Differences in the response of several cell types to inhibition of surface receptor mobility by local concanavalina binding

Concanavalin A (conA) modulates the lateral mobility of cell surface receptors differently on different cell types. This was demonstrated by using fluorescence photobleaching recovery (FPR) to measure the inhibition of the lateral mobility of conA receptors by localized binding of conA on lymphocytes, fibroblasts, and macrophages. On mouse spleen lymphocytes, binding of conA platelets above a threshold coverage (about 12% of the upper cell-surface area) reduced the diffusion coefficient of mobile TMR-SconA-receptor complexes from 3.0×10^?10 cm²/sec to 0.6× 10^?10 cm²/sec (a 5-fold decrease), and the fraction of mobile receptors was concomitantly reduced from 0.4 to 0.11. Below the threshold occupancy, no effect on either parameter was detected. On 3T3 cells, a qualitatively similar threshold phenomenon was observed: coverage of over 9% of the upper cell surface by conA platelets induced a 3-fold reduction in the diffusion coefficient of TMR-SconA-receptor complexes from 5×10^?10 cm²/sec to 1.7× 10^?10 cm²/sec. However, no effect on the mobile fraction (about 0.4) was observed. In contrast, neither the diffusion coefficient nor the mobile fraction of TMR-SconA-receptor complexes on mouse peritoneal macrophages (both resident and thioglycolate-stimulated) or on the mouse macrophage cell line P388D₁ were affected by the binding of conA platelets in amounts covering over 50% of the upper cell surface (approx. 4.6× 10^?10 cm²/sec and 0.5 for the diffusion coefficient and mobile fraction, respectively). These differences are correlated to the different cytoskeletal functions of the various cell types studied, and are discussed regarding the mechanism of the conA-induced modulation.     

Translational diffusion in the plasma membrane of sea urchin eggs

Translational diffusion in the plasma membrane of individual egg cells from the sea urchin species Paracentrotus lividus has been studied by fluorescence microphotolysis (FM). In order to probe the lipid phase of the membrane, procedures have been worked out by which the fluorescent analog 3,3′-dioctadecyl-oxatricarbocyanine (C₁₈diO) can be incorporated into the membrane. In the unfertilized egg a fraction R = 0.9 of C₁₈diO was mobile having an apparent diffusion coefficient of D = 6.0 × 10^?9 cm² sec^?1. Fifteen to twenty-five minutes after fertilization R and D were reduced to 0.8 and 2.7 × 10^?9 cm² sec^?1, respectively. In order to study diffusion of membrane proteins, procedures have been worked out by which the cell surface can be labeled with fluorescein-isothiocyanate (FITC). FITC binds to both the plasma membrane and the vitelline layer. Together with the vitelline layer two-thirds of the FITC-fluorescence could be removed from the egg surface. Gel electropherograms of isolated egg cortices showed various protein bands; however, only two of the protein bands were labeled with FITC. In the unfertilized egg a fraction R = 0.9 of the FITC-labeled membrane proteins was mobile having an apparent diffusion coefficient of D = 35 × 10^?11 cm² sem^?1. Fiteen to twenty-five minutes after fertilization R and D were reduced to 0.8 and 7.0 × 10^?11 cm² sec^?1, respectively. FITC-labeled proteins of the fertilization envelope were immobile. Our studies have shown (i) that the egg surface can be fluorescently labeled without blocking fertilization and early development, (ii) that the plasma membrane of unfertilized eggs is a fluid environment permitting a rapid movement of lipids and proteins, and (iii) that after fertilization a substantial degree of lipid and protein mobility is maintained.     

Photobleaching recovery studies of membrane events accompanying lectin stimulation of rabbit lymphocytes.

Fluorescence photobleaching recovery methods reveal marked changes in lateral mobilities of rabbit lymphocyte membrane components during the course of stimulation with succinyl concanavalin A (S Con A). The diffusion constant of S Con A receptors on T lymphocytes falls from 1.6×10^?10 cm²/sec to 6.5×10^?11 cm²/sec within 4 hr after stimulation, remains constant for 14 hr, and returns to its former value. The mobility of B cell receptors similarly falls from 1.4×10^?10 cm²/sec to 5.5×10^?11 cm²/sec but regains its unstimulated value much more slowly. In contrast, a fluorescent phospholipid analog shows constant mobilities of 1.9×10^?8 cm²/sec and 1.5×10^?8 cm²/sec in T and B cells, respectively, throughout the experiment.     

