Electrophysiological Characterization of Pancreatic Islet Cells in the Mouse Insulin Promoter-Green Fluorescent Protein Mouse
http://www.100md.com
《内分泌学杂志》
Departments of Medicine and Physiology (Y.M.L., I.A., L.S., R.G.T., N.E.D., H.Y.G.), University of Toronto, Toronto, Canada M5S 1A8
Department of Medicine (M.H.), University of Chicago, Chicago, Illinois 60637
Abstract
We recently reported a transgenic [mouse insulin promoter (MIP)-green fluorescent protein (GFP)] mouse in which GFP expression is targeted to the pancreatic islet -cells to enable convenient identification of -cells as green cells. The GFP-expressing -cells of the MIP-GFP mouse were functionally indistinguishable from -cells of normal mice. Here we characterized the ionic channel properties and exocytosis of MIP-GFP mouse islet - and -cells. -Cells displayed delayed rectifying K+ and high-voltage-activated Ca2+ channels and exhibited Na+ currents only at hyperpolarized holding potential. -Cells were nongreen and had both A-type and delayed rectifier K+ channels, both low-voltage-activated and high-voltage-activated Ca2+ channels, and displayed Na+ currents readily at –70 mV holding potential. -Cells had ATP-sensitive K+ channel (KATP) channel density as high as that in -cells, and, surprisingly, -cell KATP channels were more sensitive to ATP inhibition (IC50 = 0.16 ± 0.03 mM) than -cell KATP channels (IC50 = 0.86 ± 0.10 mM). Whereas -cells were rather uniform in size [2–4.5 picofarad (pF)], -cells varied vastly in size (2–12 pF). Of note, small -cells (<4.5 pF) showed little exocytosis, whereas medium -cells (5–8 pF) exhibited vigorous exocytosis, but large -cells (>8 pF) had weaker exocytosis. We found no correlation between -cell size and their Ca2+ channel density, suggesting that Ca2+ influx may not be the cause of the heterogeneity in exocytotic responses. The MIP-GFP mouse therefore offers potential to further explore the functional heterogeneity in -cells of different sizes. The MIP-GFP mouse islet is therefore a reliable model to efficiently examine -cell and -cell physiology and should greatly facilitate examination of their pathophysiology when the MIP-GFP mice are crossed with diabetic models.
Introduction
EACH PANCREATIC ISLET cell type, through their secretion (insulin, glucagon, somatostatin), contributes to glucose homeostasis (1, 2). Whereas the more abundant -cells (70% of islet cells) are easier to examine, -cells (20%), - and F-cells (the remaining 10%) are difficult to identify (3). This limitation has resulted in the relatively slow advances in determining the biology of non--cells, particularly the -cell, whose dysfunction is also a major contributing factor to the abnormal glucose homeostasis in diabetes (2).
A number of strategies have been used to identify each islet cell type. -Cells can be isolated from other non--cells by autofluorescence-activated cell sorting, which can yield up to 80–90% pure -cell and -cell populations (4). Cell size has been used to distinguish -cells from - and -cells because -cells are on average larger [6 picofarad (pF)] than - and -cells (3 pF) (5). Therefore, islet cells 6 pF or larger are likely to be -cells (6). Functionally, mouse -, -, and -cells possess distinct sets of ionic channels (5, 6, 7, 8, 9). -Cells have delayed rectifying K+ channels and high-voltage-activated (HVA) Ca2+ channels and exhibit voltage-gated Na+ currents only at hyperpolarized holding potential (VH). This is in contrast to -cells, which have both A-type K+ channels and delayed rectifier K+ channels, both low-voltage-activated (LVA) and HVA Ca2+ channels, and have voltage-gated Na+ currents readily at –70 mV VH. -Cells can be distinguished from -cells by lacking A-type K+ channels and T-type Ca2+ channels. This electrophysiological fingerprinting provides a reliable way to distinguish islet cell types but are nonetheless cumbersome. Furthermore, once diabetes sets in, these ion channel properties of the islet cells may become perturbed, which would make it very difficult, if not impossible, to determine the precise pathophysiological contribution of each of the islet cell to the abnormal glucose homeostasis in diabetes.
To surmount these limitations, we created a transgenic mouse that has green fluorescent protein (GFP) specifically tagged to the -cells (10), thus providing an instant visual discrimination between -cells (green) and non--cells (nongreen). These mouse insulin promoter (MIP)-GFP mice develop normally, exhibit normal glucose-tolerance and insulin secretion, and the islet -cells are morphologically and functionally similar to control -cells (10). Here we followed up our initial report by examining single islet cell function of the MIP-GFP mouse, including ion channel activities and exocytosis, by electrophysiology. Our results on the ion channel properties of the MIP-GFP mouse islet -cells and -cells are similar to previous reports on the electrophysiological descriptions of these cell types (5, 6, 7, 8, 9), which indicate this MIP-GFP model to be an excellent and valid model. Importantly, in the process of characterizing these islet cells of the MIP-GFP mice, we revealed some novel insights. First, whereas - and -cells had comparable KATP channel densities, -cell KATP channels had a 5-fold higher sensitivity to ATP inhibition than -cell KATP channels. Second, there was a large variation in the size of the green -cells ranging from 2–12 pF, which remarkably exhibited very different exocytotic capacity, suggesting the possibility of distinct subpopulations. In contrast, the nongreen non--cells, which include - and -cells, were of uniformly small size (2–4.5 pF). The MIP-GFP mouse islet cell preparation is therefore a powerful strategy to reliably and efficiently study the normal biology of not only -cells but also - and -cells. Future studies directed at crossing the MIP-GFP mice with diabetic models will further reveal the contribution of each of these islet cell types to the abnormal glucose homeostasis.
Materials and Methods
Materials
Antibodies against insulin and glucagon were purchased from Dako Cytomation (Glostrup, Denmark) and Sigma (St. Louis, MO), respectively. Tetraethylammonium-Cl (TEA), 4-aminopyridine, nifedipine, glibenclamide, and Mg-ATP were from Sigma. Tetrodotoxin (TTX) was from Alomone Labs (Jerusalem, Israel). Phosphatidylinositol-4,5-bisphosphate (PIP2) and U73122 were purchased from EMD Biosciences (San Diego, CA). PIP2 was dissolved in pipette solution [for ATP-sensitive K+ channel (KATP) current measurement] at 1 mM, sonicated for 15 min, and then used on the same day of experiment.
Islet isolation
Mouse pancreatic islets were isolated by collagenase digestion as described previously (11). Islets were dispersed to single cells with 0.05% trypsin (Sigma) in Ca2+- and Mg2+-free PBS. Islet cells were plated on glass coverslips in 35-mm dishes and cultured in RPMI 1640 medium containing 1 mM pyruvate, 11 mM glucose, 0.2% NaHCO3, 10% fetal bovine serum, 10 mM HEPES, 100 U/ml penicillin G sodium, and 100 μg/ml streptomycin sulfate (Invitrogen Canada Inc., Burlington, Ontario, Canada). Islet cells were cultured overnight before electrophysiological recordings.
Recording of Kv and Ca2+ currents
Mouse islet cells were voltage clamped in the whole-cell configuration using an EPC-9 amplifier and Pulse software (HEKA Electronik, Lambrecht, Germany) as we previously described (11). Pipette tip resistances ranged from 3 to 5 M when filled with intracellular solutions. The intracellular solution for Kv current measurement contained (in mM): 140 KCl, 1 MgCl2, 1 EGTA, 10 HEPES, and 5 MgATP (pH 7.25 adjusted with KOH). The intracellular solution for Ca2+ current measurement contained (in mM): 120 CsCl, 20 TEA-Cl, 1 MgCl2, 1 EGTA, 10 HEPES, and 5 MgATP (pH 7.25 adjusted with CsOH). The bath solution for Kv current measurements contained (in mM): 140 NaCl; 4 KCl, 1 MgCl2, 2 CaCl2, 2 D-glucose, and 10 HEPES (pH 7.3 adjusted with NaOH). The bath solution for Ca2+ current measurement was the same as above but also supplemented with 20 mM TEA-Cl and 10 μM TTX, and the concentration of CaCl2 was increased to 10 mM. After a whole-cell configuration was established, the cell was held at –70 mV or other indicated VH and subject to various experimental protocols as detailed in Results and figure legends. All experiments were performed at room temperature (22 C). Data for steady-state inactivation were fit by the Boltzmann equation: I/Imax = 1/{1 + exp [(V-V1/2)/k]}, where V1/2 is the half-maximal inactivation potential and k the slope factor.
Recording of KATP currents
The intracellular solution for measurement of KATP currents contained (in mM): 140 KCl, 1 MgCl2, 1 EGTA, 10 HEPES, and various concentrations of MgATP as indicated (pH 7.25 adjusted with KOH). Each individual cell was tested with a single concentration of ATP dialyzed through the recording pipette. The bath solution for KATP current measurements contained (in mM): 140 NaCl, 4 KCl, 1 MgCl2, 2 CaCl2, 2 D-glucose, and 10 HEPES (pH 7.4 adjusted with NaOH). In a typical recording, the -cells are identified as green cells (10), and the -cells are identified by being nongreen and having A-type K+ currents (5). After a whole-cell configuration was established in -cells, the cell was held at –80 mV, and a test pulse of –30 mV (500 ms) was given immediately to test the presence of A-type K+ currents. Subsequently the cell was then stimulated by a –140 mV hyperpolarizing voltage step (500 msec) every 10 sec. Once KATP currents reached maximum, the cell was subject to a series of voltage pulses from –140 to –40 mV (500 msec) at 20 mV increments to obtain the current-voltage relationship (I-V). All experiments were performed at room temperature (22 C). Concentration-response curves for ATP inhibition of KATP channels are fit by the Hill equation:
where I is the current in the presence of a certain pipette (ATP), Imax is the maximum current, [ATP] is the concentration of ATP inside the pipette, Kd is the apparent dissociation constant, and n is the Hill coefficient.
