Expression of the Novel Epithelial Ca 2+ Channel ECaC1 in Rat Pancreatic Islets

Abstract

The epithelial Ca²⁺ channel, ECaC1, is primarily expressed in the apical membrane of vitamin D-responsive tissues. This study characterizes for the first time the presence of this novel channel in pancreatic tissue by reverse transcriptase-polymerase chain reaction and immunohistochemistry. In addition, the expression of ECaC1 was investigated in an animal model for Type 2 diabetes mellitus, the Zucker diabetic fatty (ZDF) rat. Identical staining patterns for ECaC1 and insulin were observed, whereas no co-localization of ECaC1 with glucagon was found. ECaC1, insulin, and prohormone convertase 1 (a neuroendocrine endoprotease expressed in secretory granules) showed a similar punctate staining. ECaC1 co-localized with the Ca²⁺ binding protein calbindin-D₂8K in the β-cells. Furthermore, in contrast to wild-type rats, in ZDF rats aging led to a progressive decrease in both insulin and ECaC1 staining. Plasma 1,25-dihydroxyvitamin D₃ levels were similar in both control and ZDF rats and decreased with aging. Taken together, our findings indicate that this novel Ca²⁺ channel may play a role in the regulation of endocrine Ca²⁺ homeostasis.

THE epithelial Ca²⁺ channel ECaC1 was recently cloned from rabbit kidney and was thought to play a crucial role in Ca²⁺ homeostasis as gatekeeper of active transcellular Ca²⁺ transport (Hoenderop et al. 1999a, 2000b). ECaC1 protein expression is exclusively observed in 1,25-dihydroxyvitamin D3-responsive tissues, i.e., kidney and intestine (Hoenderop et al. 2000a). In these tissues, calcium homeostasis is mainly regulated by active transepithelial transport of Ca²⁺ from the lumen to the blood (Friedman and Gesek 1995; Hoenderop et al. 2000c), which is a three-step process consisting of passive Ca²⁺ entry across the apical membrane via ECaC1, cytosolic diffusion facilitated by 1,25-dihydroxyvitamin D₃-dependent calcium binding proteins (calbindins), and active extrusion across the basolateral membrane (Bindels 1993; Hoenderop et al. 2000c). In kidney and intestine, ECaC1 expression was indeed observed in the apical membrane and co-localized with other Ca²⁺ transport proteins, including calbindin-D (Hoenderop et al. 2000a). Recently, it was shown that ECaC1 is a selective Ca²⁺ channel, exhibiting Ca²⁺-dependent auto-regulation and being vitamin D-dependent (Vennekens et al. 2000; Hoenderop et al. 2001). Moreover, recent studies in humans demonstrated mRNA expression of ECaC1 not only in kidney and intestine but also in brain, testis, prostate, placenta, and pancreas (Müller et al. 2000).

In pancreatic β-cells, calcium plays a central role in stimulus-secretion coupling and is regulated by glucose and other secretagogues. An increase in intracellular Ca²⁺ concentrations ([Ca²⁺]_i) is necessary to induce insulin secretion (Roe et al. 1993), which occurs in a pulsatile and oscillatory manner coincident with intracellular calcium oscillations (Gilon et al. 1993). For this, a close interplay between plasma membrane ion channels (such as the ATP-regulated K⁺ channels and the voltage-gated L-type Ca²⁺ channels) and the secretory granules, containing high Ca²⁺ levels, is prerequisite (Ämmälä et al. 1991). However, it is not yet clear how the complex Ca²⁺-mediated transduction pathway resulting in stimulus-secretion coupling occurs in the β-cell and if other Ca²⁺ channels, such as ECaC1, can play a role in these processes. In Type 2 diabetes mellitus (Type 2 DM), glucose-induced insulin secretion is impaired. Both in isolated pancreatic islets exposed to high glucose (Okamoto et al. 1992) and in diabetic rats (Okamoto et al. 1995), it was observed that this impairment was due to the inability of glucose to increase the [Ca²⁺]_i.

