trans-Golgi network.

Many peptide hormones, neuropeptides, and plasma proteins are synthesized as larger precursors and undergo proteolytic processing to become mature proteins during passage through the exocytotic pathway (1). Since these precursors have paired dibasic amino acid residues at cleavage sites, it

* This work was supported in part by a grant-in-aid for scientific research from the Ministry of Education, Science, and Culture of Japan and by a fund from the Central Research Institute of Fukuoka University. 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.

$ To whom correspondence should be sent: Dept. of Biochemistry, Fukuoka University School of Medicine, 45-1, 7-chome Nanakuma, Jonan-ku, Fukuoka 814-01, Japan. Tel.: 81-092-801-1011 (ext. 3251); Fax: 81-092-865-6032.

has been postulated that a common or a family of endoproteases may be involved in the proteolytic cleavage of the precursors. Until recently the endoproteases responsible for processing precursors have remained elusive; however, there is increasing evidence that furin and related proteins (PC2, PC3) may be involved in the maturation of the precursors in the constitutive (2-5) and regulated pathway (6-9). Recently Hosaka et al. (10) proposed that the Arg-X-Lys/Arg-Arg motif is a signal for cleavage of precursors which undergo proteolytic processing along the constitutive pathway. Precursors of influenza virus hemaggulutinin (HA),’ cytomegalovirus glycoprotein B, serum albumin, complement C3, and complement C4 fall into this category.

A23187, a calcium-specific ionophore, has been used as a tool to study the physiological roles of Ca2+ in relation to the intracellular transport of macromolecules in cultured cells (11-14). Klenk et al. (14) reported that A23187 inhibits the proteolytic processing of the HA precursor. Interstingly, they showed that the intracellular transport of HA to the cell surface and its assembly into virus particle occurred under conditions in which conversion of the precursor HA to HA1 and HA2 was completely blocked, suggesting that the inhibition of processing is not caused by the arrest of the HA precursor before the site of cleavage but as a direct effect of the ionophore on the processing machinery, presumably on a putative processing endoprotease. A similar finding was reported in the processing of glycoprotein B precursor of cytomegalovirus (15). I reasoned that if proteolytic cleavages of both the viral glycoprotein precursors and plasma protein precursors were catalyzed by a common endoprotease (furin) in the constitutive pathway as proposed by Hosaka et al. (IO), A23187 would block the cleavage of plasma protein precursors in a manner similar to the viral glycoproteins.

In the present study I examined the effects of A23187 on the secretion and proteolytic cleavage of plasma protein precursors in rat hepatocytes and HepGP cells. An attempt was also made to explore a putative convertase capable of cleaving pro-C3 to its subunits in isolated Golgi membranes.

EXPERIMENTAL PROCEDURES

Material~-[~~S]Methionine (> 800 Ci/mmol) was purchased from Du Pont-New England Nuclear. Ionomycin and Pansorbin were obtained from Calbiochem; Affi-Gel-10 from Bio-Rad; aprotinin from Sigma; n-octyl-P-D-glucopyranoside from Dojin Co. (Kumamoto, Ja-

‘The abbreviations used are: HA, hemagglutinin; C3, the third component of complement; C4, the fouth component of complement; E-64, (~-3-trans-carboxyoxiran-2-carbonyl)-~-leu-agmatin; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; ER, endoplasmic reticulum; MEM, minimum essential medium; MEM - Ca2+, minimum essential medium lacking CaC1,; PAGE, polyacrylamide gel electrophoresis; TGN, trans-Golgi network; SDS, sodium dodecyl sulfate; Mes, 4-morpholineethanesulfonic acid Hepes, 4-(2-hydroxy- ethyl)-1-piperazineethanesulfonic acid.

17465

17466 Block of Proteolytic Cleavage by Calcium Depletion pan); protein A-Sepharose CL-4B from Pharmacia LKB Biotechnol- ogy Inc.; leupeptin, E-64, chymostatin, elastatinal, pepstatin A, antipain, and phosphoramidone from Peptide Institute, Inc. (Osaka, Japan). Goat antisera against rat complement component 3 (C3), human C3, and human complement component 4 (C4) were obtained from Cappel Laboratories. Antiserum against rat serum albumin was prepared as described previously (16). Antiserum against the syn- thetic propeptide (NH,-R-G-V-F-R-R-C) of albumin was prepared as described previously (17), and monospecific IgG was purified through a propeptide-coupled Affi-Gel-10 column. A23187 was a gift from Lilly Laboratories. Brefeldin A was obtained from Dr. A. Takatsuki (The Institute of Physical and Chemical Research). HepG2 cells were obtained from Dr. D. Kang (this Institute).

Cell Culture-Rat hepatocytes were prepared according to the method of Seglen (18) and inoculated in 60-mm collagen-coated plastic dishes in Eagle's minimum essential medium (MEM) supplemented with newborn calf serum (5%), insulin (lo" M), and dexa- methasone M ) as described previously (19). HepG2 cells were maintained in MEM supplemented with 5% fetal calf serum and kanamycin (60 pglml) as described previously (20).

Cell Labeling-Labeling of cells was performed essentially as described previously (19, 21). Rat hepatocytes and HepG2 cells were preincubated in methionine-free MEM for 1 h and labeled with 50- 100 pCi of [35S]methionine in this medium or methionine-free MEM lacking calcium chloride (MEM - Ca") with or without A23187, ionomycin, or EGTA. When pulse-chase experiments were performed, labeled cells were chased in the complete MEM or MEM lacking calcium chloride with or without A23187. The ionophores were dissolved in dimethyl sulfoxide (250 p~ stock), and the final concentration of dimethyl sulfoxide in the medium was 0.1% (v/v). Drugs were added to the medium at the start of the labeling and were present throughout the experiments. A t the end of labeling periods media were removed and mixed with a protease inhibitor mixture (antipain, aprotinin, chymostatin, elastatinal, leupeptin, pepstatin A, and phosphoramidone) a t a final concentration of 10 pg/ml each. Cells were dissolved in 0.5 ml of lysis buffer (10 mM Tris/HCl (pH 7.5) containing 0.5% Triton X-100,0.5% sodium deoxycholate, 150 mM NaCl, 2 mM EDTA, and the protease inhibitor mixture). Both cell lysates and media were centrifuged at 15,000 X g for 10 min, and the resultant supernatants and media were stored at -20 "C until use.

