Adenosinetriphosphate, calcium and temperature requirements for the final steps of exocytosis in Paramecium cells

The following morphological details render Paramecium trichocysts very appropriate material for the study of the final steps in exocytosis. (1) The secretory vesicle membrane is permanently positioned as closely as ∼ 15nm to the cell membrane (Plattner, Miller & Bachmann, 1973; Plattner, Wolfram, Bachmann & Wachter, 1975). (2) The multiplicity of these permanent attachment sites allows for an amplification effect during synchronous triggering. (3) The discharge involves stretching of the secretory contents (the trichocyst) which can be conveniently followed in the light microscope. When exocytosis is triggered this involves only the final exocytosis steps, i.e. membrane fusion, discharge of the contents and membrane resealing.

Trichocysts contain within their limiting membrane a highly ordered matrix (Bannister, 1972) composed mainly of one major protein (Steers, Beisson & Marchesi, 1969) in a paracrystalline arrangement (Hausmann, Stockem & Wohlfarth-Bottermann, 1972 a, b). Trichocyst discharge takes place explosively, probably within less than a millisecond (Plattner, Bilinski & Völlenklee, unpublished results), accompanied by about 7-fold stretching (Hausmann et al. 1972 a) of the contents. One can, therefore, study parameters relevant for the stretching process, even with isolated, membrane-free trichocysts, and, thus, achieve a de-coupling of membrane fusion from stretching (extrusion) phenomena.

As in other systems (Palade, 1975) the cell membrane and the membrane of the ‘extrusome’ fuse during exocytosis (Plattner et al. 1973). Specialized membrane structures were recognized at the fusion sites in freeze-fracture studies (Bachmann, Schmitt & Plattner, 1972; Plattner et al. 1973). Experiments with ionophoretic Ca²⁺ injections (Plattner, 1974, 1976) not only revealed ultrastructural intramembraneous changes during exocytosis; they also suggested that Ca²⁺-mediated ‘stimulussecretion-coupling’, a phenomenon known from other secretory cell types (Douglas, 1974; Rubin, 1974), occurs at the final exocytosis steps. These experiments were accompanied in Paramecium cells by formation of reaction products at some sites of the membranes involved and on some structural elements of the trichocyst contents (Plattner & Fuchs, 1975); as electron-microscopic X-ray microanalyses revealed the presence of Ca and P in these reaction products, this pointed to the presence of phosphate-splitting enzymes at well-defined sites. In more elaborate cytochemical analyses the occurrence of a Ca²+-dependent ATPase at the potential membrane fusion sites was ascertained (Plattner, Reichel & Matt, 1977).

Information on energy- and temperature-requirements for exocytosis is scarce (cf. Meldolesi, Borgese, DeCamilli & Ceccarelli, 1978). These aspects, in conjunction with Ca²⁺-requirements, were analysed in the present system specifically for the final steps of exocytosis.

Paramecium tetraurelia (formerly P. aurelia;Sonneborn, 1975), strain K401, was cultivated at 26 °C strictly monoxenically with Enterobacter agglomerans added in a dried lettuce medium and harvested in early stationary phase. Cultures were filtered through cheese cloth; with the use of a sieve-plate filter (30–40 μm) bacteria were removed and cells concentrated to ∼ 10⁵ cells/ml immediately before use.

A23187 (Eli Lilly). Apyrase (Sigma). Adenosinetriphosphate (ATP, Tris salt; Sigma). Cetylpyridinium chloride (CPC; Merck). Cetyltrimethylammonium bromide (CTMAB; Merck), p-chloromercuribenzoate (Sigma). D6oo (methoxyverapamil; Knoll). Dibucain (Nupercain; Ciba). EDTA (ethylenediamine tetraacetate; Merck). EGTA (ethyleneglycolbis[β-aminoethyl ether]N,N’-tetraacetate; Sigma), N-ethylmaleimide (NEM; Sigma). Firefly lantern extract (type FLE-50; Sigma). La³⁺ (chloride; Merck). Lettuce medium (Difco). Lysolecithin (L-α x-lysophosphatidylcholine, type I; Sigma). Mersalylic acid (Sigma). Polyethyleneglycol (PEG, mol. wt. ∼ 6000; Koch-Light). Salyrgan (Farbwerke Höchst). Valinomycin (Calbiochem; Serva). X-537A (Hoflmann-LaRoche). All chemicals were of the highest purity available.

