Acta Physzd S c u d 1990, 140, 383-392

Occurrence of phospholipase A, and lysophospholipase in a gastric H,K-ATPasecontaining membrane fraction, and the formation of lysophosphatidylcholine in stimulated pig parietal cells H. O L A I S S O N , G. A R V I D S O N , J.-Y. M A and S. MARDH Department of Medical and Physiological Chemistry, University of Uppsala, Sweden

OLAISSON, H., ARVIDSON,G., MA, J.-Y. & MRRDH,S. 1990. Occurrence of phospholipase A, and lysophospholipase in a gastric H,K-ATPase-containing membrane fraction, and the formation of lysophosphatidylcholine in stimulated pig parietal cells. Actu Physiol Scand 140, 383-392. Received 26 March 1990, accepted 11 June 1990. ISSN 0001-6772. Department of Medical and Physiological Chemistry, University of Uppsala, Sweden. A membrane fraction containing H,K-ATPase (EC 3.6.1 .36) was prepared from pig gastric mucosa and found to contain phospholipase A, (EC 3.1.1.4) and lysophospholipase (EC 3.1.1.5) activities. Washing the membranes decreased their protein content by 25%. Recovery profiles of H,K-ATPase, phospholipase A, and lysophospholipase were similar for membranes washed either with water or with 0.15 or 1.5 M KCI. Nearly identical distribution profiles were obtained for the three enzyme activities after centrifugation of washed vesicle membranes on a linear sucrose gradient. The phospholipase A, activity was stimulated by calcium and increased further in the presence of calmodulin. The amount of cellular radioactively labelled lysophosphatidylcholine was doubled upon cholinergic stimulation of isolated parietal cells prelabelled with [3H]glycerol or had its acyl chain in the sn-1 position, 32Pi.The liberated lys~(~~P]phosphatidylcholine which implies an activation of a phospholipase A,. These findings indicate that secretagogues which increase the cytosolic Ca2+ concentration, i.e. acetylcholine, histamine and gastrin, may activate a phospholipase A, in the parietal cell. Key words: H,K-ATPase, lysophosphatidylcholine, lysophospholipase, parietal cell, phospholipase A,.

Hydrochloric acid is produced by the protontranslocating H,K-ATPase (EC 3.6.1 .36) which is found in the secretory membranes of the parietal cell (Ganser & Forte 1973a,b, Lee et al. 1974, Chang et al. 1977). I n the resting state, these membranes form tubulovesicles which upon stimulation of the parietal cell transform into secretory canaliculi, probably by membrane Correspondence : Sven Mlrdh, Department of Medical and Physiological Chemistry, Biomedical Centre, Uppsala University, Box 575, S-751 23 Uppsala, Sweden.

fusion (Forte et al. 1977, Gibert & Hersey 1982, Jiron et al. 1984, Vial et al. 1985). T h e molecular mechanism behind the membrane transformation is not known, nor is it known how the state of the secretory membrane is regulated and coordinated with the activity of the H,KATPase. Other proteins occurring in the secretory membranes probably play important roles in this respect. T h u s evidence has been presented that a membrane-associated carbonic anhydrase is involved in the regulation of the proton pump activity (Ljungstrom et al. 1984, Vega et al. 1985). Recently, I m et a/. (1987a)