Immunoprecipitation-Immunoprecipitation of plasma proteins was performed according to Williams et al. (22) and Degen and Williams (23) with a slight modification. All procedures were performed at 4 "C. The cell lysates and media were shaken with 50 pl of fixed Staphylococcus aureus cells (Pansorbin) for 30 min and centrifuged. Precleared lysates and media were adjusted to 0.5% (w/v) skim milk powder and incubated with 3-5 pl of antiserum or 50 pg of IgG for 3 h in a shaker. Immune complexes were isolated by shaking for 1 h with 30-50 pl of a 50% suspension of protein A-Sepharose (Pharmacia) in phosphate-buffered saline. The agarose beads were washed three times with Buffer A (10 mM Tris/HCl (pH 7.5) containing 0.5% Triton X-100,0.5% sodium deoxycholate, 0.05% sodium dodecyl sulfate (SDS), and 150 mM NaCl), once with NSTE buffer (10 mM Tris/HCl (pH 7.5) containing 0.5% Nonidet P-40, 150 mM NaCl, and 2 mM EDTA), three times with NSTE buffer containing 500 mM NaCl instead of 150 mM NaCl, and twice with NSTE. The immunoisolated molecules were dissolved in Laemmli's sample buffer (24) and analyzed by SDS-polyacrylamide gel electrophoresis (PAGE, 7.5% gel) or slab gel isoelectric focusing (25), followed by fluorography (19).

Isolation of Golgi Membranes-Golgi-rich fractions were prepared from rat livers as described previously (26,27). The GF2 fraction was suspended in 2 mM Tris/HCl (pH 7.5) containing 50 mM NaCl and subjected to repeated freezing-thawing (five times) in dry-ice/acetone. Then octyl glucoside was added to the suspension at a final concentration of 0.05%. After being incubated for 30 min on ice, Golgi membranes were recovered by centrifuging the suspension at 105,000 X g., for 1 h and were resuspended in 0.25 M sucrose. The Golgi membranes were divided into small portions, quickly frozen in liquid nitrogen, and stored at -80 "C until use. Protein was determined using the method of Lowry et al. (28) with bovine serum albumin as standard.

Proteolytic Processing of Pro-c3 in Vitro-For preparation of 35S- plasma protein precursors, rat hepatocytes were labeled with 1 mCi of [35S]methionine for 5 h in methionine-free MEM - Ca2+ with 0.25 p~ A23187. After the medium was centrifuged to remove cells, the resultant supernatant was used as substrate. A complete reaction mixture consisted of 20 mM acetate buffer (pH 5.5), 10 pl of the substrate, 50 pg of Golgi membranes, 0.25% octyl glucoside, 200 p~

CaC12 in a total of 50 pl. 10 p1 of substrate contained 1,500-3,000 cpm of "S-pro-C3. The reaction was started by the addition of the Golgi membranes and was continued at 37 "C for 2 h. The reaction was stopped by the addition of 500 pl of NSTE buffer and the protease inhibitor mixture. C3 was immunoprecipitated, washed, and analyzed by SDS-PAGE as described above. When the processing of C3 was examined at various concentrations of Ca", incubations were performed using EGTA (1 mM)/Ca*+ buffer at pH 5.5. The concentration of total Ca2+ was adjusted according to the calculations described Ogawa and Kitazawa (29) to yield the final free [Ca'+] indicated.

RESULTS

Calcium-specific Ionophore Blocks Proteolytic Processings- Fig. 1 shows the effect of 1423187, a calcium-specific ionophore, on the secretion and proteolytic processing of rat C3. C3 is a heterodimeric protein consisting of a- (115 kDa) and p- (65 kDa) subunits linked together by disulfide bonds. C3 is synthesized as a single polypeptide chain precursor (pro-C3, 180 kDa), which is cleaved intracellularly at a single cleavage site with an NH,-(P)-R-R-R-R-(a)-COOH sequence (21, 30). After hepatocytes were lableled with ["S]methionine, C3 was immunoisolated from cell lysates and media and analyzed by SDS-PAGE. When A23187 was added to the medium, MEM - Ca2+ was used instead of MEM (the control culture). Generally there was a progressive decrease in the secretion of C3 as the concentration of the ionophore increased (Fig. LA), suggesting that the intracellular transport of plasma proteins is retarded in the presence of A23187. Recently Lodish and Kong (31) reported that A23187 inhibits the exit of secretory proteins from the ER (31). In the control culture most of the labeled C3 was in the pro-C3 within the cell (Fig. lB), whereas

" DO,,?., 0 125 0 :'i 0 b 0 5 2r

A23187 ( pM ) +1.8 mM Ca

0 0.0625 0.125 0.25 0.5 0 . 5 t C f t

A C M c M C M C M C M C M

I -*re-c3

- a

"B

1 . L.

FIG. 1. Effects of A23187 on the secretion and processing of C3 in rat hepatocytes. Rat hepatocytes were labeled with 50 pCi of [3'SS]methionine for 3 h in MEM (control culture) or in MEM - Ca2+ in the presence of 0.0625,0.125,0.25, or 0.5 p M A23187. Dimethyl sulfoxide was added to the control culture (0.1%). CaC12 was added at the start of the labeling (1.8 mM). C3 molecules were immunoisolated from cell lysates (C) and media (M). In A, the synthesis of C3 a t different concentrations of A23187 was determined by measuring ["Slmethionine incorporated into the total C3 (cellular and secreted C3) and expressed as percentages of that of the control cells (closed bars). The secretion of C3 was also expressed as percentages of the secreted C3 to the total C3 synthesized under each condition (hatched bars). In B, C3 molecules were analyzed by SDS-PAGE. a and p denote the a- and &subunits of mature C3, respectively.

Block of Proteolytic Cleavage by Calcium Depletion 17467

no pro-form was found in the medium, indicating that the proteolytic cleavage of pro-C3 proceeds very efficiently within a rat hepatocyte, with only mature C3 being secreted from the cell (Fig. 1B). However, as the concentration of A23187 increased, the pro-form progressively appeared in the medium, and finally only pro-C3 was secreted at an ionophore concentration above 0.25 PM. Since the hepatocytes were incubated under conditions that should render the cell depleted of Ca2+, blocking the proteolytic cleavage of pro-C3 by A23187 appears to result from a decrease in Ca2+ levels within the cell. Consistent with this idea, the addition of CaC12 into the medium restored the proteolytic activity in the cell (Fig. 1B). Since substantial amounts of plasma proteins were still secreted into the medium at a concentration of 0.25 PM (Fig. JA), the experiments described below were performed at this concentration of A23187 unless otherwise specified.

