CFTR REVIEW PAGE
* The gene that codes for CFTR was discovered by positional cloning in 1989.
* CFTR is an abbreviation for cystic fibrosis transmembrane conductance regulator, and implies that CFTR not only functions as an ion channel, but also as a regulator of other ion channels.
* CFTR was the first epithelial cell ion channel to have its primary sequence structure determined, as well as the first chloride-specific ion channel to be cloned.
* The gene coding for CFTR is highly conserved among species and has a wide tissue distribution. The gene sequence is over 300,000 base pairs in length and has 27 exons. It is found on chromosome 7, specifically at 7q31.
* The CFTR protein is 1480 amino acids long. The mRNA transcript is 6.5 KB in length. Fully glycosylated, CFTR has a molecular weight of approximately 180 KD (via SDS-PAGE it is ~135 KD). Unglycosylated, it weighs 168,173 KD.
* CFTR is N-glycosylated at two sites on extracellular loop 4, meaning that it is a "glycoprotein".
* CFTR is expressed in endothelial cells of the umbilical vein, lung microvasculature, red blood cells, pancreas, lung epithelia, sweat gland, colon, parotid gland, liver, proximal tubules (and cortex and medulla) of kidney, heart myocytes and the brain (hypothalamus). It has also been detected by PCR techniques in lymphoid cells.
* How a defect in CFTR probably causes CF: hydration (water content) of the mucus in the lung is acquired in the submucosal glands in the lower airway of the lungs. This is also the place where the large glycoprotein, mucin, is made and secreted into the mucus. Proper hydration of mucus is necessary for its timely removal by the tiny hairs (cilia) on the outside of the airway cells, otherwise, the mucus will become too sticky and as a result trap bacteria as well as the body's immune system cells (which can add to the inflammation) in the lungs. When CFTR is not working properly, chloride (and therefore water) will not flow from the bloodstream into the mucus at the submucosal glands of the lower lung. This means that in CF, the mucus deep down in the lungs is already starting out on its journey to the throat as being too sticky. In addition to this, paradoxically, in the upper airway, where the mucus normally has some of its water removed back to the bloodstream, CFTR is needed to keep the sodium ion channel, ENaC, from working. When there's no CFTR in the upper airway, as in CF, too much sodium travels back from the mucus into the cells of the upper airway, meaning that the water will also travel from the mucus back to the bloodstream. This makes the mucus lose even more of what little water it already has, and as a result becomes even more sticky and viscous than it otherwise would be.
* The family of genes CFTR is a member of is called the ABC Transporter Family (for ATP Binding Cassette Transporters; called this because they all bind and hydrolyze ATP. While most seem to function as transporters, some such as CFTR appear to function mainly as ion channels and/or regulators of other ion channels). There are over 30 known ABC transporters; some of which are found in prokaryotes as well as in eukaryotes. There is often in excess of 30% sequence similarity between the ABC family members in prokaryotes and eukaryotes. Humans have approximately 12 different ABC Transporter genes including the one for CFTR. A second ABC in humans is involved with the disease adrenoleukodystrophy. This family probably first evolved in bacteria, as E. coli alone has slightly over 25 different ABC transporters, and some appear to be involved in associating with certain binding proteins. The following is a general list of the various types of molecules and ions ABC proteins are known to transport: sugars, peptides, antibiotics, phosphate esters, inorganic phosphate and sulfate, chloride, metal cations, iron-chelator complexes, vitamins and polyamines. Some ABCs have even been found to export proteins, for example proteases. Another term which is often used for ABC Transporters is "Traffic ATPases".
* CFTR has 5 predicted domains: 2 transmembrane domains with 6 transmembrane helices each, 2 NBDs (nucleotide binding domains), and one R domain. It also appears to use its C-terminus and N-terminus for targeting and regulation, respectively. There is increasing evidence that CFTR is able to use its intracellular loops to help in gating, providing a mainly stimulatory effect in pore opening overall. Very little CFTR is apparently exposed on extracellular side of the cell membrane. It is predicted that 77% of the CFTR protein is present in the cytoplasm, 4% as extracellular loops, 19% as membrane-spanning domains. Extracellular loops are all very short in length, except perhaps extracellular loops 1 and 4.
* The gene coding for CFTR appears to have undergone a gene duplication event followed by gene fusion early in its history since the two halves of CFTR have strong sequence similarity to one another, especially at the NBD domains.
