Experimental Neurology 169, 13–22 (2001) doi:10.1006/exnr.2001.7631, available online at http://www.idealibrary.com on
Impairment of the Blood–Nerve and Blood–Brain Barriers in Apolipoprotein E Knockout Mice Stephanie M. Fullerton,* Gregory A. Shirman,† Warren J. Strittmatter,* ,† ,‡ ,§ and William D. Matthew* ,† ,‡ , ¶ *Department of Neurobiology, †Department of Medicine (Neurology), ¶Department of Pediatrics (Neonatal–Perinatal Research Institute), ‡The Deane Laboratory, and §The Joseph and Kathleen Bryan Alzheimer’s Disease Research Center, Duke University Medical Center, Durham, North Carolina 27710 Received July 31, 2000; accepted December 19, 2000
tion of a vital dye into the systemic blood circulation. A century later, the presence and physiological properties of the BBB, as well as a blood–nerve barrier (BNB), have been documented. These barriers function to regulate the passage of blood components, including water, proteins, sugars, and nutrients, into neural tissue. The most stringent of these barriers, the BBB, is highly impermeable to proteins and other macromolecules and is achieved by vascular endothelial cells of the brain. These endothelial cells have three specializations that are the essence of the BBB. First, a complex network of endothelial membrane protein interactions forms tight junctions and seals the spaces between adjacent endothelial cells preventing paracellular diffusion. Second, nonspecific transcytosis across endothelial cells is diminished due to a reduction of pinocytotic vesicles. Last, with the two routes of nonspecific influx of proteins and nutrients blocked by tight junctions and diminished transcytosis, brain endothelial cells upregulate specific transport mechanisms to maintain homeostasis within the brain. These three specializations of brain endothelial cells establish a barrier impermeable to nonspecific molecules and highly selective for required molecules. Another structure important to the BBB is the basement membrane. The subendothelial basement membrane is a matrix of collagen, laminin, and heparan sulfate proteoglycan that outlines the entire abluminal endothelial surface. The physical structure and ionic charge of the basement membrane may augment the BBB by trapping some molecules, depending on charge and size (13, 52). Although the BBB has been studied for over a century, the cells and molecules involved in the induction and maintenance of the BBB are not clearly understood. A major controversy surrounds the importance of astrocytes for induction of BBB properties. Brain capillaries are 99% enveloped on the abluminal side by specialized astrocytic processes called endfeet. Initial in vivo studies by Janzer and Raff inferred induction of endothelial BBB properties by astrocytes (36). This
Apolipoprotein E (apoE) is well characterized as a plasma lipoprotein involved in lipid and cholesterol metabolism. Recent studies implicating apoE in Alzheimer’s disease and successful recovery from neurological injury have stimulated much interest in the functions of apoE within the brain. To explore the functions of apoE within the nervous system, we examined apoE knockout (KO) mice. Previously, we showed that apoE KO mice have a delayed response to noxious thermal stimuli associated with a loss and abnormal morphology of unmyelinated fibers in the sciatic nerve. From these data, we hypothesized that apoE KO mice could have an impaired blood–nerve barrier (BNB). In this report, we demonstrate functionally impaired blood–nerve and blood– brain barriers (BBB) in apoE KO mice using immunofluorescent detection of serum protein leakage into nervous tissue as a diagnostic for decreased BNB and BBB integrity. Extensive extravasation of serum immunoglobulin G (IgG) is detected in the sciatic nerve, spinal cord, and cerebellum of apoE KO but not WT mice. In a subpopulation of apoE KO mice, IgG also extravasates into discrete cortical and subcortical locations, including hippocampus. Loss of BBB integrity was additionally confirmed by the ability of exogenously supplied Evans blue dye to penetrate the BBB and to colocalize with IgG immunoreactivity in CNS tissue. These observations support a role for apoE in maintaining the integrity of the BNB/BBB and suggest a novel relationship between apoE and neural injury. © 2001 Academic Press
Key Words: Alzheimer’s disease; apolipoprotein E (apoE); blood– brain barrier; blood–nerve barrier; knockout; Evans blue; cerebellum; sciatic nerve; spinal cord; immunoglobulin G.
The concept of a blood– brain barrier (BBB) was first developed in the late 19th century, with the observation that the brain remains unstained following injec13
0014-4886/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
FULLERTON ET AL.
FIG. 1. Immunocytochemistry (A and B) and Western blotting (C) reveal increased IgG in apoE KO sciatic nerve. IgG is normally restricted from the nerve endoneurium (region between individual nerve axons), as shown in the WT nerve (A). ApoE KO sciatic nerve demonstrates IgG immunoreactivity throughout the nerve endoneurium (B). IgG immunoreactivity in the perineurium (arrows) is evident in the WT but is increased in the apoE KO perineurium. The bright punctate regions of fluorescence (asterisks) in both the WT and the apoE KO nerve are blood vessels. (C) Elevated levels of IgG in apoE KO nerves is documented by Western blot analysis of sciatic nerve homogenates from five pairs of randomly chosen WT and apoE KO mice. Increased levels of both light chain (25 kDa) and heavy chains (50 – 60 kDa) are apparent in all the apoE KO mice versus WT controls. In all experiments, mice have been perfused with PBS prior to fixation or homogenization to remove serum IgG. Scale bar, 15 m.
study has been refuted by Holash et al. (32) and the current data support gradual formation of the BBB following brain vascularization that is induced by the developing neural environment, including neurons and/or astrocytes (8). Highlighted by this controversy are the differences between pial and intraparenchymal vessels. Unlike intraparenchymal capillaries, pial microvessels are not ensheathed by astrocytic endfeet but are closely juxtaposed to the glial limitans. Pial microvessels have been shown to have a population of tight junctions with a small intermembrane cleft (18) and to heterogeneously express a known BBB antigen (5). Nonetheless, electron microscopy studies using electron-dense tracers ranging from 900 kDa to 139 Da show that pial microvessels prevent paracellular penetration of these markers as well as brain capillaries (84). This study also showed, however, that segments of pial vessels have increased vesicular transport of proteins across the endothelium (84).
