J. Cell Set. 46, 17-40 (1980) Printed in Great Britain © Company of Biologists Limited 1080
17
GENETIC DISSECTION OF THE FINAL EXOCYTOSIS STEPS IN PARAMECIUM TETRAURELIA
CELLS: CYTOCHEMICAL
DETERMINATION OF Ca2+-ATPase ACTIVITY OVER PREFORMED EXOCYTOSIS SITES H. PLATTNER,1* K. REICHEL,1 H. MATT,1 J. BEISSON,' M. LEFORT-TRAN' AND M. POUPHILE* 1
Faculty of Biology, University of Ko>utans, P.O.B. 5560, D-7750 Konstanz, Federal Republic of Germany, 1 Department of Dermatology, University of Innsbruck, Anichstr. 4, A-bozo Innsbruck, Austria, 3 C.N.R.S., Centre de Ginitique Moliculaire, .F-9110.0 Gif-sur-Yvette, France, and * C.N.R.S., Laboratoire de la Cytophysiologie de la Photosynthhe, F-91190 Gif-sur- Yvette, France
SUMMARY
In different Paramecium tetraurelia strains the occurrence of a Ca*+-ATPase (or p-nitrophenylphosphatase) activity at the preformed attachment and exocytosis sites of the secretory vesicles (trichocysts) was analysed by electron-microscopic cytochemistry and X-ray microanalysis. In conjunction with freeze-fracture studies it was found that only those strains, which contain rosette particles, display this Ca1+-ATPase activity (jS, K 401, nd 9 (18 °C)), while other strains (nd 6,nd Q (27 °C), tarn 38) are devoid of both these characteristics. The presence (absence) of rosette particles and of Ca1+-ATPase activity at the preformed exocytosis sites is correlated with the capability (incapability) of performing exocytosis in these strains. We discuss several possible interpretations of this structure-function correlation.
INTRODUCTION
Wild-type strains of Paramecium tetraurelia cells contain specialized secretoryvesicles, called trichocysts, firmly attached to the cell membrane in a regular distribution pattern (see Ehret & McArdle, 1974; Jurand & Selman, 1968; Sonneborn, 1970). Using the freeze-fracture technique, Bachmann, Schmitt & Plattner (1972) and Janisch (1972) presented the first evidence that the cell membrane contains regular arrays of membrane-intercalated-particles (MIP) at the docking sites of trichocysts. In a more detailed analysis a variety of different membrane specializations were described to occur at these docking sites (Plattner, Miller & Bachmann, 1973). Most impressive are 300-nm large double rings of ~ 9 nm large MIP which surround a rosette of about ten ~ 13-nm large MIP, located directly over the trichocyst attachment sites (see Fig. 2). Also with the use of freeze-fracturing Beisson et al. (1976a) demonstrated that •
To whom reprint requests should be sent.
18
H. Planner and others
mutant strains which do not produce trichocysts (trichless strain of Pollack (1974)) are devoid of rosette MIP and that rings present themselves as crescent-shaped MIP aggregates ('parentheses') at these unoccupied potential trichocyst docking sites; evidently a parenthesis would be expanded to a ring only when a trichocyst becomes attached to the cell membrane. This is in agreement with the assumption that the rings lie beyond the exocytosis site proper, which they encircle, as previously concluded from a morphometric analysis performed on K 401 cells (Plattner et al. 1973). Beisson et al. (1976 a) also analysed mutant strains with temperature-dependent gene expression controlling the exocytotic system: nd 9 cells contain trichocysts always attached to the cell membrane; however, when grown at a non-permissive temperature (27 °C) they are incapable of performing discharge (in response to picric acid, commonly used in Paramecium genetics), whereas after growth at a permissive temperature (18 °C) nd 9 cells do discharge. While rings were present in permissive and non-permissive nd 9 cells, rosette particles were largely absent selectively from nonpermissive cells. Strains nd 6 and nd 7 which are non-permissive at any culture temperature behave like nd 9 (27 °C) cells. For a schematic characterization of different P. tetraurelia strains see Figs. 1 and 2. synthesis
trichless
packing transport
trigger
~\
(
nd9(27°C) 'non-permissive'
membrane fusion )
\ nd6, nd 7 (s\\ ten
discharge membrane resealing membrane detachment
7S, K401, kin 241 nd 9 (18 °C)
='permissive'
Fig. 1. Steps involved in exocytosis and their genetic inhibition at different levels (horizontal lines) in different P. tetraurelia strains. For further specifications of strains see Materials and methods of this and the following paper.
Recently a Ca2+-dependent ATPase activity was localized precisely at these preformed attachment and exocytosis sites in wild-type cells by electron-microscopic cytochemistry (Plattner, Reichel & Matt, 1977). One may envisage a variety of potential candidates for an ultrastructural equivalent of this Ca2+-ATPase, since the attachment site of trichocysts at the cell membrane displays a variety of distinct ultrastructural elements (see Fig. 2): Apart from ring and rosette MIP within the cell membrane, there occurs in addition connecting material of unidentified nature (which is assumed to connect the trichocyst and the cell membrane, holding them at a distance of ~ i 5 to 30 nm; see Plattner, Wolfram, Bachmann & Wachter, 1975;
Genetic dissection of exocytosis
19
Plattner et al. 1977; this paper) and the broad MIP annulus encircles (i.e. slightly below) the uppermost trichocyst membrane portion ('e-type' MIP, Plattner et al. 1973; Allen & Hausmann, 1976; Beisson et al. 1976a; Plattner et al. 1977). Due to diffusion of reaction product the spatial resolution of cytochemical analyses would be too poor to assign the Ca2+-ATPase activity to any one of these structures occurring in wild-type strains within this small area. Therefore, it was the aim of the present study to expand these cytochemical experiments to mutant strains which lack selectively some of these substructures at the potential exocytosis sites. The accompanying paper (Matt et al. 1980) further correlates these cytochemical findings with the capability of different mutants to perform exocytosis in response to different stimuli. MATERIALS AND METHODS Strains The following strains of Paramecium tetraurelia (formerly P. aurelia syngen 4, see Sonneborn, 1975) were used: K 401 and jS which were our reference strains of wild-type phenotype with respect to exocytosis properties, and various mutants: kin 241, tarn j8,ftb A, nd 6, nd 7, nd p. K 401 is a derivation of 8tock 51 carrying the thermolethal mutation ts 40/; jS is strain d 4,2, mating type 7. kin 241 (Beisson et al. 19766), tarn 38 (Ruiz, Adoutte, Rossignol & Beisson, 1976), nd 9 (Beisson et al. 1976a) and nd 7 were derived from stock d 4,2; nd 6 was derived from stock d 4-43 (see Sonneborn, 1974) and ftb A from stock 51 (Pollack, 1974). Except for kin 241 which has normal exocytosis properties, all other mutants are affected in trichocyst morphogenesis, attachment or secretion: tarn 38 contains only a few trichocysts of abnormal morphology which never attach to the cell membrane, ftb A displays a similar phenotype. nd 6 and nd 7 have normally attached trichocysts which cannot be secreted, nd 9 displays the same phenotype but in a thermosensitive way: cells grown at 18 °C can discharge their trichocysts while cells grown at 27 °C cannot. All these mutations affect independent loci (Rossignol & Beisson, 1975). Culture conditions All cultures were transferred to an identical culture medium. A decoct of Difco dried lettuce medium (2 g per 1. distilled water) was filtered, autoclaved (20 min at 120 °C) and inoculated with Enterobacter agglomerans. Contaminants were removed initially by a 24-h exposure to a mixture of each of 100 /tg/1. potassium-penicillin G and streptomycin sulphate. Cultures monoxenic for E. agglomerans were grown at 25 °C, except when indicated otherwise, till they reached the early stationary phase. Chemicals For previously used chemicals the specifications and sources were as indicated by Matt, Bilinski & Plattner (1978). Sigma (St. Louis, Mo.) supplied Tris-ATP, sodium-ADP, sodiumAMP, 3',s'-cAMP, its dibutyryl derivative, adenylyl imidodiphosphate, acetylphosphate, sodium-/?-glycerophosphate, atractyloside, oligomycin, ouabain (strophantin-G) and quercetin. />-nitrophenylphosphate, L ( + )-tartrate and Tiron® (1,2-dihydroxybenzene-disulphonic(3,5)Na,) were from Merck (Darmstadt, Fed. Rep. Germany), GTP was from Serva (Heidelberg, Fed. Rep. Germany). L-tetramisole was donated by Janssen Pharmaceutica (Breerse, Belgium). Baa+, Ca'+, Fe a+ , Hg'+, Mg»+, Mn'+, Srs+, Zna+, K+, Na+ were added as chlorides, La' + as nitrate. All chemicals were of the highest purity available. Durcupan ACM® was obtained from Fluka (Buchs, Switzerland).
20
H. Plattner and others
V
:•
i
wild-type 7S k401
trichless football tarn 8 tarn 38 (no attachment
rm
nd9(18°C
•0
(i.e. permiaiv
•f
nd9 {27 ° £
(i.e. non-
•
permissive)
Fig. 2.
Genetic dissection of exocytosis
21
Measurements of cation concentrations Ca, Mg, K and Na concentrations were determined by atomic absorption, colorimetrically and/or by flame photometry as outlined previously (Matt et al. 1978) in the culture medium and during the different processing steps for electron-microscopic cytochemistry. Data given in Table 1 (p. 25) were determined for K 401 cultures; practically identical values were found for cultures of all other strains. Cytochemistry Fixation. Cells were washed (3x5 min) with an excess of 10 mM Tris/HCl buffer pH yo and concentrated to 0-5 ml, containing ~ 10s cells in suspension. 05 ml glutaraldehyde, 2-5 (v/v) % in o-i M cacodylate buffer, pH 70, was added in a vigorous stream. After 1 min the fixative was diluted to 0-25 % by addition of o-i M cacodylate buffer for another 14 min. Then, cells were washed (3x5 min) in an excess of o-i M Tris/maleate buffer pH 70 and suspended in incubation media for cytochemical assays. Incubation media. Incubation was carried out for 1 h at 20 °C in 01 M Tris/maleate buffer, pH 7'O, with 2-4 mM Pb(NO3), added as a capture ion in combination with various substrates (5 mM) and cations (5 HIM); see Table 2 p. 26. We used ATP, ADP, AMP, 3'5'-cAMP, dibutyryl 3',5'-cAMP, adenylyl imidodiphosphate, GTP, p-nitrophenylphosphate, acetylphosphate, sodium-/?-glycerophosphate, some of which were combined with different cations (Ba«+, Ca«+, Cu«+, Fe1+, Hg*+, Mga+, Mn'+ Sr'+, Zn»+, K+, Na+; La3+). Some controls were run without substrates and cations added, with ATP alone, with ATP + 1 mM EGTA, with Bal+, Ca1+ or Mg1+ alone or with ATP + Ca1+ or ^-nitrophenylphosphate substrates in the absence of Pb 1+ or with Pb 1+ in the presence of a 4-fold molar excess of the Pb 1+ chelator i,2-dihydroxybenzene-disulphonic-(3,5)Na, (Tiron®). Aliquots were exposed to general ATPase inhibitors (iV-ethylmaleimide, mersalyl acid, Salyrgan®, p-chloromercuribenzoate; quercetin: 0-2 mM in DMSO) and to other inhibitors (oligomycin: io/tg/ml in ethanolic solution; ouabain, L-tetramisol, L( + )-tartrate, atractyloside). All inhibitors were also added to the last wash and, unless indicated otherwise, applied at a concentration of 5 mM. Other control aliquots were heated for 5 min to 80 °C (after fixation); others were soaked in c i M phosphate buffer, pH 7-0, before exposure to a substrate- and Ca*+-free Pb l + solution in Tris/maleate buffer, pH 7-0. In the few cases in which precipitation occurred the media were renewed. The pH was maintained at pH 7-0.
Fig. 2. Trichocyst-plasma membrane interactions and ultrastructural appearance of the trichocyst attachment sites in different P. tetraurelia strains, a = rings of membrane-intercalated particles (MIP) found within the plasma membrane around trichocyst attachment sites; am (am0) = inner (outer) alveolar membrane; c = rosette MIP; pm = plasma membrane; tm = trichocyst membrane. The dots in the upper trichocyst membrane region (e) mark annulus MIP. Cross-links drawn between rosettes and the trichocyst tip region represent schematically connecting material, thought to connect trichocyst and plasma membrane at least in permissive strains. Comparable connections are assumed to link the plasma membrane to the alveolar membrane and the latter to the trichocyst membrane, so that trichocysts can be attached to the cell periphery even in non-permissive strain nd 9 (27 °C) in the absence of rosettes and rosette-associated connecting material. Possibly trichocysts are also connected to the plasma membrane in the region of rings. The precise arrangement and extension and the chemical nature of such connecting materials has not yet been analysed. The scheme shows also the presence of a complete set of ultrastructural details in strains 7S, K401, kin 241 and nd 9 (18 °C); the lack of trichocysts in the trichless mutant; the failure to attach trichocysts in tarn 8, tarn 38 and football strains. For further specification of strains see Fig. 1 and Materials and methods of this and the accompanying contribution. Data are compiled from. Beisson et al. (1976a), Plattner et al. (1973), from unpublished work (Lefort-Tran et al. in preparation) and from the material presented here.
