Discovery of the Porosome: revealing the molecular mechanism of secretion and membrane fusion in cells

Abstract

INVITED REVIEW

Discovery of the Porosome: revealing the molecular mechanism of secretion and membrane fusion in cells

B. P. Jena *

Department of Physiology, Wayne State University School of Medicine, Detroit, Michigan, USA

Received: March 19, 2004; Accepted: March 23, 2004

* Correspondence to: Bhanu P. JENA, Ph.D. Department of Physiology, Wayne State University School of Medicine, 5245 Gordon Scott Hall, 540 E. Canfield, Detroit, MI 48201, USA; Tel.: 313-577-1532, Fax: 313-993-4177, E-Mail: bjena@med.wayne.edu

Secretion and membrane fusion are fundamental cellular processes involved in the physiology of health and disease. Studies within the past decade reveal the molecular mechanism of secretion and membrane fusion in cells. Studies reveal that membrane-bound secretory vesicles dock and fuse at porosomes, which are specialized plasma membrane structures. Swelling of secretory vesicles result in a build-up of intravesicular pressure, which allows expulsion of vesicular contents. The discovery of the porosome, its isolation, its structure and dynamics at nm resolution and in real time, its biochemical composition and functional reconstitution, are discussed. The molecular mechanism of secretory vesicle fusion at the base of porosomes, and vesicle swelling, have been resolved. With these findings a new understanding of cell secretion has emerged and confirmed by a number of laboratories.

Keywords: porosome • cell secretion • membrane fusion • SNAREs • vesicle swelling

 Source: J. Cell. Mol. Med. Vol 8, No 1, 2004 pp. 1-21


Porosome: Structure and function

To determine the morphology of the porosome at the cytosolic side of the cell, pancreatic PM preparations were used. Isolated PM in buffer when placed on freshly cleaved mica, tightly adhere to the mica surface, enabling imaging by AFM. The PM preparations reveal scattered circular disks measuring 0.5-1 m in diameter, with inverted cup-shaped structures within (Fig. 7). The inverted cups range in height from 10-15 nm. On a number of occasions, zymogen granules (ZGs) ranging in size from 0.4-1 m in diameter were found associated with one or more of the inverted cups (Fig. 7A-C). This suggested the circular disks to be pits, and the inverted cups to be fusion pores or porosomes. To determine if the cup-shaped structures in isolated PM preparations are indeed porosomes, immuno- AFM studies were carried out. Since ZGs, the membrane-bound secretory vesicles in exocrine pancreas are known to dock and fuse at the PM to release vesicular contents, it was hypothesized that if porosomes are these sites, then PM-associated t- SNAREs should localize at the base of porosomes (tip of the inverted cups). The t-SNARE protein SNAP-23, has been identified and implicated in secretion from pancreatic acinar cells [7]. A polyclonal monospecific SNAP-23 antibody recognizing a single 23kDa band in Western blot analysis of pancreatic PM fraction (Fig. 7D) was used in immuno-AFM studies. When the SNAP-23 specific antibody was added to the PM preparation during imaging with the AFM, the antibody selectively localized to the base of the cup-shaped structure, which is the tip of the inverted cup (Fig. 7E and F). No antibody labeling of the structure was detected when preimmune serum was applied. These results demonstrate that the inverted cup-shaped structures in isolated PM preparations are the porosomes observed from its cytosolic side [8]. Target membrane proteins, SNAP-25 and syntaxin (t-SNARE) and secretory vesicle-associated membrane protein (v-SNARE), are part of the conserved protein complex involved in fusion of opposing bilayers [9, 10]. Since membrane-bounded secretory vesicles dock and fuse at porosomes to release vesicular contents, suggested t-SNAREs to associate at the porosome complex. It was therefore no surprise, that the t-SNARE protein SNAP-23, implicated in secretion from pancreatic acinar cells [8], was located at the tip of the inverted cup (i.e., the base of the porosome) where secretory vesicles dock and fuse. The structure of the porosome was further demonstrated using transmission electron microscopy (TEM) [8, 11] (Fig. 8). TEM studies confirm the fusion pore to have a cup-shaped structure, with similar dimensions as determined from AFM measurement [2-5]. Additionally, TEM micrographs reveal porosomes to possess a basket-like morphology, with three lateral and a number of vertically arranged ridges [8, 11]. A ring at the base of the complex is also identified. Since porosomes are found to be stable structures at the cell PM, it was hypothesized that if ZGs, the membrane-bounded secretory vesicles in exocrine pancreas were to fuse at the base of the structure, it would be possible to isolate ZG-associated porosomes. Indeed, TEM of isolated ZG preparations reveal the presence of porosomes associated with docked vesicles [8] (Fig. 9). As observed in whole cells, vertical structures were found to originate from within the porosome complex and appear attached to its membrane. The presence of vertical ridges lining the porosome, have also been reported in NG108-15 nerve cells [6]. As discussed in further detail later in this review, studies using SNARE proteins and artificial lipid membranes demonstrated that t- and v- SNAREs located in opposing bilayers interact in a circular array to form conducting pores [12]. Since similar 45-50 nm circular structures are observed at the base of the porosome, and SNAP-23 immunoreactivity is found to localize at this site, suggests that the t-SNAREs present at the base of porosomes are possibly arranged in a ring pattern.


Discovery of a new cellular structure: The porosome

Introduction

Secretion and membrane fusion are fundamental cellular processes regulating ER-Golgi transport in protein maturation, plasma membrane recycling, cell division, sexual reproduction, acid secretion, and the release of enzymes, hormones and neurotransmitters, to name just a few. It is therefore no surprise that defects in secretion and membrane fusion give rise to diseases like diabetes, Alzheimer’s, Parkinson’s, acute gastroduodenal diseases, gastroesophageal reflux disease, intestinal infections due to inhibition of gastric acid secretion, biliary diseases resulting from malfunction of secretion from hepatocytes, polycystic ovarian disease as a result of altered gonadotropin secretion, and Gitelman disease associated with growth hormone deficiency and disturbances in vasopressin secretion, are only a few examples. Understanding cellular secretion and membrane fusion will therefore help not only to advance our understanding of these vital cellular and physiological processes, but in the development of drugs to help ameliorate secretory defects, provide insight into our understanding of cellular entry and exit of viruses and other pathogens, and in the development of smart drug delivery systems. Thus the role of secretion and membrane fusion in health and disease is profound. Studies in the last ten years reveal the molecular mechanism of secretion and membrane fusion in cells, which will be discussed in this article. Membrane-bound secretory vesicles containing their cargo, dock and fuse at specialized plasma membrane structures called porosomes. Secretory vesicles swell, resulting in a buildup of intravesicular pressure, which helps in expulsion of contents from within vesicles. The discovery of the porosome, its isolation, structure and dynamics, its biochemical composition and functional reconstitution, will be provided. Following docking of secretory vesicles at the base of porosomes, the molecular mechanism of fusion of the vesicle membrane at the porosome membrane, will be outlined. Finally, studies revealing the molecular mechanism of secretory vesicle swelling resulting in expulsion of vesicular contents, will be discussed. With these findings a new understanding of cell secretion has emerged, and confirmed by a number of laboratories.