The peripheral nerve is similarly buffered from the blood and extracellular tissue fluid by two distinct barriers that together compose the BNB. Within the nerve endoneurium, as in the brain, capillary endothelial cells form tight junctions and limit free exchange of macromolecules with the blood. Although similar to the CNS BBB, the PNS capillary endothelia are more permeable (6, 43, 49, 64), allowing the passage of some proteins, including albumin, into the endoneurium (51, 79). This difference may be due to increased transport or less adherent tight junctions. The second cellular component of the BNB is the perineurium. The perineurium consists of multiple layers of fibroblast-derived cells (15) that sheath fascicles of peripheral axons. The perineurial barrier has many of the same specialized features as the capillary endothelial cells, including layers of basement membrane and tight junctions (6, 65). These features make the perineurium virtually impermeable to proteins and severely restrict diffusion of ions and small nutrients such as glucose. By minimizing the exposure of neural tissue to cellular and macromolecular components of the blood and by transporting essential proteins and molecules, the BBB and BNB maintain homeostasis within the nervous system. Disruption of the BBB and/or BNB occurs following traumatic neural injury and in a variety of diseases, including multiple sclerosis (MS), diabetes, AIDS dementia, and Alzheimer’s disease (AD) (22, 65). It is widely held that the changes associated with increased permeability of the BBB may contribute to the development and progression of neuroinflammatory diseases and to the severity of injury after neural trauma. Apolipoprotein E (apoE) is implicated in neural health and recovery from injury, yet the exact function(s) of apoE in the nervous system is unknown. Outside the liver, the brain has the highest levels of apoE expression (24) with synthesis and secretion of apoE occurring primarily in monocyte-derived cells and astrocytes (11, 24). ApoE immunoreactivity in the brain is particularly evident in astrocytic endfeet (11). Injury to the nervous system further increases expression and secretion of apoE (11, 69, 72). Receptors for apoE belong to an expanding family of single transmembrane-domain endocytotic receptors. Brain endothelial cells and astrocytes are known to express the prototype apoE receptor, low-density lipoprotein receptor (LDLR) (21, 61, 37), while neurons express the low-density receptor-related protein (LRP) (14, 9). In vitro, apoE affects neuronal survival (29), morphology (10, 33, 34, 53, 54), and adhesion (34); induces smooth muscle cell proliferation; and inhibits inflammatory responses of microglia (45, 47). Polymorphisms in the human apoE gene produce three common apoE isoforms, apoE2, apoE3, and apoE4. The gene dosage of the APOE4 allele is associated with increased risk and decreased age of onset of familial and late-onset AD
IMPAIRED BNB AND BBB IN APOE KO MICE
FIG. 2. IgG and Evans blue–albumin (EBA) complex extravasation are present in apoE KO but not WT spinal cord. (A and B) IgG immunoreactivity in corresponding sections of WT and apoE KO cervical spinal cord ventral horn. (C and D) The identical sections using optical filters for red fluorescence to visualize EBA localization. This section of apoE KO spinal cord shows IgG immunoreactivity (B) and EBA fluorescence (D) with a blood vessel visible at the center (arrow), demonstrating extravasation from the blood vessel into the adjacent neuropil. We found similar lesions using IgG immunofluorescence alone or in combination with EBA throughout all levels of the spinal cord in both white and gray matter in apoE KO mice only. Some of the extravasated IgG and EBA appears within motor neurons (arrowhead). Uptake of extravasated proteins by motor neurons has been previously reported (25). Scale bar, 10 m.
(67). APOE4 also increases the severity of MS (26) and the risk of poor neurological outcome following stroke (70), cardiopulmonary bypass (55, 74), closed head injury (35, 71, 75), and cerebral hemorrhage (3). Taken together these data suggest a role for apoE in neural health and recovery following injury. To explore the function of apoE within the nervous system, we examined the apoE knockout (KO) mouse. In a previous report, we showed that apoE KO mice had altered morphology and reduced numbers of unmyelinated axons within the sciatic nerve (28). Since the BNB is impaired in a variety of peripheral nerve disorders in humans, we hypothesized that apoE KO mice might have an impaired BNB. To test this hypothesis we examined apoE KO sciatic nerves for evidence of decreased BNB integrity. We report here extravasation of endogenous immunoglobulin G (IgG) in the sciatic nerve of apoE KO but not WT mice using both immunocytochemistry and Western blotting. Since the structure and function of the BNB are similar to those of the BBB we also examined the integrity of the apoE KO BBB. Using the presence of peripherally injected Evans blue and/or serum-derived IgG in neural tissue
as a diagnostic for decreased BBB integrity we find extensive extravasation of these two markers in apoE KO spinal cord and cerebellum. Parenchymal penetration of these markers is also observed in scattered cortical and subcortical regions in a subpopulation of apoE KO mice. From these data we suggest a role for apoE in the maintenance of these blood–tissue barriers. MATERIALS AND METHODS
Mice. WT (C57Bl/6J) and apoE KO (60) (10th generation backbred to C57Bl/6J) mice were purchased from The Jackson Laboratory and maintained in the Bryan Research Building vivarium at Duke University Medical Center. All procedures were conducted in accordance with NIH guidelines for the care and use of laboratory animals and were approved by the Duke University Animal Use Committee. Except where noted, all mice used for these experiments were males between the ages of 8 and 12 weeks. Pairs of male WT and KO mice were matched by age with differences of
FULLERTON ET AL.
