- Email: [email protected]
Immunoelectron microscopy reveals significant granule fusion upon stimulation of electropermeabilized human neutrophils
Cellular Signalling Vol. 6, No. 1, pp. 47-58, 1994. Copyright © 1994 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0~98--6568/94 $6.00 + 0.00
Pergamon
IMMUNOELECTRON MICROSCOPY REVEALS SIGNIFICANT GRANULE FUSION UPON STIMULATION OF ELECTROPERMEABILIZED HUMAN NEUTROPHILS HANS W. M. NIESSEN,* Jos J. M. ONDERWATER,t HENK K. KOERTEN,tLEO A. GINSEL* and ARTHUR J. VERHOEVEN*§ *Central Laboratory of the Netherlands Red Cross Blood Transfusion Service and Laboratory for Experimental and Clinical Immunology of the University of Amsterdam, Amsterdam, The Netherlands; -~Laboratoryfor Electron Microscopy, University of Leiden, Leiden, The Netherlands; and ~Department of Cell Biology and Histology, University of Nijmegen, The Netherlands (Received 4 July 1993; and accepted 12 August 1993) Abstract--Although electropermeabilization has become an important tool for studying the signal requirements of exocytosis, relatively little is known about the morphological changes accompanying this response in electropermeabilized cells. In this study, we determined that electropermeabilization of human neutrophils by itself caused only minor changes in the morphology as determined by transmission electron microscopy. The structure of the plasma membrane did not show detectable changes, whereas the cytoplasm was more electron lucent as compared to intact cells. Activation of intact neutrophils with forrnyl-methionyl-leucyl-phenylalanine (FMLP), in the presence of cytochalasin B, caused the development of invaginations of the plasma membrane. In contrast, activation of electropermeabilized cells with 1 p.M Ca2+and/or 50 ~tM GTP-T-S caused the development of vacuoles that did not seem to be in contact (or had previously been in contact) with the extracellular environment. However, fusion of azurophilic and specific granules with these vacuoles clearly had taken place. The response characteristics of this fusion induced by Ca2+ and GTP-T-S were quite similar to those of the direct fusion of granules with the plasma membrane. We conclude that in electropermeabilized human neutrophils, two processes involving granule fusion can be distinguished. First, a direct fusion of granules with the plasma membrane. Secondly, the fusion of granules leading to the formation of vacuoles, not in contact with the extracellular space.
INTRODUCTION PHAGOCYTOSIS and killing of bacteria by neutrophils are important mechanisms in the host defence against invading mechanisms. The killing process is mediated by the concomitant release of lysosomal enzymes, bactericidal proteins and toxic oxygen metabolites into the phagosomes formed [ 1, 2]. For the release of proteins, fusion of intracellular granules with the phagosomal membrane is required. Recently, this degranulation process has been §Author to whom correspondenceshould be addressed at: Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, c/o Publication Secretariat, Plesmanlaan 125, 1066 CX Amsterdam,The Netherlands. Abbreviations: FMLP--formyl-methionyl-leucyl-phenylalanine; MPO~myeloporoxidase; FACS--fluorescence-aetivated cell sorter; FITC--fluoresceinisotliiocyanate;DMSO--dimethyl sulphoxide; PFA--paraformaldehyde; IgG-imrnunoglobulinG; PBS--phosphate-bufferedsaline. 47
studied in electropermeabihzed neutrophils by measuring the release of enzymes [3] or by measuring the up-regulation of specific granule membrane markers [4]. The results of these studies suggest that in electropermeabilized neutrophils the machinery of the degranulation process is intact with different requirements for degranulation between azurophilic and specific granules. Most notably, elevation of the free Ca 2÷ concentration alone to values mimicking that in stimulated neutrophils was found to be sufficient for enzyme release from the specific, but not from the azurophilic granules [4]. Knight et al. [5] have shown by electron microscopy that in electropermeabilized chromaffin cells and electropermeabilized platelets, little evidence of plasma membrane damage can be observed. Until now, the effect of the electropermeabilization procedure on the morphology of neutrophils has not been studied.
H.W.M. NmSSENet al.
48
In intact neutrophils, Hara et al. [6] have shown by electron microscopy that activation of neutrophils with the chemotactic tripeptide FMLP (in the presence of cytochalasin B) causes invaginations of the plasma membrane into the cells. The effects of activation of electropermeabilized neutrophils at the morphological level are as yet unknown. We have therefore studied the effect of electropermeabilization and subsequent activation of the neutrophils on the morphology of the cells analysing the changes accompanying degranulation. For this purpose, transmission microscopy and immunoelectron microscopy were used. MATERIALS AND METHODS
Materials FMLP and cytochalasinB (Sigma Chemical Co., St. Louis, MO, U.S.A.) were dissolved in DMSO at 1000 times the final concentration for cell incubations, and were stored at -20"C. GTPy-S was obtained from Boehringer (ingelbeim, Germany), dissolved in H20) at 100 times the final concentration for cell incubations, and was stored at -20"C. CD63 and CD67 monoelonal antibodies and FITC-labelled goat antibodies against mouse IgG were obtained from our own institute (Central Laboratory of the Blood Transfusion Service, Amsterdam, The Netherlands). All other chemicals were reagent grade. The basal incubation medium for cell suspensions consisted of 132 mM NaCI, 6 mM KC1, 1 mM CaC12, 1 mM MgSO4, 1.2 mM potassium phosphate, 20 mM Hepes, 5.5 mM glucose and 0.5% (w/v) human albumin, pH 7.4. "Electropermeabilization medium" consisted of 140 mM KCI, 1 mM MgC12, 1 mM EGTA, 20 mM Hepes and 10 mM glucose (pH 7.2).
Isolation of neutrophils Blood was obtained from healthy volunteers. Granulocytes were purified from the buffy coat of 500 ml blood anticoagulated with 0.4% (w/v) trisodium citrate (pH 7.4), as described by Verhoeven et al. [7]. After isolation, the cells were suspended in incubation medium and kept at room temperature.
Electropermeabilization and subsequent activation of neutrophils Cells were permeab'dized immeAiately before use, as the method described by Grinstein and Furuya [8]. In short, a sample of 107 cells was washed once and resuspended in 1 ml of ice-cold electropermeabilization medium. Part (0.8 nil) of this suspension was transferred to a BioRad Pulser cuvette and subjected to two discharges of 5 kV/cm2 from a 25 pFd capacitator, in the BioRad GenePulser. Between the two pulses, the cells were stirred gently with a plastic pipette. Several batches of neutrophils were electropermeabilized consecutively, pooled and stored at 4"C. Four minutes after the first batch of cells had been permeabilized, the permeabilized cells were transferred from 4oc to 37"C. Two minutes later, 1 mM ATP and cytochalasin B (5 I.tg/ml) were added. After 1 rain of further incubation, the cells were stimulated with Ca2+ and/or GTP-y-S. The free Ca2÷ concentrations in the incubations were checked by Indo-1 measurements after adding 0.1 ~M Indo-1 (free acid) to the medium [9]. At the time points indicated, 1% (v/v) PFA was added to stop the degranulation process [4].
