Báo cáo khoa học: The fate of newly synthesized V-ATPase accessory subunit Ac45 in the secretory pathway

doi:10.1046/j.1432-1033.2002.02831.x

Eur. J. Biochem. 269, 1844–1853 (2002) (cid:211) FEBS 2002

The fate of newly synthesized V-ATPase accessory subunit Ac45 in the secretory pathway

Vincent Th. G. Schoonderwoert, Eric J. R. Jansen and Gerard J. M. Martens

Department of Animal Physiology, Nijmegen Center for Molecular Life Sciences, University of Nijmegen, the Netherlands

N-terminal fragment is rapid, does not occur in lysosomes and is inhibited by brefeldin A. Both the N- and C-terminal fragment pass the medial Golgi, as they become partially endoglycosidase H resistant. The Ac45 cleavage event is a relatively slow process (half-life of intact Ac45 is 4–6 h) and takes place in the early secretory pathway, as it is not affected by brefeldin A and monensin. Tunicamycin inhibited N-linked glycosylation of Ac45 and interfered with the cleavage process, suggesting that Ac45 needs proper folding for the cleavage to occur. Together, our results indicate that Ac45 folding and cleavage occur slowly and early in the secretory pathway, and that the cleavage event may be linked to V-ATPase activation.

Keywords: acidification; regulated secretory pathway; post- translational modification; vacuolar proton ATPase; Xenopus.

The vacuolar H+-ATPase (V-ATPase) is a multimeric enzyme complex that acidifies organelles of the vacuolar system in eukaryotic cells. Proteins that interact with the V-ATPase may play an important role in controlling the intracellular localization and activity of the proton pump. The neuroendocrine-enriched V-ATPase accessory subunit Ac45 may represent such a protein as it has been shown to interact with the membrane sector of the V-ATPase in only a subset of organelles. Here, we examined the fate of newly synthesized Ac45 in the secretory pathway of a neuroendo- crine cell. A major portion of intact (cid:25) 46-kDa Ac45 was found to be N-linked glycosylated to (cid:25) 62 kDa and a minor fraction to (cid:25) 64 kDa. Trimming of the N-linked glycans gave rise to glycosylated Ac45-forms of (cid:25) 61 and (cid:25) 63 kDa that are cleaved to a C-terminal fragment of 42–44 kDa (the deglycosylated form is (cid:25) 23 kDa), and a previously not detected (cid:25) 22-kDa N-terminal cleavage fragment (the deglycosylated form is (cid:25) 20 kDa). Degradation of the

[13]. In the TGN, an acidic pH is necessary for the proper processing of proproteins [14] and for the condensation of regulated secretory proteins, which is important for their targeting to immature secretory granules [15–17]. Immature secretory granules mature and become progressively more acidic (pH of (cid:25) 5.5 [18–20]). Granular acidification further concentrates regulated proteins [21], while nonregulated proteins are sorted away into clathrin-coated vesicles that pinch off from the maturing granule [22–24]. Furthermore, the acidic granular pH is necessary for the processing enzymes to efficiently cleave the prohormones [25].

Acidification of organelles in eukaryotic cells is required for a variety of cellular processes, such as the release of ligands from receptors during endocytosis and the hydrolysis of macromolecules in lysosomes [1–3]. In the secretory path- way, the lumen gradually acidifies from endoplasmic reticulum (ER) to Golgi to secretory granules (reviewed in [4]). The pH of the lumen of the ER, Golgi, and trans-Golgi Network (TGN) is (cid:25) 7.3, (cid:25) 6.4, and (cid:25) 6.0, respectively, and is similar in regulated and nonregulated secretory cells [5–10]. The significance of the pH in the ER remains to be established, although it seems likely that ER processes such as protein glycosylation and folding depend on it. The low pH in the Golgi has been shown to be important for the regulation of protein–protein interactions [11,12] and the activity of the N-glycan processing enzyme sialyltransferase

Acidification of intracellular compartments is established and maintained by the vacuolar H+-ATPase (V-ATPase). This multimeric enzyme complex consists of at least 13 different subunits that have been classified into a membrane integral sector (V0) and a peripheral sector (V1) [26,27]. The V-ATPase V1 sector contains the catalytic site which hydrolyses ATP to translocate protons across the mem- brane by the proton-pore forming V0 sector. In the ER, the assembly of the V-ATPase starts with the V0-sector and may be completed in this compartment by the build-up of the V1 onto the V0 [28,29]. Given the pH in the ER, the V-ATPase should be considered as being essentially inactive in this part of the secretory pathway. An active V-ATPase is required further downstream in the secretory pathway but it is not known in which compartment the V-ATPase becomes active and which mechanism is involved in the targeting of the V-ATPase to the various secretory pathway compart- ments. V-ATPase interacting proteins, such as the accessory subunit Ac45, may play an important role in this targeting process, as Ac45 has been shown to interact with the

Correspondence to G. J. M. Martens, Department of Animal Physi- ology, Nijmegen Center for Molecular Life Sciences, University of Nijmegen, Geert Grooteplein Zuid 28, 193RT, 6525 GA Nijmegen, the Netherlands. Fax: + 31 24 3615317, Tel.: + 31 24 3610564, E-mail: g.martens@ncmls.kun.nl Abbreviations: Baf, bafilomycin A1; BFA, brefeldin A; EndoH, endoglycosidase H; ER, endoplasmic reticulum; NDGA, nordi- hydroguaiaretic acid; NIL, neurointermediate lobe; PC2, prohormone convertase 2; POMC, proopiomelanocortin; TGN, trans-Golgi net- work; V-ATPase, vacuolar H+-ATPase. Note: a web page is available at http://www.kun.nl/molanphys/Homepage/home.htm (Received 26 October 2001, revised 6 February 2002, accepted 8 February 2002)

The fate of Ac45 in the secretory pathway (Eur. J. Biochem. 269) 1845

(cid:211) FEBS 2002

Metabolic labeling of Xenopus neurointermediate lobes and immunoprecipitation analysis

membrane sector of the V-ATPase in only a subset of organelles [30]. Ac45 was initially isolated from bovine chromaffin granules and identified as a type I transmem- brane protein of 45 kDa [30]. However, N-terminal sequencing of the isolated 45-kDa protein and the cloning of full-length Ac45 cDNA revealed that the isolated protein represents a cleaved fragment of a larger protein [30,31]. In a differential screening strategy aimed at identifying genes that are involved in the biosynthesis and release of peptide hormones, we isolated a cDNA (X1311) encoding Ac45 of the amphibian Xenopus laevis [32]. The melanotrope cells of the Xenopus intermediate pituitary were used for this screening approach because the activity of these neuroen- docrine cells can be physiologically stimulated by placing the animal on a black background. The cellular activation results in the production and release of large amounts of the proopiomelanocortin (POMC)-derived melanophore-stimu- lating hormone, which causes pigment dispersion in dermal melanophores, thereby darkening the skin [33]. Approxi- mately 10 times more Ac45 transcripts have been found in the melanotrope cells of animals adapted to a black background compared to those of white-adapted animals [32], suggesting that Ac45 has an important role in the regulated secretory pathway of neuroendocrine cells.

