Calcium Influx in a Rat Mast Cell (RBL-2H3) Line USE OF MULTIVALENT METAL IONS TO DEFINE ITS CHARACTERISTICS AND ROLE IN EXOCYTOSIS*

An increase in

An increase in concentration of cytosolic Ca2+ ([Ca"+li) is associated with an accelerated influx of 46Ca2+ when cultured RBL-2H3 cells are stimulated with either antigen or analogs of adenosine although these agents act via different receptors and coupling proteins (Ali, H., Cunha-Melo, J. R., Saul, W. F., and Beaven, M. A. (1990) J. Biol. Chem. 265, 745-753). The same mechanism probably operates for basal Ca2+ influx in unstimulated cells and for the accelerated influx in stimulated cells. This influx had the following characteristics. 1) It was decreased when cells were depolarized with high external K'; 2) it was blocked by other cations (La3+ > Zn2+ > Cd2+ > Mn2 = Co2+ > Ba2+ > Ni2+ > Sr2+) either by competing with Caz+ at external sites (e.g. La3+ or Zn2+) or by co-passage into the cell (e.g. Mn2+ or Sr2+); and 3) the inhibition of influx by K+ and the metal ions had exactly the same characteristics whether cells were stimulated or unstimulated even though influx rates were different. The dependence of various cellular responses on influx of Ca2+ was demonstrated as follows. The stimulated influx of Ca2+, rise in [Ca2+IL, and secretion, could be blocked in a concentration-dependent manner by increasing the concentration of La3+, but concentrations of La3+ (>20 p~) that suppressed influx to below basal rates of influx markedly suppressed the hydrolysis of inositol phospholipids (levels of inositol 1,4,5-trisphosphate were unaffected). Some metal ions, e.g. Mn2+ and Sr2+, however, supported the stimulated hydrolysis of inositol phospholipid and some secretion in the absence of Ca2+. Thus a basal rate of influx of Ca2+ was required for the full activation of inositol phospholipid hydrolysis, but in addition an accelerated influx was necessary for exocytosis.
The exocytotic discharge of granules from mast cells, basophils, and cultured, cognate cell lines, in response to either immunological or nonimmunological stimuli, is associated with the mobilization of Ca2+ from extracellular (1)(2)(3)(4)(5) and intracellular (6)(7)(8)(9) sources and an increase in concentration * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ To whom corresponderve should be addressed Bldg. 10 (4,(10)(11)(12)(13). These processes have been studied in detail in the RBL-2H3 cell, which exhibits freqently, but not invariably, oscillations in [Ca'+], (8). In this cell, the increase in [Ca2+Ji probably is due in part to release of intracellular Ca2+ by inositol 1,4,5-trisphosphate (14,15), but the influx of Ca'+ is clearly essential for the sustained increase in [Ca2+Ii and exocytosis (16)(17)(18). Influx is independent of voltage-activated channels (19), and is enhanced in cholera toxin-treated cells (18,20). Entry of other metal ions may also occur (see below). Nevertheless, there is no clear description to date of the mechanism of influx of Ca'+ or of the individual contributions of intracellular and extracellular Ca'+ pools to the stimulatory process in these cells.
The physiological trigger for the stimulation of these cells is the aggregation of receptors for IgE in the plasma membrane by multivalent binding of antigen to receptor-bound IgE (21). This aggregation leads to stimulation of phospholipase C, probably via a toxin-resistant G,-like G-protein (18), to cause rapid and sustained hydrolysis of the inositol phospholipids (16). There is also extensive hydrolysis of other phospholipids (22), primarily phosphatidylcholine, through the action of phospholipase D (23), and the activation of phospholipase A2 to cause release of arachidonic acid (24). In RBL-2H3 cells, the same events are induced transiently by adenosine analogs through receptors for adenosine by a mechanism that is inhibited by both cholera toxin and pertussis toxin (18).
