- Email: [email protected]
Calcium-independent secretion by ATPγS from a permeabilized rat basophilic leukaemia cell line (RBL-23H)
Cellular Signalling Vol. 6, No. 2, pp. 223--231, 1994. Copyright © 1994 Elsevier Science Ltd Printed in Great Britain. All fights reserved 0898-6568/94 $7.00 + 0.00
Pergamon
C A L C I U M - I N D E P E N D E N T S E C R E T I O N B Y ATPyS F R O M A P E R M E A B I L I Z E D RAT BASOPHILIC LEUKAEMIA CELL LINE (RBL-2H3) RUSSELL I. LUDOWYKE* and LYNDEE L. SCURR Centre for Immunology, St. Vincent's Hospital, Victoria Road, Sydney, NSW 2010, Australia (Received 4 July 1993; and accepted 2 September 1993) Abstract--Activation of rat basophilic leukaemia cells (RBL-2H3) leads to the secretion of allergic and inflammatory mediators. These cells can be permeabilized, yet still retain their ability to secrete in response to antigen. Secretion can also be induced in permeabilized cells by the addition of the ATP analogue, ATPyS [adenosine-5"-O(3-thiotriphosphate)], which is relatively resistant to phosphatase activity. ATPyS-induced secretion (35-50% of total amine) is temperature and concentration-dependent. Calcium enhances secretion, but unlike antigen-induced secretion, it does occur in the absence of calcium and without the requirement for inositol phospholipid hydrolysis. Other ATP analogues induced secretion in the rank order AMP-PNP _>A T t ~ >>> AMP-PCP > ATPt~cS= ATP [AMP-PNP, adenylyl-imidodiphosphate; AMP-PCP, adenylyl (13,y-methylene)-diphosphonate;ATP~S, adenosine5"-O-(1-thiotdphosphate)]. At equimolar concentrations, ATP inhibits ATI~-induced secretion by 50%, but prolonged incubation in the presence of ATPyS surmounts the ATP inhibition. ADP is nearly as effective an inhibitor, but GTP and ITP are ineffective. It is likely that secretion occurs through attachment at an ATP-binding site, effectively blocking the action of a phosphatase, active later in the normal secretory pathway. Key words: ATPyS, mast cells, secretion, Ca2÷-independent, permeabilized.
INTRODUCTION
protein kinase C by d i a c y l g l y c e r o l [5, 6]. Although much is known about the early signal transduction events, little is known about the later stages. We have utilized a major attribute of the RBL-2H3 cells, i.e. their ability to be permeabilized whilst still retaining their ability to secrete when stimulated by antigen, to investigate the later stages. RBL-2H3 cells permeabilized with streptolysin-O (SLO) in the presence of buffered calcium show the same antigen-induced inositol phospholipid hydrolysis [7], secretion and myosin phosphorylation [8, 9] as intact cells. Antigeninduced phosphorylation of the light and heavy chains of myosin by protein kinase C was shown to be temporally correlated with secretion [9]. Recently, in washed, permeabilized, RBL-2H3 cells the s e c r e t o r y response to antigen was restored by the addition of free calcium and protein kinase C isozymes, principally [3 and 8 [10]. Although for antigen-induced secretion, protein kinase C appears to be a critical factor, it was suggested that other necessary pathways also exist
THE RAT basophilic leukaemia cell line RBL-2H3, a cultured analogue of rat mucosal mast cells, is an excellent model system in which to study the mechanisms of exocytosis [1-4]. The time course of secretion and some of the intervening signals which are thought to promote secretion in these cells have been well characterized. These signals are consistent with the theory that the antigeninduced hydrolysis of inositol phospholipids generates synergistic signals for secretion, namely the mobilization of calcium via the production of inositol 1,4,5-trisphosphate and the activation of *To whom correspondence should be addressed, Abbreviations: ATPyS--adenosine-5"-O-(3-thiotriphosphate); AMP-PNP--adenylyl-imidodiphosphate; AMP-PCP-adenylyl ([~,y-methylene)-diphosphonate;ATPtxS--adenosine5"-O-(1-thiotriphosphate); EGTA--ethylenebis-(oxyethylenenitrilo) tetra-acetic acid; 5HT--5-hydroxytryptamine; DNP-BSA--antigen which contained 24 molecules of dinitrophenol conjugated with 1 molecule of bovine serum albumin; PIPES--1,4-piperazinediethane-sulphonic acid; RBL-2H3-rat basophilic leukaemia cell line, sub-clone 2H3; SLO-streptolysin-O.
[lO]. 223
224
R.I. LUDOWYKEand L. L. SCURR
To investigate further the role of phosphorylation and dephosphorylation in secretion, ATPyS has been utilized in permeabilized RBL-2H3 cells. ATPyS m a y be used instead o f A T P by m a n y kinases including myosin light chain kinase [11] and protein kinase C [12], to thiophosphorylate proteins which then remain resistant to the action of most phosphatases [13]. ATPTS has been used previously to influence secretion from permeabilized cells [14-17], and we have shown in permea b i l i z e d R B L - 2 H 3 cells (R. L u d o w y k e , K. Kawasugi and P. French, manuscript submitted), that A T P ~ induces the same morphological and cytoskeletal changes as observed previously for antigen- [ 18-20] or calcium ionophore A23187induced secretion [21 ]. Results presented here indicate that ATPyS alone is sufficient to induce secretion from permeabilized R B L - 2 H 3 cells to equivalent levels as seen with antigen. ATP~/S-induced s e c r e t i o n bypasses the requirement for inositol phospholipid hydrolysis, calcium and potentially protein kinase C, but acts through an ATP-dependent mechanism. As the morphological changes seen during secretion induced by ATPTS or antigen are similar, it is believed that ATPTS is activating a later component of the normal secretory pathway.
MATERIALS AND METHODS Cell culture and labelling RBL-2H3 cells were maintained as monolayer cultures and harvested by trypsin treatment as described previously [1, 7, 9]. Cells were transferred to 24-well culture plates at a density of 2.5 x 105 cells per well in 0.4 ml Earle's modified Eagle's medium supplemented with 15% (v/v) foetal bovine serum (PA Biologicais, Sydney, Australia). For those experiments involving antigenic activation, the cells were sensitized by the addition of 75 ng/ml DNP-specific immunoglobulin E (Sigma Chemical Co., St Louis, MO, U.S.A.) and incubated overnight at 37°C with 5% CO2. To measure the secretory response, cultures were also incubated overnight with [3H]5-hydroxytryptamine ([3H]5HT) (0.5 pCi/ml, Amersham, U.K.), which becomes incorporated into the histamine-containing granules. For the measurement of the hydrolysis of inositol phospholipids, myo-[3H]inositol (5 lxCi/ml, Amersham) was added to the cells for overnight incubation.
Permeabilization and stimulation of cells Permeabilization. The cells were washed twice in a calcium-free, potassium glutamate buffer [22]: 139 mM glutamic acid (monopotassium salt); 5 mM glucose: 7 mM magnesium acetate; 1 mM [ethylenebis(oxyethylenenitrilo)tetra-acetic acid (EGTA) and 20 mM 1,4-piperazinediethanesulphonic acid (PIPES; dipotassium salt)-KOH (pH 7.2). To permeabilize the cells SLO (0.2 U/ml, Wellcome Reagents, U.K.) was added to aliquots of the buffer at 37°C [23]. CaCI2 (0.81 mM) was added as required so the free calcium concentration was 1 ~tM, which is optimal for antigenic activation in permeabilized RBL-2H3 cells [7-10]. ATP (1 mM unless otherwise indicated) was added as required and the pH adjusted accordingly after addition of all reagents.
Stimulation. To initiate antigenic secretion, DNP-BSA with 1 ~tM calcium and 1 mM ATP was added to the cells concurrently with the SLO in 200 I,tl buffer at 37°C. The specific antigen, DNP24BSA (24 molecules of DNP conjugated with 1 molecule of BSA; DNP-BSA; kindly supplied by H. Metzger, NIH, U.S.A.) was used at 100 ng/ml, a concentration optimal for secretion from permeabilized cells [7, 8]. Control wells for unstimulated cells were incubated in the presence of buffer alone. The reaction was stopped by placing the plates on ice and an aliquot of the supematant taken for the determination of the tritium content using a liquid scintillation counter (Canberra Packard, Tricarb 1500). The release of [3H]5HT into the medium was expressed as a percentage of the total cellular content of the amine as determined from lysed, unstimulated cultures or after the remaining amine within the cells had been released by lysis in 1% (v/v) Triton X-100. To initiate secretion by ATP'/S (made up fresh for each experiment) it was added to the cells in 200 pl of SLOcontaining buffer in the same manner as with antigen above. Analogues of ATPTS were used in the same way. ATP, ATP~/S and all other analogues and nucleotides were supplied from Boehringer Mannheim, Australia.
