An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Michael AC, Borland LM, editors. Electrochemical Methods for Neuroscience. Boca Raton (FL): CRC Press/Taylor & Francis; 2007.

Show details

# Chapter 15The Patch Amperometry Technique: Design of a Method to Study Exocytosis of Single Vesicles

, , and .

## Introduction: Capacitance Measurements of Membrane Fusion with Amperometric Detection of Released Molecules

Exocytosis of vesicles as small as 60 nm in diameter can be detected by cell membrane admittance measurements with the patch clamp technique in the cell attached configuration [2–4], and by amperometry with a carbon fiber electrode (CFE) [5]. The admittance measurement provides the membrane capacitance that reveals changes in surface area due to the incorporation of the vesicular membrane into the plasma membrane. Each fusing vesicle causes a stepwise increase in capacitance [2]. The admittance measurement also provides the membrane conductance and the fusion pore conductance during an exocytotic event [4]. With amperometry, released molecules that are readily oxidizable are electrochemically detected at the surface of the CFE. This technique is able to resolve release events down to a few ten-thousands of molecules [6].

In this article we describe a technique that combines high-resolution patch capacitance measurements with amperometry by placing the amperometric detector inside the patch pipette. We call this patch amperometry. It allows the simultaneous detection of membrane fusion and release of transmitter molecules originating from vesicles as small as 100 nm in diameter. In chromaffin cells it was shown that with patch amperometry, the “foot”, which sometimes precedes amperometric spike signals [7], is due to the narrow fusion pore restricting the flux of molecules in the early stages of exocytosis [8,9]. Using this technique, it was shown that even during short flickers, the entire contents of a chromaffin granule can be secreted, and that flickers are more likely to occur at high intra-pipette Ca2+-concentrations [10]. Patch amperometry was also used to investigate the regulation of quantal size and granule size in chromaffin cells [11,12]. Another recent application of patch amperometry was the measurement of the cytosolic pool of catecholamines [13].

More than 10 years ago, the first experiments combining whole-cell capacitance measurements with carbon fiber amperometry were carried out in order to reveal the correlation between fusion and release in mast cells and chromaffin cells [14–16]. The granules in mast cells are very large (greater than 0.5 μm), with an associated single granule capacitance greater than 7 fF, which is resolvable in the whole-cell configuration. One limitation of whole-cell capacitance measurements with carbon fiber amperometry is that the carbon fiber electrode does not capture exocytotic events from the entire cell surface, whereas the capacitance measurement does. As a result, for some capacitance steps, the released molecules are only partially detected or not detected at all. However, in 1993, using this approach, Alvarez de Toledo et al. [14] were the first to show a proportionality of the fusion pore conductance with the flux of molecules through that pore. They also showed that fusion of a granule can be transient (“flicker” or “kiss-and-run”) and that release can occur during transient fusion through a fusion pore of less than 1 nS.

Chromaffin granules are much smaller than mast cell granules. Thus, in whole-cell capacitance measurements from chromaffin cells, only the largest granules could be detected as capacitance steps greater than 5 fF associated with amperometric spikes [15–17] due to a noise level of 2–3 fF. From morphological studies, it is known that most chromaffin granules are much smaller, with diameters in the range of 60–450 nm (Fenwick et al. 1978), which would correspond to capacitance steps of 0.1–5.7 fF. Accordingly, whole-cell capacitance measurements capture only the largest granules as discrete steps. However, when the capacitance traces were aligned according to the amperometric spikes and averaged, an “average” step size of 2.7 fF was obtained. Whole-cell capacitance measurements did not allow determination of the fusion pore conductance in chromaffin granules.

The overall time course of amperometric spikes is still not fully understood. If release occurred instantaneously, the spike shape would be governed exclusively by diffusion from the site of release to the site of detection. In this case, the signal shape can be well described analytically [7,18]. The experimentally measured time course of amperometric spikes, on the other hand, indicates that release may not be instantaneous, but may follow a certain kinetics [19,20]. Also, from fusion pore and matrix dynamics obtained by deconvolution of amperometric recordings, it was suggested that the shape of the amperometric signals is governed by release of catecholamines from the intra-granular matrix [21]. Recent experiments employing microfabricated electrochemical detector arrays suggest that reversible binding of catecholamines to the cell surface also affects the amperometric spike shape [22].

With patch amperometry, the different features of amperometric spikes can now be directly correlated with the fusion pore restriction [8,9]. Furthermore, the cell attached configuration limits the area of observed cell membrane and measures the signals in the confined space between cell, pipette wall, and the detector. Hence, there is much less variation in signal width in a given experiment, which allows the diffusion-controlled contributions to the signal shape to be distinguished from the non-diffusion related components.

## Patch Amperometry

Patch amperometry combines the patch clamp technique with the amperometry technique by including a CFE inside the patch pipette. Several alterations to the stand-alone techniques were necessary to use this new configuration. The combination of the two techniques required the design of a novel pipette holder. To accommodate the carbon fiber electrode inside the confined space of the patch pipette, the properties of both the patch pipettes and the carbon fiber electrodes needed to be modified. Here we describe the details of these modifications as well as the analysis of the signals that are recorded with patch amperometry.

### Patch Amperometry Requires Reversed Patch Clamp Electrode Configuration

The regular patch clamp electrode configuration is shown in Figure 15.1a. The electrode, which is connected to the patch clamp amplifier, is located inside the patch pipette. The reference electrode, which is connected to ground, is placed in the bath. For conventional amperometry (Figure 15.1b), the CFE is connected to the input of the amplifier, and is held at a potential of 650–800 mV versus the Ag|AgCl reference electrode in the bath. The two techniques are combined in patch amperometry by placing the CFE inside the patch pipette (Figure 15.1c). If both the patch clamp and the amperometric electrode were located inside the pipette and the reference electrode for both measurements located in the bath, then the CFE would be separated from the reference (ground) electrode by the high giga-seal resistance, which would make the recordings unstable and would cause excessive noise. As a result, the electrode configuration for the patch clamp experiments has to be reversed. When the electrode connected to the patch clamp amplifier is located in the bath, then the CFE and the reference electrode can both be placed in the same compartment, the patch pipette. Hence, the high potential for amperometry is not applied across the cell membrane. Although the configuration of the patch clamp measurements is reversed, it still measures the current passing through the cell between the patch pipette and the electrode in the bath. Since the bath is not connected to ground, it is important to reduce its stray capacitance by reducing the surface area of the recording chamber and by using a long working distance objective (see below). Noise reduction through proper shielding of the setup and avoiding ground loops need to receive particular attention, too.

#### FIGURE 15.1

Electrode configurations.(a) Conventional cell-attached patch clamp configuration. (b) Conventional carbon fiber amperometry configuration. (c) Patch amperometry configuration, where the patch clamp electrode configuration is reversed in order to place (more...)

### Equipment and Setup

#### Data Acquisition

Figure 15.2 shows a schematic view of the signal processing devices and their connections to each other for patch amperometry. In general, a data acquisition system is required that can record five traces for at least 30 min at a sampling rate of ≥ 1 kHz. We typically used 1 kHz.

#### FIGURE 15.2

Patch amperometry set-up. For details and description of signal pathways see section “Equipment and Set-up”. (From Dernick, G. et al., Nat. Methods, 2, 699–708, 2005.)