It is uncertain whether the accelerated influx of Ca2+ in stimulated cells is a consequence of these early events, i.e. mediated by intracellular messengers, or of the direct interaction of the aggregated IgE receptors with the cromolynbinding protein (25-27) that is present in the plasma membrane of RBL-2H3 cells (28,29). In support of this latter mechanism, it has been reported that Caz+-conducting channels can be reconstituted in lipid bilayers with purified IgE receptor, IgE, and the cromolyn-binding protein and then activated with the appropriate antigen (27,30). A prediction from this model is that such channels would form only in the presence of antigen and be activated independently of the generation of intracellular messenger molecules.
The uncertainty as to the necessity for Ca2+-influx in initiating early stimulatory events in RBL-2H3 cells arises from the following seemingly contradictory observations. One set of observations suggest that the increase in [Ca2+li, hydrolysis of inositol phospholipids, and exocytosis are highly dependent on Ca2+ influx as blockade of Ca2+ entry by addition of EGTA or high concentrations of La3+ suppresses all three responses by more than 95% (4, 17, 38). Another set of observations indicates release of intracellular Ca2+ in individual RBL-2H3 cells (8) and a substantial increase in [Ca2+Ii in cell suspensions (19) when Ca" entry is blocked with high [K'],; the cells do not degranulate, but it is unknown how other other stimulatory events are affected. The studies point to differences in requirement for external Ca2+ in that low concentrations of La"+ preferentially suppress the increase in [Ca2+Ii and exocytosis without inhibiting the phosphoinositide response in RBL-2H3 cells (4, 38).
In this paper we show that the uptake mechanism, in both stimulated and unstimulated RBL-2H3 cells, is nonselective as demonstrated by the inhibition of uptake of 45Ca2+ by various metal cations into quin2-loaded and nonloaded cells. The extent and time course of quenching of the fluorescence of the quin2/Ca2+ complex (4) by some of the metal cations also indicated that some cations block entry of Ca2+ at the cell surface, whereas others compete with Ca2+ for entry into the cell. Finally, the properties of the various metal ions and [K'], reveal distinctive features of the influx mechanism and the exact requirements of the stimulatory responses for Ca2+ entry.

EXPERIMENTAL PROCEDURES
Materials-Reagents were purchased from the following sources. Nickel(I1) chloride and cesium(I1) chloride were from Aldrich; the chloride salts of other metals were of the highest available grade; "CaCl,, [1,2-"C]5-hydroxytryptamine binoxalate, and "'CeCl, were purchased from Du Pont-New England Nuclear; my0[2-~H]inositol was purchased from both Amersham Corp. and Du Pont/New England Nuclear. Other reagents were from the sources listed in our previous publications (14,18,39).
Preparation of Cells and Buffers-The preparation of cell cultures (17, 18, 38), the permeabilization of cells with streptolysin 0 (18), and the sensitization of cells with DNP-specific IgE (17) were performed exactly as described previously. The principal buffered medium consisted of 119 mM NaC1, 5 mM KC1, 1.0 mM CaC12, 0.4 mM MgC12, 5.6 mM glucose, and 25 mM PIPES buffered with NaOH for NaCl and PIPES/NaOH: KC1 and PIPES/KOH (KC1 buffer); (NaCl buffer) (14,18). Where indicated the following were substituted sodium glutamate and PIPES/NaOH (sodium glutamate buffer); and potassium glutamate and PIPES/KOH (potassium glutamate buffer). Equivalent concentrations of Na+ or K+ salts were used. Further modifications of the buffers included the omission of CaClz (Ca2+free buffer; estimated free Cazf, 5 p~) without or with the addition of 0.1 mM EGTA (Ca"-depleted medium; estimated free Ca2+, (10 nM). Other metal ions were included in the buffers as noted. The pH of these solutions was adjusted to 7.2 before use.
