CaV2.3 calcium channels control second-phase insulin release
http://www.100md.com
《临床调查学报》
1Diabetes Programme at Lund University, Lund, Sweden.
2Institute of Neurophysiology and Center of Molecular Medicine Cologne, University of Cologne, Cologne, Germany.
3Diabetes Research Laboratories, Oxford Centre for Diabetes, Endocrinology and Metabolism, the Churchill Hospital, Oxford, United Kingdom.
Abstract
Concerted activation of different voltage-gated Ca2+ channel isoforms may determine the kinetics of insulin release from pancreatic islets. Here we have elucidated the role of R-type CaV2.3 channels in that process. A 20% reduction in glucose-evoked insulin secretion was observed in CaV2.3-knockout (CaV2.3–/–) islets, close to the 17% inhibition by the R-type blocker SNX482 but much less than the 77% inhibition produced by the L-type Ca2+ channel antagonist isradipine. Dynamic insulin-release measurements revealed that genetic or pharmacological CaV2.3 ablation strongly suppressed second-phase secretion, whereas first-phase secretion was unaffected, a result also observed in vivo. Suppression of the second phase coincided with an 18% reduction in oscillatory Ca2+ signaling and a 25% reduction in granule recruitment after completion of the initial exocytotic burst in single CaV2.3–/– ? cells. CaV2.3 ablation also impaired glucose-mediated suppression of glucagon secretion in isolated islets (27% versus 58% in WT), an effect associated with coexpression of insulin and glucagon in a fraction of the islet cells in the CaV2.3–/– mouse. We propose a specific role for CaV2.3 Ca2+ channels in second-phase insulin release, that of mediating the Ca2+ entry needed for replenishment of the releasable pool of granules as well as islet cell differentiation.
See the related Commentary beginning on page 16
Introduction
Systemic glucose tolerance is orchestrated by the regulated release of insulin and glucagon from the ? and cells of the pancreatic islets of Langerhans. The and ? cells are electrically excitable and use electrical signals to couple changes in blood glucose concentration to stimulation or inhibition of hormone release. In both cell types, influx of extracellular Ca2+ through voltage-gated Ca2+ channels with resultant elevation of intracellular Ca2+ concentration ([Ca2+]i) triggers exocytosis of the hormone-containing secretory granules. Like other electrically excitable cells, both and ? cells contain several types of voltage-gated Ca2+ channel (1, 2). Assigning physiological functions to the respective Ca2+ channels is central to the understanding of electrical and secretory activities in these cells.
Voltage-gated Ca2+ channels are divided into 3 subfamilies: (a) L-type high voltage–activated (HVA) Ca2+ channel family that comprises the CaV1.1, 1.2, 1.3, and 1.4 channels and is inhibited by dihydropyridines (DHPs) (1, 3, 4); (b) non–L-type HVA channels CaV2.1 (P/Q-type), 2.2 (N-type), and 2.3 (R-type) that are sensitive to -agatoxin IVA and -conotoxin GVIA and SNX482, respectively (1, 4, 5); and (c) the low voltage–activated (LVA) T-type Ca2+ channel family (CaV3.1, 3.2, and 3.3). The latter subtype differs electrophysiologically from the HVA Ca2+ channels in opening transiently already upon modest depolarization (6, 7) and fulfilling important roles in pacemaker cells (8).
The chain of events that couples an elevation in blood glucose to initiation of ? cell electrical activity is well established and involves facilitated transport of the sugar into the ? cell and its subsequent metabolic degradation by glycolysis and mitochondrial oxidation, resulting in closure of the ATP-sensitive K+ channels (KATP channels) and ? cell depolarization, with resultant activation of voltage-gated Ca2+ channels and regulated insulin exocytosis (9). Glucose-stimulated insulin secretion consists of a rapid first-phase of insulin secretion that lasts for approximately 10 minutes before declining to near-basal levels, followed by less prominent but sustained second-phase insulin secretion that can last for several hours (10, 11). The cellular mechanisms underlying biphasic insulin release remain unclear, but consensus exists that an elevation in [Ca2+]i is required for both first- and second-phase insulin secretion (12). Type 2 diabetes is associated with a shift from biphasic to monophasic insulin release (10), and it is therefore important to establish the cell biology of insulin release kinetics.
In mouse pancreatic ? cells, 50% of the whole cell Ca2+ current exhibits properties typical for L-type channels, is inhibited by DHP channel blockers such as isradipine or nifedipine, and is activated by BayK8644. The molecular identity of the ? cell L-type Ca2+ channel involved in insulin secretion has been debated, but recent studies have established that CaV1.2 or 1C channels play a decisive role (13, 14).
In mouse ? cells, there is evidence suggesting that the secretory granules and the L-type Ca2+ channels assemble into a tight functional complex (13, 15-17). Thanks to this organization, the ? cells are capable of exocytosis at rates comparable to those encountered in chromaffin cells, although the Ca2+ channel density is only 5–10% of that in the latter cell type. This arrangement bears strong resemblance to the tight coupling of P/Q-type CaV2.1 as well as N-type CaV2.2 channels to rapid synaptic transmission (18, 19). No such interaction has been demonstrated for the R-type Cav2.3 channels. By contrast, recent evidence in neurons suggests that R-type channels are physically detached from the exocytotic machinery (20) and are not involved in rapid neurotransmission in mossy fibers (21).
About one-quarter of the ? cell whole-cell Ca2+ current is sensitive to the R-type Ca2+ channel blocker SNX482, but the role of R-type Ca2+ channels in insulin secretion remains elusive. A general CaV2.3-knockout (CaV2.3–/–) mouse has been established. It exhibits a relatively modest neurological phenotype, including altered pain responses (22), impaired spatial memory (23), and enhanced fear reaction (24). Pancreatic islets express the endocrine splice variant of CaV2.3 (25), and previous investigations of glucose homeostasis in CaV2.3–/– mice have demonstrated a slight glucose intolerance and reduced glucose-induced insulin secretion (26). Here we have extended these initial observations by performing in vivo glucose tolerance tests, dynamic measurements of phasic insulin secretion in situ, static pancreatic hormone–release experiments in isolated islets, as well as single cell recordings of whole-cell Ca2+ currents and exocytosis and ratiometric measurements of the cytoplasmic Ca2+ concentration in WT and CaV2.3–/– islets. We demonstrate that whereas R-type Ca2+ channels play a minor role in rapid insulin release, their significance becomes more apparent during second-phase secretion. These data point to an emerging picture where Ca2+ influx through different ? cell Ca2+ channels play distinct functional roles.
Results
Whole cell Ca2+ currents in and? cells from CaV2.3+/+ and CaV2.3–/– mice.
Integrated Ca2+ current versus voltage (Q-V) relations were measured in dissociated single pancreatic islet cells using the perforated-patch whole cell configuration. Insulin-releasing ? cells were identified by the absence of voltage-gated Na+ currents at membrane potentials relevant for ? cell electrical activity, that is, between the ? cell resting membrane potential (–70 mV) and the peak of the ? cell action potential (–10 mV; Figure 1A, inset). In ? cells from CaV2.3–/– mice, the integrated Ca2+ currents observed at membrane potentials less than or equal to –10 mV were reduced compared with WT cells. At –10 mV, the reduction averaged approximately 23% (P < 0.05; Figure 1B). The CaV2.3–/– mice exhibited a selective loss of a high-voltage Ca2+ current component, and current amplitudes at voltages below –10 mV were not affected. The R-type Ca2+ channel blocker SNX482 (100 nM) had no effect on voltage-gated Ca2+ currents in CaV2.3–/– mice, whereas the L-type inhibitor isradipine (2 μM) significantly reduced Ca2+ influx approximately 60% at potentials greater than or equal to –30 mV (P < 0.05; Figure 1, C and D).
Cells exhibiting Na+ currents when holding at –70 mV (Figure 1E, inset) were classified as non–? cells and likely represent glucagon-releasing cells (27). The frequency of cells was much lower in dispersed islet cells from CaV2.3–/– mice than cells made from WT islets. Counting all cells that could be functionally defined as belonging to either group, only approximately 7% of the CaV2.3–/– cells could be classified as cells versus approximately 27% of the WT cells. In 4 cells from CaV2.3–/– mice with the electrophysiological properties expected for cells, the Ca2+ current amplitude was not different from that observed in WT cells (Figure 1, E and F).
Single cell exocytosis in CaV2.3–/– and CaV2.3+/+ mice.
Ca2+-elicited exocytosis in ? cells was monitored as increases in whole-cell membrane capacitance and was elicited by trains of 10 500-ms voltage clamp depolarizations from the holding potential –70 mV to 0 applied at 1 Hz (Figure 2, A and B). In ? cells from CaV2.3–/– mice, the increase in cell capacitance during the train averaged 392 ± 47 femtofarads (fF) (n = 7). The latter value is 21% less than the exocytotic response evoked by the same stimulus in WT mice (P < 0.05; 496 ± 42 fF; n = 6). Interestingly, the early component of exocytosis (in response to the first depolarization) was not affected and averaged 55 ± 19 fF and 57 ± 22 fF in WT and CaV2.3–/– ? cells, respectively. Instead, selective suppression of the late component of exocytosis was observed. The 23% reduction of the whole cell Ca2+ current observed in the CaV2.3–/– ? cells corresponded nicely with the overall reduction in exocytosis, suggesting that exocytotic capacity as such was not affected. This idea was reinforced by the observation that exocytotic rates, measured during intracellular dialysis of a Ca2+-containing pipette solution using the standard whole-cell configuration, were identical in ? cells from both mouse strains (Figure 2, C and D). In the few cells that could be identified in CaV2.3–/– mice, exocytosis was not different from that observed in WT, either when elicited by train depolarizations (n = 3) or Ca2+ buffer dialysis (n = 3; Figure 2, E–H).
Effects of CaV2.3 ablation on intracellular Ca2+ homeostasis.
Insulin secretion is initiated by changes in the submembrane cytoplasmic free [Ca2+]i (28, 29), which are determined by ? cell electrical activity. We therefore monitored [Ca2+]i under basal conditions as well as after the addition of glucose, the KATP channel blocker tolbutamide, or following depolarization with high extracellular K+ (Figure 3). Resting [Ca2+]i measured at 5 mM glucose was identical in islets from CaV2.3–/– and WT mice and averaged approximately 90 nM. In WT islets, elevation of extracellular glucose from 5 to 15 mM evoked an initial robust elevation in [Ca2+]i after 204 ± 33 seconds (n = 7). This initial peak decayed slowly (over 4–10 minutes) toward baseline and was eventually followed by a pattern of repetitive [Ca2+]i oscillations. Subsequent depolarization by the nonmetabolizable stimuli tolbutamide and high K+ in the continued presence of glucose elicited a rapid elevation of [Ca2+]i but did not result in oscillatory activity. In CaV2.3–/– islets, the glucose-evoked initial peak in [Ca2+]i occurred 227 ± 15 seconds (n = 9) after elevating the glucose concentration. This initial amplitude was almost unaffected (7% decrease), but the time-averaged [Ca2+]i during the subsequent oscillatory phase was 17% lower than in WT islets (P < 0.01). In addition, the oscillatory activity was 29% slower in the CaV2.3–/– islets (1.8 ± 0.1 versus 2.5 ± 0.2 bursts/minute in CaV2.3–/– and WT islets, respectively; P < 0.01). By contrast, the peaks in [Ca2+]i induced by depolarization with tolbutamide or high K+ were, if anything, slightly augmented in the CaV2.3–/– islets (NS versus control). The finding that no [Ca2+]i oscillations were observed at 5 mM glucose and that elevating the glucose concentration to 15 mM increased Ca2+ signaling suggests that the signal principally reflects the behavior of the pancreatic ? cells (see ref. 30).
In vivo glucose tolerance and pancreatic hormone release in CaV2.3–/– mice.
