Mislocalization of delta F508 CFTR in cystic fibrosis sweat gland.

Misprocessing and mislocalization of the cystic fibrosis (CF) transmembrane conductance regulator (CFTR) has been described for the major CF-causing mutation (delta F508) in heterologous expression systems in vitro. We have generated monoclonal antibodies (mAbs) to CFTR with the aim of localizing the protein and its CF variants in vivo. Of the tissues where CFTR was observed, only the sweat gland is readily available and does not undergo secondary changes due to CF disease pathology. Sweat ducts from CF patients homozygous for delta F508 did not show the typical apical membrane staining seen in control biopsies. This demonstrates that the biosynthetic arrest and intracellular retention of delta F508 CFTR initially observed in vitro does occur in vivo and emphasizes the need to focus efforts on understanding the mislocalization.

Cell surface P-glycoprotein associated with multidrug resistance in mammalian cell lines.

The plasma membranes of hamster, mouse, and human tumor cell lines that display multiple resistance to drugs were examined by gel electrophoresis and immunoblotting. In every case, increased expression of a 170,000-dalton surface antigen was found to be correlated with multidrug resistance. This membrane component is of identical molecular size and shares some immunogenic homology with the previously characterized P-glycoprotein of colchicine-resistant Chinese hamster ovary cells. This finding may have application to cancer therapy.

Assignment of the gene for myelin proteolipid protein to the X chromosome: implications for X-linked myelin disorders.

Several inherited disorders in humans and in rodents result in myelin dysgenesis and a deficiency of the molecular constituents of myelin. A complementary DNA to one of the two major myelin proteins, myelin proteolipid protein (also known as lipophilin), has been used with Southern blot analysis of somatic cell hybrid DNA to map the human proteolipid protein gene to the middle of the long arm of the human X chromosome (bands Xq13-Xq22) and to assign the murine proteolipid protein gene to the mouse X chromosome. Comparison of the gene maps of the human and mouse X chromosomes suggests that myelin proteolipid protein may be involved in X-linked mutations at the mouse jimpy locus and has implications for Pelizaeus-Merzbacher disease, a human inherited X-linked myelin disorder.

Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA.

Overlapping complementary DNA clones were isolated from epithelial cell libraries with a genomic DNA segment containing a portion of the putative cystic fibrosis (CF) locus, which is on chromosome 7. Transcripts, approximately 6500 nucleotides in size, were detectable in the tissues affected in patients with CF. The predicted protein consists of two similar motifs, each with (i) a domain having properties consistent with membrane association and (ii) a domain believed to be involved in ATP (adenosine triphosphate) binding. A deletion of three base pairs that results in the omission of a phenylalanine residue at the center of the first predicted nucleotide-binding domain was detected in CF patients.

DNA-mediated transfer of multiple drug resistance and plasma membrane glycoprotein expression.

Colchicine-resistant Chinese hamster ovary (CHO) cell mutants whose resistance results from reduced drug permeability have been isolated previously in our laboratories. This reduced permeability affects a wide range of unrelated drugs, resulting in the mutants displaying a multiple drug resistance phenotype. A 170,000-dalton cell surface glycoprotein (P-glycoprotein) was identified, and its expression appears to correlate with the degree of resistance. In this study we were able to confer the multiple drug resistance phenotype on sensitive mouse L cells by DNA-mediated gene transfer of DNA obtained from the colchicine-resistant mutants. P-glycoprotein was detected in plasma membranes of these DNA transformants by staining with an antiserum raised against membranes of mutant CHO cells. These results are consistent with a causal relationship between P-glycoprotein expression and the multiple drug resistance phenotype.

Daunorubicin-resistant Chinese hamster ovary cells expressing multidrug resistance and a cell-surface P-glycoprotein.

Independent lines of Chinese hamster ovary cells resistant to the antineoplastic drug, daunorubicin, were obtained by clonal isolation in increasing drug concentrations. A single daunorubicin-resistant phenotype typified by reduced cellular drug accumulation was observed. These mutants displayed a complex phenotype of resistance to a variety of unrelated drugs. Such properties are similar to those of membrane-altered colchicine-resistant lines (V. Ling and L.H. Thompson, J. Cell. Physiol., 83: 103-116, 1974). Analysis of the plasma membrane components of the daunorubicin-resistant clones by gel electrophoresis revealed a prominent cell surface glycoprotein with a molecular weight of about 170,000. This component was immunologically cross-reactive with the cell surface P-glycoprotein of about the same molecular weight, previously identified in colchicine-resistant cells. Thus, it appears that the mechanism of resistance characterized by P-glycoprotein expression could be the basis of many drug-resistant phenotypes.