Stimulation of Cells and Measurement of Release of Radiolabeled The conditions for stimulation of the cells were exactly as described above. The reactions were terminated by placing the plates on ice, removing the medium by aspiration, and adding 300 p1 of ice-cold 15% trichloroacetic acid. The extract was clarified by brief centrifugation in a Microfuge and transfered to polypropylene tubes (Falcon 2063). The extracts were assayed by the use of the inositol-1,4,5trisphosphate radioreceptor assay kit (Du Pont/New England Nuclear, Catalog No. NEK-064). The procedures for the extraction of trichloroacetic acid and assay of the inositol 1,4,5-trisphosphate were as described in the manufacturer's instruction manual with one important exception. Following the removal of trichloroacetic acid, the aqueous phase was transferred to small ultrafiltration units (Millipore, UFC3 LGC, 10,000 molecular weight exclusion filter units). These units were centrifuged (in 1.5-ml Eppendorf tubes) in a Microfuge at 4 "C for 15 min to obtain a filtrate that was free of proteoglycans and other high molecular weight substances. As will be described elsewhere: the content of heparin, chondroitin E, and other proteoglycans in extracts of RBL-2H3 cells, basophils, and mast cells grossly interfered with the assay of the inositol trisphosphate. Although the filtration caused some loss of the trisphosphate (up to 30%), the problem was circumvented by subjecting standards and samples to the same procedures throughtout the assay.
Measurement of Uptake of 45Ca2+-In these experiments the uptake was determined in quin2-loaded cells to extend the period over which uptake was linear with time (18, 40) as well as in cultures that did not contain quin2. Cultures were prepared in 24-well cluster plates exactly as described above and incubated overnight at 37 "C with 0.5 pg/ml DNP-specific IgE. The growth medium was then replaced with warm (37 "C) growth medium that contained 30 pM quin2 acetoxymethylester or vehicle (dimethyl sulfoxide, 0.1%). The cultures were incubated for an additional 60 min. The cultures were then washed, 0.2 ml of the indicated medium was added, and the cultures were incubated for 10 min, after which the buffer was replaced by buffer that contained 4sCa2+ (1 pCi/O.2 ml), the indicated metal ion, and 20 ng/ml DNP-BSA. All solutions contained 1.0 mM CaCl,. The cultures were incubated for 5 min at 37 "C. The reactions were terminated by removal of the medium and washing the cultures twice with ice-cold medium. The attached cells were lysed in 0.5 ml of distilled water for the assay of intracellular 4sCa2+. The amount of Ca2+ taken up per culture was calculated from the specific activity of Ca2+ in the extracellular medium. In most experiments the data were expressed as a percent of the maximal uptake in the absence of metal ion. Curve fitting (second order polynominal) was done by computer program.
Because the stimulation of uptake by NECA was very transient dexamethasone. This treatment resulted in a marked enhancement of all responses to NECA (39).
Measurement of [Ca*+/;-Changes in [Ca'+Ii were measured in quin2-loaded cells exactly as described previously (18). Calibrations were performed at the end of each experiment by sequential addition of 0.5 mM Mn2+ (all concentrations noted were the final concentration in the cuvette) to quench the fluorescence of external quin2,0.5% (v/ v) Triton X-100 to liberate and quench intracellular quin2 (to give Pm,J, and finally 1.0 mM DTPA to chelate Mn'+ and to determine the maximal fluorescence of quin2/Ca2+ (to give Fmax). The order of addition of these reagents was changed for some experiments. The temperature of the cultures was maintained at 37 "C a t all times. The cells were stimulated with either 20 ng/ml DNP-BSA, or 10 or 100 FM NECA. The data were plotted as either the actual fluorescence or the calculated [Ca'+Ji, after correction for external quin2 as determined by a computer program (18).

