Ca2+-ATPases of Saccharomyces cerevisiae: diversity and possible role in protein sorting

FEMS Microbiology Letters 162 (1998) 83^91

Ca2‡ -ATPases of Saccharomyces cerevisiae: diversity and possible role in protein sorting Lev A. Okorokov

b;c;

*, Ludwig Lehle

a

a

Lehrstuhl fuër Zellbiologie und P£anzenphysiologie, Universitat Regensburg, 93040 Regensburg, Germany Laboratorio de Fisiologia e Bioquimica de Microorganismos, Centro de Biocieências e Biotechnologia, Universidade Estadual do Norte Fluminense, 28015-620, Campos dos Goytacazes, Avenida Alberto Lamego, 2000, Rio de Janeiro, Brazil c Institute of Biochemistry and Physiology of Microorganisms, Russian Academy of Sciences, Pushchino, Moscow Region, Russia b

Received 5 January 1998 ; revised 14 February 1998; accepted 3 March 1998

Abstract The PMR1 gene of Saccharomyces cerevisiae is thought to encode a putative Ca2‡ -ATPase [1]. Membranes isolated from wild-type cells and from pmr1 null mutant of S. cerevisiae were fractionated on sucrose density gradients. In the pmr1 mutant we found a decrease in activity of the P-type ATPase and of ATP-dependent, protonophore-insensitive Ca2‡ transport in light membranes, that comigrate with the Golgi marker GDPase. We conclude that the product of the PMR1 gene (Pmr1p) is indeed a Ca2‡ -ATPase of the Golgi and Golgi-like membranes. Surprisingly, the pmr1 null mutation abolished Ca2‡ -ATPase activity in Golgi and/or Golgi-like membranes only to 50% under conditions where they are separated from vacuolar membranes. This indicates that an additional Ca2‡ -ATPase is localized in Golgi and/or Golgi-like membranes. Moreover, a third Ca2‡ -ATPase is found in the ER and ER-like membranes. The data are consistent with the assumption that these Ca2‡ -ATPases are encoded by gene(s) different from PMR1. Disruption of PMR1 Ca2‡ -ATPase causes significant redistribution of enzyme activities and of total protein in compartments of the secretory pathway. A decrease in activity is observed for three integral membrane proteins: NADPH cytochrome c reductase, dolichyl phosphate mannose synthase, and Ca2‡ -ATPase, and also for total protein in Golgi, Golgi-like compartments and in vacuoles, whereas a corresponding increase of these activities is observed in endoplasmic reticulum and endoplasmic reticulum-like membranes. We assume that Ca2‡ -ATPases and sufficient Ca2‡ gradients across the organellar membranes are important for the correct sorting of proteins to the various compartments of the secretory apparatus. z 1998 Published by Elsevier Science B.V. Keywords : Ca2‡ -adenosine triphosphatase; Ca2‡ storage compartment; Protein sorting; Secretory pathway ; Saccharomyces cerevisiae

1. Introduction Ca2‡ is an important regulatory ion in all eukaryotes. Two types of Ca2‡ transporters have been described in yeast cells. (i) A Ca2‡ /H‡ antiporter * Corresponding author (b). Tel.: +55 (247) 263-724; Fax: +55 (247) 263-720 or 230160; E-mail: [email protected]

has been found in vacuolar membranes [2^4]. It is driven by the di¡erence of electrochemical H‡ potentials created by a V1 V0 H‡ -ATPase [2^5]. At high concentrations of Ca2‡ this antiporter operates with a rate comparable to that of the V1 V0 H‡ -ATPase and converts vWH‡ to vWCa2‡ [3,4]. However, its a¤nity to Ca2‡ is probably not su¤cient to regulate cytosolic free Ca2‡ at low, physiological levels, since

0378-1097 / 98 / $19.00 ß 1998 Published by Elsevier Science B.V. All rights reserved. PII S 0 3 7 8 - 1 0 9 7 ( 9 8 ) 0 0 1 0 6 - 2