K, Na, Ca and Mg concentrations in the culture medium were determined with a Beckman atomic absorption photometer, model 1248. For Ca or Mg, ‘masking’ with 1 % La₂O₃ in 1 % HNO, (final concentrations) was applied. K, Na were also determined by flame photometry and Ca, Mg by colorimetric methods. The concentrations of K, Na, Ca, Mg were 1·70, 0·40, 0·16 and 0·14 mM, respectively, in the cell suspensions used.

Trichocysts were isolated according to Fig. 1. This method was developed on the basis of data presented by Hoffmann-Belling (1961) and Anderer & Hausmann (1977). Homogenization was performed with a loosely (∼ 0·25-mm clearance) fitting Teflon pestle. On a 1·2/1·8M sucrose interface and in the presence of 1 mM EDTA and 1 M ATP we obtained pure trichocysts which were in a contracted state and devoid of a membrane envelope.

Triggering was monitored with a Reichert light microscope, model Diavar. 10-μ droplets containing ∼50 cells were analysed and the time required to trigger exocytosis in all cells was registered. For long time periods samples were stored in a moist chamber. The microscope was equipped with a Nomarski-type optical system and a temperature-controlled object holder, which can be heated or cooled; the actual specimen temperature was measured with a thermocouple. Potential trigger compounds were tested and their concentration adjusted so that complete triggering was achieved just within a ∼3-min period (20 °C, pH ∼7, free [Ca²⁺]₀ ∼ 0·1 mi). For experiments at o °C cell suspensions were rapidly cooled by shooting them in a fine jet, using 0·5 atm. (50·5 kN m^-1) pressure, through a ∼ 150-μm wide pipette on to an object holder at o °C, since the usual cooling procedures were too slow and provoked massive trichocyst discharge. Spraying did not damage the cells. When re-warmed (∼20 °C/min) after 3 min at o °C they regained their full motility and the response to the trigger compounds at 20 °C was almost fully re-established. This allowed us to analyse temperature effects on fully viable cells at 20 and 0 °C.

Isolated contracted trichocysts (contents) with 1 mM EDTA and 1 mM ATP added were also exposed to trigger compounds at 20 and o °C. Apyrase (E.C. 3.6. 1.5.), Sigma type I from potatoes, was applied at 0·33 % pH 6·7 + 3·3 mM Ca¹⁺, following the procedure of Molnar & Lorand (1961).

In principle the method of Strehler & Totter (1952) was followed.

The effect of every substance added to viable cells was checked by light microscopy. For each triggering procedure we prepared at least 4–10 pairs of triggered and untriggered aliquots. The simultaneous, pairwise comparison of stimulated and non-stimulated cells under directly comparable conditions (e.g. same time schedule, same chemicals in varying sequence, same ionic milieu and pH, etc.) was crucial to circumvent artifact hazards. Residual bacteria accounted for less than 1 % of the total ATP measured.

50 mg extract were dissolved in 5 ml 0·2 M Tris/HCl pH 7·4. After 12–18 h at 20 °C the extract was centrifuged and filtered (Millipore 0·2 µn). 0·2 ml, containing 20 mM Mg²⁺, were added to each 2-ml sample. The final [Mg²⁺] in the assay was invariably 5·8 πmM, i.e. within a range we found for maximal activation.

They were obtained from Tris-ATP solutions in 0·2 M Tris/HCl, pH 7·4, with a final [Mg²⁺] of 5·8 mM. For each type of experiment we ascertained that any compound added did not affect the calibration. See Fig. 2.

