Lateral diffusion of lipids and glycophorin in solid phosphatidylcholine bilayers. The role of structural defects

The lateral mobility of the lipid analog N-4-nitrobenzo-2-oxa-1,3 diazole phosphatidylethanolamine and of the integral protein glycophorin in giant dimyristoylphosphatidylcholine vesicles was studied by the photobleaching technique. Above the temperature of the chain-melting transition (Tm = 23 degrees C), the diffusion coefficient, Dp, of the protein [Dp = (4 +/- 2) X 10(-8) cm2/s at 30 degrees C] was within the experimental errors equal to the corresponding values DL of the lipid analog. In the P beta 1 phase the diffusion of lipid and glycophorin was studied as a function of the probe and the protein concentration. (a) At low lipid-probe content (cL less than 5 mmol/mol of total lipid), approximately 20% of the probe diffuses fast (D approximately equal to 10(-8) - 10(-9) cm2/s), while the mobility of the rest is strongly reduced (D less than 10(-10) cm2/s). At a higher concentration (cp approximately 20 mmol), all probe is immobilized (D less than 10(-10) cm2/s). (b) Incorporation of glycophorin up to cp = 0.4 mmol/mol of total lipid leads to a gradual increase of the fraction of mobile lipid probe due to the lateral-phase separation into a pure P beta 1 phase and a fraction of lipid that is fluidized by strong hydrophilic lipid-protein interaction. (c) The diffusion of the glycophorin molecules is characterized by a slow and a fast fraction. The latter increases with increasing protein content, which is again due to the lateral-phase separation caused by the hydrophilic lipid-protein interaction. The results are interpreted in terms of a fast transport along linear defects in the P beta 1 phase, which form quasi-fluid paths for a nearly one dimensional and thus very effective transport. Evidence for this interpretation of the diffusion measurements is provided by freeze-fracture electron microscopy.     

Lateral diffusion of gramicidin C in phospholipid multibilayers. Effects of cholesterol and high gramicidin concentration

We have measured the lateral diffusion coefficient (D), of active dansyl-labeled gramicidin C (DGC), using the technique of fluorescence photobleaching recovery, under conditions in which the cylindrical dimer channel of DGC predominates. In pure, hydrated, dimyristoylphosphatidylcholine (DMPC) multibilayers (MBL), D decreases from 6 X 10(-8) cm2/s at 40 degrees C to 3 X 10(-8) cm2/s at 25 degrees C, and drops 100-fold at 23 degrees C, the phase transition temperature (Tm) of DMPC. Above Tm, addition of cholesterol decreases D; a threefold stepwise drop occurs between 10 and 20 mol %. Below Tm, increasing cholesterol increases D; a 10-fold increase occurs between 10 and 20 mol % at 21 degrees C, between 20 and 25 mol % at 15 degrees C, and between 25 and 30 mol % at 5 degrees C. In egg phosphatidylcholine (EPC) MBL, D decreases linearly from 5 X 10(-8) cm2/s at 35 degrees C to 2 X 10(-8) cm2/s at 5 degrees C; addition of equimolar cholesterol reduces D by a factor of 2. Thus this transmembrane polypeptide at low membrane concentrations diffuses quite like a lipid molecule. Its diffusivity in lipid mixtures appears to reflect predicted changes of lateral composition. Increasing gramicidin C (GC) in DMPC/GC MBL broadened the phase transition, and the diffusion coefficient of the lipid probe N-4-nitrobenzo-2-diazole phosphatidylethanolamine (NBD-PE) at 30 degrees C decreases from 8 X 10(-8) cm2/s below 5 mol % GC to 2 X 10(-8) cm2/s at 14 mol % GC; D for DGC similarly decreases from 4 X 10(-8) cm2/s at 2 mol % GC to 1.4 X 10(-8) cm2/s at 14 mol % GC. Hence, above Tm, high concentrations of this polypeptide restrict the lateral mobility of membrane components.     

