Bile acids directly bind and allosterically regulate the epithelial sodium channel

Abstract

Bile acids are abundant in the biliary tree and intestinal tract, and can be elevated in the urine of patients with advanced liver disease. Our previous work showed that bile acids carrying specific moieties regulate the epithelia Na+ channel (ENaC). ENaC mediates Na+ transport in several epithelia, including the biliary epithelium and aldosterone-sensitive distal nephron and colon. Based on our work showing no dependence on bile acid physicochemical properties or membrane permeability, we hypothesize that bile acids regulate ENaC through direct binding. To test this, we performed crosslinking experiments using bile acid derivatives that facilitate uv-crosslinking and affinity purification. Using ENaC heterologously expressed in Fisher Rat Thyroid cells, we found that the deoxycholic acid derivative (p-DCA) crosslinked to the α, β, and γ subunits. To determine whether crosslinking was specific, we performed experiments in the presence of different competing bile acids. We found that crosslinking to the β subunit was significantly reduced in the presence of deoxycholic acid (DCA). We also found that crosslinking to the γ subunit was significantly reduced by taurocholic acid (t-CA). These data suggest that bile acids directly bind ENaC, with different bile acids preferentially binding different sites. We also examined the effect of subunit composition on bile acid regulation. Using Xenopus oocytes, we measured the effect of t-CA on Li+ currents conducted by different functional ENaC subunit combinations. We found that α-only channels were greatly activated by t-CA, followed by αβγ- and αβ- channels, suggesting that the α subunit is sufficient for ENaC sensitivity to t-CA. To determine whether bile acid regulation is voltage-sensitive, we measured the effect of different doses of t-CA on ENaC activation while varying voltage from -140 mV to 40 mV. We found that t-CA activated ENaC in a voltage-dependent manner, with greater activation at less negative potentials. Models of regulation having a single t-CA binding site fit the data poorly. Hypothesizing multiple binding sites, our data fit well to an allosteric model with at least two binding sites, one of which is voltage-sensitive. Our data are consistent with bile acids directly binding all three ENaC subunits at multiple sites, with one site associated with the transmembrane domains. Bile acid moieties provide for selective preferences for each of the sites, resulting in distinct effects for each bile acid.