Experiments were controlled and data acquired with an Apple Macintosh G4 or G5 computer running MacOS9 by procedures written in Igor 4 (WaveMetrics Inc., Lake Oswego, OR, U.S.A). The signals were digitized by a 16-bit analog-digital converter (ADC). We currently use the PCI-MIO- 16XE-50 DAQ board with a shielded cable and the BNC adapter BNC-2090 (National Instruments Inc., Austin, TX, U.S.A). This data acquisition board has eight differential inputs, two 12-bit digital-analog converter (DAC) outputs and eight selectable digital inputs/outputs.

Signals were passed through RG58C/U type coaxial cables with 50 Ω impedance (Newark Electronics, Chicago, IL, U.S.A.). One DAC output supplied the voltage clamp holding potential and pulses for seal formation, the other a trigger signal to synchronize the oscilloscopes with the pulses. One digital output gave a TTL signal to switch the adder and the image acquisition. For the use in our labs we created a set of Igor Procedures for acquisition (PA_Acquire) and for analysis (PA_Viewer). These are available online [1] and may be modified for the development of the reader’s own customized software.

#### Capacitance Measurements

In the following, we describe the setup for two different patch clamp amplifiers, the HEKA EPC-7 (Heka Electronics, Darmstadt, Germany) and the Axopatch 200B (Axon Instruments/Molecular Devices, Sunnyvale, CA, U.S.A.).

For patch capacitance measurements, we used a lock-in amplifier (SR 830; Stanford Research Instruments, Sunnyvale, CA, U.S.A.), whose settings were controlled from within Igor via a RS-232 serial port of the data acquisition computer. The sine wave from the lock-in amplifier (50 mV r.m.s, 20 kHz) was added to the holding potential from the DAQ board in a home-built switchable adder. This device adds two analog inputs, where a transistor–transistor logic (TTL) signal determines if only the first input goes through or if the second input is added to the first. A circuit diagram of the adder is available online [1].

The output of the adder was fed into the stimulus input of the patch clamp amplifier. For recording purposes, the stimulus monitor was low pass filtered at 500 Hz to remove the sine wave and was acquired by the ADC.

The EPC-7 was set to voltage-clamp mode, C-slow compensation to the 10-pF range; the C-slow compensation value was set to 0.2 pF (minimum setting of dial), the G-series compensation to 0.2 μS. The series resistance compensation was switched off. The compensation is chosen for a patch with a typical capacitance of 200–300 fF and an access resistance of about 5 MΩ, which results in a time constant of about 1.5 μs [23]. For high-resolution measurements, a gain of 50 mV/pA or higher must be used. In this range, the amplifier uses the high resistance feedback resistor in the head stage, which allows for low-noise recordings. During the setup of a recording and seal formation, a stimulus scaling of 0.01 is used; during the recording itself, it is set to 0.1. The input filter TR is set to 2 μs. To avoid amplifier clipping, the pipette current was filtered by the built-in 10 kHz filter (filter 1). Although a 20 kHz sine wave command voltage is used, the signal-to-noise ratio is not significantly compromised because the 10 kHz filter reduces both signal and noise around 20 kHz by the same factor [23]. This output is displayed on an oscilloscope. Additionally, the pipette current was filtered with the built-in 3 kHz filter (filter 2) from the patch clamp amplifier and acquired by the ADC. The 3 kHz filter suppresses the 20 kHz sine wave such that the net patch current can be monitored.

The Axopatch 200B is used in an analogous way. It has two command signal inputs that are added internally, one being selected or deselected by a front panel switch. With this amplifier, the signal adder is not necessary. However, the sine wave must then be switched on and off manually. The scaling factors differ for the front panel- and back panel-switched inputs. If the front panel-switched input is used, then the scaling factor is 0.02 and the lock-in output voltage must be set to 2.5 V to obtain a 50 mV r.m.s. pipet voltage signal. For patch capacitance measurements, the “whole cell β = 1” mode can be used. We were not able to achieve low noise recordings in the capacitive feedback “patch” mode. The gain setting is not critical since all gain settings use the same feedback resistor. We still use relatively high gain to minimize noise contributions from other sources, including ADC quantization noise. For the patch compensation, the SERIES RESISTANCE is set to 5 MΩ and the WHOLE CELL CAP to the minimum, which is according to the manual about 0.2 pF. The pipette compensation dials of the Axopatch 200B (termed C-slow) may not be used for the patch compensation because the time constant of C-slow is too slow. However, they may be used freely to compensate pipette and bath capacitances and to minimize the sine wave during a recording.

We have, so far, mainly used the two amplifiers described above. We have done initial tests with the EPC-10 dual amplifier (HEKA), which is suitable because filter 1 can be set independently from filter 2; this is important to avoid clipping without over-filtering the sine wave current. In its double-amplifier version it may serve for low noise recording of both the patch clamp capacitance changes and the amperometric signals (see below).

The output of either patch clamp amplifier is scaled down by a factor of 10 before input to the lock-in amplifier by a 10 k Ω :100 k Ω voltage divider. The downscaling is necessary to avoid saturation of the lock-in amplifier because the output range of the EPC-7 is ± 10 V, whereas the maximum input range of the lock-in amplifier is only ( ± 1 V). The lock-in amplifier settings can be found in Table 15.1.

#### TABLE 15.1

SR830 Lock-in Amplifier Settings

At the correct phase setting, the lock-in amplifier computes the changes in the real and the imaginary part of the patch current, which is recorded by the ADC as Y1 and Y2, respectively. Finding the correct setting of the phase is important, only then is the lock-in amplifier output Y1 proportional to the membrane conductance, and the output Y2 proportional to the membrane capacitance [23].

The correct phase for the lock-in amplifier can be found in various ways [23]. One way is to vary slightly the patch capacitance compensation (C-slow in EPC-7; WHOLE CELL CAP in Axopatch 200B), which can be done with open headstage while the sine wave is applied to the stimulus input. The amplitude of the sine wave in the pipette current is minimized to zero with the pipette compensation dials (C-fast and τ-fast in the EPC-7, C-fast and C-slow dials in the Axopatch 200B) observing it on the oscilloscope. Then, the patch capacitance compensation (C-slow in EPC-7; WHOLE CELL CAP in Axopatch 200B) is slightly varied observing the two lock-in outputs Y1 and Y2. The phase is adjusted until changes in patch capacitance compensation appear only in the Y2 channel and the Y1 channel remains unchanged. Here, an increase of the compensation is a capacitance decrease and must therefore produce a decrease of Y2. The found phase setting applies only to the used feedback resistor in the headstage and to the used filter settings. As mentioned above, the EPC-7 has two feedback resistors in the headstage. One for gain settings below 20 mV/pA and another for settings above 50 mV/pA. The Axopatch 200B has only one feedback resistor in the headstage in the “whole cell β = 1” mode, so the gain may be changed without changing the phase. In our tests the capacitive feedback in the Axopatch 200B headstage (“patch” mode) did not provide low noise in patch amperometry recordings.