Pre-embedding labelling of neutrophils Electropermeabilized neutrophils were activated with 1 ~ Ca2+and/or 50 pM GTP-T-S and subsequently fixed at different time points. Intact neutrophils were activated with FMLP (in the presence of cytochalasin B) for 1 min and subsequently fixed with 1% PFA. The cells were then incubated in a mixture of 1% PFA and 0.1% glutaraldehyde for 2 h at room temperature. Subsequently, the cells were washed with PBS and incubated at room t e ~ for 60 rain with CD63 or with CD67 antibodies. These antibodies can be used to assess independently the degranulation of azurophilic and specific granules, respectively [4, 10, 11]. The cells were then washed again with PBS, and incubate! first with rabbit anti-mouse IgG and then with protein A conjugated to gold with a diameter of about 10 nm. As a negative control, the primary antibody incubation step was omitted. Next, the pellet was
Degranulationin electropermeabilizedneutrophils post-fixed for 10 rain at room temperature in 1.5% glutaraldehyde in 0.14 M caeodylate buffer (pH 7.4) and then for 15 rain at 4"C in OsO4. After pelleting of OsO+-fixed cells in 2% Bacto-agar and subsequent dehydration in a graded series of ethanol dilutions, the cells were embedded in epon. Ultra-thin sections stained with lead hydroxide were examined in a Philips E M 201 electron microscope at 80 kV.
49
cytochalasin B). The cells were then fixed in a mixture of 1% PFA and 0.1% glutaraldehyde for 2 h at room temperature, and prepared for cryosectioning as described previously [12]. After cryosectioning, streptavidin coupled to gold (5 tam), was added. As a negative control, biotin-Nbydroxy-suceinimide ester was omitted from the pre-incubation step. The sections were examined with a Philips EM 201 at 80 kV.
Immunocytochemistry of CD63 and CD67 Activation of electropermeabilized neutrophils was stopped at different time points with 1% PFA. The cells were then fLxed a mixture of 1% PFA and 0.1% glutaraldehyde for 2 h at room temperature, and prepared for cryosectioning and immunolabelling as described previously [12]. In short, immunogold double-labelling of CD63 and MPO, or CD67 and lactoferrin was performed by incubation of the cryosections with MoAb CD63 or CD67, followed by a second antibody (rabbit anti-mouse IgG), and then protein A-conjugated colloidal gold particles with a diameter of about 10 nm. Thereafter, the cryosections were incubated with a polyclonal rabbit antibody against myeloperoxidase or lactoferrin, respectively, and protein A-coupled to gold (5 urn). As controls, the primary antibody incubation step was replaced by incubation with non-relevant mouse and rabbit IgG. The sections were examined with a Philips EM 201 at 80 kV.
lmmunocytochemistry of biotinylated cells Intact neutrophils were incubated with biotinN-hydroxy-succinimide ester (100 Ixg/ml) for 1 h at room temperature. In this way, plasma membrane proteins were efficiently biotinylated as indicated by FACS analysis after incubation of the cells with streptavidin-FITC (data not shown). Subsequently, the cells were electropermeabilized and activated with 1 lxrn Ca2÷plus 50 p_M GTP-TS for 2 min. The plasma membrane of these dec_ tropermeabilized cells still contained biotinylated proteins as established by FACS analysis. For comparison, intact cells labelled with biotin were stimulated with FMLP (in the presence of
RESULTS First, the effect of the electropermeabllization procedure on the morphology of neutrophils was evaluated. Unstimulated intact neutrophils embedtied in epon (Fig. 1A) were compared with unstimulated electropermeabilized neutrophils (Fig. 1B). The cytoplasm of electropermeabilized neutrophils was more electron-lucent as compared to the intact cells, while sometimes the permeabilized cells showed less dense nuclear material. Despite the eleclropermeabilization procedure, the granules and the plasma membrane of the permeabflized neutrophils did not show any difference with the intact neutrophils, even at a high magnification (45,000 x) (Fig. 1C). However, we established that the electropermeabilized neutrophils were permeable for propidium iodide as measured by flow cytometry [8]. Next, the morphological effects of stimulation of the electropermeabilized neutrophils with 1 ~tM Ca2+ and/or 50 lxM GTP-T-S were studied. After activation, the number of granules clearly diminished. Stimulation with 1 BM Ca2÷ plus 50 GTP-T-S diminished the total amount of granules with 89% (20 cells) as compared m unstimulated electropermeabilized neutrophils (incubated with a [Cae+]f~ of about 100 nM). Stimulation with 1 IxM Ca2+ diminished the total amount of granules with 59% (20 cells). Furthermore, after stimulation with 1 BM Ca2+ plus 50 pM GTP-T-S, large vacuoles were formed (Fig. 2A), which were not visible in unstimulated electropermeabilized neutrophils (Fig. 1B). These vacuoles were also found after stimulation with 1 BM Ca2+ (Fig. 2B) or with 50 BM GTP-y-S (not shown). There was no significant difference in number or in the
50
H.W.M. Nmss~'~et aL
volume of the vacuoles formed under the different activation conditions studied (results not shown). To assess whether the vacuoles appearing in electropermeabilized neutrophils after stimulation were in fact large invaginations of the plasma membrane, the permeabilized cells were stimulated and subsequently incubated with CD63 or CD67 monoclonal antibodies, followed by an incubation with rabbit anti-mouse IgG and protein A-conjugated colloidal gold particles (10 nm) prior to embedding in epon. In this protocol, only proteins on the plasma membrane in contact with the extracellular space will become labelled. Under all conditions of stimulation, no gold particles were detected near the membranes of the vacuoles (Fig. 2). Staining of the vacuolar membrane was also absent in samples taken as early as 30 sec after stimulation. In contrast, labelling was clearly visible at the plasma membrane (Fig. 2B). When the same procedure of in-suspension labelling was applied to intact neutrophils stimulated with the chemotactic tripeptide FMLP (in the presence of cytochalasin B), the vacuoles appearing under these conditions of stimulation contained significant numbers of gold particles bound to CD63 antibodies (Fig. 3). Positive staining of these vacuoles was also obtained with CD67 as primary antibody (results not shown). These results support the conclusion of Hara e t al. [6], that the vacuoles appearing in intact neutrophils under these conditions of stimulation are in fact invaginations of the plasma membrane, in which degranulation of both azurophilic granules (as detected by CD63 antibodies) and specific granules (as detected by CD67 antibodies) had taken place. When, in elcctropermeabilized neutrophils under the various conditions of stimulation, the number of gold particles detecting CD63 or CD67 on the plasma membrane was counted, it was established that the pattern of up-regulation was in good agreement with our earlier results obtained by FACS analysis [4]. Most notably, activation of the cells with 1 IIM Ca 2+ alone did not cause an increase in CD63 labelling at the plasma membrane (Fig. 4). However, there
