An electrodiffusion model for effects of surface glycocalyx layer on microvessel permeability

To investigate the charge effect of the endothelial surface glycocalyx on microvessel permeability, we extended the three-dimensional model developed by Fu et al. (J Biomech Eng 116: 502-513, 1994) for the interendothelial cleft to include a negatively charged glycocalyx layer at the entrance of the cleft. Both electrostatic and steric exclusions on charged solutes were considered within the glycocalyx layer and at the interfaces. Four charge-density profiles were assumed for the glycocalyx layer. Our model indicates that the overall solute permeability across the microvessel wall including the surface glycocalyx layer and the cleft region is independent of the charge-density profiles as long as they have the same maximum value and the same total charge. On the basis of experimental data, this model predicts that the charge density would be 25-35 meq/l in the glycolcalyx of frog mesenteric capillaries. An intriguing prediction of this model is that when the concentrations of cations and anions are unequal in the lumen due to the presence of negatively charged proteins, the negatively charged glycocalyx would provide more resistance to positively charged solutes than to negatively charged ones.     

TNF-alpha increases entry of macromolecules into luminal endothelial cell glycocalyx

The endothelial luminal glycocalyx has been largely ignored as a target in vascular pathophysiology even though it occupies a key location. As a model of the inflammatory response, we tested the hypothesis that tumor necrosis factor-alpha (TNF-alpha) can alter the properties of the endothelial apical glycocalyx. In the intact hamster cremaster microcirculation, fluorescein isothiocyanate (FITC)-labeled Dextrans 70, 580, and 2,000 kDa are excluded from a region extending from the endothelial surface almost 0.5 micrometer into the lumen. This exclusion zone defines the boundaries of the glycocalyx. Red blood cells (RBC) under normal flow conditions are excluded from a region extending even farther into the lumen. The cremaster microcirculation was pretreated with topical or intrascrotal applications of TNF-alpha. After infusion of FITC-dextran, FITC-albumin, or FITC-immunoglubulin G (IgG) via a femoral cannula, microvessels were observed with bright-field and fluorescence microscopy to obtain estimates of the anatomic diameters and the widths of fluorescent tracer columns and of the RBC columns (means +/- SE). After 2 h of intrascrotal TNF-alpha exposure, there was a significant increase in access of FITC-Dextrans 70 and 580 to the space bounded by the apical glycocalyx in arterioles, capillaries, and venules, but no significant change in access of FITC-Dextran 2,000. The effects of TNF-alpha could be observed as early as 20 min after the onset of topical application. TNF-alpha treatment also significantly increased the penetration rate of FITC-Dextran 40, FITC-albumin, and FITC-IgG into the glycocalyx and caused a significant increase in the intraluminal volume occupied by flowing RBC. White blood cell adhesion increased during TNF-alpha application, and we used the selectin antagonist fucoidan to attenuate leukocyte adhesion during TNF-alpha stimulation. This did not inhibit the TNF-alpha-mediated increase in permeation of the glycocalyx. These results show that proinflammatory cytokines can cause disruption of the endothelial apical glycocalyx, leading to an increased macromolecular permeation in the absence of an increase in leukocyte recruitment.     

Endothelial monolayer permeability to macromolecules

The barrier function of the endothelial monolayer has not been extensively investigated using the cultured endothelium. The in vitro approach may contribute to a more complete understanding of microvessel wall permeability. Our studies using an in vitro endothelial monolayer system have led us to the following conclusions: the endothelial monolayer is more permeable to small-molecular-weight substances than to large molecules; the permeability of albumin is different for endothelial cells derived from different vascular sites (higher for pulmonary venous than pulmonary arterial endothelium); basement membrane components may have a significant role in the permeability of albumin across the endothelium; control of endothelial monolayer permeability is determined not only by the characteristics of the macromolecule (i.e., size and charge) but also by the shape of the endothelial cells and the size of interendothelial space.     

A model for the modulation of microvessel permeability by junction strands

To investigate the effect of junction strands on microvessel permeability, we extend the previous analytical model developed by Fu et al. (1994, J. Biomech. Eng., 116, pp. 502-513), for the interendothelial cleft to include multiple junction strands in the cleft and an interface between the surface glycocalyx layer and the cleft entrance. Based on the electron microscopic observations by Adamson et al. (1998, Am. J. Physiol., 274(43), pp. H1885-H1894), that elevation of intracellular cAMP levels would increase number of tight junction strands, this two-junction-strand and two-pore model can successfully account for the experimental data for the decreased permeability to water, small and intermediate-sized solutes by cAMP.     