In vivo glucose homeostasis was investigated by intraperitoneal glucose challenges (2 g/kg body weight; Table 1). Basal (nonfasted) plasma glucose levels averaged 9.8 ± 0.6 mmol/l (n = 8) in WT CaV2.3+/+ mice and increased 67% and 135% 3 and 8 minutes after the glucose load, respectively. In CaV2.3–/– mice, basal glucose levels were elevated by 17% compared with WT (P < 0.05 versus WT; n = 9), and the plasma glucose concentrations measured 3 and 8 minutes after challenge were also elevated approximately 15% (P < 0.05 for values at 8 minutes versus WT). Basal plasma insulin levels did not differ between the CaV2.3–/– and CaV2.3+/+ strains, although the plasma glucose concentration was somewhat higher in the former mice. In WT mice, glucose increased plasma insulin concentrations 43% and 52% 3 and 8 minutes after the challenge, respectively. In CaV2.3–/– mice, the glucose-induced increase in circulating insulin was similar to that in WT mice 3 minutes after challenge, but after 8 minutes it was limited to less than 10% (P < 0.05 versus WT). These findings reinforce previous observations indicating an impaired glucose tolerance of the CaV2.3–/– mice (26).
We also monitored plasma glucagon concentrations during the glucose challenge. Basal glucagon levels in CaV2.3–/– mice were modestly decreased (approximately 10%) compared with CaV2.3+/+ littermates. In WT mice, the glucose load reduced circulating glucagon 15% already at 3 minutes after challenge, and an additional 8% decrease was observed at 8 minutes. In CaV2.3–/– mice, the inhibition of glucagon release was sluggish. At 3 minutes after challenge, plasma glucagon concentrations actually increased 18% (P < 0.05 versus basal; n = 9).
Insulin and glucagon release in vitro in pancreatic islets.
Pancreatic islets in situ receive extensive neuronal input from parasympathetic nerve endings (31). Since CaV2.3 Ca2+ channels are also expressed in neuronal tissue, their ablation might affect pancreatic hormone release in vivo by indirect mechanisms and not by directly affecting and ? cell function. Insulin and glucagon secretion was therefore also investigated in isolated islets. The first set of experiments investigated the effects of L-type and R-type channel blockers isradipine and SNX482, respectively, in WT CaV2.3+/+ islets (Table 2). Basal insulin release (1 mM glucose) was low and was unaffected by either channel blocker. Elevation of extracellular glucose to 20 mM stimulated insulin release more than 15-fold. Isradipine (2 μM) suppressed glucose-stimulated insulin release 77%, whereas SNX482 (100 nM) reduced insulin secretion 17% (Table 2). The latter value agrees favorably with the inhibition observed in capacitance measurements (see Figure 2, A and B). Glucagon release was measured under the same conditions. At 1 mM glucose glucagon release was high and remained unchanged by either isradipine or SNX482. Elevating the glucose concentration (20 mM) reduced glucagon secretion 70% (P < 0.001; low versus high glucose). At high glucose, blockade of R-type channels by SNX482 failed to affect glucagon release, whereas isradipine stimulated glucagon secretion under the same condition.
We next compared the effects of the Ca2+ channel blockers with the consequences of CaV2.3 gene ablation on pancreatic hormone release (Table 3). In WT CaV2.3+/+ islets, elevating glucose from 1 to 20 mM again stimulated insulin release more than 15-fold. In CaV2.3–/– islets, glucose-stimulated insulin release was reduced approximately 25% compared with WT. The effects of isradipine on glucose-stimulated insulin release were comparable in WT (66% reduction) and in CaV2.3–/– islets (68% suppression). Insulin secretion elicited by stimulation with high extracellular K+ (50 mM) resulted in a 75% enhancement of release relative to that observed in the presence of glucose alone in WT islets. A similar relative stimulation was observed in CaV2.3–/– islets (96%), but in absolute terms, the response in the latter type of islets was reduced 13% compared with WT islets. Surprisingly, whereas elevating glucose from 1 to 20 mM suppressed glucagon release 58% in WT CaV2.3+/+ islets (P < 0.001; low versus high glucose), the inhibitory action of the sugar was severely impaired in the CaV2.3–/– islets and amounted to a mere 27% (P < 0.01 versus WT). Isradipine exerted divergent actions in WT and CaV2.3–/– islets. As discussed above, in WT islets the L-type Ca2+ channel antagonist stimulated glucagon release 52% when applied in the presence of 20 mM glucose. By contrast, isradipine reduced glucagon secretion in CaV2.3–/– islets 32% under high-glucose conditions. Increasing extracellular K+ enhanced glucagon secretion massively, approximately 320% in WT and 220% in CaV2.3–/– islets.
Phasic insulin release measured by in situ pancreatic perfusion.
To assess the role of CaV2.3 in dynamic insulin release, we performed in situ pancreatic perfusions with fractionated sampling. When WT CaV2.3+/+ pancreata were perfused (Figure 4A) with a low-glucose solution (3.3 mM), insulin release was barely detectable. After increasing the glucose concentration to 16.7 mM (at t = 11 minutes), first-phase insulin release was initiated with a 2-minute delay. The peak in first-phase insulin secretion was attained 2 minutes after onset of release (t = 15 minutes) and measured 8.8 ± 1.7 ng/ml (n = 4). Insulin release then exhibited a transient nadir phase (t = 18–21 minutes) during which release rates averaged 4–5 ng/ml, before accelerating again during second-phase insulin secretion to approximately 18 ng/ml at t = 36 minutes and later. When the same experiment was repeated in CaV2.3–/– pancreata (Figure 4B), the peak in first-phase insulin secretion was only slightly reduced (19%) and measured 7.1 ± 2.1 ng/ml (NS; n = 4). More importantly, second-phase insulin release was markedly suppressed and averaged 9.4 ± 2.7 ng/ml at t = 40 minutes, representing a 46% reduction compared with WT (P < 0.05).
Genetic ablation of Cav2.3 may result in compensatory mechanisms resulting in rearrangements of the Ca2+ channels in the ? cell itself or neighboring and cells. In addition, we wanted to verify that the observed effects of CaV2.3 on insulin secretion are not limited to the background C57B mouse strain. Similar experiments were therefore made in pancreata from standard inbred NMRI mice, but instead CaV2.3 channel function was inhibited by SNX482. In the absence of the channel inhibitor, first-phase insulin secretion was initiated with a 2-minute delay and peaked 1 minute, 30 seconds later when it measured 15.3 ± 2.5 ng/ml (n = 6). Insulin release then decayed toward a lower plateau level of approximately 8 ng/ml and increased slightly to approximately 10 ng/ml during the second phase (Figure 4C). When SNX482 (100 nM) was included in the perfusate (Figure 4D), first-phase insulin release was largely unaffected, and peak values averaged 13.0 ± 3.7 ng/ml (n = 6). By contrast, second-phase insulin secretion was reduced more than 80% (e.g., 1.8 ± 0.6 ng/ml versus 10.0 ± 0.6 ng/ml at t = 40 minutes in the presence or absence of SNX482, respectively; P < 0.01). It was verified that the glucose concentration in the effluent medium was identical in all experiments.
Insulin and glucagon immunoreactivity in WT and CaV2.3–/– islet cells.
Islet cells were finally dispersed and investigated by confocal immunocytochemistry (Figure 5). In WT islet cells, 80% of the dispersed cells revealed immunoreactivity for insulin and approximately 20% stained positive for glucagons in double-labeling experiments. The relative frequency of and ? cells suggested by this analysis compares favorably with the electrophysiological data. Only 1 out of 230 cells investigated revealed immunoreactivity for both insulin and glucagons. A rather different picture emerged in the CaV2.3–/– islet cell preparation. Whereas 75% and 9% of the cells could readily be characterized as ? and cells, respectively, 16% of the 210 cells coexpressed insulin and glucagons.
Discussion
Insulin-producing ? cells contain multiple types of Ca2+ channel. Whereas the role of L-type Ca2+ channels in insulin secretion is amply documented, that of the non–L-type Ca2+ channels is less well understood. The advent of subtype-specific Ca2+ channel blockers in combination with the generation of transgenic knockout mice provides a unique opportunity for an in-depth analysis of the function(s) fulfilled by the different Ca2+ channel subtypes. Here we have used SNX482 and CaV2.3-null mice to study the significance of R-type CaV2.3 Ca2+ channels in islet insulin and glucagon secretion in vivo and in vitro.
The role of CaV2.3 channels for? cell [Ca2+]i homeostasis and insulin secretion.
Here we demonstrate that pharmacological inhibition of R-type CaV2.3 Ca2+ channels using SNX482 does not affect first-phase insulin secretion, but reduces second-phase release a dramatic 80%. A similar preferential effect on late-phase insulin secretion is observed in CaV2.3–/– mice (Figure 4). At the single ? cell level, close inspection of the capacitance recordings (Figure 2, A and B) reveals that the initial component of exocytosis is not much affected in CaV2.3–/– ? cells. In fact, the overall reduction results exclusively from suppression of the late component of exocytosis (elicited by the third and subsequent depolarizations). We have previously demonstrated that opening of L-type CaV1.2 Ca2+ channels is tightly associated with rapid exocytosis and first-phase secretion. Thus, it appears that Ca2+ entry via L-type CaV1.2 and R-type CaV2.3 channels have distinct intracellular effects.
CaV2.3 ablation is associated with a 23% decrease in ? cell Ca2+ current in CaV2.3–/– mice (Figure 1). This is in good agreement with the response to acute application of SNX482 in WT mice (13), suggesting that there is little compensatory upregulation of other Ca2+ channels in the knockout mice. We acknowledge that CaV2.3 channel activity may represent only part of the R-type current component (24, 32). In addition, other reports suggest that all R-type current components are not blocked by SNX482 with the same efficacy (33). The observation that SNX482 fails to affect whole cell Ca2+ currents in CaV2.3–/– ? cells (Figure 1, C and D), however, indicates at least that SNX482, at the concentration used here (100 nM), does not affect any Ca2+ channels other than CaV2.3, but we cannot exclude the contribution of an SNX482-insensitive R-type Ca2+ current component. Indeed, approximately 40% of the whole cell Ca2+ current is unaffected by either SNX482 or the L-type blocker isradipine in CaV2.3–/– ? cells. Although most of this current probably represents P/Q-type Ca2+ currents as previously shown for WT ? cells (13), it is possible that part of it reflects a SNX482-resistant portion of the R-type Ca2+ current.
How do CaV2.3 Ca2+ channels regulate second-phase insulin secretion?
Secretory granules in pancreatic ? cells can be classified according to their release competence (34). A limited (1–5%) pool is immediately available for rapid exocytosis upon stimulation (readily releasable pool, RRP). Once this pool of granules has been released, exocytosis proceeds at a slower rate determined by the supply of granules newly mobilized from a much larger reserve pool (35). The observation that only a late component of exocytosis is affected in CaV2.3–/– ? cells (Figure 2, A and B) argues that CaV2.3 channels are not tightly coupled to exocytosis but are more important for the recruitment of new granules for release. We also demonstrate that genetic ablation of the CaV2.3 Ca2+ channels results in approximately 20% reduction of the glucose-induced steady-state time-averaged [Ca2+]i (more than 5 minutes after initial response; Figure 3). Can these 2 pieces of information be reconciled? We have reported previously that a moderate (40%) but protracted elevation of [Ca2+]i results in strong enhancement of the exocytotic capacity (36). It is of interest that this effect is mediated by a global rather than a localized increase in [Ca2+]i and is operational already at concentrations as low as 200–300 nM, that is, 10-fold lower than those required to elicit fast exocytosis. Thus, even moderate reduction of Ca2+ entry and [Ca2+]i following inhibition of CaV2.3 Ca2+ channels can have strong effects on insulin secretion in the longer term by inhibition of granule recruitment. Presumably, L-type Ca2+ channels still trigger granule exocytosis, but once the RRP is depleted, the supply of new granules for release becomes rate limiting. This would explain how ablation of the approximately 20% CaV2.3 Ca2+ channel component can result in approximately 50% (knockout) to 80% (SNX482) reduction of second-phase insulin secretion. Thus, it seems justifiable to conclude that Ca2+ entry via R-type CaV2.3 Ca2+ channels is functionally (and perhaps spatially) linked to granule mobilization and priming of insulin granules for release.
CaV2.3 ablation and glucagon secretion.
