Identification of unique release kinetics of serotonin from guinea‐pig and human enterochromaffin cells
2013; Wiley; Volume: 591; Issue: 23 Linguagem: Inglês
10.1113/jphysiol.2013.259796
ISSN1469-7793
AutoresRavinarayan Raghupathi, Michael D. Duffield, Leah Zelkas, Adrian Cuda Banda Meedeniya, Simon J. Brookes, Tiong Cheng Sia, David A. Wattchow, Nick J. Spencer, Damien J. Keating,
Tópico(s)Ion Channels and Receptors
ResumoEnterochromaffin (EC) cells are enteroendocrine cells that synthesise ∼95% of the body's total serotonin (5-HT). Although 5-HT release from EC cells plays a number of important physiological roles, primary EC cells have not been studied at the single cell level. This study provides the first functional characterisation of single primary guinea-pig and human EC cells. EC cells release 5-HT from large dense core vesicles in a calcium-dependent manner with kinetics surprisingly resembling release from synaptic vesicles. 3D modelling indicates that the quantity of 5-HT released per vesicle fusion event is physiologically relevant to GI tract function in terms of the concentrations needed to activate local 5-HT receptors. These findings represent significant advances in our understanding of EC cell function and will be of broad interest to researchers in endocrine cell biology, gastroenterology, neuroscience, exocytosis and glucose control. Abstract The major source of serotonin (5-HT) in the body is the enterochromaffin (EC) cells lining the intestinal mucosa of the gastrointestinal tract. Despite the fact that EC cells synthesise ∼95% of total body 5-HT, and that this 5-HT has important paracrine and endocrine roles, no studies have investigated the mechanisms of 5-HT release from single primary EC cells. We have developed a rapid primary culture of guinea-pig and human EC cells, allowing analysis of single EC cell function using electrophysiology, electrochemistry, Ca2+ imaging, immunocytochemistry and 3D modelling. Ca2+ enters EC cells upon stimulation and triggers quantal 5-HT release via L-type Ca2+ channels. Real time amperometric techniques reveal that EC cells release 5-HT at rest and this release increases upon stimulation. Surprisingly for an endocrine cell storing 5-HT in large dense core vesicles (LDCVs), EC cells release 70 times less 5-HT per fusion event than catecholamine released from similarly sized LDCVs in endocrine chromaffin cells, and the vesicle release kinetics instead resembles that observed in mammalian synapses. Furthermore, we measured EC cell density along the gastrointestinal tract to create three-dimensional (3D) simulations of 5-HT diffusion using the minimal number of variables required to understand the physiological relevance of single cell 5-HT release in the whole-tissue milieu. These models indicate that local 5-HT levels are likely to be maintained around the activation threshold for mucosal 5-HT receptors and that this is dependent upon stimulation and location within the gastrointestinal tract. This is the first study demonstrating single cell 5-HT release in primary EC cells. The mode of 5-HT release may represent a unique mode of exocytosis amongst endocrine cells and is functionally relevant to gastrointestinal sensory and motor function. Enterochromaffin (EC) cells are enteroendocrine cells providing ∼95% of total body 5-HT (Gershon & Tack, 2007). Enteroendocrine cells collectively represent the largest endocrine organ in our body and EC cells are the major enteroendocrine cell. Gut-derived 5-HT serves diverse endocrine roles in blood clotting, liver regeneration, bone formation (Karsenty & Gershon, 2011), embryo development (Cote et al. 2007), glucose homeostasis (Sumara et al. 2012) and the increased β-cell mass that prevents gestational diabetes (Kim et al. 2010). EC cell 5-HT also serves multiple paracrine roles in the gastrointestinal (GI) tract by modulating peristaltic and secretory reflexes as well as activating extrinsic sensory nerves (Gershon & Tack, 2007; Keating & Spencer, 2010; Spencer et al. 2011). EC cells respond to luminal stimuli including distension, acid and glucose to activate 5-HT3 receptors on vagal mucosal afferent fibres (Blackshaw & Grundy, 1993; Lee et al. 2011). 