Recordings from human myenteric neurons using voltage-sensitive dyes

Journal of Neuroscience Methods 192 (2010) 240–248

Contents lists available at ScienceDirect

Journal of Neuroscience Methods journal homepage: www.elsevier.com/locate/jneumeth

Recordings from human myenteric neurons using voltage-sensitive dyes Sheila Vignali a,1 , Nadine Peter a,1 , Güralp Ceyhan b , Ihsan Ekin Demir b , Florian Zeller c , David Senseman d , Klaus Michel a , Michael Schemann a,∗ a

Human Biology, Technische Universität München, Liesel-Beckmann-Strasse 4, 85354 Freising-Weihenstephan, Germany Department of Surgery, Technsiche Universität München, München, Germany c Department of Surgery, Clinical Centre Freising, Freising, Germany d Department of Biology, UT San Antonio, San Antonio, TX, USA b

a r t i c l e

i n f o

Article history: Received 15 April 2010 Received in revised form 2 July 2010 Accepted 29 July 2010 Keywords: Enteric nervous system Submucous plexus Myenteric plexus Neuroimaging Voltage-sensitive dyes Di-8-ANEPPS Myenteric neuron culture Human tissue

a b s t r a c t Voltage-sensitive dye (VSD) imaging became a powerful tool to detect neural activity in the enteric nervous system, including its routine use in submucous neurons in freshly dissected human tissue. However, VSD imaging of human myenteric neurons remained a challenge because of limited visibility of the ganglia and dye accessibility. We describe a protocol to apply VSD for recordings of human myenteric neurons in freshly dissected tissue and myenteric neurons in primary cultures. VSD imaging of guinea-pig myenteric neurons was used for reference. Electrical stimulation of interganglionic fiber tracts and exogenous application of nicotine or elevated KCl solution was used to evoke action potentials. Bath application of the VSDs Annine-6Plus, Di-4-ANEPPS, Di-8-ANEPPQ, Di-4-ANEPPDHQ or Di-8-ANEPPS revealed no neural signals in human tissue although most of these VSD worked in guinea-pig tissue. Unlike methylene blue and FM1–43, 4-Di-2-ASP did not influence spike discharge and was used in human tissue to visualize myenteric ganglia as a prerequisite for targeted intraganglionic VSD application. Of all VSDs, only intraganglionic injection of Di-8-ANEPPS by a volume controlled injector revealed neuronal signals in human tissue. Signal-to-noise ratio increased by addition of dipicrylamine to Di-8-ANEPPS (0.98 ± 0.16 vs. 2.4 ± 0.62). Establishing VSD imaging in primary cultures of human myenteric neurons led to a further improvement of signal-to-noise ratio. This allowed us to routinely record spike discharge after nicotine application. The described protocol enabled reliable VSD recordings from human myenteric neurons but may also be relevant for the use of other fluorescent dyes in human tissues. © 2010 Elsevier B.V. All rights reserved.

1. Introduction The enteric nervous system (ENS) primarily consists of two ganglionated plexi which are embedded in the gut wall. The ENS acts as an autonomous nervous system which controls the main gut functions. This is achieved by the submucous plexus which mainly regulates mucosal functions while the myenteric plexus, located between the two muscle layers, coordinates smooth muscle activity. Electrophysiological properties of myenteric neurons have been studied with conventional microelectrode techniques, mainly in small laboratory animals (Wood, 1970, 1973; Ohkawa and Prosser, 1972; Nishi and North, 1973; Hirst et al., 1974; Furukawa et al., 1986; Brookes et al., 1988; Browning and Lees, 1996; Cornelissen et al., 2001). Such studies have advanced our knowledge on basic functions of the ENS in guinea-pigs, cats, mice, rats and pig, but

∗ Corresponding author. Tel.: +49 08161 71 5483; fax: +49 08161 71 5785. E-mail address: [email protected] (M. Schemann). 1 These authors contributed equally. 0165-0270/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jneumeth.2010.07.038

did also highlight species-specific ENS functions. There has been much less progress in the neurobiology of the human ENS as the basis to understand pathophysiology of functional, structural and inflammatory gastrointestinal diseases. This is mainly due to technical obstacles preventing the routine application of conventional microelectrode techniques to human intestinal tissue. In pioneering studies Obaid et al. (1992, 1999) and Neunlist et al. (1999), adapted multi-site optical imaging systems in combination with a voltage-sensitive dye (VSD) to record from enteric neurons in the guinea-pig. With this technique, we were able to also routinely record from human submucous neurons with a high spatial and temporal resolution that revealed fast changes in membrane potential such as action potentials and fast excitatory postsynaptic potentials (EPSPs) (Schemann et al., 2002, 2005; Michel et al., 2005; Breunig et al., 2007; Buhner et al., 2009). However, recordings from the human myenteric ganglia remained a challenge because of their limited visibility and dye accessibility. So far only two studies on the electrophysiology of human myenteric neurons have been published. They used conventional intracellular recordings to investigate basic properties of cultured myenteric neurons (Maruyama, 1981) or myenteric neu-

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

rons from freshly dissected human colon (Brookes et al., 1987). Both studies had limitations: the number of studied neurons was rather small with 2 and 43, respectively. Many years required to record from such small number of neurons are one explanation why such studies have not been followed up. In addition, the study by Maruyama used cultured fetal myenteric neurons which very likely do not reflect behaviour of mature enteric neurons. We have already demonstrated that in principle it is feasible to record from human myenteric neurons using voltage-sensitive dye imaging (Schemann et al., 2002). However, we were only successful in one tissue and reliable recordings of reproducible signals from human myenteric neurons remained a challenge. Therefore, it was the aim of our study to establish a protocol that allows the successful use of VSD imaging for recordings of human myenteric neurons. The protocols were tested in freshly dissected human tissues as well as in primary cultures of human myenteric neurons. We investigated the usefulness of several voltage-sensitive dyes both in human as well as in guinea-pig myenteric plexus. As visibility of intact, non-cultured, human myenteric ganglia is a limitation we used several vital dyes to pre-stain the ganglia. Myenteric plexus preparations from guinea-pigs were used to test whether these dyes alter electrophysiological behaviour of myenteric neurons which would exclude their use in human tissue.