383

384

H. Olaisson et al.

electrophoresis, with no contamination of plasma membrane Na,K-ATPase or of mitochondria1 membrane cardiolipin. The vesicle membranes were washed according to the procedure by Scheele et ul. (1978) with the following modifications. T o 1 ml of unwashed vesicle membranes in 0.25 M sucrose 9 ml distilled water was added. The vesicles were lyophilized and then resuspended in 9 ml of distilled water plus 1 ml of 1.5 M KCI and 20 mM EGTA, p H 7.0. T h e mixture was centrifuged for 2 h at 100000 g. The supernatant was removed and the pellet was homogenized with a Dounce homogenizer in 1 ml 0.25 xi sucrose and kept on ice for 20 min. T h e vesicle suspension was then diluted with 8 ml ice-cold water. After another 20 min 1 ml of the KCI/EGTA solution was added. T h e suspension was centrifuged as described above and the whole procedure was repeated once. After the third centrifugation the pellet was homogenized in 0.25 M sucrose and 1 mM EGTA. T h e protein concentration was measured according to I>owr?-et al. (1951) and adjusted to 5 mg ml-'. The wsicle suspension was then stored at -20 "C. For a comparison, vesicles were also washed with either KCI-free solutions or 1.5 M KCI (final concentration). Extraction and anablsis of lipids. Each sample (0.5 ml) was mixed with 4 ml chloroform/methanol (1 : 1, v / v ) containing non-radioactive carrier lipids (0.1-0.4mg of each lipid analysed). T h e clear homhl.4 T E R 1-1i d S -4N D AIE T H O D S ogeneous solution thus obtained was broken up into t N - o phases by the addition of 1.5 ml water. T h e upper The lipids l-palmitoyl-2-[ l-"C]oleoyl-sn-glycero-3phase was adjusted to pH 5 with hydrochloric acid. I n phosphocholine, l-~l-''C]oleo~-l-st~-gl~-cero-3-phosthe cell-labelling experiments the water was exchanged phocholine and 1,2-[ I-'~C]dioleo!-l-sn-glycero-3phosphocholine were from Amersham international for 1.5 ml 2 mM sodium phosphate buffer, pH 5, and the chloroform phase was then washed once by the p.1.c. (.hersham, UK). l-Palmitoyl-2-[ l-'4C]arachisuccessive additions of 2 ml methanol and 2 ml buffer. dono! I-sn-gl!-cero-3-phosphocholine, ['H]arachidonic The lipids in the lower phase were separated by acid. I ;/-,"P].4TP, "P, and ['Hlglycerol were obtained from Yew- England Nuclear Chemicals (Dreieich, one- (enzyme assays) or two-dimensional (the cellFRG). Collagenase (clostridiopeptidase A, EC labelling experiments) thin-layer chromatography 3 . 4 . 24.3), pronase (isolated from Staph.ylm.oi.sus (Olaisson rt al. 1985). In the former procedure the griseus), lipase (Rhrcopirs arrhizus, EC 3 . 1 . 1 .3), plates were developed in chloroform/methanol/25 "4) calmodulin and trifluoperazine were purchased from ammonia (60: 30: 5, v/v). In both cases thin layers of Boehringer Mannheim (Mannheim, FRG). Car- silica gel impregnated with magnesium acetate were bachol, fatty acid-free bovine albumin and I V i (.j--(6- used (Tolbert rt a / . 1980). The plates were sprayed with dichlorofluorescein and the lipids were visualized arninohexyl)-j-chloro- 1-naphthalene-sulphonamide) were from Sigma (St Louis, CS.4). Silica gel EI was under ultraviolet light. The spots nere scraped off from Merck (Darmstadt, FRG) and Medium 190 directll- into vials to which 1 ml of lO0/,, acetic acid in (Earle's salt) from Gibco (Grand Island, NJ, CS.1). methanol, 4 ml of water and 10 ml of Aquasol were Other chemicals were of analytical grade and com- added. The radioactivity was measured in a liquid scintillation counter. merciall>-available. Deionized, doubl!- quartz-distilled A4ssu,y of H,K-.4TPase. T h e ATPase activity was water was used. Vesicle membrane preparation and mushing procedure. assayed at 21°C as the release of 32P1from [y-"PIATP Vesicle membranes containing H,K-.lTPase were (Mlrdh 1975). T h e assay medium comprised 5,ug M 1 nlM [y-32P]ATP, isolated from homogenates of pig gastric mucosa as vesicular protein, 10 ~ U nigericin, ~ 10mM KCI in 1 0 m Hepes ~ buffer previously described (Ljungstrom t t a / . 1984). T h e 2 m Mgcl,, preparation is characterized by a dominating tf,K- adjusted to p H 7.4 with Tris. Incubation volume was -4TPase protein band after SDS-polyacrylamide gel 1.0 ml.

reported that a protein phosphatase influencing the transmembrane K ' transport is associated with rat heavy gastric membranes enriched with H,K-ATPase. I n our initial characterization of a preparation of H,K-.\TPase-containing vesicular membranes from pig gastric mucosa, the occurrence of low b u t significant amounts of Iysophosphatidylcholine and Iysophosphatidylethanolamine was reported (Ljungstriim rt ui. 1984). In the present report we demonstrate that this prepand aration also contains a Ca"-dependent calmodulin-stimulated phospholipase A, (PLase .\J (EC 3 . 1 . 1 .4) a n d a lysophospholipase (l>~soPLase)( E C 3 . 1 . 1 . 5 ) . T h e PLase A, is activated by Ca" concentrations in the micromolar range, which raises the possibility that P1,ase .A2 activity may be elicited in the secretory membranes of the parietal cell by secretagogues such as acetylcholine, gastrin and histamine, which all are known to increase cytosolic [Ca"] (Muallem 8r Sachs 1984, C h e w & Brown 1986, M a r d h et u l . 1987). Experiments on isolated parietal cells supported this conclusion.

Phospholipases in H,K-A TPase membranes

385

100

-

80

60

L

0

n

40

20

0 Wash no.

M

1

2

Fig. 1. Effect of washing the H,K-ATPase-containing vesicular membranes. Recoveries of protein (n),H,K-ATPase PLase A, (m) and IysoPLase (0) are given in per cent of the original amounts present in unwashed membranes. M in the figure indicates results obtained with membranes washed with 150 mM KCI. Other membranes were washed three times with either water (L) or 1.5 M KCI (H). Mean values from two experiments are given with upper half of the range.