22
H. Plattner and others
While this full set of experimental variations was applied to wild-type strains (jS, K 401) the schedule for mutant strains was restricted to the following media: no substrates or cations added, ATP or Ca1+ alone, ATP + EGTA (1 mM), ATP + Ca1+, ^-nitrophenylphosphate. The residual cation concentration in the incubation media can be seen in Table i, p. 25. Further processing. After incubation the cells were washed with 10 mM Tris/maleate, pH 70 and further processed for electron microscopy by osmication, dehydration and embedding (see below). Samples selected for X-ray microanalysis were neither osmicated nor stained. Sections from cytochemical experiments were generally viewed in the electron microscope without section staining. Osmotic shock treatment To demonstrate the presence of' connecting material' between cell membrane and trichocyst tip membrane in K 401 and nd 9 (18 °C) and its absence in nd 9 (27 °C) mutants these cells were fixed in glutaraldehyde (25 v/v %) for only 1 min. After centrifugation they were resuspended for another 10 min in cacodylate buffer, pH 7-0, at final concentrations of o-oi M, o-i or i-o M (i.e. in the absence of a fixative). Further processing included the usual osmication, dehydration, embedding and section staining procedure (see below). General electron-microscopic techniques Cells pretreated as described above for cytochemistry were postfixed in 1 % osmium tetroxide (in o-1 M cacodylate buffer, pH 7'o), washed in the same buffer for 10 min, dehydrated in 30 (v/v) %, 50, 75, 90, 95 (10 min each) and 100% acetone (3 x 10 min), impregnated overnight with acetone (2 vol. parts) plus Durcupan ACM® (1 part) and embedded in this resin. Ultrathin sections were analysed in the electron microscope without staining or after routine uranyl acetate (2 % aqueous solution, 20 min) and lead citrate staining (pH 12-0, 3 min). The acceleration voltage was 60 kV for unstained and 80 kV for stained sections; a 5O-/im objective aperture was used. Energy dispersive X-ray microanalysis Unstained, 200-nm-thick sections were prepared from non-osmicated samples, mounted on carbon-coated nylon grids (Touzart-Matignon, Paris) and covered with an additional 20-nmthick carbon layer with an electron beam evaporator. We analysed only samples from 7,!? and K 401 cells after incubation in ATP + Ca'+ or £-nitrophenylphosphate media. The operating conditions for the Jeol 100CX electron microscope (scanning transmission mode, side-entry goniometer stage), equipped with a lithium-drifted silicon-detector (Kevex®) and a spectrometer unit (Link® system 290), were as described previously (Plattner & Fuchs, 1975) and as slightly modified later (Plattner et al. 1977). Durcupan ACM® embedding is favourable because it does not exhibit a chlorine Kc^-peak (see Plattner et al. 1977). Abbreviations used Chemicals. ADP = adenosinediphosphate; AMP = adenosinemonophosphate; ATP = adenosinetriphosphate: cAMP = cyclic adenosinemonophosphate; DMSO = dimethylsulphoxide; EGTA = ethyleneglycol-Wj(/?-aminoethyl ether) A^AT-tetraacetate; GTP = guanosinetriphosphate; Tris = tris(hydroxymethyl)-aminomethane. Structures, a = 'ring' of membrane-intercalated particles within the plasma membrane around trichocyst attachment sites; am, = inner alveolar membrane; am0 = outer alveolar membrane; c = 'rosette' formed by ~ 10 membrane-intercalated particles within the plasma membrane directly over trichocyst tips; e = 'annulus' formed by membrane-intercalated particles around the trichocyst tip; MIP = membrane-intercalated particles; pm = plasma membrane; t = trichocyst; tb = trichocyst body (lower portion of a trichocyst); tm = trichocyst membrane; tt = trichocyst tip (upper portion of a trichocyst). For the arrangement of these structural elements see Fig. 2.
Genetic dissection of exocytosis
23
RESULTS
Figs. 3 to 9 are characteristic for the cytochemical reactions obtained at the preformed trichocyst attachment and exocytosis sites in P. tetraurelia wild type {?S: Figs. 3, 5, 6) and mutant strains (nd 9 (18 °C): Fig. 7; nd 9 (2/ °C): Fig. 8; tarn 38: Fi g- 9)In reactive cells reaction product is formed almost exclusively between the trichocyst and the plasma membrane; sometimes some reaction product is deposited outside this region (e.g. on the cell membrane-outer alveolar membrane complex and on the inner alveolar membrane). Reactive cells yield a reaction product equally well with ATP + Ca2+ as with />-nitrophenylphosphate. The following strains are reactive : jS, K 401 (not illustrated here; see Plattner et al. 1977) kin 241 (not illustrated) and nd 9 (18 °C). No reaction product is formed at the attachment sites of nd 6 and nd 9 (2j °C) cells or at the unoccupied potential trichocyst docking sites of tarn 38. The cytochemical reaction obtained with nd 7 cells is rather inconsistent. Frequently, a reduced amount of precipitate is found at the periphery of some trichocyst attachment sites; freeze-fracture analyses of nd 7 independently indicate the present of some large MIP within the rings, which could be rosette MIP, but without being assembled properly (unpublished observations). Reactive cells do not form a reaction product in the absence of ATP or ^>-nitrophenylphosphate or in the presence of ATP without Ca2+ added, regardless of the strain used. The background concentrations of cations during processing for cytochemistry are listed in Table 1. To achieve greater reliability all sections from cytochemical experiments were analysed without section staining. With 7 5 and K 401 cells extensive experiments were made with the use of different substrate, cation and inhibitor combinations (Table 2). Bivalent cations alone, applied after fixation, yield no reaction product. (For results obtained with Ca2+ in the presence of the fixative see Plattner & Fuchs, 1975). With ATP and Ca2+ or Ba2+ cells give a strong reaction; a less intense reaction occurs with ATP+Mg 2 + . As some spurious reaction can occur sometimes with ATP alone due to the background Ca2+-content (Table 1) we also made controls with EDTA added, which were all negative. No reaction is observed when ATP is combined with other cations (Fe 2+ , Hg2+, Mn 2+ , Sr2+, Zn2+, K+, Na+ or La3+). Table 2 also indicates the formation ot reaction product with several monophosphate substrates (acetylphosphate, AMP; less with /?-glycerophosphate) other than ^>-nitrophenylphosphate. Adenylyl imidodiphosphate, cAMP, dibutyryl cAMP, ADP, GTP gave no reaction. Further controls listed in Table 2 were negative: when Pb2+-capture ions were omitted; when Pb 2+ was chelated; when fixed cells were heat-inactivated before incubation. We also soaked fixed cells in phosphate buffer before suspension in a Pb 2+ medium (without substrates or earth alkali ions added) in order to investigate whether randomly formed Pb 2+ phosphate precipitates would selectively bind to certain structures; this was found not to occur.