Washed, permeabilized cells. For those experiments utilizing washed cells, SLO was added in buffer containing 1 mM ATP for 10 min at 37°C, and then the cells were washed twice in buffer without ATP. Activators were then added in the absence of SLO as above. It was verified by ethidium bromide staining of the nuclei that the cells had been permeabilized after 5 min incubation with SLO ([7] and data not shown).
ATPTS-inducedsecretion from RBL-2H3cells
Measurement of release of [3H]inositol-labelledproducts The cells labelled with [3H]inositol were treated exactly as described above for ceils labelled with [3H]5HT. At the appropriate times, the reactions were terminated by the addition of 0.75 ml of a mixture of chloroform:methanol:4 N HCI (100:200:2). The mixture was transferred to polypropylene tubes and 0.25 ml of both chloroform and 0.1 N HC1 were added. The organic and aqueous phases were separated by centrifugation and the total water soluble [3H]inositol phosphates were assayed by separation on Dowex-1 formate columns [24, 25]. RESULTS
ATP~S-induced secretion and its Ca2+-dependency The level o f spontaneous or basal secretion from permeabilized RBL-2H3 cells in the presence of 1-5 mM ATP has been shown previously to be between 10 and 20% [8, 9, 20, 26]. We find similar levels in the presence of 0.1-10 mM ATP (Table 1A), although the highest concentration of A T P appears to lower the level of spontaneous release. In the absence of ATP and Ca 2+, ATPyS more than doubles this basal level of secretion in a concentration-dependent manner. The addition of 1 [tM calcium induces a small (5%) increase in secretion in the presence of all concentrations of ATP, but has a marked effect upon secretion induced by low concentrations of ATPyS ( 1 2 - 1 8 % ; Table 1B). To determine the Ca 2÷i n d e p e n d e n t and Ca2+-dependent c o m p o n e n t s
225
of ATPyS-induced secretion, the values of ATP and ATP + Ca 2÷ were subtracted from the corres p o n d i n g values for A T P y S (Table 1C). This indicates there is a high level of Ca2÷-independent s e c r e t i o n i n d u c e d by A T P y S , and a l t h o u g h Ca 2÷ enhances the release, Ca 2+ has little effect at the highest c o n c e n t r a t i o n s o f A T P y S . Figure 1 shows that increasing the Ca 2÷ concentration to 10 ~tM had no g r e a t e r e f f e c t u p o n s e c r e tion induced by 1 mM ATP~3, and also, that as with antigen-induced secretion from permeabilized cells, 1 ~tM Ca 2+ was optimal for enhancement. S e c r e t i o n f r o m intact m a s t cells has b e e n shown to be temperature-dependent and to require membrane fusion [27, 28]. ATPyS-induced secretion is also t e m p e r a t u r e - d e p e n d e n t . The time course of antigen- and ATPyS-induced secretion were similarly affected by temperature with slower rates evident at the lower temperatures (data not shown).
ATPyS-induced effects on inositol phospholipid hydrolysis To determine whether ATPTS-induced secretion required the hydrolysis of inositol phospholipids as a n t i g e n - i n d u c e d s e c r e t i o n did, we analysed the inositol phospholipid hydrolysis during ATPTS activation. Figure 2 shows that the time courses of secretion observed with antigen or ATPyS are similar. As previously observed in per-
TABLE 1. SECRETIONFROM PERMEABILIZEDR B L - 2 H 3 CELLS INDUCEDBY A T P OR ATPyS iN THE PRESENCE AND ABSENCEOF CALCIUM
A % Release
ATP
B % Release
C % Release by ATPTS
or
A'rPTS (raM)
0.1 0.5 1 5 10
ATP
ATPTS
ATP + Ca 2+
ATPyS + Ca 2+
23.1±2.8 19.1±0.8 17.8±0.6 17.5±1.0 13.4±1.4
41.1±1.3 ~.3±1.0 46.0±2.9 49.5±1.7 53.0±1.3
28.4±3.8 23.2±0.8 22.4±0.5 21.0±0.6 19.3±0.5
58.8±4.0 58.5±7.6 61.4±3.8 59.9±1.9 ~.7±1.9
Ca 2+independent
18.0 27.2 28.2 32.0 39.6
Ca ~+dependent
30.4 35.3 39.0 38.9 41.4
Release was measured after 30 min incubation with the concentrationsof ATP or ATlaS given, and is given as a percentage of the total isotope present in the cells. (A) Represents release in the absence of Ca2÷and (B) release in the presence of 1 pM Ca2÷. (C) Represents the ATPTS-inducedrelease which is Ca~÷-independent(ATP'/S - ATP) and ca2+-dependent (ATPTS + Ca2÷- ATP + Ca2÷).The results are the means + S.E. of triplicates from at least three experiments.
226
R.I. LUDOWYKE and L. L. SCURR
80
60
A
mE
~|__//..._o~°
:-20(
I 0.25
0.00
I O.SO
I 0.75
I ,,/// 1.00
I S
10
CN.C' UM ( ~ )
FIG. 1. Calcium influence on secretion. Cells were permeabilized in the presence of either 1 mM ATP (O) or ATP~ (1) and increasing concentrations of calcium as described in Materials and Methods.The reaction was stopped after 30 rain and release of [3H]5HT determined as described. Release is given as a percentage of the total isotope present in the cells. The results are the means and S.E. of triplicates from at least three separate experiments. meabilized RBL-2H3 cells, the time course and levels of inositol phospholipid hydrolysis induced by antigen correlated directly with secretion [7]. H o w e v e r with ATPTS there was initially only
-[
basal inositol phospholipid hydrolysis observed. At 10 min, when secretion induced by A T P ~ was 40% there was only 6% inositol phospholipid hydrolysis. At this time, secretion and inositol phospholipid hydrolysis induced by antigen were both about 40%. After 10 rain, there was a slow increase in the inositol phospholipid hydrolysis induced by A T P ~ (20% after 30 min), but not to the extent observed with antigen (70%). Even this level of ATPTS-induced inositol phospholipid hydrolysis was reduced to control levels, by washing the cells following permeabilization and prior to addition of ATPTS (Table 2). This table also shows that antigen-induced inositol phospholipid hydrolysis is inhibited by as much as 50% by this washing procedure. Secretion induced by the two a c t i v a t o r s is c o m p l e t e l y d i f f e r e n t h o w e v e r . Antigen-induced secretion is reduced to control levels by washing, as observed previously [7, 10], whereas ATPTS-induced secretion is slightly enhanced. Calcium still enhances the ATPTS response in washed cells, but not to the same extent as in unwashed cells. It is evident then that unlike antigen, ATP~-induced secretion does not require inositol phospholipid hydrolysis and that secretion occurs as effectively in washed cells as in unwashed cells.
IiO UCRETION
p
tM
O,
o
s
,o
,s MINUTES
:o
es
so
.
o
,
•
,'o
,'s
='o
.
:'o
MINUTES
l~c. 2. ATlaS-induced secretion and inositol phospholipid hydrolysis. Cells prelabelled with either [3H]5HT or [3H]inositol were stimulated with ATP/caleium (0, ©), ATl~/S/calcium (ll, n ) or DNP-BSMcalcium (&, A) for 1-30 rain. The release of soluble inositol phosphates were measured as described in Materials and Methods and are given as described in the legend to Fig. 1.
A T l ~ - i n d u c e d secretion from RBL-2H3 cells
TABLE 2. SECRETION AND INOSITOL PHOSPHOLIPID HYDROLYSIS IN PERMEABILIZED AND UNWASHED (UV~) OR WASHED (W) RBL2H3 C~a.S
Secretion ATP + Ca 2÷ DNP-BSA ATPyS ATP3tS + Ca2+
UW W UW W UW W UW W
21.8 ± 2.6 17.3 ± 1.8 41.4 ± 2.5 18.5 ± 3.3 55.7 ± 3.6 62.4 ± 4.2 67.1±3.9 69.3 ± 2.7
PI-hydrolysis 7.5 ± 4.3 2.8 ± 1.3 48.9 ± 5.1 19.7 ± 0.4 14.7 ± 3.6 3.6 ± 0.9 18.7±3.7 4.3 ± 0.2
Cells pre-labelled with either [3H]5HT (secretion) or [3H]inositol (PI-hydrolysis) were incubated with SLO for 10 min, and
then gently washed. They were then activated with ATP + Ca 2÷, ATP'/S ± Ca 2+ or DNP-BSA+Ca 2+ for 30 rain. Secretion and hydrolysis of washed cells were compared with permeabilized cells which had not been washed. Results are given as a percentage of the total isotope present in the cells and are the means and ± S.E. of triplicates from at least three experiments.
