Get started
Work

Degenerate but indispensable: How CFTR channel activity depends on the catalytically inactive ATP binding site

Year: 2021
Type: article
Abstract: The cystic fibrosis transmembrane conductance regulator (CFTR) is an ATP binding cassette (ABC) protein which contains two transmembrane domains (TMDs) and two cytosolic nucleotide binding domains (NBDs). Whereas most ABC proteins function as active transporters, CFTR is as an anion channel whose dysfunction causes cystic fibrosis (CF). The two ABC-typical TMD-NBD units of CFTR are linked by a cytosolic regulatory (R) domain which prevents gating unless it is phosphorylated by cAMP-dependent protein kinase (PKA). Gating of phosphorylated CFTR channels is based on conformational changes conserved among ABC proteins (reviewed in (Csanády et al. 2019)). In resting ABC exporters the TMDs adopt inward-facing (IF) conformations which allow cytosolic substrate binding, while the translocation pathway is sealed at the extracellular end. ATP binding to the NBDs prompts their association into a tight head-to-tail dimer which occludes two ATP molecules in composite interfacial sites (sites 1, 2). Concomitantly, the TMDs flip into an outward-facing (OF) conformation which seals the cytosolic end of the substrate translocation pathway, but opens up its extracellular end to allow substrate release. ATP hydrolysis disrupts the NBD dimer and resets the TMDs to the IF state. Asymmetric ABC proteins, like CFTR, contain a catalytically active low-affinity site (site 2) and a catalytically inactive high-affinity site (site 1). Interestingly, among the many cryo-EM structures of ABC proteins solved in the IF state, a very large range of NBD1-NBD2 distances (∼9–50 Å) can be observed. Recent cryo-EM structures of dephosphorylated apo-CFTR in the IF state (Liu et al. 2017) and phosphorylated ATP-bound CFTR in the OF state (Zhang et al. 2018) have shown how CFTR has evolved into a channel: in the OF conformation a lateral portal opens up to bypass the closed inner gate, allowing uninterrupted transmembrane flow of anions. Thus, for CFTR the OF state corresponds to an open, the IF state to a closed pore. Furthermore, the R domain is wedged in between the NBDs when dephosphorylated, but is released upon phosphorylation. Despite these recent advances, questions remain. For instance, we can depict a dephosphorylated closed channel devoid of ATP. But how should we envisage the closed conformation that a phosphorylated channel, gating in the presence of ATP, adopts in each gating cycle? To what extent does it resemble the conformation captured in the IF apo-structure? In that structure the NBDs are completely separated and far apart, whereas functional evidence suggests that site 1 retains ATP stably bound for several gating cycles, and does not completely separate during a regular channel closed event (Tsai et al. 2010). A study published in a recent issue of The Journal of Physiology (Yeh et al. 2021), addresses this question using a functional approach, by focusing on the phenomenon of CFTR current rundown. Following phosphorylation, macroscopic CFTR channel current remains steady in the presence of ATP for tens of minutes. Moreover, when ATP is removed for a few seconds, the current vanishes but recovers promptly and fully upon ATP readmission. In contrast, when ATP is reapplied following longer (30–300 s) exposures to an ATP-free solution, the current only partially recovers and follows a bi-exponential time course. Such behaviour reports three populations of closed channels: promptly responsive channels, slowly responsive channels and channels that can no longer be opened by ATP. The latter two populations define reversible and irreversible ‘rundown’ states. The authors show that this rundown phenomenon is unrelated to dephosphorylation, as it cannot be reversed by exposure to PKA, and is also observable for channels lacking the R domain which do not require phosphorylation to open. Because upon ATP removal the time course of accumulation in the rundown states (tens of seconds) coincides with that of ATP loss from site 1 (Tsai et al. 2010), the authors hypothesize that reversible rundown reflects slow separation of site 1 followed by unbinding of ATP. Once decoupled from each other, the two NBDs progressively separate, and beyond a certain distance the process becomes irreversible causing irreversible rundown. This hypothesis generates several predictions, three of which are successfully tested in the current study. First, ATP occupancy of site 1 should prevent rundown. Exploiting the differential affinities of the two sites for ATP, the authors show that ATP effectively prevents rundown already at a concentration (10 μM) at which the occupancy of site 2, and therefore channel activity, is minimal. Second, reducing nucleotide binding affinity at site 1 should accelerate the rundown, whereas increasing the affinity should slow it. Indeed, mutations in site 1 that impair nucleotide binding (K464A, W401G) are shown to accelerate the rundown process. Conversely, N6-(2-phenylethyl)-ATP (P-ATP), a high-affinity ATP analogue, effectively slows rundown for these mutants. Moreover, the authors also demonstrate that P-ATP can delay the very fast rundown caused by the most common CF mutation (F508del), suggesting site-1 stabilization as a viable strategy for designing further clinically useful CFTR potentiator drugs. Third, when ATP binding at site 1 is weakened, occasional visits to the rundown state are expected to happen even in the continued presence of ATP. Indeed, in steady-state single-channel recordings the appearance of a distinct population of such longer closed events is demonstrated for the K464A and W401G mutants. The identification of site-1 separation with channel rundown provides incremental insight that links CFTR structures to function. As usual, the work also unearths a novel enigma that remains to be understood: how a mutation that prevents ATP hydrolysis in site 2 can almost entirely eliminate irreversible rundown. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. The author declares no conflicting interests. Sole author. EU Horizon 2020 Research and Innovation Program grant 739593, MTA Lendület grant LP2017-14/2017, and Cystic Fibrosis Foundation Research Grant CSANAD21G0.
DOI: 10.1113/jp282278
Language: English
Field: Medicine
Open Access status: hybrid
Estimated APC paid: $4,700
Award: G76517417
OpenAlex ID: W3196575185
FWCI:
Cites: 5
Cited by:
Related to: 10