* CFTR has a chloride current of between 5 and 10 pS, and a linear current to voltage (I/V) relationship under symmetrical ionic conditions. It has an ion selectivity of Br > Cl > I > F however, when bathed in asymmetric chloride concentrations, the I-V relationship appears to recitfy. It has been hypothesized that certain intracellular anions are able modify the permeability characteristics of chloride. It is possible that non-permeable anions may cause some sort of a block. CFTR also probably conducts bicarbonate and ATP anions. CFTR may form a "dual pore", or heterodimerize with other channels. CFTR significantly prefers chloride to sodium (~10:1). Conditions like buffer type, temperature, and cells used all seem to have an effect on exact permeabilities of ions thru the CFTR pore. Anion permeability is determined by the hydration energy of anions, because polyatomic anion permeability follows the lyotropic sequence, however other evidence has suggested a more complicated method of ion permeation.
* CFTR from different species have different rates of permeability. (Xenopus>human>mouse>shark).
* CFTR probably uses a multi-ion pore (i.e. its pore lets more than one anion thru at a time). Evidence for this comes from patch clamp studies where it was possible to keep intracellular gluconate ions from blocking the pore when chloride ions are present on outside of the membrane. Chloride ions appear to be able to keep gluconate from binding. There are also anomalous mole fraction effects when CFTR is bathed in symmetrical mixtures of chloride and thiocyanate anions.
* CFTR appears to have different sub-conductance states, especially when CFTR is reconstituted. Differences in experimental protocol cannot be ruled out as the cause. It is possible that modifying proteins or divalent ions in the cell cause this to occur. CFTR is activated by phosphorylation at the R domain by protein kinase A (a cyclic AMP-activated kinase. Note: adenylyl cyclase, which produces cAMP is in turn activated by G-proteins responding to the adrenogenic receptor binding to a hormone.) Recent analysis of kinetics of CFTR gating suggest that there is one open and two closed states. The first closed state, C1, is longer-lived. ATP apparently decreases this time. Once open, the channel "flickers" on and off between the open and second closed state (C2). [ Open <----> C1 <-------> C2 ]
* CFTR is predicted to help regulate the following channels: ORCC,
ROMKK+, ENaC, and Cl-/HCO3- exchanger. Note: This fact
fascinates molecular biologists and physiologists, as it is highly unusual for a
single protein like CFTR to be so important to the function of so many other
* The isoelectric point (pI) of CFTR is 8.91 In other words, the net charge on a CFTR molecule is zero when the pH is 8.91
* There is a lack of specific inhibitors of CFTR, which makes study difficult. There are currently no inhibitors of CFTR which possess high affinity or specificity. The sulfonylureas glibenclamide and tolbutamide inhibit CFTR chloride currents with half-maximal concentrations of 20 and 150 uM. They probably function as open channel blockers. Glibenclamide inhibition is voltage-dependent and enhanced when external chloride is reduced. It is believed to block the CFTR pore within a large intracellular vestibule. Gluconate, MOPS and glutamate also block CFTR approximately 30 to 60% thru pore from intracellular side. Note that all are anions.
* Aside from CF, CFTR, when overactive, is known to be responsible for the deaths of several million infants per year in developing countries due to bacterially induced diarreha (e.g. cholera, others). Other diseases caused by malfunction of ABC transporters include adrenoleukodystrophy, Tangier disease, multi-drug resistance (many cancers, as well as some antibiotic resistant bacteria), and obstetric cholestasis.
* CFTR has been difficult to overexpress in mammalian systems (CHO and BHK cells), probably due to an unusually strict quality control process during biosynthesis. CFTR has been overexpressed in other eukaryotic systems, including the baculovirus system, as well as in yeast. Overexpression in E. coli resulted in misfolding. CFTR was expressed as a stably integrated cDNA in the Drosophila S2 cell line by our group (Teintze Group, Montana State University), but at 2 to 3 orders of magnitude lower amounts of CFTR yield than the baculovirus system provides (unpublished data). There are reports of CFTR being overexpressed in vaccinia systems, as well as in bovine mammary tissue of but data is currently unavailable. Xenopus oocytes are routinely used for studies using patch clamp. Incorporation of CFTR under control of a casein promoter in mammary gland of transgenic mice resulted in expression in milk fat globule membranes, but purification has proven difficult.
* CFTR is a protein that is important enough to be found in many species. Aside from humans, the CFTR gene has been sequenced from the following organisms: mice, cows, dogfish, frogs, rats, sheep, killfish, pufferfish, and salmon (the Atlantic salmon has two CFTR genes, which are presumed to be involved in allowing this fish to adapt to salt water conditions when needed).