Secretory effects of nucleotide phosphates and ATP analogues To understand further the mechanism by which ATPTS c a u s e d secretion, a n u m b e r of other nucleotide phosphates were added to the permeabilized RBL-2H3 cells. ITP, GTP and ADP at 1 mM all showed a low level of secretion between 18 and 24% which was similar to the spontaneous secretion seen with ATP (Table 3). Up to 5 mM of these nucleotides had no greater effect upon secretion (data not shown). Analogues of ATP were also investigated and ATPcxS, which has the sulphydryl group attached to the s-phosphate of the triphosphate chain, was unable to cause any greater secretion than the nucleotide phosphates above. However, AMP-PNP which has a nitrogen group replacing the terminal bridge oxygen of the triphosphate chain was able to elicit slightly higher levels of secretion (47.2%) than that of A T P ~ (38.4%). A similar analogue, AMP-PCP, whose terminal bridge oxygen is replaced by a CH 2 group was only able to cause low levels of secretion (27.8%). To investigate whether they were affecting the same site, some of the analogues were combined as also shown in Table 3. AMP-PCP had no effect on secretion induced by A ~ or AMP-PNP. Secretion in the presence of
227
equimolar concentrations of ATPTS and AMPPNP (1 mM) was similar to that of A T P ~ alone, suggesting that ATPTS bound to the site with greater efficacy.
Inhibition of ATPyS-induced secretion by nucleotide phosphates To determine whether the other nucleotide phosphates affected ATPTS-induced secretion, cells were incubated with 1 mM ATPTS in the presence of increasing amounts of ATP, ADP, G T P and ITP. Figure 3 demonstrates that an equimolar concentration of ATP inhibits secretion to 50% of that seen with ATPTS alone. At 5 mM ATP, secretion was reduced to basal levels. ADP was also an effective inhibitor at 1 mM, but complete inhibition at higher concentrations was not observed. The two other nucleotides however, GTP and ITP, were only slightly inhibitory at 5 raM, and had little effect at 1 raM. To investigate further the inhibition of ATP'fS-induced secretion by ATP, cells were permeabilized in the presence of 1 mM ATP, 1 mM ATP'/S or 1 mM ATP + 1 mM ATP~S, for up to 90 min. Figure 4 shows that after 30 rnin (as also described in Fig. 3) release in the presence of both ATP and ATP'/S is about half of that in the presence of A T P ' ~ alone. Secretion induced by ATPTS in the presence of ATP continued to increase however, such that by 75-90 TABLE 3. SECRETION INDUCED BY VARIOUS NUCLEOTIDES AND ANALOGUF~FROMPERMEABILIZEDRBL-2H3 CELLS Nucleotide ITP GTP ADP ATP ATPctS ATI~ AMP-PNP AMP-PCP ATP'IS + AMP-PCP ATP'/S + AMP-PNP AMP-PNP + AMP-PCP
% Total release 23.9 + 2.6 20.2 ± 4.2 18.2 ± 1.6 17.8 ± 0.6 19.6 ± 2.4 38.4 ± 1.1 47.2 ± 3.0 27.8 _+ 1.1 35.9 ± 2.1 38.2 ± 3.4 46.6 ± 3.3
The concentration of each nucleotide or analogue was 1 mM and release was measured after 30 min incubation and is given as a percentage of the total isotope present in the cells. Results are means ± S.E. of triplicates from at least three separate experiments.
228
R.I. LUDOWYKE and L. L. SCURR
40
-
N)P
i CONGENTRATION
(raM)
Fio. 3. Inhibition of ATP~-induced secretion by other nucleotides. Cells were incubated with 1 mM ATP~ and increasing concentrations of ATP (IlL ADP (O), GTP (1-'1)and rrP (O) for 30 min. Results are given as a percentage of the release in the presence of ATP# alone (40-45%), and are calculated from the means of triplicates in at least three separate experiments. rain there is little difference in the level of secretion observed from cells activated by ATP~S in the presence or absence of ATP. Secretion in the presence of ATP or ATPyS alone had plateaued by about 45 rain. DISCUSSION The results presented here document some of the b i o c h e m i c a l parameters associated with A T P T S - i n d u c e d secretion, which lead us to h y p o t h e s i z e that A T P T S - i n d u c e d s e c r e t i o n 80
Z
6O
tu
m 9t ttl
40
..I
till
at
2O
0 m
.
0
|
15
-
|
30
-
|
-
45
,
60
-
a
75
•
90
MINUTES
FIG. 4. Prolonged incubation of ATP# + ATP. Cells were incubated with 1 mM ATP (O), 1 mM ATPyS (ll) or 1 mM ATP + 1 mM ATPyS (A) for up to 90 min. Release and results are given as described in the legend to Fig. 1.
bypasses the initial signal transduction mechanisms to activate a later event in the secretory pathway, which in itself is sufficient to induce secretion. We have also shown that ATPyS added to permeabilized RBL-2H3 cells, mimics the antigen-induced actin rearrangement and distinct morphological changes which occur during secretion (R. Ludowyke, K. Kawasugi and P. French, manuscript submitted). As observed previously in permeabilized RBL2H3 cells [8, 9, 20, 26], ATP alone induces a low level of secretion which is considered to be spontaneous, and calcium only enhances this release by a small amount (Table 1). We and others suggest that this is due to fusion of granules immediately adjacent to the plasma membrane, but that another signal is required for further, activated secretion [29-31]. The addition of ATP'yS to the permeabilized cells caused a concentration-dependent increase in secretion to the levels observed with antigen. As with antigen-induced secretion, this was temperature- and time-dependent (Fig. 2). At low ATPTS concentrations, Ca 2+ enhanced secretion, but at high (10 mM) concentrations there was little enhancement evident. The optimal concentration of calcium for enhancement (1 BM; Fig. 1) is similar to that reported previously for enhancement of antigen- and GTPyS-induced secretion from permeabilized RBL-2H3 cells [7-10]. A major difference between the mechanism of secretion induced by antigen and ATP~,S is observed, however, when we compare the requirement for inositol p h o s p h o l i p i d h y d r o l y s i s . Antigen-induced secretion from both intact and permeabilized RBL-2H3 cells has been clearly shown to require the hydrolysis of inositol phospholipids and activation of protein kinase C [3, 4, 7-10]. With ATP'yS-induced secretion, however, there was no prerequisite for the hydrolysis of inositol phospholipids for optimal secretion. Five minutes after the addition of A T t ~ or antigen, 25-30% of the total [3H]5HT has been secreted, compared with less than 5% in the presence of ATP (Fig. 2). At this time there is 25% hydrolysis of inositol phospholipids induced by antigen in stark contrast to the basal (< 5%) levels observed
ATP'/S-inducedsecretionfromRBL-2H3cells with ATP or A T P ~ . After 5 min, there is a slow increase in inositol phospholipid hydrolysis induced by ATP~S. Even this level of hydrolysis is reduced to control levels when the permeabilized cells are washed before the addition of A T P ~ , although this was not due to the loss of phospholipase C as there was still a significant, albeit reduced, level of hydrolysis induced by antigen. However, secretion induced by antigen is lowered to control levels by this washing procedure, as also noted previously [8, 10]. In contrast again, washing slightly enhances the ATPySinduced secretion in the presence or absence of calcium (Table 2). It was suggested to be the loss of protein kinase C during washing that accounts for the inhibiton of subsequent antigen-induced secretion [10, 32], which also suggests that protein kinase C is not involved with ATP'?S-induced secretion, or that in our hands, one of the protein kinase C isozymes may not have been totally removed by washing [10]. Other nucleotide phosphates, ADP, ITP and GTP, were, like ATP, unable to induce the levels of secretion observed with antigen or ATPyS (Table 3). ATPtxS was also unable to induce higher levels of secretion than ATP, suggesting that it was a modification of the y-phosphate that was important for secretion. However, the ATP analogue, AMP-PNP, which has an imido bond between the terminal phosphates was able to induce levels of secretion (47%) slightly higher than that of ATPyS (38%). Another ATP analogue with a modified terminal phosphate bond, AMPPCP induced only low levels of secretion (28%). Therefore the rank order of secretory activity from RBL-2H3 cells of ATP and its analogues is AMPPNP > ATF"yS >>> AMP-PCP > ATP. In comparing the analogues AMP-PNP and AMP-PCP with ATP, it was shown that the bond angle of the l?~----y phosphates of P---C--P was quite different to that of ATP, whereas the bond angle of P - - N P was closer to that of ATP [33]. The similarity to ATP then, appears to be important for secretion. When the analogues were added to the permeabilized RBL-2H3 cells, it was evident that AMP-PCP had no effect on secretion induced by either ATP'~S or AMP-PNP. However, the level of secretion in the
229
presence of ATPyS and AMP-PNP was the same as that of ATPTS alone. This suggests they are both binding to the same site and that ATP~S, assumed to have the same bond angle as ATP, has a higher efficacy for the site or that they may bind to different sites and AqTTS affects the binding to the AMP-PNP site. The inhibition of ATP'yS-induced secretion by ATP and ADP (Fig. 3) is also evidence that secretion is occurring through an ATP-driven system. At equimolar concentrations of ATP and ATP'yS there was 50% of the secretion observed with ATF~S alone. A five-fold excess of ATP reduced the ATP~S-induced secretion to baseline levels which suggests that ATP and A T P ~ are competing for the same site. ADP was nearly as active an inhibitor as ATP, but ITP and GTP were almost ineffective, except at 5 mM where about 30% inhibition was achieved. During storage, the breakdown product of ATPTS is ADP and therefore the presence of ADP in the ATPTS solution would lower the secretion observed with ATP~/S. This could explain the slight variations of secretion observed with ATPyS, depending on the duration for which a particular batch was used. Longer incubations of the cells with both ATPTS and ATP show that ATPTS can slowly overcome the inhibition by ATP (Fig. 4). These results suggest that the turnover of ATP on particular binding sites eventually sees the replacement of ATP with ATP~/S, which is not removed and subsequently induces secretion. The ~l--y phosphate bond of ATPyS may be cleaved by a few enzymes [11, 12], but AMPPNP is considered non-hydrolysable. AMP-PNP is considered a good agent to delineate mechanisms where binding to an ATP site is more important than hydrolysis of the y-phosphate [34]. Our data suggests then, that secretion is induced by these analogues through binding to an ATPbinding site and that the efficacy for the binding site is important, but that hydrolysis may not be necessary. The inability of these analogues to be removed by most phosphatases may also be a mechanism of action [13]. Alternatively ATPyS and AMP-PNP may induce secretion through different mechanisms, independently of hydrolysis.