CaV2.3 ablation is associated with disturbances of glucagon secretion. In isolated islets, the ability of glucose to suppress glucagon secretion is severely impaired (Table 3). These abnormalities are also detectable at the systemic level. In the in vivo glucose challenge test, 3-minute values for plasma glucagon were stimulated approximately 15% in CaV2.3–/– mice, whereas in WT they were reduced approximately 20% (Table 1). The observations that whole cell Ca2+ currents (Figure 1, E and F) and single cell exocytosis (Figure 2, E–H) are unaffected in CaV2.3–/– cells, taken together with the finding that the R-type blocker SNX482 fails to affect glucagon release in WT islets (Table 2), make it unlikely that these effects are the direct consequence of the loss of CaV2.3 in the cell. Rather, the explanation for the perturbation of glucose-inhibited glucagon release in CaV2.3-ablated mice appears to be due to the appearance of atypical /? cells accounting for greater than 60% of the glucagon-expressing cells in these mice (Figure 5, B and C). Taken together with the scarcity of cells with cell electrophysiological properties in preparations from CaV2.3–/– mice (7% compared with 27% in WT mice), this suggests that the /? cells electrical behavior is more like that of the ? cells. This idea is supported by the results in Table 3, showing that whereas isradipine suppresses glucagon release in CaV2.3–/– islets, it has the opposite effect in WT islets. This observation also provides a clue to the relative failure of glucose to suppress glucagon release in CaV2.3–/– islets both in vitro and (acutely) in vivo. The fact that glucagon eventually decreases after a glucose challenge in vivo may be related to neuronal input or paracrine effects that are known to be of great importance for the control of the hyperglycemic hormone (37). Surprisingly, basal glucagon release (at low glucose) is unaffected in CaV2.3–/– islets. One possibility is that the reduced population of normal cells have a very large capacity for exocytosis, but the capacitance measurements (Figure 2, E–H) provide no support for this notion. Another explanation is that the /? cells exhibit poorly regulated exocytosis and constitutively release both glucagon and insulin. This idea is reinforced by the observation that basal insulin release in CaV2.3–/– islets is higher than in WT islets (0.4 ± 0.05 versus 0.2 ± 0.02 ng/islet/hour in CaV2.3–/– and WT, respectively; Table 3). Yet another possibility is that glucagon coreleased with insulin from the /? cells potentiates further secretion by a cAMP-dependent mechanism. We have reported elsewhere that glucagon increases ? cell exocytosis 5-fold (38). The significance of paracrine mechanisms within the islet is apparent from the strong stimulatory action of isradipine on glucagon secretion in WT islets (Tables 2 and 3). The latter effect may reflect the relief from paracrine inhibition of glucagon secretion exerted by Zn2+ and GABA cosecreted with insulin from the ? cells (39, 40). Preliminary data from our laboratory indicate that R-type CaV2.3 Ca2+ channels play a decisive role in somatostatin secretion from the cells (Q. Zhang, A. Salehi, E. Renstr?m, and P. Rorsman, unpublished observations). We can exclude that the loss of physiological glucose inhibition of glucagon secretion is secondary to relief from paracrine inhibition by somatostatin, however, since the dysregulation of glucagon secretion in CaV2.3–/– islets is not mimicked by SNX482.
Role of CaV2.3 in islet cell differentiation.
A surprising finding in the present study is that the majority of the glucagon immunoreactive cells in CaV2.3–/– islets coexpress insulin, indicating that CaV2.3 channels play a role in the development of mature cells. A transient peak in CaV2.3 channel expression in glial cells along specific CNS pathways has been demonstrated to coincide with postnatal myelinization of the white matter in the rat (41). It can be speculated that CaV2.3 channel expression exerts a similar action in defining the differentiated mature islet cell lineages, but the exact underlying mechanism remains to be established.
Pathophysiological implications.
Although R-type CaV2.3 channels appear to be of little importance for first-phase insulin secretion and their contribution becomes apparent only during late exocytosis/second-phase insulin secretion, this does not mean that they are unimportant for systemic glucose homeostasis. Indeed, the CaV2.3–/– mice exhibited basal hyperglycemia (Table 1). In humans, CaV2.3 is encoded by the CACNA1E gene, which is located on chromosome 1q25–31. Interestingly, a chromosomal region around 1q25 reveals linkage to type 2 diabetes in several independent studies in different populations. These include early onset type 2 diabetes in Pima Indians (42), Utah Caucasians (43), and English sib pairs (44). In addition, the 1q25–32 region demonstrates linkage with elevated blood glucose levels in the Framingham Heart Study (45). It is also pertinent that the above region confers defective insulin secretion in type 2 diabetes in Pima Indians (46). Type 2 diabetes is a multifactorial polygenic disorder. Clearly, polymorphisms in the CACNA1E gene alone are not sufficient to trigger disease. Given the present functional data, however, it is tempting to speculate that dysfunctional CaV2.3 Ca2+ channels may frequently be involved in creating the disturbed ? cell phenotype in type 2 diabetes.
Methods
Experimental animals.
As previously reported (26), the CACNA1E gene encoding Cav2.3 was disrupted in vivo by deleting a region containing exon 2 on mating Cav2.3fl/+ and deleter mice on a C57Bl/6 background that expresses Cre recombinase constitutively under the control of the CMV promoter. Cav2.3+/– mice containing 1 cre transgene were inbred, and pups with the CaV2.3–/– genotype were selected and transferred by embryo transfer into a SPF facility. The transfer included breeding with C57Bl/6 mice and resulted in heterozygous Cav2.3+/– mice. Only cre-negative pups were selected and inbred, yielding either CaV2.3+/+ or CaV2.3–/– mice. Thus, the CaV2.3+/+ and CaV2.3–/– mice used in this study have an identical genetic background. In Figure 4, inbred NMRI mice purchased from Charles River Wiga GmbH were used.
The mice were housed at a constant temperature (22–23°C) and 12-hour light cycles (6:00 a.m.–6:00 p.m.), with access to standard pellet food and water ad libitum. All experiments were evaluated and approved by the local ethical committee Malm?/Lund djurf?rs?ksetiska n?mnd, Lund District Court, Sweden.
Islet isolation and islet cell preparation.
The mice were sacrificed by cervical dislocation, and collagenase was administered into the pancreas by retrograde injection via the pancreatic duct. After 15–20 minutes’ incubation at 37°C, the islet suspension was washed 4 times with HBSS (4°C) before being manually collected. The islets were cultured in RPMI-1640 medium (Invitrogen Corp.) supplemented with 10% FCS, penicillin, and streptomycin. For the preparation of single cells, islets obtained by collagenase digestion were dissociated by incubation and gentle triculation in Ca2+-free medium. The resulting cell suspension was centrifuged, the supernatant discarded, and the pellet (containing the cells) was resuspended in RPMI-1640 medium, plated on plastic Nunc 35-mm Petri dishes, and maintained in tissue culture for up to 2 days.
Electrophysiology.
The measurements were conducted using an EPC-10 patch-clamp amplifier in conjunction with the PULSE software suite (version 8.53; HEKA Elektronik). Whole-cell Ca2+ currents were measured in intact cells using the perforated-patch whole-cell approach (Figure 1) using a pipette solution consisting of 76 mM Cs2SO4, 10 mM NaCl, 10 mM CsCl, 1 mM MgCl2, 5 mM HEPES (pH 7.35 with KOH), and 0.24 mg/ml amphotericin B (47). Exocytosis was monitored as increases in cell capacitance using the sine + DC mode of the lock-in amplifier included in the PULSE software and the standard whole cell configuration. When eliciting exocytosis by trains of ten 500-ms voltage clamp depolarizations (Figure 2, A and E), the pipette solution consisted of 125 mM Cs-glutamate, 10 mM CsCl, 10 mM NaCl, 1 mM MgCl2, 5 mM HEPES, 3 mM Mg-ATP, 0.1 mM cAMP, and 0.05 mM EGTA (pH 7.2 with CsOH). In Figure 2, C and G, this pipette solution was slightly modified to include 10 mM EGTA and 9 mM CaCl2, and all Cs+ salts were replaced by corresponding K+ salts. The resulting free intracellular Ca2+ concentration of this Ca2+/EGTA buffer was estimated to 1.5 μM using the binding constants of Martell and Smith (48, 49). The extracellular bath solution contained 138 mM NaCl, 5.6 mM KCl, 2.6 mM CaCl2, 1.2 mM MgCl2, 5 mM glucose, and 5 mM HEPES (pH 7.4 with NaOH). In the recordings of whole-cell Ca2+ currents (Figure 1) and depolarization-evoked exocytosis (Figure 2, A and E), 20 mM of NaCl was equimolarly replaced by the K+ channel blocker TEA-Cl to facilitate the separation of the small voltage-gated Ca2+ currents from the large outward K+ current. The DHP isradipine (Pfizer Inc.) was prepared as stock solution in DMSO (final concentration less than or equal to 0.1%). The R-type blocker SNX482 (Peptide Institute Inc.) was dissolved directly in the extracellular medium. All other reagents were from Sigma-Aldrich. Effects were determined in the steady state. The bath (approximately 1.5 ml) was continuously perfused (6 ml/min) and the temperature maintained at approximately 32°C.
In vivo glucose challenges.
For the in vivo studies, glucose (11.1 mmol [equivalent to 2 g]/kg body weight) was dissolved in 0.9% NaCl and delivered by intraperitoneal injection. Blood sampling, detection of plasma insulin by RIA, and enzymatic determination of plasma glucose concentrations were performed as described previously (50).
In situ/ex vivo pancreatic perfusion.
Experiments were performed in the morning at 10:00 a.m. in nonfasted mice. Anesthesia was given by intraperitoneal injection of midazolam (Hofmann-La Roche AG; 0.4 mg/25 g body weight) and fentanyl (Janssen Pharmaceuticals Inc.; 0.02 mg/25 g body weight). The experimental procedures were essentially identical to those described by Bonnevie-Nielsen et al. (51). Briefly, the mice were kept on a heating pad during the entire experiment. After opening the abdominal cavity and ligating the renal, hepatic, and splenic arteries, the aorta was tied off above the level of the pancreatic artery. The pancreas was perfused with modified Krebs-Ringer HEPES buffer preheated to 37°C (1 ml/min) via a silicone catheter placed in the aorta. The perfusate was collected via a silicone catheter from the portal vein at 30- or 60-second intervals, as indicated, in 2.5-ml Eppendorf tubes containing 25 μl Trasylol. Insulin and glucose concentrations in the effluent medium were detected by RIA and the glucose oxidase method, respectively.
In vitro pancreatic hormone release.
Insulin release in vitro was measured in static incubations. Briefly, freshly isolated islets were preincubated for 30 minutes at 37°C in a Krebs-Ringer bicarbonate buffer (pH 7.4) consisting of 120 mM NaCl, 25 mM NaHCO3, 4.7 mM KCl, 1.2 mM MgSO4, 2.5 mM CaCl2, 1.2 mM KH2PO, 1 mM glucose, and 10 mM HEPES (pH 7.4). The medium was gassed with 95% O2 and 5% CO2 to obtain constant pH and oxygenation. Groups of 10 islets were then incubated in 1 ml for 60 minutes at 37°C in Krebs-Ringer buffered solution supplemented with either glucose, the L-type Ca2+ channel blocker isradipine, the R-type Ca2+ channel inhibitor SNX482, tolbutamide, or high K+, as specified in the text and figures. Immediately after incubation, a 25-μl aliquot of the medium was removed for assay of insulin and glucagon radioimmunoassay as described previously (50).
Immunocytochemistry.
Insulin and glucagon immunoreactivities were visualized in dissociated islet cells by indirect immunocytochemistry using a Carl Zeiss AG 510 LSM confocal microscope and a x100 Plan-Apochromat x100/1.4 oil objective. After fixation with 3% paraformaldehyde and permeabilization by 0.1% Triton-X, the cells were incubated with normal donkey serum to reduce unspecific staining. The primary insulin and glucagon Abs, raised in guinea pigs and sheep, respectively, were incubated at 1:1,000. To prevent cross-talk between the channels, the secondary Cy3-conjugated anti–guinea pig and the Cy5-conjugated anti-sheep Abs (1:600) were excited in the multitrack mode using the 543-nm and 633-nm lines of the HeNe lasers, and emitted light was collected using greater than 560-nm and greater than 650-nm long-pass filters, respectively.
Microfluorimetry.
