Diabetes Mellitus: A Fundamental and Clinical Text
3rd Edition
© 2004 Lippincott Williams & Wilkins
←↑→
1
The Two Phases of Glucose-Stimulated Insulin Secretion Mechanisms and Controls
Susanne G. Straub
Geoffrey W. G. Sharp
Biphasic Insulin Secretion
Glucose-stimulated biphasic insulin secretion has intrigued researchers and clinicians alike since the phenomenon was first observed. This has been so, despite the fact that it is a rare occurrence under physiologic conditions because an abrupt increase in the blood glucose concentration is required to elicit the response—and such an increase does not usually occur following a meal. Fig. 1.1 shows a glucose-induced biphasic response of isolated rat islets. After a change in glucose from a basal concentration of 2.8 mM to the stimulatory 16.7 mM, there is a short delay prior to the onset of the first phase of secretion. The rate of insulin secretion increases to a peak over the next 4 minutes and then declines over a further 4 minutes to a nadir that marks the end of the first phase. The second phase arising from the nadir is characterized by a steadily increasing rate until it reaches the plateau. While more information is available concerning the mechanisms and control of the first phase than that of the second phase, our knowledge of both is still rudimentary. Because the magnitude of the first phase response is reduced in people with type 2 diabetes, and can be reduced in people prior to the development of overt type 1 diabetes (1,2,3,4), the importance of understanding the biphasicity is obvious.
Models
Historically, modeling the biphasic response led to a compartmental model and to a signal-related feedback model (5,6,7,8). Both reproduce the biphasic pattern of insulin secretion by perfused pancreas under stimulation by glucose. The two-compartment model is shown in Fig. 1.2. The two compartments are often assumed to be pools of insulin-containing granules, but the compartments could equally be signals for secretion. In the model, an abrupt increase in the concentration of glucose induces the first phase of release. Because the magnitude of the first phase response to glucose is proportional to the concentration, the small compartment must be heterogeneous and contain granules or signals with different sensitivities to glucose (6). The second phase is produced by a gradually increasing excitatory signal up to the plateau of the response, whereupon the signal remains constant. This two-compartment model, first described in 1972 (6), has been remarkably resilient over time. The excitor-inhibitor or feedback model is shown in Fig. 1.3. In this case, the first phase is caused by an increase in an excitatory signal or signals, followed by a feedback inhibitory signal. The concentration dependency of the first phase to glucose (5,6) is presumed to be due to concentration-dependent increases in the excitatory signal followed by the feedback inhibition. The second phase of insulin secretion in this model is due to the same potentiating mechanism as that in the two-compartment model, that is, a gradually increasing excitatory signal. As will be seen as this chapter unfolds, a two-compartment model consisting of two pools of docked insulin-containing granules combined with excitatory signals that trigger and augment exocytosis are sufficient to explain biphasic insulin release. Only for long-term sustained release is a multicompartment model required.
β-cell Granule Pools
Insulin-containing granules in the β-cell are either docked at the plasma membrane or free to move in the cytosol. The latter are defined as reserve granules (9). The amount of insulin present in the endocrine pancreas is huge relative to the needs of the body at any particular time, and following a meal, only a small fraction of the granules in the pancreas will be released. These two major pools of granules are complex, and the docked granules have different potential for release. On arrival at the membrane after translocation from the reserve pool, they dock and are subsequently primed for release readiness. Therefore, the docked granules may be primed or unprimed. The primed granules also may be subdivided into a group that can be defined as readily releasable and a smaller group (with the lowest threshold for release) that can be defined as immediately releasable (10,11,12). The reserve pool is also complex because some of the newly synthesized granules are preferentially secreted relative to older granules (13,14,15).
Figure 1.1. Biphasic insulin secretion. Rat pancreatic islets were perfused with Krebs-Ringer buffer containing 2.8 mM glucose for 40 minutes before the glucose concentration was changed to 16.7 mM at the 10-minute point. After a brief delay, the islets respond to the increased glucose concentration with a rapid increase in the rate of insulin secretion that takes only 4 minutes to the peak and declines to a nadir after a further 4 minutes. This constitutes the first phase. Following the nadir, the rate of secretion rises to a plateau that is maintained to the end of the glucose stimulation (the second phase). The results are shown as means ± SEM from five separate experiments.
P.4

The number of granules in the mouse β-cell was first estimated by quantitative morphometric analysis as 13,000 (16). A later estimate puts the number at 9,000, with the morphologically docked granules comprising approximately 7% of the total (12). For the calculations that follow, a mean value of 11,000 granules per β-cell is used. The number of immediately releasable granules in the docked pool, as shown by capacitance studies, is in the range of 50 to 60 granules (11,12), or only 0.5% of the number of granules available. This is similar to the percentage content released during insulin secretion experiments on mouse islets. Assuming an “average” β-cell and granules of the same size and insulin content, the number of granules released by the β-cell during the first phase is 55 (0.5% of 11,000). Thus, the immediately releasable granules in the morphologically docked pool are presumed responsible for the first phase of glucose-stimulated insulin release. For a first phase of 5 minutes duration, the average rate of granule release would be only one every 5 seconds or so. The remaining granules in the morphologically docked pool are likely to be largely responsible for the rising period of the second phase of release and the early part of the plateau. However, for sustained release over longer periods of time, granules from the reserve pool must be translocated to the plasma membrane to replace those that are released. Given the slow rate of β-cell granule exocytosis, the rate at which granules from the reserve pool dock at the plasma membrane to replace those lost from the docked pool must be equally slow. A scheme of the β-cell granule pools is shown in Fig. 1.4.
Figure 1.2. Two-compartment model of insulin secretion. (For mathematical and quantitative development, see refs. 5 and 7.) A: Glucose stimulates first-phase release by rapidly depleting insulin (or a metabolic signal similarly depleted during secretion). Glucose also stimulates a potentiating signal (P) that accumulates gradually and drives the increasing second-phase secretion. B: Results when a partial, but constant, defect for release from the small compartment is introduced. Response to glucose is initially inhibited. With time, P causes a compensating increased content in the small compartment; release is now regulated solely by P and is normal. (From
Grodsky GM. An update on implications of phasic insulin secretion. In: Flatt PR, Lenzen S, eds. Insulin secretion and pancreatic B-cell research. London: Smith-Gordon Press, 1994:421–430
, with permission.)
Figure 1.3. Feedback or excitor-inhibitor model of insulin secretion. (For mathematical and quantitative development, see refs. 5 and 7.) A: Glucose initially stimulates transient first-phase release by increasing an excitor (E) (B), followed quickly by a rise in an inhibitor (I) (C). Release is a function of the difference between E and I. Glucose also stimulates the same potentiator (P) as described in the compartmental model (Fig. 1.2), which accumulates gradually and, in this case, causes second-phase release by increasing the excitor (E) (D). I, which increased only with glucose, remains constant. (From
Grodsky GM. Kinetics of insulin secretion: underlying metabolic events. In: LeRoith D, Taylor SI, Olefsky JM, eds. Diabetes mellitus: a fundamental and clinical text, 2nd ed. Philadelphia: Lippincott Williams & Wilkins
, with permission.)