5-HT3 receptor antagonists are used clinically to reduce the nausea and vomiting caused by chemotherapy-induced surges in EC cell 5-HT release that activate mucosal vagal afferent fibres innervating the brainstem vomiting centres (Gershon & Tack, 2007). Altered EC cell 5-HT levels have been implicated in functional gastrointestinal disorders such as irritable bowel syndrome (IBS) and inflammatory bowel disease (IBD). 5-HT4 receptor agonists have been used to treat chronic constipation (Gershon & Tack, 2007) and inhibition of TPH1, the rate-limiting enzyme in gut-derived serotonin biosynthesis, has clinical benefits in patients with non-constipating IBS (Brown et al. 2011). EC cell 5-HT release is increased in inflammatory bowel disorders such as Crohn's disease (Kidd et al. 2009) or experimental models of colitis (Bertrand et al. 2010) and 5-HT availability is a negative effector of the severity of inflammation in rodent models of IBD (Bischoff et al. 2009; Ghia et al. 2009; Haub et al. 2010). Despite their importance, however, primary EC cells have yet to be studied at the single cell level. Previous investigations utilised cell lines derived from pancreatic carcinomas (Kim et al. 2001; Braun et al. 2007) or the human small intestinal carcinoid-derived neoplasia (Kidd et al. 2007), but how closely their function represents primary EC cell function is questionable. Results from studies of primary EC cell function using whole tissue or isolated crypts (Lomax et al. 1999; Nozawa et al. 2009; Keating & Spencer, 2010) are confounded by indirect effects from non-EC cells, such as neurons, epithelial cells and myocytes, in these preparations, or by gut wall contraction, which is a major stimulus of EC cell 5-HT release (Keating & Spencer, 2010). While enzyme-linked immunosorbent assay (ELISA) assays of 5-HT release from primary human EC cell cultures have illustrated responses to a number of physiological stimuli (Kidd et al. 2008), no study has yet demonstrated the cellular mechanisms by which 5-HT is released from primary single EC cells. The aims of this present study are to characterise the mechanisms underlying the release of 5-HT from primary EC cells, and to compare these mechanisms between the human and the commonly studied animal model of GI function, the guinea-pig, and to compare this with release from other endocrine cells. We additionally have developed sophisticated three-dimensional models of the diffusion of 5-HT released from EC cell populations throughout the GI tract to understand what the implications of the EC cell release mechanisms would be for activation of 5-HT receptors located on mucosal nerve endings. Adult guinea-pigs were killed humanely by stunning with a blow to the head followed by severing of the carotid arteries, as approved by the Flinders University Animal Welfare Committee. Four to six centimetres of distal colon was removed and EC cell isolation and purification was conducted using an approach modified from those previously published (Schafermeyer et al. 2004; Kidd et al. 2006). The mucosal layer of the colon was scraped off, minced and washed once in Buffer A (in mm: 140 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 10 Hepes, 5 d-glucose, pH 7.4). The tissue was then digested in trypsin–EDTA (0.05%) with collagenase A (1 mg ml−1, Roche Diagnostics GmbH, Mannheim, Germany) at 37°C for 30 min with continuous agitation. An equal volume of growth medium (DMEM containing 10% FBS, 1%l-glutamine and 1% penicillin–streptomycin) stopped this reaction. The suspension was filtered through a 40 μm steel mesh filter and centrifuged (Sigma, 6K15, USA) at 1000 ×g. The resultant pellet was resuspended in 1 ml of growth medium and layered onto a Percoll density gradient formed according to manufacturer's instructions. After centrifuging at 1100 ×g for 8 min with slow braking, EC cells were harvested at a density of approximately 1.07 g l−1, washed once with growth medium and plated onto 6 cm2 pre-treated cell culture dishes (Iwaki, from VWR Intl Pty Ltd, Murrarrie, QLD, Australia). EC cells were maintained in growth medium for 2–4 days. All chemicals were from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated. Human colon tissue samples were obtained with prior informed consent from patients undergoing elective colectomy at the Flinders Medical Centre under the approval of the Southern Adelaide Clinical Human Research Ethics