was a clear increase in the amount of labelling with CD63 after addition of 50 p.M GTP-~-S + 1 lxM Ca ~'. In contrast, a clear increase in the amount of gold particles detecting CD67 was detected on the plasma membrane after activation with 1 ktM Ca 2+as the sole stimulus, indicating the fusion of specific granules with the plasma membrane. The absence of CD63 and CD67 labelling in the vacuoles formed upon activation might be due to the absence of fusion of granules with these structures. We therefore prepared frozen sections of electropermeabilized neutrophils after stimulating the cells with 1 pNI Ca 2÷ and/or 50 JIM GTP)'-S. Subsequently, the MAbs CD63 or CD67 were added to the frozen sections, followed by the addition of rabbit anti-mouse IgG and protein A gold (10 nm). In addition, anti-MPO or anti-lactoferrin, visualized by addition of 5 nm protein A gold, were added to the frozen sections labelled with
300 --
T
¢j
"~
200
q~
100
o CD63
CD67
FIG. 4. Amount of protein colloidal gold particles on the plasma membrane of electropermeabilized cells activated with Ca 2+ and/or GTP-T-S. Electropermeabifized neutrophils (107/ml) were preincubated at 37"C for 3 rain in the presence of cytochalasin B (5 ~tg/ml) and subsequently activated with 1 ~tM Ca2+ and/or 50 p.M GTP-y-S for 8 rain and fixed with 1% PFA. Then the cells were labelled with CD63 and CD67. Subsequently, the cells were prepared for epon (MAbs in suspension) as described in Materials and Methods. The results are presented as the mean (+ S.E.M.) of 20 cells. ( I ) Unsfimulated cells; ([~) 1 pM Ca2+; ([~) 1 ~tM Ca2÷plus 50 ~ GTP-T-S.
FIG. 1. Morphology of intact and electropermeabilized neutrophils. Intact and electropermeabilized neutrophils (lO’/ml) were incubated at 37°C in the presence of cytochalasin B (5 pg/ml) for 4 min and then fixed with 1% PFA. The cells were then prepared for epon as described in Materials and Methods. (A) Intact neutrophils; bar: 0.5 pm. (B) Electropermeabilized neutrophils; bar: 0.5 Frn.
51
FIG. 1. Morphology of intact and ¢lectropermeabilize,d neutrophils. Intact and electropermeabilized neutrophils (107/ml) were incubated at 37"C in the presence of cytochalasin B (5 gg/ml) for4 min and then fixed with 1% PFA. The cells were then prepared for epon as described in Materials and Methods. (C) Electropermeabilized neutrophils; bar: 0.1 gm.
52
FIG. 2. Morphology and in-suspension labelling of activated electropermeabilized neutrophils. Electropermeabilized neutrophils (107/ml) were preincubated for 3 min at 37°C in the presence of cytochalasin B (5 ~tg/ml) and subsequently stimulated for 8 min with (A) 1 I.tM Ca 2÷ plus 50 [tM GTP-y-S or with (B) 1 paM Ca 2÷, and subsequently fixed with 1% PFA. The cells were then prepared for epon as described in Material and Methods. Gold particles in (B) represent CD67 antibodies, which had been added prior to embedding in epon. Bar: 0.5 lxm. 53
FIG. 3. CD63 Labelling in suspension of activated neutrophils. Intact neutrophils (10711111)were preincubated for 3 min at 37°C in the presence of cytoehalasin B (5 ~tg/ml) and subsequently activated with FMLP for 1 min and fixed with 1% PFA. After labelling with CD63, the cells were prepared for epon (MAbs in suspension) as described in Materials and Methods. Bar: 0.1 I.tm; arrow: plasma membrane.
~A
FIG. 5. Frozen sections of activated electropermeabilized neutrophils. Electropermeabilized neutrophils (107/ml) were preincubated at 370C for 3 min in the presence of cytochalasin B (5 ~tg/ml) and subsequently activated for 8 min with (A) 1 l.tM Ca 2÷ plus 50 ktM GTP-T-S or (B) 1 IxM Ca 2÷. Thereafter, the cells were fixed with 1% PFA. The cells were then treated for frozen sectioning as described in Materials and Methods. The cryosections were incubated with MoAb CD63 (detected by 10 nm gold particles) and anti-MPO antiserum (detected by 5 nm gold particles). Bar: 0.1 lam. 55
FIG. 6. Frozen sections of intact and electropermeabilized cells labelled with biotin. Intact neutrophils (107/ml) were preincubated for 1 h with biotin-N-hydroxy-succinimide ester (100 pg/ml) at room temperature. After washing, one part of the cells was activated for 1 min with FMLP (in the presence of cytochalasin B), and another part was electropermeabilized and subsequently stimulated with Ca*+plus GTP-YS for 2 min, in the presence of cytochalasin B. After cryosectioning, streptavidin A coupled to gold (5 nm) was added. (A) Intact neutrophils; bar 0.1 pm. (B) Electropetmeabilized neutrophils, bar: 0.1 urn.
56
Degranulation in electropermeabilized neutrophils
MAb CD63 or CD67, respectively. With this experimental procedure, there was a clear labelling of CD63 at the membranes of the vacuoles, accompanied by gold particles detecting MPO (Fig. 5A). To achieve positive staining of the vacuolar membranes with CD63 antibodies, the permeabilized cells had to be stimulated with 1 laM Ca 2÷ plus 50 pNl GTP-T-S (Fig. 5A) or with 50 ttlVl GTP-T-S (results not shown). Almost no gold labelling of CD63 and/or MPO was detected in the vacuoles after stimulation with 1 BM Ca2÷as the sole stimulus (Fig. 5B). Under the latter condition of stimulation, there was, however, gold labelling of CD67 at the membranes of the vacuoles, accompanied by gold particles detecting lactoferrin (results not shown), which was comparable with cells activated with 1 pM Ca 2+ plus 50 JaM GTP-T-S. Hence, the labelling pattern of CD63 and CD67 on the membranes of the vacuoles induced after stimulation with Ca 2÷ and/or GTP-T-S, was similar to that found on the plasma membrane presented above (Fig. 4). The binding of the monoclonal antibodies to the vacuoles in the frozen sections of the activated cells and the absence of binding to the vacuoles when the cells were stained with antibodies in suspension prior to embedding, would suggest that these vacuoles are not in contact with the extracellular environment. To investigate whether there had been contact with the extracellular space, we determined the presence of plasma membrane proteins in the vacuolar membrane. For this purpose, intact cells were biotinylated prior to the elcctropcrmcabilization and subsequently activated for 2 rain with I B M Ca 2+ plus 50 IxM GTP7-S. As a positive control, intact cells labelled with biotin were stimulated with F M L P (in the presence of cytochalasin B). Subsequent to cryosectioning, streptavidin coupled to gold was added. Indeed, the vacuoles observed after activation of intact cells contained plasma membrane proteins (Fig. 6A). However, the vacuoles of the clectropermeabilized neutrophils were not labelled (Fig. 6B), supporting the hypothesis that these vacuoles were formed intracellularlywithout contacting the plasma membrane.
