High channel-mediated water permeability in rabbit erythrocytes: characterization in native cells and expression in Xenopus oocytes.

Erythrocytes from several mammalian species contain mercurial-sensitive water transporters. By a stopped-flow light scattering technique, osmotic water permeability (Pf) was exceptionally high in rabbit erythrocytes (0.053 +/- 0.002 cm/s) and reversibly inhibited by 98% by p-(chloromercuri)benzenesulfonate (pCMBS). The activation energy (Ea) was 4.6 kcal/mol (15-37 degrees C). pCMBS inhibition was half-maximal at 0.1 mM (60-min incubation); at 1 mM pCMBS, half-maximal inhibition occurred in 8 min. Pf was also inhibited by HgCl2 and pCMB with greater than 90% inhibition in 5 min. There was no inhibition by high concentrations of phloretin, DNDS, cytochalasin B, amiloride, ouabain, furosemide, and several proteases. In defolliculated Xenopus oocytes microinjected with 50 nL of water or unfractionated mRNA (1 mg/mL) from rabbit reticulocytes, oocyte Pf assayed at 10 degrees C after 72-h incubation increased from (4 +/- 1) X 10(-4) cm/s (water injected) to (18 +/- 2) X 10(-4) cm/s (mRNA injected). Pf increased linearly with [mRNA] (0-75 ng/oocyte) and was inhibited slowly and reversibly by pCMBS and immediately by HgCl2 but not by cytochalasin B, phloretin, or DNDS. Ea was 9.6 kcal/mol (water injected) and 2.6 kcal/mol (mRNA injected). These results demonstrate that rabbit erythrocytes have the highest Pf and the greatest percentage inhibition of Pf by mercurials of any mammalian erythrocyte studied. The characteristics of the expressed and native water channels were similar, suggesting that the erythrocyte water channel is a membrane protein suitable for expression cloning.     

Cytochalasin B-binding proteins in rabbit erythrocyte membranes and their post-natal change in relation to the glucose carrier activity

Two distinct, carrier-mediated glucose uptake processes, a fast, cytochalasin B-sensitive and a slow, cytochalasin B-insensitive flux are identified in parallel in newborn rabbit erythrocytes. The fast, cytochalasin B-sensitive carrier function disappears as rabbits age, and only the slow cytochalasin B-insensitive carrier function is observed with adult rabbit erythrocytes.Three different cytochalasin B binding sites are distinguished in newborn rabbit erythrocytes; a glucose-sensitive site (site I), a cytochalasin E-sensitive site (site II), and a site insensitive to both glucose and cytochalasin E. With adult rabbit erythrocytes, only a cytochalasin E-sensitive site is detected. The glucose-sensitive site disappears as rabbits age, with a time course which is comparable to that of the disappearance of the cytochalasin B-sensitive glucose carrier function. The cytochalasin E-sensitive cytochalasin B binding site does not increase during this change, thus the disappearance of the glucose-sensitive site is not due to its conversion to a cytochalasin E-sensitive site. Sodium dodecyl sulfate polyacrylamide gel electrophoresis of rabbit erythrocyte ghosts revealed a partial decrease in each of the membrane polypeptides of approximate molecular weights of 240 000, 160 000 and 50 000 as rabbits aged. It is concluded that the cytochalasin B-sensitive glucose carrier of fetal rabbit erythrocytes, like that of the human erythrocyte, is tightly associated with the site I cytochalasin B-binding protein, while the cytochalasin B-insensitive glucose carrier, operative in adult rabbit erythrocytes, is not.     