Committee. Cultures were obtained from four female patients aged 54–88 years old. Approximately 2 cm2 of mucosa was dissected from areas of colon well removed from the site of tumours. As the surgery is performed to remove all neoplasia from the patient, we are confident that obtaining tissue so distal from the neoplasia represents normal tissue. This tissue was processed as described above and EC cells plated and studied in the first 24 h post culture. EC cell viability was measured by incubating cells with Trypan Blue (0.2% final concentration) for 5–10 min at 37°C followed by a cell count on a haemocytometer. Cells were considered viable if they completely excluded the dye. All results are from at least three separate cell cultures. EC cells were grown for 24 h on glass coverslips previously coated with 10 μg ml−1 each of poly-d-lysine and laminin (Sigma-Aldrich) in growth medium. Cells were fixed for 18–20 h in Zamboni's fixative at 4°C followed by serial 5 min washes as follows: 4 × 80% EtOH, 2 × 100% EtOH, 3 × DMSO, 4 × PBS. Fixed cells were incubated for 30 min in 10% normal donkey serum diluted in antibody diluent (290 mm NaCl, 7.5 mm Na2HPO4, 2.6 mm Na2HPO4.2H2O, 0.1% NaN3 in distilled water, pH 7.1), followed by incubation for 24 h in a humid chamber with goat monoclonal antibody against 5-HT (Jackson Immunoresearch, West Grove, PA, USA, 1:400) or sheep monoclonal antibody against TPH-1 (Millipore, Billerica, MA, USA, Cat. No. AB1541, 1:200). After three washes with PBS, the cells were incubated with donkey anti-goat IgG or donkey anti-sheep IgG tagged with Cy3 (Jackson Immunoresearch, 1:200 and DAPI (Sigma-Aldrich, 1:500) for 2 h in a humid chamber. Cells were then washed 3 times with PBS and the coverslips mounted onto glass slides in buffered glycerol. Fluorescence was visualised on a Leica TCS SP5 Spectral confocal microscope. The purity of the EC cell culture was calculated by assessing the proportion of DAPI-positive cells which were 5-HT-positive. Using this approach we calculated that the cells in these cultures were 98% 5-HT-positive. Release of 5-HT from single EC cells was measured using carbon fibre amperometry (Keating et al. 2008). A carbon fibre electrode (ProCFE, Dagan Corporation, Minneapolis, MN, USA) was placed directly above an EC cell and +400 mV applied to the electrode under voltage clamp conditions as this is the oxidation peak for 5-HT. Current due to 5-HT oxidation was recorded using an EPC-9 amplifier and Pulse software (HEKA Electronic, Germany) with the current sampled at 100 kHz and low-pass filtered at 3 kHz. For quantitative analysis files were converted to Axon Binary Files (ABF Utility, version 2.1, Synaptosoft, USA) and secretory spikes analysed (Mini Analysis, version 6.0.1, Synaptosoft, USA). The standard bath solution was Buffer A. High K+-containing solution was the same as Buffer A except that 70 mm K+ replaced an equimolar amount of NaCl. To analyse the effect of Ca2+ entry on 5-HT release, cells were exposed to 70 mm K+ in the absence of external Ca2+. All solutions were applied to cells using a gravity perfusion system at 34–37°C. For recordings from EC cells, amperometric spikes were selected for analysis of event frequency if spike amplitude exceeded 10 pA. For spike frequency calculations we counted spikes occurring within 60 s of the start of stimulation. For kinetic analysis of spikes only those events that were not overlapping were included. Rise time of each spike was calculated from the 50–90% rising phase. Chromaffin cells and chromaffin cell amperometry data were obtained from adult mice as previously described (Zanin et al. 2011). All spikes that met our selection criteria were included in calculating the median values of each spike parameter for each recording. The averages of these median values were then used to compare each parameter between cell populations (Colliver et al. 2001). This provides a parametric data set which was subsequently tested for statistical differences using an unpaired Student's t test. For comparisons of spike kinetics in unstimulated and stimulated cells, we pooled all the spike data as we could not obtain a meaningful median value in many unstimulated cell recordings due to the low number of spikes. These data sets were compared using a Mann–Whitney test for non-parametric