Scintillation counting was performed with a Packard Tricarb liquid scintillation spectrometer, model 3375, using the β-canal, 50% coincidence and a detector temperature of 8 °C; 20 s after addition of firefly lantern extract, samples were counted for 0·i min. The timing was precise enough to minimize any effect of impulse rate decay. The final pH value of the samples, determined immediately after counting, was 7·50 ± 0·02. This lies within the optimal pH range which was between 7·2–7·8 under our experimental conditions.

As a control for ATP-hydrolysis measurements by scintillation counting we assayed also the corresponding P₁-increase according to Taussky & Shorr (1953).

The method of Lowry, Rosebrough, Farr & Randall (1951) was used in the Eppendorf manual modification.

The oxygen pressure (pO₂)was measured at 25 °C with a Gilson oxygraph, model K.I.C., equipped with a 2-ml chamber and a Clark electrode.

Cells and isolated cell fractions were analysed in a conventional electron microscope. After fixation in 2% glutardialdehyde and washing (0·1 M cacodylate buffer pH 7·0), isolated cell fractions were either spread on coated support grids (with or without 1 % aqueous uranyl acetate) or processed by routine methods (1 % OsO₄; acetone dehydration; Durcupan ACM-Fluka-embedding) to ultrathin sections, which were contrasted for 20 min with 7·5 % aqueous unbuffered magnesium uranyl acetate and for 3 min with lead citrate, pH 12·0.

Freeze-fracturing was carried out routinely as indicated previously (Plattner et al. 1973), i.e. without any pretreatment, and with the use of a Balzers BAF 300 unit at— 100 °C. Some pellets were cooled at ∼ 20 °C/mm till frozen (to provoke massive trichocyst discharge) and then plunged into liquid nitrogen.

From estimates of the cell volume and surface (by fight microscopy), of the possible number of trichocysts (from freeze-fracture data) and of the trichocyst volume (from electron micrographs of longitudinally cut organelles) we obtained a rough estimate of the relative volume fraction of trichocysts in an average cell (Table 2).

Various chemical agents listed in Table 1 trigger the exocytosis of trichocysts very efficiently: (1) ATPase-inhibitors (p-chloromercuribenzoate, mersalylic acid, La³⁺, N-ethylmaleimide, Salyrgan), (2) the Ca²⁺-antagonist D600, (3) Ca²⁺-transporting ionophores (X-537A, A23187), (4) lipolytic agents (lysolecithin, acetone), (5) the fusogenic compound polyethyleneglycol, (6) the local anaesthetic Dibucain (Figs. 3, 4) and (7) the cationic detergents cetyltrimethylammonium bromide (CTMAB) and cetylpyridinium chloride (CPC). With most compounds the triggering effect depends on the availability of extracellular Ca²⁺ (Ca²⁺), at least in the concentrations present in the medium (see below); exceptions are La³⁺ and Dibucain, while D6∞, CTMAB and CPC gave variable results. With X-537A and A23187 [Ca²⁺] has to be enhanced to accelerate exocytosis. With all compounds used trichocyst discharge is preceded by a strong ciliary reversal reaction. Only with La³⁺ is ciliary reversal rare and all ciliary movement is, in contrast to other agents, rapidly abolished.

None of these compounds is effective at 0 °C (Table 1), even though cells fully respond to these agents after re-warming to 20 °C, first by ciliary reversal and then by trichocyst exocytosis.