H. Plattner and others
L
Genetic dissection of exocytosis
25
Furthermore, Table 2 contains experiments with different inhibitors. As expected, ouabain (inhibitor for K + , Na+-transport ATPase), L-tetramisole (inhibitor for alkaline phosphatase), atractyloside (inhibitor for nucleotide transport) or oligomycin (selective inhibitor for mitochondrial fx ATPase) do not inhibit reaction product formation in the presence of ATP + Ca2+. As Table 2 shows general ATPase inhibitors, like iV-ethylmaleimide, p-chloromercuribenzoate, mersalyl acid, Salyrgan and quercetin lead to only a partial reduction but not to a total abolition of reactivity with ATP + Ca2+ or />-nitrophenylphosphate media. This might be due to the stabilizing effect of the prefixation, but this effect was not analysed in detail. Table 1. Cation concentrations (OTM) in media during culturing and processing of P. tetraurelia cells for electron microscopic cytochemistry Culture and preparation conditions
Mg'+
K+
Na+
Ca*+
i-8o
o-34 ±0-05
018
018
±o-io
±003
±0-06
(B) Sterilized culture medium (C) Early stationary culture
178
o-39
016
0-16
167
040
016
0-14
(D) (C) after washing with 10-fold excess of 20 ITLM Tris buffer PH70
018
006
0-026
0163
(E) 100-fold excess of wash buffer
009
0-07
0015
0068
(F) 300- to 1000-fold excess of wash buffer
0-05
019
0048
0-041
0-035 ±0021
0042 ±0009
0-035 ±0003
0052 ±0006
47-7X
95X
4-6x
27X
(A) 14-day permanent culture
(G) Like (F) but subsequently fixed with glutaraldehyde 3 times washed with o-i M Tris buffer PH70
(G/C) Dilution factor
In X-ray spectrograms only P and Pb but no Ca peaks are detectable (Fig. 4). This indicates that Pb 2+ and not Ca2+ acted as a capture ion (see Discussion). This is in accordance with the presumed presence of a Ca2+-stimulated ATPase activity. Table 3 summarizes the cytochemical results obtained with different P. tetraurelia strains and with the use of different media. In Table 4 these results are put in relation to the ultrastructural organization of the trichocyst attachment region: Strains K 401, 7S, kin 241, and nd 9 (18 °C), i.e. all strains in which a rosette is present in freeze-fracture replicas and with the capability of extruding their trichocysts (see Fig. 3. Tangentional section of a ?S cell incubated in a ATP + Ca1+ medium. Positions I-J mark trichocyst attachment sites cut at different levels. Positions 1-4 contain the attachment zone proper so that they contain reaction product; in positions j , 6 and 7 trichocysts are cut at lower levels so that a decreasing amount of reaction product is encountered. The cell axis is oriented perpendicular to the micrograph (2-5, 1-4, 3-7). Note that a few deposits occur also along the plasma membrane and the outer alveolar membrane. No section staining, x 48000. 3
cm. 46
26
H. Plattner and others
accompanying paper) display a Ca 2+ -ATPase and ^-nitrophenylphosphatase reaction over the trichocysts; all these characteristics are absent from non-permissive cells. An important new ultrastructural detail is documented in Fig. 10. When permissive nd 9 cells (cultured at 18 °C) and non-permissive nd g cells (cultured at 27 °C) are exposed to a hypertonic treatment after very brief fixation (see Materials and methods) the trichocyst-cell membrane attachment sites present a very different appearance: with nd 9 (2/ °C) cells the plasma membrane is frequently detached from the trichocyst membrane so that a bleb is formed selectively over the trichocyst attachment Table 2. Electron-microscopic cytochemistry of trichocyst-to-plasmalemma zones in P. tetraurelia wild-type strains (K 401 and yS) Incubation medium.1 ATP-cation combinations Tris-ATP Baa+, Ca'+ or Mg*+ Tris-ATP + K+ + Na+ Tris-ATP + Bas+ or Caa+ Tris-ATP+ Mg'+ Tris-ATP + Fe'+ or Hga+, Mn'+, Sr»+ or Zn2+ Tris-ATP + La3 + Monophosphate substrates ^-nitrophenylphosphate, acetylphosphate, sodium-AMP Sodium-/?-glycerophosphate Other substrates Sodium-ADP, adenylyl imidodiphosphate, 3',5'-cAMP or dibutyryl-3 ',5 '-cAMP Sodium-ADP + Ca'+ Omission of Pb2+-capture ions Tris-ATP + Ca'+ £-nitrophenylphosphate Chelation of PbJ+-capture ions 1,2-dihydroxybenzene- disulphonic-(3,5)Na, + Tris-ATP + CaJ+ Pre-incubation with phosphate buffer Pb'+-m.edium without substrate and other cations Heat inactivation 5 min 80 °C, Tris-ATP + Ca*+ Non-ATPase inhibitors (type of function inhibited) Tris-ATP+ Caa+ + ouabain (K+, Na+-ATPase) or L-tetramisole (alkaline phosphatase), L( + )-tartrate (acid phosphatase) or atractyloside (nucleotide transport) Mitochondrial ATPase inhibitor Tris-ATP + Cas+ + oligomycin General ATPase inhibitors Tris-ATP + Ca*+ + iV-ethylmaleimide or ^-chloromercuribenzoate, mersalyl acid, Salyrgan or quercetin ^-nitrophenylphosphate+p-chloromercuribenzoate or quercetin or Salyrgan®
attachment
Reactivity* o o o + (+) o o + (+) o o o o o o o
+ + (+) (+)
1 Pb 1+ as a capture ion at pH 7-0; for concentrations and further details see Materials and methods. a o = no or only spurious reaction; ( + ) = weak or inconsistent reaction; + = full reactivity.
Genetic dissection of exocytosis
27
site; this does not occur when nd 9 {18 °C) (or K 401) cells are treated in an identical way. Concomitantly only permissive cells contain some moderately electron-dense, morphologically ill-defined and biochemically unidentified material between both these membranes, whereas this space appears electron-translucent in nd 9 (2/ °C) cells. In permissive cells the space between cell membrane and trichocyst membrane is always ~ 15 to ~ 30 nm wide, even after osmotic shock treatment. In this strain
15 r10
I
5
-
8
a a. T
£
c
Q.
JS
CD
» n
AAi \
CD
8!
