Logo of endoArchiveHomepageTES HomepageSubscriptionsSubmissionAbout
Endocrinology. 2009 Jul; 150(7): 3067–3075.
Published online 2009 Feb 12. doi:  10.1210/en.2008-0475
PMCID: PMC2703535

Impaired Insulin Exocytosis in Neural Cell Adhesion Molecule−/− Mice Due to Defective Reorganization of the Submembrane F-Actin Network


The neural cell adhesion molecule (NCAM) is required for cell type segregation during pancreatic islet organogenesis. We have investigated the functional consequences of ablating NCAM on pancreatic β-cell function. In vivo, NCAM−/− mice exhibit impaired glucose tolerance and basal hyperinsulinemia. Insulin secretion from isolated NCAM−/− islets is enhanced at glucose concentrations below 15 mM but inhibited at higher concentrations. Glucagon secretion from pancreatic α-cells evoked by low glucose was also severely impaired in NCAM−/− islets. The diminution of insulin secretion is not attributable to defective glucose metabolism or glucose sensing (documented as glucose-induced changes in intracellular Ca2+ and KATP-channel activity). Resting KATP conductance was lower in NCAM−/− β-cells than wild-type cells, and this difference was abolished when F-actin was disrupted by cytochalasin D (1 μM). In wild-type β-cells, the submembrane actin network disassembles within 10 min during glucose stimulation (30 mM), an effect not seen in NCAM−/− β-cells. Cytochalasin D eliminated this difference and normalized insulin and glucagon secretion in NCAM−/− islets. Capacitance measurements of exocytosis indicate that replenishment of the readily releasable granule pool is suppressed in NCAM−/− α- and β-cells. Our data suggest that remodeling of the submembrane actin network is critical to normal glucose regulation of both insulin and glucagon secretion.

The neural cell adhesion molecules (NCAMs) are members of the Ig superfamily that mediate Ca2+-independent cell-cell and cell-substratum interactions by homophilic and heterophilic interactions. The three major isoforms of NCAM (NCAM 120, 140, and 180) are generated via alternative splicing of a single gene (1). The 140- and 180-kDa isoforms of NCAM contain intracellular domains that bind to the cytoskeleton, and may be involved in actin filament turnover, assembly, and remodeling (2,3,4).

Whereas NCAM is widely expressed during development, the expression in adults is confined to the neuromuscular junction and certain endocrine cells, including chromaffin cells of the adrenal medulla (5), thyroid cells (6), and the islets of Langerhans (7,8). NCAM-deficient mice exhibit a variety of defects in the central nervous system (CNS), such as decreased brain size, smaller olfactory bulbs, memory defects, and aggression (9,10,11). They also demonstrate neuromuscular transmission deficits that are attributed to defective synaptic vesicle trafficking (9,12,13). Indeed, a recent study provides evidence of impaired exocytosis in the NCAM-deficient chromaffin cell (14).

In pancreatic islets, NCAM is essential for islet cell type segregation, and the subcellular distribution of F-actin and cadherins (8,15). Remodeling of F-actin plays an important role in glucose-stimulated insulin secretion (16). F-actin depolymerizes in response to glucose, and insulin secretion is facilitated by disruption of the actin network (17). Collectively, these data raise the interesting possibility that F-actin limits the access of insulin granules to the plasma membrane and that its disassembly facilitates the movement of secretory granules to the plasma membrane (18).

Here, we have examined the functional consequences of NCAM ablation in pancreatic β-cells. We demonstrate that NCAM-deficient mice exhibit basal hyperinsulinemia and impaired glucose tolerance, features that correlated with abnormal glucose regulation of both insulin and glucagon secretion. Our findings indicate that this involves defective reorganization of the submembrane F-actin network, and we show that depolymerization of the actin network using cytochalasin D restores normal secretory responses.

Materials and Methods


NCAM knockout and wild-type mice were bred on a C57/BL6 background (9). Male and female heterozygous NCAM+/− mice were used for breeding homozygous NCAM−/− (knockout) and NCAM+/+ (wild type) mice. In the NCAM-deficient mice, most parts of exons 3 and 4 are deleted, leading to ablation of all NCAM isoforms. Genotyping was performed according to previously described procedures (9). Only female mice, aged 16–22 wk, were used in this study. The mice were killed by cervical dislocation, and pancreatic islets were isolated by collagenase digestion. Animal husbandry and procedures for killing the animals were approved by the ethical committee at Lund University.

For electrophysiology the islets were dissociated into single cells using a Ca2+-free solution as previously described (19). The resulting cell suspension was plated on Corning petri dishes (Corning, Inc., Corning, NY) and maintained in tissue culture for 6–30 h in RPMI 1640 medium containing 10% (vol/vol) fetal calf serum, 100 μg/ml streptomycin, and 100 IU/ml penicillin.

Measurements of intracellular Ca2+ concentration ([Ca2+]i)

Changes in [Ca2+]i were recorded by dual-wavelength microfluorimetry (20) in intact islets loaded with 3 μm fura-2 in the presence of 0.007% wt/vol pluronic acid (Invitrogen Corp., Carlsbad, CA) for 40 min at 37 C. Experiments were performed as outlined elsewhere (21), and the bath solution contained (in mm): 140 NaCl, 3.6 KCl, 2 NaHCO3, 0.5 NaH2PO4, 0.5 MgSO4, 5 HEPES (pH 7.4 with NaOH), 2.6 CaCl2, and glucose as indicated. The measurements were performed at approximately +32 C to allow comparison with electrophysiological data.