As in kidney, in pancreatic β-cells Na⁺/Ca²⁺ exchanger (NCX), parathyroid hormone-related peptide (PTHrP), calbindin-D_28K, and receptors for 1,25-dihydroxyvitamin D₃ have been found (Clark et al. 1980; Drucker et al. 1989; Johnson et al. 1994; Eylen et al. 1997). It is believed that in the β-cell, NCX participates in the control of [Ca²⁺]_i and that both NCX and PTHrP modulate insulin release (Eylen et al. 1997; Villanueva-Peñacarrillo et al. 1999). Moreover, vitamin D has been postulated to modulate calcium metabolism of the β-cell via an effect at the transcriptional level, including the production of calbindin-D_28K (Lee et al. 1984). In addition, in vitamin D-deficient rats, insulin secretion is impaired (Norman et al. 1980) but is restored by vitamin D treatment (Ishida et al. 1983). Sooy et al. (1999) suggested that calbindin-D_28K could control the rate of insulin release via regulation of [Ca²⁺]_i. Altogether, this indicates that in the pancreatic β-cell Ca²⁺ homeostasis is a complex process in which many Ca²⁺ transport proteins play a role.

The aim of the present study was to investigate by RT-PCR and immunohistochemistry whether ECaC1 is expressed in rat pancreatic tissue. In our studies, we included the Zucker diabetic fatty (ZDF) rat, an animal model for Type 2 DM. The ZDF rat becomes mildly hyperglycemic between 7 and 10 weeks of age and thereafter spontaneously develops overt diabetes (Clark et al. 1983; Tokuyama et al. 1995). We used the ZDF rat to investigate whether ECaC1 expression is affected during the progression of diabetes. We therefore studied ECaC1 expression longitudinally by IHC in control and ZDF rats 6–24 weeks of age. Furthermore, we measured plasma 1,25-dihydroxyvitamin D₃ levels to examine if ECaC1 expression in the pancreas is dependent on vitamin D concentrations.

Materials and Methods

Experimental Animals

Male Zucker diabetic fatty (ZDF/Gmi-fa/fa or ZDF) and lean Zucker control (ZDF/Gmi-+/? or control) rats were obtained from Genetic Models (Indianapolis, IN). All rats were maintained on a 12-hr light/dark cycle (lights on at 0700, lights off at 1900 hr), and had ad libitum access to tapwater and chow (RMH-B: 22.8% protein, 5.1% fat, 4.2% fiber and 67.9% other carbohydrates; Hope Farms, Woerden, The Netherlands). Animals were studied at 6, 12, and 24 weeks of age. For immunofluorescence studies, three control and three ZDF rats were examined at each time point. All animal experiments were performed according to the principles of laboratory animal care (NIH publication no. 85–23, revised 1985) and Dutch laws on the protection of animals.

Blood Sampling and 1,25-Dihydroxyvitamin D₃ Determination

Blood samples for 1,25-dihydroxyvitamin D₃ determinations were obtained from control and ZDF rats under isoflurane anesthesia at 6, 12, 21, and 24 weeks of age by orbital puncture and collected in precooled tubes containing dry lithium-heparin additive (30 UP U/tube; Vacutainer, Becton Dickinson, Etten-Leur, The Netherlands). Tubes were gently shaken and centrifuged for 10 min at 1500 × g (4C). Plasma samples were stored at −80C until assayed. Plasma 1,25-dihydroxyvitamin D₃ levels were measured using a radioreceptor assay after extraction of the samples, followed by paper chromatography as described previously (Hoof et al. 1993).

