Inositol hexakisphosphate stimulates non-Ca²⁺-mediated and primes Ca²⁺-mediated exocytosis of insulin by activation of protein kinase C

Abstract

d-myo-inositol 1,2,3,4,5,6-hexakisphosphate (InsP₆), formed via complex pathways of inositol phosphate metabolism, composes the main bulk of inositol polyphosphates in the cell. Relatively little is known regarding possible biological functions for InsP₆. We now show that InsP₆ can modulate insulin exocytosis in permeabilized insulin-secreting cells. Concentrations of InsP₆ above 20 μM stimulated insulin secretion at basal Ca²⁺-concentration (30 nM) and primed Ca²⁺-induced exocytosis (10 μM), both effects being due to activation of protein kinase C. Our results suggest that InsP₆ can play an important modulatory role in the regulation of processes such as exocytosis in insulin-secreting cells. The specific role for InsP₆ can then be to recruit secretory granules to the site of exocytosis.

Highly phosphorylated inositol phosphates are the major inositol phosphates present in the cell (1). Specifically, the intracellular concentration of d-myo-inositol 1,2,3,4,5,6-hexakisphosphate (InsP₆) reaches levels corresponding to 40–60 μM in several cell types, including insulin-secreting cells (2–4). In some cell types the concentration of these inositol polyphosphates can be rapidly altered upon cell stimulation (5, 6).

Inositol polyphosphates may be involved in the control of several cellular functions (1), one of which is regulation of vesicle trafficking, processes of endocytosis and vesicle recycling. d-myo-inositol 1,3,4,5-tetrakisphosphate, d-myo-inositol 1,3,4,5,6-pentakisphosphate (InsP₅) and InsP₆ bind to the C2B domain of synaptotagmin (7), a protein believed to be a main Ca²⁺-sensor in regulated exocytosis in neurons. The injection of inositol polyphosphates into squid giant synapse preterminal caused blocking of synaptic transmission, due to binding of injected compounds to the C2B domain of synaptotagmin (8). Inositol polyphosphates binding to clathrin assembly proteins, adaptor proteins AP-2 (9), and AP-3 (10) inhibit clathrin assembly, suggesting the influence of the compounds upon vesicle trafficking (10).

The aim of the present study was to evaluate a possible role for InsP₆ in the regulation of insulin exocytosis.

MATERIALS AND METHODS

Materials.

InsP₅ and InsP₆ were from Calbiochem–Novabiochem. Streptolysin O was obtained from Difco. All other reagents were from Sigma.

Cell Culture.

Hamster insulinoma tumor HIT T15 cells were cultured at 37°C in RPMI medium 1640 containing 11 mM glucose and supplemented with 10% (vol/vol) heat-inactivated fetal calf serum, 2 mM glutamine, 100 μg/ml streptomycin, 100 units/ml penicillin, 10⁻⁷ M selenous acid, and 10 μg/ml glutathione, as described (11). All experiments were carried out in HIT T15 cells of passage numbers between 73–80.

Cell Permeabilization and Measurements of Insulin Secretion.

Two different procedures were used to permeabilize cells, namely permeabilization by electric field and streptolysin O-induced permeabilization. The permeabilization buffer consisted of 140 mM potassium glutamate, 5 mM NaCl, 1.2 mM MgCl₂, 10 mM EGTA, 25 mM Hepes, 0.025% albumin (pH 7.0). For streptolysin O-induced permeabilization, 1 mM dithiothreitol was added to the buffer. The electropermeabilization and permeabilization with streptolysin O were performed as described (12, 13).

Both types of permeabilization have their own benefits and drawbacks. Electropermeabilization creates only small holes in the plasma membrane, that prevents leakage of molecules >5 kDa from the cell. These holes reseal with time (14). By contrast, cells permeabilized with streptolysin O are leaking not only small molecules but also large proteins. Cells permeabilized by this procedure do not reseal (15). The use of one or the other of the procedures was dependent on the aim of each particular experiment.

Measurements of insulin secretion were performed in a permeabilization buffer containing additionally 2 mM MgATP and an ATP-regenerating system, 2 mM creatine phosphate, 10 unit/ml creatine phosphokinase, and Ca²⁺ concentrations in the range from 3 × 10⁻⁸ M to 1 × 10⁻⁵ M. Priming effects of InsP₆ on insulin secretion were measured as follows. Cells were treated with different concentrations of InsP₆ for 15 min in ATP-containing buffer and then insulin secretion was measured in ATP-free buffer without InsP₆ at 30 nM or at 10 μM Ca²⁺ during 15 min. Insulin content was determined by radioimmunoassay using rat insulin as a standard. The level of insulin secretion at 30 nM Ca²⁺, ranging from 30–50 μunits/10⁵ cells/15 min in different experiments, was taken as 100%.