Lipid diffusibility in the intact erythrocyte membrane

The lateral diffusion of fluorescent lipid analogues in the plasma membrane of intact erythrocytes from man, mouse, rabbit, and frog has been measured by fluorescence photobleaching recovery (FPR). Intact cells from dystrophic, normoblastic, hemolytic, and spherocytotic mouse mutants; from hypercholesterolemic rabbits and humans; and from prenatal, neonatal, and juvenile mice have been compared with corresponding normals. The lateral diffusion coefficient (D) for 3,3'-dioctadecylindodicarbocyanine iodide (DiI[5]) in intact normal human erythrocytes is D = 8.2 +/- 1.2 X 10(-9) cm2/s at 25 degrees C and D = 2.1 +/- 0.4 X 10(-8) cm2/s at 37 degrees C, and varies approximately 50-fold between 1 degree and 42 degrees C. The diffusion constants of lipid analogue rhodamine-B phosphatidylethanolamine (RBPE) are about twice those of DiI[5]. The temperature dependence and magnitude of D vary by up to a factor of 3 between species and are only influenced by donor age in prenatals. DiI[5] diffusibility is not perturbed by the presence of calcium or local anesthetics or by spectrin depletion (via mutation). However, lipid-analogue diffusibility in erythrocyte ghosts may differ from intact cells. Dietary hypercholesterolemia in rabbits reduces the diffusion coefficient and eliminates the characteristic break in Arrhenius plots of D found in all other cells studied except frog.     

Lateral diffusion of epidermal growth factor complexed to its surface receptors does not account for the thermal sensitivity of patch formation and endocytosis

The patching and endocytosis of EGF (epidermal growth factor) bound to A-431 cells (a human epidermoid carcinoma line) are temperature-sensitive processes which are completely inhibited at 4 degrees C. Receptor-mediated endocytosis generally occurs through coated regions, and EGF bound to its membrane receptor must diffuse laterally to these points of internalization. In this work we investigated the thermal sensitivity of the lateral diffusion of EGF receptor complexes and the thermal sensitivity of the patching and endocytosis of the hormone receptor complexes. Using the fluorescence photobleach recovery technique, we measured the lateral diffusion coefficients of a fluorescent derivative of EGF as a function of temperature. The lateral diffusion coefficient (D) increased gradually from 2.8 X 10(-10) cm2/s at 5 degrees C to 8.5 X 10(-10) cm2/s at 37 degrees C, and no phase transition was detected. Neither was a phase transition detected when we measured the diffusion coefficient of fluorescent lipid probes over this temperature range. From a calculation of the collision frequency of the occupied EGF receptors with coated regions using our measured values of D at 5 and 37 degrees C, we conclude that diffusion is not the rate-limiting step for either endocytosis or patching.     

Detection and characterization of actin monomers, oligomers, and filaments in solution by measurement of fluorescence photobleaching recovery

Fluorescence photobleaching recovery (FPR) was measured to determine the diffusion coefficient of fluorescein-labeled G-actin in low-salt buffer. The result obtained, 7.15 +/- 0.35 X 10(-7) cm2/s, is in good agreement with that computed from the molecular weight, partial specific volume, and sedimentation coefficient, but is higher than previously obtained values. It is demonstrated from theory that at low ionic strength, the electrostatic contribution to the intrinsic viscosity leads to an overestimate of the hydrodynamic eccentricity of G-actin. Data from FPR, sedimentation, and fluorescence polarization experiments all indicate that the true low-salt form of the actin monomer has an axial ratio less than or equal to 3.0. The G-F transformation of actin was also observed by measurement of FPR during the assembly phase, in the steady state, and in the presence of ligands such as cytochalasin and aldolase. Each FPR record in general yields three data: relative proportion of rapidly and slowly diffusing actin, diffusion coefficient for the high-mobility fraction, and a mean diffusion coefficient for the low-mobility fraction. A relation between the mean low-mobility diffusion coefficient and the number-average filament length is derived and applied to the analysis of FPR data. Under typical conditions, the average filament length was much greater than 10 micron in the steady state. Cytochalasin D was found to decrease filament length and total amount of filament proportionally; total filament number was not greatly affected. In all polymerizations of G-actin, the high-mobility material observed in situ was found to be essentially monomeric actin. Relatively stable oligomers of actin were separated by fractionating G-AF-actin by gel filtration in 50 microM MgCl2 at 4 degrees C. On the basis of the diffusion coefficient, we conclude that monomer and dimer constitute the major particle types present under these conditions. Sedimentation of labeled actin polymerized in 1.0 mM MgCl2 yielded a graded supernatant that contained actin oligomers significantly larger than the monomer.     