Inhibition of hexose transport by adenosine derivatives in human erythrocytes

The human erythrocyte membrane carriers for hexoses and nucleosides have several structural features in common. In order to assess functional similarities, the effects of adenosine derivatives on hexose transport and cytochalasin B binding sites were studied. Adenosine inhibited zero-trans uptake of 3-O-methylglucose half-maximally at 5 mM, while more hydrophobic adenosine deaminase-resistant derivatives were ten- to 20-fold more potent transport inhibitors. However, degradation of adenosine accounted for very little of this difference in potency. Hexose transport was rapidly inhibited by N6-(L-2-phenylisopropyl)adenosine at 5 degrees C in a dose-dependent fashion (EC50 = 240 microM), to lower the transport Vmax without affecting the Km. A direct interaction with the carrier protein was further indicated by the finding that N6-(L-2-phenylisopropyl)adenosine competitively inhibited [3H]cytochalasin B binding to erythrocytes (Ki = 143 microM) and decreased [3H]cytochalasin B photolabeling of hexose carriers in erythrocyte ghosts. The cross-reactivity of adenosine and several of its derivatives with the hexose carrier suggests further homologies between the carriers for hexoses and nucleosides, possibly related to their ability to transport hydrophilic molecules through the lipid core of the plasma membrane.     

Cytochalasin B does not serve as a marker of glucose transport in rabbit erythrocytes

Glucose inhibitable cytochalasin B binding to erythrocyte membranes has been used as a marker of the glucose transporter. Glucose transport and cytochalasin B binding in rabbit erythrocytes differ from those activities found in human erythrocytes. We evaluated the uptake of 3-0-methylglucose and found similar Km (4.81 +/- 1.20 mM (SEM) and 6.59 +/- 0.72 mM) though significantly different Vmax (5.2 +/- 0.7 nM . min-1/10(9) cells and 234 +/- 13 nM X min -1/10(9) cells, p less than 0.001) for rabbit and human erythrocytes, respectively. Equilibrium binding of cytochalasin B to human erythrocyte membranes demonstrates a high affinity cytochalasin B binding site (Kd 38.6 +/- 22.7 nM) which is displaced by glucose. No comparable glucose inhibitable cytochalasin B site exists for rabbit erythrocyte membranes. Photoaffinity labeling of cytochalasin B confirms the presence of a glucose inhibitable cytochalasin B binding site in human, but not rabbit erythrocyte membranes. Cytochalasin B binding is a useful method in the identification of the glucose transporter in human cells, but the technique may be less useful in other species.     

Comparison of the effects of human erythrocyte ghosts and intact erythrocytes on platelet interactions with subendothelium in flowing blood

We investigated whether ghosts behaved similarly to intact erythrocytes to maintain regular primary hemostasis under flow conditions. To this end we performed perfusion experiments with whole blood in which erythrocytes were replaced by pink ghosts, and platelet interaction with the subendothelial surface of a damaged vessel was morphometrically evaluated. The same objective was sought by means of studies with a platelet function analyzer (PFA-100(TM) instrument). Perfusions performed with control blood reconstituted with intact erythrocytes gave rise to 0.4+/-0.2% contact but not spread platelets, 10.8+/-3.4% adhering and spread platelets, 16.3+/-4.6% platelets in thrombi, with 27.5+/-7.4% of the surface covered. Even though the average diameter of the ghosts was smaller than that of intact erythrocytes (5.3 microm vs. 7.7 microm), the values obtained in perfusions performed with ghosts were similar to those of the erythrocyte controls. Studies performed with the PFA-100(TM) analyzer were consistent with those observed in perfusion studies. The viscosity of control blood was compared with that of blood reconstituted with ghosts. At shear rates lower than 450 s(-1), the viscosity of the ghost samples was higher than that of the controls, but the difference progressively decreased as shear rate increased up to 750 s(-1) (3.61+/-0.15 and 3.71+/-0.17 cP, respectively). In conclusion, the results of our study showed that ghosts behaved similarly to intact erythrocytes in maintaining a normal platelet interaction with digested subendothelium, under conditions of moderate shear rate and constant hematocrit (40%). The rheological activity of ghosts, bodies that are metabolically less active, was sufficient for them to satisfactorily act as substitutes for intact erythrocytes in our system.     

Functional reconstitution of the isolated erythrocyte water channel CHIP28.