To check for a specific Ca²⁺ effect on exocytosis triggering we administered in another approach different ionophore-cation combinations (Fig. 5). Ionophore experiments were designed to reverse the high [K⁺]₁- and low [Ca²⁺]₁-values which determine the normal surface potential in Paramecium cells (Naitoh & Eckert, 1974). Values of [K⁺, Na⁺, Ca²⁺, Mg²⁺]₀ which are constantly around 1·70, 0·40, 0·16 and 0·14 mM in the culture medium, were correspondingly diluted 2 or 3 times in these experiments unless added in excess. K+ is unable to exert a trigger effect, even when valinomycin is incorporated and [K⁺]₀ then raised to levels exceeding those assumed for [K⁺]₁ (~2O mM; Naithoh & Eckert, 1974), provided Ca₀²⁺ is chelated by EDTA or EGTA. Both X-537A and A23187 are about 7–8 times more effective with exogenous Ca²⁺ than with Mg²⁺.

Temperature effects on trichocyst discharge were studied in different ways. Very rapid cooling to o °C, performed as indicated in Methods, abolishes the reaction to any trigger compound (Table 1) and, concomittantly, does not entail spontaneous trichocyst discharge. However, slow cooling (∼2o to 30 °C/min) always results in massive exocytosis as a temperature of about 12–10 °C is reached (Fig. 6). Warming ∼80 ºC/min) results in a similar reaction. Within a temperature range (26–12 °C tested) where spontaneous trichocyst expulsion is rare, exocytosis mediated by ionophoretic Ca²⁺ injection is temperature dependent (Fig. 6).

When exocytosis is triggered by slowly cooling the cells to o °C and the pellets are then freeze-fractured, they contain abundant free, i.e. discharged, stretched trichocysts. Those membrane-intercalated particles which are normally randomly distributed in the cell membrane (‘h-type’ particles; see Plattner et al. 1975), become mostly clustered by lateral segregation as in Tetrahymena membranes (Speth & Wunderlich, 1973). The regular double ‘rings’ of cell membrane particles which surround every trichocyst attachment site (Plattner et al. 1973), persist; they become only slightly displaced so that small groups of particles no longer lie within an ideal ring (Fig. 7). ‘Central granule patches’ (‘fusion rosettes’) become rare or absent. Nevertheless, one never sees an open exocytosis canal within these exocytosis sites. These, evidently, undergo resealing, even when the temperature is progressively reduced.

When the ATP content of cells is measured after massive synchronous trichocyst exocytosis, the ATP-content is significantly reduced in comparison with untriggered controls (Figs. 8, 9). This holds true for all the different triggering procedures, except for polyethyleneglycol, notwithstanding its strong triggering capacity. Values of controls prepared in different ways are not significantly different from each other, regardless of whether trichocyst discharge is inhibited by the addition of Mg²⁺ or whether cells are inactivated with trichloroacetic acid. The ATP loss obtained with different types of trigger agents was slightly different; this could be due to slightly different amounts of trichocysts expelled; so far, there are unfortunately no means for a more precise quantitation of trichocyst exocytosis and we always aimed at a maximal trigger effect under light-microscopical control.

As a control for the involvement of ATPase-activity during chemically (Dibucain) induced exocytosis we measured whether there occurs a concomitant increase of free phosphate (P₁). In all experiments P₁ increased in parallel with the hydrolysis of ATP.

At 22 °C a Paramecium cell consumes 0·21 × 10⁻¹² mol O₂ per min. Massive synchronous trichocyst expulsion was not paralleled by enhanced respiration (Fig. 16) under any experimental conditions used. Nevertheless, even within a triggering period of io s cells could compensate for some of the ATP consumed during exocytosis. Therefore, the ATP-consumption determined after triggering might be an underestimate. pO measurements also indicate that cells were fully viable during the trigger phase.

Trichocysts were separated from pellicles in a sucrose gradient, supplemented with EDTA and ATP (Fig. 1). Pure trichocysts (Fig. 10) are obtained which are devoid of the membrane and of the thin ‘outer lamellar sheath’ (Bannister, 1972); they contain the ‘inner lamellar sheath ‘and the whole matrix in a contracted state (Figs. 11, 12). Isolation of trichocysts without addition of ATP resulted in a variable percentage of stretched trichocysts (Bilinski & Plattner, unpublished observations).