[Ca2+]i in intact pancreatic islets was measured by dual-wavelength microfluorimetry using fura-2 and a D104 PTI microfluorimetry system. The temperature of the experimental chamber was +32°C to allow comparison with the electrophysiological data. Procedures for loading and calibration of the fluorescence signal were as described previously (52).
Statistical analysis.
All data are given as means ± SEM. Statistical significance was evaluated using absolute values only. A paired Student’s t test was used when comparing responses in the same cell. For comparisons between groups with independent observations we used independent Student’s t tests or, when comparisons involved more than 2 groups, ANOVA.
Acknowledgments
We thank Kristina Borglid and Britt-Marie Nilsson for expert technical assistance. This work was supported by the Swedish Research Council (grant numbers 13509, 12239, and 6589), the European Foundation for the Study of Diabetes, European Community network grant "Neuronal Ca2+ Channels," the Albert P?hlsson Foundation, the NovoNordisk Foundation, the Center of Molecular Medicine Cologne (BMBF 01 KS 9502), and the Deutsche Forschungsgemeinschaft (SCH 387/9-1). P. Rorsman is a recipient of the G?ran Gustafsson Award for Research in the Natural Sciences and Medicine.
Footnotes
Nonstandard abbreviations used: [Ca2+]i, intracellular Ca2+ concentration; CaV2.3–/–, CaV2.3-knockout; DHP, dihydropyridine; fF, femtofarad(s); HVA, high voltage–activated; KATP, ATP-sensitive K+; LVA, low voltage–activated; RRP, readily releasable pool.
Conflict of interest: The authors have declared that no conflict of interest exists.
References
Catterall, W.A. 1998. . Structure and function of neuronal Ca2+ channels and their role in neurotransmitter release. Cell Calcium. 24::307-323.
D’Ascenzo, M. et al. 2004. . Electrophysiological and molecular evidence of L-(Cav1), N- (Cav2.2), and R- (Cav2.3) type Ca2+ channels in rat cortical astrocytes. Glia. 45::354-363.
Dolphin, A.C. 1999. . L-type calcium channel modulation. Adv. Second Messenger Phosphoprotein Res. 33::153-177.
Catterall, W.A. 2000. . Structure and regulation of voltage-gated Ca2+ channels. Annu. Rev. Cell Dev. Biol. 16::521-555.
Reid, C.A., Bekkers, J.M., and Clements, J.D. 2003. . Presynaptic Ca2+ channels: a functional patchwork. Trends Neurosci. 26::683-687.
Perez-Reyes, E. 2003. . Molecular physiology of low-voltage-activated T-type calcium channels. Physiol. Rev. 83::117-161.
Heady, T.N., Gomora, J.C., Macdonald, T.L., and Perez-Reyes, E. 2001. . Molecular pharmacology of T-type Ca2+ channels. Jpn. J. Pharmacol. 85::339-350.
Triggle, D.J. 1998. . The physiological and pharmacological significance of cardiovascular T-type, voltage-gated calcium channels. Am. J. Hypertens. 11::80S-87S.
Ashcroft, F.M., Proks, P., Smith, P.A., ?mm?l?, C., Bokvist, K., and Rorsman, P. 1994. . Stimulus-secretion coupling in pancreatic ? cells. J. Cell. Biochem. 55(Suppl.)::54-65.
Del Prato, S., Marchetti, P., and Bonadonna, R.C. 2002. . Phasic insulin release and metabolic regulation in type 2 diabetes. Diabetes. 51(Suppl. 1)::S109-S116.
Grodsky, G.M., and Bolaffi, J.L. 1992. . Desensitization of the insulin-secreting ? cell. J. Cell. Biochem. 48::3-11.
Henquin, J.C., Ravier, M.A., Nenquin, M., Jonas, J.C., and Gilon, P. 2003. . Hierarchy of the ? cell signals controlling insulin secretion. Eur. J. Clin. Invest. 33::742-750.
Schulla, V. et al. 2003. . Impaired insulin secretion and glucose tolerance in ? cell-selective Ca(v)1.2 Ca2+ channel null mice. EMBO J. 22::3844-3854.
Sinnegger-Brauns, M.J. et al. 2004. . Isoform-specific regulation of mood behavior and pancreatic ? cell and cardiovascular function by L-type Ca2+ channels. J. Clin. Invest. 113::1430-1439. doi:10.1172/JCI200420208.
Wiser, O. et al. 1999. . The voltage sensitive Lc-type Ca2+ channel is functionally coupled to the exo-cytotic machinery. Proc. Natl. Acad. Sci. U. S. A. 96::248-253.
Barg, S., Eliasson, L., Renstr?m, E., and Rorsman, P. 2002. . A subset of 50 secretory granules in close contact with L-type Ca2+ channels accounts for first-phase insulin secretion in mouse ? cells. Diabetes. 51(Suppl. 1)::S74-S82.
Barg, S. et al. 2001. . Fast exocytosis with few Ca(2+) channels in insulin-secreting mouse pancreatic B cells. Biophys. J. 81::3308-3323.
Mochida, S. et al. 2003. . Requirement for the synaptic protein interaction site for reconstitution of synaptic transmission by P/Q-type calcium channels. Proc. Natl. Acad. Sci. U. S. A. 100::2819-2824.
Mochida, S., Sheng, Z.H., Baker, C., Kobayashi, H., and Catterall, W.A. 1996. . Inhibition of neurotransmission by peptides containing the synaptic protein interaction site of N-type Ca2+ channels. Neuron. 17::781-788.
Wu, L.G., Borst, J.G., and Sakmann, B. 1998. . R-type Ca2+ currents evoke transmitter release at a rat central synapse. Proc. Natl. Acad. Sci. U. S. A. 95::4720-4725.
Dietrich, D. et al. 2003. . Functional specialization of presynaptic Cav2.3 Ca2+ channels. Neuron. 39::483-496.
Saegusa, H. et al. 2000. . Altered pain responses in mice lacking 1E subunit of the voltage-dependent Ca2+ channel. Proc. Natl. Acad. Sci. U. S. A. 97::6132-6137.
Kubota, M. et al. 2001. . Intact LTP and fear memory but impaired spatial memory in mice lacking Ca(v)2.3 ((IE)) channel. Biochem. Biophys. Res. Commun. 282::242-248.
Lee, S.C. et al. 2002. . Molecular basis of R-type calcium channels in central amygdala neurons of the mouse. Proc. Natl. Acad. Sci. U. S. A. 99::3276-3281.
Vajna, R. et al. 1998. . New isoform of the neuronal Ca2+ channel 1E subunit in islets of Langerhans and kidney—distribution of voltage-gated Ca2+ channel 1 subunits in cell lines and tissues. Eur. J. Biochem. 257::274-285.
Pereverzev, A. et al. 2002. . Disturbances in glucose-tolerance, insulin-release, and stress-induced hyperglycemia upon disruption of the Ca(v)2.3 ( 1E) subunit of voltage-gated Ca(2+) channels. Mol. Endocrinol. 16::884-895.
Barg, S., Galvanovskis, J., G?pel, S.O., Rorsman, P., and Eliasson, L. 2000. . Tight coupling between electrical activity and exocytosis in mouse glucagon-secreting cells. Diabetes. 49::1500-1510.
Rutter, G.A., Theler, J.M., Li, G., and Wollheim, C.B. 1994. . Ca2+ stores in insulin-secreting cells: lack of effect of cADP ribose. Cell Calcium. 16::71-80.
Bokvist, K., Eliasson, L., ?mm?l?, C., Renstr?m, E., and Rorsman, P. 1995. . Co-localization of L-type Ca2+ channels and insulin-containing secretory granules and its significance for the initiation of exocytosis in mouse pancreatic B cells. EMBO J. 14::50-57.
Nadal, A., Quesada, I., and Soria, B. 1999. . Homologous and heterologous asynchronicity between identified -, ?- and delta cells within intact islets of Langerhans in the mouse. J. Physiol. 517::85-93.
Satin, L.S. 2000. . Localized calcium influx in pancreatic ? cells: its significance for Ca2+-dependent insulin secretion from the islets of Langerhans. Endocrine. 13::251-262.
Wilson, S.M. et al. 2000. . The status of voltage-dependent calcium channels in 1E knock-out mice. J. Neurosci. 20::8566-8571.
Tottene, A., Volsen, S., and Pietrobon, D. 2000. . (1E) subunits form the pore of three cerebellar R-type calcium channels with different pharmacological and permeation properties. J. Neurosci. 20::171-178.
Neher, E. 1998. . Vesicle pools and Ca2+ microdomains: new tools for understanding their roles in neurotransmitter release. Neuron. 20::389-399.
Rorsman, P., and Renstr?m, E. 2003. . Insulin granule dynamics in pancreatic ? cells. Diabetologia. 46::1029-1045.
Gromada, J. et al. 1999. . CaM kinase II-dependent mobilization of secretory granules underlies acetylcholine-induced stimulation of exocytosis in mouse pancreatic B cells. J. Physiol. 518::745-759.
Nandi, J. et al. 2002. . Central mechanisms involved with catabolism. Curr. Opin. Clin. Nutr. Metab. Care. 5::407-418.
Gromada, J., Ding, W.G., Barg, S., Renstr?m, E., and Rorsman, P. 1997. . Multisite regulation of insulin secretion by cAMP-increasing agonists: evidence that glucagon-like peptide 1 and glucagon act via distinct receptors. Pflugers Arch. 434::515-524.
Ishihara, H., Maechler, P., Gjinovci, A., Herrera, P.L., and Wollheim, C.B. 2003. . Islet ? cell secretion determines glucagon release from neighbouring cells. Nat. Cell Biol. 5::330-335.
Wendt, A. et al. 2004. . Glucose inhibition of glucagon secretion from rat cells is mediated by GABA released from neighboring ? cells. Diabetes. 53::1038-1045.
Chen, S., Ren, Y.Q., Bing, R., and Hillman, D.E. 2000. . A 1E subunit of the R-type calcium channel is associated with myelinogenesis. J. Neurocytol. 29::719-728.
Hanson, R.L. et al. 1998. . An autosomal genomic scan for loci linked to type II diabetes mellitus and body-mass index in Pima Indians. Am. J. Hum. Genet. 63::1130-1138.
Elbein, S.C., Hoffman, M.D., Teng, K., Leppert, M.F., and Hasstedt, S.J. 1999. . A genome-wide search for type 2 diabetes susceptibility genes in Utah Caucasians. Diabetes. 48::1175-1182.
Wiltshire, S. et al. 2001. . A genomewide scan for loci predisposing to type 2 diabetes in a U.K. population (the Diabetes UK Warren 2 Repository): analysis of 573 pedigrees provides independent replication of a susceptibility locus on chromosome 1q. Am. J. Hum. Genet. 69::553-569.
Jun, G., Song, Y., Stein, C.M., and Iyengar, S.K. 2003. . An autosome-wide search using longitudinal data for loci linked to type 2 diabetes progression. BMC Genet. 4(Suppl. 1)::S8.
Hanson, R.L. et al. 2001. . Family and genetic studies of indices of insulin sensitivity and insulin secretion in Pima Indians. Diabetes Metab. Res. Rev. 17::296-303.
Renstr?m, E., Eliasson, L., Bokvist, K., and Rorsman, P. 1996. . Cooling inhibits exocytosis in single mouse pancreatic B cells by suppression of granule mobilization. J. Physiol. 494::41-52.
Martell, A.E., and Smith, R.M. 1971. Amino acids. Volume 1 of Critical stability constants. Plenum Press. New York, New York, USA..
Martell, A.E., and Smith, R.M. 1971. Amines. Volume 2 of Critical stability constants. Plenum Press. New York, New York, USA..
Salehi, A. et al. 1999. . Dysfunction of the islet lysosomal system conveys impairment of glucose-induced insulin release in the diabetic GK rat. Endocrinology. 140::3045-3053.
Bonnevie-Nielsen, V., Steffes, M.W., and Lernmark, A. 1981. . A major loss in islet mass and B cell function precedes hyperglycemia in mice given multiple low doses of streptozotocin. Diabetes. 30::424-429.