Characterization of heterogeneous solutions using laser light scattering: study of the tubulin system

Laser light scattering has been applied to a systematic study of a heterogeneous solution of tubulin at low temperature--conditions under which tubulin assembly into microtubules does not take place. Methods of analyzing laser light scattering results obtained from solutions containing multiple components are discussed. Data analysis techniques are described and their application to the determination of diffusion constants from experimental data is extensively illustrated. Multiple components were found under the conditions that the tubulin was studied. We have identified one component having D20,w = 4.41 X 10(-7) cm2/s (sigma = 0.54 X 10(-7) cm2/s) which has the expected value for tubulin dimer. In addition, we have found two components which are significantly larger than tubulin. One large component has D 20,w approximately 0.55 X 10(-7) cm2/s and is present in all samples at 4 degrees C even after centrifugation to remove components greater than 10 S. Another large component having 3.2 X 10(-7) cm2/s greater than or equal to D20,w greater than or equal to 1.5 X 10(-7) cm2/s has been found to sediment with 10 S less than or equal to s less than 20 S.     

Rapid lateral diffusion of the variant surface glycoprotein in the coat of Trypanosoma brucei

The membrane form of the variant surface glycoprotein (mfVSG) is anchored in the plasma membrane of Trypanosoma brucei by a dimyristoylphosphatidylinositol residue connected via a glycan to the COOH-terminal amino acid. The glycoprotein molecules are tightly packed, forming a coat that is impenetrable to lytic serum components. Lateral diffusion of mfVSG was measured by the fluorescence recovery after photobleaching technique. mfVSG labeled on the cell surface with rhodamine-conjugated anti-VSG Fab fragments showed a diffusion coefficient of 1 X 10(-10) cm2/s at 37 degrees C and of 0.7 X 10(-10) cm2/s at 27 degrees C. About 80% of the molecules were mobile. Affinity-purified mfVSG molecules implanted into the plasma membrane of baby hamster kidney cells exhibited a similar mobility to that found in the trypanosome coat [D = (0.4-0.7) X 10(-10) cm2/s at 4 degrees C]. Phospholipid mobility in the plasma membrane of trypanosomes was characterized by a diffusion coefficient of 2.2 X 10(-9) cm2/s at 37 degrees C. It is concluded that mfVSG mobility in the surface coat of the parasite is rapid and comparable to that of other membrane-bound glycoproteins but slower than that of phospholipids.     

Dynamic light scattering investigations of human erythrocyte spectrin.

Dynamic light scattering measurements were performed on spectrin from human erythrocytes in 25 mM Tris buffer at pH 7.6 with 100 mM NaCl and 5 mM EDTA. Measurements were made on spectrin solutions prepared as dimers and tetramers over the temperature range from 23 to 41 degrees C, as a function of the square of the scattering vector (K2) over the range of 0.7 x 10(10) cm-2 less than or equal to K1 less than or equal to 20 x 10(10) cm-2. Analysis of the autocorrelation functions collected for these solutions revealed the presence of two predominant motional components over the entire range of K2. Plots of the diffusion coefficients (D20) of these components, with viscosity and temperature corrected to water at 20 degrees C, as a function of K2 indicated three rather distinct regions, flat regions at low and high K2 joined by a sloping intermediate region. At small K2 (less than or equal to 4 x 10(10) cm-2) the D20 values were (7.3 +/- 2.0) x 10(-8) cm2/s for the slow component and (20.3 +/- 2.0) x 10(-8) cm2/s for the fast component. At large K2 (greater than or equal to 10 x 10(10) cm-2) the values increased to (13.0 +/- 2.0) x 10(-8) cm2/s for the slow component and (39.4 +/- 2.0) x 10(-8) cm2/s for the fast component. In the intermediate K2 region, D20 is a linear function of K2 and appears as a transition between the low and high K2 regions.(ABSTRACT TRUNCATED AT 250 WORDS)     

