Search

The blood-brain barrier (BBB) to water-soluble drugs and macromolecules can be opened in vivo by infusing a hypertonic solution of arabinose or mannitol into the carotid artery for 30 sec. Opening involves widening of tight junctions between endothelial cells of the cerebrovasculature and is mediated by endothelial cell shrinkage, vascular dilatation associated with removal of water from brain, and modulation of the contractile state of the endothelial cytoskeleton and junctional proteins by increased intracellular calcium. A 10-fold increase in BBB permeability to intravascular substances, lasting about 10 min following osmotic exposure, reflects both increased diffusion and bulk fluid flow from blood into brain. Furthermore, recent evidence indicates that the duration of peak BBB opening can be extended beyond 30 min, by pre-treatment with a Na(+)/Ca(2+) channel blocker. In experimental animals, the osmotic method has been used to grant wide access to brain of water-soluble drugs, peptides, antibodies, boron compounds for neutron capture therapy, viral vectors for gene therapy and enzymes. Ongoing multi-centre clinical studies suggest that the method, when used with intra-arterially administered anticancer drugs, can prolong survival in patients with malignant brain tumours, with minimal morbidity. However, controlled clinical trials are critical to see if the osmotic procedure with intra-arterial drugs enhances survival in brain tumour patients compared with intra-arterial drug alone.

1. Osmotic opening of the blood-brain barrier by intracarotid infusion of a hypertonic arabinose or mannitol solution is mediated by vasodilatation and shrinkage of cerebrovascular endothelial cells, with widening of the interendothelial tight junctions to an estimated radius of 200 A. The effect may be facilitated by calcium-mediated contraction of the endothelial cytoskeleton. 2. The marked increase in apparent blood-brain barrier permeability to intravascular substances (10-fold for small molecules) following the osmotic procedure is due to both increased diffusion and bulk fluid flow across the tight junctions. The permeability effect is largely reversed within 10 min. 3. In experimental animals, the osmotic method has been used to grant wide access to the brain of water-soluble drugs, peptides, antibodies, boron compounds for neutron capture therapy, and viral vectors for gene therapy. The method also has been used together with anticancer drugs to treat patients with metastatic or primary brain tumors, with some success and minimal morbidity.

The cholinergic agonist, arecoline, was used to examine the effects of cholinergic stimulation upon incorporation of radiolabeled arachidonic acid from blood into cerebral microvessels of awake rats. Animals received a single i.p. injection of arecoline (1 mg/kg) followed 3 to 5 minutes later by a 5 minute intravenous infusion of [1-14C]arachidonic acid (AA) (170 microCi/kg) via the femoral vein. Timed arterial blood samples were collected over 20 minutes following the start of infusion, after which the animal was killed, and the brain was removed. The incorporation coefficient k* for [1-14C]AA was approximately 2-fold higher in microvessels isolated from arecoline-injected than from sham-injected animals. The data demonstrate in an in vivo paradigm, that activation of cholinergic pathways within the rat CNS stimulates arachidonic acid turnover in cerebral microvessels. This suggests a direct involvement of this fatty acid in second messenger function within microvessel endothelial cells and possibly attached pericytes.

Brain fatty acid incorporation into phospholipids can be measured in vivo following intravenous injection of fatty acid tracer. However, to calculate a cerebral incorporation rate, knowledge is required of tracer specific activity in the final brain precursor pool. To determine this for one tracer, unesterified [3H]arachidonate was infused intravenously in pentobarbital-anesthetized rats to maintain constant plasma specific activity for 1-10 min. At the end of infusion, animals were killed by microwave irradiation and analyzed for tracer specific activity and concentration in brain phospholipid, neutral lipid, and lipid precursor, i.e., unesterified arachidonate and arachidonoyl-CoA, pools. Tracer specific activity in brain unesterified arachidonate and arachidonoyl-CoA rose quickly (t1/2 < 1 min) to steady-state values that averaged < 5% of plasma specific activity. Incorporation was rapid, as > 85% of brain tracer was present in phospholipids at 1 min of infusion. The results demonstrate that unesterified arachidonate is rapidly taken up and incorporated in brain but that brain phospholipid precursor pools fail to equilibrate with plasma in short experiments. Low brain precursor specific activity may result from (a) dilution of label with unlabeled arachidonate from alternate sources or (b) precursor pool compartmentalization. The results suggest that arachidonate turnover in brain phospholipids is more rapid than previously assumed.

