Mast cell activation initiated by antigen-mediated crosslinking of IgE receptors results in stimulated exocytosis of secretory lysosomes in the process known as degranulation. Much has been learned about the molecular mechanisms important for this process, including the crucial role of Ca2+ mobilization, but spatio-temporal relationships between stimulated Ca2+ mobilization and granule exocytosis are incompletely understood. Here we use a novel imaging-based method that uses fluorescein isothiocyanate (FITC)–dextran as a reporter for granule exocytosis in RBL mast cells and takes advantage of the pH sensitivity of FITC. We demonstrate the selectivity of FITC–dextran, accumulated by fluid-phase uptake, as a marker for secretory lysosomes, and we characterize its capacity to delineate different exocytotic events, including full fusion, kiss-and-run transient fusion and compound exocytosis. Using this method, we find strong dependence of degranulation kinetics on the duration of cell to substrate attachment. We combine imaging of degranulation and Ca2+ dynamics to demonstrate a spatial relationship between the sites of Ca2+ wave initiation in extended cell protrusions and exocytosis under conditions of limited antigen stimulation. In addition, we find that the spatially proximal Ca2+ signaling and secretory events correlate with participation of TRPC1 channels in Ca2+ mobilization.
In response to antigen-mediated crosslinking of cell-surface IgE receptors (IgE–FcεRI), mast cells undergo a series of intracellular signaling events that result in fusion of secretory granules with the plasma membrane to release a number of inflammatory mediators including histamine, proteoglycans, proteases and lysosomal hydrolases, which trigger allergic and inflammatory reactions that can have systemic effects (Blank and Rivera, 2004; Gilfillan and Tkaczyk, 2006). In this process, stimulated tyrosine phosphorylation of FcεRI leads to inositol 1,4,5-trisphosphate (IP3) production by phospholipase Cγ (PLCγ), which causes Ca2+ depletion from endoplasmic reticulum stores to induce store-operated Ca2+ entry (SOCE) (Di Capite and Parekh, 2009). Activation of SOCE elicits oscillatory cytosolic Ca2+ elevations that, together with activated protein kinase C, trigger granule exocytosis (Kim et al., 1997; Ma and Beaven, 2009).
Although the early biochemical events of mast cell activation have been studied extensively (Rivera and Gilfillan, 2006), the mechanisms by which degranulation occurs and is regulated in these cells are not fully understood, calling for new experimental approaches. In particular, delineating the relationships between degranulation and Ca2+ mobilization requires high spatial and temporal resolution of both cytoplasmic Ca2+ changes and individual exocytotic events. We recently showed that antigen-mediated crosslinking of IgE–FcεRI initiates Ca2+ responses in mast cells in the form of a fast Ca2+ wave that usually begins in the tip of an extended protrusion, then moves down the length of the cell over a time period of several seconds (Cohen et al., 2009). At low doses of antigen, initiation of this wave depends on Ca2+ influx by the canonical transient receptor potential channel TRPC1. The initiating waves are typically followed by periodic Ca2+ oscillations that depend on Ca2+ influx through SOCE (Ma and Beaven, 2009; Vig et al., 2008). These oscillations encode temporal information relevant to degranulation events (Kim et al., 1997), but they do not manifest the spatial localization displayed by wave propagation (Cohen et al., 2009). Very little is known about the physiological function of the extended protrusions in mast cells where Ca2+ waves initiate. We hypothesize that, in vivo, mast cells protrude through epithelial cell barriers to provide distal sensors for relevant ligands. In this context, the capacity for granule exocytosis from these protrusions might be physiologically significant.
A number of methods have been developed to measure individual granule exocytotic events in mast cells (Hohman and Dreskin, 2001). Early studies used Acridine Orange, a fluorescent weak base that accumulates in acidified vesicles (Kawasaki et al., 1991; Williams and Webb, 2000). However, a recent study demonstrated that photosensitization of this dye during imaging limits its utility (Jaiswal et al., 2007). Other strategies have used granule-localized serotonin fluorescence imaged using multiphoton microscopy, together with complementary membrane and lysosome-labeling probes (Williams et al., 1999). Although these methods allow detection of individual stimulated granule exocytotic events in mast cells, they are technically demanding and therefore limited in their usage.