FEMSLE 8116 27-4-98

84

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

its apparent Km is around 40 WM [3]. (ii) A reticulum-type Ca2‡ -ATPase is found in intracellular Golgi-like membranes [6]. There is evidence that Ca2‡ ATPases of the same type also occur in other compartments of the secretory apparatus of Saccharomyces cerevisiae [7,8]. Because of their high a¤nity to Ca2‡ (Km = 110 nM) they could be involved in the ¢ne regulation of cytosolic Ca2‡ [6]. Ca2‡ -ATPase can be distinguished from the Ca2‡ /H‡ antiporter, since Ca2‡ -ATPase is blocked by orthovanadate [6,8] while the antiporter is inhibited by protonophores [2^6]. A yeast putative Ca2‡ -ATPase is encoded by the PMR1 gene [1]. It was reported that Pmr1p is most likely localized to a novel Golgi-like compartment or even to various compartments of the secretory pathway [9]. Vacuoles, the endoplasmic reticulum (ER) and plasma membrane were de¢nitely excluded as possible candidates [9]. Direct biochemical proof that Pmrp1 is a Ca2‡ -ATPase was lacking. While this article was in preparation, a ¢rst report appeared presenting e¡orts to biochemically characterize Pmr1p and its subcellular localization [10] (see Section 4). Our earlier attempt to demonstrate a di¡erence in Ca2‡ -ATPase activity of total membrane preparations isolated from PMR1 wild-type S. cerevisiae and a pmr1 null mutant failed [6]. The present work aims to investigate whether fractionation of total membranes into subcellular compartments can reveal di¡erences in Ca2‡ transport activities between these two strains. We show now that Pmr1p is indeed a Ca2‡ -ATPase of the Golgi and/or Golgilike membranes. Furthermore, we also demonstrate the occurrence of two additional Ca2‡ -ATPases located in the ER and the Golgi and/or Golgi-like membranes, respectively. Finally, it is observed that disruption of PMR1 causes alterations in the distribution of proteins within the compartments of the secretory machinery.

ura3-52; AA274: ade2, his3-200, leu2-3,112, lys2-201, ura3-52, pmr1-1: :LEU2. Standard medium (YEPD) containing 1% yeast extract, 2% bactopeptone and 2% glucose was used for cultivation.

2. Materials and methods

2.3. Marker enzymes

2.1. Yeast strains and culture conditions

Marker enzymes were tested as described: dolichyl phosphate mannose synthase [11] and NADPH cytochrome c reductase for ER [12], GDPase for Golgi [13] and ATPase inhibited by 100 WM orthovanadate

The genotype of the yeast strains used is as follows: AA255: ade2, his3-200, leu 2-3,112, lys2-201,

2.2. Isolation of spheroplasts (SP) and membranes SP were isolated as described [7] with some modi¢cations. Cells were suspended in SP bu¡er (1.4 M sorbitol in 50 mM Tris-HCl, pH 7.4 containing 30 mM mercaptoethanol). Lyophilized stomach juice from Helix pomatia was used as a lytic enzyme. For 1 g of cells (wet weight) 5 ml of SP bu¡er and 50 mg of enzyme were added. To monitor formation of SP 10 Wl of the suspension of SP and cells were injected in 1 ml of water, mixed and the optical density at 578 nm was determined after 2 min. After 40^ 50 min incubation at 37³C with occasional stirring the suspension was cooled in ice, EDTA and PMSF were added to a ¢nal concentration of 1 mM. 10^20ml aliquots of SP suspension were layered onto 15^ 20 ml of SP bu¡er containing 1 mM PMSF but without mercaptoethanol. SP were collected by 5 min centrifugation (1500Ug) and resuspended in 10^30 ml of lysis bu¡er (365 mM sucrose, 20 mM MOPS-Na, pH 7.4, 1 mM DTT, 1 mM PMSF, 1 mM benzamidine and 1 Wl of a cocktail containing antipain, aprotinin, chymostatin, leupeptin and pepstatin (each 1 mg ml31 ) for each 1 ml of the lysis bu¡er). After lysis of SP with a Dounce homogenizer (20 strokes), cell walls, unlysed SP and cells were removed by 5 min centrifugation (1500Ug). The total membrane preparation was then sedimented from the supernatant by centrifugation (15 min 40 000Ug). Membranes were resuspended in lysis bu¡er and layered (0.8^1.0 ml) onto a sucrose density gradient (see details in the ¢gure legends). Sucrose solutions were prepared in 10 mM HEPES-Na (pH 7.2). After centrifugation for 165 min at 174 000Ug, fractions were collected from the top and analyzed or frozen at 320³C.

FEMSLE 8116 27-4-98

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

for plasma membrane [14]. ATPase was measured in the presence of 5 mM sodium azide, 50 mM potassium nitrate and 0.2 mM ammonium molybdate to block mitochondrial and vacuolar ATPases as well as unspeci¢c phosphatases, correspondingly. The difference of ATPase activity measured in the absence and the presence of 100 WM orthovanadate was taken as the activity of H‡ -ATPase of plasma membranes. Membrane-bound pyrophosphatase, a marker of vacuolar membrane [15], was measured in the presence of 100 mM sorbitol with 1 mM pyrophosphate and 2.5 mM MgSO4 at pH 7.2 and 30³C [15]. 2.4.