Permeability-surface area products (PAs) of the blood-nerve barrier (BNB) and blood-brain barrier (BBB) to 125I-labeled native bovine serum albumin (nBSA, pI approximately 4), and to 2 cationized albumins (cBSA) of differing pI (pI approximately 8 and 11), were quantitatively determined in awake rats, using an i.v. bolus injection technique. Mean PAs of the BNB and BBB to 125I-nBSA, after a circulation time of up to 120 min, were (0.17 +/- 0.23) and (0.09 +/- 0.05) x 10(-5) ml/s.g. wet wt, respectively (n = 12 rats), and were not significantly different from 0 (P greater than 0.05). Mean PAs of the BNB and BBB to 125I-cBSA (pI approximately 8), after circulation time of 12 min, were (1.9 +/- 0.1) and (1.7 +/- 0.1) x 10(-5) ml/s.g wet wt, respectively (n = 8). Significant greater PAs, at both the BNB and BBB to 125I-cBSA (pI approximately 11) [(8.2 +/- 1.8) and (3.0 +/- 0.6) x 10(-5) ml/s.g wet wt, respectively (n = 12)], than both PA's of nBSA and cBSA (pI approximately 8) were found. The accumulation of 125I-cBSA in epi-perineurial tissues also was higher than that of 125I-nBSA, and was related to the degree of cationization. Our results indicate that, as at the BBB, the transfer of cationized serum albumin is enhanced over that of native albumin at the BNB of the mammalian peripheral nerve.

The therapeutic efficacy of many anticancer drugs against intracerebral tumors is limited by poor uptake into the central nervous system. One way to enhance brain delivery is to design agents that are transported into the brain by the saturable nutrient carriers of the blood-brain barrier. In this paper, we describe a nitrogen mustard amino acid, DL-2-amino-7-bis[(2-chloroethyl)amino/bd-1,2,3,4-tetrahydro-2-napthoi c acid, that is taken up into brain with high affinity by the large neutral amino acid carrier of the blood-brain barrier. Brain transport of DL-2-amino-7-bis[(2-chloroethyl)aminol-1,2,3,4-tetrahydro-2-naphth oic acid in the rat was found to be rapid (cerebrovascular permeability-surface area product approximately 2 x 10(-2) ml/s/g), saturable and inhibitable by large neutral amino acids. Maximal influx rate (Vmax) and half-saturation (Km) constants equaled 0.26 nmol/min/g and 0.19 microM, respectively, in the parietal cortex. Regional brain uptake of acid exceeded that of the clinical analogue, melphalan, by greater than 20-fold. The results demonstrate that drug modification to produce high-affinity ligands for the cerebrovascular nutrient carriers is a viable means to enhance drug delivery to brain for the treatment of brain tumors and other central nervous system disorders.

The blood-brain barrier (BBB) limits the ability of many hydrophilic solutes to pass from the circulatory system into the nervous system and significantly restricts brain drug delivery. This study examined brain uptake of liposomal formulations in rats. Radiolabeled liposomes less than 1 micron in diameter, prepared from phosphatidylcholine, cholesterol and containing p-aminophenyl-alpha-mannopyranoside or sulfatide, or covalently linked to cationized albumin were injected into the right carotid arteries of rats and organ distribution determined 10 min and 1 h after injection. In all formulations, liposomal retention in brain did not exceed one-tenth of 1% of the injected dose. It seems unlikely that liposomes can significantly bind to or cross the BBB; however, apparent brain uptake could result from liposomal entrapment in the small capillaries of the brain by liposomes larger than 1 micron in diameter.

Unanesthetized adult male rats were infused intravenously with solutions containing 54Mn (II) and one of six concentrations of stable Mn(II). The infusion was timed to produce a near constant [Mn] in plasma for up to 20 min. Plasma was collected serially and on termination of the experiment, samples of CSF, eight brain regions, and choroid plexus (CP) were obtained. Influx of Mn (JMn) was calculated from uptake of 54Mn into tissues and CSF at two different times. Plasma [Mn] was varied 1,000-fold (0.076-78 nmol/ml). Over this plasma concentration range, JMn increased 123 times into CP, 18-120 times into brain, and 706 times into CSF. CP and brain JMn values fit saturation kinetics with Km (nmol/ml) equal to 15 for CP and 0.7-2.1 for brain, and Vmax (10(-2) nmol.g-1.s-1) of 27 for CP and 0.025-0.054 for brain. Brain JMn except at cerebral cortex had a nonsaturable component. CSF JMn varied linearly with plasma [Mn]. These findings suggest that Mn transport into brain and CP is saturable, but transport into CSF is nonsaturable.

Regional blood-brain barrier permeability-surface area products (PAs) of cationized bovine serum albumin (cBSA) with isoelectric point (pI) approximately 8 or greater than or equal to 11 and of native bovine serum albumin (nBSA;pI approximately 4) were determined in awake male Sprague-Dawley rats after bolus intravenous injection. The albumins were labeled with 125I. Brain uptakes were assessed by autoradiography and by direct assay of radioactivity in brain regions. nBSA uptake into brain was statistically insignificant over 3 h, whereas cBSA uptake was significantly even at 6 min. Mean PA values of cBSA with pI approximately 11 (1.69-2.65 x 10(-5) ml.s-1.g-1) in most brain regions were twofold higher than PAs of cBSA with pI approximately 8 (0.98-1.37 x 10(-5) ml.s-1.g-1), whereas mean PA for nBSA did not differ significantly from zero. Autoradiographs of brain slices and net distributions in brain compartments at 6 and 30 min after injection suggested that cBSA entered the brain parenchyma via blood vessels and cerebrospinal fluid but that the former was the main route. The results quantitate for the first time regional brain PA values for cationized proteins and suggest specific mechanisms at cerebral blood vessels that distinguish transport of cationized from noncationized macromolecules.