2. Materials and methods 2.1. Guinea-pig myenteric plexus preparations For the experiments we used male Dunkin Hartley guineapigs (Charles River Laboratories, Kisslegg, Germany; Harlan GmbH, Borchen, Germany). After killing the animals by cervical dislocation followed by exsanguination, the ileum was quickly removed and dissected in ice cold Krebs solution to obtain longitudinal muscle myenteric plexus preparations. The procedures were according to the German animal protection law. During preparation the tissue was constantly perfused with ice cold Carbogen (5% CO2 , 95% O2 ) gassed Krebs solution (pH 7.4) containing (in mM): 117 NaCl, 4.7 KCl, 1.2 MgCl2 , 1.2 NaH2 PO4 , 25 NaHCO3 , 2.5 CaCl2 and 11 glucose. The preparation was placed in a recording chamber and continuously perfused with 37 ◦ C Krebs solution (pH 7.4) containing (in mM): 117 NaCl, 4.7 KCl, 1.2 MgCl2 , 1.2 NaH2 PO4 , 20 NaHCO3 , 2.5 CaCl2 and 11 glucose. All chemicals were obtained from Sigma–Aldrich (Schnelldorf, Germany).

2.1.1. Screening vital dyes to visualize human myenteric ganglia To check if the dyes used for pre-staining human myenteric ganglia had any unspecific effects on the electrophysiological behaviour, we performed the following tests in freshly dissected guinea-pig longitudinal muscle myenteric plexus preparations. After staining the ganglia with 20 ␮M of the voltage-sensitive dye Di-8-ANEPPS, we evoked control fast EPSPs and compound action potentials by electrical stimulation of interganglionic fiber tracts with supramaximal stimulus strength using a Teflon-coated platinum electrode with a diameter of 25 ␮m (101R-1T; Science Products GmbH, Hofheim, Germany). The tissues were then incubated with the vital dyes methylene blue (50 ␮M for 15 min), FM1–43 (4 ␮M for 10 min) which labels nerve terminals during synaptic activity (Betz and Bewick, 1992) or 4-Di-2-ASP (2.5 ␮M for 10 min) which is a cationic mitochondrial dye staining nerve terminals (Magrassi et al., 1987). After the incubation we performed another electrical fiber tract stimulation to investigate changes in compound action potentials and fast EPSPs.

241

2.1.2. Screening different voltage-sensitive dyes to record from human myenteric ganglia In our previous experiments in guinea-pig and human submucous neurons we exclusively used the voltage-sensitive dye Di-8-ANEPPS (Neunlist et al., 1999; Schemann et al., 2002). As it is known that membrane properties of different cell types determine the suitability of voltage-sensitive dyes (Cohen and Salzberg, 1978; Waggoner, 1979; Fluhler et al., 1985; Loew et al., 1992; Loew, 1996; Wu et al., 1998; Zochowski et al., 2000) we tested several voltagesensitive dyes for their use in human myenteric neurons. All these dyes were first tested in guinea-pig myenteric plexus preparations to check tissue dye accessibility and their suitability to record action potentials. In guinea-pig preparations we used 2.5–50 ␮M ANNINE-6Plus which is a water soluble voltage-sensitive dye displaying strong binding to lipid membranes (Kuhn and Fromherz, 2003; Fromherz et al., 2008). We used Di-4-ANEPPDHQ which is less phototoxic, more water soluble and leads to larger fluorescence signals than Di-8-ANEPPS (Obaid et al., 2004) at concentrations of 7.5–75 ␮M. Di-4-ANEPPS, which incorporates into the cell membrane faster than Di-8-ANEPPS (Fluhler et al., 1985; Fromherz and Lambacher, 1991; Bedlack et al., 1992) was used at 10 ␮M. Another analogue of Di-8-ANEPPS is the more hydrophobic dye Di-8-ANEPPQ (Tsau et al., 1996) which was used at a concentration of 20 ␮M. ANNINE-6Plus and Di-4-ANEPPDHQ were tested in addition with and without Pluronic F-127 (Invitrogen, Karlsruhe, Germany). To increase the signal-to-noise ratio for experiments in freshly dissected human myenteric plexus preparations we used DPA (dipicrylamine) as a resonance-energy-transfer acceptor as described elsewhere (Bradley et al., 2009; Sjulson and Miesenböck, 2008; Chanda et al., 2005a,b). DPA is a non-fluorescent hydrophobic anion, which localizes at the lipid–aqueous interface and can be used as a classical Förster-energy-transfer acceptor from variety of donor fluorophores (Chanda et al., 2005a). All voltage-sensitive dyes were purchased from Invitrogen except methylene blue (Sigma–Aldrich). ANNINE-6plus was generously supplied by Prof. Dr. Peter Fromherz (Department of Membrane and Neurophysics; Max Planck Institute, Munich). 2.2. Human myenteric plexus preparations The studies on human myenteric neurons were performed using surgical specimens from patients undergoing abdominal surgery in the Departments of Surgery at the Medical Clinic in Freising and at the Medical Clinic of the Technische Universität München. Samples were taken from macroscopically normal, unaffected areas as determined by visual inspection of the pathologist. The protocol was approved by the ethic committee of the Technische Universität München (project approval: 1746/07). 2.2.1. Freshly dissected human tissue After removal, the surgical specimens were placed in ice cold, oxygenated, sterile Krebs solution and immediately transferred to the laboratory. We isolated the myenteric plexus by gently removing the mucosa, submucosa and the circular and longitudinal muscle layer. During the dissection the tissue was constantly perfused with ice cold Carbogen gassed Krebs solution. The myenteric plexus preparation was then pinned onto a silicone ring (Down Cornig, Midland, TX, USA), placed in the recording chamber and continuously perfused with 37 ◦ C Krebs solution. To visualize the ganglia, we first incubated the tissue for 10–15 min with 4-Di-2-ASP (2.5 ␮M) followed by 5–15 min wash out time in order to visualize the ganglia. For human myenteric plexus we used three staining protocols to label ganglia with voltage-sensitive dyes. The first approach was to stain individual ganglia by local pressure application of the dye