57
DISCUSSION In this study we have investigated the degranulation process in electropermeabilized human neutrophils using transmission (immuno-)electron microscopy. Similar to the results .found by Knight and Scrutton [5] in electropcrrneabilized chromaffin cells and electropcrmeabilized platelates, we observed that the plasma membrane of the electropermcabilized neutrophils did not show any visible changes. However, the cytoplasm in the pcrmeabilizcd neutrophils was more electron lucent than in intact cells.This was also found by Smolcn et al. [13] in neutrophils pcrmeabilized with digitonin. Furthermore, sometimes the nucleus appeared less electron dense in the permeabilized neutrophils. However, differences in morphological appearance were much more apparent after activation. In intact cells treated with cytochalasin B, FMLP caused the development of invaginations of the plasma membrane intruding into the cells concurrently with degranulation. These invaginations have been reported previously by Hara et al. [6]. These authors used Ruthenium Red staining to prove contact between the vacuolar content and the extracellular medium. Esaguy et al. [14] suggested that in intact neutrophils activated with PMA vacuoles are formed, but these authors did not distinguish between vacuoles formed intracellularly or invaginations of the plasma membrane. Similarly, Hoffstein et al. [15] suggested that intact neutrophils activated with A23187 show invaginations, but also these authors did not prove whether these were true invaginations or intracellular vacuoles. In electropermeabilized neutrophils treated with cytochalasin B, we found that after activation with Ca 2+ and/or GTP-T-S, the cells also developed vacuoles. In these cells however, antibodies against relevant membrane markers were not able to bind to the vacuolar membrane after addition to the cells in suspension, indicating that these vacuoles were not in contact with the extracellular medium. Furthermore, the vacuoles seemed not to be derived from an endocytotic process (Fig. 6B). Instead, the vacuoles seemed to be derived from
58
H.W.M. Nmss~'~et al.
granule fusion as indicated by the presence of CD67 and/or CD63 antigens in the vacuolar membrane (Fig. 5), the presence being dependent on the stimulation conditions. The characteristics of the fusion of the azurophilic and specific granules with the vacuolar membrane were identical with that of the granule fusion with the plasma membrane, as measured by the number of gold particles of CD63 and CD67 on the vacuolar and the plasma membrane, respectively. This pattern was also in agreement with the measured up-regulation of CD63 and CD67 antigens as measured by FACS analysis [4]. The occurrence of degranulation towards vacuoles not in contact with the external medium implicate that in electropermeabilized cells the amount of enzymes released into the medium [3] or the extent of upregulation to the plasma membrane of granule membrane markers [4] underestimates the total degranulation. The amount of granule enzyme release into these vacuoles cannot be determined from the electron microscopy experiments presented here. Tentatively, it might amount to 20-30% of the total granule content, because stimulation of electropermeabilized neutrophils with 1 laM Ca 2+ plus 50 lxM GTP-T-S results in only 55 ± 6% release of [3-hexosaminidase (n = 3, mean ± S.E.M.), while about 90% of all granules disappear under these conditions of stimulation. At present it is not clear why in electropermeabilized neutrophils, significant granule fusion takes place without exocytosis. It might be caused by inadequate activation by Ca 2+ and/or GTP-7-S as compared to the receptor agonist FMLP. However, in electropermeabilized neutrophils, even in the presence of 1 OM Ca 2÷, we did not find any up-regulation of granule markers to the plasma membrane upon FMLP addition, nor did we observe the formation of vacuoles under these conditions (results not shown). Alternatively, there could be an inhibitory effect of the electropermeabilization procedure on the machinery required for fusion of the vacuoles with the plasma membrane. Hoffstein et al. [16] have suggested that microtubuli might "line up" granules
prior to fusion and by doing so direct the formation of invaginations. Indeed, Rothwell et al. [ 17] have found that granule-microtubule interactions are stimulated in neutrophils as a consequence of cell activation. Further studies are warranted to elucidate the role of the cytoskeleton in the
degranulation process. Acknowledgements--We thank PROF. D. Roos for critically reading the manuscript and MR L. D. C. VERSCHRAO~for preparing the photographic material.
REFERENCES 1. Malech H. L. and Gallin J. I. (1987) N. Engl. J. Med. 317, 687-694. 2. Bellavite P. (1988) Free tad. Biol. Med. 4, 225-261. 3. Smolen J. E. and Sandborg R. R. (1990) Biochim. biophys. Acta 1052, 133-142. 4. Niessen H . W . M . and Verhoeven A. J. (1992) Cell. Signal. 4, 501-509. 5. Knight D. E. and Scrutton M. S. (1986) Biochem. J. 234, 497-506. 6. Hara Y., Tsukada M., Komiyama A. and Akabane T. (1989) Fur. J. Haematol. 42, 474-480. 7. Verhoeven A. J., Van Schalk M. L. J., Roos D. and Wecning R. S. (1988) B/ood 71, 505-507. 8. Grinstein S. and Furuya W. (1988) J. biol. Chem. 263, 1779-1783. 9. Grynkiewicz G., Poeni M. and Tsien R. Y. (1985) J. biol. Chem. 260, 3440-3450. 10. Kuijpers T. W., Tool A. T. J., Van der Schoot C. E., Ginsel L. A., Onderwater J. J. M., Roos D. and Verhoeven A. J. (1991) Blood78, 1105-1111. 11. Jost C. R., GaiUard M. L., Fransen J. A. M., Daha M. R. and Ginsel L. A. (1991) Blood 78, 3030-3036. 12. Ginsel L. A., Onderwater J. J. M., Fransen J. A. M., Verhoeven A. J. and Roos D. (1990) B/ood 76, 2105-2116. 13. Smolen J. E., Stoehr S. J. and Boxer L. A. (1986) Biochim. biophys. Acta 886, 1-17. 14. Esaguy N., Aguas A. P., Vilanova M. and Silva M. T. (1991) J. Leuk. Biol. 50, 4s.s. s.52. 15. Hoffstein S. and Weissrnann G. (1978) J. Cell Biol. 78, 769-781. 16. Hoffstein S., Goldstein I. M. and Weissmann G. (1977) J. Cell Biol. 73, 242-256. 17. Rothwell S. W., Nath J. and Wright D. G. (1989) J. Cell Biol. 108, 2313-2326.