Measurements of water permeability indicate the existence of a facilitated water transporting pathway in erythrocytes, kidney tubules and amphibian urinary bladder. Two lines of evidence suggest that one type of water channel is an approximately 30-kDa protein: the approximately 30-kDa target size determined by radiation inactivation (van Hoek, A. N., Hom, M. L., Luthjens, L. H., de Jong, M. D., Dempster, J. A., and van Os, C. H. (1991) J. Biol. Chem. 266, 16633-16635) and the increased water permeability in oocytes that express mRNA encoding a 28-kDa erythrocyte protein (CHIP28, Preston, B. M., Carroll, T. P., Guggino, W. B., and Agre, P. (1992) Science 256, 385-387). We report direct evidence that CHIP28 is the erythrocyte water channel. Osmotic water permeability (Pf) remained high (0.029 cm/s, 37 degrees C) when erythrocyte membranes were stripped of nearly all proteins except for CHIP28. N-terminal sequence analysis confirmed that the 28-kDa protein was CHIP28. Pf in proteoliposomes reconstituted with solubilized CHIP28 was high (Pf = 0.03 cm/s, 37 degrees C), the activation energy was low (2.2 kcal/mol), and Pf was decreased by greater than 50-fold by mercurial sulfhydryl reagents and Me2SO. The single-channel water permeability was approximately 10(-13) cm3/s, slightly higher than that of the gramicidin A channel. The water channel excluded the small solute urea. These data establish a procedure to reconstitute functional water channels into liposomes and demonstrate that CHIP28 is the erythrocyte water channel.     

Phospholipid asymmetry in human erythrocyte ghosts

Using phospholipase digestion and the fluorescent probe merocyanine 540 the maintenance of phospholipid asymmetry in the plasma membrane of human erythrocyte ghosts was investigated. Digestion with phospholipase A2 indicated that ghosts prepared in the presence of Mg++ as the only divalent cation retained the normal phospholipid asymmetry characteristic of intact erythrocytes. These ghosts, like normal erythrocytes, also failed to stain with merocyanine 540. However, the presence of as little as 5-10 microM Ca++ during ghost preparation resulted in ghosts in which lipid asymmetry had been abolished, as indicated by phospholipase digestion. Moreover, these ghosts stained with merocyanine 540. In contrast to ghosts, intact erythrocytes treated with ionophore required millimolar levels of Ca++ ions to disrupt membrane lipid asymmetry. To discover the reason for this difference in behavior between ghosts and intact cells, ghosts were prepared from preswollen cells using only small volumes of buffer for lysis. These experiments demonstrated that as the cellular contents of erythrocytes are diluted, the asymmetric arrangement of phospholipids becomes more sensitive to disruption by Ca++.     

Dynamics of membrane structure of frog erythrocyte ghosts measured with a nanosecond fluorometer

The dynamics of membrane microstructure was studied as molecular motions of phospholipids for bullfrog erythrocyte ghosts by the DPH fluorescence depolarization technique with a nanosecond fluorometer. The bullfrog erythrocyte ghosts were obtained by hypotonic lysis and collagenase treatment. The constituents of membrane proteins were confirmed by the disk gel electrophoresis. The viscosity of erythrocyte membrane ghosts was estimated to be 3.3 +/- 1.0 at 10 degrees C, and 2.1 +/- 0.1 at 20 degrees C and 1.3 +/- 0.2 at 30 degrees C in the unit of poise and the wobbling angle of lipid molecule was 35 +/- 1, 41 +/- 1 and 43 +/- 1 degree at the respective temperatures on an average and +/- S.D. The viscosity is lower than that of human erythrocytes. The relatively low viscous phospholipid bilayer may be one of the factors for the deformability of bullfrog erythrocytes.     

Fluidity of human erythrocyte ghosts.

The fluidity, defined by its two components, the order parameter, S, and the rotation correlation time, tau c, was studied on healthy human erythrocytes ghosts. We also measured ghost protein, cholesterol and phospholipid contents as well as acetylcholinesterase activities. No statistically significant difference was evidenced between erythrocyte ghosts from men and women. Whereas tau c values did not significantly vary among sample elements, variations of ghost order parameters about the mean were explained at 61% by changes in cholesterol contents and, to a lesser extent, in protein contents. No relationship was evidenced between ghost order parameter values and those of corresponding acetylcholinesterase activities. Liposomes prepared from ghost lipid extracts had much lower order parameter values than did corresponding ghosts. A few experiments were performed in the same way on ghosts from sickle blood. This disease appeared to decrease the bilayer lipid motionnal freedom as an increase of the order parameter values was evidenced.     