2-Amino-3-(methylamino)-propanoic acid (BMAA) is a neurotoxic, excitatory amino acid which has been linked through cycad use and consumption with the onset of a variant of amyotrophic lateral sclerosis occurring with high incidence in the western Pacific region. We have studied BMAA pharmacokinetics, oral bioavailability and blood-brain barrier permeability in the rat in an attempt to better define the possible role for BMAA in this disease. To evaluate its kinetics and uptake, BMAA (25-400 mg/kg) was administered to rats, either acutely or chronically, and then plasma and brain concentrations were determined at various times thereafter by combined gas chromatography mass spectrometry. After single dose i.v. injection, BMAA was cleared from plasma in a rapid distribution phase (Vd approximately 16 liters/kg) followed by a slower elimination phase (t1/2 approximately 1 day). Brain uptake was limited by a low blood-brain barrier permeability-surface area product of 2 to 5 x 10(-5) ml/sed/g. Brain BMAA levels peaked within 8 hr after injection, and then declined with a t1/2 similar to that of plasma. After two weeks of continuous infusion (100 mg/kg/day), steady-state brain concentrations equalled 10 to 30 micrograms/g, and only moderately exceeded those in plasma. The results suggest that BMAA may reach potentially toxic levels in brain (i.e., greater than 250 microM) after large doses (greater than 100 mg/kg). However, such doses are orders of magnitude greater than those available from dietary or medicinal use of cycads.

To evaluate the potential contribution of circulating kynurenines to brain kynurenine pools, the rates of cerebral uptake and mechanisms of blood-brain barrier transport were determined for several kynurenine metabolites of tryptophan, including L-kynurenine (L-KYN), 3-hydroxykynurenine (3-HKYN), 3-hydroxyanthranilic acid (3-HANA), anthranilic acid (ANA), kynurenic acid (KYNA), and quinolinic acid (QUIN), in pentobarbital-anesthetized rats using an in situ brain perfusion technique. L-KYN was found to be taken up into brain at a significant rate [permeability-surface area product (PA) = 2-3 x 10(-3) ml/s/g] by the large neutral amino acid carrier (L-system) of the blood-brain barrier. Best-fit estimates of the Vmax and Km of saturable L-KYN transfer equalled 4.5 x 10(-4) mumol/s/g and 0.16 mumol/ml, respectively. The same carrier may also mediate the brain uptake of 3-HKYN as D,L-3-HKYN competitively inhibited the brain transfer of the large neutral amino acid L-leucine. For the other metabolites, uptake appeared mediated by passive diffusion. This occurred at a significant rate for ANA (PA, 0.7-1.6 x 10(-3) ml/s/g), and at far lower rates (PA, 2-7 x 10(-5) ml/s/g) for 3-HANA, KYNA, and QUIN. Transfer for KYNA, 3-HANA, and ANA also appeared to be limited by plasma protein binding. The results demonstrate the saturable transfer of L-KYN across the blood-brain barrier and suggest that circulating L-KYN, 3-HKYN, and ANA may each contribute significantly to respective cerebral pools. In contrast, QUIN, KYNA, and 3-HANA cross the blood-brain barrier poorly, and therefore are not expected to contribute significantly to brain pools under normal conditions.

The permeability-surface area product (PA) of [3H]- or [14C]sucrose at the blood-nerve barrier (BNB) of the sciatic nerve; and at the blood-brain barrier (BBB), were determined in Fischer-344 rats at 3, 11 and 31 months of age. PA was determined by using an in vivo i.v. bolus injection of radiotracer with two-time point graphic and quantitative autoradiographic methods. Vascular space and water content of the tibial nerve of these rats also were determined using quantitative morphometry and dry and wet weight ratios, respectively. There was no significant difference between mean PA(BNB) in any age group [(PA(BNB) at 3 months = 1.2 +/- 0.1 (mean +/- S.E.), at 11 months = 1.8 +/- 0.3; and at 31 months = 1.4 +/- 0.2 x 10(5) ml/s . g wet wt; n = 5-8 rats], nor any difference in PA(BBB). The mean ratio (%) of surface area of endoneurial blood vessels/nerve cross-section of the tibial nerve also did not differ between any group [3 months: 16 +/- 2 vessels; mean surface area ratios = 2.20 +/- 0.10%, n = 5; 11 months: 22 +/- 3 vessels and 2.48 +/- 0.21%, n = 5; 11 months: 22 +/- 3 vessels and 2.48 +/- 0.21%, n = 5; and at 31 months: 26 +/- 1 vessels and 2.40 +/- 0.23%, n = 4). The mean nerve water in rats at 31 months was 64.8 +/- 1.1% wet wt and did not differ from that at 11 months (66.0 +/- 0.6% wet wt) or at 3 months (65.1 +/- 1.0% wet wt) (n = 5-8 nerves). Our results indicate that BBB and BNB integrities are not altered in senescent Fischer-344 rats.