Earlier electrophysiological studies on mast cells suggested the existence of fusion pores at the cell plasma membrane (PM), which became continuous with the secretory vesicle membrane after stimulation of secretion [1]. Atomic force microscopy (AFM) has confirmed the existence of the fusion pore and its structure and dynamics in both exocrine [2,3] and neuroendocrine cells [4,5] at near nm resolution and in real time. Fusion pores in NG108-15 nerve cells have also been reported [6]. Isolated live pancreatic acinar cells in physiological buffer, when imaged with the AFM [2,3], reveal at the apical PM a group of circular ‘pits’ measuring 0.4-1.2 μm in diameter which contain smaller ‘depressions’ (Fig. 1). Each depression averages between 100 and 150 nm in diameter, and typically 3-4 depressions are located within a pit. The basolateral membrane of acinar cells is however, devoid of either pits or depressions. High-resolution AFM images of depressions in live cells further reveal a cone-shaped morphology (Fig. 1). The depth of each depression cone measures 15-30 nm. Similarly, growth hormone (GH) secreting cells of the pituitary gland and chromaffin cells possess pits and depression structures at their PM [4,5], suggesting their universal presence in secretory cells. Exposure of pancreatic acinar cells to a secretagogue (mastoparan) results in a time-dependent increase (20-35%) in depression diameter, followed by a return to resting size on completion of secretion [2, 3] (Fig. 2). No demonstrable change in pit size is detected following stimulation of secretion [2]. Enlargement of depression diameter and an increase in its relative depth after exposure to secretagoguescorrelated with increased secretion. Conversely, exposure of pancreatic acinar cells to cytochalasin B, a fungal toxin that inhibits actin polymerization, results in a 15-20% decrease in depression size and a consequent 50-60% loss in secretion [2]. Results from these studies suggest that depressions are the fusion pores in pancreatic acinar cells. Furthermore, these studies demonstrate the involvement of actin in regulation of both the structure and function of depressions. Analogous to pancreatic acinar cells, examination of resting GH secreting cells of the pituitary [4] and chromaffin cells of the adrenal medulla [5] also reveal the presence of pits and depressions at the cell PM (Fig. 3). Depressions in resting GH cells measure 154 4.5 nm (mean ± SE) in diameter. Exposure of GH cells to a secretagogue results in a 40% increase in depression diameter (215 ± 4.6 nm; p Fig. 4). These studies confirm depressions to be the fusion pores or porosomes in pancreatic acinar cells where membrane-bound secretory vesicles dock and fuse to release vesicular contents. Similarly, in somatotrophs of the pituitary, gold-tagged growth hormone-specific antibody is found to selectively localize at depressions following stimulation of secretion [4], again identifying depressions in GH cells as fusion pores or porosomes. Furthermore, recent studies in the laboratory using both AFM and EM, reveal the presence of porosomes in neurons,  β-cells of the endocrine pancreas, and in mast cells (Fig. 5 and 6, unpublished).


Porosome: Isolation, composition, and reconstitution

In the last decade, a number of studies have demonstrated the involvement of cytoskeletal proteins in secretion, and some studies implicate direct interaction of cytoskeleton protein with SNAREs [2, 13- 17]. Furthermore, actin and microtubule-based cytoskeleton have been implicated in intracellular vesicle traffic [2, 15]. Fodrin, which was previously implicated in exocytosis [13], has recently been shown to directly interact with SNAREs [16]. Studies demonstrate α-fodrin to regulate exocytosis via its interaction with t-SNARE syntaxin family of proteins [16]. The c-terminal coiled coil region of syntaxin interacts with α-fodrin, a major component of the submembranous cytoskeleton. Similarly, vimentin filaments interact with SNAP23/25 and hence are able to control the availability of free SNAP23/25 for assembly of the SNARE complex [14]. All these findings suggested that vimentin, α- fodrin, actin, and SNAREs may be part of the porosome complex. Additional proteins such as v- SNARE (VAMP or synaptobrevin), synaptophysin and myosin, may associate when the porosome establishes continuity with the secretory vesicle membrane. The globular tail domain of myosin V contains binding site for VAMP, which is bound in a calcium independent manner [17]. Further interaction of myosin V with syntaxin requires both calcium and calmodulin. It has been suggested that VAMP acts as a myosin V receptor on secretory vesicles and regulates formation of the SNARE complex [17]. Interaction of VAMP with synapto- physin and myosin V has also been observed [18]. In agreement with earlier findings in other tissues, our studies have demonstrated the association of SNAP- 23, syntaxin 2, cytoskeletal proteins actin, α-fodrin, and vimentin, and calcium channels β3 and α1c, together with the SNARE regulatory protein NSF, in porosomes [8, 11] (Fig. 10). Additionally, chloride ion channels ClC2 and ClC3 were also identified as part of the porosome complex [8]. Isoforms of the various proteins identified in the porosome complex, were subsequently demonstrated using 2D-BAC gels electrophoresis [11]. Three isoforms each of the calcium ion channel and vimentin were clearly identifiable. Although multiple spots were identified in several of the immunoblots, the low molecular weight spots may represent proteolytic degredation of the parent molecule [11]. Using yeast 2-hybrid analysis, our study confirms the presence and further reveals the interactions of some of these proteins with t-SNAREs within the porosome complex (unpublished). The size and shape of the immunoisolated porosome complex when examined using both negative staining electron microscopy and AFM, was revealed in greater detail. The images of the immunoisolated porosome obtained by both EM and AFM were super-imposable (Fig. 11) [11]. To further test whether the immunoisolated supramolecular complex was indeed the porosome, the complex was reconstituted into artificial liposomes, and the liposome-reconstituted complex examined using TEM (Fig. 12) [11]. Transmission electron micrographs reveal a 150-200-nm cupshaped basket-like structure as observed of the porosome when co-isolated with ZGs. The important question then remained, are such reconstituted porosomes functional? To answer this question, the complex was also reconstituted into lipid membranes and challenged with isolated secretory vesicles (ZGs) in an electrophysiological bilayer setup (EPC9). Both the electrical activity of the reconstituted membrane as well as the transport of vesicular contents from the cis to the trans compartment, was monitored in the EPC9 electrophysiologicallipid membrane setup. Results of these experiments demonstrate that the lipid membrane-reconstituted porosomes are functional supramolecular complexes (Fig. 13) [11]. When the supramolecular porosome complexes is reconstituted into the lipid bilayer membrane in the EPC9 setup (Fig. 13A), ZG fused with the bilayer as demonstrated by an increase in capacitance and conductance, and a time-dependent release of amylase (one of the major contents of ZGs) from cis to the trans compartment of the chamber. Amylase is detected using immunoblot analysis of the buffer in the cis and trans chambers (Fig. 13B), using a previously characterized amylase specific antibody [3]. As observed in immunoblot assays of isolated porosomes (Fig. 10), chloride channel activities is additionally demonstrated within the reconstituted supramolecular porosome complex (Fig. 13C). Further, the chloride channel inhibitor DIDS, was found to inhibit current activity in the porosomereconstituted bilayer. Contrarily, although our immunoblot analysis of porosomes demonstrate the association of calcium channels with the complex, we were unable to detect calcium channel activities in the reconstituted membrane. This may have been due to inactivation of the associated calcium channels in the complex as a result of low pH wash during immunoisolation, since recently we have been able to recover this activity (unpublished). The role of the chloride channel in the porosome complex remains unknown at this time. In summary, these studies demonstrate that the porosome is a 100-150- nm in diameter supramolecular cup-shaped lipoprotein basket at the cell PM, where membrane-bound secretory vesicles dock and fuse to release vesicular contents.


