Basics of Electrodes for Single Cell Measurements

Basic Amperometry

Electrochemical detection of biogenic amines is typically carried out in either the amperometric or voltammetric mode. Voltammetry is discussed in the next section. Amperometric experiments involve holding the electrode at high enough potentials to oxidize molecules in the solution in a diffusion-limited fashion. In the amperometric mode, changing concentration of easily oxidized species results in a current vs. time trace that changes with the amount of oxidation. By integrating the current transient, it is possible via Faraday’s Law to determine the number of molecules oxidized (N = Q/nF, where N is the number of moles detected, Q is the integrated charge, n is the number of electrons transferred in the oxidation reaction, and F is the Faraday constant).

Amperometry has been used to study exocytosis in primary cultures [3,10,18,19], immortalized cell lines [20,21], brain slices [22], and at intact neurons in vivo [23,24]. In general, candidate cell systems have been limited to those that release an oxidizable substance, usually a catecholamine-, serotonin-, or tyrosine/tryptophan-containing peptide. Immortalized cultures allow for single, isolated cells to be studied, while primary cultures offer the advantage of modeling cells that can be studied in the context of belonging to a network, continuing to receive information from adjacent cells. Since amperometry can also be used within an intact brain, the technique can potentially be applied in a setting where cell communication remains relatively undisturbed. However, difficulty in production and placement of ultrasmall electrodes limits in vivo amperometry to measurements of the extracellular “overflow” catecholamine flux, not intrasynaptic concentrations. Ultrasmall electrodes have not yet been constructed to fit into a synapse. Amperometry is well suited to measuring secretion from cells because of its ability to quantify release from vesicles on a millisecond time scale (Figure 14.1) [25]. To carry out amperometry at single cells, a small electrode is placed near the cell and held at a potential where oxidation is diffusion limited. Stimulant is applied with a small pipette (Figure 14.1a) resulting in current transients for each exocytosis event. The first experiment to measure exocytosis used adrenal cells and detected the catechols norepinephrine and epinephrine [2,3]. The general oxidation reaction for catechols is shown in Figure 14.1b and a typical current–time trace is shown in Figure 14.1c. The number of molecules detected from a release event can be calculated using Faraday’s law (Figure 14.1d). Experimental determination of charge allows calculation of the number of moles of molecules detected by rearranging the Faraday equation. During these experiments, the electrode is positioned near a cell, where exuded neurotransmitter can be detected (Figure 14.1a). Calculations have been carried out that estimate that when the electrode is placed flush against the cell, there is a small solution filled gap of about 300 nm, between the cell and the electrode [26,27]. The characteristic shape of amperometric peaks, aids in understanding the various aspects of the exocytosis event. The peak shape is dictated by the flow of electroactive agents to the surface of the electrode. Since the electrode is held at a constant potential, 650 mV (vs. SSCE), the oxidation of the molecule generates a current that is measured by a potentiostat [3,28]. This current vs. time trace can be analyzed to provide information about exocytosis (Figure 14.1c). Specifically, the half-width of the peak (full width at half maximum) measures the duration of the exocytotic event. The rise time, or the time it takes to go from 10% to 90% of the peak height, relates to the time it takes the fusion pore to open. Thus, the basic amperometric measurement provides data on the amount of transmitter released, the duration of each event, and the opening of the fusion pore.

FIGURE 14.1

Basic introduction to amperometric detection of exocytosis at single cells. (a) Foreground image is the typical setup for amperometry for single cell. Exocytosis is stimulated by a pipette containing a stimulant, and the release is monitored by a carbon (more...)

Fast Scan Rate Voltammetry

Fast scan cyclic voltammetry can be used to provide information about the identity of the molecule in a solution by examining characteristic oxidation–reduction peaks [25]. Cyclic voltammetry has therefore been adopted in order to measure easily oxidized substances in and at single cells. In contrast to amperometry in which the signal is proportional to the total number of molecules oxidized, in cyclic voltammetry experiments, the signal is proportional to the local concentration of electroactive molecules; it has also been shown to have nanomolar detection limits. The advent of microelectrodes has lead to the ability to scan at much higher scan rates than with conventional electrodes. High scan rates provide a means to obtain voltammograms in 10 ms or less, thus, this procedure gives time resolution necessary to monitor rapid biological events. To detect catecholamines, the voltage sweep typically begins at − 0.4 V (vs. a sodium-saturated calomel reference electrode), ramps to + 1.1 V and then returns back to − 0.4 V. A typical scan rate is in the range of 800 V/s [25]. The rapid sweep rate creates a high background current and, therefore, background subtraction becomes very important in identifying what can often be a very small signal in a large background.

Exocytosis at Adrenal Cells

Adrenal chromaffin cells, which synthesize, store and release epinephrine and norepinephrine have been the most widely used cell model in amperometric studies to understand exocytosis. One interesting feature of these cells is a dense core at the center of the vesicles. This dense core is composed of, among other things, an anionic polypeptide called chromogranin A. The cationic messenger molecules are thought to associate with this polypeptide. During exocytosis, the vesicle fuses, beginning with the formation of a fusion pore. When the fusion pore forms, an influx of extracellular solution changes the pH of the vesicular solution. As the solution in the vesicle becomes more basic, chromogranin A releases the neurotransmitters associated in the matrix and they move into the synapse. Chromogranin A swelling has been hypothesized to force open the fusion pore during full exocytosis [29].