45

Ca2 + uptake

The standard incubation mixture contained in a total volume of 200 Wl: 10 mM HEPES-K pH 7.2; 150 mM KCl, 5 mM magnesium acetate, 1 mM sodium ATP, 9.8 WM EGTA, 10.4 WM CaCl2 , 0.5 WCi 45 Ca2‡ (speci¢c activity 10 Ci g31 ; NEN-Dupont) and 1 WM FCCP to block Ca2‡ /H‡ antiporter [2^ 6]. The reaction was initiated by adding 10^30 Wl of suspension of membrane vesicles (3^20 Wg protein). After 10 min incubation at 30³C 180-Wl aliquots were injected into 10 ml of stop solution (150 mM KCl, 5 mM Mg(CH3 COO)2 , HEPES-K, pH 7.2, ¢ltered on 0.45-mm nitrocellulose ¢lters (Millipore) and washed with 10 ml of the same bu¡er. Radioactivity retained on the ¢lters was measured by scintillation counting. Protein content was determined according to Bradford using serum albumin as standard.

3. Results We have previously reported that Ca2‡ transport activity does not di¡er signi¢cantly in total membranes isolated from a PMR1 strain as compared to a pmr1 null mutant [6]. However, fractionation of the membranes on a continuous sucrose density gradient revealed signi¢cant di¡erences in Ca2‡ transport activity in both light and heavy membrane fractions (Fig. 1A). ATP-dependent and FCCP-insensitive Ca2‡ transport activity in pmr1 membranes was decreased in fractions 1^18 and is partly compensated by a corresponding increase in fractions 24^44. In the absence of ATP, a low binding of

85

Ca2‡ was observed in wild-type (not shown) and mutant membranes (Fig. 1A). The decrease in Ca2‡ transport activity was observed mainly in fractions with high GDPase activity, a marker for the Golgi compartment (Fig. 1B, fractions 5^19). It should be mentioned that in fractions 2^5 vacuolar and secretory vesicles would also migrate (unpublished data). The increase of Ca2‡ transport activity is found in pmr1 membranes (Fig. 1A), which comigrate with the ER marker dolichyl phosphate mannose synthase (Fig. 1C; fractions 24^40). We have recently shown that the Ca2‡ uptake measured in such conditions is due to the activity of Ca2‡ -ATPase(s) [6^8]. In order to investigate the e¡ect of the pmr1 null mutation on the Ca2‡ ATPase(s), we also measured the activity of P-type ATPases in the gradient. For this purpose mitochondrial and vacuolar H‡ -ATPases and unspeci¢c phosphates were inhibited by azide, nitrate and molybdate, respectively. The largest ATPase activity in membranes from the wild-type strain was found in fractions 20^41 with the maximum in fraction 32 (Fig. 1D). 90% of this activity was inhibited by 100 WM orthovanadate, indicating that the ATPase activity is mainly due to the plasma membrane H‡ ATPase. In agreement with [11] the plasma membrane is only partly separated from the ER (Fig. 1C,D). P-type ATPase activity was also found in light membranes in both wild-type and pmr1 mutant (Fig. 1D, fractions 1^17). In the latter, both P-type ATPase and Ca2‡ transport activity were decreased in a similar fashion. This ¢nding further supports the conclusion that Pmr1p is indeed a Ca2‡ -ATPase and that it is localized to the Golgi and/or Golgi-like membranes. It is noteworthy that Ca2‡ transport activity in these membranes did not fully disappear in the pmr1 mutant (Fig. 1A). The residual 50% activity may indicate that these membranes are equipped with an additional Ca2‡ -ATPase encoded by a gene di¡erent from PMR1. On the other hand Ca2‡ transport activity increased in the ER and the ER-like membranes as a result of the pmr1 null mutation. Since ATPase activity insensitive towards 100 WM orthovanadate is simultaneously increased in the same region, one may assume that this activity re£ects a reticulum type of Ca2‡ -ATPase [6^8] rather than plasma membrane H‡ - or Ca2‡ -ATPase [14,16].

FEMSLE 8116 27-4-98

86

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

Fig. 1. Fractionation of yeast membranes in a continuous density gradient of sucrose. Total membranes isolated from yeast harvested at the beginning of the stationary phase of growth were applied onto a 30^56% continuous sucrose gradient and centrifugated for 2 h 45 min at 174 000Ug in an SW-41 rotor. Samples were collected from the top with an automatic fractionation system. A: 45 Ca2‡ uptake insensitive to 1 WM FCCP was determined after 10 min incubation under standard conditions. 45 Ca2‡ binding by membranes was measured in the absence of ATP. B : GDPase activity, marker of Golgi. C: Dolichyl phosphate mannose synthase, marker of ER. D: ATPase measured in the presence of 5 mM azide, 50 mM nitrate, 0.1 mM molybdate with 100 WM orthovanadate and without it. E: Protein distribution between di¡erent membranes. All data were normalized for 1 mg of protein loaded onto the gradient.

