32

Btochimica et Biophysica Acta, 1053 (1990) 32-36 Elsevier

BBAMCR 12721

Rat pancreatic acini permeabilised with streptolysin O secrete amylase at Ca 2+ concentrations in the micromolar range, when provided with ATP and GTPyS J. Michael Edwardson, Craig Vickery and Leslie J. Christy Department of Pharmacology, Unwersity of Cambridge, Cambridge ( U.K.) (Received 9 October 1989) (Revised manuscript received 7 February 1990)

Key words: Pancreatic acinus; Permeabilization; Exocytosis regulation

The aim of this study was to define the conditions required for exocytosis in pancreatic acini permeabUised with the bacterial toxin streptolysin O. Treatment of a suspension of acini with streptolysin O caused the release of both the cytoplasmic enzyme lactate dehydrogenase and the zymogen granule enzyme amylase. The release of amylase occurred more quickly than that of lactate dehydrogenase and was smaller in magnitude. In addition, a component of amylase release occurred only in the presence of Ca2+ (at concentrations in the micromolar range), ATP and GTP~,S. We conclude that this component represents an exocytotic event, but that the release of lactate dehydrogenase occurs through toxin-generated lesions. The concentrations of Ca2+, ATP and GTP3,S causing half-maximal exocytosis were 0.7 #M, 0.2 mM and 10 pM, respectively. This system should permit a study of the mechanisms underlying regulated exocytosis in this cell type.

Introduction Despite intense interest over the last few years, the molecular mechanisms underlying the final step in the process of regulated exocytosis, the fusion of the membrane of the secretory granule with the plasma membrane, remain unknown [1]. Permeabilised cell systems have been developed as way of dissecting the exocytotic process, and have been used to examine the roles of Ca 2. [2,3], G-proteins [4,5,6] and ATP [2,5]. One promising approach is to use permeabilised cells to screen antibodies to membrane and cytosolic proteins in order to identify proteins involved in exocytosis. This technique has recently been used to demonstrate that the cytosolic proteins a-fodrin and calpactin are both involved in exocytosis in adrenal chromaffin cells [7,8]. We set out to develop a similar system based on the pancreatic acinus. In the past, pancreatic acinar cells have been permeabilised both by electric discharge [9] and by the use

Abbreviation: LDH, lactate dehydrogenase. Correspondence: J.M. Edwardson, Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge, CB2 IQJ,

U.K.

of the detergents saponin and digitonin [10]. We have used as a permeabilising agent the bacterial toxin, streptolysin O, which is known to generate lesions of about 12 nm in diameter [11] (i.e., large enough to allow antibodies into cells), while preserving basic subcellular organization. In this report we describe the properties of these permeabilised cells, and in particular the requirements for exocytotic release of the granule enzyme, amylase. Materials and Methods Streptolysin O was obtained as a freeze-dried preparation of a partially purified culture filtrate from Wellcome Diagnostics, Dartford, U.K. and was used without further purification. All other chemicals were obtained from Sigma, Poole, U.K. Acini were prepared from rat pancreas by collagenase digestion as described previously by Rogers et al. [12]. For experimental incubations, 500 #1 aliquots of acini in a modified Krebs solution (pH 7.2) [12] were added to 500 /.tl of a buffer containing 73 mM NaCI and 50 mM sodium phosphate (pH 6.5) together with appropriate additions. Streptolysin O was reconstituted immediately before use. ATP was added together with a regenerating cocktail of creatine phosphate (8 mM) and creatine phosphokinase (7 IU/ml). Samples were in-

0167-4889/90/$03.50 .% 1990 Elsevier Science Publishers B,V. (Biomedical Division)

33

cubated at 37 ° C with gentle end-over-end mixing, usually for 15 min. Acini were then cooled on ice and pelleted by centrifugation at 30 rpm for 2 min. Supernatants were assayed for amylase [13] and lactate dehydrogenase [14]. Total cell content of these enzymes was determined after treatment of the samples with 0.2% Triton X-100. Enzyme release was expressed as a percentage of the total cellular enzyme content. Data shown are usually taken from a single experiment. In all cases, however, they are representative of results from at least three experiments. Ca 2+ was buffered using 1 mM EGTA. Free Ca 2÷ concentrations were calculated using an iterative computer program based on that of Fabiato and Fabiato

[151.