If the effect of A23187 on the proteolytic cleavage of pro- C3 is indeed caused by a decrease in intracellular Ca2+ levels, it is expected that ionomycin, a calcium ionophore, would also block the processing of pro-C3. Fig. 2 shows that this is the case. When hepatocytes were labeled in MEM - Ca2+ with ionomycin, pro-C3 was secreted into the medium. In contrast, only the mature C3 with the a- and &subunits was found in the medium supplemented with CaCL

To test whether A23187 affects proteolytic processing of other plasma protein precursors, the processing of rat proalbumin was examined in the absence or presence of A23187 (Fig. 3). Serum albumin is synthesized as a precursor (proalbumin) with an extra hexapeptide extension (NH2-R-G-V-F- R-R-) at the NH2 terminus of serum albumin (32). Proalbu- min undergoes proteolytic processing to become serum albumin on its way to a final discharge into the medium (21,33). To analyze the processing of proalbumin, immunoprecipitates were subjected to isoelectric gel focusing. This method takes advantage of the fact that proalbumin has 3 additional argi- nine residues compared with serum albumin and hence has a higher PI value than that of serum albumin (34). In the control culture proalbumin was a major form found within the cell (Fig. 3A, lane 1 ), whereas only serum albumin was secreted into the medium (Fig. 3A, lane 2), indicating that proalbumin is completely converted to serum albumin prior to secretion. However, in the presence of A23187 no serum albumin was detectable, and instead proalbumin was found to be secreted into the medium (Fig. 3A, lane 6), showing that the proteolytic

0 0.0625 0.125 0.25 0~25+&?+

C M C M C M C M C M

1

"p

FIG. 2. Effect of ionomycin on the processing of pro43 in rat hepatocytes. Rat hepatocytes were labeled with 50 pCi of [35S] methionine for 3 h in MEM (control culture) or in MEM - Caz+ in the presence of 0.0625, 0.125, and 0.25 p~ ionomycin. Dimethyl sulfoxide was added to the control culture (0.1%). CaC12 was added at the start of the labeling (1.8 mM). C3 molecules were immunoisolated from cell lysates (C) and media (M) and subjected to SDS- PAGE. CY and @ denote the CY- and @-subunits of mature C3, respectively.

A 1 2 3 4 5 6 -

- PA

- SA

B Antl-RSA Antl-propeptlde

"" 1 2 3 4 5 6 7 8

-"-4

- PA

- S A

+ FIG. 3. Effect of A23187 on the processing of proalbumin

in rat hepatocytes. A, rat hepatocytes were labeled with 50 pCi of [3sS]methionine for 2 h in MEM (lanes 1 and 2) , MEM - Ca2+ (lanes 3 and 4 ) , or MEM - Ca2+ in the presence of 0.25 p~ A23187 (lanes 5 and 6). Rat albumin was immunoisolated from cell lysates (odd numbers) and media (euen numbers) with antiserum against rat serum albumin. Albumin molecules were analyzed by isoelectric gel focusing (pH 5-8). PA and SA denote proalbumin and serum albumin, respectively. E , rat hepatocytes were labeled with 50 pCi of ["S]methionine for 2.5 h in MEM - Ca2+ in the presence of 0.25 p~ A23187. CaCh was added at the start of labeling at a final concentration of 1.8 mM (lanes 3, 4, 7, and 8). Each cell lysate (odd numbers) and medium (euen numbers) was divided into two parts, and then each sample was subjected to immunoisolation either with antiserum against rat serum albumin (RSA) or with anti-propeptide IgG. Albumin molecules were analyzed by isoelectric gel focusing (pH 5-8).

cleavage of proalbumin to serum albumin was completely blocked by A23187. Immunoprecipitation using anti-propeptide IgG, which specifically recognizes proalbumin (17, 35), confirmed that proalbumin per se was secreted into the medium in the presence of A23187 (Fig. 3B, lanes 2 and 6). Again, the addition of CaC12 to the medium restored the processing activity in the cell (Fig. 3B, lanes 4 and 8). These results indicate that the proteolytic cleavage of proalbumin is blocked by a decrease in Ca2+ levels within the cell, as in the case of pro-C3.

Fig. 4 shows the effects of A23187 on the processing of human pro-C3 and pro-C4 in HepG2 cells. C4 is a hetrotri- meric protein consisting of a- (95 kDa), 0- (75 kDa), and y- (31 kDa) chains linked together by disulfide bonds and synthesized as a single polypeptide chain precursor (pro-C4,200 kDa) (36, 37). Pro-C4 is cleaved into three subunits at two cleavage sites with NH2-(P)-R-K-K-R-(a) or (a)-R-R-R-R- (7)-COOH sequences. Proteolytic processing of plasma protein precursors in the hepatoma cell was not as efficient as that in the hepatocyte (Fig. l), and significant amounts of the precursors were secreted with mature proteins (Fig. 4, lanes 2 and 8). Besides uncleaved precursor, intermediate p-a- and a-y-chains were released from HepGP cells in the case of C4 (Fig. 4, lane 8). Only the precursors were found to be secreted into the medium in the presence of A23187 (Fig. 4, lanes 6 and 12). The addition of CaClz restored processing in the cell (data not shown). It is noteworthy that calcium depletion hhibited two proteolytic events required to convert pro-C4 to mature C4. Taken together, these data imply that the block of proteolytic cleavage of the plasma protein precursors with an R-X-K/R-R motif by calcium depletion is a general phe- nomenon of hepatocytes or hepatoma cells.

EGTA Blocks Proteolytic Processing-When the hepatocytes were preincubated for 2 h in MEM-Ca2+ in the absence of A23187, pro-C3 was found to be released into the medium (Fig. 5, lane 4 ) . This finding suggests that prolonged incuba-

17468 Block of Proteolytic Cleavage by Calcium Depletion

c 3 c4 I 2 3 4 5 6 7 8 91011 12 I

pro.c3- a-I -P'O.C4 -

b-a - a-y

"a

P- --b

"I

FIG. 4. Effects of A23187 on the processing of pro-C3 and pro-C4 in HepG2 cells. HepG2 cells were labeled with ["'Slmethi- onine for 4 h in MEM (lanes 1, 2, 7, and 8) , MEM - Ca'+ (lanes 3, 4, 9, and IO) or MEM - Ca2+ in the presence of 0.25 p~ A23187 (lanes 5, 6, 11, and 12). Each cell lysate (odd numbers) and medium (euen numbers) was divided into two parts. Then each sample was subjected to immunoisolation either with antiserum against C3 or with antiserum against C4. C3 and C4 molecules were analyzed by SDS-PAGE. Left side, a and denote the a- and 8-subunits of mature C3, respectively; right side, 8-a and a-y denote two-chain C4 molecules with uncleaved &a- and a,y-subunits, respectively. a, p, and y denote a-, 8-, and y-subunits of mature C4, respectively. Bands on a fluorograph were assigned to each C4-derived processing intermediate and subunit according to Chan and Atkinson (37).