1 ~
mi
a 1
r"
lati
CD
CD
en
0
1
•
i
2
3
1
4
5
Energy, keV
Fig. 4. Energy spectrogram obtained by X-ray microanalysis of reaction product formed over trichocyst attachment sites in wild-type cells after incubation in ATP + Ca*+ medium. For further details see Materials and methods. Note the presence of P-Kaj, and Pb-Ma peaks in the absence of a Ca-Ka peak which would be at 3-690 keV; the Si-peak left from P-Ka^ is an artifact.
the presence (absence) of this material correlates with the presence (absence) of rosette MIP in the cell membrane (see Discussion), with the capability of performing exocytosis (see following paper) and with the presence (absence) of a Ca2+-ATPase at the trichocyst attachment and exocytosis sites. Due to our limited knowledge on this material between trichocyst and cell membrane we refer to it as to the ' connecting material'. So far we have not yet analysed systematically other mutant strains. The present findings on nd 9 cells, however, were recently confirmed by a physiological approach with the use of quite different methods (Beisson, Cohen, Lefort-Tran, Pouphile & Rossignol, 1980). 3-2
H. Plattner and others
28
Table 3. Electron-microscopic cytochemistry of trichocyst-to-plasmalemma zones in P. tetraurelia cells1
attachment
Incubation medium
Strain
ATP
ATP
+
+ EGTA
p-nitrophenylphosphate
No additives
ATP
Ca'+
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
±'
0
±«
K 401 7S kin 241 nd 9 (iS °C) nd 9 (27 °C) nd6 tarn 38 nd 7
Ca«+
+ + + +
+ + + +
0 0 0 0
1
All substrate and ion concentrations were 5 mM; EGTA was 1 mM. All experiments were performed at 20 °C; when nd 9 (18 °C) or nd 9 {27 °C) cells were incubated at their culture temperatures or at reversed temperatures, results were not significantly different, o = no or only spurious reaction; + = full reactivity. For further experimental details see Materials and methods. 1 Inconsistent results with no reactivity or with some reaction product near the ring regions instead of over the full trichocyst attachment site. Table 4. Electron-microscopic morphology and cytochemical Cd*+-ATPase reactivity of preformed exocytosis sites (trichocyst-to-plasma membrane attachment zones) in P. tetraurelia Freeze-fracture morphology
ATPase cytochemistry j.
Strain
Ring
K4017Sb kin 241° nd9(r8°C)b
Present Present Present Present Present Present Present" Present
nd 9 (27 °C) b nd6A tarn 38* nd 7*
Rosette Present Present Present Present
Annulus
ATP + Caa+
p-nitrophenylphosphate
Present Present Present Present Present Present
Positive Positive Positive Positive Negative Negative Negative 1 Inconsistent
Positive Positive Positive Positive Negative Negative Negative Inconsistent'
Absent Absent Absent
Absent
Inconsistent
Present
* Freeze-fracture data from Bachmann et al. (1972) and Plattner et al. (1973). b Freeze-fracture data from Beisson et al. (1976a). 0 Freeze-fracture data from Lefort-Tran et al. (unpublished results). d Freeze-fracture data from Lefort-Tran, Pouphile, Rossignol, Aufderheide & Beisson (in preparation). ° Rings are present in a compressed form (parentheses). ' Fewer particles than usually and no assembly of a rosette proper; reaction product poor and occasionally concentrated near the periphery of rings. Fig. 5. ?S cell incubated in a£-nitrophenylphosphate medium. Heavy depositions of reaction product occur at the attachment sites of trichocysts (arrowheads). Only a little reaction product surrounds the lower regions of the trichocyst tip (tt) membrane and none is present on the trichocyst body (tb). No section staining, x 30000.
Genetic dissection of exocytosis
tt
29
H. Plattner and others
3° DISCUSSION
Specificity of cytochemical reaction
As to be expected, inhibitors directed against non ATP-splitting enzymes do not reduce the reactivity with Ca2+ +ATP (Table 2). Furthermore, any interference with adenylate cyclase is excluded since adenylyl imidodiphosphate is a substrate for adenylate cyclase (Howell & Whitfield, 1972), but not for ATPase (Yount, Babcock, Ballantyne & Ojala, 1971), and cannot substitute for ATP in our experiments. Sur-
6
!
tt
tt
I*.
Fig. 6. Longitudinal sections through trichocyst tips (tt) of 7S celk. Left: ATP + Ca1+ medium. Right :p-nitrophenylphosphate medium. Arrowheads point towards reaction product in the cleft between trichocyst tip and plasma membrane. The dots outline the trichocyst tips. No section staining, x 60000.
prisingly, general ATPase inhibitors (for references see Plattner et al. 1977; for quercetin see: Kuriki & Racker, 1976, and Fewtrell & Gomperts, 1977a) do not totally abolish the reaction, although they reduce it. Evidently ATPase molecules are sufficiently stabilized by the brief prefixation with glutaraldehyde; possibly sulphhydryl groups, where inhibitors interfere with the ATPase activity, are already exhausted by the fixative with which they also react (Habeeb & Hiramoto, 1968). This prefixation is also sufficient to abolish the triggering effect of ATPase inhibitors (as well as of other agents) on trichocyst discharge (Matt & Plattner, unpublished observations).
Genetic dissection of exocytosis
31
Precise values of local intracellular cation concentrations after the fixation procedure used are, of course not known. A rough estimate can be attempted on the basis of extracellular (Table 1) and intracellular cation concentrations (K+, Ca2+; Naitoh & Eckert, 1974), known for Paramecium cells, and from measurements on the rate of cation equilibration during glutaraldehyde fixation determined for other cell types (Penttila, Kalimo & Trump, 1974). [Na+, Ca2+, Mg 2+ ]! can be estimated to be about o-i rail, while [K+], could perhaps be as high as ~ 10 mM. Therefore, the precise cation requirements (expecially of K+) for the reaction could be quite different from the concentrations in the medium. As we found no differences in the cultures of
tt
Fig. 7. Trichocyst attachment sites in tangentional {top) and longitudinal (bottom) sections. Permissive nd g (18 °C) cells incubated in ATP + Ca2+ (left), in ^-nirrophenylphosphate (middle) or in a medium without additives (right). In the latter case no reaction product is formed. Dots outline the trichocyst tips (tt). No section staining, x 60000.