230
R.I. LUDOWYKEand L. L. SCURR
We would hypothesize that in the unstimulated RBL-2H3 cell, phosphatase activity maintains the cell in a non-secreting state. Activating a kinase by antigen-stimulation, or overcoming the effect o f this phosphatase with agents like ATPTS or AMP-PNP, allows secretion to occur. As GTP did not affect ATPyS-induced secretion, the point at which ATP'~S acts must be downstream from the putative G-protein, Ge [35], or be activating a separate p a t h w a y . O u r p r e l i m i n a r y e v i d e n c e has shown that few proteins are thiophosphorylated with ATp73sS, similar to that observed in chromaffin cells [15], but that one of these proteins is m y o s i n . O u r r e c e n t d a t a (R. L u d o w y k e , K. Kawasugi and P. French, manuscript submitted) also showed that ATP'I,S and antigen both initiated the same morphological and cytoskeletal changes during secretion. This suggests their pathways are linked and that ATPTS is activating a later component of the normal secretory pathway. Acknowledgements--R.I.L. is supported by the E. Steinberg Leukemia Research Fellowship, and L.L.S. is supported by a grant from N.H. & M.R.C. We are grateful for the technical assistance from MARTIN O'StJLLIVAN who was supported by a grant from the Clive and Vera Ramaciotti Foundation. We are very grateful for the helpful discussions with MICHAELA. BEAVEN, N.I.H., U.S.A., and RON PENNY, PETER FRENCH, MARGARETCOOLEY,JAMES WILLIAMSONand others of the Centre for Immunology.
REFERENCES 1. Barsumian E. L., Isersky C., Petrino M. G. and Siraganian R. P. (1981) Eur. J. lmmun. 11, 317-323. 2. Metzger H. G., Alcarez G., Hohman A. R., Kinet J.-P., Pribluda V., and Quarto R. (1986) A. Rev. Immun. 4, 419-470. 3. Beaven M. A. and Ludowyke R. I. (1989) Advances in Regulation of Cell Growth, Vol. 1: Regulation of Cell Growth and Activation (Mond J. J., Cambier J. C. and Weiss A., Eds), pp. 245-285. Raven Press, New York. 4. Beaven M. A. and Cunha-Melo J. R. (1988) Progress in Allergy (Becker E. L., Ed), Vol. 42, pp. 123-184. S. Karger AG, Basel, Switzerland. 5. Berridge M. J. (1987) A. Rev. Biochem. 56, 159-193.
6. Nishizuka Y. (1988) Nature 334, 661-665. 7. Ali H., Cunha-Melo J. R. and Beaven M. A. (1989) Biochim. biophys. Acta 1010, 88-99. 8. Ali H., Collado-Escobar D. M., and Beaven M. A. (1989) J. Immun. 143, 2626--2633. 9. Ludowyke R. I., Peleg I., Beaven M. A. and Adelstein R. S. (1989) J. biol. Chem. 264, 12492-12501. 10. Ozawa K., Szallais Z., Kazanietz M. G., Blumberg P. M., Mischak H., Mushinski J. F. and Beaven M. A. J. biol. Chem. 268, 1749-1756. 11. Haeberle J. R., Sutton T. A. and Trockman B. A. (1988) J. biol. Chem. 263, 442 a. A,A,29. 12. Giambalvo C. T. (1988) Biochem. Pharmac. 37, 4001-4007. 13. Eckstein F. (1985) A. Rev. Biochem. 54, 367402. 14. Brooks J. C., Treml S. and Brooks M. (1984) Life Sci. 35, 569-574. 15. Brooks J. C. and Brooks M. (1985) Life Sci. 37, 1869-1875. 16. Wagner P. D. and Vu N-D. (1989) J. biol. Chem. 264, 19614-19620. 17. Wagner P. D. and Vu N-D. (1990) J. biol. Chem. 265, 10352-10357. 18. Oliver J. M., Seagrave J-C., Stump R. F,, Pfeiffer J. R. and Deanin G. G. (1988) Progress in Allergy, Vol. 42 (Ishizaka K., KaUos P., Lachmann P. J. and Waksman B. H., Eds), pp. 185-245. Karger, Basel, Switzerland. 19. Pfeiffer J. R., Seagrave J. C., Davis B. H., Deanin G. G. and Oliver J. M. (1985) J. Cell Biol. 101, 2145-2155. 20. Deanin G. G., Pfeiffer J. R., Cutts J. L., Fore M. L. and Oliver J. M. (1991) Cell Reg. 2, 627-640. 21. Sahara N., Siraganian R. P. and Oliver C. (1990) J. Histochem. Cytochem. 38, 975-983. 22. Knight D. E. and Baker P. F. (1982) J. Membrane Biol. 68, 107-140. 23. Howell T. W. and Gomperts B. D. (1987) Biochim. biophys. Acta 927, 177-183. 24. Berridge M. J., Dawson R. M., Downes C. P., Heslop J. P. and Irvine R. F. (1983) Biochem. J. 212, 473-482. 25. Beaven M. A., Moore J. P., Smith G. A., Hesketh T. R. and Metcalfe J. C. (1984) J. biol. Chem. 259, 7137-7142. 26. De Matteis M. A., Di Tullio G., Buccione R., and Luini A. (1991) J. biol. Chem. 266, 10452-10460. 27. Lagunoff D. and Wan H. (1974) J. Cell Biol. 61, 809-811. 28. LagunoffD. (1973) J. Cell Biol. 57, 252-259. 29. Sontag J.-M., Aunis D. and Bader M.-F. (1988) Eur. J. Cell Biol. 46, 316--326. 30. Burgoyne R. D., Morgan A. and O'Sullivan A. J. (1989) Cell. Signal. 1, 323-334.
ATP~-induced secretion from RBL-2H3cells 31. Koffer A., Tatham P. E. R. and Gomperts B. D. (1990) J. Cell Biol. 111, 919-927. 32. Cunha-Melo J. R., Gonzaga H. M. S., Ali H., Huang F. L., Huang K.-P. and Beaven M. A. (1989) J. Immun. 143, 2617-2625.
