Inositol 1,4,5-trisphosphate binding sites copurify with the putative Ca-storage protein calreticulin in rat liver

Cd/ C&urn (lQ93) 14. 485-492 0 LongmanGroupUK Lid 1893

Inositol 1,4,5=trisphosphate binding sites copurify with the putative Ca-storage protein calreticulin in rat liver P. ENYEDI’, G. SZABADKAI’, K.-H. KRAUSE2, D.P. LEW2 and A. SPAT’

Department of Physiology, Semmelweis University of Medicine, Budapest, Hungary, 2 Division of /nfectious Diseases, Hapita/ Cantonal Univetsitaire, Geneva, Switzerland

’

Abstract - Rat liver was homogenized and subjected to differential centrlfugatlon. When the low speed nuclear pellet was processed on a Percoll gradient, plasma membrane markers and lns(l,4,5)P3 binding activity purified together. The high speed (microsomal) fraction was subfractionated by sucrose density gradient centrlfugation, resuttlng in l@fold enrichment of js2P]-lns(1 ,4,5)Ps binding. In the sucrose density gradient fractions there was an inverse relationship between the enrichment of plasma membrane markers and Ins(l,4,5)Ps binding sites. Endoplasmic reticulum markers showed a moderate enrichment in the fractions displaying high Ins(l,4,5)Ps binding activity. Calcium binding protelns in the homogenate and in the microsomal subfractions were separated by SDS/PAGE. A 50 kD protein, stained metachromatically with Stains-All was identified as calretlculin with immunoblotting. lts enrichment pattern was similar to that of Ins(l,4,5)Ps blnding sites, indicating the cosxistence of these two elements of Ca2+-metabolism in the same intracellular compartment in the liver.

Ca2+-signal,generated in response to different hormones, neurotransmitters and growth factors, is a known regulator of cellular functions. Demonstration of Ins(l,4,5)P3Anduced Ca2’ release from a non-mitochondrial intracellular store was a key observation in our understanding of the mode of action of these agents. Based on the first studies on Ca2’ release [ 1, 21 and on the detection of specific Ins(1,4,5)P3 receptors in the microsomal fraction [3], the source of Ca2+iu this process was suggested to be the endoplasmic reticulum. Further observations, however, challenged this assumption and in spite of several immunocytochemical and cell frac-

tionation studies the source of Ca2’ release induced by Ins( 1,4,5)P3is still a question of debate. The exceptional abundance of Ins(1,4,5)P3 receptors allowed their immunochemical localization at electronmicroscopic level in cerebellar Purkinje cells [4-6]. Immunoreactivity, albeit of various intensity, was detected in several intracellular membranous structums, including smooth and rough endoplasmic reticulum 14.7, 81, subplasmalemmal cistemae [4,7], and the nuclear envelope [4,7, 81, but not in plasma membrane. Considering that the structure and function of neurons differ significantly from non-excitable cells, data obtained on Purkhije