This conclusion is also supported by the absence of the Ca2‡ translocase activity of this enzyme type in our membrane preparations ([8] and unpublished results). We conclude that the increase of Ca2‡ transport activity in the ER and ER-like membranes of the pmr1 mutant is due to a Ca2‡ -ATPase di¡erent from the PMR1 enzyme. The pmr1 mutation a¡ects not only Ca2‡ -ATPase distribution but also that of other proteins of the secretory pathway. In the case of dolichyl phosphate mannose synthase an increase of its activity in the ER region and a simultaneous decrease in the Golgi region were observed (Fig. 1B). In the case of GDPase an increase of the activity takes place in the majority of the gradient fractions of the pmr1 mutant with the exception of fractions 1^6 (Fig. 1B). GDPase increases mainly in the ER and in compartments migrating between ER and Golgi in the gradient. We assume that these changes are consequences of protein mis-sorting and/or perturbation of the secretory apparatus caused by disruption of the PMR1 Ca2‡ -ATPase (see Section 4). Changes were even manifest at the level of the total protein:

FEMSLE 8116 27-4-98

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

87

Fig. 2. Time course of SP formation from the wild-type strain and pmr1 mutant. After addition of lytic enzyme cell suspension was incubated at 37³C and 10-Wl samples were diluted in 1 ml of water. OD578 was measured after 2 min and its decrease was taken as an indication for the lysis of SP.

the amount of protein in the mutant was lower in light membranes (fractions 1^17), while it was increased in denser membranes (fractions 30^44) (Fig. 1E). An additional interesting ¢nding is the faster formation of spheroplasts from the pmr1 mutant cells in comparison to wild-type (Fig. 2). This is in agreement with the ¢nding of truncated mannosylation of proteins in the pmr1 mutant [1] and presumably also with a mis-sorting of some proteins of the cell wall, leading to an altered cell wall. It has been shown for X2180 wild-type S. cerevisiae that a sucrose multi-step gradient instead of a linear one resolves Ca2‡ transport activity into several distinct peaks [17]. Therefore, in an e¡ort to achieve a better separation of membrane vesicles derived from various compartments we applied such a type of gradient also for the analysis of the strains used in this study. As shown in Fig. 3A, Ca2‡ transport activity was detectable in several peaks of the gradient in agreement with our previous ¢ndings for the S. cerevisiae X-2180 membranes [17]. The activity decreased in four peaks of light membranes of pmr1 mutant (fractions 1^6, 7^13, 14^18 and 18^22), while it simultaneously increased in three peaks of denser membranes (fractions 27^33, 33^38 and 38^42). This was also true for NADPH cytochrome c reductase (Fig. 3B). GDPase activity increased in the majority of peaks with an exception of fractions 1^6 and fractions 37^44 (Fig. 3C). Interestingly, this gradient technique revealed that the yeast Golgi is not a ho-

Fig. 3. Fractionation of yeast membranes in multi-step sucrose density gradient. Total membranes isolated from yeast collected at the beginning of the stationary phase of growth were applied onto a multi-step sucrose gradient (56, 52, 48, 45, 42, 39%, each step of 1.33 ml and 36, 33, 30%, each step of 1 ml. Each step received 2 Wl of polypeptide inhibitors of proteases). After centrifugation for 2 h 45 min at 174 000Ug in an SW41 rotor samples were collected from the top of the gradient. A: 45 Ca2‡ uptake (10 min) insensitive to 1 WM FCCP. B: NADPH cytochrome c reductase. C: GDPase activity.

mogeneous entity. The main peak of GDPase activity (Fig. 3C, fractions 6^18) cofractionates with two peaks of Ca2‡ transport activity (fractions 7^13 and

FEMSLE 8116 27-4-98

88

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

2^7, 10, 11 and 21^23 showed a small increase of GDPase activity. Thus, the uncommon distribution of GDPase, a luminal Golgi enzyme, is only observed in the stationary growth phase of the pmr1 mutant while membrane enzymes, Ca2‡ -ATPase and NADPH cytochrome c reductase show an uncommon distribution irrespective of the growth phase. Fig. 4B shows that the vacuolar membrane enzyme PPi ase [15] peaks together with the lightest membrane fractions (maximum in fraction 4) and is separated from GDPase. It is noteworthy that the PPi ase activity of fractions 7, 8 and 11, 12 amounted to 12% and 8%, respectively, of the activity of fraction 4, while the residual activity of Ca2‡ -ATPase in those fractions is the same as or even higher than in fraction 4. This can be regarded as evidence that a main part of the residual activity of Ca2‡ -ATPase is of Golgi and/or Golgi-like membrane origin rather than a vacuolar contamination.