Collateral sensitivity of multidrug resistant cells to narcotic analgesics is due to effects on the plasma membrane.

It has previously been demonstrated that opiates interact directly with P-glycoprotein in drug resistant Chinese hamster ovary (CHO) cells (Callaghan, R. and Riordan, J.R. (1993) J. Biol. Chem. 268, 16059-16064). In this study we have examined the effects of several opiates on the growth of drug sensitive and resistant CHO and human MCF7 cell lines. The growth of P-glycoprotein expressing cells was inhibited by the opiates pentazocine, pethidine and naloxone to a greater extent than in drug sensitive cells. Since P-glycoprotein is localised at the plasma membrane the effects of opiates on membrane biophysical properties were investigated. The opiates caused a fluidizing effect in membranes from P-glycoprotein expressing cells and decreased the basal level of P-glycoprotein phosphorylation. In addition, they were able to increase the leakage of the membrane impermeant compound 6-carboxyfluorescein entrapped in model membrane vesicles. The ability to alter membrane biophysical properties correlated with the inhibitory effects on growth of drug resistant cells. These results suggest that the collateral sensitivity of P-glycoprotein expressing cell lines to opiates is mediated by the drugs' effects on the plasma membrane.

CFTR, a channel with the structure of a transporter.

The cystic fibrosis transmembrane conductance regulator (CFTR) belongs to a superfamily of active transport molecules. However, when expressed in a wide variety of heterologous cell systems and when purified to homogeneity and reconstituted in planar lipid bilayers, it exhibits low conductance chloride channel activity. We postulate that the active transport capability of the molecule has been adapted to provide very stringent metabolic control of this channel which is responsible for chloride secretion and hydration of wet epithelial surfaces.

Isolation and characterization of the plasma membranes of cultured lymphoblasts from patients with cystic fibrosis and normal individuals.

Plasma membranes were isolated from two cystic fibrosis lymphoblastoid cell lines and two age and sex-matched normal cell lines after controlled cell disruption. The preparations were enriched approx. 20-fold in plasma membrane markers and were essentially free of markers for other organelles except the Golgi-localized enzyme, galactosyltransferase. Sodium dodecyl sulphate polyacrylamide gel electrophoresis showed virtually identical polypeptide profiles after staining with Coomassie blue. Membranes from normal cells contained 0.49--0.67 mumol/mg of total phospholipid/mg membrane protein and those from cystic fibrosis cells, 0.53-0.67 mumol. The cholesterol content was 0.35-0.41 ummol/mg protein for normal, and 0.43--0.49 mumol for cystic fibrosis cell membranes. The proportions of individual phospholipid classes and their fatty acid compositions as well as that of the total membrane lipid also did not differ significantly in the two cell types. The bulk lipid fluidity as reflected by the fluorescence polarization of beta-parinaric acid and diphenylhexatriene was also not changed. The carbohydrate composition of cell surface glycopeptides released with papain was not consistently different with the normal and cystic fibrosis cells. The quantity of fucose in intact plasma membranes was also quantitated enzymatically and found to be very similar in both cell types.

Small copper-binding proteins from normal and copper-loaded liver.

The same three low molecular weight copper-binding proteins which have been purified from the soluble fractions of Cu2+-loaded livers (200 mug Cu2+ per g wet weight) were also found to be present in livers of normal rats (4 mug Cu2+ per g wet weight) but in much smaller quantities. Therefore, Cu2+ loading enhances the amount of preexisting proteins rather than inducing the synthesis of new proteins. Amino acid analyses of each of the three showed them to be different from the metallothioneins which are present on loading with other trace metals including Cd2+ and Zn2+.

Binding of [3H]ctyochalasin B and [3H]colchicine to isolated liver plasma membranes.