485

CELLCALCIUM

486

cells may be generalized with precaution only. Studies on receptors in peripheral tissues gave controversial results, raising the possibility that more than one orgauelle may contain Ins(1,4,5)P3 receptors [9]. Endoplasmic reticulum was found to constitute an inhomogeneous calcium store, and only one of its subfractions may be responsive to Ins(1,4,5)P3 [9, lo]. Ins(l,4,5)Ps induced Ca2’ flux and/or Ins(1,4,5)Pa binding were found by several authors in the plasma membrane fraction [9,11-151, but negative data were also published [13, 161. Ins(1,4,5)P3 binding and Ca2’ release were also found in nuclear preparations [17, 181suggesting an additional site of Ius(l,4,5)P3 action. A hitherto unrecognized Ins(1,4,5)P3 sensitive Ca2’ store was described in neutrophil granulocytes [19], HL-60 cells [20] and bovine adrenal cortex [13]. This new subcellular entity, which is different from plasma membrane and endoplasmic reticulum, was termed calciosome [ZO]. Any calcium storing vesicle, capable of rapid Ca2’ mobilization, is expected to contain high capacity Ca2’-binding pro teins. Indeed calciosomes were marked by an antibody raised against calsequestrin [20], a calcium binding protein characteristic for the terminal cisterns of sarcoplasmic reticulum Membrane bound vacuoles containing calsequestrin immunoreactivity, but devoid of endoplasmic reticuhrm markers, were detected in the cytoplasm of rat hepatocytes and pancreatic acinar cells [21]. In cerebellar Purkinje cells, immunoreactive calsequestrin was heavily concentrated in membrane bound structures assumed to be calciosomes [22]. In a later study, calsequestrin could not be detected in liver microsomes [23], and mom recently, as studied by means of immunoblotting of microsomal extracts of HL-60 cells [24], liver and brain [25], the calsequestrin-like protein has been identified as caheticulin (CR), a 60 kD calcium-binding protein described first in longitudiual tubules of the sarcoplasmic reticulum In the liver, immunogold labelling for CR decorated vacuolar profiles indistinguishable from calciosomes [25]. The presence of Ca-ATPase, an enzyme responsible for calcium uptake was also demonstrated on calciosomes [21,22] confirming their physiological role in the control of calcium metabolism. Until now, the subcellular colocalization of Ins( 1,4,5)Ps binding sites aud CR was described in

the tumor cell line IX-60 only [26]. Here, for the first time, we present data on the colocalization of Ins(1,4,5)P3binding sites and CR in a pamnchymal tissue, the liver.

Materialsand methods [32p]-Ins(1,4,5)P3(20-200 Ci/mmol) was obtained from Du Pont New England Nuclear, u&be&d Ins(1,4,5)Ps from Calbiochem [‘4C]-glucose 6phosphate and [3H+adenosine S-monophosphate were from Amersham, [a-3%]-ATP from Izinta, reinforced cellulose nitrate membrane (BAS 85) from Schleicher & Schuell. Other chemicals were of analytical grade, obtained from Bio-Bad Laboratories and Sigma. Preparationof subcellularfractions

Livers from 230-270 g female Wistar rats were perfused with ice cold homogenization medium (HM) composed of sucrose (250 mM), EGTA (1 n&l), dithiothreitol (1 mM), HJZPBS/KOH(10 mM, pH 7.4) and protease inhibitors: phenylmethyl sulfonyl fluoride (0.6 mM), leupeptine (1 pM) and aprotinine (0.1 pM), and all the following procedures were performed at 4°C. The tissue was minced and homogenized in 9 vol of HM with Dounce homogenizer (loose pestle, 10 strokes) followed by handdriven Potter (loose pestle, 4 strokes). The nuclear pellet fraction (1000 g x 10 min pellet) was processed on a self generating Percoll gradient (20% v/v) in a Beckman JA20 rotor spun at 35 000 g x 30 min. The upper membranous fraction was withdrawn, resuspended with a hand-driven Potter in HM and mcentrifuged in a Beckman Ti 50.2 rotor at 100 000 g x 60 min. In order to eliminate DNA contamination, the pellet was treated with DNAase [27] and finally resuspended in HM. This suspension is termed low-speed fraction. The postnuclear supematant was centrifuged first in a JA20 rotor for 8000 g x 25 min. then the supernatant for 35 000 g x 25 min. The microsomal pellet thus obtained was resuspended in HM. The concentration of sucrose was adjusted with additional sucrose to 1.2 M. The suspension was applied as the iutermediate phase to a step-wise sucrose gra-

JNOSITOL 1,4,5TRISPHOSPHATE BINDING SITES & CALRETICULIN

diem (0.25, 1.2 and 2 M sucrose, respectively, in HM). After centrlfugation (200 000 g x 60 min in a Beckman SW41 rotor) the 0.2X.2 M interphase (fraction 1). the 1.2 M phase (fraction 2) and the 1.212 M interphase (traction 3) were separated, diluted with HM, pelleted (200 000 g x 60 min) and resuspended in HM. Of the 107.4 + 8.6 mg protein/ g tissue in the homogenate, the fractionation yielded 0.18 + 0.04 mg protein/g tissue in fraction LO.48 + 0.09 mg/g in fraction 2 and 0.58 f 0.20 mg/g in fraction 3. AIiquoted samples were stored at -70°C.

