The sieving of spheres during agarose gel electrophoresis: quantitation and modeling

By use of agarose gel electrophoresis, the sieving of spherical particles in agarose gels has been quantitated and modeled for spheres with a radius (R) between 13.3 and 149 nm. For quantitation, the electrophoretic mobility has been determined as a function of agarose percentage (A). Because a previously used model of sieving [D. Rodbard and A. Chrambach (1970) Proc. Natl. Acad. Sci. USA 65, 970-977] was found incompatible with some of these data, alternative models have been tested. By use of an underivatized agarose, two models, both based on the assumption of a single effective pore radius (PE) for each A, were found to yield PE values that were independent of R and that were in agreement with values of PE obtained independently (PE = 118 nm X A-0.74): sieving by altered hydrodynamics in a cylindrical tube of radius, PE, and sieving by steric exclusion from a circular hole of radius, PE. The same analysis applied to a 6.5% hydroxyethylated commercial agarose yielded a steeper PE vs A plot and also agreement of the above two models with the data. The PE vs A plot was significantly altered by both further hydroxyethylation and factors that cause variation in the electro-osmosis found in commercial agarose.     

Effect of 100% O2 on passage of uncharged dextrans from blood to lung lymph

Since charge as well as size may influence the passage of plasma proteins from blood to lung lymph, we used uncharged dextrans as tracers to study the effects of hyperoxic lung injury on the molecular sieving properties of the pulmonary microcirculation in unanesthetized sheep. Polydisperse [3H]dextran was infused intravenously into five sheep before and after the animals breathed 100% O2 until lymph flow increased threefold (66-84 h). Lymph-to-plasma concentration ratios (L/P) were determined for [3H]dextran fractions of graded molecular sizes (1.6-8.4 nm effective radius) from samples obtained during the infusions. Before hyperoxia the blood-lymph barrier was highly restrictive to transport of [3H]dextrans above 5.0 nm in radius; steady-state L/P for these molecules averaged 0.03 or less. After the sheep breathed 100% O2, [3H]dextrans as large as 8.4 nm radius appeared in the lymph. Posthyperoxia, the L/P were significantly increased relative to prehyperoxia base-line values for every [3H]dextran fraction larger than 2.0 nm radius (P less than 0.05). In contrast, neither the L/P for albumin or total protein changed significantly. At autopsy, electron microscopy showed widespread damage to the endothelium of the alveolar capillaries with infrequent gaps between endothelial cells. In two control sheep, inhalation of compressed air for 96 h had no effect on lymph flow or L/P for the [3H]dextrans. We conclude that O2 poisoning reduced the selective sieving of uncharged dextrans across the blood-lymph barrier of the lungs and allowed larger dextrans to enter the lymph. These larger molecules may have leaked from the pulmonary microcirculation via disruptions in the continuity of the endothelial lining.     

Passage of uncharged dextrans from blood to lung lymph in awake sheep

To examine how molecular size alone influences the passage of macromolecules from the pulmonary microcirculation into lymph collected from the caudal mediastinal lymph node of the sheep, we infused polydisperse uncharged [3H]dextrans intravenously at a constant rate over a period of 7.5 h in nine awake sheep with lung lymph fistulas. Lymph and plasma were collected during hours 5.5-7.5 of the infusions, and the [3H]dextrans were separated by molecular sieve chromatography into fractions that ranged from 1.6 to 8.4 nm in effective molecular (Stokes-Einstein) radius. Lymph-to-plasma (L/P) ratios for [3H]dextrans were near 1.0 at 1.6-nm radius, decreased with increasing molecular size, and approached zero at radii above 5.0 nm. We confirmed that these L/P ratios represented steady-state values by extending the duration of the infusion to approximately 30 h in two of the nine sheep and finding that the L/P ratios remained unchanged. These results were consistent with molecular sieving through a homoporous membrane with cylindrical pores of 5.0-nm radius. We also found that the L/P ratio for albumin [0.76 +/- 0.13 (SE)] in five of the same sheep was much higher than that for the [3H]dextran fraction of the same effective molecular radius [0.11 +/- 0.02 (SE)]. These results suggest that the movement of macromolecules from the pulmonary microcirculation into pulmonary lymph collected from the caudal mediastinal node of the sheep is influenced by both molecular size and molecular charge and that, compared with uncharged dextrans, the steady-state passage of anionic endogenous proteins from plasma to lymph is enhanced.     

Casein micelle size from elastic and quasi-elastic light scattering measurements.