Fig. 4. Fractionation of yeast membranes in multi-step density gradient of sucrose. Total membranes were isolated from yeast collected in the middle of the exponential stage of growth and fractionated as in Fig. 3. A: 45 Ca uptake (10 min) insensitive to 1 WM FCCP. B : GDPase and PPi ase activity.

14^18). One may suppose that membrane fractions 7^13 and 14^18 derive from di¡erent Golgi subcompartments of wild-type cells and that membrane fractions 18^22 represent a pre-Golgi compartment. The results presented in Figs. 1 and 3 were obtained with membranes isolated from the cells harvested at the beginning of the stationary phase of growth. We next asked whether the ¢ndings described above depend on the growth stage of the cells. Fractionation of membranes isolated from yeast harvested in the middle of the exponential growth phase revealed a similar pattern, namely the increase of Ca2‡ transport activity in the ER and a simultaneous decrease of the activity in Golgi and/or Golgi-like membranes (Fig. 4A). That is also the case for NADPH cytochrome c reductase (not shown). However, GDPase activity of the pmr1 membranes did not change signi¢cantly compared to wild-type membranes (Fig. 4B). Only fractions

4. Discussion A comparative biochemical analysis of membranes of the pmr1 null mutant and the isogenic wild-type strain makes it possible to conclude the following four points. First, PMR1 encodes a Ca2‡ -ATPase of the Golgi and/or Golgi-like membranes. Secondly, the presence of an additional Ca2‡ -ATPase in the Golgi and/or Golgi-like membranes is demonstrated. Thirdly, evidence for the existence of a Ca2‡ -ATPase in the ER and/or ER-like membranes was obtained. Fourthly, the importance of the PMR1 Ca2‡ ATPase for correct addressing of proteins to organelles of the secretory pathway is illustrated. Our ¢rst conclusion follows from the ¢nding that ATP-dependent, FCCP-insensitive Ca2‡ transport activity and vanadate-sensitive ATPase decreased in a similar fashion in the Golgi and/or Golgi-like membranes of the pmr1 null mutant compared to wild-type (Figs. 1, 3 and 4). Our data on the localization of the PMR1 Ca2‡ -ATPase in the Golgi and/or Golgi-related membranes are in agreement with the ¢nding of Antebi and Fink [9], who used anti-Pmr1p antibodies and concluded that Pmr1p is localized in a novel subcompartment of the Golgi, di¡erent from the KEX2 and SEC7 subcompartments [19^22]. The

FEMSLE 8116 27-4-98

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

heterogeneity of light membranes with respect to PMR1 Ca2‡ -ATPase (Figs. 1, 3 and 4) can be taken as evidence that several (sub)compartments of the Golgi and/or Golgi-like membranes are serviced by this pump. This interpretation is in agreement with the concept that the yeast Golgi is not a homogeneous entity [18^21] and with the assumption that a proportion of Pmr1p resides in multiple secretory compartments [9]. The di¡erent ratios of the Ca2‡ transport, GDPase and NADPH cytochrome c reductase activities in di¡erent peaks of light membranes (Figs. 1, 3 and 4) can be taken as an indication that these membranes derive from various (sub)compartments. Since we measured Ca2‡ transport activity, our data do not allow us to distinguish between compartments which have lost the PMR1 Ca2‡ -ATPase (for example the Golgi) and those that are located downstream in a secretory pathway and probably did not receive all copies of their Ca2‡ ATPases (for example, vacuoles) and therefore show a decrease of the activity as a result of the protein mis-sorting (see below). Altogether our data allow us to assume that the PMR1 Ca2‡ -ATPase serves in at least two subcompartments of Golgi (Fig. 3C; fractions 7^13 and 14^17) and `pre-Golgi', presumably an ER/Golgi intermediate compartment (fractions 18^22). It has been reported recently that an overexpression of PMR1 leads to the appearance of Ca2‡ ATPase activity in the Golgi membranes [10], suggesting that Pmr1p has a potential property to pump Ca2‡ . The authors stressed that PMR1 is indeed a Golgi-speci¢c Ca2‡ pump. However, Western blot analysis shows, in contradiction to the data of Antebi and Fink [9], localization of Pmr1p to the vacuole instead of to the Golgi (see Fig. 2C in [10]). This discrepancy needs further clari¢cation. Golgi-localized Ca2‡ -ATPase activity in wild-type S. cerevisiae, as demonstrated here, was not reported in [10]. A surprising ¢nding of the present work is the residual Ca2‡ -ATPase activity detectable in the Golgi after disrupting PMR1 (Fig. 1A, Figs. 3A and 4A). At present it cannot be distinguished whether this activity is localized to the same Golgi membranes containing the PMR1 Ca2‡ -ATPase or to other membranes comigrating with the Golgi. This question needs further investigation. However, our data show de¢nitely that most of the residual