*\-a

1 2 3 4 5 6 7 8 910l112 - Fm.0

i ! "B

~

~ I

FIG. 5. Effect of EGTA on the processing of pro-C3 in rat hepatocytes. Rat hepatocytes were preincubated in MEM (lanes 1 and 2 ) or MEM - Ca" (lanes 3-12) for 2 h and labeled with ["SI methionine for 3 h in MEM (lanes 1 and 2), MEM - Ca2+ (lanes 3 and 4 ) , or MEM - Ca2+ in the presence of 50 p~ EGTA (lanes 5 and 61, 100 p~ EGTA (lanes 7 and 8) , 250 p~ EGTA (lanes 9 and lo), and 500 p~ EGTA (lanes 11 and 12). EGTA was added at the start of labeling. C3 molecules were immunoisolated from cell lysates (odd numbers) and media (even numbers) and subjected to SDS-PAGE. a and 8 denote the a- and &subunits of mature C3, respectively.

tion of hepatocytes in the medium lacking Ca2+ leads to a decrease in Ca2+ levels within the cell, which is reflected in the efficiency with which the processing of pro-C3 proceeds. The addition of EGTA instead of the ionophores blocked the processing to such an extent that almost no mature C3 was released into the medium (Fig. 5, lanes 5-12). Thus, EGTA mimics A23187 in relation to the inhibition of the proteolytic processing of pro-C3. These results support the idea that the block in processing plasma protein precursors by A23187 or ionomycin described above (Figs. 1-4) may indeed be ascribed to calcium depletion facilitated by the ionophores and not to side effects of the ionophores themselves.

Calcium Depletion Blocks Proteolytic Cleavage Occurring at the Golgi-In previous studies I reported that weakly basic amines such as chloroquine, methylamine, and Tris caused the distension of the Golgi and concomitantly inhibited proteolytic cleavage of proalbumin and pro-C3 in the rat hepatocyte, resulting in the release of the precursors into the medium (21, 38, 39). These findings suggest that processing the plasma protein precursors occurs late in transport, presumably in the amine-sensitive acidic Golgi compartments (trans-Golgi cisterna and/or trans-Golgi network). If so, calcium depletion should affect the Golgi/TGN along the secre-

tory pathway such that the proteolytic processing of plasma protein precursors no longer occurs, and as a result, only the precursors are secreted into the medium. To address this question, the following experiments were carried out. First, hepatocytes which had been pulse-labeled with [35S]methionine for 10 min were further incubated at 15 "C for 1 h. Incubating the cells at a low temperature is known to arrest the intracellular transport of newly synthesized proteins at the ER/tubular vesicular structures (40-43). Then the cells were chased at 37 "C in MEM or MEM - Ca2+ with or without A23187, thus allowing one to examine the effect of calcium depletion on a later stage in transport subsequent to egress from the ER/tubular vesicular structure. As shown in Fig. 6, only pro-C3 was detected within the cell after a 1-h incubation at 15 "C (Fig. 6, lane 1 ). Once the blockade was lifted by incubating the cells at 37 "C, the precursor became processed and was secreted as mature C3 with a- and @-subunits into the MEM (Fig. 6, lane 4 ) , whereas the precursor was secreted in the presence of A23187 (Fig. 6, lane 8), indicating that Ca2+ is needed for effective processing of pro-C3 after the precursor left the ER. Second, HepG2 cells were labeled with [%I methionine in the presence of brefeldin A. Brefeldin A is known to dissociate the Golgi complex presumably by inter- fering with an anterograde pathway, eventually leading to the fusion of the Golgi components with the ER and nuclear envelope (44-48). When the brefeldin A-treated cells are incubated in fresh medium lacking brefeldin A, the Golgi stack is reported to reassemble rapidly, and transport to the cell surface resumes (46,48). In this experiment HepG2 cells were used instead of rat hepatocytes in which blockade by brefledin A of secretion of plasma proteins is temporary because of the short duration time of the effect of the drug (20, 44). After a 1-h incubation pro-C3 was the only form found within the cell, and no mature form was detected (Fig. 7, lane 1 ). Recovery from the intoxication by brefeldin A was studied by incubating the brefeldin A-treated cells either in MEM or in MEM - Ca2+ with or without A23187. Mature C3 was secreted in the absence of the ionophore (Fig. 7, lane 4 ) to a degree that was comparable to that of the untreated HepG2 cell (Fig. 4, lane 2 ) ) suggesting that the processing event of the presursor is closely related to the appearance of the Golgi stack. In contrast, only pro-C3 was released into the medium in the presence of A23187 (Fig. 7, lane 8), suggesting that reassembly of the Golgi stack and the resultant reformation of secretory pathway within the cell are not sufficient for the processing of pro-C3 unless Ca2+ is afforded. These data are in agreement with the notion that the processing of pro-C3 occurs in the Golgi/TGN along the secretory

1 2 3 4 5 6 7 8

FIG. 6. Processing and secretion of pro-C3 in rat hepatocytes at reduced temperature. Each rat hepatocyte culture in quadruplicate was pulse labeled with 100 pCi of [%]methionine for 10 min and further incubated for 1 h at 15 "C. Cell lysate and medium were prepared from one set of the culture (lanes 1 and 2). Other sets of the culture were further incubated at 37 "C for 3 h in MEM (lanes 3 and 4 ) , in MEM - Ca2+ (lanes 5 and 6) , or in MEM - Ca2+ in the presence of 0.25 p~ A23187 (lanes 7 and 8). c 3 molecules were immunoisolated from cell lysates (odd numbers) or media (euen numbers) and subjected to SDS-PAGE. a and 8 denote the a- and 8- subunits of mature C3, respectively.