Here, we examined in detail the fate of the Ac45 protein in the melanotrope cells of Xenopus intermediate pituitary. We found that in these cells, the folding and proteolytic cleavage of intact Ac45 is slow, and occurs in the early secretory pathway where activation of V-ATPases is required.

Fig. 1. Antigenic epitopes used to produce Ac45 region-specific antisera. Recombinant proteins comprising residues Gly68 to Pro388 and Pro208-Ser381, and a synthetic peptide corresponding to the 12 C-terminal amino-acid residues of Xenopus Ac45 were used to produce rabbit polyclonal antisera 1311N, 1311C, and 1311NC, respectively.

M A T E R I A L S A N D M E T H O D S

Animals

South-African clawed toads, Xenopus laevis, were bred and reared in the aquarium facility of the Department of Animal Physiology of the University of Nijmegen. Animals were adapted to a black background by keeping them in black buckets under constant illumination for at least three weeks at 22 (cid:176)C. All experiments were carried out under the guidelines of the Dutch law concerning animal welfare.

Biochemicals and antibodies

Neurointermediate lobes (NILs) from black-adapted Xen- opus laevis were dissected and preincubated in methionine- and cysteine-free culture medium [6.7 mL L15 medium (Gibco-BRL, Gaithersburg, MD, USA), 3 mL milli-Q water, 10 lgÆmL)1 kanamycin, 1% antibiotic-antimycotic solution (Gibco-BRL), 8 mg CaCl2, 3 mg bovine serum albumin and 2 mg glucose] for 30 min at 22 (cid:176)C. Pulse labeling of newly synthesized proteins was performed by incubating the lobes in methionine/cysteine-free culture medium containing 5 mCiÆmL)1 [35S]Met/Cys (Promix, Amersham, Buckinghamshire, UK) for 1 h at 22 (cid:176)C. Subsequent chase incubations were in culture medium supplemented with 5 mM L-methionine, 2.5 mM L-cysteine and 10% fetal bovine serum. BFA (2.5 lgÆmL)1) was present during the pre-, pulse and chase incubations, unless stated otherwise. NDGA (30 lM) was present only during the chase incubation. In some experiments, lobes were first incubated overnight in the absence or presence of 10 lgÆmL)1 tunicamycin in culture medium containing 10% fetal bovine serum (Gibco-BRL). For immunoprecipi- tation analysis, lobes were homogenized on ice in lysis buffer (50 mM Hepes pH 7.2, 140 mM NaCl, 10 mM EDTA, 1% Tween-20, 0.1% Triton X-100, 0.1% deoxycholate) containing 1 mM phenylmethanesulfonyl fluoride and 0.1 mgÆmL)1 soybean trypsin inhibitor. Homogenates were cleared by centrifugation (10 000 g, 7 min at 4 (cid:176)C), and used for protein deglycosylation (see below), or directly supplemented with 0.1 volume of 10% SDS and diluted 10-fold in lysis buffer before addition of anti-Ac45 antise- rum (1 : 500 dilution). Immune complexes were precipitated with protein-A–Sepharose (Pharmacia Biotech, Uppsala, Sweden) and subjected to SDS/PAGE [35]. Gels were processed for fluorography and radiolabeled proteins were detected by autoradiography.

Immunoblotting

NILs dissected from black-adapted Xenopus laevis were incubated overnight at 22 (cid:176)C in culture medium with 10% fetal bovine serum in the absence or presence of drugs, or directly homogenized in lysis buffer containing 1 mM phenylmethanesulfonyl fluoride and 0.1 mgÆmL)1 soybean trypsin inhibitor. Lysates were cleared by centrifugation (10 000 g, 7 min, 4 (cid:176)C) and used for protein deglycosylation (see below) or immediately denatured in sample buffer at

Rabbit polyclonal antisera 1311C and 1311N, directed against a synthetic peptide comprising the 12 C-terminal amino-acid residues of Xenopus Ac45 and against a recombinant fragment of Xenopus Ac45 (comprising ami- no-acid residues Gly68 to Pro388 with a hexahistidine tail at its N-terminus; numbering according to [34]), respectively, have been described previously [34] (Fig. 1). Rabbit poly- clonal antiserum 1311NC was raised against a recombinant polypeptide corresponding to amino-acid residues Pro208– Ser381 (numbering according to [34]) of Xenopus Ac45 expressed in E. coli as a fusion protein with a hexahistidine tag at its C-terminus (Cogon, Hilden, Germany) (Fig. 1). Brefeldin A (BFA), monensin, nordihydroguaiaretic acid (NDGA), chloroquine, and tunicamycin were purchased from Sigma (St Louis, MO, USA). Leupeptin was from Roche Diagnostics (Mannheim, Germany) and bafilomycin A1 (Baf) from Wako Pure Chemical Industries (Osaka, Japan).

1846 V. Th. G. Schoonderwoert et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

4 h and 8 h, giving rise to products of (cid:25) 61 and (cid:25) 63 kDa (Fig. 2, compare lane 1 with 2 and 10 with 11). This minor shift in mobility is likely due to a change in the N-linked sugars (possibly oligosaccharide trimming), as deglycosyla- tion of these proteins again yielded a product of (cid:25) 46 kDa

95 (cid:176)C for 5 min Proteins were separated by SDS/PAGE and electrotransferred to nitrocellulose (Schleicher & Schuell, Dassel, Germany). Membranes were blocked and washed with blocking buffer (100 mM NaCl; 100 mM Na2PO4; 1% Tween-20) containing 5% low-fat dry milk. Blocking buffer with 2% low-fat dry milk was used for further washing steps and incubations with primary and secondary antibodies. The secondary antibody used was an peroxidase-conjugated anti-(rabbit IgG) Ig (Sigma, St Louis, MO, USA) at a dillution of 1 : 3000. Peroxidase activity was detected using the Lumilight system (Roche Diagnostics, Mannheim, Germany).

Deglycosylation of proteins

Proteins were treated with endoglycosidase H (EndoH) (Roche Diagnostics, Mannheim, Germany) to remove high- mannose N-glycans from glycoproteins. Lysates were boiled for 10 min in 50 mM Na-citrate buffer (pH 5.5) containing 0.1% SDS, gradually cooled to RT, and incubated overnight in the absence or presence of 40 mUÆmL)1 EndoH at 37 (cid:176)C. Proteins were deglycosylated by N-glycosidase F (Roche Diagnostics, Mannheim, Germany) to remove both high-mannose and complex oligosacchar- ides. For this purpose, protein lysates were boiled for 10 min in 10 mM Hepes (pH 7.4) containing 0.1% SDS, cooled down to RT, supplemented with 0.5% Nonidet P-40, 100 lM phenylmethanesulfonyl fluoride and 100 lgÆmL)1 soybean trypsin inhibitor, and incubated overnight at 37 (cid:176)C with or without 5 U N-glycosidase F per mL.