H. Plattner and others
Genetic dissection of exocytosis
33
different strains, we conclude that the ionic milieu was identical in all cases and, thus, is not relevant for the different cytochemical reactions found in different strains. Phosphate ions could theoretically be precipitated by Ca2+, which invades cells during fixation, if any type of phosphate-splitting enzyme were present, regardless of any stimulating effect of Ca2+. Since we detected only P and Pb in the reaction product, when submitted to electron-microscopic X-ray microanalysis, Ca s+ is assumed to act as a stimulating ion rather than as an incidental capture ion. This is also in accordance with the lack of reaction product formation without Pb 2+ added or after addition of a Pb2+-chelator (Table 2). As is generally known, the fixation procedure used accounts largely for the selectivity of cytochemical ATPase reactions (Essner, 1973). The preferable Ca2+-ATPase reactivity of trichocyst attachment sites does not therefore, imply the absence of other ATPases from other regions of the cell surface. Structural correlation
The intermembrane space between trichocyst tip and cell membrane is the site where reaction product is deposited upon incubation with ATP + Ca2+ (and a few other bivalent cations) or monophosphate media. According to Table 4 these cytochemical results are closely correlated only with the presence of rosette MIP. Therefore, other structural elements, such as annulus and ring MIP can be excluded from the list of candidates for the structural equivalents for the enzyme activity found at the preformed exocytosis sites. Probably, b-particles are not candidates for the Ca2+ATPase, since they are recognized also in nd g (2/ °C) and tarn 8 cells (Beisson et al. 1976a), which are all non-reactive. Annulus MIP are unlikely candidates also because they are located at a distance from the uppermost trichocyst tip region (Allen & Hausmann, 1976; Beisson et al. 1976a). Concerning strain nd 7, which gave inconsistent results (Table 4), recent unpublished results (Lefort-Tran et al. in preparation) indicate the following. No rosettes of the usual type are assembled; instead membrane particles of the size of rosette particles, but less in number than in rosettes of K 401, 7S, kin 241 or nd 9 (18 °C), are scattered throughout the area surrounded by the rings. Simultaneously the reactivity for Caa+-ATPase and ^-nitrophenylphosphatase is either absent or weak and then appears to be concentrated along the periphery of the rings. Despite their inconsistency, these results are also in favour of rosette particles being the most likely ultrastructural counterparts of these enzyme activities at the trichocyst attachment sites. It also appears possible that they have to be present in a minimal amount to be functionally active. Preliminary results indicate the presence of so far unidentified 'connecting materials' in permissive cells and its absence (or presence in reduced amount) from some Fig. 8. Non-permissive nd g (2; °C) cells incubated in ATP + CaI+. tb, trichocyst body; tt, trichocyst tip. All trichocyst attachment sites (arrows and circles) are unreactive, in longitudinal (top) as well as in tangentional section (bottom). No section staining. Top, x 7500; bottom, x 16700.
H. Plattner and others
34
tb
Sl^^Sf
Genetic dissection of exocytosis
35
non-permissive mutants (see above). Among several possibilities, soluble proteins between the membranes engaged in exocytosis could be involved in the assembly or dispersal of rosette MIP or in regulating the local fusion capability (see below). Possible functional aspects
Recent critical evaluations of the present knowledge on the final steps of exocytosis reveal considerable uncertainties (see Plattner, 1978; Stossel et al. 1978). For reasons outlined in the Introduction, Paramecium cells offer a unique possibility for structurefunction correlation with regard to regions involved selectively in the final steps of exocytosis. In view of the rather selective effect of Ca2+ on exocytotic performance in general (Douglas, 1974) and in Paramecium cells in particular (Plattner, 1974; Matt, Bilinski & Plattner, 1978) it is obvious that the bivalent cation-stimulated ATPase, which we localized, acts probably as a Ca2+-ATPase. The reactivity with monophosphate substrates indicates the presence of an intermediate phosphorylationdephosphorylation step as known with other ATPase systems, including various membrane-bound ATPase (Korenbrot, 1977). Given the fact that only exocytosis-capable Paramecium strains display a Ca2+ATPase at precisely that portion of the cell membrane where exocytosis takes place, what functions could be involved? At present we envisage several possibilities: a function of rosette particles as a Ca2+-translocator and/or a Ca2+-channel; a system for the regulation of local ATP and/or Ca2+ bound to the membrane; protein phosphorylation related to fusion; a possible involvement of surface proteins, such as 'connecting material', and possibly yet others. As outlined in the accompanying paper it would be reasonable to assume a special Ca2+-regulating system at these strategic areas. By compensating for the Ca2+leakage into the cell periphery a local Ca2+-regulating system would be required to enable the cells to maintain the trichocysts in a position ready-for-discharge. If so, this system would permanently invest energy in order to maintain the ready-fordischarge state (much like the costly contemporary strategic missile systems). It would then be logical to assume also that rosette MIP would act not only as a Ca2+-pump but also as a Ca2+-channel in order to allow for Ca2+-entry precisely at the sites where trichocyst firing has to be ignited. In non-permissive strains both functions could be disturbed simultaneously; this could find its morphological equivalent in a disturbance or lack of rosette assembly. As indicated above, this disturbance could also involve a soluble (Ca2+-?) regulating factor. Both functions, a Ca2+-ATPase active translocator and an ionophoretic channel function, are intimately coupled with each other also in other biomembranes (mast cells: Fewtrell & Gomperts, 1977a, b; sarcoplasmic reticulum; Shamoo & MacLennan, 1974). In the sarcoplasmic reticulum these functions are also performed by membrane-integrated proteins which present themselves as MIP in freeze-fracture replicas (Baskin, 1974) Fig. 9. tarn 38 cell. Only a few trichocysts are found (tb, trichocyst body) but none is attached to the cell membrane; arrowheads indicate the potential docking sites for trichocysts which are more clearly recognized in tangentional section {inset). No section staining, x 21000.
Fig. 10. Longitudinal sections through nd g trichocyst attachment sites after osmotic shock treatment. Arrowheads outline the uppermost trichocyst tip ( t t ) region. Large arrows indicate the zones to be compared. L f t , nd 9 (18O C ) ; note the presence of moderately electron-dense connecting material between trichocyst and cell membrane and that both membranes remain closely attached to each other despite the hyperosmotic treatment. Middle and right, nd g (27 "C) cells showing 2 appearances found after osmotic shock treatment, that on the right being more commonly found. I n both examples no electron-dense connecting material is present between trichocyst and cell membrane. Right, both membranes become frequently separated from each other due to osmotic shrinkage of the cell body, so that the cell membrane forms a bleb selectively over the trichocyst attachment site. Note the strict maintenance of membrane-to-membrane attachments in the regions where rings (a)and annulus (e) occur in freeze-fracture experiments, and also the occurrence of a central elevation on the uppermost trichocyst tip region. Section staining. x 125000.