89

Ca2‡ -ATPase activity found in the Golgi region of the gradient cannot be due to contamination by vacuolar membranes (Fig. 4A). Therefore, a gene encoding this enzyme activity has to be identi¢ed. The ¢nding of a third Ca2‡ -ATPase localized to the ER and/or ER-like membranes is noteworthy. The increase of Ca2‡ transport activity in the ER of the pmr1 mutant (Fig. 1A, Figs. 3A and 4A) may signify that the enzyme is upregulated upon deletion of the PMR1 Ca2‡ -ATPase. The increase could also be a consequence of mis-addressing of Ca2‡ -ATPases of the Golgi, secretory vesicles, vacuoles and so on (see below). In any case it shows that the main Ca2‡ transport activity of the ER and ERlike membranes is due to Ca2‡ -ATPase(s), encoded by a gene(s) di¡erent from PMR1. Taking all the results as a whole one can conclude that yeast intracellular membranes are equipped with at least four Ca2‡ -ATPases. Besides the enzyme of the ER and two Ca2‡ -ATPases of the Golgi and/or Golgi-like membranes, the PMC1 Ca2‡ -ATPase of vacuoles [22] should be mentioned (see also Fig. 1A, Figs. 3A and 4A; fractions 1^5). The conclusion is in agreement with a recent ¢nding that Ca2‡ -ATPase activity is a characteristic of various compartments of S. cerevisiae X-2180. [17]. Inhibitor analysis also made it possible to assume the existence of four Ca2‡ -ATPases in the yeast ([17] and manuscript in preparation). This body of biochemical evidence of multiplicity of Ca2‡ -ATPases in yeast intracellular membranes is in agreement with genetic data, which predict the existence of various Ca2‡ -ATPases [23,24]. They are also in agreement with the present knowledge on the existence of three genes encoding ¢ve isoforms of Ca2‡ -ATPases of reticulum type in animal cells [25]. It is known that the pmr1 mutation causes a truncated glycosylation of secretory invertase presumably due to a Golgi bypass [1]. The data of the present work show a faster formation of spheroplasts in the pmr1 mutant, compared to an isogenic wild-type strain (Fig. 2). This may mean that Ca2‡ -ATPase(s) and Ca2‡ gradients are involved in biosynthesis and then (indirectly) in assembly of cell wall components. Several facts indicate a possible mis-targeting of some proteins of the secretory pathway caused by the disruption of PMR1. We observed a decrease of NADPH cytochrome c reductase, dolichyl phos-

FEMSLE 8116 27-4-98

90

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

phate mannose synthase and of total protein in the pmr1 membranes in Golgi region beside of a decreased Ca2‡ -ATPase activity (Figs. 1, 3 and 4). This alteration in compartments of the secretory pathway is a result of the loss of Pmr1p and it points to the importance of a su¤cient lumenal Ca2‡ concentration or/and of the presence of Ca2‡ -ATP itself for the correct transport and/or retention of resident proteins. It was shown earlier that ionophore A-23187 blocks the exit of secretory proteins from the rough ER in mammalian cells [26]. Since the ionophore collapses both the Ca2‡ and H‡ electrochemical gradients across any biological membrane, the conclusion regarding the role of Ca2‡ -ATPase(s) in the correct targeting of proteins [26] was not unambiguous. The importance of a V1 V0 H‡ -ATPase and a H‡ gradient across an yeast tonoplast was previously demonstrated for vacuolar biogenesis [27]. Our report shows that the loss of the PMR1 Ca2‡ -ATPase of Golgi and Golgi-like membranes causes signi¢cant changes of enzyme activities of compartments of the secretory machinery. Since the changes are even manifested at the level of total Golgi and ER proteins (Fig. 1D) one may speculate that other Golgi and vacuolar enzymes are also not properly sorted. A signi¢cant increase of activity of K-mannosidase, a non-integral membrane protein, has previously been found in all compartments of the secretory pathway even in the early exponential phase [9]. This is analogous to the behavior of GDPase in early stationary phase (Figs. 1B and 3C). A signi¢cant decrease of activity of the membrane enzyme of the Golgi, Kex2p, was also found in the early exponential phase of growth of the pmr1 mutant [9]. The activity of Kex2p did not increase, however, in ER and ER-like membranes (what one may expect from our model of the Ca2‡ -ATPase(s)-dependent sorting of proteins) probably because of the growth phase or because of the possible toxicity of the protease for ER and therefore its inactivation. All the data can be taken as evidence for a redistribution of proteins within compartments of the secretory pathway as a consequence of the disappearance of the PMR1 Ca2‡ -ATPase. One may speculate that incorrectly sorted proteins fall within two groups: integral membrane proteins will partly arrive in their compartments and partly accumulate in ER