The relative affinity of six anticancer amino acid drugs for the neutral amino acid carrier of the blood-brain barrier was examined in rats using an in situ brain perfusion technique. Affinity was evaluated from the concentration-dependent inhibition of L-[14C]-leucine uptake into rat brain during perfusion at tracer leucine concentrations and in the absence of competing amino acids. Of the six drugs tested, five, including melphalan, azaserine, acivicin, 6-diazo-5-oxo-L-norleucine, and buthionine sulfoximine, exhibited only low affinity for the carrier, displaying transport inhibition constants (Ki, concentrations producing 50% inhibition) ranging from 0.09 to 4.7 mM. However, one agent - D,L-2-amino-7-bis[(2-chloroethyl)amino]- 1,2,3,4-tetrahydro-2-naphthoic acid (D,L-NAM) - demonstrated remarkably high affinity for the carrier, showing a Ki value of approximately 0.2 microM. The relative affinity (1/Ki) of D,L-NAM was greater than 100-fold that of the other drugs and greater than 10-fold that of any compound previously tested. As the blood-brain barrier penetrability of most endogenous neutral amino acids is related to their carrier affinity, the results suggest that D,L-NAM may be a promising agent which may show enhanced uptake and distribution to brain tumors.

A mathematical model describing drug uptake into brain tumors, directly from blood and indirectly from neighboring tissue, is presented. The model quantitatively describes uptake into tumor, brain surrounding tumor (BST), and normal brain and uptake following reversible osmotic blood-brain barrier (BBB) and blood-tumor barrier disruption. It employs published data on the time course for reclosure of the BBB following osmotic treatment and on the brain and tumor uptake of [14C]alpha-aminoisobutyric acid by Walker 256 carcinomas and C6 gliomas implanted into the rat brain. Constant infusion and bolus injection infusion schedules are considered. In untreated brain, the BST acts as a sink, reducing the integrated exposure of the adjacent tumor to the drug, whereas following osmotic treatment, tumor exposure to drug is enhanced, not only by increased delivery from blood but also by diffusion (and bulk flow) from neighboring brain. The model provides a quantitative framework for examining the efficacy of osmotic treatment to enhance chemotherapy of brain tumors.

Infusion of 25% mannitol in saline into the internal carotid artery of dogs disrupts the blood-brain barrier (BBB) in a controlled and reproducible manner (osmotic disruption), whereas mannitol infused through the common carotid artery produces variable disruption of the BBB. Compared to controls, infusion of methotrexate in dogs after osmotic disruption produced significantly higher drug levels in brain. We report very preliminary data on the effects of osmotic disruption with mannitol in five patients harboring malignant brain tumors.

Neutral amino acid (NAA) transport across the blood-brain barrier was examined in pentobarbital-anesthetized rats with an in situ brain perfusion technique. Fourteen of 16 plasma NAAs showed measurable affinity for the cerebrovascular NAA transport system. Values of the transport constants (Vmax, Km, KD) were determined for seven large NAAs from saturation studies, whereas Km values for five small NAAs were estimated from inhibition studies. These data, together with our previous work, provide a complete set of constants for prediction of NAA influx from plasma. Among the NAAs, Vmax varied at least fivefold and Km varied approximately 700 fold. The apparent affinity (1/Km) of each NAA was related linearly (r = 0.910) to the octanol/water partition coefficient, a measure of NAA side-chain hydrophobicity. Predicted influx values from transport constants and average plasma concentrations agree well with values measured using plasma perfusate. These results provide accurate new estimates of the kinetic constants that determine NAA transport across the blood-brain barrier. Furthermore, they suggest that affinity of a L-alpha-amino acid for the transport system is determined primarily by side-chain hydrophobicity.

The blood-brain barrier in the rhesus monkey was opened to intravascular Evans blue-albumin, without causing brain edema or altering brain electrolytes, by perfusing 2.5 molal recrystallized D,L-lactamide into the internal carotid artery for 20--30 sec. Gross neurological and behavioral sequelae were absent in 7 of 8 animals with barrier opening, and 2 days after perfusion no statistically significant changes were observed in sodium, potassium or water contents of perfused as compared to unperfused gray and white matters of brains of the 7 normal animals. Brain endema may not have developed because parenchymal albumin was excreted or metabolized by 2 days. It is suggested also that closure of the barrier after several hours prevents salt from accompanying plasma fluid into the brain. Entry of fluid without salt would reduce, before measurable edema developed, any transcapillary osmotic gradient established by prior entry of plasma albumin.

Right brain regions of anesthetized rats were loaded with [3,5,7-3H]methotrexate ([3H]MTX) or with [14C]sucrose by infusing the tracers into the right carotid artery, after the blood-brain barrier had been opened by right carotid infusion of a hypertonic arabinose solution. During the 6 hr following the procedure, the [3H]MTX concentration in 7 right-sided brain regions, when normalized to the plasma concentration integral during tracer infusion, fell, with an average half-time of 4.8 hr as compared to less than 20 min for the initial rate of loss [14C]sucrose. Right-left brain concentration differences 3 hr after treatment were statistically significant (p less than 0.05) for [3H]MTX but not for [14C]sucrose. The results indicate that intracerebral [3H]MTX is lost more slowly than is intracerebral [14C]sucrose, possibly because [3H]MTX enters brain cells, whereas [14C]sucrose remains largely extracellular.