As shown in Table 1 and Figs. 13–15, isolated trichocysts can be stretched with Dibucain, CTMAB and CPC. Stretching is also achieved with exogenous ATPase (Apyrase). Trichocysts respond to all these agents at o °C as well as at 20 °C.

Paramecium cells contain a free internal Ca²⁺ concentration ([Ca²⁺]₁) of ≤ 10⁻⁷ M (Naitoh & Eckert, 1974), i.e. significantly lower than [Ca²⁺]₀ (1·6×10⁻⁴M). Ionophoretic Ca²+ injections indicated that an increase of free [Ca²⁺]₁ stimulates exocytosis in this system (Planner, 1974; Planner & Fuchs, 1975). These observations were now extended to different triggering procedures which -except for a few -depend upon the availability of Cao²⁺ (Table 1). With all triggering procedures used a ciliary reversal reaction, which is initiated by a [Ca²⁺]₁ of ∼ 10⁻⁶ M (Naitoh & Kaneko, 1972) precedes exocytosis. Free [Ca²+]i necessary for the final exocytosis steps must, therefore, be in the range of 10⁻⁵ to 10⁻⁴M. Interestingly, fusion of isolated secretory granules needs just about 10⁻⁶ M Ca²⁺ (Dahl & Gratzl, 1976).

The surface potential of paramecia is determined by a low [Ca²⁺]₁ and a high [K⁺]₁ of ∼20mM (Naitoh & Eckert, 1974). When the K+-selective ionophore valinomycin (Pressman, 1976) is incorporated and [K⁺]₀ raised to > [K⁺]₁ (after chelation of Ca²⁺) cells become immobilized, but no exocytosis takes place (Fig. 5). Thus, reversal of the K⁺-gradient, which participates in the surface potential formation, does not suffice to initiate exocytosis. This is in line with the experience of electrophysiologists (Ogura, personal communication). Among the other ionophores used, only A23187 is rather selective for bivalent cations, while X-537A transports also K+ and other monovalent cations (Pressman, 1976). Both promote trichocyst discharge considerably more with increased [Ca²⁺]₀ than with Mg²⁺ or K⁺ added. These experiments underscore the fact that a [Ca²⁺]₁-increase is the crucial event for the final exocytosis steps.

A variety of ATPase inhibitors trigger exocytosis (Table 1), but only when free Ca₀²⁺ is present. This means that the Ca²⁺ influx over the cell surface would no longer be compensated by the Ca²⁺-pumping system(s) of the cell surface. As to be expected, cells are not triggered by other inhibitors which attack only selective types of ATPases (100μg/ml oligomycin; 5 MM ouabain), which inhibit nucleotide transport (5 mM atractyloside or carboxyatractyloside) or the hydrolysis of monophosphates beyond pH 7 (5 mM L(+)tartrate or L-tetramisole) (Plattner, unpublished results).

D600 acts as a potent inhibitor to potential-dependent Ca²⁺-influx (Fleckenstein, Nakayama, Fleckenstein-Grün & Byon, 1975) and thus stops the secretory activity in a variety of cells (cf. Carafoli, Clementi, Drabikowski & Margreth, 1975). Its triggering effect on paramecia was, therefore, unexpected. However, at the relatively high concentrations used (575 μM), this compound acts rather like a local anaesthetic and releases bound Ca²⁺ (Dörrscheidt-Käfer, 1977). In fact, the local anaesthetic Dibucain also strongly triggers trichocyst (Table 1) or mucocyst discharge (Tetra-hymena: Sati?, 1977), possibly by liberating membrane-bound Ca²⁺ (Feinstein, 1964; Nicolson, Smith & Poste, 1976). This could explain why triggering takes place also in the absence of free Ca₀²⁺. It is uncertain whether the cationic detergents used would exert a similar effect when incorporated into membranes. La³⁺ also triggers regardless of the [Ca²⁺]₀. La³⁺ inhibits various Ca²⁺ pumps; as it seems to penetrate certain cells (Batra, 1973) it could also augment free [Ca²⁺]i by depressing intracellular Ca²⁺-segregating systems.