less than 1 week. The number of mice examined for this study exceeded 30 matched pairs. Western blots. Sciatic nerves from saline-perfused, age-matched WT and apoE KO mice were dissected and frozen on dry ice. Nerves were weighed, homogenized with a ground-glass homogenizer in 10 l/mg tissue ice-cold buffer (8 M urea, 0.5% SDS, 2% ␤-mercaptoethanol, 1% NP-40, 50 mM Tris, pH 7.6, Boehringer Mannheim Protease Cocktail, and PMSF), sonicated, and centrifuged (10 4 g, 15 min, 4°C). Protein concentrations of the nerve supernatants were determined using the Pierce Protein Assay. Samples were diluted with Laemmli sample buffer (Bio-Rad) and 20 g of protein was loaded onto each well of a 12% SDS– acrylamide gel (Bio-Rad). Molecular weights were estimated from electrophoretic mobility of protein standards (Gibco BRL). Following SDS–PAGE, proteins were transferred to PVDF membrane. Membranes were fixed in methanol–acetic acid, stained with Coomassie blue, washed, blocked with 5% milk in Trisbuffered saline with 0.05% Tween 20, and probed with anti-IgG (1:1000 of HRP– goat anti-mouse IgG, heavy and light chain; Boehringer Mannheim). Membranes were washed, incubated in chemiluminescence reagent (ECL; Amersham), and exposed to X-ray film. Immunocytochemistry. Mice were anesthetized with 0.06 ml of ketamine cocktail (50 mg/ml ketamine, 2.6 mg/ml xylazine, and 0.5 mg/ml acepromazine) and transcardially perfused with 20 ml of phosphate-buffered saline (PBS) followed by 20 ml of 4% paraformaldehyde in 150 mM phosphate buffer, pH 7.4. Brain, spinal cord, and/or sciatic nerve were removed immediately following perfusion, postfixed for 1 h, and then placed in 30% sucrose in PBS for at least 48 h. Cryoprotected tissue was mounted on chucks with TissueTek OCT compound and cut into 20-m cryostat sections. Sections were transferred to Superfrost Plus slides (Fisher) and dried at room temperature. Slides were rinsed with PBS, blocked with 2% goat serum in PBS for 1 h, and then incubated for 1.5 h with fluorescently labeled (Alexa 488 or Oregon Green 488) goat anti-mouse IgG (H⫹L) (Molecular Probes) diluted 1:250 in 2% goat serum–PBS. Following antibody incubation, slides were washed three times with PBS, coated with Vectashield (Vector Laboratories), and coverslipped. Sections were then examined for IgG immunoreactivity using a fluorescence microscope. Evans blue injections. In primary experiments (i.e., Figs. 4A– 4C), three pairs of WT and apoE KO mice were given an ip injection of 1 ml of 4% w/v Evans blue in PBS (38). In subsequent experiments (i.e., Figs. 2, 3, 4D, and 4E), mice were given an iv injection of 200 l of 4% w/v Evans blue while under methoxyflurane anesthesia (Schering–Plough). We found no difference in the patterns of extravasation produced by ip versus iv injections and therefore adopted iv injections as stan-
dard protocol as this permitted reduced Evans blue dosage. Mice were sacrificed at 1 h postinjection. In one pair of mice, an ip injection of 1 ml 0.4% Evans blue circulated for 24 h prior to sacrifice (i.e., Fig. 5). Mice were perfused, dissected, and prepared for fluorescence microscopy exactly as described for immunocytochemistry. RESULTS
IgG immunoreactivity in apoE KO sciatic nerves. To assess BNB integrity, apoE KO and WT sciatic nerves were examined for infiltration by IgG, an abundant serum protein that is normally restricted from the nerve endothelium, using both immunocytochemistry (Figs. 1A and 1B) and Western blotting (Fig. 1C). In both assays, mice were perfused with PBS to minimize contamination from IgG in the blood. Figure 1 demonstrates extensive IgG immunoreactivity in apoE KO sciatic nerves compared to WT. Fluorescent light microscopy shows IgG immunoreactivity in both WT and apoE KO perineurium. Staining is more intense in apoE KO perineurium and additional immunoreactivity is observed throughout the apoE KO nerve. This IgG immunoreactivity pervades the entire endoneurium but does not localize to axoplasm or myelin. Western blot analysis of homogenized sciatic nerve probed for IgG also demonstrates increased IgG (heavy and light chains) in sciatic nerve from five apoE KO mice versus five age- and sex-matched WT mice. From these data, we conclude that the apoE KO mice have an impaired BNB. IgG immunoreactivity in apoE KO spinal cord. To determine the BBB integrity in the apoE KO mice we examined IgG immunoreactivity in apoE KO spinal cord. IgG extravasation was consistently observed in apoE KO spinal cord (Fig. 2B) with numerous discrete patches of IgG immunoreactivity throughout the length of the apoE KO spinal cord in both white and gray matter. IgG extravasation was not observed in WT control mice. In some apoE KO spinal cord sections, extravasation was particularly evident at the edges of the tissue and in the dorsal horn. Areas of intraparenchymal extravasation were typically 30 – 60 m in diameter and centered around a blood vessel. Extravasation in apoE KO spinal cord. To confirm the loss of BBB integrity in the apoE KO mice we used a second method to assess BBB permeability. Evans blue dye is an azo dye that binds with high affinity to albumin and has intense red fluorescence under rhodamine optics. Once Evans blue is introduced into the circulatory system, either through an ip or an iv injection, it binds serum albumin (19, 76). The Evans blue– albumin (EBA) complex (MW 69 kDa) then penetrates organs lacking blood–tissue barriers but is excluded from the CNS by the BBB. Thus, the presence of EBA in neural tissue, detected by fluorescence microscopy,
IMPAIRED BNB AND BBB IN APOE KO MICE
is evidence of increased permeability of the BBB. As shown in Fig. 2, EBA extravasates into apoE KO (Fig. 2D) but not WT (Fig. 2C) spinal cord. Furthermore, EBA colocalizes with IgG immunoreactivity and reveals the same patterns of extravasation in the spinal cord (Fig. 3). In this figure, IgG immunofluorescence (Fig. 3A) and EBA (Fig. 3B) demarcate a large region of extravasation that follows a blood vessel in the dorsal column. Double exposure of the film through both the red and the green fluorescence filters reveals the overlap of these two markers in yellow (Fig. 3C). Extravasation in apoE KO cerebellum. Intravenous injection of Evans blue reveals macroscopic evidence of BBB breakdown in apoE KO mice (Fig. 4A). The apoE KO cerebellum, but not the WT cerebellum, is stained blue from EBA penetrating the BBB. EBA penetration within the cerebellar neuropil is shown in the photomicrographs in Fig. 4. In WT cerebella (Figs. 4B and 4D), faint EBA demarcates pial and intraparenchymal blood vessels. In the apoE KO mice (Figs. 4C and 4E), however, EBA extravasates into the adjacent neuropil from pial vessels. EBA extravasation occurred primarily near pial vessels situated in the sulci of the cerebellum. In affected mice, large regions of extravasation were often immediately adjacent to regions without extravasation. For this reason, care was taken to examine many tissue sections spanning the breadth of both the WT and the apoE KO cerebellum. Cerebellar extravasation was detected in approximately half of the apoE KO mice and in none of the WT mice. We observed identical patterns of extravasation with IgG immunostaining (data not shown) both in the absence and in the presence of EBA. EBA in apoE KO hippocampus. Several of the apoE KO mice examined for IgG or EBA fluorescence showed faint extravasation in cortical and subcortical regions, consistent with Western blot analysis of saline-perfused brains demonstrating elevated levels of heavy and light chain IgG (data not shown). To show extravasation using fluorescence microscopy we allowed a low dose of Evans blue (0.4% w/v) to circulate for 24 h in a pair of WT and apoE KO mice. Figure 5 demonstrates EBA extravasation into apoE KO hippocampus using this protocol. The pial vessel within the hippocampal fissure appears to be the primary source of EBA extruding into the neuropil. DISCUSSION