Free Ca²⁺-concentration in solutions was measured with a Ca²⁺-selective electrode (model 93-20, Orion, Boston). As standard solutions, Ca²⁺-buffers (World Precision Instruments, Sarasota, FL) with known Ca²⁺ concentrations were used.

Measurements of Protein Kinase C (PKC) Activity.

PKC activity was assayed by measuring the transfer of ³²P to a specific enzyme substrate. PKC Assay system (Promega) provided high specificity of the assay by using biotinylated peptide substrate and streptavidin-coated filters. Briefly, cells were washed several times in PBS buffer, resuspended in extraction buffer (25 mM Tris⋅HCl, pH 7.4/0.5 mM EDTA/0.5 mM EGTA/10 mM 2-mercaptoethanol/1 μg/ml leupeptin/1 μg/ml aprotinin/0.5 mM phenylmethylsulfonyl fluoride) and sonicated. Enzyme activity was assayed in the presence of phospholipids (0.35 mg/ml phosphatidylserine and 0.035 mg/ml diacylglycerol) and either without (2.5 mM EGTA and 0.1 mM EDTA) or with Ca²⁺ (4.5 mM CaCl₂, 2.5 mM EGTA, and 0.1 mM EDTA) for 8 min at 30°C. Reactions were started by addition [γ-³²P]ATP and terminated by spotting on streptavidin-coated filters. Incorporation of ³²P into biotinylated substrate was assessed by liquid scintillation counting.

Presentation of Results.

Data analysis was performed using the program sigma plot 1.02 for Windows (Jandel, San Rafael, CA). All results are expressed as means ± SEM for the indicated number of observations. Statistical significance of differences between means were assessed by Student’s unpaired t test. Differences were considered significant at P < 0.05.

RESULTS

The effects of 20 μM InsP₅ and 20 μM InsP₆ on insulin secretion were tested at free Ca²⁺-concentrations ranging from 30 nM to 10 μM in electropermeabilized HIT T15 cells (Fig. 1). Only at basal (30 nM) free Ca²⁺ concentration, both compounds stimulated exocytosis of insulin. The stimulatory effect could not be detected at higher Ca²⁺-concentrations. Since InsP₆ is present at high concentrations in insulin-secreting cells (4), this inositol polyphosphate was chosen for the investigation of the mechanism underlying the stimulatory effects of inositol polyphosphates on insulin secretion.

Figure 1.

Effects of 20 μM InsP₅ (A) and 20 μM InsP₆ (B) on insulin secretion in electropermeabilized HIT T15 cells at different free Ca²⁺-concentrations. Secretion without inositol polyphosphates is denoted by □ and with inositol polyphosphates by ▪. Data represent mean ± SEM for 16 observations from three separate experiments. ∗∗, P < 0.01 and ∗∗∗, P < 0.001, relative to insulin secretion without inositol polyphosphates.

Stimulation of non-Ca²⁺-mediated insulin secretion by InsP₆ was concentration-dependent and at 50 μM InsP₆, insulin release reached ≈150% of secretion obtained in the absence of inositol polyphosphate (Fig. 2A). The stimulatory effects of InsP₆ on insulin secretion could not be detected when the Ca²⁺-concentration was increased (Fig. 2B). To check if the observed effects of InsP₆ were due to possible changes in free Ca²⁺-concentration induced by the compound, measurements of the ambient free Ca²⁺-concentration in the permeabilization buffer were performed (Fig. 2 C and D). At a concentration of 20 μM, InsP₆ did not alter resting Ca²⁺-concentration (30 nM) (Fig. 2C). Under this condition 50 μM of InsP₆ only reduced resting Ca²⁺-concentration by 20%. InsP₆ (20 and 50 μM) decreased stimulating Ca²⁺-concentration not more than 7–10% (Fig. 2D). Although InsP₆ has a well-documented ability to chelate Ca²⁺-ions (16), the Ca²⁺-buffers used in this study have enough capacity to prevent significant changes in free Ca²⁺-concentration in the presence of InsP₆. Another Ca²⁺-chelator, 100 μM EGTA, added to the incubation buffer produced analogous slight decrease in free Ca²⁺-concentration but did not significantly affect insulin secretion measured at 30 nM Ca²⁺ (data not shown). Thus, the effects of InsP₆ on insulin secretion cannot be accounted for by changes in the free Ca²⁺-concentration.