Measurement of diffusion coefficient and electrophoretic mobility with a quasielastic light-scattering–band-electrophoresis apparatus

We have constructed an apparatus for the simultaneous measurement of electrophoretic mobility, μ, and diffusion coefficient, D, of macromolecules and cells. It combines band electrophoresis in a vertical, sucrose-gradient stabilized column, with quasielastic laser light-scattering determination of the diffusion coefficient of the species within the band. The entire electrophoresis cell is scanned through the laser beam of the quasielastic laser light-scattering apparatus by a vertical translation stage. Total intensity light-scattering measurement at each point in the cell gives the macromolecular concentration at that point. Solvent viscosity and electrical potential are measured at each point in the cell. Application of this apparatus to resealed red blood cell ghosts and to bovine hemoglobin indicates that measurements of field, viscosity, and migration distance are reliable, and that electroosmosis is insignificant. Application to T4D bacteriophage gives μ_20,w = (?1.05 ± 0.05) × 10^?4 cm²/V sec and D_20,w = (3.35 ± 0.10) × 10^?8 cm²/sec for fiberless particles, and μ_20,w = ?(0.59 ± 0.03) × 10^?4 cm²/V sec and D_20,w = (2.86 ± 0.09) × 10^?8 cm²/sec for whole phage with 6 fibers. Approximate analysis of these results with the Henry electrophoresis theory for spheres in dicates that each fiber contributes about 193 positive charges to the phage particle, compared with 327 from amino-acid analysis. The advantages and disadvantages of this apparatus, relative to conventional electrophoresis and to electrophoretic light scattering, are discussed.     

Hydrodynamics,size, and shape of bacteriophage T4D tails and baseplates

We have used translational diffusion coefficient measurements and subunit hydrodynamic theory to determine the dimensions and shape of bacterioophage T4D baseplates and tails. The diffusion coefficient of the baseplate, measured by quasielastic laser light scattering (QLS), was determined previously by Wagenknecht and Bloomfield to be D = 8.56 × 10^?8 cm²/s. For the tail, we found D = 5.88 × 10^?8 cm²/s by QLS, and D = 6.02 × 10^?8 cm²/s by combining sedimentation coefficient and molecular weight in the Svedberg equation. These values, which have an uncertainty of ±2.7%, when combined with subunit hydrodynamic theory, enabled us to refine estimates of dimensions obtained by electron microscopy. For the hexagonal baseplate, the vertex-to-vertex distance is about 480 Å, the thickness is 160 Å, and there are six extended short fibers 320-Å long and 40 Å in diameter. When a baseplate of these dimensions is attached to a tail tube-sheath-connector complex 1050-Å long and 240 Å in diameter, the calculated D is 5.93 × 10^?8 cm²/s, within 1% of experiment. This combined use of electron microscopy and hydrodynamics, using the former to ascertain shape, and the latter to obtain solution dimensions, is a powerful approach to the structure of biomolecular complexes.     

The effect of ligand affinity on integrins’ lateral diffusion in cultured cells

The role of ligand affinity in altering αPS2CβPS integrins’ lateral mobility was studied using single particle tracking (SPT) with ligand-functionalized quantum dots (QDs) and fluorescence recovery after photobleaching (FRAP) with fluorescent protein tagged integrins. Integrins are ubiquitous transmembrane proteins that are vital for numerous cellular functions, including bidirectional signaling and cell anchorage. Wild-type and high ligand affinity mutant (αPS2CβPS-V409D) integrins were studied in S2 cells. As measured by SPT, the integrin mobile fraction decreased by 22 % and had a 4× slower diffusion coefficient for αPS2CβPS-V409D compared to wild-type integrins. These differences are partially the result of αPS2CβPS-V409D integrins’ increased clustering. For the wild-type integrins, the average of all diffusion coefficients measured by SPT was statistically similar to the ensemble FRAP results. A 75 % slower average diffusion coefficient was measured by SPT compared to FRAP for αPS2CβPS-V409D integrins, and this may be the result of SPT measuring only ligand-bound integrins, in contrast all ligand-bound and ligand-unbound integrins are averaged in FRAP measurements. Specific binding of the ligand-functionalized QDs was 99 % for integrin expressing cells. The results prove that the ligand binding affinity affects the lateral dynamics of a subset of integrins based on the complementary SPT and FRAP data.     