These observations of EBA and IgG in apoE KO nervous tissue demonstrate that apoE KO mice have an impaired BBB and BNB. Extravasation of IgG is most evident in peripheral nerve, spinal cord, and cerebellum, with occasional evidence of leakage in cortical and subcortical regions. In the apoE KO sciatic nerve, IgG was observed throughout the endoneurium and is elevated above WT levels in Western blot analysis. In
the spinal cord, discrete regions of extravasation were found in all segments of the cord and in both white and gray matter. The cerebellum was the most affected region within the brain. In the cerebellum, extravasation occurred primarily from pial vessels and extended into the adjacent neuropil. These patterns of extravasation in the apoE KO CNS were observed using EBA alone or in combination with immunostaining for endogenous IgG. We did not find evidence of general permeability increases in the entire vascular system of apoE KO mice nor did we find evidence of EBA or IgG penetration into WT neural tissue. These data imply a novel role for apoE in maintaining BBB and BNB integrity. Consequences of IgG accumulation in neural tissue. Several peripheral neuropathies are associated with BNB defects. Thus, our finding of altered morphology and functional impairment of apoE KO unmyelinated sensory nerves (28) led us to propose that apoE KO mice may have an impaired BNB. In this study, we have shown that there is an impaired BNB in apoE KO mice by demonstrating increased IgG in peripheral nerve. We have not determined if there is a causal association between IgG accumulation and altered nerve function. It is known, however, that there are several potentially damaging consequences of immunoglobulin penetration into neural tissue. Immunoglobulins may identify neural proteins as foreign and initiate an autoimmune reaction causing complement fixation and immune-cell activation. Such a response can cause demyelination, disruption of neural transmission, and cell death (30, 56, 59, 73). Recent work also demonstrates that immunoglobulins have an intrinsic ability to produce reactive oxygen species by converting molecular oxygen to hydrogen peroxide (83). In the appropriate setting, this reaction would enhance oxidative killing of pathogens, but in a disease state, it could lead to oxidative cellular and tissue damage. On balance, iv injections of IgG are used therapeutically for a variety of autoimmune diseases, including MS and Guillian–Barre´ syndrome. Improved clearance of myelin debris, blockade of Fc receptors on activated immune-effector cells, neutralization of cytokines, and anti-idiotypic binding of autoimmune antibodies are the proposed beneficial effects of this treatment (1, 2, 44). These disparate findings emphasize the importance of subsequent work to determine the consequences of IgG in the brain, spinal cord, and peripheral nerves of apoE KO mice. Considerations of the apoE KO mouse. ApoE KO mice are especially vulnerable to neural trauma and pathogenic infection. In one ischemic stroke model, apoE KO mice have increased infarct volumes and higher mortality rates than WT controls (46). In closed head injury paradigms, apoE KO mice have decreased levels of antioxidant compounds (48) and impaired
FULLERTON ET AL.
FIG. 3. IgG and EBA fluorescence colocalize in apoE KO tissue sections. Three images of apoE KO spinal cord showing a capillary traversing the dorsal column are presented. Both IgG (A) and EBA (B) show extravasation from this vessel. Imaging this tissue using double exposure through both red and green fluorescence filters reveals the overlap of these two serum protein markers in yellow (C). The colocalization of EBA and IgG provides evidence that IgG within nervous tissue originates from the serum and that EBA is a highly sensitive method of demonstrating extravasation. Scale bar, 10 m.
functional recovery and neuronal loss (46). ApoE KO mice also have increased mortality and elevated levels of the proinflammatory cytokine TNF␣ following lipopolysaccharide injections (47) and infection with Klebsiella pneumoniae (20) and Listeria monocytogenes (66). Infection and inflammatory cytokines alter endothelial permeability (22). Thus, exposure to pathogens and environmental stressors may be particularly detrimental to apoE KO mice, since their ability to regulate the inflammatory response and to protect the nervous system from circulating pathogens, cytokines, and immune cells is innately impaired. The altered BNB and BBB integrity observed by us may be, therefore, either due directly to the absence of apoE or due to secondary metabolic effects. Variability in apoE KO mice. In humans, the APOE4 allele predisposes one to, but does not determine, the development of AD. Similarly, the apoE4 genotype does not alter the incidence of stroke or cardiopulmonary bypass but does affect neurological recovery following these traumas. These findings imply that additional environmental or genetic factors influ-
FIG. 4. Macroscopic and microscopic evidence of EBA penetration into apoE KO cerebellum. (A) EBA is evident in apoE KO cerebellum but not in WT. (B and C) Fluorescence micrograph of the cerebella depicted in A reveals extravasation of EBA from a pial vessel into adjacent neuropil in the apoE KO brain (C). The corresponding WT region has EBA fluorescence that is precisely limited to the vessel (B). (D and E) Evidence of similar extravasation from aged mice. Key: Arrows point to vessels situated within cerebellar sulci and also indicate both the angle and the direction of the sulci. Scale bar, 25 m.
IMPAIRED BNB AND BBB IN APOE KO MICE
FIG. 5. EBA infiltrates apoE KO but not WT hippocampus after 24-h in vivo incubation. Both images show sagittal sections of hippocampus. In the WT tissue (A), EBA is restricted to the blood vessel within the hippocampal fissure. In the apoE KO (B), EBA extravasates from this pial vessel into the hippocampus. Microvessels within the parenchyma also stain. Key: Arrows indicate the position of the hippocampal fissure. The dotted line marks the dentate gyrus. Scale bar, 25 m.