Reaction of an exofacial sulfhydryl group on the erythrocyte hexose carrier with an impermeant maleimide. Relevance to the mechanism of hexose transport

The presence of a reactive exofacial sulfhydryl on the human erythrocyte hexose carrier was used to test several predictions of the alternating conformation or one-site model of transport. The cell-impermeant glutathione-maleimide-I (GS-Mal) irreversibly inhibited hexose entry by decreasing the transport Vmax. This effect was potentiated by phloretin and maltose but decreased by cytochalasin B, indicating that under the one-site model the external sulfhydryl is on the outward-facing carrier but that it does not overlap with the exofacial substrate-binding site. Incubation of erythrocytes with maltose competitively inhibited the binding of [3H]cytochalasin B to the inward-facing carrier (Ki = 40 mM). Furthermore, both equilibrium cytochalasin B binding and its photolabeling of the band 4.5 carrier protein were decreased in ghosts prepared from GS-Mal-treated cells. Thus induction of an outward-facing carrier conformation with either maltose or GS-Mal caused the endofacial substrate-binding site to disappear. Dose-response studies of GS-Mal treatment of intact cells suggested that some functional carriers lack a reactive external sulfhydryl, which can be partially regenerated by pretreatment with excess cysteine. These data provide direct support for the one-site model of transport and further define the role of the external sulfhydryl in the transport mechanism.     

Membrane transport in resealed haemoglobin-containing human erythrocyte 'ghosts' prepared by a dialysis procedure

Resealed haemoglobin-containing erythrocyte ‘ghosts’ have been proposed as $\text{in}\text{vivo}$ carriers for enzyme replacement therapy. Transport of substrates and metabolites into and out of the ‘ghost’ has been suggested to be a limiting factor in such therapy. Studies of the transport of L-phenylalanine and of uric acid in normal human erythrocytes and prepared ‘ghosts’, in which the transport of sodium ions and D-glucose was intact, have shown that transport characteristics of ‘ghosts’ are identical to those of normal erythrocytes with transport not being a quantitatively limiting factor.     

Water exchange through erythrocyte membranes: p-choloromercuribenzene sulfonate inhibition of water diffusion in ghosts studied by a nuclear magnetic resonance technique

A comparison of water diffusion in human erythrocytes and ghosts revealed a longer relaxation time in ghosts, A comparison of water diffusion in human erythrocytes and ghosts revealed a longer relaxation time in ghosts, corresponding to a decreased exchange rate. However, the diffusional permeability of ghosts was not significantly different from that of erythrocytes . The changes in water diffusion following exposure to p-chloromercuribenzene sulfonate (PCMBS) have been studied on ghosts suspended in isotonic solutions. It was found that a significant inhibitory effect of PCMBS on water diffusion occurred only after several minutes of incubation at 37°C. No inhibition was noticed after short incubation at 0°C as previously used in some labelling experiments. This indicates the location in the membrane interior of the SH groups involved in water diffusion across human erythrocyte membranes. The nuclear magnetic resonance ( n . m . r . ) method appears as a useful tool for studying changes in water diffusiofl in erythrocyte ghosts with the aim of locating the water channel.     

Purification and characterization of human erythrocyte glucose transporter in decylmaltoside detergent solution.

The facilitative glucose transporter from human erythrocyte membrane, Glut1, was purified by a novel method. The nonionic detergent decylmaltoside was selected for solubilization on the basis of its efficiency to extract Glut1 from the erythrocyte membrane and its ability to maintain the protein in a monodisperse state. A positive, anion-exchange chromatography protocol produced a Glut1 preparation of 95% purity with little copurified lipid. This protein preparation exhibited cytochalasin B binding in detergent solution, as measured by tryptophan fluorescence quenching. The transporter existed as a monomer in decylmaltoside, with a Stokes radius of 50 A and a molecular mass of 147 kDa for the protein-detergent complex. We screened detergent, pH, additive, and lipid and have found conditions to maintain Glut1 monodispersity for 8 days at 25 degrees C or over 5 weeks at 4 degrees C. This Glut1 preparation represents the best available material for two- and three-dimensional crystallization trials of the human glucose transporter protein.     