487

tion V, 1 mg/ml). CaC12was added in order to convert low affinity binding sites to high affinity ones [33]. After adding [32p]-Ins(1,4.5)P3(2-5 nCi), the incubation was continued for a further 15 min. Bound and unbound ligands were separated by filtration through GF/B filters. Non-specific binding was determined in the presence of 1 p.M Ins(1,4,5)P3. Specific binding in the homogenate was 0.95 + 0.08% of the added tracer by the 100 pg protein sample. Analysisof protein

Marker enqvne determinations

Protein content of the samples was detemrined by NADPH cytochrome c reductase assay was per- Bradford’s method [34], using bovine serum alformed according to Ullrich and Wollheim [28]. bumin as standard. In order to avoid erroneous estiThe activity of 5’-nucleotidase was measured with mation of protein content resulting from the soluble the method of Avmch and Wallace [29] using [3H]- or particulate character of the samples, parallel AMP as substrate [28]. determinations were carried out with the methods of Glucose 6-phosphatase activity was measured by Bradford [34] and Peterson 1351in two experiments. the method of Aronson and Touster [30] modified as The results showed no systematic deviation. follows. [14C]-glucose6-phosphate (3 nCi/sample, Analysis of proteins by SDS/PAGE was per1 mM) was used as substrate. It was separated from formed using the discontinuous buffer system of glucose (formed during the incubation) by passing Laemmli [36] with gradient gel (5-1695 acrylathe samples through 0.5 ml Bio-Rad AG 1 x 8 anion mide). Gels were stained with the cationic carboexchange columns (formate form). Samples of 20- cyanine dye Stains-All as described [37]. Relative 100 pg protein were measured. Substrate consump- .molecular masses were estimated on the basis of tion during the 30 min incubation period (37°C) mobilities of authentic prestained standards (Sigma). never exceeded 10%. When PAGE was followed by immunoblotting, The activity of adenylyl cyclase was measured in separation of the proteins was performed on 10% the presence of 0.1 mM forskolin as previously de- acrylamide gel. Electrophoretic transfer to reinscribed [311. forced cellulose nitrate membrane was performed Marker enzyme activities (per mg protein) in the following a standard method described previously homogenate were as follows. NADPH cytochrome [26]. The transfers were blocked with phosphate c reductase activity was 1.85 f 0.15 pmol/30 min, buffered saline containing ‘I&en-20 (0.05%) and 5’-nucleotidase activity was 1.94 + 0.26 pmol/30 non-fat dry milk (OS-l%) prior to incubation with min. glucose 6-phosphatase activity was 0.29 f 0.03 the CR antibody which had been raised by immunipmoV30 min and adenylyl cyclase activity was 1.81 sation of rabbits with synthetic peptides correspond+ 0.26 nrnoYl0 min. respectively. ing to the 20 N-terminal amino acids of human calrcticulin, coupled to keyhole limpet haemocyanin Ins(l,4J)P3 binding 1261. Immunoblots were developed with [35S]-Protein A. To quantitate the CR content of the differ[3sP]-Ins(1,4,5)Pjbinding was measured by a modi- ent subcellular fractions, 4 different aliquots of each fication of a previously described method [32]. sample were tested. The amount of radioactivity Samples (100 pg protein) were incubated in 0.5 ml was detected by autoradiography using a storage buffer at 4°C containing KCl (100 mM), NaCl (20 phosphor screen and quantified with the Phosphor mM), BDTA (1 mM), CaClz (1 mM), HEPBS/KOH Imager (Molecular Diagnostics). The relationship (25 mM, pH 7.4) and bovine serum albumin (frac- between the amount of protein and the pixel value

488

appeared to be non-linear, therefore a power curve was fitted to the data using the equation y = axb, where y represented the pixel value, x the amount of protein in the aliquot, whereas a and b were the constauts to be calculated. After curve fitting for each sample, the exponent was averaged (b = 0.68 f 0.06, n = 14) and each slope, a, was recalculated by substituting 0.68 for b. Enrichment of CR was expressed as the ratio of a in the respective fraction and in its corresponding homogenate.