Molecular mechanism of membrane fusion

As eluded earlier, membrane fusion is mediated via a specialized set of proteins in the secretory vesicles and the plasma membrane. Three soluble N-ethylmaleimide- sensitive factor (NSF)-attachment protein receptors (SNAREs) have been implicated in membrane fusion [10]. Target membrane proteins, SNAP-25 and syntaxin (t-SNARE) and secretory vesicle-associated membrane protein (v-SNARE), are part of the conserved protein complex involved in fusion of opposing bilayers [9, 10]. Although the structure of SNARE complex formed by interacting native [19] or recombinant [20, 21] t- and v- SNAREs was known from studies using electron microscopy [19, 20] and x-ray crystallography [21], the molecular mechanism of the involvement of SNAREs to bring about membrane fusion remained unknown until two years ago [12]. To determine the molecular mechanism of SNARE-induced membrane fusion, the structure and arrangement of SNAREs associated with lipid bilayers were examined using atomic force microscopy. The bilayer electrophysiological setup (EPC9) allowed measurements of membrane conductance and capacitance, prior to and after t- SNARE or v-SNARE reconstitution, and following exposure of v-SNARE or t-SNARE reconstituted vesicles. These studies demonstrate that the interaction of t-/v-SNARE proteins to form a fusion pore is dependent on the presence of t-SNAREs and v- SNARE in opposing bilayers. Addition of purified recombinant v-SNARE to a t-SNARE-reconstituted lipid membrane increased only the size of the globular t-SNARE oligomer without influencing the electrical properties of the membrane (Fig. 14). However when t-SNARE vesicles were added to a v-SNARE membrane, SNAREs assembles in a ring pattern (Fig. 15) and a stepwise increase in capacitance, and increase in conductance were observed (Fig. 16). Thus, t- and v-SNAREs are required to reside in opposing bilayers to allow appropriate t-/v-SNARE interactions leading to membrane fusion in the presence of calcium [12]. Fusion of membrane-bounded secretory vesicles with the target membrane of the porosome is a highly regulated event, where a large number of proteins participate [8, 11]. Among them the SNAREs, which were suggested to be the minimal fusion machinery [10]. Studies however report calcium (Ca2+) to be a major fusogen, whereas SNAREs promote Ca2+ sensitivity to the fusion process [22-24]. Studies further reveal that micro domains of high Ca2+ concentrations co-localize at the fusion site [25,26]. Ca2+ ion channels have been found to associate with the SNARE complex [27] and with the porosome complex [8, 11]. Furthermore, in the presence of Ca2+, t- SNAREs and v-SNARE in opposing bilayers interact in a circular array to form conducting pores [12]. Finally, several Ca2+-binding proteins such as synaptotagmin and syncollin, which interact with SNAREs in a Ca2+-dependent manner, have also been identified [28, 29], further supporting the involvement of Ca2+ in membrane fusion. Hence, there was growing evidence supporting Ca2+ as a key player in membrane fusion. To determine the role of Ca2+ in SNARE-induced membrane fusion, the fusion of t- /v-SNARE-reconstituted liposomes was investigated using various approaches [30]. Results from these studies lead to the same conclusion i.e., (i) a low fusion rate (t=16 min) between t- and v-SNAREreconstituted liposomes in the absence of Ca2+; and (ii) exposure of t-/v-SNARE liposomes to Ca2+, drives vesicle fusion on a near physiological relevant time-scale (t ~10s), demonstrating an essential role of Ca2+ in membrane fusion (Fig. 17, 18). This study also supports earlier findings on the role of Ca2+ and SNAREs in cortical vesicle fusion in sea urchin eggs [22, 24], where Ca2+ was found to acts downstream of SNAREs. Since the Ca2+ effect on membrane fusion in SNARE-reconstituted liposomes is downstream of SNAREs, suggests a regulatory role for Ca2+-binding proteins in membrane fusion in the physiological state [30]. These studies further demonstrate that in the physiological state in cells, both SNAREs and Ca2+ operate as the minimal fusion machinery [30]. Native and synthetic vesicles exhibit a significant negative surface charge primarily due to the polar phosphate head groups. These polar head groups produce a repulsive force, preventing aggregation and fusion of apposing vesicles. SNAREs bring opposing bilayers closer to within a distance of 2-3 Å, allowing Ca2+ to bridge them [30]. The bound Ca2+ would then lead to expulsion of water between the bilayers at the bridging site, allowing lipid mixing and promote membrane fusion. Hence SNAREs, besides bringing opposing bilayers closer, dictate the site and size of the fusion area (fusion occurs within the circular array [12] formed by the t-/v-SNARE complex) during secretion. The size of the t-/v-SNARE complex forming the pore is found dependent on the size of t-SNARE or v-SNARE reconstituted vesicles used. The smaller the vesicles, the smaller the pores formed (unpublished observation).