and ER-like membranes; non-integral membrane proteins (K-mannosidase) and lumenal proteins (GDPase) will primarily accumulate in their resident compartment, probably in an attempt to compensate for a decrease of their activity at lower concentrations of Ca2‡ [13]. Additionally, non-integral membrane proteins could show a di¡erent response also depending on the growth phase (for example, Kmannosidase). As discussed above the pmr1 mutation causes mislocalization of activity of several enzymes of compartments which have lost the PMR1 Ca2‡ -ATPase and of compartments which are further downstream in the secretory pathway. The increase of catalytic activity of those enzymes in other compartment(s) indicates that enzymes accumulate there in a functionally active form and not in a silent state. Therefore, it is likely that other compartments (in this case ER) can partly take over some functions (e.g. Ca2‡ uptake) of compartment(s) (e.g. Golgi or vacuoles) that did not receive a complete set of own enzymes. Future experiments have to answer the question whether the increase of activity of several enzymes in the ER of the pmr1 mutant is just a prelude to their proteolysis or whether enzymes have changed compartments to ful¢l their function.

Acknowledgments We are grateful to Dr. W. Tanner for the interest in this work and for helpful discussions. We also thank Dr. H. Rudolph for yeast strains. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (SFB 521).

References [1] Rudolph, H.K., Antebi, A., Fink, G.R., Buckley, C.M., Dorman, T.E., Le Vitre, J., Davidow, L.S., Jen-i-Mao and Moir, D.T. (1989) The yeast secretory pathway is perturbed by mutation in PMR1, a member of Ca2‡ -ATPase family. Cell 55, 133^145. [2] Ohsumi, Y. and Anraku, Y. (1983) Calcium transport driven by a proton motive force in vacuolar membrane vesicles of Saccharomyces cerevisiae. J. Biol. Chem. 258, 5614^5617. [3] Okorokov, L.A., Lichko, L.P. and Kulakovskaya, T.V. (1985) H‡ -ATPase and H‡ -/ion antiporters of the vacuolar mem-

FEMSLE 8116 27-4-98

L.A. Okorokov, L. Lehle / FEMS Microbiology Letters 162 (1998) 83^91

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

brane of Saccharomyces cerevisiae yeast. In: Biochemistry and Function of Vacuolar Adenosine Triphosphatase in Fungal and Higher Plant Cells (Marin, B.P., Ed.), pp. 203^212. Springer Verlag, Berlin. Okorokov, L.A., Martinez, E., Lichko, L.P., Kulakovskaya, T.V. and Predeina, E.A. (1988) Ca2‡ /H‡ antiporter of the vacuolar membrane of Saccharomyces carlsbergensis yeast. Biol. Membr. 5, 1156^1160 (in Russian). Okorokov, L.A. and Lichko, L.P. (1983) The identi¢cation of proton pump on vacuoles of yeast Saccharomyces carlsbergensis: ATPase is electrogenic H‡ -translocase. FEBS Lett. 155, 102^106. Okorokov, L.A., Tanner, W. and Lehle, L. (1993) A novel primary Ca2‡ transport system from Saccharomyces cerevisiae yeast. Eur. J. Biochem. 216, 573^577. Okorokov, L.A. (1994) Several compartments of Saccharomyces cerevisiae yeast are equipped with Ca2‡ -ATPase(s). FEMS Microbiol. Lett. 117, 311^318. Okorokov, L.A., Kuranov, A.Ju., Kuranova, E.V. and de Santos Silva, R. (1997) Ca2‡ -transporting ATPase(s) of the reticulum type in intracellular membranes of Saccharomyces cerevisiae: biochemical identi¢cation. FEMS Microbiol. Lett. 39^46. Antebi, A. and Fink, G. (1992) The yeast Ca2‡ -ATPase homologue, PMR1, is required for normal Golgi function and localizes in a novel Golgi-like distribution. Mol. Biol. Cell 3, 633^654. Sorin, A., Rosas, G. and Rao, R. (1997) PMR1, a Ca2‡ -ATPase in yeast Golgi, has properties distinct from sarco/endoplasmic reticulum and plasma membrane calcium pumps. J. Biol. Chem. 272, 9895^9901. Guënther, R., Brauer, C., Janetzky, B., Forsters, H.-H., Ehbrecht, J.-M., Lehle, L. and Kuëntzel, H. (1991) The Saccharomyces cerevisiae TRG1 gene is essential for growth and encodes a luminal ER glycoprotein involved in the maturation of vacuolar carboxypeptidase J. Biol. Chem. 266, 24557^ 24563. Feldman, R.J., Bernstein, M. and Scheckman, R. (1987) Product of SEC53 is required for folding and glycosylation of secretory proteins in the lumen of the yeast endoplasmic reticulum. J. Biol. Chem. 262, 9332^9339. Abeijon, C., Orlean, P., Robbins, P.W. and Hirschberg, C.B. (1989) Topography of glycosylation in yeast: Characterization of GDP mannose transport and luminal guanosine diphosphatase activities in Golgi-like vesicles. Proc. Natl. Acad. Sci. USA 86, 6935^6939. Serrano, R. (1988) H‡ -ATPase from plasma membranes of