(m),

PLase A, assay. In the routine assay l-palmitoyl-2[ l-'4C]oleoyl-sn-glycero-3-phosphocholine was mixed with dioleoyl glycerophosphocholine to a specific activity of 1.7 x 10' d.p.m. mmol-'. The phospha-

tidylcholine (PC) mixture was suspended in distilled water and dispersed by sonication (MSE Soniprep 150) at 0 "C for 45 min. The sample was subjected to 45 cycles which each consisted of 1 min sonication and 1 min cooling. The PLase A, activity was assayed at the indicated concentrations of Ca2+ in 50 mM Tris/HCI buffer, pH 7.5. In each sample 0.25 mg vesicle membrane protein was incubated with 50 nmol sonicated PC for 1 h at 37 "C. The volume was 0.5 ml. The reaction was stopped by the addition of chloroform/methanol (1 : 1, V/V)and the extracts were treated as described above. The distribution of radioactivity between the fatty acid and PC fractions on the thin layers was determined. Blank values for samples without vesicle membranes were subtracted. The amount of PC that had been hydrolysed was calculated according to the expression : (c.p.m. in liberated fatty acids/c.p.m. in added PC) x (total amount of PC [exogenous endogenous]). In these calculations the amount of endogenous PC was 300 nmol mg-' protein. This figure was obtained after phospholipid analysis of washed vesicular membranes (data not shown). LysoPLase assay. 1-[ 1-'4C]Oleoyl-sn-glycero-3phosphocholine (IysoPC) with a specific activity of

+

1.7 x 10' d.p.m. mmol-' was dissolved in water. The substrate (40 nmol) was preincubated with fatty acidfree albumin (8 nmol) for 10 min at 37 "C in 0.45ml 50 mM Tris-HC1 buffer, pH 9.0. Inclusion of this amount of albumin in the assay medium was found to yield optimal enzyme activity. The -action was started by the addition of H,K-ATPase-containing vesicular membranes (0.25 mg protein in 50 pi 0.25 M sucrose). After 2 min at 37 "C the reaction was stopped with 4 ml chloroform/methanol (1 : 1, v/v), and the samples were then treated and analysed as described above. The distribution of radioactivity between the fatty acid and lysoPC fractions was determined, and the amount of lysoPC that had been hydrolysed was calculated in the same way as the hydrolysis of PC (see above). The amount of endogenous IysoPC according to phospholipid analysis was 17 nmol mg-' protein.

RESULTS T h e protein content of the vesicle membranes was reduced by approximately 25 yo after the first wash with 150 mM KCl (Fig. 1). Only negligible losses occurred during the following two washes. T h e phospholipid/protein ratio increased in proportion to the protein loss but the phospholipid composition remained essen-

386

H. Olaisson et af.

tially unchanged as well as the phospholipid/ cholesterol ratio (data not shown). All of the H,K->iTPase activity was recowxed even after three washes with 150 m51 KC1, while the activities of both PLase A, and l!-soPI,ase were reduced by approximately 2 j o C ,after the third wash (Fig. 1). Washing the vesicle membranes three times with either water or 1.5 X I KCI yielded similar recoveries of protein, PLase -I2 and lysoP1,ase as after washing with 150 mxi KC1 (Fig. 1). C p o n centrifugation of washed xsicles on a linear sucrose gradient both PLase -4?and the 1ysoPLase migrated with the H,K-r\TPase (Fig. 2 ) , which is the specific marker of the purified membranes accounting for 7O0, of the membrane protein (Ljungstrom rt cii. 1984). I n initial experiments and during the course of this study the PLase A, was characterized in various ways. T h e amount of added, unlabelled dioleoyl glycerophosphocholine did not affect the reaction rate when the amount of exogenous PC per incubation ranged from 0.2 (no dioleol-I glycerophosphocholine added) to 400 nmol. T h e reaction remained constant during at least 60 min of incubation under standard conditions. I n the range tested, i.e. 0.1-1 .0 mg protein per incubation, the reaction rate was directly proportional to the amount of vesicular protein in the incubation mixture provided that the endogenous PC was taken into account in the calculations. T h e specific activities of PLase -I2 varied between different preparations. Over a period of several years the mean 1-alue& SD was 15 6 nmol mg-' protein h-' for 17 different preparations. There was no indication that 1ysoPLase activity in the vesicle membranes interfered when the PLase A , activity was assayed at p H 7 . 5 as described above. T h e pH optimum of the PLase A , was in the range of 7.5-10. T h e widely used PLase -4,inhibitors, pbromophenacyl bromide and mepacrine, were efficient inhibitors of the enzyme (I& of 2 x 10-' M and 4 x 2.1 respectively). T h e PLase A, was La"-dependent (Fig. 3 ) . Calmodulin amplified the activating effect of La2+(Fig. 3B and C). T h e calmodulin inhibitors trifluoperazine and W7 abolished the stimulator!effect of calmodulin at all Ca" concentrations studied. T h e PLase .4, activity, howewr, was reduced to zero only at the lower Ca'+ concentrations (Fig. 3 C). At 1 mM Ca" in the presence of 2 p" calmodulin and 100 , u trifluoperazine, ~

-

'_

E cn

* c

3 0

E m U

al L

al

>

0

V

a L l L 0

+ E

20

al

ID

-

V L

al

a

Fract.no.