RT-PCR

Under isoflurane anesthesia of the rats, the pancreas was rapidly removed and cleared of fat. Small parts of the pancreas were immediately frozen in liquid nitrogen. For RNA isolation, the tissue was ground with a mortar and pestle that were pretreated with 70% ethanol in diethyl pyrocarbonate (DEPC; Sigma Chemical, St Louis, MO). Total RNA was isolated from the resulting powder using TRIzol (Gibco BRL; Gaithersburg, MD) according to the manufacturer's instructions, except that an additional acidic phenol-chloroform extraction was performed to remove excessive protein. For cDNA synthesis, 2–5 μg of total RNA was incubated with 5 mU Pd(N)₆ (Pharmacia; Roosendaal, The Netherlands) at 70C for 10 min. Thereafter, the RNA was reverse transcribed (RT) in a final volume of 20 μl containing 10 mM dithiothreitol, 0.5 mM dNTPs, 40 U RNase inhibitor (Promega; Madison, WI), and 100 U of Superscript II (Gibco BRL) for 1.5 hr at 37C. The samples were then incubated at 70C for 10 min and subsequently diluted to a final volume of 50 μl. For RT-PCR analysis, 2.5 μl of cDNA was used as a template in a 25-μl amplification mixture containing Sybr Green (PE Applied Biosystems Benelux; Nieuwerkerk a/d IJssel, The Netherlands), 1 mM dNTPs, 3 mM MgCl₂, 0.6 U Taq Gold polymerase (PE Applied Biosystems) and 0.6 μM of each primer. The primer sets used in the PCR were: for insulin (118-bp fragment), sense 5′-gggaacgtggtttcttctacaca-3′, antisense 5′-tccagtgccaaggtctgaagat-3′; for ECaC1 (158 bp), sense 5′-agcaatagccaccgtggatg-3′, antisense 5′-atgtcccagggtgtttcgac-3′; for calbindin-D_28K (70 bp), sense 5′-ggagctgcagaacttgatccag-3′, antisense 5′-ctcaggtgatagctccaatccag-3′; and for β-actin (84 bp), sense 5′-accctaaggccaaccgtga-3′, anti-sense 5′-cagcctggatggctacgtacat-3′. After an initial step at 95C for 10 min, a real-time semiquantitative PCR of 40 cycles was performed (PE Applied Biosystems), each cycle consisting of 15-sec denaturation at 95C and 1-min annealing at 60C. Cycle threshold (CT) values were determined using the PE Applied Biosystems software. CT is defined as the cycle number at which the amount of amplified target passes a fixed threshold above baseline. As a control for the amount of RNA used, β-actin was used.

Antisera

The guinea pig anti-rabbit ECaC1 polyclonal antiserum was directed against a peptide corresponding to the 10 carboxy-terminal amino acid residues 698–707 (NH2-SHRG-WEILRQ-COOH) of ECaC1. Anti-insulin and anti-glucagon polyclonal antisera were obtained from Dakopatts (Copenhagen, Denmark). The anti-prohormone convertase 1 (PC1) polyclonal antiserum was kindly provided by Dr. W.J.M. Van de Ven (Leuven, Belgium) and has been extensively characterized (Jackson et al. 1997). The polyclonal antiserum against calbindin-D_28K was a generous gift of Dr. S. Christakos (New Jersey). Antisera were diluted in TNB buffer [0.15 M NaCl, 0.1 M Tris-HCl, pH 7.5, 0.5% (w/v) blocking reagent from NEN Life Science Products (Zaventem, Belgium)]. The guinea pig antiserum to ECaC1 was used at a dilution of 1:3000, the guinea pig antiserum to human insulin at 1:12,000, the rabbit antiserum to human glucagon 1:500, the rabbit antiserum to human PC1 and the rabbit antiserum to rat calbindin-D_28K both at 1:600. As negative controls, pre-immune serum and TNB buffer were used instead of the primary and secondary antisera. Antiserum pre-absorbed for 1 hr with 10 μg/ml ECaC1 peptide was used as a negative control for the ECaC1 antiserum.