CELLcALcfuM

+

Data are expressed as mean + SEIvI. For calculating the correlation with marker proteins, the logarithm of binding values was taken.

l5

s Q)

E r

Statistics

1

10

0 .L

t 5 l&l

Results and discussion While in most Ca2+ release studies Ins( 1,4,5)P3 acted on intracellular vesicles, Ins(1,4,5)Pa binding sites were usually locahzed in the plasma membrane fraction. In these binding studies plasma membrane was purified from the nuclear pellet [11-G]. In preliminary experiments we purified a membrane rich fraction (termed low-speed fraction) from the nuclear pellet of rat liver. Similarly to previous reports, this preparation was also enriched in Ins(1,4,5)P3 binding sites (Fig. 1). Recalling the proposal that Ins(1,4,5)P3responsive vesicles are attached to plasma membrane rather than being its constituent [38], we tested the possibility that plasma membrane and the binding sites can be detected separately. The low-speed fraction usually contains big sheets of plasma membrane, while small plasma membrane vesicles, formed during more forceful fragmentation, sediment in the microsomal fraction. Since cytoskeletal elements, responsible for the postulated attachment of calcium stores to the plasma membrane, are probably less preserved after such a fragmentation, we expected to isolate Ins(1,4,5)4 binding sites separate from plasma membrane in the high-speed microsomal fraction. Microsomal fraction prepared from the posmuclear supematant displayed substantial Ins( 1,4,5)P3 binding activity (Fig. 2A). Considering that this fraction consists of different subcellular particles, in-

0

-

A

B

Fig. 1 Enrichment (fold incmase over homogenate) of glucose 6-phosphatase (A). 5’-nucleotidase (B) and [3~]-Ins(1,4,5)~ binding sites (C) in the membrane fraction prepared from the low-speed nuclear pellet (for details of preparation, see Materials and me&ods), Mean f SEM are shown for 3 experiments.

eluding plasma membrane, endoplasmic reticulum, and possibly also calciosomes, the identity of Ins(1,4,5)Pa responsive vesicles was not obvious. Therefore, the microsomes were processed further by sucrose density gradient centrifugation. Binding of tracer amounts of [3aP]-Ins(1,4,5)P3(per mg pm tein) increased about 2.5, 4 and lo-fold in fractions 1, 2 and 3, respectively, as compared to that in the homogenate (Fig. 2A). In a pmliminaty experiment receptor density, as estimated by Scatchard analysis, was 61 tinol/mg protein in the homogenate and 477 fmol/mg protein in fraction 3. Binding affinity of the receptor did not change during the purification process (& = 1.5 uM in the homogenate, 1.8 nM in fraction 3), therefore binding of the tracer was regamed as an indication of receptor density. With

489

INOSITOL 1.4.5’RISPHOSPHA’IE BINDING SITES & CAUBTICULIN

IO 5

+c

0

L

q A

HMl

s

23

10 5 0

HM123

W

HM123

Enrichment (fold increase over homogenate) of [‘+]-Ins(1,4.S)p3 binding sites (A), calmticulin immunoreactivity (B), rotenoneinsensitive NADPH cytochrome c reductase.activity (C), glucose 6-phosphatase (D) and S-nucleotidase (E) in subcellular fractions of rat liver homogenates. Mean f SEM are shown for 5 (A, D 8t E) or 3 (B dt C) experiments. H, homogenate: M, microsomal fraction (35 OCO g pellet); 1, 0.2V1.2 M interphase; 2, 1.2 M phase; 3, 1.212 M interphase of the sucrose density gradient.