Osmotic opening of the blood-brain barrier (BBB) most likely is mediated by modification of interendothelial tight junctions, subsequent to shrinkage of cerebrovascular endothelial cells, and not by stimulation of transendothelial vesicular transport or by channel formation. This paper summarizes evidence for this conclusion: osmotic BBB opening is mediated by endothelial cell shrinkage, electron microscopy, with single or serial thin sections, demonstrates penetration of intravascular tracer into brain via tight junctional complexes, the BBB remains open after the brain is fixed, osmotic BBB opening is rapidly reversible, and is insensitive to phenothiazines, and BBB closure following osmotic treatment is size-dependent, indicative of a sieve (pore) mechanism with bulk flow. The entire mechanism of vesicular transport in normal tissue is, furthermore, in doubt, because vesicles that are found in tissue layers connected by tight junctions (e.g., frog nerve perineurium and capillary endoneurium) do not support macromolecular transport.

Inadequate drug delivery to the brain, caused by an intact or partially intact blood-brain barrier (BBB), probably accounts for poor therapeutic responsiveness to cytoreductive drugs by malignant metastatic and primary brain tumors. Drug delivery can be enhanced in normal brains and brains with tumors by administering drugs into the carotid or vertebral circulation after osmotic opening of the BBB. The osmotic procedure in humans involves infusion into the carotid or vertebral arteries of a 25% mannitol solution for 30 s. The procedure is reversible, can be accomplished without long-term neurological deficits, and can be monitored in dogs and humans by means of enhanced computerized tomography. Osmotic BBB disruption, when combined with multiple drug administration, has proved effective in treating brain tumors in a small number of clinical cases.

The reclosure of the blood-brain barrier following osmotic opening was investigated by determining the product of permeability times surface area for two neutral, water-soluble compounds differing widely in molecular size. [14C]Sucrose (Mr 340 Da, radius 5A) and [3H]dextran (Mr 200,000 Da, radius 100 A) were simultaneously injected i.v., and their regional permeability times surface areas were calculated at 6, 10, 35, and 55 min after the blood-brain barrier was opened by a 30-s infusion of 1.8 m L(+)-arabinose into the right external carotid artery. The control permeability times surface area product was about 10(-5) cm3 s-1 g-1 brain for sucrose and negligible for dextran. It increased to 4 X 10(-4) cm3 s-1 g-1 brain and 10(-4) cm3 s-1 g-1 brain for sucrose and dextran, respectively, at 6 min after opening of the blood-brain barrier. Thereafter, permeability-surface area products for both substances declined. Dextran had significantly lower (P less than 0.05) values than sucrose at all times. The ratios of permeability times surface areas of [14C]sucrose to those to [3H]dextran were consistent with restricted diffusion through pores or slits at 35 and 55 min after blood-brain barrier opening, but at 6 and 10 min these ratios were less than the ratio of their free diffusion coefficients, indicative of bulk fluid flow with solute drag from blood to brain. A previously measured increase in brain water content following opening of the blood-brain barrier together with the present results, suggest the creation of slits approximately 400 A in width after osmotic treatment. Reduction in bulk fluid flow from blood to brain appears to be the major cause for the reduction of permeability times surface areas for both sucrose and dextran as the blood-brain barrier recloses.

The effects of osmotic opening of the blood-brain barrier (BBB) on local cerebral glucose utilization (LCGU) were studied in rats with the carbon 14-deoxyglucose method. The BBB was opened to Evans blue dye by unilateral carotid artery perfusion with hypertonic mannitol solution. (14C)-Deoxyglucose was injected 15 minutes or two to three hours later. Osmotic opening of the BBB resulted in focal areas of intense LCGU in affected areas of the perfused hemisphere. In the contralateral hemisphere, glucose utilization was diminished, especially in cerebral cortical areas. The effects were reversible. LCGU was normal in both hemispheres when permeability to Evans blue returned to normal two to three hours after carotid artery perfusion. In 3 of 5 experiments the administration of high doses of intravenous diazepam immediately following carotid artery perfusion with mannitol prevented an increase in LCGU in the perfused hemisphere but did not prevent lowering of LCGU in the contralateral hemisphere. Increased LCGU following osmotic opening of the BBB was not accompanied by a rise in local cerebral blood flow as measured by the iodoantipyrine method. Blood flow was, in fact, significantly decreased.

Unidirectional L-leucine influx across cerebral capillaries was measured at different concentrations with an in situ rat brain perfusion technique, which has several advantages over presently-used methods such as the Brain Uptake Index (BUI) technique. The maximal influx rate (Vmax) and half-saturation concentration (Km) equaled 1.07 +/- 0.02 X 10(-3) mumol X s-1 X g-1 and 0.026 +/- 0.002 mumol X ml-1, respectively, for the saturable component, and the constant for non-saturable influx equaled 6.8 +/- 1.4 X 10(-5) s-1. These values differ by 3-4-fold from respective values obtained with the BUI technique.