Immunofluorescence Studies

Under isoflurane anesthesia of the rats, the pancreas was rapidly removed and cleared of fat. Small parts of the pancreas were fixed overnight at 4C in 4% phosphate-buffered formaldehyde (pH 7.4). The tissues were then rinsed in PBS, dehydrated in a graded series of ethanol, rinsed in xylene two times for 30 min each, and impregnated twice for 30 min with Paraplast Plus (Oxford Labware; St Louis, MO) under vacuum. Paraffin sections 4 μm thick were mounted on SuperFrost Plus glass slides (Merck; Darmstadt, Germany) and kept overnight at 37C. For immunofluorescence studies, tissue sections were dewaxed with two changes of xylene, rehydrated through graded ethanol, and then washed with PBS (pH 7.4). Thereafter, sections were boiled for 2 min in citrate buffer (0.01 M C₅H₆Na₃O₇ and 0.01 M C₆H₈O₇, pH 6), cooled to room temperature (RT), and thoroughly rinsed with TN buffer (0.15 M NaCl, 0.1 M Tris-HCl, pH 7.5). Sections were incubated with TNB buffer for 30 min before overnight incubation at 4C with the primary antibodies. Then the sections were rinsed three times for 5 min in TNT buffer [TN-buffer containing 0.05% (v/v) Tween-20]. Sections stained for ECaC1 and insulin were incubated for 1 hr at RT with goat anti-guinea pig Alexa 594-conjugated anti-IgG (1:300) (Molecular Probes; Eugene, OR). Sections double stained for ECaC1 and glucagon were incubated with both goat anti-guinea pig Alexa 594-conjugated anti-IgG (1:300) and goat anti-rabbit Alexa 488-conjugated anti-IgG (1:300) (both from Molecular Probes). The sections double stained with ECaC1 and PC1 or with ECaC1 and calbindin-D_28K were incubated for 30 min at RT with pig anti-rabbit biotinylated antiserum (1:1500) (NEN Life Science Products; Boston, MA), rinsed three times in TNT buffer, and then incubated for 30 min at RT with streptavidin-HRP antiserum (1:100) (NEN Life Science Products). After three 5-min rinses in TNT buffer, the peroxidase was visualized by incubation with fluorescent tyramides (1:50 in amplification solution) (NEN Life Science Products) for 7 min. For detection of ECaC1, these double stained sections were incubated for 1 hr at RT with goat anti-guinea pig Alexa 594-conjugated anti-IgG (1:300) (Molecular Probes). All sections were rinsed thoroughly with TNT buffer, rinsed in graded methanol, and mounted in Mowiol (Hoechst; Frankfurt, Germany) containing 2.5% (w/v) NaN3. Photographs were taken with a Zeiss Axioskop microscope equipped for epifluorescence illumination using Kodak EPH P1600X film.

Results

ECaC1 mRNA Expression in Rat Pancreatic Tissue

To establish if ECaC1 is expressed in the rat pancreas, we performed RT-PCR analysis on RNA isolated from total pancreatic tissue of 6-week-old control and ZDF rats. In addition, mRNA expression of insulin and calbindin-D_28K was analyzed in the pancreas of both control and ZDF rats. In control rats, mRNA expression was found for ECaC1 (C_T 33 ± 1), insulin (C_T 21 ± 1), calbindin-D_28K (C_T 34 ± 1), and β-actin (C_T 22 ± 1). In addition, between control and ZDF rats no significant differences in C_T values were observed, suggesting that mRNA expression levels of these proteins were similar in both rat strains. No amplification was found when the RT step was omitted or in the water controls. Analysis of β-actin mRNA was used as a control for RNA isolation and cDNA synthesis.

Immunofluorescence Localization of ECaC1 in Rat Pancreatic Islets

To localize ECaC1 protein expression in pancreatic tissue, immunofluorescence experiments were performed. In the rat pancreas, identical labelling patterns were found for ECaC1 and insulin, i.e., immunopositive staining was present in the pancreatic islets, whereas no staining was observed in the exocrine pancreas (Figure 1A). At higher magnification, punctate ECaC1 and insulin staining was found within the cytoplasm of the endocrine cells, suggesting a localization of ECaC1 in the secretory granules of the β-cells (Figure 1B). Sections incubated with ECaC1 antiserum pre-absorbed for 1 hr with ECaC1 peptide or pre-immune antiserum were devoid of any immunostaining (Figure 2D, left and middle columns).