J%3 2

as observed in separate experiments, where enrichment of 5’-nucleotidase activity (Table) was comparable to that shown in Figum 2. Accordingly, these binding sites may neither be constituents of the sinusoidal plasma membrane. It should be recalled, however, that copurification of Ins(1,4,5)Ps binding sites with plasma membrane markers could be prevented by disrupting cytoskeletal elements when the starting material was the nuclear fraction of rat liver [38]. Therefore, it is possible that our low-speed fraction contains plasma membrane to which Ins(1,4,5)Ps responsive vesicles are loosely attached and after detachment these vesicles sediment in the microsomal fraction. Ins(1,4,5)P3 receptors were recently detected both in plasma membraue and in intracellular membranes of Jurkat cells and it was also hypothesized ithat different receptor isoforms are present in these membrane systems [39]. With this concept in mind, it could be argued that our low-speed fraction contains plasma membrane with integral Ins(1,4,5)Ps receptor molecules, while they are primarily the intracellular membranes that contain binding sites in the micmsomal fraction. In this case, fraction 1 (after sucrose density gradient centrifugation) should contain binding activity at least as much as corresponding to its plasma membrane content. Enrichment of Ins(1,4,5)P3 binding sites in fraction 1 is much lower than that in the purified low-speed fraction, in spite of comparable 5’-nucleotidase activity in the two fractions (Figs 1 and 2) indicating that canalicular plasma membrane is not the site of Ins(1,4,5)P3 action. Our data are not conclusive

Table Enrichment (fold increase over homogenate) of

this provision it could also be calculated that the highest density fraction (traction 3), in which the highest purification was attained, contained 3.8 and 2.3-fold as many binding sites as fraction 1 and 2, respectively. Activity of 5’-nucleotidaseshowed an inverse relationship with Ins(l,4,5)P3 binding (Fig. 2A,E) indicating that Ins(1,4,5)Ps binding sites in fraction 3 are not constituents of the canalicular plasma membrane. Fraction 3 was also poor in adenylyl cyclase

5’-nucleotidase and forskolin (100 p&i) stimulated adenylyl cyclase in liver subcellular fractions. Mean f SEMare given for 3 experiments. 5’-nucleotidase

Adenylyl cyclme

Homogenate Low-speed fraction Microsomes Sucmiie fraction 1 sucxosefra&m2 sucrose fraction 3

1.00 15.13 f 3.3 1 f 11.43 i 4.94 f 1.20 f

3.42 0.44 2.45 0.43 0.23

1.00 7.79 f 1.44 f 3.50 f 3.08 f 1.04 f

2.03 0.29 0.85 0.80 0.33

490

ular

Molec

mass kDa

84

St E’ig. 3 Ekctmphoretic

H

M

1

2

3

H

StH

M

12

separation of proteins. Subcellubx fractions (70 pg per lane) were separated by SDS/PAGE.

3 Molecular weights

were estimated using pxestained standards (St). The left hand gel was stained with Stains-All and photographed with a red iXer which reduces the visibility of isochromatically SDS/PAGE (wing 546%

stained (red) bands.

The middle panel shows immunological

detection of calt&culin

after

gradient) and blotting. The right hand gel was stained with Coomassie Blue. For abbreviations see caption to

Figure 2.

coucerning the presence of Ins( 1,4,5)Ps binding sites in the sinusoidal plasma membrane. Although somewhat less abundant in traction 1, NADPHcytochrome c reductase and glucose 6phosphatase, two endoplasmic reticulum markers, were present with comparable activity in all 3 microsomal subfractions, showing an about 2.54 fold enrichment, with respect to the homogenate (Fig. 2C.D). Since Ins(1,4,5)P3 binding sites enriched 2.5-fold more than endoplasmic reticulum markers in fraction 3, the binding sites are certainly present in a compartment much smaller than total endoplasmic reticulum, yet possibly iu one of its specialized compartments. Proteins from different subcellular fractions were separated by SDS/PAGE. Calcium binding proteins were detected by blue metachromatic staining with Stains-All. Several blue bands attained high con-