I

5

I0

I5

Fig. 2. Protein and enzyme distribution profiles after

centrifugation of washed H,K-ATPase-containing vesicular membranes on a linear sucrose gradient. Five milligrams of membrane protein in 2 ml sucrose was layered onto a 14-ml sucrose gradient ranging from 15 to 50°, (\\-/I-)sucrose in 10 mM Hepes buffer, p€I 7.5. The gradient was centrifuged for 5 h at 4 "C and 23000 r.p.m. in a Beckman SW 28.1 rotor. Fractions were collected and aliquots were taken for protein and enzyme assays. The densities of the fractions were calculated from their refractive index. (A). Density of the recovered fractions; (B) protein; (C) €I,K-.lTPase; (D) PI,ase A,2; (E) IysoPLase. or 100 /(\I U'7, the PLase A, actikity approached that obserked in the presence of 1 mM Ca2+only (Fig. 3.4 and C). Similar PLasc A, activities ere obtained also when vesicles were incubated with 100 / ( X I trifluoperazine or 100 ,UM W7 in the presence of 1 mM Ca2+ but without calmodulin (data not shown). Vesicle membranes were isolated from mucosal cells previously incubated with ["HIarachidonic acid. T h e lipids of these vesicles

Phospholipases in H,K-ATPase membranes

tc

2.0 A

I .5

F

h

E

s

-

21

387

1.0

0

L

2,

I

0.5

0 EGTA

8

6

4

EGTA

8

6

4

EGTA

8

6

4

- l o g t C P + 1. M

Fig. 3. Effect of Ca2+and calmodulin on the activity of PLase A, in H,K-ATPase-containing vesicular membranes. Vesicle membranes from three different preparations were incubated for 20 min at 37 "C in 25 mM Tris/HCl buffer, pH 7.5, with 0.2nmol l-palmitoyl-2-[1''C]arachidonoyl-sn-glycero-3-phosphocholine, in the absence (A) or presence of 0.5 p~ (B) or 2.0 ,UM calmodulin (C). Each sample contained 30 pg protein and the final volume was 0.25 ml. The Ca2+levels were maintained with Ca2+/EGTA buffers (Bartfai 1979). 'EGTA' signifies samples containing 2 mM EGTA without addition of Ca2+. The actual concentration of free Ca+ was determined with quin-2 (Tsien et al. 1982) in two of the experiments ( .,0)in (A), (B) and (C). These determinations were carried out on all the samples except for those with the highest Ca2+ and M). The concentration of free Ca2+in the 'EGTA' samples concentrations ~~ (C) the activity of the PLase A, was was 1 6 f 5 nM (mean+ SD; n = 12). At 2 . 0 , calmodulin determined also in the presence of 100 p~ trifluoperazine ( x ) or 100 p~ W7 (W). Trifluoperazine and W7 were dissolved in ethanol. The final concentration of ethanol in the incubation mixture was 1% (v/v).

were radioactively labelled with the following distribution of the [3H]arachidonic acid : P E 44%, PC 31%, PI 21%, PS 2 % and PA 1%. [3H]Arachidonic acid was released from the PC (Fig. 4) but not from the other phospholipids (not shown) when the vesicle membranes were incubated in the presence of 1 mM Ca2+. T h e hydrolysis did not depend on the vesicle concentration and no hydrolysis occurred in the presence of E G T A (Fig. 4). Incubation of parietal cells with "Pi or [3H]glycerol resulted in a radioactive labelling of their phospholipids. T h e phospholipids were separated on thin-layer plates, and the percentage distribution of radioactivity between the main fractions (above 1yo)was for the 32Piincubates : phosphatidylinositol (PI) 43 yo, phosphatidylethanolamine (PE) 17%, PC 16%, phosphatidic acid 14%, unidentified fraction on the sample origin 8 % and phosphatidylserine (PS) 2%; and for the [3H]glycerol incubates: PI 40%, PC