Inhibition by nucleosides of glucose-transport activity in human erythrocytes.

The interaction of nucleosides with the glucose carrier of human erythrocytes was examined by studying the effect of nucleosides on reversible cytochalasin B-binding activity and glucose transport. Adenosine, inosine and thymidine were more potent inhibitors of cytochalasin B binding to human erythrocyte membranes than was D-glucose [IC50 (concentration causing 50% inhibition) values of 10, 24, 28 and 38 mM respectively]. Moreover, low concentrations of thymidine and adenosine inhibited D-glucose-sensitive cytochalasin B binding in an apparently competitive manner. Thymidine, a nucleoside not metabolized by human erythrocytes, inhibited glucose influx by intact cells with an IC50 value of 9 mM when preincubated with the erythrocytes. In contrast, thymidine was an order of magnitude less potent as an inhibitor of glucose influx when added simultaneously with the radioactive glucose. Consistent with this finding was the demonstration that glucose influx by inside-out vesicles prepared from human erythrocytes was more susceptible to thymidine inhibition than glucose influx by right-side-out vesicles. These data, together with previous suggestions that cytochalasin B binds to the glucose carrier at the inner face of the membrane, indicate that nucleosides are capable of inhibiting glucose-transport activity by interacting at the cytoplasmic surface of the glucose transporter. Nucleosides may also exhibit a low-affinity interaction at the extracellular face of the glucose transporter.     

Optical measurement of osmotic water transport in cultured cells. Role of glucose transporters

Methodology was developed to measure osmotic water permeability in monolayer cultured cells and applied to examine the proposed role of glucose transporters in the water pathway (1989. Proc. Natl. Acad. Sci. USA. 86:8397-8401). J774 macrophages were grown on glass coverslips and mounted in a channel-type perfusion chamber for rapid fluid exchange without cell detachment. Relative cell volume was measured by 45 degrees light scattering using an inverted microscope; measurement accuracy was validated by confocal imaging microscopy. The time required for greater than 90% fluid exchange was less than 1 s. In response to a decrease in perfusate osmolality from 300 to 210 mosM, cells swelled without lag at an initial rate of 4.5%/s, corresponding to a water permeability coefficient of (6.3 +/- 0.4) x 10(-3) cm/s (SE, n = 20, 23 degrees C), assuming a cell surface-to-volume ratio of 4,400 cm-1. The initial rate of cell swelling was proportional to osmotic gradient size, independent of perfusate viscosity, and increased by amphotericin B (25 micrograms/ml), and had an activation energy of 10.0 +/- 1 kcal/mol (12-39 degrees C). The compounds phloretin (20 microM) and cytochalasin B (2.5 micrograms/ml) inhibited glucose transport by greater than 85% but did not influence Pf in paired experiments in which Pf was measured before and after inhibitor addition. The mercurials HgCl2 (0.1 mM) and p-chloromercuribenzoate (1 mM) did not inhibit Pf. A stopped-flow light scattering technique was used to measure Pf independently in J774 macrophages grown in suspension culture. Pf in suspended cells was (4.4 +/- 0.3) x 10(-3) cm/s (assuming a surface-to-volume ratio of 8,800 cm-1), increased more than threefold by amphotericin B, and not inhibited by phloretin and cytochalasin B under conditions of strong inhibition of glucose transport. The glucose reflection coefficient was 0.98 +/- 0.03 as measured by induced osmosis, assuming a unity reflection coefficient for sucrose. These results establish a quantitative method for measurement of osmotic water transport in adherent cultured cells and provide evidence that glucose transporters are not involved in the water transporting pathway.     

