Glucose regulates free cytosolic Zn²⁺ concentration, Slc39 (ZiP), and metallothionein gene expression in primary pancreatic islet β-cells

Abstract

Zn²⁺ is an important cofactor for insulin biosynthesis and storage in pancreatic β-cells. Correspondingly, polymorphisms in the SLC30A8 gene, encoding the secretory granule Zn²⁺ transporter ZnT8, are associated with type 2 diabetes risk. Using a genetically engineered (FRET)-based sensor (eCALWY-4), we show here that elevated glucose time-dependently increases free cytosolic Zn²⁺ ([Zn²⁺](cyt)) in mouse pancreatic β-cells. These changes become highly significant (853 ± 96 pm versus 452 ± 42 pm, p < 0.001) after 24 h and are associated with increased expression of the Zn²⁺ importer family members Slc39a6, Slc39a7, and Slc39a8, and decreased expression of metallothionein 1 and 2. Arguing that altered expression of the above genes is not due to altered [Zn²⁺](cyt), elevation of extracellular (and intracellular) [Zn²⁺] failed to mimic the effects of high glucose. By contrast, increases in intracellular cAMP prompted by 3-isobutyl-1-methylxanthine and forskolin partially mimicked the effects of glucose on metallothionein, although not ZiP, gene expression. Modulation of intracellular Ca²⁺ and insulin secretion with pharmacological agents (tolbutamide and diazoxide) suggested a possible role for changes in these parameters in the regulation of Slc39a6 and Slc39a7 but not Slc39a8, nor metallothionein expression. In summary, 1) glucose induces increases in [Zn²⁺](cyt), which are then likely to facilitate the processing and/or the storage of insulin and its cocrystallization with Zn²⁺, and 2) these increases are associated with elevated expression of zinc importers. Conversely, a chronic increase in [Zn²⁺](cyt) following sustained hyperglycemia may contribute to β-cell dysfunction and death in some forms of diabetes.

FIGURE 1.

Imaging the effect of elevated glucose on [Zn²⁺]_cyt with eCALWY-4 in primary mouse β-cells. A, CD1 mouse islets were dissociated into single cells or clusters of cells, infected with eCALWY-4-expressing adenovirus, and plated on 24-mm coverslips. Cells were imaged with an Olympus IX microscope and excited at 433 nm, and the emission was recorded at 465 nm and 535 nm for cerulean and citrine, respectively (22). B, partially dissociated mouse islets seeded on converslips, were infected with eCALWY-4-expressing adenovirus, formaldehyde-fixed, permeabilized, and immunostained with anti-insulin antibody and Alexa Fluor 568-coupled secondary antibody, whereas DAPI was used for nuclear staining. Cell clusters containing insulin-positive cells were imaged by confocal microscopy. Scale bar = 10 μm. C, representative trace showing the changes in fluorescence ratio of a cell infected with eCALWY-4 adenovirus or eCALWY-nonbinding (D) and perifused with the indicated solutions. KRB, Krebs-Ringer buffer; TPEN, 50 μm; ZnPyr, 5 μm pyrithione plus 100 μm ZnCl₂. E and F, analysis of free [Zn²⁺]_cyt concentration after 2 and 24 h, respectively, of high glucose treatment. Bars represent mean ± S.E. Number of cell imaged (n) on three different days of experiments is given in brackets under each bar. ***, p < 0.001.

FIGURE 2.

Effect of elevated glucose concentrations on the expression of the genes implicated in Zn²⁺ homeostasis in mouse pancreatic islets. Mouse islets were incubated for 1 h at 3 mm glucose before incubation for 24 h (A) or 2 h (B) with either 3 mm (white bars) or 16.7 (black bars) glucose. Total RNA was extracted, and qPCR analysis of Slc39a1–12, Slc30a1–10, Mt-1, and Mt-2 was performed. The mRNA levels were normalized to those of a housekeeping gene (cyclophilin A). Bars represent mean ± S.E. (n = 4). *, p < 0.05.

FIGURE 3.

Effect of elevated glucose concentrations on the expression of proteins implicated in Zn²⁺ homeostasis in mouse pancreatic islets. A, total cell lysates were loaded onto 12% SDS-PAGE, which was subsequently transferred onto a PVDF membrane (see “Experimental Procedures”). The membrane was blotted for ZiP6 (1:200), ZiP7 (1:200), and β-tubulin (1:1000) and with an HRP-lined anti-rabbit (1:5000) secondary antibody. B, quantification of three different immunoblot analyses for ZiP6 and ZiP7. The same area of interest was drawn around the specific bands for ZiP, ZiP7 as well as the corresponding tubulin band, and the intensity was measured using ImageJ software. The ratio of intensity between ZiP or ZiP7 signals and tubulin were plotted. *, p < 0.05.