Block of Proteolytic Cleavage by Calcium Depletion I 2 3 4 5 6 7 8 A 1 2 3 4 5 6 7 8 9 1011

FIG. 7. Processing and secretion of pro-C3 in the brefeldin A-treated HepG2 cells. Each HepG2 cell culture in quadruplicate was pulse labeled with 50 pCi of ["'S]methionine for 1 h in the presence of brefledin A (2.5 pglml). Cell lysate and medium were prepared from one set of the culture (lanes 1 and 2). Other sets of the culture were further incubated for 4 h in MEM (lanes 3 and 4 ) , MEM - Ca2+ (lanes 5 and 6), or MEM - Caz+ in the presence of 0.25 p M A23187 (lanes 7 and 8). C3 molecules were immunoisolated from cell lysates (odd numbers) or media (euen numbers) and subjected to SDS-PAGE. a and B denote the a- and @-subunits of mature C3, respectively.

pathway (49, 50). Moreover, the data suggest that Ca2+ is necessary for processing pro-C3 in the Golgi/TGN. However, one might argue that reassembly of the Golgi stack did not occur at all in the cells depleted of Ca2+ and that plasma protein precursors were released directly into the medium. When cells that had been incubated with 6-((N-(7-nitrobenz- 2-oxa-1,3-diazol-4-yI)amino)-caproyl)sphingosine-ceramide to specifically stain the Golgi stacks (51) were exposed to brefeldin A and then incubated in brefeldin A-free medium and examined under a fluorescence microscope, the Golgi stacks were found to reassemble even in the presence of A23187 (data not shown).

Processing Pro-C3 in Vitro-The finding that plasma proteins are secreted exclusively as precursor forms into the medium from hepatocytes depleted of Ca2+ provides a conven- ient method to obtain plasma protein precursors, which can be used as substrates for assaying the Golgi enzymes responsible for processing plasma protein precursors. In the present study a mixture of [35S]methionine-labeled precursors secreted from Ca2+-depleted hepatocytes was added to a reaction mixture as a substrate and incubated with Golgi membranes in the presence of detergent followed by immunoprecipitation with anti-C3 serum. As shown in Fig. 8A, "S-pro-C3 was found to be cleaved into three bands ( l a n e 2), two of which were indistinguishable from the a- and &subunits of the mature C3 on SDS-polyacrylamide gel (lanes 2 and 11 ). To study the relationship between these pro43-derived molecular species a time course study of in vitro cleavage of "S-pro- C3 was carried out. Fig. 8B shows that pro-C3 was directly cleaved into the a- and P-subunits, and another band appeared only after a 30-min lag time (Fig. 8B, lane 5), suggesting that pro-C3 was efficiently cleaved by a putative converting enzyme into the subunits and that the faster migrating band is a degradation product of the subunit(s). In support of this interpretation, phenylmethylsulfonyl fluoride did not affect the processing of pro-C3 to the subunits but inhibited the degradation of the subunit(s) (Fig. 8B, lane 9). EGTA abol- ished the processing of pro-C3 completely (Fig. 8A, lane 3) , pointing to an absolute requirement of Ca2+ for the converting enzyme. For optimal activity 1-5 mM free Ca2+ was needed (Fig. 9A). The converting enzyme exhibited its maximal activity at pH 5.5-6.5 (Fig. 9B), consistent with its potential role in the acidic trans-Golgi compartments (38, 52, 53). The endoprotease was inhibited by p-chloromercuribenzoate (Fig. 8A, lane 9), suggesting that the thiol group is located in the active site or in the vicinity of the active site. However, resistance to leupeptin (Fig. 8A, lane 5 ) and E-64 (Fig. 8A, lane 6 ) rules out the possibility that cathepsin B or related

17469

- mc3

-01

- B

1 1-0

I 1

FIG. 8. Proteolytic cleavage of pro-C3 in v i t ro . A, 35S-plasma protein precursors were incubated in the complete reaction mixture for 2 h ( l a n e 2) as described under "Experimental Procedures." Incubations were performed for 2 h in the absence of Golgi membranes (lane 1 ) or in the presence of EGTA (lane 3,2 mM), leupeptin (lane 5,30 pglml), E-64 (lane 6,30 pglml), soybean trypsin inhibitor (lane 7, 30 pg/ml), p-amidinophenylmethylsulfonyl fluoride ( l a n e 8, 1 mM), andp-chloromercuribenzoate ( l a n e 9, 1 mM). Octyl glucoside was omitted from the reaction mixture (lane 4 ) . C3 molecules were immunoisolated and subjected to SDS-PAGE. Pro-C3 (lane 10) and mature C3 (lane 11 ) were immunoisolated from rat hepatocytes which had been labeled with [%]methionine. B, %-plasma protein precursors were incubated with Golgi membranes in the complete reaction mixture for 0 min (lane I), 5 min (lane 2), 10 min ( l a n e 3), 15 min ( l a n e 4 ) , 30 min (lane 5), 60 min (lane 6), 90 min ( l a n e 71, and 120 min (lanes 8-10). Phenylmethylsulfonyl fluoride (lane 9, 2 mM) and aprotinin (lane 10, 30 pg/ml) were added to the reaction mixtures, respectively. C3 molecules were immunoisolated and subjected to SDS-PAGE. A mixture of pro-C3 and mature C3 was immunoisolated from [35S]methionine-labeled rat hepatocytes (lane 11). a and p denote the a- and &subunits of mature C3, respectively.

lysosomal thiol proteinases are involved in the processing of pro-C3. Furthermore, the inability of soybean trypsin inhibitor (Fig. 8A, lane 7) and aprotinin (Fig. 8B, lane 8 ) to inhibit the endoprotease activity argues for the Golgi enzyme being distinct from trypsin or related serine proteases. Thus, these results as a whole strongly suggest that the Golgi endoprotease responsible for the processing of pro-C3 is a Ca2+-dependent enzyme.