* The total number of mutations in the CFTR gene worldwide is approaching 1000 (I've read numbers as high as 1300). (www.genet.sickkids.on.ca/cftr) It should be noted that, since many mutations in CFTR have been discovered by partial sequencing, they may in fact be neutral polymorphisms, and the actual CF-causing mutation remains undetected.
* There is as yet no x-ray crystal structure or NMR structure of CFTR. In September, 2001 a high-resolution structure of a distant bacterial relative of CFTR, MsbA, appeared in the journal Science (Chang and Roth, Vol 293, 9/7/01, 1793-1800). Previously, the only structure of an ABC protein was a low-resolution one of P-glycoprotein (Rosenberg et al., J. Biol. Chem. 272, 10685, 1997).
* CFTR is activated in a dose-dependent manner by the drugs genistein, apigenin, 8-cyclopentyl-1,3-dipropylxanthine, IBMX, 8-methoxypsoralen, and milrinone
More Facts: Regulation
Gating refers to the way in which ion channels open and close. It is perhaps the most important and yet least understood characteristic of CFTR. Mutagenesis data in the past has been ambiguous and difficult to interpret. For a more thorough discussion about gating and CFTR, please see the CFTR Review Diagram.
* CFTR is the only member of the ABC family that has an R-domain included in it's structure. The R-domain is situated between the two NBD domains in the primary structure and is known to be involved in gating. The two NBD domains are also involved and it is probably the role of the R-domain to modulate the function of the two NBDs. It is also possible that NBD2 regulates the R-domain and helps to deactivate the channel by encouraging phosphatase enzymes to dephosphorylate the R-domain when it is time for the channel to deactivate. The bottom line is that there appears to be a great deal of "cross-talk" taking place among the R and NBD domains during channel gating, and perhaps cross-talk among these domains and the intracellular loops as well.
* The inorganic phosphate (Pi) analogs are VO4 and BeF3. Both greatly prolong duration of bursts of activity of CFTR when it is already opened. They may bind to the site where Pi is released following ATP hydrolysis. Alone, they do not keep the channel open. It has been speculated that they may be "locking" CFTR into the open configuration by inhibiting ATP hydrolysis at the NBDs. PPi (pyrophosphate) also prolongs channel bursts. PPi appears to affect two steps during gating: first, it increases the rate that CFTR is made to open, and second, it delays the rate of opened channels to closing.
* In 1995, Gunderson and Kopito found two distinct conductance states in CFTR open state (O1 and O2). It turned out to be due to a block of the channel by the buffer MOPS, occurring 50% of the way thru the channel. MOPS binding may be altered by ATP hydrolysis at the NBDs by changing the conformation of the pore. Their conclusion was that ATP hydrolisis is an input of energy which changes the conformation of the pore. "These data indicate that ATP hydrolysis by the NBDs drives a series of asymmetric transitions in the gating cycle." They based this on patch clamp tracings which showed that "..the two gating states showed a strong asymmerty during bursts of activity; the first opening into a burst was most frequently into the O1 state, whereas the last opening before exit from a burst tended to be in the O2 state. Not all transitions proceeded in the sequence C-O1-O2-C, suggesting the reversibility of some transitions.....interventions which slowed or blocked hydrolysis prevented the transition to the O2 state and prolonged the O1 state." (therefore an asymmetric transition in gating cycle)
* CFTR, when expressed in CHO and BHK cells, does not become activated without PKC-dependent phosphorylation. CFTR activation does not seem to depend on intracellular calcium concentration. cGMP-dependent protein kinases (cGK) are of at least two types: type Ia and type II (cGKII). They are isotypes, expressed in lung and intestine. Both have been shown to phosphorylate CFTR, but only cGKII is able to activate CFTR. The result was similar to PKA activation, but slower.
* TNR-CFTR is the truncated version consisting only of the first 6 TM helices, NBD1 and R domains. Has been found as an isoform expressed in rat trachea, lung and kidney. Believed to be important in intracellular organelles.
* NBD1 may be necessary for forming part of pore of CFTR and takes part in conduction as well as ion selectivity. NBD1 and R-domain are needed for regulation of ORCC. Expression of CFTR in MDCK cells results in appearance of a novel ATP channel with properties different than CFTR. But deletion of R-domain leaves the CFTR chloride channel pore intact, while eliminating the ATP channel. (Sugita et.al 1998)
* "Tabcharani (1993) provided evidence that CFTR functions as a multi-ion pore. If channels translocate several ions simultaneously, the destabilization resulting from ion-ion repulsions allows rapid ion movement despite high affinity binding.