Modulation of venular microvessel permeability by calcium influx into endothelial cells.

It has been proposed that calcium ion influx into endothelial cells modulates the permeability of venular microvessels via a calcium-dependent contractile process. The results of recent investigations using permeabilized endothelial cell monolayers conform to this hypothesis by demonstrating a calcium-dependent interaction of endothelial actin and myosin during the retraction of adjacent endothelial cells exposed to inflammatory agents. Little is known about the pathway for calcium influx into endothelial cells after exposure to mediators of inflammation, but evidence suggests that the properties of the calcium entry pathways are similar to the calcium entry pathways that regulate the release of endothelium-derived relaxing factor (EDRF). Substances that stimulate EDRF release from arterial endothelium also increase venular microvessel permeability. Recently developed methods to measure cytoplasmic calcium concentration in the endothelial cells forming the walls of individually perfused microvessels enable a direct investigation of the modulation of the permeability of venular microvessels by calcium influx. These experiments demonstrate that the magnitude of the initial increase in the permeability of microvessels after exposure to an agent that increases permeability, such as a calcium ionophore, is determined by the magnitude of calcium ion influx into the endothelial cells. Furthermore, the magnitude of the calcium influx into endothelial cells is modulated by the membrane potential of the endothelial cells. Depolarization of the endothelial cell membrane reduces calcium influx and attenuates increases in permeability whereas hyperpolarization of the endothelial membrane increases calcium influx and potentiates increases in permeability. These data conform to the hypothesis that a passive conductance channel for calcium is a major pathway for calcium ion flux responsible to eliciting an increase in the permeability of the endothelial barrier in microvessels.     

Hypoxia induces permeability in brain microvessel endothelial cells via VEGF and NO

In this study, an in vitro model of the blood-brain barrier,consisting of porcine brain-derived microvascular endothelial cells(BMEC), was used to evaluate the mechanism of hypoxia-induced hyperpermeability. We show that hypoxia-induced permeability in BMECwas completely abolished by a neutralizing antibody to vascular endothelial growth factor (VEGF). In contrast, under normoxic conditions, addition of VEGF up to 100 ng/ml did not alter monolayer barrier function. Treatment with either hypoxia or VEGF under normoxicconditions induced a twofold increase in VEGF binding sites and VEGFreceptor 1 (Flt-1) mRNA expression in BMEC. Hypoxia-induced permeability also was prevented by the nitric oxide (NO) synthase inhibitor N^G-monomethyl-L-arginine,suggesting that NO is involved in hypoxia-induced permeability changes,which was confirmed by measurements of the cGMP level. During normoxia,treatment with VEGF (5 ng/ml) increased permeability as well as cGMPcontent in the presence of several antioxidants. These results suggestthat hypoxia-induced permeability in vitro is mediated by the VEGF/VEGFreceptor system in an autocrine manner and is essentially dependent onreducing conditions stabilizing the second messenger NO as the mediatorof changes in barrier function of BMEC.     

A possible model for cell-cell recognition via surface macromolecules.

Alternative possibilities for the establishment of the proper cell distribution during embryogenesis are summarized at the beginning, followed by an assessment of the examples known so far where cell-cell recognition is known to be mediated via cell surface components. In the second part the species-specific recognition process which occurs during the sorting-out of dissociated sponge cells is analysed since it may serve as a possible model for cell-cell recognition in higher animals. Three possible mechanisms for the establishment of proper cell distribution are considered. These include, first, chemotaxis: secondly, guidance of cell or cell sheet movement by extracellular matrix or by surrounding cells and thirdly, random movement followed by recognition at the final point of destination. Recognition is necessary for both of the two latter processes, i.e. for cell guidance as well as for locking the cells into their final position after random movement. Two basically different recognition mechanisms should be distinguished from each other. On the one hand cells may recognize each other with the help of macromolecules situated in or just outside of the plasmamembrane which fit to each other like enzymes and substrates or antibodies and antigens. On the other hand, cells may exchange information by exchanging cytoplasmatic components via vesicles or gap junctions. The species-specific aggregation of dissociated sponge cells is considered to be a possible model for cell-cell recognition in higher animals. A proteoglycan-like intercellular macromolecule called aggregation factor seems to mediate recognition of a given species of cells in the reaggregation process of dissociated cells. The data available at the present time suggest that a monovalent surface macromolecule (baseplate) may mediate the recognition process probably by recognizing the carbohydrate side chains of the multivalent proteoglycan aggregation factor. A cell-free system was devised to mimic this aggregation process. Addition of aggregation factor to baseplate-coated sepharose beads of approximately the size of the original sponge cells has essentially the same characteristics as the cellular system. Macromolecule-coded surface information for the recognition between cells has not been established during the embryogenesis of higher animals and remains an interesting challenge.     

