Crystal structures of the gastric proton pump

Abstract

The gastric proton pump-the H, K-ATPase-is a P-type ATPase responsible for acidifying the gastric juice down to pH 1. This corresponds to a million-fold proton gradient across the membrane of the parietal cell, the steepest known cation gradient of any mammalian tissue. The H, K-ATPase is an important target for drugs that treat gastric acid-related diseases. Here we present crystal structures of the H, K-ATPase in complex with two blockers, vonoprazan and SCH28080, in the luminal-open state, at 2.8 Å resolution. The drugs have partially overlapping but clearly distinct binding modes in the middle of a conduit running from the gastric lumen to the cation-binding site. The crystal structures suggest that the tight configuration at the cation-binding site lowers the pK value of Glu820 sufficiently to enable the release of a proton even into the pH 1 environment of the stomach.

A single K+-binding site in the crystal structure of the gastric proton pump

“…To our surprise, however, the Tyr799Trp mutant shows the highest ATPase activity in the absence of K + and the activity decreases with increasing K + concentration, in 20 marked contrast to the K + -dependent increase in ATPase activity of the wild-type enzyme ( Figure S1). The observed K + -independent ATPase activity of Tyr799Trp in the absence of K + indicates that the luminal gate of this mutant is spontaneously closed, like previously reported constitutive-active mutants (12), i.e., the Tyr799Trp mutation exerts a molecular signal that induces E2P dephosphorylation when K + is bound to the cation binding site. Decreasing ATPase 25 activity of Tyr799Trp with increasing K + concentration (K0.5,cyto = ~5 mM) can be interpreted as being due to K + -occlusion from the cytoplasmic side, which is also observed in the wild-type enzyme albeit with much lower affinity (K0.5,cyto = ~200 mM, 13).…”

“…The charge-conserved mutant of Glu343Asp shows no detectable ATPase activity due to its inability to induce K + -dependent dephosphorylation (26), indicating that having a negative charge at this position is not essential for the K + -coordination. Glu343Gln reduces K + -affinity (12), but to a far lesser extent than Glu343Asp, suggesting that Glu343 is 30 likely protonated in the crystal structure. The short distance between the Glu343 carboxyl and Val341 carbonyl (2.6 Å) suggest protonation of Glu343.…”

“…Wild-type-like K + -dependence can be restored by additional mutagenesis based on the observed structure of YW(K + )E2-P. We therefore conclude that the luminal-closed molecular conformation spontaneously induced by the Try799Trp mutation is not an artifact, and the driving force for the gate closure is essentially the same as in the wild-type enzyme. 40 Comparison of the luminal-open E2P ground state [(von)E2BeF structure with bound vonoprazan, a specific inhibitor for H + ,K + -ATPase (12)] and K + -occluded and luminal-closed E2-P transition state [YW(K + )E2-MgF4 2structure] reveals several key conformational rearrangements upon luminal gate closure ( Figure S4). Gate closure brings the luminal portion of TM4 (TM4L) close to TM6, effectively capping the cation-binding site from the luminal side of 45 the membrane.…”

“…The short distance between the Glu343 carboxyl and Val341 carbonyl (2.6 Å) suggest protonation of Glu343. The juxtaposition of Glu795 and Glu820 indicates that at least one of the carboxyls must be protonated (12,27). Based on the almost identical K + -affinity of WT and the charge-neutralized mutant Glu795Gln, Glu795 is the one expected to be protonated (12).…”

Cryo-EM structures capturing the entire transport cycle of the P4-ATPase flippase

“…6). While ATP8A1 shares the similar catalytic reaction of 5 ATP hydrolysis with the ion-transporting P2-type ATPases, such as SERCA (28, 29), Na + /K + -ATPase (37), and H + /K + -ATPase (38), there are significant differences in their transport mechanisms. S3D, E and S13B).…”