CFTR - THE NUCLEOTIDE-BINDING FOLDS REGULATE THE ACCESSIBILITY AND STABILITY OF THE ACTIVATED STATE

The functional roles of the two nucleotide binding folds, NBF1 and NBF 2, in the activation of the cystic fibrosis transmembrane conductance regulator (CFTR) were investigated by measuring the rates of activatio n and deactivation of CFTR Cl- conductance in Xenopus oocytes. Activat ion of wild-type CFTR in response to application of forskolin and 3-is obutyl-1-methylxanthine (IBMX) was described by a single exponential. Deactivation after washout of the cocktail consisted of two phases: an initial slow phase, described by a latency, and an exponential declin e, Rate analysis of CFTR variants bearing analogous mutations in NBF1 and NBF2 permitted us to characterize amino acid substitutions accordi ng to their effects on the accessibility and stability of the active s tate. Access to the active state was very sensitive to substitutions f or the invariant glycine (G551) in NBF1, where mutations to alanine (A ), serine (S), or aspartic acid (D) reduced the apparent on rate by mo re than tenfold. The analogous substitutions in NBF2 (G1349) also redu ced the on rate, by twofold to 10-fold, but substantially destabilized the active state as well, as judged by increased deactivation rates. In the putative ATP-binding pocket of either NBF, substitution of alan ine, glutamine (Q), or arginine (R) for the invariant lysine (K464 or K1250) reduced the on rate similarly, by two- to fourfold. III contras t, these analogous substitutions produced opposite effects on the deac tivation rate. NBF1 mutations destabilized the active state, whereas t he analogous substitutions in NBF2 stabilized the active state such th at activation was prolonged compared with that seen with wild-type CFT R. Substitution of asparagine (N) for a highly conserved aspartic acid (D572) in the ATP-binding pocket of NBF1 dramatically slowed the on r ate and destabilized the active state. In contrast, the analogous subs titution in NBF2 (D1370N) did not appreciably affect the on rate and m arkedly stabilized the active state. These results are consistent with a hypothesis for CFTR activation that invokes the binding and hydroly sis of ATP at NBF1 as a crucial step in activation, while at NBF2, ATP binding enhances access to the active state, but the rate of ATP hydr olysis controls the duration of the active state. The relatively slow time courses for activation and deactivation suggest that slow process es modulate ATP-dependent gating.