Cystic Fibrosis Transmembrane Conductance Regulator and Adenosine Triphosphate

Open Access
Publisher Website
Google Scholar
Abstract
The observations by Abraham et al . ([1][1]), that P-glycoprotein and the cystic fibrosis transmembrane conductance regulator (CFTR) are associated with adenosine triphosphate (ATP) movement across the plasma membrane, have been supported by other studies using patch clamp and bulk ATP measurements of systems expressing CFTR ([2][2], [3][3]). On the other hand, Reddy et al . ([4][4]) and Li et al . ([5][5]) did not detect ATP channel activity with electrophysiological methods or with radioactive ATP in reconstituted vesicles that contain CFTR. They state that CFTR does not conduct ATP, but that it might act as a regulator of an associated channel or transport system for ATP, a possibility also raised by Al-Awqati ([6][6]) and Higgins ([7][7]). Although there is conjecture that CFTR can form a complex with channels for anions ([3][3]) and cations ([8][8]), it is not excluded that CFTR and other ABC proteins are themselves directly involved in ATP movement. The resolution of this issue is hampered by disagreement about the mechanism of the ATP movement associated with ABC transporters: whether it is electrodiffusional through a channel or by facilitated transport. A further complication is that the form of ATP that moves across the membrane may vary, depending on the experimental conditions, between the fully charged ATP−4 anion and the uncharged ATP−3(Mg+2·L+) molecule, where L+ is a cotransported solute. Both ATP and adenosine diphosphate (ADP) are capable of forming complexes with P-glycoprotein substrates in cell-free electrochemical models (Fig. [1][9]), where positively charged doxorubicin moves toward the positive electrode in the presence of ATP and ADP. In related electrophoresis experiments, camptothecin (which is not a substrate for P-glycoprotein) is unaffected by ATP, while positively charged but structurally similar topotecan (a P-glycoprotein substrate) is cotransported with ATP in the electric field ([9][10]). Cotransport of ATP with cationic molecules can result in nearly electroneutral ATP efflux across membranes. Net electroneutrality would also occur if CFTR functions as an anion exchanger (for example, chloride-ATP). Under these circumstances, the ability to detect ATP currents in patch clamp studies would be more difficult than measurement of bulk ATP movement. ![Fig. 1.][11] Fig. 1. Comparison of the effect of various adenylates on the movement of a homogeneously doxorubicin (200 μg/ml) impregnated polyacrylamide gel. The gel was exposed to 100 volts for 30 min. Well 1 was loaded with 30 μl of tris-glycine buffer (control). Loading was 30 μl of 10 mM ATP in well 2, 30 μl of 10 mM ADP in well 3, 30 μl of 10 mM adenosine monophosphate (AMP) in well 4, 30 μl of 10 mM adenosine in well 5, and 30 μl of 10 mM phosphate in well 6. All the doxorubicin traveled upward toward the cathode in lane 1. Flow of doxorubicin toward the anode was seen in the ATP loaded lane and to a lesser extent in the ADP loaded lane. Adenosine and phosphate had doxorubicin mobilities similar to control. Experiments were performed at pH = 7, 20°C. The experiments demonstrate the potential of ATP-dependent cotransport to explain the removal of cationic and zwitterionic drugs mediated by P-glycoprotein and MRP. This mechanism, based on electrostatic coupling, requires that both ATP and drug are transported through P-glycoprotein. The structural similarities of P-glycoprotein and CFTR suggest that both proteins are involved in ATP transport in the same fashion. This view is supported by the observation (Fig. [2][12]) that cells overexpressing CFTR or P-glycoprotein are associated with increased release of ATP. Furthermore, exposure of cells overexpressing CFTR, sulfonylurea receptor (SUR), or multidrug resistance-associated protein (MRP) to antisense oligonucleotides specific for the respective ABC protein ([10][13]) results in reduced bulk ATP exit as measured by bioluminescence (Fig. [3][14]). These observations indicate that several ABC proteins, all of which share similar structures ([7][7]), are associated with ATP transport. ![Fig. 2.][11] Fig. 2. Comparison between steady-state light output generated by luciferin/luciferase-treated T84 cells, colon carcinoma cells overexpressing CFTR, compared with SW620, colon carcinoma cells with low detectable expression of P-glycoprotein and no detectable CFTR. Light detection was performed using a high-speed CCD camera (Princeton Instruments LN/CCD-1024TK[B]) mounted on an Olympus inverted microscope. Cells were plated in 96 well Wallac Inc. (LB96PMP) plates. Each 6-mm diameter well was partitioned using stainless steel dividers. The images are 3 by 3 mm and 1000 cells were plated on each side of the divider in a volume of 150 μl. Fifty-microliter aliquots of luciferin/Luciferase Assay Mix (Sigma FL-AAM, Lot 45H8000) were added to both sides of the well. The CFTR overexpressing T84 cells have greater light output during a 10-min exposure. Light output is a power function of ATP concentration and the assay is specific for ATP. Other nucleotide phosphates do not produce light. Experiments were also performed as with SW620/AD300 cells, overexpressing P-glycoprotein, and control SW620 cells (data not shown); 104 cells of each cell line were plated in 150 μl on opposite sides of the divider. The P-glycoprotein overexpressing SW620/AD300 cells, like the CFTR overexpressing cells, have greater light output compared to the control population. ![Fig. 3.][11]![Fig. 3.][11] Fig. 3. ( A ) Steady-state extracellular ATP and extracellular ATP accumulation after ecto-ATPase inhibition [by added 0.5 mM cytidine triphosphate (CTP); Pharmacia Biotech] were measured from cells with tandem 18-mer sense and anti-sense oligonucleotides. The tandem oligonucleotides were directed at the first 36 residues starting at the ATG site of the relevant ABC gene ([14][15]), and were replaced at 20 μM every 12 hours for 36 hours. T84...