Other methods to find the correct phase include the application of suction pulses to a patched cell followed by adjustments of the phase until only changes in Y2 are observed [23], and the utilization of a capacitance dither switch on the patch clamp amplifier. Maladjustments of the phase may be corrected off-line, after the recording. This is only possible if the amplifier was not saturated (clipping). Unfortunately, the clipping indicator LED’s on the amplifiers may fail to indicate slight amplifier saturation at the extrema of the sine wave current. Thus, it is imperative to display the pipette current on an oscilloscope and to observe the sine wave amplitude during all operations.

The application of a defined capacitance compensation change, such as, 200 fF, with a capacitance dither may also be used to calibrate the recording. However, in this case such a large change may saturate the amplifier. It would then be necessary to use 0.01 stimulus scaling, effectively generating a calibration signal ten times smaller. In the EPC-7, for convenience, a switchable 20 Ω resistor was inserted in series to the C-slow potentiometer which provides a 20 fF capacitance compensation change for calibration. In the Axopatch 200B, the built-in 100 fF capacitance dither may be used, but a low enough gain must be selected to avoid clipping. Instructions to modify the dither circuit of the Axopatch 200B for a 10 or 20 fF capacitance compensation change can be obtained from Axon Instruments upon request. However, the small resistors used may result in an imprecise calibration. Note that the resistor for calibration is added to the compensation in the EPC-7, while it is subtracted (switched off) in the Axopatch 200B. The orientation of the calibration pulses wil thus be downwards for th EPC-7 but upwards in the Axopatch 200B.

After the recording, the calibration pulses were used to convert units of the Y1 and Y2 traces from raw counts into pS and fF, respectively. This is described in the data analysis section of this article.

#### Amperomety

The VA-10 (NPI Electronics, Tamm, Germany) or a patch clamp amplifier may be used as an amperometric amplifier. Also, a simple home-built amplifier can be used; a circuit diagram may be obtained from the Alvarez de Toledo laboratory. As mentioned before, initial tests indicate that the EPC-10 dual amplifier (HEKA Electronics) is a practical solution for patch amperometry because it may serve as both an amperometric and a patch clamp amplifier. Capacitance measurements with a 20 kHz sine wave have been implemented in the new software for this instrument. This instrument will be tested in the future to determine its noise performance in patch amperometry recordings.

When using the VA-10 or the EPC-10, a special version of the head stage must be ordered, in which the BNC shield is grounded. This is available upon request from the manufacturers. The modification is necessary because in patch amperometry, the reference electrode for amperometry is connected to the shield of the BNC. The standard head stages where the BNC shield is connected to Vref will not work for patch amperometry.

The amplifier typically applies a potential of +700 mV to the CFE relative to the reference electrode. The amperometric current is low pass filtered at 500 Hz and acquired by the ADC. Since both the patch clamp amplifier and the amperometric amplifier use the same reference electrode that is connected to the ground of the amperometric amplifier headstage, it is essential to connect the signal-ground connectors of the two amplifiers to each other, but to make no connections to the ground pin of the patch damp headstages to avoid ground loops.

#### Microscopy

The experiment was observed through an inverted microscope (Axiovert 100; Zeiss, Oberkochen, Germany), using an objective with long working distance (Zeiss 20×0.5 Plan Neofluar) to minimize stray capacitance. Objectives with short working distance, and in particular immersion objectives, may not be used because the stray capacitance between such an objective and the bath (which is not grounded in patch amperometry) cannot be compensated for.

As for other patch clamp applications, the microscope was mounted on a vibration-isolated table (Physik Instrumente, Waldbronn, Germany) hooked up to house pressurized air. Coarse positioning of the self-built microscope stage and the electrode holder arm was done with tower-mounted linear micrometers (Spindler&Hoyer, Göttingen, Germany, now LINOS Photonics). For the fine positioning of the patch pipette, a hydraulic high-resolution micromanipulator (Narishige, Tokyo, Japan) was used.

Images were taken with a CCD camera (SSC M-370, Sony, Japan), which was mounted on the microscope. Any USB or FireWire camera may be used for this purpose. Images were taken automatically during acquisition of electrophysiological data with a PixelBuffer frame grabber (Perceptics, Knoxville, TN, U.S.A) in an additional Apple Macintosh computer. As explained earlier, during data acquisition, a TTL signal given from the DAQ board switches on the sine wave from the lock-in amplifier. The same signal was fed into the trigger input of the frame grabber. A script in IPLab (Scanalytics Inc., Fairfax, VA, U.S.A.) took pictures at 60 s intervals whenever the trigger was enabled. The images were time stamped for off-line synchronization with the recorded data.

### Design of the Electrode Holders

The pipette holder for patch amperometry had to combine several different features in order to measure fusion and release from single chromaffin granules simultaneously in the cell-attached configuration. The design was guided by the two techniques, patch clamp and amperometry, which were combined. The patch clamp part of the holder had to accommodate a silver wire for the electrical connection as well as the patch pipette. It needed to be airtight, so that suction could be applied to it via a small tube to obtain a pipette-membrane seal [24]. For the patch amperometry configuration, the CFE as well as the reference electrode had to be inserted into the pipette. As already mentioned, the electrode configuration for the patch clamp technique was reversed such that the reference electrode located inside the pipette was connected to ground. Therefore, the BNC shield of the pipette holder had to be connected to ground on the amplifier headstage. In addition, the CFE needed to be movable with respect to the patch pipette, or the pipette relative to the CFE, to adjust the variations in length of both the patch pipettes and CFEs. The CFE has to be positioned as close as possible to the patch to minimize diffusional broadening of the amperometric signals [25].

Two different types of electrode holders were developed: one with a fixed CFE and a manually adjustable pipette, the other with a fixed pipette and a CFE that was moved by a motor inside the electrode holder.

#### FIGURE 15.3

Manually adjustable electrode holder. (a) Photograph of electrode holder. The lower part can be adjusted against the upper part by the fine-pitched thread. The CFE is mounted fixed and the pipette is moved in respect to it. The asterisk denotes parts (more...)

The connection to the amperometric amplifier was made with a BNC connector (Figure 15.3, top). A silver wire (0.25 mm in diameter; Goodfellow, Cambridge, England) was soldered to the inner pin of the BNC plug and chlorinated. The wire both guided the CFE and made electrical contact with the carbon fiber via the 3M KCl solution inside the CFE. A short piece of polyethylene tubing attached the CFE mechanically to the BNC plug. The reference electrode (chlorinated silver wire, 0.25 mm in diameter) was soldered to a small screw in the housing of the holder and was in this way connected to ground. The cap was screwed onto the housing with a fine pitched thread. O-rings in the housing and the cap of the holder ensured that the junction was airtight and provided additional stability to prevent the tilting of the cap against the housing. The pipette was clamped onto the holder in a manner similar to conventional patch pipette holders. A step in the insertion hole provided a stop position for the pipette. An O-ring in the small screwcap tightened and held the pipette in place. Suction was applied to the entire lumen of the holder through the metal tube close to the BNC plug.