To distinguish and identify epinephrine and norepinephrine release at adrenal chromaffin cells, researchers have exploited the chemical identification ability of fast scan cyclic voltammetry. Epinephrine and norepinephrine differ only by a methyl group on the amine side chain and have identical electrophores (catechol) and electrochemical formal potentials for oxidation. Hence, they are difficult to discriminate from one another. However, cyclic voltammetry can be used to discriminate these substances if one examines the re-reduction of oxidized catechol and considers that in the oxidized state these molecules undergo a cyclization reaction involving the nitrogen ending side chain to form an adrenochrome that is more difficult to reduce, resulting in a second reduction peak. The reaction rate for epinephrine is considerably faster than that for norepinephrine, hence, the adrenochrome reduction peak is larger for epinephrine. Formation of the adrenochrome from oxidized epinephrine also shifts the equilibrium potential for oxidation, providing a second means to discriminate it from norepinephrine. Figure 14.2 shows representative time traces obtained at a single adrenal cell following stimulation with nicotine [30]. In this case, the stimulated cell releases both epinephrine and norepinephrine as indicated by the voltammetry at 10 V/s. Examining several cells, 75% of the cells release primarily either norepinephrine or epinephrine, while 25% release both.

FIGURE 14.2

Cosecretion of epinephrine and norepinephrine from an adrenal cell. (a) Oxidation current (solid line), ΔE p (dotted line), and adrenochrome current (dashed line) vs. time. (b) E p (solid line), ΔE p (dotted line), and i cyc/iox (dashed (more...)

Amperometric Detection of Release via the Fusion Pore

It has been suggested that docked vesicles undergoing exocytosis transcend an intermediate state when a small fusion pore is open but cannot immediately expand (for a discourse on the forces stabilizing this fusion pore see Ref [31].), and that this pore allows a small but measurable amount of neurotransmitter to leak out [32]. Transport of chemical messengers through the constricted pore manifests itself as a pre-spike feature or “foot” in the amperometric trace as these messengers are oxidized at the electrode. Fusion pore formation appears to be highly regulated, as numerous proteins have been implicated in this step in exocytosis [33]. Therefore, it is of interest as a mechanism by which the cell could regulate release of neurotransmitter and be involved in synaptic plasticity.

A combination of patch clamp capacitance measurements with carbon fiber amperometry was used to correlate the fusion of individual vesicles with the release of neurotransmitter [32]. Alvarez de Toledo and co-workers showed that membrane capacitance appeared to transiently increase and decrease prior to the detection of amperometric events. They attributed this phenomenon to transient, reversible vesicle fusion prior to subsequent full fusion. Furthermore, they showed that the duration of the foot correlates with vesicle radius, and that the percentage of neurotransmitter released in the foot increases with decreasing vesicle size [32]. Amperometric methods were used to quantify the amount of neurotransmitter released at various points during each capacitance trace on a sub-millisecond time scale.

Insulin Release from Pancreatic Beta Cells

Although most experiments apply amperometry to catecholaminergic cells, it is not limited to them. Kennedy’s group has shown that amperometry can also be used to monitor the release of peptides from cells via the same mechanisms. Highly sensitive detection (detection with a limit of 100 n M ) of insulin (a peptide with 51 amino acids) from human pancreatic beta-cells required coating the electrode with a ruthenium oxide/cyanoruthenate film [34]. Smaller tyrosine/tryptophan-containing peptides (with 8–15 amino acids) have been detected with uncoated carbon fibers [35–37], albeit at higher limits of detection.

It is believed that peptides are released by large dense core vesicles (LDCVs), while classical neurotransmitters are sequestered in small synaptic vesicles (SSVs) in the brain, making peptide detection potentially more feasible. Larger vesicles give rise to larger and more easily detected peaks. However, those neuropeptides, lacking oxidizable amino acid residues in their composition, would not be detected.

Exocytosis at Mast Cells

As part of the immune system, peritoneal mast cells secrete histamine and serotonin during an allergic response. Mast cells extracted from one variety of mouse, the beige mouse, have unusually large vesicles (mast cell vesicle diameters have been reported as 770 nm for normal mice and 2000 nm for beige mouse [32,38]). Beige mouse mast cells are an excellent model system for performing dual simultaneous measurements of exocytosis using amperometry and patch clamp for capacitance [32,38]. Individual vesicles are large enough to be monitored by light microscopy, and fusing of single vesicles generates discrete steps in the capacitance trace. Beige mouse mast cells have a smaller intravesicular “halo”-like lumen surrounding the core than those seen in other cell types (i.e., PC12 cells and “active” vesicles from chromaffin cells [19]) but respond to many of the same agents that affect exocytosis from chromaffin cells [39].

Previous data have suggested that histamine and serotonin are co-released from single vesicles in mast cells; however, the evidence for this was indirect. Using electrochemically pretreated carbon fiber electrodes and a high positive potential scan limit, two oxidative waves for histamine were observed, as well as an oxidation wave for serotonin [40]. This allowed the simultaneous detection of the two analytes, leading to the conclusion that indeed both histamine and serotonin are released from the same vesicles in mast cells.

Pheochromocytoma (PC12) Cells, Undifferentiated and Differentiated

The adrenal pheochromocytoma (PC12) cell line was originally isolated from a tumor in the adrenal medulla of a rat in 1976 [41]. They resemble the phenotype of sympathetic ganglion neurons upon differentiation (with nerve growth factor, NGF) and can be subcultured indefinitely. They possess slightly smaller LDCVs (75–120 nm radius compared to 170 nm [16,41,42] for chromaffin cells) and generally contain dopamine, which gives rise to spikes similar to, but smaller than, those of isolated chromaffin cells following stimulation to exocytosis [43].

The first demonstration of zeptomole detection was shown by Chen et al. for amperometric detection of dopamine during exocytosis at undifferetiated PC12 cells [20]. The average single PC12 vesicle from this study contained 190 zmol of catecholamine. Using previous electron microscopy data [42] that estimated the average vesicular radius to be 74 nm; the concentration of catecholamine in a single vesicle was estimated to be 0.11 M . Chen et al. reported a limit of detection of 31 zmol with the amperometric technique, which is within the range needed to detect exocytosis at mammalian neurons. Mammalian neurons contain between 1.8 and 96.4 zmol of catecholamine in a synaptic vesicle [44].