To analyze in more detail the localization of ECaC1 in the pancreatic islets, we performed double immunofluorescence studies. Labeling of rat pancreatic tissue with anti-ECaC1 and anti-glucagon antibodies revealed that ECaC1 did not co-localize with glucagon, indicating that ECaC1 is not expressed in pancreatic α-cells (Figure 2A). Double immunofluorescence experiments showed complete co-localization in the β-cells of ECaC1 and PC1, a neuroendocrine endoprotease expressed in secretory granules (Figure 2B). In agreement with previous studies (Bourlon et al. 1996), calbindin-D_28K staining was detected in both α-and β-cells, with the clearest staining being found in the α-cells. Double staining of ECaC1 with calbindin-D_28K revealed their co-localization in the β-cells but not in the α-cells, suggesting that, in the pancreas, ECaC1 expression is restricted to the β-cells (Figure 2C). A control was performed in which as secondary antibody a goat anti-rabbit IgG was employed in combination with the guinea pig anti-ECaC1 antibody as the primary antibody to detect possible crossreactivity (Figure 2D, right). No immunopositive staining was observed under these experimental conditions and crossreactivity can therefore be excluded.

Figure 1

Immunofluorescence staining of ECaC and insulin in rat pancreatic islets. ECaC and insulin staining is present in the cytoplasm of cells in the pancreatic islets and absent in the exocrine pancreatic cells (A). Bar = 100 μm. At higher magnification (B), the staining in the endocrine cells appears to be punctate, both for insulin and ECaC. Bars = 10 μm.

Figure 2

Double immunofluorescence staining in rat pancreatic islets of ECaC with glucagon, prohormone convertase 1 (PC1) and calbindin-D_28K. ECaC staining was observed in the β-cells and glucagon staining in the α-cells. No co-localization of ECaC and glucagon was found, as shown by the superimposed picture (A). PC1 staining was observed in the pancreatic β-cells. The co-localization of ECaC and PC1 was highlighted by the yellow color in the β-cells in the superimposed picture (B). Calbindin-D_28K staining was present in both α- and β-cells, the most intense staining being found in the α-cells. The superimposed picture revealed co-localization of ECaC and calbindin-D_28K in the β-cells (C). Sections incubated with ECaC1 antiserum pre-absorbed for 1 hr with ECaC1 peptide (D, left) or pre-immune antiserum (D, middle). Control was performed in which as secondary antibody a goat anti-rabbit IgG was employed in combination with the guinea pig anti-ECaC1 antibody as the primary antibody (D, right). Bars = 25 μm.

ECaC1 Expression During the Progression of Diabetes in ZDF Rats

Because glucose-induced insulin secretion is impaired in Type 2 DM, probably due to the inability of glucose to increase intracellular Ca²⁺ concentrations, we studied the expression of ECaC1 in ZDF rats, an animal model for Type 2 DM. In control rats between 6 and 24 weeks of age, no difference in the amount of immunostaining of both ECaC1 and insulin was observed (Figure 3A). In 6-week-old ZDF rats, the amount of ECaC1 and insulin staining was about the same as that in control rats of the same age (Figure 3B). However, in ZDF rats, aging affected the staining patterns of both ECaC1 and insulin, i.e., from 12 weeks onwards a decrease in the amount of immunostaining of these two proteins was observed. At 24 weeks of age the amount of insulin staining had progressively decreased in the ZDF rat compared to that in 6-week-old ZDF rats. Remarkably, the amount of immunostaining of ECaC1 in 24-week-old ZDF rats was also decreased compared to the 6-week-old ZDF rats (Figure 3B).

Plasma 1,25-Dihydroxyvitamin D₃ Concentrations in Control and ZDF Rats

To establish if the reduction of ECaC1 expression in the ZDF rats was accompanied by a change in 1,25-dihydroxyvitamin D₃ levels, we measured plasma vitamin D concentrations in both control and ZDF rats from 6 to 24 weeks of age. Plasma 1,25-dihydroxyvitamin D₃ levels were identical in control and ZDF rats and decreased with aging (Figure 4). At 6 weeks of age, in control and ZDF rats 1,25-dihydroxyvitamin D₃ levels were 548 ± 15 pmol/liter and 556 ± 27 pmol/liter, respectively. These levels decreased progressively to 295 ± 26 pmol/liter and 320 ± 20 pmol/liter in control and ZDF rats of 24 weeks of age, respectively, showing that although a decrease in plasma vitamin D concentrations was observed in the ZDF rats, the decrease in ECaC1 expression in these rats was not due to a reduction in vitamin D levels (Figure 4).