Whole cell currents and membrane capacitance (as an indicator of exocytosis) were recorded from single cells using an EPC-9 patch-clamp amplifier and Pulse software (HEKA Electronics, Lambrecht/Pfalz, Germany). Pancreatic α- and β-cells were functionally identified by the inactivation properties of the Na+ current (22). Whole cell currents were recorded in a solution containing (in mm): 138 NaCl, 5.6 KCl, 2.6 CaCl2, 1.2 MgCl2, 5 HEPES (pH 7.4 with NaOH), and glucose as specified. For the recordings of Ca2+currents, this solution was supplemented with 20 mm tetraethylammonium chloride (TEA) to block delayed rectifier K+ channels (NaCl correspondingly reduced to maintain iso-osmolarity). Voltage-gated Ca2+ and K+ currents were measured as previously reported (23,24).

The maximum whole cell KATP conductance was estimated using the standard whole cell configuration after intracellular dialysis with (in mm) 120 KCl, 1 MgCl2, 0.3 Mg-ATP, 0.3 Mg-ADP, 5 EGTA, 2 CaCl2, and 10 HEPES (pH 7.15), and applying ±20 mV depolarizing pulses (500 msec) from a holding potential of −70 mV.

Changes in whole cell-resting K+ conductance were assessed in metabolically intact β-cells using the perforated patch technique. Cells were initially held at −70 mV, and currents were elicited by 200 msec voltage ramps between −110 and 0 mV applied every 3 sec. The recording pipettes contained (in mm): 76 K2SO4, 10 NaCl, 10 KCl, 1 MgCl2, and 5 HEPES (pH 7.35 with KOH). Electrical contact to the cell interior was established by the pore-forming antibiotic amphotericin B (21). Glucose was added at the indicated concentrations to the standard TEA-free extracellular solution. When the effect of cytochalasin D was evaluated, the drug was added to the culture medium 15 min before measurements.

Exocytosis was monitored using the standard whole cell configuration in conjunction with capacitance measurements using the TEA-containing extracellular solution. Some of the experiments were performed after overnight culture in the presence of 250 μm of the KATP-channel opener diazoxide (Sigma-Aldrich Corp., St. Louis, MO) to prevent spontaneous electrical activity with resultant possible suppression of exocytosis due to degranulation. The sine wave had a frequency of 500 Hz and a peak amplitude of 20 mV. Secretion was elicited by trains of 10 500-msec voltage-clamp depolarizations from −70 to 0 mV. The pipette solution consisted of (in mm): 125 Cs-glutamate, 10 CsCl, 10 NaCl, 1 MgCl2, 0.05 EGTA, 0.1 cAMP, 3 Mg-ATP, and 5 HEPES (pH 7.15 using CsOH). All electrophysiological measurements were performed at approximately +32 C.

Measurements of hormone release

Hormone release was studied using intact islets (21). The amount of insulin and glucagon in the incubation media was determined using an in-house RIA (25). Glucose was included as indicated in the legends. Cytochalasin D was prepared as a concentrated stock solution in dimethylsulfoxide (DMSO). Cytochalasin was added both to the preincubation and incubation media, and was present throughout. The final concentration of DMSO was 0.1%, and DMSO was included also in the control experiments.

In vivo glucose tolerance test and insulin sensitivity assay

All in vivo experiments were conducted on animals with free access to food. For the in vivo glucose challenge, glucose [11.1 mmol (equivalent to 2 g/kg body weight)] was dissolved in 0.9% NaCl and delivered by ip injection. For the insulin sensitivity assay, 0.5 IU insulin/kg body weight (Actrapid; Novo Nordisk, Bagsværd, Denmark) was dissolved in 0.9% NaCl and delivered by ip injection. Blood sampling in nonanesthetized mice (with ethical consent), detection of plasma insulin by RIA, and enzymatic determination of plasma glucose concentrations were performed as described previously (26).

Glucose oxidation

The rate of glucose oxidation was studied by incubating batches of 30 islets for 60 min at 37 C in 100 μl KRB-HEPES buffer containing d-[14C(U)] glucose (specific activity 310 mCi/mmol; New England Nuclear Life Science Products, Boston, MA). Glucose oxidation rates were calculated from the formation of 14CO2 during the incubation period. The incubation was terminated by addition of 100 μl 7% trichloroacetic acid, and released 14CO2 was trapped by addition of 300 μl benzethonium hydroxide and determined by liquid scintillation.


All incubations were made at 4 C unless otherwise stated. Islets were fixed for 1 h in 4% paraformaldehyde, washed in PBS, and then permeabilized in 0.1% Triton X-100 for 30 min. After several washes in PBS, nonspecific binding was blocked for 1 h in PBS containing 5% goat serum. Insulin was probed by overnight incubation with a guinea pig anti-insulin antibody (Abcam plc, Cambridge, UK) at a 1:100 dilution in blocking solution. After several washes in PBS, secondary staining was performed for 2 h at room temperature using fluorescein-conjugated goat anti-guinea pig antibody (Vector Laboratories, Burlingame, CA). The islets were washed thoroughly, and F-actin was stained by incubation at room temperature for 45 min in PBS containing 0.66 μm rhodamine phalloidin (Molecular Probes, Inc., Eugene OR). Localization of F-actin was visualized by confocal microscopy (Carl Zeiss, Jena, Germany) using 545 nm (He-Ne) excitation. Emission was detected above 600 nm.