Membrane capacitance measurement
Recording electrodes were coated with orthodontic wax (Butler, Guelph, Ontario, Canada) close to the tips and heat polished. Resistances ranged from 3 to 5 M when pipettes were filled with the intracellular solution, which contained (in mM): 125 K-glutamate, 10 KCl, 10 NaCl, 1 MgCl2, 5 HEPES, 0.05 EGTA, 0.1 cAMP, and 4 MgATP (pH to 7.1 by KOH). The extracellular solution contained (in mM): 140 NaCl, 4 KCl, 1 MgCl2, 2 CaCl2, 5 D-glucose, and 10 HEPES (pH 7.3 adjusted with NaOH). Membrane capacitance (Cm) was estimated by the Lindau-Neher technique, implementing the Sine + DC feature of the Lock-in module (40 mV peak to peak and a frequency of 500 Hz) in the standard whole-cell configuration. Recordings were conducted using an EPC9 patch clamp amplifier and Pulse software. Exocytotic events were elicited by a train of eight 500-msec depolarizing pulses (1-Hz stimulation frequency) from –70 to 0 mV. All Cm measurements were performed at 28 C.
Confocal immunofluorescence microscopy
Laser confocal immunofluorescence microscopy was performed as previously described (11). Dispersed pancreatic islet cells were plated on glass coverslips coated with 0.01% poly-L-lysine. After 2 h, the cells were fixed in 2% formaldehyde for 30 min, then treated with 5% goat serum and 0.1% saponin for 1 h, and finally immunolabeled with either mouse monoclonal antiglucagon (1:2000) or guinea pig monoclonal antiinsulin (1:100) for 2 h. The coverslips were rinsed with 0.1% saponin in PBS and then incubated with appropriate fluorescent-labeled secondary antibodies (either Texas Red or rhodamine red) for 1 h. After rinsing once more, coverslips were mounted on slides in a fading retarder (0.1% p-phenylenediamine in glycerol) and examined using a laser-scanning confocal imaging system (LSM 510, Zeiss, Oberkochen, Germany).
Statistics
Results are presented as means ± SEM. ANOVA was used for multiple-group comparisons, and statistical significance was determined by Student-Newman-Keuls test. P < 0.05 was considered statistically significant.
Results
Visualization of green -cells and nongreen islet cells of MIP-GFP mice
Characterization of voltage-gated K+ (Kv) and Na+ currents of cells in intact islets
Whole islets were used in these experiments. It is often difficult to distinguish the nongreen cells from the green cells in large islets because the green fluorescence of the GFP-containing -cells in a deeper layer could emit through the transparent nongreen cells on the islet surface. We therefore chose smaller islets and rolled them around until at some angle we saw exclusively green cells or exclusively nongreen cells, and then we recorded one of them.
When the membrane VH was at –70 mV, depolarizing the -cell triggered outward K+ currents but not inward Na+ currents (Fig. 3A; inset shows the expanded very early time frame; note absence of Na+ currents). The K+ currents were sensitive to TEA. When VH was at –120 mV, depolarizing the -cell triggered both TEA-sensitive K+ currents and Na+ currents (Fig. 3B and inset). The latter were resistant to TEA block. We then showed that these -cell Na+ currents could be completely abolished by 10 μM TTX, which expectedly did not affect -cell Ca2+ currents (Fig. 4). Steady-state inactivation experiments showed that the voltage at which half of the -cell Na+ channels were inactivated (V1/2) was –100.4 ± 1.9 mV (Fig. 3D). In Fig. 3C, nongreen -cells could be distinguished functionally from -cells by displaying both Na+ currents and TEA-sensitive K+ currents upon depolarization at a physiological VH (–70 mV). Accordingly, V1/2 of -cell Na+ channel inactivation was –46.8 ± 5.3 mV (Fig. 3D). Another distinctive feature of -cells was the rapidly inactivating TEA-insensitive A-type K+ currents (Fig. 3C, asterisk). Addition of 100 μM 4-aminopyridine abolished most of the A-type K+ currents (not shown).
Characterization of Kv and Na+ currents of single cells
As indicated above, it is often difficult to distinguish green and nongreen cells in intact islets. Furthermore, the ion channel properties could be influenced by the confounding paracrine effects of secreted hormones from neighboring cells within an intact islet. We therefore carried out the rest of our characterization studies in dispersed single islet cells (Fig. 5). When VH was at –70 mV, depolarizing the -cell triggered TEA-sensitive outward K+ currents but not inward Na+ currents (Fig. 5A and inset). However, when VH was at –120 mV, depolarizing the -cell triggered both K+ currents and Na+ currents (Fig. 5B and inset). V1/2 for Na+ channel inactivation was –98 ± 2.8 mV (Fig. 5E). Single dispersed -cells displayed TEA-sensitive K+ currents but not Na+ currents upon depolarization (VH = –70 mV). The reason for this discrepancy is unclear, but the loss of Na+ channels in the dispersed -cells but not -cells may in part be due to the fact that all -cells are located on the islet surface and therefore might be more vulnerable to enzymatic and mechanical stresses during islet isolation and dispersion. -Cells are located in the islet core and might be less susceptible. Nevertheless, single -cells could still be distinguished functionally from other islet cell types by displaying A-type K+ currents at physiological VH (–70 mV) (Fig. 5C, asterisk). The latter became larger at hyperpolarized VH (–120 mV) (Fig. 5D). Accordingly, V1/2 for A-type K+ channel inactivation was –75.7 ± 0.3 mV (Fig. 5F). We found that about 65% of dispersed nongreen cells exhibited A-type K+ currents and were therefore -cells (7, 9); the remaining presumably were -cells, which we did not further characterize.
Characterization of single -cell exocytosis and voltage-dependent Ca2+ channels (VDCC)
In our characterization of -cell Na+ channels and Kv channels so far, we studied -cells of medium size range (4–8 pF) because they were the most numerous (Fig. 2). With the MIP-GFP mouse islet cell preparation, we became increasingly aware that the -cells were of a large range of sizes (Figs. 1D and 2). We then asked whether there was a functional heterogeneity in -cells of different sizes by examining their exocytotic capacity. When small single -cells (3–4.5 pF) were stimulated by a train of eight depolarizations, there were only very small changes in Cm, indicating that small -cells were poor secretors (Fig. 6A, left). -Cells of medium size range (5–8 pF) showed vigorous exocytosis (Fig. 6A, middle). Large -cells (>8 pF; Fig. 6A, right) did not exhibit larger exocytosis than the medium-size -cells but instead showed generally smaller responses. When Cm increases (Fig. 6B, left) and normalized Cm increases (Fig. 6B, right) are plotted against cell sizes, there is a bell-shaped relationship, in which it becomes obvious that medium -cells were the most vigorous secretors. Exocytosis by small cells (10.7 ± 0.7 fF/pF, n = 4) was much weaker (P < 0.05) than that by medium-size cells (76.4 ± 18.9 fF/pF, n = 8). Exocytosis by large cells (24.6 ± 8.9 fF/pF, n = 4) was generally much weaker than that by medium-size cells.
Because the opening of VDCC provides the Ca2+ influx to trigger exocytosis, we examined the VDCC in single -cells to see whether a possible difference of VDCC property might be the underlying cause of the heterogeneity in exocytotic capacity observed between the different size -cells. Figure 7A shows the Ca2+ currents in -cells of different sizes. Figure 7B shows the I-V relationship of the Ca2+ currents triggered in three groups of -cells based on different sizes. The small increase in current density with cell size is not statistically significant. When peak current density of each individual cell was plotted against cell size, there was no significant correlation (Fig. 7C; r = 0.40; P = 0.14). Therefore, the difference in exocytotic capacity in -cells of different sizes may not be due to differences in VDCC density per se. Previous reports showed that mouse -cells possess only HVA Ca2+ channels (i.e. L-, P/Q-, and R-type VDCC) (8). Consistently we observed that in -cells of any size group, the activation threshold of the VDCC was from –30 to –20 mV (Fig. 7B), indicating the presence of only HVA Ca2+ channels but not LVA Ca2+ channels. LVA Ca2+ channels may be inactivated by VH of –70 mV or greater. However, when VH was –90 mV, activation threshold of any -cell was still from –30 to –20 mV (data not shown), ruling out the possibility of presence of LVA Ca2+ channels in -cells. Nifedipine (10 μM, selective L-type VDCC blocker) inhibited 64 ± 4 and 58 ± 6% (n = 3; P < 0.05) of Ca2+ currents in small and medium -cells, respectively, indicating the majority of HVA Ca2+ channels were L-type.