data sets. P < 0.05 was taken as the lowest level of statistical significance. All data presented are shown as means ± SEM and all data are from at least three different cell cultures. Changes in EC cell Ca2+ level were measured at 34–37°C using previously described methods (Zanin et al. 2011). Whole-cell perforated patch clamp was performed using a EPC-10 patch clamp amplifier and PatchMaster software (HEKA Electronik, Lambrecht/Pfalz, Germany). Patch pipettes were pulled from borosilicate glass and fire polished, with a resistance of 3–5 MΩ. Patch clamping was performed in the whole-cell configuration for measurement of Ca2+ currents, using an internal solution containing (in mm): 135 caesium glutamate, 9 NaCl, 10 Hepes, 0.5 TEA-Cl, pH 7.2. External solution contained (in mm): 150 NaCl, 2.8 KCl, 10 Hepes, 2 MgCl2, 10 CaCl2, 10 glucose, (pH 7.4). Calcium currents were elicited in the voltage-clamp mode using a voltage step protocol, in which voltage was stepped from a holding potential of −80 mV to voltages between −80 and 50 mV (10 mV increments) for 100 ms. Series resistance was compensated at least 70%. EC cells were treated with Fluo-4 (4 μm) in serum-free DMEM for 30 min at 37°C. Cells were washed twice in Buffer A, which also served as the standard bath buffer. Cells were stimulated for 60 s with the same solution as Buffer A except that 70 mm K+ replaced an equimolar amount of NaCl. An electron multiplied (EM-CCD) camera (Cascade II 512, Roper Scientific, Tucson, AZ, USA) was used to record dynamic changes in fluorescence activity from EC cells. Data were acquired at 2 Hz, All imaging data were acquired and analysed using Imaging Workbench (Version 6.0; Indec Biosystems; Santa Clara; CA, USA). All experiments were carried out at 34–37°C. Tissues from pre-selected regions of guinea-pig gastrointestinal tract were dissected free and placed in Krebs solution (in mm: 117 NaCl, 5 KCl; 1.2MgSO4, 25 NaHCO3, 1.2 NaH2PO4, 2.5 CaCl2, 10 glucose, bubbled with 95% O2–5% CO2, pH 7.4) containing 1 μm nicardipine. Samples were taken from the mid-oesophagus (oesophagus) and the lower oesophageal sphincter region (cardia), the fundus, corpus and antrum of the stomach, the antral–pyloric border, the pyloric sphincter region and the duodenal–pyloric border, the mid-duodenum, the proximal, mid- and distal ileum, the caecum, the proximal and distal colon and the rectum. The tissues were pinned on Sylgard blocks with maximal stretching and taken through the fixation, clearing and impregnation processes while pinned to these blocks. Tissue was fixed overnight in modified Zamboni's fixative (2% formaldehyde and 15% picric acid in 0.1 m phosphate buffer, pH 7.0) and cleared in dimethyl sulphoxide (DMSO) for 30 min with the solution changed every 10 min. Samples were then dehydrated through an alcohol series comprising 50%, 70% and 95% ethanol for 15 min in each solution and in 100% ethanol for 1 h. The tissues were then impregnated with polyethylene glycol (Sigma-Aldrich, USA), by immersion in PEG 200 followed by PEG 400, for 1 h in each solution at room temperature. The tissues were then placed in a solution of PEG 1000 for 60 min, in vacuo, and with a desiccant (silica gel) at 45–48°C. The tissues were unpinned from the Sylgard blocks, trimmed, oriented in PEG 1400 in a cryo-mould and then embedded in PEG 1400 by hardening at −20°C for 5 min, in vacuo. The blocks were sectioned at 60 μm thickness on a rotary microtome at room temperature. The sections were placed on agarose (Sigma-Aldrich, USA) sheets 3% in phosphate-buffered saline (PBS) and the agarose sheets were placed on chrome alum slides with the sections contacting the slides. The slides, with the agarose sheets, were then dehydrated, in vacuo, with P2O5 for 20 min, placed in PBS on a rocker until the agarose sheets floated free, leaving the sections adhering to the slides. The slides were kept overnight, at room temperature in fresh PBS to further enhance the removal of the PEG from the sections. Immunoreactivity for 5-HT was revealed by indirect immunofluorescence with sections incubated overnight in goat anti-5-HT (108072, 1:1000, Incstar) primary antibody in a humidified chamber. The primary antisera were visualised using CY3 conjugated donkey anti-goat IgG (Jackson, Pennsylvania, USA, code: 26035, 1:300). The