In the present study we use FITC–dextran, which accumulates in secretory lysosomes by fluid-phase pinocytosis (Dragonetti et al., 2000), to monitor exocytosis of individual granules in RBL mast cells. The pH sensitivity of FITC (Ohkuma and Poole, 1978) reveals fusion events between granules and the plasma membrane that result in exposure to the extracellular medium. Such exposure relieves the low-pH quenching of FITC fluorescence in the lumen of the secretory lysosomes that constitute exocytic granules in these cells (Xu et al., 1998). By this means, we can simultaneously monitor granule exocytosis and cytosolic Ca2+ elevation in individual cells. We find that, at limiting concentrations of antigen, exocytosis occurs preferentially along extended cell protrusions where Ca2+ waves initiate.
Labeling of secretory lysosomes with fluorescent dextrans
As previously described (Dragonetti et al., 2000), fluorescent dextran conjugates accumulate in secretory lysosomes of RBL mast cells after fluid-phase pinocytosis. Fig. 1A shows RBL cells transfected with CD63–GFP and incubated with TxRed–dextran together with 5-hydroxytryptamine (5-HT) for 24 hours. These cells exhibited a high degree of colocalization of TxRed–dextran in the lumen of large intracellular vesicles with CD63–GFP at their periphery (Fig. 1A, inset). CD63, also known as LAMP3, is a tetraspanin membrane protein containing a lysosome-targeting domain (Levy and Shoham, 2005), and it is commonly used as a secretory granule marker in mast cells (Blott and Griffiths, 2002; Amano et al., 2001). CD63 has been used previously to monitor secretory granule exocytosis in RBL cells by total internal reflection fluorescence microscopy (Wu et al., 2007).
As evident in Fig. 1A, we frequently observed a high density of granules in tips of extended protrusions. This accumulation of granules was also seen often with EGFP–VAMP7, which is a v-SNARE for secretory lysosomes (Puri et al., 2003) that was also highly colocalized with TxRed–dextran incorporated during overnight loading (Fig. 1B). Cross-correlation analysis revealed a moderately high value for Pearson's coefficient (>0.7 for n = 8 cells analyzed) for this label pair. Values for Pearson's coefficient greater than 0.2 are usually considered to indicate significant colocalization (Pyenta et al., 2001). Co-labeling with FITC–dextran and TxRed–dextran further showed almost complete co-incidence in intracellular vesicles enlarged by 5-HT loading (supplementary material Fig. S1), with a Pearson's coefficient of 0.85 (n = 12). Thus, similar localization patterns for TxRed–dextran with both CD63–GFP and EGFP–VAMP7 provide strong evidence that FITC-tagged dextrans selectively label secretory lysosomes in these cells.
Granule fusion events
To image single exocytosis events in RBL cells that are stimulated by antigen, IgE-sensitized cells were loaded with FITC–dextran and 5-HT, and then antigen was delivered locally by micropipette as described in the Materials and Methods. At a high concentration of antigen in the pipette (1.7 µg/ml), degranulation events, detected as local bursts of FITC fluorescence, were observed as soon as 20–40 seconds after antigen addition, and they continued for several minutes (supplementary material Movie 1). The representative images in Fig. 2 illustrate three variants of these events. Based on previous results (Williams and Webb, 2000), we interpret the rapid increase in FITC fluorescence to be pH neutralization occurring in the granule lumen as it fuses with the plasma membrane and is exposed to the extracellular medium, and the rapid decrease is due to diffusion or dilution in the extracellular medium. No bursts of FITC–dextran release in response to antigen were observed in cells that were not sensitized with specific IgE (data not shown).
In addition to rapidly releasing individual events (supplementary material Movie 1; Fig. 2A,B), which were the most frequent manifestation of exocytosis, a smaller percentage of the events exhibited a sharp initial burst of FITC fluorescence that more slowly returned to baseline levels. As exemplified in Fig. 2C,D, a rapid rise and partial decrease in fluorescence was followed by a second increase and then a gradual decrease that lasted for >100 seconds. The sustained decreases are probably due to reclosing of the partially released granule in a ‘kiss-and-run’ mechanism, followed by re-acidification of the lumen (Williams and Webb, 2000). We also observed sequential or compound exocytosis (Pickett et al., 2005), resulting from granule–granule fusion that followed an initial granule–plasma-membrane fusion event (Fig. 2E,F), as previously reported for mast cells (Alvarez de Toledo and Fernandez, 1990; Guo et al., 1998).