Figure 2.

Concentration-dependent effects of InsP₆ on insulin secretion in electropermeabilized cells in the presence of either 30 nM (A) or 10 μM (B) Ca²⁺ and on free Ca²⁺-concentration in the permeabilization buffer containing either ≈30 nM (C) or ≈10 μM (D) Ca²⁺. For measurements of insulin secretion, data represent mean ± SEM for 18 observations from three separate experiments. ∗∗∗, P < 0.001, relative to insulin secretion without InsP₆. For measurements of Ca²⁺, representative experiments of three are presented.

InsP₆ is not the last member in the array of inositol polyphosphates. Higher phosphorylated inositol polyphosphate pyrophosphates, PP-InsP₅ and (PP)₂-InsP₄, formed from InsP₆ and InsP₅, have also been described (17). It has been proposed that inositol polyphosphate pyrophosphates are intracellular energy stores because of the high energy hydrolysis of pyrophosphoryl residues (18). Therefore, we tested the specificity of the effect of InsP₆ on exocytosis. The process of formation of pyrophosphates can be activated by NaF, through the inhibition of phosphatases responsible for pyrophosphate dephosphorylation (17). Addition of 5 mM NaF, this concentration maximally increased PP-InsP₅ concentration but did not significantly activate G-proteins (17), instead of increasing slightly decreased InsP₆-stimulated insulin release (139 ± 3% increase in secretion with 50 μM InsP₆ vs. 127 ± 2% with 50 μM InsP₆ and NaF, 15 observations from three separate experiments). This supports the suggestion that InsP₆ rather than PP-InsP₅ or (PP)₂-InsP₄ stimulated exocytosis.

One of the reported effects of inositol polyphosphates, is activation of PKC isoenzymes as well as the PKC-related kinase, PRK1 (19). The kinase activation was observed only in a Ca²⁺-free medium and at inositol polyphosphate concentrations similar to those that stimulated insulin secretion in our experiments. Therefore, we examined the possibility that stimulation of exocytosis by inositol polyphosphates can be explained by activation of PKC. In HIT T15 cells, 50 μM InsP₆ activated PKC in the absence of free Ca²⁺ and did not significantly affect Ca²⁺-induced activation of the enzyme (Fig. 3A).

Figure 3.

Involvement of PKC activation in the stimulation of insulin secretion in HIT T15 cells by InsP₆. (A) Modulation of PKC activity by 50 μM InsP₆ (▪) in the absence and the presence of free Ca²⁺ as described in Materials and Methods. (□) The enzyme activity without InsP₆. Data represent mean ± SEM of four experiments. ∗, P < 0.05, relative to PKC activity in the absence of free Ca²⁺ and InsP₆. (B) Inhibitor of PKC, 1.5 μM Calphostin C (Calph. C), blocked increase in insulin secretion induced by 50 μM InsP₆. (C) Requirement of ATP-hydrolysis for the development of the stimulatory effect of 50 μM InsP₆ on insulin secretion. ATP (2 mM) was substituted by 2 mM AMP-PCP, a nonhydrolyzable analog of ATP. (D) Additive effect of 50 μM InsP₆ and 1 μM okadaic acid (OA) on stimulation of insulin secretion and the absence of additive effect of 50 μM InsP₆ and 100 nM TPA. In B–D insulin secretion was measured in buffer with basal Ca²⁺ concentration (30 nM). (□) Insulin secretion without InsP₆ and black columns illustrate insulin secretion with 50 μM InsP₆. The presence of other compounds is indicated in the figure. For measurements of insulin secretion, data represent mean ± SEM for 17 (B), 18 (C), and 20 (D) observations from three separate experiments. ∗∗∗, P < 0.001, relative to insulin secretion at 30 nM Ca²⁺. ###, P < 0.001, relative to insulin secretion in the presence of 50 μM InsP₆. §§§, P < 0.001, relative to insulin secretion in the presence of 1 μM OA.

