* The Granulocyte Research Laboratory, Department of Hematology, Rigshospitalet, Copenhagen, Denmark;
School of Applied and Engineering Physics, Cornell University, Ithaca, New York; and
Division of Cell Biology, The Netherlands Cancer Institute, Amsterdam
Correspondence: Karsten Lollike, M.D., Ph.D., Novo Nordisk A/S, Krogshøjvej 53A, 2880 Bagsværd, Denmark. E-mail: kalo{at}novonordisk.com
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Key Words: signal transduction • patch-clamp • capacitance
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In principle, and as seen in other cell types [4 , 5 ], release of multiple granules can be envisioned to proceed through three different mechanisms (Fig. 1 ) or a combination of these. Granules fuse independently of each other with the plasma membrane (Fig. 1A) . We call this simple exocytosis. In Figure 1B , granules fuse first with other granules in the cytosol prior to their subsequent fusion with the plasma membrane. This is known as compound exocytosis [4 ]. In Figure 1C , one granule fuses initially with the plasma membrane, followed by fusion of a second granule to the membrane of the granule already engaged in fusion, thus forming a degranulation sac. This is termed cumulative fusion [4 ].
 View larger version (32K): [in a new window] |
Figure 1. Possible mechanisms for exocytosis of individual granules. (A) Two granules fuse independently of each other with the plasma membrane, giving rise to two distinct capacitance steps. We call this simple exocytosis. (B) Two granules fuse with each other in the cytosol prior to fusion with the plasma membrane. This will result in the appearance of a single step with the size equivalent to the area of the combined granules. This is termed compound exocytosis. (C) One granule fuses with the plasma membrane. Subsequently, a second granule fuses with the membrane of the first granule, thus forming a degranulation sac. In capacitance recordings, this will give rise to two distinct capacitance steps. This is termed cumulative fusion.
|
Neutrophils are formed in the bone marrow from myeloid precursors and are closely related to eosinophilic granulocytes. Electron micrographs have suggested that rabbit neutrophils stimulated by formyl-Met-Leu-Phe (fMLP) in combination with cytochalasin B secrete predominantly by cumulative fusion [7 ]. However, electron microscopy (EM) gives only a static picture and cannot distinguish cumulative and compound exocytosis.
We have used the high-resolution cell-attached patch-clamp capacitance technique [8 9 10 11 ] to measure the size of single capacitance steps corresponding to exocytosis of individual granules in neutrophils. As cell capacitance is proportional to plasma membrane area, changes in plasma membrane area associated with exocytosis of single granules or compounds can be well resolved with high temporal resolution with this technique. Compound exocytosis manifests itself in such experiments by capacitance steps significantly larger than those associated with exocytosis of individual granules. The method also allows for resolution of the initial fusion pore, the first aqueous connection between the vesicle lumen and the extracellular space forming during exocytic fusion. We have also captured morphological manifestations of transient exocytic events by EM, corresponding to the functional results obtained by capacitance measurements.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Cell-attached, patch-clamp capacitance measurements
Isolated human neutrophils were suspended in ES buffer (140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM glucose, 10 mM HEPES, pH 7.2–7.3) at 4°C and were used for patching within 3 h.
