Towards an understanding of CFTR regulation by ATP hydrolysis

The cystic fibrosis transmembrane conductance regulator (CFTR) belongs to the protein family of ATP binding cassette (ABC) transporters. However, it is not known as a transporter but functions as an anion channel. Channel opening depends on phosphorylation of a regulatory domain and is stimulated in the presence of ATP which is bound and hydrolyzed at the nucleotide binding domains. To address the poorly understood link between conductance and ATP hydrolysis, our aim was to measure CFTR-ATPase in living cells using microphysiometry in order to monitor CFTR function in a native environment. We investigated how phosphorylation via protein kinase A (PKA) is reflected in this assay. We detected an increased ATP consumption upon PKA stimulation that depends on the presence of functional CFTR and is sensitive to PKA inhibition. As our results could be described by a model for PKA activation, we concluded that activation of CFTR by PKA was monitored, although the signal is a mixture of ATP consumption by PKA and CFTR. We could also show that microphysiometry can be used to detect the presence of functional CFTR and could thus be used as a label-free assay for detection of functional CFTR.
The second part of this study focuses on inhibitors and potentiators of CFTR chloride conductance. These compounds supposedly bind to CFTR directly and some are known as substrates of other ABC transporters. We could show that channel inhibitors caused concentration dependent stimulation of ATP consumption only in presence of functional CFTR. In contrast to phosphorylation agents, an indirect mechanism via PKA could be excluded. The observed effects could be described by a two-site binding model developed earlier for the Pgp-ATPase activity. These findings suggested that we were indeed monitoring ATPase activity of CFTR. The results for potentiators were more ambiguous, as the readout was dominated by unspecific effects. However, we could conclude that ATPase activity was not stimulated. As our assumptions are based on similarities among ABC transporters, knowledge about ATPase modulation of other transporters (Pgp, BCRP) by the tested compounds was essential. In inside-out vesicles, all compounds caused bell-shaped ATPase stimulation in at least one of the tested transporters and supported earlier findings that amphiphilic neutral compounds stimulate Pgp-ATPase whereas non-amphiphilic compounds inhibit. Substrate binding in these transporters is a two-step process and compounds need to partition into the membrane where they bind to the transporter from the inner leaflet of the lipid bilayer. We estimated lipid-water partitioning from measured air-water partition coefficients in order to calculate actual transporter-lipid binding constants. Based on these parameters the observed effects on CFTR can be reasonably explained by the same mechanism for substrate-transporter interaction as in Pgp or BCRP and CFTR inhibitors are amphiphilic compounds while most potentiators are rather non-amphiphilic. We therefore proposed a model for ATPase activity modulation of CFTR and suggested that CFTR resembles Pgp as the ATPase is stimulated by amphiphilic compounds whereas activity is inhibited by non-amphiphilic ones. From a comparison with effects on channel conductance, we could furthermore conclude that stimulation of ATPase activity might be the common mechanism for inhibition by the structurally divergent inhibitors. In the last part of this study, we investigated the effect of steroid hormones known as Pgp substrates on CFTR. Both potentiation and inhibition of CFTR currents had been reported for estradiol and the estrogen receptor antagonist tamoxifen before, while no data was available for cortisol and progesterone. We could distinguish between CFTR ATPase stimulation, reversible unspecific effects and an additional irreversible cytotoxic effect only seen for tamoxifen. Cortisol and progesterone resembled CFTR inhibitors in terms of ATPase stimulation while estradiol caused reversible unspecific effects as seen before for potentiators. This supports the soundness of the assumptions made above, as again more pronounced stimulation of CFTR-ATPase was observed for more amphiphilic compounds. We propose that cortisol and progesterone could be inhibitors while estradiol potentiates at low and inhibits at high concentrations which is in line with aforementioned findings for estradiol.
In summary, we could establish label-free assays to monitor phosphorylation of CFTR and CFTR-ATPase activity in living cells. We could show that modulators most likely bind to CFTR via the same mechanisms known for substrate binding in other ABC transporters. Comparison of the results with effects on channel conductance finally allowed us to propose that potentiation and inhibition are based on drug induced ATPase activity changes according to the two-site binding model introduced earlier for Pgp.