231
33. Yount R. G., Babcock D., BaUantyne W. and Ojala D. (1971) Biochemistry 10, 2484-2489. 34. Yount R. G., Ojala D. and Babcock D. (1971) Biochemistry 10, 2490-2496. 35. Tatham P. E. R. and Gomperts B. D. (1989) Biosci. Rep. 9, 99-109.
Pergamon
C A L C I U M - I N D E P E N D E N T S E C R E T I O N B Y ATPyS F R O M A P E R M E A B I L I Z E D RAT BASOPHILIC LEUKAEMIA CELL LINE (RBL-2H3) RUSSELL I. LUDOWYKE* and LYNDEE L. SCURR Centre for Immunology, St. Vincent's Hospital, Victoria Road, Sydney, NSW 2010, Australia (Received 4 July 1993; and accepted 2 September 1993) Abstract--Activation of rat basophilic leukaemia cells (RBL-2H3) leads to the secretion of allergic and inflammatory mediators. These cells can be permeabilized, yet still retain their ability to secrete in response to antigen. Secretion can also be induced in permeabilized cells by the addition of the ATP analogue, ATPyS [adenosine-5"-O(3-thiotriphosphate)], which is relatively resistant to phosphatase activity. ATPyS-induced secretion (35-50% of total amine) is temperature and concentration-dependent. Calcium enhances secretion, but unlike antigen-induced secretion, it does occur in the absence of calcium and without the requirement for inositol phospholipid hydrolysis. Other ATP analogues induced secretion in the rank order AMP-PNP _>A T t ~ >>> AMP-PCP > ATPt~cS= ATP [AMP-PNP, adenylyl-imidodiphosphate; AMP-PCP, adenylyl (13,y-methylene)-diphosphonate;ATP~S, adenosine5"-O-(1-thiotdphosphate)]. At equimolar concentrations, ATP inhibits ATI~-induced secretion by 50%, but prolonged incubation in the presence of ATPyS surmounts the ATP inhibition. ADP is nearly as effective an inhibitor, but GTP and ITP are ineffective. It is likely that secretion occurs through attachment at an ATP-binding site, effectively blocking the action of a phosphatase, active later in the normal secretory pathway. Key words: ATPyS, mast cells, secretion, Ca2÷-independent, permeabilized.
INTRODUCTION
protein kinase C by d i a c y l g l y c e r o l [5, 6]. Although much is known about the early signal transduction events, little is known about the later stages. We have utilized a major attribute of the RBL-2H3 cells, i.e. their ability to be permeabilized whilst still retaining their ability to secrete when stimulated by antigen, to investigate the later stages. RBL-2H3 cells permeabilized with streptolysin-O (SLO) in the presence of buffered calcium show the same antigen-induced inositol phospholipid hydrolysis [7], secretion and myosin phosphorylation [8, 9] as intact cells. Antigeninduced phosphorylation of the light and heavy chains of myosin by protein kinase C was shown to be temporally correlated with secretion [9]. Recently, in washed, permeabilized, RBL-2H3 cells the s e c r e t o r y response to antigen was restored by the addition of free calcium and protein kinase C isozymes, principally [3 and 8 [10]. Although for antigen-induced secretion, protein kinase C appears to be a critical factor, it was suggested that other necessary pathways also exist
THE RAT basophilic leukaemia cell line RBL-2H3, a cultured analogue of rat mucosal mast cells, is an excellent model system in which to study the mechanisms of exocytosis [1-4]. The time course of secretion and some of the intervening signals which are thought to promote secretion in these cells have been well characterized. These signals are consistent with the theory that the antigeninduced hydrolysis of inositol phospholipids generates synergistic signals for secretion, namely the mobilization of calcium via the production of inositol 1,4,5-trisphosphate and the activation of *To whom correspondence should be addressed, Abbreviations: ATPyS--adenosine-5"-O-(3-thiotriphosphate); AMP-PNP--adenylyl-imidodiphosphate; AMP-PCP-adenylyl ([~,y-methylene)-diphosphonate;ATPtxS--adenosine5"-O-(1-thiotriphosphate); EGTA--ethylenebis-(oxyethylenenitrilo) tetra-acetic acid; 5HT--5-hydroxytryptamine; DNP-BSA--antigen which contained 24 molecules of dinitrophenol conjugated with 1 molecule of bovine serum albumin; PIPES--1,4-piperazinediethane-sulphonic acid; RBL-2H3-rat basophilic leukaemia cell line, sub-clone 2H3; SLO-streptolysin-O.
[lO]. 223
224
R.I. LUDOWYKEand L. L. SCURR
To investigate further the role of phosphorylation and dephosphorylation in secretion, ATPyS has been utilized in permeabilized RBL-2H3 cells. ATPyS m a y be used instead o f A T P by m a n y kinases including myosin light chain kinase [11] and protein kinase C [12], to thiophosphorylate proteins which then remain resistant to the action of most phosphatases [13]. ATPTS has been used previously to influence secretion from permeabilized cells [14-17], and we have shown in permea b i l i z e d R B L - 2 H 3 cells (R. L u d o w y k e , K. Kawasugi and P. French, manuscript submitted), that A T P ~ induces the same morphological and cytoskeletal changes as observed previously for antigen- [ 18-20] or calcium ionophore A23187induced secretion [21 ]. Results presented here indicate that ATPyS alone is sufficient to induce secretion from permeabilized R B L - 2 H 3 cells to equivalent levels as seen with antigen. ATP~/S-induced s e c r e t i o n bypasses the requirement for inositol phospholipid hydrolysis, calcium and potentially protein kinase C, but acts through an ATP-dependent mechanism. As the morphological changes seen during secretion induced by ATPTS or antigen are similar, it is believed that ATPTS is activating a later component of the normal secretory pathway.
MATERIALS AND METHODS Cell culture and labelling RBL-2H3 cells were maintained as monolayer cultures and harvested by trypsin treatment as described previously [1, 7, 9]. Cells were transferred to 24-well culture plates at a density of 2.5 x 105 cells per well in 0.4 ml Earle's modified Eagle's medium supplemented with 15% (v/v) foetal bovine serum (PA Biologicais, Sydney, Australia). For those experiments involving antigenic activation, the cells were sensitized by the addition of 75 ng/ml DNP-specific immunoglobulin E (Sigma Chemical Co., St Louis, MO, U.S.A.) and incubated overnight at 37°C with 5% CO2. To measure the secretory response, cultures were also incubated overnight with [3H]5-hydroxytryptamine ([3H]5HT) (0.5 pCi/ml, Amersham, U.K.), which becomes incorporated into the histamine-containing granules. For the measurement of the hydrolysis of inositol phospholipids, myo-[3H]inositol (5 lxCi/ml, Amersham) was added to the cells for overnight incubation.
Permeabilization and stimulation of cells Permeabilization. The cells were washed twice in a calcium-free, potassium glutamate buffer [22]: 139 mM glutamic acid (monopotassium salt); 5 mM glucose: 7 mM magnesium acetate; 1 mM [ethylenebis(oxyethylenenitrilo)tetra-acetic acid (EGTA) and 20 mM 1,4-piperazinediethanesulphonic acid (PIPES; dipotassium salt)-KOH (pH 7.2). To permeabilize the cells SLO (0.2 U/ml, Wellcome Reagents, U.K.) was added to aliquots of the buffer at 37°C [23]. CaCI2 (0.81 mM) was added as required so the free calcium concentration was 1 ~tM, which is optimal for antigenic activation in permeabilized RBL-2H3 cells [7-10]. ATP (1 mM unless otherwise indicated) was added as required and the pH adjusted accordingly after addition of all reagents.
Stimulation. To initiate antigenic secretion, DNP-BSA with 1 ~tM calcium and 1 mM ATP was added to the cells concurrently with the SLO in 200 I,tl buffer at 37°C. The specific antigen, DNP24BSA (24 molecules of DNP conjugated with 1 molecule of BSA; DNP-BSA; kindly supplied by H. Metzger, NIH, U.S.A.) was used at 100 ng/ml, a concentration optimal for secretion from permeabilized cells [7, 8]. Control wells for unstimulated cells were incubated in the presence of buffer alone. The reaction was stopped by placing the plates on ice and an aliquot of the supematant taken for the determination of the tritium content using a liquid scintillation counter (Canberra Packard, Tricarb 1500). The release of [3H]5HT into the medium was expressed as a percentage of the total cellular content of the amine as determined from lysed, unstimulated cultures or after the remaining amine within the cells had been released by lysis in 1% (v/v) Triton X-100. To initiate secretion by ATP'/S (made up fresh for each experiment) it was added to the cells in 200 pl of SLOcontaining buffer in the same manner as with antigen above. Analogues of ATPTS were used in the same way. ATP, ATP~/S and all other analogues and nucleotides were supplied from Boehringer Mannheim, Australia.