Data analysis

Off-line image analysis was performed using LSM 510 software (Zeiss) and Microcal Origin (version 7; OriginLab Corp., Northampton, MA). For each condition, six to eight islets from two independent experiments were stained. Cortical F-actin disassembly was determined by line scanning the periphery of cells using published methods (27). These scans were then subtracted from background fluorescence and were classified as being continuous or discontinuous. Disruption of F-actin was quantified by measuring the sd of the fluorescence intensity in a 1-μm band below the plasma membrane. To compensate for variations in dye labeling, the sd was normalized to mean fluorescence intensity.

All data are presented as mean values ± sem of the indicated number of experiments (n). The statistical significance for the difference between two means was evaluated using two-way ANOVA and the Student’s t test, paired or unpaired as appropriate.


NCAM−/− mice exhibit impaired glucose tolerance in vivo

We determined the consequences of genetic ablation of NCAM on systemic glucose homeostasis and insulin release in fed animals (Fig. 1A1A).). NCAM−/− mice were essentially normoglycemic but exhibited 5-fold higher basal plasma insulin levels. NCAM−/− mice were glucose intolerant during an ip glucose challenge. Whereas the plasma glucose concentration returned to basal within approximately 15 min in control animals, glucose remained elevated for more than 50 min in NCAM−/− mice. Both wild-type and NCAM−/− mice responded to the glucose challenge with an enhanced insulin secretion. However, plasma insulin levels measured at 3 and 8 min, when plasma glucose was around 23 mm, did not differ between wild-type and NCAM−/− mice.

Figure 1
Impaired regulation of insulin and glucagon secretion, as well as glucose tolerance but normal insulin tolerance in NCAM−/− mice. A, A glucose challenge (2 g/kg body weight ip) was applied at time zero in wild-type (NCAM+/+ ...

The high basal plasma insulin levels in NCAM−/− mice do not appear to be a consequence of insulin resistance (Fig. 1B1B).). If anything, these animals have improved insulin sensitivity. In fed mice, insulin (0.5 IU/kg ip) reduced plasma glucose by approximately 6 mm in both wild-type and NCAM−/− mice. Whereas plasma glucose returned to the prestimulatory level within 70 min in wild-type mice, it remained low for at least 70 min in NCAM−/− mice.

Defective in vitro regulation of insulin and glucagon secretion by glucose in NCAM−/− islets

In vivo experiments are difficult to interpret because glucose homeostasis in vivo reflects insulin secretion, insulin clearance, and glucose-sensing mechanisms in the CNS (28). To elucidate whether the observed glucose intolerance involves defective insulin secretion, islets were isolated from wild-type and NCAM−/− mice. The islets were exposed to increasing glucose concentrations (1–30 mm). Basal insulin secretion (1–5 mm glucose) was elevated by 50–80% in NCAM−/− islets relative to that observed in control islets (Fig. 1C1C).). Insulin secretion was also approximately 2-fold higher in NCAM−/− than wild-type islets at 10 and 15 mm glucose. However, a paradoxical decrease was observed at glucose concentrations more than 15 mm, and less insulin was released at 30 mm glucose than at 15 mm. In wild-type islets, glucose produced a sigmoidal stimulation of insulin secretion with an approximate EC50 of 17 mm.

Glucagon secretion was also affected in NCAM−/− islets (Fig. 1D1D).). Whereas exposure to increasing glucose concentrations led to a concentration-dependent reduction of glucagon secretion that was half-maximal at less than 5 mm in wild-type animals, no such decrease was seen in the NCAM−/− islets. In NCAM−/− islets, glucagon secretion at 1 mm glucose was already strongly reduced compared with that seen in wild-type islets, and no further suppression was produced by high glucose.

Intracellular Ca2+ handling and electrical activity are unperturbed in NCAM−/− islets

Basal [Ca2+]i (measured in 1 mm glucose) averaged 100 ± 7 (n = 17) and 103 ± 5 nm (n = 23) in wild-type and NCAM−/− islets, respectively (Fig. 22).). In islets, exposed to 5 mm glucose, [Ca2+]i averaged 111 ± 8 (n = 10) and 82 ± 7 nm (n = 10; P < 0.05 vs. wild-type) in wild-type and NCAM−/− islets, respectively (data not shown). After stimulation with 15 mm glucose, [Ca2+]i increased to a peak average value of 347 ± 19 nm in control islets and 408 ± 23 nm (not statistically different) in islets from the knockout mice. After the initial peak, [Ca2+]i oscillations with a frequency of 2.34 ± 0.24 and 3.26 ± 0.45 min−1 were observed in wild-type and NCAM−/− islets, respectively. The average [Ca2+]i at 15 mm glucose amounted to 198 ± 14 nm in NCAM+/+ and 265 ± 15 nm (P < 0.01) in NCAM−/− islets. When glucose was increased to 30 mm, the frequency of the oscillations increased to 6.4 ± 1.4 min−1 in wild-type and 8.1 ± 2.7 min−1 in NCAM−/− islets. The average [Ca2+]i at 30 mm glucose amounted to 309 ± 25 and 354 ± 14 nm in wild-type and NCAM−/− islets, respectively.

Figure 2
Normal glucose regulation of [Ca2+]i in NCAM−/− islets. A, [Ca2+]i measured in a wild-type islet when glucose was increased from 1–15 and 30 mm as indicated by the arrows. B, As in ...