Regulation of secretory vesicle swelling: Involvement in expulsion of vesicular contents

In the last decade, the molecular mechanism of vesicle swelling [31-33] and its involvement in the regulated expulsion of vesicular contents [Jena, et al., unpublished observation] has been established. Secretory vesicle swelling is critical for secretion [34-37], however, the underlying mechanism of vesicle swelling was largely unknown until recently [31-33]. In mast cells, an increase in secretory vesicle volume after stimulation of secretion has previously been suggested from electrophysiological measurements [38]. However, direct evidence of secretory vesicle swelling in live cells was first demonstrated in pancreatic acinar cells using the AFM (Fig. 19) [39]. Isolated zymogen granules (ZGs), the membrane-bound secretory vesicles in exocrine pancreas and parotid glands, possess Cland ATP-sensitive, K+-selective ion channels at the vesicle membrane whose activities have been implicated in vesicle swelling [31]. Additionally, secretion of ZG contents in permeabilized pancreatic acinar cells requires the presence of both K+ and Cl- ions. In vitro ZG-pancreatic plasma membrane fusion assays further demonstrate potentiation of fusion in the presence of GTP [39]. Gai protein has been implicated in the regulation of both K+ and Cl- ion channels in a number of tissues. Analogous to the regulation of K+ and Cl- ion channels at the cell plasma membrane, their regulation at the ZG membrane by a Gai3 protein was demonstrated [31]. Isolated ZGs from exocrine pancreas swell rapidly in response to GTP (Fig. 20) [31]. These studies suggested the involvement of rapid water entry into ZGs following exposure to GTP. Therefore, when the possible involvement of water channels or aquaporins in ZG swelling was explored [32], results from the study demonstrate the presence of aquaporin-1 (AQP1) at the ZG membranes and its participation in GTP-mediated vesicle water entry and swelling (Fig. 21) [32]. To further understand the molecular mechanism of secretory vesicle swelling, the regulation of AQP1 in the ZG was investigated [33]. Detergent-solubilized ZGs immunoprecipitated with monoclonal AQP-1 antibody, co-isolates AQP-1, PLA2, Gai3, potassium channel IRK-8, and the chloride channel ClC-2 [33]. Exposure of ZGs to either the potassium channel blocker glyburide, or the PLA2 inhibitor ONO-RS-082, blocked GTP-induced ZG swelling. RBC, known to possess AQP-1 at the plasma membrane, also swell on exposure to the Gai-agonist mastoparan, and respond similarly to ONO-RS-082 and glyburide, as do ZGs. Additionally, liposomes reconstituted with the AQP-1 immunoisolated complex from solubilized ZGs, also swell in response to GTP. Glyburide or ONO-RS-082 abolished the GTP effect in reconstituted liposomes. Furthermore, immunoisolate-reconstituted planar lipid membrane demonstrate conductance, which is sensitive to glyburide and an AQP-1 specific antibody. These results demonstrate a Gai3-PLA2 mediated pathway and potassium channel involvement in AQP-1 regulation (Fig. 22) [33], contributing to our understanding of the molecular mechanism of ZG swelling. Although secretory vesicle swelling is involved in membrane fusion [37], our recent studies demonstrate that its primary role is in the expulsion of vesicular contents during secretion (unpublished observation).


A new understanding of cell secretion

Fusion pores or porosomes are present in all secretory cells examined. From exocrine, endocrine, neuroendocrine, to neurons, where membranebound secretory vesicles dock and transiently fuse to expel vesicular contents. Porosomes in pancreatic acinar or GH-secreting cells are cone-shaped structures at the plasma membrane, with a 100- to 150-nm-diameter opening. Membrane-bound secretory vesicles ranging in size from 0.2 to 1.2 μm in diameter dock and fuse at porosomes to release vesicular contents. Following fusion of secretory vesicles at porosomes, only a 20-35% increase in porosome diameter is demonstrated. It is therefore reasonable to conclude that secretory vesicles “transiently” dock and fuse at the site. In contrast to accepted belief, if secretory vesicles were to completely incorporate at porosomes, the PM structure would distend much wider than what is observed. Furthermore, if secretory vesicles were to completely fuse at the plasma membrane, there would be a loss in vesicle number following secretion. Examination of secretory vesicles within cells before and after secretion demonstrates that the total number of secretory vesicles remains unchanged following secretion. However, the number of empty and partially empty vesicles increases significantly, supporting the occurrence of transient fusion [39-43]. Earlier studies on mast cells also demonstrated an increase in the number of spent and partially spent vesicles following stimulation of secretion, without any demonstrable increase in cell size [44]. Similarly, secretory granules are recaptured largely intact after stimulated exocytosis in cultured endocrine cells [45]. Other supporting evidence favoring transient fusion is the presence of neurotransmitter transporters at the synaptic vesicle membrane. These vesicle-associated transporters would be of little use if vesicles were to fuse completely at the plasma membrane to be compensatorily endocytosed at a later time. In further support, a recent study reports that single synaptic vesicles fuse transiently and successively without loss of vesicle identity [46]. Although the fusion of secretory vesicles at the cell plasma membrane occurs transiently, complete incorporation of membrane at the cell plasma membrane would occur when cells need to incorporate signaling molecules like receptors, second messengers, or ion channels.

Conclusion

The discovery of the porosome, and an understanding of the molecular mechanism of membrane fusion and the swelling of secretory vesicles required for expulsion of vesicular contents, finally provides a clear understanding of secretion and membrane fusion in cells. These findings have prompted many laboratories to work in the area and further confirm these findings. Thus, the porosome is a supramolecular structure universally present in secretory cells, from the exocrine pancreas to the neurons, and in the endocrine to neuroendocrine cells, where membrane-bound secretory vesicles transiently dock and fuse to expel vesicular contents. Hence, the secretory process in cells is a highly regulated event, orchestrated by a number of biomolecules.

Acknowledgement

I wish to thank Won-Jin Cho for help in preparation of figures. Supported by Grants DK-56212 and NS-39918 from the National Institutes of Health (BPJ).