Genetic dissection of exocytosis
37
and a transient phosphorylation step also takes place (Korenbrot, 1977). In our system the results from trigger analysis with permissive and non-permissive Paramecium strains (see accompanying communication) with the use of ATPase inhibitors could be interpreted in the sense of a simultaneous presence or absence of both functions. Both could depend on the assembly of rosette MIP. However, there is no direct experimental evidence for a possible function of rosette MIP as Ca2+-channels so far. Evidence presented by Satir & Oberg (1978) must be interpreted in quite different terms, i.e. as a solvent induced side-effect, as discussed in the following paper. We have obtained some hints of the possible relevance of membrane-connecting material, presumably a protein, for exocytosis permissivity (this paper and Beisson et al. in press). It could be involved in rosette assembly or, for example in conferring Ca2+-sensitivity on a local Ca2+-pumping and/or channel system (see above). Along these lines it is interesting to note that certain soluble proteins enhance the membrane fusion rate at physiologically low [Ca2+], in chromafRn cells (Pollard, Pazoles, Creutz & Zinder, 1979) and that a soluble protein, identified as Ca2+-dependent regulator protein (calmodulin), regulates the Ca2+-ATPase activity in various other systems (see Larsen & Vincenti, 1979; Cheung, 1980). Recently, Baker & Whitaker (1980) presented some evidence for the possible relevance of calmodulin for exocytosis. Does non-permissivity imply a defect on this basis? If so, non-assembly of rosette MIP, deficiency of local Ca2+-ATPase activity and lack of exocytosis performance in non-permissive strains would be secondary features of a primary defect of the kind indicated above. Recent microinjection studies by Beisson et al. (1980) provide some indirect evidence for the functional importance of a soluble protein for exocytosis performance. A soluble protein of the type of' connecting material' could affect also the following possible functions of a local Ca2+-ATPase (p-nitrophenylphosphatase). Firstly, according to Poste & Allison (1973) membrane-bound Ca2+ and ATP have to be locally removed for membrane fusion to occur. Secondly, Zakai, Kulka & Loyter (1976) give some experimental evidence that insoluble calcium phosphate precipitate enhances the fusion suceptibility of membranes in an unknown way. Thirdly, there is an increasing body of evidence that some membrane proteins have to be phosphorylated before membrane fusion can occur (DeLorenzo, Freedman, Yohe & Maurer, 1979)Although it will need further analyses to ascertain which of these possible Ca2+ATPase functions govern permissivity or non-permissivity for exocytosis in Paramecium cells, the present and the following paper give important clues on the involvement of clearly defined structural and functional parameters under genetic control in the regulation of the final steps of exocytosis in Paramecium. We are grateful to Drs J. Benger, F. Tiefenbrunner and H. Winkler for providing laboratory equipment, to Dr M. Schinner for help with the atomic absorption measurements, to Mr A. Gora for giving us access to a Jeol 100 CX X-ray microanalysis unit, to Mrs M. Matt for photometric cation measurements, to Miss A. T. Orque for ultrathin sectioning, to Jansson Pharmaceutica for a gift of tetramisole. This work was supported by the Deutsche
38
H. Plattner and others
Forschung8gemeinschaft and SFB 138, by the Osterreichische Forschungsfonds, by a shortterm grant from the European Molecular Biology Organization (to H.P.) for a journey at Gif-sur-Yvette and by grant no. 77.7.0267 from the Delegation Generale a la Recherche Scientifique et Technique (to M.L.T.).
REFERENCES ALLEN, R. D. & HAUSMANN, K. (1976). Membrane behavior of exocytic vesicles. I. The ultrastructure of Paramecium trichocysts in freeze-fracture preparations. J. Ultrastruct. Res. 54, 224-234. BACHMANN, L., SCHMITT, W. W. & PLATTNER, H. (1972). Improved cryofixation: Demonstrated on freeze-etched solutions, cell fractions and unicellular organisms. Proc. 5th Eur. reg. Congr. Electron Microsc. (ed. V. E. Cosslett), pp. 244-245. London and Bristol: The Institute of Physics. BAKER, P. F. & WHITAKER, M. J. (1980). Trifluoperazine inhibits exocytosis in sea-urchin eggs. jf. Physiol., Lond. 298, 55 P. BASKIN, R. J. (1974). Ultrastructure and calcium transport in microsomes from developing muscle. J. Ultrastruct. Res. 49, 348-371. BEISSON, J., COHEN, J., LEFORT-TRAN, M., POUPHILE, M. & ROSSIGNOL, M. (1980). Control of membrane fusion in exocytosis: Physiological studies on a Paramecium mutant blocked in the final step of the trichocyst extrusion process. J. Cell Biol. 85, 213-227. BEISSON, J., LEFORT-TRAN, M., POUPHILE, M., ROSSIGNOL, M. & SATIR, B. (1976a). Genetic analysis of membrane differentiation in Paramecium. Freeze-fracture study of the trichocyst cycle in wild-type and mutant strains. J. Cell Biol. 69, 126—143. BEISSON, J. & ROSSIGNOL, M. (1975). Movements and positioning of organelles in Paramecium aurelia. In Nucleocytoplasmic Relationships during Cell Morphogenesis in Some Unicellular Organisms (ed. S. Puiseux-Dao), pp. 291-294. Amsterdam, New York, London: Elsevier. BEISSON, J., ROSSIGNOL, M., RUIZ, F., ADOUTTE, A. & GRANDCHAMP, S. (19766). Genetic analysis of morphometric processes in Paramecium: A mutation affecting cortical pattern and nucleolar reorganization. J. Protozool. 23, 3A. CHEUNG, W. Y. (1980). Calmodulin plays a pivotal role in cellular regulation. Science, N.Y. 207, 19-27. DELORENZO, R. J., FREEDMAN, S. D., YOKE, W. B. & MAURER, S. C. (1979). Stimulation of Ca1+-dependent neurotransmitter release and presynaptic nerve terminal protein phosphorylation by calmodulin and a calmodulin-like protein isolated from synaptic vesicles. Proc. natn. Acad. Sci. U.S.A. 76, 1838-1842. DOUGLAS, W. W. (1974). Involvement of calcium
in exocytosis and the exocytosis-vesiculation sequence. Biochem. Soc. Symp. 39, 1-28. EHRET, C. F. & MCARDLE, E. W. (1974). The structure of Paramecium as viewed from its constituent levels of organization. In Paramecium. A Current Survey (ed. W. J. Van Wagtendonk), pp. 263-338. Amsterdam, London, New York: Elsevier. ESSNER, E. (1973). Phosphatases. In Electron Microscopy of Enzymes, Principles and Methods (ed. M. A. Hayat), pp. 44-76. New York etc.: Van Nostrand Reinhold. FEWTRELL, C M . S. & GOMPERTS, B. D. (1977a). Effect of flavone inhibitors of transport ATPases on histamine secretion from rat mast cells. Nature, Lond. 265, 633-636. 1+ FEWTRELL, C. M. S. & GOMPERTS, B. D. (19776) Quercetin: A novel inhibitor of Ca influx and exocytosis in rat peritoneal mast cells. Biochim. biophys. Acta 469, 52-60. HABEEB, A. F. S. & HntAMOTO, R. (1968). Reaction of proteins with glutaraldehyde. Archs Biochem. Biophys. 126, 16-26. HOWELL, S. L. & WHITFIELD, M. (1972). Cytochemical localization of adenyl cyclase activity in rat islet of Langerhans. J. Histochem. Cytochem. 20, 873-879. JANISCH, R. (1972). Pellicle of Paramecium caudatum as revealed by freeze etching. J. Protozool. 19, 470-472. JURAND, A. & SELMAN, G. G. (1968). The Anatomy of Paramecium aurelia. London: MacMillan. KORENBROT, J. I. (i977). Ion transport in membranes: Incorporation of biological iontranslocating proteins in model membrane systems. A. Rev. Physiol. 39, 19—49.