242

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

through a fine tipped glass pipette which was positioned inside an interganglionic fiber tract close to the site where it entered the ganglion. Pressure pulses of several 100 ms were used to inject the dye. The staining was followed by a 5–10 min incubation period to allow the dye to incorporate into the cell membrane before starting the experiments. The second approach was to use a volume controlled injector (UltraMicroPump II, World precision instruments, Sarasota, FL, USA) to label single ganglia by local intraganglionic perfusion for up to 10 s. Application was followed by a 10–30 min incubation time to allow the dye to incorporate into the cell membrane before starting the experiments. For the third approach we labeled individual ganglia by local intraganglionic perfusion with a mixture of 20 ␮M Di-8-ANEPPS and 0.1 ␮M dipicrylamine (DPA). Recordings were started after a 20 min incubation time. Tissues were constantly perfused with Krebs solution containing 0.5 ␮M DPA. In freshly dissected human myenteric plexus preparations we used 30 ␮M Di-8-ANEPPS with different concentrations of Pluronic F-127 (0.0135–0.0675%), 2.5–10 ␮M ANNINE-6Plus, 20 ␮M Di-4ANEPPS, 20 ␮M Di-8-ANEPPQ or 75 ␮M of Di-4-ANEPPDHQ. Interganglionic fiber tracts were electrically stimulated with 2 V for 600 ␮s (400 ␮s in guinea-pig myenteric plexus preparation) using a Teflon-coated platinum wire electrode with a diameter of 75 ␮m (101R-3T; Science Products GmbH) placed onto an interganglionic fiber tract. 2.2.2. Primary culture of the human myenteric plexus Primary cultures of human myenteric neurons were prepared using the previously published protocol for guinea-pig small intestine with a few modifications (Vanden Berghe et al., 2000). Briefly, we first dissected the tissue in ice cold Krebs solution under sterile conditions to obtain the myenteric plexus as described above. The dissected plexus was then digested in an enzymatic solution containing 0.9 mg ml−1 protease from bovine pancreas (Type I, crude) and 1.2 mg ml−1 collagenase from clostridium histolyticum (Type II) (both from Gibco, Karlsruhe, Germany). After a 25–45 min incubation at 37 ◦ C the enzymatic reaction was stopped by adding ice cold sterile Krebs solution. Following the centrifugation of the suspension at 500 × g, 15 ganglia were picked and plated in culture dishes (␮-dish 35 mm; Ibidi, Martinsried, Germany) in which they adhered to the bottom. The culture medium was composed of the growth medium M199 which was enriched with 10% fetal bovine serum (FBS; both from Gibco) and 50 or 100 ng ml−1 mouse nerve growth factor 7 s (Alomone Laboratories, Jerusalem, Israel) and was changed every 2–3 days. The culture medium contained an elevated glucose concentration (30 mM) and 100 U ml−1 penicillin and 100 ␮g ml−1 streptomycin (both from Gibco). The culture wells were kept in an incubator at 37 ◦ C, 5% CO2 and 100% humidity. In order to prevent the proliferation of glial cells and fibroblasts the culture medium was supplemented with 2 ␮M cytosine ␤-darabinfuranoside (Sigma–Aldrich). Experiments were performed in at least 14-day-old cultures which were transferred to a recording chamber. During the dye imaging experiments the cells were constantly perfused with a 37 ◦ C warm modified Krebs solution containing (in mM): 150 NaCl, 5 KCl, 1 MgCl2 , 2 CaCl2 , 10 glucose, 10 HEPES (all from Sigma–Aldrich). With NaOH the pH of the solution was adjusted to 7.4. 2.3. Pharmacology Hexamethonium (200 ␮M; Sigma–Aldrich) was perfused for 20 min to block action of acetylcholine at nicotinic receptors. Tetrodotoxin (0.5 ␮M; TTX; Tocris Bioscience, Bristol, UK) was perfused for 20 min to investigate neural origin of the signals.

Local microejection of nicotine and Krebs solution containing high concentrations of potassium chloride (50–100 mM) were used to evoke spike discharge. The drugs were applied to single ganglia by pressure ejection from micropipettes (3–20 psi, up to 800 ms duration, ejection speed 55 ± 27 nl s−1 , approximately 200 ␮m from the ganglion) closely positioned to myenteric ganglia. We previously showed that any substance applied via pressure ejection pulses will be diluted by 1:200 (20 ms pulse duration) to 1:8 (400 ms pulse duration) once it reaches the ganglion (Breunig et al., 2007) 2.4. Voltage-sensitive dye (VSD) imaging The multi-site optical recording system to image neural activity in human tissue was previously described in detail (Neunlist et al., 1999; Schemann et al., 2002). Briefly, the freshly dissected tissues (2 cm × 1 cm) were pinned onto a silicone ring that was placed in a recording chamber with a 42 mm diameter glass bottom (130–170 ␮m thickness, Sauer, Reutlingen, Germany) and continuously perfused with Carbogen gassed 37 ◦ C Krebs solution. The recording chamber was mounted onto an epifluorescence Olympus IX 71 microscope (Olympus, Hamburg, Germany) equipped with a 75 W xenon arc lamp (Optosource, Cairn Research Ltd., Faversham, UK). Controlled illumination of the preparation was achieved by a software operated shutter (Uniblitz D122, Vincent Associates, New York, USA). Ganglia stained with voltage-sensitive dyes were visualized with a 40× oil immersion objective (UAPO/340, NA = 1.4; Olympus) by using a fluorescence filter cube consisting of a 450 ± 50 nm excitation interference filter, a 565 nm dichroic mirror and a 515 nm barrier filter (Olympus) for ANNINE-6Plus (Fromherz et al., 2008) and a fluorescence filter cube consisting of a 545 ± 15 nm excitation interference filter, a 565 nm dichroic mirror and a 580 nm barrier filter (AHF Analysentechnik, Tübingen, Germany) for all other dyes. For the recordings of Di-8-ANEPPS in combination with dipicrylamine (DPA) measurements we used a FITC-filter cube equipped with a 464.5–499.5 nm excitation interference filter, a BS506 dichroic mirror and a 516–556 nm barrier filter (Olympus). Signals were acquired with a frequency of 1.6 kHz, which enables the detection of action potentials, and were processed by a cooled charge-coupled device camera made of 80 × 80 pixels (RedShirt Imaging, Decatur, GA, USA). The VSD imaging setup allows measurements of relative changes in the fluorescence (F/F, the denominator F indicates resting light level while F is the change in fluorescence) which is linearly related to changes in the membrane potential (Neunlist et al., 1999). The Neuroplex software enabled signal filtering. A high pass Butterworth filter was used to reduce mainly dye bleaching and some minor mechanical noise. A low pass Butterworth filter was used to reduce high frequency electrical noise. The number of action potentials was not affected by filtering. 2.5. Immunohistochemistry To determine whether recordings were obtained from enteric neurons we performed immunohistochemistry in the freshly dissected human tissues as well as in the human primary cultures. Both were fixed over night at room temperature in a 37% phosphatebuffered formalin solution followed by several washes (3× 10 min) in phosphate buffer. The samples were then incubated for 1 hr in the pre-incubation solution containing phosphate-buffered saline (PBS) plus Triton X-100 (0.5%), PBS/NaN3 -solution (0.1%) and normal horse serum (4%). After three washing steps (3× 10 min), the primary antibodies, diluted in the pre-incubation solution, were added to the samples. Primary antibodies used to visualize enteric neurons were sheep anti-protein gene product 9.5 (PGP; The Binding Site, Birmingham, UK) and rabbit anti-neuron-specific enolase