R E S U L T S

Intact newly synthesized Ac45 is N-linked glycosylated

To study the biosynthesis of Ac45, we raised, in addition to the previously produced antisera 1311N and 1311C (Fig. 1; [34]), a third anti-Ac45 antiserum (1311NC; against a recombinant protein comprising Xenopus Ac45 residues 208–388; Fig. 1). Following a 1-h pulse labeling of neuro- intermediate lobes (NILs) from black-adapted Xenopus, the 1311N and 1311C antisera detected a newly synthesized protein of (cid:25) 62 kDa and a less abundant protein of (cid:25) 64 kDa (Fig. 2, lanes 1 and 10). Both proteins represent intact forms of Ac45 and vary only in the degree of N-linked glycosylation, as deglycosylation of these radiolabeled proteins by N-glycosidase F led to an (cid:25) 46-kDa protein (Fig. 2, lanes 4 and 13). The mobility of the two intact forms increased slightly during subsequent chase incubations of

Fig. 2. Deglycosylation allows detection of the Ac45 processing products by region-specific polyclonal antisera. NILs from black-adapted Xenopus were pulsed for 1 h with [35S]Met/Cys and then chased for the indicated time periods. Total lobe extracts were directly subjected to immunoprecipitation with antisera 1311N, 1311C or 1311NC, or deglycosylated by N-glycosidase F or EndoH prior to immunopre- cipitation. Precipitated proteins were resolved by SDS/PAGE and visualized by fluorography. Migration positions of intact and processed forms of Ac45 are indicated. Lane 18 with increased contrast is depicted in lane 18¢. Note that some of the immunoprecipitates contain 37-kDa glycosylated or 35-kDa deglycosylated POMC that bound nonspecifically (asterisk).

The fate of Ac45 in the secretory pathway (Eur. J. Biochem. 269) 1847

(cid:211) FEBS 2002

(Fig. 2, lanes 5, 6, 14 and 15). The amount of 61- and 63- kDa Ac45 decreased during the 8-h chase (half-life 4–6 h) as a result of a cleavage event (see below). Both the 61- and 63- kDa forms are not immunoprecipitated by the 1311NC antiserum (Fig. 2, lanes 19–21). Presumably, the N-linked glycans prevent detection of these forms as their removal by N-glycosidase F results in the immunoprecipitation of the deglycosylated 46-kDa intact form by this antiserum (Fig. 2; lanes 22–24). These results show that newly synthesized Ac45 is N-linked glycosylated to a major product of (cid:25) 62 kDa and a minor product of (cid:25) 64 kDa that are subsequently processed to (cid:25) 61- and (cid:25) 63-kDa products.

Newly synthesized Ac45 is cleaved

(see below) are not immunoprecipitated with the two antisera raised against more C-terminally located regions of Ac45 (1311C and 1311NC, Figs 1 and 2, lanes 10–15 and 19–24). Finally, the N-terminal Ac45 fragment contains one potential N-linked glycosylation site (Asn128; numbering according to [34]), and this site appears to be used, as N-glycosidase F treatment of the NIL lysate prior to immu- noprecipitation causes a shift in the mobility of the 22-kDa product to (cid:25) 20 kDa (Fig. 2, lanes 4–6). The amount of the N-terminal fragment would be expected to increase during the chase period because of the progressive cleavage of intact Ac45. However, during the chase incubation a decrease in the level of the N-terminal fragment was found, suggesting that this cleavage product may be subjected to an intracellular degradation process. This circumstance may also explain why the N-terminal Ac45 fragment has not been detectable by Western blotting [34].

Transport of newly synthesized Ac45 to the Golgi

(Fig. 2,

In the melanotrope cells of the Xenopus NIL, intact N-linked glycosylated Ac45 is intracellularly cleaved to a C-terminal fragment of (cid:25) 40 kDa. Although the (cid:25) 40-kDa product could be detected by Western blotting with the C-terminally directed anti-Ac45 serum 1311C, this antise- rum did not immunoprecipitate the newly synthesized form of this fragment [32]. However, after optimization of the immunoprecipitation conditions, we detected the newly synthesized C-terminal product with antiserum 1311C as a diffuse band of 42–44 kDa (Fig. 2, lanes 11 and 12). With antisera 1311N and 1311NC, we could not precipitate this product (Fig. 2, lanes 2 and 3, and 20 an 21), possibly because of the presence of numerous N-linked glycans in this region of the protein (Fig. 1). Indeed, after removal of the N-linked glycans by N-glycosidase F, all three antisera (1311N, 1311NC and 1311C) immunoprecipitated this fragment in its deglycosylated forms, namely as proteins of (cid:25) 23 and (cid:25) 24 kDa (Fig. 2, lanes 5, 14 and 23, and 6, 15 and 24, respectively). During the chase incubation, the mobility of the deglycosylated C-terminal fragment shifted from (cid:25) 23 kDa to (cid:25) 24 kDa (Fig. 2, lane 14 and 15), probably as the result of a post-translational modification. the deglycosylated C-terminal Ac45 The amount of cleavage fragment, with a size of (cid:25) 23 kDa after 4 h of chase and (cid:25) 24 kDa after 8 h of chase, increased during the chase incubation (Fig. 2, lane 14 and 15), as was expected because of the progressive cleavage of intact 61/ 63-kDa Ac45. Thus, from these data, we conclude that newly synthesized (cid:25) 61/63-kDa Ac45 is cleaved to C-terminal products of 42–44 kDa (with deglycosylated forms of (cid:25) 23 and (cid:25) 24 kDa).

Identification of the N-terminal Ac45 cleavage product

In the medial Golgi, N-linked oligosaccharides can be modified to two broad classes, namely complex oligosac- charides and high-mannose oligosaccharides. Both types of oligosaccharides can be removed completely from proteins by treating them with N-glycosidase F. In contrast, endoglycosidase H (EndoH) removes only high-mannose oligosaccharides. The acquisition of resistance of an N-glycosylated protein to EndoH, which requires the action of glycosylation enzymes localized in the medial Golgi, can thus be used to determine whether a glycosylated protein has entered the medial compartment [36]. We determined whether the intact or the cleavage products of Ac45 acquire resistance to digestion with EndoH. Extracts of pulse- chased NILs were subjected to EndoH before immunopre- cipitation with antisera 1311N, 1311C or 1311NC. All three anti-Ac45 antisera immunoprecipitated from the EndoH-treated NIL lysate a newly synthesized product of (cid:25) 46 kDa. This product corresponds with the intact newly synthesized deglycosylated Ac45 protein that was immuno- precipitated from NIL lysates that were treated with N-glycosidase F, indicating that intact Ac45 is sensitive to EndoH. This finding implies that intact Ac45 is cleaved in a compartment before the medial Golgi. Antisera 1311N and 1311C immunoprecipitated from the EndoH-treated and the N-glycosidase F-treated lysates similar amounts of the (cid:25) 46-kDa product lanes 7–9 and 16–18). In contrast, the 1311NC antiserum precipitated a considerably lower amount of this product from the EndoH-treated than from the N-glycosidase F-treated lysates (Fig. 2, compare lanes 22–24 with 25–27). Probably, the presence of the N-acetylglucosamine residues remaining after EndoH digestion [37], but removed by N-glycosidase F [38], lowers the affinity of the Ac45 product for the 1311NC antiserum. This possibility may also explain why this antiserum was not able to detect significant amounts of the (cid:25) 23-kDa C-terminal cleavage product in EndoH-treated lysates (Fig. 2, lanes 17 and 18).