centration in fraction 2, and especially in fraction 3 (Fig. 3). One of them with a relative molecular mass of 100 kD may correspond to endoplasmin [4O],while another one of 60 kD was presumed to be CR. This assumption was confirmed by immunoblotting with an anti-CR antibody. Considering that CR is the calcium binding protein in calciosomes, we quantitated immunoreactive CR in the different fractions. The antibody labelled a single protein band with a relative molecular mass of 60 kD (Fig. 3). The enrichment of CR showed a pattern similar to that of Ins(1,4,5)4 binding (r = 0.83, P < 0.05) and attained its highest value (7.60 rt 0.71-fold) in fraction 3 (Fig. 1B). The fact that the enrichment of CR was slightly lower in &action 3 than that of the Ins(1,4,5)Ps binding sites may be accounted for by the presence of immunoreactive CR in the processing particles, endoplasmic reticu-

INOSITOL 1.4.5TRISPHOSPHA’LE BINDING SITES & CALRETICULIN

lum and Golgi [41]. In addition, it is also possible that CR but not binding sites were lost into the supemataut from the lumen of damaged vesicles. (The average loss of the two proteins during density gradient centifugation attained 56% and 20%. respectively.) These results are compatible with the assumption that calciosomes are the Ins(1,4,5)P3 responsive Ca2+-storing vesicles in the liver; nevertheless, it may not be ruled out that these vesicles constitute a specialized subfraction of endoplasmic reticulum.

Acknowledgements The skilful technical assistance of Mrs Zsuzsa Lkgeza is

highly appreciated. The work was supported by grants from the Hungarian National Science Foundation (OTKA grant no. 1111) and the Hungarian Council for Medical Sciences (BTK grant no. T-17611990). The expenses of PE during his stay in Ge.neva were covered by a joint grant to C.B. Wollheim and AS from the Swiss Nationai Foundation (no. 70 UP-029786). The Budapest group gratefully thanks the Howard Florey Institute of Physiology and Experimental Medicine. University of Melbourne, Victoria, Australia for their generous gift of a Beckman SW-28 ulhacentrifuge rotor and Du Pont NEN for supplying [nP&Ins( 1,4,5)P3.

References 1. Streb H. Irvine RF. Berridge. MJ. Schtthz I. (1983) Release of Ca2+from a nomnitochondrial httracelhtkr store in pancreatic a&tar cells by inositol-1,4,5-trisphosphate. Nature, 306.67-68. 2. Prentki M. Biden TJ. Janjic D. Irvine RF. Berridge MJ. Wollheim CB . (1984) Rapid mobilization of Ca2’ from rat insulinoma microsomes by in~itol-1,4,5-~isphosphate. Nature, 309,562-564. 3. Spilt A. Fabiato A. Rubin RP. (1986) Binding of inositol trisphosphate by a liver microsomal fraction. Biocbem. I., 232929-932. 4. Ross CA. Meldolesi J. Milner TA. Satoh T. Supattapone S. Snyder SH. (1989) Inositol 1,4,5-trisphosphate receptor localized to endoplasmic reticulum in cembellar Purkinje neurons. Nature, 339.468-470. 5. M&nary GA. Siidhof TC. Takei K. De Camilli P. (1989) Putative receptor for inositol-1,4,5-trisphosphate similar to ryanodine receptor. Nature, 342.192-195. 6. Snyder SH. Supattapone S. (1989) Isolation and functional characterization of an inositol trisphosphate receptor from brain. Cell Cakium, 10, 337-342.