The KATP-channel conductance is reduced in NCAM−/− β-cells

Perforated patch whole cell recordings, in which cell metabolism is maintained, were used to estimate GK,ATP [the glucose-sensitive resting K+ conductance of the β-cell (29)] by applying voltage ramps between −110 and 0 mV (Fig. 33,, A and B). Membrane conductance was estimated between −100 and −50 mV, i.e. the (approximate) linear range of the current-voltage relationship. The resting GK,ATP measured at 1 mm glucose in wild-type β-cells averaged 3.6 ± 0.9 nanosiemens (nS) (n = 8). The corresponding value in NCAM−/− β-cells averaged 1.6 ± 0.3 nS (P < 0.05). Increasing glucose to 15 mm decreased GK,ATP to 0.77 ± 0.15 and 0.40 ± 0.07 nS (P < 0.05) in wild-type and NCAM−/− β-cells, respectively. In the presence of 30 mm glucose, GK,ATP averaged 0.54 ± 0.12 and 0.40 ± 0.06 nS in the control and knockout islets, respectively. Values obtained at 30 mm glucose are similar to those obtained in the presence of 100 μm tolbutamide [0.56 ± 012 nS (n = 3) in wild-type β-cells]. In both wild-type and NCAM−/− β-cells, the deviations from linearity observed at voltages above −30 mV are due to the opening of voltage-gated Ca2+ channels (downward deflections) and activation of voltage-gated K+ channels [upward deflections (30)]. These conductances do not activate at voltages more negative than −50 mV and, therefore, do not contaminate the KATP-conductance values.

Figure 3
Reduced β-cell whole cell KATP conductance and normal glucose oxidation in NCAM−/− islets. A, Whole cell KATP conductance measured at 1 (black), 15 (dark gray), and 30 mm glucose (light gray) in NCAM+/+ (middle ...

The maximum whole cell KATP conductance averaged 12 ± 1.5 (n = 10) and 9.5 ± 2.1 nS (n = 9; not statistically different) in wild-type and NCAM−/− β-cells, respectively (data not shown). In separate experiments it was ascertained that there was no difference in the voltage dependence or magnitude of the voltage-gated Ca2+ or K+ currents between NCAM+/+ and NCAM−/− β-cells (data not shown).

Glucose oxidation is unaltered in NCAM −/− islets

The rates of glucose oxidation at 1, 15, and 30 mm glucose were approximately the same in wild-type and NCAM−/− islets (Fig. 3C3C).

Strong suppression of exocytosis in NCAM−/− α- and β-cells

High-resolution capacitance measurements of exocytosis were applied to single α- and β-cells from wild-type and knockout animals. Exocytosis was triggered by a train of 10 500-msec depolarizations from −70 to 0 mV. In wild-type β-cells, the train produced a total capacitance increase of 890 ± 250 fF. The responses were much smaller in NCAM-deficient β-cells, and the total capacitance increase was limited to 240 ± 90 fF, a reduction of 73% (P < 0.05; Fig. 44,, A and B). When exocytosis was expressed as capacitance increase per pulse, it became evident that exocytosis in wild-type β-cells peaked at 150 fF/pulse (during the second pulse) and then proceeded at a fairly stable rate of approximately 90 fF/pulse during the following six pulses. Exocytosis in NCAM−/− cells proceeded at a much lower speed, being approximately 50 fF during the first pulse and approximately 20 fF/pulse during the remainder of the train (Fig. 4B4B).). These experiments were conducted on β-cells cultured in the presence of diazoxide (250 μm) to prevent degranulation. In cells not exposed to diazoxide, responses were smaller, but the difference between wild-type and knockout mice remained. The total increase in cell capacitance evoked by the train under these experimental conditions averaged 525 ± 86 (n = 10) and 195 ± 46 fF (n = 7) in wild-type and NCAM−/− β-cells, respectively (P < 0.01).

Figure 4
Reduced exocytotic capacity of NCAM−/− β-cells. A, Capacitance increases (ΔCm) elicited in wild-type (NCAM+/+) and NCAM-deficient (NCAM−/−) β-cells stimulated by trains consisting ...

Exocytosis in α-cells was also compromised in NCAM−/− mice (Fig. 44,, C and D). In wild-type α-cells, a train of 10 500-msec depolarizations from −70 to 0 mV resulted in a capacitance increase of 1890 ± 260 fF. The responses were much smaller in NCAM−/− α-cells and only amounted to 880 ± 180 fF, thus a reduction of more than 50% (P < 0.01). As was the case in β-cells, exocytosis was suppressed to roughly the same extent throughout the train in the knockout α-cells compared with their wild-type counterparts (Fig. 4D4D).

We also compared exocytosis in wild-type and NCAM null β-cells dialyzed with 1.5 μm free Ca2+ and 0.1 mm cAMP. The magnitude of the exocytotic response using this protocol principally reflects the mobilization of granules from the reserve pool (31). The steady-state rate of capacitance increase (ΔC/Δt; measured 20–100 sec after establishment of the whole cell configuration) averaged 34 ± 6 fF/sec in wild-type β-cells but only 17 ± 2 fF/sec in knockout cells (data not shown).

We analyzed the ultrastructure of wild-type and NCAM−/− β- and α-cells. The data are summarized in supplemental Table 1, which is published as supplemental data on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org. In brief, there was an approximate 20% reduction of the number of granules per granule density in β-cells from NCAM−/− mice. This echoes the insulin content, which was reduced by approximately 10% (from 158 ± 30 and 147 ± 27 ng/islet in control and NCAM−/− islets, respectively). Islet glucagon content was reduced by 38% in NCAM mice, from 15.5 ± 1.7 pg/islet (n = 4) in wild-type islets to 9.6 ± 0.9 pg/islet (n = 5; P < 0.025) in NCAM−/− islets, and the number of granules per α-cell was down approximately 25% (supplemental Table 1).