RESULTS
Suppression of Uptake of 4sCa'+ by Various Metal Ions and High [K+J,-Both quin2-loaded and nonloaded cells were used to monitor uptake of 4sCa2+. As in previous studies (18, 39), the uptake of 45Ca2+ into nonloaded cells occured at rapidly diminishing rates to give no further net increase by 5-6 min. In quin2-loaded cells, the rate of uptake did not markedly diminish until 5 min after the addition of stimulant (data not shown). Under either condition, multivalent metal ions inhibited uptake, in a concentration-dependent manner, in antigen-stimulated and unstimulated cells (Fig. 1). The rank order of inhibitory potencies was (Table I). The discrimination in values among Mn2+, Co2+, and Ba2+ in stimulated cells, however, was less evident in unstimulated cells (Table I,   Ce3+, Co'+, Ni2+ and Sr2+, >1 for Cd2+ and Mn"; and >2 for Zn2+ (Table I). 3 The most significant observation was the similarity in the pattern of inhibition by the various metal ions in unstimulated and antigen-stimulated cells, even though uptake of Ca2+ was accelerated by 3-4-fold in antigen-stimulated cells (Table I; compare also panels A and B in Fig. 1). Also of interest was the marked attenuation of the inhibitory effects of La3+ in quin2-loaded cells (compare panels A and C in Fig. 1) probably due to chelation of the La3+ by quin2 (41) that leaks from RBL-2H3 cells (4). A concentration of La3+ of 1 /IM was required to suppress uptake of 4sCa2+ by 50% (ICso) in quin2loaded cells whereas 0.5 /IM La3+ was sufficient to inhibit uptake by 50% in nonloaded cells (Table I). The concentration of external quin2 was between 1 and 2 WM in these studies. Other metal ions were not affected by the external quin2, presumably because they inhibit uptake only at greater than micromolar concentrations. Thus external quin2 would have only a trivial effect a t these concentrations.
As reported by others (19,33), the acceleration of influx of 45Ca2+ in antigen-stimulated cells was substantially reduced by high [K' ],. In contrast to the previous reports (19, 33), however, influx was reduced in unstimulated cells as well. In either case, however, the impairment of uptake by high [K+I0 was not as complete as that observed with La3+ (Fig. 2).
Therefore, the only difference between antigen-stimulated and unstimulated cells in these experiments was in the rate of influx of Ca2+. Both the ICbo values and the Hill coefficients for inhibition by metal ions and the effects of high [K+]" were virtually independent of stimulation.
Selectivity of the Ca2+ Influx Pathway-To determine whether metal ions blocked 4sCa2+ uptake at the sites of entry on the cell surface or competed with Ca2+ for entry, experiments were conducted in quin2-loaded cells. Entry of Mn2+, Zn2+, and La3+ could be assessed because all three metal ions had higher affinity for quin2 than Ca2+ and quenched quin2 fluorescence (41). These experiments demonstrated that MnZ+, but not Zn2+ or La3+, entered the cells after taking into account the spontaneous release of quin2 from the cells at an approximate rate of 1 %/min.
Typical experiments (e.g. Fig. 3), indicated an immediate Ce3+ was tested because the availability of radionuclides of Ce:'+ provided the prospect of determining the kinetics of its binding on RBL-2H3 cells. There was, however, a high degree of nonspecific binding t.o cells, BSA in the medium, and even some types of plastic ware, and estimates of saturable binding were suspiciously high to give an apparent 70 X lo6 sites/cell. quenching of fluorescence of external quin2 upon addition of Mn2+, Zn2+, or La3+. In unstimulated cells, there was then a slow, steady decline in fluorescence which could be due to continuous leakage of the quin2 and its quenching by the metal ions. However, because the decay in fluorescence was more rapid (>2.5%/min) in the presence of Mn2+ (Fig. 3A) than in the presence of La3+ (-l%/min) and Zn2+ (1-1.5%/ min) (Fig. 3, B and C ) , Mn2+ probably entered the cells but at a very slow rate.
