Figure 1.
Wild type and Cx36-antisense MIN6 cells differ by the expression of Cx36 and cell-to-cell coupling.
(A) Immunoblots of total cell extracts showed that AS Cx36 (gray bar) in comparison to WT MIN6 cells (light gray bar) featured a 80% reduction in Cx36 levels, relative to those of β smooth muscle actin. Data are means + SE of three independent experiments. **p<0.01; (B) Immunostaining showed that the expression Cx36, but not of GLUT-2, was lower in AS Cx36 than in WT MIN6 cells. Bar, 10 µm; (C) Microinjection of either EB or LY showed that AS MIN6 (gray bar) were significantly less coupled than WT cells (light gray bar). Values are means + SE of four independent experiments. ***p<0.001. The global evaluation of coupling, provided by computing a total coupling index, showed that the coupling of AS MIN6 was ∼3% of that of WT cells.
Table 1.
Coupling of MIN6 cells, as investigated by tracer microinjection.
Figure 2.
Glibenclamide and quinine differentially modulate the intercellular synchronization of glucose-induced Ca2+ oscillations.
(A) During stimulation by 20 mM glucose in the presence of 15 mM TEA, MIN6 cells exposed for 24 h to 10 µM glibenclamide showed increased proportion of synchronous cells (gray bars) compared to control, but decreased proportions of asynchronous (white bars) and silent cells (black bars); (B) Under the same conditions, MIN6 cells exposed for 24 h to 10 µM quinine, showed a reduced proportion of synchronous cells (gray bars). Bars show means + SE values of three independent experiments. *p<0.05 versus values of WT MIN6 cells; (C) The synchrony index of WT MIN6 cells, as evaluated using a NIPKOW confocal microscope, was about 0.6. This value was decreased in AS MIN6 cells, and increased after exposure of the cells to glibenclamide; (D) Similar values of synchrony index data were semi-automatically calculated, in the same cell types, using an ImageXpress microscope. Data are means + SE of seven clusters measured for each condition in C, and 87 clusters measured for each condition in D. *p<0.05, **p<0.01 and ***p<0.001 versus values of WT MIN6 cells.
Figure 3.
Primary and secondary screenings identify several molecules affecting the synchrony index of MIN6 cells.
(A) The primary screening of 1040 compounds revealed that several molecules increased or decreased the synchrony index. Each point of the curve shows the percentage change in synchrony index induced by one drug, relative to the value evaluated in WT MIN6 cells exposed to DMSO, which was the vehicle of all compounds, and served as internal control. Of note, glibenclamide increased the synchrony index by ∼12%, whereas quinine, and its stereoisomer quinidine, decreased it by 22 and 34%, respectively; (B) 17 drugs increasing (white bars) and 13 drugs decreasing the synchrony index (black bars) were repeatedly tested, and their effect on the synchrony index plotted relative to that of WT MIN6 cells exposed to DMSO. 97% of the drugs reproduced the effects observed in the primary screening. The six drugs pointed by the arrows were selected for further experiments. Data are means ± SE of four independent experiments; (C) Dose-response experiments were conducted with the six selected drugs. The effect of these drugs on the synchrony index was plotted relative to that of WT MIN6 cells exposed to DMSO. The effects on the synchrony index varied with the concentration tested. Glibenclamide, norcantharidin and zaprinast increased the synchrony index for concentrations ≤10 µM, whereas quinine, mebeverine and gedunin decreased it. Several drugs showed cytotoxicity at concentrations >10 µM. All drugs were tested at the following concentrations: 0.1 µM, 1 µM, 10 µM, 50 µM and 100 µM. “†” indicates concentrations toxic for MIN6 cells, as evaluated by cell detachment from the support. Data are means ± SE of at least four independent experiments.
Figure 4.
Molecules altering the synchrony index, modify Cx36-dependent coupling.
(A–B) Microinjection of either EB (A) or LY (B) revealed a similar extent of coupling in MIN6 cells exposed to DMSO (gray bars), the solvent used to dissolve all drugs, and untreated controls (hatched bars). All the drugs increasing the synchrony index (white bars) increased the extent of coupling, whether the permeability of Cx36 channels was tested with EB (A) of LY (B). In contrast, all the drugs decreasing the synchrony index (black bars) also decreased the extent of coupling tested with either EB (A) or LY (B). Data are means + SE of the number of injections indicated in Table 1. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0007 versus values obtained in cells exposed to DMSO; (C) Computing of total coupling index, assuming independent permeability of Cx36 channels to EB and LY, showed that coupling was significantly increased by all the drugs enhancing the synchrony index (white bars) and reduced by all the drugs decreasing this index (black bars). *p<0.05, **p<0.01, ****p<0.0007 versus values obtained in cells exposed to DMSO, using a chi-square test on the number of injections indicated in Table 1.
Figure 5.
Molecules altering the synchrony index, modify Cx36 expression and the size of Cx36 plaques.
(A) Immunolabeling of total protein extracts revealed that the steady state levels of Cx36 were increased in the MIN6 cells exposed to compounds enhancing the synchrony index, but unchanged in the MIN6 cells exposed to compounds reducing this index. 20 µg proteins were loaded per lane. The bars show means + SE ratio of the Cx36 to the β-actin signal, as evaluated by densitometric analysis of 5 independent experiments. *p<0.05 versus DMSO-exposed controls; (B) The volume density (Vv) of Cx36 was increased by some of the drugs that enhanced MIN6 cell synchronization, and decreased by all the drugs decreasing this synchronization. Data show means + SE of three independent experiments. *p<0.05; **p<0.01; ***p<0.001 versus DMSO-exposed controls; (C) The Vv changes were mostly accounted for by differential alterations in the length of Cx36 plaques, defined as the span of the longest chord through the object outline. Data show means + SE of three independent experiments. *p<0.05; **p<0.01; ***p<0.001 versus DMSO-exposed controls; (D) The length of “small” Cx36 plaques was reduced in cells exposed to drugs decreasing the synchrony index (left panel), whereas that of “large” plaques increased in cells exposed to drugs enhancing this index (right panel). Data show means + SE of three independent experiments. *p<0.05 and ***p<0.001 versus DMSO-exposed controls. For all parameters, no difference was observed between untreated and DMSO-treated cells.
Figure 6.
Drugs altering the synchrony index modulate the function of MIN6 cells.
(A) Insulin release was tested during a 30 min exposure to either 2.8 mM (left panel) or 16.8 mM glucose (right panel) in cells that had been exposed for 24 h to 10 µM the drugs selected in the secondary screening. No drug significantly altered the basal insulin release (left panel). All drugs, but glibenclamide and quinine, also increased to control levels their fractional insulin release (expressed as percentage of insulin content), when challenged with a high glucose level (right panel). (B) The insulin content of MIN6 cells was reduced by all drugs decreasing the synchrony index and by glibenclamide. Data are means + SE of three independent experiments. ***P<0.001 versus DMSO-exposed controls.