Limited rotational diffusion of DPH in human erythrocyte membranes

The rotational diffusion of diphenylhexatriene (DPH) determines its fluorescence depolarization. Time-resolved polarization measurements were used to calculate the coefficient of diffusion of this probe in human crythrocyte ghost membranes on the basis of a diffusion theory of limited rotation. The diffusion coefficient is 5.9 × 10⁷ sec^?1 at 37°C; this was compared with the diffusion coefficient of DPH in liquid paraffin for an estimation of the microviscosity of the membrane bilayer.     

Lateral mobility of plasma membrane lipids in Xenopus eggs: Regional differences related to animal/vegetal polarity become extreme upon fertilization

Regional differences in the lateral mobility properties of plasma membrane lipids have been studied in unfertilized and fertilized Xenopus eggs by fluorescence photobleaching recovery (FPR) measurements. Out of a variety of commonly used lipid probes only the aminofluorescein-labeled fatty acids HEDAF (5-(N-hexadecanoyl)-aminofluorescein) and TEDAF (5-(N-tetradecanoyl)-aminofluorescein) appear to partition into the plasma membrane. Under all experimental conditions used these molecules show partial recovery upon photobleaching indicating the existence of lipidic microdomains. In the unfertilized egg the mobile fraction of plasma membrane lipids (～50%) has a fivefold smaller lateral diffusion coefficient (D = 1.5 × 10^?8 cm²/sec) in the animal than in the vegetal plasma membrane (D = 7.6 × 10^?8 cm²/sec). This demonstrates the presence of an animal/vegetal polarity within the Xenopus egg plasma membrane. Upon fertilization this polarity is strongly (>100×) enhanced leading to the formation of two distinct macrodomains within the plasma membrane. At the animal side of the egg lipids are completely immobilized on the time scale of FPR measurements (D ? 10^?10 cm²/sec), whereas at the vegetal side D is only slightly reduced (D = 4.4 × 10^?8 cm²/sec). The immobilization of animal plasma membrane lipids, which could play a role in the polyspermy block, probably arises by the fusion of cortical granules which are more numerous here. The transition between the animal and the vegetal domain is sharp and coincides with the boundary between the presumptive ecto- and endoderm. The role of regional differences in the plasma membrane is discussed in relation to cell diversification in early development.     

The concentration dependence of the hemoglobin mutual diffusion coefficient

Laser correlation Spectroscopy was used to measure the mutual diffusion coefficient, D, of human cyanomethemoglobin (Fe⁺⁺⁺:CN) at varying protein concentrations. These measurements were male at 20°C in a 0.1 M phosphate buffer solution at pH 7.0. For low protein concentrations we find D = (6.43 ± 0.26) × 10^?7 cm²/S and that there is a near linear decrease from this value at higher concentrations. The linear relation between the diffusion coefficient and protein concentration allows us to deduce the value of the linear frictional volume fraction coefficient, K_f= 7.75. and to extrapolate to hemoglobin concentrations equivalent to that in the red blood cell where we estimate D = 4.25 × 10^?7 cm²/s Various theoretical predictions of the dependence of the mutual diffusion coefficient on concentration are tested; we find that the generalized Stokes-Einstein relation can be made to fit our high concentration data if we assume a hard-sphere model and if we include a term involving a hydrodynamic interaction integral.     

Lateral mobility of phospholipid molecules in thin liquid films

The Fluorescence Recovery After Photobleaching (FRAP) method was applied to measure the lateral mobility of the fluorescent lipid analog, dioctadecylindocarbocyanine perchlorate (Dil-C18), in microscopic thin liquid films (Foam Films (FFs)). The foam film structures were comprised of two phosphatidylcholine monolayers adsorbed at air/water interfaces which sandwiched a thin liquid core. Lateral diffusion of the DiI molecules in the plane of the monolayers was determined as a function of the thickness of the thin liquid core of the film between the FF monolayers. The results obtained indicated that the diffusion coefficient was strongly dependent both on the distance between the FF monolayers in the range 4 nm to 85 nm (corresponding to the FF thickness) and on the film type. The applicability of the FRAP method for studying the molecular mobility in phospholipid FFs was demonstrated. Considerable differences in the surface diffusion coefficient of Dil were observed, ranging between 2 × 10^–8 cm²/s and 22 × 10^–8 cm²/s in so called yellow, gray, common black and Newton black FFs. The effect of the presence of polyethylene glycol (PEG-400) in the liquid core of lecithin FFs on surface diffusion was also studied. The surface diffusion results from the FF studies were compared with data from black lipid membranes (BLMs). These structures are related in thickness terms but the molecular orientation in FFs is the reverse of that in BLMs. Correspondence to: Z. Lalchev     

