Effect of the Insulin Mimetic L-783,281 on Intracellular [Ca²⁺] and Insulin Secretion From Pancreatic β-Cells:

The insulin mimetic L-783,281 and its inactive analog L-767,827 were prepared at Merck Laboratories as described elsewhere (1). All chemicals for islet isolation and cell culture were from Life Technologies. All other chemicals were from Sigma Chemical (St. Louis, MO) unless otherwise stated and were of highest purity available. Experiments were performed with cells bathed in a modified Krebs-Ringer buffer (KRB) consisting of (in mmol/l): 118 NaCl, 5.4 KCl, 2.4 CaCl₂, 1.2 MgSO₄, 1. KH₂PO₄, 3.0 d-glucose, and 20 HEPES. Drugs were dissolved in DMSO and diluted in KRB to give the desired final concentration. A stock solution of LY294002 was dissolved in ethanol before diluting in KRB.

Single β-cells were prepared as previously described (24). Briefly, male CD1, C57Bl/6 IRS-1^+/+, or C57Bl/6 IRS-1^−/− mice (25), weighing 20–30 g, were killed by cervical dislocation, and islets were isolated by collagenase digestion and hand-picked under a stereomicroscope. Islets were dispersed into single cells using a 0.025% trypsin/EDTA solution for 6 min at 37°C and plated on 35-mm tissue culture dishes for electrochemical experiments or 25-mm glass coverslips for Ca²⁺ imaging experiments. Cells were maintained in RPMI 1640 culture medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin, stored at 37°C in 5% CO₂, and used on days 2–4 following isolation.

Cells were prepared for secretion measurements by incubating in RPMI medium supplemented with 1 mmol/l 5-hydroxytryptamine (5-HT) for 16 h before experiments. This treatment has previously been shown to cause 5-HT to accumulate in secretory vesicles of β-cells (26). During exocytosis, the accumulated 5-HT is co-released with insulin (27), providing a marker of secretion that is reliably detected by amperometry using carbon-fiber microelectrodes. The validity of this method as a means to measure insulin exocytosis has been verified in several reports (27–29). To perform secretion measurements, RPMI medium containing 5-HT was replaced with KRB, and the culture plate was mounted on the stage of a Zeiss Axiovert 100 microscope equipped with a microincubator (Medical Systems, Greenvale, NY) to maintain temperature at 37°C. 5-HT released from the cells was detected with a 9-μm diameter carbon fiber disc electrode prepared as described elsewhere (30). For amperometric measurements, the electrode was poised at −0.65 V versus a sodium-saturated calomel reference electrode using a battery and voltage divider. Amperometric current was collected at 500 Hz and low-pass filtered at 100 Hz using a AI 403 current amplifier (Axon Instruments, Foster City, CA), a personal computer (Gateway 2000 P5–166), and a data acquisition board (Digidata 1200B; Axon Instruments). Amperometric traces were analyzed using software written in-house. For analysis, spikes with signal-to-noise ratio >10 were counted as significant.

A picospritzer (General Valve, Fairfield, NJ) was used to deliver test solutions to the cells via a glass micropipette positioned ∼30 μm from the cell, as described elsewhere (31). Independent experiments have confirmed that this method of stimulation delivers drug at the stated concentration to the cell for the duration of stimulation. After stimulation, the drug is quickly diluted to negligible levels by diffusion into the surrounding medium. For all secretion experiments, 10 μmol/l L-783,281 in KRB was applied to isolated cells for 40 s, while secretion was measured by amperometry. To compare the amount of exocytotic events induced by different secretagogues, 100 nmol/l insulin or 18 mmol/l glucose was also applied to single β-cells for 40 s. A control solution consisting of drug carrier was applied to the same cell under investigation to ensure amperometric spikes were not due to an artifact of stimulation. To determine LY294002 sensitivity, an individual β-cell was stimulated with 10 μmol/l L-783,281 twice, with 10 min between stimulations, to demonstrate cell activity. After successful stimulations, the cell was incubated with 25 μmol/l LY294002 for 10 min and the same cell stimulated with L-783,281 again to compare the amount of secretory events. Control experiments were performed in which cells were incubated with drug carrier using a similar protocol.

