Number and Volume of Islets Transplanted in Immunobarrier Devices

Immunobarrier devices may prevent immune destruction of transplanted islets, but there are concerns about survival within such devices. Islets were transplanted in diffusion chambers that employed two laminated polytetrafluoroethylene membranes held together with titanium rings. Five hundred syngeneic mouse islets placed in devices were transplanted into the epididymal fat pads of streptozotocin (STZ) diabetic mice (B6AF1). After 2 wk the devices were removed. Sections were made parallel to the membrane surface. Eight to 13 systematically selected sections of each device were analyzed by planimetry to determine the area of the device space and of the islets within that space. From these data we estimated total volume of the device, volume of islets, and number of islets in a device. The data were segregated into two groups: group I (blood glucose less than 100 mg/dL 2 wk after implantation), and group II (over 150 mg/dL). The volume (mean + SE) of devices implanted for 2 wk was 2.1 + 0.4 μL in group I and 2.2 + 0.2 μL in group II. The islet volume and number within devices were 0.30 + 0.06 and 0.17 + 0.01 μL, or 340 + 50 and 230 + 20 islets in group I and group II, respectively. The volume of fibrous tissue in devices was about 0.50 μL. About 10% of the islet tissue had central necrosis. The beta cell volume in a membrane device needed for cure is comparable to that required with islets under the kidney capsule (0.25-0.80 μL). The mass of islets contained within membrane devices needed to cure diabetes is equivalent to that of a graft in an optimal transplant site such as under the kidney capsule.

Implantable Biohybrid Artificial Organs

Biohybrid artificial organs encompass all devices which substitute for an organ or tissue function and incorporate both synthetic materials and living cells. This review concerns implantable immunoisolation devices in which the tissue is protected from immune rejection by enclosure within a semipermeable membrane. Two critical areas are discussed in detail: (i) Device design and performance as it relates to maintenance of cell viability and function. Attention is focussed on oxygen supply limitation and how it is affected by tissue density and the development of materials that induce neovascularization at the host tissue-membrane interface; and (ii) Protection from immune rejection. Our current knowledge of the mechanisms that may be operative in immune rejection in the presence of a semipermeable membrane barrier is limited. Nonetheless, recent studies shed light on the role played by membrane properties in preventing immune rejection, and many studies demonstrate substantial progress towards clinically useful implantable immunoisolation devices.

Oxygen supply by photosynthesis to an implantable islet cell device

Transplantation of islet cells is an effective treatment for type 1 diabetes with critically labile metabolic control. However, during islet isolation, blood supply is disrupted, and the transport of nutrients/metabolites to and from the islet cells occurs entirely by diffusion. Adequate oxygen supply is essential for function/survival of islet cells and is the limiting factor for graft integrity. Recently, we developed an immunoisolated chamber system for transplantation of human islets without immunosuppression. This system depended on daily oxygen supply. To provide independence from this external source, we incorporated a novel approach based on photosynthetically-generated oxygen. The chamber system was packed sandwich-like with a slab of immobilized photosynthetically active microorganisms (Synechococcus lividus) on top of a flat light source (LEDs, red light at 660 nm, intensity of 8 μE/m(2)/s). Islet cells immobilized in an alginate slab (500-1,000 islet equivalents/cm(2)) were mounted on the photosynthetic slab separated by a gas permeable silicone rubber-Teflon membrane, and the complete module was sealed with a microporous polytetrafluorethylene (Teflon) membrane (pore size: 0.4 μm) to protect the contents from the host immune cells. Upon illumination, oxygen produced by photosynthesis diffused via the silicone Teflon membrane into the islet compartment. Oxygen production from implanted encapsulated microorganisms was stable for 1 month. After implantation of the device into diabetic rats, normoglycemia was achieved for 1 week. Upon retrieval of the device, blood glucose levels returned to the diabetic state. Our results demonstrate that an implanted photosynthetic bioreactor can supply oxygen to transplanted islets and thus maintain islet viability/functionality.

A Study of the Mass Transport Resistance of Glucose Across Rat Capsular Membranes

The interaction between a polymer or other foreign surface and soft tissue is determined by a variety of materials and tissue factors. After failing to engulf the foreign body, the classical response is to wall it off. First the site is invaded by macrophages and giant cells and then fibrous connective tissue is laid down. This fibrous connective tissue gradually replaces the cellular matrix and forms the capsule. The composition is mostly collagen and mucopolysaccharides with few cells in the mature capsule. It contains 75–80% water.

When the implant surface represents a sensor and the transport of low molecular weight species across the capsule is necessary for meaningful measurement and response time, the mass transport resistance of the capsule may become a critical factor. This study represents an initial attempt to characterize the diffusion of glucose through fibrous capsules grown around silicone elastomer implants in a rat. Specifically, the study was designed to develop techniques to measure mass transport properties of tissue capsules, to use these techniques to determine effective glucose transport properties at two weeks, four weeks, and ten weeks after implant; and, to use these results along with histological examinations to gain an understanding of the factors which influence mass transport.

Rat islet cell aggregates are superior to islets for transplantation in microcapsules

AIMS/HYPOTHESIS

Islet transplantation is a promising treatment for type 1 diabetes but is hampered by a shortage of donor human tissue and early failure. Research on islet cell transplantation includes finding new sources of cells and immunoisolation to protect from immune assault and tumourigenic potential. Small islet cell aggregates were studied to determine if their survival and function were superior to intact islets within microcapsules because of reduced oxygen transport limitation and inflammatory mediators.

METHODS

Islet cell aggregates were generated by dispersing rat islets into single cells and allowing them to re-aggregate in culture. Rat islets and islet cell aggregates were encapsulated in barium alginate capsules and studied when cultured in low (0.5% or 2%) or normal (20%) oxygen, or transplanted into mice.

RESULTS

Encapsulated islet cell aggregates were able to survive and function better than intact islets in terms of oxygen-consumption rate, nuclei counts, insulin-to-DNA ratio and glucose-stimulated insulin secretion. They also had reduced expression of pro-inflammatory genes. Islet cell aggregates showed reduced tissue necrosis in an immunodeficient transplant model and a much greater proportion of diabetic xenogeneic transplant recipients receiving islet cell aggregates (tissue volume of only 85 islet equivalents) had reversal of hyperglycaemia than recipients receiving intact islets.

CONCLUSIONS/INTERPRETATION

These aggregates were superior to intact islets in terms of survival and function in low-oxygen culture and during transplantation and are likely to provide more efficient utilisation of islet tissue, a finding of importance for the future of cell therapy for diabetes.