Abnormal F-actin distribution in NCAM−/− islets explains the secretory defect

The subcellular distribution of F-actin is perturbed in NCAM−/− mice (8). The cortical actin web plays an important role in regulating exocytosis, and actin filaments both facilitate and inhibit the transport of secretory granules (32). To visualize the intracellular actin distribution, islets were stained with phalloidin-rhodamine (Fig. 5A5A).). It was ascertained that the actin-containing cells were also positive for insulin (Fig. 5B5B).

Figure 5
Abnormal distribution of F-actin in NCAM-deficient islet cells. A, Confocal images obtained using phalloidin-rhodamine to visualize actin in knockout (NCAM−/−) and wild-type (NCAM+/+) islets. Islets were preincubated for ...

We determined the integrity of the cortical actin network in wild-type NCAM−/− β-cells. Figure 5C5C summarizes how the integrity of the cortical actin web varies with the glucose concentration and in response to cytochalasin D. Again, it is evident that glucose in wild-type islets led to a concentration-dependent disintegration of the actin network, that 30 mm glucose was nearly as potent as cytochalasin D, and that glucose had no effect in NCAM-deficient cells.

We validated these observations by calculating the sd of the phalloidin-rhodamine fluorescence intensity in a 1-μm band below the plasma membrane. In wild-type cells, there was a glucose-dependent increase in sd (normalized to the mean fluorescence intensity) that was observable within 10 min stimulation and remained stable for at least 60 min (Fig. 5D5D).). No such changes were observed in cells exposed to 1 mm glucose or in NCAM−/− β-cells, regardless of whether the cells were exposed to 1 or 30 mm glucose (Fig. 5E5E).). Collectively, these analyses suggest that high glucose in wild-type cells causes almost as much disassembly of the cortical web as cytochalasin D and that this effect is abolished in NCAM-deficient cells.

Disruption of F-actin network restores normal glucose regulation of insulin and glucagon secretion in NCAM−/− islets

Does impaired remodeling of the actin network in NCAM−/− islets contribute to the failure of glucose to stimulate insulin secretion further when increasing the concentration beyond 15 mm? We tested this by measuring insulin release at 1, 15, and 30 mm glucose in the absence or presence of cytochalasin D in wild-type (Fig. 6A6A)) and NCAM−/− islets (Fig. 6B6B).). In wild-type islets, cytochalasin D had a moderate (40%) stimulatory effect at 15 mm glucose but exerted no additive effect when the islets were stimulated by 30 mm glucose. In NCAM−/− islets exposed to 30 mm glucose (but not 15 mm), cytochalasin D produced a strong 2.6-fold stimulation of insulin secretion from NCAM−/− islets. Under these conditions, insulin secretion at 30 mm glucose was greater than that observed at 15 mm glucose, and the paradoxical suppression of insulin secretion obtained when glucose was elevated from 15–30 mm was removed.

Figure 6
Normalization of glucose-induced insulin secretion from NCAM−/− mice after disruption of cortical actin network. A, Insulin release measured during 1 h static incubations at 1, 15, and 30 mm glucose in the absence (○) or presence ...

We also compared the effects of cytochalasin D on glucagon secretion from wild-type and NCAM−/− islets. In wild-type islets, cytochalasin D increased glucagon secretion at 1 mm glucose by approximately 30%, but it did not interfere with the inhibitory action of 15 mm glucose (Fig. 6C6C).). In agreement with the data of Fig. 1D1D,, glucagon secretion at 1 mm glucose was strongly (−62%) inhibited in NCAM−/− islets compared with wild-type data, and elevation of glucose resulted in no further suppression (Fig. 6D6D).). Glucagon secretion at 1 mm glucose increased more than 2-fold after pretreatment with cytochalasin, and under these conditions, 15 mm glucose inhibited glucagon secretion by 42% (Fig. 6D6D),), similar to the 46% inhibition seen in wild-type islets under the same experimental conditions.

KATP-channel activity is known to be influenced by actin (33,34). We tested whether the denser submembrane actin network in NCAM−/− β-cells (Fig. 55)) may account for the lower resting K+ conductance. Whole cell-resting K+ conductance was measured in wild-type and knockout β-cells preincubated in the presence of 1 μm cytochalasin D or vehicle alone (0.1% DMSO) for 15 min. The resting (tolbutamide sensitive) GK,ATP at 1 mm glucose averaged 5.1 ± 1 nS in NCAM+/+ cells pre-exposed to DMSO alone (n = 5). In NCAM−/− β-cells, the corresponding value was 2.3 ± 0.5 nS (n = 5; P < 0.05), thus 55% less than in wild-type cells and in close agreement with the data in Fig. 3B3B.. The latter value increased to 4.8 ± 1.2 nS (n = 6) in NCAM−/− β-cells pretreated with cytochalasin D (not statistically different from control cells). When cytochalasin D pretreated NCAM−/− β-cells were exposed to 100 μm tolbutamide, the KATP conductance decreased to 0.86 ± 0.12 nS (n = 5).


Here, we show that genetic ablation of NCAMs results in impaired glucose tolerance. Our studies on isolated pancreatic islets and single islet cells indicate that the loss of NCAM is associated with functional changes that can be expected to have profound effects on islet hormone secretion and, thus, explain the phenotype of the mouse.