#### Motorized Electrode Holder

Although successful recordings were made with the manually adjustable holder, it has some draw-backs. The holder has to be removed from the head stage every time a new pipette is mounted, and the adjustments under the inverted microscope must to be done very carefully to avoid breaking the newly mounted pipette. The most serious drawback of the manually adjustable holder is that the distance between patch and CFE cannot be changed after the seal is formed. In some cases the CFE was too far away, resulting in very small and broad amperometric signals that were hard to analyze. In other cases, the CFE was positioned too close to the pipette tip preventing seal formation when upon applying suction, the patch touched the CFE before a giga-seal was achieved. The motorized electrode holder improved convenience of handling, control and variation of CFE-patch distance during an experiment and revealed information about the release mechanism of catecholamines after the fusion pore expanded completely.

An electrode holder that allows adjustment of the CFE-patch distance during an experiment required a stationary pipette to keep the seal intact. It had to be robust and airtight like the manually adjustable holder. Inside, a small device was needed to move the CFE, while the electrical connection between the CFE and the Ag |AgCl wire via the KCl-Solution was maintained. These considerations yielded the design of a new electrode holder, which accommodated a small motor to move the carbon fiber (Nanomotor, Klocke, Nanotechnik, Aachen, Germany). This type of piezo-driven linear motor has a wide spectrum of applications in micro-positioning and micro-manipulation, ranging from scanning tunneling microscopes to micro-assembly of electronics and telecommunication devices [26]. The motor was mounted inside the holder by the manufacturer. The first model of the motorized holder is shown in Figure 15.4a and b. Improvements in air-tightness and sturdiness were achieved for the second model, shown in Figure 15.4c and d.

#### FIGURE 15.4

Motorized electrode holders. (a) Photograph of the first model of the motorized electrode holder. (b) Schematic sketch of the first model of the motorized electrode holder. (c) Photograph of the second, improved model. (d) Sketch of the second, improved (more...)

As the manually adjustable holder, the motorized holder’s connection to the amperometric amplifier was made by a BNC plug. A teflon-coated silver wire (0.125 mm in diameter; World Precision Instruments, Sarasota, Florida, U.S.A.) soldered to the inner pin of the BNC plug connects to the CFE. The teflon coat serves three purposes: (1) it insulates the wire against the housing when the wire is ducted through the movable inner sheath of the nanomotor; (2) since the CFE is moved against the wire, the teflon coat minimizes friction; and (3) it prevents the 3 M KCl solution in the CFE from ascending to the motor where the solution could cause short circuits and crystals from dried KCl solution could damage the motor. The CFE was attached mechanically to the inner moving sheath of the nanomotor via a tightly fitting, short (approximately 5 mm in length) piece of silicon tubing. In the first model (Figure 15.4a and b), the motor received power via two jacks, one on each side of the holder. A hole in the housing of the holder, which is required for mounting the motor, was closed by a tightly fitting brass ring that slides over the hole. This ring also holds a small pin to connect the reference electrode (silver wire, 0.25 mm in diameter) to the grounded holder housing. The holder is further sealed with two O-rings. The perspex cap resembles that of commercially available electrode holders and was screwed onto the brass housing. The features of this part are similar to those of the manually adjustable holder described above.

In the improved second model of the motorized holder (Figure 15.4c and d), the perspex cap screws against the rim above the mounting hole, thus improving the air-tightness and stability. Because of this modification, the reference electrode pin had to be moved beyond this rim. The pin was soldered into the hole in the housing to make it airtight. The two contact jacks were placed on the same side of the holder making it sleeker and fit better into the setup.

The controller for the nanomotor was built according to a circuit design provided by the manufacturer. The motor moves by means of a high frequency control signal, which is switched on briefly to move the CFE. It is switched off to stop CFE movement and during recording. The controller should be placed inside the faraday cage of the setup to minimize noise pick-up.

### Patch Pipettes

Patch pipettes were pulled from borosilicate glass (outer diameter of 2.0 mm, inner diameter of 1.4 mm, length of 85 mm, ends fire polished; Hilgenberg GmbH, Malsfeld, Germany) with a programmable horizontal puller (P-97; Sutter Instruments, Novato, CA, U.S.A). The puller was set up with a box filament, which had a box size of 3×3 mm and a width of 3 mm. The settings for the puller are given in the following in the arbitrary units which are used to program the device. The air pressure was set to 500, the air before pull and the air after pull were 5 s each. The ramp test for these capillaries was 655. A single line program looped four times and had the parameters as follows: heat 615, pull 0, velocity 50, and time 200. Pipettes ideally had a tip diameter of 2–3 μm, an opening angle of at least 20°, and a shank length of 5 ± 0.5 mm. A large opening angle was necessary to bring the tip of the CFE close to the pipette tip. For larger angles the velocity was sometimes reduced. However, a compromise had to be found between a large enough opening angle and a small enough tip diameter. After inspection under the microscope, only those pipettes which fulfilled the requirements were used. Pipette tips were dipped in melted wax to reduce stray capacitance (Sticky Wax; Kerr Inc., Orange, CA, U.S.A.). The wax coating procedure is quicker than using sylgard, and formation of seals appeared to be facilitated with wax-coated pipettes. Right before use, pipettes were fire polished with a microforge (microscope: Axiovert 25; Zeiss, Oberkochen, Germany; microforge: CPM-2; ALA Scientific Instruments Inc., Westbury, NY, U.S.A.). The pipettes were placed near a heated wire. As a result, the wax in the tip melted and flowed out of the pipette tip. Subsequently the tip narrowed due to the hot filament (polishing). Pipettes were backfilled with pipette solution, which was filtered through a 0.22 μm syringe filter and applied with a hypodermic needle. The pipette solution was connected to ground via the Ag|AgCl reference electrode in the holder. Typical pipette resistance was approximately 1 M Ω.

### Carbon Fiber Electrode Fabrication

The procedure followed the method described by Chow and V. Rüden [18], but was modified so that the CFE fit inside the patch pipette close to the tip. Briefly, the entire electrode was made from polyethylene tubing, no glass capillary was used, and the big and the small drop-like beads formed during fabrication were manipulated to minimize their size. The differences between a conventionally fabricated CFE and one produced with our method are shown in Figure 15.5.

#### FIGURE 15.5

Carbon fiber electrodes.(a) Conventionally fabricated CFE. Due to symmetrical pulling, a big PE-bead is present on the left. Inside the big bead the silver wire (0.25 mm in diameter) can be seen. (b) Magnification of the box in (a) with the small bead. (more...)

#### Carbon Fiber Fabrication Setup

Fabrication of CFEs was performed under a dissecting microscope (Stemi 2000, Zeiss, Oberkochen, Germany). A Pt/Ir wire (0.4 mm in diameter) measuring about 50 mm in length served as a heating filament. The wire was bent in the middle into a loop with a diameter of about 6 mm. The ends were held by an insulated clamp and connected to a variable power supply (12 V, greater than 3 A) via a foot switch. The wire loop was positioned under the dissecting microscope in the middle of the field of view. The 0.4 mm wire was chosen as a trade off between stability of the loop geometry and required heating current.

Pulling CFEs requires some practice. The beginner may train by pulling just the PE tubing without the inserted carbon fiber until appropriate heat settings are found and the desired shape, as described below, can be pulled consistently. Heat settings were chosen, such that the PE-tubing in the loop softened within 10–15 s and melted within 20–25 s. Upon melting, the PE tubing became clearly transparent. It may be helpful to set up a programmable CFE puller as described (Supplementary Note of [1]).