Characterization of single -cell exocytosis and VDCC
We next examined the VDCC in -cells, which have been shown to possess both LVA and HVA Ca2+ channels (5, 9). As shown in Fig. 8, A and B, when the -cell was held at –100 mV and depolarized to different potentials in the presence of 10 μM TTX, Ca2+ currents could readily be triggered at –40 mV and were transient (fast activation and inactivation) in nature, characteristic of LVA Ca2+ channels or the T-type VDCC (Fig. 8A, left). Ca2+ currents triggered at stronger depolarization (–10 or 0 mV) began to show, in addition to the transient current, a sustained component, indicative of HVA Ca2+ channels. When VH was at –50 mV, which would readily inactivate LVA Ca2+ channels, depolarization expectedly only triggered the slow-inactivating HVA Ca2+ currents (Fig. 8A, middle). The difference between the Ca2+ currents obtained at the two VH yields the LVA Ca2+ currents (Fig. 8A, right). The results are summarized in the I-V curves shown in Fig. 8B (n = 4). Thus, the total currents (VH at –100 mV) were the sum of HVA (VH at –50 mV) and LVA Ca2+ currents (the difference currents). The HVA and LVA Ca2+ currents typically peaked at +10 mV and –20 mV, respectively. We then examined -cell exocytosis. Figure 8C shows a representative -cell exocytotic response (275 ± 35 fF at eighth stimulation, n = 3), which was within the range previously reported (5, 12, 13).
KATP channel sensitivity to ATP block was higher in - than -cells
KATP channels are present in not only -cells but also -cells (5, 14, 15, 16). By dialyzing 0.3 mM ATP into mouse -cells, Barg et al. (5) reported a low KATP channel density. Using whole-cell configuration, we here reassessed the -cell KATP current density by dialyzing different concentrations of ATP into each cell via the pipette (each individual cell was dialyzed with only one concentration of ATP). When a low concentration of ATP (0.05 mM) was dialyzed into the -cell, large currents were indeed triggered (Fig. 9Ai). The currents were sensitive to bath application of 1 μM glibenclamide (Fig. 9Aii), indicating that these are indeed KATP channel currents. Pipette ATP concentration-dependently inhibited the channel so that 5 mM ATP completely blocked the channel (Fig. 9Aiii–v). Some -cells did not develop any KATP currents at all when dialyzed with 1 mM ATP (Fig. 9Aiv, lower trace). When we dialyzed a low concentration of ATP (0.05 mM) into -cells, we also detected large glybenclamide-sensitive currents (Fig. 9Bi and Bii). Pipette ATP inhibited the channel in a concentration-dependent manner such that 5 mM ATP completely blocked the channel. (Fig. 9Biii–v). Of note, 0.3 mM ATP, which triggers relatively small currents in -cells (Fig. 9Aiii), was in fact the maximal concentration triggering -cell KATP currents (Fig. 9Biii). When KATP channel densities of -cells and -cells are plotted against different concentrations of ATP (Fig. 9C), we observed that KATP channel densities in - and -cells were in fact comparable. The high KATP channel density was, however, not apparent in -cells at 0.3 mM ATP, which would be in agreement with the findings by Barg et al. (5). If normalized current densities are plotted against ATP concentrations, it is clear that -cell KATP channels were significantly (P < 0.05) more sensitive (IC50 = 0.16 ± 0.03 mM) to ATP inhibition than -cell KATP channels (IC50 = 0.86 ± 0.10 mM) (Fig. 9D).
What is the mechanism for the higher sensitivity of the -cell KATP channels to ATP PIP2 has been known to open KATP channel by decreasing the channel sensitivity to ATP (17, 18). We therefore questioned whether there may be a lack of PIP2 in the -cells to explain the heightened ATP sensitivity of the -cell KATP channels. If so, increasing PIP2 concentration in -cells by either direct introduction of exogenous PIP2 or inhibition of phospholipase C (by U73122) to increase endogenous PIP2 would reduce the KATP channel sensitivity, thereby increasing the KATP current density. At 0.3 mM ATP, -cell KATP channels were only slightly opened by only approximately 30% (see Fig. 9D). Here we dialyzed 100 μM PIP2 into the -cells in the presence of 0.3 mM intracellular ATP, but PIP2 failed to increase KATP channel opening (Fig. 9E). Consistently 10 μM U73122 inhibition of phospholipase C also failed to increase -cell KATP channel density in the presence of 0.3 mM intracellular ATP (Fig. 9E). These results indicate that the heightened ATP sensitivity of the -cell KATP channels is unlikely a consequence of a possible lower abundance of PIP2 at the -cell membrane. The lack of sensitivity of -cell KATP channels to PIP2 was surprising, and further work will therefore be required to determine the factor(s) responsible for the enhanced sensitivity of -cell KATP channels to ATP.
Discussion
The MIP-GFP transgenic mouse was created by the expression of GFP under MIP to specifically tag the islet -cells so that the -cells can be easily identified as green cells (10). We had already shown that the MIP-GFP mice develop normally and exhibit normal glucose homeostasis comparable with their control littermates (10). -Cells from MIP-GFP and control mice show similar Ca2+ signal response to glucose (10). Here we showed that MIP-GFP -cells exhibited exocytotic responses and ion channel properties (Kv channels, Na+ channels and VDCC) qualitatively and quantitatively similar to those of nontransgenic mice previously described (5, 6, 7, 8, 9). Therefore, MIP-GFP mouse -cells offer a reliable and efficient strategy to further examine novel aspects of -cell biology and electrophysiology.
In previous studies of -cells, large cells were deliberately chosen to avoid mixing with the smaller islets cells that may be - or -cells (5, 6, 7, 8, 9). Such studies might not have considered the biology of small and medium-size -cell subpopulations. With the MIP-GFP mouse islet cell preparation, examination of -cells of different sizes was made possible and indeed led to the realization that -cells could be variable in size ranging from 2 to 12 pF. We confirmed that green -cells were on average (5.5 pF) larger than -cells (2.8 pF) and any cell greater than 4.5 pF must be green -cells. We here reported a novel finding that the large variation in -cell sizes was associated with functional heterogeneity: small -cells were very poor secretors; -cells of medium-sizes exhibited vigorous secretion, whereas large -cells (>8 pF) had reduced exocytosis. This did not appear to be due to Ca2+ channel defects in smaller or larger cells because VDCC density remained relatively constant over the different cell sizes. Moreover, regardless of sizes, all -cells did not have LVA and only possessed HVA, the majority being L type.
Interestingly, -cell subpopulations have been reported to have different sensitivity to glucose stimulation (19). -Cells are known to undergo hypertrophy in type 2 diabetes as is the case in Zucker fa/fa rats, and insulin secreted from these large -cells was dysfunctional (20). It is therefore possible that in diabetes, a redistribution to a higher proportion of larger -cells may contribute to an overall poorer insulin secretory capacity and response to the increased demand. The precise mechanism to explain the heterogeneity in exocytotic capacity of the different subpopulations of islet -cells in health and in diabetes remained to be further explored. Of interest, the levels of exocytotic proteins (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) in the Zucker fa/fa rat islets were much lower than normal rat islets (20), which raises the possibility that perhaps these subpopulations of -cells in the MIP-GFP mouse might also exhibit different levels of exocytotic proteins, which would in part explain the differences in exocytotic capacity.
With the MIP-GFP islet cell preparation, - and -cells can be easily distinguished from small -cells by being nongreen. The -cell can then be distinguished from -cells (and -cells) by displaying A-type K+ channels (Figs. 3 and 5) and T-type Ca2+ channels (Fig. 8) (9). These channels have low activation threshold (A-type K+ channels, –60 mV; T-type Ca2+ channels, –40 mV) and rapid activation and inactivation rate. Whereas T-type Ca2+ channels perform the pacemaker function, A-type K+ channels provide repolarizng currents (6). -Cells, -cells, and -cells may also be distinguished by Na+ channel-gating properties (5, 6, 7, 8, 9). V1/2 values of steady-state inactivation of mouse -, - and -cell Na+ channel are –100, –47, and –28 mV, respectively (5, 6, 7, 8, 9) (Fig. 3). These values are far apart enough to distinguish between the mouse islet cells.
The differentiation between - and -cells using these functional markers (A-type K+ channels and T-type Ca2+ channels in -cells) discussed above may become elusive under pharmacological manipulation or during pathological states, particularly diabetes. For instance, activation of protein kinase A and protein kinase C has been shown to down-regulate A-type K+ channels in neurons or expressed in Xenopus oocytes (21, 22). Whereas T-type Ca2+ channels are not present in control mouse -cells, they have been reported to express in -cells of nonobese diabetic mice (23). Therefore, electrophysiological fingerprinting alone does not appear to be sufficient for accurate identification of islet cell type under certain situations, particularly diabetes. Tagging of GFP to the -cells therefore provides an independent yet complementary tool for convenient -cell and non--cell (predominantly -cells) identification, even during disease states.
Future developments of other transgenic animals having GFP-tagged -cells by crossing diabetic mice models with the MIP-GFP mice, or having GFP (or other fluorophore) genetically tagged to - or -cells, would be very welcome by the research community to then be able to reliably determine the contribution of each islet cell to the abnormal glucose homeostasis in diabetes. For example, MIP-GFP mice could be crossbred with gene knockouts or other transgenic animals, which exhibit selective islet cell (-cell) hyperplasia and dysfunction (24) or altered channel expression (15, 25). An alternative to transgenic technology is infection of islet cells with replication-defective recombinant adenovirus expressing GFP under insulin promoter control and has been successful in yielding more than 95% pure human -cells (26). However, there may be untoward toxic effects of the virus on the -cell channel properties, which would be confounding to those effects being examined and caused by the disease states (i.e. diabetes, glucolipotoxicity).