Dysregulation of insulin and glucagon secretion

Whereas insulin secretion at high (>15 mm) glucose concentrations was reduced in NCAM-deficient islets relative the wild-type islets, the opposite was observed at glucose concentrations that are normally subthreshold. The KATP channels play a central role in the glucose sensing of the pancreatic β-cells (35). Therefore, it is pertinent that the resting whole cell KATP conductance was reduced by 60% in NCAM−/− β-cells. The finding that the maximum whole cell KATP current was essentially the same in both wild-type and knockout β-cells argues that the reduced resting GK,ATP in intact β-cells cannot be accounted for by a lowered number of channels per β-cells. Our observation suggests that the KATP channels are under stronger resting inhibition in NCAM−/− β-cells. This may result from an acceleration of β-cell metabolism/ATP production that in turn would lead to greater inhibition of the KATP channels. However, this possibility seems less likely because our measurements of glucose oxidation revealed no statistically significant alterations in NCAM−/− islets (Fig. 3C3C).). Thus, there is no evidence suggesting that glucose metabolism and ATP generation are enhanced in NCAM−/− islets at lower glucose concentrations that may explain the shift in insulin secretion toward lower concentrations of the sugar.

As shown in Fig. 55,, the submembrane actin network is denser in NCAM−/− than NCAM+/+ β-cells. Depolymerization of actin filaments has resulted in activation of KATP channels, whereas stabilization has the opposite effect (33,34). Possibly, the denser actin network in NCAM−/− cells leads to reduced KATP-channel activity. Because channel activity is already low in NCAM−/− β-cells, even small further (glucose induced) decrements in channel activity can be expected to evoke β-cell electrical activity with resultant stimulation of insulin secretion. This scenario would explain the stimulation of insulin secretion seen at glucose concentrations 15 mm or less and the increase in steady-state average [Ca2+]i in NCAM−/− islets at 15 mm glucose. It is not immediately evident why insulin secretion at 1–5 mm glucose is elevated in NCAM−/− islets compared with wild-type islets. [Ca2+]i was not higher in NCAM−/− islets, so, although resting KATP conductance is reduced, this is clearly not sufficient to evoke electrical activity at these low glucose concentrations. Nevertheless, increased β-cell responsiveness to glucose may contribute to the finding that the basal plasma insulin levels measured in vivo were 5-fold higher in knockout than wild-type mice, but it should be noted that insulin secretion in vitro was only approximately 2-fold higher in NCAM−/− than wild-type islets at approximately 9 mm glucose (basal plasma glucose in fed animals), and, therefore, additional (systemic) factors must also be involved.

The high basal insulin levels in NCAM−/− mice were not associated with insulin resistance (Fig. 1B1B).). It seems possible that the latter effect reflects the failure of low glucose to stimulate glucagon secretion (Fig. 1D1D).). In addition, glucose sensing in the CNS may be affected and contribute to both the basal hyperinsulinemia and defective counter-regulation in NCAM−/− mice (28). Indeed, there is evidence that glucagon secretion in vivo depends on glucose sensing in hypothalamic neurons (36).

Reduced exocytosis in NCAM-deficient α- and β-cells

Although insulin secretion at subthreshold glucose concentrations is enhanced in NCAM-deficient islets, insulin secretion is paradoxically inhibited at concentrations above 15 mm of the sugar. The strong suppression (70%) of insulin secretion from NCAM−/− islets at a maximally stimulatory glucose concentration (30 mm) is comparable to the inhibition of exocytosis seen in the capacitance measurements during trains of 500-msec depolarizations (60–75% depending on whether the experiments were conducted in β-cells cultured in the absence or presence of diazoxide). Thus, the reduction of secretion at the higher glucose concentrations may result from the impairment of exocytosis. The inhibition of exocytosis is likely to occur at all glucose concentrations, and the lack of detectable inhibition at glucose concentrations 15 mm or less we attribute to a leftward shift of the dose-response curve due to the reduction of the whole cell KATP conductance.

Disassembly of cortical actin network required for maximum glucose-induced insulin secretion

Secretory vesicles in endocrine cells, including pancreatic β-cells, can functionally be separated into two compartments that vary with regard to release competence: a release-ready vesicle pool, and a reserve pool (19,37). Close inspection of the capacitance measurements during trains of depolarizations reveals that exocytosis was only moderately reduced (~50%) during the first 500-msec pulse, and that the suppression became stronger during the second and third pulse (>80%). This behavior is in agreement with observations in chromaffin cells indicating that the transition from a readily releasable granule pool into an immediately releasable pool is impaired in NCAM-deficient animals (14). Our experiments on glucagon-producing pancreatic α-cells reveal that exocytosis in this cell type is also strongly reduced. Again, there was a strong suppression of exocytosis during the latter part of the stimulation train (pulses 5–10 in Fig. 4D4D).). Collectively, these data indicate, together with the published reports on chromaffin cells (14) and neuromuscular junction (12,13), that NCAM plays a general role in vesicle release and trafficking in neurons as well as endocrine cells.

In chromaffin cells, F-actin forms a cortical network that blocks the approach of the secretory granules toward the plasma membrane, and controls the size of the release-ready vesicle pool and the initial rate of exocytosis (27). A dense network of actin filaments, approximately 50–300-nm thick, is also situated beneath the plasma membrane in β-cells, and its destruction is associated with stimulation of insulin secretion (38). In β-cells, the cortical actin network in β-cells functions as a barrier preventing passive diffusion of insulin granules to the plasma membrane. Glucose has depolymerized F-actin (17,18). Here, we demonstrate that no glucose-induced remodeling of the actin network occurs in NCAM-deficient β-cells, whereas the effects of glucose in wild-type β-cells were almost as strong as those of cytochalasin D (Fig. 55).). The difference was particularly pronounced at glucose concentrations more than 15 mm, i.e. the range of concentrations where the paradoxical inhibition of insulin secretion in NCAM−/− islets was observed.