#### Carbon Fiber Pulling

To handle the rigid but brittle carbon fibers without breaking them, the tips of two pairs of watch-makers forceps (#5; TechniTools, Switzerland) were coated with 0.5 cm of polyethylene tubing (0.80 mm outer diameter, 0.40 mm inner diameter; Sims-Portex Ltd. Hythe, Kent, England). The same type of polyethylene (PE) tubing was used to fabricate the CFEs. A piece 10 cm in length was dipped in methanol and allowed to fill completely with the solvent by capillary action. This reduced the static attraction between the plastic and the carbon fiber during the next (cannulation) step. A single carbon fiber (Thornell T650/42, 5 μm in diameter; Amoco Co., Atlanta, GA, U.S.A.) of at least 5 cm in length was taken from a strand of carbon fibers with the watchmakers forceps and its tip was inserted into the methanol-filled PE-tubing under a dissecting microscope using bright illumination. The carbon fiber was then pushed into the PE-tubing completely. The solvent was wicked away by tapping the tubing onto a piece of tissue paper, followed by drying for 30 min at 50°C. It should be noted that production of the 5 μm carbon fibers was discontinued by the manufacturer but they are still available from the Lindau laboratory upon request.

#### Testing the CFE in the Patch Amperometry Configuration

CFEs were tested prior to use. When CFEs were immersed in Ringer saline with the reference electrode, the amperometric current ideally decayed to a few pA within a minute or two, and the current noise level was significantly below 1 pA r.m.s. These properties were determined by observing the amperometric current on an oscilloscope. CFEs that did not fulfill these criteria were discarded. If a CFE shows rapid current spikes in this configuration, it may help to rechlorinate the silver wires thoroughly and to fill the pipette and CFE such that only the chlorinated parts of the wires are immersed into the respective solutions.

It is important, that the CFE is insulated all the way to the tip and that its surface is working properly. This can be tested in the patch amperometry configuration: With the pipette immersed in the bath solution, but no cell attached to it, the CFE is positioned close to the pipette tip. Positive pressure is applied to the pipette, and a drop of dopamine solution is added to the bath (e.g., 10 μl of 1 M dopamine to 100 μl of bath solution). When the pressure is released from the pipette, dopamine diffuses into the pipette tip and a large oxidation current (IA) appears. When positive pressure is applied again, IA quickly decreases to background current level. The CFE is then moved further away from the tip. Now, when pressure is released, it takes longer for the dopamine to diffuse into the small orifice of the pipette, so the onset of the signal is delayed. CFE’s were usually tested in this way prior to the experiments. If a CFE was used for more than a few days its functionality was tested again in this way.

CFE’s can be stored and used for a several weeks. If a CFE is not in use it is advisable to withdraw the KCl solution from the inside to avoid crystallization, which may damage the CFE.

### Recording Chamber Design

Because of the reversed electrode configuration, the bath is not held at ground potential as usual, but at the potential applied by the patch clamp amplifier. If the bath surface is large, then the stray capacitance between the bath on one side and the microscope stage and objective on the other side cannot be compensated for with the patch clamp amplifier compensation circuit. A smaller bath surface also reduces noise. We found that a bath camber with a diameter of 10 mm and an amount of bath solution of 100–150 μL gave good results.

Recording chambers were prepared by drilling a 10 mm hole into a petri dish that fit into the respective patch clamp rig. A coverslip measuring 12 mm in diameter or more, was glued to the bottom of the dish with Sylgard. Sylgard is non-toxic and broken coverslips are easily exchanged. Only the inner 10 mm diameter well formed by the hole with the coverslip bottom serves as recording chamber.

A small coverslip with adherent cells (e.g., chromaffin cells) is placed in this inner well and 100–150 μL of bath solution is added. Care must be taken not to spill solution over the edge of the inner well to avoid large and fluctuating bath capacitance. The small volume requires that bath solution and cells are exchanged at least every hour because the bath solution’s evaporation leads to unacceptable changes in concentration.

We performed experiments mostly with chromaffin cells. Bath and pipette solutions were typically based on the following compositions, with alterations according to experimental conditions:

• Bath solution (for chromaffin cells):140 mM NaCl, 5 mM KCl, 5 mM CaCl2, 1 mM MgCl2, 10 mM HEPES–NaOH, 10 mM glucose (pH 7.3).
• Pipet solution (for chromaffin cells): 50 mM NaCl, 100 mM tetraethylammonium ion(TEA)-Cl, 5 mM KCl, 10 mM CaCl2, 1 mM MgCl2, 10 mM HEPES–NaOH (pH 7.3).
• The osmolality of the solutions was checked such that bath and pipette solution had matching osmolalities not exceeding 300 mosmol.

### Patching of Chromaffin Cells

Experiments began with mounting a CFE and testing its quality as described above. During a series of experiments, the fiber typically remained attached to the electrode holder unless it broke or became noisy or leaky. A coverslip with cells from the incubator was washed once with bath solution and placed in the recording chamber. Under the microscope, a suitable cell for an experiment was chosen. A patch pipette was fire polished and filled with pipette solution as described above, and mounted on the electrode holder.

The manual holder always needed to be removed from the setup for mounting the patch pipette and for coarse adjustment of the CFE, close to the tip of the pipette, under the dissecting microscope. The fine adjustment of the carbon fiber to the tip of the pipette was done on the manual holder with the pipette just barely immersed in the bath and by observing it through the inverted microscope. Great care must be taken not to move the holder too much while turning the thumbscrew of the holder, which could break the pipette tip by touching the bath chamber. This was rather tedious because the pipette tip very often comes out of focus during the adjustment.

Adjustment of the CFE in the motorized holder was much easier. With some practice and good lighting, the pipette could be mounted on the motorized holder while the holder remained on the head stage in the setup. Some caution had to be taken not to break the CFE. While observing the pipette tip immersed in the bath through the inverted microscope, the CFE was then adjusted inside the pipette using the nanomotor, first at the highest speed, i.e., with the highest frequency selected at the motor control unit. We always operated the controller at maximal voltage amplitude and varied only the frequency to change speed. As soon as the CFE came into the field of view, the frequency of the control signal was reduced at the motor controller. With the smallest possible frequency (but still with maximum amplitude) the CFE was slowly positioned approximately 15 μm back from the pipette tip.

Seal formation followed standard patch clamp procedures as described elsewhere [27]. A test pulse was applied and the pipette current monitored on an oscilloscope with a gain of 2 mV/pA and a stimulus scaling of 0.01 on the EPC-7 (similarly, a low gain is used in the Axopatch 200B; the built-in test pulse capability can be used for pipette and seal resistance determination). If during application of suction to the pipette for seal formation the patch moved too far into the pipette near the CFE, then the CFE was retracted with the motorized holder. When working with the manual holder, a CFE too close to the tip sometimes prevented seal formation when the patch was sucked too far into the pipette. Sometimes increasing of the holding potential facilitated seal formation. Note that because the electrode configuration is reversed, the polarity of the applied voltage is reversed.