The binding to isolated hepatocyte plasma membranes of radioactively labelled inhibitors of microfilamentous and microtubular protein function ([3H]cytochalasin B and [3H]colchicine, respectively) was studied as one means of assessing the degree of association of these proteins with cell surface membranes. [3H]Cytochalasin B which behaved identically to the unlabelled compound with respect to binding to these membranes was prepared by reduction of cytochalasin A with NaB3H4. The binding was rapid, readily reversible, proportional to the amount of membrane and relatively insensitive to changes of pH or ionic strength. At 10(-6) M [3H]cytochalasin B, glucose of p-chloromercuribenzoate, an inhibitor of glucose transport inhibited binding by about 20%; treatment of membranes with 0.6 M KI which depolymerizes F actin to G actin caused about 60% inhibition of binding. These two types of inhibition were additive indicating two separate classes of binding sites, one associated with sugar transport and one with microfilaments. Filamentous structures with the diameter of microfilaments (50 A) were seen in electron micrographs of thin sections of the membranes. At concentrations greater than 10(-5) M [3H]cytochalasin B, binding was proportional to drug concentration, characteristic of non-specific adsorption or partitioning. Intracellular membranes of the hepatocyte also bound [3H]cytochalasin B, those of the smooth endoplasmic reticulum to a greater extent than plasma membranes. [3H]Colchicine bound to plasma membranes in proportion to the amount of membrane and at a rate compatible with binding to tubulin. However, other properties of the binding including effects of temperature, drug concentration and antisera against tubulin were different from those of binding to tubulin. Hence, no evidence was obtained for association of microtubular elements with these membranes. Despite this there appeared to be an interdependence between microtubule and microfilament inhibitors: vinblastine sulfate stimulated [3H]cytochalasin B binding and cytochalasin B stimulated 3H colchicine binding. [3H]Colchicine also bound to intracellular membranes, especially smooth microsomes.

Induction of alkaline phosphatase in cultured human fibroblasts. Comparison of normal cells and those from patients with cystic fibrosis.

The membrane glycoprotein enzyme, alkaline phosphatase was induced in cultured human fibroblasts by dibutyryl cyclic AMP, sodium butyrate, the serum glycoprotein fetuin, the Tamm-Horsfall urinary glycoprotein, and by a number of inhibitors of DNA synthesis. The uninduced basal enzyme activity increased at later stages of growth when the cells became confluent. Induction by dibutyryl cyclic AMP or fetuin was most effective when the agents were added after the cells had reached stationary phase and was maximal after at least two days of exposure. The levels of induction resulting from the addition of pairs of the agents, dibutyryl cyclic AMP, n-butyrate and fetuin were additive indicating that these have different modes of action. The inhibitors of DNA synthesis, cytosine arabinoside, hydroxyurea, and methothrexate were less effective inducers. Bromodeoxyuridine which also has non-DNA mediated effects induced to the same extent as dibutyryl cyclic AMP. Similar experiments with sex- and age-matched cell strains derived from patients with cystic fibrosis failed to detect differences in the levels of induction from those observed in normal cells. In addition, the combined inductive effects of Tamm-Horsfall glycoprotein, isoproterenol and theophylline, were similar with normal and cystic fibrosis cells.

Plasma membrane lipids of human diploid fibroblasts from normal individuals and patients with cystic fibrosis.

The lipid composition of isolated plasma membranes of human skin fibroblasts is described for the first time. Plasma membranes from a number of strains of fibroblasts from patients with cystic fibrosis and matched normals were isolated by a recently described procedure and analysed for major phospholipid classes, cholesterol and fatty acids. No differences in the quantities of these compounds were detected between cells of the two different origins. The fetal calf serum used to supplement the growth medium contained relatively more palmitoleate and oleate but less stearate than the membranes. There were also no consistent differences between cystic fibrosis and normal membranes in terms of the fatty acid compositions of their individual phospholipid classes. Consistent with this lack of chemical change in the lipids of membranes of cystic fibrosis cells, the degree of fluorescence polarization of diphenylhexatriene, an index of fluidity, was also unchanged.

Influence of phosphorylation by protein kinase A on CFTR at the cell surface and endoplasmic reticulum.