lane 15),

In addition to the (cid:25) 23-kDa product, the 1311C anti- serum detected also a low-abundant product of (cid:25) 26-kDa in the EndoH-treated lysate (Fig. 2, lanes 17 and 18/18¢ ). This product was not detected in the N-glycosidase F-treated lysate (Fig. 2, indicating that it represents a C-terminal Ac45 cleavage form of which most, but not all,

In contrast to what holds for the C-terminal cleavage fragment of Ac45 [30,34], the N-terminal cleavage fragment has not been identified yet. However, after optimization of the immunoprecipitation conditions, from newly synthe- sized Xenopus NIL proteins we precipitated with antiserum 1311N a low-abundant product of (cid:25) 22 kDa (Fig. 2, lanes 1–3). Because of the following we conclude that this (cid:25) 22- kDa product is the glycosylated form of the N-terminal Ac45 cleavage fragment. First, the size of this product is in line with the predicted size of the N-terminal fragment that remains following cleavage of intact 61/63-kDa Ac45 to the 42–44-kDa C-terminal product. Furthermore, both the (cid:25) 22-kDa fragment and its deglycosylated (cid:25) 20-kDa form

1848 V. Th. G. Schoonderwoert et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

BFA inhibits the degradation of the N-terminal Ac45 fragment

N-glycans are sensitive to EndoH. The amount of the (cid:25) 23-kDa product in the EndoH-treated lysates remained constant during the chase, whereas the analysis of the N-glycosidase F-treated samples clearly indicated an increase in the total amount of this fragment (Fig. 2, compare lanes 14 and 15 with 17 and 18). These findings suggest that at first, all the N-linked sugars on the C-terminal cleavage product are sensitive to EndoH (EndoH treatment gives an (cid:25) 23-kDa product), and that during the chase some of the N-glycans on the C-terminal cleavage product become resistant to EndoH (resulting in an (cid:25) 26-kDa product). The N-linked sugar on the N-terminal cleavage product also acquired resistance to EndoH, as we found a faint band of (cid:25) 22 kDa in the EndoH-treated extracts that is absent in the total lysates of these samples (Fig. 2, lanes 8 and 9, and data not shown).

Western blot analysis was employed to study the steady state levels of EndoH-sensitive and EndoH-resistant forms of Ac45. In line with the results of biosynthetic studies, EndoH treatment of the NIL lysate prior to Western blot analysis with the 1311C antibody again resulted in the detection of an (cid:25) 23-kDa and an (cid:25) 26-kDa product (Fig. 3, lane 2). The intensity of the (cid:25) 23-kDa band is higher than that of the (cid:25) 26-kDa band, indicating that in the steady state situation the (cid:25) 23-kDa product is the major form in the EndoH-treated lysate. As expected, deglycosylation by N-glycosidase F resulted in the detec- tion of the (cid:25) 23-kDa C-terminal cleavage product (Fig. 3, lane 3). As this product is more abundant in the EndoH- treated NIL lysate than the (cid:25) 26-kDa product, we conclude that at steady state, most of the glycosylated 42–44-kDa C-terminal cleavage products contain N-linked glycans that are sensitive to EndoH.

in this compartment

Together, these results demonstrate that the cleavage of intact 61/63-kDa Ac45 occurs before the medial Golgi, and that the N-glycan on the N-terminal and some of the N-glycans on the C-terminal cleavage product are converted to complex oligosaccha- rides.

As the N-terminal fragment was not detected by immu- noblotting, we hypothesized that it may be degraded intracellularly. We examined this possibility by affecting Ac45 transport through the secretory pathway via drugs that interfere with intracellular protein transport, namely the fungal metabolite brefeldin A (BFA) and the sodium ionophore monensin. BFA causes fusion of Golgi mem- branes with the ER and the retention of newly synthesized proteins in a lumenal milieu characteristic of the early compartments of the secretory pathway [39]. In addition, BFA blocks the exit of proteins from the TGN [40] as we have recently shown for several regulated secretory proteins in Xenopus melanotrope cells [41]. Monensin interferes with protein transport between Golgi compartments [42]. Fur- thermore, to examine if the N-terminal cleavage product is degraded in lysosomes, a number of compounds that interfere with lysosomal function were used. Leupeptin is a thiol protease inhibitor that inhibits degradation of proteins in lysosomes [43]. The weak base chloroquine and the V-ATPase-specific inhibitor bafilomycin A1 (Baf) are known to inhibit lysosomal and endosomal enzymes by disturbing the intralumenal pH [1]. Baf may also affect the transport of intact Ac45 or its cleavage products in a post- TGN compartment, e.g. in Xenopus melanotrope cells [41]. To examine the effects of the above-mentioned drugs, Xenopus NILs were incubated overnight in the absence or presence of a drug, and the lobes were lysed and subjected to Western blotting with the 1311N or 1311C antiserum. The N-terminal cleavage fragment of Ac45 did not accumulate when NILs were incubated in the presence of monensin or the lysosomal inhibitors leupeptin, chloroquine and Baf. However, in NILs incubated in the presence of BFA, an (cid:25) 22-kDa product had clearly accumulated (Fig. 4A). This product could be deglycosylated with N-glycosidase F to (cid:25) 20 kDa and was not detected with the 1311C antibody (data not shown), indicating that this product represents the N-terminal fragment of Ac45. The drugs used did not significantly change the amount of the 42–44-kDa

Fig. 4. Effect of inhibitors of intracellular transport and lysosomal function on the degradation of the N-terminal 22-kDa Ac45 cleavage fragment. NILs dissected from black-adapted Xenopus were incubated overnight in medium with no drugs, brefeldin A (BFA, 2.5 lgÆmL)1), chloroquine (Chl, 100 lM), bafilomycin A1 (Baf, 1 lM), leupeptin (Leu, 100 lgÆmL)1), or monensin (Mon, 100 nM). Proteins were extracted from these NILs, separated by SDS/PAGE, transferred to nitrocellulose and probed with the anti-Ac45 serum 1311N to detect the 22-kDa N-terminal fragment (A) or 1311C to detect the 42 to 44-kDa C-terminal fragment (B). Fig. 3. Steady-state levels of EndoH-sensitive and -resistant forms of the C-terminal Ac45 cleavage fragment. Total NIL extracts from black- adapted Xenopus were incubated overnight with no enzyme (lane 1), EndoH (lane 2), or N-glycosidase F (lane 3). Reactions were stopped by adding SDS sample buffer, and the samples were subjected to SDS/ PAGE and immunoblotting, using 1311C.

The fate of Ac45 in the secretory pathway (Eur. J. Biochem. 269) 1849

(cid:211) FEBS 2002

C-terminal Ac45 product (Fig. 4B), suggesting that cleavage of intact Ac45 was not affected.

in the next 4-h chase period with BFA, the amount of the cleaved N-terminal product did not increase, presumably because this fragment was degraded (Fig. 5, lane 3). Thus, BFA leads to an accumulation of the N-terminal fragment, but does not prevent the degradation process. These data indicate that the BFA-induced transport block of proteins out of the ER still allows cleavage of intact Ac45, and support our notion that cleavage of Ac45 occurs in the early secretory pathway. We also conclude that degradation of the N-terminal (cid:25) 22-kDa cleavage fragment is inhibited by BFA, and seems to occur after the N-linked sugar acquires EndoH resistance and not in the endosomal-lysosomal system.