Sustained stimulation of insulin release results in granule translocation to the plasma membrane, docking, priming in
P.5

preparation for release, and exocytosis. As already stated, the reserve pool is large relative to the size of the immediately releasable pool and to release rates. Similarly, the morphologically docked pool is large relative to the size of the immediately releasable pool that is contained within it. Most of the morphologically docked granules are primed and readily releasable; a minority are primed and immediately releasable. However, a single immediately releasable pool does not explain what we know of β-cell responsiveness. As has been known for many years (5,6,17), the number of granules released from the immediately releasable pool during the first phase is related to the concentration of glucose, most likely because of concentration-dependent changes in depolarization, Ca2+ channel activation and [Ca2+]i. The size of this pool can be increased by time-dependent potentiation (TDP), and by agents such as glucagon-like peptide-1 (GLP-1) and pituitary adenylate cyclase-activating polypeptide (PACAP) or acetylcholine and cholecystokinin, agonists that raise intracellular cyclic adenosine monophosphate (cAMP) and diacylglycerol (DAG) levels, respectively. The effects of these agents occur so promptly that they must be acting on granules that are “readily releasable” and rapidly transform them to the immediately releasable state. Thus, the docked granules are heterogeneous and exist in several different states of readiness for release. Furthermore, in addition to being responsible for the first phase of glucose-stimulated release, the docked granule pools are most likely responsible for the second phase. Assuming there are approximately 770 granules in the docked pool (7% of 11,000), it can be calculated that they are sufficient to maintain the second phase of glucose-stimulated insulin secretion by the mouse β-cell for over 2 hours. Information is not available on granule numbers for the rat or human β-cell. However, if similar to the mouse β-cell, it would mean that the rising portion of the second-phase response following the nadir between the two phases, which lasts for only 20 to 30 minutes, is also derived from the docked, readily releasable granules. Therefore, the rate-controlling step for the second phase of glucose-stimulated insulin release will be on the conversion of docked, readily releasable granules to the immediately releasable pool. It is obvious that during the second phase of release, translocation of granules from the reserve pool to the plasma membrane will occur to replace those released. However, granule translocation from the reserve pool appears to be neither necessary nor rate limiting for the rising portion of the second-phase response. The rate-limiting step in the second phase of glucose-stimulated insulin secretion is the conversion of readily releasable granules into immediately releasable granules, whereupon in a stimulated β-cell they would immediately undergo exocytosis. Despite the apparent differences in kinetics between the first and second phases, the mechanisms involved in the preparation and release of the granules are presumably
P.6

the same. After insulin synthesis and granule formation, the granules move to the plasma membrane, where they dock and are primed and sorted into the readily releasable pool prior to joining the smaller, immediately releasable pool for exocytosis on activation of the cell. It should be noted that while all granules are prepared for release in a similar manner, the order in which the granules are released is not strictly sequential. There is evidence for the release of newly formed granules during sustained secretion. Therefore, it is possible for granules to “jump the queue.” As will be described later, following the elevation of [Ca2+]i and first phase of release, the second phase is due to elevated [Ca2+]i and additional glucose-induced signals that augment the response to [Ca2+]i. These signals are currently unknown.
Figure 1.4. β-cell granule pools. Scheme of the minimal description of β-cell granule pools under resting and activated conditions. In the resting, nonstimulated state (A), it is assumed that most of the docked granules are primed and present in the readily releasable and immediately releasable pools. The numbers of granules in the different pools are calculated from a mean of 11,000 granules per β cell (12,16), of which 0.5% are in the immediately releasable pool (11,12). This is similar to the percentage content released during insulin secretion experiments on islets. Upon activation of the β cell by glucose (B), the immediately releasable granules are discharged as the first phase of release. Subsequently, readily releasable granules are converted (or move) to the immediately releasable pool, whereupon they are immediately released. The conversion to the immediately releasable state is presumed to occur at an increasing rate during the rising portion of the second phase until it reaches its maximum rate at the plateau. Granules from the reserve pool will replace the granules released in order to prepare for sustained secretion if required. However, the number of predocked, readily releasable granules is sufficient for both the first and second phases of release as described in the text.
The Rate-Limiting Steps
The patterns of insulin secretion during the two phases of glucose-stimulated insulin secretion are obviously different, and both phases have separate rate controls. Following an initial increase in [Ca2+]i, the rate of exocytosis of the immediately releasable granules is controlled by a step between the recognition of the Ca2+ signal and the mechanism of exocytosis itself. Whatever the mechanism of this step might be, it has a faster rate than the rate-limiting step for the second phase. At the nadir between the two phases, the rate of exocytosis is low relative to the peak rates of either phase. Therefore, at this time, the release of insulin-containing granules is rate limited. The granules in the immediately releasable pool have been discharged, and the pool has to be resupplied. The replacement granules are most likely from the readily releasable pool of docked and primed granules. As soon as the replacements become immediately releasable, in an activated β-cell, they will undergo exocytosis. Therefore, the rate-limiting step for the second phase is the rate of conversion of granules to the immediately releasable state. In the rat and human islet, after the second phase is first observed following the nadir between the two phases, the rate of exocytosis increases over time until it reaches a plateau. Over this period, the effect of glucose is to gradually accelerate the rate-limiting step. This crucial step in the preparation of granules for the second phase of release has not been identified, nor has the mechanism by which glucose affects it. In seeking the rate-limiting step for the second phase of release, one has to consider the sequence of events from granule docking to exocytosis. Conceivably, docking, priming, conversion to immediate releasability, translocation on the membrane to an exocytotic site, or movement of the site or critical components to the granule could be rate limiting. However, there are good reasons to think that neither docking nor priming is rate limiting. Docking seems to be an unlikely rate-limiting step because, as stated earlier, there is an excess of docked granules that is sufficient to sustain glucose-stimulated insulin secretion for 1 to 2 hours. Priming also seems unlikely because it can occur at a much faster rate than granule release. While the term priming has not been defined clearly, adenosine triphosphate (ATP) has a role in the preparation of granules to fusion competence (18,19). In the mouse pancreatic β-cell, elevation of [Ca2+]i in the cell by photo release from caged Ca2+ in the presence of ATP evokes a biphasic stimulation of exocytosis as measured by capacitance changes (18). Exocytosis occurred within 45 msec of the elevation of [Ca2+]i, which represents the time taken by the granules and plasma membrane to recognize the Ca2+ signal and undergo fusion. The first phase under these conditions was complete in 200 msec. The second phase was slower and lasted at least 10 seconds. The presence of adenosine 51-[beta, gamma-methylene] triphosphate (AMP-PCP), a non-hydrolyzable analogue of ATP, abolished the second phase but did not affect the first phase. Thus, the first phase was due to a pool of primed granules that was immediately available for release. The pool of granules responsible for the second phase, the release of which was abolished by AMP-PCP, were either not primed and required ATP to proceed to fusion competence, or lost their primed state in the presence of AMP-PCP. In further experiments, the time required for ATP priming was measured. After removal of endogenous ATP, the photo release of ATP in the presence of elevated [Ca2+]i resulted in exocytosis within 400 msec (18). Therefore, the time required for granules to undergo ATP-dependent priming and reach the immediately releasable state should not impose any rate limitation. In another study in which β-cells were stimulated by caged Ca2+, again two rates were observed. ATP both accelerated and augmented the exocytosis occurring with the faster rate. ATP-γ-S, another nonhydrolyzable analogue of ATP, was more effective than ATP. Inhibitors of adenylyl cyclase and of cAMP both blocked the effect of ATP. The investigators concluded that there is a distal ATP-sensing mechanism for Ca2+-dependent exocytosis that involves adenylyl cyclase and PKA, and that rapid exocytosis involves both Ca2+ and protein phosphorylation (19). Summarizing these data, it seems unlikely that priming is involved in the rate-limiting step for the second phase of insulin release. ATP-dependent priming is rapid (400 msec) and much faster than the rate of granule release during the second-phase response. In addition, the β-cell contains a reserve pool of primed granules that is more than sufficient for the rising portion of the second phase when the rate-limiting step is accelerating. Therefore the rate-limiting step occurs after the priming event.
With respect to the role of GTP in exocytosis, both GTP and GTP-γ-S induced a concentration-dependent increase in exocytosis, as measured by capacitance change, in the absence of intracellular Ca2+ (9). The responses were unaffected by cAMP. The Hill equation for the relationship between the maximum exocytotic rate and [Ca2+] was not changed by guanosine triphosphate (GTP), so GTP had no effect on the Ca2+ sensitivity of exocytosis. Guanosine diphosphate (GDP)-β-S inhibited exocytosis induced by either GTP-γ-S or Ca2+, demonstrating the involvement of G proteins in both processes. However, GDP-β-S was without effect on exocytosis evoked by depolarization-mediated Ca2+ entry. Thus, an immediately releasable pool of granules could be discharged by Ca2+ entry in a GTP-independent manner. These results indicate separate and common components in the secretory pathways stimulated by Ca2+ and GTP, with at least two sites of action for Ca2+. The time course of GTP-γ-S–stimulated exocytosis following rapid elevation of GTP-γ-S by photolysis of a caged precursor was dependent on the intracellular Ca2+ and cAMP concentrations (9). In summary, ATP, GTP, and Ca2+ play important roles in preparing granules for exocytosis, the nature of the interactions they control is not known, but they are not rate limiting for the second phase of insulin secretion.