Human islet oxygen consumption rate and DNA measurements predict diabetes reversal in nude mice

There is a need for simple, quantitative and prospective assays for islet quality assessment that are predictive of islet transplantation outcome. The current state-of-the-art athymic nude mouse bioassay is costly, technically challenging and retrospective. In this study, we report on the ability of 2 parameters characterizing human islet quality: (1) oxygen consumption rate (OCR), a measure of viable volume; and (2) OCR/DNA, a measure of fractional viability, to predict diabetes reversal in nude mice. Results demonstrate that the probability for diabetes reversal increases as the graft's OCR/DNA and total OCR increase. For a given transplanted OCR dose, diabetes reversal is strongly dependent on OCR/DNA. The OCR and OCR/DNA (the 'OCR test') data exhibit 89% sensitivity and 77% specificity in predicting diabetes reversal in nude mice (n = 86). We conclude that the prospective OCR test can effectively replace the retrospective athymic nude mouse bioassay in assessing human islet quality prior to islet transplantation.

2-Aminoethoxydiphenyl borate as a common activator of TRPV1, TRPV2, and TRPV3 channels

2-Aminoethoxydiphenyl borate (2APB) had been depicted as a universal blocker of transient receptor potential (TRP) channels. While evidence has accumulated showing that some TRP channels are indeed inhibited by 2APB, especially in heterologous expression systems, there are other TRP channels that are unaffected or affected very little by this compound. More interestingly, the thermosensitive TRPV1, TRPV2, and TRPV3 channels are activated by 2APB. This has been demonstrated both in heterologous systems and in native tissues that express these channels. A number of 2APB analogs have been examined for their effects on native store-operated channels and heterologously expressed TRPV3. These studies revealed a complex mechanism of action for 2APB and its analogs on ion channels. In this review, we have summarized the current results on 2APB-induced activation of TRPV1-3 and discussed the potential mechanisms by which 2APB may regulate TRP channels.

High-density culture of human islets on top of silicone rubber membranes

Islet culture has emerged as a standard practice prior to clinical transplantation. However, culturing large numbers of islets requires low islet density (number of islets per unit surface area) and, consequently, 20 to 30 flasks per pancreas in order to avoid hypoxia-induced death (HID). There is a need for a simple, practical, small-footprint culture vessel that will accommodate aseptic maintenance of entire human islet isolations while avoiding HID. In this communication, we examine the hypothesis that by improving oxygen transfer through culture of islets on silicone rubber membranes (SRM), we may increase islet surface coverage and reduce the number of flasks required while avoiding HID. Our results demonstrate that islets cultured for up to 48 hours in vessels with SRM bottoms at 2000 to 4000 islet equivalents (IE)/cm(2), a surface coverage 10- to 20-fold higher than the standard culture protocol, displayed no significant loss of viability. In contrast, islets cultured for 48 hours at 4000 IE/cm(2) in flasks with gas-impermeable bottoms suffered a 60% to 70% reduction in viability. The data suggest that it is possible to culture all islets isolated from a human pancreas on SRM in a single, standard-sized vessel while maintaining the same viability as with the current, standard culture protocols that require 20 to 30 flasks. This approach may lead to substantial improvements in islet culture for research and clinical transplantation.

Pancreas Oxygenation Is Limited During Preservation With the Two-Layer Method

BACKGROUND

The two-layer method (TLM) for pancreas preservation reportedly improves islet yield and transplantation outcome relative to previous methods. Increased ATP concentrations support the hypothesis that these improvements are related to better oxygenation from the perfluorocarbon solution. However, there are limited direct measurements of oxygen partial pressure, (pO(2)) in pancreata preserved with the TLM. Theory predicts that only a small fraction of a human pancreas can be oxygenated externally. In this report we examine pancreas oxygenation with the TLM using theory and direct pO(2) measurements.

METHODS

pO(2) profiles in cylindrical pancreata were calculated at various temperatures with a diffusion-reaction model. The pO(2) was measured using fiber optic sensors in the core of porcine pancreatic tissue preserved with the TLM in media saturated with 100% oxygen.

RESULTS

The model predicts that at 8 degrees C, even in the absence of an external pO(2) gradient, oxygen penetration depth is about 1 mm and insensitive to pancreas diameter, while the oxygenated volume fraction is about 15% for a 2.5-cm-diameter pancreas. Experimental measurements verified that pO(2) is virtually zero in the core of a 1-cm-thick pancreatic piece preserved with the TLM. Penetration of solution around the sensor may be responsible for the observed lag and for the previously reported nonzero pO(2) measurements. Reoxygenation of heat-treated tissue took several hours.

CONCLUSIONS

The TLM can oxygenate only a small volume fraction of a human pancreas. Pancreas oxygenation through the native vasculature should be explored to further improve yield of viable islets.

NMR spectroscopy in beta cell engineering and islet transplantation

Islet transplantation is a promising method for restoring normoglycemia and alleviating the long term complications of diabetes. Widespread application of islet transplantation is hindered by the limited supply of human islets and requires a large increase in the availability of suitable insulin secreting tissue as well as robust quality assessment methodologies that can ensure safety and in vivo efficacy. We explore the application of nuclear magnetic resonance (NMR) spectroscopy in two areas relevant to beta cell engineering and islet transplantation: (1) the effect of genetic alterations on glucose metabolism, and (2) quality assessment of islet preparations prior to transplantation. Results obtained utilizing a variety of NMR techniques demonstrate the following: (1) Transfection of Rat1 cells with the c-myc oncogene (which may be involved in cell proliferation and cell cycle regulation) and overexpression of Bcl-2 (which may protect cells from stresses such as hypoxia and exposure to cytokines) introduce a wide array of alterations in cellular biochemistry, including changes in anaerobic and oxidative glucose metabolism, as assessed by 13C and 31P NMR spectroscopy. (2) Overnight incubation of islets and beta cells in the bottom of centrifuge tubes filled with medium at room temperature, as is sometimes done in islet transportation, exposes them to severe oxygen limitations that may cause cell damage. Such exposure, leading to reversible or irreversible damage, can be observed with NMR-detectable markers using conventional 13C and 31P NMR spectroscopy of extracts. In addition, markers of irreversible damage (as well as markers of hypoxia) can be detected and quantified without cell extraction using high-resolution magic angle spinning 1H NMR spectroscopy. Finally, acute ischemia in a bed of perfused beta cells leads to completely reversible changes that can be followed in real time with 31P NMR spectroscopy.