ence an individual’s chance of developing AD or having poor recovery following neural injury. The variability among apoE KO mice in the extent of BBB and BNB deficits also suggests that other determinants, such as environmental stressors or pathogens, contribute to the variation among the apoE KO mice. Another factor that affects apoE KO phenotype is age. The inability of these mice to properly metabolize cholesterol and lipid leads to elevated serum cholesterol levels and the formation of atherosclerotic lesions that develop progressively with age (63). To this date, no alterations in expression of the apoE receptors have been reported in apoE KO mice and normal cholesterol metabolism can be restored by genetically reintroducing hepatic apoE expression (77). To avoid complications associated with age-dependent changes in the apoE KO vascular system, the mice used for this study were young adults (8 –12 weeks). At this age, blood vessel thickness (62) and cerebral blood flow are comparable to those of WT (7). To determine the possible effect of age/cholesterol on the extent of BBB integrity we examined EBA penetration in three pairs of apoE KO and WT mice age ⬎10 months. We found no difference in the incidence or pattern of cerebral extravasation in aged versus young adult apoE KO mice (Figs. 4C and 4E). Regional difference in BBB permeability. Variation in the extent of extravasation throughout the apoE KO brain, spinal cord, and sciatic nerve suggests regional differences in the barrier properties of the endothelium throughout the nervous system. The BNB is more permeable than the BBB (43, 49, 64). The extensive extravasation of IgG we observed in apoE KO sciatic nerve suggests that the BNB may also be more suscep-
tible to changes in the systemic environment. Regional differences between brain and spinal cord BBB have also been documented (80, 81). While some of these differences can be attributed to active transport mechanisms, transcellular entry of proteins into the spinal cord could proceed through the dorsal roots, since sensory ganglia lack a blood–nerve or perineurial barrier (4). The striking difference between extravasation in the cerebellum versus the other brain regions may be associated with developmental differences in the BBB. The cerebellar BBB develops postnatally, after the rest of the BBB is already intact (12). The cerebellar BBB is also more sensitive to increases in proinflammatory cytokines. Transgenic mice overexpressing interleukin-6 do not develop a cerebellar BBB, although all other regions of the brain acquire a functional BBB (12). The adult rodent cerebellar BBB is also more permeable than other brain regions after a variety of injuries, including scrapie infection (78), hyperammonemia (85), the chemical toxin soman (58), experimental allergic encephalomyelitis (57), and forced swimming stress (68). These differences in the development and vulnerability of the rodent cerebellar BBB may explain the high degree of BBB impairment in the cerebellum of apoE KO mice. Alternatively, the susceptibility of the apoE KO cerebellum to protein extravasation may be associated with the increased presence of pial vessels within this tissue. As noted, some pial vessels, although maintaining bona fide tight junctions, have increased transcytosis of a variety of markers (84). The extensive folding of the cerebellum provides significant volume of pial vasculature that is absent from other brain regions (31) and may explain the extensive extravasation in the apoE KO cerebellum. This observation implies that apoE influences BBB permeability through interactions with transcellular transport mechanisms within endothelial cells. Such a mechanism is particularly intriguing given the established role of apoE in lipid and cholesterol transport. Other potential mechanisms of interaction between ApoE and the BNB/BBB. We are investigating several other hypotheses regarding the mechanism of apoE involvement with the BBB and BNB. ApoE could exert a direct effect on BBB/BNB endothelial cell tight junctions and basement membranes. Communication between endothelial cells and astrocytes may be important for tight junction formation and/or maintenance. Astrocytes produce and secrete apoE and increase their production of apoE after injury (11, 69, 72). Astrocytes and brain endothelial cells express LDLR (21, 37, 61). ApoE secreted from astrocytic endfeet may be part of a signaling cascade establishing tight junctions. Astrocytes also contribute to basement membrane production. Evidence suggest that the basement membrane
FULLERTON ET AL.
has a role in preventing protein extravasation into the parenchyma by binding some proteins depending on size and charge (13, 52). Endothelial cells throughout the nervous system are ensconced in a complex meshwork of basement membrane. ApoE has been identified in basement membrane (24) and binds numerous proteins including laminin (34) and heparan sulfate proteoglycan (16, 82), two primary components of the basement membrane. ApoE may bind extravasated proteins, ultimately preventing them from entering the neuropil, and/or aid in the clearance of extravasated proteins from the parenchyma. ApoE may also have indirect effects on the BNB and BBB. ApoE inhibits proliferation of lymphocytes (17) and suppresses inflammatory responses of microglia (45, 47). Since endothelial cell permeability changes in response to many proinflammatory cytokines, apoE may affect the BBB/BNB through interactions with the immune system. Relationship to human neural injury and Alzheimer’s disease. The association of apoE genotype with the inheritance of AD, severity of MS, and functional outcome following stroke, cardiopulmonary bypass, and closed head injury is also consistent with a role for apoE in the maintenance of the BBB. These acute and chronic neural traumas disrupt the cerebral vasculature and alter perfusion of the brain. Susceptibility to and extent of damage from these injuries are likely to depend on the health of the BBB. CNS vascular and BBB abnormalities have been extensively documented in AD brains (23, 27, 39 – 42, 50). Despite this evidence, the causal relationship of these vascular and BBB defects to the etiology of AD is controversial. Linking apoE, a protein associated with AD, to the proper function of the BBB may provide insight into the connection between vascular and AD pathology. SUMMARY
In this study we found extensive extravasation of serum proteins into apoE KO sciatic nerve, spinal cord, and cerebellum and occasional extravasation into cortex and subcortex, providing evidence of impaired BNB and BBB in apoE KO mice. The functional consequences of serum protein extravasation on apoE KO neural function are unknown, although we have previously demonstrated impaired sensory axons in apoE KO sciatic nerves. Our data also show variability in the extent and localization of BBB deficits in apoE KO mice, suggesting that multiple factors, genetic and/or environmental, are involved. Taken together, our findings present a novel relationship between apoE and neural health and injury by implicating apoE as a key component of a functional BBB/BNB.
ACKNOWLEDGMENTS This work was supported by funding from Glaxo Wellcome, The Deane Laboratory, The Zeist Foundation, and the Duke University Medical Center Neonatal–Perinatal Research Institute. We thank Dr. Catherine Gutman for helpful ideas throughout all stages of this work and Jeffrey Mance for excellent technical assistance.