Simultaneous detection of FITC–dextran and TxRed–dextran release from single granules
To distinguish whether a transient burst of FITC–dextran fluorescence was due to exocytotic release or transient, intracellular neutralization of acidic granules, we loaded the granules with both FITC–dextran and TxRed–dextran, then stimulated these cells as above. As shown by the images in Fig. 3, the transient burst of FITC–dextran fluorescence correlated temporally and spatially with the pH-insensitive release of TxRed–dextran from the granule. Release of both fluorophores was followed by rapid diffusion and dilution in the extracellular medium. This representative example of events detected as simultaneous changes in FITC–dextran and TxRed–dextran fluorescence provides evidence that FITC–dextran transient bursts are due to exocytosis of the granule contents at the time of FITC dequenching.
Combined amperometry and FITC–dextran imaging in individual RBL cells
To evaluate further whether FITC–dextran dequenching is a reliable marker for granule exocytosis in RBL cells, we simultaneously monitored FITC–dextran fluorescence and 5-HT release that was detected amperometrically by a locally positioned carbon fiber electrode. As previously demonstrated for mast cells, amperometric detection of 5-HT release provides high temporal resolution of exocytotic events (Kim et al., 1997; Jaffe et al., 2001). As shown in Fig. 4, we found a strong temporal correlation between bursts of FITC fluorescence and amperometric spikes that represent degranulation events. Multiple, spatially separated exocytotic events, detected as FITC fluorescence bursts, were often seen to occur during the same amperometric spike, suggesting that the initiation of these separate events is temporally correlated, possibly to the peak of a Ca2+ oscillation (Kim et al., 1997).
Cell attachment influences RBL mast cell degranulation kinetics
The pH sensitivity of FITC–dextran release permits measurements of exocytosis from suspended mast cells by monitoring with steady-state fluorimetry. As shown in Fig. 5, top panel, addition of antigen to IgE-sensitized RBL cells loaded with FITC–dextran caused a time-dependent increase in fluorescence that depended on extracellular Ca2+. Subsequent addition of the Ca2+ ionophore A23187 caused a further increase in FITC fluorescence such that the combined response for antigen and ionophore approached ∼60% of the total FITC fluorescence detected following cell lysis with Triton X-100, indicating that a large percentage of FITC–dextran is released from the secretory granules under these conditions. The response to antigen was relatively slow under these conditions, with a half-time of ∼10 minutes in multiple experiments (data not shown).
The degranulation response to antigen for cells that remained attached after overnight loading with FITC–dextran occurred somewhat more rapidly, with a half-time of ∼5–6 minutes (data not shown). This degranulation response for attached cells was only about half that measured when 1.7 ng/ml antigen was used to stimulate these adherent cells, suggesting that antigen binding and crosslinking are not rate limiting under these conditions (data not shown). These differences in antigen-stimulated degranulation kinetics for adherent compared with suspended cells prompted us to investigate their dependence on the duration of cell adhesion. As shown in Fig. 5, bottom panel, there was a strong dependence of both the rate and number of antigen-stimulated exocytotic events on the time that cells had been attached before stimulation, with substantial increases over 20 to 90 minutes of attachment. During 300 seconds of stimulation, cells pre-attached for 90 minutes exhibited similar or even larger numbers of degranulation events compared with cells attached overnight, but this response was somewhat slower, with a more substantial lag time before initiation. Shorter attachment periods of 20 and 60 minutes resulted in substantially longer lag times. These results demonstrate that the kinetics of the degranulation response to antigen strongly depends on the duration of cell attachment.