Discussion

In pancreatic β-cells, calcium plays an important role in insulin secretion since an increase in [Ca²⁺]_i is a prerequisite to facilitate stimulus-secretion coupling. At present, the complex calcium-mediated transduction pathway in the β-cells is not clear, but a close in terplay between plasma membrane ion channels and secretory granules may be needed (Ämmälä et al. 1991). In this respect, our new finding that the epithelial Ca²⁺ channel ECaC1 is expressed in the endocrine pancreatic islets is of interest. Thus far, ECaC1 expression was mainly observed at the apical membrane of 1,25-dihydroxyvitamin D₃-responsive tissues, i.e., in kidney and intestine (Hoenderop et al. 2000a). In the cytoplasm of the β-cells, the immunostaining of both ECaC1 and insulin was punctate, suggesting that ECaC1 is expressed in secretory granules, probably in the granular membrane. In addition, ECaC1 co-localized with PC1, an endoprotease present in secretory granules and involved in the conversion of proinsulin to insulin (Bailyes et al. 1992). In contrast, in kidney and intestine ECaC1 is localized to the plasma membrane (Hoenderop et al. 2000a). Our observation that in pancreatic β-cells ECaC1 expression is restricted to secretory granules suggests that it might play a role in the regulation of endocrine Ca²⁺ homeostasis.

Figure 3

Immunofluorescence staining of ECaC and insulin in pancreatic islets of 6-week- and 24-week-old control (A) and 6-week- and 24-week-old ZDF (B) rats. In the control rats, aging did not have an effect on the amount of immunostaining of both ECaC and insulin (A). However, in the β-cells of 24-week-old ZDF rats the amount of staining of both ECaC and insulin was decreased compared to that in 6-week-old ZDF rats (B). Bars = 25 μm.

Figure 4

Plasma levels of 1,25-dihydroxyvitamin D₃ in control (open circles) and ZDF (filled circles) rats between 6 and 24 weeks of age. Data are presented as mean ± SEM (four or five animals).

The secretory granules of β-cells contain high levels of Ca²⁺ implicated in the sorting and processing of precursor proteins (Davidson et al. 1988). In the pancreatic β-cell, ECaC1 may serve as an efflux pathway for Ca²⁺ from the secretory granules into the cytoplasm to increase [Ca²⁺]_i. Increased cytoplasmic Ca²⁺ concentrations play an essential role in triggering exocytosis (Bokvist et al. 1995). To release Ca²⁺ via an efflux pathway, hyperpolarization of the granular membrane would be needed. In this respect it is interesting to note that hyperpolarization of the Xenopus laevis oocyte plasma membrane indeed allows Ca²⁺ transport via ECaC1 (Hoenderop et al. 1999b). In an alternative mechanism, ECaC1 may act at the plasma membrane after the exocytotic event, e.g., providing an influx pathway for extracellular Ca²⁺ to retain high cytoplasmic Ca²⁺ concentrations. Clearly, further studies are needed to establish the exact role of ECaC1 in the pancreatic β-cell.