Acute hypertension, induced either by intravenous injection of metaraminol bitartrate (Aramine), infusion of isotonic saline into the common carotid artery, or a combination of both procedures did not in the rhesus monkey lead to breakdown of the blood-retinal barrier. Whereas the cerebral vasculature was made permeable to blood-borne dye at carotid pressure above 160 mm Hg, the retinal blood vessels were intact even at pressures as high as 310 mm Hg. Hypertensive blood-brain barrier opening was associated with neurologic defects and brain edema. The results indicate that the retina is more resistant to acute hypertension than is the brain. The greater resistance in the retina may be due to the high number of contractile, perivascular mural cells counteracting increased intravascular hydrostatic pressure. An alternative or supplementary explanation is that choroidal and retinal blood vessels are better protected from surges in blood pressure than are brain blood vessels. Differences between the innervation of brain and ocular blood vessels could account for this.

There is conflicting evidence as to whether dimethyl sulfoxide (DMSO) can reversibly open the blood-brain barrier and augment brain uptake of water-soluble compounds, including anticancer agents. To investigate this, 125I-human serum albumin, horseradish peroxidase, or the anticancer drug melphalan was administered iv to rats or mice, either alone or in combination with DMSO. Some animals received an additional ip priming dose of DMSO. The regional brain concentrations of each substance were measured after the animals were killed. DMSO administration did not significantly increase the brain uptake of any of the compounds as compared to control uptakes. These results do not support prior reports that DMSO increases the permeability of water-soluble agents across the blood-brain barrier.

The blood-brain barrier of rats was opened reversibly by infusing a hyperosmotic solution of arabinose into the external carotid artery. Permeability was increased maximally in the first 15 min and remained slightly elevated at 1 hr. Osmotic barrier opening significantly increased brain uptake of intravenously injected alpha-mannosidase (alpha-D-mannoside mannohydrolase, EC 3.2.1.2.4) (derived from human placenta) and horseradish peroxidase (donor:hydrogen-peroxide oxidoreductase, EC 1.11.1.7). By injection of 4 X 10(5) units of alpha-mannosidase into an animal, brain activity rose to about twice the normal control activity of the enzyme. After 30 min, activity of administered enzyme in the extracellular space of the brain was calculated to be 30% of the serum concentration. Biochemical and histological studies with horseradish peroxidase showed that exogenously administered enzyme entered the brain extracellular space immediately after barrier opening and was incorporated within neuronal lysosomal packets during the next 24 hr. Measurable peroxidase activity was found in brain as much as 72 hr after osmotic treatment. The results demonstrate that the blood-brain barrier can be reversibly opened to enzymes, that a glycoprotein enzyme is incorporated into neuronal lysosomes, and that the brain may now be considered a potential target for enzyme replacement therapy in heritable metabolic disorders.

Osmotic opening of the blood-brain barrier in the rat, by intracarotid infusion of 1.6 molal (m) arabinose solution, increases cerebrovascular permeability to 3H-methotrexate by a factor of about seven. PA (product of permeability and capillary surface area) increases from a mean of 3.3 X 10(-5) sec-1 in brains of control rats to as much as 28 X 10(-5) sec-1 at the cerebral hemisphere ipsilateral to carotid infusion. If tracer is introduced into the carotid artery after osmotic treatment, brain uptake is increased by a net factor of 50 (a factor of 70 due to elevation of PA, multiplied by 7 due to infusion by the carotid route) as compared to uptake by normal, untreated brain with infusion into a peripheral vein. Osmotic barrier opening followed by carotid drug administration may be of use in experimental chemotherapy of nervous system disease.

Melphalan has been reported to be actively transported into tumor cells by two amino acid carrier systems. As amino acids are transported across cerebral capillaries by a facilitated mechanism, studies were undertaken to assess whether or not melphalan was transported similarly, and additionally to determine melphalan's plasma and brain pharmacokinetics. The brain uptake of [14C]melphalan was measured by an in situ brain perfusion technique in the anesthetized rat utilizing [14C]-melphalan. The cerebrovascular permeability-surface area product of [14C]melphalan was calculated at cold melphalan concentrations from O to 16.3 mumol/ml. The permeability-surface area product was concentration dependent and decreased from 10.8 +/- 0.6 (+/- SE) X 10(-4)S-1 at 0.02 mumol/ml melphalan to 5.4 +/- 0.3 X 10(-4)S-1 at 16.3 mumol/ml. The system became saturated at a concentration in excess of 0.1 mumol/ml. The Michaelis-Menten parameters Vmax and Km, determined by nonlinear regression analysis of the permeability-surface area product data, equaled 0.9 +/- 0.3 X 10(-4) mumol/s/g and 0.15 +/- 0.06 mumol/ml, respectively, for the saturable component of melphalan's brain uptake. The Kd of the nonsaturable component was 5.3 +/- 0.03 X 10(-4)S-1. Addition of the amino acid 1-phenylalanine to the brain perfusate inhibited the saturable component of melphalan's brain uptake. The analysis of the plasma and brain concentrations of melphalan by high-performance liquid chromatography, following i.v. melphalan administration, demonstrated that approximately 15% of the drug that was present in plasma entered the brain. These data suggest that the brain uptake of melphalan is facilitated, demonstrating concentration-dependent uptake, saturation, and inhibition, and that melphalan shares the large neutral amino acid carrier system at the blood-brain barrier.