Simultaneous imaging of Ca2+ and degranulation responses
Exocytosis and Ca2+ mobilization are strongly coupled, and their temporal and spatial relationships have been extensively characterized in both excitable and non-excitable cells (Kasai, 1999; Blank et al., 2001). Our previous studies demonstrated that Ca2+ responses to antigen in RBL mast cells usually initiate as waves that begin at the tips of extended cell protrusions (Cohen et al., 2009), and these results prompted us to investigate whether degranulation is influenced by Ca2+ waves. To monitor Ca2+ mobilization and degranulation simultaneously in the same cells, we used Fura Red as the Ca2+ indicator in cells that were pre-loaded with FITC–dextran. Fig. 6A shows a representative cell with an extended protrusion in which the distribution of detectable FITC–dextran is shown from a single frame taken 75 seconds after the stimulation was initiated (1.7 µg/ml in pipette; left image). The middle image in Fig. 6A shows Fura Red from the same frame and indicates the region selected for the kymograph of Ca2+ changes in the cell (yellow line with arrow). The right image shows the exocytotic bursts from a series of FITC fluorescence images integrated over the stimulation time period of 75 seconds. Fig. 6B shows the kymograph of this cell that compares the time course of Ca2+ elevation (top panel) with the time course of degranulation events corresponding to FITC–dextran bursts (middle panel). Fura Red exhibited a decrease in fluorescence when Ca2+ levels were elevated, such that darker intervals in this timeline correspond to increased cytoplasmic Ca2+, and brighter red intervals correspond to lower Ca2+ levels. The slanted dashed line in Fig. 6B (top panel) indicates the initiation of the Ca2+ wave and its propagation from the protrusion towards the cell body. As seen in the bottom panel in Fig. 6B, degranulation events detected with FITC–dextran initiated frequently during intervals of elevated cytoplasmic Ca2+ that correspond to the peaks of Ca2+ oscillations. To quantify this comparison, we assessed the temporal distribution of granule exocytosis events relative to the peaks of oscillatory elevations in cytoplasmic Ca2+ as depicted in Fig. 6C. This analysis is shown in Fig. 6D,E, and it revealed that the majority of exocytotic events occur in conjunction with, and typically just following, the peaks of Ca2+ oscillations.
In our initial investigation of the spatial distributions of exocytotic events, we compared these distributions for cells that initiate Ca2+ waves from protrusions (‘P-wave’ cells) with cells initiating Ca2+ waves from the cell body (‘B-wave’ cells) when a high dose of antigen (1.7 µg/ml) was used for stimulation. As shown in Fig. 7A, left side, only a small percentage of exocytotic events occurred along protrusions compared with the cell body at this high antigen dose, and the difference in total number of events for cells that initiate Ca2+ responses as P-waves versus B-waves was small. These cells were shown to initiate Ca2+ waves from protrusions with a higher probability at low doses of antigen (Cohen et al., 2009), and we investigated whether a low dose of antigen (1.7 ng/ml) caused a higher probability of exocytosis along these extended protrusions than seen with a high dose of antigen. When this low dose of antigen was used to stimulate degranulation, the average number of exocytotic events per cell was less than at the high antigen dose, and this difference was particularly evident for cells with Ca2+ waves that initiated in the cell body (Fig. 7A, right). However, at this low dose of antigen, we found that there was a substantially larger percentage of exocytotic events that occurred along protrusions for cells that initiate Ca2+ responses at these protrusions. More than 50% of the exocytotic events occurred in protrusions under these conditions, and this percentage was greater than that for cells stimulated with the high dose of antigen. When the time course for exocytosis events was compared for cells that initiate Ca2+ responses as P-waves, events in protrusions at early times (≤120 seconds after stimulation) occurred at similar frequencies for cells at high and low antigen doses. Interestingly, the larger number of events along protrusions at the low dose of antigen was seen primarily at longer times after stimulation (Fig. 7B). This suggests that the P-wave somehow primes that region of the cell for exocytotic events that occurs many tens of seconds after the initiating wave. In addition, we found that the increased occurrence of exocytosis events along the protrusions following P-waves correlated with appearance of one or more local Ca2+ transients, observed as an abrupt and isolated change in Fura Red fluorescence before wave initiation (Fig. 7C). This observation suggests that local exocytosis is facilitated by these transient and spatially restricted Ca2+ elevations.
TRPC1 knockdown inhibits antigen-stimulated granule exocytosis
We previously found that Ca2+ entry through TRPC1 is important for Ca2+-dependent P-wave initiation, which is particularly manifest at low doses of antigen (Cohen et al., 2009). To determine whether TRPC1 contributes to antigen-stimulated degranulation in RBL cells, we compared the degranulation response of these cells in which TRPC1 is knocked down by expression of a vector containing a TRPC1-interfering small hairpin (sh) RNA sequence with one containing a control shRNA sequence. As shown in Fig. 8A–D, TRPC1 shRNA caused a significant reduction in the degranulation response and was particularly effective in inhibiting degranulation events in protrusions. These results provide evidence that TRPC1 channels contribute to the degranulation response in RBL mast cells. Furthermore, they are consistent with results in Fig. 7A showing enhanced exocytosis from protrusions at low antigen concentration, conditions in which TRPC1 is important for Ca2+ wave initiation from these morphological features (Cohen et al., 2009).