Olofsson, C.S. et al. 2002. . Fast insulin secretion reflects exocytosis of docked granules in mouse pancreatic B cells. Pflugers Arch. 444::43-51.(Xingjun Jing, Dai-Qing Li)
2Institute of Neurophysiology and Center of Molecular Medicine Cologne, University of Cologne, Cologne, Germany.
3Diabetes Research Laboratories, Oxford Centre for Diabetes, Endocrinology and Metabolism, the Churchill Hospital, Oxford, United Kingdom.
Abstract
Concerted activation of different voltage-gated Ca2+ channel isoforms may determine the kinetics of insulin release from pancreatic islets. Here we have elucidated the role of R-type CaV2.3 channels in that process. A 20% reduction in glucose-evoked insulin secretion was observed in CaV2.3-knockout (CaV2.3–/–) islets, close to the 17% inhibition by the R-type blocker SNX482 but much less than the 77% inhibition produced by the L-type Ca2+ channel antagonist isradipine. Dynamic insulin-release measurements revealed that genetic or pharmacological CaV2.3 ablation strongly suppressed second-phase secretion, whereas first-phase secretion was unaffected, a result also observed in vivo. Suppression of the second phase coincided with an 18% reduction in oscillatory Ca2+ signaling and a 25% reduction in granule recruitment after completion of the initial exocytotic burst in single CaV2.3–/– ? cells. CaV2.3 ablation also impaired glucose-mediated suppression of glucagon secretion in isolated islets (27% versus 58% in WT), an effect associated with coexpression of insulin and glucagon in a fraction of the islet cells in the CaV2.3–/– mouse. We propose a specific role for CaV2.3 Ca2+ channels in second-phase insulin release, that of mediating the Ca2+ entry needed for replenishment of the releasable pool of granules as well as islet cell differentiation.
See the related Commentary beginning on page 16
Introduction
Systemic glucose tolerance is orchestrated by the regulated release of insulin and glucagon from the ? and cells of the pancreatic islets of Langerhans. The and ? cells are electrically excitable and use electrical signals to couple changes in blood glucose concentration to stimulation or inhibition of hormone release. In both cell types, influx of extracellular Ca2+ through voltage-gated Ca2+ channels with resultant elevation of intracellular Ca2+ concentration ([Ca2+]i) triggers exocytosis of the hormone-containing secretory granules. Like other electrically excitable cells, both and ? cells contain several types of voltage-gated Ca2+ channel (1, 2). Assigning physiological functions to the respective Ca2+ channels is central to the understanding of electrical and secretory activities in these cells.
Voltage-gated Ca2+ channels are divided into 3 subfamilies: (a) L-type high voltage–activated (HVA) Ca2+ channel family that comprises the CaV1.1, 1.2, 1.3, and 1.4 channels and is inhibited by dihydropyridines (DHPs) (1, 3, 4); (b) non–L-type HVA channels CaV2.1 (P/Q-type), 2.2 (N-type), and 2.3 (R-type) that are sensitive to -agatoxin IVA and -conotoxin GVIA and SNX482, respectively (1, 4, 5); and (c) the low voltage–activated (LVA) T-type Ca2+ channel family (CaV3.1, 3.2, and 3.3). The latter subtype differs electrophysiologically from the HVA Ca2+ channels in opening transiently already upon modest depolarization (6, 7) and fulfilling important roles in pacemaker cells (8).
The chain of events that couples an elevation in blood glucose to initiation of ? cell electrical activity is well established and involves facilitated transport of the sugar into the ? cell and its subsequent metabolic degradation by glycolysis and mitochondrial oxidation, resulting in closure of the ATP-sensitive K+ channels (KATP channels) and ? cell depolarization, with resultant activation of voltage-gated Ca2+ channels and regulated insulin exocytosis (9). Glucose-stimulated insulin secretion consists of a rapid first-phase of insulin secretion that lasts for approximately 10 minutes before declining to near-basal levels, followed by less prominent but sustained second-phase insulin secretion that can last for several hours (10, 11). The cellular mechanisms underlying biphasic insulin release remain unclear, but consensus exists that an elevation in [Ca2+]i is required for both first- and second-phase insulin secretion (12). Type 2 diabetes is associated with a shift from biphasic to monophasic insulin release (10), and it is therefore important to establish the cell biology of insulin release kinetics.
In mouse pancreatic ? cells, 50% of the whole cell Ca2+ current exhibits properties typical for L-type channels, is inhibited by DHP channel blockers such as isradipine or nifedipine, and is activated by BayK8644. The molecular identity of the ? cell L-type Ca2+ channel involved in insulin secretion has been debated, but recent studies have established that CaV1.2 or 1C channels play a decisive role (13, 14).
In mouse ? cells, there is evidence suggesting that the secretory granules and the L-type Ca2+ channels assemble into a tight functional complex (13, 15-17). Thanks to this organization, the ? cells are capable of exocytosis at rates comparable to those encountered in chromaffin cells, although the Ca2+ channel density is only 5–10% of that in the latter cell type. This arrangement bears strong resemblance to the tight coupling of P/Q-type CaV2.1 as well as N-type CaV2.2 channels to rapid synaptic transmission (18, 19). No such interaction has been demonstrated for the R-type Cav2.3 channels. By contrast, recent evidence in neurons suggests that R-type channels are physically detached from the exocytotic machinery (20) and are not involved in rapid neurotransmission in mossy fibers (21).
About one-quarter of the ? cell whole-cell Ca2+ current is sensitive to the R-type Ca2+ channel blocker SNX482, but the role of R-type Ca2+ channels in insulin secretion remains elusive. A general CaV2.3-knockout (CaV2.3–/–) mouse has been established. It exhibits a relatively modest neurological phenotype, including altered pain responses (22), impaired spatial memory (23), and enhanced fear reaction (24). Pancreatic islets express the endocrine splice variant of CaV2.3 (25), and previous investigations of glucose homeostasis in CaV2.3–/– mice have demonstrated a slight glucose intolerance and reduced glucose-induced insulin secretion (26). Here we have extended these initial observations by performing in vivo glucose tolerance tests, dynamic measurements of phasic insulin secretion in situ, static pancreatic hormone–release experiments in isolated islets, as well as single cell recordings of whole-cell Ca2+ currents and exocytosis and ratiometric measurements of the cytoplasmic Ca2+ concentration in WT and CaV2.3–/– islets. We demonstrate that whereas R-type Ca2+ channels play a minor role in rapid insulin release, their significance becomes more apparent during second-phase secretion. These data point to an emerging picture where Ca2+ influx through different ? cell Ca2+ channels play distinct functional roles.
Results
Whole cell Ca2+ currents in and? cells from CaV2.3+/+ and CaV2.3–/– mice.
Integrated Ca2+ current versus voltage (Q-V) relations were measured in dissociated single pancreatic islet cells using the perforated-patch whole cell configuration. Insulin-releasing ? cells were identified by the absence of voltage-gated Na+ currents at membrane potentials relevant for ? cell electrical activity, that is, between the ? cell resting membrane potential (–70 mV) and the peak of the ? cell action potential (–10 mV; Figure 1A, inset). In ? cells from CaV2.3–/– mice, the integrated Ca2+ currents observed at membrane potentials less than or equal to –10 mV were reduced compared with WT cells. At –10 mV, the reduction averaged approximately 23% (P < 0.05; Figure 1B). The CaV2.3–/– mice exhibited a selective loss of a high-voltage Ca2+ current component, and current amplitudes at voltages below –10 mV were not affected. The R-type Ca2+ channel blocker SNX482 (100 nM) had no effect on voltage-gated Ca2+ currents in CaV2.3–/– mice, whereas the L-type inhibitor isradipine (2 μM) significantly reduced Ca2+ influx approximately 60% at potentials greater than or equal to –30 mV (P < 0.05; Figure 1, C and D).
Cells exhibiting Na+ currents when holding at –70 mV (Figure 1E, inset) were classified as non–? cells and likely represent glucagon-releasing cells (27). The frequency of cells was much lower in dispersed islet cells from CaV2.3–/– mice than cells made from WT islets. Counting all cells that could be functionally defined as belonging to either group, only approximately 7% of the CaV2.3–/– cells could be classified as cells versus approximately 27% of the WT cells. In 4 cells from CaV2.3–/– mice with the electrophysiological properties expected for cells, the Ca2+ current amplitude was not different from that observed in WT cells (Figure 1, E and F).
Single cell exocytosis in CaV2.3–/– and CaV2.3+/+ mice.
Ca2+-elicited exocytosis in ? cells was monitored as increases in whole-cell membrane capacitance and was elicited by trains of 10 500-ms voltage clamp depolarizations from the holding potential –70 mV to 0 applied at 1 Hz (Figure 2, A and B). In ? cells from CaV2.3–/– mice, the increase in cell capacitance during the train averaged 392 ± 47 femtofarads (fF) (n = 7). The latter value is 21% less than the exocytotic response evoked by the same stimulus in WT mice (P < 0.05; 496 ± 42 fF; n = 6). Interestingly, the early component of exocytosis (in response to the first depolarization) was not affected and averaged 55 ± 19 fF and 57 ± 22 fF in WT and CaV2.3–/– ? cells, respectively. Instead, selective suppression of the late component of exocytosis was observed. The 23% reduction of the whole cell Ca2+ current observed in the CaV2.3–/– ? cells corresponded nicely with the overall reduction in exocytosis, suggesting that exocytotic capacity as such was not affected. This idea was reinforced by the observation that exocytotic rates, measured during intracellular dialysis of a Ca2+-containing pipette solution using the standard whole-cell configuration, were identical in ? cells from both mouse strains (Figure 2, C and D). In the few cells that could be identified in CaV2.3–/– mice, exocytosis was not different from that observed in WT, either when elicited by train depolarizations (n = 3) or Ca2+ buffer dialysis (n = 3; Figure 2, E–H).
Effects of CaV2.3 ablation on intracellular Ca2+ homeostasis.
Insulin secretion is initiated by changes in the submembrane cytoplasmic free [Ca2+]i (28, 29), which are determined by ? cell electrical activity. We therefore monitored [Ca2+]i under basal conditions as well as after the addition of glucose, the KATP channel blocker tolbutamide, or following depolarization with high extracellular K+ (Figure 3). Resting [Ca2+]i measured at 5 mM glucose was identical in islets from CaV2.3–/– and WT mice and averaged approximately 90 nM. In WT islets, elevation of extracellular glucose from 5 to 15 mM evoked an initial robust elevation in [Ca2+]i after 204 ± 33 seconds (n = 7). This initial peak decayed slowly (over 4–10 minutes) toward baseline and was eventually followed by a pattern of repetitive [Ca2+]i oscillations. Subsequent depolarization by the nonmetabolizable stimuli tolbutamide and high K+ in the continued presence of glucose elicited a rapid elevation of [Ca2+]i but did not result in oscillatory activity. In CaV2.3–/– islets, the glucose-evoked initial peak in [Ca2+]i occurred 227 ± 15 seconds (n = 9) after elevating the glucose concentration. This initial amplitude was almost unaffected (7% decrease), but the time-averaged [Ca2+]i during the subsequent oscillatory phase was 17% lower than in WT islets (P < 0.01). In addition, the oscillatory activity was 29% slower in the CaV2.3–/– islets (1.8 ± 0.1 versus 2.5 ± 0.2 bursts/minute in CaV2.3–/– and WT islets, respectively; P < 0.01). By contrast, the peaks in [Ca2+]i induced by depolarization with tolbutamide or high K+ were, if anything, slightly augmented in the CaV2.3–/– islets (NS versus control). The finding that no [Ca2+]i oscillations were observed at 5 mM glucose and that elevating the glucose concentration to 15 mM increased Ca2+ signaling suggests that the signal principally reflects the behavior of the pancreatic ? cells (see ref. 30).
In vivo glucose tolerance and pancreatic hormone release in CaV2.3–/– mice.