A dense actin network may finally, via reduction of exocytosis (Fig. 44),), also explain why glucagon secretion at 1 mm glucose is strongly reduced in NCAM−/− islets (Figs. 11 and 66).). This inhibition is stronger than can be accounted for by the reduced glucagon content. The role of the cytoskeleton in glucagon secretion has not been extensively studied. However, it is of interest that cytochalasin enhanced glucagon secretion in response to low glucose in both wild-type and NCAM−/− islets but did not interfere with the inhibitory action of 15 mm glucose. It is tempting to speculate that glucagon secretion and exocytosis are strongly reduced in NCAM−/− islets/α-cells because the dense submembrane actin network does not allow an adequate supply of secretory granules to the release site.

In conclusion, our observations in NCAM−/− mice highlight the importance of remodeling of the submembrane actin network in the metabolic control of both insulin and glucagon secretion. However, we acknowledge that the precise molecular link between NCAM and the appearance of a dense actin network in islet cells remains to be established.

Supplementary Material

[Supplemental Data]


Supported by Wellcome Trust and the Swedish Research Council. P.R. is Royal Society-Wolfson Merit Award Research Fellow. C.S.O. holds a postdoctoral fellowship from the Swedish Research Council.

Disclosure Summary: The authors have nothing to disclose.

First Published Online February 12, 2009

Abbreviations: CNS, Central nervous system; DMSO, dimethylsulfoxide; NCAM, neural cell adhesion molecule; TEA, tetraethylammonium chloride.