tissues were incubated overnight with the secondary antibodies. The sections were then mounted in bicarbonate-buffered glycerol and analysed. EC cells from each region were clearly identified by their immunoreactivity for 5-HT and counted in three guinea-pigs. A standard sample area of 200 μm × 200 μm was used for cell counts on a Vanox Olympus microscope fitted with epifluorescence. Images were processed using Image J (Bethesda, MD, USA). Sample areas were taken from all levels in the mucosa, oriented to exclude muscularis mucosa or regions that lacked tissue. Ten random samples were taken from each area, from each animal, and cell density per unit volume of tissue calculated in each region. Using a section thickness approximately twelve times the diameter of the EC cells and accurately defining cell morphology minimised sampling errors (Coggeshall & Lekan, 1996). An explicit finite-differences model (described in more detail below), incorporating release of serotonin from EC cells, diffusion of serotonin, and uptake of serotonin by the SERT was used to simulate 5-HT concentrations within the mucosa of different GI regions. Single-cell models Single-cell release models were of dimensions 50 μm per side, discretised in 1 μm steps (see equations below), with the entire volume representing the mucosa layer, and with a single release site at the centre of the volume. This system was simulated for a sufficient amount of time for numerous release events to occur. Tissue models Tissue models consisted of a cuboid region of dimensions between 400–1000 μm per side, discretised in 10 μm steps (see equations below: Δx, Δy, Δz= 10 μm). For these models a mucosal cell was assumed to be of dimensions 10 μm × 10 μm × 10 μm, consistent with the approximate measured dimensions of single EC cells. A cuboid layer representing mucosal tissue, of an appropriate thickness for the GI tract region being modelled, was placed within this volume, with a submucosal layer ‘below’ and luminal layer ‘above’. 5-HT release points were randomly distributed throughout the mucosal layer at a density consistent with the experimentally determined EC cell density. 5-HT entry and movement into the submucosal and luminal regions was purely diffusional, with no release sites placed within these regions. To determine steady-state 5-HT concentrations in the mucosa of different GI regions under stimulated or non-stimulated conditions simulations with an initial 5-HT concentration of 0 m at all points were allowed to run until the average mucosal concentration of 5-HT had stabilised, using release frequencies obtained from our own experiments in EC cells. Contraction models To model the effects of GI contraction it was assumed that contraction of the tissue would result in a change in EC cell state from ‘unstimulated’ to ‘stimulated’, with a resultant change in 5-HT release rate, and the reverse during relaxation (Spencer et al. 2011). Therefore modelling of this process consisted of changing the 5-HT release parameters at a rate consistent with the contraction–relaxation rate described experimentally (Table 1). These simulations were initially run in the ‘relaxed’ (non-stimulated release) state until steady state had been achieved, at which point a series of contraction and relaxation events were simulated. Model parameters Parameters utilised for the modelling are shown in Table 1, and are all based on experimentally determined parameters from the literature or from the current paper. To account for the diffusion properties of the complex media, which is the extracellular space of the GI layers, we have included both tortuosity and volume fraction effects in our model (Sykova & Nicholson, 2008). The effective diffusion coefficient for 5-HT in tissue was calculated based on the 5-HT diffusion in CNS tissue, which found a tortuosity constant of 1.65 for 5-HT (representing an effective diffusion coefficient in tissue approximately 3× smaller than that in aqueous solution) (Rice & Nicholson, 1986). This effective tissue diffusion constant was used for the mucosal and submucosal layers of the model, whilst diffusion within the luminal layer utilised a diffusion coefficient consistent with that reported for 5-HT in aqueous solution. Within the cellular layers of the mucosa and submucosa a volume fraction