P.7

Cyclic AMP enhances Ca2+-stimulated exocytosis while having no effect in the absence of stimulated exocytosis (20). Under standard whole-cell conditions, cAMP augmented insulin secretion induced by a single depolarization, and this was unaffected by inhibition of protein kinase A (PKA). When exocytosis was stimulated by the release of caged Ca2+, cAMP had PKA-dependent and PKA-independent effects. There was a rapid effect that was PKA independent and a late effect that was PKA-dependent. It was concluded that cAMP potentiates insulin secretion by increasing the release probability of granules in the immediately releasable pool and by increasing the rate at which the pool is refilled (21). Therefore, its major effects appear to be immediately prior to exocytosis per se (22). This confirms our belief that prior to the immediately releasable pool is a readily releasable pool that supplies it. This pool can be mobilized by activation of PKA or by increased DAG levels. The manner in which PKA acts is likely different from the manner by which the KATP channel-independent pathways act. However, the action of DAG could be similar to that of the KATP channel-independent pathways if, as has been suggested, glucose stimulates the production of DAG (23,24,25,26). The effect of acetylcholine to stimulate insulin release results from two mechanisms (see reference 27 for review). One mechanism involves an increase in [Ca2+]i and the other involves an increase in the level of DAG. Increased DAG results in an augmentation of Ca2+-stimulated exocytosis that could be mediated by protein kinase C (PKC) isoforms or other DAG-binding proteins. Thus, PKA-, PKC-, and DAG-binding proteins are involved in the augmentation of exocytosis, and they act via several different mechanisms.
Signaling Pathways in Glucose-Stimulated Insulin Secretion
In 1984, the ATP-sensitive potassium channel (KATP channel) was identified (28) and found to be responsive to glucose (29). Exposure of the β-cell to a glucose concentration that stimulates insulin release closes the channel and depolarizes the cell. Depolarization results in increased Ca2+ entry through voltage-dependent Ca2+ channels, an increase in [Ca2+]i, and stimulation of insulin release (29,30,31). Thus, the KATP channel is the link between glucose metabolism and the stimulation of insulin secretion. This signaling pathway is now known as the KATP channel–dependent or “triggering” pathway (10,32). A second glucose-signaling pathway that augments the secretory response to increased [Ca2+]i was demonstrated in 1992 (33,34,35). This is currently named the KATP channel–independent augmentation pathway. The KATP channel–dependent pathway triggers the release of insulin by increasing [Ca2+]i, and the KATP channel–independent pathway enhances the response to the increased [Ca2+]i. The two pathways, dependent and independent, are synergistic. A KATP channel–independent pathway has been demonstrated that enhances insulin release even in the absence of extracellular Ca2+ and in the absence of an increase in [Ca2+]i. This was achieved by the use of phorbol esters or agonists such as carbachol that increase intracellular DAG levels, and forskolin or other agonists that increase intracellular cAMP levels (36,37,38,39). This signaling pathway can be distinguished from the pathway that acts in the presence of Ca2+ by depletion of cellular GTP levels. Stimulation of the Ca2+-independent pathway is eliminated by reducing GTP levels in the islets with mycophenolic acid, while maintaining ATP levels with adenine. Under these conditions, the Ca2+-dependent pathway was untouched (40). It is not known whether the pathway that can operate in the absence of an increase in [Ca2+]i has a physiologic role, although for it not to have a role would require that it be specifically shut down in the presence of Ca2+. If it has a role, it would occur when the pancreatic islets are exposed to nutrients and multiple agonists that increase cAMP and DAG levels simultaneously, as happens following a mixed meal. Such an effect has been shown experimentally (39). The KATP channel–dependent and –independent pathways are illustrated in Fig. 1.5.
Mechanisms Underlying the Biphasicity of Glucose-Stimulated Insulin Release
The First Phase
The sequence of events that trigger the first phase of insulin release involve the rapid equilibration of glucose between the outside and the inside of the β-cell, an event that is facilitated by an excess of glucose transporters in the plasma membrane. Glucose metabolism, initiated and controlled by its interaction with glucokinase (41), leads to an elevation of the ATP/adenosine diphosphate (ADP) ratio and closure of KATP channels (29). The overall rate of insulin secretion and the amount of insulin released during both phases are set by glucokinase activity and the rate of conversion of glucose to glucose-6-phosphate (41). Depolarization of the β-cell leads to gating of the primarily voltage-dependent calcium channels. Increased Ca2+ entry results in increased [Ca2+]i and exocytosis of the immediately releasable pool of granules. Recent evidence suggests that number and location of the Ca2+ channels in the β-cell may have a major influence over the secretion rate. The concentration of Ca2+ close to the mouth of a Ca2+ channel will be higher than is the case deeper into the cell or laterally from the channel mouth because of diffusion and Ca2+ buffering and sequestering mechanisms. Additionally, if Ca2+ channels are clustered, the concentration of Ca2+ close to the cluster will be even higher. There is evidence for “hot spots” of exocytotic activity in specific locations on the cell membrane (42), which is presumed to be caused by the presence of aggregates of Ca2+ channels, granules, and exocytotic machinery (12,43,44,45). Thus, the immediately releasable granules that make up the first phase of insulin secretion would be present in these specialized areas at the membrane. It seems likely that after they have been discharged, at the nadir between the two phases, the conversion of readily releasable granules to the immediately releasable state may require a biochemical modification of the granule; equally, it might be necessary to reassemble the aggregates of channels, granules, and exocytotic machinery. One of these processes (11,12,43,44,45) is presumed to be the rate-limiting step during the second phase.
The size of the immediately releasable pool is determined by the KATP channel–independent pathway (and TDP). The docked granules exist in a state of equilibrium between the
P.8

readily releasable pool of docked, primed granules and the pool of immediately releasable granules. The equilibrium position depends on the ambient glucose concentration, the activity of the KATP channel–independent pathway, and previous exposure to glucose (the effect of TDP). Activation of the KATP channel–independent pathways and TDP drive the equilibrium position to the right, and in the case of TDP to increase the size of the immediately releasable pool. Low ambient glucose concentrations, as occurs during fasting for example, allow the equilibrium position to move to the left so that the size of the immediately releasable pool and the first phase of insulin release is decreased.
Figure 1.5. The KATP channel–dependent (triggering) and KATP channel–independent (augmentation) pathways of glucose signaling. Following exposure to a stimulatory concentration of glucose, depolarization of the cell by the KATP channel–dependent pathway increases [Ca2+]i to stimulate insulin release. The KATP channel–independent pathways then augment the rate of Ca2+-stimulated insulin release. Two other augmentation pathways are shown. Acetylcholine and other agonists that act via G protein–coupled receptors and phospholipase C increase [Ca2+]i and DAG. Other agonists like GLP-1 and PACAP working via G protein–coupled receptors activate adenylyl cyclase and raise cyclic AMP levels. All these augmentation pathways must act on the rate-limiting step between the readily releasable and immediately releasable granule pools, although this is not the only site at which they act in the cell. Not shown is the KATP channel–independent pathway that operates in the absence of an increase in [Ca2+]i. This pathway interacts with DAG-stimulated release and is further enhanced by cyclic AMP.
The Second Phase
In human and rat pancreas, the second phase of glucose-stimulated insulin release is due to the KATP channel–independent pathways and TDP acting to potentiate the effect of the KATP channel–dependent pathway (45,46,47). The underlying mechanisms are not known. Multiple possible mechanisms and second messengers have been suggested. These include changes in [Ca2+]i that parallel (and drive) the biphasic response, Ca2+-calmodulin and calmodulin kinase II, cAMP via both PKA and PKA-independent reactions, cyclic GMP, DAG activation of PKC isoforms and DAG-binding proteins, nitric oxide, phospholipase A2, and phosphatidylinositol 3-kinase (PI3-kinase), increased malonyl coenzyme A (CoA) and cytosolic long-chain acyl CoA, changes in adenine and guanine nucleotide levels, and increased glutamate levels.