Characterization of cell detachment from hollow fiber affinity membranes

Hollow fiber affinity membranes have potential for use in cell separations because they offer many advantages over currently available techniques. An understanding of the parameters controlling cell attachment and detachment from the surface is vital to the success of the separation. Cell attachment was probed with transmission electron microscopy (TEM), which revealed that gravity settling must be used to bring the cells to the surface of the membrane, because forward pressurization caused cells to infiltrate the pores of the membranes. Fluorescence microscopy studies showed that viable target cells could be recovered from the surface through the use of either back pressurization or liquid drainage and that 98% of the cells could be recovered from the surface through the sequential use of both. This ability to reproducibly detach target cells from the surface suggests that cell recovery will not be a limiting factor for cell separations with hollow fiber membranes.

Islets in alginate macrobeads reverse diabetes despite minimal acute insulin secretory responses

BACKGROUND

Encapsulation of islets has been widely investigated as a treatment for diabetes. The characteristics and dynamics of insulin secretion by encapsulated islets in response to glucose and other secretagogues are not well understood.

METHODS

In our study, macroencapsulated syngeneic islets at 3-4 wk after transplantation were studied for insulin release in response to i.v. glucose (hyperglycemic clamps at 250 or 350 mg/dl plasma glucose), arginine (i.v. bolus, 100 mg/kg), glucagon-like peptide-1 (i.v. infusion for 20 min, 2.2 pmol/kg/min), and meal challenge. Syngeneic islets (6000 islets) were encapsulated in alginate macrobeads (2-3 mm diameter) with or without poly-L-lysine coating and transplanted into the peritoneal cavity of STZ-diabetic Lewis rats. Normal (nontransplanted) and diabetic Lewis rats transplanted with "naked" islets under the kidney capsule served as controls.

RESULTS

Animals transplanted with macrobeads displayed subnormal insulin responses to glucose, arginine, and glucagon-like peptide-1 despite achieving normoglycemia faster than animals with renal subcapsular islet transplants. Plasma insulin responses to meal challenges were blunted in animals with macrobeads resulting in increased plasma glucose excursions.

CONCLUSIONS

We conclude that, after transplantation into diabetic Lewis rats, macroencapsulated islets have significantly impaired insulin secretion despite achieving normal fed glycemic levels.

Improved vascularization of planar membrane diffusion devices following continuous infusion of vascular endothelial growth factor

Improving blood vessel formation around an immunobarrier device should improve the survival of the encapsulated tissue. In the present study we investigated the formation of new blood vessels around a planar membrane diffusion device (the Baxter Theracyte System) undergoing a continuous infusion of vascular endothelial growth factor through the membranes and into the surrounding tissue. Each device (20 microl) had both an inner immunoisolation membrane and an outer vascularizing membrane. Human recombinant vascular endothelial growth factor-165 was infused at 100 ng/day (low dose: n = 6) and 500 ng/day (high dose: n = 7) for 10 days into devices implanted s.c. in Sprague-Dawley rats; noninfused devices transplanted for an identical period were used as controls (n = 5). Two days following the termination of VEGF infusion, devices were loaded with 20 microl of Lispro insulin (1 U/kg) and the kinetics of insulin release from the lumen of the device was assessed. Devices were then explanted and the number of blood vessels (capillary and noncapillary) was quantified using morphometry. High-dose vascular endothelial growth factor infusion resulted in two- to threefold more blood vessels around the device than that obtained with the noninfused devices and devices infused with low-dose vascular endothelial growth factor. This increase in the number of blood vessels was accompanied by a modest increase in insulin diffusion from the device in the high-dose vascular endothelial growth factor infusion group. We conclude that vascular endothelial growth factor can be used to improve blood vessel formation adjacent to planar membrane diffusion devices.

In situ electrochemical oxygen generation with an immunoisolation device

The viability and function of transplanted tissue encapsulated in immunobarrier devices is subject to oxygen transport limitation. In this study, we have designed and used an in situ electrochemical oxygen generator which decomposes water electrolyticaly to provide oxygen to the adjacent planer immunobarrier diffusion chamber. The rate of oxygen generation, which increases linearly with electrical current, was accurately controlled. A theoretical model of oxygen diffusion was also developed and was used to calculate the oxygen profiles in some of the experimental systems. In vitro culture experiments were carried out with beta TC3 cells encapsulated in titanium ring devices. The growth and viability of cells with or without in situ oxygen generation was studied. We found that under otherwise similar culturing conditions, the thickness of the cell layer and the viability of cells was the highest in devices cultured in stirred media with oxygen generation, even though the thickness had not reached the theoretically predicted value, and lowest in those unstirred and without oxygen generation.

Reversal of hyperglycemia in mice after subcutaneous transplantation of macroencapsulated islets

BACKGROUND

Macroencapsulated islets can reverse hyperglycemia in diabetic animals when transplanted i.p. or into the fat pad. The s.c. space is an attractive site for such transplantation because macrocapsules can be implanted with local anesthesia and be easily removed or reloaded with fresh islets.

METHODS

Immunoprotective 20 microl ported devices were transplanted under the skin of Streptozocin-diabetic nude mice. Devices were loaded with 1200 rat islets in culture medium or in alginate. Empty devices were implanted for 2 weeks and then loaded with islets. Normal mice and mice with islets transplanted under the renal capsule or under the skin were used as controls. Seven weeks after transplantation, an intraperitoneal glucose tolerance test (IPGTT) was performed, followed by implant removal.

RESULTS

Three weeks after transplantation, normal blood glucose levels were observed in all animals. Compared with those of normal controls, IPGTTs showed accelerated blood glucose clearance in mice transplanted with islets either within devices or beneath the kidney capsule. Fasted transplanted mice were hypoglycemic before glucose injection and 2 hr later. After removal of the implants, all recipient mice returned to hyperglycemia. Histological evaluation revealed viable islet cells and a network of close vascular structures outside the devices.

CONCLUSIONS

Macroencapsulated islets transplanted into the s.c. space were able to survive and regulate blood glucose levels in mice. The observed differences in glucose metabolism between normal and transplanted mice may be attributed to the site of transplantation and to the use of rat islets, which have a different set point for glucose induced insulin release.