of 0.17, again based on properties of 5-HT diffusion in the CNS (Rice & Nicholson, 1986), was used for calculation of the 5-HT concentration within the extracellular compartment of the tissue. All concentration values reported within the current paper refer to concentrations within this extracellular compartment, which are the relevant concentrations with regard to 5-HT uptake and receptor activation. Models were written in the Python programming language (http://www.python.org), utilising the NumPy/SciPy libraries (Jones et al. 2001). In all models the spatial discretisation parameters were determined as described previously, and then the time discretisation, Δt, was determined to ensure that this condition was fulfilled at all times. All data are presented as means ± SEM. Error bars represent SEM. Analyses of significant differences between means were performed using two-tailed Student's t tests or the Mann–Whitney test when comparing two groups of parametric or non-parametric data sets. n indicates the number of independent cultures or animals used. In all cases, significant differences are indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001. We purified EC cells from guinea-pig colon and maintained them in primary culture (Fig. 1A). Immunolabelling confirmed these cells as 5-HT-containing EC cells (Fig. 1B) with a purity of >98% (Fig. 1C). 5-HT expression in the cytoplasm is punctate (Fig. 1D) and these cells also contain Tph1 (Fig. 1E), further confirming them as EC cells. Cell viability was >98% for the first 48 h in culture (Fig. 1F). Isolation and purification of primary EC cells Low magnification of EC cells in culture labelled with a 5-HT antibody and viewed under brightfield (A) and immunofluorescence (B). Red arrows illustrate 5-HT staining in these cells. C, cells staining positively for both 5-HT and the nuclear marker DAPI, demonstrating >98% pure EC cell culture (n= 3 cultures, ***P < 0.001). D, higher magnification of these cells using confocal microscopy observed in brightfield (top left), the nuclear stain DAPI (blue), punctuate cytoplasmic 5-HT immunoreactivity (red) and the merged image. Scale = 8 μm. E, these cells also contain the EC cell marker Tph1. Scale = 5 μm. F, viability assay demonstrates cells are healthy in culture (n= 4 cell cultures). Whole-cell patch clamp demonstrated voltage-gated Ca2+ channels in EC cells (Fig. 2A) which peak at 20–30 mV (Fig. 2B). Ca2+ imaging in cells loaded with Fluo-4 demonstrated that membrane depolarisation with either 70 mm K+ or acetylcholine (10 μm) induced Ca2+ entry (Fig. 2C). Thus, EC cell stimulation causes the entry of Ca2+, the main physiological trigger of secretion. We then used carbon fibre amperometry to measure 5-HT release from single EC cells. Carbon fibre probes were held at the oxidation peak for 5-HT, +400 mV (Fig. 2E, inset), and placed adjacent to a single EC cell. Individual amperometric peaks indicative of single vesicle release of 5-HT were observed under basal and stimulatory (70 mm K+) conditions (Fig. 2D). Similar results were observed with acetylcholine (10 μm, Fig. 3D). To confirm our oxidation currents represented 5-HT release, we incubated EC cells for 24 h with an inhibitor of Tph, LP533401 (1 μm, Dalton USA), which results in almost complete suppression of 5-HT synthesis (Yadav et al. 2010). These cells (Fig. 2E) showed significantly reduced release (Fig. 2F, P < 0.001), confirming that these oxidation peaks represent single 5-HT release events from primary EC cells. Stimulation of primary EC cells causes Ca2+ entry and 5-HT release A, Ca2+ currents elicited from −80 mV holding potential, stepped for 100 ms to −20 mV to +30 mV in 10 mV increments. B, current density–voltage relationship of these Ca2+ currents (n= 6). C, example trace of Ca2+ entry in a single cell stimulated with 70 mm K+ (dashed line, Scale = 10 s and 5 fluorescence points). Inset, average EC cell fluorescence change in response to 70 mm K+ or acetylcholine (ACh, 10 μm, n= 13 cells for both groups). D, amperometry measures 5-HT release from single EC cells. 