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

243

(NSE; Polysciences, Epplheim, Germany) at dilutions of 1: 5000 and 1:2000, respectively. The incubation with the primary antibodies lasted in the case of the tissues 40 h and for the primary cultures 12–16 h. After the incubation the samples were rinsed three times for 10 min with PBS and then exposed to the species-specific secondary antibodies, also diluted in the pre-incubation solution: Cy2-labeled donkey anti-rabbit (1:200) and Cy5-labeled donkey anti-sheep (1:500) (both from Dianova, Hamburg, Germany). The incubation with the secondary antibodies lasted 3.5 h for dissected human tissue and 1.5–2 h for primary culture. Finally, samples were rinsed three times for 10 min with PBS and then mounted in Citifluor AF 1 (Citifluor Ltd, London, United Kingdom). The staining was evaluated with a fluorescence microscope (BX61WI; Olympus, Tokyo, Japan) using appropriate filter blocks. The microscope was equipped with a camera (F-View) and image software (analySIS; both from Soft imaging system, Münster, Germany) for acquisition, processing and editing. 2.6. Data analysis and statistics Individual cells can be identified since the dye incorporates into the membrane revealing the outline of individual cell bodies (Michel et al., 2005). The overlay of signals and ganglion image allowed the analysis of the responses from single neurons. We thereby identified and analysed the number of cells per ganglion in the field of view and the number of responding cells per ganglion. We also analysed the amplitude of compound action potentials and fast EPSPs. For signal analysis we used Neuroplex 8.3 (RedShirt Imaging), Igor Pro 6.03 (Wavemetrics Inc, Lake Oswego, OR, USA). All the statistic analyses were performed with the software Sigmastat 3.1 and Sigmaplot 9.0 software (both from Systat Software Inc, Erkrath, Germany). All data are presented as mean ± standard deviation or as median value together with the 25% and 75% quartiles, if they are not normally distributed. To test the differences in the amplitudes of compound action potentials and fast EPSPs as well as in the spike frequencies after drug application we performed Student’s ttest or rank sum test if the data were not normally distributed. For all analysis the difference between two data groups was significant when p was <0.05. The statistical analysis of the signal-to-noise ratio was done as follows. For each action potential the minimum and maximum values were determined. Similar procedure was used to measure noise amplitude. The resulting values were averaged, divided and presented as mean ± standard deviation.

Fig. 1. Control studies performed in guinea-pig myenteric neurons demonstrate that 4-Di-2-ASP is a useable tool for the visualization of the myenteric plexus. (A) Pre-staining with 4-Di-2-ASP did not change the amplitudes of compound action potentials or fast EPSPs. Compound action potentials are responsible for the initial spike, while the succeeding depolarization represents fast EPSPs. (B) After the prestaining with methylene blue the amplitudes of compound action potentials and fast EPSPs were significantly reduced. (C) Similar to the findings with methylene blue, FM1–43 caused a significant reduction in the amplitudes of compound action potentials and fast EPSPs. () Compound action potential (AP); ( ) fast EPSP. *p < 0.001. The representative traces illustrate in each case the response of one myenteric nerve cell body to the electrical stimulation (arrow) of interganglionic fiber tracts with 2 V for 400 ␮s prior (control) and after (treated) the pre-staining with a vital dye. All traces have been filtered with a low pass filter at 200 Hz.

3. Results 3.1. Visualization (pre-staining) of the human myenteric plexus Guinea-pig myenteric ganglia can be easily identified in living tissue by conventional light microscopy. In contrast, myenteric ganglia cannot be visualized in freshly dissected, non-stained vital human tissue preparations. This however is crucial to load ganglia by intraganglionic pressure ejection. In initial experiments we found that bulk loading of the tissues by long term dye incubation did not lead to any specific staining, very likely because of limited dye penetration and accessibility to the neurons. Therefore, we first had to find a dye that stains the human myenteric ganglia without affecting basic electrophysiological properties of myenteric neurons. We therefore performed experiments in Di-8-ANEPPS labeled guinea-pig myenteric plexus preparations after pre-staining the ganglia with the vital dyes 4-Di-2-ASP, methylene blue or FM1–43. All of them have in common that they revealed the ganglionic

network. We used electrical stimulation of interganglionic fiber tracts to evoke nerve signals. The signal consisted of two components as previously demonstrated (Schemann et al., 2002). The first component was a compound action potential caused by axonal spikes while the second component was due to synaptic release of acetylcholine, which activates postsynaptic nicotinic receptors to generate fast EPSPs. The results of these experiments are summarized in Fig. 1. Methylene blue (3 tissues, 60 neurons) and FM1–43 (4 tissues, 107 neurons) are both not recommendable as they caused a significant decrease in the amplitudes of the compound action potentials and fast EPSPs (p < 0.001). In contrast 4-Di-2-ASP (4 tissues, 81 neurons) did not alter the amplitudes of the signals (p = 0.870 for compound action potentials and p = 0.622 for fast EPSPs). Therefore, we concluded that 4-Di-2-ASP is the dye of choice to visualize human myenteric ganglia.