Pergamon
IMMUNOELECTRON MICROSCOPY REVEALS SIGNIFICANT GRANULE FUSION UPON STIMULATION OF ELECTROPERMEABILIZED HUMAN NEUTROPHILS HANS W. M. NIESSEN,* Jos J. M. ONDERWATER,t HENK K. KOERTEN,tLEO A. GINSEL* and ARTHUR J. VERHOEVEN*§ *Central Laboratory of the Netherlands Red Cross Blood Transfusion Service and Laboratory for Experimental and Clinical Immunology of the University of Amsterdam, Amsterdam, The Netherlands; -~Laboratoryfor Electron Microscopy, University of Leiden, Leiden, The Netherlands; and ~Department of Cell Biology and Histology, University of Nijmegen, The Netherlands (Received 4 July 1993; and accepted 12 August 1993) Abstract--Although electropermeabilization has become an important tool for studying the signal requirements of exocytosis, relatively little is known about the morphological changes accompanying this response in electropermeabilized cells. In this study, we determined that electropermeabilization of human neutrophils by itself caused only minor changes in the morphology as determined by transmission electron microscopy. The structure of the plasma membrane did not show detectable changes, whereas the cytoplasm was more electron lucent as compared to intact cells. Activation of intact neutrophils with forrnyl-methionyl-leucyl-phenylalanine (FMLP), in the presence of cytochalasin B, caused the development of invaginations of the plasma membrane. In contrast, activation of electropermeabilized cells with 1 p.M Ca2+and/or 50 ~tM GTP-T-S caused the development of vacuoles that did not seem to be in contact (or had previously been in contact) with the extracellular environment. However, fusion of azurophilic and specific granules with these vacuoles clearly had taken place. The response characteristics of this fusion induced by Ca2+ and GTP-T-S were quite similar to those of the direct fusion of granules with the plasma membrane. We conclude that in electropermeabilized human neutrophils, two processes involving granule fusion can be distinguished. First, a direct fusion of granules with the plasma membrane. Secondly, the fusion of granules leading to the formation of vacuoles, not in contact with the extracellular space.
INTRODUCTION PHAGOCYTOSIS and killing of bacteria by neutrophils are important mechanisms in the host defence against invading mechanisms. The killing process is mediated by the concomitant release of lysosomal enzymes, bactericidal proteins and toxic oxygen metabolites into the phagosomes formed [ 1, 2]. For the release of proteins, fusion of intracellular granules with the phagosomal membrane is required. Recently, this degranulation process has been §Author to whom correspondenceshould be addressed at: Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, c/o Publication Secretariat, Plesmanlaan 125, 1066 CX Amsterdam,The Netherlands. Abbreviations: FMLP--formyl-methionyl-leucyl-phenylalanine; MPO~myeloporoxidase; FACS--fluorescence-aetivated cell sorter; FITC--fluoresceinisotliiocyanate;DMSO--dimethyl sulphoxide; PFA--paraformaldehyde; IgG-imrnunoglobulinG; PBS--phosphate-bufferedsaline. 47
studied in electropermeabihzed neutrophils by measuring the release of enzymes [3] or by measuring the up-regulation of specific granule membrane markers [4]. The results of these studies suggest that in electropermeabilized neutrophils the machinery of the degranulation process is intact with different requirements for degranulation between azurophilic and specific granules. Most notably, elevation of the free Ca 2÷ concentration alone to values mimicking that in stimulated neutrophils was found to be sufficient for enzyme release from the specific, but not from the azurophilic granules [4]. Knight et al. [5] have shown by electron microscopy that in electropermeabilized chromaffin cells and electropermeabilized platelets, little evidence of plasma membrane damage can be observed. Until now, the effect of the electropermeabilization procedure on the morphology of neutrophils has not been studied.
H.W.M. NmSSENet al.
48
In intact neutrophils, Hara et al. [6] have shown by electron microscopy that activation of neutrophils with the chemotactic tripeptide FMLP (in the presence of cytochalasin B) causes invaginations of the plasma membrane into the cells. The effects of activation of electropermeabilized neutrophils at the morphological level are as yet unknown. We have therefore studied the effect of electropermeabilization and subsequent activation of the neutrophils on the morphology of the cells analysing the changes accompanying degranulation. For this purpose, transmission microscopy and immunoelectron microscopy were used. MATERIALS AND METHODS
Materials FMLP and cytochalasinB (Sigma Chemical Co., St. Louis, MO, U.S.A.) were dissolved in DMSO at 1000 times the final concentration for cell incubations, and were stored at -20"C. GTPy-S was obtained from Boehringer (ingelbeim, Germany), dissolved in H20) at 100 times the final concentration for cell incubations, and was stored at -20"C. CD63 and CD67 monoelonal antibodies and FITC-labelled goat antibodies against mouse IgG were obtained from our own institute (Central Laboratory of the Blood Transfusion Service, Amsterdam, The Netherlands). All other chemicals were reagent grade. The basal incubation medium for cell suspensions consisted of 132 mM NaCI, 6 mM KC1, 1 mM CaC12, 1 mM MgSO4, 1.2 mM potassium phosphate, 20 mM Hepes, 5.5 mM glucose and 0.5% (w/v) human albumin, pH 7.4. "Electropermeabilization medium" consisted of 140 mM KCI, 1 mM MgC12, 1 mM EGTA, 20 mM Hepes and 10 mM glucose (pH 7.2).
Isolation of neutrophils Blood was obtained from healthy volunteers. Granulocytes were purified from the buffy coat of 500 ml blood anticoagulated with 0.4% (w/v) trisodium citrate (pH 7.4), as described by Verhoeven et al. [7]. After isolation, the cells were suspended in incubation medium and kept at room temperature.
Electropermeabilization and subsequent activation of neutrophils Cells were permeab'dized immeAiately before use, as the method described by Grinstein and Furuya [8]. In short, a sample of 107 cells was washed once and resuspended in 1 ml of ice-cold electropermeabilization medium. Part (0.8 nil) of this suspension was transferred to a BioRad Pulser cuvette and subjected to two discharges of 5 kV/cm2 from a 25 pFd capacitator, in the BioRad GenePulser. Between the two pulses, the cells were stirred gently with a plastic pipette. Several batches of neutrophils were electropermeabilized consecutively, pooled and stored at 4"C. Four minutes after the first batch of cells had been permeabilized, the permeabilized cells were transferred from 4oc to 37"C. Two minutes later, 1 mM ATP and cytochalasin B (5 I.tg/ml) were added. After 1 rain of further incubation, the cells were stimulated with Ca2+ and/or GTP-y-S. The free Ca2÷ concentrations in the incubations were checked by Indo-1 measurements after adding 0.1 ~M Indo-1 (free acid) to the medium [9]. At the time points indicated, 1% (v/v) PFA was added to stop the degranulation process [4].