Patch-clamp capacitance measurements were performed in the cell-attached configuration as described previously [8
, 9
]. The correct phase setting was inspected in the recorded traces. If capacitance steps in the imaginary part (Im) were accompanied consistently by a projection into the real part (Re) or if conductance changes were accompanied by a projection into the Im trace, then the phase was rotated until the phase setting was correct. This phase shift was usually within ±5°. When the fusion pore conductance is small, the signal in the Im trace is reduced compared with the size that is obtained after full exocytosis. Simultaneously, a signal appears in the Re trace as long as the fusion pore conductance Gp is comparable with 
CV, with CV being the capacitance of the exocytosed vesicle and being 2
f (f=sine-wave frequency). For convenience of notation, we use Re and Im for the changes in the real and imaginary trace after baseline subtraction. These changes are given by Re = [(CV)2/Gp]/[1+(CV/Gp)2], and Im = CV/[1+(CV/Gp)2] and thus can be used to calculate the true values of CVand Gp [13
]: CV = [(Re2+Im2)/Im]/ (1); Gp = (Re2+Im2)/Re (2). To avoid division by zero, CV was set to zero for Im = 0. Gp was set to zero for Re = 0 and Im < CV/2. Determination of fusion pore conductance was discontinued when Re 
0 and Im > CV/2. We used a specific capacitance of 8 fF/µm [2
, 14
] to convert capacitance to diameter (d) according to: d = [step size (fF) · µm2/8 fF · ]1/2 (3). Fusion pore conductance (Gp) was converted to approximate pore radius (r) according to [15
, 16
]: r = {[100 
cm · 15 nm · Gp (pS)]/}1/2 (4). Exocytosis was stimulated by ejecting ionomycin (200 µM; Sigma Chemical Co., St. Louis, MO) from a pipette positioned approximately 100 µm from the cell. Only a small volume of ionomycin was ejected into the cell chamber containing 200 µl ES, and the final ionomycin concentration was consequently much less than 200 µM. This stimulation was adjusted to achieve [Ca2+]i, equivalent to that obtained when stimulating with 2 µM ionomycin (final concentration) [8
]. Thus, the final concentration of ionomycin will be approximately 2 µM but will vary from experiment to experiment because of variance in stimulation pipette resistance and distance between stimulation pipette and patch-clamped cell. All recordings were performed at room temperature with a green filter in the light patch to diminish damage by light.
EM
Human neutrophils (isolated as above) were incubated for 1, 5, or 10 min with 5 µM ionomycin or for 10 min with 1 µg/ml phorbol 12-myristate 13-acetate (PMA) at room temperature and then were fixed with 2.5% glutaraldehyde (v/v) in 0.1 M cacodylate buffer (pH 7.2) for 1 h and post-fixed in 1% (w/v) osmium tetroxide in the same buffer for 1 h, block-stained with uranyl acetate, dehydrated, and embedded in LX-112. Thin sections were stained with uranyl acetate and lead citrate and were examined with a CM10 transmission electron microscope (Philips, Eindhoven, The Netherlands).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
 View larger version (10K): [in a new window] |
Figure 2. Recording from human neutrophil stimulated with ionomycin to induce exocytosis visualized as single capacitance steps. Recording from a human neutrophil showing the capacitance (Im) and conductance (Re) trace. Following a 20 s ionomycin pulse, the capacitance trace starts to increase in discrete steps of varying amplitude, indicating exocytosis of single granules. Conversely, the conductance trance remains steady, except for transient changes associated with the beginning of individual capacitance steps, indicating the opening of narrow fusion pores.
|
 View larger version (12K): [in a new window] |
Figure 3. Capacitance traces showing compound and normal exocytosis. Capacitance traces with the two outputs from the lock-in amplifier (Im and Re corresponding to capacitance and conductance, respectively). (A) A large capacitance step (19.7 fF) is seen followed by four steps of ordinary size. Only the large first step and the last step have associated conductance changes. (B) A large capacitance step is seen in the middle of eight steps of ordinary size. This recording has a lower noise level than in (A), and all steps have associated conductance changes.
|
 View larger version (13K): [in a new window] |
Figure 4. Analysis of fusion pores from capacitance steps. (A and B) Detailed analysis of the two large capacitance steps from Figure 3A
and 3B
, respectively. Top panels show the capacitance (Im) and conductance (Re) traces as recorded. (A) The arrowhead indicates the last time point in which Re still deviates from the base line (dashed line), and this is the last time point for which we can calculate a reliable fusion pore conductance (two lower traces). CV shows calculated vesicle capacitance. This calculation reveals a single step in capacitance, confirming that these events reflect exocytosis of single, large granules. The two lower traces show calculation of the fusion pore conductance (Gp) but on different scales so that the initial (upper) and final (lower) fusion pore conductance can be recognized. (C and D) The fusion pore conductance (Gp) of the two last steps from Figure 3B
on two different scales.