244

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

Fig. 2. Potential of different voltage-sensitive dyes for recording signals in human myenteric neurons. This figure illustrates the ability or failure of a particular voltagesensitive dye to stain guinea-pig (GP) or human (HU) myenteric ganglia and to reveal signals. The images illustrate the dye labeling and the traces show signals from single neurons in response to exogenous application (time of application indicated by bars underneath the traces) of 50 mM KCl containing Krebs solution, 100 ␮M nicotine or after electrical stimulation of interganglionic fiber tracts. (A) ANINNE-6Plus revealed a strong staining of guinea-pig and human myenteric ganglia together with a low background staining. But only in guinea-pig myenteric neurons it was possible to observe spike discharge after the application of 50 mM KCl (trace filtered with a low pass filter at 199 Hz and a high pass filter at 6 Hz). (B) With Di-4-ANEPPDHQ action potential discharge after nicotine application was only observed in guinea-pig tissues (trace filtered with a low pass filter at 70 Hz and a high pass filter at 45 Hz). (C) No signals could be detected with Di-4-ANEPPS because of immediate internalization of the dye in both guinea-pig and human myenteric neurons. (D) Using Di-8-ANEPPQ action potential discharge after nicotine application was only observed in guinea-pig myenteric neurons (trace filtered with a low pass filter at 199 Hz and a high pass filter at 6 Hz). Panels (E) and (F) illustrate the superior properties of Di-8-ANEPPS in comparison to the other dyes. (E) Image of a Di-8-ANEPPS labeled guinea-pig myenteric ganglion with clearly visible outlines of individual neurons. The top trace next to the image shows the response of one neuronal cell body (marked by white arrow in the image) to electrical stimulation of an interganglionic fiber tract (time of stimulation indicated by arrow). The first peak represents a compound action potential while the second signal is a subthreshold fast EPSPs. The bottom trace shows spike discharge in response to nicotine application. (F) Image of a Di-8-ANEPPS labeled human myenteric ganglion. Although not as clear as in guinea-pig myenteric ganglia Di-8-ANEPPS reveals the outline of ganglion cells. The top trace next to the image shows the response of one neuronal cell body (marked by white arrow in the image) to electrical stimulation of an interganglionic fiber tract (time of stimulation indicated by arrow). The stimulation evoked two spikes. The bottom trace shows spike discharge in response to nicotine application (two spikes marked by arrows).

3.2. Examination of voltage-sensitive dyes to record from myenteric neurons We tested various dyes in both guinea-pig and human myenteric neurons to find the one best suited for recordings in human myenteric neurons (Fig. 2). Compared to the other tested dyes, we achieved the highest staining intensity of the ganglia together with a low background staining with ANNINE-6Plus when supplemented with 0.0135% Pluronic F-127. However, only in the

guinea-pig myenteric plexus we could record compound action potentials, fast EPSPs, KCl and nicotine evoked spike discharge (Fig. 2A). Individual ganglia were stained with Di-4-ANEPPDHQ by tissue incubation or by intraganglionic application. In guinea-pig myenteric plexus we were able to record compound action potentials, fast EPSPs and nicotine evoked spike discharge. In human myenteric plexus the staining was insufficient and no signals could be recorded (Fig. 2B).

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

245

Fig. 3. The signal-to-noise ratio of Di-8-ANEPPS stained human myenteric neurons is significantly lower than in guinea-pig myenteric neurons. Signal-to-noise ratios of spontaneous (A) as well as nicotine evoked action potential discharge (B) were significantly less in human compared to guinea-pig myenteric neurons, although the addition of 0.1 ␮M DPA to Di-8-ANEPPS improved the signal-to-noise ratio and single action potentials can be clearly identified (compare to trace in Fig. 2F) The trace labeled “control” illustrates lack of spontaneous action potential discharge of the neuron where nicotine was applied (trace above control trace) Traces have been filtered with a low pass filter at 199 Hz and high pass filter at 6 Hz. Bars below traces mark exogenous application of nicotine. Arrows mark action potentials in traces from human myenteric neurons.

In guinea-pig as well as in human myenteric plexus we could not record any signals after staining with Di-4-ANEPPS due to the almost immediate internalization of this dye (Fig. 2C). We also tested Di-8-ANEPPQ and Di-8-ANEPPS on guinea-pig and human myenteric plexus preparations. In guinea-pig myenteric neurons we could routinely record compound action potentials, fast EPSPs and nicotine evoked spike discharge with Di-8-ANEPPQ (Fig. 2D) as well as with Di-8-ANEPPS (Fig. 2E). In human myenteric neurons Di-8-ANEPPQ revealed no signals (Fig. 2D). Applying the same staining protocol as in the guinea-pig revealed no signals from Di-8-ANEPPS stained ganglia. By elevating the Di-8-ANEPPS concentration to 30 ␮M it was possible to record signals in human myenteric neurons. However, the responses to nicotine application showed a relatively poor signal-to-noise ratio of ≤1. There was also no improvement after increasing the Pluronic concentration to 0.027%, 0.0405%, 0.054% or 0.0675%, respectively. We did not use higher concentrations as they impair signal generation in enteric neurons (Neunlist et al., 1999). We therefore labeled single ganglia with the volume controlled injector for 10 s with a speed of 1000 nl s−1 resulting in an injected volume of about 10 ␮l. This approach revealed a stronger staining and an improved signal-to-noise ratio (1.5 ± 0.28) which enabled us to record nicotine and KCl evoked spike discharge as well as compound action potentials and fast EPSP triggered action potentials in human myenteric neurons (Fig. 2F). Although the signal-to-noise ratio could be improved from 0.98 ± 0.16 to 1.5 ± 0.28, it still remained significantly lower than in preparations from guinea-pigs (2.66 ± 1.18, Fig. 4). It is unlikely that pre-staining with 4-Di-2-ASP negatively affected signal-to-noise ratio as dye labeling disappeared after a 10 min wash out. Best results were achieved by intraganglionic injection of a mixture of 20 ␮M Di-8-ANEPPS and 0.1 ␮M DPA with the volume controlled injector for 10 s (Fig. 3B). The signal-to-noise ratio could be increased to 2.4 ± 0.62 and nicotine evoked action potentials could be clearly recorded (Fig. 3B). Consequently, Di-8-ANEPPS was the most suitable dye to record from human myenteric neurons. This is further supported by the electrical stimulation of interganglionic fiber tracts which often evoked two peaks (Figs. 2F and 4B). The immunohistochemical staining for the pan-neuronal marker NSE revealed that the recorded signals indeed originated