In vivo glucose homeostasis was investigated by intraperitoneal glucose challenges (2 g/kg body weight; Table 1). Basal (nonfasted) plasma glucose levels averaged 9.8 ± 0.6 mmol/l (n = 8) in WT CaV2.3+/+ mice and increased 67% and 135% 3 and 8 minutes after the glucose load, respectively. In CaV2.3–/– mice, basal glucose levels were elevated by 17% compared with WT (P < 0.05 versus WT; n = 9), and the plasma glucose concentrations measured 3 and 8 minutes after challenge were also elevated approximately 15% (P < 0.05 for values at 8 minutes versus WT). Basal plasma insulin levels did not differ between the CaV2.3–/– and CaV2.3+/+ strains, although the plasma glucose concentration was somewhat higher in the former mice. In WT mice, glucose increased plasma insulin concentrations 43% and 52% 3 and 8 minutes after the challenge, respectively. In CaV2.3–/– mice, the glucose-induced increase in circulating insulin was similar to that in WT mice 3 minutes after challenge, but after 8 minutes it was limited to less than 10% (P < 0.05 versus WT). These findings reinforce previous observations indicating an impaired glucose tolerance of the CaV2.3–/– mice (26).
We also monitored plasma glucagon concentrations during the glucose challenge. Basal glucagon levels in CaV2.3–/– mice were modestly decreased (approximately 10%) compared with CaV2.3+/+ littermates. In WT mice, the glucose load reduced circulating glucagon 15% already at 3 minutes after challenge, and an additional 8% decrease was observed at 8 minutes. In CaV2.3–/– mice, the inhibition of glucagon release was sluggish. At 3 minutes after challenge, plasma glucagon concentrations actually increased 18% (P < 0.05 versus basal; n = 9).
Insulin and glucagon release in vitro in pancreatic islets.
Pancreatic islets in situ receive extensive neuronal input from parasympathetic nerve endings (31). Since CaV2.3 Ca2+ channels are also expressed in neuronal tissue, their ablation might affect pancreatic hormone release in vivo by indirect mechanisms and not by directly affecting and ? cell function. Insulin and glucagon secretion was therefore also investigated in isolated islets. The first set of experiments investigated the effects of L-type and R-type channel blockers isradipine and SNX482, respectively, in WT CaV2.3+/+ islets (Table 2). Basal insulin release (1 mM glucose) was low and was unaffected by either channel blocker. Elevation of extracellular glucose to 20 mM stimulated insulin release more than 15-fold. Isradipine (2 μM) suppressed glucose-stimulated insulin release 77%, whereas SNX482 (100 nM) reduced insulin secretion 17% (Table 2). The latter value agrees favorably with the inhibition observed in capacitance measurements (see Figure 2, A and B). Glucagon release was measured under the same conditions. At 1 mM glucose glucagon release was high and remained unchanged by either isradipine or SNX482. Elevating the glucose concentration (20 mM) reduced glucagon secretion 70% (P < 0.001; low versus high glucose). At high glucose, blockade of R-type channels by SNX482 failed to affect glucagon release, whereas isradipine stimulated glucagon secretion under the same condition.
We next compared the effects of the Ca2+ channel blockers with the consequences of CaV2.3 gene ablation on pancreatic hormone release (Table 3). In WT CaV2.3+/+ islets, elevating glucose from 1 to 20 mM again stimulated insulin release more than 15-fold. In CaV2.3–/– islets, glucose-stimulated insulin release was reduced approximately 25% compared with WT. The effects of isradipine on glucose-stimulated insulin release were comparable in WT (66% reduction) and in CaV2.3–/– islets (68% suppression). Insulin secretion elicited by stimulation with high extracellular K+ (50 mM) resulted in a 75% enhancement of release relative to that observed in the presence of glucose alone in WT islets. A similar relative stimulation was observed in CaV2.3–/– islets (96%), but in absolute terms, the response in the latter type of islets was reduced 13% compared with WT islets. Surprisingly, whereas elevating glucose from 1 to 20 mM suppressed glucagon release 58% in WT CaV2.3+/+ islets (P < 0.001; low versus high glucose), the inhibitory action of the sugar was severely impaired in the CaV2.3–/– islets and amounted to a mere 27% (P < 0.01 versus WT). Isradipine exerted divergent actions in WT and CaV2.3–/– islets. As discussed above, in WT islets the L-type Ca2+ channel antagonist stimulated glucagon release 52% when applied in the presence of 20 mM glucose. By contrast, isradipine reduced glucagon secretion in CaV2.3–/– islets 32% under high-glucose conditions. Increasing extracellular K+ enhanced glucagon secretion massively, approximately 320% in WT and 220% in CaV2.3–/– islets.
Phasic insulin release measured by in situ pancreatic perfusion.
To assess the role of CaV2.3 in dynamic insulin release, we performed in situ pancreatic perfusions with fractionated sampling. When WT CaV2.3+/+ pancreata were perfused (Figure 4A) with a low-glucose solution (3.3 mM), insulin release was barely detectable. After increasing the glucose concentration to 16.7 mM (at t = 11 minutes), first-phase insulin release was initiated with a 2-minute delay. The peak in first-phase insulin secretion was attained 2 minutes after onset of release (t = 15 minutes) and measured 8.8 ± 1.7 ng/ml (n = 4). Insulin release then exhibited a transient nadir phase (t = 18–21 minutes) during which release rates averaged 4–5 ng/ml, before accelerating again during second-phase insulin secretion to approximately 18 ng/ml at t = 36 minutes and later. When the same experiment was repeated in CaV2.3–/– pancreata (Figure 4B), the peak in first-phase insulin secretion was only slightly reduced (19%) and measured 7.1 ± 2.1 ng/ml (NS; n = 4). More importantly, second-phase insulin release was markedly suppressed and averaged 9.4 ± 2.7 ng/ml at t = 40 minutes, representing a 46% reduction compared with WT (P < 0.05).
Genetic ablation of Cav2.3 may result in compensatory mechanisms resulting in rearrangements of the Ca2+ channels in the ? cell itself or neighboring and cells. In addition, we wanted to verify that the observed effects of CaV2.3 on insulin secretion are not limited to the background C57B mouse strain. Similar experiments were therefore made in pancreata from standard inbred NMRI mice, but instead CaV2.3 channel function was inhibited by SNX482. In the absence of the channel inhibitor, first-phase insulin secretion was initiated with a 2-minute delay and peaked 1 minute, 30 seconds later when it measured 15.3 ± 2.5 ng/ml (n = 6). Insulin release then decayed toward a lower plateau level of approximately 8 ng/ml and increased slightly to approximately 10 ng/ml during the second phase (Figure 4C). When SNX482 (100 nM) was included in the perfusate (Figure 4D), first-phase insulin release was largely unaffected, and peak values averaged 13.0 ± 3.7 ng/ml (n = 6). By contrast, second-phase insulin secretion was reduced more than 80% (e.g., 1.8 ± 0.6 ng/ml versus 10.0 ± 0.6 ng/ml at t = 40 minutes in the presence or absence of SNX482, respectively; P < 0.01). It was verified that the glucose concentration in the effluent medium was identical in all experiments.
Insulin and glucagon immunoreactivity in WT and CaV2.3–/– islet cells.
Islet cells were finally dispersed and investigated by confocal immunocytochemistry (Figure 5). In WT islet cells, 80% of the dispersed cells revealed immunoreactivity for insulin and approximately 20% stained positive for glucagons in double-labeling experiments. The relative frequency of and ? cells suggested by this analysis compares favorably with the electrophysiological data. Only 1 out of 230 cells investigated revealed immunoreactivity for both insulin and glucagons. A rather different picture emerged in the CaV2.3–/– islet cell preparation. Whereas 75% and 9% of the cells could readily be characterized as ? and cells, respectively, 16% of the 210 cells coexpressed insulin and glucagons.
Discussion
Insulin-producing ? cells contain multiple types of Ca2+ channel. Whereas the role of L-type Ca2+ channels in insulin secretion is amply documented, that of the non–L-type Ca2+ channels is less well understood. The advent of subtype-specific Ca2+ channel blockers in combination with the generation of transgenic knockout mice provides a unique opportunity for an in-depth analysis of the function(s) fulfilled by the different Ca2+ channel subtypes. Here we have used SNX482 and CaV2.3-null mice to study the significance of R-type CaV2.3 Ca2+ channels in islet insulin and glucagon secretion in vivo and in vitro.
The role of CaV2.3 channels for? cell [Ca2+]i homeostasis and insulin secretion.
Here we demonstrate that pharmacological inhibition of R-type CaV2.3 Ca2+ channels using SNX482 does not affect first-phase insulin secretion, but reduces second-phase release a dramatic 80%. A similar preferential effect on late-phase insulin secretion is observed in CaV2.3–/– mice (Figure 4). At the single ? cell level, close inspection of the capacitance recordings (Figure 2, A and B) reveals that the initial component of exocytosis is not much affected in CaV2.3–/– ? cells. In fact, the overall reduction results exclusively from suppression of the late component of exocytosis (elicited by the third and subsequent depolarizations). We have previously demonstrated that opening of L-type CaV1.2 Ca2+ channels is tightly associated with rapid exocytosis and first-phase secretion. Thus, it appears that Ca2+ entry via L-type CaV1.2 and R-type CaV2.3 channels have distinct intracellular effects.
CaV2.3 ablation is associated with a 23% decrease in ? cell Ca2+ current in CaV2.3–/– mice (Figure 1). This is in good agreement with the response to acute application of SNX482 in WT mice (13), suggesting that there is little compensatory upregulation of other Ca2+ channels in the knockout mice. We acknowledge that CaV2.3 channel activity may represent only part of the R-type current component (24, 32). In addition, other reports suggest that all R-type current components are not blocked by SNX482 with the same efficacy (33). The observation that SNX482 fails to affect whole cell Ca2+ currents in CaV2.3–/– ? cells (Figure 1, C and D), however, indicates at least that SNX482, at the concentration used here (100 nM), does not affect any Ca2+ channels other than CaV2.3, but we cannot exclude the contribution of an SNX482-insensitive R-type Ca2+ current component. Indeed, approximately 40% of the whole cell Ca2+ current is unaffected by either SNX482 or the L-type blocker isradipine in CaV2.3–/– ? cells. Although most of this current probably represents P/Q-type Ca2+ currents as previously shown for WT ? cells (13), it is possible that part of it reflects a SNX482-resistant portion of the R-type Ca2+ current.
How do CaV2.3 Ca2+ channels regulate second-phase insulin secretion?
Secretory granules in pancreatic ? cells can be classified according to their release competence (34). A limited (1–5%) pool is immediately available for rapid exocytosis upon stimulation (readily releasable pool, RRP). Once this pool of granules has been released, exocytosis proceeds at a slower rate determined by the supply of granules newly mobilized from a much larger reserve pool (35). The observation that only a late component of exocytosis is affected in CaV2.3–/– ? cells (Figure 2, A and B) argues that CaV2.3 channels are not tightly coupled to exocytosis but are more important for the recruitment of new granules for release. We also demonstrate that genetic ablation of the CaV2.3 Ca2+ channels results in approximately 20% reduction of the glucose-induced steady-state time-averaged [Ca2+]i (more than 5 minutes after initial response; Figure 3). Can these 2 pieces of information be reconciled? We have reported previously that a moderate (40%) but protracted elevation of [Ca2+]i results in strong enhancement of the exocytotic capacity (36). It is of interest that this effect is mediated by a global rather than a localized increase in [Ca2+]i and is operational already at concentrations as low as 200–300 nM, that is, 10-fold lower than those required to elicit fast exocytosis. Thus, even moderate reduction of Ca2+ entry and [Ca2+]i following inhibition of CaV2.3 Ca2+ channels can have strong effects on insulin secretion in the longer term by inhibition of granule recruitment. Presumably, L-type Ca2+ channels still trigger granule exocytosis, but once the RRP is depleted, the supply of new granules for release becomes rate limiting. This would explain how ablation of the approximately 20% CaV2.3 Ca2+ channel component can result in approximately 50% (knockout) to 80% (SNX482) reduction of second-phase insulin secretion. Thus, it seems justifiable to conclude that Ca2+ entry via R-type CaV2.3 Ca2+ channels is functionally (and perhaps spatially) linked to granule mobilization and priming of insulin granules for release.
CaV2.3 ablation and glucagon secretion.