In pancreatic β-cells, the maintenance of intracellular Ca²⁺ homeostasis is a complex process because in these cells many proteins, such as NCX, PTHrP, calbin-din-D_28K, receptors for 1,25-dihydroxyvitamin D₃, and now also ECaC1, have been found that are postulated to play a role in Ca²⁺ homeostasis and insulin release. On the basis of the analysis of calbindin-D_28K knockout mice, Sooy et al. (1999) suggested that calbindin controls the rate of insulin release via regulation of [Ca²⁺]_i. In addition, 1,25-dihydroxyvitamin D₃ can enhance insulin secretion (Clark et al. 1981), indicating that both calbindin-D_28K and vitamin D, together with calcium, may control β-cell function. However, Bourlon et al. (1996) showed that, in β-cells, 1,25-dihydroxyvitamin D₃ can stimulate insulin release but independently of calbindin-D_28K. In contrast, these authors observed that in α-cells vitamin D negatively regulates both calbindin-D_28K expression and glucagon release, suggesting that the role of calbindin and vitamin D in Ca²⁺ handling is different in α- and β-cells. In agreement with the study of Bourlon et al. (1996), we observed calbindin-D_28K immunostaining in both α- and β-cells, the clearest staining being present in the α-cells. Thus far, ECaC1 has always been found to co-localize with calbindin-D (Hoenderop et al. 2000a; Müller et al. 2000). Our observation that ECaC1 and calbindin-D_28K co-localization is exclusively found in the β-cells but not in the α-cells suggests that the modulation of [Ca²⁺]_i, in which both ECaC1 and calbindin-D_28K may play a role, is different in α- and β-cells. This is in line with the notion of Bourlon et al. (1996).

In ZDF rats, progression of diabetes is accompanied by a reduced expression of insulin (Pick et al. 1998). We now find that the expression of ECaC1 also is decreased in ZDF rats. The decrease in both insulin and ECaC1 was found at about 12 weeks of age and proceeded in a progressive way, at least until the age of 24 weeks. In age-matched control rats, the decrease in insulin and ECaC1 expression was not observed, which suggests that this reduction is due to diabetes-related factors. One of the major characteristics of Type 2 DM is the impairment of glucose-induced insulin secretion. However, the underlying mechanism leading to this impairment is not known but may be based on either defective steps proximal to the elevation of [Ca²⁺]_i or the exocytotic process distal to the increase in [Ca²⁺]_i (Okamoto et al. 1995). As mentioned above, we propose a role for ECaC1 in cellular Ca²⁺ handling. The decrease in ECaC1 expression in 24-week-old ZDF rats detected in the present study might lead to an insufficient increase of [Ca²⁺]_i which, in turn, may impair insulin secretion. However, not only was ECaC1 expression decreased during development of diabetes but also insulin expression, indicating that the expression levels of several proteins of the β-cell are affected. Although multiple alterations in various protein levels in the β-cell have been shown to contribute to the dysfunction of this cell type, the primary defect is still not known. Male homozygous ZDF rats become hyperglycemic at about the age of 7 weeks and thereafter develop overt diabetes (Clark et al. 1983). Because no differences were detected in ECaC1 expression between 6-week-old control and ZDF rats and because the decrease occurs from 12 weeks of age onwards, this reduction in ECaC1 expression is probably not the primary event leading to β-cell dysfunction in diabetes.

As already mentioned, vitamin D together with calcium may modulate β-cell function. Our data show that aging of both control and ZDF rats is accompanied by a decrease in plasma 1,25-dihydroxyvitamin D₃ levels, which confirms previous observations (Johnson et al. 1995). The reduction was similar in control and ZDF rats, indicating that the decrease of ECaC1 expression in the ZDF rats was not caused by the decrease of plasma 1,25-dihydroxyvitamin D₃ concentrations and that, in the ZDF rats, the impaired β-cell function is not due to changes in vitamin D₃ concentrations.

In conclusion, ECaC1 is abundantly expressed in rat pancreatic islets. Punctate cytoplasmic ECaC1 expression was restricted to the β-cells, suggesting a localization of ECaC1 to the membranes of secretory granules. We propose a function of ECaC1 in the maintenance of intracellular Ca²⁺ homeostasis either as part of a Ca²⁺ efflux pathway from granules to the cytoplasm or as a component of a Ca²⁺ influx pathway for extracellular Ca²⁺ into the cytoplasm. Furthermore, in ZDF rats a diabetes-related reduction of ECaC1 expression, independent of plasma 1,25-dihydroxyvitamin D₃ levels, was demonstrated. This decrease in ECaC1 expression in the ZDF rat may contribute to the impaired stimulus-secretion coupling observed in diabetes. Altogether, the present study suggests that ECaC1 may play an important role in the regulation of Ca²⁺ homeostasis in pancreatic β-cells.