DISCUSSION

The results presented in this paper demonstrate an essential role for Ca2+ in the proteolytic cleavage of plasma protein precursors such as proalbumin, pro-C3, and pro-C4 in rat hepatocytes and HepG2 cells. This conclusion is based on a series of experiments in which the effects of calcium depletion on the secretion and processing of the precursors were examined via in vivo application of drugs which cause an alter- ation of intracellular Ca2+ levels. When hepatocytes or HepG2 cells were incubated with the calcium-specific ionophore A23187 or ionomycin in the medium lacking CaC12, proteolytic cleavage of the precursors was disturbed, and consequently the precursors were found to be secreted into the medium (Figs. 1-4). These findings suggest that calcium depletion causes the inhibition of the proteolytic processing of plasma protein precursors. Calcium ionophores have been used widely to study the role of Ca2+ in intracellular transport and accom-

17470 Block of Proteolytic Cleavage by Calcium Depletion

A I 2 3 4 5 6 7 8 9 IO

F - Pm." - a

I - p

B 1 2 3 4 5 6 7 8 9 I

1 ip 1

FIG. 9. Characterization of a pro-C3 convertase activity. A, 3sS-plasma protein precursors were incubated for 2 h in the presence of 1 mM EGTA (lane I), 1 pM (lane Z ) , 10 pM (lane 3 ) , 100 p M ( l a n e 4 ) , 500 pM (lune 5) , 1 mM ( h n e 61, 5 mM (lane 71, 10 mM ( h n e 8,) and 50 mM ( l a n e 9) free Ca2'. EGTA (1 mM)/Ca2+ buffers were used to yield free Ca2+ indicated as described under "Experimental Proce- dures." Lane 10, a mixture of pro-C3 and mature C3 immunoisolated from [35S]methionine-labeled rat hepatocytes. B, 35S-plasma protein precursors were incubated for 2 h at pH 4.5 ( l a n e 2, acetate buffer), at pH 5.0 ( l a n e 3, acetate buffer), at pH 5.5 (lane 4, acetate buffer), at pH 6.0 ( l a n e 5, Mes buffer), at 6.5 (lane 6, Hepes buffer), at 7.0 (lane 7, Hepes buffer), and at 7.5 ( l a n e 8, Tris buffer). C3 molecules were immunoisolated and subjected to SDS-PAGE. Lane 1, no incubation; lane 9, a mixture of pro-C3 and mature C3 immunoisolated from [3sS]methionine-labeled rat hepatocytes. a and @ denote the a- and @-subunits of mature C3, respectively.

panying processing of various proteins (11-15, 31, 54, 55). However, as discussed by Beckers et al. (55), the results obtained through the in vivo application of calcium ionophores should be interpreted with caution. Since Ca'+ is involved in a wide variety of physiological processes in cells, modulation of Ca'+ levels in intact cells by the addition of the ionophores may cause multiple effects on cellular metabolism which are not simple to interpret. It is likely that the effects of the ionophores may vary from one experiment to another depend- ing on the concentration used, the presence of EGTA, and the time of exposure to the ionophores. Moreover, possible side effects of the ionophores themselves other than simple calcium-mediated actions cannot be dismissed. Lodish and Kong (31) have recently reported that A23187, but not ionomycin, inhibits the terminal glycosylation of plasma proteins in HepG2 cells (31). It is unlikely that the inhibitory effects of the calcium ionophores on proteolytic processing are caused by side effects of the ionophores. Pro-C3 appeared in the medium at as low a concentration as 0.0625 p M of the ionophores (Figs. 1 and 2). No preincubation step is needed for the ionophores to manifest their inhibitory effects (Fig. 1). Moreover, when the cells were exposed to the medium lacking Ca2+ for 2 h before the start of labeling, the presursor appeared in the medium (Fig. 5), excluding the possibility that inhibition of the proprotein cleavage is a result of side effects of the ionophores themselves. In support of these data, EGTA, the impermeable calcium-specific chelator, which indirectly facil- itates a reduction in intracellular Ca2+ levels, was found to strongly inhibit proteolytic processing (Fig. 5).

Subfractionation studies on rat liver and HepG2 cells

showed that mature serum albumin and C3 are enriched in the Golgi fractions, whereas the precursors are largely recovered in the microsomal fractions (49,50,56,57), indicating that proteolytic processing of proalbumin and pro-C3 occurs in the Golgi prior to secretion. Moreover, weak bases are known to inhibit the proteolytic cleavage of proalbumin (21, 25), pro-C3 (38, 58), and pro-C4 (58) in rat hepatocytes and HepG2 cells, suggesting that the acidic trans-Golgi compartments (trans-Golgi and/or TGN) are the sites at which plasma protein precursors undergo proteolytic cleavage. It is therefore reasonable to speculate that calcium depletion somehow affects the Golgi/TGN along the secreory pathway in such a way that plasma protein precursors are no longer processed to mature forms and are secreted in uncleaved forms. Three lines of evidence support this idea. First, pro-C3 was the only form found within the cells when the transport of newly synthesized C3 was arrested in a pre-Golgi compartment at a reduced temperature (Fig. 6) or in the mixed ER/Golgi upon brefeldin A treatment (Fig. 7), implying that pro-C3 is cleaved at later stages in the secretory pathway. Second, Ca'' is necessary for pro-C3 to undergo proteolytic cleavage at stages in transport subsequent to egress from the ER (Fig. 6). Third, reassembly of the Golgi complex and the resultant reorgani- zation of the secretory pathway during recovery from brefeldin A intoxication are sufficient for the passage of pro-C3 through later compartments of the secretory pathway but not for processing pro-C3. Pro-C3 is not processed to the mature C3 unless Ca2+ is afforded (Fig. 7). Taken together these findings strongly suggest that Ca2+ is needed for the proteolytic processing which takes place at the Golgi/TGN.

The mechanism by which calcium depletion leads to a blockage of this particular processing among other processing occurring in the Golgi/TGN remains unknown. However, the finding that Golgi membranes possess a Ca'+-dependent endoprotease capable of converting pro-C3 to the a- and p- subunits in vitro may be a clue to understanding the mechanism. Inhibitory spectra with class-directed protease inhibi- tors indicate that the pro-C3-converting enzyme is distinct from cathepsin B or related lysosomal thiol proteinases, typ- ical serine proteases, and trypsin (Fig. 8), precluding a possible lysosomal origin of the converting enzyme. The enzyme ex- hibits its maxmal activity at slightly acidic pH (Fig. 9 B ) , consistent with its role at the acidic trans-Golgi compartments. EGTA completely abolishes the converting activity in the Golgi fraction, demonstrating that the converting enzyme is indeed a Ca"-dependent endoprotease (Fig. 8A). The converting enzyme requires a mM range of free Ca2+ for full activity. At present no information is available with regard to both the physiological concentration of Ca2+ and the size of the Ca" pool in the Golgi lumen, especially those in the acidic trans-Golgi compartments. However, there are a few reports pertinent to Ca'+ store in the Golgi. Virk et al. (59) reported Ca2+-dependent ATPase activity in the Golgi vesicles prepared from rat mammary glands, Fasolato et al. (60) recently reported an inositol 1,4,5-triphosphate-insensitive ionomycin-sensitive Ca'' pool in PC12 cells which may represent a Ca'' pool in the Golgi. It is therefore conceivable that Ca'+ is pumped continuously into the Golgi lumen by a specific Ca2+- ATPase located on Golgi membranes to attain millimolar concentrations of Ca'+ as in the case of the ER (61). Notably, the converting activity is negligible at (1 mM free Ca'+ (Fig. 9 A), leading to speculation that calcium depletion causes a reduction in Ca2+ levels in the Golgi below a threshold concentration of 1 mM, resulting in inactivation of the pro-c3- converting enzyme in vivo. If this speculation is correct, a similar Ca'+-dependent endoprotease(s) should be involved in processing not only pro-C3 but also proalbumin and pro-C4, since the processing of both proalbumin and pro-C4 is also