The average molecular weight, particle radius and size distribution of particles in skim milk from eight cows in mid-lactation have been measured by means of elastic and quasi-elastic light scattering techniques. The properties of sub-micellar casein particles in the milk of each cow were also studied. Particular attention has been given to the effects of particle size heterogeneity in the interpretation of results. The weight average molecular weight of the particles from different cows varied from 2.6-10(8) to 15-10(8) and the corresponding average particle radius varied between 90 and 130 nm. An unusual feature of these particles is their high water content, which was found to vary from 2.4 to 6.4 ml/g with a positive correlation between average particle density and average particle mass. Variations in particle water content can be most readily understood in terms of a gel-like casein micelle.     

High density lipoprotein particle size distribution in subjects with obstructive jaundice

High density lipoproteins (HDL) from 14 patients with obstructive jaundice were examined by gradient gel electrophoresis to determine the effect of obstruction on particle size distribution. HDL from 7 of these patients were fractionated by gel permeation chromatography and further characterized by electron microscopy, SDS gel electrophoresis, apolipoprotein A-I and apolipoprotein A-II immunoturbidimetry, and analysis of chemical composition. In addition, lecithin:cholesterol acyltransferase (LCAT) activity was measured and correlated with plasma apolipoprotein A-I concentration and particle size distribution. HDL were abnormal in all patients regardless of severity, cause, or duration of obstruction. The major HDL subfraction in normal subjects, HDL3a (radius 4.1-4.3 nm) was either absent or considerably diminished, and HDL2b (radius 5.3 nm) was also frequently absent. Very small particles comparable in size to normal HDL3c (radius 3.8 nm) were prominent. In patients with a bilirubin concentration greater than 250 mumol/l, normal HDL had totally disappeared and were replaced by large discoidal particles of radius 8.5 nm and small spherical particles of radius 3.6-3.7 nm. Both populations of particles were markedly depleted of cholesteryl ester and enriched in free cholesterol and phospholipid. The discoidal particles were rich in apolipoproteins E, A-I, A-II, and C, while the small spherical particles contained predominantly apolipoprotein A-I. LCAT activity was diminished in all subjects to 8-54% of normal, and was strongly positively correlated (r = 0.91 P less than 0.05) with plasma apolipoprotein A-I levels.     

Size separation of polystyrene sulfate particles (189 to 1085 nm radius) in solutions of methyl-hydroxypropyl-cellulose of different chain lengths

Polystyrene sulfate particles of 189 to 1,085 nm radius are retarded in their electrophoretic migration in solutions of methyl-hydroxypropyl-cellulose (4,000 and 15,000 cP viscosity for a 2% solution). The retardation is proportional to polymer concentration and increases inversely with particle size. This suggests a separation mechanism based on size-dependent exclusion from the polymer network, as found previously in other liquid polymer media. In the size range of approximately 200 to 600 nm radius, the 4,000 cP polymer exhibits a steeper decrease of retardation with particle size than the 15,000 cP polymer, but fails to exhibit any size dependent retardation beyond that range.     

Improvements in procedures for electrophoresis in dilute agarose gels

Procedures have been developed for performing electrophoresis in agarose gels with agarose concentrations as low as 0.035%. Using these procedures, agarose gel electrophoresis of the following has been performed: (a) bacteriophage T7 missing its tail fibers; no detectable sieving of this spherical particle (radius = 30.1 nm) occurred below 0.075% agarose, (b) duplex DNAs with molecular weights between 26.5 × 10⁶ and 110 × 10⁶.     

Determination of the size distribution of liposomes by SEC fractionation,and PCS analysis and enzymatic assay of lipid content

In this study, small liposomes obtained by high-pressure homogenization were fractionated according to their particle sizes by size exclusion chromatography (SEC). The subfractions were analyzed by photon correlation spectroscopy (PCS) as well as enzymatic phosphatidylcholine (PC) assay for their particle sizes and lipid contents, respectively. For small egg PC-liposomes, a size range of 15 nm to 60 nm was found, with 80% of the vesicles being smaller than 30 nm in size. This is in contradiction to a mean size of 85±32 nm as indicated by PCS without fractionation. The PCS technique appears to underestimate very small particles below 30 nm if (few) bigger particles are present. The PCS particle size analysis of unfractionated hydrogenated egg PC/cholesterol-liposomes (2:1, mole/mole) by PCS did not yield any significant results. On fractionation, however, a particle size range of 40 nm to 120 nm was determined in a reproducible manner. Our results indicate that the combination of size exclusion fractionation with subsequent photon correlation spectroscopic particle size analysis and enzymatic PC assay can give both more detailed and more reliable insight into the particle size distribution of small liposomes than PCS alone. Published: May 15, 2002.     