The most prominent hypothesis, first proposed in 1989, is that glucose, via anaplerosis, increases the amount of mitochondrial citrate that results in increased levels of citrate and malonyl-CoA in the cytosol (23). As malonyl-CoA inhibits carnitine palmitoyl transferase 1 (CPT1), it exerts a major control over the disposition of fatty acids. In this case, diversion of fatty acids from β-oxidation in the mitochondria is expected to increase the amount of long-chain acyl-CoA moieties in the cytoplasm. These long-chain acyl-CoAs and their ability to generate other potential signaling molecules (23,24,25,26) are presumed responsible for the second phase of release. This hypothesis, despite a great deal of evidence in its favor, is now controversial. Some of the evidence, for and against, is as follows:
  • The stimulation of insulin release by glucose is associated with increased citrate levels, increased malonyl-CoA, inhibition of CPT1, and decreased fatty acid oxidation (23,24,25,26). Against this assertion is the fact that much of the evidence in favor of the hypothesis has been obtained from work on cloned β-cell lines. When the HIT-TI5, INS-1, and βHC9 cell lines were examined for expression of the KATP channel– independent pathway, it was found that it did not exist in these cells (48). The KATP channel–independent (Ca2+-dependent) pathway has been demonstrated classically by the use of diazoxide to activate the KATP channel (thus eliminating the effect of glucose via the KATP channel– dependent pathway) and KCl to depolarize the cell (33,34). Under these conditions, glucose augments the KCl-induced secretion by KATP channel–independent mechanisms. A demonstration of this type of experiment on isolated islets of the rat is shown in Fig. 1.6. Under the same conditions, responses of HIT-TI5, INS-1, and βHC9 cell lines to KCl, diazoxide, and high glucose concentrations were no different from their responses to KCl, diazoxide, and basal nonstimulatory glucose concentrations (48). Thus, the insulin secretory response of these cell lines is solely due to the glucose-induced increased in [Ca2+]i, and there is no KATP channel-independent effect of glucose. However, as described later, a
    P.9

    cell line has now been developed that manifests the independent pathway.
    Figure 1.6. KATP channel–independent pathway.This pathway was first demonstrated by the use of diazoxide, which opens the KATP channel, thus nullifying the effect of glucose on the channel, and a depolarizing concentration of KCl to replace the depolarizing action of glucose (32,33). Under these conditions, with no effect of glucose on the KATP channel, additional effects of glucose were sought. Thus, in the figure, both control and test islets from rats were treated with 250 μM diazoxide and 40 mM KCl, and 2.8 mM (○) or 16.7 mM glucose (●), respectively. Both control and test islets responded similarly over the first few minutes due to the stimulatory effect of depolarization by KCl. Subsequently, the response of the control islets decreased to low values, typical of the response to KCl. In contrast, the test islets responded with a “second-phase–like” release of insulin. The effect of 16.7 mM glucose under these conditions is independent of any effect of glucose on the KATP channel and has been termed the KATP channel–independent or augmentation pathway. The results are shown as means ± SEM from five separate experiments.
  • Hydroxycitrate, an inhibitor of citrate lyase, prevents the build-up of citrate and malonyl CoA and inhibits glucose-stimulated insulin secretion (49). Notably, the provision of palmitate under these conditions eliminated the inhibition of secretion by hydroxycitrate. Against this assertion is the fact that others have not confirmed these findings with hydroxycitrate and the reversal of its effects by palmitate (50,51).
  • Pharmacologic inhibition of CPT1 potentiates glucose stimulated-insulin secretion (49,52).
  • β-cells in which acetyl-CoA carboxylase (ACC) is decreased by stable expression of antisense ACC have reduced secretory responses to glucose but not to depolarizing concentrations of KCl (53). The reasons for these results are not clear, because the clonal cell line used does not manifest the KATP channel–independent pathway.
  • Acute stimulation of permeabilized INS-1 cells with long-chain acyl-CoA enhances Ca2+-stimulated insulin release (54). Again, INS-1 cells do not have the KATP channel pathway.
The absence of the KATP channel–independent pathway from clonal cell lines does not mean that evidence favorable to the malonyl CoA long-chain acyl-CoA hypothesis that has been obtained from these cells is of no value. The lack of a response could result from the absence of a distal mechanism in the pathway, and proximal mechanisms may be valid. However, evidence against the hypothesis that is derived from cell lines that do not possess the pathway cannot be considered valid. For example, malonyl-CoA decarboxylase was expressed in INS-1 cells to decrease malonyl-CoA levels and test the hypothesis. The maneuver reduced the effect of glucose to decrease the oxidation of fatty acid oxidation and decrease the incorporation of fatty acids into lipids. Nevertheless, there was no effect on glucose-stimulated insulin secretion (50). Triacsin C, an inhibitor of long-chain acyl-CoA synthetase, reduced fatty acid oxidation and incorporation into lipids. Again there was no effect on glucose-stimulated insulin release (50). These results would be expected in a cell that has no KATP channel–independent pathway and therefore provide no evidence against the hypothesis. Future studies on the KATP channel–independent pathway should be performed on pancreatic islets or on cell lines that do manifest the pathway. Such clonal cell lines do now exist. INS-1 cell clones, such as the INS-1 832/13 cells, that do express the KATP channel–independent pathway have been developed (55). These cells respond to KCl, diazoxide, and 2.8 mM and 16.7 mM glucose with a similar first phase of response. However, the cells treated with 16.7 mM glucose have a second-phase response that is larger than that in the presence of 2.8 mM glucose, that is, they have a KATP channel–independent pathway. When human malonyl-CoA decarboxylase lacking both its mitochondrial and peroxisomal targeting sequences was expressed in these cells, the glucose-induced increase in malonyl-CoA was blocked and the inhibitory effect of glucose on fatty acid oxidation was reduced. Again there was no effect on glucose-stimulated insulin release (56). Also in these cells, triacsin C did not impair glucose-stimulated insulin release. These findings are not consistent with the malonyl-CoA/long-chain acyl-CoA hypothesis. An additional piece of evidence against this hypothesis is that the KATP channel–independent pathway of glucose signaling can be mimicked by 2-aminobicyclo[2.2.1]heptane-2-carboxylic acid (BCH), which is a nonmetabolizable analogue of leucine and which activates glutamate dehydrogenase to provide α-ketoglutarate to the tricarboxylic acid (TCA) cycle. Importantly, while BCH mimics the effect of glucose on the KATP channel–independent pathway, it does so without having any effect on the oxidation of palmitate and without any detectable effect on lipid esterification (Liu, Cheng, Sharp, and Straub, in press, 2003).
A second hypothesis holds that in the same manner that the ATP/ADP ratio controls the KATP channel–dependent pathway, it may well be involved in the control of the KATP channel–independent pathway. In such a situation, these two synergistic pathways would be perfectly coordinated (57,58,59). Certainly there is a correlation between the rate of insulin secretion and changes in adenine and guanine nucleotide concentrations (57,58,59). Glucose induces concentration-dependent increases in the ATP/ADP and GTP/GDP ratios. The rate of insulin secretion is inversely correlated with ADP and GDP levels and positively with the ATP/ADP and GTP/GDP ratios. Because the ATP/ADP ratio determines the KATP channel–dependent pathway, the possibility exists that adenine and guanine
P.10

nucleotides control both glucose-signaling pathways. Such regulation of both pathways by the same adenine nucleotide concentrations would provide perfect coordination between them. However, as for all the other hypotheses raised thus far, there is evidence against this hypothesis also (60). Exposure of pancreatic islets to leucine for 24 hours results in an impaired insulin secretory response to glucose that is associated with increased ADP levels. There was a 50% decrease in the ATP/ADP ratio in the presence of glucose relative to control islets. However, under the same conditions, when the KATP channel–independent pathway was examined in the presence of diazoxide and KCl, the response was not changed (60). Thus, the KATP channel-independent pathway is unaffected by the decrease in the ATP/ADP ratio. Therefore, it appears unlikely that the ATP/ADP ratio has any major role in the control of the KATP channel–independent pathway. Additionally, there is evidence against a role for both GTP and the GTP/GDP ratio. When the GTP requirements of the Ca2+-dependent and the Ca2+-independent KATP channel–independent pathways were compared, mycophenolic acid, which blocks the synthesis of GTP, decreased the GTP content of islets by 40%. Under these conditions the Ca2+-dependent KATP channel–independent pathway was unaffected (40). In contrast, the Ca2+-independent KATP channel–independent pathway was eliminated. Mycophenolic acid also decreased the GTP/GDP ratio in islets, and again the KATP channel–independent augmentation pathway was intact (61). Thus, neither ATP nor GTP, nor the ATP/ADP and GTP/GDP ratios, play a major role in the augmentation pathway.