* ATP levels have to be close to physiological levels (~5mM). The open probability increases with increasing levels of ATP from 0.1 to 3 mM. Note: since PKA has a Km for ATP in the tens of micromolar range, activation is probably due to CFTR binding ATP. ADP inhibits at half-maximal concentrations of 10 mM. AMP does not inhibit at all. ATP hydrolysis is necessary for channel opening. Nonhydrolyzable analogs such as ATP-gamma-S or magnesium-free ATP do not activate it. Note: ATP-gamma-S is able to donate a phosphate when catalyzed by PKA, however. It isn't able to take part in ATPase reactions, though. It has been suggested that ATP hydrolysis is only necessary for the initial step in activation because nonhydrolyzable analogs are able to stabilize CFTR in the absence of ATP once the channel is open.
* Vmax of NBD1 is ~30 nmol/mg/min, which is considered low compared to other ATPases (20 times lower than Calcium-ATPase). As of 1996, there have been no known ATPase inhibitors which work on CFTR. It's possible it uses mechanisms other than those used by known ATPases. AMP-PNP is a nonhydrolyzable analog of ATP and can "lock" CFTR in the open configuration.
* The exact identity of the phosphatases which deactivate CFTR is in doubt. It's possible they may differ depending on the tissue-type.
* PKA phosphorylates each CFTR 5-6 times in vitro. Most phosphates end up on six of nine dibasic consensus sites. Mutation to alanine of between 1 and 3 of the 4 primary sites reduces the open probability slightly, while removal of all 4 decreases the activity significantly. Removal of all ten results in ~30% activity. The consensus sites for PKC allow for ~2 phosphates to be added to the R-domain. Other kinases may include tyrosine kinase p60(c-src) and cGMP-dependent protein kinase.
* SDS-PAGE shows a mass at 130 KDa (band A) and a diffusely migrating 150-170 band that represents the mature glycosylated form (band C). There is also a band at 135 KDa (band B) and this is the core glycosylated fraction.
* Sodium butyrate stimulate gene transcription.
* Inorganic phosphate analogues and polyphosphates lock CFTR into the open configuration once it is phosphorylated and ATP hydrolyzed.
* NS004 is a novel substituted benzimidazolone drug which activates mutant CFTR. When given with cAMP agonists, it produces anion efflux similar to wild-type. CPX (8-cyclopentyl-1,3,-dipropylxanthine) is an adenosine receptor antagonist but stimulates chloride efflux only in mutant deltaF508 cells.
* CFTR and perhaps Pgp are the only ABC proteins which function as a channel. It is not the only one which regulates other ion channels. Recently, it has been found that the ABC transporter SUR (sulfonylurea receptor) forms an oligomeric complex with inward rectifier K+ channel and regulates it's function. (4 SUR to one tetrameric K+ channel).
* "Recent analysis of particle sizes for a range of recombinant membrane proteins, using FFR-EM of oocyte membranes, led to the conclusion that CFTR exists as a dimer. (cell biol. 11235-40 1998). Moreover, preliminary electrophysiological analysis of concatenated CFTR dimers, comprising linked monomers with different gating characteristics, has now prompted Zerhusen and Ma (Pediatr.Pulmonol.Suppl. 17: 207, 1998) to propose that a single CFTR channel contains two CFTR polypeptides. Interestingly, complementation and reconstitution studies of the E.coli ABC transporter Ars, responsible for arsenical extrusion, had already led to a model in which the transporter functions as the equivalent of a dimer in which the NH2-terminal NBD of each monomer interacts with teh acid terminal NBD of the other to form a total of only 2 catalytic sites (JBC 271: 25247-52 1996).
* The open probability of CFTR in epithelia is 0.4-0.5, while from cardiac myocytes, it's 0.7-0.8
* Channel Blockers: sulfonylureas adn diarylsulfonylureas, disulfonic stilbenes, arylaminobenzoates.
* Channel Openers: xanthines, phosphatase inhibitors, isoflavones and flavones (genstein), benzimidazolones, benzoxazoles (chlorzoxazone), psoralens.
* The phosphatase PP2A completely dephorphorylates CFTR in vitro. PP1 and PP2B were much less effective. PP2C-alpha can dephorphorylate both intact CFTR and R-domain PKA-phosphorylated peptides.
* Perhaps the most convincing evidence to date for ATP hydrolysis of the NBDs being coupled to channel gating was done by Li, et.al in 1996. They found that purified and reconstituted CFTR hydrolyzes ATP at about the same rate that the channel opens and closes (~1/sec).