Genetic dissection of exocytosis
39
+
Y. & RACKEH, E. (1976). Inhibition of (Na , K+) adenosine triphosphatase and its partial reactions by quercetin. Biochemistry, N.Y. 15, 4951-4956. LARSEN, F. L. & VINCENTI, F. F. (1979). Calcium transport across the plasma membrane: Stimulation by calmodulin. Science, N. Y. 204, 306-309. MATT, H., BILINSKI, M. & PLATTNER, H. (1978). Adenosinetriphosphate, calcium and temperature requirements for the final steps of exocytosis in Paramecium cells. J. Cell Sci. 32, 67-86. MATT, H., PLATTNER, H., REICHEL, K., LEFORT-TRAN, M. & BEISSON, J. (1980). Genetic dissection of the final exocytosis steps in Paramecium tetraurelia cells: Trigger analyses. J. Cell Sci. 46, 41-60. NAITOH, Y. & ECKERT, R. (1974). The control of ciliary activity in protozoa. In Cilia and Flagella (ed. M. A. Sleigh), pp. 305-352. London, New York: Academic Press. PENTTILA, A., KALIMO, H. & TRUMP, B. F. (1974). Influence of glutaraldehyde and/or osmium tetroxide on cell volume, ion content, mechanical stability, and membrane permeability of Ehrlich ascites tumor cells. J. Cell Biol. 63, 197-214. PLATTNER, H. (1974). Intramembraneous changes upon cationophore-triggered exocytosis in Paramecium. Nature, Lond. 252, 722-724. PLATTNER, H. (1978). Fusion of cellular membranes. In Transport of Macromolecules in Cellular Systems, Life Sci. Res. Rep. 11 (ed. S. C. Silverstein), pp. 465-488. Dahlem Konferenzen (Berlin). PLATTNER, H. & FUCHS, S. (1975). X-ray microanalysis of calcium binding sites in Paramecium. With special reference to exocytosis. Histochemistry 45, 23-47. PLATTNER, H., MILLER, F. & BACHMANN, L. (1973). Membrane specializations in the form of regular membrane-to-membrane attachment sites in Paramecium. A correlated freezeetching and ultrathin-sectioning analysis. J. Cell Sci. 13, 687—719. PLATTNER, H., REICHEL, K. & MATT, H. (1977). Divalent-cation-stimulated ATPase activity at preformed exocytosis sites in Paramecium coincides with membrane-intercalated particle aggregates. Nature, Lond. 267, 702-704. PLATTNER, H., WOLFRAM, D., BACHMANN, L. & WACHTER, E. (1975). Tracer and freezeetching analysis of intra-cellular membrane junctions in Paramecium. With a note on a new heme-nonapeptide tracer. Histochemistry 45, 1-21. POLLACK, S. (1974). Mutations affecting the trichocysts in Paramecium aurelia. I. Morphology and description of the mutants. J. Protozool. 21, 352-362. POLLARD, H. B., PAZOLES, C. J., CREUTZ, C. E. & ZINDER, O. (1979). The chromaffin granule and possible mechanisms of exocytosis. Int. Rev. Cytol. 58, 159-197. POSTE, G. & ALLISON, A. C. (1973). Membrane fusion. BiocHm. biophys. Acta 300, 421-465. Ruiz, F., ADOUTTE, A., ROSSIGNOL, M. & BEISSON, J. (1976). Genetic analysis of morphogenetic processes in Paramecium. I. A mutation affecting trichocyst formation and nuclear division. Genet. Res. 57, 109-122. J+ SATIR, B. H. & OBERG, S. G. (1978). Paramecium fusion rosettes: Possible function as Ca gates. Science, N.Y. 199, 536-538. a SHAMOO, A. E. & MACLENNAN, D. H. (1974). A Ca +-dependent and -selective ionophore as part of the (Ca1+ + Mg8+)-dependent adenosinetriphosphatase of sarcoplasmic reticulum. Proc. natn. Acad. Sci. U.S.A. 71, 3522-3526. SONNEBORN, T. M. (1970). Methods in Paramecium research. In Methods in Cell Physiology (ed. D. M. Prescott), pp. 241-339. New York and London: Academic Press. SONNEBORN, T. M. (1974). Paramecium aurelia. In Handbook of Genetics, vol. 2 (ed. R. Kung), pp. 469-594. New York: Plenum Press. SONNEBORN, T. M. (1975). The Paramecium aurelia complex of 14 sibling species. Trans. Am. microsc. Soc. 94, 155-178. KURIKI,
STOSSEL, T. P., BRETSCHER, M. S., CECCARELLI, B., DALES, S., HJELENIUS, A., HEUSER, J. E., HUBBARD, A. L., KARTENBECK, J., KINNE, R., PAPAHADJOPOULOS, D., PEARSE, B., PLATTNER, H., POLLARD, T. D., REUTER, W., SATIR, B. H., SCHLIWA, M., SCHNEIDER, Y. J., SILVERSTEIN, S. C. & WEBER, K. (1978). Membrane dynamics. In Transport of Macromolecules in
Cellular Systems, Life Sci. Res. Rep. n (ed. S. C. Silverstein), pp. 503-516. Dahlem Konferenzen (Berlin).
40
H. Plattner and others
R. G., BABCOCK, D., BALLANTYKE, W. & OJALA, D. (1971). Adenylyl imidodiphosphate, an adenosine triphosphate analog containing a P-N-P linkage. Biochemistry, N. Y. io, 2484-2489. ZAKAI, N., KULKA, R. G. & LOYTER, A. (1976). Fusion of human erythrocyte ghosts promoted by the combined action of calcium and phosphate ions. Nature, Lond. 263, 696-699.
YOUNT,
(Received 26 October 1979)