|
Isolated human neutrophilic granulocytes are usually contaminated slightly by eosinophils, normally less than 5%. Human eosinophilic granulocytes do contain large granules and potentially might be responsible for the large steps we record. We recorded compound steps in 11 out of 35 recordings, which corresponds to 31% of the recordings, a much higher fraction than what would be expected by random recording of eosinophils from a pool of mixed cells. Furthermore, the large granules of eosinophils can be seen by differential interference contrast (DIC) microscopy, and such cells were avoided in our recordings. Finally, all compound steps were surrounded by capacitance steps of ordinary size as expected for human neutrophils, which would not be expected for human eosinophils that have predominantly much larger steps [4 , 19 ]. For these reasons, we rule out the possibility that our compound exocytic capacitance steps might originate from contaminating eosinophils. Could the compound steps be the result of fusion of other organelles than granules with the membrane? Mature, human neutrophils contain very small amounts of organelles other than granules (see Fig. 6 ), and because the fusion pores of the compound steps behaved in a manner similar to fusion pores in single granules, it appears unlikely that the compound steps should not result from exocytosis of granules. Recently, cytoplasmic granule-granule fusion giving rise to compound exocytosis has been implicated in cells ordinarily not associated with exocytosis of very large granules [20 , 21 ]. Based on these arguments, we conclude that the large steps we record are the result of compound exocytosis of multiple granules in normal human neutrophils.
 View larger version (112K): [in a new window] |
Figure 6. Electron micrograph showing compound exocytosis. Human neutrophils were incubated for 1 min with 1 µM ionomycin (A) or for 10 min with 1 µg/ml PMA (B) at room temperature and then were fixed and processed for EM. (A) The thin section of neutrophil shows granules scattered over the cell and a large granule (marked area) that at higher magnification (inset) can be seen, which is composed of two granules (arrows) fused to one another. The granule complex is associated closely with the plasma membrane (arrowhead). n, Nucleus. Original bar = 200 nm. (B) The thin section of neutrophil shows granules scattered all over the cell and some contacting the plasma membrane (arrows). Inset, Higher magnification of the marked area showing two granules (arrows) fused to one another. The granule complex is closely associated with the plasma membrane (arrowhead). Original bar = 500 nm; inset, 200 nm.
|
0.9 fF), for example, will result in a combined capacitance of less than 5 fF (0.9 fF+0.9 fF=1.8fF). In a total of 1321 exocytic steps from 35 recordings of ionomycin-stimulated cells, we recorded 20 steps >5 fF, ranging from 5 fF to 100.6 fF with a mean step size of 20.5 fF ± 5.2fF (±SE). When the mean size of compound steps is 20 fF, and the largest granules have a capacitance of 4 fF, we can calculate that the mean number of granules contributing to compound steps must be at least 5. Because this calculation holds true only if the largest granules participate exclusively in compound formation, the actual number of granules contributing to compound granules is probably higher. Although the number of compound events constituted only 1.5% of the total exocytic events, the capacitance contribution resulting from the compound events was 22% of the total capacitance increase as a result of all steps. This indicates that at least 22% of granule contents is secreted by compound exocytosis. When timed from the start of each electrophysical recording, the average time point for occurrence of an exocytic capacitance step was 329 s, whereas the average time point for a compound exocytic step was 513 s. That the large capacitance steps are "delayed" as compared with single capacitance steps indicates that they will have had time to undergo granule-granule fusion prior to their fusion with the plasma membrane. Resting human neutrophils have, by EM, never been found to have large granules equivalent to our compound steps [22 , 23 ]. These two pieces of information would be consistent with the idea that compound formation is induced by stimulation.
Cumulative fusion
It can be very difficult to distinguish between simple exocytosis and cumulative exocytosis by capacitance measurements, as both forms of exocytosis will result in stepwise increases in capacitance. In horse eosinophils, large degranulation sacs are formed that can be detected in electrophysiological measurements as a result of the slow time constant for charging the large capacitance of the degranulation through a narrow fusion pore [4
]. The complex changes in the electrical equivalent circuit cannot be tracked with single frequency sine wave stimulation. It requires multifrequency or pulse stimulation, which allows for determination of the time course of many equivalent circuit parameters. In neutrophil patches, degranulation sacs will still be rather small, such that the charging time constant associated with a fusion of more than 10 nS conductance [4
] is too fast to be resolved.