Another novel finding here is that -cell KATP channels had a 5-fold higher sensitivity to ATP inhibition than -cell KATP channels. This difference did not appear to be due to a possible lower abundance of PIP2 in the -cell plasma membrane because maneuvers that raised plasma membrane PIP2 concentration in -cells either exogenously or endogenously failed to increase -cell KATP channel opening. What additional information can we learn from the ATP concentration-inhibition curves (particularly the physiologically relevant range, 1–5 mM) Raising ATP from 1 to 5 mM increased the inhibition of -cell KATP currents from 65 to 100% and would expectedly depolarize the cell. By contrast, 1 mM ATP almost completely (93%) blocked -cell KATP currents. It has been estimated that ATP concentration inside the -cell is already higher than 1 mM at low glucose (27), suggesting that the fraction of closed KATP channels in -cells may actually exceed 93%. Because high glucose does raise ATP concentration in -cells (28), a minority of -cells may be expected to be depolarized by high glucose after a meal. This is consistent with the observation by Liu et al. (29) in which the KATP channel blocker tolbutamide depolarized only a minority of -cells. All these observations are also in agreement with the demonstration that high glucose causes only a very mild depolarization in -cells, which nonetheless, may be able to inactivate T-type Ca2+ channels and Na+ channels (15, 16).
In summary, this work on the electrophysiological characterization of ion channels and exocytosis of - and -cells of MIP-GFP mouse islets showed similar results as previous electrophysiological descriptions of mouse islet cells (5, 6, 7, 8, 9) and therefore fully validated the MIP-GFP mouse as an excellent model to greatly facilitate the examination of islet cell biology. Of note, the size-dependent functional heterogeneity among -cells may have important implication on furthering our understanding of islet secretory dysfunction in diabetes and may even be applicable to islet selection during islet transplantation (30). Further work will be required to examine these distinct subpopulations of -cells in health and diabetes and also determine the PIP2-independent factor that enhances the -cell KATP channel sensitivity to ATP. The groundwork we have done here on the MIP-GFP mouse islet cell preparation, including some novel insights we have gained in - and -cell biology, should serve the community in using this model as a powerful and highly reliable tool to further islet cell studies.
Footnotes
This work was supported by Juvenile Diabetes Research Foundation Grants 1-2005-1112 (to H.Y.G.), Canadian Institutes of Health Research Grant CIHR-MOP-69083 (to H.Y.G. and R.G.T.) and Grant MOP-36499 (to N.E.D. and H.Y.G.) and equipment grants from the Banting and Best Diabetes Centre (to H.Y.G. and R.G.T.) and CIHR (to H.Y.G.). Y.M.L. was supported by a fellowship award from the Canadian Diabetes Association in honor of the late Evelyn J. Parker.
Abbreviations: Cm, Membrane capacitance; GFP, green fluorescent protein; HVA, high-voltage-activated; I-V, current-voltage relationship; KATP channel, ATP-sensitive K+ channel; KV channel, voltage-gated K+ channel; LVA, low-voltage-activated; MIP, mouse insulin promoter; pF, picofarad; PIP2, phosphatidylinositol-4,5-bisphosphate; TEA, tetraethylammonium; TTX, tetrodotoxin; VDCC, voltage-dependent Ca2+ channel; VH, holding potential.
References
Williams RH, Ensinck JW 1966 Secretion, fates and actions of insulin and related products. Diabetes 15:623–654
Unger RH 1985 Glucagon physiology and pathophysiology in the light of new advances. Diabetologia 28:574–578
Baetens D, Malaisse-Lagae F, Perrelet A, Orci L 1979 Endocrine pancreas: three-dimensional reconstruction shows two types of islets of Langerhans. Science 206:1323–1325
Van De Winkel M, Pipeleers D 1983 Autofluorescence-activated cell sorting of pancreatic islet cells: purification of insulin-containing B-cells according to glucose-induced changes in cellular redox state. Biochem Biophys Res Commun 114:835–842
Barg S, Galvanovskis J, Gopel SO, Rorsman P, Eliasson L 2000 Tight coupling between electrical activity and exocytosis in mouse glucagon-secreting -cells. Diabetes 49:1500–1510
Kanno T, Gopel SO, Rorsman P, Wakui M 2002 Cellular function in multicellular system for hormone-secretion: electrophysiological aspect of studies on -, - and -cells of the pancreatic islet. Neurosci Res 42:79–90
Gopel SO, Kanno T, Barg S, Rorsman P 2000 Patch-clamp characterisation of somatostatin-secreting-cells in intact mouse pancreatic islets. J Physiol. 528(Pt 3):497–507
Gopel S, Kanno T, Barg S, Galvanovskis J, Rorsman P 1999 Voltage-gated and resting membrane currents recorded from B-cells in intact mouse pancreatic islets. J Physiol. 521(Pt 3):717–728
Gopel SO, Kanno T, Barg S, Weng XG, Gromada J, Rorsman P 2000 Regulation of glucagon release in mouse-cells by KATP channels and inactivation of TTX-sensitive Na+ channels. J Physiol. 528(Pt 3):509–520
Hara M, Wang X, Kawamura T, Bindokas VP. Dizon R, Alcoser S, Magnuson MA, Bell GI2003 Transgenic mice with green fluorescent protein-labeled pancreatic -cells. Am J Physiol 284:E177–E183
Leung YM, Kang YH, Gao, X, Xia F, Xie H, Sheu L, Tsuk S, Lotan I, Tsushima R, Gaisano HY 2003 Syntaxin 1A binds to the cytoplasmic C-terminus of Kv2.1 to regulate channel gating and trafficking. J Biol Chem 278:17532–17538
Gopel S, Zhang Q, Eliasson L, Ma XS, Galvanovskis J, Kanno T, Salehi A, Rorsman P 2004 Capacitance measurements of exocytosis in mouse pancreatic -, - and -cells within intact islets of Langerhans. J Physiol. 556(Pt 3):711–726
Ma X, Zhang Y, Gromada J, Sewing S, Berggren PO, Buschard K, Salehi A, Vikman J, Rorsman P, Eliasson L 2005 Glucagon stimulates exocytosis in mouse and rat pancreatic -cells by binding to glucagons receptors. Mol Endocrinol 19:198–212
Bokvist K, Olsen HL, Hoy M, Gotfredsen CF, Holmes WF, Buschard K, Rorsman P, Gromada J 1999 Characterisation of sulphonylurea and ATP-regulated K+ channels in rat pancreatic A-cells. Pflugers Arch 438:428–436
Gromada J, Ma X, Hoy M, Bokvist K, Salehi A, Berggren PO, Rorsman P 2004 ATP-sensitive K+ channel-dependent regulation of glucagon release and electrical activity by glucose in wild-type and SUR1–/– mouse -cells. Diabetes 53(Suppl 3):S181–S189
Franklin I, Gromada J, Gjinovci A, Theander S, Wollheim CB 2005 -Cell secretory products activate -cell ATP-dependent potassium channels to inhibit glucagon release. Diabetes 54:1808–1815
Baukrowitz T, Schulte U, Oliver D, Herlitze S, Krauter T, Tucker SJ, Ruppersberg JP, Fakler B 1998 PIP2 and PIP as determinants for ATP inhibition of KATP channels. Science 282:1141–1144
Shyng SL, Nichols CG 1998 Membrane phospholipid control of nucleotide sensitivity of KATP channels. Science 282:1138–1141
Pipeleers DG 1992 Heterogeneity in pancreatic -cell population. Diabetes 41:777–781
Chan CB, MacPhail RM, Sheu L, Wheeler MB, Gaisano HY 1999 -Cell hypertrophy in fa/fa rats is associated with basal glucose hypersensitivity and reduced SNARE protein expression. Diabetes 48:997–1005
Schrader LA, Anderson AE, Mayne A, Pfaffinger PJ, Sweatt JD 2002 PKA modulation of Kv4.2-encoded A-type potassium channels requires formation of a supramolecular complex. J Neurosci 22:10123–10133
Hoffman DA, Johnston D 1998 Downregulation of transient K+ channels in dendrites of hippocampal CA1 pyramidal neurons by activation of PKA and PKC. J Neurosci 18:3521–3528
Wang L, Bhattacharjee A, Fu J, Li M 1996 Abnormally expressed low-voltage-activated calcium channels in -cells from NOD mice and a related clonal cell line. Diabetes 45:1678–1683
Gelling RW, Du XQ, Dichmann DS, Romer J, Huang H, Cui L, Obici S, Tang B, Holst JJ, Fledelius C, Johansen PB, Rossetti L, Jelicks LA, Serup P, Nishimura E, Charron MJ 2003 Lower blood glucose, hyperglucagonemia, and pancreatic cell hyperplasia in glucagon receptor knockout mice. Proc Natl Acad Sci USA 100:1438–1443
Jing X, Li DQ, Olofsson CS, Salehi A, Surve VV, Caballero J, Ivarsson R, Lundquist I, Pereverzev A, Schneider T, Rorsman P, Renstrom E 2005 CaV2.3 calcium channels control second-phase insulin release. J Clin Invest 115:146–154
Meyer K, Irminger JC, Moss LG, de Vargas LM, Oberholzer J, Bosco D, Morel P, Halban PA 1998 Sorting human -cells consequent to targeted expression of green fluorescent protein. Diabetes 47:1974–1977
Detimary P, Dejonghe S, Ling Z, Pipeleers D, Schuit F, Henquin JC 1998 The changes in adenine nucleotides measured in glucose-stimulated rodent islets occur in cells but not in cells and are also observed in human islets. J Biol Chem 273:33905–33908
Ravier MA, Rutter GA 2005 Glucose or insulin, but not zinc ions, inhibit glucagon secretion from mouse pancreatic -cells. Diabetes 54:1789–1797
Liu YJ, Vieira E, Gylfe E 2004 A store-operated mechanism determines the activity of the electrically excitable glucagon-secreting pancreatic -cell. Cell Calcium 35:357–365
Hara M, Yin D, Dizon RF, Shen J, Chong AS, Bindokas VP 2004 A mouse model for studying intrahepatic islet transplantation. Transplantation 78:615–618(Yuk M. Leung, Ishtiaq Ahm)
Department of Medicine (M.H.), University of Chicago, Chicago, Illinois 60637
Abstract
We recently reported a transgenic [mouse insulin promoter (MIP)-green fluorescent protein (GFP)] mouse in which GFP expression is targeted to the pancreatic islet -cells to enable convenient identification of -cells as green cells. The GFP-expressing -cells of the MIP-GFP mouse were functionally indistinguishable from -cells of normal mice. Here we characterized the ionic channel properties and exocytosis of MIP-GFP mouse islet - and -cells. -Cells displayed delayed rectifying K+ and high-voltage-activated Ca2+ channels and exhibited Na+ currents only at hyperpolarized holding potential. -Cells were nongreen and had both A-type and delayed rectifier K+ channels, both low-voltage-activated and high-voltage-activated Ca2+ channels, and displayed Na+ currents readily at –70 mV holding potential. -Cells had ATP-sensitive K+ channel (KATP) channel density as high as that in -cells, and, surprisingly, -cell KATP channels were more sensitive to ATP inhibition (IC50 = 0.16 ± 0.03 mM) than -cell KATP channels (IC50 = 0.86 ± 0.10 mM). Whereas -cells were rather uniform in size [2–4.5 picofarad (pF)], -cells varied vastly in size (2–12 pF). Of note, small -cells (<4.5 pF) showed little exocytosis, whereas medium -cells (5–8 pF) exhibited vigorous exocytosis, but large -cells (>8 pF) had weaker exocytosis. We found no correlation between -cell size and their Ca2+ channel density, suggesting that Ca2+ influx may not be the cause of the heterogeneity in exocytotic responses. The MIP-GFP mouse therefore offers potential to further explore the functional heterogeneity in -cells of different sizes. The MIP-GFP mouse islet is therefore a reliable model to efficiently examine -cell and -cell physiology and should greatly facilitate examination of their pathophysiology when the MIP-GFP mice are crossed with diabetic models.