BFA leads to the accumulation of the newly synthesized C-terminal Ac45 cleavage fragment

Next, we sought to determine whether not only the steady state levels but also the amount of the newly synthesized N- terminal cleavage fragment is affected by BFA. For this purpose, we pulse-chased NILs in the absence or presence of BFA, and performed immunoprecipitation analyses with antibodies 1311N and 1311C. In line with the Western blotting results, the presence of BFA did not affect the cleavage of intact Ac45. However, in the presence of BFA, intact Ac45 does not migrate as a 61/63-kDa product, but rather as a single product of (cid:25) 61 kDa (Fig. 5, lanes 2 and 3). Apparently, the redistribution of Golgi enzymes to the ER induced by BFA [44] results in the premature trimming of the N-linked sugars. BFA-treatment also led to the accumulation of the (cid:25) 22-kDa N-terminal cleavage frag- ment during the first 4 h of chase (Fig. 5, lane 2). However,

notion, we used the

lipoxygenase

At steady state, the C-terminal Ac45 cleavage fragment is the predominant form of Ac45 present in the melanotrope cells of the NIL (Fig. 3, lane 2). However, in the biosyn- thetic studies the amount of newly synthesized C-terminal cleavage product was lower than one would expect on the basis of the amount of intact glycosylated Ac45 that is cleaved to the C-terminal product. Surprisingly, immuno- precipitates from extracts of BFA-incubated NILs (Fig. 5) show, in addition to the newly synthesized N-terminal fragment, a high amount of newly synthesized C-terminal Ac45 cleavage product, much higher than detected in NILs that were incubated in the absence of BFA (Fig. 2, lane 1–3, 10–12). Possibly, the region of the C-terminal cleavage fragment to which the 1311C antibody was directed (the cytoplasmic tail of Ac45) is more accessible to the antibody in the presence of BFA. A binding candidate may be COPI, as BFA is known to dissociate COPI from Golgi mem- branes [45,46]. Alternatively, and more likely, BFA led to the accumulation of the C-terminal cleavage fragment of Ac45 in the ER-Golgi, thereby preventing the C-terminal fragment from obtaining its normal conformation or from associating with its normal partner (e.g. the V-ATPase enzyme complex). In case of the possibility of epitope unmasking, one would expect to find equal amounts of the C-terminal cleavage product to be immunoprecipitable from radiolabeled NILs when BFA is either present constantly or added at a later stage of the chase period. However, from NILs pulse-chased in the continuous presence of BFA (Fig. 6, lane 1) or chased first in the presence and then in the absence of BFA (Fig. 6, lane 2), the amount of immunoprecipitated C-terminal cleavage prod- uct is much higher than from NILs chased first in the absence and then in the presence of BFA (Fig. 6, lane 3). Therefore, we conclude that the more efficient detection of the C-terminal cleavage product of Ac45 in the presence of BFA can not be attributed to an unmasking of the 1311C epitope by, e.g. COPI-dissociation. To further support this inhibitor nordihydroguaiaretic acid (NDGA), a drug acting similar to BFA but preventing dissociation of COPI from Golgi membranes [47]. As for BFA, the presence of NDGA during the chase incubation allowed the efficient detection of the newly synthesized C-terminal product (Fig. 7, lane 1 and 2) and thus COPI dissociation is not involved. Together, we conclude that inhibition of ER to Golgi transport prevents the C-terminal Ac45 cleavage product

Fig. 5. BFA allows immunoprecipitation of the N- and C-terminal Ac45 cleavage fragments. NILs from black-adapted Xenopus were pulsed for 1 h with [35S]Met/Cys and subsequently chased for the indicated time periods in the presence of BFA. Ac45 products were immunoprecipi- tated with both the 1311N and 1311C antibody. Precipitated proteins were resolved by SDS/PAGE and visualized by fluorography. Migration positions of intact and processed forms of Ac45 are indicated. Note that some of the immunoprecipitates contain 37-kDa POMC and (cid:25) 70 kDa prohormone convertase PC2 that bound nonspecifically (asterisk).

1850 V. Th. G. Schoonderwoert et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

Fig. 6. BFA leads to the accumulation of the C-terminal cleavage product of Ac45. NILs from black-adapted Xenopus were pulsed for 1 h with [35S]Met/Cys in the presence of BFA, and chased for two subsequent periods of 4 h in the absence or presence BFA. Ac45 products were immunoprecipitated with antibody 1311C, separated by SDS/PAGE and visualized by fluorography. The migration positions of intact and processed forms of Ac45 are indicated. Note that some of the immunoprecipitates contain 37-kDa POMC and (cid:25) 70 kDa PC2 that bound nonspecifically (asterisks).

from adopting its normal conformation or from interacting with its binding partner.

Tunicamycin inhibits N-linked glycosylation and cleavage of intact Ac45

N-linked glycosylation of intact Ac45. Interestingly, the processing of the (cid:25) 46-kDa unglycosylated intact form of Ac45 was clearly affected (Fig. 8, lanes 4–6). Even after 8 h of chase a high amount of the (cid:25) 46-kDa unglycosylated intact form of Ac45 is still present. These findings demon- strate that tunicamycin not only inhibits N-linked glycosy- lation but also cleavage of Ac45, suggesting that N-linked glycosylation of intact Ac45 is necessary to allow its cleavage.

Fig. 7. NDGA allows immunoprecipitation of the C-terminal fragment of Ac45. NILs from black-adapted Xenopus were pulsed for 1 h with [35S]Met/Cys and subsequently chased for the indicated time periods in the presence of NDGA or BFA. Ac45 products were immunoprecip- itated with antibody 1311C. Precipitated proteins were resolved by SDS/PAGE and visualized by fluorography. Migration positions of intact and processed forms of Ac45 are indicated. Note that some of the immunoprecipitates contain nonspecifically bound 37-kDa POMC and (cid:25) 70 kDa PC2 (asterisks).