CFTR possesses a large cluster of strict dibasic consensus sites for phosphorylation by protein kinase A (PKA) in the R-domain and an obligatory dependence on phosphorylation is a hallmark of CFTR Cl(-) channel function. Removal of as many as 11 of these sites reduces the conformational change in the R-domain and the degree of channel activation in response to PKA. However, until recently a completely PKA-unresponsive CFTR variant has not been reported, leaving open the possibility that the residual response may be mediated by associating ancillary phosphoproteins. We traced the residual PKA-catalyzed (32)P-labelling of the variant with 11 sites mutagenized (11SA) to distinct CNBr phosphopeptides within the R-domain. Mutagenesis of 4 additional monobasic sites in these segments produced a 15SA variant in which Cl(-) channel response to PKA was abolished. Therefore, it can be concluded that ancillary phosphoproteins do not contribute to CFTR activation by PKA. Notably, however, the 15SA protein did exhibit a low level of constitutive channel activity not dependent on PKA, which might have reflected a down-regulating effect of phosphorylation of one or two of the 15 sites as suggested by others. However, this did not prove to be the case.Since immature CFTR has been claimed to be active in the endoplasmic reticulum (ER), we also examined whether it can be phosphorylated in cells and what influence if any this might have on its susceptibility to degradation. Teleologically, activation by phosphorylation of CFTR Cl(-) channels in the ER might be undesirable to the cell. Using various phosphorylation site mutants and kinase and phosphatase inhibitors in pulse-chase experiments, we have found that although nascent CFTR can be phosphorylated at the ER, this is without effect on its ability to mature and avoid proteolysis. Furthermore, we found that microsomes from cells expressing CFTR processing mutants such as DeltaF508 do not generate Cl(-) active channels when fused with planar bilayers unless maturation is promoted, e.g. by growth of cells at reduced temperature or other means. We conclude that the ER-retained mutant nascent chains which are incapable of maturation may be phosphorylated but do not form active channels. Stimulation by PKA of the insertion of CFTR containing vesicles into the plasma membrane as part of the mechanism of stimulation of chloride secretion has been reported, as has an influence of CFTR on the balance between endocytosis and exocytosis but these findings have not been universally confirmed.

Dual effects of ADP and adenylylimidodiphosphate on CFTR channel kinetics show binding to two different nucleotide binding sites.

The CFTR chloride channel is regulated by phosphorylation by protein kinases, especially PKA, and by nucleotides interacting with the two nucleotide binding domains, NBD-A and NBD-B. Giant excised inside-out membrane patches from Xenopus oocytes expressing human epithelial cystic fibrosis transmembrane conductance regulator (CFTR) were tested for their chloride conductance in response to the application of PKA and nucleotides. Rapid changes in the concentration of ATP, its nonhydrolyzable analogue adenylylimidodiphosphate (AMP-PNP), its photolabile derivative ATP-P3-[1-(2-nitrophenyl)ethyl]ester, or ADP led to changes in chloride conductance with characteristic time constants, which reflected interaction of CFTR with these nucleotides. The conductance changes of strongly phosphorylated channels were slower than those of partially phosphorylated CFTR. AMP-PNP decelerated relaxations of conductance increase and decay, whereas ATP-P3-[1-(2-nitrophenyl)ethyl]ester only decelerated the conductance increase upon ATP addition. ADP decelerated the conductance increase upon ATP addition and accelerated the conductance decay upon ATP withdrawal. The results present the first direct evidence that AMP-PNP binds to two sites on the CFTR. The effects of ADP also suggest two different binding sites because of the two different modes of inhibition observed: it competes with ATP for binding (to NBD-A) on the closed channel, but it also binds to channels opened by ATP, which might either reflect binding to NBD-A (i.e., product inhibition in the hydrolysis cycle) or allosteric binding to NBD-B, which accelerates the hydrolysis cycle at NBD-A.

Voltage-dependent block of the cystic fibrosis transmembrane conductance regulator Cl- channel by two closely related arylaminobenzoates.

The gene defective in cystic fibrosis encodes a Cl- channel, the cystic fibrosis transmembrane conductance regulator (CFTR). CFTR is blocked by diphenylamine-2-carboxylate (DPC) when applied extracellularly at millimolar concentrations. We studied the block of CFTR expressed in Xenopus oocytes by DPC or by a closely related molecule, flufenamic acid (FFA). Block of whole-cell CFTR currents by bath-applied DPC or by FFA, both at 200 microM, requires several minutes to reach full effect. Blockade is voltage dependent, suggesting open-channel block: currents at positive potentials are not affected but currents at negative potentials are reduced. The binding site for both drugs senses approximately 40% of the electric field across the membrane, measured from the inside. In single-channel recordings from excised patches without blockers, the conductance was 8.0 +/- 0.4 pS in symmetric 150 mM Cl-. A subconductance state, measuring approximately 60% of the main conductance, was often observed. Bursts to the full open state lasting up to tens of seconds were uninterrupted at depolarizing membrane voltages. At hyperpolarizing voltages, bursts were interrupted by brief closures. Either DPC or FFA (50 microM) applied to the cytoplasmic or extracellular face of the channel led to an increase in flicker at Vm = -100 mV and not at Vm = +100 mV, in agreement with whole-cell experiments. DPC induced a higher frequency of flickers from the cytoplasmic side than the extracellular side. FFA produced longer closures than DPC; the FFA closed time was roughly equal (approximately 1.2 ms) at -100 mV with application from either side. In cell-attached patch recordings with DPC or FFA applied to the bath, there was flickery block at Vm = -100 mV, confirming that the drugs permeate through the membrane to reach the binding site. The data are consistent with the presence of a single binding site for both drugs, reached from either end of the channel. Open-channel block by DPC or FFA may offer tools for use with site-directed mutagenesis to describe the permeation pathway.

