The amiloride-sensitive epithelial Na⁺ channel, ENaC, composed of α, β and γ subunits controls the fluid lining the airways. Overexpression of ENaC is linked to the pathogenesis associated with cystic fibrosis (Mall et al., 2004). We have shown that the AMP mimetic, AICAR, activates the metabolic sensor AMP-activated protein kinase (AMPK) and inhibits amiloride-sensitive Na⁺ transport in H441 human lung epithelial cells (Woollhead et al., 2005). Its mechanism of action is unknown. As PI(4,5)P₂ is reported to be required for ENaC activation (Kunzelmann et al., 2005), we have investigated the hypothesis that AMPK converges on PI(4,5)P₂ and alters ENaC-PI(4,5)P₂ interactions. Monolayers were mounted in Ussing chambers in physiological salt solution. AICAR (2 mM) rapidly reduced amiloride-sensitive short circuit currents (I_SC-Amil) by 49.0 ± 7.6 % (P < 0.05, n = 8). Compound C, an inhibitor of AMPK, rescued I_SC-Amil. Purinergic stimulation by UTP (100 μM) inhibited I_SC-Amil by 49.2 ± 7.5 % (P < 0.05, n = 8) via depletion of membrane PI(4,5)P₂. Neomycin (5 mM), which sequesters PI(4,5)P₂, inhibited I_SC-Amil by 59.7 ± 12.6 % (P < 0.05, n = 8) and there was no further inhibition with UTP. Western blotting showed that there was no change in apical abundance of the α, β or γ ENaC subunits. U-73122 (10 μM), an inhibitor of phospholipase C, elevated I_SC-Amil 1.3 fold (P = 0.13, n = 4) and AICAR inhibited the elevated I_SC-Amil by 47.7 ± 5.6 % (P < 0.05, n = 4). Treatment with either UTP or neomycin followed by AICAR further reduced I_SC-Amil to 72.9 ± 4.0 % and 77.1 ± 7.1 % (both P < 0.05, n = 4), respectively. In these situations, the abundance of αENaC was diminished by 75 ± 7.6 % and that of βENaC by 61 ± 7.6 % (both P < 0.01, n = 4). There was no change to the apical abundance of the γ ENaC subunit. PI(4,5)P₂ co-immunoprecipitated α, β and γ ENaC subunits. However, following treatment of monolayers with AICAR, PI(4,5)P₂ only co-immunoprecipitated αENaC and αENaC was barely detectable following treatment with UTP. The data indicate that activation of AMPK by AICAR or compromising the ENaC-PI(4,5)P₂ interaction inhibit the open probability of ENaC. Together, AICAR and the absence of the ENaC-PI(4,5)P₂ interaction inhibit amiloride-sensitive Na⁺ transport which is paralleled by α and β ENaC subunit retrieval. PI(4,5)P₂ interacts with all three ENaC subunits and we hypothesise that this interaction is important for subunit surface expression. We propose that channel subunit retrieval by AMPK is dependent on the PI(4,5)P₂ concentration being low, as would be the case in physiological situations in which PLC signaling pathways are operative. It would appear that using PI(4,5)P₂ to regulate activity as well as trafficking means that ENaC would only be active at the membrane and not during insertion into or retrieval from the cell surface.