After seal formation, the gain was increased to 50 mV/pA for low noise, and the stimulus scaling set to 0.1. A gain as low as 10 mV/pA could be used with the Axopatch 200B if a high-resolution data acquisition system was used. With the test pulses still switched on, fast transients of the pipette current were minimized with the pipette compensation dials (C-fast and τ-fast dials on the EPC-7; dials in the “C-slow” area on the Axopatch 200B). Then the test pulses were switched off, the sine wave was switched on, and the amplitude of the sine wave in the pipette current minimized again using the respective pipette compensation dials.

### Capacitance Calibration

After starting the data acquisition, the recording was calibrated with the capacitance dither to determine the amplitude factor at the chosen amplifier and filter settings. For calibration, we typically applied three capacitance calibration pulses for easy identification during analysis by transiently changing the capacitance compensation by 20 fF. Larger changes may saturate the amplifier and a reduced stimulus scaling would have needed to be used. Chromaffin cell exocytosis was stimulated during seal formation, such that the exocytotic events typically occurred at the beginning of a recording. It was consequently advisable to perform all the adjustments between seal formation and start of the data acquisition in less than 30 s.

During the recording, the amplitude of the sine wave in the pipette current had to be observed on the oscilloscope to prevent clipping of the amplfiers. The clipping indicator LED’s were not reliable when a 20 kHz sine wave was used. The amplitude of the patch clamp amplifier output should never exceed plus or minus 10 V to avoid saturation of the lock-in amplifier. When the sine wave of the patch clamp amplifier output exceeded 5 V (100 pA at 50 mV/pA gain) we re-adjusted the pipette capacitance compensation to minimize the sine wave current amplitude. Occasionally checking of the seal resistance was appropriate, particularly when recordings became very unstable. For seal check, the sine wave was switched off momentarily and the test pulses switched on. During such a seal check capacitance recordings were invalid because of the absence of the sine wave.

The end of a recording is either marked by loosing the seal or by the cell going to the whole cell configuration. In the latter case, the cytosolic catecholamines diffused out of the cell and caused a rather slow but fairly big spike in the oxidation current [13]. Sometimes a cell went whole cell and then resealed to the cell-attached configuration. In such cases, some chromaffin granules may have been freely floating outside the patch in the pipette tip. When they touched the CFE, they gave very fast amperometric signals that were not accompanied by capacitance steps.

Capacitance steps without amperometric signals may reflect exocytosis of granules that do not contain catecholamines [11]. However, if all events in a patch lack amperometric signals, then the CFE may not function properly and should be checked. In a given experiment it may be helpful to break the patch at the end and check the CFE by eliciting spikes from bursting granules sucked out of the cell.

## Analysis of Exocytotic Events

### Vesicle Capacitance and Fusion Pore Conductance

The practical approach and the underlying theoretical treatment of capacitance measurements of individual chromaffin vesicles is described in detail elsewhere [4,28,38,39]. Here we discuss the steps in analysis of patch amperometry recordings. We wrote a set of procedures for data analysis that are available online [1] and may serve as a guide to the development of custom data analysis procedures.

After a recording, the five traces were loaded into Igor, where the “fifo” file format from the acquisition was converted into “waves”. Next, the traces V (stimulus voltage monitor), I (filtered pipette current) and A (ampermetric current) were converted into mV and pA, respectively. This required entering the gain settings that were used for the amplifiers during the recording. In the next step, the admittance traces were inspected. The phase was initially set by visual inspection of the calibration pulses. The calibration pulses should only occur in the imaginary part of the pipette current (Im-trace, Y2). If a projection was seen in the conductance trace (Re-trace, Y1), then the phase was adjusted according to the following relation:

$(YGYC)=(cos(Δϕ)-sin(Δϕ)sin(Δϕ)cos(Δϕ))(Y1Y2)$
15.1

until the calibration pulses did not show any projection into the real part of (Re-trace, Y1) anymore.

This relation was converted into an Igor procedure as a function as follows:

Function Phaseshift()

Variable Y1, Y2

Wave Y1_ = Y1_, Y2_ = Y2_

nvar Ph_ch = Ph_ch

Variable wavecount = 0

Variable endpoint = numpnts(Y1_)

Variable cp = cos(Ph_ch*2*Pi/360)

Variable sp = sin(Ph_ch*2*Pi/360)

Do

Y1 = (Y1_[wavecount]*cp)+(Y2_[wavecount]*sp)

Y2 = (Y2_[wavecount]*cp)− (Y1_[wavecount]*sp)

Y1_[wavecount] = Y1

Y2_[wavecount] = Y2

wavecount += 1

while (wavecount< endpoint)

End

This function applied phase changes to the entire length of the Y1 and Y2 traces. Phase changes were made absolute (not relative) by first resetting the phase to the originally recorded traces. After the phase adjustment, the calibration pulses were measured and the Y1 and Y2 traces converted into fF (1 femto Farad = 10−15 F) and nS (1 nano Siemens = 10−9 S = 109 Ω ). The real and imaginary traces were displayed at the same “magnification”: The axis range for the imaginary trace was set by the user, and the axis range for the real trace calculated using the relation:

$Re(nS)=ωIm 10-6(fF)$
15.2

If the phase is set correctly, the real and imaginary part of the pipette current can be measured at the two outputs of the lock-in amplifier. In this ideal case:

$Re=(ωCV)2/Gp1+(ωCV/Gp)2$
15.3
$Im=ωCV1+(ωCV/GP)2$
15.4

The true vesicle capacitance and the true conductance of the fusion pore can be calculated as [4,28]:

$CV=[(Re2+Im2)/Im]/ω$
15.5
$GP=(Re2+Im2)/Re$
15.6

The expected trace appearance, as well as the parameter calculations, for these traces is illustrated for an idealized exocytotic event in Figure 15.6. Here we assume that the imaginary part of the patch admittance (Figure 15.6a, solid line) increases linearly until it reaches the level of the full vesicle capacitance ( ω CV). The real part of the patch admittance was calculated from the imaginary part by substituting Equation 15.5 and Equation 15.6 into Equation 15.3 and is shown in the same panel as a dashed line. As the imaginary part increases, the real part increases steeply, becomes shallower, and reaches its maximum when GP = ω CV. At this point, the real and imaginary part are equal to ω CV/2 (Re = Im = 1/2 ω CV). As the imaginary part increases further towards the actual vesicle capacitance (ω CV), the real part returns to baseline. From such lock-in amplifier output traces, the time course of CV and GP are calculated using Equation 15.5 and Equation 15.6, as shown in Figure 15.6b. CV increases stepwise at fusion pore opening. While CV is constant, the fusion pore conductance GP increases in a time-dependent manner and eventually becomes immeasurably large.

#### FIGURE 15.6

Theoretical traces for two cases of fusion pore expansion. Linear increase of imaginary part (a),(b); linear increase of fusion pore conductance (c),(d). Abscissae: units of Ω CV; ordinate: arbitrary units of time. For details see text.

Figure 15.6c and d show the traces one would obtain if the fusion pore conductance increased linearly to 10× ω CV and then increased rapidly to infinity (full flattening in the membrane). From the two traces for GP and CV (Figure 15.6d), the real and the imaginary parts were calculated as they were expected to be measured at the outputs of the lock-in for such a pore (Figure 15.6c). In actual recordings, the time course of GP is usually more complex but a flat CV is a good indication that the fusion pore analysis is reasonable.