Function and survival of macroencapsulated syngeneic islets transplanted into streptozocin-diabetic mice

BACKGROUND

Macroencapsulation is a strategy to protect transplanted islets from rejection and autoimmune attack. This study addresses questions about the survival and function of macroencapsulated syngeneic islets.

METHODS

Planar immunobarrier membrane diffusion devices were used for syngeneic islet transplantation. After being mixed with a 1% alginate solution, a total of 250, 500, 750 or 1000 islets were loaded into the devices, which were implanted into the epididymal fat pad(s) of streptozocin diabetic mice.

RESULTS

The success rate for restoration of normoglycemia at week 4 was highest for the recipients receiving two devices, each with 500 islets. Loading 750 or 1000 islets provided no improvement over loading 500 islets in a single device. Devices containing 250 islets were rarely successful. There was a striking tendency of transplants to either bring glucose levels into the near normal range or to fail with marked hyperglycemia. After an overnight fast at 1 and 4 weeks, but not at 12 weeks, hypoglycemia was found. The insulin content of devices from animals with normalized glucose values was higher than the insulin content in failed devices. Islet volume was maintained for 12 weeks, and fibrosis did not increase.

CONCLUSIONS

A relatively small mass of macroencapsulated islet tissue can survive and function well enough to normalize glucose levels for at least 12 weeks. Maintenance of glucose levels in the near-normal range seems to have a beneficial influence on graft success. The finding of fasting hypoglycemia raises important clinical questions about islet dysfunction. Important limitations in the requirements for islet packing density in macroencapsulation have been defined. New approaches for improving islet packing density must be developed to make diffusion-dependent macroencapsulation more practical.

Number and volume of islets transplanted in immunobarrier devices

Immunobarrier devices may prevent immune destruction of transplanted islets, but there are concerns about survival within such devices. Islets were transplanted in diffusion chambers that employed two laminated polytetrafluoroethylene membranes held together with titanium rings. Five hundred syngeneic mouse islets placed in devices were transplanted into the epididymal fat pads of streptozotocin (STZ) diabetic mice (B6AF1). After 2 wk the devices were removed. Sections were made parallel to the membrane surface. Eight to 13 systematically selected sections of each device were analyzed by planimetry to determine the area of the device space and of the islets within that space. From these data we estimated total volume of the device, volume of islets, and number of islets in a device. The data were segregated into two groups: group I (blood glucose less than 100 mg/dL 2 wk after implantation), and group II (over 150 mg/dL). The volume (mean +/- SE) of devices implanted for 2 wk was 2.1 +/- 0.4 microL in group I and 2.2 +/- 0.2 microL in group II. The islet volume and number within devices were 0.30 +/- 0.06 and 0.17 +/- 0.01 microL, or 340 +/- 50 and 230 +/- 20 islets in group I and group II, respectively. The volume of fibrous tissue in devices was about 0.50 microL. About 10% of the islet tissue had central necrosis. The beta cell volume in a membrane device needed for cure is comparable to that required with islets under the kidney capsule (0.25-0.80 microL). The mass of islets contained within membrane devices needed to cure diabetes is equivalent to that of a graft in an optimal transplant site such as under the kidney capsule.

Microcinematographic studies of flow patterns in the excised rabbit aorta and its major branches

Arterial fluid mechanics may play a role as a localizing factor for early atherosclerosis. Flow patterns in natural rabbit aortas rendered transparent were studied using a microcinematographic visualization technique. The aortic arch exhibited a single cell of clockwise-rotating helical secondary flow along the ventral and inner walls. Flow separation occurred proximal to the two arch branches with flow reversal proximal to the brachiocephalic artery. Sinusoidal flow rendered the helical motion more pronounced in systole, while the reverse flow zone periodically expanded and contracted. Steady flow in the abdominal aorta revealed streamlines which follow slow looping trajectories lateral to ostia before tracing helical paths into the branches. Flow separation was present along the dorsal wall of the aorta opposite the superior mesenteric artery. With the exception of the left renal artery, steady flow wall shear stresses were higher distal to ostia than proximal. Spatial gradients of wall shear stress were larger around branches than elsewhere. Similar to observed flow patterns, sites of enhanced macromolecular permeability, as observed previously in the normal rabbit aorta, follow a clockwise helical pattern in the arch and exhibit a distribution around ostia that correlates to some degree with regions of elevated shear stress gradients.

Measurement of flow rates through aortic branches in the anesthetized rabbit

The volumetric flow rates, mean and pulsatile, in the aorta and its major branches were measured in nonfed, anesthetized rabbits, using a transit time Doppler ultrasonic flowmeter. Anesthesia was maintained with isoflurane, and a vasodilator was applied topically during the measurements to avoid introducing additional flow resistance due to vasoconstriction. The cranial mesenteric and celiac arteries received the bulk of the aortic flow, (mean +/- SD) 29.5 +/- 6.6% and 23.3 +/- 5.8%, respectively, for mean flow. The brachiocephalic artery received as much as 14.7 +/- 3.2%, while each of the other branches received a considerably smaller fraction: 7.1 +/- 2.5% for the left subclavian artery, 6.2 +/- 2.6% and 5.1 +/- 2.2%, respectively, for the right and left renal arteries, and 6.0 +/- 2.5% for each of the two iliac arteries. Flow divisions were nearly the same in paired vessels. Peak pulsatile flow divisions were similar to their steady flow counterparts in the brachiocephalic, left subclavian, celiac, and cranial mesenteric arteries, but were smaller in the renal and iliac arteries, although the difference was not statistically significant. Reverse flow from one or more of the branches back into the aorta occurred in diastole in seven of eight rabbits studied.

Time course of membrane microarchitecture-driven neovascularization

The host response to a microporous material that induces neovascularization at the material-tissue interface was studied in terms of the number and types of cells invading the membrane, the degree of vascularization at the material-tissue interface, and the characteristics of the surrounding connective tissue as a function of time following implantation. Millipore-MF mixed esters of cellulose membranes with a nominal pore diameter of 8.0 microns were implanted subcutaneously into male Sprague-Dawley rats and explanted at 3, 5, 7, 10, 21 and 329 days post-implantation. Two samples from each of two devices at each implantation time were embedded in paraffin, sectioned to a thickness of 5 microns, and stained with haematoxylin and eosin for light microscopic observation. The density of cells in the membrane increased up to 7 days following implantation, then remained roughly constant through 21 days and decreased at the 329 day time point. The vascularity of the material-tissue interface increased up to 10 days and remained at this level even at 329 days post-implantation. The connective tissue was disorganized, loose and avascular at 3 days, resembled granulation tissue at 5 days, and underwent fibrous capsule formation and maturation starting at 7 days following implantation.