Adsorption of globular proteins on locally planar surfaces: models for the effect of excluded surface area and aggregation of adsorbed protein on adsorption equilibria.

Equilibrium statistical-thermodynamic models are presented for the surface adsorption of proteins modeled as regular convex hard particles. The adsorbed phase is treated as a two-dimensional fluid, and the chemical potential of adsorbed protein is obtained from scaled particle theory. Adsorption isotherms are calculated for nonassociating and self-associating adsorbing proteins. Area exclusion broadens adsorption isotherms relative to the Langmuir isotherm (negative cooperativity), whereas self-association steepens them (positive cooperativity). The calculated isotherm for adsorption of hard spheres using scaled particle theory for hard discs agrees well with that calculated from the hard disc virial expansion. As the cross section of the adsorbing protein in the plane of the surface becomes less discoidal, the apparent negative cooperativity manifested in the isotherm becomes more pronounced. The model is extended to the case of simultaneous adsorption of a tracer protein at low saturation and a competitor protein with a different size and/or shape at arbitrary fractional saturation. Area exclusion by competitor for tracer (and vice versa) is shown to substantially enhance the displacement of tracer by competitor and to qualitatively invalidate the standard interpretation of ligand competition experiments, according to which the fractional displacement of tracer by competitor is equal to the fractional saturation by competitor.     

Apolipoprotein A-I localization and dipalmitoylphosphatidylcholine dynamics in reconstituted high density lipoproteins

The structure and molecular dynamics of recombinant high density lipoproteins (rHDL) were studied by non-radiative energy transfer (NRET), fluorescence anisotropy and intensity measurements. The rHDL particles contained human plasma apolipoprotein (apo) A-I and dipalmitoylphosphatidylcholine (DPPC). Fluorescent cis- and trans-parinaric acids were used both as probes of molecular motion in the particle lipid phase and as acceptors in the Forster's energy transfer from apo A-I tryptophan residues to determine particle dimensions, apolipoprotein localization and lipid dynamics. The probes are sensitive to thermal wobbling (macromobility) and conformational deformations (micromobility) of phospholipid acyl chains. The experimental data fitted to various models of the particle structure are compatible with the following: (a) at T < Tt the particles appeared as lens-like discs with a radius of the lipid phase of 5 nm and a mean thickness of 4 nm, the value being more by 20% in the particle centre, the alpha-helices of about 1 nm thickness were located around the edge of the lipid core. Compared to liposomes, both macro- and micromobility of DPPC molecules in rHDL were more rapid due to a significant disorder of the boundary lipid molecules close to the apo A-I molecule. This disorder led to the increase of the specific surface area per one lipid molecule, S(o). The lipid phase can be divided into three regions: (i) zone I of the most tightly packed lipid (0-1.7 nm from the disc axis) with a S(o) value small as 0.5 nm2; (ii) intermediate zone II (from 1.7 to 4.0 nm); and (iii) boundary lipid zone III (4-5 nm) of significantly disordered lipid with a S(o) value large as 0.65 nm2. (b) at T> Tt the S(o) heterogeneity disappeared, the radius of the lipid phase did not increase significantly, not exceeding 5.2-5.4 nm, but protein-induced immobilization of lipid molecules which affected about half or more of the total lipid, became remarkable. The overall effect was the suppression of the transition amplitude in rHDL compared to liposomes. The structural inhomogeneity might underlie the function of the native plasma HDL as the key component of the transport and metabolism of plasma lipids.     

Branchial arches of suspension–feeding Oreochromis esculentus: sieve or sticky filter?

Only 4% of 600 particles slid on or adhered to buccopharyngeal surfaces of Oreochromis esculentus . Only a minute quantity of mucus was visible during feeding (0·6% of 68 224 video frames; 30 frames s^–1), and very few particles were retained in this mucus. Filtration by mucus entrapment on the branchial arches is rejected as a mechanism of particle retention in this species. Since particles contacted the branchial arches infrequently, and the inter–raker spaces are too large to retain particles less than approximately 50 μm, sieving by gill rakers as a particle retention mechanism is rejected also. Unlike the suspension–feeding congener O. niloticus, O. esculentus does not use branchial arches to retain food particles by sieving or by mucus entrapment.     