37%, P E 20% and PA 2% (Table 1). T h e content of IysoPC was less than 1% in the two different incubates. However, treatment of the [32P]phosphate- or [3H]glycerol-labelled parietal cells with carbachol for 5 min doubled the isotope content of the small fraction of lysoPC (Table 1). Relative changes of this magnitude were not seen among the other lipids. T h e structure of the IysoPC formed in the parietal cells stimulated by carbachol was investigated. Parietal cells were labelled with 32Piand then incubated with carbachol for 5 min as described in the legend to Table 2, except that 37 MBq 32Piwas used. T h e lipids were extracted and 1-oleoyl-sn-glycero-3-phosphocholine was added to the extracted lipids, which then were separated by two-dimensional thin-layer chromatography. T h e 32P-labelled lysoPC in the mixture with the added unlabelled lysoPC was recovered from the silica gel as previously described (Arvidson 1968). After determination of the

388

H. Olaisson et al.

- 0 0

30 Time, rnin

60

specific radioactivity the isolated IysoPC was incubated with lipase from Rhzzopus arrhizus, which specifically hydrolyses the ester bond at the sn-1 position of glycerophospholipids (Fischer et al. 1973). I n two different preparations in which 30 and 54% of the lysoPC had been hydrolysed by the Rhizopus lipase after an incubation of 1 h, the specific radioactivity of the remaining lysoPC differed by less than 14% from that of the unhydrolysed lysoPC. T h e added 1-acyl-sn-glycero-3-phosphocholine and the 32P-labelled lysoPC were thus hydrolysed at approximately the same rate by the Rhizopus enzyme. This indicates that the labelled lysoPC which was formed in the stimulated parietal cells had its acyl chain in the sn-1 position, as would be expected for a product of a PLase A,. Under standard assay conditions the reaction rate of lysoPLase in the H,K-ATPase-containing membranes remained constant during the first 15 min of incubation. T h e rate increased linearly with the protein concentration up to 2.0 mg protein ml-' provided that the amount of endogenous substrate was taken into account in the calculations of the reaction rate as was done for PLase A,. Addition of albumin to the 1ysoPLase assay mixture influenced the rate of hydrolysis, which was maximal when the molar quotient albumin/lysoPC was 0.2&0.25(Fig. 5 A). Microsoma1 1ysoPLases from brain, liver and kidney in the rat were previously reported to attain maximal activity at an albumin/lysoPC molar ratio of 0.5 (Leibovitz-BenGershon & Gatt 1976). As shown by Leibovitz-BenGershon et al. (1972) the effect of albumin may be to protect the lysoPLase from inactivation by its own substrate, lysoPC. This is consistent with our finding that without albumin the reaction rate reached a maximum at low substrate concentrations and then rapidly declined as the substrate concentration is increased (Fig. 5B). T h e p H optimum of the 1ysoPLase was 8.5-9.5.

Fig. 1. Ca"-dependent release of [3H]arachidonic acid from labelled H,K-,4TPase-containing vesicular membranes. Mucosal cells were isolated from pig gastric mucosa as previously described (Mirdh et al. 1984). Forty million cells were incubated with 7.10 kBq [3H]arachidonic acid for 2 h in 10 ml Medium 199 at 37 OC in a shaking water bath and in a humidified atmosphere of oxygen. T h e labelled cells were pelleted by centrifugation, resuspended in 20 ml ice-cold 7 5 mM sucrose, containing 1 mM EGTh, and then homogenized in a Dounce homogenizer. The sucrose concentration was adjusted to 0.25 M and the homogenate was centrifuged for 10 min at 600g. Labelled H,K-ATPase-containing vesicular membranes were isolated from the supernatant as previously described (Ljungstrom et a / . 1984), and subsequently washed (see Materials and methods). T h e labelled membranes were suspended in 50 mM Tris/HCl buffer, p H 7.5, containing 0.25 M sucrose and incubated at 37 "C in the presence of 1 mM Ca2- (filled symbols) or 0.5 mM EGTA (open symbols) (A). In (A) the amount of vesicle membrane protein was 3 5 p g in a final volume of 0.5 ml. In (B) 35 pup vesicle membrane protein was incubated in 5.0ml final volume. T h e lipids were extracted with chloroform/methanol (1 : 1, v/v) and the distribution of radioactivity between the DISCUSSION lipids was determined after two-dimensional thinlayer chromatography. Total radioactivity of PC (0,Various phospholipid deacylating enzymes have @) and unesterified fatty acids (0, ). is shown. previously been reported to be present in the Each point represents the mean value of duplicate gastric mucosa of the rat, yet not in association determinations in one experiment. with the H,K-ATPase (Wassef et al. 1978, Lin et

al. 1979, Wassef & Horowitz 1981, Hirohara et a l . 1987, Grataroli et al. 1987). Our results indicate that PLase A, and 1ysoPLase are closely

389

Phospholipases in H,K-ATPase membranes

Table 1. Formation of lysophosphatidylcholine in ["P]phosphate- and [3H]glycerol-labelled parietal cells stimulated by carbachol 32P-labelledcells

[3H]Glycerol-labelled cells

Carbachol

Carbachol

-.