The fusogenic lipid lysolecithin (Lucy, 1974) also induces exocytosis, but common lipid solvents, like acetone (Table 1), ethanol, diethylether, chloroform, etc., do as well. As their effect always depends upon Ca₀²⁺ they might trigger not simply by perturbation of the membrane lipids but also by a simultaneous Ca²⁺-influx. Concomitantly, according to Ahkong, Fisher, Tampion & Lucy (1975), cell fusion induced by various lipid-soluble fusogenic agents also depends upon Ca₀²⁺. It is extensively documented that Ca²⁺ influx from outside stimulates exocytosis in various systems (Douglas, 1974; Rubin, 1974). It remains open, to what extent intracellular Ca²⁺ stores, like ‘calcium-storing vacuoles’ (Plattner & Fuchs, 1975) or mitochondria, would contribute to a [Ca²⁺]₁-increase for trichocyst exocytosis under physiological conditions.

As Table 1 shows, none of the trigger compounds becomes effective when cells are very rapidly chilled to o °C. This holds true also for those agents, like Dibucain, which transform isolated trichocyst contents from the contracted to the expanded state even at o °C. We assume, therefore, that the temperature sensitivity of the exocytosis process in vivo is due to a membrane effect. At o °C not only Ca²+-pumping systems (Browning & Nelson, 1976) but also Ca²⁺ channels for Ca²⁺ entry appear to be blocked in paramecia.

When slowly cooled to ∼ 12 to 10 °C or below, paramecia display considerable trichocyst exocytosis. Could an increase of free [Ca²⁺]i again be involved? One could argue that solidification of the membrane lipids (their phase transition point could be around 12–10 °C but was not determined precisely) could entail a deactivation of the membrane-bound Ca²⁺-ATPase system (Seelig & Hasselbach, 1971) which then would fail to maintain the normal, low [Ca²⁺]i. We found morphological evidence by freeze-cleaving that in these experiments membranes are (or again become temporarily) sufficiently fluid for undergoing subsequent resealing. Above the critical temperature region, Ca²⁺-ionophore-mediated exocytosis depends directly from the temperature. We presume that this pattern of temperature-dependence is analogous to the triphasic temperature response of Ca²⁺-dependent miniature endplate potentials in neuromuscular junctions (Duncan & Statham, 1977).

Too little is known about the biochemistry of membranes involved in exocytosis in paramecia to make it worth while to try to project into our system all the detailed knowledge on membrane fusion which has been obtained by analysing well-defined synthetic systems. All our evidence underscores the necessity for lipids to be in a fluid state, as in cell fusion experiments (Ahkong et al. 1973; Papahadjopoulos, Poste & Schaeffer, 1973; VanDerBosch, Schudt & Pette, 1973).

At what steps is energy needed when a trichocyst is discharged from a Paramecium cell? In their well-known review article Poste & Allison (1973) indicate reasons why locally bound ATP would have to be hydrolysed at the membrane contact sites. If Ca²⁺ for exocytosis would possibly be mobilized from internal pools such mechanisms could also require energy. One also has to envisage actin- or actomyosin-like materials around some secretory granules (Palade, 1975) and around trichocysts (Bannister, 1972). Similar material is visible on the ∼15-nm-wide contact zone between the cell membrane and a trichocyst (Plattner et al. 1973, 1975, 1977)> which could act as a ‘mechanocomplex’ postulated by Poste & Allison (1973). Similar structures were recently recognized to ‘tie’ secretory granules to the cell membrane in different systems. The actin- or actomyosin-nature of such structures has not yet been proved in Paramecium cells. In other systems actin and myosin bind to secretory granule membranes in vitro (Burridge & Phillips, 1975) and display Ca²⁺-ATPase activity (Clarke & Spudich, 1977). It is unknown whether such structures are candidates for energy transduction during exocytosis of trichocysts, but they could indeed facilitate in different ways the attraction of membranes to be fused. The trichocyst tip is flanked by a tightly fitting microtubular collar (Bannister, 1972; Plattner et al. 1973). In pituitary cells ‘dynein-like crossbridges’ connect the secretory granules with microtubules (Sherline, Lee & Jacobs, 1977). Although it is unclear on the one hand whether they display ATPase activity (cf. Sherline et al. 1977) and on the other hand whether such connecting structures exist on Paramecium trichocysts, it seems noteworthy, that we found calcium phosphate deposits in these regions, which neatly reflected the arrangement of microtubules (Plattner & Fuchs, 1975).