In antigen-stimulated cells, the addition of Mn2+ resulted in an immediate decline in fluorescence due to external quin2 and then a decline to a level lower than that observed in unstimulated cells (Fig. 3A), a probable indication that influx of Mn'+ was enhanced in stimulated cells. In contrast, after the addition of Zn2+ or La3+ (Fig. 3, B and C), the fluorescence never declined to levels below that in unstimulated cells. The conclusion drawn from these experiments was that, after antigen stimulation, additional Mn2+ but not La3+ or Zn2+ was taken up irreversibly by the cells.
The entry of Mn2+ and lack of entry of La3+ and Zn2+ in antigen-stimulated cells was further demonstrated by the effect of chelating agents. The addition of DTPA, which has been used to chelate selectively MnZ+ (42), resulted in only partial restoration of fluorescence to levels less than those observed in the absence of Mn'+ in antigen-stimulated cells (Fig. 3A). Therefore, a fraction of the Mn2+ was now inaccessible to the chelator. In contrast, the addition of EDTA or EGTA to cell suspensions that contained Zn2+ (Fig. 3B) or  DTPA (panel A ) ; Zn2+ in contrast did not penetrate cells and could be removed by addition of Ca-EDTA to restore fluorescence to stimulated levels (panel B ) ; and similarly La3+ did not enter into the cells, but could be removed by addition of Ca-EGTA to fully restore fluorescence to stimulated levels (panel C). Note that the immediate drop of fluorescence is due to quenching of external quin2. The slow decay in fluorescence in unstimulated cells is due to leakage of quin2 into the medium and subsequent quenching by the external metal ions. DTPA is a selective chelator for Mn2+ (42).
Similar experiments with NECA-stimulated cells indicated that the increase in [Ca2+Ii was reduced to the same extent as in antigen-stimulated cells by high [K+], and EGTA (Fig. 4B) and was barely detectable (e50 nM) in the presence of the same concentrations of metal ions (data not shown). The studies thus indicated that the same impediments to Caz+ influx resulted in markedly diminished increases in [Ca'+]; in response to antigen or NECA and that high [K'], was less effective than La3+ or EGTA in suppressing the mobilization of Ca2+ in these cells. Also the cells were dependent on a non- voltage-gated influx mechanism with either stimulant. Ability of Metal Ions to Substitute for Ca" in Supporting Cellular Responses-Because the influx mechanism conducted Mn2+, and probably other metal ions, in addition to Ca2+, the various metal ions were tested for their ability to substitute for Ca2+ in supporting antigen-induced responses in intact cells (Fig. 5 ) . As noted previously (4, 17, 38), hydrolysis of inositol phospholipids was much reduced and exocytosis blocked by the omission of Ca2+. The presence of 1 mM Mn2+ or ST2+, however, partially restored both responses to antigen although the extent of exocytosis was still small when compared with the response in 1 mM Ca2+. At a concentration of 10 mM, Sr"+ now promoted substantial secretion and a full phosphoinositide response (Fig. 5, inset). The presence of 10 mM Ba2+ also resulted in a small increase in secretion (6%) but it should be noted that high concentrations of this ion are generally required to replace the need for Ca2+, e.g. >10 mM with antigen-stimulated rat mast cells (1). The presence of other metal ions (Cd' ' , Zn2+, Co2+, and La3+) suppressed the residual antigen-induced hydrolysis of inositol phospholipids in the intact cell (Fig. 5 ) . Because none of these metal ions, at concentrations up to 50 p~, suppressed antigen-induced responses in permeabilized cells (Fig.   6); Studies with fura 2, which exhibits similar fluorescence spectra in the presence of Ca'+, Ba", or Sr", indicated that the free intracellular concentration of Ba'+ and Sr2' remained below 5 PM  they probably acted by interfering with entry of the residual Ca2+ that remained in the Ca2+-free medium (no EGTA was added in these experiments). These results established that Mn2+ and Sr2+ entered the cells and could substitute for Ca2+ to support hydrolysis of inositol phospholipids totally, and exocytosis partially, whereas other metal ions, by exclusion of Ca'+, inhibited antigen-induced responses.