The minute dimensions of microelectrodes have been exploited to explore zones of exocytotic release on differentiated PC12 cells. Zerby and Ewing used 5 μm diameter carbon fiber electrodes to probe local zones of dopamine exocytosis at PC12 cells that were induced to sprout varicose neurites via exposure to NGF [45]. These experiments showed that quantal size from differentiated PC12 cell varicosities is not significantly different from undifferentiated PC12 cell quantal size, and exocytotic events were absent from the cell body of the NGF treated cells.

Release from the Cell Body of a Processed Neuron in an Animal System

Invertebrate systems offer an advantage as a model for neuronal communication in the form of larger cell dimensions than their mammalian counterparts. The Nobel prize research of Hodgkin and Huxley on membrane potential would not have been possible without the macroscopic squid axon as a research tool [46]. More recently, Planorbis corneus (the freshwater pond snail) has been used to study exocytosis. Planorbis has a giant dopamine cell from which tens of thousands of events can be elicited per stimulation. These events are derived from LDCVs and putative small synaptic vesicles (SSVs) [23,47,48]. This was pioneering work to measure individual exocytosis events from an intact neuron in a living organism.

Release of Small Amounts of Catecholamine from the Leech Neuron

The Retzius cell of Hirudo medicinalis (the freshwater leech) has also been used to study exocytosis. This cell synthesizes, stores and releases serotonin. Bruns and Jahn detected release from two populations of vesicles (large and small) from this cell type, with the current transients from the small population averaging 3.0 fC in area (about 5000 molecules) and 595 μs wide, while the larger vesicles had an average area of 47.4 fC (about 1million molecules) and width of 3.66 ms [49].

Release from Mammalian Cells

Neurons used to study exocytosis with amperometry have been isolated from the midbrain, hippo-campus, retina, primary sensory ganglia, neostriatum, nucleus accumbens, and substantia nigra [15,18,19,50–54]. Unlike immortalized cell lines, which can be used within a few days of plating, primary neuronal cultures require 3–6 weeks of recovery after plating [15]. In addition, primary cultures are a heterogenous phenotype requiring post-experiment immunohistochemistry to confirm the presence of critical enzymes (i.e., tyrosine hydroxylase) for phenotype identification. The many brain areas, thus far examined in this way, share one characteristic: they are known to have a high density of catecholamine releasing cells. In the future, tailored genotypes of mice expressing GFP in catecholaminergic cells will allow amperometry to be applied to areas where such cells are diffuse.

Kiss-and-Run Measurements

In some cases, during exocytosis, the fusion pore is thought to open and close without a full exocytosis event occurring. This phenomenon, termed kiss-and-run has recently been observed with amperometry at mammalian cells by Sulzer’s group [55]. If these events represent a significant communication pathway, then the cell’s ability to modify vesicle parameters that affect the fusion pore could be a mechanism of neuronal plasticity (learning and memory).

Measurements of Cholesterol in the Cell Membrane

Although enzyme-modified electrodes are not being discussed explicitly in this review, a new application of them warrants mentioning. Devadoss et al. have developed a cholesterol oxidase modified electrode that has been used to detect cholesterol in the plasma membranes of Xenopus oocytes [56]. A platinum electrode modified with a lipid bilayer containing cholesterol oxidase [57] is operated in the amperometric mode and placed adjacent to an oocyte (Figure 14.3a and Figure 14.3b). As the electrode moves closer to the cell surface, the amperometric current rises (Figure 14.3c). When the electrode is placed in contact with the cell surface, cholesterol is extracted from the cell membrane, moves across the thin membrane layer, and partitions into the electrode-supported enzyme-modified lipid bilayer. Subsequent enzyme catalyzed oxidation of cholesterol by molecular oxygen produces hydrogen peroxide, which is oxidized at the platinum electrode and detected. This is an efficient and simple method to determine cholesterol levels in the plasma membrane of cells, which should find applicability in determining the role cholesterol plays in the organization of cellular membranes [58].

FIGURE 14.3

( See color insert following page 272.) Photographs and amperometric detection of cholesterol in the plasma membrane of a Xenopus oocyte. Photographs showing the electrode (a) positioned about 5 μm from the plasma membrane and (b) contacting the (more...)

Patch Amperometry

Although electrochemical methods alone are powerful tools for the detection and analysis of single exocytotic events, they can be combined with electrophysiological methods, patch clamp for instance, to provide more detailed information about the events. This technique is known as patch amperometry. Patch amperometry is typically accomplished by inserting a carbon fiber, which functions as the electrochemical sensor, into a glass patch clamp pipette and performing cell attached voltage clamp experiments simultaneously with electrochemical experiments. This is shown in Figure 14.4a through Figure 14.4c.

FIGURE 14.4

(a) Arrangement of a CFE inside a patch pipette. I A , Amperometric current; I C , sine wave current used to measure capacitance changes. (b) and (c) Chromaffin cell with attached patch pipette containing CFE at the beginning (b) and end (c) of the experiment. (more...)

This configuration allows the simultaneous monitoring of membrane capacitance, as well as detection of the opening of individual fusion pores and the monitoring of catecholamine release kinetics from the same vesicle [59]. Membrane capacitance is directly proportional to membrane area; therefore, discrete jumps in capacitance signify fusion of individual vesicles with the plasma membrane and allow estimation of the size of the fusing vesicles (Figure 14.4d). These experiments were used with currently accepted models of the structure of the fusion pore to estimate the diameter of the pore preceding full release in chromaffin cells to be less than 3 nm. Additionally, this technique was used to determine that release of transmitter through the pore in adrenal cells is much faster than release through the pore of mast cells. In other studies on mast cells, Tabares et al. demonstrated that the size of the initial fusion pore does not determine the rate of neurotransmitter efflux during the early or late stages of full exocytotic release [60]. This technique has also been used to elucidate the coregulation of vesicle membrane area and the amount of neurotransmitter contained in vesicles in response to pharmacological treatment [61]. Patch amperometric experiments have reinforced the conclusion put forth by Colliver et al. that neurotransmitter concentration in vesicles remains constant when vesicles are induced to grow or shrink by drug treatment [21]. While the cell attached configuration is the most prevalent configuration, patch amperometry has also been applied to excised membrane patches, thereby allowing control over the environment on both sides of the plasma membrane [62].