The selective inhibitor of PKC, Calphostin C (1.5 μM) abolished stimulation of insulin secretion by 50 μM InsP₆ (Fig. 3B). To confirm that protein kinase activation is involved in the stimulatory effect of InsP₆ on insulin secretion, experiments substituting ATP with its non-hydrolyzable analog β,γ-methyleneadenosine 5′-triphosphate (AMP-PCP), which cannot serve as a substrate for protein kinase-induced phosphorylation, were performed. AMP-PCP blocked InsP₆-induced exocytosis (Fig. 3C). Finally, 1 μM okadaic acid, an inhibitor of protein phosphatases type 1, 2A, and 3 (20), potentiated the effect of 50 μM InsP₆ on insulin release, whereas 100 nM 12-O-tetradecanoylphorbol 13-acetate (TPA), a potent activator of PKC, did not (Fig. 3D). This suggests that activation of PKC rather than inhibition of protein phosphatases is responsible for the stimulation of exocytosis by InsP₆.

Experiments presented so far described acute effects of InsP₆ on non-Ca²⁺-mediated insulin exocytosis. To evaluate possible effects of InsP₆ on Ca²⁺-mediated exocytosis, cells were preincubated with the inositol polyphosphate in the absence of Ca²⁺ and subsequently stimulated with high concentrations of the ion. We have employed streptolysin O-permeabilized cells for this type of experiments, since the stability of the formed holes in the plasma membrane during prolonged time of incubation was essential. Permeabilized cells were first incubated in a buffer containing ATP, an ATP-regenerating system, 30 nM Ca²⁺, and different concentrations of InsP₆. This buffer was replaced by a buffer containing 30 nM or 10 μM Ca²⁺ without InsP₆ and ATP. Insulin content in the latter buffer was measured. Changing of permeabilization procedure did not perturb the ability of InsP₆ to affect insulin exocytosis. InsP₆ (20 and 50 μM) also primed non-Ca²⁺-mediated exocytosis (Fig. 4A). The effect of InsP₆ on Ca²⁺-mediated exocytosis was pronounced. In this case, both 20 μM and 50 μM of the compound stimulated insulin secretion by 45% (Fig. 4B). Calphostin C (1.5 μM) suppressed insulin secretion induced by 50 μM InsP₆ (Fig. 4C).

Figure 4.

Priming of insulin secretion by InsP₆ in streptolysin O-permeabilized cells. Cells were pretreated with or without different concentrations of InsP₆ for 15 min in ATP-containing buffer and then insulin secretion in ATP-free buffer was measured during 15 min at 30 nM (A) or at 10 μM Ca²⁺ in the absence of InsP₆ (B and C). In C, 1.5 μM Calphostin C (Calph. C) was introduced together with InsP₆. Data represent mean ± SEM for 15 observations from three separate experiments. ∗∗∗, P < 0.001, relative to insulin secretion without InsP₆ and at appropriate Ca²⁺ concentration. ###, P < 0.001, relative to insulin secretion in the presence of InsP₆.

DISCUSSION

Permeabilized insulin-secreting cells were used to investigate a possible role for InsP₆ in regulated exocytosis. InsP₆ stimulated non-Ca²⁺-mediated and primed Ca²⁺-mediated exocytosis of insulin, effects mediated through the activation of PKC. The involvement of PKC was verified by experiments showing that the InsP₆ effect was blocked with Calphostin C and substitution of ATP with the non-hydrolyzable analog AMP-PCP. The potentiation of insulin secretion by InsP₆ in the presence of 1 μM okadaic acid, a concentration known to inhibit protein phosphatases type 1, 2A, and 3 (20), suggests that the observed stimulatory activity of InsP₆ on insulin release cannot be explained by an inhibitory activity of the compound on these protein phosphatases. Activation of PKC by InsP₆ in insulin-secreting cells, in the absence of Ca²⁺, is consistent with the data that InsP₆, at concentrations higher than 20 μM, activates several isoenzymes of PKC as well as PKC-related kinase (19).

Processes of protein phosphorylation play a central role in the regulation of exocytosis in the pancreatic β-cell (21, 22). Activation of PKC by InsP₆ may lead to increased availability of releasable secretory granules, by promoting the recruitment and transport of granules to the site of exocytosis. The increasing amount of granules associated with the plasma membrane and ready-to-fuse would lead to the subsequent increase in insulin exocytosis. In addition, InsP₆ may directly affect granule fusion since the conformation of proteins responsible for fusion events in exocytosis could be controlled by PKC activity (23, 24).