Regional differences in the lateral mobility of plasma membrane lipids in a molluscan embryo

Regional and temporal differences in plasma membrane lipid mobility have been analyzed during the first three cleavage cycles of the embryo of the polar-lobe-forming mollusc Nassarius reticulatus by the fluorescence photobleaching recovery (FPR) method, using 1,1'-ditetradecyl 3,3,3',3'-tetramethylindocarbocyanine iodide (C14diI) as a fluorescent lipid probe. During this period of development the lateral diffusion coefficient of membrane lipids is consistently greater in the vegetal polar lobe area as compared to the animal plasma membrane area (on average 30%), demonstrating the existence of an animal-vegetal polarity in plasma membrane properties. At third cleavage, the differences between animal and vegetal plasma membrane region become even more pronounced; in the four animal micromeres the diffusion coefficient (D) and mobile fraction (MF) are 2.9 +/- 0.2 X 10(-9) cm2/sec and 51 +/- 2%, respectively, while in the four vegetal macromeres D = 5.0 +/- 0.3 X 10(-9) cm2/sec and MF = 78 +/- 2%. Superimposed upon the observed animal-vegetal polarity, the lateral diffusion in the polar lobe membrane area shows a cell-cycle-dependent modulation. The highest mean values for D are reached during the S phase (ranging from 7.0 to 7.8 X 10(-9) cm2/sec in the three cycles measured), while at the end of G2 phase and during early mitosis mean values for D have decreased significantly (ranging from 5.0 to 5.9 X 10(-9) cm2/sec). Diffusion rates in the animal membranes of the embryo are constant during the three successive cell cycles (D = 4.3-5.0 X 10(-9) cm2/sec), except for a peak at the S phase of the first cell cycle (D = 6.0 X 10(-9) cm2/sec). These results are discussed in relation with previously observed ultrastructural heterogeneities in the Nassarius egg plasma membrane. It is speculated that the observed animal-vegetal polarity in the organization of the egg membrane might play an important role in the process of cell diversification during early development.     

Rapid diffusion of spectrin bound to a lipid surface.

Human erythrocyte spectrin was labelled with the probe 5, 5'-disulfato-1-(6-hexanoic acid N-hydroxysuccinimide ester)-1'-ethyl-3,3,3',3'-tetramethylindocarbocyanine (Cy3). Cy3-spectrin was bound to the outer surface of dimyristoylphosphatidylcholine (DMPC) multilamellar vesicles and its diffusion measured by fluorescence recovery after photobleaching (FRAP). It was found that at 30 degrees C, above the lipid gel to liquid-crystalline phase transition of the lipids, Cy3-spectrin had an unexpectedly high diffusion coefficient D=(2.1+/-0.6)x10(-7)) cm2/s. At the phase transition, diffusion of Cy3-spectrin was only slightly lower; D=(1.3+/-0.3)x10(-7) cm2/s, whereas at 14 degrees C, well below the lipid phase transition, diffusion was found to be much slower with D=(3.1+/-0.12)x10(-9) cm2/s. The fast diffusion of Cy3-spectrin on the lipid surface implies that the individual bonds which bind spectrin to the lipid surface must rapidly be made and broken. In the light of these results, spectrin-lipid interactions alone appear unlikely to have any significant role in supporting the cell membrane. Probably, the interactions serve only to localise the spectrin at the inner lipid surface in order to facilitate formation of the cytoskeleton.     