A third hypothesis holds that a glucose-induced increase in the export of mitochondrial glutamate increases cytosolic glutamate and sensitizes the granules so that they more readily undergo exocytosis (62,63,64,65). This hypothesis is also controversial (66,67,68,69). The initial data were the report that glucose generated glutamate from the mitochondria, that a membrane-permeant glutamate analogue enhanced the secretory response to glucose, and that permeabilized INS-1 cells responded to glutamate with increased Ca2+-stimulated insulin secretion (62). Also, inhibitors of vesicular glutamate transport reduced the effect of glutamate. Thus, glutamate produced by glucose is presumed to exit the mitochondria and enter the granules to increase the likelihood of their release. Further evidence in favor of the hypothesis is that overexpression of glutamate decarboxylase (GAD) in INS-1 cells to reduce cytosolic glutamate levels reduced the secretory response to glucose without affecting the response to KCl. Similar results were obtained in perfused rat pancreatic islets (63). Despite these data, there is considerable evidence against this hypothesis. Others report that glucose does not increase the levels of intracellular glutamate (66,67,68,69) and that glucose can suppress glutamate levels (69). Glutamine increases β-cell glutamate levels up to 10-fold without stimulating or augmenting insulin release (66). Dimethyl glutamate similarly has no effect on secretion, as already reported (66,69) However, the fuel effect of both glutamine and dimethyl glutamate can be detected when GDH is activated by leucine or BCH (69).
A fourth hypothesis holds that glucose activates protein acyl transferase activity. The potential role of protein acylation in the second phase of glucose-stimulated insulin secretion came from studies on isolated rat pancreatic islets. The protein acylation inhibitor cerulenin had no effect on the rate of basal insulin secretion but profoundly inhibited both phases of glucose-stimulated insulin secretion (70,71). The effect to inhibit the first phase of release suggested that the operation of either the KATP channels or voltage-dependent Ca2+ channels had been compromised. To determine if this was the case, the effect of cerulenin on the response to a depolarizing concentration of KCl was studied. Cerulenin had no effect on the response to KCl. The Ca2+ channels are operating normally in response to depolarization, and the subsequent secretory response to elevated [Ca2+]i is also unaffected. Therefore, it was possible to study the effect of cerulenin on the KATP channel–independent pathway in isolation by activating the KATP channels with diazoxide, depolarizing the cell with KCl, and then measuring the effect of glucose on insulin release. Under these conditions, as anticipated, the response to KCl in the presence of diazoxide was unaffected. However, the effect of glucose was inhibited. Thus, cerulenin inhibits the KATP channel–independent pathway of glucose signaling. More directly, the KATP channel–independent effect of palmitate was studied under the same conditions. The response to palmitate also was completely blocked by cerulenin (70). Because neither glucose oxidation nor the effect of glucose to inhibit fatty acid oxidation is affected by cerulenin (71), these data strongly suggest that protein acylation is involved in the KATP channel–independent pathway of glucose signaling and the second phase of the response. Caution has to be heeded to possible other specific effects (by inhibition of acylation) or nonspecific effects that cerulenin might exert in the β-cell, where it clearly inhibits the response to glucose. Additional potential targets for inhibition by cerulenin are enzymatic activities (e.g., fatty acid synthase) (72,73), although its activity is very low in the β-cell (74) and acetyl CoA carboxylase (75). Whereas an inhibition of acetyl CoA carboxylase cannot be excluded as a cause of the inhibition of the response to glucose, the suppression of fatty acid oxidation by glucose was not affected by cerulenin (71). Also, the enzyme was bypassed in experiments in which palmitate was used as the stimulus both in the presence and the absence of Ca2+ and in which cerulenin still exerted its full inhibitory capacity (70,71). Protein acylation is a posttranslational event that usually links palmitate in the form of a fatty acyl-CoA, as the preferred substrate, to a cysteine residue through a thioester linkage in a variety of proteins (e.g., Gα subunits), ras (76), and Ca2+ channels (77), and it has only recently been established that an enzymatic activity (protein S-acyltransferase) is likely to represent the predominant mechanism for thioacylation (78). Because it is a reversible modification with dynamic cycles of acylation and deacylation, it is eminently capable of playing a role in signal transduction. One possible target for acylation has been recently identified as a PKC of 80 kD whose translocation to membrane bilayers was facilitated upon palmitoylation (79). Moreover, acylation has also been shown to occur on proteins directly linked to exocytosis like synaptotagmin (80), which is a putative Ca2+ sensor, and SNAP-25 (81), a component of the SNARE complex, and it therefore might reflect a distal event in secretion. This assumption is supported by a study in HIT-T15 cells, where an enhancement of Ca2+-induced insulin secretion by long-chain acyl-CoAs was reported (54). However, nonspecific
P.11

interactions of acyl-CoAs at high concentrations with cell membranes and subsequent facilitation of fusion processes cannot yet be excluded.
In summary, the mechanisms by which glucose activates the KATP channel–independent signaling pathways are still unknown. However, as proposed earlier in this review, their primary action is exerted on the pool of readily releasable docked granules where they increase the rate at which the granules are transferred to the immediately releasable pool. Of the mechanisms proposed as being responsible for these changes, we do not favor the hypotheses that glutamate or adenine and guanine nucleotides are directly involved. Similarly, we do not see a role for decreased fatty acid oxidation and increased cytosolic acyl-CoA. However, we do think that anaplerosis is critical to the KATP channel–independent signaling pathways and that there is a possible role for protein acylation.
Time-Dependent Potentiation
In addition to the effect of glucose to stimulate insulin secretion, glucose imprints a memory into the β-cell that persists after the glucose stimulus has been removed and that affects subsequent secretory responses. The phenomenon, first identified in 1969 by Grodsky et al. (82), is termed priming or TDP. Since then, TDP has been demonstrated in isolated rat islets and perfused rat pancreata (83,84,85,86), rabbits (87), spiny mice (88), and humans (89,90,91). A demonstration of TDP in rat islets is illustrated in Fig. 1.7. In this study, pancreatic islets were exposed to two pulses of 16.7 mM glucose, separated by a rest period of 30 minutes in 2.8 mM glucose. Clearly, the second response to glucose is larger than the first. Similar TDP can be induced by several nutrients, including leucine, α-ketoisocaproate (KIC), and D-glyceraldehyde (92,93,94,95,96). As might be expected, the concentration of glucose used, the duration of the first exposure to glucose, and the interval between the two pulses of glucose are all-important determinants of the second response (82,83). In addition, TDP is critically dependent on the intracellular pH, the priming effect being inhibited by elevated pH and enhanced by low pH (95). It is noteworthy that both TDP and the rising second phase of glucose-induced insulin release (which reflects the KATP channel–independent augmentation pathway) are prominent in the rat and absent in the mouse (97,98).