The fact that the stimulatory effect of InsP₆ on insulin exocytosis disappeared at elevated Ca²⁺-concentrations, may be explained by the conversion of InsP₆ into the inactive Ca²⁺-bound form, each molecule of InsP₆ binding two or three Ca²⁺-ions (16). Such a conversion is in agreement with what has previously been suggested in the context of the disappearance of the stimulatory effect of inositol hexakisphosphate on PKC activity at high concentrations of Ca²⁺ (19). In the latter study, the authors suggested that the negative charge of phosphate groups plays a significant role in the effects of InsP₆.

As discussed above, InsP₆ readily chelates Ca²⁺-ions. This complicates any investigation of the role of the compound in Ca²⁺-mediated exocytosis. We have separated in time the preincubation of cells with InsP₆ in ATP-containing medium and subsequent incubation of cells in the buffer containing high Ca²⁺-concentration in the absence of ATP and InsP₆. The results obtained show that in HIT T15 cells Ca²⁺ triggers exocytosis of insulin in an ATP-independent manner, as described before for exocytosis of biogenic amines in PC12 cells (25–27). By applying this type of incubation procedure, we found that preincubation with InsP₆ led to a markedly pronounced stimulatory effect on Ca²⁺-induced insulin exocytosis. It is likely that InsP₆ under these conditions potentiated insulin release by priming Ca²⁺-mediated fusion of secretory granules with the plasma membrane. Hence, inositol polyphosphates may keep granule transport active under basal conditions and thereby increasing the number of granules available at the site of exocytosis, explaining both the stimulatory effect of InsP₆ on non-Ca²⁺-mediated- and the priming of Ca²⁺-mediated insulin release (Fig. 5).

Figure 5.

Scheme illustrating the role of InsP₆ in insulin exocytosis. A denotes an intracellular pool of insulin containing granules that are not activated for fusion, and B denotes a pool of granules close to the plasma membrane that undergoes Ca²⁺-activated fusion. The exocytotic process of insulin can thus be separated into ATP-dependent vesicle priming and ATP-independent Ca²⁺-activated fusion. According to the scheme the rate of insulin secretion is determined by the amount of primed vesicles and the probability of the fusion event for each granule. Through activation of PKC, InsP₆ may influence both of these steps of insulin secretion. Hence, the processes discussed in this paper may be presented in the following way. (i) Insulin secretion at resting Ca²⁺ concentration and in the absence of InsP₆. Insulin release is restricted by the amount of primed vesicles and a low probability of vesicle fusion at resting Ca²⁺ concentration. (ii) Activation of insulin secretion at resting Ca²⁺ concentration by InsP₆. InsP₆ activates PKC, which leads to an increased priming of secretory vesicles and maybe in addition direct activation of the fusion process. Modest stimulation of insulin release is observed relative to i. (iii) Ca²⁺-induced exocytosis in the absence of InsP₆. Probability of vesicle fusion is high. It gives rise to a pronounced stimulation of insulin secretion, but the stimulation is still not maximal and limited by the availability of primed vesicles. (iv) Priming of Ca²⁺-induced exocytosis by InsP₆. Similar to ii, InsP₆ increased priming of secretory vesicles, an effect mediated by PKC activation. A high probability of vesicle fusion together with an increased amount of vesicles at the site of exocytosis lead to pronounced insulin secretion.

Intracellular levels of inositol polyphosphates vary, being particularly high in several cell types like the insulin-secreting cell. By influencing the rates of InsP₆ synthesis and conversion, InsP₆ concentration may be rapidly changed, which should in turn give rise to modulation of insulin exocytosis. We have shown that InsP₆, in a physiological concentration range, is able to modulate exocytosis. At 20–50 μM concentrations, this compound stimulated insulin secretion at basal Ca²⁺-concentration and primed Ca²⁺-mediated exocytosis. Both effects of InsP₆ were dependent on the activation of PKC. This suggests that inositol polyphosphates can function as signaling molecules in stimulus-secretion coupling in pancreatic β-cells.

ABBREVIATIONS

InsP₅

d-myo-inositol 1,3,4,5,6-pentakisphosphate

InsP₆

d-myo-inositol 1,2,3,4,5,6-hexakisphosphate

PKC

protein kinase C

PP-InsP₅ and (PP)₂-InsP₄

inositol polyphosphate pyrophosphates

TPA

12-O-tetradecanoylphorbol 13-acetate

AMP-PCP

β,γ-methyleneadenosine 5′-triphosphate