Quartz Crystal Microbalance Measurements

The development of the quartz crystal microbalance (QCM) started from the predictions of Sauerbrey in the 1950s that the physical nature of a quartz crystal resonator results in a proportionality between the resonant frequency and the total mass of the crystal [63]. Therefore, a change in the crystal mass can be detected by measuring the changes in crystal resonant frequency. Further development of QCM equipment included the electrochemical quartz crystal microbalance (E-QCM), which is a widely used instrument for measurement of attached surface mass to the front side of the quartz crystal, as a working electrode [64]. This has been used for simultaneous detection of ion transfer processes that occur in adsorbed films. This has been a very useful technique in studying electrochemical deposition, etching, and electropolymerization. A QCM combined with a dissipation measuring (QCM-D) technique has been developed [65]. This approach involves simultaneous measurement of the dissipation and the resonant frequency. The QCM-D technique allows the study of surface phenomenon of adsorbed non-rigid materials in cases where the Sauerbrey equation is not entirely valid, such as living cells. Measurements of the mechanical properties of cells can also be achieved including viscosity, elasticity, density, and thickness of the adsorbing material. Cans and co-workers used the QCM-D method to monitor the mass change and rigidity of populations of excitable cells during exocytosis and endocytosis. NG108-15 and PC12 cells were grown to confluence separately on piezoelectric quartz crystals and were stimulated to exocytosis using elevated potassium solution [66]. This resulted in an increase in the frequency response corresponding to loss of mass as vesicles underwent release and a subsequent mass gain during retrieval of dense-core vesicles. The two cell lines were examined separately and used to demonstrate differences in release and retrieval for cells of different morphology, size, and number of dense-core vesicles. The amplitude and peak area in the frequency response corresponding to mass change with stimulated release was larger for PC12 cells than for NG108-15 cells, suggesting that a greater number and larger size of vesicles in PC12 cells results in a greater amount of release from these cells vs. NG108-15 cells. However, the initial rate constants for the frequency responses were similar, suggesting that the recycling of vesicles utilizes similar fusion/retrieval mechanisms in both cell types. The QCM-D measurements also revealed that vesicle retrieval had a rapid onset, masking exocytosis, and implying a rapid retrieval mechanism in the early stages of release. This demonstrates that complex, dynamic processes relating to dense-core vesicle release and retrieval can be measured simultaneously using the QCM-D technique. The QCM technique, with further development using QCM in an array format and with miniaturization of individual resonators, might offer a promising approach to study the processes of neuronal communication in neuronal networks.

Electrochemistry of Material Released from Cells in Microvials

Electrochemical detection of cellular behavior in small volume containers requires fabrication methods capable of defining micrometer size structures from a suitable substrate. Initially, Bowyer et al. sandwiched silver and platinum foil between layers of Tefzel film and glass to form an electrochemical cell with auxiliary, reference, and working band electrodes, the smallest of which measured 4 μm thick [67]. This electrochemical cell was used to investigate the differential pulse, normal pulse, and cyclic voltammetric behavior of ferrocence in aqueous solutions with volumes ranging from 2 mL to 50 nL.

More recently, however, photolithography has been employed to define and fabricate vials with a capacity as small as 1 pL [68,69]. Small vials, described by Clark et al., have been fabricated using a silicon template photolithographically, then transferring the template pattern into polystyrene using a hot press method. This results in arrays of transparent microvials. An alternative approach to producing transparent sample vials has been taken by Bratten and co-workers wherein a polyimide chamber is photolithographically defined on a transparent glass slide [70–72]. These structures differed from those of Clark and Ewing in that the reference, auxiliary, and working electrodes were formed by in situ gold deposition on the bottom surface of the container.

A third method of vial fabrication uses a combination of screen printing and laser ablation [73,74]. Briefly, carbon ink was screen printed through a patterned stencil onto a ceramic substrate. An insulating, dielectric ink was screen printed on top of the carbon layer and a Ag/AgCl ink was printed on top of the dielectric ink. A final layer of dielectric ink was printed on the outermost layer. After laser ablation, the carbon and Ag/AgCl layers eventually formed the working and reference electrodes of the device. Formation of the vials was accomplished by Kr–F excimer laser ablation at 248 nm and resulted in vial with a volume of 7.2 nL. Figure 14.5 shows scanning electron micrographs of the three types of vials described above.

FIGURE 14.5

Scanning electron micrographs of three types of microvials. The leftmost images are polystyrene microvials fabricated with a hot press method. The center images are vials photolithographically defined polyimide wells. The rightmost image is a vial prepared (more...)

There are two prominent electrochemical detection schemes when working in small volume vials. The simplest is to use micromanipulators to position the working and reference electrodes into the vial. This is the approach taken by Clark and Ewing, as well as some groups conducting biological electrochemical experiments in microvials, which will be discussed later in this section. Secondly, electrodes can be formed in situ in the chambers by traditional microfabrication and photolithography techniques [70–72,75] or by a combination of screen printing and laser ablation [73,74]. The drawback to some of these fabrication methods is that the materials that form the sides and bottom of the vials are not always the same, which could lead to differential adsorption of analytes and erroneous voltammograms.