Increased permeability of primary cultured brain microvessel endothelial cell monolayers following TNF-alpha exposure.

TNF-alpha is a cytokine that produces increased permeability in the peripheral vasculature; however, little is known about the effects of TNF-alpha on the blood-brain barrier (BBB). Using primary cultured bovine brain microvessel endothelial cells (BBMEC) as an in vitro model of the BBB, this study shows that TNF-alpha produces a reversible increase in the permeability of the brain microvessel endothelial cells. The BBMEC monolayers were pre-treated with 100 ng/ml of TNF-alpha for periods ranging from 2 to 12 hours. Permeability was assessed using three molecular weight markers, fluorescein (376 MW), fluorescein-dextran (FDX-4400; 4400 MW), and FDX-70000 (MW 70000). The permeability of BBMEC monolayers to all three fluorescent markers was increased two-fold or greater in the TNF-alpha treatment group compared to control monolayers receiving no TNF-alpha. Significant changes in permeability were also observed with TNF-alpha concentrations as low as 1 ng/ml. These results suggest that TNF-alpha acts directly on the brain microvessel endothelial cells in a dynamic manner to produce a reversible increase in permeability. Exposure of either the lumenal or ablumenal side of BBMEC monolayers to TNF-alpha resulted in similar increases in permeability to small macromolecules, e.g. fluorescein. However, when a higher molecular weight marker was used (e.g. FDX-3000), there was a greater response following lumenal exposure to TNF-alpha. Together, these studies demonstrate a reversible and time dependent increase in brain microvessel endothelial cell permeability following exposure to TNF-alpha. Such results appear to be due to TNF's direct interaction with the brain microvessel endothelial cell.     

Ability of human T lymphocytes to adhere to vascular endothelial cells and to augment endothelial permeability to macromolecules is linked to their state of post-thymic maturation

The accumulation of mononuclear cells at sites of chronic inflammation is dependent on a number of factors including localized adherence of lymphocytes to vascular endothelial cells (EC), cytokine-mediated increased adhesiveness of endothelium, chemotactic factors and endothelial permeability. The present study investigates two of the above attributes of lymphocyte-EC interaction: namely, the ability of maturationally distinct subpopulations of human T lymphocytes to adhere to vascular EC and to increase vascular endothelial permeability to macromolecules in an in vitro model. Thus, human T lymphocytes were separated into CD4+ CD8-helper/inducer, CD4- CD8+ cytotoxic/suppressor, CD29+ CD45RA- CD45RO+ memory, and CD29- CD45RA+ CD45RO- naive/virgin T subpopulations, were activated with PHA and PMA, and then examined for their adherence to EC and also for their effect on endothelial permeability. Upon activation, cells within each of the above four subpopulations exhibited increased adherence to EC. In contrast, resting CD29+ CD45RA- CD45RO+ memory T lymphocytes exhibited two to three times greater ability to adhere to EC than their CD29- CD45RA+ CD45RO- naive/virgin counterparts. Consistent with their increased adherence to EC, CD29+ CD45RO+ memory T lymphocytes, when activated, significantly increased endothelial permeability to albumin. Although activated CD45RA+ naive T lymphocytes exhibited increased adherence to EC, these cells failed to increase significantly endothelial permeability. Similar to their polyclonal counterparts, Ag-specific CD4+ CD29+ CD45RO+ T cell clones, but not their actively released mediators, also increased endothelial permeability via a noncytolytic mechanism(s). This ability of CD29+ CD45RO+ memory T lymphocytes to augment endothelial permeability may facilitate their transendothelial migration into extravascular space. These observations may provide additional insights into molecular mechanism(s) underlying pathophysiology of localized chronic inflammatory responses in general and more specifically selective accumulation of CD29+/CD45RO+ memory T lymphocytes at sites of chronic inflammation such as rheumatoid synovium.     