Lateral diffusion of lipid probes in the surface membrane of human platelets. An electron-electron double resonance (ELDOR) study.

Electron-electron double resonance (ELDOR) techniques employing [14N], [15N] 16-Doxylstearate spin-label pairs have been used to measure the lateral diffusion constant, D, of lipids in the surface membrane of intact human blood platelets. For freshly prepared platelets, D is 1.0 X 10(-8) cm2/s at 37 degrees C and for platelets stored for 3 d at room temperature under accepted routine blood bank conditions, D is 2.6 X 10(-8) cm2/s at 37 degrees C. This is the first time that D in the surface membrane of platelets is reported. The marked increase in D for stored platelets may be attributed at least partly to loss of cholesterol during storage, suggesting a correlation between lipid lateral diffusion and cholesterol levels in cell membranes.     

Physical characteristics of 16 S rRNA under reconstitution conditions

The hydrodynamic shape and conformation of the 16 S ribosomal RNA in reconstitution buffer at both 4 degrees C and 37 degrees C were determined and compared with the corresponding properties of the 30 S ribosomal subunit at 4 degrees C in order to understand the role of the RNA molecule in the assembly of the 30 S subunit. At 4 degrees C, the 16 S rRNA has a sedimentation coefficient s020,w of 21.0 S, a diffusion coefficient D020,w of 1.72 X 10(-7) cm2/s, a frictional coefficient f/fmin of 2.37, and a hydrodynamic radius of 125 A. At 37 degrees C, the 16 S rRNA has a sedimentation coefficient s020,w of 18.4 S, a diffusion coefficient D020,w of 1.39 X 10(-7) cm2/s, a frictional coefficient f/fmin of 2.91, and a hydrodynamic radius of 153 A. At 4 degrees C, the 30 S subunit has a sedimentation coefficient s020,w of 31.8 S, a diffusion coefficient D020,w of 1.97 X 10(-7) cm2/s, a frictional coefficient f/fmin of 1.77, and a hydrodynamic radius of 109 A. These results suggested that the free RNA in solution at 4 degrees C is less folded than the RNA in the ribosomal subunit. At 37 degrees C, the free 16 S rRNA is unfolded when compared to the structure of the same RNA at 4 degrees C. This implies that the folding step accompanying the RI to RI transformation in the assembly process needs the presence of both the RNA and core proteins.     

Surface diffusion in human serum lipoproteins

From the viscosity dependence of the 31P NMR signals, the diffusion coefficients DT of phospholipid molecules in the surface monolayer of HDL, LDL and VLDL have been determined. DT for HDL3 and HDL2 are found to be 2.3 X 10(-8) cm2/s and 1.8 X 10(-8) cm2/s, respectively. These values are similar to values reported for diffusion of phospholipid molecules in phospholipid bilayers above the gel to liquid crystalline phase transition temperature. Viscosity dependence of [16,16,16-2H3]phosphatidylcholine incorporated into HDL2 yielded a value similar to that determined by 31P (DT = 1.9 X 10(-8) cm2/s). Slower diffusion coefficients were measured for LDL2 and VLDL. VLDL had a value DT = 9.1 X 10(-9) cm2/s. The diffusion coefficient for LDL2 was 1.4 X 10(-9) cm2/s. Thus, diffusion of phospholipids in LDL2 is a full order of magnitude slower at 25 degrees C than diffusion of phospholipids in the HDLs.     