Pre-embedding labelling of neutrophils Electropermeabilized neutrophils were activated with 1 ~ Ca2+and/or 50 pM GTP-T-S and subsequently fixed at different time points. Intact neutrophils were activated with FMLP (in the presence of cytochalasin B) for 1 min and subsequently fixed with 1% PFA. The cells were then incubated in a mixture of 1% PFA and 0.1% glutaraldehyde for 2 h at room temperature. Subsequently, the cells were washed with PBS and incubated at room t e ~ for 60 rain with CD63 or with CD67 antibodies. These antibodies can be used to assess independently the degranulation of azurophilic and specific granules, respectively [4, 10, 11]. The cells were then washed again with PBS, and incubate! first with rabbit anti-mouse IgG and then with protein A conjugated to gold with a diameter of about 10 nm. As a negative control, the primary antibody incubation step was omitted. Next, the pellet was
Degranulationin electropermeabilizedneutrophils post-fixed for 10 rain at room temperature in 1.5% glutaraldehyde in 0.14 M caeodylate buffer (pH 7.4) and then for 15 rain at 4"C in OsO4. After pelleting of OsO+-fixed cells in 2% Bacto-agar and subsequent dehydration in a graded series of ethanol dilutions, the cells were embedded in epon. Ultra-thin sections stained with lead hydroxide were examined in a Philips E M 201 electron microscope at 80 kV.
49
cytochalasin B). The cells were then fixed in a mixture of 1% PFA and 0.1% glutaraldehyde for 2 h at room temperature, and prepared for cryosectioning as described previously [12]. After cryosectioning, streptavidin coupled to gold (5 tam), was added. As a negative control, biotin-Nbydroxy-suceinimide ester was omitted from the pre-incubation step. The sections were examined with a Philips EM 201 at 80 kV.
Immunocytochemistry of CD63 and CD67 Activation of electropermeabilized neutrophils was stopped at different time points with 1% PFA. The cells were then fLxed a mixture of 1% PFA and 0.1% glutaraldehyde for 2 h at room temperature, and prepared for cryosectioning and immunolabelling as described previously [12]. In short, immunogold double-labelling of CD63 and MPO, or CD67 and lactoferrin was performed by incubation of the cryosections with MoAb CD63 or CD67, followed by a second antibody (rabbit anti-mouse IgG), and then protein A-conjugated colloidal gold particles with a diameter of about 10 nm. Thereafter, the cryosections were incubated with a polyclonal rabbit antibody against myeloperoxidase or lactoferrin, respectively, and protein A-coupled to gold (5 urn). As controls, the primary antibody incubation step was replaced by incubation with non-relevant mouse and rabbit IgG. The sections were examined with a Philips EM 201 at 80 kV.
lmmunocytochemistry of biotinylated cells Intact neutrophils were incubated with biotinN-hydroxy-succinimide ester (100 Ixg/ml) for 1 h at room temperature. In this way, plasma membrane proteins were efficiently biotinylated as indicated by FACS analysis after incubation of the cells with streptavidin-FITC (data not shown). Subsequently, the cells were electropermeabilized and activated with 1 lxrn Ca2÷plus 50 p_M GTP-TS for 2 min. The plasma membrane of these dec_ tropermeabilized cells still contained biotinylated proteins as established by FACS analysis. For comparison, intact cells labelled with biotin were stimulated with FMLP (in the presence of
RESULTS First, the effect of the electropermeabllization procedure on the morphology of neutrophils was evaluated. Unstimulated intact neutrophils embedtied in epon (Fig. 1A) were compared with unstimulated electropermeabilized neutrophils (Fig. 1B). The cytoplasm of electropermeabilized neutrophils was more electron-lucent as compared to the intact cells, while sometimes the permeabilized cells showed less dense nuclear material. Despite the eleclropermeabilization procedure, the granules and the plasma membrane of the permeabflized neutrophils did not show any difference with the intact neutrophils, even at a high magnification (45,000 x) (Fig. 1C). However, we established that the electropermeabilized neutrophils were permeable for propidium iodide as measured by flow cytometry [8]. Next, the morphological effects of stimulation of the electropermeabilized neutrophils with 1 ~tM Ca2+ and/or 50 lxM GTP-T-S were studied. After activation, the number of granules clearly diminished. Stimulation with 1 BM Ca2÷ plus 50 GTP-T-S diminished the total amount of granules with 89% (20 cells) as compared m unstimulated electropermeabilized neutrophils (incubated with a [Cae+]f~ of about 100 nM). Stimulation with 1 IxM Ca2+ diminished the total amount of granules with 59% (20 cells). Furthermore, after stimulation with 1 BM Ca2+ plus 50 pM GTP-T-S, large vacuoles were formed (Fig. 2A), which were not visible in unstimulated electropermeabilized neutrophils (Fig. 1B). These vacuoles were also found after stimulation with 1 BM Ca2+ (Fig. 2B) or with 50 BM GTP-y-S (not shown). There was no significant difference in number or in the
50
H.W.M. Nmss~'~et aL
volume of the vacuoles formed under the different activation conditions studied (results not shown). To assess whether the vacuoles appearing in electropermeabilized neutrophils after stimulation were in fact large invaginations of the plasma membrane, the permeabilized cells were stimulated and subsequently incubated with CD63 or CD67 monoclonal antibodies, followed by an incubation with rabbit anti-mouse IgG and protein A-conjugated colloidal gold particles (10 nm) prior to embedding in epon. In this protocol, only proteins on the plasma membrane in contact with the extracellular space will become labelled. Under all conditions of stimulation, no gold particles were detected near the membranes of the vacuoles (Fig. 2). Staining of the vacuolar membrane was also absent in samples taken as early as 30 sec after stimulation. In contrast, labelling was clearly visible at the plasma membrane (Fig. 2B). When the same procedure of in-suspension labelling was applied to intact neutrophils stimulated with the chemotactic tripeptide FMLP (in the presence of cytochalasin B), the vacuoles appearing under these conditions of stimulation contained significant numbers of gold particles bound to CD63 antibodies (Fig. 3). Positive staining of these vacuoles was also obtained with CD67 as primary antibody (results not shown). These results support the conclusion of Hara e t al. [6], that the vacuoles appearing in intact neutrophils under these conditions of stimulation are in fact invaginations of the plasma membrane, in which degranulation of both azurophilic granules (as detected by CD63 antibodies) and specific granules (as detected by CD67 antibodies) had taken place. When, in elcctropermeabilized neutrophils under the various conditions of stimulation, the number of gold particles detecting CD63 or CD67 on the plasma membrane was counted, it was established that the pattern of up-regulation was in good agreement with our earlier results obtained by FACS analysis [4]. Most notably, activation of the cells with 1 IIM Ca 2+ alone did not cause an increase in CD63 labelling at the plasma membrane (Fig. 4). However, there
was a clear increase in the amount of labelling with CD63 after addition of 50 p.M GTP-~-S + 1 lxM Ca ~'. In contrast, a clear increase in the amount of gold particles detecting CD67 was detected on the plasma membrane after activation with 1 ktM Ca 2+as the sole stimulus, indicating the fusion of specific granules with the plasma membrane. The absence of CD63 and CD67 labelling in the vacuoles formed upon activation might be due to the absence of fusion of granules with these structures. We therefore prepared frozen sections of electropermeabilized neutrophils after stimulating the cells with 1 pNI Ca 2÷ and/or 50 JIM GTP)'-S. Subsequently, the MAbs CD63 or CD67 were added to the frozen sections, followed by the addition of rabbit anti-mouse IgG and protein A gold (10 nm). In addition, anti-MPO or anti-lactoferrin, visualized by addition of 5 nm protein A gold, were added to the frozen sections labelled with
300 --
T
¢j
"~
200
q~
100
o CD63
CD67
FIG. 4. Amount of protein colloidal gold particles on the plasma membrane of electropermeabilized cells activated with Ca 2+ and/or GTP-T-S. Electropermeabifized neutrophils (107/ml) were preincubated at 37"C for 3 rain in the presence of cytochalasin B (5 ~tg/ml) and subsequently activated with 1 ~tM Ca2+ and/or 50 p.M GTP-y-S for 8 rain and fixed with 1% PFA. Then the cells were labelled with CD63 and CD67. Subsequently, the cells were prepared for epon (MAbs in suspension) as described in Materials and Methods. The results are presented as the mean (+ S.E.M.) of 20 cells. ( I ) Unsfimulated cells; ([~) 1 pM Ca2+; ([~) 1 ~tM Ca2÷plus 50 ~ GTP-T-S.