+

-

Lipid fraction

c.p.m. lOP cells

Phosphatidylcholine

4017 5142 33 39 4389 5667 17 20 556 715 5 20 11312 13950 3810 4597 10 38 1980 2634

Difference

-

(yo)

c.p.m.

-7 -6 $136 +I36 -2 -3

149763 164736 194 237 81612 88630 161 219 1633 1869 0 0 163699 177058 7799 8694 41 68 2785 2797

+

Difference

cells

~

Lysophosphatidylcholine" Phosphatidylethanolamine Lysophosphatidylethanolamine Phosphatidylserine

L ysophosphatidylserine Phosphatidylinositol Phosphatidic acid Sphingomyelin Unidentified

3740 4828 78 92 4298 5508 24 30 538 672 10 11 10698 12887 3729 4693 8 39 1945 2519

$41

+SO -3 -6

-5 -8 -2 +2

136295 146536 389 420 80030 90279 132 244 1639 1835 5 0 159375 171141 7662 8321 18 45 2711 2771

-9 -11 + 100 77 -2 +2 - 18 $11 0 -2

+

-3 -3 -2 -4

~

*

The samples were counted for a sufficiently long time to reach a standard deviation of 1.5%. Parietal cells were isolated and purified from pig gastric mucosae with a purity of 74 and 92% (Mlrdh et al. 1984 and 1987). One-half of the cells in each preparation was labelled with 5 MBq 32P,and the other half with 7.4 MBq [3H]glycerol, at 37 "C in 0.5 ml Medium 199. After 90 min incubation in a shaking water bath and in an oxygenated atmosphere, each cell suspension was diluted 10 times and centrifuged at 600 g for 5 min. The cell pellet thus obtained was suspended in Medium 199 (2 x lo6 cells/ml). After 30 min an aliquot of 50 pI was extracted with chloroform/methanol (1 : 1, v/v) and another 450 p1 of the cell suspension was incubated in the M carbachol for an additional 5 min and then extracted with chloroform/methanol absence or presence of (1 : 1, v/v). The lipids were washed and analysed by two-dimensional thin-layer chromatography (see Materials and Methods). Data from two different preparations are shown.

associated with the H,K-ATPase-containing membranes of pig gastric vesicles. All three enzymes appear to be firmly bound to the vesicle membrane. The PLase A, reported in the present study seemed to be saturated with endogenous substrate in the membranes. I m et al. (1987a) recently observed that PC and particularly PE were extensively degraded when rat heavy gastric vesicles enriched with H,K-ATPase were incubated with 0.5 mM Ca2+ for only 5 min. I n the pig gastric vesicles we found a Ca2+-dependent release of arachidonic acid from the PC of the vesicles. However, our most conspicuous finding

was the high calcium sensitivity of the PLase A,, which was significantly activated by micromolar concentrations of Ca2+. Generally, higher levels of Ca2' are required for the activation of PLase A, in cell-free systems, and it has been proposed that enhanced PLase A, activity in vivo would require some kind of signal amplification in addition to calcium mobilization (Rubin 1986). O u r findings imply that the increase in cytosolic [Ca2+]which acetylcholine, histamine and gastrin elicit (Muallem & Sachs 1984, Chew & Brown 1986, M i r d h et al. 1987) would be sufficient per se for the activation of the PLase A, in the H,KATPase-containing vesicular membranes of par-

390

H. Olaisson et al. B 15

20

'c .E

-* 1,

10

E

10

X

0

f

5

0

I

0

I .o

0.5

albumin/lyeo-PC,

mol/mol

0 0.4

0.8

t r u b s t r a t e j , mM

Fig. 5 . LysoPLase actii-it!- in H,K-ATPase-containing vesicular membranes. (A) Effect of albumin; (B) effect of substrate in the absence (0) or presence ( 0 )of 0.2 mol albumin mol-' exogenous substrate. Albumin was incubated with exogenous IysoPC for 10 min at 37 "C before the addition of vesicle membranes.

ietal cells stimulated by these secretagogues. This conclusion is supported by our present experiments on isolated parietal cells. Cholinergic stimulation of these cells j-ielded unequivocal signs of an increased PLase A, activity. LysoPC and unsaturated fatty acids, which are the p r o d x t s generated by PLase A,, could influence the HCI secretory process in several ways. Specific effects of the products have been reported. T h u s lysoPC stimulates K' transport in rat heavy gastric membranes enriched with Ii,K-riTPase v-ithout abolishing the steep, transmembrane pH gradient (Im et a / . 1987b). Inhibitory effects on the H,K-A'TPase were also reported both for IysoPC (Im et al. 1987 b) and for unsaturated fatty acids ( I m & Blakeman 1982). LysoPC is an amphiphile with soap-like properties (Brentel et nl. 1987). T h i s amphiphile and the unsaturated fatty acids are known as membrane-perturbing, fusogenic agents (Lucy 1978). Their presence, although at low concentrations, in the membranes of the tubulovesicles could be crucial for the extensive fusion of these vesicles which is thought to take place when the parietal cell is activated (Forte et nl. 1977, Gibert & Hersey 1982, Jiron et a/. 1984, Vial et rrl. 1985). Recent studies in our laboratory indicate that gastric vesicles may fuse in e'itro under

conditions that stimulate PLase A, (Olaisson et 01. 1990). This work was supported by the Swedish Natural Science Research Council, the Swedish Medical Research Council, project 4x4965, and the Swedish Society for Medical Research.