The stretching of isolated trichocyst contents upon addition of Apyrase (Table 1) indicates that stretching possibly involves the hydrolysis of matrix-bound ATP. This could be expected from earlier work (Hoffmann-Berling, 1961) and also explains, why contracted trichocyst contents can be isolated only with ATP added (Anderer & Hausmann, 1977). We had previously obtained some indications by X-ray microanalysis for a sudden Ca²+ influx into trichocysts under trigger conditions and for the occurrence of phosphate-splitting sites within trichocysts (Plattner & Fuchs, 1975). To confirm this assumption spectrophotometric ATPase assays will have to be done. A note on ATPase activity of trichocyst contents was given by Hoffmann-Berling (1961). The impressive ultrastructural changes of matrix proteins upon discharge (Hausmann et al. 1972 a, b;Bannister, 1972) indicate that conformational changes act as a driving force for their expulsion. Since trichocysts can be stretched in the absence of a boundary membrane, this is a strong argument against the osmotic gradient theory (which presumes membrane-integrated particle aggregates at exocytosis sites to act as gap-junction analogues) proposed by Satir (1974), and in favour of our previous evidence against this theory which was obtained from tracer analyses (Plattner et al.

Trichocysts stretch explosively during discharge, i.e. within ∼0 ·5ms according to high-speed microkinematographic analyses (Plattner, Bilinski & Völlenklee, unpublished results). Given the large extent of random coil organization in adrenergic vesicles (Smith & Winkler, 1967; Sharp & Richards, 1977) and the presumable absence of expandable proteins in cholinergic transmitter vesicles (cf. Meldolesi et al. in the press) there does not appear to exist a common feature for extrusion. However, exocytosis in mast cells (Douglas, 1974) and in nerve terminals (Heuser, 1977) is perhaps as rapid and vigorous.

In Paramecium cells exocytosis stimulation by all kinds of triggers strongly reduces the ATP content of cells. For the only exception observed, polyethyleneglycol, we have no explanation so far. As indicated in the Results section, the ATP-loss determined after triggering might be somewhat underestimated. Evidently the whole exocytotic process needs energy. In most systems some energy is invested into resynthesis of secretory materials during triggering and for transcellular transport (Palade, 1975; Meldolesi et al. in the press). Data recently obtained for lacrymal gland cells by Herzog, Sies & Miller (1976) on the increase of O₂ consumption during exocytosis stimulation allow for some rough calculations. The basal O₂ consumption per unit time and cell volume (37 °C) is about twice as high as in Paramecium cells (22 ^?C). In contrast, the energy needed per one discharge event in paramecia appears to be only a very small percentage of that which one can estimate for the lacrymal gland cells. Therefore, we believe that the energy required for the final exocytosis steps would be much smaller than that consumed during the preceding steps of secretory activity.

We thank Professor H. Benger and Dr F. Tiefenbrunner for making available to us some laboratory equipment and several firms for gifts of chemicals (Eli Lilly: A23187; Hofimann-LaRoche: X-537A; Knoll AG: D600). Part of this work is from a PhD thesis by M. B. This work was supported by the österreichische Fonds zur Förderung der wissenschaftlichen For?chung.