The actual calculation of recorded traces requires correct setting of base lines in the real and the imaginary traces in the section preceding an event. If that is not possible, one can also set a base line for only the imaginary trace, use CV obtained from the imaginary part at late times, and calculate the fusion pore conductance and the expected real part. The expected and measured real part should not show significant differences.

The vesicle capacitance can be converted into membrane area by using an appropriate conversion factor (9 fF/μm2 for chromaffin granules) [8]. Assuming spherical shape, this can also be transformed into vesicle volume.

The fusion pore may be characterized by various parameters. We found it helpful to use the initial fusion pore conductance, initial expansion rate, and fusion pore lifetime [9]. The initial opening is marked by the first increase in Re and Im trace above the noise. Interestingly, this initial fusion pore conductance is not randomly distributed; it shows a distribution with peak indicating the most likely fusion pore conductance for a particular cell type. The initial fusion pore conductance represents a sudden transition from a closed to an open state of the fusion pore, similar to that of ion channels. However, the distribution of initial fusion pore conductances is wider than for most ion channels. We quantified the initial fusion pore expansion rate following fusion pore opening using the slope of GP ( t ) over the next 25 ms [9]. As fusion pore lifetime, we defined the time between fusion pore opening and the time where GP exceeded 1 nS. This value may be compared with the foot duration compared in conventional amperometric recordings.

The calculation of fusion pore dimensions from the conductance is somewhat uncertain. Although the thickness of the pore can be assumed to be about two lipid bilayers (10 nm), the specific conductance of the solution in the pore is not precisely known. Usually the conductance of physiological saline is assumed for fusion pore calculations [4,9,29,30], however, the pore will be filled with a mixture of extracellular and intravesicular solution. Chromaffin granules contain about 1 M catecholamines, which are mostly monovalent cations. Mast cell granules contain the divalent cation histamine at a concentration of hundreds of m M. In contrast, the ion content of neutrophil azurophil granules and esosinophil granules may be much lower. This may explain why the fusion pore conductance in mast cells [29] and chromaffin cells [9] is somewhat higher than in neutrophils [4] and eosinophils [30].

Fusion pores may open transiently (for review see [31]). In several studies, amperomtrically observed changes in quantal size were attributed to transient vs. full fusion pore openings or to changes in the vesicular transmitter content. Patch amporemetry can directly distinguish between these possibilities [12].

### Release of Molecules

The amperometric spikes that are recorded upon release of molecules from single vesicles were analyzed by several groups already and the methods have been discussed elsewhere [18,32–34]. Frequently used parameters are the presence or absence of a foot signal, foot duration, foot amplitude and foot charge; the maximum amplitude of a spike and its width at half the maximum (half width or thalf); and the integrated current. The latter yields the charge in coulombs. Assuming that z electrons are transferred during the electrochemical oxidation of catecholamines, the number of released molecules can be calculated from the charge with the following relation:

$Q=∫I dt=zFMNA=eM$
15.7

Here, Q represents the total charge involved in the redox reaction, which is obtained by integration of the current ( I ) transient; M represents the number of molecules reacted; z is the number of electrons transferred per oxidized molecule; F is Faraday’s constant, 96485 C/mol; NA is Avogadro’s number, 6.023×1023; and e is the elementary charge 1.602×10−19 C. For catechol-amines, z = 2 is generally accepted, and thus the number of molecules per vesicle can be estimated.

### Quantal Size, Vesicle Size, and the Fusion Pore

From the capacitance step the vesicle surface area can be determined as described above. This can be converted into diameter or volume of the vesicle. The associated amperometric spike provides the charge from which the number of molecules can be calculated. From these two parameters the concentration of individual vesicles can be determined. Typical catecholamine concentrations in vesicles are on the order of 1 M [8]. It was shown that the drugs l-Dopa and reserpine, which modulate vesicle content, also modulate vesicle membrane area, keeping the concentration virtually constant [12]. However, exocytosis of empty vesicles that do not contain catecholamines has also been reported [11].

Correlating the foot signal with fusion pore conductance [8,9,14,35], it was shown that the fusion pore conductance controls the flow of molecules through the fusion pore. Patch amperometry may be useful to investigate the effect of fusion pore mutants that may differentially affect fusion pore conductance and selectivity as determined from the correlation between fusion pore conductance and flux of catecholamines.

## Summary and Discussion

We developed patch amperometry, a new technique that combines high-resolution cell-attached capacitance measurements with simultaneous amperometric detection by placing the amperometric electrode inside the patch pipette. For this purpose we designed a new electrode holder, which accommodates the connections for two electrodes and has the capability to change the position of the CFE tip versus the pipette tip. The method makes it possible to determine vesicular concentrations, to distinguish transient and full fusion events, and to analyze individual fusion pore openings and the role of the fusion pore in limiting release. While it is now clear how the size of the fusion pore determines release from chromafin granules during the amperometric foot signal, it is not yet well understood what determines the shape of the spike after the fusion pore expansion. After the fusion pore dilates, the molecules are released rapidly, giving rise to the amperometric spike. It is likely that in this phase release is not hindered by the fusion pore restriction any longer. The vesicular contents should be able to diffuse freely to the detector. However, when a CFE is placed very close to the cell surface, spikes are broader than expected from diffusion theory if the release giving rise to the amperometric spike were instantaneous [18–20,36,37]. Using patch amperometry it was shown that in mast cells the upstroke of the amperometric spike is not limited by the fusion pore [35]. It is widely believed that the time course of the amperometric spike reflects dissociation of bound molecules from the matrix followed by diffusion to the CFE. However, recent amperometric measurements from chromaffin cells using four-electrode electrochemical detector arrays have suggested that diffusion near the membrane is very slow, presumably due to reversible binding of catecholamines to the cell surface [22]. Whether a larger fusion pore in chromaffin cells is also participating in limiting release during the amperometric spike is still to be answered.

## Acknowledgments

We thank M. Montesinos, R. Borges, R. Staal, and D. Sluzer for their contributions, in particular in the use of the Axopatch 200B for patch amperometry. This work was supported by grants form NIH (RO1 NS38200-01A2) and the Nanobiotechnology Center (an STC program of NSF, Agreement No. ECS-9876771) to M. L., and a grant from the Ministerio de Educación y Cultura, Spain to G.A.d.T.