Synthesis of tumor necrosis factor alpha and interleukin-1 receptor antagonist, but not interleukin-1, by human mononuclear cells is enhanced by exposure of whole blood to shear stress

Extracorporeal circulation exposes blood to shear stress. In many studies, researchers reported effects of shear stress on morphology and function of various blood cells, but effects on cytokine synthesis have not been studied. The authors investigated the effect of shear stress on the synthesis of interleukin-1 beta, interleukin-1 alpha, tumor necrosis factor alpha, and interleukin-1 receptor antagonist by human peripheral blood mononuclear cells. Whole heparinized blood at room temperature was exposed to shear stresses of 50, 200, or 500 dyne/cm2 for 5 min or 30 sec, and to 980 dyne/cm2 for 5 sec. Peripheral blood mononuclear cells were then separated from sheared blood and cultured for 24 hrs with or without lipopolysaccharide or Staphylococcus epidermidis. Total (intra + extracellular) cytokine synthesis was measured by specific radioimmunoassay. Viability of cultured peripheral blood mononuclear cells, determined by trypan blue exclusion and lactate dehydrogenase release, was not significantly affected by shear stress. Shear stress without lipopolysaccharide or S. epidermidis stimulation did not affect synthesis of interleukin-1 or tumor necrosis factor alpha but did enhance synthesis of interleukin-1 receptor antagonist. Lipopolysaccharide- or S. epidermidis- induced synthesis of interleukin-1 was not significantly altered by shear stress. In contrast, lipopolysaccharide-induced tumor necrosis factor alpha synthesis increased with increasing shear stress and was significantly elevated over unsheared controls, whereas S. epidermidis-induced tumor necrosis factor alpha and lipopolysaccharide- or S. epidermidis-induced interleukin-1 receptor antagonist synthesis were not significantly enhanced by shear. Therefore, sublytic trauma, such as exposure to shear stress, affects in vitro responses of peripheral blood mononuclear cells to secondary stimuli.

Effect of hypoxia on insulin secretion by isolated rat and canine islets of Langerhans

The effect of pO2s reduced below physiological levels on GSIR by isolated islets of Langerhans was investigated with a microperifusion apparatus that provided control of pO2 and rapid dynamic response. Second-phase insulin secretion was reduced substantially by hypoxia. The response to lower pO2 was rapid and reversible. Although the steady, normoxic (pO2 = 142 mmHg) second-phase secretion rate varied widely from one islet preparation to another, the ratio of Sx to S142 for each preparation could be represented by a single curve that exhibited a continuous reduction with decreasing pO2. For rat islets perifused 1 day after isolation, the secretion rate was nearly 100% of the normoxic value at a pO2 of 60 mmHg, 50% at 27 mmHg (P50, the pO2 at which the S142 is reduced by 50%), and approximately 2% at 5 mmHg. Oxygen sensitivity of second-phase secretion rate declined after 1 wk of in vitro culture: P50 was 13 mmHg after 1 wk and remained at 10 mmHg after 2-5 wk of culture. Canine islets exhibited a P50 of 16 mmHg after 1 wk of culture. The reduction in insulin secretion is thought to be associated with the existence of pO2 gradients outside and inside the isolated islets, resulting in exposure of islet cells to low pO2 levels that decrease radially from the periphery to the core. We hypothesize that the effect of low pO2 on S is manifested through depletion of the energy stores of the beta-cells. The effect of hypoxia on S may be an important factor in some in vitro secretion studies and may play a critical role in the effectiveness of transplanted islets before their revascularization and of immunoisolated islet implantation devices.

Topographical mapping of sites of enhanced HRP permeability in the normal rabbit aorta

The spatial distribution of sites of enhanced permeability to the macromolecule horseradish peroxidase (HRP) in the normal rabbit aorta after one min circulation was studied using image analysis. These sites, referred to as "HRP spots," exhibit a nonuniform distribution that is qualitatively similar in all rabbits studied. The density of HRP spots is highest in the aortic arch, decreases distally, reaches a minimum in the lower descending thoracic aorta, and then increases again in the abdominal aorta. The region of highest spot density follows a clockwise helical pattern in the aortic arch and outside the arch occurs in streaks largely oriented in the bulk flow direction. The streaks in the abdominal aorta localize along the anatomical right lateral wall and occasionally along the left lateral wall proximal to the celiac artery and along the ventral wall between the celiac and superior mesenteric arteries. The density of spots is high in the immediate vicinity of aortic ostia with the most elevated density being distal to ostia in most cases. At a short distance from the ostium edge of the celiac and superior mesenteric branches the proximal density is comparably high, and no preferred spot orientation is observed around the brachiocephalic vessel. These results are consistent with an influence of localizing factors such as detailed hemodynamic phenomena and/or arterial wall structural and/or functional variations.

Permeability of dialyzer membranes to TNF alpha-inducing substances derived from water bacteria

Pro-inflammatory cytokine-inducing substances derived from cultured E. coli have previously been shown to pass across low-flux regenerated cellulosic dialyzer membranes. In the present study, a sterile filtrate of Pseudomonas maltophilia grown from standard bicarbonate dialysis fluid was used to test the permeability of various dialyzer membranes (regenerated cellulose, cellulose triacetate, polyacrylonitrile, polysulfone and polyamide) to TNF alpha-inducing bacterial substances. Pyrogen-free tissue culture medium (MEM) was recirculated for 60 minutes in the dialysate compartment of a closed-loop dialysis system, then P. maltophilia filtrate was added and recirculation was continued for a further hour. Samples from the dialysate (MEM) and the blood side (containing 10% human plasma in MEM) were incubated with donor mononuclear cells (MNC) for 18 hours and TNF alpha release was measured in MNC supernatants by radioimmunoassay. Five minutes after the addition of P. maltophilia filtrate, mean TNF alpha-inducing activity in the dialysate increased from (mean +/- SEM) 0.10 +/- 0.02 to 18.2 +2- 1.5 (ng/2.5 x 10(6) MNC/18 hr). TNF alpha-inducing activity in the blood side increased with regenerated cellulose from 0.10 +/- 0.01 to 4.57 +/- 1.55 (N = 8; P less than 0.001); with cellulose triacetate from 0.20 +/- 0.05 to 0.44 +/- 0.10 (N = 5; P less than 0.05), and with polyacrylonitrile from 0.10 +/- 0.02 to 1.16 +/- 0.45 (N = 5; P less than 0.03). No increased TNF alpha-inducing activity was observed in the blood side of polysulfone (N = 5) or polyamide dialyzers (N = 5).(ABSTRACT TRUNCATED AT 250 WORDS)