491

7. Walton PD. AixeyJA. Sutko JL. et al. (1991) Ryanodhk and inositol trisphcaphate receptors coexist in avian cembellar Puthbtje neurons. J. Cell Bkl., 113, 1145-1158. 8. 0tsuH.Y amamoto A. Ma& N. Mikoshii K. Tashim Y. (1990) Imnumogold kcalizatkn of inosito]- 1,4,5-t&phosphate (In&) receptor in mouse cembelhu Purkinje cells using three monocknal antibodies. Cell Struc& Fun&. 15.163-173. 9. Sharp AH. Snyder SH. Nigam SK. (1992) Inositol-1,4,5-t&phosphate receptors. Localization in epithelial tissue. J. Biol. Chem.. 267,7444-7449. 10. Gill DL. Ghost TK. Mullaney JM. (1989) Cakium signaIling mechanisms in endoplasmic reticuhtm activated by inositol-1,4,5-trisphosphate and GTP. Cell Calcium, 10, 363-374. 11. Guilkmette G. Balla T. Beukal AJ. Catt KJ. (1988) Chamcbxization of inositol- 1,4,5&sphosphate receptors and calcium mobilization in a hepatic plasma membrane ftaction J. Biol. C&m.. 263.45414548. 12. Dunkp ME. La&ins RG. (1988) GTP and inosiml- 1.4.5~trisphosphate-induced release of “Ca2+ from a membrane store co-localized with pancreatic-iskt-cell plasma membtane. Biochem. J., 253.67-72. 13. Rossier MF. Capponi AM. Vallotton MB. (1989) The inositol-1,4,5-trisphosphate-binding site in adrenal cortical cells is distinct from the endoplasmic reticulum J. Biol. Chem.. 264, 14078-14084. 14. Feng L. Pet&a B. Kraua-Friedmann N. (1992) Different localization of inositol-1,4,5&sphosphate and ryanodine binding sites in rat liver. Cell Calcium, 13,79-87. 15. Khan AA. Steiner JP. Snyder SH. (1992) maSma membrane inositol-1.4.5~t&phosphate receptor of lymphocytes: Selective emichment in sialic acid and unique binding specificity. Pmt. Natl. Acad. Sci. USA 89.2849-2853. 16. At&i KS. (1992) Localization of the &]I& binding site on human platelet intracellular membranes isolated by high-voltage free-flow ekctrophomsis. FEBS Lett. 298, 173-176. 17. Malviya AN. Rogue P. Vincendon G. (1990) Stemcapecifk inositol-1,4,5-p2p]trisphoephats binding to isolated rat liver nuclei: Evidence for inositol t&phosphate receptor mediated calcium telease from the nucleus. Ptuc. Natl. Acad. Sci. USA. 87,927~9274. 18. Nicotera P. Ormnius S. Nilsson T. BerFn P. (1990) An inositol- 1,4,5-t&phosphate-sensitive Ca pool in liver nuclei. Proc. Natl. Acad. Sci. USA. 87,6858-6862. 19. Kmuse KH. Lew PD. (1987) Subcellular disuibution of Ca” pumping sites in human neutmphils. J. Chn. Invest., 80. 107-l 16. 20. Volpe P. Krause K-H. Hashimoto S. et al. (1988) “Cal&some”, a cytoplasmic organelle: tbe inositol-1,4,5-trisphosphate.-sensitive Ca2’ store of nonmuscle cells? Proc. Natl. Acad. Sci. USA, 85, 1091-1095. 21. Ha&in&o S. Bruno B. Lew DP. Pozran T. Volpe P. Meldolesi J. (1988) Immunocytochemistry of cakiosomes in liver and pawxeas. J. Cell Biol., 107.2523-2531. 22. Villa A. Podini P. Clegg DO. Pozzan T. Meldokai J. (1991) Inuacelhdar Ca2+stores in chicken Purkhtje neurons: differential distribution of the low afEnity-high capacity

CELL CALCIUM

492

Ca2+ binding protein,

23.

24.

25.

26.

27.

28.

29.

30.

31. 32.