CaV2.3 ablation is associated with disturbances of glucagon secretion. In isolated islets, the ability of glucose to suppress glucagon secretion is severely impaired (Table 3). These abnormalities are also detectable at the systemic level. In the in vivo glucose challenge test, 3-minute values for plasma glucagon were stimulated approximately 15% in CaV2.3–/– mice, whereas in WT they were reduced approximately 20% (Table 1). The observations that whole cell Ca2+ currents (Figure 1, E and F) and single cell exocytosis (Figure 2, E–H) are unaffected in CaV2.3–/– cells, taken together with the finding that the R-type blocker SNX482 fails to affect glucagon release in WT islets (Table 2), make it unlikely that these effects are the direct consequence of the loss of CaV2.3 in the cell. Rather, the explanation for the perturbation of glucose-inhibited glucagon release in CaV2.3-ablated mice appears to be due to the appearance of atypical /? cells accounting for greater than 60% of the glucagon-expressing cells in these mice (Figure 5, B and C). Taken together with the scarcity of cells with cell electrophysiological properties in preparations from CaV2.3–/– mice (7% compared with 27% in WT mice), this suggests that the /? cells electrical behavior is more like that of the ? cells. This idea is supported by the results in Table 3, showing that whereas isradipine suppresses glucagon release in CaV2.3–/– islets, it has the opposite effect in WT islets. This observation also provides a clue to the relative failure of glucose to suppress glucagon release in CaV2.3–/– islets both in vitro and (acutely) in vivo. The fact that glucagon eventually decreases after a glucose challenge in vivo may be related to neuronal input or paracrine effects that are known to be of great importance for the control of the hyperglycemic hormone (37). Surprisingly, basal glucagon release (at low glucose) is unaffected in CaV2.3–/– islets. One possibility is that the reduced population of normal cells have a very large capacity for exocytosis, but the capacitance measurements (Figure 2, E–H) provide no support for this notion. Another explanation is that the /? cells exhibit poorly regulated exocytosis and constitutively release both glucagon and insulin. This idea is reinforced by the observation that basal insulin release in CaV2.3–/– islets is higher than in WT islets (0.4 ± 0.05 versus 0.2 ± 0.02 ng/islet/hour in CaV2.3–/– and WT, respectively; Table 3). Yet another possibility is that glucagon coreleased with insulin from the /? cells potentiates further secretion by a cAMP-dependent mechanism. We have reported elsewhere that glucagon increases ? cell exocytosis 5-fold (38). The significance of paracrine mechanisms within the islet is apparent from the strong stimulatory action of isradipine on glucagon secretion in WT islets (Tables 2 and 3). The latter effect may reflect the relief from paracrine inhibition of glucagon secretion exerted by Zn2+ and GABA cosecreted with insulin from the ? cells (39, 40). Preliminary data from our laboratory indicate that R-type CaV2.3 Ca2+ channels play a decisive role in somatostatin secretion from the cells (Q. Zhang, A. Salehi, E. Renstr?m, and P. Rorsman, unpublished observations). We can exclude that the loss of physiological glucose inhibition of glucagon secretion is secondary to relief from paracrine inhibition by somatostatin, however, since the dysregulation of glucagon secretion in CaV2.3–/– islets is not mimicked by SNX482.
Role of CaV2.3 in islet cell differentiation.
A surprising finding in the present study is that the majority of the glucagon immunoreactive cells in CaV2.3–/– islets coexpress insulin, indicating that CaV2.3 channels play a role in the development of mature cells. A transient peak in CaV2.3 channel expression in glial cells along specific CNS pathways has been demonstrated to coincide with postnatal myelinization of the white matter in the rat (41). It can be speculated that CaV2.3 channel expression exerts a similar action in defining the differentiated mature islet cell lineages, but the exact underlying mechanism remains to be established.
Pathophysiological implications.
Although R-type CaV2.3 channels appear to be of little importance for first-phase insulin secretion and their contribution becomes apparent only during late exocytosis/second-phase insulin secretion, this does not mean that they are unimportant for systemic glucose homeostasis. Indeed, the CaV2.3–/– mice exhibited basal hyperglycemia (Table 1). In humans, CaV2.3 is encoded by the CACNA1E gene, which is located on chromosome 1q25–31. Interestingly, a chromosomal region around 1q25 reveals linkage to type 2 diabetes in several independent studies in different populations. These include early onset type 2 diabetes in Pima Indians (42), Utah Caucasians (43), and English sib pairs (44). In addition, the 1q25–32 region demonstrates linkage with elevated blood glucose levels in the Framingham Heart Study (45). It is also pertinent that the above region confers defective insulin secretion in type 2 diabetes in Pima Indians (46). Type 2 diabetes is a multifactorial polygenic disorder. Clearly, polymorphisms in the CACNA1E gene alone are not sufficient to trigger disease. Given the present functional data, however, it is tempting to speculate that dysfunctional CaV2.3 Ca2+ channels may frequently be involved in creating the disturbed ? cell phenotype in type 2 diabetes.
Methods
Experimental animals.
As previously reported (26), the CACNA1E gene encoding Cav2.3 was disrupted in vivo by deleting a region containing exon 2 on mating Cav2.3fl/+ and deleter mice on a C57Bl/6 background that expresses Cre recombinase constitutively under the control of the CMV promoter. Cav2.3+/– mice containing 1 cre transgene were inbred, and pups with the CaV2.3–/– genotype were selected and transferred by embryo transfer into a SPF facility. The transfer included breeding with C57Bl/6 mice and resulted in heterozygous Cav2.3+/– mice. Only cre-negative pups were selected and inbred, yielding either CaV2.3+/+ or CaV2.3–/– mice. Thus, the CaV2.3+/+ and CaV2.3–/– mice used in this study have an identical genetic background. In Figure 4, inbred NMRI mice purchased from Charles River Wiga GmbH were used.
The mice were housed at a constant temperature (22–23°C) and 12-hour light cycles (6:00 a.m.–6:00 p.m.), with access to standard pellet food and water ad libitum. All experiments were evaluated and approved by the local ethical committee Malm?/Lund djurf?rs?ksetiska n?mnd, Lund District Court, Sweden.
Islet isolation and islet cell preparation.
The mice were sacrificed by cervical dislocation, and collagenase was administered into the pancreas by retrograde injection via the pancreatic duct. After 15–20 minutes’ incubation at 37°C, the islet suspension was washed 4 times with HBSS (4°C) before being manually collected. The islets were cultured in RPMI-1640 medium (Invitrogen Corp.) supplemented with 10% FCS, penicillin, and streptomycin. For the preparation of single cells, islets obtained by collagenase digestion were dissociated by incubation and gentle triculation in Ca2+-free medium. The resulting cell suspension was centrifuged, the supernatant discarded, and the pellet (containing the cells) was resuspended in RPMI-1640 medium, plated on plastic Nunc 35-mm Petri dishes, and maintained in tissue culture for up to 2 days.
Electrophysiology.
The measurements were conducted using an EPC-10 patch-clamp amplifier in conjunction with the PULSE software suite (version 8.53; HEKA Elektronik). Whole-cell Ca2+ currents were measured in intact cells using the perforated-patch whole-cell approach (Figure 1) using a pipette solution consisting of 76 mM Cs2SO4, 10 mM NaCl, 10 mM CsCl, 1 mM MgCl2, 5 mM HEPES (pH 7.35 with KOH), and 0.24 mg/ml amphotericin B (47). Exocytosis was monitored as increases in cell capacitance using the sine + DC mode of the lock-in amplifier included in the PULSE software and the standard whole cell configuration. When eliciting exocytosis by trains of ten 500-ms voltage clamp depolarizations (Figure 2, A and E), the pipette solution consisted of 125 mM Cs-glutamate, 10 mM CsCl, 10 mM NaCl, 1 mM MgCl2, 5 mM HEPES, 3 mM Mg-ATP, 0.1 mM cAMP, and 0.05 mM EGTA (pH 7.2 with CsOH). In Figure 2, C and G, this pipette solution was slightly modified to include 10 mM EGTA and 9 mM CaCl2, and all Cs+ salts were replaced by corresponding K+ salts. The resulting free intracellular Ca2+ concentration of this Ca2+/EGTA buffer was estimated to 1.5 μM using the binding constants of Martell and Smith (48, 49). The extracellular bath solution contained 138 mM NaCl, 5.6 mM KCl, 2.6 mM CaCl2, 1.2 mM MgCl2, 5 mM glucose, and 5 mM HEPES (pH 7.4 with NaOH). In the recordings of whole-cell Ca2+ currents (Figure 1) and depolarization-evoked exocytosis (Figure 2, A and E), 20 mM of NaCl was equimolarly replaced by the K+ channel blocker TEA-Cl to facilitate the separation of the small voltage-gated Ca2+ currents from the large outward K+ current. The DHP isradipine (Pfizer Inc.) was prepared as stock solution in DMSO (final concentration less than or equal to 0.1%). The R-type blocker SNX482 (Peptide Institute Inc.) was dissolved directly in the extracellular medium. All other reagents were from Sigma-Aldrich. Effects were determined in the steady state. The bath (approximately 1.5 ml) was continuously perfused (6 ml/min) and the temperature maintained at approximately 32°C.
In vivo glucose challenges.
For the in vivo studies, glucose (11.1 mmol [equivalent to 2 g]/kg body weight) was dissolved in 0.9% NaCl and delivered by intraperitoneal injection. Blood sampling, detection of plasma insulin by RIA, and enzymatic determination of plasma glucose concentrations were performed as described previously (50).
In situ/ex vivo pancreatic perfusion.
Experiments were performed in the morning at 10:00 a.m. in nonfasted mice. Anesthesia was given by intraperitoneal injection of midazolam (Hofmann-La Roche AG; 0.4 mg/25 g body weight) and fentanyl (Janssen Pharmaceuticals Inc.; 0.02 mg/25 g body weight). The experimental procedures were essentially identical to those described by Bonnevie-Nielsen et al. (51). Briefly, the mice were kept on a heating pad during the entire experiment. After opening the abdominal cavity and ligating the renal, hepatic, and splenic arteries, the aorta was tied off above the level of the pancreatic artery. The pancreas was perfused with modified Krebs-Ringer HEPES buffer preheated to 37°C (1 ml/min) via a silicone catheter placed in the aorta. The perfusate was collected via a silicone catheter from the portal vein at 30- or 60-second intervals, as indicated, in 2.5-ml Eppendorf tubes containing 25 μl Trasylol. Insulin and glucose concentrations in the effluent medium were detected by RIA and the glucose oxidase method, respectively.
In vitro pancreatic hormone release.
Insulin release in vitro was measured in static incubations. Briefly, freshly isolated islets were preincubated for 30 minutes at 37°C in a Krebs-Ringer bicarbonate buffer (pH 7.4) consisting of 120 mM NaCl, 25 mM NaHCO3, 4.7 mM KCl, 1.2 mM MgSO4, 2.5 mM CaCl2, 1.2 mM KH2PO, 1 mM glucose, and 10 mM HEPES (pH 7.4). The medium was gassed with 95% O2 and 5% CO2 to obtain constant pH and oxygenation. Groups of 10 islets were then incubated in 1 ml for 60 minutes at 37°C in Krebs-Ringer buffered solution supplemented with either glucose, the L-type Ca2+ channel blocker isradipine, the R-type Ca2+ channel inhibitor SNX482, tolbutamide, or high K+, as specified in the text and figures. Immediately after incubation, a 25-μl aliquot of the medium was removed for assay of insulin and glucagon radioimmunoassay as described previously (50).
Immunocytochemistry.
Insulin and glucagon immunoreactivities were visualized in dissociated islet cells by indirect immunocytochemistry using a Carl Zeiss AG 510 LSM confocal microscope and a x100 Plan-Apochromat x100/1.4 oil objective. After fixation with 3% paraformaldehyde and permeabilization by 0.1% Triton-X, the cells were incubated with normal donkey serum to reduce unspecific staining. The primary insulin and glucagon Abs, raised in guinea pigs and sheep, respectively, were incubated at 1:1,000. To prevent cross-talk between the channels, the secondary Cy3-conjugated anti–guinea pig and the Cy5-conjugated anti-sheep Abs (1:600) were excited in the multitrack mode using the 543-nm and 633-nm lines of the HeNe lasers, and emitted light was collected using greater than 560-nm and greater than 650-nm long-pass filters, respectively.
Microfluorimetry.