The present study provides several lines of evidence to establish the use of FITC–dextran as a reliable marker for live-cell degranulation imaging. Our real-time imaging method permits simultaneous monitoring of Ca2+ mobilization and secretory granule exocytosis with high spatial and temporal resolution, and thus allows direct comparison of these co-stimulated processes. We found that granule exocytosis occurs more frequently along extended cell protrusions when Ca2+ waves initiate from these structures, particularly at limiting antigen concentration. Under these conditions, Ca2+ influx through TRPC1 contributed significantly to granule exocytosis in these cells, consistent with our previous results indicating its participation in Ca2+ wave initiation from protrusions (Cohen et al., 2009) and with recent evidence for its role in stimulated degranulation in mouse bone-marrow-derived mast cells (Suzuki et al., 2010).
FITC–dextran is a reliable marker for real-time detection of granule fusion in mast cells
Consistent with previous results (Dragonetti et al., 2000), we found that pinocytosed fluorescent dextrans colocalize to a large extent with CD63 and VAMP7, two membrane markers for secretory lysosomes, and these secretory lysosomes frequently concentrated at the tips of extended protrusions (Fig. 1). Stimulated exocytosis events were monitored by local increases in the pH-sensitive fluorescence of FITC–dextran, and these are observed as either transient full fusion events (Fig. 2A,B) or more sustained changes in FITC–dextran fluorescence, which represents either kiss-and-run (Fig. 2C,D) or compound exocytosis (Fig. 2E,F) events. It is possible that these different degranulation dynamics represent distinct subsets of secretory granules that use different molecular machineries to promote the membrane fusion. However, we saw little evidence for granule heterogeneity by other criteria, and the time course for stimulated FITC–dextran exocytosis events was similar to that for β-hexosaminidase release under optimal conditions (data not shown). Simultaneous labeling of these granules with FITC–dextran and TxRed–dextran resulted in stimulated exocytosis events in which the transient increase in FITC fluorescence corresponded in time and space to the loss of TxRed fluorescence (pH insensitive) from the same granule (Fig. 3), providing direct evidence for the correspondence of transient FITC fluorescence increases with the emptying of granule contents. In addition, the secretory events detected by a local burst of FITC–dextran fluorescence temporally correlated with exocytosis events detected by 5-HT amperometry (Fig. 4). Together, these results provide strong evidence for the identification of stimulated FITC–dextran fluorescence bursts as granule exocytosis events.
Degranulation is significantly influenced by the duration of cell adhesion
The FITC–dextran loading method provides not only the means for single-cell-based imaging, but also provides an easy, time-resolved and sensitive approach to examine population-based exocytosis studies. With that in mind, we monitored antigen-stimulated degranulation of suspended RBL cells as the net increase in FITC–dextran fluorescence detected by steady state fluorimetry (Fig. 5, top panel), which revealed a somewhat slower time course for this process relative to that observed for attached cells in imaging experiments. Therefore, we monitored the time course for antigen-stimulated exocytotic events detected as FITC–dextran fluorescence bursts as a function of RBL cell attachment time. We observed a strong dependence of the onset and rate of granule exocytosis on the time of cell attachment in the range of 20–90 minutes (Fig. 5, bottom panel). The extent of stimulated degranulation was near-maximal after cells are adherent for 90 minutes, and the rate was much slower for attachment durations of 60 minutes or less. Because Ca2+ mobilization is not rate limiting for degranulation of suspended cells (Pierini et al., 1997), other aspects of signaling evidently provide this kinetic regulation that depends on cell adherence and morphology. A role for cell adherence in stimulated exocytosis was demonstrated previously for human eosinophils (Fujiu et al., 2002), as well as for RBL mast cells (Apgar, 1997), but the time dependence of this enhancement was not previously characterized. Although the molecular basis for the time needed for cell attachment to achieve optimal degranulation of RBL mast cells remains to be determined, it could be related to the time needed for development of stable cell protrusions.