Sometimes the exocytic process can be reversible, such that an open fusion pore closes again and the granule is retained within the cell. When looking at individual granule exocytosis by capacitance measurements, this is a well-described phenomenon known as flickering [8 , 16 , 24 ]. In Figure 5 , one rare event is shown where a granule not only flickers (the fusion pore opens at 1 and closes at 5), but a second granule fuses with the first granule (at 2), and when the fusion pore closes at 5, the capacitance contribution from the second granule also disappears, indicating that the second granule must have fused with the membrane of the first granule. Additionally, yet another granule flickers (from 3 to 4) in this trace, but we cannot determine whether this granule also fuses with the first granule or somewhere else in the patch. This event shows that cumulative fusion can occur in human neutrophils. This was the only such event we observed in 37 recordings from different cells. However, flickers are generally very rare in neutrophils [9 ], and so we cannot determine how frequently exocytosis occurs by cumulative fusion in neutrophils.
 View larger version (18K): [in a new window] |
Figure 5. Recording showing reversible and cumulative fusion. The capacitance trace (Im) shows three upward capacitance steps and two downward steps. At 3, a vesicle opens up its fusion pore and closes it again at 4. This is an example of a reversible fusion event, also known as a capacitance flicker. The steps at 3 and 4 have the same size. At 1, a vesicle opens up its fusion pore (as can be seen by the associated conductance changes in Re), and at 5, the pore closes again. Meanwhile, another vesicle has fused with the membrane of the first vesicle, because when the first vesicle closes its fusion pore at 5, the capacitance contribution from the second vesicles also disappears. The fusion pore analysis is shown in Gp.
|
As PMA is known to induce delayed exocytosis compared with ionomycin, we tried to stimulate human neutrophils with PMA for 10 min and looked for compound events. Figure 6B shows a human neutrophil stimulated with PMA for 10 min to undergo exocytosis, and several granules can be seen touching the plasma membrane (arrows), indicating an early stage of fusion. All these are single fusion events, without any indication of degranulation sacs or compound exocytosis. Figure 6B (inset) illustrates two granules that have actually fused with each other. The granules are very close to the plasma membrane (docked to the membrane) and thus are expected to undergo exocytosis. Alternatively, they might already be connected to the plasma membrane in another plane out of focus, and thus it could be the result of compound as well as cumulative fusion. In this case, however, we would have expected release of granular contents that would be associated with loss of the electron dense material inside the degranulation sac.
Additionally, we have looked for granule-granule fusion in connection with phagosomes in human neutrophils stimulated with immunoglobulin G-coated latex beads (similar to experiments presented in Sørensen et al. [25 ]). Although many single granules could be seen fusing with the phagosome, we never could observe granule-granule fusion (unpublished results). However, as concluded in this manuscript, EM is not an optimal method for detection of granule-granule fusion, and thus we cannot rule out that granule-granule fusion does occur in connection with phagosomes.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Molecular mechanism of complex exocytosis
Eosinophils, the close cousin of neutrophils, have been known for many years to be able to secrete by compound exocytosis. Recently, a growing number of other cells, previously not associated with compound exocytosis, have been demonstrated to have the capacity to undergo compound exocytosis [20
, 21
, 26
]. The universally distributed fusiogenic snap receptors (SNARE) proteins [27
] are also thought to be essential in controlling exocytosis in human neutrophils [28
]. v-SNAREs are present on cargo vesicles (granules). They specifically interact with cognate t-SNAREs residing on the plasma membrane. In theory, this should ensure a specific fusion process. However, in some cells t-SNAREs are also present on vesicles [29
], giving rise to the possibility that vesicles might fuse with each other. SNARE proteins are also present in human neutrophils, and the v-SNARE, VAMP-2 seems to be restricted to secretory vesicles and granules, the v-SNARE SNAP-23 is present on gelatinase and specific granules, and the t-SNAREs syntaxin 4 and syntaxin 6 are present on the plasma membrane [28
, 30
, 31
]. SNAP proteins are usually considered to be t-SNAREs, but SNAP-23 proteins have been suggested to function as a v-SNARE in human neutrophils and were demonstrated to be able to interact with syntaxin 6 [28
]. If t-SNAREs are present occasionally on granules, a scenario could be envisioned where granules accumulated close to the plasma membrane might undergo fusion with each other prior to their fusion with the plasma membrane. Recently, different isoforms of syntaxin have been found to mediate granule-plasma membrane fusion and granule-granule fusion (compound exocytosis) in pancreatic acinar cells [20
]. Accordingly, human neutrophils seem to be equipped with an array of SNARE proteins that might well be the molecular mechanism governing simple as well as compound exocytosis.