Block of Proteolytic Cleavage by Calcium Depletion 17471

inhibited by calcium depletion (Figs. 3 and 4). In favor of this speculation, proalbumin is reported to be converted in vitro into mature serum albumin by a Ca2+-dependent Golgi endoprotease with properties similar to the pro-C3 convertase (27, 62), although it is not determined whether the processing of proalbumin and p r o 4 3 is catalyzed by a single enzyme or different enzymes.

A unique family of Ca2+-dependent endoproteases is emerg- ing as candidates for the processing enzymes involved in proteolytic cleavage of a wide range of proproteins. The cleavage of pro-a factor in the yeast is clearly shown to be catalyzed by a Ca2+-dependent subtilisin-like serine protease, the product of the KEX2 gene (63-65). Ca2+-dependent proteases in secretory granules of insulinoma cells are reported to convert proinsulin and chromogranin A to insulin and betagranin, respectively (66, 67). Moreover, fur and the related PC2 and PC3 genes, mammalian homologues of the yeast KEX gene, are found to cleave proproteins correctly when cotransfected with the genes for proproteins in mammaliam cells (2-9). In particular, furin, the product of the fur gene, is capable of converting proalbumin and pro-C3 to serum albumin and mature C3, respectively, when coexpressed with proalbumin or pro-C3 in the COS-1 cells ( 5 ) , raising the possibility that the pro-C3 convertase described in this study may be identical to furin. This possibility is currently being investigated.

Klenk et al. (14), in their pioneer work, showed that calcium depletion causes the inhibition of proteolytic cleavage of the HA precursor of the influenza virus. The present study ex- tends their observations to the proteolytic cleavage of plasma protein precursors. In light of these findings, it is possible that calcium depletion could block the proteolytic cleavage of a wide range of proproteins which are catalyzed by a family of Ca2+-dependent endoproteases. Thus, calcium depletion would provide a simple test to determine the involvement of a Ca2+-dependent protease in any proteolytic cleavage events of interest. Moreover, the precursors accumulated in response to calcium depletion could be used as substrates for in vitro assay systems of processing enzymes, as demonstrated in this study.

Acknowledgments-I am grateful to Dr. Y. Ikehara for support throughout this study. I would like to acknowledge Dr. Y. Sakamoto for suggesting the use of the program for formulations of the EGTA/ Caz+ buffers. I thank F. Arakawa for assistance in calculation using a computer. I also thank Drs. Y. Misumi and E. Tsuji for supplying isolated rat hepatocytes in some experiments. I thank Drs. A. Tak- atsuki and D. Kang for providing brefeldin A and HepG2 cells, respectively.

REFERENCES

Ikemizu, J., Murakami, K., and Nakayama, K. (1991) J. Biol. Chem. 266,12127-12130

11. Tartakoff, A. M., and Vassali, P. (1977) J. Exp. Med. 146 , 1332-1345 12. Uchida, N., Smilowitz, H., and Tanzer, M. L. (1979) Proc. Natl. Acad. Sci.

13. Johnson, D. C.. and Schlesinger, M. J. (1980) Virology 103,407-424 14. Klenk, H.-D., Garten, W., and Rott, R. (1984) EMBO J . 3,2911-2915 15. Spaete, R. R., Saxena, A,, Scott, P. I., Song, G. J., Probert, W. S., Britt, W.

J., Gibson, W., Rasmussen, L., and Pachl, C. (1990) J. Virol. 6 4 , 2922- 2931

16. Ikehara, Y., Oda, K., and Kato, K. (1977) J. Biochem. (Tokyo) 8 1 , 1293-

U. 5'. A. 7 6 , 1868-1872

1297 17. Oda,-K., Fujiwara, T., Ogata, S., and Ikehara, Y. (1990) J . Biochem. (Tokyo)

18. Se len P 0 (1976) Methods Cell Biol. 13 , 29-83 19. Od%, K., Misumi, Y., and Ikehara, Y. (1983) Eur. J. Biochem. 136 , 209-

108,549-553

316 20. O&:K., Fujiwara, T., and Ikehara, Y. (1990) Biochem. J . 266,161-167 21. Oda, K., Ogata, S., Koriyama, Y., Yamada, E., Mifune, K., and Ikehara, Y.

22. Williams, D. B., Borriello, F., Zeff, R. A., and Nathenson, S. G. (1988) J.

23. Degen, E., and Williams, D. B. (1991) J. Cell Biol. 112 , 1099-1115 24. Laemmli, U. K. (1970) Nature 227,680-685 25. Oda, K., and Ikehara, Y. (1985) Eur. J . Biochem. 162,605-609 26. Oda, K., Ikehara, Y., Ishikawa, T., and Kato, K. (1979) Biochim. Biophys.

27. Oda, K., and Ikehara, Y. (1988) J. Biochem. (Tokyo) 104 , 159-161 28. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J.

29. Ogawa, Y., and Kitazawa, T. (1975) Seikagaku Jikken KOZQ 16 , 325-337

30. Morris, K. M., Goldberger, G., and Colten, H. R. (1982) Science 216 , 399-

(1988) J. Biol. Chem. 263 , 12576-12583

Bwl. Chem. 263,4549-4560

Acta 662,212-224

Biol. Chem. 193,265-275

(in Japanese)

Ann 31. Lodish, H. F., and Kong, N. (1990) J. Biol. Chem. 2 6 6 , 10893-10899 32. Russell, J. H., and Geller, D. M. (1975) J. Biol. Chem. 260,3409-3413 33. Judah, J. D. and Quinn, P. S. (1976) Trends Biochem. Sci. 1, 107-109 34. Geller, D. hi., Judah, J. D., and Nicholls, M. R. (1972) Biochem. J. 127,

#5-R74 35. Oda, K., Misumi, Y., Sohda, M., Takami, N., Sakaki, Y., and Ikehara, Y.