A microperifusion system with environmental control for studying insulin secretion by pancreatic tissue

A continuous flow reactor (perifusion system) was fabricated and tested for measuring the kinetics of insulin secretion from isolated pancreatic islets of Langerhans in response to step changes in the glucose concentration and oxygen partial pressure in the perfusate flowing around the islets. The system was capable of making rapid changes in perfusate glucose concentration and pO2, had rapid dynamic response for measuring the change in insulin secretion rate as a result of these changes in perfusate, and was suitable for studying very small volumes of tissue. Initial experiments with this system demonstrated that (1) the response of isolated rat islets to glucose stimulation was very fast, with the first phase peak occurring in as little as about 10 s, (2) bulk perfusate oxygen partial pressure levels of 30 mmHg or less reduced the second-phase insulin secretion rate in graded fashion, (3) the reduction in secretion rate began within 1 min following an oxygen partial pressure decrease, and (4) the reduction in secretion rate was reversible, with a burst of insulin secretion occurring during the first minute after partial pressure restoration.

Determining molecular weight distributions of antigen-antibody complexes by quasi-elastic light scattering

Physiological properties of soluble antigen-antibody (Ag-Ab) complexes depend in part on the size of the complexes. In previous work, the size distribution and structure of model Ag-Ab complexes were determined by electron microscopy. In this study, we used constrained regularization analysis of quasi-elastic light scattering data to estimate molecular weight distributions of model Ag-Ab complexes. A conformational model was necessary to determine appropriate correlations between molecular weight and diffusion coefficient, and to estimate particle structure factors. Porod-Kratky theory proved to be an adequate conformational model for these purposes. The molecular weight distributions determined by constrained regularization compared favorably with distributions obtained either by electron microscopy or by thermodynamic modeling.

Bioengineering in development of the hybrid artificial pancreas

The hybrid artificial pancreas for treatment of diabetes consists of insulin-secreting pancreatic tissue which is surrounded by a membrane that protects the tissue from rejection by the immune system following implantation. In this paper, we review the alternative therapeutic approaches for diabetes under study and then discuss the technical requirements that must be met by a hybrid device useful to humans. Previous work on intravascular and extravascular immunoisolation devices is reviewed from the standpoint of these requirements, and three critical unresolved issues are discussed: biocompatibility, oxygen supply limitations, and prevention of immune rejection.

Elution conditions and degradation mechanisms in long-term immunoadsorbent use

The limited life of immunoadsorbents used for the large-scale purification of biological macromolecules poses a significant limitation to the more widespread application of this technology. In this study, the binding activity of a monoclonal antibody (MAb) to bovine serum albumin (BSA) was measured as a function of pH, ionic strength, and varying concentrations of KSCN, ethylene glycol, or DMSO. Low pH (2.5) and 3 M KSCN each reduced the antibody binding constant below 6 x 10(5) L/mol, meeting criteria derived from a simple chromatographic model for identifying effective eluents. A panel of six MAb to BSA was exposed repeatedly to adsorption conditions and the two eluents. Four MAb lost less than 50% of their initial binding capacity over 100 cycles. The other two lost 75% of their initial capacity. One MAb was stable when exposed to low pH but lost binding capacity with KSCN. In all cases, the equilibrium constant was unchanged. The loss of capacity was also shown to be a strong function of antibody loading: at 14.5 mg/mL, 98% of the initial binding capacity of one MAb was lost within 40 cycles, versus 75% loss at 1 mg/mL. Antibody leakage and nonspecific adsorption of contaminants were not responsible for significant loss of antibody activity over time.

Size and structure of antigen-antibody complexes: thermodynamic parameters

The role of antigen-antibody (Ag-Ab) complexes in the immune response depends, in part, on the size of the complexes. Previously, we combined electron microscopy with classical and quasi-elastic light scattering to characterize the molecular weight distribution and the conformation of Ag-Ab complexes made from bovine serum albumin (BSA) and pairs of anti-BSA monoclonal antibodies at a single concentration and Ag:Ab molar ratio. In this report, the molecular weight distribution of Ag-Ab complexes was determined by classical light scattering at a single Ag:Ab ratio and over a range of concentrations, and binding of BSA to pairs of MAb was determined by radioimmunoassay at several Ag:Ab molar ratios. A thermodynamic model was developed for the equilibrium size distribution of Ag-Ab complexes formed between a pair of MAb, each with unique affinity and specificity, and an Ag containing a single epitope for each of the pair of MAb. The combined experimental data were used in conjunction with the model to determine the values of cyclization and polymerization constants. Successful determination of the parameters required data from both classical light scattering and electron microscopy. Cyclization constants were lower than those reported in other studies of Ag-Ab complexes; this may be attributable to our use of a protein Ag, as compared to a divalent hapten. In two out of three cases, cyclization constants increased with increasing number of Ab in the complex, in contrast to previous assumptions. The validity of the thermodynamic model was further shown by its ability, in combination with conformational and hydrodynamic model, to predict the hydrodynamic radius of the complexes over a wide range of experimental conditions.

In vitro and in vivo testing of an electrocatalytic glucose sensor

A prerequisite for the development of an implantable artificial pancreas is the availability of a stable, long-life glucose sensor. Platinum (Pt) catalyzed electrodes have been demonstrated in vitro to show high sensitivity to glucose and long cycle life but are more sensitive to co-reactants compared with enzymatic methods. The authors developed a special data processing method (compensated net charge ratio, or CNCR) in which the measured electrode response is very sensitive to glucose, completely insensitive to urea, and only moderately sensitive to amino acids. Other endogenous and exogenous co-reactants show only minor interferences. The CNCR method involves the determination of the ratio of net oxidation charge to total charge during one complete cycle of a cyclic voltammogram. Prototype electrodes tested in vitro in spiked plasma have shown typical sensitivities of greater than 2 x 10(-4) CNCR units per 1 mg/dl change in glucose concentration, with linear response up to 400 mg/dl. For in vivo testing, a modified 5 F vascular catheter with membrane covered surface mounted electrodes was used at a vena cava site in swine. Several sensor designs were tested in vivo, with sensitivities of 1-5 x 10(-4) CNCR units (mg/dl).