If compound exocytosis is such a frequent event in human neutrophils, why has it evaded detection until now? As a result of the fast and transient nature of granule-granule fusion, these events are expected to be difficult to capture on static electron micrographs. This is also confirmed by our EM pictures, where we were able to find only two examples of compound exocytosis. Also, the whole-cell capacitance technique does not have sufficient resolution to reveal exocytosis of small, single granules from human neutrophils. We therefore we conclude that the cell-attached patch-clamp technique is superior for detection of complex granule exocytosis in cells with small (<500 nm in diameter) granules.
Granule-granule fusion
It would be interesting to examine wheter one kind of granule fuses only with identical granules (e.g., azurophilic granules with azurophilic granules) or whether all types of granules might fuse with each other. Unfortunately, the cell-attached patch-clamp technique by itself does not allow for this. Theoretically, it might be possible if the carbon electrode in the patch amperometry technique [32
] (combining cell-attached patch-clamp capacitance with amperometry) could be used in the voltametric mode and could thus distinguish different granules by differences in the charge of the matrix content and at the same time resolve compound exocytosis by capacitance.
We found a large range in the size of compound steps (5–100.6 fF), which we attribute to large intercell variations. The size of the patch pipette differs between experiments as will the area of the neutrophil that is covered by the pipette, which will lead to differences in the calcium concentrations at the site of fusion. Furthermore, there might be hot spots on the plasma membrane where calcium channels or SNARE receptors are concentrated. Arbitrarily placing the patch pipette on the membrane will be predicted to lead to variability.
It would be ideal to be able to measure calcium in the patch in order to determine the concentration, which is needed for compound exocytosis. Unfortunately, the patch pipette will distort the fluorescence signal in the patch, such that relevant local calcium concentrations cannot be determined for technical reasons.
Function of compound exocytosis
Compound exocytosis and cumulative fusion might serve to target release focally, which could be important in obtaining high, local concentrations of microbicidal products. Recently, it has been shown that cleavage of the human cathelicidin hCAP18 to the functional antimicrobial peptide LL-37 by proteinase 3 only takes place extracellularly [25
]. As hCAP18 and proteinase 3 reside in different granule populations (specific and azurophilic granules, respectively), compound exocytosis might facilitate important processing of the cathelicidin in this case. In pancreatic acini cells, cumulative fusion has been suggested to reduce the need for transport of granules to the plasma membrane [26
]. This might be an important mechanism for long-lived cells. However, neutrophils are short-lived, undergo apoptosis rapidly, and thus do not need to save resources for subsequent exocytotic events.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received March 4, 2001; revised January 12, 2002; accepted January 14, 2002.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Neef, D. Khimich, P. Pirih, D. Riedel, F. Wolf, and T. Moser Probing the Mechanism of Exocytosis at the Hair Cell Ribbon Synapse J. Neurosci., November 21, 2007; 27(47): 12933 - 12944. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Stie, A. V. Jesaitis, C. I. Lord, J. M. Gripentrog, R. M. Taylor, J. B. Burritt, and A. J. Jesaitis Localization of hCAP-18 on the surface of chemoattractant-stimulated human granulocytes: analysis using two novel hCAP-18-specific monoclonal antibodies J. Leukoc. Biol., July 1, 2007; 82(1): 161 - 172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gomez-Gaviro, M. Dominguez-Luis, J. Canchado, J. Calafat, H. Janssen, E. Lara-Pezzi, A. Fourie, A. Tugores, A. Valenzuela-Fernandez, F. Mollinedo, et al. Expression and Regulation of the Metalloproteinase ADAM-8 during Human Neutrophil Pathophysiological Activation and Its Catalytic Activity on L-Selectin Shedding J. Immunol., June 15, 2007; 178(12): 8053 - 8063. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Lominadze, D. W. Powell, G. C. Luerman, A. J. Link, R. A. Ward, and K. R. McLeish Proteomic Analysis of Human Neutrophil Granules Mol. Cell. Proteomics, October 1, 2005; 4(10): 1503 - 1521. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Hartmann, S. Scepek, I. Hafez, and M. Lindau Differential Regulation of Exocytotic Fusion and Granule-Granule Fusion in Eosinophils by Ca2+ and GTP Analogs J. Biol. Chem., November 7, 2003; 278(45): 44929 - 44934. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