Oxygen diffusion in biological and artificial membranes determined by the fluorochrome pyrene

Quenching of pyrene fluorescence by oxygen was used to determine oxygen diffusion coefficients in phospholipid dispersions and erythrocyte plasma membranes. The fluorescence intensity and lifetime of pyrene in both artificial and natural membranes decreases about 80% in the presence of 1 atm O2, while the fluorescence excitation and emission spectra and the absorption spectrum are unaltered. Assuming the oxygen partition coefficient between membrane and aqueous phase to be 4.4, the diffusion coefficients for oxygen at 37 degrees C are 1.51 X 10(-5) cm2/s in dimyristoyl lecithin vesicles, 9.32 X 10(-6) cm2/s in dipalmitoyl lecithin vesicles, and 7.27 X 10(-6) cm2/s in erythrocyte plasma membranes. The heats of activation for oxygen diffusion are low (less than 3 kcal/degree-mol). A dramatic increase in the diffusion constant occurs at the phase transition of dimyristoyl and dipalmitoyl lecithin, which may result from an increase in either the oxygen diffusion coefficient, partition coefficient, or both. The significance of the change in oxygen diffusion below and above the phase transition for biological membranes is discussed.     

Lateral mobility of a lipid analog in the membrane of irreversible sickle erythrocytes

The major feature of sickle cell anemia is the tendency of erythrocytes to sickle when exposed to decreased oxygen tension and to unsickle when reoxygenated. Irreversible sickle cells (ISCs) are sickle erythrocytes which retain bipolar elongated shapes despite reoxygenation. ISCs are believed to owe their biophysical abnormalities to acquired membrane alterations which decrease membrane deformability. While increased membrane surface viscosity has been measured in ISCs, the lateral dynamics of membrane lipids in these cells have not heretofore been examined. We have measured the lateral diffusion of the lipid analog 3,3'-dioctadecylindocyanine iodide (DiI) in the plasma membrane of intact normal erythrocytes, reversible sickle cells (RSCs), and irreversible sickle cells by fluorescence photobleaching recovery (FPR). The diffusion coefficients +/- standard errors of the mean of DiI in intact normal red blood cells (RBCs), RSCs, and ISCs at 37 degrees C are (8.06 +/- 0.29) X 10(-9) cm2 X s-1, (7.74 +/- 0.22) X 10(-9) cm2 X s-1, and (7.29 +/- 0.24) X 10(-9) cm2 X s-1, respectively. A similar decrease in the diffusion coefficient of DiI in the plasma membranes of the three cell types was observed at 4, 10, 17, 23, and 30 degrees C. ANOVA analysis of the changes in DiI diffusion showed significant differences between the RBC and ISC membranes at all temperatures examined. The characteristic breaks in Arrhenius plots of the diffusion coefficients for the RBCs, RSCs, and ISCs occurred at 20, 19, and 18.6 degrees C, respectively. Photobleaching recovery data were used to estimate (Boullier, J.A., Melnykovich, G. and Barisas, B.G. (1982) Biochim. Biophys. Acta 692, 278-286) the microviscosities of the plasma membranes of the three cell types at 25 degrees C. We find significant differences between our microviscosity values and those obtained in previous fluorescence depolarization studies. However, both methods indicate qualitatively similar differences in membrane microviscosity among the various cell types.     

Lateral diffusivity of lipid analogue excimeric probes in dimyristoylphosphatidylcholine bilayers.

The lateral mobility of pyrenyl phospholipid probes in dimyristoylphosphatidylcholine (DMPC) vesicles was determined from the dependence of the pyrene monomeric and excimeric fluorescence yields on the molar probe ratio. The analysis of the experimental data makes use of the milling crowd model for two-dimensional diffusivity and the computer simulated random walks of probes in an array of lipids. The fluorescence yields for 1-palmitoyl-2-(1'-pyrenedecanoyl)phosphatidylcholine (py10PC) in DMPC bilayers are well fitted by the model both below and above the fluid-gel phase transition temperature (Tc) and permit the evaluation of the probe diffusion rate (f), which is the frequency with which probes take random steps of length L, the host membrane lipid-lipid spacing. The lateral diffusion coefficient is then obtained from the relationship D = fL2/4. In passing through the fluid-gel phase transition of DMPC (Tc = 24 degrees C), the lateral mobility of py10PC determined in this way decrease only moderately, while D measured by fluorescence photobleaching recovery (FPR) experiments is lowered by two or more orders of magnitude in gel phase. This difference in gel phase diffusivities is discussed and considered to be related either to (a) the diffusion length in FPR experiments being about a micrometer or over 100 times greater than that of excimeric probes (approximately 1 nm), or (b) to nonrandomicity in the distribution of the pyrenyl probes in gel phase DMPC. At 35 degrees C, in fluid DMPC vesicles, the diffusion rate is f = 1.8 x 10(8) s-1, corresponding to D = 29 microns2 s-1, which is about three times larger than the value obtained in FPR experiments. The activation energy for lateral diffusion in fluid DMPC was determined to be 8.0 kcal/mol.     