Washed, permeabilized cells. For those experiments utilizing washed cells, SLO was added in buffer containing 1 mM ATP for 10 min at 37°C, and then the cells were washed twice in buffer without ATP. Activators were then added in the absence of SLO as above. It was verified by ethidium bromide staining of the nuclei that the cells had been permeabilized after 5 min incubation with SLO ([7] and data not shown).
ATPTS-inducedsecretion from RBL-2H3cells
Measurement of release of [3H]inositol-labelledproducts The cells labelled with [3H]inositol were treated exactly as described above for ceils labelled with [3H]5HT. At the appropriate times, the reactions were terminated by the addition of 0.75 ml of a mixture of chloroform:methanol:4 N HCI (100:200:2). The mixture was transferred to polypropylene tubes and 0.25 ml of both chloroform and 0.1 N HC1 were added. The organic and aqueous phases were separated by centrifugation and the total water soluble [3H]inositol phosphates were assayed by separation on Dowex-1 formate columns [24, 25]. RESULTS
ATP~S-induced secretion and its Ca2+-dependency The level o f spontaneous or basal secretion from permeabilized RBL-2H3 cells in the presence of 1-5 mM ATP has been shown previously to be between 10 and 20% [8, 9, 20, 26]. We find similar levels in the presence of 0.1-10 mM ATP (Table 1A), although the highest concentration of A T P appears to lower the level of spontaneous release. In the absence of ATP and Ca 2+, ATPyS more than doubles this basal level of secretion in a concentration-dependent manner. The addition of 1 [tM calcium induces a small (5%) increase in secretion in the presence of all concentrations of ATP, but has a marked effect upon secretion induced by low concentrations of ATPyS ( 1 2 - 1 8 % ; Table 1B). To determine the Ca 2÷i n d e p e n d e n t and Ca2+-dependent c o m p o n e n t s
225
of ATPyS-induced secretion, the values of ATP and ATP + Ca 2÷ were subtracted from the corres p o n d i n g values for A T P y S (Table 1C). This indicates there is a high level of Ca2÷-independent s e c r e t i o n i n d u c e d by A T P y S , and a l t h o u g h Ca 2÷ enhances the release, Ca 2+ has little effect at the highest c o n c e n t r a t i o n s o f A T P y S . Figure 1 shows that increasing the Ca 2÷ concentration to 10 ~tM had no g r e a t e r e f f e c t u p o n s e c r e tion induced by 1 mM ATP~3, and also, that as with antigen-induced secretion from permeabilized cells, 1 ~tM Ca 2+ was optimal for enhancement. S e c r e t i o n f r o m intact m a s t cells has b e e n shown to be temperature-dependent and to require membrane fusion [27, 28]. ATPyS-induced secretion is also t e m p e r a t u r e - d e p e n d e n t . The time course of antigen- and ATPyS-induced secretion were similarly affected by temperature with slower rates evident at the lower temperatures (data not shown).
ATPyS-induced effects on inositol phospholipid hydrolysis To determine whether ATPTS-induced secretion required the hydrolysis of inositol phospholipids as a n t i g e n - i n d u c e d s e c r e t i o n did, we analysed the inositol phospholipid hydrolysis during ATPTS activation. Figure 2 shows that the time courses of secretion observed with antigen or ATPyS are similar. As previously observed in per-
TABLE 1. SECRETIONFROM PERMEABILIZEDR B L - 2 H 3 CELLS INDUCEDBY A T P OR ATPyS iN THE PRESENCE AND ABSENCEOF CALCIUM
A % Release
ATP
B % Release
C % Release by ATPTS
or
A'rPTS (raM)
0.1 0.5 1 5 10
ATP
ATPTS
ATP + Ca 2+
ATPyS + Ca 2+
23.1±2.8 19.1±0.8 17.8±0.6 17.5±1.0 13.4±1.4
41.1±1.3 ~.3±1.0 46.0±2.9 49.5±1.7 53.0±1.3
28.4±3.8 23.2±0.8 22.4±0.5 21.0±0.6 19.3±0.5
58.8±4.0 58.5±7.6 61.4±3.8 59.9±1.9 ~.7±1.9
Ca 2+independent
18.0 27.2 28.2 32.0 39.6
Ca ~+dependent
30.4 35.3 39.0 38.9 41.4
Release was measured after 30 min incubation with the concentrationsof ATP or ATlaS given, and is given as a percentage of the total isotope present in the cells. (A) Represents release in the absence of Ca2÷and (B) release in the presence of 1 pM Ca2÷. (C) Represents the ATPTS-inducedrelease which is Ca~÷-independent(ATP'/S - ATP) and ca2+-dependent (ATPTS + Ca2÷- ATP + Ca2÷).The results are the means + S.E. of triplicates from at least three experiments.
226
R.I. LUDOWYKE and L. L. SCURR
80
60
A
mE
~|__//..._o~°
:-20(
I 0.25
0.00
I O.SO
I 0.75
I ,,/// 1.00
I S
10
CN.C' UM ( ~ )
FIG. 1. Calcium influence on secretion. Cells were permeabilized in the presence of either 1 mM ATP (O) or ATP~ (1) and increasing concentrations of calcium as described in Materials and Methods.The reaction was stopped after 30 rain and release of [3H]5HT determined as described. Release is given as a percentage of the total isotope present in the cells. The results are the means and S.E. of triplicates from at least three separate experiments. meabilized RBL-2H3 cells, the time course and levels of inositol phospholipid hydrolysis induced by antigen correlated directly with secretion [7]. H o w e v e r with ATPTS there was initially only
-[
basal inositol phospholipid hydrolysis observed. At 10 min, when secretion induced by A T P ~ was 40% there was only 6% inositol phospholipid hydrolysis. At this time, secretion and inositol phospholipid hydrolysis induced by antigen were both about 40%. After 10 rain, there was a slow increase in the inositol phospholipid hydrolysis induced by A T P ~ (20% after 30 min), but not to the extent observed with antigen (70%). Even this level of ATPTS-induced inositol phospholipid hydrolysis was reduced to control levels, by washing the cells following permeabilization and prior to addition of ATPTS (Table 2). This table also shows that antigen-induced inositol phospholipid hydrolysis is inhibited by as much as 50% by this washing procedure. Secretion induced by the two a c t i v a t o r s is c o m p l e t e l y d i f f e r e n t h o w e v e r . Antigen-induced secretion is reduced to control levels by washing, as observed previously [7, 10], whereas ATPTS-induced secretion is slightly enhanced. Calcium still enhances the ATPTS response in washed cells, but not to the same extent as in unwashed cells. It is evident then that unlike antigen, ATP~-induced secretion does not require inositol phospholipid hydrolysis and that secretion occurs as effectively in washed cells as in unwashed cells.
IiO UCRETION
p
tM
O,
o
s
,o
,s MINUTES
:o
es
so
.
o
,
•
,'o
,'s
='o
.
:'o
MINUTES
l~c. 2. ATlaS-induced secretion and inositol phospholipid hydrolysis. Cells prelabelled with either [3H]5HT or [3H]inositol were stimulated with ATP/caleium (0, ©), ATl~/S/calcium (ll, n ) or DNP-BSMcalcium (&, A) for 1-30 rain. The release of soluble inositol phosphates were measured as described in Materials and Methods and are given as described in the legend to Fig. 1.
A T l ~ - i n d u c e d secretion from RBL-2H3 cells
TABLE 2. SECRETION AND INOSITOL PHOSPHOLIPID HYDROLYSIS IN PERMEABILIZED AND UNWASHED (UV~) OR WASHED (W) RBL2H3 C~a.S
Secretion ATP + Ca 2÷ DNP-BSA ATPyS ATP3tS + Ca2+
UW W UW W UW W UW W
21.8 ± 2.6 17.3 ± 1.8 41.4 ± 2.5 18.5 ± 3.3 55.7 ± 3.6 62.4 ± 4.2 67.1±3.9 69.3 ± 2.7
PI-hydrolysis 7.5 ± 4.3 2.8 ± 1.3 48.9 ± 5.1 19.7 ± 0.4 14.7 ± 3.6 3.6 ± 0.9 18.7±3.7 4.3 ± 0.2
Cells pre-labelled with either [3H]5HT (secretion) or [3H]inositol (PI-hydrolysis) were incubated with SLO for 10 min, and
then gently washed. They were then activated with ATP + Ca 2÷, ATP'/S ± Ca 2+ or DNP-BSA+Ca 2+ for 30 rain. Secretion and hydrolysis of washed cells were compared with permeabilized cells which had not been washed. Results are given as a percentage of the total isotope present in the cells and are the means and ± S.E. of triplicates from at least three experiments.