β-cells attached to coverslips were incubated with 2 μmol/l fluo-4 acetoxymethyl ester or 2 μmol/l fura-2 acetoxymethyl ester in RPMI medium for 30 min at 37°C in 5% CO₂ before experiments to load dye into the cells (32). Cells were washed three times and bathed with KRB. Coverslips containing the cells formed the bottom of a microchamber kept at 37°C throughout all experiments via a microincubator (Medical Systems, Greenvale, NY).

Fluo-4 measurements were made using a Nikon RCM8000 confocal microscope that consisted of a Nikon Diaphot 100 inverted microscope and an Ar⁺ laser (INNOVA Enterprise 622; Coherent, Santa Clara, Calif). The 488-nm line from the Ar⁺ laser was focused onto cells by a 40×, 1.15 numerical aperture water immersion objective and scanned across the image plane via galvanometers. Emission was collected through the same objective, passed through a 520 ± 10 nm bandpass filter and onto a photomultiplier tube. Images were taken at 1 Hz and stored on an optical disk recorder (Panasonic TQ-FH332). For analysis, images were played back and the ratio (F₁/F₀) of each image (F₁) versus the first image (F₀) of the series was recorded (14).

Fura-2 fluorescence was measured using a SPEX CMX cation measurement system and DM3000M data acquisition software (Instruments SA Group, Edison, NJ) with a Zeiss Axiovert 100 microscope as previously described (18). Briefly, fura-2 loaded β-cells were excited at 1 Hz with 340 and 380 nm light focused onto cells by a 40×, 1.3 numerical aperture oil immersion objective. Emission was collected via the same objective, passed through a 400-nm long pass filter, 510 ± 10 nm bandpass filter, and 20 μm spatial filter and onto a photomultiplier tube. [Ca²⁺]_i was quantified as described elsewhere (33).

For [Ca²⁺]_i recordings, test solutions were applied to cells using the same protocol as described for secretion measurements. To test the sensitivity of L-783,281–induced Ca²⁺ to drugs, cells were incubated in KRB containing 100 nmol/l wortmannin, 25 μmol/l LY294002, 20 μmol/l nifedipine, or 1 μmol/l thapsigargin for 10 min before applying 10 μmol/l L-783,281 for 40 s. Only cells that responded to 200 μmol/l tolbutamide in the presence of wortmannin, LY294002, or thapsigargin were used in the study. The L-783,281–evoked increase in fluo-4 fluorescence recorded in the presence of drugs was compared with that obtained for identical stimulations in the absence of drugs.

All values are reported as means ± SE. Statistical significance of the difference between means was determined by a two-tailed Student’s t test.

Application of L-783,281 to isolated β-cells in the presence of 3 mmol/l glucose resulted in a concentration-dependent increase in [Ca²⁺]_i, whereas no increase for the inactive structural analog, L-767,827, was observed (Fig. 1). At 10 μmol/l L-783,281, the average maximal increase in fluo-4 fluorescence was 178 ± 10% (n = 18 cells) with an average latency between stimulation and onset of the increase of 22 ± 17 s. The maximal increase was comparable to that for glucose and tolbutamide stimulations ( Fig. 2). The increase in [Ca²⁺]_i was long-lived and generally lasted for the entire recording time of 200 s. Further measurements revealed that the [Ca²⁺]_i stayed within 10% of the maximum stimulated level more than 5 min after the end of stimulation. For comparison, a typical trace obtained with application of 200 μmol/l tolbutamide to a single cell is shown in Fig. 1. This trace is typical of traces obtained from six different cells.

