Fig. 7. Control and Cx-transduced pseudo-islets show similar insulin content but different secretion patterns. (A) Total insulin content was similar in control pseudo-islets (solid columns; uninfected and GFP-transduced pseudo-islets were pooled) and in pseudo-islets infected with lentiviral vectors coding for Cx32 (horizontal hatched columns), Cx36 (diagonal hatched columns) or Cx43 (vertical hatched columns). Data are means plus s.e.m. of four experiments. (B) Pseudo-islets expressing Cx43 (vertical hatched columns) increased their insulin secretion in response to glucose similarly to the control pseudo-islets (solid columns)–i.e. they increased their insulin secretion when the glucose concentration was raised from the basal value of 5.6 mM to the stimulatory concentration of 16.8 mM. By contrast, pseudo-islets transduced for either Cx36 (diagonal hatched columns) or Cx32 (horizontal hatched columns) did not significantly raise their insulin release in response to the glucose increase. All connexin-transduced pseudo-islets increased their insulin secretion when exposed to increased intracellular cAMP concentrations. Data are means plus s.e.m. of three experiments, each testing in parallel all the four different groups of pseudo-islets. (C) When the secretion data were expressed as a percentage of the normalized control value (100%), it was apparent that expression of Cx43 did not alter insulin secretion in response to glucose. By contrast, glucose-induced insulin secretion was significantly reduced (P<0.005) by over-expression of either Cx32 or Cx36. The secretion induced by glucose plus IBMX and FSK was lowered in all Cx-transduced pseudo-islets. Data are means plus s.e.m. of three experiments.