Secretory effects of nucleotide phosphates and ATP analogues To understand further the mechanism by which ATPTS c a u s e d secretion, a n u m b e r of other nucleotide phosphates were added to the permeabilized RBL-2H3 cells. ITP, GTP and ADP at 1 mM all showed a low level of secretion between 18 and 24% which was similar to the spontaneous secretion seen with ATP (Table 3). Up to 5 mM of these nucleotides had no greater effect upon secretion (data not shown). Analogues of ATP were also investigated and ATPcxS, which has the sulphydryl group attached to the s-phosphate of the triphosphate chain, was unable to cause any greater secretion than the nucleotide phosphates above. However, AMP-PNP which has a nitrogen group replacing the terminal bridge oxygen of the triphosphate chain was able to elicit slightly higher levels of secretion (47.2%) than that of A T P ~ (38.4%). A similar analogue, AMP-PCP, whose terminal bridge oxygen is replaced by a CH 2 group was only able to cause low levels of secretion (27.8%). To investigate whether they were affecting the same site, some of the analogues were combined as also shown in Table 3. AMP-PCP had no effect on secretion induced by A ~ or AMP-PNP. Secretion in the presence of
227
equimolar concentrations of ATPTS and AMPPNP (1 mM) was similar to that of A T P ~ alone, suggesting that ATPTS bound to the site with greater efficacy.
Inhibition of ATPyS-induced secretion by nucleotide phosphates To determine whether the other nucleotide phosphates affected ATPTS-induced secretion, cells were incubated with 1 mM ATPTS in the presence of increasing amounts of ATP, ADP, G T P and ITP. Figure 3 demonstrates that an equimolar concentration of ATP inhibits secretion to 50% of that seen with ATPTS alone. At 5 mM ATP, secretion was reduced to basal levels. ADP was also an effective inhibitor at 1 mM, but complete inhibition at higher concentrations was not observed. The two other nucleotides however, GTP and ITP, were only slightly inhibitory at 5 raM, and had little effect at 1 raM. To investigate further the inhibition of ATP'fS-induced secretion by ATP, cells were permeabilized in the presence of 1 mM ATP, 1 mM ATP'/S or 1 mM ATP + 1 mM ATP~S, for up to 90 min. Figure 4 shows that after 30 rnin (as also described in Fig. 3) release in the presence of both ATP and ATP'/S is about half of that in the presence of A T P ' ~ alone. Secretion induced by ATPTS in the presence of ATP continued to increase however, such that by 75-90 TABLE 3. SECRETION INDUCED BY VARIOUS NUCLEOTIDES AND ANALOGUF~FROMPERMEABILIZEDRBL-2H3 CELLS Nucleotide ITP GTP ADP ATP ATPctS ATI~ AMP-PNP AMP-PCP ATP'IS + AMP-PCP ATP'/S + AMP-PNP AMP-PNP + AMP-PCP
% Total release 23.9 + 2.6 20.2 ± 4.2 18.2 ± 1.6 17.8 ± 0.6 19.6 ± 2.4 38.4 ± 1.1 47.2 ± 3.0 27.8 _+ 1.1 35.9 ± 2.1 38.2 ± 3.4 46.6 ± 3.3
The concentration of each nucleotide or analogue was 1 mM and release was measured after 30 min incubation and is given as a percentage of the total isotope present in the cells. Results are means ± S.E. of triplicates from at least three separate experiments.
228
R.I. LUDOWYKE and L. L. SCURR
40
-
N)P
i CONGENTRATION
(raM)
Fio. 3. Inhibition of ATP~-induced secretion by other nucleotides. Cells were incubated with 1 mM ATP~ and increasing concentrations of ATP (IlL ADP (O), GTP (1-'1)and rrP (O) for 30 min. Results are given as a percentage of the release in the presence of ATP# alone (40-45%), and are calculated from the means of triplicates in at least three separate experiments. rain there is little difference in the level of secretion observed from cells activated by ATP~S in the presence or absence of ATP. Secretion in the presence of ATP or ATPyS alone had plateaued by about 45 rain. DISCUSSION The results presented here document some of the b i o c h e m i c a l parameters associated with A T P T S - i n d u c e d secretion, which lead us to h y p o t h e s i z e that A T P T S - i n d u c e d s e c r e t i o n 80
Z
6O
tu
m 9t ttl
40
..I
till
at
2O
0 m
.
0
|
15
-
|
30
-
|
-
45
,
60
-
a
75
•
90
MINUTES
FIG. 4. Prolonged incubation of ATP# + ATP. Cells were incubated with 1 mM ATP (O), 1 mM ATPyS (ll) or 1 mM ATP + 1 mM ATPyS (A) for up to 90 min. Release and results are given as described in the legend to Fig. 1.
bypasses the initial signal transduction mechanisms to activate a later event in the secretory pathway, which in itself is sufficient to induce secretion. We have also shown that ATPyS added to permeabilized RBL-2H3 cells, mimics the antigen-induced actin rearrangement and distinct morphological changes which occur during secretion (R. Ludowyke, K. Kawasugi and P. French, manuscript submitted). As observed previously in permeabilized RBL2H3 cells [8, 9, 20, 26], ATP alone induces a low level of secretion which is considered to be spontaneous, and calcium only enhances this release by a small amount (Table 1). We and others suggest that this is due to fusion of granules immediately adjacent to the plasma membrane, but that another signal is required for further, activated secretion [29-31]. The addition of ATP'yS to the permeabilized cells caused a concentration-dependent increase in secretion to the levels observed with antigen. As with antigen-induced secretion, this was temperature- and time-dependent (Fig. 2). At low ATPTS concentrations, Ca 2+ enhanced secretion, but at high (10 mM) concentrations there was little enhancement evident. The optimal concentration of calcium for enhancement (1 BM; Fig. 1) is similar to that reported previously for enhancement of antigen- and GTPyS-induced secretion from permeabilized RBL-2H3 cells [7-10]. A major difference between the mechanism of secretion induced by antigen and ATP~,S is observed, however, when we compare the requirement for inositol p h o s p h o l i p i d h y d r o l y s i s . Antigen-induced secretion from both intact and permeabilized RBL-2H3 cells has been clearly shown to require the hydrolysis of inositol phospholipids and activation of protein kinase C [3, 4, 7-10]. With ATP'yS-induced secretion, however, there was no prerequisite for the hydrolysis of inositol phospholipids for optimal secretion. Five minutes after the addition of A T t ~ or antigen, 25-30% of the total [3H]5HT has been secreted, compared with less than 5% in the presence of ATP (Fig. 2). At this time there is 25% hydrolysis of inositol phospholipids induced by antigen in stark contrast to the basal (< 5%) levels observed
ATP'/S-inducedsecretionfromRBL-2H3cells with ATP or A T P ~ . After 5 min, there is a slow increase in inositol phospholipid hydrolysis induced by ATP~S. Even this level of hydrolysis is reduced to control levels when the permeabilized cells are washed before the addition of A T P ~ , although this was not due to the loss of phospholipase C as there was still a significant, albeit reduced, level of hydrolysis induced by antigen. However, secretion induced by antigen is lowered to control levels by this washing procedure, as also noted previously [8, 10]. In contrast again, washing slightly enhances the ATPySinduced secretion in the presence or absence of calcium (Table 2). It was suggested to be the loss of protein kinase C during washing that accounts for the inhibiton of subsequent antigen-induced secretion [10, 32], which also suggests that protein kinase C is not involved with ATP'?S-induced secretion, or that in our hands, one of the protein kinase C isozymes may not have been totally removed by washing [10]. Other nucleotide phosphates, ADP, ITP and GTP, were, like ATP, unable to induce the levels of secretion observed with antigen or ATPyS (Table 3). ATPtxS was also unable to induce higher levels of secretion than ATP, suggesting that it was a modification of the y-phosphate that was important for secretion. However, the ATP analogue, AMP-PNP, which has an imido bond between the terminal phosphates was able to induce levels of secretion (47%) slightly higher than that of ATPyS (38%). Another ATP analogue with a modified terminal phosphate bond, AMPPCP induced only low levels of secretion (28%). Therefore the rank order of secretory activity from RBL-2H3 cells of ATP and its analogues is AMPPNP > ATF"yS >>> AMP-PCP > ATP. In comparing the analogues AMP-PNP and AMP-PCP with ATP, it was shown that the bond angle of the l?~----y phosphates of P---C--P was quite different to that of ATP, whereas the bond angle of P - - N P was closer to that of ATP [33]. The similarity to ATP then, appears to be important for secretion. When the analogues were added to the permeabilized RBL-2H3 cells, it was evident that AMP-PCP had no effect on secretion induced by either ATP'~S or AMP-PNP. However, the level of secretion in the
229
presence of ATPyS and AMP-PNP was the same as that of ATPTS alone. This suggests they are both binding to the same site and that ATP~S, assumed to have the same bond angle as ATP, has a higher efficacy for the site or that they may bind to different sites and AqTTS affects the binding to the AMP-PNP site. The inhibition of ATP'yS-induced secretion by ATP and ADP (Fig. 3) is also evidence that secretion is occurring through an ATP-driven system. At equimolar concentrations of ATP and ATP'yS there was 50% of the secretion observed with ATF~S alone. A five-fold excess of ATP reduced the ATP~S-induced secretion to baseline levels which suggests that ATP and A T P ~ are competing for the same site. ADP was nearly as active an inhibitor as ATP, but ITP and GTP were almost ineffective, except at 5 mM where about 30% inhibition was achieved. During storage, the breakdown product of ATPTS is ADP and therefore the presence of ADP in the ATPTS solution would lower the secretion observed with ATP~/S. This could explain the slight variations of secretion observed with ATPyS, depending on the duration for which a particular batch was used. Longer incubations of the cells with both ATPTS and ATP show that ATPTS can slowly overcome the inhibition by ATP (Fig. 4). These results suggest that the turnover of ATP on particular binding sites eventually sees the replacement of ATP with ATP~/S, which is not removed and subsequently induces secretion. The ~l--y phosphate bond of ATPyS may be cleaved by a few enzymes [11, 12], but AMPPNP is considered non-hydrolysable. AMP-PNP is considered a good agent to delineate mechanisms where binding to an ATP site is more important than hydrolysis of the y-phosphate [34]. Our data suggests then, that secretion is induced by these analogues through binding to an ATPbinding site and that the efficacy for the binding site is important, but that hydrolysis may not be necessary. The inability of these analogues to be removed by most phosphatases may also be a mechanism of action [13]. Alternatively ATPyS and AMP-PNP may induce secretion through different mechanisms, independently of hydrolysis.