In some cases of non–insulin-dependent diabetes mellitus (NIDDM), when the insulin response to glucose is impaired, TDP remains and can improve the response (88,90). In other cases, both the first phase response and TDP are impaired (99,100,101,102). The fact that the KATP channel–independent pathway and TDP have several similarities suggests that they may have similar mechanisms. It is known that metabolism of glucose is required for both. TDP is generated by glucose even when the secretory response is blocked by agents such as somatostatin (85,103) or diazoxide (46,103) or by the absence of extracellular Ca2+ (96). Thus, the induction of TDP is independent of the KATP channels, as is glucose-induced augmentation of insulin release. As already mentioned, the rising second- phase response that is due to the KATP channel–independent pathway is seen in rats and humans but is not seen in mice (97,98), indicating that augmentation and TDP may involve similar mechanisms. The knowledge currently available on the mechanisms of TDP include the facts that (a) TDP provides a general enhancement of the secretory response due to different secretory pathways; (b) TDP is not dependent on insulin biosynthesis, KATP channel function, or elevation of cAMP; (c) TDP is critically dependent on [Ca2+]i and/or [H+]i; and (d) TDP may be induced by anaplerosis.
The Relationship Between the KATP Channel–Independent Pathways of Glucose Signaling, the Second Phase of Glucose-Stimulated Insulin Secretion, and Time-Dependent Potentiation
It is now thought that the KATP channel–independent pathway is largely responsible for the second phase of glucose-stimulated insulin secretion. Furthermore, TDP appears to be responsible for the rising portion of the second phase that is seen in rats and humans but appears not to be involved in mice. Thus, there are at least two types of second phase, and the underlying mechanisms have to be carefully defined. In the rat and human, the rate of insulin release during the second phase increases over time from a nadir after the first phase to reach a higher plateau. In the mouse, after the first phase has declined, the second phase is essentially a plateau from that point on (97,98) so that the rising phase is slight or nonexistent. Similarly, TDP is pronounced in rats and humans and slight or nonexistent in mice. Recent work has shown a close association between the nutrients that induce TDP and those that support the KATP channel–independent pathway in rats (Gunawardana, Liu, Cheng, Sharp, and Straub, unpublished observations). All these findings have led to the concept that the KATP channel–independent pathway is largely responsible for the second phase and that TDP is an additional mechanism that is responsible for the rising portion of the second phase.
Conclusion
Glucose stimulates insulin secretion through multiple signaling pathways. In the KATP channel–dependent pathway, a glucose-induced increase in the ATP/ADP ratio is involved in closure of the KATP channels and depolarization of the β-cell. The resulting increase in Ca2+ entry and rise in [Ca2+]i triggers the stimulation of insulin release. The KATP channel–independent pathways act in synergy with the KATP channel–dependent pathway to augment the secretory responses to increased [Ca2+]i. The signals underlying this effect are derived from anaplerosis and may involve increased activity of protein acyl transferase. Increased [Ca2+]i, resulting from activation of the KATP channel–dependent pathway, discharges an immediately releasable pool of insulin-containing granules that comprise the first phase of glucose-stimulated insulin release. The rate-limiting step for this lies between Ca2+ sensing and exocytosis. The rate is fast and the result is a spike of insulin release. The second phase of glucose-stimulated insulin secretion is due to the KATP channel–independent pathways. The rate-limiting step here is in the conversion of the already primed, readily releasable
P.12

granules to the immediately releasable state. In the rat and human β-cell, the KATP channel–independent pathways induce a time-dependent potentiation of the rate of this step that results in the typical rising second-phase response.
Figure 1.7. Glucose-induced time-dependent potentiation (TDP) of insulin secretion. The data shown are derived from paired batches of rat pancreatic islets. All islets were perfused with Krebs-Ringer bicarbonate buffer (KRB) containing 2.8 mM glucose for 35 minutes before one of the pairs was challenged with 16.7 mM glucose (at the 30-minute point). After exposure to 16.7 mM glucose for 15 minutes, the islets were returned to KRB containing 2.8 mM glucose for a period of 30 minutes. At this time (the 80-minute point) both the control and test islets were exposed to 16.7 mM glucose until the end of the experiment. The test islets responded to the final glucose challenge with a larger first phase of release and a more rapid increase to the plateau of the second phase (i.e., they demonstrated TDP). The control islets, exposed to glucose only during the second challenge, subsequently reach the same rate of insulin release during the second phase as the test islets, due to simultaneous stimulation and TDP by glucose. The results are presented as means ± SEM from five separate experiments.
Acknowledgments
The authors’ work in this area is supported by a Career Development Award from the Juvenile Diabetes Research Foundation International (to S.G.S.), and by National Institutes of Health Grants DK54243 and DK56737 (to G.W.G.S.).
References
1. Simpson RG, Benedetti A, Grodsky GM, et al. Early phase of insulin release. Diabetes 1968;17:684.
2. Cerasi E, Luft R, Efendic S. Decreased sensitivity of the pancreatic β cells to glucose in prediabetic and diabetic subjects: a glucose dose-response study. Diabetes 1972;21:224.
3. Hales CN, Greenwood FC, Mitchell FL, et al. Blood-glucose, plasma-insulin and growth hormone concentrations of individuals with minor abnormalities of glucose tolerance. Diabetologia 1968; 4:73.
4. Turner RC, Hattersley AT, Shaw JTE, et al. Type II diabetes: clinical aspects of molecular biological studies. Diabetes 1995;44:1.
5. O’Connor MDL, Landahl H, Grodsky GM. Comparison of storage- and signal-limited models of pancreatic insulin secretion. Am J Physiol 1980;238:R378.
6. Grodsky GM. Threshold distribution hypothesis for packet storage of insulin and its mathematical modeling. J Clin Invest 1972;51: 2047.
7. Cerasi E, Fick G, Rudemo M. A mathematical model for the glucose-induced insulin release in man. Eur J Clin Invest 1974;4:267.
8. Porte D, Pupo AA. Insulin response to glucose: evidence for a two pool system in man. J Clin Invest 1969;48:2309.
9. Proks P, Eliasson L, Ammala C, et al. Ca2+- and GTP-dependent exocytosis in mouse pancreatic β-cells involves both common and distinct steps. J Physiol 1996;496:255.
10. Bratanova-Tochkova TK, Cheng H, Daniel S, et al. Triggering and augmentation mechanisms, granule pools and biphasic insulin secretion. Diabetes 2002;51(suppl):83.
11. Barg S, Eliasson L, Renstrom E, et al. A subset of 50 secretory granules in close contact with L-type Ca2+ channels accounts for first-phase insulin secretion in mouse β-cells. Diabetes 2002;51(suppl): 74.
12. Olofsson CS, Gopel SO, Barg S, et al. Fast insulin secretion reflects exocytosis of docked granules in mouse pancreatic B-cells. Pflugers Arch 2002;444:43.
13. Gold G, Landahl HD, Gishizky ML, et al. Heterogeneity and compartmental properties of insulin storage and secretion in rat islets. J Clin Invest 1982;69:554.
14. Halban P. Differential rates of release of newly synthesized and of stored insulin from pancreatic islets. Endocrinology 1982;110:1183.
15. Wang SY, Halban PA, Rowe JW. Effects of aging on insulin synthesis and secretion: differential effects on preproinsulin messenger RNA levels, proinsulin biosynthesis, and secretion of newly made and preformed insulin in the rat. J Clin Invest 1988;81:176.
16. Dean PM. Ultrastructural morphometry of the pancreatic β-cell. Diabetologia 1973;9:115.
17. Grodsky GM. A threshold distribution hypothesis for packet storage of insulin: II. Effect of calcium. Diabetes 1972;21(suppl 2):584.
18. Eliasson L, Renstrom E, Ding WG, et al. Rapid ATP-dependent priming of secretory granules precedes Ca2+-induced exocytosis in mouse pancreatic β-cells. J Physiol 1997;503:399.
19. Takahashi N, Kadowaki T, Yazaki Y. Post-priming actions of ATP on Ca2+-dependent exocytosis in pancreatic β-cells. Proc Natl Acad Sci USA 1999;96:760.
20. Sharp GWG. The adenylate cyclase-cAMP system in islets of Langerhans and its role in the control of insulin release. Diabetologia 1979;16:287.
21. Renstrom E, Eliasson L, Rorsman P. Protein kinase A–dependent and –independent stimulation of exocytosis by cAMP in mouse pancreatic β-cells. J Physiol 1997;502:105.