The Importance of Ca2+ Influx in Promoting Stimulatory and Secretory Responses in RBL-2H3 Cells-Advantage was taken of the ability to manipulate influx of Ca2+, by use of high [K'], or different concentrations of metal ions, to determine the precise requirement of each response on influx. In either chloride or glutamate buffers, high [K+l0 did not suppress breakdown of the inositol phospholipids, but it did suppress exocytosis (Fig. 7). Although this suppression was never complete, and varied from 50 to 80% in different experiments, it did indicate that a diminution of stimulated influx (i.e. Fig. 2) and of the increase in [Ca"Ii (i.e. Fig. 4A) will impair exocytosis but not necessarily the phosphoinositide response.
A partial uncoupling of responses was also evident in studies with different concentrations of various metal ions. Exocytosis appeared to be more susceptible to inhibition than hydrolysis of the phospholipids with most metal ions. The exceptions were Sr2+, which inhibited neither response, and Ba2+, which suppressed exocytosis to a limited extent (Fig. 8). These exceptions were not unexpected because, as noted above, both these ions were reported to enter mast cells and substitute for Ca2+ (1).
The most incisive information was obtained by increasing the concentration of La3+ to decrease Ca2+ influx. Exocytosis was impaired when influx of Ca2+ was reduced by more than metal ions was based arhitarily on the assumption that the actual levels of free metal ion in the cytosol in the experi,ments shown in Fig. 5 did not exceed 50 PM. 50%, but hydrolysis of the phospholipids was impaired only when the concentration of La3+ was sufficient to reduce the rate of Ca2+ influx to below that observed in unstimulated cells. Similar effects were observed on exocytosis and phospholipid hydrolysis in cells stimulated with either antigen or NECA (Fig. 9). It should be noted, however, that even when influx of Ca'+ was blocked to a degree sufficient to impair significantly the hydrolysis of inositol phospholipids, the production of inositol 1,4,5-trisphosphate was not compromised. Additional experiments showed that in the presence of high concentrations of La3+ (33 and 100 pM) or even in Ca*+-free medium (i.e. no added Caz+ with 0.1 mM EGTA), conditions that resulted in 90-96% reduction in secretion and up to 70% decrease in the generation of [3H]inositol phosphates, did not affect the normal increase in levels of inositol 1,4,5-trisphosphate in antigen-stimulated cells. These levels increased from -2 pmol/1O6 cells to 5-7 pmol/1O6 cells (range of values, 10 experiments) within 5 min and remained elevated for at least 15 min. In no experiment was any significant decrease observed in these levels in the presence of La3+ or absence of Cd+.

DISCUSSION
Although the increase in [Ca2+]i provides an essential signal for the exocytotic release of granules from rat mast cells (43) and RBL-2H3 cells (16), additional synergistic signals are clearly necessary. Low concentrations of the Ca2+-specific ionophores, for example, can induce substantial increases in [Ca'+Ii without stimulating RBL-2H3 cells to secrete (44). Because antigen stimulation results in the translocation of protein kinase C from the cytosol to the membrane fraction (45) as well as protein kinase C-dependent phosphorylation of myosin (46), the activation of protein kinase C may provide the second synergistic signal for exocytosis in RBL-2H3 cells.
There are, however, indications of an additional synergistic signal that is independent of protein kinase C (44,47) as well as evidence for substantial hydrolysis of phosphatidylcholine by a phospholipase D enzyme, which may supply messenger molecules (23) in addition to those generated by the hydrolysis of the inositol phospholipids (16). Despite the uncertainty of the relationship of the various stimulatory events to exocytosis, the full expression of all stimulatory responses and exocytosis in RBL-2H3 cells and its variants are, without exception, highly dependent on the presence of external Ca'+ and, presumably, the influx of Ca'+ (16).