Spatial localization of the initiation of Ca2+ mobilization influences the spatial distribution of granule exocytosis
Our simultaneous imaging approach allowed us to determine how the subcellular spatial and/or temporal properties of stimulated granule exocytosis correlate with spatially localized Ca2+ mobilization. Our results confirm that granule exocytosis is temporally correlated to the peaks of Ca2+ oscillations, as previously established using Ca2+ indicators and amperometry (Fig. 6) (Kim et al., 1997). In addition, our results provide new evidence that the site of Ca2+ wave initiation strongly influences the region of granule exocytosis at low doses of antigen (Fig. 7). Under these conditions, more than half of the exocytosis events were localized along protrusions when the initiating Ca2+ wave begins in this region, and this compares with less than 20% of these events along protrusions when the high dose of antigen was used. Exocytosis along protrusions was also enhanced at low doses of antigen when Ca2+ waves initiate from the cell body, but this effect was substantially less than that for cells with Ca2+ waves initiating from protrusions (Fig. 7A). These results suggest that the protrusions are involved both in antigen sensing distal to the cell body and in secretion of mediators at these extended sites. It is possible, for example, that these protrusions extend through the epithelial cell layer of intestinal villi to sense luminal antigen and to deliver mediators at that location in response to IgE receptor activation.
The mechanism that correlates Ca2+ wave initiation in protrusions and enhanced exocytosis events in those regions is not yet clear because these events typically occur tens of seconds after the initiating Ca2+ wave. It is possible that Ca2+ wave initiation in protrusions primes this region for enhanced exocytosis, or, alternatively, this localization of Ca2+ wave initiation might reflect more effective coupling of stimulus and secretion in this region. The roles of TRPC1 in promoting Ca2+ wave initiation along protrusions (Cohen et al., 2009), as well as in enhancing exocytosis at limiting doses of antigen (Fig. 8) might underlie this correlation, and further experiments are in progress to test these possibilities.
Materials and Methods
The GCaMP2 construct was provided by Michael Kotlikoff, Cornell University College of Veterinary Medicine, Ithaca, NY. The CD63–GFP and EGFP–VAMP7 constructs were gifts from Juan Bonifacino and Paul Roche (NIH, Bethesda, MD), respectively. The shRNA constructs were from OriGene (Rockville, MD).
Chemicals and reagents
Fura-Red–AM and Texas Red (TxRed)–dextran (10 kDa) were purchased from Invitrogen/Molecular Probes (Eugene, OR). FITC–dextran (150 kDa), 5-hydroxytryptamine (5-HT), thapsigargin and A23187 were purchased from Sigma (St Louis, MO).
RBL-2H3 cells were maintained in monolayer culture in Minimum Essential Medium supplemented with 20% fetal bovine serum (Atlanta Biologicals, Norcross, GA) and 10 µg/ml gentamicin sulfate. All tissue culture reagents were obtained from Gibco (Grand Island, NY) unless otherwise noted.
Single-cell degranulation imaging
RBL-2H3 cells were plated as previously described (Cohen et al., 2009) in 35 mm MatTek (Dover, MA) coverslip dishes in medium containing 0.5 µg/ml anti-2,4-dinitrophenyl (DNP) IgE (Posner et al., 1992) and FITC–dextran (1 mg/ml). 5-hydroxytryptamine (5-HT, 0.2 mM) was also added to the medium in some experiments to increase the diameter of secretory granules for better visualization (Williams et al., 1999); we found that this addition has no significant effects on the kinetics of granule exocytosis measured either by FITC–dextran or by β-hexosaminidase release. Transfection of RBL-2H3 cells with GFP fusion proteins and shRNA plasmids was carried out as described previously (Cohen et al., 2009). 24 hours after transfection, cells were washed with fresh medium and incubated for 1–2 hours at 37°C. In some experiments, cells were loaded with Fura Red at 0.5 µM for 10 minutes at 37°C. Just before imaging, cells were washed into buffered salt solution (BSS, 135 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5.6 mM glucose, 20 mM HEPES, pH 7.2, 1 mg/ml BSA), and individual MatTek dishes were mounted on a Zeiss LSM 510 Meta confocal microscope with a heated 40× (NA 1.4) oil-immersion objective at 37°C. The fluorophores of GFP fusion proteins and FITC–dextran were excited using the 488 nm line of a krypton-argon laser and viewed with a 505–550 nm band-pass filter. TxRed–dextran and Fura Red were excited using the 561 nm and 488 nm laser lines, and viewed with 575 nm and 650 nm long-pass filters, respectively.