Microvials are attractive environments for single cell experiments because of their limited volumes and the possibility of creating integrated devices with electrodes fabricated by photolithographic techniques inside the vials. Such studies include the monitoring of purine release during cardiac cell ischemia [72], monitoring metabolic flux due to stress response [76], and monitoring dopamine transport into cells [77]. Microvials with spatially distinct electrodes have been fabricated by Dias et al. and have been used to determine the spatial distribution of exocytosis events on the surface of adrenal chromaffin cells [75]. In a study by Troyer and Wightman, HEK-293 cells transfected to express the dopamine transporter were placed in vials ranging from 100 to 200 pL [77]. Dopamine was then injected into the vials and the rate of clearance of dopamine by the cell was monitored with fast scan rate cyclic voltammetry. Monitoring dopamine concentration change with time allows the calculation of the kinetics of the dopamine transporter, specifically the initial transport rate, V _max. Using the single cell microvial approach, V _max was determined to be 55 ± 17 amol/s-cell. In populations of cells, V _max was determined to 18.9 ± 1.4 amol/s-cell by rotating disk electrode voltammetry [78]. The discrepancy in V _max measured by competing techniques may be due to the difference in the cell-volume/solution-volume ratio. In the microvial experiments, this ratio is larger, meaning, there is less extracellular volume per cell and thus, fewer dopamine molecules are transported into the cell. Fewer transported molecules generally prevents the cytosolic concentration from reaching a high enough concentration for reverse transport to occur. However, in some cells in microvial experiments, reverse transport did occur, and was determined by a nonzero dopamine concentration plateau. Cells exhibiting this behavior were not taken into account in the calculation of V _max, as the calculated V _max from these experiments appeared lower.

Artificial Cells (Liposomes) as Models of Exocytosis

Exocytosis is a complex process that involves protein–protein and protein–lipid interactions, membrane fusion, and release of neurotransmitter. A model system that mimics this process in a protein free fashion is advantageous to study the role of the lipid bilayer in exocytosis. A totally lipidic system is useful to study the effects of altered membrane physical properties, typically accomplished by altering the components that form the membrane. Thus, an “artificial cell” composed of two liposomes with a lipid membrane nanotube connection, and one inside the other (Figure 14.6), was developed [79].

FIGURE 14.6

( See color insert following page 272.) Formation and release of vesicles in an artificial cell. (a) through (d) Schematics of a microinjection pipette electroinserted into the interior of a unilamellar liposome and then through the opposing wall, pulled (more...)

The liposome inside a liposome configuration is analogous to a vesicle inside a secretory cell during the last stages of the release process. Microelectroporation [80] assisted insertion and careful micromanipulation of a micropipette is required for formation of the artificial cell. Briefly, a micropipette filled with the redox molecule of choice (typically catechol) is positioned next to a surface immobilized giant unilamellar vesicle (GUV) with an attached multilamellar vesicle. At the opposite side of the vesicle is the electroporation counter electrode. When a voltage pulse is applied across the GUV, the membrane is transiently destabilized, allowing insertion of the micropipette. The lipid membrane is allowed to seal around the micropipette tip before the pipette is rapidly withdrawn into the interior. This forms a nanotube connecting the tip and the liposome. When solution flows from the micropipette into the lipid nanotube, a small vesicle begins to form. When the vesicle has grown to the size where it nearly contacts the larger exterior vesicle, there is only a short nanotube separating the two liposome compartments. This geometry mimicks that of a cell undergoing the later stages of exocytosis, with a small vesicle fused to a larger vesicle and an aqueous pore connecting the vesicular contents with the extracellular space. At this point, the remaining toroidal shaped fusion pore of the artificial cell is of an energetically unfavorable lipid membrane geometry, and this results in spontaneous opening of the lipid pore and the interior vesicle emptying its content to the surrounding solution in a process that resembles exocytosis. The nanotube remains adhered to the pipette tip, allowing successive release events without pipette reinsertion. The size of the fusing vesicle can be adjusted by varying the distance from the pipette tip to the large vesicle membrane, and release from vesicles ranging from 4 μm in diameter to larger than 30 μm has been measured.

Kinetics of Membrane Distention during Artificial Cell Exocytosis

Monitoring the material released is accomplished by amperometrically detecting the redox molecule contained in the smaller interior vesicle as it is released. Typically, a carbon fiber electrode measuring 33 μm in diameter is employed for amperometric detection. The resulting current transient appears qualitatively similar to those recorded at living secretory cells, such as, mast cells, chromaffin cells, and PC12 cells (Figure 14.7a and Figure 14.7b). The obvious difference in time scale, amplitude, and area of the artificial cell current spike is due to the fact that the vesicle is larger and contains more electroactive molecules than those in living cells. However, release from the smallest vesicles measured (approximately 4 μm diameter) is similar in time course to events measured from the large vesicles of the beige mouse mast cell (which average 700 nm in diameter) [81]. In Figure 14.7c, the relationship between vesicle radius and full width at half maximum (half width) is shown for experiments with artificial cells. The fit is nearly perfectly cubic, meaning that release kinetics scale linearly with vesicle volume.

FIGURE 14.7

Amperometric monitoring of repeated exocytosis events at artificial cells and cells. (a) Amperometric detection of continuous exocytosis of three vesicles from an artificial cell. (Scale bars are 40 pA and 3000 ms.) (b) Amperometric detection of dopamine (more...)

Modeling Release via the Fusion Pore (Lipid Nanotube)

An interesting feature of this system is the transport or leakage of catechol through the nanotube prior to full release. This can be thought of as analogous to neurotransmitter leakage through the fusion pore formed before full release in living cells [8,32]. Trans-pore catechol transport is apparent on the amperometric traces as a prespike rise in current, which in live cells is termed a “foot” (Figure 14.8a and Figure 14.8b).

FIGURE 14.8

Amperometric monitoring of release via an artificial fusion pore. (a) Amperometric detection of release from a vesicle with a radius of 5 μm, showing prespike feet (arrows), indicating catechol transport through the lipid nanotube or fusion pore. (more...)