D I S C U S S I O N

Acidification of organelles is important for numerous intracellular processes. In the regulated secretory pathway, acidification is mainly required for the sorting of proteins and processing of prohormones [1]. The lumen of the organelles of the regulated secretory pathway gradually

As N-linked glycosylation of intact Ac45 precedes its cleavage, we wondered whether inhibition of N-glycosyla- tion by tunicamycin would affect the cleavage event. In the tunicamycin, newly synthesized Ac45 was absence of detected with the 1311N antiserum as the intact glycosylated (cid:25) 62–64-kDa form, with the mobility shifting to (cid:25) 61– 63 kDa during the subsequent chase period. The cleavage process caused the amount of the intact glycosylated form of Ac45 to decrease during the chase period (Fig. 8, lane 1– 3). In the presence of tunicamycin, Ac45 is immunoprecip- itated as a product of (cid:25) 46 kDa (Fig. 8, lane 4–6). The size of this (cid:25) 46-kDa unglycosylated product is similar to the size of intact Ac45 deglycosylated with N-glycosidase F indicating that tunicamycin prevents (Fig. 2,

lane 4–6),

The fate of Ac45 in the secretory pathway (Eur. J. Biochem. 269) 1851

(cid:211) FEBS 2002

processed to a C-terminal cleavage product of (cid:25) 40 kDa [34]. The results obtained in the present study allow us to propose the following more detailed model for the synthesis, processing and transport of Ac45 in Xenopus intermediate pituitary cells. Ac45 is synthesized as an intact protein of (cid:25) 46 kDa that is N-linked glycosylated to (cid:25) 62- and (cid:25) 64-kDa products. Trimming of the N-glycans in the ER gives rise to products of (cid:25) 61 and (cid:25) 63 kDa. As most oligomeric complexes are assembled in the ER [50], the association of Ac45 with the V-ATPase V0 sector may well be established already in this compartment. The intact glycosylated (cid:25) 61/63-kDa Ac45 protein was found to be cleaved to an (cid:25) 22-kDa N-terminal and a 42 to 44-kDa C-terminal product. The cleavage takes place in the ER or cis-Golgi, as it is not inhibited by BFA, and occurs before the cleavage products acquire EndoH resistance in the medial Golgi. When N-linked glycosylation was prevented by tunicamyin, the cleavage of Ac45 was inhibited, suggesting that the protein needs proper folding or associ- ation with the pump before cleavage can occur. However, we can not exclude the possibility that tunicamycin inter- fered with the activity of the elusive Ac45 cleavage enzyme. The extensive time between glycosylation and cleavage of Ac45 may indicate that its folding and assembly with the V-ATPase is a complex process. Following cleavage of intact glycosylated Ac45, both cleavage products pass the medial Golgi, as the single N-linked glycan on the N-terminal fragment and some of the N-glycans on the C-terminal fragment acquire resistance to EndoH. Subse- quently, the N-terminal cleavage fragment is degraded by a mechanism that is independent of the endosomal-lysosomal system, as the degradation process is not affected by drugs that disturb the acidification of these compartments or that inhibit hydrolytic lysosomal enzymes. The C-terminal cleavage fragment increases (cid:25) 1 kDa in size by an unknown type of modification and is likely transported to secretory granules, as in bovine Ac45 has been found to be associated with the chromaffin granular V-ATPase [30]. The bovine Ac45 C-terminal fragment (222 amino-acid residues) starts with Val209, suggesting that the intact molecule is proteo- lytically cleaved between Val208 and Val209 (numbering according to [34]) [31]. Remarkably, this presumptive cleavage site is not conserved in Ac45 of Xenopus and other species [34]. Therefore, we hypothesize that the site of cleavage in Ac45 is located in a more conserved region N-terminally of Val208/Val209, and that following cleavage the N-terminal portion of the C-product is subjected to exoproteolytic processing. Exoproteolytic trimming would explain why in Xenopus the size of the (deglycosylated) C-terminal cleavage fragment ((cid:25) 23 kDa) is smaller than expected on the basis of the sizes of intact (deglycosylated) Ac45 ((cid:25) 46 kDa) and the (deglycosylated) N-terminal cleavage product ((cid:25) 20 kDa). Exoproteolytic processing is not unusual, as it has also been described for several cathepsins and the light chain of myeloperoxidase [51], and for lactase-phlorizin hydrolase (LPH) [52].

acidifies from the ER to Golgi to secretory granules. Responsible for the acidification is the activity of the multimeric V-ATPase enzyme complex that translocates protons across membranes at the expense of ATP [26,48]. Several mechanisms have been proposed that may explain how the lumen of an organelle acquires its specific pH. The membranes of the organelles may differ in their permeability for protons, the composition or assembly state of the V-ATPase enzyme itself may vary between the different organelles, or organelle-specific proteins/factors may regu- late the V-ATPase. Evidence has been presented for all of these mechanisms (reviewed in [4,27,49]), suggesting that they may work simultaneously or in a cell type-specific manner. In the chromaffin cells of the bovine adrenal medulla, secretory granules have been found to contain a V-ATPase that is associated with the accessory subunit Ac45 [30]. This neuroendocrine-enriched subunit of (cid:25) 45 kDa may play a role in targeting or controlling the activity of the V-ATPase in the regulated secretory pathway [30,32]. Deglycosylation experiments and N-terminal sequencing of bovine Ac45 showed that the isolated protein is a proteolytically cleaved fragment [30,31]. We have recently shown that Xenopus Ac45 is synthesized as an subsequently N-glycosylated intact protein which is

Thus far, the only indication of a possible involvement of a cleavage enzyme in the regulation of V-ATPase activity comes from yeast mutant studies. A yeast mutant for the endoprotease Kex2p shows phenotypic character- istics similar to those of V-ATPase mutants, indicating that the Kex2p endoprotease is necessary for V-ATPase activity in vivo. A model has been proposed in which

Fig. 8. Tunicamycin inhibits the glycosylation and cleavage of intact Ac45. Lobes dissected from black-adapted Xenopus were preincubated overnight, pulsed for 1 h with [35S]Met/Cys and chased for the indi- cated time periods. The incubations were performed either in the absence or presence of 10 lgÆmL)1 tunicamycin. Radiolabeled proteins were immunoprecipitated from lobe extracts using antibody 1311N. Immunoprecipitates were resolved by SDS/PAGE and visualized by fluorography. Migration positions of glycosylated (61–64 kDa) and unglycosylated (46-kDa) intact Ac45, as well as the 22-kDa N-terminal Ac45 cleavage product are indicated.

1852 V. Th. G. Schoonderwoert et al. (Eur. J. Biochem. 269)

(cid:211) FEBS 2002

regulator of

refer

6. Kim, J.H., Lingwood, C.A., Williams, D.B., Furuya, W., Manolson, M.F. & Grinstein, S. (1996) Dynamic measurement of the pH of the Golgi complex in living cells using retrograde transport of the verotoxin receptor. J. Cell Biol. 134, 1387–1399. 7. Demaurex, N., Furuya, W., D’Souza, S., Bonifacino, J.S. & Grinstein, S. (1998) Mechanism of acidification of the trans-Golgi network (TGN). In situ measurements of pH using retrieval of TGN38 and furin from the cell surface. J. Biol. Chem. 273, 2044– 2051.

Kex2p would cleave a negative the V-ATPase, thereby activating the pump. Ac45 has been suggested to be this negative regulator [53] and, in a region just N-terminal of the N-terminus of the bovine C-terminal cleavage product, Ac45 contains a conserved sequence to (Arg183-Pro-Ser-Arg186; numbers Xenopus [34]); that could act as a recognition site for furin (consensus of furin cleavage site is RX(K/R)R [54]); [55], the vertebrate Kex2p ortholog [56,57]. However, it is unlikely that Ac45 represents the negative regulator, as yeast does not seem to contain an Ac45 ortholog [26]. Furthermore, Kex2p cleaves proproteins in the late Golgi, whereas we found that Xenopus Ac45 is cleaved in the ER or cis-Golgi.

8. Miesenbock, G., DeAngelis, D.A. & Rothman, J.E. (1998) Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394, 192–195.