References



1. Monck J.R., Oberhauser A.F., Fernandez J.M., The exocytotic fusion pore interface: a model of the site of neurotransmitter release, Mol. Memb. Biol.,12:151-156, 1995

2. Schneider S.W., Sritharan K.C., Geibel J.P., Oberleithner H., Jena B.P., Surface dynamics in living acinar cells imaged by atomic force microscopy: identifi- cation of plasma membrane structures involved in exocytosis, Proc. Natl. Acad. Sci. USA, 94:316-321, 1997

3. Cho S.-J., Quinn A.S., Stromer M.H., Dash S., Cho J., Taatjes D.J., Jena B.P., Structure and dynamics of the fusion pore in live cells, Cell Biol. Int., 26:35-42, 2002

4. Cho S.-J., Jeftinija K., Glavaski A., Jeftinija S., Jena B.P., Anderson L.L., Structure and dynamics of the fusion pores in live GH-secreting cells revealed using atomic force microscopy, Endocrinology, 143:1144-1148, 2002

5. Cho S.-J., Wakade A., Pappas G.D., Jena B.P., New structure involved in transient membrane fusion and exocytosis, New York Acad. Sci., 971:254-256, 2002

6. Tojima T., Yamane Y., Takagi H., Takeshita T., Sugiyama T., Haga H., Kawabata K., Ushiki T., Abe K., Yoshioka T., Ito E., Three-dimensional characterization
of interior structures of exocytotic apertures of nerve cells using atomic force microscopy, Neuroscience, 101:471- 481, 2000

7. Gaisano H.Y., Sheu L., Wong P.P., Klip A., Trimble W.S., SNAP-23 is located in the basolateral plasma membrane of rat pancreatic acinar cells, FEBS Lett., 414:298-302, 1997

8. Jena B.P., Cho S-J., Jeremic A., Stromer M.H., Abu- Hamdah R., Structure and composition of the fusion pore, Biophys. J., 84:1337-1343, 2003

9. Rothman J.E., Mechanism of intracellular protein transport, Nature, 372:55-63, 1994

10. Weber T., Zemelman B.V., McNew J.A., Westerman B., Gmachl M., Parlati F., Sollner T.H., Rothman J.E., SNAREpins: minimal machinery for membrane fusion, Cell, 92:759-772, 1988

11. Jeremic A.M., Kelly M., Cho S-J., Stromer M.H., Jena B.P., Reconstituted fusion pore, Biophys. J., 85:2035- 2043, 2003

12. Cho S.-J., Kelly M., Rognlien K.T., Cho J., Hoerber J.K.H., Jena B.P., SNAREs in opposing bilayers interact in a circular array to form conducting pores, Biophys. J., 83:2522-2527, 2002

13. Bennett V., Spectrin-based membrane skeleton: a multipotential adaptor between plasma membrane and cytoplasm, Physiol. Rev., 70:1029-1065, 1990

14. Faigle W., Colucci-Guyon E., Louvard D., Amigorena S., Galli T., Vimentin filaments in fibroblasts are a reservoir for SNAP-23, a component of the membrane fusion machinery, Mol. Biol. Cell., 11:3485-3494, 2000

15. Goodson H.V., Valetti C., Kreis T.E., Motors and membrane traffic, Curr. Opin. Cell Biol., 9:18-28, 1997

16. Nakano M., Nogami S., Sato S., Terano A., Shirataki H., Interaction of syntaxin with -fodrin, a major component of the submembranous cytoskeleton, Biochem. Biophys. Res. Commun., 288:468-475, 2001

17. Ohyama A., Komiya Y., Igarashi M., Globular tail of myosin-V is bound to vamp/synaptobrevin, Biochem. Biophys. Res. Commun., 280:988-991, 2001

18. Prekereis R., Terrian D.M., Brain myosin V is a synaptic vesicle-associated motor protein: evidence for a Ca2+- dependent interaction with the synaptobrevin-synaptophysin complex, J. Cell Biol., 137:1589-1601, 1997

19. Jeong E.-H., Webster P., Khuong C.Q., Sattar A.K.M.A., Satchi M., Jena B.P., The native membrane fusion machinery in cells, Cell Biol. Int., 22:657-670,
1998

20. Hanson P.I., Roth R., Morisaki H., Jahn R., Heuser J.E., Structure and conformational changes in NSF and its membrane receptor complexes visualized by quickfreeze/ deep-etch electron microscopy, Cell, 90:523-535, 1997

21. Sutton R.B., Fasshauer D., Jahn R., Brunger A.T., Crystal structure of a SNARE complex involved in synaptic exocytosis at 2.4 Å resolution, Nature, 395:347-353, 1998

22. Coorssen J.R., Blank P.S., Tahara M., Zimmerberg J., Biochemical and functional studies of cortical vesicle fusion: the SNARE complex and Ca2+ sensitivity, J. Cell Biol., 143:1845-1857, 1998

23. Tahara M., Coorssen J.R., Timmers K., Blank P.S., Whalley T., Scheller R., et al., Calcium can disrupt the SNARE protein complex on sea urchin egg secretory vesicles without irreversibly blocking fusion, J. Biol. Chem., 273:33667-33673, 1998

24. Zimmerberg J., Blank P.S., Kolosova I., Cho M.S., Tahara M., Coorssen J.R., A stage-specific preparation to study the Ca2+-triggered fusion steps of exocytosis: rationale and perspectives, Biochimie, 82:303-314, 2000

25. Llinas R., Sugimori M., Silver R.B., Microdomains of high calcium concentration in a presynaptic terminal, Science, 256:677-679, 1992

26. Llinas R., Sugimori M., Silver R.B., Presynaptic calcium concentration microdomains and transmitter release, J. Physiol. Paris, 86:135-138, 1992
27. Sheng Z.H., Rettig J., Cook T., Catterall W.A., Calcium-dependent interaction of N-type calcium channels with the synaptic core complex, Nature, 379:451-454, 1996

28. Edwardson J.M., An S., Jahn R., The secretory granule protein syncollin binds to syntaxin in a Ca2+-sensitive manner, Cell, 90:325-333, 1997