from myenteric nerve cell bodies (Fig. 4A). Further we noticed the patchy distribution of nerve cells in human myenteric ganglia. Neurons were found in clusters rather than being evenly distributed throughout the ganglia as it is the case in guinea-pig myenteric ganglia. Also different to guinea-pig myenteric plexus preparations, we have not observed subthreshold fast EPSPs in human myenteric neurons. The blockade of the neuronal conductance with 0.5 ␮M TTX for 20 min fully abolished both peaks, supporting the neural origin of the signals (Fig. 4). The propagation velocity of electrically evoked compound action potentials were 0.17 ± 0.12 m/s which corresponds to unmyelinated C-fibers. To check whether the second peak is a fast EPSP triggered action potential, the tissue was perfused for 20 min with 200 ␮M Hexamethonium. However, even if this second peak was reduced in some experiments, no complete blockade could be achieved. Although we were able to record action potentials in myenteric neurons of freshly dissected tissues after application of 50 to 100 mM KCl or 100 ␮M nicotine the signal-to-noise ratio was never as good as for guinea-pig myenteric neurons. Thus, it was our aim to establish primary cultures of human myenteric neurons in order to further improve dye accessibility and the signal-to-noise ratio. 3.3. Responses of myenteric neurons in human primary cultures to nicotine In Fig. 5 a 14-day-old primary culture of human myenteric neurons is illustrated. After bulk staining of the neurons with 10 ␮M Di-8-ANEPPS for 10 min (Fig. 5A), we were able to record action potential discharge after microejection of 100 ␮M nicotine (Fig. 5B). Both the staining intensity and the signal-to-noise ratio were superior to those observed in freshly dissected tissue. The nicotine evoked response was reversibly blocked by hexamethonium (Fig. 5D and E). Those cells which fired action potentials were always immunoreactive for the neuronal marker PGP (Fig. 5C). 4. Discussion This is the first report on reproducible recordings from human myenteric neurons with voltage-sensitive dye imaging paving the way for further in depth studies on the neurobiology of human myenteric neurons.

246

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

Fig. 4. Verification of the neuronal origin of signals induced by electrical stimulation of interganglionic fiber tracts. (A) Human myenteric ganglion stained with Di-8-ANEPPS (upper image) followed by immunohistochemical demonstration of the pan-neuronal marker NSE (lower image). NSE-positive nerve cell bodies (arrows) are located in the upper right part of the ganglion. Next to the images are the responses of the two neurons marked by the white arrows to electrical stimulation of interganglionic fiber tracts (black arrows indicate time of electrical stimulation). Panel (B) shows the response in another myenteric neuron to electrical stimulation of an interganglionic fiber tract (black arrows indicate time of electrical stimulation). The upper trace shows the response consisting of two peaks under control (Ctr) condition. Both peaks are fully blocked in the presence of 0.5 ␮M tetrodotoxin (TTX, middle trace) and recover 45 min after wash out of TTX (lower trace). The graph (C) is a summary of such experiments in 7 tissues and 9 ganglia. TTX fully abolished the neural signals. *p < 0.001. All traces have been filtered with a low pass filter at 199 Hz and a high pass filter at 6 Hz.

Fig. 5. Voltage-sensitive dye recordings in primary cultures of human myenteric neurons. (A) Primary culture of the human myenteric plexus neurons after incubation with Di-8-ANEPPS. The neuron which responded to nicotine application is marked by an arrow. (B) Action potential discharge of the marked neuron in panel A after 200 ms application of nicotine (indicated by the bar below the trace; signals filtered with a low pass filter at 200 Hz and a high pass filter at 20 Hz). (C) The cell marked in panel A was immunoreactive for the neuronal marker PGP (marked by arrow). Despite the slightly distortion due to fixation and immunohistochemical processing the neuron marked in A can be clearly identified as PGP 9.5 positive. (D) The action potential discharge to nicotine (200 ms application marked by bar below the trace) was fully abolished by hexamethonium (Hexa) and recovered after wash out (signals filtered with a low pass filter at 130 Hz and a high pass filter at 19 Hz). (E) Statistical analysis of such experiments in 6 cultures and 11 neurons. Hexamethonium significantly reduced the action potential frequency. *p < 0.001 indicates significant smaller responses in Hexa compared to control and wash out.

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

We described a protocol that improves visibility of human myenteric ganglia in intact tissue by pre-staining with 4-Di-2-ASP. Of all the tested vital dyes 4-Di-2-ASP was the only dye that had no negative effect on the electrophysiological behaviour of guinea-pig myenteric neurons. These observations are in agreement with previous findings reporting that 4-Di-2-ASP has the ability to visualize the nerve plexi in guinea-pig and human intestine with no negative effects on electrophysiological behaviour of guinea-pig enteric neurons (Hanani, 1992; Hanani et al., 1993). Methylene blue caused significant reduction in the compound action potential and fast EPSP amplitude probably due to toxic effects indicated by nerve terminal swelling and loss of intracellular recorded postsynaptic potentials (Purves and Lichtman, 1987). FM1–43 has been used to track the recycling of synaptic vesicles at intact motor nerve terminals (Ryan and Smith, 1995; Ryan et al., 1996; Harata et al., 2001). It is a powerful tool to study synaptic transmission in the central, peripheral and also the enteric nervous system (Kavalali et al., 1999; Vanden Berghe et al., 2008). However, this dye was not useful to label and identify human myenteric ganglia and in addition it also reduced the amplitudes of compound action potentials and fast EPSPs. We have no final explanation for its negative effects on electrically induced signals in myenteric neurons except that the dye bleaching (Stanton et al., 2001) may have impacted on the viability of myenteric neurons. Pre-staining with 4-Di-2-ASP clearly revealed the ganglionic network in the human myenteric plexus. This allowed us to target individual ganglia in order to apply the voltage-sensitive dye directly into the ganglion. Intraganglionic injections have the advantage of low background staining as the dye labels ganglionic structures only (Schemann et al., 2002). With this protocol we were able to record compound action potentials in human myenteric plexus preparations after electrical stimulation of single nerve fiber strands. In some tissues the compound action potential was followed by a second peak, which could be partially blocked by the nicotinic acetylcholine-receptor blocker hexamethonium. This suggested that the second peak may represent fast EPSP triggered action potentials which have a cholinergic component. Galligan and Bertrand (1994) have revealed for the guinea-pig myenteric plexus that only in 25% of the investigated myenteric neurons the fast EPSPs were completely inhibited by blockers of nicotinic acetylcholine receptors. In the remaining neurons, the fast EPSPs were only partially blocked, which suggested that other transmitters were involved in the fast synaptic excitation as well. Consequently, most of the myenteric neurons received a mixed fast excitatory synaptic input, which is primarily mediated through nicotinic acetylcholine and purinergic P2X receptors (Galligan and Bertrand, 1994; LePard and Galligan, 1999; Galligan et al., 2000). Therefore, further experiments should be performed in human myenteric plexus preparations to investigate the pharmacology of fast EPSPs. Despite these promising results, the signal-to-noise ratio achievable in intact tissue was below the one that can be achieved in guinea-pig myenteric plexus preparations. In addition, the success rate to record signals from myenteric neurons in human tissue is much lower than in guinea-pig tissue. One explanation is that human myenteric ganglia contain only few neurons which in addition are clustered within the ganglion (Hanani et al., 2004). Although human myenteric ganglia were clearly dye labeled with our protocol it was not possible to identify the exact location of neuronal cell bodies. This is also one likely explanation for the finding that nicotine application did not always evoke spike discharge as we probably applied in some cases nicotine onto ganglionic areas that lack neuronal cell bodies. Another explanation could also be the basal lamina around the ganglia which as a barrier may prevent rapid drug delivery to the neurons (Gabella, 1972).