This transport was characterized at artificial cells as a function of vesicle size and as a function of the pressure applied to the micropipette used to inflate the vesicles (Figure 14.8c through Figure 14.8e). The pressure applied to the inflation micropipette correlates with the flow rate of catechol solution from the micropipette. The duration of the foot is strongly dependent on vesicle size for larger vesicles (size of vesicle just before release) and is also dependent on the flow rate. However, the number of molecules detected during the foot portion of the event (foot area) is not dependent on solution flow rate. The ratio of area to the duration of the foot is dependent on both vesicle size and flow rate. These electrochemical data, coupled with models of membrane and fluid dynamics, lead to a cogent explanation of leakage through the membrane nanotube (Figure 14.8f). As the vesicle is inflated, its volume grows at a constant rate, whereas the rate of surface area growth slows as the vesicle grows larger. Additional membrane required to accommodate the growing volume is drawn along the nanotube, thus inducing shear flow of the solution inside the nanotube. Consequently, shear flow has a high velocity when the vesicle is small, and slows as the surface area to volume ratio decreases. Opposing shear flow is Poiseuille flow resulting from the pressure difference across the nanotube. Additionally, as the vesicle grows, the nanotube connecting it to the larger vesicle shrinks, and as a result, Poiseuille flow velocity increases. When Poiseuille flow overcomes shear flow, catechol solution is transported out of the nanotube and is amperometrically detected as a prespike foot.

In this sense, transport through the nanotube is analogous to leakage through an elongated fusion pore in living cells. Intravesicular pressure and membrane transfer through the fusion pore have been suggested previously as significant driving forces for exocytosis [33,82–84]. Finally, the most significant finding from this work is that membrane mechanics alone seem to be sufficient to drive exocytosis-like behavior at biological time scales without protein intervention.

Artificial Synapses with Liposome Models

In a separate set of experiments, the artificial cell model was used to develop models of coulometric efficiency as a function of electrode size and to determine the size of the space between the electrode and the membrane [85]. In this case, coulometric efficiency is defined as the ratio of the total number of molecules detected from a vesicle to the total number of molecules released by the vesicle. These studies employed two different size electrodes (5 and 33 μm in diameter) to detect release and compared the electrochemical responses as a function of electrode and vesicle size.

Model of the Synapse

When a vesicle is released from the artificial cell, its size relative to the detection electrode is a strong determinant of the amount of vesicular content detected, as well as the spike amplitude. However, the measurement of the kinetics of release is not affected by the size of the electrode used for detection. To model coulometric efficiency and determine the size of the electrode membrane gap, a simple geometric representation of the system was considered (Figure 14.9). The space between the electrode and the membrane was modeled as a cylinder, and the ratio of the vesicle volume to cylinder volume determines the coulometric efficiency. If the vesicular volume is smaller than the cylinder volume, 100% oxidation is expected. For an electrode measuring 33 μm in diameter, the best fit of the data is obtained when the height of the cylinder is 300 nm, suggesting that even when the electrode is in direct contact with the membrane there is a small volume of solution trapped between the electrode and the membrane. It should be pointed out that even for somewhat smaller vesicles, a significant portion of the released molecules escape undetected. This is probably caused by solution flow out of the membrane–electrode gap due to volume limitation not diffusion, i.e., the volume of the gap cannot accommodate the total volume of liquid released from the vesicle.

FIGURE 14.9

Simple model of coulometric efficiency for artificial exocytosis. This first-stage model assumes that the efficiency of oxidation for material released is simply the ratio of the membrane-electrode space (calculated as φr_e² h ) over the volume (more...)

The first model presented in this work was quite simplistic as it did not take into account the distinct stages of release that are characteristic of the artificial cell model. Therefore, a second model with distinct stages of release was proposed; each stage was represented as having a distinct geometric shape (Figure 14.10). In the first stage of release, the nanotube begins to dilate and some vesicular material escapes the vesicle as revealed by fluorescence imaging of release events (Figure 14.10b). The next stage is characterized by a membrane structure that is between a spherical vesicle and a frustum (Figure 14.10c). The final stage is after full release, when the small vesicle has become fully incorporated into the larger liposome membrane and the nanotube remains attached to the pipette tip (Figure 14.10d). During the time the vesicle is in the frustum state (calculated as twice the spike half-width for each vesicle) the charge passed is estimated from the integrated Cottrell equation.

FIGURE 14.10

A more complete model of the release process during exocytosis basing on the fluorescence observations. This model assumes that as the vesicle opens, it has a transitory period where mass transport of catechol to the electrode is via diffusion from a (more...)

where n is the number of electrons transferred in the electrochemical reaction, F is Faraday’s constant, A is the effective electrode area which is taken to be the area defined by the opening of the frustum or the electrode area (which ever is smaller), D is the diffusion coefficient (6× 10⁻⁶ cm² /s), C * is the concentration of catechol in the vesicle, and t is the time the frustum is open (taken to be twice the peak half-width). During the next stage, after full exocytosis, the charge passed by oxidation of catechol in the electrode-membrane space is given by the equation

where r_e is the average radius of the beveled electrode, h is the distance between the membrane and the electrode, and C′ is the concentration of the remaining unoxidized catechol in the membrane– electrode space, which is calculated by subtracting the number of moles oxidized during the previous (frustum) stage by the equation

where V_v is the volume of the vesicle and V _flow is the volume of flow out of the pipette during the release event. Flow during the event is calculated by dividing the vesicle volume by the time between events to get a volume flow rate in cm³ /s then multiplying by twice the event half-with. Combining Faraday’s Law (N = Q/nF) and the integrated Cottrell equation, Q = 2nFAD ^1/2 C * t ^1/2/φ ^1/2 (Equation 14.1), the number of moles oxidized during the frustum stage is given by

Oxidation of all the catechol in the vesicle prior to release results in the following coulometric charge:

So, the predicted coulometric efficiency is the ratio of the sum of the charge from the two stages of opening over Q _max. The resulting equation that predicts percent coulometric efficiency is given as

Only 25% of the Cottrell equation is used in the final expression to correct for the diminished diffusion in the narrowing frustum. This model correctly predicts the shape of the percent coulometric efficiency vs. vesicle radius; however, the predicted magnitude is 2.6times larger than that observed. To correct for this a factor of 0.38 has been applied to the theoretical prediction. Experimental data, predicted coulometric efficiency and the best fit of the experimental data is shown in Figure 14.11.