Permeability of wild-type and mutant cystic fibrosis transmembrane conductance regulator chloride channels to polyatomic anions.

Permeability of the cystic fibrosis transmembrane conductance regulator (CFTR) chloride channel to polyatomic anions of known dimensions was studied in stably transfected Chinese hamster ovary cells by using the patch clamp technique. Biionic reversal potentials measured with external polyatomic anions gave the permeability ratio (P/P) sequence NO > Cl > HCO > formate > acetate. The same selectivity sequence but somewhat higher permeability ratios were obtained when anions were tested from the cytoplasmic side. Pyruvate, propanoate, methane sulfonate, ethane sulfonate, and gluconate were not measurably permeant (P/P < 0.06) from either side of the membrane. The relationship between permeability ratios from the outside and ionic diameters suggests a minimum functional pore diameter of approximately 5.3 A. Permeability ratios also followed a lyotropic sequence, suggesting that permeability is dependent on ionic hydration energies. Site-directed mutagenesis of two adjacent threonines in TM6 to smaller, less polar alanines led to a significant (24%) increase in single channel conductance and elevated permeability to several large anions, suggesting that these residues do not strongly bind permeating anions, but may contribute to the narrowest part of the pore.

A two-domain model for the R domain of the cystic fibrosis transmembrane conductance regulator based on sequence similarities.

CFTR belongs to a group of proteins sharing the structural motif of six transmembrane helices and a nucleotide binding domain. Unique to CFTR is the R domain, a charged cytoplasmic domain. Comparison of R domain sequences from ten species revealed that the N-terminal third is highly conserved, while the C-terminal two-thirds is poorly conserved. The R domain shows no strong sequence similarity to known proteins; however, 14 viral pol proteins show limited similarity to fragments of the R domain. Analysis revealed a relationship between the N- and C-terminal fragments of the R domain and two discontinuous fragments of the pol protein. These observations support a two-domain model for the R domain.

Regulation of CFTR ion channel gating by MgATP.

Single channel currents of wild-type CFTR reconstituted in lipid bilayers were recorded to study the temperature dependence of channel gating between +20 degrees C and +40 degrees C. The opening of the channel was highly temperature dependent and required an activation energy of about 100 kJ/mol. Closing of the channel was only weakly temperature dependent with an activation energy close to that of diffusion in water. We found no significant difference in the free energy between the open and closed states. Most of the excess energy needed to activate channel opening is used to diminish the entropy of the open state. This structural reorganization is initiated by ATP binding followed by interconversion to the open channel structure as the CFTR-ATP-Mg complex passes to the transition state for hydrolysis. The energy of the CFTR-ATP-Mg interaction in the transition state is responsible for the CFTR ion channel opening rather than the energy of ATP hydrolysis. Channel closing is a diffusion limited process and does not require additional ATP binding.

Inhibition of epithelial Na+ currents by intracellular domains of the cystic fibrosis transmembrane conductance regulator.

Cystic fibrosis is characterized by an impaired cyclic adenosine 3,5-monophosphate (cAMP) activated Cl- conductance in parallel with an enhanced amiloride sensitive Na+ conductance (ENaC) of the respiratory epithelium. Very recently, acute downregulation of ENaC by the cystic fibrosis transmembrane conductance regulator (CFTR) was demonstrated in several studies. The mechanism, however, by which CFTR exerts its inhibitory effect on ENaC remains obscure. We demonstrate that cytosolic domains of human CFTR are sufficient to induce inhibition of rat epithelial Na+ currents (rENaC) when coexpressed in Xenopus oocytes and stimulated with 3-isobutyl-1-methylxanthine (IBMX). Moreover, mutations of CFTR, which occur in cystic fibrosis, abolish CFTR-dependent downregulation of rENaC. Yeast two hybrid analysis of CFTR domains and rENaC subunits suggest direct interaction between the proteins. Enhanced Na+ transport as found in the airways of cystic fibrosis patients is probably due to a lack of CFTR dependent downregulation of ENaC.