FIG. 1. Morphology of intact and electropermeabilized neutrophils. Intact and electropermeabilized neutrophils (lO’/ml) were incubated at 37°C in the presence of cytochalasin B (5 pg/ml) for 4 min and then fixed with 1% PFA. The cells were then prepared for epon as described in Materials and Methods. (A) Intact neutrophils; bar: 0.5 pm. (B) Electropermeabilized neutrophils; bar: 0.5 Frn.
51
FIG. 1. Morphology of intact and ¢lectropermeabilize,d neutrophils. Intact and electropermeabilized neutrophils (107/ml) were incubated at 37"C in the presence of cytochalasin B (5 gg/ml) for4 min and then fixed with 1% PFA. The cells were then prepared for epon as described in Materials and Methods. (C) Electropermeabilized neutrophils; bar: 0.1 gm.
52
FIG. 2. Morphology and in-suspension labelling of activated electropermeabilized neutrophils. Electropermeabilized neutrophils (107/ml) were preincubated for 3 min at 37°C in the presence of cytochalasin B (5 ~tg/ml) and subsequently stimulated for 8 min with (A) 1 I.tM Ca 2÷ plus 50 [tM GTP-y-S or with (B) 1 paM Ca 2÷, and subsequently fixed with 1% PFA. The cells were then prepared for epon as described in Material and Methods. Gold particles in (B) represent CD67 antibodies, which had been added prior to embedding in epon. Bar: 0.5 lxm. 53
FIG. 3. CD63 Labelling in suspension of activated neutrophils. Intact neutrophils (10711111)were preincubated for 3 min at 37°C in the presence of cytoehalasin B (5 ~tg/ml) and subsequently activated with FMLP for 1 min and fixed with 1% PFA. After labelling with CD63, the cells were prepared for epon (MAbs in suspension) as described in Materials and Methods. Bar: 0.1 I.tm; arrow: plasma membrane.
~A
FIG. 5. Frozen sections of activated electropermeabilized neutrophils. Electropermeabilized neutrophils (107/ml) were preincubated at 370C for 3 min in the presence of cytochalasin B (5 ~tg/ml) and subsequently activated for 8 min with (A) 1 l.tM Ca 2÷ plus 50 ktM GTP-T-S or (B) 1 IxM Ca 2÷. Thereafter, the cells were fixed with 1% PFA. The cells were then treated for frozen sectioning as described in Materials and Methods. The cryosections were incubated with MoAb CD63 (detected by 10 nm gold particles) and anti-MPO antiserum (detected by 5 nm gold particles). Bar: 0.1 lam. 55
FIG. 6. Frozen sections of intact and electropermeabilized cells labelled with biotin. Intact neutrophils (107/ml) were preincubated for 1 h with biotin-N-hydroxy-succinimide ester (100 pg/ml) at room temperature. After washing, one part of the cells was activated for 1 min with FMLP (in the presence of cytochalasin B), and another part was electropermeabilized and subsequently stimulated with Ca*+plus GTP-YS for 2 min, in the presence of cytochalasin B. After cryosectioning, streptavidin A coupled to gold (5 nm) was added. (A) Intact neutrophils; bar 0.1 pm. (B) Electropetmeabilized neutrophils, bar: 0.1 urn.
56
Degranulation in electropermeabilized neutrophils
MAb CD63 or CD67, respectively. With this experimental procedure, there was a clear labelling of CD63 at the membranes of the vacuoles, accompanied by gold particles detecting MPO (Fig. 5A). To achieve positive staining of the vacuolar membranes with CD63 antibodies, the permeabilized cells had to be stimulated with 1 laM Ca 2÷ plus 50 pNl GTP-T-S (Fig. 5A) or with 50 ttlVl GTP-T-S (results not shown). Almost no gold labelling of CD63 and/or MPO was detected in the vacuoles after stimulation with 1 BM Ca2÷as the sole stimulus (Fig. 5B). Under the latter condition of stimulation, there was, however, gold labelling of CD67 at the membranes of the vacuoles, accompanied by gold particles detecting lactoferrin (results not shown), which was comparable with cells activated with 1 pM Ca 2+ plus 50 JaM GTP-T-S. Hence, the labelling pattern of CD63 and CD67 on the membranes of the vacuoles induced after stimulation with Ca 2÷ and/or GTP-T-S, was similar to that found on the plasma membrane presented above (Fig. 4). The binding of the monoclonal antibodies to the vacuoles in the frozen sections of the activated cells and the absence of binding to the vacuoles when the cells were stained with antibodies in suspension prior to embedding, would suggest that these vacuoles are not in contact with the extracellular environment. To investigate whether there had been contact with the extracellular space, we determined the presence of plasma membrane proteins in the vacuolar membrane. For this purpose, intact cells were biotinylated prior to the elcctropcrmcabilization and subsequently activated for 2 rain with I B M Ca 2+ plus 50 IxM GTP7-S. As a positive control, intact cells labelled with biotin were stimulated with F M L P (in the presence of cytochalasin B). Subsequent to cryosectioning, streptavidin coupled to gold was added. Indeed, the vacuoles observed after activation of intact cells contained plasma membrane proteins (Fig. 6A). However, the vacuoles of the clectropermeabilized neutrophils were not labelled (Fig. 6B), supporting the hypothesis that these vacuoles were formed intracellularlywithout contacting the plasma membrane.