The influx of Ca2+ into RBL-2H3 cells can be slowed by La"+ (4) or depolarization of the plasma membrane with high [K'], (19). As in mast cells (48,49), influx is not blocked by organic blockers of voltage activated Ca2+-channels (48,49). As reported here, the process allows entry of Mn'+, Sr2+, and possibly other divalent cations. It is blocked competitively by Zn'+ and La3+ outside the cell. The interaction of some metal ions with the influx mechanism exhibit a pseudo-Hill coefficient of greater than one to indicate possible cooperative actions at the sites of uptake. Finally, the characteristics of the influx mechanism are the same in unstimulated and stimulated cells. This could indicate that either additional channels (or carriers) are recruited or the opening times of existing channels are extended in stimulated cells. It is highly unlikely, therefore, that the influx mechanisms operate by the interaction of the antigen-IgE-receptor complex with a membrane protein (26,30). In any case, the same influx mechanism appears to be activated by NECA, which acts independently of IgE receptors (18). 5 The earlier work with rat peritoneal mast cells provides sufficient description to suggest that mast cells and RBL-2H3 cells utilize similar mechanisms of uptake. La3+ and Mn'+, for example, inhibit Ca'+-dependent histamine release in a reversible, competitive manner in stimulated mast cells to give values for ICs0 that are comparable with those reported here for inhibition of uptake of Ca2+ (35). In addition, Sr2+ and Ca2+ compete for uptake in unstimulated mast cells (34), and Sr2+ inhibits Ca2+-dependent release of histamine in stimulated mast cells to give a value for ICso of about 4 mM (1). The data, in fact, show remarkable qualitative, and even quantitative, similarities.
Our intention was to utilize the metal ions to obtain a kinetic description of the uptake mechanism in RBL-2H3 cells and, thereby, provide an experimental model for comparison with uptake mechanisms in other types of cells. Of specific interest was the non-voltage-gated, second messenger-operated channel because of the presumed ubiquitous distribution of these channels in cells (50). This category of channels, as opposed to voltage-gated or receptor-operated channels, has not been well defined, however, because of the lack of selective and diagnostic blockers of these channels. We thought instead that the metal ions could be useful in view of their well defined kinetics of inhibition of influx.
The evidence for non-voltage-activated influx of cations in tissue cells, including mast cells, comes from studies with patched-clamped cells and from studies with metal ions in intact cells. Electrophysiological studies indicate that discrete cation channels can be activated in at least three cellular systems, namely T lymphocytes (51), neutrophils (52), and rat peritoneal mast cells (31). These channels, whose activation is independent of membrane potential, are of low conductance (5-10 picosiemens) and become apparent several seconds after the addition of stimulant. In mast cells and T lymphocytes the channels carry Ba2+ and are blocked by Cd".
In neutrophils, the channels are permeable to Na+ and K+.
Nonselective cation channels of high conductance (50 pS) are also activated in stimulated rat mast cells (32). All these studies suggest the possible involvement of messenger molecules or G-proteins in that the opening of the channels is delayed and the same channel activities or conductances can be activated by inositol 1,4,5-trisphosphate, Ca2+, or GTPrS.