50

0

40

30

20

_c "1" Q ,-J

10

o~

~s

5

Results After incubation of pancreatic acini at 3 7 ° C for 15 min, 12% of the total cellular lactate dehydrogenase (LDH) was present in the supernatant. In the presence of streptolysin O, this value rose in a dose-dependent manner, up to a maximum of 60% at 0.5 U / m l strep-

'

o

5

1

time o

'

15

(rain)

70

k

6o

t

so c

Fig. 2. Effect of ATP and G T P y S on release of LDH and amylase from permeabilised pancreatic acini. Acini were incubated with streptolysin O (0.2 U / m l ) for various times at 37°C, in the presence of Ca 2+ (3/xM), and in the absence ( 0 ) or presence (Q) of both ATP (3 mM) and GTPyS (30/.tM). The release of both LDH (a) and amylase (b) was then measured. Values are from a single experiment.

3o

_c "T"

2

2O i 1o

o o

b

10

A c

_c

=;

0!2

o!,

016

0!8

,'0

/

tolysin O (Fig. la). These figures, which are the mean values from 13 separate experiments, conceal considerable variation between experiments in the extent of the release of LDH caused by streptolysin O. The cause of this variation is not known, although it may be a consequence of differences in the properties of the acini 16

i,0 tn

00

01,

0'2 [streptolysin

013 O]

o,'

0'0

(U/ml)

Fig. 1. Dependence of enzyme release on streptolysin O concentration. Pancreatic acini were incubated with streptolysin O in the presence of Ca 2 + (10 ~tM), ATP (1 mM) and GTPTS (100 ~M) for 15 rain at 37 ° C, The acini were then pelleted by centrifugation and the supernatants assayed for LDH (a) and amylase (b). Values in (a) are means :l: S.E. from 13 separate experiments; values in (b) are from a single experiment.

10g [Ca2"] (M) Fig. 3. Dependence of amylase release on Ca 2+ concentration. Acini were incubated with streptolysin O (0.2 U / m l ) for 15 rain at 3 7 ° C at various concentrations of Ca 2+ in the presence of GTP'tS (100 p.M) and in the presence (•) or absence ( O ) of ATP (1 mM). Values are from a single experiment.

34

o~ ca c

/

=___.__....----- /

>.



I o01

zero

l OO3

OI I

DI3

[ATP] (mM)

Fig. 4. Dependence of amylase release on A T P concentratton. Acini were incubated with streptolysin O (0.2 U / m l ) for 15 min at 3 7 ° C at various concentrations of A T P in the presence of Ca 2÷ (10 jaM) and in the presence (~) or absence ( , ) of G T P y S (100 #M). Values are from a single experiment.

(e.g., mean size and viability) produced by collagenase digestion of the pancreas. In the presence of Ca 2÷ (10 #M), ATP (1 mM) and G T P y S (100 #M), streptolysin O brought about a dose-dependent release of amylase, from 6% initially to

D

5

I 5

zero

2

~ z

b

o

I 5

l I~

I 10

I ~0

l ~

I 100

I 100

[GTP'I' S] (~M)

Fig. 5. Dependence of amylase release on GTPTS concentration. Acini were incubated with streptolysin O (0.1 U / m l ) for 15 min at 3 7 ° C at various concentrations of GTP'rS in the presence of Ca 2. (10 # M ) and in the presence (a) or absence (b) of A T P (1 mM). Values are from single experiments.