REFERENCES ARVIDSON,G.A.E. 1968. Structural and metabolic heterogeneity of rat liver glycerophosphatides. Eur J Biochetn 4, 478486. B A R T F . ~T~. , 1979. Preparation of metal-chelate complexes and the design of steady-state kinetic experiments involving metal nucleotide complexes. In: G. Brooker, P. Greengard, & G.A. Robison (eds.) .4druncer in Cyclic Nucleottde Research, vol. 10, pp. 219-242. Raven Press, New York. BRENTEI., I., ARVIDSON,G. & LINDBLOM, G. 1987. Phase equilibria of the ternary system l-palmitoylsn-glycero-3-phosphocholine/oleic acid/water studied by NMR. Biochim Rioph.ys '4ctu 904, 401404. CIIANG, €I.,SACCOMANI, G., RABON, E, SCHACKMANN, G. 1977. Proton transport by gastric R. & SACHS, membrane vesicles. Riochim Rioph,ys Actu 464, 3 13-327.

Phospholipases in H,K-ATPase membranes CHEW, C.S. & BROWN, M.R. 1986. Release of intracellular Ca2+ and elevation of inositol trisphosphate by secretagogues in parietal and chief cells isolated from rabbit gastric mucosa. Biochim Biophys Acta 888, 116-125. FISCHER,W., HEINZ,E. & ZEUS, M. 1973. The suitability of lipase from Rhizopus arrhizus delemar for analysis of fatty acid distribution in dihexosyldiglycerides, phospholipids and plant sulfolipids. Hoppe-Seyler’s Z Physiol Chem 354, 1115-1 123. FORTE,T.M., MACHEN,T.E. & FORTE,J.G. 1977. Ultrastructural changes in oxyntic cells associated with secretory function: A membrane-recycling hypothesis. Gastroenterology 73, 941-955. J.G. 1973a. K+-stimulated GANSER, A. L. & FORTE, ATPase in purified microsomes of bullfrog oxyntic cells. Biochim Biophys Acta 307, 169-180. GANSER, A.L. & FORTE, J.G. 1973b. Ionophoretic stimulation of K+-ATPase of oxyntic cell microsomes. Biochem Biophys Res Commun 54, 69&696. GIBERT,A. J. & HERSEY,S.J. 1982. Morphometric analysis of parietal cell membrane transformations in isolated gastric glands.3Membr Biol67, 113-124. GRATAROLI, R., CHARBONNIER, M., LiONARDI, J., GRIMAUD, J.-C., LAFONT, H. & NALBONE, G. 1987. Phospholipase A, activity of rat stomach. Arch Biochem Biophys 258, 77-84. HIROHARA, J., SUGATANI, J., OKUMURA, T., SAMESHIMA,Y. & SAITO,K. 1987. Properties and localization of phospholipase A, activity in rat stomach. Biochim Biophys Acta 919, 231-238. IM, W.B. & BLAKEMAN, D.P. 1982. Inhibition of gastric (H+ K+)-ATPase by unsaturated longchain fatty acids. Biochim Biophys Acta 692, 355-360. IM,W. B., BLAKEMAN, D.P., BLEASDALE, J.E. & DAVIS, J.P. 1987a. A protein phosphatase associated with rat heavy gastric membranes enriched with (H+-K+)-ATPase influences membrane K+ transport activity. 3 Biol Chem 262, 9865-9871. IM, W.B., BLAKEMAN, D.P. & DAVIS,J.P. 1987b. Effect of lysophosphatidylcholine on K+ transport in rat heavy gastric membranes enriched with (H+K+)-ATPase. Biochem Biophys Res Commun 146, 84k848. F. 1984. A JIRON,C., ROMANO, M. & MICHELANGELI, study of dynamic membrane phenomena during the gastric secretory cycle : Fusion, retrieval and recycling of membranes. 3 Membr Biol79, 119-134. G. & SCHOLES, P. 1974. An ATPase LEE,J., SIMPSON, from dog gastric mucosa: Changes of outer pH in suspensions of membrane vesicles accompanying ATP hydrolysis. Biochem Biophys Res Commun 60, 825-8 32. LEIBOVITZ-BENGERSHON, Z., KOBILER, I. & GATT,S. 1972. Lysophospholipases of rat brain. 3Biol Chem 247, 684k-6847.