Introduction
EACH PANCREATIC ISLET cell type, through their secretion (insulin, glucagon, somatostatin), contributes to glucose homeostasis (1, 2). Whereas the more abundant -cells (70% of islet cells) are easier to examine, -cells (20%), - and F-cells (the remaining 10%) are difficult to identify (3). This limitation has resulted in the relatively slow advances in determining the biology of non--cells, particularly the -cell, whose dysfunction is also a major contributing factor to the abnormal glucose homeostasis in diabetes (2).
A number of strategies have been used to identify each islet cell type. -Cells can be isolated from other non--cells by autofluorescence-activated cell sorting, which can yield up to 80–90% pure -cell and -cell populations (4). Cell size has been used to distinguish -cells from - and -cells because -cells are on average larger [6 picofarad (pF)] than - and -cells (3 pF) (5). Therefore, islet cells 6 pF or larger are likely to be -cells (6). Functionally, mouse -, -, and -cells possess distinct sets of ionic channels (5, 6, 7, 8, 9). -Cells have delayed rectifying K+ channels and high-voltage-activated (HVA) Ca2+ channels and exhibit voltage-gated Na+ currents only at hyperpolarized holding potential (VH). This is in contrast to -cells, which have both A-type K+ channels and delayed rectifier K+ channels, both low-voltage-activated (LVA) and HVA Ca2+ channels, and have voltage-gated Na+ currents readily at –70 mV VH. -Cells can be distinguished from -cells by lacking A-type K+ channels and T-type Ca2+ channels. This electrophysiological fingerprinting provides a reliable way to distinguish islet cell types but are nonetheless cumbersome. Furthermore, once diabetes sets in, these ion channel properties of the islet cells may become perturbed, which would make it very difficult, if not impossible, to determine the precise pathophysiological contribution of each of the islet cell to the abnormal glucose homeostasis in diabetes.
To surmount these limitations, we created a transgenic mouse that has green fluorescent protein (GFP) specifically tagged to the -cells (10), thus providing an instant visual discrimination between -cells (green) and non--cells (nongreen). These mouse insulin promoter (MIP)-GFP mice develop normally, exhibit normal glucose-tolerance and insulin secretion, and the islet -cells are morphologically and functionally similar to control -cells (10). Here we followed up our initial report by examining single islet cell function of the MIP-GFP mouse, including ion channel activities and exocytosis, by electrophysiology. Our results on the ion channel properties of the MIP-GFP mouse islet -cells and -cells are similar to previous reports on the electrophysiological descriptions of these cell types (5, 6, 7, 8, 9), which indicate this MIP-GFP model to be an excellent and valid model. Importantly, in the process of characterizing these islet cells of the MIP-GFP mice, we revealed some novel insights. First, whereas - and -cells had comparable KATP channel densities, -cell KATP channels had a 5-fold higher sensitivity to ATP inhibition than -cell KATP channels. Second, there was a large variation in the size of the green -cells ranging from 2–12 pF, which remarkably exhibited very different exocytotic capacity, suggesting the possibility of distinct subpopulations. In contrast, the nongreen non--cells, which include - and -cells, were of uniformly small size (2–4.5 pF). The MIP-GFP mouse islet cell preparation is therefore a powerful strategy to reliably and efficiently study the normal biology of not only -cells but also - and -cells. Future studies directed at crossing the MIP-GFP mice with diabetic models will further reveal the contribution of each of these islet cell types to the abnormal glucose homeostasis.
Materials and Methods
Materials
Antibodies against insulin and glucagon were purchased from Dako Cytomation (Glostrup, Denmark) and Sigma (St. Louis, MO), respectively. Tetraethylammonium-Cl (TEA), 4-aminopyridine, nifedipine, glibenclamide, and Mg-ATP were from Sigma. Tetrodotoxin (TTX) was from Alomone Labs (Jerusalem, Israel). Phosphatidylinositol-4,5-bisphosphate (PIP2) and U73122 were purchased from EMD Biosciences (San Diego, CA). PIP2 was dissolved in pipette solution [for ATP-sensitive K+ channel (KATP) current measurement] at 1 mM, sonicated for 15 min, and then used on the same day of experiment.
Islet isolation
Mouse pancreatic islets were isolated by collagenase digestion as described previously (11). Islets were dispersed to single cells with 0.05% trypsin (Sigma) in Ca2+- and Mg2+-free PBS. Islet cells were plated on glass coverslips in 35-mm dishes and cultured in RPMI 1640 medium containing 1 mM pyruvate, 11 mM glucose, 0.2% NaHCO3, 10% fetal bovine serum, 10 mM HEPES, 100 U/ml penicillin G sodium, and 100 μg/ml streptomycin sulfate (Invitrogen Canada Inc., Burlington, Ontario, Canada). Islet cells were cultured overnight before electrophysiological recordings.
Recording of Kv and Ca2+ currents
Mouse islet cells were voltage clamped in the whole-cell configuration using an EPC-9 amplifier and Pulse software (HEKA Electronik, Lambrecht, Germany) as we previously described (11). Pipette tip resistances ranged from 3 to 5 M when filled with intracellular solutions. The intracellular solution for Kv current measurement contained (in mM): 140 KCl, 1 MgCl2, 1 EGTA, 10 HEPES, and 5 MgATP (pH 7.25 adjusted with KOH). The intracellular solution for Ca2+ current measurement contained (in mM): 120 CsCl, 20 TEA-Cl, 1 MgCl2, 1 EGTA, 10 HEPES, and 5 MgATP (pH 7.25 adjusted with CsOH). The bath solution for Kv current measurements contained (in mM): 140 NaCl; 4 KCl, 1 MgCl2, 2 CaCl2, 2 D-glucose, and 10 HEPES (pH 7.3 adjusted with NaOH). The bath solution for Ca2+ current measurement was the same as above but also supplemented with 20 mM TEA-Cl and 10 μM TTX, and the concentration of CaCl2 was increased to 10 mM. After a whole-cell configuration was established, the cell was held at –70 mV or other indicated VH and subject to various experimental protocols as detailed in Results and figure legends. All experiments were performed at room temperature (22 C). Data for steady-state inactivation were fit by the Boltzmann equation: I/Imax = 1/{1 + exp [(V-V1/2)/k]}, where V1/2 is the half-maximal inactivation potential and k the slope factor.
Recording of KATP currents
The intracellular solution for measurement of KATP currents contained (in mM): 140 KCl, 1 MgCl2, 1 EGTA, 10 HEPES, and various concentrations of MgATP as indicated (pH 7.25 adjusted with KOH). Each individual cell was tested with a single concentration of ATP dialyzed through the recording pipette. The bath solution for KATP current measurements contained (in mM): 140 NaCl, 4 KCl, 1 MgCl2, 2 CaCl2, 2 D-glucose, and 10 HEPES (pH 7.4 adjusted with NaOH). In a typical recording, the -cells are identified as green cells (10), and the -cells are identified by being nongreen and having A-type K+ currents (5). After a whole-cell configuration was established in -cells, the cell was held at –80 mV, and a test pulse of –30 mV (500 ms) was given immediately to test the presence of A-type K+ currents. Subsequently the cell was then stimulated by a –140 mV hyperpolarizing voltage step (500 msec) every 10 sec. Once KATP currents reached maximum, the cell was subject to a series of voltage pulses from –140 to –40 mV (500 msec) at 20 mV increments to obtain the current-voltage relationship (I-V). All experiments were performed at room temperature (22 C). Concentration-response curves for ATP inhibition of KATP channels are fit by the Hill equation:
where I is the current in the presence of a certain pipette (ATP), Imax is the maximum current, [ATP] is the concentration of ATP inside the pipette, Kd is the apparent dissociation constant, and n is the Hill coefficient.