Equivalent pore estimate for the alveolar-airway barrier in isolated dog lung

In high-pressure pulmonary edema, lung interstitial and air space edema liquids have equal protein concentrations (Am. J. Physiol. 231: 1466, 1976). This suggests that the alveolar-airway barrier separating the air and interstitial spaces is relatively unrestrictive, even without apparent epithelial injury. To estimate the equivalent pore population of the alveolar-airway barrier we inflated each of 18 isolated dog lung lobes for 1 h with a solution of colored tracer of uniform radius. Tracer radii ranged from 1.3 to 405 nm. After freezing the lobes in liquid N2, we measured interstitial tracer concentrations in frozen perivascular cuffs or in samples thawed after dissection from frozen cuffs. Relative to the concentrations instilled, interstitial concentrations ranged from 0.34 for the smallest particles (1.3 and 3.5 nm radius) to zero for particles with radii of 405 nm. From the results we designed a pore model of the alveolar-airway barrier to reproduce the concentrations we measured. No single-pore model could be obtained, although a three-pore model fit the data well. The model results predict that pores with radii of 1, 40, and 400 nm would account for 68, 30, and 2% of total liquid flux, respectively. The majority of liquid flux (68%) would occur through passageways smaller than the smallest tracer we used (1.3 nm radius). We believe the alveolar-airway barrier consists not only of tight intercellular junctions that allow passage of only water and electrolytes but also of a smaller number of large leaks that allow passage of particles up to nearly 400 nm in radius.     

Porin from bacterial and mitochondrial outer membranes

The outer membrane of gram-negative bacteria acts as a molecular filter with defined exclusion limit for hydrophilic substances. The exclusion limit is dependent on the type of bacteria and has for enteric bacteria like Escherichia coli and Salmonella typhimurium a value between 600 and 800 Daltons, whereas molecules with molecular weights up to 6000 can penetrate the outer membrane of Pseudomonas aeruginosa. The molecular sieving properties result from the presence of a class of major proteins called porins which form trimers of identical subunits in the outer membrane. The porin trimers most likely contain only one large but well-defined pore with a diameter between 1.2 and 2 nm. Mitochondria are presumably descendents of gram-negative bacteria. The outer membrane of mitochondria contains in agreement with this hypothesis large pores which are permeable for hydrophilic substances with molecular weights up to 6000. The mitochondrial porins are processed by the cell and have molecular weights around 30,000 Daltons. There exists some evidence that the pore is controlled by electric fields and metabolic processes.     

Evaluation of a sterile filtration process for viral vaccines using a model nanoparticle suspension

There is growing interest in the development of new vaccines based on live‐attenuated viruses (LAVs) and virus‐like particles. The large size of these vaccines, typically 100–400 nm, significantly complicates the use of sterile filtration. The objectives of this study are to examine the performance of several commercial sterile filters for filtration of a cytomegalovirus vaccine candidate (referred to as the LAV) and to develop and evaluate the use of a model nanoparticle suspension to perform a more quantitative assessment. Data obtained with a mixture of 200‐ and 300‐nm fluorescent particles provided yield and pressure profiles that captured the behavior of the viral vaccine. This included the excellent performance of the Sartorius Sartobran P filter, which provided greater than 80% yield of both the vaccine and model particles even though the average particle size was more than 250 nm. The particle yield for the Sartobran P was independent of filtrate flux above 200 L/m²/h, but increased with increasing particle concentration, varying from less than 10% at concentrations around 10⁷ particles/ml to more than 80% at concentrations above 10¹⁰ particles/ml due to saturation of particle capture/binding sites within the filter. These results provide important insights into the factors controlling transmission and fouling during sterile filtration of large vaccine products.     

Temperature-induced fusion of small unilamellar vesicles formed from saturated long-chain lecithins and diheptanoylphosphatidylcholine

Small unilamellar vesicles which form when gel-state long-chain phosphatidylcholines are mixed with micellar short-chain lecithins undergo an increase in size as the long-chain species melts to its liquid-crystalline form. Analysis of the vesicle population with quasi-elastic light scattering shows that the particle size increases from 90-A radius to greater than 5000-A radius. Resonance energy transfer experiments show total mixing of lipid probes with unlabeled vesicles only when the Tm of the long-chain phosphatidylcholine is exceeded. This implies that the large size change represents a fusion process. Aqueous compartments are also mixed during this transition. 31P NMR analysis of the vesicle mixtures above the phase transition shows a great degree of heterogeneity with large unilamellar particles coexisting with oligo- and multilamellar structures. Upon cooling the vesicles below the Tm, the original size distribution (e.g., small unilamellar vesicles) is obtained, as monitored by both quasi-elastic light scattering and 31P NMR spectroscopy. This temperature-induced fusion of unilamellar vesicles is concentration dependent and can be abolished at lower total phospholipid concentrations. It occurs over a wide range of long-chain to short-chain ratios and occurs with 1-palmitoyl-2-stearoylphosphatidylcholine and dimyristoylphosphatidylcholine as well. Characterization of this fusion event is used to understand the anomalous kinetics of water-soluble phospholipases toward these unusual vesicles.     