247

Labeling ganglia with the volume controlled injector showed an improved signal-to-noise ratio which enabled us to record action potentials after nicotine application. A possible explanation could be that this staining method is generally more gentle as only low pressure is used to inject the dye into the ganglion. In addition the dye has more time to incorporate into the cell membrane because the volume injection fills the ganglion with a much slower speed compared to the pressure controlled ejection. The significant improvement of the signal-to-noise ratio after addition of DPA can be explained by the properties of DPA. As described elsewhere DPA can be used as a Förster-energy-transfer (FRET) acceptor and in our case Di-8-ANEPPS as a donor fluorophore. In response to a change in membrane potential, DPA moves to the inner membrane and the larger distance of the molecules reduces the FRET. This results in a stronger fluorescence of Di-8-ANEPPS (Chanda et al., 2005a,b). Recordings in primary cultures of human myenteric neurons revealed signal-to-noise ratios that were comparable with those achievable in guinea-pig myenteric plexus preparations. It was possible to reliably detect action potential discharge in response to nicotine. We described protocols to use voltage-sensitive dye recordings in human myenteric neurons. Although there is still room for improvement we can conclude that the method can be used to record postsynaptic potentials in intact tissue as well as in primary culture of human myenteric neurons. This is an important step to study the neurobiology and neuropharmacology of human myenteric neurons. Some steps described in our staining protocol may also be relevant for labeling human tissues with other fluorescent dyes that monitor cell functions. Acknowledgements We thank Prof. Dr. Peter Fromherz for the gift of ANINNE-6-Plus. This work was supported by DFG (International Research Training Group 1373). References Bedlack Jr RS, Wie M, Loew LM. Localized membrane depolarizations and localized calcium influx during electric field-guided neurite growth. Neuron 1992;9:393–403. Betz WJ, Bewick GS. Optical analysis of synaptic vesicle recycling at the frog neuromuscular junction. Science 1992;255:200–3. Bradley J, Luo R, Otis TS, DiGregorio DA. Submillisecond optical reporting of membrane potential in situ using a neuronal tracer dye. J Neurosci 2009;29:9197–209. Breunig E, Michel K, Zeller F, Seidl S, von Weyhern CWH, Schemann M. Histamine excites neurones in the human submucous plexus through activation of H1, H2, H3 and H4 receptors. J Physiol 2007;583:731–42. Brookes SJH, Ewart WR, Wingate DL. Intracellular recordings from myenteric neurones in the human colon. J Physiol 1987;390:305–18. Brookes SJH, Ewart WR, Wingate DL. Intracellular recordings from cells in the myenteric plexus of the rat duodenum. Neuroscience 1988;24:297–307. Browning KN, Lees GM. Myenteric neurons of the rat descending colon: electrophysiological and correlated morphological properties. Neuroscience 1996;73:1029–47. Buhner S, Li Q, Vignali S, Barbara G, De Giorgio R, Stanghellini V, et al. Activation of human enteric neurons by supernatants of colonic biopsy specimens from patients with irritable bowel syndrome. Gastroenterology 2009;137:1425–34. Chanda B, Asamoah OK, Blunck R, Roux B, Bezanilla F. Gating charge displacement in voltage-gated ion channels involves limited transmembrane movement. Nature 2005a;436:852–6. Chanda B, Blunck R, Faria LC, Schweizer FE, Mody I, Bezanilla F. A hybrid approach to measuring electrical activity in genetically specified neurons. Nat Neurosci 2005b;8:1619–26. Cohen LB, Salzberg BM. Optical measurement of membrane potential. Rev Physiol Biochem Pharmacol 1978;83:35–88. Cornelissen W, de Laet A, Kroese AB, van Bogaert PP, Scheuermann DW, Timmermans JP. Excitatory synaptic inputs on myenteric Dogiel type II neurones of the pig ileum. J Comp Neurol 2001;432:137–54. Fluhler E, Burnham VG, Loew LM. Spectra, membrane binding, and potentiometric responses of new charge shift probes. Biochemistry 1985;24:5749–55.