Effects of human neutrophil chemotaxis across human endothelial cell monolayers on the permeability of these monolayers to ions and macromolecules

We have developed a method for studying the permeability properties of human endothelia in vitro. Human umbilical vein endothelial cells (HUVEC) were cultured on a substrate of human amnion. Confluent monolayers of these cells demonstrated 6-12 delta.cm2 of electrical resistance (a measure of their permeability to ions) and restricted the transendothelial passage of albumin from their apical to their basal surface. To determine whether leukocyte emigration alters endothelial permeability in this model, we examined the effects of migrating human polymorphonuclear leukocytes (PMN) on these two parameters. Few PMN migrated across the HUVEC monolayers in the absence of chemoattractants. In response to chemoattractants, PMN migration through HUVEC monolayers was virtually complete within 10 minutes and occurred at random locations throughout the monolayer. PMN migrated across the monolayer via the paracellular pathway. Although one PMN migrated across the monolayer for each HUVEC, PMN migration induced no change in electrical resistance or albumin permeability of these monolayers. At this PMN:HUVEC ratio, these permeability findings were correlated morphologically to measurements that HUVEC paracellular pathway size increases by less than 0.22% with PMN migration. This increase is insufficient to effect a measurable change in the electrical resistance of the endothelial cell monolayer. These findings demonstrate that increased permeability of cultured endothelial cell monolayers is not a necessary consequence of PMN emigration.     

Microrheology of hyaluronan solutions: implications for the endothelial glycocalyx

The endothelial glycocalyx (EG) is a complex biopolymer network produced by vascular endothelial cells that forms a layer with multiple functions at the luminal side of blood vessels. The EG acts as an anti-adhesive protection layer, as a molecular sieve, as a chemical sensor site, and as a mechanotransducer of fluid shear stress to the underlying cell layer. A major component involved in these processes is the highly hydrated glycosaminoglycan (GAG) hyaluronan (HA). Here we used laser interferometry to measure the broadband mechanical response of reconstituted HA solutions at close to physiological conditions. HA showed rheological behavior consistent with that of a flexible polymer. The elastic behavior observed for entangled HA networks showed reptational relaxation with a large distribution of time scales, which disappeared quickly (15 min) with the addition of hyaluronidase (HAase). We conclude that the broadband mechanical probing of model systems (HA solutions) provides quantitative data that are crucial to understand the mechanical response of the EG in vivo and its role in mechanosensing.     

Reversibility of increased microvessel permeability in response to VE-cadherin disassembly

We determined the role of vascular endothelial (VE)-cadherin complex in regulating the permeability of pulmonary microvessels. Studies were made in mouse lungs perfused with albumin-Krebs containing EDTA, a Ca(2+) chelator, added to study the VE-cadherin junctional disassembly. We then repleted the perfusate with Ca(2+) to restore VE-cadherin integrity. Confocal microscopy showed a disappearance of VE-cadherin immunostaining in a time- and dose-dependent manner after Ca(2+) chelation and reassembly of the VE-cadherin complex within 5 min after Ca(2+) repletion. We determined the (125)I-labeled albumin permeability-surface area product and capillary filtration coefficient (K(fc)) to quantify alterations in the pulmonary microvessel barrier. The addition of EDTA increased (125)I-albumin permeability-surface area product and K(fc) in a concentration-dependent manner within 5 min. The permeability response was reversed within 5 min after repletion of Ca(2+). An anti-VE-cadherin monoclonal antibody against epitopes responsible for homotypic adhesion augmented the increase in K(fc) induced by Ca(2+) chelation and prevented reversal of the response. We conclude that the disassembled VE-cadherins in endothelial cells are mobilized at the junctional plasmalemmal membrane such that VE-cadherins can rapidly form adhesive contact and restore microvessel permeability by reannealing the adherens junctions.     

Structural mechanisms of acute VEGF effect on microvessel permeability

To investigate the ultrastructural mechanisms of acute microvessel hyperpermeability by vascular endothelial growth factor (VEGF), we combined a mathematical model (J Biomech Eng 116: 502-513, 1994) with experimental data of the effect of VEGF on microvessel hydraulic conductivity (L(p)) and permeability of various-sized solutes. We examined the effect of VEGF on microvessel permeability to a small solute (sodium fluorescein, Stokes radius 0.45 nm), an intermediate solute (alpha-lactalbumin, Stokes radius 2.01 nm), and a large solute [albumin (BSA), Stokes radius 3.5 nm]. Exposure to 1 nM VEGF transiently increased apparent permeability to 2.3, 3.3, and 6.2 times their baseline values for sodium fluorescein, alpha-lactalbumin, and BSA, respectively, within 30 s, and all returned to control within 2 min. On the basis of L(p) (DO Bates and FE Curry. Am J Physiol Heart Circ Physiol 271: H2520-H2528, 1996) and permeability data, the prediction from the model suggested that the most likely structural changes in the interendothelial cleft induced by VEGF would be a approximately 2.5-fold increase in its opening width and partial degradation of the surface glycocalyx.     