Urea permeability of human red cells

The rate of unidirectional [14C]urea efflux from human red cells was determined in the self-exchange and net efflux modes with the continuous flow tube method. Self-exchange flux was saturable and followed simple Michaelis-Menten kinetics. At 38 degrees C the maximal self-exchange flux was 1.3 X 10(-7) mol cm-2 s-1, and the urea concentration for half-maximal flux, K1/2, was 396 mM. At 25 degrees C the maximal self-exchange flux decreased to 8.2 X 10(-8) mol cm-2 s-1, and K1/2 to 334 mM. The concentration-dependent urea permeability coefficient was 3 X 10(-4) cm s-1 at 1 mM and 8 X 10(-5) cm s-1 at 800 mM (25 degrees C). The latter value is consonant with previous volumetric determinations of urea permeability. Urea transport was inhibited competitively by thiourea; the half-inhibition constant, Ki, was 17 mM at 38 degrees C and 13 mM at 25 degrees C. Treatment with 1 mM p-chloromercuribenzosulfonate inhibited urea permeability by 92%. Phloretin reduced urea permeability further (greater than 97%) to a "ground" permeability of approximately 10(-6) cm s-1 (25 degrees C). This residual permeability is probably due to urea permeating the hydrophobic core of the membrane by simple diffusion. The apparent activation energy, EA, of urea transport after maximal inhibition was 59 kJ mol-1, whereas in control cells EA was 34 kJ mol-1 at 1 M and 12 kJ mol-1 at 1 mM urea. In net efflux experiments with no extracellular urea, the permeability coefficient remained constantly high, independent of a variation of intracellular urea between 1 and 500 mM, which indicates that the urea transport system is asymmetric. It is concluded that urea permeability above the ground permeability is due to facilitate diffusion and not to diffusion through nonspecific leak pathways as suggested previously.     

Electric field-induced redistribution and postfield relaxation of low density lipoprotein receptors on cultured human fibroblasts

The lateral mobility of unliganded low density lipoprotein-receptor (LDL-R) on the surface of human fibroblasts has been investigated by studying the generation and relaxation of concentration differences induced by exposure of the cultured cells to steady electric fields. The topographic distribution of receptors was determined by fluorescence microscopy of cells labeled with the intensely fluorescent, biologically active LDL derivative dioctadecylindolcarbocyanine LDL (dil(3)-LDL), or with native LDL and anti-LDL indirect immunofluorescence. Exposure of the LDL-receptor-internalization defective J. D. cells (GM2408A) to an electric field of 10 V/cm for 1 h at 22 degrees C causes greater than 80% of the cells to have an asymmetric distribution of LDL-R; receptors accumulate at the more negative pole of the cell. In contrast, only 20% of LDL-internalization normal GM3348 cells exposed to identical conditions have asymmetrical distributions. Phase micrographs taken during electric-field exposure rule out cell movement as the responsible mechanism for the effect. In both cell types, postfield labeling with the F-actin-specific fluorescent probe nitrobenzoxadiazole-phallacidin shows that no topographic alteration of the actin cytoskeleton accompanies the redistribution of cell surface LDL-Rs, and indirect immunofluorescence labeling of the coat protein clathrin shows that coated pits do not redistribute asymmetrically. Measurements of the postfield relaxation in the percentage of GM2408A cells showing an asymmetric distribution allow an estimate of the effective postfield diffusion coefficient of the unliganded LDL-R. At 37 degrees C, D = 2.0 X 10(-9) cm2/s, decreasing to 1.1 X 10(-9) cm2/s at 22 degrees C, and D = 3.5 X 10(-10) cm2/s at 10 degrees C. These values are substantially larger than those measured by photobleaching methods for the LDL-R complexed with dil(3)-LDL on intact cells, but are comparable to those measured on membrane blebs, and are consistent with diffusion coefficients measured for other unliganded integral membrane receptor proteins by postfield-relaxation methods.     