36. Goldberger, G., and Colten, H. R. (1980) Nature 286 , 514-516 (1991) Biochem. Biophys. Res. Commun. 176,690-696

37. Chan, A. C., and Atkinson, J. P. (1983) J. Clin. Inuest. 72 , 1639-1649 38. Oda, K., Koriyama, Y., Yamada, E., and Ikehara, Y. (1986) Biochem. J.

39. Koriyama, Y., Oda, K., and Yamada, E. (1989) Biomed. Res. 10 , 341-351 40. Saraste, J., and Kuismanen, E. (1984) Cell 38,535-549 41. Tartakoff, A. M. (1986) EMBO J . 6, 1477-1482 42. Balch, W. E., Elliott, M. M., and Keller, D. S. (1986) J . Biol. Chem. 2 6 1 ,

14681-14689 43. Lodish, H. F., Kong, N., Hirani, S., and Rasmussen, J. (1987) J. Cell Biol.

104,221-230 44. Fujiwara, T. K., Oda, K., Yokota, S., Takatsuki, A,, and Ikehara, Y. (1988)

J . Biol. Chem. 263,18545-18552 45. Fujiwara, T., Oda, K., and Ikehara, Y. (1989) Cell Struct. Funct. 14, 605-

616 46. Li pincott-Schwartz, J., Yuan, L. C., Bonifacino, J. S., and Klausner, R. b. (1989) Cell 66,801-813 47. Li incott-Schwartz J Donaldson J. G. Schweizer A. Berger E., Hauri,

V-P., Yuan, L. C.: a i d Klausner: R. D: (1990) Ceh 60,821-636 48. Doms, R. W., Russ, G. and Yewdell, J. W. (1989) J. Cell Biol. 109,61-72 49. Lodish, H. F., Kong, N., Snider, M., and Strous, G. J. A. M. (1983) Nature

"_ _. -

240 , 739-745

2nd. WI-R.? 50. Oda, K., Miki, K., Hirose, S., Takami, N., Misumi, Y., and Ikehara, Y.

52. Anderson, R. G. W., and Pathak, R. K. (1985) Cell 40,635-640 51. Lipsky, N. G., and Pagano, R. E. (1985) Science 228 , 745-747

53. Schwartz A. Strous, G. J. A. M., Slot, J. W., and Geuze, H. J. (1985)

54. Booth, C., and Koch, G. L. E. (1989) Cell 69,729-737 55. Beckers C. J. M Plutner H. Davidson, H. W., and Balch, W. E. (1990)

56. Oda, K., Ikehara, Y., and Kato, K. (1978) Biochim. Biophys. Acta. 636,97-

1 - - , - - " (1986) J . Biochem. (Tokyo) 100,1669-1675

EMBO'J. i, 899-904

J. Bio). Chem. 2 6 6 , 18288-!8310

1 nA

2. Wise, R. J., Barr, P. J., Wong, P. A., Kiefer, M. C., Brake, A. J., and 1. Docherty, K., and Steiner, D. F. (1982) Annu. Reu. Physiol. 44,625-638

3. van de Ven, W. J. M., Voorberg, J., Fontijn, R., Pannekoek, H., van den Kaufman, R. J. (1990) Proc. Natl. Acad. Scr. U. 5'. A. 87,9378-9382

Ouweland, A. M. W., van Duijnhoven, H. L. P., Roebroek, A. J. M., and Siezen, B. J. (1990) Mol. Biol. Rep. 1 4 , 265-275

4. Bresnahan, P. A., Leduc, R., Thomas, L., Thorner, J., Gibson, H. L., Brake, A. J., Barr, P. J., and Thomas, G. (1990) J . Cell Biol. 11 1, 2851-2859

5. Misumi, Y., Oda, K., Fujiwara, T., Takami, N., Tashiro, K., and Ikehara,

6. Hatsuzawa, K., Hosaka, M., Nakagawa, T., Na ase, M., Shoda, A., Mura- Y. (1991) J . Biol. Chem. 266,16954-16959

7. Nakayama, K., Hosaka, M., Hatsuzawa, K., and Murakami, K (1991) J . kami, K., and Nakayama, K. (1990) J. Biol. &em. 266,22075-22078

8. Benjannet S., Rondeau, N., Day R., Chretien, M., and Seidah, N. G. (1991) Biochem. (Tokyo) 109,803-806

9. Thomas, L., Leduc, R., Thorne B. A., Smeekens, S. P. Steiner D. F., and Proc. N&l. Acad. Sci. U. S. A.'88, 3564-3568

Thomas G. (1991) Proc. Nati. Acd. Sci. U. S. A. 88,'5297-5$01 67. Hutton, J. C., Davidson, H. W., and Peshavaria, M. (1987) Biochem. J. 10. Hosaka, M., Nagahama, M., Kim, W.-S., Watanabe T., Hatsuzawa, K., 244,457-464

57. %-&an, C. M., Banerjee, D., Manning, C., Haung, C. Y., and Green, K.

58. Hortin, G., and Strauss, A. W. (1986) Biochem. Biophys. Res. Commun.

59. Virk, S. S., Kirk, C. J., and Shears, S. B. (1985) Biochem. J. 2 2 6 , 741-748 60. Fasolato, C., Zottini, M., Clementi, E., Zaccheti, D., Meldolesi, J., and

61. Carafoli, E. (1987) Annu. Reu. Biochem. 66,395-433 62. Brennan, S. O., and Peach, R. J. (1988) FEES Lett. 229 , 167-170 63. Julius, D., Brake, A., Blair, L., Kunisawa, R., and Thorner, J. (1984) Cell

64. Mizuno, K., Nakamura, T., Ohshima, T., Tanaka S., and Matsuo, H.(1988)

65. Fuller, R. S. Brake, A,, and Thorner, J. (1989) Proc. Natl. Acad. Sei.

66. Davidson, H. W., Rhodes, C. J., and Hutton, J. C. (1988) Nature 333,93-

(1978) J. Cell Biol. 77,400-416

136,603-609

Possan, T. (1991) J . Biol. Chem. 2 6 6 , 20159-20167

37,1075-1089

Biochem. Biophys. Res. Commun. 166 , 246-2i4

U. S. A. 86,1434-1438

9fi

Documents

Calcium Depletion Blocks Proteolytic Cleavages of Plasma Protein