To determine if the increase in [Ca²⁺]_i was associated with Ca²⁺ entry through voltage-dependent Ca²⁺ channels, the effect of nifedipine was investigated on the L-783,281–evoked [Ca²⁺]_i response. As shown in Fig. 2, pretreatment with 20 μmol/l nifedipine reduced the L-783,281–induced Ca²⁺ increase 33 ± 6% (n = 6 cells, P < 0.05). At this concentration, nifedipine completely abolishes the [Ca²⁺]_i increases because of stimulation by K⁺ or tolbutamide (data not shown), compounds that evoke an influx of extracellular Ca²⁺ through voltage-dependent Ca²⁺ channels. To determine if the rise in [Ca²⁺]_i was associated with release from intracellular Ca²⁺ stores, the effect of thapsigargin was examined. Thapsigargin has been shown to deplete intracellular Ca²⁺ stores by inhibiting sarco/endoplasmic reticulum Ca²⁺-ATPases (34). As shown in Fig. 2, pretreatment with 1 μmol/l thapsigargin for 10 min reduced the L-783,281–induced [Ca²⁺]_i change by 82 ± 4% (n = 6 cells, P < 0.001). (Although thapsigargin increases intracellular Ca²⁺ as it depletes the stores, this effect is transient, and measurements were made after [Ca²⁺]_i had returned to baseline.) Combined, these results suggest that most of the increase in [Ca²⁺]_i was due to intracellular release from thapsigargin-sensitive stores.

Application of L-783,281 to β-cells in the presence of 3 mmol/l glucose resulted in a series of current spikes in the amperometric trace indicative of exocytosis ( Fig. 3A). An average of 12.2 ± 2.1 exocytotic events per stimulation (n = 14 cells) were detected following application of L-783,281, whereas only 1.2 ± 0.3 exocytotic events per stimulation (n = 8 cells) were detected with 10 μmol/l L-767,827, the inactive structural analog. For comparison, stimulation with insulin or glucose for an identical period of time resulted in a similar number of spikes detected by amperometry as with L-783,281 ( Fig. 3B). To confirm that this response was due to effects at the insulin receptor, cells were preincubated with 100 nmol/l insulin for 10 min before application of 10 μmol/l L-783,281. Application of 10 μmol/l L-783,281 to these cells caused 0 ± 0 exocytotic events per stimulation, even though the cells were still viable as application of 40 mmol/l K⁺ to the same cells resulted in 14.8 ± 0.5 spikes (n = 4 cells, P < 0.05).

The potential role of IRS-1 in mediating the insulin secretion and increase in [Ca²⁺]_i evoked by L-783,281 was investigated by comparing the effect of the drug on β-cells from IRS-1^–/– and IRS-1^+/+ C57Bl/6 mice. As seen in Fig. 4, stimulation of IRS-1^–/– cells with L-783,281 resulted in no significant increases in [5Ca²⁺]_i, even though 18 mmol/l glucose was able to evoke substantial increases in [Ca²⁺]_i in these cells. The average increase induced by L-783,281 was 8 ± 3% (n = 5 cells) over basal values, which was not significantly different from the effect of drug carrier alone (data not shown) and significantly less (P < 0.05) than the increase of 119 ± 25% (n = 5 cells) observed in the IRS-1^+/+ β-cells from the same type of mouse. Stimulation of β-cells from IRS-1^–/– mice with L-783,281 resulted in 1.3 ± 0.1 (n = 5 cells) exocytotic events, which was significantly less (P < 0.05) than the 11.0 ± 2.0 (n = 5 cells) observed in control cells. The IRS-1^–/– cells still responded to stimulation with 18 mmol/l glucose, with an average of 6.8 ± 0.1 (n = 8 cells) exocytotic events detected per cell (see Fig. 4B). These results indicate that the presence of IRS-1 is required for L-783,281 to evoke increases in [Ca²⁺]_i and insulin secretion in β-cells.