Membrane capacitance measurement
Recording electrodes were coated with orthodontic wax (Butler, Guelph, Ontario, Canada) close to the tips and heat polished. Resistances ranged from 3 to 5 M when pipettes were filled with the intracellular solution, which contained (in mM): 125 K-glutamate, 10 KCl, 10 NaCl, 1 MgCl2, 5 HEPES, 0.05 EGTA, 0.1 cAMP, and 4 MgATP (pH to 7.1 by KOH). The extracellular solution contained (in mM): 140 NaCl, 4 KCl, 1 MgCl2, 2 CaCl2, 5 D-glucose, and 10 HEPES (pH 7.3 adjusted with NaOH). Membrane capacitance (Cm) was estimated by the Lindau-Neher technique, implementing the Sine + DC feature of the Lock-in module (40 mV peak to peak and a frequency of 500 Hz) in the standard whole-cell configuration. Recordings were conducted using an EPC9 patch clamp amplifier and Pulse software. Exocytotic events were elicited by a train of eight 500-msec depolarizing pulses (1-Hz stimulation frequency) from –70 to 0 mV. All Cm measurements were performed at 28 C.
Confocal immunofluorescence microscopy
Laser confocal immunofluorescence microscopy was performed as previously described (11). Dispersed pancreatic islet cells were plated on glass coverslips coated with 0.01% poly-L-lysine. After 2 h, the cells were fixed in 2% formaldehyde for 30 min, then treated with 5% goat serum and 0.1% saponin for 1 h, and finally immunolabeled with either mouse monoclonal antiglucagon (1:2000) or guinea pig monoclonal antiinsulin (1:100) for 2 h. The coverslips were rinsed with 0.1% saponin in PBS and then incubated with appropriate fluorescent-labeled secondary antibodies (either Texas Red or rhodamine red) for 1 h. After rinsing once more, coverslips were mounted on slides in a fading retarder (0.1% p-phenylenediamine in glycerol) and examined using a laser-scanning confocal imaging system (LSM 510, Zeiss, Oberkochen, Germany).
Statistics
Results are presented as means ± SEM. ANOVA was used for multiple-group comparisons, and statistical significance was determined by Student-Newman-Keuls test. P < 0.05 was considered statistically significant.
Results
Visualization of green -cells and nongreen islet cells of MIP-GFP mice
Characterization of voltage-gated K+ (Kv) and Na+ currents of cells in intact islets
Whole islets were used in these experiments. It is often difficult to distinguish the nongreen cells from the green cells in large islets because the green fluorescence of the GFP-containing -cells in a deeper layer could emit through the transparent nongreen cells on the islet surface. We therefore chose smaller islets and rolled them around until at some angle we saw exclusively green cells or exclusively nongreen cells, and then we recorded one of them.
When the membrane VH was at –70 mV, depolarizing the -cell triggered outward K+ currents but not inward Na+ currents (Fig. 3A; inset shows the expanded very early time frame; note absence of Na+ currents). The K+ currents were sensitive to TEA. When VH was at –120 mV, depolarizing the -cell triggered both TEA-sensitive K+ currents and Na+ currents (Fig. 3B and inset). The latter were resistant to TEA block. We then showed that these -cell Na+ currents could be completely abolished by 10 μM TTX, which expectedly did not affect -cell Ca2+ currents (Fig. 4). Steady-state inactivation experiments showed that the voltage at which half of the -cell Na+ channels were inactivated (V1/2) was –100.4 ± 1.9 mV (Fig. 3D). In Fig. 3C, nongreen -cells could be distinguished functionally from -cells by displaying both Na+ currents and TEA-sensitive K+ currents upon depolarization at a physiological VH (–70 mV). Accordingly, V1/2 of -cell Na+ channel inactivation was –46.8 ± 5.3 mV (Fig. 3D). Another distinctive feature of -cells was the rapidly inactivating TEA-insensitive A-type K+ currents (Fig. 3C, asterisk). Addition of 100 μM 4-aminopyridine abolished most of the A-type K+ currents (not shown).
Characterization of Kv and Na+ currents of single cells
As indicated above, it is often difficult to distinguish green and nongreen cells in intact islets. Furthermore, the ion channel properties could be influenced by the confounding paracrine effects of secreted hormones from neighboring cells within an intact islet. We therefore carried out the rest of our characterization studies in dispersed single islet cells (Fig. 5). When VH was at –70 mV, depolarizing the -cell triggered TEA-sensitive outward K+ currents but not inward Na+ currents (Fig. 5A and inset). However, when VH was at –120 mV, depolarizing the -cell triggered both K+ currents and Na+ currents (Fig. 5B and inset). V1/2 for Na+ channel inactivation was –98 ± 2.8 mV (Fig. 5E). Single dispersed -cells displayed TEA-sensitive K+ currents but not Na+ currents upon depolarization (VH = –70 mV). The reason for this discrepancy is unclear, but the loss of Na+ channels in the dispersed -cells but not -cells may in part be due to the fact that all -cells are located on the islet surface and therefore might be more vulnerable to enzymatic and mechanical stresses during islet isolation and dispersion. -Cells are located in the islet core and might be less susceptible. Nevertheless, single -cells could still be distinguished functionally from other islet cell types by displaying A-type K+ currents at physiological VH (–70 mV) (Fig. 5C, asterisk). The latter became larger at hyperpolarized VH (–120 mV) (Fig. 5D). Accordingly, V1/2 for A-type K+ channel inactivation was –75.7 ± 0.3 mV (Fig. 5F). We found that about 65% of dispersed nongreen cells exhibited A-type K+ currents and were therefore -cells (7, 9); the remaining presumably were -cells, which we did not further characterize.
Characterization of single -cell exocytosis and voltage-dependent Ca2+ channels (VDCC)
In our characterization of -cell Na+ channels and Kv channels so far, we studied -cells of medium size range (4–8 pF) because they were the most numerous (Fig. 2). With the MIP-GFP mouse islet cell preparation, we became increasingly aware that the -cells were of a large range of sizes (Figs. 1D and 2). We then asked whether there was a functional heterogeneity in -cells of different sizes by examining their exocytotic capacity. When small single -cells (3–4.5 pF) were stimulated by a train of eight depolarizations, there were only very small changes in Cm, indicating that small -cells were poor secretors (Fig. 6A, left). -Cells of medium size range (5–8 pF) showed vigorous exocytosis (Fig. 6A, middle). Large -cells (>8 pF; Fig. 6A, right) did not exhibit larger exocytosis than the medium-size -cells but instead showed generally smaller responses. When Cm increases (Fig. 6B, left) and normalized Cm increases (Fig. 6B, right) are plotted against cell sizes, there is a bell-shaped relationship, in which it becomes obvious that medium -cells were the most vigorous secretors. Exocytosis by small cells (10.7 ± 0.7 fF/pF, n = 4) was much weaker (P < 0.05) than that by medium-size cells (76.4 ± 18.9 fF/pF, n = 8). Exocytosis by large cells (24.6 ± 8.9 fF/pF, n = 4) was generally much weaker than that by medium-size cells.
Because the opening of VDCC provides the Ca2+ influx to trigger exocytosis, we examined the VDCC in single -cells to see whether a possible difference of VDCC property might be the underlying cause of the heterogeneity in exocytotic capacity observed between the different size -cells. Figure 7A shows the Ca2+ currents in -cells of different sizes. Figure 7B shows the I-V relationship of the Ca2+ currents triggered in three groups of -cells based on different sizes. The small increase in current density with cell size is not statistically significant. When peak current density of each individual cell was plotted against cell size, there was no significant correlation (Fig. 7C; r = 0.40; P = 0.14). Therefore, the difference in exocytotic capacity in -cells of different sizes may not be due to differences in VDCC density per se. Previous reports showed that mouse -cells possess only HVA Ca2+ channels (i.e. L-, P/Q-, and R-type VDCC) (8). Consistently we observed that in -cells of any size group, the activation threshold of the VDCC was from –30 to –20 mV (Fig. 7B), indicating the presence of only HVA Ca2+ channels but not LVA Ca2+ channels. LVA Ca2+ channels may be inactivated by VH of –70 mV or greater. However, when VH was –90 mV, activation threshold of any -cell was still from –30 to –20 mV (data not shown), ruling out the possibility of presence of LVA Ca2+ channels in -cells. Nifedipine (10 μM, selective L-type VDCC blocker) inhibited 64 ± 4 and 58 ± 6% (n = 3; P < 0.05) of Ca2+ currents in small and medium -cells, respectively, indicating the majority of HVA Ca2+ channels were L-type.
Characterization of single -cell exocytosis and VDCC
We next examined the VDCC in -cells, which have been shown to possess both LVA and HVA Ca2+ channels (5, 9). As shown in Fig. 8, A and B, when the -cell was held at –100 mV and depolarized to different potentials in the presence of 10 μM TTX, Ca2+ currents could readily be triggered at –40 mV and were transient (fast activation and inactivation) in nature, characteristic of LVA Ca2+ channels or the T-type VDCC (Fig. 8A, left). Ca2+ currents triggered at stronger depolarization (–10 or 0 mV) began to show, in addition to the transient current, a sustained component, indicative of HVA Ca2+ channels. When VH was at –50 mV, which would readily inactivate LVA Ca2+ channels, depolarization expectedly only triggered the slow-inactivating HVA Ca2+ currents (Fig. 8A, middle). The difference between the Ca2+ currents obtained at the two VH yields the LVA Ca2+ currents (Fig. 8A, right). The results are summarized in the I-V curves shown in Fig. 8B (n = 4). Thus, the total currents (VH at –100 mV) were the sum of HVA (VH at –50 mV) and LVA Ca2+ currents (the difference currents). The HVA and LVA Ca2+ currents typically peaked at +10 mV and –20 mV, respectively. We then examined -cell exocytosis. Figure 8C shows a representative -cell exocytotic response (275 ± 35 fF at eighth stimulation, n = 3), which was within the range previously reported (5, 12, 13).