Selected cells were stimulated with antigen delivered in a micropipette as previously described (Cohen et al., 2009). Briefly, cells were approached with a ∼5 µm diameter pulled glass capillary, typically positioned within 100 µm from the cell and pre-filled with stimulating solution. Cells were imaged at 10–30 Hz while applying a puff of 10 seconds at 5 p.s.i. from the pipette. To evaluate the contribution of TRPC1 in antigen-stimulated degranulation, RBL cells were transfected with either TRPC1 shRNA or a control shRNA vector as previously described (Cohen et al., 2009), together with a monomeric red fluorescent protein (mRFP)-containing vector at a 2.5 µl DNA ratio. Transfected cells were sensitized with IgE, loaded with FITC–dextran and 5-HT, and mRFP-positive cells were evaluated for degranulation responses stimulated with 7 ng/ml antigen in the micropipette.
Fluorimetry-based degranulation assay
106 cells in 2 ml of full medium were cultured in the presence of 1–2 mg/ml FITC–dextran and IgE (0.5 µg/ml) for 24 hours at 37°C, then harvested, washed, and resuspended in 2 ml of BSS. 0.5 ml of these cells was diluted to 2 ml in a stirred acrylic cuvette, and FITC fluorescence (ex, 490 nm; em, 520 nm) was monitored at 37°C using an SLM 8100C steady-state fluorimeter (SLM Instruments, Urbana, IL) in a time-based acquisition mode. Lysis of cells by addition of 0.1% Triton X-100 at the end of each experiment provided unquenched FITC fluorescence that was used to normalize the other fluorescence measurements in the same sample.
Combined single-cell amperometry and imaging
Cells were sensitized with IgE and loaded with FITC–dextran and 5-HT in overnight culture as described above. Amperometry was performed using custom-made carbon fiber electrodes and a patch-clamp amplifier (EPC-8, HEKA Elektronik). Electrodes were positioned in contact with the cell surface, and voltage was maintained at +700 mV using a reference Ag/AgCl electrode. The measured current was low-pass filtered at 500 Hz using the built-in analog filter of the EPC-8 amplifier. Cells were stimulated using a glass pipette containing 1.7 µg/ml antigen positioned ∼40 µm away from the cell in conjunction with a pressure application system (PicoSpritzer II, Parker-Hannifin/General Valve Corporation). Amperometric recordings were collected for up to 10 minutes after stimulation, and the data were digitized at a rate of 2 kHz by a 16-bit resolution NIDAQ board (BNC- 2090, National Instruments). Recordings were analyzed as previously described (Mosharov and Sulzer, 2005). Spikes with amplitude <10 pA, with half-width >300 mseconds, or with overlapping areas were excluded from the analysis.
Fluorescence images of individual cells used in the analyses are representative of several experiments. Off-line image analysis was carried out using Zeiss ZEN image analysis software (Carl Zeiss) and ImageJ (NIH, Bethesda, MD). Image processing of FITC–dextran degranulation for noise and background reduction was achieved in two steps as previously described (Demuro and Parker, 2005). Briefly, ratio images were formed by dividing each ‘raw image’ frame in the time series stack by an average of the initial 10–20 frames acquired before applying the stimulus. The fluorescence signal at any pixel in the resulting ratio image thus represents the change in fluorescence as a fraction of the resting fluorescence (ΔF/Fo). To further minimize background noise, a highly smoothed copy of each frame was formed by applying a Gaussian blur with a width of 10 pixels, and this was subtracted from its original ratio image to form the final, corrected image in the time sequence.
Degranulation kinetics were analyzed either by manual counting of FITC fluorescence bursts, or by using the SparkFinder plug-in for ImageJ (Babraham Bioinformatics). Data were processed and plotted using Origin 8 (OriginLab) and Excel (Microsoft). Statistical comparisons between experiments were performed using the Student's t-test.
We thank Khajak Berberian and Manfred Lindau for help with the amperometry measurements and for valuable discussions, and we thank Hong-Tao Ma and Michael Beaven (NHLBI, NIH, Bethesda, MD) for cDNA constructs.
This work was supported by National Institutes of Health [grant number AI022449]; and by the Nanobiotechnology Center at Cornell funded in part by the National Science Foundation [grant number ECS-9876771]. Deposited in PMC for release after 12 months.