22. Ammala C, Ashcroft FM, Rorsman P. Calcium-independent potentiation of insulin release by cyclic AMP in single β-cells. Nature 1993;363:356.
23. Corkey BE, Glennon MC, Chen KS, et al. A role for malonyl CoA in glucose-stimulated insulin secretion from clonal pancreatic β-cells. J Biol Chem 1989;264:21608.
24. Prentki M, Vischer S, Glennon MC, et al. Malonyl CoA and long chain acyl CoA esters as metabolic coupling factors in nutrient-induced insulin secretion. J Biol Chem 1992;267:5802.
25. Prentki M. New insights into pancreatic β-cell metabolic signaling in insulin secretion. Eur J Endocrinol 1996;134:272.
26. Corkey BE, Deeney JT, Yaney GC, et al. The role of long-chain fatty acyl-CoA esters in β-cell signal transduction. J Nutr 2000;130 (suppl):299.
27. Gilon P, Henquin JC. Mechanisms and physiological significance of the cholinergic control of pancreatic β-cell function. Endocr Rev 2001;22:565.
P.13

28. Cook DL, Hales CN. Intracellular ATP directly blocks K+ channels in pancreatic β-cells. Nature 1984;311:271.
29. Ashcroft FM, Harrison DE, Ashcroft SJH. Glucose induces closure of single potassium channels in isolated rat pancreatic β-cells. Nature 1984;312:446.
30. Wollheim CB, Sharp GWG. The regulation of insulin release by calcium. Physiol Rev 1981;61:914.
31. Hoenig M, Sharp GWG. Glucose induces insulin release and a rise in cytosolic calcium concentration in a transplantable rat insulinoma. Endocrinology 1986;119:2502.
32. Henquin J-C. Triggering and amplifying pathways of regulation of insulin secretion by glucose. Diabetes 2000;49:1751.
33. Sato Y, Aizawa T, Komatsu M, et al. Dual functional role of membrane depolarization/Ca2+ flux in rat pancreatic β-cell. Diabetes 1992;41:438.
34. Gembal M, Gilon P, Henquin JC. Evidence that glucose can control insulin release independently from its action on ATP-sensitive K+ channels in mouse B cells. J Clin Invest 1992;89:1288.
35. Best L, Yates AP, Tomlinson S. Stimulation of insulin secretion by glucose in the absence of diminished 86Rb+ permeability. Biochem Pharmacol 1992;43:2483.
36. Aizawa T, Komatsu M, Asanuma N, et al. Glucose action “beyond ionic events” in the pancreatic β-cell. Trends Pharmacol Sci 1998; 19:496.
37. Komatsu M, Schermerhorn T, Aizawa T, et al. Glucose stimulation of insulin release in the absence of extracellular Ca2+ and in the absence of any rise in intracellular Ca2+ in rat pancreatic islets. Proc Nat Acad Sci U S A 1995;92:10728.
38. Komatsu M, Schermerhorn T, Noda M, et al. Augmentation of insulin release by glucose in the absence of extracellular Ca2+. New insights into stimulus-secretion coupling. Diabetes 1997;46:1928.
39. Komatsu M, Schermerhorn T, Straub SG, et al. Pituitary adenylate cyclase-activating peptide, carbachol and glucose stimulate insulin release in the absence of an increase in intracellular Ca2+. Mol Pharmacol 1996;50:1047.
40. Komatsu M, Noda M, Sharp GWG. The two augmentation pathways in glucose stimulus-secretion coupling, Ca2+-dependent and Ca2+-independent, can be distinguished by their requirement for GTP studies on rat pancreatic islets. Endocrinology 1998;139:1172.
41. Matschinsky FM. A lesson in metabolic regulation inspired by the glucokinase sensor paradigm. Diabetes 1996;45:223.
42. Paras CD, Qian W, Lakey JR, et al. Localized exocytosis detected by spatially resolved amperometry in single pancreatic β-cells. Cell Biochem Biophys 2000;33:227.
43. Barg S, Ma X, Eliasson L, Galvanovskis J, et al. Fast exocytosis with few Ca2+ channels in insulin-secreting mouse pancreatic B cells. Biophys J 2001;81:3308.
44. Qian WJ, Kennedy RT. Spatial organization of Ca2+ entry and exocytosis in mouse pancreatic β-cells. Biochem Biophys Res Commun 2001;286:315.
45. Wiser O, Trus M, Hernandez A, et al. The voltage sensitive Lc-type Ca2+ channel is functionally coupled to the exocytotic machinery. Proc Natl Acad Sci USA 1999;96:248.
46. Taguchi N, Aizawa T, Sato Y, et al. Mechanism of glucose-induced biphasic insulin release: physiological role of adenosine triphosphate-sensitive K+ channel-independent glucose action. Endocrinology 1995;136:3942.
47. Aizawa T, Komatsu M, Asanuma N, et al. Glucose action “beyond ionic events” in the pancreatic β-cell. Trends Pharmacol Sci 1998; 19:496.
48. Straub SG, Sharp GWG. Glucose-stimulated signaling pathways in biphasic insulin secretion. Diabetes Metab Res Rev 2002;18:451.
49. Chen S, Ogawa A, Ohneda M, et al. More direct evidence for a malonyl-CoA-carnitine palmitoyltransferase I interaction as a key event in pancreatic β-cell signaling. Diabetes 1994;43:878.
50. Antinozzi PA, Segall L, Prentki M, et al. Molecular or pharmacologic perturbation of the link between glucose and lipid metabolism is without effect on glucose-stimulated insulin secretion. A re-evaluation of the long chain acyl-CoA hypothesis. J Biol Chem 1998; 273:16146.
51. Sener A, Malaisse WJ. Hexose metabolism in pancreatic islets. Effect of (-)-hydroxycitrate upon fatty acid synthesis and insulin release in glucose-stimulated islets. Biochimie 1991;73:1287.
52. Bliss CR, Sharp GWG. Glucose-induced insulin release in islets of young rats: time-dependent potentiation and effects of 2-bromostearate. Am J Physiol 1992;263:E890.
53. Zhang S, Kim K-H. Essential role of acetyl-CoA carboxylase in the glucose-induced insulin secretion in a pancreatic β-cell line. Cell Signal 1998;10:35.
54. Deeney JT, Gromada J, Hoy M, et al. Acute stimulation with long chain acyl-CoA enhances exocytosis in insulin secreting cells (HIT T-15 and NMRI β-cells). J Biol Chem 2000;275:9363.
55. Hohmeier HE, Mulder H, Chen G, et al. Isolation of INS-1-derived cell lines with robust ATP-sensitive K+ channel-dependent and -independent glucose-stimulated insulin secretion. Diabetes 2000;49: 424.
56. Mulder H, Lu D, Finley J, et al. Overexpression of a modified human malonyl-CoA decarboxylase blocks the glucose-induced increase in malonyl-CoA level but has no impact on insulin secretion in INS-1-derived (832/13) β-cells. J Biol Chem 2001;276:6479.
57. Sato Y, Henquin J-C. The KATP channel-independent pathway of regulation of insulin secretion by glucose. In search of the underlying mechanism. Diabetes 1998;47:1713.
58. Detimary P, Van den Berghe G, Henquin JC. Concentration dependence and time course of the effects of glucose on adenine and guanine nucleotides in mouse pancreatic islets. J Biol Chem 1996;271: 20559.
59. Detimary P, Dejonghe S, Ling Z, et al. The changes in adenine nucleotides measured in glucose-stimulated rodent islets occur in β cells but not in α cells and are also observed in human islets. J Biol Chem 1998;273:33905.
60. Anello M, Ucciardello V, Piro S, et al. Chronic exposure to high leucine impairs glucose-induced insulin release by lowering the ATP-to-ADP ratio. Am J Physiol 2001;281:E1082.
61. Meridith M, Rabaglia M, Metz S. Evidence of a role for GTP in the potentiation of Ca2+-induced insulin secretion from intact rat islets. J Clin Invest 1995;96:811.
62. Maechler P, Wollheim CB. Mitochondrial glutamate acts as a messenger in glucose-induced insulin exocytosis. Nature 1999;402:685.