9. Wu, M.M., Llopis, J., Adams, S., McCaffery, J.M., Kulomaa, M.S., Machen, T.E., Moore, H.P. & Tsien, R.Y. (2000) Organelle pH studies using targeted avidin and fluorescein-biotin. Chem. Biol. 7, 197–209.

10. Wu, M.M., Grabe, M., Adams, S., Tsien, R.Y., Moore, H.P. & Machen, T.E. (2001) Mechanisms of pH regulation in the regu- lated secretory pathway. J. Biol. Chem. 276, 33027–33035.

11. Wilson, D.W., Lewis, M.J. & Pelham, H.R. (1993) pH-dependent binding of KDEL to its receptor in vitro. J. Biol. Chem. 268, 7465– 7468.

12. Scheel, A.A. & Pelham, H.R. (1996) Purification and character- ization of the human KDEL receptor. Biochemistry 35, 10203– 10209.

The question arises concerning the possible role of the Ac45 cleavage event. Recently, a model for acidification in the regulated secretory pathway has been proposed [10]. In this model, the gradual decrease in the pH value of the organelles of the secretory pathway is attributed to a decrease in the proton permeability from the ER to the mature secretory granules, concomitant with a gradual increase in the number of active V-ATPases from the ER to the Golgi. How the number of active H+-pumps increases from the ER to the Golgi is not clear from this model. Ac45 could be a key player in this process. Intact glycosylated Ac45 may interact with the V-ATPase and thereby keeping the pump inactive in the ER. Following Ac45 cleavage, the V-ATPase would become active, whereby the cleavage may have allowed the dissociation of the (inhibiting) N-terminal cleavage fragment.

13. Yilla, M., Tan, A., Ito, K., Miwa, K. & Ploegh, H.L. (1993) Involvement of the vacuolar H(+)-ATPases in the secretory pathway of HepG2 cells. J. Biol. Chem. 268, 19092–19100. 14. Xu, H. & Shields, D. (1994) Prosomatostatin processing in permeabilized cells. Endoproteolytic cleavage is mediated by a vacuolar ATPase that generates an acidic pH in the trans-Golgi network. J. Biol. Chem. 269, 22875–22881. 15. Burgess, T.L. & Kelly, R.B. (1987) Constitutive and regulated secretion of proteins. Annu. Rev. Cell Biol. 3, 243–293.

Altogether, we conclude that N-linked glycosylated intact Ac45 is cleaved to an (cid:25) 22-kDa N-terminal and a 42–44- kDa C-terminal cleavage fragment in the ER or cis-Golgi, where activation of the V-ATPase is necessary. Following passage through the Golgi, the N-terminal fragment is degraded and, together with the V-ATPase, the C-terminal fragment is targeted to secretory granules.

16. Chanat, E. & Huttner, W.B. (1991) Milieu-induced, selective aggregation of regulated secretory proteins in the trans-Golgi network. J. Cell Biol. 115, 1505–1519.

17. Kim, T., Tao-Cheng, J., Eiden, L.E. & Loh, Y.P. (2001) Chro- mogranin A, an (cid:212)on/off(cid:213) switch controlling dense-core secretory granule biogenesis. Cell 106, 499–509.

A C K N O W L E D G E M E N T S

18. Cool, D.R., Normant, E., Shen, F.S., Chen, H.C., Pannell, L., Zhang, Y. & Loh, Y.P. (1997) Carboxypeptidase E is a regulated secretory pathway sorting receptor: genetic obliteration leads to endocrine disorders in Cpe (fat) mice. Cell 88, 73–83.

19. Urbe, S., Dittie, A.S. & Tooze, S.A. (1997) pH-dependent pro- cessing of secretogranin II by the endopeptidase PC2 in isolated immature secretory granules. Biochem. J. 321, 65–74.

We would like to thank Ron Engels for taking care of Xenopus, and Peter Cruijsen for technical assistance. This work was supported by grant 805-33-212 from the Netherlands Organization for Scientific Research-Earth and Life Sciences (NWO-ALW), and by European Union-Training and Mobility Researchers network ERBFMR- XCT960023. 20. Laine, J. & Lebel, D. (1999) E(cid:129)cient binding of regulated secre- tory protein aggregates to membrane phospholipids at acidic pH. Biochem. J. 338, 289–294.

R E F E R E N C E S

21. Kuliawat, R. & Arvan, P. (1994) Distinct molecular mechanisms for protein sorting within immature secretory granules of pan- creatic beta-cells. J. Cell Biol. 126, 77–86.

1. Mellman, I., Fuchs, R. & Helenius, A. (1986) Acidification of the endocytic and exocytic pathways. Annu. Rev. Biochem. 55, 663– 700.

2. Arai, K., Shimaya, A., Hiratani, N. & Ohkuma, S. (1993) Puri- fication and characterization of lysosomal H(+)-ATPase. An anion-sensitive v-type H(+)-ATPase from rat liver lysosomes. J. Biol. Chem. 268, 5649–5660. 22. Tooze, J. & Tooze, S.A. (1986) Clathrin-coated vesicular transport of secretory proteins during the formation of ACTH-containing secretory granules in AtT20 cells. J. Cell Biol. 103, 839–850. 23. Dittie, A.S., Hajibagheri, N. & Tooze, S.A. (1996) The AP-1 adaptor complex binds to immature secretory granules from PC12 cells, and is regulated by ADP-ribosylation factor. J. Cell Biol. 132, 523–536.

3. Clague, M.J., Urbe, S., Aniento, F. & Gruenberg, J. (1994) Vacuolar ATPase activity is required for endosomal carrier vesicle formation. J. Biol. Chem. 269, 21–24. 4. Schoonderwoert, V.T.G. & Martens, G.J.M.

24. Klumperman, J., Kuliawat, R., Gri(cid:129)th, J.M., Geuze, H.J. & Arvan, P. (1998) Mannose 6-phosphate receptors are sorted from immature secretory granules via adaptor protein AP-1, clathrin, and syntaxin 6-positive vesicles. J. Cell Biol. 141, 359–371. 25. Halban, P.A. & Irminger, J.C. (1994) Sorting and processing of secretory proteins. Biochem. J. 299, 1–18. (2001) Proton pumping in the secretory pathway. J. Membr. Biol. 182, 159–169. 5. Seksek, O., Biwersi, J. & Verkman, A.S. (1995) Direct measure- ment of trans-Golgi pH in living cells and regulation by second messengers. J. Biol. Chem. 270, 4967–4970.

The fate of Ac45 in the secretory pathway (Eur. J. Biochem. 269) 1853

(cid:211) FEBS 2002

42. Tartakoff, A.M. (1983) Perturbation of vesicular tra(cid:129)c with the carboxylic ionophore monensin. Cell 32, 1026–1028. 26. Nelson, N. & Harvey, W.R. (1999) Vacuolar and plasma mem- brane proton-adenosinetriphosphatases. Physiol. Rev. 79, 361– 385.

43. Ascoli, M. (1979) Inhibition of the degradation of receptor-bound human choriogonadotropin by leupeptin. Biochim. Biophys. Acta. 586, 608–614. 27. Forgac, M. (2000) Structure, mechanism and regulation of the clathrin-coated vesicle and yeast vacuolar H(+)-ATPases. J. Exp. Biol. 203, 71–80.