248

S. Vignali et al. / Journal of Neuroscience Methods 192 (2010) 240–248

Fromherz P, Hübener G, Kuhn B, Hinner MJ. ANNINE-6plus, a voltage-sensitive dye with good solubility, strong membrane binding and high sensitivity. Eur Biophys J 2008;37:509–14. Fromherz P, Lambacher A. Spectra of voltage-sensitive fluorescence of styryl-dye in neuron membrane. Biochim Biophys Acta 1991;1068:149–56. Furukawa K, Taylor GS, Bywater RAR. An intracellular study of myenteric neurons in the mouse colon. J Neurophysiol 1986;55:1395–406. Gabella G. Fine structure of the myenteric plexus in the guinea-pig ileum. J Anat 1972;111:69–97. Galligan JJ, LePard KJ, Schneider DA, Zhou X. Multiple mechanisms of fast excitatory synaptic transmission in the enteric nervous system. J Auton Nerv Syst 2000;81:97–103. Galligan JJ, Bertrand PP. ATP mediates fast synaptic potentials in enteric neurons. J Neurosci 1994;14:7563–71. Hanani M, Fellig Y, Udassin R, Freund HR. Age-related changes in the morphology of the myenteric plexus of the human colon. Auton Neurosci 2004;113:71–8. Hanani M, Udassin R, Ariel I, Freund HR. A simple and rapid method for staining the enteric ganglia: application for Hirschsprung’s disease. J Pediatr Surg 1993;28:939–41. Hanani M. Visualization of enteric and gallbladder ganglia with a vital fluorescent dye. J Auton Nerv Syst 1992;38:77–84. Harata N, Ryan TA, Smith SJ, Buchanan J, Tsien RW. Visualizing recycling synaptic vesicles in hippocampal neurons by FM 1–43 photoconversion. Proc Natl Acad Sci USA 2001;98:12748–53. Hirst GDS, Holman ME, Spence I. Two types of neurones in the myenteric plexus of duodenum in the guinea-pig. J Physiol 1974;236:303–26. Kavalali ET, Klingauf J, Tsien RW. Properties of fast endocytosis at hippocampal synapses. Phil Trans R Soc Lond 1999;354:337–46. Kuhn B, Fromherz P. Anellated hemicyanine dyes in a neuron membrane: molecular Stark effect and optical voltage recording. J Phys Chem B 2003;107:7903–12. LePard KJ, Galligan JJ. Analysis of fast synaptic pathways in myenteric plexus of the guinea pig ileum. Am J Physiol 1999;276:G529–38. Loew LM. Potentiometric dyes: imaging electrical activity of cell membranes. Pure Appl Chem 1996;68:1405–9. Loew LM, Cohen LB, Dix J, Fluhler EN, Montana V, Salama G, et al. A naphthyl analog of the aminostyryl pyridinium class of potentiometric membrane dyes shows consistent sensitivity in a variety of tissue, cell, and model membrane preparations. J Membr Biol 1992;130:1–10. Magrassi L, Purves D, Lichtman JW. Fluorescent probes that stain living nerve terminals. J Neurosci 1987;7:1207–14. Maruyama T. Two types of spike generation of human Auerbach’s plexus cells in culture. Neurosci Lett 1981;25:143–8. Michel K, Zeller F, Langer R, Nekarda H, Kruger D, Dover TJ, et al. Serotonin excites neurons in the human submucous plexus via 5-HT3 receptors. Gastroenterology 2005;128:1317–26. Neunlist M, Peters S, Schemann M. Multisite optical recording of excitability in the enteric nervous system. Neurogastroenterol Motil 1999;11:393–402. Nishi S, North RA. Intracellular recordings from the myenteric plexus of the guineapig ileum. J Physiol 1973;231:471–91.

Obaid AL, Koyano T, Lindstrom J, Sakai T, Salzberg BM. Spatiotemporal patterns of activity in an intact mammalian network with single-cell resolution: optical studies of nicotinic activity in an enteric plexus. J Neurosci 1999;19:3073–93. Obaid AL, Loew LM, Wuskell JP, Salzberg BM. Novel naphthylstyryl-pyridinium potentiometric dyes offer advantages for neural network analysis. J Neurosci Methods 2004;134:179–90. Obaid AL, Zou DJ, Rohr S, Salzberg BM. Optical recording with single cell resolution from a simple mammalian nervous system: electrical activity in ganglia from the submucous plexus of the guinea-pig ileum. Biol Bull 1992;183: 344–6. Ohkawa H, Prosser CL. Functions of neurons in enteric plexuses of cat intestine. Am J Physiol 1972;222:1420–6. Purves D, Lichtman JW. Synaptic sites on reinnervated nerve cells visualized at two different times in living mice. J Neurosci 1987;7:1492–7. Ryan TA, Smith SJ, Reuter H. The timing of synaptic vesicle endocytosis. Proc Natl Acad Sci USA 1996;93:5567–71. Ryan TA, Smith SJ. Vesicle pool mobilization during action potential firing at hippocampal synapses. Neuron 1995;14:983–9. Schemann M, Michel K, Ceregrzyn M, Zeller F, Seidl S, Bischoff SC. Human mast cell mediator cocktail excites neurons in human and guinea-pig enteric nervous system. Neurogastroenterol Motil 2005;17:281–9. Schemann M, Michel K, Peters S, Bischoff SC, Neunlist M. Cutting-edge technology III. Imaging and the gastrointestinal tract: mapping the human enteric nervous system. Am J Physiol Gastrointest Liver Physiol 2002;282:G919–25. Sjulson L, Miesenböck G. Rational optimization and imaging in vivo of genetically encoded optical voltage reporter. J Neurosci 2008;28:5582–93. Stanton PK, Heinemann U, Müller W. FM1–43 imaging reveals cGMP-dependent long-term depression of presynaptic transmitter release. J Neurosci 2001;21:1–6. Tsau Y, Wenner P, O’Donovan MJ, Cohen LB, Loew LM, Wuskell JP. Dye screening and signal-to-noise ratio for retrogradely transported voltage-sensitive dyes. J Neurosci Methods 1996;70:121–9. Vanden Berghe P, Tack J, Coulie B, Andrioli A, Bellon E, Janssens J. Synaptic transmission induces transient Ca2+ concentration changes in cultured myenteric neurones. Neurogastroenterol Motil 2000;12:117–24. Vanden Berghe P, Tack J, Boesemans W. Highlighting synaptic communication in the enteric nervous system. Gastroenterology 2008;135:20–40. Waggoner AS. Dye indicators of membrane potential. Annu Rev Biophys Bioeng 1979;8:47–68. Wood JD. Electrical activity from single neurons in Auerbach’s plexus. Am J Physiol 1970;219:159–69. Wood JD. Electrical discharge of single enteric neurons in guinea pig small intestine. Am J Physiol 1973;225:1107–13. Wu JY, Lam YW, Falk CX, Cohen LB, Fang J, Loew L, et al. Voltage-sensitive dyes for monitoring multineuronal activity in the intact central nervous system. Histochem J 1998;30:169–87. Zochowski M, Wachowiak M, Falk CX, Cohen LB, Lam YW, Antic S, et al. Imaging membrane potential with voltage-sensitive dyes. Biol Bull 2000;198: 1–21.