63. Rubi B, Ishihara H, Hegardt FG, et al. GAD65-mediated glutamate decarboxylation reduces glucose-stimulated insulin secretion in pancreatic β cells. J Biol Chem 2001;276:36391.
64. Maechler P, Wollheim CB. Mitochondrial signals in glucose-stimulated insulin secretion in the β-cell. J Physiol 2000;529:49.
65. Hoy M, Maechler P, Efanov AM, et al. Increase in cellular glutamate levels stimulates exocytosis in pancreatic β-cells. FEBS Lett 2002; 531;199.
66. Yamada S, Komatsu M, Sato Y, et al. Glutamate is not a major conveyer of ATP-sensitive K+ channel-independent glucose action in pancreatic islet β-cell. Endocr J 2001;48;391.
67. MacDonald MJ, Fahien LA. Glutamate is not a messenger in insulin secretion. J Biol Chem 2000;275:712.
68. Bertrand G, Ishuama N, Nenquin N, et al. The elevation of glutamate content and the amplification of insulin secretion in glucose-stimulated pancreatic islets are not causally related. J Biol Chem 2002;277:32883.
69. Li C, Najafi H, Daikhin Y, Nissim I, et al. Regulation of leucine stimulated insulin secretion and glutamine metabolism in isolated rat islets. J Biol Chem 2003;278:2853.
70. Straub SG, Yajima H, Komatsu M, et al. The effects of cerulenin, an inhibitor of protein acylation, on the two phases of glucose-stimulated insulin secretion. Diabetes 2002;51(suppl):91.
71. Yajima H, Komatsu M, Yamada S, et al. Cerulenin, an inhibitor of protein acylation, selectively attenuates nutrient stimulation of insulin release. A study in rat pancreatic islets. Diabetes 2000;49:712.
72. Kuhajda FP, Jenner K, Wood FD, et al. Fatty acid synthesis: a potential selective target for antineoplastic therapy. Proc Natl Acad Sci USA 1994;91:6379.
P.14

73. Pizer ES, Thupari J, Han WF, et al. Malonyl-coenzyme-A is a potential mediator of cytotoxicity induced by fatty-acid synthase inhibition in human breast cancer cells and xenografts. Cancer Res 2000; 60:213.
74. Brun T, Roche E, Kim KH, et al. Glucose regulates acetyl-CoA carboxylase gene expression in a pancreatic β-cell line (INS-1). J Biol Chem 1993;268:18905.
75. Jochen AL, Hays J, Mick G. Inhibitory effects of cerulenin on protein palmitoylation and insulin internalization in rat adipocytes. Biochem Biophys Acta 1995;1259:65.
76. Resh MD. Fatty acylation of proteins: new insights into membrane targeting of myristoylated and palmitoylated proteins. Biochim Biophys Acta 1999;1451:1.
77. Hurley JH, Cahill AL, Currie KP, et al. The role of dynamic palmitoylation in Ca2+ channel inactivation. Proc Natl Acad Sci USA 2000;97:9293.
78. Dunphy JT, Schroeder H, Leventis R, et al. Differential effects of acyl-CoA binding protein on enzymatic and non-enzymatic thioacylation of protein and peptide substrates. Biochem Biophys Acta 2000;1485:185.
79. Ford DA, Horner CC, Gross RW. Protein kinase C acylation by palmitoyl coenzyme A facilitates its translocation to membranes. Biochemistry 1998;37:11953.
80. Chapman ER, Blasi J, An S, et al. Fatty acylation of synaptotagmin in PC12 cells and synaptosomes. Biochem Biophys Res Commun 1996; 225:326.
81. Gonzalo S, Greentree WK, Linder ME. SNAP-25 is targeted to the plasma membrane through a novel membrane-binding domain. J Biol Chem 1999;274:21313.
82. Grodsky GM, Curry D, Herbert L, et al. Further studies on the dynamic aspects of insulin release in vitro with evidence for a two-compartmental storage system. Acta Diabetol Latina 1969;6(suppl 1): 554.
83. Grill V, Rundfeldt M. Effects of priming with D-glucose on insulin secretion from rat pancreatic islets: increased responsiveness to other secretagogues. Endocrinology 1979;105:980.
84. Ashby JP, Shirling D. The priming effect of glucose on insulin secretion from isolated islets of Langerhans. Diabetologia 1981;21:230.
85. Grill V, Adamson U, Cerasi E. Immediate and time-dependent effects of glucose on insulin release from rat pancreatic tissue: evidence for different mechanisms of action. J Clin Invest 1978;61:1034.
86. Grill V, Rundfeldt M. Effects of priming with D-glucose on insulin secretion from rat pancreatic islets: increased responsiveness to other secretagogues. Endocrinology, 1979;105:980.
87. Fichaux F, Bonnafous R. Responsiveness and memory of the pancreatic B-cells to the insulin secretagogues D-glucose and L-arginine in prediabetic and diabetic rabbits. Pancreas 1992;7:585.
88. Nesher R, Abramovitch E, Cerasi E. Correction of diabetic pattern of insulin release from islets of the spiny mouse (Acomys cahirinus) by glucose priming in vivo. Diabetologia 1985;28:233.
89. Cerasi E. Potentiation of insulin release by glucose in man. II. Role of the insulin response, and enhancement of stimuli other than glucose. Acta Endocrinol 1975;79:502.
90. Cerasi E. Potentiation of insulin release by glucose in man. III. Normal recognition of glucose as a potentiator in subjects with low insulin response and in mild diabetics. Acta Endocrinol 1975;79:511.
91. Cerasi E. Potentiation of insulin release by glucose in man. I. Quantitative analysis of the enhancement of glucose-induced insulin secretion by pretreatment with glucose in normal subjects. Acta Endocrinol 1975;79:483.
92. Grill V, Adamson U, Cerasi E. Immediate and time-dependent effects of glucose on insulin release from rat pancreatic tissue: evidence for different mechanisms of action. J Clin Invest 1978;61:1034.
93. Grill V. Nutrient-induced priming of insulin and glucagon secretion. Effect of α-ketoisocaproic acid. Endocrinology 1982;110:1013.
94. Grill V, Adamson U, Rundfeldt M, et al. Glucose memory of pancreatic B and A2 cells: evidence for common time-dependent actions of glucose on insulin and glucagon secretion in the perfused rat pancreas. J Clin Invest 1979;700.
95. Zawalich WS. Time-dependent potentiation of insulin release induced by α-ketoisocaproate and leucine in rats: possible involvement of phosphoinositide hydrolysis. Diabetologia 1988;31:435.
96. Gunawardana SC, Sharp GWG. Intracellular pH plays a critical role in glucose-induced time-dependent potentiation of insulin release in rat islets. Diabetes 2002;51:105.
97. Berglund O. Different dynamics of insulin secretion in the perfused pancreas of mouse and rat. Acta Endocrinol 1980;93:54.
98. Zawalich WS, Zawalich KC. Species differences in the induction of time-dependent potentiation of insulin secretion. Endocrinology 1996;137:1664.
99. Grill V, Rundfeldt M. Abnormalities of insulin responses after ambient and previous exposure to glucose in streptozotocin-diabetic and dexamethazone-treated rats: role of hyperglycemia and increased B-cell demands. Diabetes 1986;35:44.
100. Grill V, Efendic S. Loss of priming effect of glucose on A and D cell secretion in perfused pancreas from alloxan-diabetic rats: role of insulin and alloxan. Diabetologia 1983;24:47.
101. Nadi AB, Malaisse WJ. Altered metabolic priming by D-glucose in pancreatic islets from Goto-Kakizaki rats. Int J Mol Med 2000;5:625.
102. Efendic S, Lins PE, Cerasi E. Potentiation and inhibition of insulin release in man following priming with glucose and with arginine—effect of somatostatin. Acta Endocrinol 1979;90:259.
103. Taguchi N, Aizawa T, Hashizume K. ATP-sensitive K+ channel closure is not an obligatory step for glucose-induced priming of pancreatic B-cell. Adv Exp Med Biol 1997;426:29.
←↑→