Role of the cAMP sensor Epac as a determinant of KATP channel ATP sensitivity in human pancreatic beta-cells and rat INS-1 cells

Abstract

Protein kinase A (PKA)-independent actions of adenosine 3',5'-cyclic monophosphate (cAMP) are mediated by Epac, a cAMP sensor expressed in pancreatic beta-cells. Evidence that Epac might mediate the cAMP-dependent inhibition of beta-cell ATP-sensitive K(+) channels (K(ATP)) was provided by one prior study of human beta-cells and a rat insulin-secreting cell line (INS-1 cells) in which it was demonstrated that an Epac-selective cAMP analogue (ESCA) inhibited a sulphonylurea-sensitive K(+) current measured under conditions of whole-cell recording. Using excised patches of plasma membrane derived from human beta-cells and rat INS-1 cells, we now report that 2'-O-Me-cAMP, an ESCA that activates Epac but not PKA, sensitizes single K(ATP) channels to the inhibitory effect of ATP, thereby reducing channel activity. In the presence of 2'-O-Me-cAMP (50 microM), the dose-response relationship describing ATP-dependent inhibition of K(ATP) channel activity (NP(o)) is left-shifted such that the concentration of ATP producing 50% inhibition (IC(50)) is reduced from 22 microM to 1 microM for human beta-cells, and from 14 microM to 4 microM for rat INS-1 cells. Conversely, when patches are exposed to a fixed concentration of ATP (10 microM), the administration of 2'-O-Me-cAMP inhibits channel activity in a dose-dependent and reversible manner (IC(50) 12 microM for both cell types). A cyclic nucleotide phosphodiesterase-resistant ESCA (Sp-8-pCPT-2'-O-Me-cAMPS) also inhibits K(ATP) channel activity, thereby demonstrating that the inhibitory actions of ESCAs reported here are unlikely to arise as a consequence of their hydrolysis to bioactive derivatives of adenosine. On the basis of such findings it is concluded that there exists in human beta-cells and rat INS-1 cells a novel form of ion channel modulation in which the ATP sensitivity of K(ATP) channels is regulated by Epac.

Figure 1. Inhibition of K_ATP channel activity by 2′-O-Me-cAMP in INS-1 cells

A, 2′-O-Me-cAMP was applied at a concentration of 50 μm to an inside-out patch. The inhibition of channel activity by 2′-O-Me-cAMP is illustrated on a compressed time scale (top trace) or on an expanded time scale (traces a, b, and c) in order to depict channel activity prior to (a), during (b), and following washout (c) of 2′-O-Me-cAMP. Unitary currents measured in this patch resulted from the activity of K_ATP channels because they were abolished during application of 1 mm ATP (top trace). Arrows at the top indicate when the patch was exposed to each test solution. Patch potential was −100 mV here and in subsequent figures. B, all-points amplitude histogram depicting K_ATP channel activity prior to (dotted line) or during (continuous line) application of 50 μm 2′-O-Me-cAMP to the same patch depicted in panel A. C, population study summarizing the action of 50 μm 2′-O-Me-cAMP to decrease values of NP_o in excised patches under conditions in which the ATP concentration was fixed at 10 μm. Channel activity prior to and following washout of 2′-O-Me-cAMP is depicted by the left-most and right-most bars, respectively. Values of NP_o for each bar are the mean ±s.e.m. for 16 patches and were not subjected to normalization. *The value of NP_o was different (P < 0.05 here and in subsequent figures).

Figure 2. Dose–response relationship for INS-1 cell K_ATP channel inhibition by 2′-O-Me-cAMP

A, K_ATP channel activity was measured under conditions in which a progressively higher concentration of 2′-O-Me-cAMP was administered while maintaining the ATP concentration at 10 μm. Illustrated are continuous data subdivided into five traces. The top trace indicates control activity in the absence of 2′-O-Me-cAMP. The bottom trace indicates recovery of channel activity after wash out of 2′-O-Me-cAMP. B, cumulative dose–response relationship describing the inhibition of K_ATP channel activity, as generated using the experimental design illustrated in panel A. For each concentration of 2′-O-Me-cAMP the value of NP_o was normalized relative to a value of 1.0, which represents the relative current measured when patches were exposed to 10 μm ATP in the absence of 2′-O-Me-cAMP. Each data point (squares) is the mean ±s.e.m. for 5 patches. Dashed lines indicate the IC₅₀ concentration of 2′-O-Me-cAMP.

Figure 3. INS-1 cell K_ATP channel activity is inhibited by cAMP but not 2′-O-Me-cGMP

A, cAMP (300 μm) or 2′-O-Me-cGMP (100 μm) was applied under conditions in which an excised patch was continuously exposed to 10 μm ATP. B, the same experiment illustrated in A but depicted on an expanded time scale. Traces a, b and c correspond to the time periods indicated by the horizontal bars in the trace illustrated in A. C, population study summarizing the action of cAMP (300 μm) and the lack of action of 2′-O-Me-cGMP (100 μm) to alter values of NP_o in excised patches exposed to 10 μm ATP. Values of NP_o for each bar are the mean ±s.e.m. for 6 patches and were not normalized.