57
DISCUSSION In this study we have investigated the degranulation process in electropermeabilized human neutrophils using transmission (immuno-)electron microscopy. Similar to the results .found by Knight and Scrutton [5] in electropcrrneabilized chromaffin cells and electropcrmeabilized platelates, we observed that the plasma membrane of the electropermcabilized neutrophils did not show any visible changes. However, the cytoplasm in the pcrmeabilizcd neutrophils was more electron lucent than in intact cells.This was also found by Smolcn et al. [13] in neutrophils pcrmeabilized with digitonin. Furthermore, sometimes the nucleus appeared less electron dense in the permeabilized neutrophils. However, differences in morphological appearance were much more apparent after activation. In intact cells treated with cytochalasin B, FMLP caused the development of invaginations of the plasma membrane intruding into the cells concurrently with degranulation. These invaginations have been reported previously by Hara et al. [6]. These authors used Ruthenium Red staining to prove contact between the vacuolar content and the extracellular medium. Esaguy et al. [14] suggested that in intact neutrophils activated with PMA vacuoles are formed, but these authors did not distinguish between vacuoles formed intracellularly or invaginations of the plasma membrane. Similarly, Hoffstein et al. [15] suggested that intact neutrophils activated with A23187 show invaginations, but also these authors did not prove whether these were true invaginations or intracellular vacuoles. In electropermeabilized neutrophils treated with cytochalasin B, we found that after activation with Ca 2+ and/or GTP-T-S, the cells also developed vacuoles. In these cells however, antibodies against relevant membrane markers were not able to bind to the vacuolar membrane after addition to the cells in suspension, indicating that these vacuoles were not in contact with the extracellular medium. Furthermore, the vacuoles seemed not to be derived from an endocytotic process (Fig. 6B). Instead, the vacuoles seemed to be derived from
58
H.W.M. Nmss~'~et al.
granule fusion as indicated by the presence of CD67 and/or CD63 antigens in the vacuolar membrane (Fig. 5), the presence being dependent on the stimulation conditions. The characteristics of the fusion of the azurophilic and specific granules with the vacuolar membrane were identical with that of the granule fusion with the plasma membrane, as measured by the number of gold particles of CD63 and CD67 on the vacuolar and the plasma membrane, respectively. This pattern was also in agreement with the measured up-regulation of CD63 and CD67 antigens as measured by FACS analysis [4]. The occurrence of degranulation towards vacuoles not in contact with the external medium implicate that in electropermeabilized cells the amount of enzymes released into the medium [3] or the extent of upregulation to the plasma membrane of granule membrane markers [4] underestimates the total degranulation. The amount of granule enzyme release into these vacuoles cannot be determined from the electron microscopy experiments presented here. Tentatively, it might amount to 20-30% of the total granule content, because stimulation of electropermeabilized neutrophils with 1 laM Ca 2+ plus 50 lxM GTP-T-S results in only 55 ± 6% release of [3-hexosaminidase (n = 3, mean ± S.E.M.), while about 90% of all granules disappear under these conditions of stimulation. At present it is not clear why in electropermeabilized neutrophils, significant granule fusion takes place without exocytosis. It might be caused by inadequate activation by Ca 2+ and/or GTP-7-S as compared to the receptor agonist FMLP. However, in electropermeabilized neutrophils, even in the presence of 1 OM Ca 2÷, we did not find any up-regulation of granule markers to the plasma membrane upon FMLP addition, nor did we observe the formation of vacuoles under these conditions (results not shown). Alternatively, there could be an inhibitory effect of the electropermeabilization procedure on the machinery required for fusion of the vacuoles with the plasma membrane. Hoffstein et al. [16] have suggested that microtubuli might "line up" granules
prior to fusion and by doing so direct the formation of invaginations. Indeed, Rothwell et al. [ 17] have found that granule-microtubule interactions are stimulated in neutrophils as a consequence of cell activation. Further studies are warranted to elucidate the role of the cytoskeleton in the
degranulation process. Acknowledgements--We thank PROF. D. Roos for critically reading the manuscript and MR L. D. C. VERSCHRAO~for preparing the photographic material.
REFERENCES 1. Malech H. L. and Gallin J. I. (1987) N. Engl. J. Med. 317, 687-694. 2. Bellavite P. (1988) Free tad. Biol. Med. 4, 225-261. 3. Smolen J. E. and Sandborg R. R. (1990) Biochim. biophys. Acta 1052, 133-142. 4. Niessen H . W . M . and Verhoeven A. J. (1992) Cell. Signal. 4, 501-509. 5. Knight D. E. and Scrutton M. S. (1986) Biochem. J. 234, 497-506. 6. Hara Y., Tsukada M., Komiyama A. and Akabane T. (1989) Fur. J. Haematol. 42, 474-480. 7. Verhoeven A. J., Van Schalk M. L. J., Roos D. and Wecning R. S. (1988) B/ood 71, 505-507. 8. Grinstein S. and Furuya W. (1988) J. biol. Chem. 263, 1779-1783. 9. Grynkiewicz G., Poeni M. and Tsien R. Y. (1985) J. biol. Chem. 260, 3440-3450. 10. Kuijpers T. W., Tool A. T. J., Van der Schoot C. E., Ginsel L. A., Onderwater J. J. M., Roos D. and Verhoeven A. J. (1991) Blood78, 1105-1111. 11. Jost C. R., GaiUard M. L., Fransen J. A. M., Daha M. R. and Ginsel L. A. (1991) Blood 78, 3030-3036. 12. Ginsel L. A., Onderwater J. J. M., Fransen J. A. M., Verhoeven A. J. and Roos D. (1990) B/ood 76, 2105-2116. 13. Smolen J. E., Stoehr S. J. and Boxer L. A. (1986) Biochim. biophys. Acta 886, 1-17. 14. Esaguy N., Aguas A. P., Vilanova M. and Silva M. T. (1991) J. Leuk. Biol. 50, 4s.s. s.52. 15. Hoffstein S. and Weissrnann G. (1978) J. Cell Biol. 78, 769-781. 16. Hoffstein S., Goldstein I. M. and Weissmann G. (1977) J. Cell Biol. 73, 242-256. 17. Rothwell S. W., Nath J. and Wright D. G. (1989) J. Cell Biol. 108, 2313-2326.