Anin vitro model for endothelial permeability: Assessment of monolayer integrity

Summary An essential component of anyin vitro model for endothelial permeability is a confluent cell monolayer. The model reported here utilizes primary human umbilical vein endothelial cells (HUVEC) cultured on recently developed polyethylene terephthalate micropore membranes. Using a modification of the Wright-Giemsa stain, confluent HUVEC monolayers grown on micropore membranes were routinely assessed using light microscopy. Determination of confluence using this method was confirmed by scanning electron microscopy. Transendothelial electrical resistance of HUVEC monolayers averaged 27.9±11.4 Ω · cm², 10 to 21% higher than literature values. Studies characterizing the permeability of the endothelial cell monolayer to³H-inulin demonstrated a linear relationship between the luminal concentration of³H-inulin and its flux across HUVEC monolayers. The slope of the flux versus concentration plot, which represents endothelial clearance of³H-inulin, was 2.01±0.076 × 10⁻⁴ ml/min (r²=.9957). The permeability coefficient for the HUVEC monolayer-micropore membrane barrier was 3.17±0.427×10⁻⁶ cm/s with a calculated permeability coefficient of the HUVEC monolayer alone of 4.07±0.617×10⁻⁶ cm/s. The HUVEC monolayer reduced the permeability of the micropore membrane alone to³H-inulin (1.43±0.445×10⁻⁵ cm/s) by 78%. Evans blue dye-labeled bovine serum albumin could not be detected on the abluminal side without disruption of the HUVEC monolayer. These results demonstrate a model for endothelial permeability that can be extensively assessed for monolayer integrity by direct visualization, transendothelial electrical resistance, and the permeability of indicator macromolecules.     

An electrodiffusion model for the blood-brain barrier permeability to charged molecules

The endothelial surface glycocalyx layer (SGL) and the basement membrane (BM) are two important components of the blood-brain barrier (BBB). They provide large resistance to solute transport across the BBB in addition to the tight junctions in the cleft between adjacent endothelial cells. Due to their glycosaminoglycan compositions, they carry negative charge under physiological conditions. To investigate the charge effect of the SGL and BM on the BBB permeability to charged solutes, we developed an electrodiffusion model for the transport of charged molecules across the BBB. In this model, constant charge densities were assumed in the SGL and in the BM. Both electrostatic and steric interaction and exclusion to charged molecules were considered within the SGL and the BM and at their interfaces with noncharged regions of the BBB. On the basis of permeability data for the positively charged ribonuclease (+4,radius=2.01?nm) and negatively charged α-lactalbumin (-10,radius=2.08?nm) measured in intact rat mesenteric and pial microvessels, our model predicted that the charge density in both SGL and BM would be ～30?mEq/L, which is comparable to that in the SGL of mesenteric microvessels. Interestingly, our model also revealed that due to the largest concentration drop in the BM, there is a region with a higher concentration of negatively charged α-lactalbumin in the uncharged inter-endothelial cleft, although the concentration of α-lactalbumin is always lower than that of positively charged ribonuclease and that of a neutral solute in the charged SGL and BM.     

Role of a glycocalyx on coronary arteriole permeability to proteins: evidence from enzyme treatments