A model of random fiber networks that describes the sieving of particles during gel electrophoresis.

A model is presented to describe the sieving of particles during gel electrophoresis by considering the movement of a spherical particle through a random network of straight, rigid fibers. The movement of the particle through the network is approximated by a discrete model of the network composed of parallel planes containing fibers through which the particle must pass. Unlike previous models this model does not assume that the rate of movement is proportional to the proportion of cross-sectional area available to the particle. The results provide a new justification for approximately linear Ferguson plots and suggest that for large particles, Ferguson plots may become nonlinear.     

Neutron scattering studies of chromatosomes

Neutron scattering data establish that the radius of gyration of the DNA in chicken erythrocyte chromatosome particles is significantly higher, by about 0.3 nm, than the radius of gyration of the DNA in the core particle. Corresponding information of the radius of gyration of the protein component in the chromatosomes (3.75 nm) indicated an enlargement, compared to the radius of gyration of the octamer of histone proteins both in core particles and in the histone octamer stabilised in 2 M NaCl (3.25 nm). From the latter data, we could calculate the distance in the chromatosome between the centre of mass of the linker histone and the histone octamer as 5.5 nm. These results impose severe limitations for the organisation of the 22 bp extra DNA and the possible location of H1/H5 in the chromatosome, implying that the H1/H5 is close to the centre turn of the core particle DNA.     

Size and Dynamics of the Vibrio cholerae Porins OmpU and OmpT Probed by Polymer Exclusion

The trimeric OmpU and OmpT porins form large, triple-barrel hydrophilic channels in the outer membrane of the pathogen Vibrio cholerae. They have distinct pore properties, such as conductance, block by deoxycholic acid, and sensitivity to acidic pH. Their three-dimensional structures are unknown, but they share significant sequence homologies. To gain insight into the molecular basis for the distinct functional properties of these two similar porins, we carried out polymer exclusion experiments using planar lipid bilayer and patch-clamp electrophysiology. By studying the partitioning of polyethylene glycols (PEGs) of different molecular weights into each porin, we determined an effective radius of 0.55 nm and 0.43 nm for OmpU and OmpT respectively, and found an increased OmpU effective radius at acidic pH. PEGs or high buffer ionic strength promotes the appearance of single step closures in OmpU similar to the acidic-pH induced closures we documented previously. In addition, these closing events can be triggered by nonpenetrating PEGs applied asymmetrically. We believe our results support a model whereby acidic pH, high ionic strength, or exposure to PEGs stabilizes a less conductive state that corresponds to the appearance of an additional resistive element on one side of the OmpU protein and common to the three monomers.     

High-efficiency intracellular magnetic labeling with novel superparamagnetic-Tat peptide conjugates

A biocompatible, dextran coated superparamagnetic iron oxide particle was derivatized with a peptide sequence from the HIV-tat protein to improve intracellular magnetic labeling of different target cells. The conjugate had a mean particle size of 41 nm and contained an average of 6.7 tat peptides. Derivatized particles were internalized into lymphocytes over 100-fold more efficiently than nonmodified particles, resulting in up to 12.7 x 10(6) particles/cell. Internalized particles localized in cytoplasm and nuclear compartments as demonstrated by fluorescence microscopy and immunohistochemistry. Labeled cells were highly magnetic, were detectable by NMR imaging, and could be retained on magnetic separation columns. The described method has potential applications for in vivo tracking of magnetically labeled cells by MR imaging and for recovering intracellularly labeled cells from organs.     

Interaction between DNA and charged colloids could be hydrophobically driven

The interaction of DNA with amino-functionalized polystyrene particles has been studied by using a dynamic light scattering (DLS) technique. In 10 mM NaBr solution the particles have a hydrodynamic radius of 76 nm and the DNA macromolecule investigated (double stranded) has a hydrodynamic radius of 107 nm. At very low DNA concentrations, DNA adopts a flat conformation on the particle surface. If the DNA concentration is increased above 0.1 microg/mL, the thickness of the DNA layer increases, suggesting the presence of large loops and tails. Although the particles contain primary amino groups, they have a negative net charge under the conditions used in this work. Thus, the driving force for DNA adsorption is not of electrostatic origin but rather due to a hydrophobic effect. Addition of cationic surfactant to the DNA-precoated amino-functionalized particles induces changes in the adsorbed layer conformation, in agreement with the coadsorption of cationic surfactant.