Abe, Y., A. Horiuchi, M. Miyake, and S. Kimura. 1994. Anticytokine nature of natural human immunoglobulin: One possible mechanism of intravenous immunoglobulin therapy. Immunol. Rev. 139: 5–19. Achiron, A., Y. Barak, M. Goren, U. Gabbay, S. Miron, Z. Rotstein, S. Noy, and I. Sarova-Pinhas. 1996. Intravenous immunoglobulin in multiple sclerosis: Clinical and neuroradiologic results and implications for possible mechanisms of action. Clin. Exp. Immunol. 104(Suppl. 1): 67–70. Alberts, M. J., C. Graffagnino, C. McClenny, D. DeLong, W. J. Strittmatter, A. M. Saunders, and A. D. Roses. 1995. ApoE genotype and survival from intracerebral hemorrhage. Lancet 346: 575. Allen, D. T., and J. A. Kiernan. 1994. Permeation of proteins from the blood into peripheral nerves and ganglia. Neuroscience 59: 755–764. Allt, G., and J. G. Lawrenson. 1997. Is the pial microvessel a good model for blood– brain barrier studies? Brain Res. Rev. 24: 67–76. Allt, G., and J. G. Lawrenson. 2000. The blood–nerve barrier: Enzymes, transporters and receptors—A comparison with the blood– brain barrier. Brain Res. Bull. 52: 1–12. Bart, R. D., H. Sheng, D. T. Laskowitz, R. D. Pearlstein, and D. S. Warner. 1998. Regional cerebral bloodflow in apolipoprotein E-deficient and wild-type mice during focal cerebral ischemia. NeuroReport 9: 2615–20. Bauer, H. C., and H. Bauer. 2000. Neural induction of the blood– brain barrier: Still an enigma. Cell. Mol. Neurobiol. 20: 13–28. Beffert, B., M. Danik, P. Krzywkowski, C. Ramassamy, F. Berrada, and J. Poirer. 1998. The neurobiology of apolipoproteins and their receptors in the CNS and Alzheimer’s disease. Brain Res. Rev. 27: 119 –142. Bellosta, S., B. P. Nathan, M. Orth, L. M. Dong, R. W. Mahley, and R. E. Pitas. 1995. Stable expression and secretion of apolipoprotein E3 and E4 in mouse neuroblastoma cells produces differential effects on neurite outgrowth. J. Biol. Chem. 270: 27063–27017. Boyles, J. K., R. E. Pitas, E. Wilson, R. W. Mahley, and J. M. Taylor. 1985. Apolipoprotein E associated with astrocytic glia of the central nervous system and with nonmyelinating glia of the peripheral nervous system. J. Clin. Invest. 76: 1501–1513. Brett, F. M., A. P. Mizisin, H. C. Powell, and I. L. Campbell. 1995. Evolution of neuropathologic abnormalities associated with blood– brain barrier breakdown in transgenic mice expressing interleukin-6 in astrocytes. J. Neuropathol. Exp. Neurol. 54: 766 –775. Brightman, M. W., and M. Kaya. 2000. Permeable endothelium and the interstitial space of the brain. Cell. Mol. Neurobiol. 20: 111–130. Bu, G., E. A. Maksymovitch, J. M. Nerbonne, and A. L. Schwartz. 1995. Expression and function of the low density lipoprotein receptor-related protein (LRP) in mammalian central nervous system. J. Biol. Chem. 269: 18521–18528.
IMPAIRED BNB AND BBB IN APOE KO MICE 15.
Bunge, M. B., P. M. Wood, L. B. Tynan, M. L. Bates, and J. R. Sanes. 1989. Perineurium originates from fibroblasts: Demonstration in vitro with a retroviral marker. Science 43: 229 –231. Cardin, A. D., N. Hirose, D. T. Blankenship, R. L. Jackson, J. A. Harmony, D. A. Sparrow, and J. T. Sparrow. 1986. Binding of a high reactive heparin to human apolipoprotein E: Identification of two heparin-binding domains. Biochem. Biophys. Res. Commun. 134: 783–789. Cardin, A. D., T. L. Bowlin, and J. L. Krstenansky. 1988. Inhibition of lymphocyte proliferation by synthetic peptides homologous to human plasma apolipoproteins B and E. Biochem. Biophys. Res. Commun. 154: 741–745. Casella, J. P., J. G. Lawrenson, and J. A. Firth. 1997. Development of endothelial paracellular clefts and their tight junctions in the pial microvessels of the rat. J. Neurocytol. 26: 567–575. Clasen, R. A., S. Pandolfi, and G. M. Hass. 1970. Vital staining, serum albumin and the blood– brain barrier. J. Neuropathol. Exp. Neurol. 29: 266 –284. de Bont, N., M. G. Netea, P. N. Demacker, I. Verschueren, B. J. Kullberg, K. W. van Dijk, J. W. M. van der Meer, and A. Stalenhoef. 1999. Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae infection. J. Lipid Res. 40: 680 – 685. Dehouck, B., M.-P. Dehouck, J. C. Fruchart, and R. Cecchelli. 1994. The upregulation of the low density lipoprotein receptor at the blood– brain barrier: Intercommunication between brain capillary endothelial cells and astrocytes. J. Cell Biol. 126: 465– 473. De Vries, H., J. Kuiper, A. G. De Boer, T. J. C. Van Berkel, and D. D. Breimer. 1997. The blood– brain barrier in neuroinflammatory disease. Pharmacol. Rev. 49: 143–152. Dorheim, M. A., W. R. Tracey, J. S. Pollock, and P. Grammas. 1994. Nitric oxide synthase activity is elevated in brain microvessels in Alzheimer’s disease. Biochem. Biophys. Res. Commun. 205: 659 – 665. Elshourbagy, N. A., W. S. Liao, R. W. Mahley, and J. M. Taylor. 1985. Apolipoprotein E mRNA is abundant in the brain and adrenals, as well as in the liver, and is present in other peripheral tissues of rats and marmosets. Proc. Natl. Acad. Sci. USA 8: 203–207. Fabian, R. H. 1988. Uptake of plasma IgG by CNS motoneurons: Comparison of antineuronal and normal IgG. Neurology 38: 1755–1780. Fazekas, F., S. Strasser-Fuchs, H. Schmidt, C. Enzinger, S. Ropele, A. Lechner, E. Flooh, R. Schmidt, and H.-P. Hartung. 2000. Apolipoprotein E genotype related difference in brain lesions of multiple sclerosis. J. Neurol. Neurosurg. Psychiatry 69: 25–28. Fischer, V. W., A. Siddiqi, and Y. Yusufaly. 1990. Altered angioarchitecture in selected areas of brains with Alzheimer’s disease. Acta Neuropathol. 79: 672– 679. Fullerton, S. M., W. J. Strittmatter, and W. D. Matthew. 1998. Peripheral sensory nerve defects in apolipoprotein E knockout mice. Exp. Neurol. 153: 156 –163. Gutman, C. R., W. J. Strittmatter, K. H. Weisgraber, and W. D. Matthew. 1997. Apolipoprotein E binds to and potentiates the biological activity of ciliary neurotrophic factor. J. Neurosci. 17: 6114 – 6121. He, X. P., M. Patel, K. D. Whitney, S. Janumpalli, A. Tenner, and J. O. McNamara. 1998. Glutamate GluR3 antibodies and the death of cortical cells. Neuron 20: 153–163. Holash, J. A., K. Sugamori, and P. A. Stewart. 1990. The difference in vascular volume between the cerebrum and cerebellum is the pia mater. J. Cereb. Blood Flow Metab. 10: 432– 434.