[Ca2+]i in intact pancreatic islets was measured by dual-wavelength microfluorimetry using fura-2 and a D104 PTI microfluorimetry system. The temperature of the experimental chamber was +32°C to allow comparison with the electrophysiological data. Procedures for loading and calibration of the fluorescence signal were as described previously (52).
Statistical analysis.
All data are given as means ± SEM. Statistical significance was evaluated using absolute values only. A paired Student’s t test was used when comparing responses in the same cell. For comparisons between groups with independent observations we used independent Student’s t tests or, when comparisons involved more than 2 groups, ANOVA.
Acknowledgments
We thank Kristina Borglid and Britt-Marie Nilsson for expert technical assistance. This work was supported by the Swedish Research Council (grant numbers 13509, 12239, and 6589), the European Foundation for the Study of Diabetes, European Community network grant "Neuronal Ca2+ Channels," the Albert P?hlsson Foundation, the NovoNordisk Foundation, the Center of Molecular Medicine Cologne (BMBF 01 KS 9502), and the Deutsche Forschungsgemeinschaft (SCH 387/9-1). P. Rorsman is a recipient of the G?ran Gustafsson Award for Research in the Natural Sciences and Medicine.
Footnotes
Nonstandard abbreviations used: [Ca2+]i, intracellular Ca2+ concentration; CaV2.3–/–, CaV2.3-knockout; DHP, dihydropyridine; fF, femtofarad(s); HVA, high voltage–activated; KATP, ATP-sensitive K+; LVA, low voltage–activated; RRP, readily releasable pool.
Conflict of interest: The authors have declared that no conflict of interest exists.
References
Catterall, W.A. 1998. . Structure and function of neuronal Ca2+ channels and their role in neurotransmitter release. Cell Calcium. 24::307-323.
D’Ascenzo, M. et al. 2004. . Electrophysiological and molecular evidence of L-(Cav1), N- (Cav2.2), and R- (Cav2.3) type Ca2+ channels in rat cortical astrocytes. Glia. 45::354-363.
Dolphin, A.C. 1999. . L-type calcium channel modulation. Adv. Second Messenger Phosphoprotein Res. 33::153-177.
Catterall, W.A. 2000. . Structure and regulation of voltage-gated Ca2+ channels. Annu. Rev. Cell Dev. Biol. 16::521-555.
Reid, C.A., Bekkers, J.M., and Clements, J.D. 2003. . Presynaptic Ca2+ channels: a functional patchwork. Trends Neurosci. 26::683-687.
Perez-Reyes, E. 2003. . Molecular physiology of low-voltage-activated T-type calcium channels. Physiol. Rev. 83::117-161.
Heady, T.N., Gomora, J.C., Macdonald, T.L., and Perez-Reyes, E. 2001. . Molecular pharmacology of T-type Ca2+ channels. Jpn. J. Pharmacol. 85::339-350.
Triggle, D.J. 1998. . The physiological and pharmacological significance of cardiovascular T-type, voltage-gated calcium channels. Am. J. Hypertens. 11::80S-87S.
Ashcroft, F.M., Proks, P., Smith, P.A., ?mm?l?, C., Bokvist, K., and Rorsman, P. 1994. . Stimulus-secretion coupling in pancreatic ? cells. J. Cell. Biochem. 55(Suppl.)::54-65.
Del Prato, S., Marchetti, P., and Bonadonna, R.C. 2002. . Phasic insulin release and metabolic regulation in type 2 diabetes. Diabetes. 51(Suppl. 1)::S109-S116.
Grodsky, G.M., and Bolaffi, J.L. 1992. . Desensitization of the insulin-secreting ? cell. J. Cell. Biochem. 48::3-11.
Henquin, J.C., Ravier, M.A., Nenquin, M., Jonas, J.C., and Gilon, P. 2003. . Hierarchy of the ? cell signals controlling insulin secretion. Eur. J. Clin. Invest. 33::742-750.
Schulla, V. et al. 2003. . Impaired insulin secretion and glucose tolerance in ? cell-selective Ca(v)1.2 Ca2+ channel null mice. EMBO J. 22::3844-3854.
Sinnegger-Brauns, M.J. et al. 2004. . Isoform-specific regulation of mood behavior and pancreatic ? cell and cardiovascular function by L-type Ca2+ channels. J. Clin. Invest. 113::1430-1439. doi:10.1172/JCI200420208.
Wiser, O. et al. 1999. . The voltage sensitive Lc-type Ca2+ channel is functionally coupled to the exo-cytotic machinery. Proc. Natl. Acad. Sci. U. S. A. 96::248-253.
Barg, S., Eliasson, L., Renstr?m, E., and Rorsman, P. 2002. . A subset of 50 secretory granules in close contact with L-type Ca2+ channels accounts for first-phase insulin secretion in mouse ? cells. Diabetes. 51(Suppl. 1)::S74-S82.
Barg, S. et al. 2001. . Fast exocytosis with few Ca(2+) channels in insulin-secreting mouse pancreatic B cells. Biophys. J. 81::3308-3323.
Mochida, S. et al. 2003. . Requirement for the synaptic protein interaction site for reconstitution of synaptic transmission by P/Q-type calcium channels. Proc. Natl. Acad. Sci. U. S. A. 100::2819-2824.
Mochida, S., Sheng, Z.H., Baker, C., Kobayashi, H., and Catterall, W.A. 1996. . Inhibition of neurotransmission by peptides containing the synaptic protein interaction site of N-type Ca2+ channels. Neuron. 17::781-788.
Wu, L.G., Borst, J.G., and Sakmann, B. 1998. . R-type Ca2+ currents evoke transmitter release at a rat central synapse. Proc. Natl. Acad. Sci. U. S. A. 95::4720-4725.
Dietrich, D. et al. 2003. . Functional specialization of presynaptic Cav2.3 Ca2+ channels. Neuron. 39::483-496.
Saegusa, H. et al. 2000. . Altered pain responses in mice lacking 1E subunit of the voltage-dependent Ca2+ channel. Proc. Natl. Acad. Sci. U. S. A. 97::6132-6137.
Kubota, M. et al. 2001. . Intact LTP and fear memory but impaired spatial memory in mice lacking Ca(v)2.3 ((IE)) channel. Biochem. Biophys. Res. Commun. 282::242-248.
Lee, S.C. et al. 2002. . Molecular basis of R-type calcium channels in central amygdala neurons of the mouse. Proc. Natl. Acad. Sci. U. S. A. 99::3276-3281.
Vajna, R. et al. 1998. . New isoform of the neuronal Ca2+ channel 1E subunit in islets of Langerhans and kidney—distribution of voltage-gated Ca2+ channel 1 subunits in cell lines and tissues. Eur. J. Biochem. 257::274-285.
Pereverzev, A. et al. 2002. . Disturbances in glucose-tolerance, insulin-release, and stress-induced hyperglycemia upon disruption of the Ca(v)2.3 ( 1E) subunit of voltage-gated Ca(2+) channels. Mol. Endocrinol. 16::884-895.
Barg, S., Galvanovskis, J., G?pel, S.O., Rorsman, P., and Eliasson, L. 2000. . Tight coupling between electrical activity and exocytosis in mouse glucagon-secreting cells. Diabetes. 49::1500-1510.
Rutter, G.A., Theler, J.M., Li, G., and Wollheim, C.B. 1994. . Ca2+ stores in insulin-secreting cells: lack of effect of cADP ribose. Cell Calcium. 16::71-80.
Bokvist, K., Eliasson, L., ?mm?l?, C., Renstr?m, E., and Rorsman, P. 1995. . Co-localization of L-type Ca2+ channels and insulin-containing secretory granules and its significance for the initiation of exocytosis in mouse pancreatic B cells. EMBO J. 14::50-57.
Nadal, A., Quesada, I., and Soria, B. 1999. . Homologous and heterologous asynchronicity between identified -, ?- and delta cells within intact islets of Langerhans in the mouse. J. Physiol. 517::85-93.
Satin, L.S. 2000. . Localized calcium influx in pancreatic ? cells: its significance for Ca2+-dependent insulin secretion from the islets of Langerhans. Endocrine. 13::251-262.
Wilson, S.M. et al. 2000. . The status of voltage-dependent calcium channels in 1E knock-out mice. J. Neurosci. 20::8566-8571.
Tottene, A., Volsen, S., and Pietrobon, D. 2000. . (1E) subunits form the pore of three cerebellar R-type calcium channels with different pharmacological and permeation properties. J. Neurosci. 20::171-178.
Neher, E. 1998. . Vesicle pools and Ca2+ microdomains: new tools for understanding their roles in neurotransmitter release. Neuron. 20::389-399.
Rorsman, P., and Renstr?m, E. 2003. . Insulin granule dynamics in pancreatic ? cells. Diabetologia. 46::1029-1045.
Gromada, J. et al. 1999. . CaM kinase II-dependent mobilization of secretory granules underlies acetylcholine-induced stimulation of exocytosis in mouse pancreatic B cells. J. Physiol. 518::745-759.
Nandi, J. et al. 2002. . Central mechanisms involved with catabolism. Curr. Opin. Clin. Nutr. Metab. Care. 5::407-418.
Gromada, J., Ding, W.G., Barg, S., Renstr?m, E., and Rorsman, P. 1997. . Multisite regulation of insulin secretion by cAMP-increasing agonists: evidence that glucagon-like peptide 1 and glucagon act via distinct receptors. Pflugers Arch. 434::515-524.
Ishihara, H., Maechler, P., Gjinovci, A., Herrera, P.L., and Wollheim, C.B. 2003. . Islet ? cell secretion determines glucagon release from neighbouring cells. Nat. Cell Biol. 5::330-335.
Wendt, A. et al. 2004. . Glucose inhibition of glucagon secretion from rat cells is mediated by GABA released from neighboring ? cells. Diabetes. 53::1038-1045.
Chen, S., Ren, Y.Q., Bing, R., and Hillman, D.E. 2000. . A 1E subunit of the R-type calcium channel is associated with myelinogenesis. J. Neurocytol. 29::719-728.
Hanson, R.L. et al. 1998. . An autosomal genomic scan for loci linked to type II diabetes mellitus and body-mass index in Pima Indians. Am. J. Hum. Genet. 63::1130-1138.
Elbein, S.C., Hoffman, M.D., Teng, K., Leppert, M.F., and Hasstedt, S.J. 1999. . A genome-wide search for type 2 diabetes susceptibility genes in Utah Caucasians. Diabetes. 48::1175-1182.
Wiltshire, S. et al. 2001. . A genomewide scan for loci predisposing to type 2 diabetes in a U.K. population (the Diabetes UK Warren 2 Repository): analysis of 573 pedigrees provides independent replication of a susceptibility locus on chromosome 1q. Am. J. Hum. Genet. 69::553-569.
Jun, G., Song, Y., Stein, C.M., and Iyengar, S.K. 2003. . An autosome-wide search using longitudinal data for loci linked to type 2 diabetes progression. BMC Genet. 4(Suppl. 1)::S8.
Hanson, R.L. et al. 2001. . Family and genetic studies of indices of insulin sensitivity and insulin secretion in Pima Indians. Diabetes Metab. Res. Rev. 17::296-303.
Renstr?m, E., Eliasson, L., Bokvist, K., and Rorsman, P. 1996. . Cooling inhibits exocytosis in single mouse pancreatic B cells by suppression of granule mobilization. J. Physiol. 494::41-52.
Martell, A.E., and Smith, R.M. 1971. Amino acids. Volume 1 of Critical stability constants. Plenum Press. New York, New York, USA..
Martell, A.E., and Smith, R.M. 1971. Amines. Volume 2 of Critical stability constants. Plenum Press. New York, New York, USA..
Salehi, A. et al. 1999. . Dysfunction of the islet lysosomal system conveys impairment of glucose-induced insulin release in the diabetic GK rat. Endocrinology. 140::3045-3053.
Bonnevie-Nielsen, V., Steffes, M.W., and Lernmark, A. 1981. . A major loss in islet mass and B cell function precedes hyperglycemia in mice given multiple low doses of streptozotocin. Diabetes. 30::424-429.
Olofsson, C.S. et al. 2002. . Fast insulin secretion reflects exocytosis of docked granules in mouse pancreatic B cells. Pflugers Arch. 444::43-51.(Xingjun Jing, Dai-Qing Li)