29. Sudhof T.C., Rizo J., Synaptotagmins: C2-domain proteins that regulate membrane traffic, Neuron, 17:379-388, 1996

30. Jeremic A., Kelly M., Cho J.-H., Cho S.-J., Horber J.K.H., Jena B.P., Calcium drives fusion of SNAREapposed bilayers, Cell Biol. Int., 28:19-31, 2004
31. Jena B.P., Schneider S.W., Geibel J.P., Webster P., Oberleithner H., Sritharan K.C., Gi regulation of secretory vesicle swelling examined by atomic force
microscopy, Proc. Natl. Acad. Sci. USA, 94:13317-13322, 1997

32. Cho S.-J., Sattar A.K., Jeong E.H., Satchi M., Cho J.A., Dash S., Mayes M.S., Stromer M.H., Jena B.P., Aquaporin 1 regulates GTP-induced rapid gating of water in secretory vesicles, Proc. Natl. Acad. Sci. USA, 99:4720- 4724, 2002

33. Abu-Hamdah R., Cho W.-J., Cho S.-J., Jeremic A., Kelly M., Ilie A.E., Jena B.P., Regulation of the water channel aquaporin-1: isolation and reconstitution of the regulatory complex, Cell Biol. Int., 28:7-17, 2004

34. Alvarez de Toledo G., Fernandez-Chacon R., Fernandez J.M., Release of secretory products during transient vesicle fusion, Nature, 363:554-558, 1993

35. Curran M.J., Brodwick M.S., Ionic control of the size of the vesicle matrix of beige mouse mast cells, J. Gen. Physiol., 98:771-790, 1991

36. Monck J. R., Oberhauser A.F., Alvarez de Toledo G., Fernandez J.M., Is swelling of the secretory granule matrix the force that dilates the exocytotic fusion pore? Biophys. J., 59:39-47, 1991

37. Sattar A.K.M., Boinpally R., Stromer M.H., Jena B.P., Gi3 in pancreatic zymogen granule participates in vesicular fusion, J. Biochemistry, 131:815-820, 2002

38. Fernandez J.M., Villalon M., Verdugo P., Reversible condensation of the mast cell secretory products in vitro, Biophys. J., 59:1022-1027, 1991

39. Cho S.-J., Cho J., Jena B.P., The number of secretory vesicles remains unchanged following exocytosis, Cell Biol. Int., 26:29-33, 2002

40. Jena B.P., Fusion pore in live cells, NIPS, 17:219-222, 2002

41. Jena B.P., Fusion pore: structure and dynamics, J. Endocrinology, 176:169-174, 2003

42. Jena B.P., Exocytotic fusion: total or transient, Cell Biol. Int., 21: 257-259, 1997

43. Lee J.-S., Mayes M.S., Stromer M.H., Scanes C.G., Jeftinija S., Anderson L.L., Number of secretory vesicles in growth hormone cells of the pituitary remains
unchanged after secretion, Exp. Biol. Med. (in press, 2004)

44. Lawson D., Fewtrell C., Gomperts B., Raff M., Antiimmunoglobulin induced histamine secretion by rat peritoneal mast cells studied by immuno ferritin electron microscopy, J. Exp. Med., 142: 391-402, 1975

45. Taraska J.W., Perrais D., Ohara-Imaizumi M., Nagamatsu S., Almers W., Secretory granules are recaptured largely intact after stimulated exocytosis in cultured endocrine cells, Proc. Natl. Acad. Sci. USA, 100:2070- 2075, 2003

46. Aravanis A.M., Pyle J.L., Tsien R.W., Single synaptic vesicles fusing transiently and successively without loss of identity, Nature, 423: 643-647, 2003


Figures

Figures

..................................................

Fig. 1. On the far left is an AFM micrograph depicting ‘pits’ (yellow arrow) and ‘depressions’ within (blue arrow), at the plasma membrane in live pancreatic acinar cells. On the right is a schematic drawing depicting depressions, at the cell plasma membrane, where membrane-bound secretory vesicles dock and fuse to release vesicular contents [2].

 

figure 1

 ..................................................

Fig. 2. Dynamics of depressions following stimulation of secretion. The top panel shows a number of depressions within a pit in a live pancreatic acinar cell. The scan line across three depressions in the top panel is represented graphically in the middle panel and defines the diameter and relative depth of the depressions; the middle depressions is represented by red arrowheads. The bottom panel represents percent of total cellular amylase release in the presence and absence of the secretagogue Mas 7. Notice an increase in the diameter and depth of depressions, correlating with an increase in total cellular amylase release at 5 min after stimulation of secretion. At 30 min after stimulation of secretion, there is a decrease in diameter and depth of depressions, with no further increase in amylase release over the 5 min time point. No significant increase in amylase secretion or depressions diameter were observed in resting acini or those exposed to the nonstimulatory mastoparan analog Mas 17 [2, 40].

 

figure 2

 ..................................................

Fig. 3. Depressions in live secretory cell of the exocrine pancreas (A,B), the growth hormone secreting cell of the pituitary (C), and in the chromaffin cell (D). Note the ‘pit’ (white arrow heads) with four depressions (yellow arrow head). A high resolution AFM micrograph is shown in Fig. B. Bars = 40 nm for Fig. A and B. [4,5,41].

 

figure 3

 ..................................................

Fig. 4 Depressions are fusion pores or porosomes. Porosomes dilate to allow expulsion of vesicular contents. (A and B) AFM micrographs and section analysis of a pit and two out of the four fusion pores or porosomes, demonstrating enlargement following stimulation of secretion. (C) Exposure of live cells to gold conjugated-amylase antibody (Ab) results in specific localization of immuno-gold to the porosome opening. Amylase is one of the proteins within secretory vesicles of the exocrine pancreas. (D) AFM micrograph of a fixed pancreatic acinar cell, demonstrating a pit and porosomes within, immunogold-labeling amylase at the site. Blue arrowheads point to immunogold clusters and the yellow arrowhead points to a porosome [3].

 

figure 4

 ..................................................

Fig. 5. Electron micrograph of Porosomes in neurons. A. Electron micrograph of a synaptosome demonstrating the presence of 40-50 nm synaptic vesicles. B-D. Electron micrograph of neuronal porosomes which are 10-15 nm cup-shaped structures at the presynaptic membrane (yellow arrow head), where synaptic vesicles transiently dock and fuse to release vesicular contents. [Kelly, M.L., Cho, W-J., Jeremic, A., Lazrishvili, I.L., Bikashvili, T.Z., Zhvania, M., and Jena, B.P., unpublished observation].

 

figure 5

 ..................................................