* A simplified model: There is a strict sequence of ATP binding, ADP and/or Pi release which must take place for opening and closing of channel. In this model, one NBD interacts with the other to "help" it gain or lose it's nucleotide or hydrolysis products. The interactions between the two NBDs may be similar to the way G-proteins bind other protein factors which encourage binding, hydrolysis, and release of GTP and hydrolysis products.
* There is a somewhat stronger influence of temperature on channel opening rather than channel closing.
* The novel synthesized xanthine derivative 3, 7-dimethyl-1-isobutylxanthine (X-33) is an activator of the CFTR channel in Calu-3 cells. Manipulating the chemical structure of xanthine derivatives offers an opportunity to identify further specific activators of CFTR in airway cells.
* Apigenin (4',5,7-trihydroxyflavone) is an activator of cystic fibrosis transmembrane conductance regulator (CFTR)-mediated Cl(-) currents across epithelia at low concentrations and a blocker at high concentrations. , block of Cl(-) current by apigenin was not affected by forskolin stimulation. Perhaps apigenin binds to a stimulatory and an inhibitory binding site, which could be distinguished by their affinities and the molecular interactions during binding.
* ORCC has not been identified as of November 2000
* Phloxine B slows deactivation and, at high concentrations, inhibited I(CFTR) weakly. Phloxine B modulates I(CFTR) by increasing channel activity and slowing channel deactivation; at high concentrations inhibition dominates. The effects may be mediated by direct interactions with CFTR from the inside of the cell.
* Several therapeutic modalities have been proposed for CF patients, including the phytoestrogen genistein. Experiments with radioactive photoactivatable estrogen derivatives demonstrated that these compounds bind directly to CFTR expressed in insect cells. Taken together, the data suggest that estrogens (like17beta-estradiol) can interact directly with CFTR to alter anion transport.
* Linsdell & Hanrahan (1998) showed that the CFTR can pass very large anions. In the presence of ATP the direction of flux is from the inside of cells to the outside. In the absence of ATP the channel appears to permit movement in either direction. The movement of organic molecules from inside cells to the outside is the basic activity of another ABC protein, P-glycoprotein; a protein which leads to multi-drug-resistance.
* The roles of CFTR as a Cl- channel and a conductance regulator are not mutually exclusive, since one function can be eliminated while the other is preserved. Schwiebert showed that "The first TMD of CFTR is an essential part of the Cl- channel pore, necessary for proper Cl- conduction by CFTR and its selectivity. Conversely, the first NBD and the regulatory domain are essential for the ability of CFTR to regulate ORCCs.
* Sugita et al. showed that expression of CFTR in MDCK cells results in the appearance of a novel ATP channel with different properties from CFTR.
* In addition to it's ability to transport Cl-, ATP and Na+, CFTR appears to lead to a multidrug resistance phenotype, perhaps due to the removal of some types of drugs. Interestingly, it appears that CFTR and P-glycoprotein have complementary patterns of expression.
* The known number of missense, nonsense and frameshift mutations within CFTR is now past 600, and yet a 3-base pair deletion in exon 10 is responsible for approximately 70% of the mutant alleles. This results in the loss the amino acid phenylalanine at codon 508 (i.e. it is a DF508 mutation). Not all individuals who have this mutation have the same course of disease. This mutation changes the structure, function and folding of the NBD1 domain, which leads to defective processing of the CFTR protein on its way through the endoplasmic reticulum. This causes a reduced level of protein to be found on the plasma membrane of epithelia.
* The non- DF508 mutations are considered relatively rare (from <1% to 10%). In the first half of the gene almost 60% of the mutations are substitutions or deletions of amino acids, while in the second half nearly 80% are nonsense, frameshift or splice mutations.
* Monovalent pseudohalides like C(CN)3 bind tightly within the ion conducting pathway of CFTR and block the flow of chloride, but analysis of block is complicated by fact that they also permeate.
* Relative conductance rates for CFTR indicate no other physiological ion conducts faster than chloride. There are "sticky" ions like SCN- and ClO4- which exhibit a reduced conductance (so does iodide).
* Arylaminobenzoates block within the pore of CFTR and therefore may be useful as probes of pore structure. Ex: NPPB, DPC.
* Recent studies show channels may be formed only from C-terminal half, however they were not selective for chloride over iodide.
* Freeze-Fracture and Lectin-Gold-labeling studies have been done: Note that FFR alone indicated in 1998 CFTR exists as dimers. Lectin-gold labeling of the single glycosylaton site in P-glycoprotein (at 25A resolution) indicated a monomeric form.
* n-acetyl-L-cysteine has been known for decades to help CF patients. It is believed to help activate the channel directly.