Holash, J. A., D. M. Noden, and P. A. Stewart. 1993. Reevaluating the role of astrocytes in blood– brain barrier induction. Dev. Dyn. 197: 14 –25. Holtzman, D. M., R. E. Pitas, J. Kilbridge, B. Nathan, R. W. Mahley, G. Bu, and A. L. Schwartz. 1995. Low density lipoprotein receptor-related protein mediates apolipoprotein E-dependent neurite outgrowth in a central nervous system-derived neuronal cell line. Proc. Natl. Acad. Sci. USA 92: 9480 –9484. Huang, D. Y., K. H. Weisgraber, W. J. Strittmatter, and W. D. Matthew. 1995. Interaction of apolipoprotein E with laminin increases neuronal adhesion and alters neurite morphology. Exp. Neurol. 136: 251–257. Jordan, B. D., N. Relkin, L. D. Ravdin, A. R. Jacobs, A. Bennett, and S. Gandy. 1997. Apolipoprotein E epsilon-4 associated with chronic traumatic brain injury in boxing. JAMA 278: 136 –140. Janzer, R. C., and M. C. Raff. 1987. Astrocytes induce blood– brain barrier properties in endothelial cells. Nature 325: 253– 257. Jung-Testas, I., H. Weintraub, D. Dupuis, B. Eychenne, E. E. Baulieu, and P. Robel. 1992. J. Steroid Biochem. Mol. Biol. 42: 597– 605. Kakinuma, Y., H. Hama, F. Sugiyama, K.-I. Yagami, K. Goto, K. Murakami, and A. Fukamizu. 1998. Impaired blood– brain barrier function in angiotensinogen deficient mice. Nat. Med. 4: 1078 –1080. Kalaria, R. N., and P. Hedera. 1995. Differential degeneration of the cerebral microvasculature in Alzheimer’s disease. NeuroReport 6: 477– 480. Kalaria, R. N., and A. B. Pax. 1995. Increased collagen content of cerebral microvessels in Alzheimer’s disease. Brain Res. 705: 349 –352. Kalaria, R. N. 1996. Cerebral vessels in aging and Alzheimer’s disease. Pharmacol. Ther. 72: 193–214. Kalaria, R. N., D. L. Cohen, and D. R. Premkumar. 1996. Apolipoprotein E alleles and brain vascular pathology in Alzheimer’s disease. Ann. N.Y. Acad. Sci. 777: 266 –270. Kiernan, J. A. 1996. Vascular permeability in the peripheral autonomic and somatic nervous systems: Controversial aspects and comparisons with the blood– brain barrier. Microsc. Res. Tech. 35: 122–136. Kuhlmann, T., and W. Bruck. 1999. Immunoglobulins induce increased myelin debris clearance by mouse macrophage. Neurosci. Lett. 275: 191–194. Laskowitz, D. T., S. Goel, E. R. Bennet, and W. D. Matthew. 1997. Apolipoprotein E suppresses glial secretion of TNF␣. J. Neuroimmunol. 76: 70 –74. Laskowitz, D. T., H. Sheng, R. Bart, K. Joyner, A. D. Roses, and D. Warner. 1997. Apolipoprotein E deficient mice have increased susceptibility to focal cerebral ischemia. J. Cereb. Blood Flow Metab. 17: 753–758. Laskowitz, D. T., W. D. Matthew, E. R. Bennett, D. Schmechel, M. H. Herbstreith, S. Goel, and M. K. McMillian. 1998. Endogenous apolipoprotein E suppresses LPS-stimulated microglial nitric oxide production. NeuroReport 9: 615– 618. Lomnitski, L., R. Kohen, Y. Chen, E. Shohami, V. Trembolver, T. Vogel, and D. Michaelson. 1997. Reduced levels of antioxidants in the brains of apolipoprotein E-deficient mice following closed head injury. Pharmacol. Biochem. Behav. 56: 669 – 673. Malmgren, L. T., and Y. Olsson. 1980. Difference between the peripheral and central nervous system in permeability to sodium fluorescein. J. Comp. Neurol. 191: 103–117. Mancardi, G. L., F. Perdelli, C. Rivano, A. Leonardi, and O. Bugiani. 1980. Thickening of the basement membrane of cortical capillaries in Alzheimer’s disease. Acta Neuropathol. (Berlin) 49: 79 – 83.
FULLERTON ET AL. Mata, M., J. Staple, and D. J. Fink. 1987. The distribution of serum albumin in rat peripheral nerve. J. Neuropathol. Exp. Neurol. 46: 485– 494. Muldoon, L. L., M. A. Pagel, R. A. Kroll, S. Roman-Goldstein, R. S. Jones, and E. A. Neuwelt. 1999. A physiological barrier distal to the anatomic blood– brain barrier in a model of transvascular delivery. Am. J. Neuroradiol. 20: 217–222. Nathan, B. P., S. Bellosta, D. A. Sanan, K. H. Weisgraber, R. W. Mahley, and R. E. Pitas. 1994. Differential effects of apolipoproteins E3 and E4 on neuronal growth in vitro. Science 264: 850 – 852. Nathan, B. P., K.-C. Chang, S. Bellosta, E. Brisch, N. Ge, R. W. Mahley, and R. E. Pitas. 1995. The inhibitory effect of apolipoprotein E4 on neurite outgrowth is associated with microtubule depolymerization. J. Biol. Chem. 270: 19791–19799. Newman, M. F., N. D. Croughwell, J. A. Blumenthal, E. Lowry, W. D. White, W. Spillane, R. D. Glower, L. R. Smith, and E. P. Mahanna. 1995. Predictors of cognitive decline after cardiac operation. Ann. Thoracic Surg. 59: 1326 –1330. Noseworthy, J. H., C. Lucchinetti, M. Roderiguez, and B. G. Weinshenker. 2000. Medical progress: Multiple sclerosis. N. Engl. J. Med. 343: 938 –952. Paul, C., and C. Bolton. 1995. Inhibition of blood– brain barrier disruption in experimental allergic encephalomyelitis by shortterm therapy with dexamethasone or cyclosporin A. Int. J. Immunopharmacol. 17: 497–503. Petrali, J. P., D. M. Maxwell, D. E. Lenz, and K. R. Mills. 1981. Effect of an anticholinesterase compound on the ultrastructure and function of the rat blood– brain barrier: A review and experiment. J. Submicrosc. Cytol. Pathol. 23: 331–338. Piddlesden, S. J., H. Lassman, F. Zimprich, B. P. Morgan, and C. Linington. 