9% at 0.2 U / m l streptolysin O (Fig. lb). This concentration of streptolysin O was used in most further experiments. In order to gain some insight into the processes underlying the release of the two enzymes, we examined for both enzymes the time-course of release and its dependence on Ca 2÷, ATP and GTPyS. Fig. 2 shows the time-courses of release of LDH and amylase in the presence and absence of an A T P / G T P y S cocktail. It can be seen that the amount of LDH in the supernatant was still rising at the 15 min time point and that enzyme release at all time-points was unaffected by the presence of cocktail (Fig. 2a). Amylase release, on the other hand, consisted of two components, one cocktail-independent and the other cocktail-dependent, that accounted for 7% and 6% of the total cellular enzyme, respectively. Both components of amylase release were almost complete by 5 rain (Fig. 2b). These results indicate that the release of the two enzymes is occurring by different mechanisms. The effect of varying the Ca 2 ÷ concentration on the size of the A T P / G T P v S - d e p e n d e n t component of amylase release is shown in Fig. 3. In this experiment the release of amylase into the supernatant rose to a maximum at 3 ttM Ca 2+ and began to fall again at 10 #M Ca 2". The concentration of Ca 2+ giving half-maximal stimulation of release was 0.7/~M. When ATP was omitted from the 10 /~M Ca 2÷ sample, there was no enhancement of amylase release. Hence, the stimulatory effect of Ca 2+ is absolutely dependent on ATP. There was some variation between experiments in the Ca 2+ sensitivity of amylase release; for example, the Ca'** concentration at which stimulation of amylase release was maximal ranged from 1 /~M up to 10 ~M. This variation may reflect differences in the extent of the permeabilisation of the cells (see above), that would affect the efficiency of EGTA as a buffer of intracellular Ca 2~. In a separate experiment we found that the release of lactate dehydrogenase and the A T P / GTP~,S-independent component of amylase release were unaffected by reduction of the Ca 2+ concentration to 10 nM (data not shown). Hence, only the A T P / GTPyS-dependent component of amylase release requires Ca 2+. Finally, we examined the concentration-dependence of the effects of ATP and GTP-/S on amylase release. At 10 ~M Ca 2+, ATP stimulated amylase release in the presence, but not the absence, of 100 #M G T P y S (Fig. 4); half-maximal stimulation occurred at about 0.2 mM ATP. Interestingly, there is an indication in Fig. 4 that G T P y S is able to to stimulate amylase release to some extent even in the absence of ATP, an effect that was not seen in the experiment illustrated in Fig. 3. This finding suggests that on some occasions intracellular ATP may be depleted either slowly or incompletely. In the reciprocal experiment (Fig. 5), G T P y S stimulated

35 amylase release in the presence (Fig. 5a), but not the absence (Fig. 5b), of ATP (1 mM), with a half-maximal effect at 10/tM.

Discussion In this paper we show that streptolysin O-permeabilised pancreatic acini release amylase in a manner that is dependent on the presence of Ca 2~, buffered in the micromolar range, and also of ATP and GTPyS. These requirements are strongly indicative of exocytotic secretion. Release of LDH, on the other hand, occurs in the absence of these agents and is likely to represent leakage through toxin-generated pores. The nature of the background release of amylase is at present unclear. The requirements for exocytotic release of amylase are different from those reported for exocytosis from acinar cells permeabilised by electric discharge [9] or by saponin or digitonin [10]. When acinar cells were permeabilised by electric discharge [9], release occurred in the presence of Ca 2÷ alone, but was enhanced by the phorbol ester 12-O-tetradecanoylphorbol 13-acetate (TPA). The effect of Ca 2÷ was half-maximal at 2 ~M and peaked at 10 ~M. The differences between these results and our own may be a consequence of the use of single cells rather than acini. Alternatively, they may be related to the fact that the two methods used to achieve permeabilisation are known to produce cells with significantly different properties [16]. For example, electric discharge allows the leakage of ions and small molecules, whereas streptolysin O causes the loss of large proteins. When acini were permeabilised by saponin [10], the Ca 2÷ effect peaked only at 1 mM and was half-maximal between 10 and 100 ~M. Here the effect of Ca 2÷ was enhanced by several agents, such as cAMP and TPA. ATP and GTP'tS were not used, which may explain the relative insensitivity of exocytosis to Ca 2÷ An effect of GTPyS on exocytosis in pancreatic acinar cells has been reported previously by Maruyama [17]. Exocytosis was detected through changes in cell membrane capacitance, measured using a whole-cell patch-clamp technique. It was found that as well as enhancing the actions of the agonists acetylcholine and cholecystokinin octapeptide, GTP~,S itself produced oscillatory changes in membrane capacitance. As in our experiments, the effects of GTP~,S were seen at high concentrations (50-100 ~tM) and required Ca 2÷ at concentrations around 1/~M. Stimulatory effects of GTP~,S on exocytosis have also been demonstrated in permeabilised mast cells and neutrophils by Gomperts and co-workers [4-6]. To account for his findings, Gomperts has proposed the existence of a G-protein (GE) that controls exocytosis directly [18]. At the moment we are unable to say whether the effect of GTP~,S seen in our system is mediated through G E, or whether it is a