Fig. 6. AFM micrographs of porosomes in β-cell of the endocrine pancreas (A,B) and mast cell (C,D). Note the 100-130 nm porosomes in the β-cell (B) and the 70-80 nm porosomes in the mast cell (D), demonstrated using AFM section analysis. [Jena, B.P., unpublished observation].

 

figure 6

 ..................................................

Fig. 7. Morphology of the cytosolic side of the porosome revealed in AFM studies on isolated pancreatic plasma membrane (PM) preparations. (A). AFM micrograph of isolated PM preparation reveals the cytosolic end of a pit with inverted cup-shaped structures, the porosome. Note the 600 nm in diameter ZG at the left hand corner of the pit. (B). Higher magnification of the same pit showing clearly the 4-5 porosomes within. (C). The cytosolic end of a single porosome is depicted in this AFM micrograph. (D). Immunoblot analysis of 10μg and 20 μg of pancreatic PM preparations, using SNAP-23 antibody, demonstrates a single 23 kDa immunoreactive band. (E,F). The cytosolic side of the PM demonstrating the presence of a pit with a number of porosomes within, shown prior to (E) and following addition of the SNAP-23 antibody (F). Note the increase in height of the porosome cone base revealed by section analysis (bottom pannel), demonstrating localization of SNAP-23 antibody at the base of the porosome [8].

 

figure 7

 ..................................................

Fig. 8. Electron micrograph depicting a porosome (red arrow head) close to a microvilli (MV) at the apical plasma membrane (PM) of a pancreatic acinar cell. Note association of the porosome membrane (yellow arrow head), and the zymogen granule membrane (ZGM) (red arrow head) of a docked zymogen granule (ZG), the membranebound secretory vesicle of exocrine pancreas. Also a cross section of the ring at the mouth of the porosome is seen (blue arrow head) [11].

 

figure 8

 ..................................................

Fig. 9. Transmission electron micrograph of porosome associated with zymogen granules, the secretory vesicle in exocrine pancreas. (A) An isolated zymogen granule associated with a porosome, reveals (B) clearly the lateral and vertical structures in the complex. The porosome membrane (PM) and the vesicle membrane (VM) of the zymogen granule is clearly seen. Scale = 100 nm. [8, 11].

 figure 9

 ..................................................

Fig. 10. SNAP-23 associated proteins in pancreatic acinar cells. Total pancreatic homogenate was immunoprecipitated using the SNAP-23 specific antibody. The precipitated material was resolved using 12.5% SDS-PAGE, electrotransferred to nitrocellulose membrane and then probed using antibodies to a number of proteins. Association of SNAP-23 with syntaxin2, with cytoskeletal proteins actin, α-fodrin, and vimentin, and calcium channels β3 and α1c, together with the SNARE regulatory protein NSF, is demonstrated (arrow heads). Lanes showing more than one arrowhead suggest presence of isomers or possible proteolytic degradation of the specific protein [8].

 

figure 10

 ..................................................

Fig. 11. Negatively stained electron micrograph and atomic force micrograph of the immunoisolated porosome complex. (A). Negatively stained electron micrograph of an immunoisolated porosome complex from solubilized pancreatic plasma membrane preparations, using a SNAP-23 specific antibody. Note the three rings and the 10 spokes that originate from the inner smallest ring. This structure represents the protein backbone of the porosome complex, since the three rings and the vertical spikes are observed in electron micrographs of cells and porosome co-isolated with ZGs. Bar = 30 nm. (B). The electron micrograph of the fusion pore complex, cut out from (A), and (C) an outline of the structure presented for clarity. (D-F). Atomic force micrograph of the isolated pore complex in near physiological buffer. Bar = 30 nm. Note the structural similarity of the complex, imaged both by EM (G) and AFM (H). The EM and AFM micrographs are superimposable (I). [11].

 

figure 11

 ..................................................

Fig. 12. Electron micrographs of reconstituted Porosome or fusion pore complex in liposomes, showing a cup-shaped basket-like morphology. (A). A 500 nm vesicle with an incorporated porosome is shown. Note the spokes in the complex. The reconstituted complex at greater magnification is shown in Fig. B-D. Bar = 100 nm. [11].

 figure 12

 ..................................................

Fig. 13. Lipid bilayer-reconstituted Porosome complex is functional. (A). Schematic drawing of the bilayer setup for electrophysiological measurements. (B). Zymogen Granules (ZGs) added to the cis side of the bilayer fuse with the reconstituted Porosomes, as demonstrated by an increase in capacitance and current activities, and a concomitant time dependent release of amylase (a major ZG content) to the trans side of the membrane. The movement of amylase from the cis to the trans side of the chamber was determined by immunoblot analysis of the contents in the cis and the trans chamber over time. (C). As demonstrated by immunoblot analysis of the immunoisolated complex, electrical measurements in the presence and absence of chloride ion channel blocker DIDS, demonstrate the presence of chloride channels in association with the complex [11].

 

figure 13

 ..................................................

Fig. 14. AFM micrographs and force plots of mica and lipid surface and of SNAREs on lipid membrane. (A) AFM performed on freshly cleaved mica (left) and on lipid membrane formed on the same mica surface (right), demonstrating differences in the force-versus-distant curves. Note the curvilinear shape exhibited in the force-versus-distant curves of the lipid surface in contrast to mica. Three dimensional AFM micrographs of neuronal t-SNAREs deposited on the lipid membrane (B), and after the addition of v-SNARE (C). Section analysis of the SNARE complex in (B) and (C) is depicted in (D). Note that the smaller curve belonging to the t-SNARE complex in (B) is markedly enlarged after addition of v-SNARE. Artificial bilayer lipid membranes are nonconducting either in the presence or absence of SNAREs (E, F). Current verses time traces of bilayer membranes containing proteins involved in docking and fusion of synaptic vesicles while the membranes are held at -60 mV (current/reference voltage). (E) When t-SNAREs are added to the planar lipid bilayer containing the synaptic vesicle protein, VAMP-2, no occurrence of current spike for fusion event at the bilayer membrane is observed (n = 7). (F) Similarly, no current spike is observed when t-SNAREs (syntaxin 1A-1 and SNAP25) are added to the cis side of a bilayer chamber, following with VAMP-2. Increasing the concentration of t-SNAREs and VAMP-2 protein. [12].

 figure 14

 ..................................................