  • Cunningham BA, Hemperly JJ, Murray BA, Prediger EA, Brackenbury R, Edelman GM 1987 Neural cell adhesion molecule: structure, immunoglobulin-like domains, cell surface modulation, and alternative RNA splicing. Science 236:799–806 [PubMed]
  • Goridis C, Brunet JF 1992 NCAM: structural diversity, function and regulation of expression. Semin Cell Biol 3:189–197 [PubMed]
  • Suter DM, Forscher P 1998 An emerging link between cytoskeletal dynamics and cell adhesion molecules in growth cone guidance. Curr Opin Neurobiol 8:106–116 [PubMed]
  • Juliano RL 2002 Signal transduction by cell adhesion receptors and the cytoskeleton: functions of integrins, cadherins, selectins, and immunoglobulin- superfamily members. Annu Rev Pharmacol Toxicol 42:283–323 [PubMed]
  • Grant NJ, Leon C, Aunis D, Langley K 1992 Cellular localization of the neural cell adhesion molecule L1 in adult rat neuroendocrine and endocrine tissues: comparisons with NCAM. J Comp Neurol 325:548–558 [PubMed]
  • Vargas F, Tolosa E, Sospedra M, Catálfamo M, Lucas-Martín A, Obiols G, Pujol-Borrell R 1994 Characterization of neural cell adhesion molecule (NCAM) expression in thyroid follicular cells: induction by cytokines and over-expression in autoimmune glands. Clin Exp Immunol 98:478–488 [PMC free article] [PubMed]
  • Langley OK, Aletsee-Ufrecht MC, Grant NJ, Gratzl M 1989 Expression of the neural cell adhesion molecule NCAM in endocrine cells. J Histochem Cytochem 37:781–791 [PubMed]
  • Esni F, Täljedal IB, Perl AK, Cremer H, Christofori G, Semb H 1999 Neural cell adhesion molecule (N-CAM) is required for cell type segregation and normal ultrastructure in pancreatic islets. J Cell Biol 144:325–337 [PMC free article] [PubMed]
  • Cremer H, Lange R, Christoph A, Plomann M, Vopper G, Roes J, Brown R, Baldwin S, Kraemer P, Scheff S, Barthels D, Rajewsky K, Wille W 1994 Inactivation of the N-CAM gene in mice results in size reduction of the olfactory bulb and deficits in spatial learning. Nature 367:455–459 [PubMed]
  • Holst BD, Vanderklish PW, Krushel LA, Zhou W, Langdon RB, McWhirter JR, Edelman GM, Crossin KL 1998 Allosteric modulation of AMPA-type glutamate receptors increases activity of the promoter for the neural cell adhesion molecule, N-CAM. Proc Natl Acad Sci USA 95:2597–2602 [PMC free article] [PubMed]
  • Tomasiewicz H, Ono K, Yee D, Thompson C, Goridis C, Rutishauser U, Magnuson T 1993 Genetic deletion of a neural cell adhesion molecule variant (N-CAM-180) produces distinct defects in the central nervous system. Neuron 11:1163–1174 [PubMed]
  • Rafuse VF, Polo-Parada L, Landmesser LT 2000 Structural and functional alterations of neuromuscular junctions in NCAM-deficient mice. J Neurosci 20:6529–6539 [PubMed]
  • Polo-Parada L, Bose CM, Landmesser LT 2001 Alterations in transmission, vesicle dynamics, and transmitter release machinery at NCAM-deficient neuromuscular junctions. Neuron 32:815–828 [PubMed]
  • Chan SA, Polo-Parada L, Landmesser LT, Smith C 2005 Adrenal chromaffin cells exhibit impaired granule trafficking in NCAM knockout mice. J Neurophysiol 94:1037–1047 [PubMed]
  • Cirulli V, Baetens D, Rutishauser U, Halban PA, Orci L, Rouiller DG 1994 Expression of neural cell adhesion molecule (N-CAM) in rat islets and its role in islet cell type segregation. J Cell Sci 107:1429–1436 [PubMed]
  • Howell SL, Tyhurst M 1982 Microtubules, microfilaments and insulin-secretion. Diabetologia 22:301–308 [PubMed]
  • Thurmond DC, Gonelle-Gispert C, Furukawa M, Halban PA, Pessin JE 2003 Glucose-stimulated insulin secretion is coupled to the interaction of actin with the t-SNARE (target membrane soluble N-ethylmaleimide-sensitive factor attachment protein receptor protein) complex. Mol Endocrinol 17:732–742 [PubMed]
  • Nevins AK, Thurmond DC 2003 Glucose regulates the cortical actin network through modulation of Cdc42 cycling to stimulate insulin secretion. Am J Physiol Cell Physiol 285:C698–C710 [PubMed]
  • Eliasson L, Renström E, Ding WG, Proks P, Rorsman P 1997 Rapid ATP-dependent priming of secretory granules precedes Ca(2+)- induced exocytosis in mouse pancreatic B-cells. J Physiol 503:399–412 [PMC free article] [PubMed]
  • Grynkiewicz G, Poenie M, Tsien RY 1985 A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 260:3440–3450 [PubMed]
  • Olofsson CS, Salehi A, Holm C, Rorsman P 2004 Palmitate increases L-type Ca2+ currents and the size of the readily releasable granule pool in mouse pancreatic β-cells. J Physiol 557:935–948 [PMC free article] [PubMed]
  • Göpel S, Kanno T, Barg S, Galvanovskis J, Rorsman P 1999 Voltage-gated and resting membrane currents recorded from B-cells in intact mouse pancreatic islets. J Physiol 521:717–728 [PMC free article] [PubMed]
  • Bokvist K, Rorsman P, Smith PA 1990 Effects of external tetraethylammonium ions and quinine on delayed rectifying K+ channels in mouse pancreatic β-cells. J Physiol 423:311–325 [PMC free article] [PubMed]
  • Schulla V, Renström E, Feil R, Feil S, Franklin I, Gjinovci A, Jing XJ, Laux D, Lundquist I, Magnuson MA, Obermüller S, Olofsson CS, Salehi A, Wendt A, Klugbauer N, Wollheim CB, Rorsman P, Hofmann F 2003 Impaired insulin secretion and glucose tolerance in β cell-selective Ca(v)1.2 Ca2+ channel null mice. EMBO J 22:3844–3854 [PMC free article] [PubMed]
  • Panagiotidis G, Salehi AA, Westermark P, Lundquist I 1992 Homologous islet amyloid polypeptide: effects on plasma levels of glucagon, insulin and glucose in the mouse. Diabetes Res Clin Pract 18:167–171 [PubMed]
  • Salehi A, Carlberg M, Henningson R, Lundquist I 1996 Islet constitutive nitric oxide synthase: biochemical determination and regulatory function. Am J Physiol 270(6 Pt 1):C1634–C1641 [PubMed]
  • Vitale ML, Seward EP, Trifaró JM 1995 Chromaffin cell cortical actin network dynamics control the size of the release-ready vesicle pool and the initial rate of exocytosis. Neuron 14:353–363 [PubMed]
  • Marty N, Dallaporta M, Thorens B 2007 Brain glucose sensing, counterregulation, and energy homeostasis. Physiology (Bethesda) 22:241–251 [PubMed]
  • Ashcroft FM, Rorsman P 1989 Electrophysiology of the pancreatic β-cell. Prog Biophys Mol Biol 54:87–143 [PubMed]
  • Rorsman P, Trube G 1986 Calcium and delayed potassium currents in mouse pancreatic β-cells under voltage-clamp conditions. J Physiol 374:531–550 [PMC free article] [PubMed]
  • Rorsman P, Eliasson L, Renström E, Gromada J, Barg S, Göpel S 2000 The cell physiology of biphasic insulin secretion. News Physiol Sci 15:72–77 [PubMed]
  • Rorsman P, Renström E 2003 Insulin granule dynamics in pancreatic β cells. Diabetologia 46:1029–1045 [PubMed]
  • Harvey J, Hardy SC, Irving AJ, Ashford ML 2000 Leptin activation of ATP-sensitive K+ (KATP) channels in rat CRI-G1 insulinoma cells involves disruption of the actin cytoskeleton. J Physiol 527:95–107 [PMC free article] [PubMed]
  • Terzic A, Kurachi Y 1996 Actin microfilament disrupters enhance K(ATP) channel opening in patches from guinea-pig cardiomyocytes. J Physiol 492:395–404 [PMC free article] [PubMed]
  • Ashcroft FM 2005 ATP-sensitive potassium channelopathies: focus on insulin secretion. J Clin Invest 115:2047–2058 [PMC free article] [PubMed]
  • Miki T, Liss B, Minami K, Shiuchi T, Saraya A, Kashima Y, Horiuchi M, Ashcroft F, Minokoshi Y, Roeper J, Seino S 2001 ATP-sensitive K+ channels in the hypothalamus are essential for the maintenance of glucose homeostasis. Nat Neurosci 4:507–512 [PubMed]
  • Olofsson CS, Göpel SO, Barg S, Galvanovskis J, Ma X, Salehi A, Rorsman P, Eliasson L 2002 Fast insulin secretion reflects exocytosis of docked granules in mouse pancreatic B-cells. Pflugers Arch 444:43–51 [PubMed]
  • Orci L, Gabbay KH, Malaisse WJ 1972 Pancreatic β-cell web: its possible role in insulin secretion. Science 175:1128–1130 [PubMed]

Articles from Endocrinology are provided here courtesy of The Endocrine Society

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...