Hydrodynamic properties of mushroom tyrosinase

The hydrodynamic properties of mushroom tyrosinase were determined at pH 6.5 using a Sephadex G-200 column. From the comparison of its gel-filtration behaviour with those of standard proteins, the following parameters were calculated: MW (122 500 ± 1%), Stokes' radius (42.75 × 10^?8 cm²/sec), diffusion coefficient (5.048 × 10^?7 cm²/sec) and frictional ratio (1.26). These values suggest a globular conformation of this enzyme.     

Cholesterol modulation of nicotinic acetylcholine receptor surface mobility

Nicotinic acetylcholine receptor (AChR) function and distribution are quite sensitive to cholesterol (Chol) levels in the plasma membrane (reviewed by Barrantes in J Neurochem 103 (suppl 1):72–80, 2007). Here we combined confocal fluorescence recovery after photobleaching (FRAP) and confocal fluorescence correlation spectroscopy (FCS) to examine the mobility of the AChR and its dependence on Chol content at the cell surface of a mammalian cell line. Plasma membrane AChR exhibited limited mobility and only ~55% of the fluorescence was recovered within 10 min after photobleaching. Depletion of membrane Chol by methyl-β-cyclodextrin strongly affected the mobility of the AChR at the plasma membrane; the fraction of mobile AChR fell from 55 to 20% in Chol-depleted cells, whereas Chol enrichment by methyl-β-cyclodextrin-Chol treatment did not reduce receptor mobility at the cell surface. Actin depolymerization caused by latrunculin A partially restored receptor mobility in Chol-depleted cells. In agreement with the FRAP data, scanning FCS experiments showed that the diffusion coefficient of the AChR was about 30% lower upon Chol depletion. Taken together, these results suggest that membrane Chol modulates AChR mobility at the plasma membrane through a Chol-dependent mechanism sensitive to cortical actin.     

Physical properties of type i collagen in solution: Structure of α-Chains and β- and γ-components and two-component mixtures

The structure of thermally denatured Type I collagen has been studied using laser light scattering. The results indicate that the diffusion coefficients of α-chains and β- and γ-components are 1.550 ± 0.08 × 10^?7, 1.000 ± 0.05 × 10^?7, and 0.835 ± 0.04 × 10^?7 cm²/sec, respectively, at temperatures between 20 and 40°C. It is concluded from diffusion data that these species have hydrodynamic radii of about 13.8 nm (α-chain), 21.5 nm (β-component), and 25.7 nm (γ-component), consistent with previous studies of thermal denaturation by light scattering. It is also concluded, based on volume calculations, that a large volume increase occurs when the triple helix unfolds. Homodyne correlation functions for two component mixtures of α-chains and β-and γ-components appeared to decay exponentially. In all but one case discussed the correlation function could be fitted with a single component having a translational diffusion coefficient which was an intensity weighted average of the diffusion coefficient of each component present.     

The two-cross immunodiffusion technique: diffusion coefficients and precipitating titers of IgG in human serum and rabbit serum antibodies.

The two-cross technique, a new two-dimensional double-diffusion technique in gelplates, has been applied for simultaneous determination of precipitating titers and diffusion coeffients of antigen and antibody in body fluids. The advantage of this technique is that it works without using any standard solution and ensures conditions of “time-invariant sink”. The theory of the technique has been verified by experimental results on the precipitating system human serum-rabbit anti-human IgG in phosphate-buffered saline solution at pH 7.4. The results obtained using several modes of calculations from experimental parameters have been compared and found satisfactory. The accuracy and reproducibility of the results have been confirmed. It has been found that at 20°C the diffusion coefficient of human IgG in 10-times-diluted serum is (4.4 ± 0.2) × 10^?7 cm² s^?1, while the diffusion coefficient of rabbit anti-human IgG in a purified preparation is (2.9 ± 0.2) × 10^?7 cm² s^?1. The critical precipitating concentration of human IgG against rabbit anti-human IgG is invariable to concentration and amounts to 0.174 ± 0.03 mg/100 ml at pH 7.4.     