Low-density lipoprotein transport in blood vessel walls of squirrel monkeys

Transmural accumulations of low-density lipoprotein (LDL) were examined in the blood vessel walls of four squirrel monkeys. Vascular wall concentrations of LDL were measured using quantitative autoradiography after 125I-labeled LDL circulation for 30 min. Profiles of relative tissue concentration from different sections in the same region were similar to each other, and there was little animal-to-animal variation. Concentrations were highest near the luminal endothelium, lower near the medial-adventitial border, and lowest within the media. Profiles from different regions fell into three groups: 1) aortic samples had steep intimal concentration gradients and near-zero media concentrations; 2) the iliac, femoral, popliteal, and common carotid arteries had higher intimal concentrations than group 1 but had similar concentrations deep within the media; and 3) the cerebral and coronary arteries, inferior vena cava, and pulmonary artery had intimal concentrations that were similar to group 2, but the concentrations deep within the media were greater than either groups 1 or 2. Arterial bifurcation profiles from the inner wall and the outer walls were similar to each other and to profiles from the upstream and downstream areas. Out of 280 total sites examined, 15 examples of profiles with substantially increased concentrations near the luminal endothelium were found scattered throughout the cardiovascular system, demonstrating that there are focal regions throughout the cardiovascular system which have greatly increased 125I-LDL transendothelial permeability.

Effect of oxygen on isolated pancreatic tissue

One approach to insulin replacement therapy is transplantation of islets of Langerhans immunoisolated from host tissue by a semipermeable membrane. In this state, islets depend on diffusion of nutrients and wastes to and from the beta-cell to provide a suitable environment for survival and secretion. A perifusion system was constructed to test glucose-stimulated (100-300 mg/100 ml) insulin secretion from whole islets, or small (5-10 cell) aggregates, under controlled pO2. First phase insulin secretion from adult rat islets was unaffected by hypoxic levels of pO2, but second phase secretion was rapidly reduced at pO2 levels below 60 mmHg in the bulk media. Secretion from single-cell aggregates was unaffected until pO2 levels dropped to 12 mmHg, at which point secretion progressively decreased with falling pO2. A theoretical reaction/diffusion model was developed to correlate intraislet pO2 with reduced insulin secretion. Oxygen limited secretion was reversible, and not a result of decreased cell viability, as ascertained by both long-term static culture and trypan blue staining. Insulin secretion is more sensitive to hypoxia than is cell viability, in part because O2 uptake increases with glucose stimulation. These results indicate that O2 may be the limiting factor in the ability of immunoisolated islets to respond to blood glucose changes. We conclude that maintenance of a sufficiently high islet pO2 for maximal insulin secretion may be an important issue for graft design and implant site selection.

Is the rise in plasma beta-2-microglobulin seen during hemodialysis meaningful?

Beta 2-Microglobulin (beta 2M) plasma levels and levels of a second low-molecular-weight protein (myoglobin) were studied during a 2- to 4-hour sham dialysis period (no dialysate flow, no weight loss) and during a 4- to 5-hour hemodialysis (HD) with a Cuprophan capillary dialyzer. While no rise of the beta 2M or myoglobin levels occurred during sham dialysis, a rise of 22.1 +/- (SD) 8.5% (beta 2M) or 19.9 +/- 12.1% (myoglobin) was seen during regular HD. The increases of both molecules showed a significant correlation (r = 0.44; p less than 0.03). Both rises could not be completely abolished using correction factors for hemoconcentration. The rises occurred irrespectively of the dialysate buffer. The results suggest that neither the Cuprophan membrane nor the extracorporeal circuit were responsible for the rise of both molecules during HD. It seems more likely that changes of the extracellular volume and extra- to intracellular water shifts are involved and account for the majority of the rise. However, the possibility of minor increase in the extracellular mass of beta 2M or myoglobin cannot be excluded completely.

Measurement of 125I-low density lipoprotein uptake in selected tissues of the squirrel monkey by quantitative autoradiography

A recently developed technique of absolute quantitative light microscopic autoradiography of 125I-labeled proteins in biologic specimens was used to measure 125I-low density lipoprotein (125I-LDL) concentration levels in various tissues of the squirrel monkey after 30 minutes of in vivo LDL circulation. Liver and adrenal cortex exhibited high 125I-LDL concentrations, presumably because of binding to specific cell surface receptors and/or internalization in vascular beds with high permeability to LDL. High tissue concentrations of LDL were associated with the zona fasciculata and reticularis of the adrenal cortex and the interstitial cells of Leydig in the testis; significantly lower levels of 125I-LDL were observed in the adrenal medulla, the zona glomerulosa, and germinal centers of the testis. Contrary to previous reports, low 125I-LDL concentrations were observed throughout the gastrointestinal tract and in lymph nodes. In addition, multiple arterial intramural focal areas of high 125I-LDL concentrations were identified in arteries supplying the adrenal gland, lymph node, small bowel, and liver.

Size and structure of antigen-antibody complexes. Electron microscopy and light scattering studies

Size parameters of model antigen-antibody (Ag-Ab) complexes formed by the interaction of bovine serum albumin (BSA) and pairs of monoclonal anti-BSA antibodies (mAb) were evaluated by quasielastic light scattering, classical light scattering, and electron microscopy (EM). Mean values for the hydrodynamic radius, radius of gyration, and molecular weight were determined by light scattering. Detailed information regarding the molecular weight distribution and the presence of cycles or open chains was obtained with EM. Average molecular weights were calculated from the EM data, and the Porod-Kratky wormlike chain theory was used to model the conformational behavior of the Ag-mAb complexes. Ag-mAb complexes prepared from three different mAb pairs displayed significantly different properties as assessed by each of the techniques employed. Observations and size parameter calculations from EM photomicrographs were consistent with the results from light scattering. The differences observed between the mab pairs would not have been predicted by idealized thermodynamic models. These results suggest that the geometric constraints imposed by the individual epitope environment and/or the relative epitope location are important in determining the average size of complexes and the ratio of linear to cyclic complexes.