1992. The demyelinating potential of antibodies to myelin oligodendrocytes glycoprotein is related to their ability to fix complement. Am. J. Pathol. 143: 555–564. Piedrahita, J. A., S. H. Zhang, J. R. Hagaman, P. M. Oliver, and N. Maeda. 1992. Generation of mice carrying a mutant apolipoprotein E gene inactivated by gene targeting in embryonic stem cells. Proc. Natl. Acad. Sci. USA 89: 4471– 4475. Pitas, R. E., J. K. Boyles, S. H. Lee, D. Foss, and R. W. Mahley. 1987. Astrocytes synthesize apolipoprotein E and metabolize apolipoprotein E lipoproteins. Biochem. Biophys. Acta 917: 148 –161. Plump, A. S., J. D. Smith, T. Hayek, K. Aalto-Setala, A. Walsh, J. G. Verstuyft, E. M. Rubin, and J. L. Breslow. 1992. Severe hypercholesterolemia and atherosclerosis in apolipoprotein Edeficient mice created by homologous recombination in ES cells. Cell 71: 343–353. Plump, A. S., and J. L. Breslow. 1995. Apolipoprotein E and the apolipoprotein E-deficient mouse. Annu. Rev. Neurosci. 15: 495–518. Poduslo, J. F., G. F. Curran, and C. T. Berg. 1994. Macromolecular permeability across the blood–nerve and blood– brain barriers. Proc. Natl. Acad. Sci. USA 91: 2705–5709. Rechthand, E., and S. I. Rapoport. 1987. Regulation of the microenvironment of the peripheral nerve: Role of the blood– nerve barrier. Prog. Neurobiol. 28: 303–343. Roselaae, S. E., and A. Daugherty. 1998. Apolipoprotein E-deficient mice have impaired innate immune responses to Listeria monocytogenes in vivo. J. Lipid Res. 39: 1740 –1743. Saunders, A. M., W. J. Strittmatter, D. Schmechel, P. GeorgeHyslop, M. Pericek-Vance, S. Joo, B. Rosi, J. Gusella, D. Crapper-MacLachlan, M. Alberts, et al. 1993. Association of apolipoprotein E allele E4 with late-onset familial and sporadic Alzheimer’s disease. Neurology 43: 1467–1472.
Sharma, H. S., J. Cervos-Navarro, and P. K. Dey. 1991. Increased blood– brain barrier permeability following acute shortterm swimming exercise in conscious normotensive young rats. Neurosci. Res. 10: 211–221. Skene, J. H. P., and E. M. Shooter. 1983. Denervated sheath cells secrete a new protein after nerve injury. Proc. Natl. Acad. Sci. USA 80: 4169 – 4173. Slooter, A. J., M. X. Tang, C. M. van Duijin, Y. Stern, A. Ott, K. Bell, M. M. Breteler, C. Van Broeckhoven, T. K. Tatemichi, B. Tycko, A. Hofman, and R. Mayeux. 1997. Apolipoprotein E epsilon-4 and the risk of dementia with stroke. A population based investigation. JAMA 277: 818 – 821. Sorbi, S., B. Nacmias, and S. Piacentini. 1996. Apolipoprotein E genotypes and outcome after post-traumatic coma. Neurology 46: A307. Stoll, G., and H. W. Muller. 1986. Macrophages in the peripheral nervous system and astroglia in the central nervous system of rat commonly express apolipoprotein E during development but differ in their response to injury. Neurosci. Lett. 72: 233– 238. Storch, M. K., S. Piddlesden, M. Haltia, M. Iivanainen, P. Morgan, and H. Lassmann. 1998. Multiple sclerosis: In situ evidence for antibody and complement mediated demyelination. Ann. Neurol. 43: 465–571. Tardiff, B. E., M. F. Newman, A. M. Saunders, W. J. Strittmatter, J. A. Blumenthal, W. D. White, N. Croughwell, R. D. Davis, A. D. Roses, J. G. Reves, et al. 1997. Preliminary report of a genetic basis for cognitive decline after cardiac operations. Ann. Thoracic Surg. 64: 715–720. Teasdale, G. M., J. A. R. Nicoll, G. Murray, and H. Fiddes. 1997. Association of apolipoprotein E polymorphisms with outcome after head injury. Lancet 350: 1069 –1071. Tschirgi, R. D. 1950. Protein complexes and the impermeability of the blood– brain barrier to dyes. Am. J. Physiol. 163: 756. Tsukamoto, K., R. Tangirala, S. H. Chun, E. Pure, and D. J. Rader. 1999. Rapid regression of atherosclerosis induced by liver-directed gene transfer of apoE in apoE-deficient mice. Arterioscler. Thromb. Vasc. Biol. 19: 2162–2170. Vorbrodt, A. W., D. H. Dobrogowska, M. Tarnawski, H. C. Meeker, and R. I. Carp. 1997. Immunocytochemical evaluation of blood– brain barrier to endogenous albumin in scrapie-infected mice. Acta Neuropathol. (Berlin) 93: 341–348. Weerasuriya, A., G. L. Curran, and J. F. Poduslo. 1989. Blood– nerve transfer of albumin and its implications for the endoneurial microenvironment. Brain Res. 494: 114 –121. Weihong, W., W. A. Banks, and A. J. Kastin. 1997. Permeability of the blood– brain and blood–spinal cord barriers to neurotrophins. J. Neuroimmunol. 76: 105–111. Weihong, W., W. A. Banks, and A. Kastin. 1998. Permeability of the blood– brain barrier to neurotrophins. Brain Res. 788: 87– 94. Weisgraber, K. H., S. C. Rall, Jr., R. W. Mahley, R. W. Milne, Y. L. Marcel, and J. T. Sparrow. 1986. Human apolipoprotein E. Determination of the heparin binding sites of apolipoprotein E3. J. Biol. Chem. 261: 2068 –2076. Wentworth, A. D., L. H. Jones, P. Wentworth, Jr., K. D. Janda, and R A. Lerner. 2000. Antibodies have the intrinsic capacity to destroy antigens. Proc. Natl. Acad. Sci. USA 97: 10930 –10935. Westergaard, E., and M. W. Brightman. 1973. Transport of proteins across normal cerebral arterioles. J. Comp. Neurol. 152: 17– 44. Ziylan, Y. Z., G. Uzum, G. Bernard, A. S. Diler, and J. M. Bourre. 1993. Changes in the permeability of the blood– brain barrier in acute hyperammonemia. Effect of dexamethasone. Mol. Chem. Neuropathol. 20: 203–218.