Positron emission tomography (PET) was employed to examine time-dependent changes in blood-brain barrier (BBB) permeability to [68Ga]ethylenediaminetetraacetate (EDTA) in the rhesus monkey, following reversible barrier opening by intracarotid infusion of a hypertonic mannitol solution. The PET technique, when combined with measurements of plasma radioactivity, provided a quantitative measure of the cerebrovascular permeability-area product (PA) at different times following mannitol infusion. Hypertonic mannitol treatment reversibly increased PA to [68Ga]EDTA more than 10-fold; much of the barrier effect was over by 10 min after mannitol treatment. The results show that PET can be used to measure transient changes in BBB integrity in specific brain regions, under in vivo, noninvasive conditions.

We investigated the localization of alkaline phosphatase (AP) in the peripheral and central nervous systems of the frog (Rana pipiens) and rat. In the frog sciatic nerve, AP reaction product was seen as a precipitate within caveolae and vesicular profiles of perineurial cells, and frequently filled the extracellular space. In the rat peripheral nerve, AP reaction product appeared as small tufts on the cell surfaces and within vesicular profiles of endoneurial blood vessels. AP reaction product was not detected in the rat perineurium or in endoneurial blood vessels of the frog. In the frog central nervous system, AP reaction product was detected in the arachnoid membrane adjacent to the subarachnoid space, but not in the cerebral or pial vessels, whereas in the rat it was detected in the outer arachnoid membrane and in the cerebral and pial blood vessels. Biochemical analysis indicated a sevenfold higher AP activity in the frog perineurium over the endoneurium, whereas in the rat, threefold more activity was measured in the endoneurium over the perineurium. Levamisole, an AP inhibitor, decreased the enzyme activity by 95% in rat tissues, and by 70% in frog tissues and in plasma from both animals. Similar decrements were observed cytochemically. This study suggests that: (1) the distribution of AP varies between species, but that it is always present in at least one component of the blood-brain and blood-nerve barriers, (2) because barrier tissues of the nervous system have enzymatic activity, they may biochemically modify the adjacent environment, (3) vesicular profiles and caveolae in the blood vessels and perineurium may function as microenvironments for enzymatic activity, and (4) in the rat and frog, different isozymes of AP may be present.

A three-compartment model was derived to analyze solute exchange among plasma, peripheral nerve epineurium, and endoneurium. The model was fit to measured tibial-nerve epineurial and endoneurial contents of [14C]sucrose after intravenous bolus injection of tracer in pentobarbital sodium-anesthetized rats. The transfer constant (K) for tracer at epineurial vessels approximated 1.1 X 10(-3) ml.s-1.g whole-nerve-1. K at the blood-nerve barrier (BNB) equaled 1.36-1.51 X 10(-5) ml.s-1.g endoneurium-1. The endoneurial uptake data were analyzed also by a simplified two-compartment model incorporating solute exchange between plasma and endoneurium. K at the BNB (1.46 X 10(-5) ml.s-1.g endoneurium-1), determined by multiple uptake time graphic analysis, was only threefold greater than K at the blood-brain barrier. The transfer constant at the perineurium alone was determined in situ and equaled (1 X 10(-6) ml.s-1.g endoneurium-1. Calculated permeability coefficients (X 10(-8) cm/s) equaled 3.2 at the perineurium, 23 at the endoneurial capillaries, and 3.2 at the brain vasculature. The results demonstrate that for hydrophilic nonelectrolytes 1) flux across the perineurium contributes significantly to solute uptake from plasma into the endoneurium; 2) although nerve capillaries are more permeable than brain capillaries, the tissues of the BNB function as a unit to markedly restrict solute diffusion into the endoneurium; and 3) a two-compartment analysis accurately describes hydrophilic solute transfer from plasma into peripheral nerve.

Conscious rats were exposed to 100% O2 at 2.5 ATA for 90 min for up to 10 consecutive days, or to an N2-O2 mixture (PO2 = 0.3 ATA) under the same conditions (control animals). Cerebrovascular permeability to 14C-sucrose in the experimental animals was not altered by hyperbaric O2 exposure when compared to the value in control animals. These results differ with other reports that similar hyperbaric O2 exposure increases cerebrovascular permeability to ferritin and to a protein enzyme.