Quasi-elastic light scattering of antigen-antibody complexes

Many biological properties of immune complexes (IC) depend upon their size. Quasi-elastic light scattering (QLS) was used to measure a mean equivalent hydrodynamic radius (Rh) and variance of the distribution of model IC composed of bovine serum albumin (BSA) as antigen (Ag) and combinations of two or three well-characterized monoclonal antibodies (MAb) which bound noncompetitively to unique epitopes on BSA. With the molar ratio (X) of each MAb to Ag fixed, Rh increased with concn. Rh was maximal at equivalence (X = 0.5) with two MAb and at slight MAb excess (X = 0.67) with three MAb. The largest Rh with two MAb was about 200 A, and Rh was uniformly different amongst the three combinations of two MAb IC. The largest hydrodynamic radius of individual complexes which formed with two MAb was estimated to be about 400 A; even larger individual complexes were formed with three MAb. Size changes following alteration of solution concns were also followed with QLS. Kinetics of two MAb IC association were too fast to observe; dissociation following large dilution (40-fold) required 5-10 min to attain a new steady state, much less at small dilution. With three MAb, Rh dropped sharply in 5 min and became steady after 1-2 hr. These results suggest that conventional chromatographic and ultracentrifugation techniques for studying IC size, involving large dilution and long measurement time, provide misleading results. Association of three MAb produced a rapid initial increase of Rh in several min, followed by diverse behavior which depended upon concn. From high to low concn, these included (1) exponential growth of Rh with time and appearance of visible macroscopic particles; (2) metastable states for several hr followed by slow growth to large size over several days, leading to formation of particles; and (3) rapid growth to steady state conditions with no visible particles. This heretofore unobserved equilibrium and kinetic behavior of model IC in solution may be reflected in the behavior of more complex, naturally occurring IC.

Monoclonal antibodies to bovine serum albumin: affinity and specificity determinations

A panel of 12 monoclonal antibodies (MAb) to bovine serum albumin (BSA) was developed and characterized as to their physiochemical and immunological properties. Affinity constants of the MAb varied over a wide range from 10(5) to 10(8) M-1. MAb were assembled into several groups of non- or minimally interacting antibodies by analysis of competitive binding experiments, and BSA domain and subdomain specificities of the MAb were assigned by analysis of results of MAb binding to purified BSA fragments. Further fine specificity delineation was accomplished by examination of cross-reactivity patterns to several mammalian albumins. The data suggest that some of the low affinity MAb recognize sites on different portions of the BSA molecule, indicating that similar epitopes exist on different domains of the BSA molecule.

Endothelial cell perturbation and low-density lipoprotein. Quantitative autoradiography

The focal entry and accumulation of LDL within the arterial wall of the normal animal may represent an early stage in the development of the atherosclerotic plaque. Concentrations of LDL 10 to 100 times normal medial concentrations might be difficult to clear from the arterial wall, permitting accumulation of lipid. Elevated LDL concentrations, in proximity to smooth muscle cells, appear to stimulate SMC proliferation. High LDL concentrations might also enhance mononuclear cell adhesion to endothelium. Since LDL has a high affinity for heparin and heparin for growth factors, LDL accumulation may be a mechanism for the concentration of such materials in the intima. The observation of markedly enhanced macromolecular permeability foci could be related to several potential mechanisms of initiation of atherosclerosis. This observation is of particular note when the focal occurrence of atherosclerosis is considered. Although atherosclerosis is seen as a generalized thickening of the intima, it is the focal narrowing of the lumen that is often responsible for the stenosis which produces symptoms such as angina or myocardial infarction.

Absolute quantitative autoradiography of low concentrations of [125I]-labeled proteins in arterial tissue

We developed a method for absolute quantitative autoradiographic measurement of very low concentrations of [125I]-labeled proteins in arterial tissue using Kodak NTB-2 nuclear emulsion. A precise linear relationship between measured silver grain density and isotope concentration was obtained with uniformly labeled standard sources composed of epoxy-embedded gelatin containing glutaraldehyde-fixed [125I]-albumin. For up to 308-day exposures of 1 micron-thick tissue sections, background grain densities ranged from about two to eight grains/1000 micron 2, and the technique was sensitive to as little as about one grain/1000 micron 2 above background, which correspond to a radioactivity concentration of about 2 x 10(4) cpm/ml. A detailed statistical analysis of variability was performed and the sum of all sources of variation quantified. The half distance for spatial resolution was 1.7 micron. Both visual and automated techniques were employed for quantitative grain density analysis. The method was illustrated by measurement of in vivo transmural [125I]-low-density lipoprotein [( 125I]-LDL) concentration profiles in de-endothelialized rabbit thoracic aortic wall.

Analysis of membrane processes for blood purification

Physical phenomena play an important role in membrane processes for blood purification. They largely determine the separation performance of these devices and they interact with chemical and biological phenomena to determine their biocompatibility, or lack thereof, in the clinical setting. In the first part of this paper, analyses of physical phenomena which determine the separation and purification characteristics are reviewed for several processes, including hemodialysis, hemofiltration, combined hemodialysis and ultrafiltration, and membrane plasmapheresis with cross-flow microfiltration. Special attention is given to transport of high-molecular weight solutes in hemodialysis, for use in subsequent analyses, and to the factors which determine filtrate flux in membrane plasmapheresis, because recent findings in this area provide an understanding of filtration processes in general. The second part concerns the problem of biocompatibility, especially as manifested in renal prostheses. After reviewing some of the pathways to bioincompatibility, exploratory analyses are presented using relatively simple models. The objective of these analyses is to provide an initial quantitative framework for examining the likelihood of monocyte secretion of interleukin-1 being stimulated by various routes. Issues examined, for which illustrative calculations are presented, include (1) transport of endotoxin fragments across regenerated cellulose and other membranes, (2) anaphylatoxin C5a concentrations in conventional hemodialysis and (3) the effects of equilibrium and reaction phenomena, ultrafiltration, diffusive membrane permeation and membrane adsorption on the disposition of C5a which is generated at the membrane surface.

Local variation in arterial wall permeability to low density lipoprotein in normal rabbit aorta

Normal rabbits were injected intravenously with horseradish peroxidase (HRP) and 125I-labeled human low density lipoprotein (LDL), and the aortas were perfusion-fixed. Subsequent visualization of HRP in the aortas was produced by reaction of the tissue with diaminobenzidine and hydrogen peroxide. The luminal surface of the aortas showed many small punctate foci of brown reaction product to the HRP, which represented penetration of the HRP into the vessel wall. The foci were scattered over the luminal surface, and most of the focal areas were less than 1 mm in diameter. The concentration of LDL was up to 47 times greater in these focal areas than in surrounding noncolored regions not showing increased permeability to HRP. Small circumscribed foci of heightened permeability to LDL may predispose to the local accumulation of lipid and ultimately to the formation of atherosclerotic plaques.