KATP channel sensitivity to ATP block was higher in - than -cells
KATP channels are present in not only -cells but also -cells (5, 14, 15, 16). By dialyzing 0.3 mM ATP into mouse -cells, Barg et al. (5) reported a low KATP channel density. Using whole-cell configuration, we here reassessed the -cell KATP current density by dialyzing different concentrations of ATP into each cell via the pipette (each individual cell was dialyzed with only one concentration of ATP). When a low concentration of ATP (0.05 mM) was dialyzed into the -cell, large currents were indeed triggered (Fig. 9Ai). The currents were sensitive to bath application of 1 μM glibenclamide (Fig. 9Aii), indicating that these are indeed KATP channel currents. Pipette ATP concentration-dependently inhibited the channel so that 5 mM ATP completely blocked the channel (Fig. 9Aiii–v). Some -cells did not develop any KATP currents at all when dialyzed with 1 mM ATP (Fig. 9Aiv, lower trace). When we dialyzed a low concentration of ATP (0.05 mM) into -cells, we also detected large glybenclamide-sensitive currents (Fig. 9Bi and Bii). Pipette ATP inhibited the channel in a concentration-dependent manner such that 5 mM ATP completely blocked the channel. (Fig. 9Biii–v). Of note, 0.3 mM ATP, which triggers relatively small currents in -cells (Fig. 9Aiii), was in fact the maximal concentration triggering -cell KATP currents (Fig. 9Biii). When KATP channel densities of -cells and -cells are plotted against different concentrations of ATP (Fig. 9C), we observed that KATP channel densities in - and -cells were in fact comparable. The high KATP channel density was, however, not apparent in -cells at 0.3 mM ATP, which would be in agreement with the findings by Barg et al. (5). If normalized current densities are plotted against ATP concentrations, it is clear that -cell KATP channels were significantly (P < 0.05) more sensitive (IC50 = 0.16 ± 0.03 mM) to ATP inhibition than -cell KATP channels (IC50 = 0.86 ± 0.10 mM) (Fig. 9D).
What is the mechanism for the higher sensitivity of the -cell KATP channels to ATP PIP2 has been known to open KATP channel by decreasing the channel sensitivity to ATP (17, 18). We therefore questioned whether there may be a lack of PIP2 in the -cells to explain the heightened ATP sensitivity of the -cell KATP channels. If so, increasing PIP2 concentration in -cells by either direct introduction of exogenous PIP2 or inhibition of phospholipase C (by U73122) to increase endogenous PIP2 would reduce the KATP channel sensitivity, thereby increasing the KATP current density. At 0.3 mM ATP, -cell KATP channels were only slightly opened by only approximately 30% (see Fig. 9D). Here we dialyzed 100 μM PIP2 into the -cells in the presence of 0.3 mM intracellular ATP, but PIP2 failed to increase KATP channel opening (Fig. 9E). Consistently 10 μM U73122 inhibition of phospholipase C also failed to increase -cell KATP channel density in the presence of 0.3 mM intracellular ATP (Fig. 9E). These results indicate that the heightened ATP sensitivity of the -cell KATP channels is unlikely a consequence of a possible lower abundance of PIP2 at the -cell membrane. The lack of sensitivity of -cell KATP channels to PIP2 was surprising, and further work will therefore be required to determine the factor(s) responsible for the enhanced sensitivity of -cell KATP channels to ATP.
Discussion
The MIP-GFP transgenic mouse was created by the expression of GFP under MIP to specifically tag the islet -cells so that the -cells can be easily identified as green cells (10). We had already shown that the MIP-GFP mice develop normally and exhibit normal glucose homeostasis comparable with their control littermates (10). -Cells from MIP-GFP and control mice show similar Ca2+ signal response to glucose (10). Here we showed that MIP-GFP -cells exhibited exocytotic responses and ion channel properties (Kv channels, Na+ channels and VDCC) qualitatively and quantitatively similar to those of nontransgenic mice previously described (5, 6, 7, 8, 9). Therefore, MIP-GFP mouse -cells offer a reliable and efficient strategy to further examine novel aspects of -cell biology and electrophysiology.
In previous studies of -cells, large cells were deliberately chosen to avoid mixing with the smaller islets cells that may be - or -cells (5, 6, 7, 8, 9). Such studies might not have considered the biology of small and medium-size -cell subpopulations. With the MIP-GFP mouse islet cell preparation, examination of -cells of different sizes was made possible and indeed led to the realization that -cells could be variable in size ranging from 2 to 12 pF. We confirmed that green -cells were on average (5.5 pF) larger than -cells (2.8 pF) and any cell greater than 4.5 pF must be green -cells. We here reported a novel finding that the large variation in -cell sizes was associated with functional heterogeneity: small -cells were very poor secretors; -cells of medium-sizes exhibited vigorous secretion, whereas large -cells (>8 pF) had reduced exocytosis. This did not appear to be due to Ca2+ channel defects in smaller or larger cells because VDCC density remained relatively constant over the different cell sizes. Moreover, regardless of sizes, all -cells did not have LVA and only possessed HVA, the majority being L type.
Interestingly, -cell subpopulations have been reported to have different sensitivity to glucose stimulation (19). -Cells are known to undergo hypertrophy in type 2 diabetes as is the case in Zucker fa/fa rats, and insulin secreted from these large -cells was dysfunctional (20). It is therefore possible that in diabetes, a redistribution to a higher proportion of larger -cells may contribute to an overall poorer insulin secretory capacity and response to the increased demand. The precise mechanism to explain the heterogeneity in exocytotic capacity of the different subpopulations of islet -cells in health and in diabetes remained to be further explored. Of interest, the levels of exocytotic proteins (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) in the Zucker fa/fa rat islets were much lower than normal rat islets (20), which raises the possibility that perhaps these subpopulations of -cells in the MIP-GFP mouse might also exhibit different levels of exocytotic proteins, which would in part explain the differences in exocytotic capacity.
With the MIP-GFP islet cell preparation, - and -cells can be easily distinguished from small -cells by being nongreen. The -cell can then be distinguished from -cells (and -cells) by displaying A-type K+ channels (Figs. 3 and 5) and T-type Ca2+ channels (Fig. 8) (9). These channels have low activation threshold (A-type K+ channels, –60 mV; T-type Ca2+ channels, –40 mV) and rapid activation and inactivation rate. Whereas T-type Ca2+ channels perform the pacemaker function, A-type K+ channels provide repolarizng currents (6). -Cells, -cells, and -cells may also be distinguished by Na+ channel-gating properties (5, 6, 7, 8, 9). V1/2 values of steady-state inactivation of mouse -, - and -cell Na+ channel are –100, –47, and –28 mV, respectively (5, 6, 7, 8, 9) (Fig. 3). These values are far apart enough to distinguish between the mouse islet cells.
The differentiation between - and -cells using these functional markers (A-type K+ channels and T-type Ca2+ channels in -cells) discussed above may become elusive under pharmacological manipulation or during pathological states, particularly diabetes. For instance, activation of protein kinase A and protein kinase C has been shown to down-regulate A-type K+ channels in neurons or expressed in Xenopus oocytes (21, 22). Whereas T-type Ca2+ channels are not present in control mouse -cells, they have been reported to express in -cells of nonobese diabetic mice (23). Therefore, electrophysiological fingerprinting alone does not appear to be sufficient for accurate identification of islet cell type under certain situations, particularly diabetes. Tagging of GFP to the -cells therefore provides an independent yet complementary tool for convenient -cell and non--cell (predominantly -cells) identification, even during disease states.
Future developments of other transgenic animals having GFP-tagged -cells by crossing diabetic mice models with the MIP-GFP mice, or having GFP (or other fluorophore) genetically tagged to - or -cells, would be very welcome by the research community to then be able to reliably determine the contribution of each islet cell to the abnormal glucose homeostasis in diabetes. For example, MIP-GFP mice could be crossbred with gene knockouts or other transgenic animals, which exhibit selective islet cell (-cell) hyperplasia and dysfunction (24) or altered channel expression (15, 25). An alternative to transgenic technology is infection of islet cells with replication-defective recombinant adenovirus expressing GFP under insulin promoter control and has been successful in yielding more than 95% pure human -cells (26). However, there may be untoward toxic effects of the virus on the -cell channel properties, which would be confounding to those effects being examined and caused by the disease states (i.e. diabetes, glucolipotoxicity).