FIGURE 4.

Effect of extracellular Zn²⁺ on [Zn²⁺]_cyt and on the expression of Slc39a6, Slc39a7, Slc39a8, Mt-1, and Mt-2. A, dissociated mouse islets were infected with eCALWY-4-expressing adenovirus and subsequently incubated for 24 h in medium containing 16.7 mm glucose or 3 mm glucose with or without 50 μm ZnCl₂ before imaging on an Olympus IX microscope. [Zn²⁺]_cyt was calculated as described under “Results” and in the legend to Fig. 1. Bars represent mean ± S.E. The number of cells imaged (n) on three different days of experiments is given in brackets under each bar. **, p < 0.01; ***, p < 0.001. B, mouse islets were incubated for 1 h at 3 mm glucose before culture at either 3, 10, or 16.7 mm glucose for 24 h in the presence (black bars) or absence (white bars) of 50 μm ZnCl₂. qPCR analysis of Slc39a6, Slc39a7, Slc39a8, Mt-1, and Mt-2 was performed. Values of expression were normalized to that of a housekeeping gene (cyclophillin A). The fold changes compared with the values obtained at 3 mm glucose are plotted. Bars represent mean ± S.E. (n = 4). *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 5.

Effect of an L-type voltage-gated Ca²⁺ channel blocker on glucose-induced [Zn²⁺]_cyt increase. Dissociated mouse islets were infected with eCALWY-4 expressing adenovirus before incubation with 3 or 16.7 mm glucose with or without 20 μm verapamil for 24 h before imaging on an Olympus IX microscope. [Zn²⁺]_cyt was then calculated as explained under “Results” and in Fig. 1. Bars represent mean ± S.E. The number of cells imaged (n) on three different days of experiments is given in brackets under each bar. **, p < 0.01; ***, p < 0.001.

FIGURE 6.

Effect of elevated intracellular cAMP levels on Slc39a6, Slc39a7, Slc39a8, Mt-1, and Mt-2 expression. Mouse islets were incubated for 1 h at 3 mm glucose before incubation at either 3,10, or 16.7 mm glucose for 24 h in the presence (black bars) or absence (white bars) of 50 μm 3-isobutyl-1-methylxanthine and 10 μm forskolin. qPCR analysis of Slc39a6, Slc39a7, Slc39a, Mt-1, and Mt-2 was performed. Values of expression were normalized to that of cyclophillin A. The fold changes compared with the values at 3 mm glucose were plotted. Bars represent mean ± S.E. (n = 4). **, p < 0.01; ***, p < 0.001.

FIGURE 7.

Effect of the K_ATP channel opener diazoxide on gene expression. CD1 mouse islets were incubated for 1 h at 3 mm glucose before a further 24 h in the presence of either 3 mm (white bars) or 16.7 mm (black bars) glucose in the presence or absence of 325 μm diazoxide. RNA was extracted, and Slc39a6, Slc39a7, Slc39a8, Mt-1, and Mt-2 gene expression was analyzed by qPCR. Values were normalized to the one of cyclophillin A, and the fold changes compared with control experiments at 3 mm glucose were plotted. Bars represent mean ± S.E. (n = 3). *, p < 0.05.

FIGURE 8.

Effect of the K_ATP channel opener diazoxide on [Zn²⁺]_cyt and the expression of the marker of cellular stress, TxNIP. A, mouse islets were incubated for 24 h in the presence of either 3 or 16.7 mm glucose supplemented (black bars) or not (white bars) with 325 μm diazoxide before RNA extraction and analysis of TxNIP gene expression by quantitative real-time PCR. Values were normalized to the one of cyclophillin A, and the fold changes compared with control experiments at 3 mm glucose were plotted. Bars represent mean ± S.E. (n = 3). *, p < 0.05. B, dissociated islets were infected with eCALWY-4-expressing adenovirus for 24 h and subsequently incubated for a further 24 h supplemented with 3 or 16.7 mm glucose in the presence or absence of 325 μm diazoxide before imaging on an Olympus IX microscope. [Zn²⁺]_cyt was then calculated as explained under “Results” and in the legend to Fig. 1. The number of cells imaged (n) on three different days of experiments is given in brackets under each bar. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 9.

Proposed model of the effect of glucose on the regulation of Zn²⁺ homeostasis in pancreatic β-cells. Shown are glucose-induced increases in the levels of ZiP6 and ZiP7, likely to be both at the plasma membrane and intracellular membranes. ZiP8 levels may conceivably be increased both on lysosomes and plasma membrane. Metallothionein protein levels were unchanged at 24 h but seem likely to decrease at later time points following substantial decrease in Mt-1 and Mt-2 mRNA. Please see the Discussion for other details. The image was generated using Servier Medical Art.