[15]

[16] [17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

91

Saccharomyces cerevisiae and Avena sativa roots and its reconstitution. Methods Enzymol. 157, 533^544. Lichko, L.P. and Okorokov, L.A. (1991) Puri¢cation and some properties of membrane bound and soluble pyrophosphatases of yeast vacuoles. Yeast 7, 805^812. Carafoli, E. (1991) Calcium pump of the plasma membrane. Annu. Rev. Biochem. 56, 395^433. Okorokov, L.A., Prasad, R., Zvjagilskaia, R.A., Lichko, L.P., Kulakovskaya, T.V., Yurina, V.P. and Odintzova, M.S. (1996) Isolation of pure membrane fractions for lipid analysis. In: Manual of Lipids (Prasad, R., Ed.), pp. 12^31. Springer Verlag, Berlin. Franzuso¡, A. and Schekman, R. (1989) Functional compartments of the yeast Golgi apparatus are de¢ned by the sec7 mutation. EMBO J. 8, 2695^2702. Cunningham, K.W. and Winckner, W.T. (1989) Yeast KEX2 protease and mannosyltransferase I are localized to distinct compartments of the secretory pathway. Yeast 5, 25^33. Graham, T.R. and Emr, S.D. (1991) Compartmental organization of Golgi-speci¢c protein modi¢cation and vacuolar protein sorting events de¢ned in a yeast sec 18 (NSF) mutant. J. Cell. Biol. 114, 207^218. Bowser, R. and Novick, P. (1991) Sec 15 protein, an essential component of the exocytotic apparatus, is associated with the plasma membrane and with a soluble 19.5S particle. J. Cell Biol. 112, 1117^1131. Cuningham, K. and Fink, G. (1994) Calcineurin-dependent growth control in Saccharomyces cerevisiae mutants lacking PMC1, a homolog of plasma membrane Ca2‡ -ATPases. J. Cell Biol. 124, 351^363. Catty, P. and Go¡eau, A. (1996) Identi¢cation and phylogenetic classi¢cation of eleven putative P-type calcium transport ATPase genes in the yeast Saccharomyces cerevisiae and Schizosaccharomyces pombe. Biosci. Rep. 16, 75^85. Catty, P., de Kerchove d'Exaerde, A. and Go¡eau, A. (1997) The complete inventory of the yeast Saccharomyces cerevisiae P-type transport ATPases. FEBS Lett. 409, 325^332. Lytton, J., Westlin, M., Burk, S.E., Schull, G.E. and Maclennan, D.H. (1992) Functional comparisons between isoforms of the sarcoplasmic or endoplasmic reticulum family of calcium pumps. J. Biol. Chem. 267, 14483^14489. Lodish, H.F. and Kong, N. (1990) Perturbation of cellular calcium blocks exit of secretory proteins from the rough endoplasmic reticulum. J. Biol. Chem. 265, 10893^10899. Kane, P.M. and Stevens, T.H. (1992) Subunit composition, biosynthesis and assembly of the yeast vacuolar proton-translocating ATPase. J. Bioenerg. Biomembr. 24, 383^393.

FEMSLE 8116 27-4-98