230
R.I. LUDOWYKEand L. L. SCURR
We would hypothesize that in the unstimulated RBL-2H3 cell, phosphatase activity maintains the cell in a non-secreting state. Activating a kinase by antigen-stimulation, or overcoming the effect o f this phosphatase with agents like ATPTS or AMP-PNP, allows secretion to occur. As GTP did not affect ATPyS-induced secretion, the point at which ATP'~S acts must be downstream from the putative G-protein, Ge [35], or be activating a separate p a t h w a y . O u r p r e l i m i n a r y e v i d e n c e has shown that few proteins are thiophosphorylated with ATp73sS, similar to that observed in chromaffin cells [15], but that one of these proteins is m y o s i n . O u r r e c e n t d a t a (R. L u d o w y k e , K. Kawasugi and P. French, manuscript submitted) also showed that ATP'I,S and antigen both initiated the same morphological and cytoskeletal changes during secretion. This suggests their pathways are linked and that ATPTS is activating a later component of the normal secretory pathway. Acknowledgements--R.I.L. is supported by the E. Steinberg Leukemia Research Fellowship, and L.L.S. is supported by a grant from N.H. & M.R.C. We are grateful for the technical assistance from MARTIN O'StJLLIVAN who was supported by a grant from the Clive and Vera Ramaciotti Foundation. We are very grateful for the helpful discussions with MICHAELA. BEAVEN, N.I.H., U.S.A., and RON PENNY, PETER FRENCH, MARGARETCOOLEY,JAMES WILLIAMSONand others of the Centre for Immunology.
REFERENCES 1. Barsumian E. L., Isersky C., Petrino M. G. and Siraganian R. P. (1981) Eur. J. lmmun. 11, 317-323. 2. Metzger H. G., Alcarez G., Hohman A. R., Kinet J.-P., Pribluda V., and Quarto R. (1986) A. Rev. Immun. 4, 419-470. 3. Beaven M. A. and Ludowyke R. I. (1989) Advances in Regulation of Cell Growth, Vol. 1: Regulation of Cell Growth and Activation (Mond J. J., Cambier J. C. and Weiss A., Eds), pp. 245-285. Raven Press, New York. 4. Beaven M. A. and Cunha-Melo J. R. (1988) Progress in Allergy (Becker E. L., Ed), Vol. 42, pp. 123-184. S. Karger AG, Basel, Switzerland. 5. Berridge M. J. (1987) A. Rev. Biochem. 56, 159-193.
6. Nishizuka Y. (1988) Nature 334, 661-665. 7. Ali H., Cunha-Melo J. R. and Beaven M. A. (1989) Biochim. biophys. Acta 1010, 88-99. 8. Ali H., Collado-Escobar D. M., and Beaven M. A. (1989) J. Immun. 143, 2626--2633. 9. Ludowyke R. I., Peleg I., Beaven M. A. and Adelstein R. S. (1989) J. biol. Chem. 264, 12492-12501. 10. Ozawa K., Szallais Z., Kazanietz M. G., Blumberg P. M., Mischak H., Mushinski J. F. and Beaven M. A. J. biol. Chem. 268, 1749-1756. 11. Haeberle J. R., Sutton T. A. and Trockman B. A. (1988) J. biol. Chem. 263, 442 a. A,A,29. 12. Giambalvo C. T. (1988) Biochem. Pharmac. 37, 4001-4007. 13. Eckstein F. (1985) A. Rev. Biochem. 54, 367402. 14. Brooks J. C., Treml S. and Brooks M. (1984) Life Sci. 35, 569-574. 15. Brooks J. C. and Brooks M. (1985) Life Sci. 37, 1869-1875. 16. Wagner P. D. and Vu N-D. (1989) J. biol. Chem. 264, 19614-19620. 17. Wagner P. D. and Vu N-D. (1990) J. biol. Chem. 265, 10352-10357. 18. Oliver J. M., Seagrave J-C., Stump R. F,, Pfeiffer J. R. and Deanin G. G. (1988) Progress in Allergy, Vol. 42 (Ishizaka K., KaUos P., Lachmann P. J. and Waksman B. H., Eds), pp. 185-245. Karger, Basel, Switzerland. 19. Pfeiffer J. R., Seagrave J. C., Davis B. H., Deanin G. G. and Oliver J. M. (1985) J. Cell Biol. 101, 2145-2155. 20. Deanin G. G., Pfeiffer J. R., Cutts J. L., Fore M. L. and Oliver J. M. (1991) Cell Reg. 2, 627-640. 21. Sahara N., Siraganian R. P. and Oliver C. (1990) J. Histochem. Cytochem. 38, 975-983. 22. Knight D. E. and Baker P. F. (1982) J. Membrane Biol. 68, 107-140. 23. Howell T. W. and Gomperts B. D. (1987) Biochim. biophys. Acta 927, 177-183. 24. Berridge M. J., Dawson R. M., Downes C. P., Heslop J. P. and Irvine R. F. (1983) Biochem. J. 212, 473-482. 25. Beaven M. A., Moore J. P., Smith G. A., Hesketh T. R. and Metcalfe J. C. (1984) J. biol. Chem. 259, 7137-7142. 26. De Matteis M. A., Di Tullio G., Buccione R., and Luini A. (1991) J. biol. Chem. 266, 10452-10460. 27. Lagunoff D. and Wan H. (1974) J. Cell Biol. 61, 809-811. 28. LagunoffD. (1973) J. Cell Biol. 57, 252-259. 29. Sontag J.-M., Aunis D. and Bader M.-F. (1988) Eur. J. Cell Biol. 46, 316--326. 30. Burgoyne R. D., Morgan A. and O'Sullivan A. J. (1989) Cell. Signal. 1, 323-334.
ATP~-induced secretion from RBL-2H3cells 31. Koffer A., Tatham P. E. R. and Gomperts B. D. (1990) J. Cell Biol. 111, 919-927. 32. Cunha-Melo J. R., Gonzaga H. M. S., Ali H., Huang F. L., Huang K.-P. and Beaven M. A. (1989) J. Immun. 143, 2617-2625.
231
33. Yount R. G., Babcock D., BaUantyne W. and Ojala D. (1971) Biochemistry 10, 2484-2489. 34. Yount R. G., Ojala D. and Babcock D. (1971) Biochemistry 10, 2490-2496. 35. Tatham P. E. R. and Gomperts B. D. (1989) Biosci. Rep. 9, 99-109.