Fig. 15. Pore-like structures are formed when t-SNAREs and v-SNARE in opposing bilayers interact. (A) Unfused v-SNARE vesicles on t-SNARE reconstituted lipid membrane. (B) Dislodgement or fusion of v- SNARE-reconstituted vesicles with a t-SNARE-reconstituted lipid membrane, exhibit formation of pore-like structures due to the interaction of v- and t-SNAREs in a circular array. The size of the pores range between 50 and 150 nm (B-D). Several 3D AFM amplitude images of SNAREs arranged in a circular array (C) and some at higher resolution (D), illustrating a pore-like structure at the center is depicted. Scale bar is 100 nm. Recombinant t-SNAREs and v-SNARE in opposing bilayers drive membrane fusion. (E) When t-SNARE vesicles were exposed to v-SNARE reconstituted bilayers, vesicles fused. Vesicles containing nystatin/ergosterol and t- SNAREs were added to the cis side of the bilayer chamber. Fusion of t-SNARE containing vesicles with the membrane observed as current spikes that collapse as the nystatin spreads into the bilayer membrane. To determine membrane stability, the transmembrane gradient of KCl was increased, allowing gradient driven fusion of nystatin-associated vesicles [12].

 

figure 15

 ..................................................

Fig. 16. Opposing bilayers containing t- and v-SNAREs respectively, interact in a circular array to form conducting pores. (A) Schematic diagram of the bilayer-electrophysiology setup. (B) Lipid vesicle containing nystatin channels (red) and both vesicles and membrane bilayer without SNAREs, demonstrate no significant changes in capacitance and conductance. Initial increase in conductance and capacitance may be due to vesicle-membrane attachment. To demonstrate membrane stability (both bilayer membrane and vesicles), the transmembrane gradient of KCl was increased to allow gradient-driven fusion and a concomitance increase of conductance and capacitance. (C) When t- SNARE vesicles were added to a v-SNARE membrane support, the SNAREs in opposing bilayers arranged in a ring pattern, forming pores (as seen in the AFM micrograph on the extreme right) and there were seen stepwise increases in capacitance and conductance (-60 mV holding potential). Docking and fusion of the vesicle at the bilayer membrane, opens vesicle-associated nystatin channels and SNAREinduced pore formation, allowing conductance of ions from cis to the trans side of the bilayer membrane. Then further addition of KCl to induce gradient-driven fusion resulted in little or no further increase in conductance and capacitance, demonstrating that docked vesicles have already fused [12].

 

figure 16

 ..................................................

Fig. 17. Fluorimetric fusion assays demonstrate the ability of Ca2+ to induce rapid lipid mixing of plain (AV) and SNARE-associated vesicles. Addition of 5 mM Ca2+ to liposomal solution significantly increases the fusion of plain and SNARE-associated vesicles (+P0.1, Student t-test between AV and t- /v-SNARE-AV, n=5). Incorporation of t-/v-SNAREs at the vesicles membrane increases the overall yield but does not alter the rate of Ca2+-induced membrane fusion (A). The graph depicts the first-order kinetics of SNAREs vesicle fusion in the presence and absence of Ca2+ (B). [30].

 

figure 17

 ..................................................

Fig. 18. Conductance and capacitance measurements of SNARE-reconstituted lipid bilayers. The EPC9-electrophysiological set up is shown (A). In the presence of 5 mM EGTA, t-SNARE-associated vesicles containing nystatin channels at their membrane (represented as red structures at the vesicle membrane) interact with v-SNARE-reconstituted lipid bilayer without fusing (B). Note no change in conductance or capacitance following exposure of SNAREassociated lipid vesicles to the bilayer. The vesicles fuse, however, when 3 mM KCl is applied, demonstrating fusion of docked vesicles and presence of an intact bilayer (B). The green arrowhead indicates when the stirring is switched on to mix the addition. In the presence of 1 mM CaCl2, the t-SNARE-associated vesicles fuse with the v- SNAREreconstituted bilayer as depicted in a consequent increase in conductance and capacitance. Since a large majority of docked vesicles have fused, addition of 3 mM KCl has no further e.ect (C). Traces (B-C) are representative profiles from one of five separate experiments [30].

 

figure 18

 ..................................................

Fig. 19. AFM micrograph or live pancreatic acinar cell demonstrating the size of ZGs within the cell, in resting (A) and following stimulation of secretion (B). Note the increase in size of the same granules immediately following a secretory stimuli (section analysis of two such ZGs are shown). [39].

 

figure 19

 ..................................................

Fig. 20. Increase in size of ZGs in the presence of GTP. (A-C) Two-dimensional AFM images of the same granules after exposure to 20 mM GTP at time 0 (A), 5 minutes (B), and 10 minutes (C). (D-F) The same granules are shown in three-dimensions: the three-dimensional image of the granules at time 0, 5 minutes, and 10 minutes, respectively, after exposure to GTP. (G-I) The GTP-induced increase in size of another group of ZGs observed by confocal microscopy. Confocal images of the same ZGs at time 0, 5 minutes, and 10 minutes after GTP exposure are shown. (Bar = 1 μm.) Values represent one of three representative experiments. [31].

 figure 20

 ..................................................

Fig. 21. AQP1-specific antibody binds to the ZG membrane and blocks water traffic. (A) Immunoblot assay demonstrating the presence of AQP1 antibody in SLO-permeabilized ZG. Lanes: 1, AQP1 antibody alone; 2, nonpermeable ZG exposed to antibody; 3, permeable ZG exposed to AQP1 antibody. Immunoelectron micrographs of intact ZGs exposed to AQP1 antibody demonstrate little labeling (B and C). (Bar = 200 nm.) Contrarily, SLO-treated ZG demonstrate intense gold labeling at the luminal side of the ZG membrane (D and E). AQP1 regulates GTP-induced water entry in ZG. (F) Schematic diagram ofZGmembrane depicting AQP1-specific antibody binding to the carboxyl domain of AQP1 at the intragranular side to block water gating. (G,H, and K) AQP1antibody introduced into ZG blocks GTP-induced water entry and swelling (from G to H, after GTP exposure). (I-K) However, only vehicle introduced into ZG retains the stimulatory effect of GTP (from I to J, after GTP exposure). [32].

 

figure 21

 ..................................................

Fig. 22. Schematic outline depicting the regulatory pathway and the main players involved in water channel aquaporin-1 regulation. [33].

figure 22

..................................................

Source: J. Cell. Mol. Med. Vol 8, No 1, 2004 pp. 1-21


http://www.biology-online.org/articles/discovery_porosome_revealing_molecular/abstract.html