Whereas the glycocalyx of endothelial cells has been shown to influence solute flux from capillary microvessels, little is known about its contribution to the movement of macromolecules across the walls of other microvessels. We evaluated the hypothesis that a glycocalyx contributes resistance to protein flux measured in coronary arterioles. Apparent solute permeability (P(s)) to two proteins of different size and similar charge, alpha-lactalbumin (alpha-lactalb) and porcine serum albumin (PSA), was determined in arterioles isolated from the hearts of 43 female Yucatan miniature swine. P(s) was assessed in arterioles with an "intact" glycocalyx under control conditions and again after suffusion with adenosine (Ado, 10(-5) M, n = 42 arterioles, N = 29 pigs). In a second set of experiments (n = 21 arterioles, N = 21 pigs) arteriolar P(s) was determined before and after perfusion with enzyme (pronase or heparinase), which was used to digest the glycocalyx. P(s) was assessed a third time on those microvessels after exposure to Ado. Consistent with the hypothesis, P(s) for PSA (P(PSA)(s)) and P(s) for alpha-lactalb (P(alpha-lactalb)(s)) increased from basal levels following enzyme treatment. Subsequent suffusion with Ado, a significant metabolite known to alter coronary vascular smooth muscle tone and permeability, resulted in a significant reduction of basal P(alpha-lactalb)(s) in both untreated and enzyme-treated arterioles. Furthermore, in untreated arterioles, P(PSA)(s) was unchanged by Ado suffusion, whereas Ado induced a pronounced reduction in P(PSA)(s) of enzyme-treated vessels. These data demonstrate that in intact coronary arterioles an enzyme-sensitive layer, most likely at the endothelial cell surface, contributes significantly to net barrier resistance to solute flux.     

Effect of glycocalyx on shear-dependent albumin uptake in endothelial cells

The glycocalyx layer on the surface of an endothelial cell is an interface barrier for uptake of macromolecules, such as low-density lipoprotein and albumin, in the cell. The shear-dependent uptake of macromolecules thus might govern the function of the glycocalyx layer. We therefore studied the effect of glycocalyx on the shear-dependent uptake of macromolecules into endothelial cells. Bovine aorta endothelial cells were exposed to shear stress stimulus ranging from 0.5 to 3.0 Pa for 48 h. The albumin uptake into the cells was then measured using confocal laser scanning microscopy, and the microstructure of glycocalyx was observed using electron microscopy. Compared with the uptake into endothelial cells under static conditions (no shear stress stimulus), the albumin uptake at a shear stress of 1.0 Pa increased by 16% and at 3.0 Pa decreased by 27%. Compared with static conditions, the thickness of the glycocalyx layer increased by 70% and the glycocalyx charge increased by 80% at a shear stress of 3.0 Pa. The albumin uptake at a shear stress of 3.0 Pa for cells with a neutralized (no charge) glycocalyx layer was almost twice that of cells with charged layer. These findings indicate that glycocalyx influences the albumin uptake at higher shear stress and that glycocalyx properties (thickness and charge level) are involved with the shear-dependent albumin uptake process.     

Membrane permeability to macromolecules mediated by the membrane attack complex

A simple and well-defined system of purified phospholipids and human complement proteins was used to study membrane permeability to macromolecules mediated by the membrane attack complex (MAC) of complement. Large unilamellar vesicles (LUVs) of phosphatidylcholine (PC) or phosphatidylserine (PS) containing trapped macromolecules [bovine pancreatic trypsin inhibitor (BPTI), thrombin, glucose-6-phosphate dehydrogenase (G6PD), and larger molecules] were used to monitor permeability. Membrane permeability to macromolecules was measured by thrombin inhibition by an external inhibitor or by separation of released molecules by gel filtration. Membrane-bound intermediates (C5b-8 or C5b-93) were stable for hours, and macromolecular permeability occurred without fragmentation, fusion, or aggregation of the vesicles. Quantitative membrane binding by C5b-7 as well as essentially quantitative release of thrombin was obtained for PS vesicles. MAC binding to PS-LUVs approximated the theoretical Poisson distribution curve for full release of vesicle contents by one complex per vesicle. Reactions with PC-LUVs occurred with some fluid-phase MAC assembly. Therefore, results from experiments with these vesicles were interpreted in a relative manner. However, the values obtained closely corroborated those obtained with PS-LUVs. At low C9/C5b-8 ratios, the size of the lesion was proportional to the C9 content of the MAC. Half-maximum release of BPTI, thrombin, and G6PD, by a single MAC per vesicle, required approximately 3,5, and 7 C9/C5b-8 (mol/mol), respectively. Larger molecules (greater than or equal to 118-A diameter) were not released from the vesicles. Release of G6PD (95.4-A diameter) required 45% of saturating C9. Therefore, it appeared that the last half of the bound C9 molecules did not increase pore size and the pore which released G6PD approached the diameter of the closed circular lesion measured (by others) in electron micrographs (approximately 100 A). The results were consistent with the formation of a stable membrane pore by a single complex per vesicle in which C9 molecules line only one side of the pore at low C9/C5b-8 ratios and maximum pore size is attained by incomplete, noncircular polymers of C9.