Rapid GLUT-1 mediated glucose transport in erythrocytes from the grey-headed fruit bat (Pteropus poliocephalus)

D-Glucose entry into erythrocytes from adult grey-headed flying fox fruit bats (Pteropus poliocephalus) was rapid and showed saturation at high substrate concentrations. Kinetic parameters were estimated from the concentration dependence of initial rates of zero-trans D-glucose entry at 5.5 degrees C as Michaelis constant (K(m)) 1. 64+/-0.56 mM, and maximal velocity (V(max)) 1162+/-152 micromol.l. cell water(-1).min(-1). D-Glucose entry was inhibited by cytochalasin B; mass law analysis of D-glucose-displaceable cytochalasin B binding gave values of K(d) 37.1+/-5.0 nM and B(max) 361.2+/-9.1 pmol/mg membrane protein. Entry of 2-deoxy-D-glucose, and 3-O-methyl-D-glucose, into P. poliocephalus red cells was rapid, entry of D-fructose was very slow. Glucose transporter polypeptides were identified on immunoblots as a band M(r) 47000-54000 and their identity confirmed by D-glucose-sensitive photolabeling of membranes with [3H]-cytochalasin B. Peptide-N-glycanase F digestion of both human and bat erythrocyte membranes generated GLUT-1-derived bands M(r) 37000. Trypsin digestion of human and fruit bat erythrocyte membranes generated fragmentation patterns consistent with similar GLUT-1 polypeptide structures in both species. Erythrocytes from adult Australian ghost bats (Macroderma gigas), a carnivorous microchiropteran bat, also expressed high levels of GLUT-1.     

Reconstitution studies of the human erythrocyte nucleoside transporter

The human erythrocyte nucleoside transporter has been identified as a band 4.5 polypeptide (Mr 45,000-66,000) on the basis of reversible binding and photoaffinity labeling experiments with the nucleoside transport inhibitor, nitrobenzylthioinosine (NBMPR). In the present study, the NBMPR-binding protein was extracted from protein-depleted human erythrocyte "ghosts" with Triton X-100 and reconstituted into soybean phospholipid vesicles by a freeze-thaw-sonication procedure. The reconstituted proteoliposomes exhibited nitrobenzylthioguanosine (NBTGR)-sensitive [14C]uridine transport. A partially purified preparation of the NBMPR-binding protein, consisting largely of band 4.5 polypeptides, was also shown to have nucleoside transport activity. This band 4.5 preparation exhibited a 10-fold increase in uridine transport activity and a 7-fold increase in NBMPR-binding activity relative to the crude membrane extract. Uridine transport by the reconstituted band 4.5 preparation was saturable (apparent Km = 0.21 mM; Vmax = 9 nmol/mg of protein/5 s) and was inhibited by dipyridamole, dilazep, adenosine, and inosine. The vesicles reconstituted with the band 4.5 preparation also exhibited stereospecific glucose transport which was inhibited by cytochalasin B, but unaffected by NBTGR. In contrast, cytochalasin B was a poor inhibitor of NBTGR-sensitive uridine transport. These experiments implicate band 4.5 polypeptides in both nucleoside and sugar permeation.     

TRPC6 contributes to the Ca(2+) leak of human erythrocytes.

Human erythrocytes express cation channels which contribute to the background leak of Ca(2+), Na(+) and K(+). Excessive activation of these channels upon energy depletion, osmotic shock, Cl(-) depletion, or oxidative stress triggers suicidal death of erythrocytes (eryptosis), characterized by cell-shrinkage and exposure of phosphatidylserine at the cell surface. Eryptotic cells are supposed to be cleared from circulating blood. The present study aimed to identify the cation channels. RT-PCR revealed mRNA encoding the non-selective cation channel TRPC6 in erythroid progenitor cells. Western blotting indicated expression of TRPC6 protein in erythrocytes from man and wildtype mice but not from TRPC6(-/-) mice. According to flow-cytometry, Ca(2+) entry into human ghosts prepared by hemolysis in EGTA-buffered solution containing the Ca(2+) indicator Fluo3/AM was inhibited by the reducing agent dithiothreitol and the erythrocyte cation channel blockers ethylisopropylamiloride and amiloride. Loading of the ghosts with antibodies against TRPC6 or TRPC3/6/7 but neither with antibodies against TRPM2 or TRPC3 nor antibodies pre-adsorbed with the immunizing peptides inhibited ghost Ca(2+) entry. Moreover, free Ca(2+) concentration, cell-shrinkage, and phospholipid scrambling were significantly lower in Cl(-)-depleted TRPC6(-/-) erythrocytes than in wildtype mouse erythrocytes. In conclusion, human and mouse erythrocytes express TRPC6 cation channels which participate in cation leak and Ca(2+)-induced suicidal death.