Oxygen Diffusion from the Roots of Rice Grown under Non-waterlogged Conditions

Three rice varieties, cv. Norin 36, cv. Norin 37 and cv. Yubae, were grown in a loam with a 20 cm water-table which gave aerobic conditions to a depth of not less than 15 to 17 cm. Under these conditions Norin 36 grew more vigorously and tillered more frequently than the other two varieties. The rates of oxygen diffusion at 23°C from roots up to 11 cm in length were however appreciably lower for Norin 36 (4.3 × 10^?8g · cm^?2 of root surface · min^?1) than for Norin 37 or Yubae (c. 7.8 × 10^?8g). A considerable increase (up to 200 %) in the oxygen diffusion rate (ODR) from the roots occurred if they were cooled to 3°C, and at this temperature differences in ODR between the varieties were not significant. For a purely physical system, because of the decrease in the diffusion coefficient of oxygen in water, and, the increase in oxygen solubility, a drop of c. 20 % in ODR should accompany the above 20°C drop in temperature. A 16 % drop was recorded for artificial ‘roots’ under these conditions. It was concluded that respiratory activity at the higher temperature must have been responsible for the low readings and intervarietal differences observed at 23°C. By increasing the 3°C values by 25 % a mean value of 14.2 × 10^?8g · cm^?2 of root surface · min^?1 was recorded for the three varieties, being the probable ODR at 23°C in the absence of a respiratory factor. Calculations show that respiratory activity removed enough oxygen to reduce the ODR for Norin 36 by more than 9 × 10^?8g, and for Norin 37 and Yubae by c. 6.7 × 10^?8g · cm^?2 of root surface · min^?1. Anatomical investigations showed that cortical breakdown was always extensive at 4 to 4.5 cm from the apex of the roots. In some cases however breakdown had not occurred in the basal segment of the root. No opinion could be formed as to whether differences in the amount of cortical breakdown between the varieties might have occasioned the respiratory differences observed. An interesting feature of the root anatomy was the failure of breakdown in those regions of the roots through which lateral roots emerged.     

Lateral diffusion and patch formation of H-2Kk antigens on mouse spleen lymphocytes

We have studied the diffusion and aggregation of H-2K^k antigens labeled with a fluorescent anti-H-2K^k monoclonal antibody (IgG) on mouse splenic lymphocytes, employing fluorescence photobleaching recovery and fluorescence microscopy. The H-2K^k antigens were initially distributed homogeneously on all lymphocytes. Upon antibody binding, sub-micron patches were formed on 50–60% of the cells. A lateral diffusion coefficient, D, of 7.1·10^?10 cm²/s and a mobile fraction of 0.73 were found for H-2K^k antigens on diffusely-labeled cells, while these antigens were immobile (D?5·10^?12 cm²/s) on patched cells. The patched and nonpatched sub-populations did not correspond to B- and T-lymphocytes. Subjection to low temperature or treatment with NaN₃ or cytoskeleton-disrupting drugs did not affect the diffusion or patching of H-2K^k, indicating no involvement of metabolic energy or drug-sensitive cytoskeletal components. These findings could be related to the interactions of H-2 antigens on the cell surface, and to the different susceptibilities of various cells to lysis by cytotoxic T-cells.     

Antigen-specific mouse lymphocyte stimulation by DNP-conjugated T-independent antigens studied by photobleaching recovery

Fluorescence photobleaching recovery techniques have allowed us to measure the lateral mobility of T-independent antigens bound to antigen-specific mouse B cells. The in vitro immunogenicity or tolerogenicity of antigens we have examined, DNP-polymerized flagellin (DNP-POL), and DNP-linear dextran (DNP-DEX), depend upon the antigen dose and epitope density. These factors also determine the mobility of antigen bound to B cell surfaces. For DNP-POL bound to DNP-specific cells, the observed diffusion constants D decrease monotonically with increasing antigen dose and epitope density. Values of D range from 10.4 × 10^?11 cm² sec^?1 for DNP_0.4-POL at 0.15 μg/ml to 0.8 × 10^?11cm² sec^?1for DNP_3.5-POL at 30 μg/ml. For receptor-bound DNP-DEX, D depends strongly on antigen epitope density but not observably on antigen concentration. For epitope densities of 1.2 or less, D is close to the value of 21 × 10^?11cm²sec^?1 observed for single slg receptors. By an epitope density of 4.8, D has fallen to 2.1 × 10^?11cm²sec^?1. Peak immunogenicities for DNP-POL and DNP-DEX arc observed when antigen- receptor aggregates have mobilities 14-fold and 3-fold lower, respectively, than a single slg molecule.     