Another novel finding here is that -cell KATP channels had a 5-fold higher sensitivity to ATP inhibition than -cell KATP channels. This difference did not appear to be due to a possible lower abundance of PIP2 in the -cell plasma membrane because maneuvers that raised plasma membrane PIP2 concentration in -cells either exogenously or endogenously failed to increase -cell KATP channel opening. What additional information can we learn from the ATP concentration-inhibition curves (particularly the physiologically relevant range, 1–5 mM) Raising ATP from 1 to 5 mM increased the inhibition of -cell KATP currents from 65 to 100% and would expectedly depolarize the cell. By contrast, 1 mM ATP almost completely (93%) blocked -cell KATP currents. It has been estimated that ATP concentration inside the -cell is already higher than 1 mM at low glucose (27), suggesting that the fraction of closed KATP channels in -cells may actually exceed 93%. Because high glucose does raise ATP concentration in -cells (28), a minority of -cells may be expected to be depolarized by high glucose after a meal. This is consistent with the observation by Liu et al. (29) in which the KATP channel blocker tolbutamide depolarized only a minority of -cells. All these observations are also in agreement with the demonstration that high glucose causes only a very mild depolarization in -cells, which nonetheless, may be able to inactivate T-type Ca2+ channels and Na+ channels (15, 16).
In summary, this work on the electrophysiological characterization of ion channels and exocytosis of - and -cells of MIP-GFP mouse islets showed similar results as previous electrophysiological descriptions of mouse islet cells (5, 6, 7, 8, 9) and therefore fully validated the MIP-GFP mouse as an excellent model to greatly facilitate the examination of islet cell biology. Of note, the size-dependent functional heterogeneity among -cells may have important implication on furthering our understanding of islet secretory dysfunction in diabetes and may even be applicable to islet selection during islet transplantation (30). Further work will be required to examine these distinct subpopulations of -cells in health and diabetes and also determine the PIP2-independent factor that enhances the -cell KATP channel sensitivity to ATP. The groundwork we have done here on the MIP-GFP mouse islet cell preparation, including some novel insights we have gained in - and -cell biology, should serve the community in using this model as a powerful and highly reliable tool to further islet cell studies.
Footnotes
This work was supported by Juvenile Diabetes Research Foundation Grants 1-2005-1112 (to H.Y.G.), Canadian Institutes of Health Research Grant CIHR-MOP-69083 (to H.Y.G. and R.G.T.) and Grant MOP-36499 (to N.E.D. and H.Y.G.) and equipment grants from the Banting and Best Diabetes Centre (to H.Y.G. and R.G.T.) and CIHR (to H.Y.G.). Y.M.L. was supported by a fellowship award from the Canadian Diabetes Association in honor of the late Evelyn J. Parker.
Abbreviations: Cm, Membrane capacitance; GFP, green fluorescent protein; HVA, high-voltage-activated; I-V, current-voltage relationship; KATP channel, ATP-sensitive K+ channel; KV channel, voltage-gated K+ channel; LVA, low-voltage-activated; MIP, mouse insulin promoter; pF, picofarad; PIP2, phosphatidylinositol-4,5-bisphosphate; TEA, tetraethylammonium; TTX, tetrodotoxin; VDCC, voltage-dependent Ca2+ channel; VH, holding potential.
References
Williams RH, Ensinck JW 1966 Secretion, fates and actions of insulin and related products. Diabetes 15:623–654
Unger RH 1985 Glucagon physiology and pathophysiology in the light of new advances. Diabetologia 28:574–578
Baetens D, Malaisse-Lagae F, Perrelet A, Orci L 1979 Endocrine pancreas: three-dimensional reconstruction shows two types of islets of Langerhans. Science 206:1323–1325
Van De Winkel M, Pipeleers D 1983 Autofluorescence-activated cell sorting of pancreatic islet cells: purification of insulin-containing B-cells according to glucose-induced changes in cellular redox state. Biochem Biophys Res Commun 114:835–842
Barg S, Galvanovskis J, Gopel SO, Rorsman P, Eliasson L 2000 Tight coupling between electrical activity and exocytosis in mouse glucagon-secreting -cells. Diabetes 49:1500–1510
Kanno T, Gopel SO, Rorsman P, Wakui M 2002 Cellular function in multicellular system for hormone-secretion: electrophysiological aspect of studies on -, - and -cells of the pancreatic islet. Neurosci Res 42:79–90
Gopel SO, Kanno T, Barg S, Rorsman P 2000 Patch-clamp characterisation of somatostatin-secreting-cells in intact mouse pancreatic islets. J Physiol. 528(Pt 3):497–507
Gopel S, Kanno T, Barg S, Galvanovskis J, Rorsman P 1999 Voltage-gated and resting membrane currents recorded from B-cells in intact mouse pancreatic islets. J Physiol. 521(Pt 3):717–728
Gopel SO, Kanno T, Barg S, Weng XG, Gromada J, Rorsman P 2000 Regulation of glucagon release in mouse-cells by KATP channels and inactivation of TTX-sensitive Na+ channels. J Physiol. 528(Pt 3):509–520
Hara M, Wang X, Kawamura T, Bindokas VP. Dizon R, Alcoser S, Magnuson MA, Bell GI2003 Transgenic mice with green fluorescent protein-labeled pancreatic -cells. Am J Physiol 284:E177–E183
Leung YM, Kang YH, Gao, X, Xia F, Xie H, Sheu L, Tsuk S, Lotan I, Tsushima R, Gaisano HY 2003 Syntaxin 1A binds to the cytoplasmic C-terminus of Kv2.1 to regulate channel gating and trafficking. J Biol Chem 278:17532–17538
Gopel S, Zhang Q, Eliasson L, Ma XS, Galvanovskis J, Kanno T, Salehi A, Rorsman P 2004 Capacitance measurements of exocytosis in mouse pancreatic -, - and -cells within intact islets of Langerhans. J Physiol. 556(Pt 3):711–726
Ma X, Zhang Y, Gromada J, Sewing S, Berggren PO, Buschard K, Salehi A, Vikman J, Rorsman P, Eliasson L 2005 Glucagon stimulates exocytosis in mouse and rat pancreatic -cells by binding to glucagons receptors. Mol Endocrinol 19:198–212
Bokvist K, Olsen HL, Hoy M, Gotfredsen CF, Holmes WF, Buschard K, Rorsman P, Gromada J 1999 Characterisation of sulphonylurea and ATP-regulated K+ channels in rat pancreatic A-cells. Pflugers Arch 438:428–436
Gromada J, Ma X, Hoy M, Bokvist K, Salehi A, Berggren PO, Rorsman P 2004 ATP-sensitive K+ channel-dependent regulation of glucagon release and electrical activity by glucose in wild-type and SUR1–/– mouse -cells. Diabetes 53(Suppl 3):S181–S189
Franklin I, Gromada J, Gjinovci A, Theander S, Wollheim CB 2005 -Cell secretory products activate -cell ATP-dependent potassium channels to inhibit glucagon release. Diabetes 54:1808–1815
Baukrowitz T, Schulte U, Oliver D, Herlitze S, Krauter T, Tucker SJ, Ruppersberg JP, Fakler B 1998 PIP2 and PIP as determinants for ATP inhibition of KATP channels. Science 282:1141–1144
Shyng SL, Nichols CG 1998 Membrane phospholipid control of nucleotide sensitivity of KATP channels. Science 282:1138–1141
Pipeleers DG 1992 Heterogeneity in pancreatic -cell population. Diabetes 41:777–781
Chan CB, MacPhail RM, Sheu L, Wheeler MB, Gaisano HY 1999 -Cell hypertrophy in fa/fa rats is associated with basal glucose hypersensitivity and reduced SNARE protein expression. Diabetes 48:997–1005
Schrader LA, Anderson AE, Mayne A, Pfaffinger PJ, Sweatt JD 2002 PKA modulation of Kv4.2-encoded A-type potassium channels requires formation of a supramolecular complex. J Neurosci 22:10123–10133
Hoffman DA, Johnston D 1998 Downregulation of transient K+ channels in dendrites of hippocampal CA1 pyramidal neurons by activation of PKA and PKC. J Neurosci 18:3521–3528
Wang L, Bhattacharjee A, Fu J, Li M 1996 Abnormally expressed low-voltage-activated calcium channels in -cells from NOD mice and a related clonal cell line. Diabetes 45:1678–1683
Gelling RW, Du XQ, Dichmann DS, Romer J, Huang H, Cui L, Obici S, Tang B, Holst JJ, Fledelius C, Johansen PB, Rossetti L, Jelicks LA, Serup P, Nishimura E, Charron MJ 2003 Lower blood glucose, hyperglucagonemia, and pancreatic cell hyperplasia in glucagon receptor knockout mice. Proc Natl Acad Sci USA 100:1438–1443
Jing X, Li DQ, Olofsson CS, Salehi A, Surve VV, Caballero J, Ivarsson R, Lundquist I, Pereverzev A, Schneider T, Rorsman P, Renstrom E 2005 CaV2.3 calcium channels control second-phase insulin release. J Clin Invest 115:146–154
Meyer K, Irminger JC, Moss LG, de Vargas LM, Oberholzer J, Bosco D, Morel P, Halban PA 1998 Sorting human -cells consequent to targeted expression of green fluorescent protein. Diabetes 47:1974–1977
Detimary P, Dejonghe S, Ling Z, Pipeleers D, Schuit F, Henquin JC 1998 The changes in adenine nucleotides measured in glucose-stimulated rodent islets occur in cells but not in cells and are also observed in human islets. J Biol Chem 273:33905–33908
Ravier MA, Rutter GA 2005 Glucose or insulin, but not zinc ions, inhibit glucagon secretion from mouse pancreatic -cells. Diabetes 54:1789–1797
Liu YJ, Vieira E, Gylfe E 2004 A store-operated mechanism determines the activity of the electrically excitable glucagon-secreting pancreatic -cell. Cell Calcium 35:357–365
Hara M, Yin D, Dizon RF, Shen J, Chong AS, Bindokas VP 2004 A mouse model for studying intrahepatic islet transplantation. Transplantation 78:615–618(Yuk M. Leung, Ishtiaq Ahm)