Pancreatic β-cells were tested to determine if L-783,281–induced [Ca²⁺]_i changes were sensitive to the PI3-K inhibitors LY294002 and wortmannin. When cells were pretreated with 25 μmol/l LY294002, application of L-783,281 still increased fluo-4 fluorescence; however, the increase was just 31 ± 3% (n = 9 cells) of that evoked without LY294002. Similarly, application of L-783,281 in the presence of 100 nmol/l wortmannin caused an increase that was 48 ± 10% (n = 6 cells) of that observed without wortmannin. Both of these increases were significantly less (P < 0.05) than that observed without PI3-K inhibition.

Secretion was also sensitive to LY294002, as the number of detected exocytotic events was reduced to 21 ± 6% (n = 4 cells, P < 0.05) of the L-783,281–induced secretion in the presence of no LY294002 at a given cell (see Fig. 5). Without application of LY294002, successive applications of L-783,281 at a 10-min interval resulted in no reduction of spike number as the second stimulation yielded 97 ± 5% of the spikes from the first stimulation (n = 8 cells).

Our data support the conclusions that L-783,281 evokes increases in [Ca²⁺]_i and insulin exocytosis by an IRS-1/PI3-K sensitive mechanism at a nonstimulatory glucose concentration. Given the known ability of L-783,281 to activate IRTK and its downstream cascade, including IRS-1 and PI3-K in CHO cells, and the similarity of L-783,281 and insulin effects on β-cell [Ca²⁺]_i and secretion, it is reasonable to conclude that the effects of L-783,281 on β-cells are mediated by insulin receptor activation.

The effects of insulin signaling on β-cell [Ca²⁺]_i and exocytosis have been reported previously (13–15,18). These data support the idea that insulin binding to the insulin receptor promotes insulin secretion and results in increases in [Ca²⁺]_i via release of Ca²⁺ from the endoplasmic reticulum by a pathway that involves IRS-1 and PI3-K. Most relevant to the present study is the report of direct application of 100 nmol/l insulin to β-cells, which resulted in increases in [Ca²⁺]_i and exocytosis at β-cells (14,18).

Several similarities are observed in the acute effects of direct application of exogenous insulin and L-783,281 on β-cell [Ca²⁺]_i and secretion. In both cases, the [Ca²⁺]_i increase is characterized by a latency after application of stimulus, 20 ± 17 s and 12 ± 10 s for L-783,281 and insulin, respectively, followed by a slow increase to a maximum value (∼30 s) that plateaus and persists for over 120 s. This time course is in contrast to stimulation with depolarizing agents, such as tolbutamide, which have been shown to have a latency of 1 to 2 s, a peak in [Ca²⁺]_i at ∼5 s, and a return to near baseline within a minute after stimulation (Fig. 1) (14). The pharmacology of the L-783,281–induced [Ca²⁺]_i increase also resembles that initiated by insulin. Nifedipene has no effect on insulin-induced increases (14) and decreases the L-783,281–evoked increase by just 33%. In contrast, thapsigargin has strong effects, resulting in a decrease of 75 and 82% for insulin (14) and L-783,281–initiated [Ca²⁺]_i increases, respectively. PI3-K inhibitors such as wortmannin also significantly inhibit the [Ca²⁺]_i increases evoked by both insulin and L-783,281. Insulin and L-783,281 had essentially no effect on β-cell [Ca²⁺]_i in the IRS-1 knockout model. Likewise, exocytosis evoked by insulin and L-783,281 was strongly inhibited by PI3-K inhibitors and was nonexistent in β-cells from IRS-1–null mice. The similarities between the effects of L-783,281 and insulin on β-cell secretion and [Ca²⁺]_i support the idea that these compounds exert their effects through a similar mechanism that involves activation of IRS-1/PI3-K.