FIGURE 14.11

Coulometric efficiencies for a data set obtained with a beveled 33 μm electrode (♦) compared to data obtained with a 5 μm electrode (▪) for release of catechol measured from a range of vesicle sizes. This is compared to (more...)

Stages of Release during Exocytosis Elucidated with Amperometry

Exocytosis of dense core vesicles occurs in stages with the opening of the fusion pore, expansion to full distention, and dissociation of transmitter from the internal dense core matrix all playing important roles. All of these have been investigated with amperometry experiments. The rising portion of a current spike has been correlated to the opening of the fusion pore. Based on models that predict steeper rise times for current spikes than observed experimentally, the Amatore and Wightman groups conclude that the rise time is “due to a separate kinetic step that is temporally located between the initially formed fusion pore, where the majority of catecholamines are tightly associated with the matrix, and the final stage, where the matrix is dissociated and release of catecholamines has fully developed” [89]. Using this interpretation of the rising portion of the spike, it is therefore feasible that some manipulation of the fusion pore and/or rate of catecholamine –core-dissociation may result in changes to spike rise time.

Prespike foot-like events are elicited under hypertonic conditions in chromaffin cells [90,91]. During vesicle fusion, the dense core degranulates and releases a large volume of bound neurotransmitter, producing a spike in the amperometric trace. However, under hyperosmotic (approximately 730 mOsm, compared to the normal 330 mOsm) conditions, the core appears to remain intact and only a small amount of uncomplexed neurotransmitter in the halo is released, producing flat, broad peaks about 1% of normal peak amplitude, as measured by cyclic voltammetry. These events resemble feet without associated peaks [91]. Although stimulation elicits a few small foot-like events, momentarily restoring isotonic conditions produces the massive release expected for full exocytosis from many vesicles. This is purportedly because vesicles are stalled in mid-fusion until the osmotic gradient is restored. Isotonic restoration-induced release is not accompanied by a calcium rise, supporting the idea that vesicles have already fused but cannot expel their contents. Like the patch–clamp and amperometry measurements of Alvarez de Toledo et al., the amperometric feet observed under hyperosmotic conditions appear to result from traces of neurotransmitter diffusing through the fusion pore prior to full fusion. Osmolarity is thus presented as a tool for separating initial vesicle fusion from full fusion [91].

In a follow-up to Alvarez de Toledo’s conclusion that foot duration varies inversely with vesicle size, Sombers et al. presented evidence that pharmacologically altered vesicle sizes exhibit different sized feet in PC12 cells [92]. Decreasing the amount of dopamine and the vesicular volume by treating with reserpine results in a greater portion of vesicle contents released through the constricted fusion pore, which manifests itself as a foot in the amperometric trace. Conversely, loading vesicles with L-DOPA to increase the dopamine content and their volumes, leads to a smaller fraction of total vesicular contents released through the fusion pore [92]. Evidence that a greater fraction of vesicle contents is released through a constricted fusion pore as vesicular volume decreases is consistent with recent reports that a significant portion of neurotransmitter released at synapses is through a “kiss-and-run” mechanism [93,94]. “Kiss-and-run” exocytosis involves the expulsion of neurotransmitter through a transient, constricted fusion pore without full fusion, similar to the “foot” portion of amperometric events; however, this is hypothesized to take place under physiological conditions and without full exocytosis occurring.

Theories on the Forces Involved in Opening the Fusion Pore

The rate of release of transmitter during full exocytosis is dependent on the rate of opening of the fusion pore. The dynamics of the pore have been studied by both patch–clamp [7,62,95] and amperometry [2,8,96]. Based on amperometric studies of exocytosis from adrenal chromaffin cells, a physicochemical model accurately describing the dynamics of full-fusion events from dense core vesicles has been proposed [31,82,97]. This model predicts that the vesicular matrix has a determining role in the opening of the vesicle during the final stage of exocytosis. As soon as the fusion pore opens, the concentration gradient of catecholamine cations (mostly epinephrine in chromaffin cells) provokes their outward diffusion. To maintain electroneutrality within the matrix, this is accompanied by the entry of extracellular, hydrated, monovalent cations, such as Na⁺ and H ⁺ , into the intravesicular matrix. It is usually considered that during this phase, the fusion pore has not yet expanded, but rather fluctuates between open and closed phases [8,32]. Its geometry may be imposed by a specific architecture, which is still a matter of debate, possibly implicating SNARE protein complexes, or phospholipids, or a combination of both [98–100]. However, at this step, it may be considered that the pore energy is necessarily controlled by two opposite phenomena. The first one is positive and tends to close the pore. It reflects the edge energy of the pore rim. The second one is negative and tends to open the pore. It stems from Laplace tensions of the cell and vesicle membranes [31,82,97,101]. Initially, the edge energy is larger than the surface energy so that the pore remains stable. Owing to the continuous cation exchange, which occurs as soon as the pore is opened, the matrix swells. In dense core vesicles, the matrix expansion likely increases the pressure within the vesicle and provokes a continuous increase of the vesicle membrane tension. Unless the pore closes before, the surface energy of a toroidal pore ineluctably increases until reaching the point where it compensates for the pore edge energy [31,82,97]. When this point is reached, the pore becomes unstable and enlarges irreversibly [31,82,97,102,103].

The pore expansion initiates a massive release of catecholamines reflected by a sharp increase of current detected at the electrode immediately after the foot current. The ensuing membrane fusion ends by fully exposing the vesicular matrix to the external medium. Afterwards, the progressive depletion of catecholamines by diffusion is observed by a decrease in current. This scheme is proposed for the exocytosis of dense core vesicles that constitute the main category of vesicles within chromaffin cells. Vesicles without a dense core, as in nerve terminals [104], or with a smaller dense core versus the total vesicle volume, as observed in PC12 cells [21], may proceed differently in the fusion pore expansion. The main difference should lie in the physicochemical origin of the building up of the internal pressure within the vesicle and, consequently, of the surface tension of the fusion pore.