+

391

LEIBOVITZ-BENGERSHON, Z. & GATT,S. 1976. LYSOlecithinase activity in subcellular fractions of rat organs. Biochem Biophys Res Commun 69, 592-598. LIN,Y.N., WASSEF,M.K. & HOROWITZ, M.I. 1979. Lysophospholipase and transacylase activities of rat gastric mucosa. Arch Biochem Biojhys 193, 213-220. LJUNGSTROM, M., NORBERG, L., OLAISSON, H., WERNSTEDT, C., VEGA,F.V., ARVIDSON,G. & MKRDH,S. 1984. Characterization of proton-transporting membranes from resting pig gastric mucosa. Biochim Biophys Acta 769, 209-219. LOWRY,O.H., ROSEBROUGH, N.J., FARR,A.L. & R.J. 1951. Protein measurement with the RANDALL, fohn phenol reagent. 3 Biol Chem 193, 265-275. LUCY,J.A. 1978. Mechanisms of chemically induced cell fusion. In: G. Poste & G.L. Nicolson, (eds.) Membrane Fusion, pp. 267-304, Elsevier/NorthHolland Biomedical Press, Amsterdam. MUALLEM, S. & SACHS, G. 1984, Changes in cytosolic free Ca2+in isolated parietal cells: different effects of secretagogues. Baochim Biophys Acta. 805, 181-185. MKRDH, S. 1975. Bovine brain Na+,K+-stimulated ATP phosphohydrolase studied by a rapid-mixing K+-stimulated liberation of technique : [32P]orthophosphatefrom [32P]phosphoenzymeand resolution of the dephosphorylation into two phases. Biochim Biophys Acta 391, 448463. M~RDH s.,, NORBERG, L., LJUNGSTROM, M., HUMBLE, C. 1984. Preparation of L., BORG,T. & CARLSSON, cells from pig gastric mucosa: Isolation of parietal cells by isopycnic centrifugation on linear density gradients of Percoll. Acta Physiol Scand 122, 607-613. MiRDH, s.,SONG, Y.-H., CARLSSON, c. & BJORKMAN, T. 1987. Mechanisms of stimulation of acid production in parietal cells isolated from the pig gastric mucosa. Acta Physiol Scand 131, 589-598. OLAISSON,H., MKRDH,S. & ARVIDSON, G. 1985. Phospholipid organization in H,K-ATPase-containing membranes from pig gastric mucosa. 3Biol Chem 260, 11262-11267. OLAISSON,H., MKRDH,S. & ARVIDSON, G. 1990. Calcium and calmodulin stimulate phospholipase A, and fusion of H,K-ATPase-containing membrane vesicles isolated from pig gastric mucosa. Acta Physiol Scand 140, 393400. RUBIN,R.P. 1986. Inositol lipids and cell secretion. In: J.G. Venter, & L.C. Harrison, (series eds.) Receptor Biochemistry and Methodology vol. 7, J.W. Putney Jr (ed.) Phosphoinositides and Receptor Mechanisms, pp. 149-162. Alan R. Liss, New York. SCHEELE, G.A., PALADE,G.E. & TARTAKOFF, A.M. 1978. Cell fractionation studies on the guinea pig pancreas : Redistribution of exocrine proteins during tissue homogenization. 3CeN Biol78, IIG130.

392

H . Olaisson et al.

TOLBERT, M.E.M., WHITE,.4.C., ASPRY,K., CUTTS, membranes isolated from pig gastric mucosa. Acta J. & FAIN,J.N. 1980. Stimulation by vasopressin Physiol Scand 124, 573-579. J. & GONZALEZ, A. 1985. T h e and r-catecholamines of phosphatidylinositol for- VIAL,D.V., GARRIDO, mation in isolated rat liver parenchymal cells. 3 Biol early changes of parietal cell structure in the course of secretory activity in the rat. A m 3 Anat 172, Chrm 255, 1938-1944. 29 1-306. TSIEN, R.Y., POZZAN, T. & RINK,T.J. 1982. Calcium M.K., LIN,Y.N. & HOROWITZ, M. I. 1978. homeostasis in intact lymphocytes : Cytoplasmic WASSEF, Phospholipid-deacylating enzymes of rat stomach free calcium monitored with a new, intracellularljmucosa. Biochim Biaphys Acta 528, 318-330. trapped fluorescent indicator. j’ Cell Btol 94, WASSEF, M.K. & HOROWITZ, M.I. 1981. Degradation 325-334. of phosphatidylinositol by soluble enzymes of rat VEGA, F.V., OLAISSON, H. & >T.~RDH, S. 1985. gastric mucosa. Biochim Biophys ‘4cta 665,234-243. Distribution of carbonic anh!-drase in cells and