Rates of Shrinkage and Growth of Air Bubbles Entrained in Wheat Flour Dough

The rates of shrinkage at constant temperature, and growth under a temperature rise below 100°C, of bubbles entrained in wheat flour dough were analyzed and compared with those of a bubble in water. The rate of shrinkage of bubbles in flour dough was controlled by the diffusion of dissolved air from the surface of bubbles to the bulk of flour dough. The apparent diffusion coefficient of the dissolved air in wheat flour dough with the water fraction of 0.49 calculated from the shrinkage of bubbles, was (3.2 ± 1.5) × 10^1?1 m²/sec (19°C), and (6.4 ± 2.0) × 10^?11 m²/sec (42°C). However, the growth behavior of bubbles in flour dough under a temperature rise was very different from that predicted from the diffusion theory. The critical radius of bubbles to grow was larger than that estimated from the diffusion theory. The mechanism of growth of bubbles in wheat flour dough, which was different from that of a bubble in water, is a subject that needs to be clarified.     

Synthesis,characterization, DNA interaction,and cytotoxicity of novel Pd(II) and Pt(II) complexes

Four complexes [Pd(L)(bipy)Cl]·4H₂O (1), [Pd(L)(phen)Cl]·4H₂O (2), [Pt(L)(bipy)Cl]·4H₂O (3), and [Pt(L)(phen)Cl]·4H₂O (4), where L = quinolinic acid, bipy = 2,2’-bipyridyl, and phen = 1,10-phenanthroline, have been synthesized and characterized using IR, ¹H NMR, elemental analysis, and single-crystal X-ray diffractometry. The binding of the complexes to FS-DNA was investigated by electronic absorption titration and fluorescence spectroscopy. The results indicate that the complexes bind to FS-DNA in an intercalative mode and the intrinsic binding constants K of the title complexes with FS-DNA are about 3.5?×?10⁴ M^?1, 3.9?×?10⁴ M^?1, 6.1?×?10⁴ M^?1, and 1.4?×?10⁵ M^?1, respectively. Also, the four complexes bind to DNA with different binding affinities, in descending order: complex 4, complex 3, complex 2, complex 1. Gel electrophoresis assay demonstrated the ability of the Pt(II) complexes to cleave pBR322 plasmid DNA.     

Li+ counterion self-diffusion in ordered DNA

The anisotropic self-diffusion coefficient of ⁷Li⁺ (I = 3/2) counterions has been studied in hydrated, macroscopically oriented Li-(B)DNA fibers at relatively high water contents, corresponding to approximate DNA-DNA helix axis distances of 22–35 Å, using the pulsed field gradient hmr spin-echo method. Self-diffusion coefficients parallel (D_∥) and perpendicular (D_?) to the DNA helix axis increase with increasing salt content and with increasing DNA-DNA helix axis distance. The observed anisotropy D_∥/D_? decreases from 1.6 to 1.2 with the DNA-DNA separation increasing from 22 to 35 Å in the salt-free sample. This result can be understood by the obstruction effect caused by the DNA molecules themselves. The values of the Li⁺ self-diffusion coefficients in the most water-rich system with no added salt (corresponding to an approximate distance of 35 Å between the DNA helix axes) were D_∥ ～ 1.15 × 10^?10 m² s^?1 and D_? ～ 0.98 × 10^?10 m² s^?1, compared to 9.14 × 10^?10 m² s^?1 for the diffusion of Li⁺ in an aqueous solution of LiCl (～ 2.1M). The possible occurrence of restriction effects in the DNA fibers have also been studied by determining the self-diffusion coefficient at different effective diffusion times. The self-diffusion coefficient of Li⁺ in the sample with the largest DNA-DNA helix axis distance seems to be independent of the effective diffusion time, which indicates that the lithium ions are not trapped within impermeable barriers. The possibility of diffusion through permeable barriers has also been investigated, and is discussed. © 1994 John Wiley & Sons, Inc.