Effects of Snare Protein Manipulation

Amperometry on PC12 cells has been used to elucidate protein interactions that regulate cellular machinery. The part of the cellular machinery that recruits vesicles to the plasma membrane is called the SNARE (SNAP receptor, where SNAP is soluble N -ethylmaleimide-sensitive factor-attachment protein) complex [105]. A syntaxin-binding protein, Munc18, has been shown to associate with the SNARE complex and is thought to play a role in regulating exocytosis [106]. Previous models have shown that Munc18 dissociates from syntaxin 1 prior to SNARE-complex formation and, therefore, it cannot exert any control on the later stages of exocytosis.

Based on data presented regarding a point mutation in a Munc18 homologus in Drosophila melanogaster [107], Burgoyne et al. transfected PC12 cells with a mutated Munc18 protein [105]. Munc18 was believed to dissociate from syntaxin, and arginine residue 39 was identified to interact directly with syntaxin. The arginine residue was changed to a cysteine to reduce this interaction [108]. Cells transfected with the mutation in Munc18 displayed faster vesicle fusion during exocytosis. This was noted through a decrease in both amplitude and half width of the amperometric signals. The authors suggested that the reduction in amperometric signal resulted from kiss and run fusion. The conclusion drawn from this work is that Munc18 interacts with the SNARE complex and augments the later stages of exocytosis [105]. It is possible that opening of the fusion pore is kinetically hindered by the SNARE complex needed for its formation. Free Munc18 after dissociation from syntaxin might facilitate disassembly of the SNARE complex once the pore is formed, thereby leading to faster opening.

Effects of Calcium Stores on Exocytosis

Amperometry has been used to investigate the role of Ca²⁺ and ryanodine receptors in mast cells. Ryanodine receptors have been identified to modulate exocytosis by triggering a release of internal Ca²⁺ stores. Previous studies of mast cells reported that histamine release was inhibited by the presence of ryanodine, which would suggest that exocytosis was modulated by ryanodine receptors [109]. Preliminary data showed that small concentrations of ryanodine and caffeine increased intracellular Ca²⁺ levels in mast cells with no external Ca²⁺, suggesting an increase in exocytosis. Jaffe et al. employed amperometry to investigate the role of ryanodine and caffeine in regulating serotonin release from mast cells [111]. A control group of mast cells was stimulated with a mast cell activator, 48/80, which stimulates exocytosis. These control cells were compared to mast cells treated with both a low (1 μM ) and high (50 μM ) dose of ryanodine. The low dose showed an increase in both amplitude and frequency when compared to control, while the higher dose greatly reduced the half width and spike area but not peak amplitude. Furthermore, addition of caffeine increased exocytosis. Caffeine (20 m M ) was added to mast cells and compared to control conditions; the caffeine addition caused an increase in the number of events. Further experiments were conducted, and it was determined that caffeine increases exocytosis by increasing the release of intracellular Ca²⁺ stores. These data suggest that the combination of 1 μM ryanodine and 20 m M caffeine work by ryanodine opening ryanodine receptor channels and caffeine entering the cell through the open receptor channels, which facilitates the release of Ca ²⁺ from intracellular stores. This study shows that there is a direct correlation between the ryanodine receptor and exocytosis [110].

Regulation of Insulin Exocytosis from Pancreatic Beta Cells

One set of experiments conducted with pancreatic β cells was that of Takahashi et al. to examine both fast and slow exocytosis [111]. This reportedly takes place in two different types of vesicles: small clear vesicles, and large, dense core vesicles. Small clear vesicles have been reported to undergo fast release; where as, large dense core containing vesicles tend towards slower release. Dense core vesicles contain, among other things, the polyanionic peptide called chromogranin A. As discussed above, it is hypothesized that neurotransmitters in these vesicles are bound inside the dense core. Upon initial fusion of the vesicle to the plasma membrane, the dense core, which is in the low pH environment in the vesicle, is exposed to the more basic environment outside the cell, causing it to dissociate and release the neurotransmitter molecules. The working hypothesis is that the unraveling of the dense core takes finite time and, thus, increases the time for exocytosis. In order to differentiate between the two types of vesicles, amperomeric spikes were compared. The average time difference between a group of fast- and a group of slow-release events was 1.5 s, an easily differentiated time difference as amperometry has submillisecond resolution. Ca²⁺ plays a large role in cell signaling for exocytosis, and in order to control the Ca²⁺ influx in this experiment, photo releasable caged Ca²⁺ was used. Slow and fast release events were also monitored using patch–clamp techniques.

Vesicle Size Changes with Messenger Amount

Adding the dopamine precursor L-DOPA to cell environments has been shown to affect quantal size in many cell systems [19,21,61,94,112]. L-DOPA is currently used to treat Parkinson’s disease by apparently increasing the amount of dopamine released by neurons. Treating with L-DOPA leads to larger amounts of transmitter detected per vesicle by amperometry [21]. Reserpine, an ancient anti-hypertensive drug, has the opposite effect. Reserpine blocks the vesicular monoamine transporter (VMAT) and depletes vesicles of neurotransmitter. Thus, quantal size is reduced [21,61]. L-DOPA is now frequently employed as a positive control to increase quantal size when examining novel effects on quantal size [15,19,61,94].

Interestingly, when PC12 cells are incubated with L-DOPA to increase the dopamine level in the cells and to provide larger signals [21,112], transmission electron microscopy experiments show that vesicle size increases. Vesicles in PC12 cells exposed to reserpine have less dopamine and are smaller [21]. This surprising result has been corroborated by Lindau and co-workers using patch amperometry on adrenal cells [113] and suggests that cells regulate vesicular volume to maintain a relatively constant neurotransmitter concentration.