44. Doms, R.W., Russ, G. & Yewdell, J.W. (1989) Brefeldin A redistributes resident and itinerant Golgi proteins to the endo- plasmic reticulum. J. Cell Biol. 109, 61–72. 28. Myers, M. & Forgac, M. (1993) Assembly of the peripheral domain of the bovine vacuolar H(+)-adenosine triphosphatase. J. Cell Physiol. 156, 35–42.

45. Donaldson, J.G., Lippincott, S.J., Bloom, G.S., Kreis, T.E. & Klausner, R.D. (1990) Dissociation of a 110-kD peripheral membrane protein from the Golgi apparatus is an early event in brefeldin A action. J. Cell Biol. 111, 2295–2306. 29. Graham, L.A., Hill, K.J. & Stevens, T.H. (1998) Assembly of the yeast vacuolar H+-ATPase occurs in the endoplasmic reticulum and requires a Vma12p/Vma22p assembly complex. J. Cell Biol. 142, 39–49.

30. Supek, F., Supekova, L., Mandiyan, S., Pan, Y.C., Nelson, H. & Nelson, N. (1994) A novel accessory subunit for vacuolar H (+) - ATPase from chroma(cid:129)n granules. J. Biol. Chem. 269, 24102– 24106. 46. Orci, L., Tagaya, M., Amherdt, M., Perrelet, A., Donaldson, J.G., Lippincott-Schwartz, J., Klausner, R.D. & Rothman, J.E. (1991) Brefeldin A, a drug that blocks secretion, prevents the assembly of non-clathrin-coated buds on Golgi cisternae. Cell 64, 1183–1195.

31. Getlawi, F., Laslop, A., Schagger, H., Ludwig, J., Haywood, J. & Apps, D. (1996) Chroma(cid:129)n granule membrane glycoprotein IV is identical with Ac45, a membrane-integral subunit of the granule’s H+-ATPase. Neurosci. Lett. 219, 13–16. 47. Fujiwara, T., Takami, N., Misumi, Y. & Ikehara, Y. (1998) Nordihydroguaiaretic acid blocks protein transport in the secre- tory pathway causing redistribution of Golgi proteins into the endoplasmic reticulum. J. Biol. Chem. 273, 3068–3075.

32. Holthuis, J.C., Jansen, E.J., van-Riel, M.C. & Martens, G.J. (1995) Molecular probing of the secretory pathway in peptide hormone-producing cells. J. Cell Sci. 108, 3295–3305. 48. Stevens, T.H. & Forgac, M. (1997) Structure, function and reg- ulation of the vacuolar (H+) -ATPase. Annu. Rev. Cell Dev. Biol. 13, 779–808.

49. Futai, M., Oka, T., Sun-Wada, G., Moriyama, Y., Kanazawa, H. & Wada, Y. (2000) Luminal acidification of diverse organelles by V-ATPase in animal cells. J. Exp. Biol. 203, 107–116. 50. Hurtley, S.M. & Helenius, A. (1989) Protein oligomerization in the 33. Jenks, B.G.J.L.H., Martens, G.J.M. & Roubos, E.W. (1993) Adaptation physiology: the functioning of pituitary melanotrope cells during background adaptation of the amphibian Xenopus laevis. Zoo. Sci. 10, 1–11. endoplasmic reticulum. Annu. Rev. Cell Biol. 5, 277–307.

34. Holthuis, J.C., Jansen, E.J., Schoonderwoert, V.T., Burbach, J.P. & Martens, G.J. (1999) Biosynthesis of the vacuolar H+-ATPase accessory subunit Ac45 in Xenopus pituitary. Eur. J. Biochem. 262, 484–491. 51. Hasilik, A. (1992) The early and late processing of lysosomal enzymes: proteolysis and compartmentation. Experientia 48, 130–151.

52. Mesonero, J.E., Gloor, S.M. & Semenza, G. (1998) Processing of human intestinal prolactase to an intermediate form by furin or by a furin-like proprotein convertase. J. Biol. Chem. 273, 29430– 29436.

35. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 36. Maley, F., Trimble, R.B., Tarentino, A.L. & Plummer, T.H. Jr (1989) Characterization of glycoproteins and their associated oligosaccharides through the use of endoglycosidases. Anal. Biochem. 180, 195–204.

53. Oluwatosin, Y.E. & Kane, P.M. (1998) Mutations in the yeast KEX2 gene cause a Vma(–)-like phenotype: a possible role for the KEX2 endoprotease in vacuolar acidification. Mol. Cell. Biol. 18, 1534–1543.

37. Tarentino, A.L., Plummer, T.H. Jr & Maley, F. (1974) The release of intact oligosaccharides from specific glycoproteins by endo- beta-N-acetylglucosaminidase H. J. Biol. Chem. 249, 818–824. 38. Plummer, T.H. Jr & Tarentino, A.L. (1981) Facile cleavage of complex oligosaccharides from glycopeptides by almond emulsin peptide: N-glycosidase. J. Biol. Chem. 256, 10243–10246. 54. Molloy, S.S., Bresnahan, P.A., Leppla, S.H., Klimpel, K.R. & Thomas, G. (1992) Human furin is a calcium-dependent serine endoprotease that recognizes the sequence Arg-X-X-Arg and e(cid:129)ciently cleaves anthrax toxin protective antigen. J. Biol. Chem. 267, 16396–16402.

55. Matthews, D.J., Goodman, L.J., Gorman, C.M. & Wells, J.A. (1994) A survey of furin substrate specificity using substrate phage display. Protein Sci. 3, 1197–1205.

39. Lippincott-Schwartz, J., Yuan, L., Tipper, C., Amherdt, M., Orci, L. & Klausner, R.D. (1991) Brefeldin A’s effects on endosomes, lysosomes, and the TGN suggest a general mechanism for regu- lating organelle structure and membrane tra(cid:129)c. Cell 67, 601–616. 40. Thorens, B., Gerard, N. & Deriaz, N. (1993) GLUT2 surface expression and intracellular transport via the constitutive pathway in pancreatic beta cells and insulinoma: evidence for a block in trans-Golgi network exit by brefeldin A. J. Cell Biol. 123, 1687– 1694. 56. Roebroek, A.J., Schalken, J.A., Leunissen, J.A., Onnekink, C., Bloemers, H.P. & Van de Ven, W.J. (1986) Evolutionary con- served close linkage of the c-fes/fps proto-oncogene and genetic sequences encoding a receptor-like protein. EMBO J. 5, 2197– 2202.

57. Barr, P.J., Mason, O.B., Landsberg, K.E., Wong, P.A., Kiefer, M.C. & Brake, A.J. (1991) cDNA and gene structure for a human subtilisin-like protease with cleavage specificity for paired basic amino acid residues. DNA Cell Biol. 10, 319–328. 41. Schoonderwoert, V.T., Holthuis, J.C., Tanaka, S., Tooze, S.A. & Martens, G.J. (2000) Inhibition of the vacuolar H+-ATPase per- turbs the transport, sorting, processing and release of regulated secretory proteins. Eur. J. Biochem. 267, 5646–5654.