Figure 4. Contrasting actions of Rp-cAMPS and 6-Bnz-cAMP in INS-1 cells

A, pretreatment of patches with 100 μm Rp-cAMPS failed to prevent the inhibition of K_ATP channel activity by 50 μm 2′-O-Me-cAMP. Arrows indicate the time intervals during which test substances were applied. B, treatment of patches with 100 μm 6-Bnz-cAMP stimulated K_ATP channel activity in a reversible manner. C, population study summarizing findings presented in A and B. The actions of ATP (10 μm), 6-Bnz-cAMP (100 μm), Rp-cAMPS (100 μm), and 2′-O-Me-cAMP (50 μm) to affect K_ATP channel activity in INS-1 cells are illustrated. Values of NP_o were normalized relative to the channel activity measured in an ATP-free solution which was assigned a relative current value of 1.0. Relative current values for each bar are the mean ±s.e.m. for the number of patches indicated.

Figure 5. Interaction of ATP and 2′-O-Me-cAMP to inhibit K_ATP channels in INS-1 cells

A, K_ATP channel activity was inhibited by 2′-O-Me-cAMP (50 μm) in the absence of ATP. Note that the inhibition of channel activity developed slowly. Channel activity recovered upon wash out of 2′-O-Me-cAMP. B, in the presence of ATP (30 μm), the inhibitory action of 2′-O-Me-cAMP (50 μm) was faster in onset and of greater magnitude. C, population study summarizing the interaction of ATP and 2′-O-Me-cAMP to decrease values of NP_o. For each patch, ATP was tested at a concentration of 0, 1 and 30 μm in the absence or presence of 50 μm 2′-O-Me-cAMP. Values of NP_o for each bar in C are the mean ±s.e.m. for 7 patches and were normalized relative to the channel activity measured in an ATP-free solution, which was assigned a value of 1.0.

Figure 6. 2′-O-Me-cAMP increases the ATP sensitivity of INS-1 cell K_ATP channels

A, experimental design for establishment of the dose–response relationship describing the interaction of ATP and 2′-O-Me-cAMP to inhibit K_ATP channels. The K_ATP channel activity is illustrated for a single excised patch under conditions in which no 2′-O-Me-cAMP was present (traces labelled as a, b and c) or during administration of 50 μm 2′-O-Me-cAMP (right series of traces labelled as d, e and f). Traces a and c as well as d and f illustrate channel activity in an ATP-free solution. Traces b and e illustrate channel activity during exposure of the patch to 10 μm ATP. Each test solution was administered in the order a through e and the duration of exposure to each test solution was 45 s. Note that channel activity was inhibited by ATP in a reversible manner, and that the inhibitory effect of ATP was stronger under conditions in which the patch was also exposed to 2′-O-Me-cAMP. Dashed lines indicate the pipette current corresponding to the closed state of the channels. B, dose–response relationship describing the action of ATP to inhibit K_ATP channel activity under conditions in which excised patches were not exposed to 2′-O-Me-cAMP (•) or when patches were exposed to 50 μm 2′-O-Me-cAMP (□). • indicate values of NP_o normalized relative to a value of 1.0, which represents the relative current measured in the presence of 0.3 μm ATP alone. □ indicate values of NP_o normalized relative to a value of 1.0, which represents the relative current measured in the combined presence of 0.3 μm ATP and 50 μm 2′-O-Me-cAMP. Each data point is the mean ±s.e.m. for 5 patches. Dashed lines indicate the method by which the IC₅₀ concentration of ATP was estimated. For patches exposed to 2′-O-Me-cAMP, it was confirmed that similar values of channel activity existed prior to and following washout of the ESCA.

Figure 7. Interaction of ATP and 2′-O-Me-cAMP to inhibit K_ATP channels in human pancreatic β-cells

K_ATP channel activity is illustrated for an excised patch under conditions in which no 2′-O-Me-cAMP was present (traces a, b and c) or during administration of 50 μm 2′-O-Me-cAMP (right series of traces d, e and f). Arrows indicate the current levels at which channels are closed (0) or when one or more channels are open (1–5). Dashed lines indicate the pipette current corresponding to the closed state of the channels. Traces a and c as well as d and f illustrate channel activity in an ATP-free solution. Traces b and e illustrate channel activity during exposure of the patch to 10 μm ATP. Each solution was administered in the order a through e and the duration of exposure to each solution was 45 s. Note that channel activity was inhibited by ATP in a reversible manner, and that the inhibitory effect of ATP was stronger under conditions in which the patch was exposed to 2′-O-Me-cAMP.

Figure 8. Inhibition of K_ATP channels by PDE-resistant Sp-8-pCPT-2′-O-Me-cAMPS

A, Sp-8-pCPT-2′-O-Me-cAMPS (100 μm) was applied under conditions in which a patch was continuously exposed to 10 μm ATP. B, the same experiment illustrated in A but illustrated on an expanded time scale. Traces a, b and c correspond to the time periods indicated by the horizontal bars in the trace illustrated in A. C, population study summarizing the action of Sp-8-pCPT-2′-O-Me-cAMPS (100 μm) to decrease values of NP_o in excised patches exposed to 10 μm ATP. Values of NP_o for each histogram bar are the mean ±s.e.m. for 8 patches and were not normalized.