33.

cahiequeatrin, of Ca2+-ATPase and of the ER hnnenal protein, bip. J. Cell Biol., 113,779-791. Van PN. Peter F. Soling HD. (1989) Four intmcistemal calcium-binding glycopmteins from rat liver micmsomes with high affimity for calcium. No indication for calsequesttin-like proteins in inositol- 1,4,5-t&phosphate-sensitive calcium sequestering rat liver vesicles. J. Biol. Chem., 264, 17494-17501. Ktause K-H. Simmerman HKB. Jones LR. Campbell KP. (1990) Sequence simikity of caheticulin with a Ca2+-binding pmtein that co-purifies with an Ins(l,4,5)Ps-sensitive Ca2’ store in HL-60 cells. Biocbem. J., 270,545-548. Tmves S. Mattei MD. Lanfmdi M. et al. (1990) Caheticuhn is a candidate for a calsequestrin-like function in Ca2+-storage compartments (calciommes) of liver and brain Biochem. J.. 271,473~480. Van Del&n C. Favm C. Sptit A. &my E. Krause K-H. Lew DP. (1992) PuriEcation of an inositol-1,4,5-trisphosphate -binding calreticuhn-containing intracellular compartment of HL60 cells. Biockm. J., 281.651-656. Berezney R. Funk KL. Crane FL. (1970) The isolation of nuclear membrane from a large-scale preparation of bovine liver nuclei. Biocbim. Biophys. Acta, 203,531-546. Ulhich S. Wollheim CB. (1989) Characterization of alpha 2 admnoceptom in a plasma membrane enriched iiaction from the insulin-secreting cell line RINm5F. Acta Endocrinol. Copenh., 121.525-532. Avmch J. Wallace DF. (1971) Preparation and properties of plasma membranes and endopknnic reticulum &gnumts from isolated tat fat cells. Biochim. Biophys. Acta. 233, 334-347. Amnson NN. Touster 0. (1974) Isolation of tat liver plasma membrane. fragments in isotonic sucmse. Meth. Enzymol., 31.90-102. Salomon Y. (1979) Adenylate cyclase assay. Adv. Cyclic Nucleotide Res.. 10,35-55. Enyedi P. Williams GH. (1988) Hetemgenous inositol tetmkisphosphate binding sites in the adrenal cortex. J. Biol. Chem., 263.7940-7942. Pietri F. Hilly M. Mauger J-P. (1990) Calcium mediates the intemonversion between two states of the liver

34.

35.

36.

37.

38.

39.

40.

41.

inositol-1,4,5+isphosphate mceptor. J. Biol. Ckm., 265, 17478-17485. Bradford MM. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biocbem., 72, 248-254. Peterson GL. (1977) A simplification of the protein assay method of Lowry et al. which is mom generally applicable. Anal. B&hem., 83.346-356. Laemmli UK. (1970) Cleavage of stmctural proteins during the assembly of the head of bacteriophage T4. Nature, 227. 680-685. Campkll KP. Macknnan DH. Jorgensen AO. (1983) Staining of the Ca2’ binding proteins, calsequestrin, cabnodulin, tmponin C, and S-100, with the cationic carbocyanine dye “Stains-all”. J. Biol. Chem., 258, 11267-11273. Rossier MF. St J Bird G. Putney Jr JW. (1991) Subcellular distribution of the calcium-storing inositol-1,4,5-trisphosphate-sensitive organelle in rat liver. Possible lit&age to the plasma membmne tbmugh the actin micmfilammts. Biochem. J., 274,643-650. Khan AA. Steiner JP. Klein MO. Schneider MF. Snyder SH. (1992) IP3 receptors Localization to plasma membrane of T cells and cocapping with the T cell receptor. Science, 257,815-818. Koch G. Smith M. Mater D. Webster P. (1986) Endoplasmic mticuhun contains a common abundant calcium-binding glycopmtein, endoplasmin. J. Cell Sci., 86, 217-232. Pietmkn D. Di Vhgilio F. Pozzan T. (1990) Structuml and functional aspects of calcium homeostasis in eukaryotic cells. Eur. J. B&km., 193.599-622.

Please send reprint requests to : Dr Peter Enyedi, Department of physiology, Semmelweis University of Medicine, H-1444 Budapest PG Box 259, Hungary Received : 28 August 1992 Revised : 14 December 1992 Accepted : 6 January 1993