Nerve water content and the permeability-surface area product (PA) to [3H]-or [14C]sucrose at the blood-nerve barrier were determined in unanesthetized control rats fed a normal diet and in rats fed galactose with or without an aldose reductase inhibitor (Statil or AL 1576) or a thromboxane synthetase inhibitor (CGS 12970). Nerve water content was determined by taking the difference between dry and wet weights of whole tibial nerves. PA was determined by an intravenous bolus injection of radiotracer with multiple-time-point graphic and quantitative autoradiographic methods. The mean nerve water content in galactosemic rats was 15% higher than in control rats after 7-11 mo on the diet. Statil and AL 1576 prevented nerve edema, but CGS 12970 was only partially effective in preventing an increase in nerve water content in galactose-fed rats. In galactosemic rats, the mean PA to sucrose at the blood-nerve barrier, calculated from nerve dry weight, was twofold higher than in control rats. Treatment with Statil, AL 1576, or CGS 12970 prevented increased PA. Our results suggest that nerve edema and increased blood-nerve barrier PA are secondary to polyol production and can be prevented by inhibiting aldose reductase.

Recent studies have shown regulation of central nervous system [Ca] after chronic hypo- and hypercalcemia. To investigate the mechanism of this regulation, 3-week-old rats were fed diets for 8 weeks that contained low or normal levels of Ca. Plasma [Ca] was 40% less in rats fed the low Ca diet than in animals fed normal diet. Unidirectional transfer coefficients for Ca (KCa) and Cl (KCl) into cerebrospinal fluid (CSF) and brain were determined from the 10 min uptake of intravenously injected 45Ca and 36Cl in awake animals. KCa for CSF was 68% greater in low-Ca rats than in normal rats. Likewise, the values of KCa for brain regions with areas adjacent to the ventricles like the hippocampus and pons-medulla were 50% higher than in normal animals. On the other hand, KCas for parietal cortex, a brain region distant from the choroid plexus and not expected to be influenced by Ca entry into CSF, were similar between the groups. Comparison of the regional ratios of KCa/KCl revealed that a selective increase of Ca transport occurred into CSF and all brain regions except the parietal cortex in Ca-deficient rats. The results suggest that Ca homeostasis of CSF and brain [Ca] during chronic hypocalcemia is due to increased transfer of Ca from blood to brain, and that the regulation occurs via the CSF, possibly at the choroid plexus, but not via the cerebral capillaries.

This paper reviews that relation between regional cerebral blood flow (rCBF) and local cerebral glucose utilization (LCGU) under normal conditions of altered functional activity, following disruption of the blood-brain barrier by osmotic stress or hypertension, and during development and maturation of the rat brain. rCBF usually increases in parallel with LCGU when neuronal activity in the central nervous system increases. However, disruption of the blood-brain barrier disturbs this coupling between rCBF and LCGU. the blood-brain barrier, which is located at the continuous cerebrovascular endothelium, protects against brain edema and helps to maintain normal cerebral metabolism. When the barrier is disrupted by osmotic stress or by hypertension, LCGU increases markedly, due to increased neuronal activity, and brain edema develops. rCBF does not increase, however, possibly because brain edema prevents cerebral vasodilatation. rCBF and LCGU follow separate time courses during development and maturation of the rat brain. Both parameters increase between the ages of 1 and 3 months, but rCBF continues to rise from 3 to 12 months when LCGU simultaneously falls. The divergent time courses of these two parameters of brain functional activity may reflect an age-related change in the sensitivity of the cerebrovascular bed to metabolic factors that regulate rCBF.

Protein-deprivation does not increase the permeability of the blood-brain barrier in rats aged 3.5--12 months. PA, the product of cerebrovascular permeability (P) to 14C-sucrose and of cerebral capillary surface area (A), is very low in mature rats that have been maintained either on an 8% or 25% casein diet, and equals about 8 X 10(-6) sec-1 in both groups. There is a tendency for the calculated distribution volume of 14C-sucrose within the brain to decline in protein-deprived rats. Conscious, partially immobilized, protein-deprived rats have the same mean blood pressure, heart rate, arterial plasma pH and adreno-sympathetic response to stress (as measured by plasma norepinephrine and epinephrine concentrations) as do normally fed animals.

Pharmacokinetic parameters of human lymphoblastoid interferon (IFN-alpha) delivery to normal rat brain were examined. IFN-alpha concentrations in brain parenchyma could only be detected 120 min after its intravascular administration, and were 0.003% per gram of the administered dose. The mean cerebrovascular permeability-surface area (permeability x surface area) product to IFN, 120 min after infusion, was 0.35 x 10(-6) sec-1, which is not significantly different from zero. Neither i.v. nor intracarotid IFN-alpha administration significantly affected delivery to brain. Intrathecal administration of IFN-alpha, via the cisterna magna, resulted in undetectable concentrations in brain tissue and plasma at 30 and 60 min. However, osmotic blood-brain barrier opening significantly increased IFN-alpha delivery to brain after its carotid administration. A maximum concentration of 0.18% per gram of the total administered dose was achieved at 120 min, and the cerebrovascular permeability-surface area was increased to 30.8 x 10(-6) sec-1. Intracerebral IFN-alpha concentrations did not decline significantly during the 240 min study. Osmotic blood-brain barrier opening increased the area under the brain concentration vs. time curve, measured between 30 and 240 min, from 0.012 x 10(6) U.min/g, in controls, to 1.24 x 10(6) U.min/g, at least 100-fold. This study indicates that osmotic blood-brain barrier opening significantly increases the delivery of IFN-alpha into brain, and that delivered remains within the brain for many hours.