Diffusion-limited forward rate constants in two dimensions. Application to the trapping of cell surface receptors by coated pits.

A variety of receptors are known to aggregate in specialized cell surface structures called coated pits, prior to being internalized when the coated pits close off. At 37 degrees C on human fibroblasts, as well as on other cell types, a recycling process maintains a constant number of coated pits on the cell surface. In this paper, we explore implications for receptor aggregation and internalization of the two types of recycling models that have been proposed for the maintenance of the coated pit concentration. In one model, coated pits alternate between accessible and inaccessible states at fixed locations on the cell surface, while in the other model, coated pits recycle to random locations on the cell surface. We consider receptors that are randomly inserted in the membrane, move by pure diffusion with diffusion coefficient D, and are instantly and irreversibly trapped when they reach a coated pit boundary (the diffusion limit). For such receptors, we calculate for each of the two models: the mean time tau to reach a coated pit, the forward rate constant k+ for the interaction of a receptor with a coated pit, and the fraction phi of receptors aggregated in coated pits. We show that for the parameters that characterize coated pits on human fibroblasts, the way in which coated pits return to the surface has a negligible effect on the values of tau, k+, and phi for mobile receptors, D greater than or equal to 1.0 X 10(-11) cm2/s, but has a substantial effect for "immobile" receptors, D much less than 1 X 10(-11) cm2/s. We present numerical examples to show that it may be possible to distinguish between these models if one can monitor slowly diffusing receptors (D less than 1 X 10(-11) cm2/s) on cells whose coated pits have relatively short lifetimes (less than or equal to 1 min). Finally, we show that for the low-density lipoprotein (LDL) receptor on human fibroblasts (D = 4.5 X 10(-11) cm2/s), the predicted and observed values of K+ and phi are in close agreement. Therefore, even for slowly diffusing LDL receptor, unaided diffusion as the transport mechanism of receptors to coated pits is consistent with measured rates of LDL internalization.     

Cellular effects of endotoxin in vitro: mobility of endotoxin in the plasma membrane of hepatocytes and neuroblastoma cells

Lipopolysaccharide labeled with fluorescein isothiocyanate (FITC-LPS) was used to examine interactions between endotoxin and plasma membrane in isolated rat hepatocytes and mouse neuroblastoma NB41A3 cells. At the same endotoxin to cell ratio, hepatocytes bound more toxin than did neuroblastoma cells. At a dose of 12 micrograms/mg dry wt, a bound mobile fraction of between 60 and 75% of FITC-LPS was found on hepatocytes at 25 degrees C with a lateral diffusion coefficient (D) of 4.0 X 10(-9) cm2/s. In neuroblastoma cells, the mobile fraction was larger (85-90%), with D 1.0 X 10(-8) cm2/s. D was temperature-dependent between 10 and 37 degrees C and increased from 1.8 X 10(-9) to 1.0 X 10(-8) cm2/s in hepatocytes and from 9.4 X 10(-9) to 1.9 X 10(-8) cm2/s in neuroblastoma cells. In both types of cell, nonviable (cells which did not exclude Trypan blue) as compared to viable cells showed different recovery patterns and 100% of the probe molecules were mobile. These results suggest that: (1) endotoxin binding to mammalian cells consists of two subpopulations with different mobilities; (2) binding of the immobile fraction is dependent on cellular integrity; and (3) the differences in binding, lateral mobility, and size of the immobile fraction in hepatocytes and neuroblastoma cells may be due to variations in membrane composition and/or number of binding sites.