## References

1.
Dernick G, et al. Patch amperometry: High resolution measurements of single vesicle fusion and release. Nat Methods. 2005;2:699–708. [PubMed: 16118641]
2.
Neher E, Marty A. Discrete changes of cell membrane capacitance observed under conditions of enhanced secretion in bovine adrenal chromaffin cells. Proc Natl Acad Sci USA. 1982;79:6712–6716. [PMC free article: PMC347199] [PubMed: 6959149]
3.
Lollike K, Borregaard N, Lindau M. Capacitance flickers and “pseudoflickers” of small granules, measured in the cell attached configuration. J Biophys. 1998;75:53–59. [PMC free article: PMC1299679] [PubMed: 9649367]
4.
Lollike K, Borregaard N, Lindau M. The exocytotic fusion pore of small granules has a conductance similar to an ion channel. J Cell Biol. 1995;129:99–104. [PMC free article: PMC2120381] [PubMed: 7535305]
5.
Leszczyszyn DJ, et al. Nicotinic receptor-mediated catecholamine secretion from individual chromaffine cells. J Biol Chem. 1990;265:14736–14737. [PubMed: 2394692]
6.
Bruns D, Jahn R. Real-time measurement of transmitter release from single synaptic vesicles. Nature. 1995;377:62–65. [PubMed: 7659162]
7.
Chow RH, Rüden Lv, Neher E. Delay in vesicle fusion revealed by electrochemical monitoring of single secretory events in adrenal chromaffin cells. Nature. 1992;356:60–63. [PubMed: 1538782]
8.
Albillos A, et al. The exocytotic event in chromaffin cells revealed by patch amperometry. Nature. 1997;389:509–512. [PubMed: 9333242]
9.
Dernick G, Alvarez De Toledo G, Lindau M. Exocytosis of single chromaffin granules in cell-free inside-out membrane patches. Nat Cell Biol. 2003;5:358–362. [PubMed: 12652310]
10.
Alés E, et al. High calcium concentrations shift the mode of exocytosis to the kiss-and-run mechanism. Nat Cell Biol. 1999;1:40–44. [PubMed: 10559862]
11.
Tabares L, et al. Exocytosis of catecholamine-containing and catecholamine-free granules in chromaffin cells. J Biol Chem. 2001;276:39974–39979. [PubMed: 11524425]
12.
Gong LW, Alvarez De Toledo G, Lindau M. Secretory vesicles membrane area is regulated in tandem with quantal size in chromaffin cells. J Neurosci. 2003;23(21):7917–7921. [PMC free article: PMC6740609] [PubMed: 12944522]
13.
Mosharov EV. Intracellular patch electrochemistry: Regulation of cytosolic catecholamines in chromaffin cells. J Neurosci. 2003;23:5835–5845. [PMC free article: PMC6741260] [PubMed: 12843288]
14.
Alvarez de Toledo G, Fernández-Chacón R, Fernandez JM. Release of secretory products during transient vesicle fusion. Nature. 1993;363:554–558. [PubMed: 8505984]
15.
Chow RH, et al. Mechanisms determining the time course of secretion in neuroendocrine cells. Neuron. 1996;16:369–376. [PubMed: 8789951]
16.
Oberhauser AF, Robinson IM, Fernandez JM. Simultaneous capacitance and amperometric measurements of exocytosis: A comparison. Biophys J. 1996;71:1131–1139. [PMC free article: PMC1233568] [PubMed: 8842250]
17.
Robinson IM, et al. Colocalization of calcium-entry and exocytotic release sites in adrenal chromaffin cells. Proc Natl Acad Sci USA. 1995;92:2474–2478. [PMC free article: PMC42240] [PubMed: 7708668]
18.
Chow RH, Rüden Lv. Electochemical Detection of Secretion from Single Cells, in Single Channel Recording. Sakmann B, Neher E, editors. Plenum Press; New York: 1995.
19.
Wightman RM, et al. Time course of release of catecholamines from individual vesicles during exocytosis at adrenal medullary cells. Biophys J. 1995;68:383–390. [PMC free article: PMC1281698] [PubMed: 7711264]
20.
Schroeder TJ, et al. Temporally resolved, independent stages of individual exocytotic secretion events. Biophys J. 1996;70(2):1061–1068. [PMC free article: PMC1225008] [PubMed: 8789125]
21.
Amatore C, et al. Interplay between membrane dynamics, diffusion and swelling pressure governs individual vesicular exocytotic events during release of adrenaline by chromaffin cells. Biochimie. 2000;82(5):481–496. [PubMed: 10865134]
22.
Hafez I, et al. Electrochemical imaging of fusion pore openings by electrochemical detector arrays. Proc Natl Acad Sci USA. 2005;102:13879–13884. [PMC free article: PMC1236545] [PubMed: 16172395]
23.
Debus K, Lindau M. Resolution of patch capacitance recordings and of fusion pore conductances in small vesicles. Biophys J. 2000;78:2983–2997. [PMC free article: PMC1300882] [PubMed: 10827977]
24.
Hamill OP, et al. Improved patch-clamp technique for high-resolution current recording from cells and cell-free membrane patches. Pflügers Arch Eur J Physiol. 1981;391:85–100. [PubMed: 6270629]
25.
Schroeder TJ, et al. Analysis of diffusional broadening of vesicular packets of catecholamines released from biological cells during exocytosis. Anal Chem. 1992;64:3077–3083. [PubMed: 1492662]
26.
Klocke V. Atomic precision and millimeter range. F & M. 1996;104 (4):274.
27.
Penner R. A Practical Guide to Patch Clamping, in Single Channel Recording. Sakmann B, Neher E, editors. Plenum Press; New York: 1995.
28.
Lindau M. Time-resolved capacitance measurements: Monitoring exocytosis in single cells. Q Rev Biophys. 1991;24:75–101. [PubMed: 2047522]
29.
Breckenridge LJ, Almers W. Currents through the fusion pore that forms during exocytosis of a secretory vesicle. Nature. 1987;328:814–817. [PubMed: 2442614]
30.
Hartmann J, Lindau M. A novel Ca2+-dependent step in exocytosis subsequent to vesicle fusion. Fed Eur Biochem Soc Lett. 1995;363:217–220. [PubMed: 7737405]
31.
Lindau M, Alvarez de Toledo G. The fusion pore. Biochim Biophys Acta. 2003;1641:2–3. 167–173. [PubMed: 12914957]
32.
Gomez JF, et al. New approaches for analysis of amperometrical recordings. Ann N Y Acad Sci. 2002;971:647–654. [PubMed: 12438200]
33.
Travis ER, Wightman MR. Spatio-temporal resolution of exocytosis from individual cells. Ann Rev Biophys Biomol Struct. 1998;27:77–103. [PubMed: 9646863]
34.
Mosharov EV, Sulzer D. Analysis of exocytotic events recorded by amperometry. Nat Methods. 2005;2(9):651–658. [PubMed: 16118635]
35.
Tabares L, Lindau M, Alvarez De Toledo G. Relationship between fusion pore opening and release during mast cell exocytosis studied with patch amperometry. Biochem Soc Trans. 2003;31(4):837–841. [PubMed: 12887317]
36.
Jankowski JA, Finnegan JM, Wightman RM. Extracellular ionic composition alters kinetics of vesicular release of catecholamines and quantal size during exocytosis at adrenal medullary cells. J Neurochem. 1994;63:1739–1747. [PubMed: 7931329]
37.
Walker A, Glavinovic MI, Trifaró JM. Time course of release of content of single vesicles in bovine chromaffin cells. Pflügers Arch Eur J Physiol. 1996;431:729–735. [PubMed: 8596723]
38.
Lindan M, Neher E. Patch-clamp techniques for time-resolved capacitance measurements in single cells. Pflugers Arch. 1988;411(2):137–146. [PubMed: 3357753]
39.
Gillis KD. Techniques for membrane capacitance measured, in single channel recording. Sakmann B, Neher E, editors. Plenum Press; New York: 1995.
Bookshelf ID: NBK2563PMID: 21204378

### Related information

• PMC
PubMed Central citations
• PubMed

### Similar articles in PubMed

See reviews...See all...