Another similarity involved in insulin signaling by insulin and L-783,281 is in the effect of regulating insulin gene expression in β-cells. Insulin upregulates its transcription in an autocrine manner (12). Similarly, L-783,281 acts to enhance insulin transcription by acting through the insulin receptor/IRS-1 pathway (13). Thus, insulin signaling helps to regulate its bioavailability and secretion.

Our data indicate that the majority of the [Ca²⁺]_i increase that occurs during application of L-783,281 comes from thapsigargin-sensitive stores. Generally, these intracellular stores are considered to be the endoplasmic reticulum and are regulated by IP₃ or ryanodine receptors; however, [Ca²⁺]_i can also be increased by inhibition of SERCA (sarco/endoplasmic reticulum calcium ATPase) pumps, resulting in passive diffusion of Ca²⁺ out of the endoplasmic reticulum. Several lines of evidence point to SERCA inhibition resulting from activation of IRTK or its downstream effectors, including the ability of IRS-1 to interact with SERCA pumps (35), increased basal [Ca²⁺]_i by IRS-1 inhibition of SERCA pumps, thereby inhibiting Ca²⁺ uptake into ER (15), and regulation of SERCA gene expression by activation of insulin signaling (13). The present study agrees with these previous reports in that deletion of IRS-1 inhibits the L-783,281–stimulated [Ca²⁺]_i increase. The observations of the effects of insulin signaling on [Ca²⁺]_i are consistent with the hypothesis that activation of the insulin receptor, whether by insulin or L-783,281, releases Ca²⁺ from the endoplasmic reticulum via an interaction involving IRS-1 and SERCA. Whether this Ca²⁺ is responsible for activating exocytosis has yet to be proven.

Initial studies of L-783,281 demonstrated that this compound enhanced glucose uptake in vitro and was antidiabetic inasmuch as it restored normoglycemia in db/db and ob/ob mice (1). It is tempting to speculate that part of the reason for an antidiabetic effect of L-783,281 was an increase in insulin secretion due to activation of the pathway observed here; however, it would be premature to reach this conclusion, since the experiments in this report involved only transient application of L-783,281 in the presence of 3 mmol/l glucose. Further experiments are necessary to determine if L-783,281 affects insulin secretion at stimulatory glucose levels, with longer-term stimulation, in whole islets, and in vivo. Demonstration of enhanced insulin secretion in vivo may be difficult to demonstrate because of multiple effects of the drug, and in vivo glucose homeostasis mechanisms may mask effects at the level of β-cells. For example, L-783,281 lowers glucose levels by activating glucose transport and potently increases insulin sensitivity in peripheral tissues (1); therefore, insulin demand during a glucose challenge is likely less, which could result in a net decrease in insulin secretion even if the drug has a stimulatory effect on insulin secretion by direct action on β-cells.

Type 2 diabetes is characterized by both insulin resistance and impaired insulin secretion, with the possible link between these two symptoms remaining obscure; however, the recent discovery of active insulin receptors on β-cells that promote insulin secretion, insulin synthesis, and β-cell growth has led to the hypothesis that impaired insulin signaling could account for both symptoms (36,37,38,39). We have demonstrated that the insulin mimetic L-783,281, a fungal metabolite previously shown to increase glucose uptake in peripheral tissues, also induces an increase in [Ca²⁺]_i and insulin secretion from pancreatic β-cells at nonstimulatory glucose concentrations. The effects appear to involve insulin signaling, since they are blocked by knockout of IRS-1 and are sensitive to inhibitors of PI3-K. Other groups have demonstrated that L-783,281 promotes insulin synthesis (13). These results suggest that the actions of insulin mimetics in β-cells should be further investigated as a possible tool in studying insulin signaling in β-cells and for a potential role in diabetes treatments.

This work was supported by National Institutes of Health (NIH) Grant DK46960. R.N.K. acknowledges support from Clinician Scientist Development Award (K08) and DK31036 (to C.R.K.) and from NIH DK02885-01.