70 mm K+ solution (dashed line) triggers 5-HT release from single vesicles as indicated by individual current spikes. E, inset, oxidation current in 5-HT (10 μm) when voltage is ramped from 0 to 0.8 V using cyclic voltammetry demonstrates +400 mV as the peak oxidation current for 5-HT. Scale bar = 100 pA. Cells treated for 24 h with the Tph inhibitor LP533401 (1 μm) have almost no 5-HT release (E) and this decrease is significant (F; ***P < 0.001, n= 9–12 cells). Scale bars in C and D represent 100 pA and 20 s. EC cell 5-HT release is dependent on external Ca2+ entry via voltage-gated Ca2+ channels K+ at 70 mm (dashed line) triggers 5-HT release in the presence of external Ca2+ (A) and in the same cell when external Ca2+ is removed (B). C, a similar effect is seen in cells exposed to the L-type Ca2+ channel antagonist nicardipine (2 μm). Scale bars in A represent 20 pA and 10 s and apply to A, B and C. D, quantification of spike frequency demonstrates that 70 mm K+ (n= 18 paired recordings, *P < 0.05) and acetylcholine (ACh, n= 16 paired recordings, **P < 0.01) increase the number of 5-HT release events and high K+-induced release is reduced in the absence of external Ca2+ (n= 6 paired recordings, **P < 0.01) or presence of nicardipine (n= 8 paired recordings, **P < 0.01). A reduced number of 5-HT release events were observed when EC cells were stimulated in the absence of external Ca2+ (Fig. 3A and B). Blocking L-type Ca2+ channels with nicardipine (2 μm) also reduced release event frequency (Fig. 3C). Thus, 5-HT release from EC cells is largely dependent on Ca2+ entry, with one major avenue being through L-type Ca2+ channels (Fig. 3D). Individual current peaks in amperometric recordings represent the release of oxidisable substances from a single vesicle (Fig. 4A). Vesicle size is a major determinant of vesicle release kinetics (Albillos et al. 1997; Zhang & Jackson, 2010) so we compared single EC cell events to those in the more commonly studied endocrine cell, the adrenal chromaffin cell, which releases catecholamines from similarly sized large dense core vesicles (LDCVs; Nilsson et al. 1987; Pothos et al. 2002). This comparison demonstrates that the amount released per exocytosis event was much smaller in EC cells (Fig. 4B). Comparisons of frequency distribution of spike width at half-maximum height (half-width, Fig. 4C), rise time (Fig. 4D) and decay time (Fig. 4E) further confirmed the lack of overlap between these groups. Unexpectedly, the amount of 5-HT released per fusion event was similar to that seen for release of dopamine from much smaller synaptic vesicles (Staal et al. 2004; Table 2). EC cell 5-HT release occurs with synaptic-like release kinetics A, individual release events from EC cells are more rapid than those in adrenal chromaffin cells. An example spike from each cell type is overlaid for comparison B, a comparison of spike area distribution illustrates that the amount of EC cell 5-HT released per fusion event represents a separate population to release events from chromaffin cells. Faster release kinetics in EC cells are further illustrated by comparing the frequency distribution of spike half-width (C), rise time (D) or decay time (E) in both cell types. n= 2113 spikes from 24 EC cell recordings and 781 spikes from 16 chromaffin cell recordings. Red bars = EC cell data, black bars = chromaffin cell data. We also isolated and cultured human colonic EC cells and measured single 5-HT release events (Fig. 5A). Spike frequency in unstimulated and stimulated cells (Fig. 5B) was similar to that in guinea-pig. The relative number of release events over time was identical in human and guinea-pig EC cells, and was significantly reduced in the absence of external Ca2+ or when L-type Ca2+ channels were blocked (Fig. 5C). Single exocytosis events also occurred in human EC cells (Fig. 5D) and the distribution of spike area was strikingly similar to guinea-pig EC cells (Fig. 5E). To test whether partial release via kiss-and-run fusion might explain the synaptic-like kinetics of 5-HT release in EC cells we inhibited Tph activity for 24 h with LPS533401 (1 μm) to reduce the availability of 5-HT loading into vesicles. If EC cell vesicles release 5-HT via full fusion, this treatment would result in reduced spike charge (Colliver et al. 2000; Pothos et al. 2002; Gong et al. 2003; Sombers et al. 2004). The mean charge of release events was unaltered by this treatment (Fig. 5F), indicating that rapid kiss-and-run, rather than full fusion, may be occurring in EC cells
Referência(s)