Studies with intact cells also indicate that the production of the inositol phosphates and the mobilization of intracellular Ca2+ are often associated with the activation of nonselective, non-voltage-activated influx of cations (50). Some patterns emerge that are analagous to the present studies. The trivalent lanthanides block uptake of Ca2' not only in mast cells (9) but in many other types of cells (Ref. 53 and citations therein). The alkaline metal ions, Sr2+ and Ba", can permeate cells and activate responses or substitute for Caz+ in many cell types (Ref. 53 and citations therein) in addition to mast cells. In two types of cells, receptor-stimulated influx of these metal ions occurs (53,54), and in one instance Sr2+, but not Ba'+, was shown to enter intracellular stores of Ca2+ (53). Of the transition metal ions, Mn'+ has been widely used to study Our recent studies with RBL-2H3 cells that have been transfected with the gene for the muscarinic M1 receptor (P. Jones, 0. Choi, and M. A. Beaven) indicate that both carbachol and antigen activate the same, or very similar, influx mechanisms to produce virtually the same profiles as those shown in Figs. 1 and 2  influx of cations because this ion quenches fluorescence of intracellular probes for Caz+ (42). The effects of Ni2+ and Co2+ are variable, but Ni2+ does not permeate lachrimal acinar cells and blocks entry of Ca2+ (53). Most studies provide only qualitative descriptions of the effects of the metal ions, and because of this, the influx mechanism cannot be directly compared with that in mast cells. In addition there may be different mechanisms of uptake even within the same cell (42).
The indications from patch-clamp studies with mast cells of the involvement of second messengers in cation influx are consistent with all that is known about the stimulatory events in RBL-2H3 cells. For example, a transient stimulation of hydrolysis of the inositol phospholipids in RBL-2H3 cells via adenosine receptors is associated with a similarly transient influx of Ca2+ and rise in [Ca"];, whereas these responses are sustained in antigen-stimulated RBL-2H3 cells (18). In addition all manipulations that depress or enhance stimulated breakdown of inositol phospholipids invariably result in identical changes in the magnitude of the increase in [Ca2+Ii (16,17) and influx of Caz+ (39). As the present results demonstrate, the converse situation does not apply. Indeed, the uncoupling of Ca2+ influx from breakdown of inositol phospholipids ( Fig. 9) provides the clearest indication to date that this breakdown is an early event in the stimulatory pathway for exocytosis in RBL-2H3 cells or mast cells.
Even though stimulated breakdown of the inositol phospholipids was not nearly as dependent on influx of Ca2+ as exocytosis, there was a threshold for rate of influx below which this breakdown could no longer be sustained. This threshold appeared to be equal to or possibly slightly lower than that in unstimulated cells (basal influx). Thus high [K'], did not inhibit stimulated Ca2+ influx to less than basal rates of influx (Fig. 2), nor did it inhibit the stimulated breakdown of the phospholipids. Both the increase in [CaZ+Ji and exocytosis, however, were impaired (Figs. 4 and 7). In contrast, suppression to below basal rates of influx with >10 p~ La3+ (Fig. 2) caused significant impairment of hydrolysis of the phospholipids (Fig. 9). These rates might have been insufficient for maintenance of adequate levels of Ca2+ in intracellular stores and cytosol.
In contrast to the hydrolysis of inositol phospholipids in total, the generation of inositol 1,4,5-trisphosphate is less dependent on external Ca2+ and probably accounts for the transient increases in [Ca*+]; in the absence of external Caz+ in individual RBL-2H3 cells (8). The data add support to the previous suggestion that much of the hydrolysis of the inositol phospholipids is not dependent on the formation of inositol 1,4,5-trisphosphate (14). Whether the extensive hydrolysis of inositol phospholipids in the presence of Ca2+ is due to a change in specificity of phospholipase C with elevated [Ca2+]; or to recruitment of additional isozymes of phospholipase C is unknown.
We conclude that full activation of early stimulatory events requires external Ca2+ and a basal influx of Ca", but that an acceleration of influx is necessary for exocytosis. The homeostasis of Ca2+ is important, not only in stimulated RBL-2H3 cells to maintain high [Ca2+Ii (4), but may be equally critical in unstimulated cells to maintain [Ca2+Ii high enough to support activation mechanisms and low enough not to activate exocytosis. Here too the synergistic signals come into play, as increases of 0.1 pM free Ca2+ will elicit substantial exocytosis in permeabilized RBL-2H3 cells when this coincides with the activation of phospholipase C, but without this activation concentrations of free Ca2+ >1 PM are necessary for exocytosis (55).