consequence of stimulation of a receptor-coupled Gprotein at the plasma membrane. Recently, we have reconstituted in a cell-free system the exocytotic interaction between zymogen granules and pancreatic plasma membranes [19]. The interaction is specific to pancreatic membranes and involves membrane fusion, leading to amylase release from the granules. We find that the granule-membrane interaction is stimulated by activators of G-proteins, such as GTP3,S, but does not require either Ca 2÷ or ATP. The fact that the concentration of GTP~,S that causes half-maximal stimulation of the granule-membrane interaction (15 #M) is very similar to the concentration (10 /~M) that causes half-maximal stimulation of amylase release from permeabilised cells suggests that we may be looking at the same process in the two systems. On the basis of these results, it is tempting to speculate that exocytosis in the exocrine pancreas requires Ca 2*, ATP and Gprotein activation, but that only the G-protein effect is directly involved in membrane fusion. In this scenario Ca 2+ and ATP would operate upstream of fusion; for example, they may be involved in uncoupling the granules from the cytoskeleton, as has been suggested previously [1]. The exocytotic release of amylase from streptolysin O-permeabilised cells typically accounts for about 6% of total cellular amylase. This should be compared with a value of around 12% obtained with the muscarinic agonist bethanechol in unpermeabilised cells prepared in the same way [12]. There are several factors that may account for the reduced response seen in the permeabilised cells. Firstly, all of the cells are not permeabilised, as judged by the release of LDH. Secondly, permeabilisation may damage cells and thereby render them unresponsive. Finally, other agents may be required in order to bring about a maximal response. It is possible, for example, that activation of protein kinase C may have an additional stimulatory effect, since it is known that TPA is effective in the permeabilised cell systems described above [9,10]. A full understanding of the molecular details of regulated exocytosis will require the identification of proteins on the zymogen granule membrane, on the plasma membrane and in the cytosol that are involved in this process. The permeabilised cell system described here provides us with a way of screening antibodies raised against both membrane and cytosolic proteins, and should therefore make possible such an identification.

Acknowledgements We thank the MRC for financial support, and Dr. J. Rogers of this Department for advice about preparation of acini. C.V. was a recipient of a vacation studentship

36 from Fitzwilliam College, Cambridge. L.J.C. was a Rotary Foundation Scholar. References 1 Lindstedt, A.D. and Kelly, R.B. (1987) Trends Neurosci. 10, 446-448. 2 Baker, P.F. and Knight, D.E. (1978) Nature 276, 620-622. 3 Howell, T.W. and Gomperts. B.D. (1987) Biochim. Biophys. Acta 927, 177-183. 4 Barrowman, M.M., Cockcroft, S. and Gomperts, B.D. (1986) Nature 319, 504-507. 5 Howell, T.W., Cockcroft, S. and Gomperts, B.D. (1987) J. Cell Biol. 105, 191-198. 6 Cockcroft, S., Howell, T.W. and Gomperts, B.D. (1987) J. Cell Biol. 105, 2745-2750. 7 Perrin, D., Langley, O.K. and Aunis, D. (1987) Nature 326, 498-501. 8 Ali, S.M., Geisow, M.J. and Burgoyne, R.D. (1989) Nature 340, 313-315.

9 Knight, D.E. and Koh, E. (1984) Cell Calcium 5, 401-418. 10 Kimura, T., Imamura, K., Eckhardt, L. and Schulz, I. (1986) Am. J. Physiol. 250, G-698-G708. 11 Buckingham, L. and Duncan, J. (1983) Biochim. Biophys. Acta 729, 115-122. 12 Rogers, J., Hughes, R.G. and Matthews, E.K. (1988) J. Biol. Chem. 263, 3713-3719. 13 Rinderknecht, H., Wilding, P. and Haverback, B.J. (1967) Experientia 23, 805. 14 Bergmeyer, H.U., Bernt, E. and Hess, B. (1963) in Methods of Enzymatic Analysis (Bergmeyer, H.U., ed.), pp. 736-743, Academic Press, New York. 15 Fabiato, A. and Fabiato. F. (1979) J. Physiol. (Paris) 75, 463-505. 16 Ahnert-Hilger, G. and Gratzl, M. (1988) Trends Pharmacol. Sci. 9, 195-197. 17 Maruyama, Y. (1988) J. Physiol. 406, 299-313. 18 Gomperts, B.D. (1986) Trends Biochem. Sci. 11,290-292. 19 Nadin, C.Y., Rogers, J., Tomlinson, S. and Edwardson, J.M. (1989) J. Cell Biol. 109, 2801-2808.

Rat pancreatic acini permeabilised with streptolysin O secrete amylase at Ca2+ concentrations in the micromolar range, when provided with ATP and GTP gamma S.

The aim of this study was to define the conditions required for exocytosis in pancreatic acini permeabilised with the bacterial toxin streptolysin O. ...
353KB Sizes 0 Downloads 0 Views