Effect of intravascular-to-extravascular water exchange on the determination of blood-to-tissue transfer constant by magnetic resonance imaging.

Water exchange across capillary walls couples intra- and extravascular (IV-EV) protons and their magnetization. A bolus i.v. injection of an extracellular MRI contrast agent (MRCA) causes a large increase in the spin-lattice relaxation rate, R1, of water protons in the plasma and blood cells within the capillaries and changes the effective relaxation rate R1eff in tissue via IV-EV water exchange. An analysis of the effect of plasma-red cell and IV-EV water exchange on the MRI-measured influx and permeability of capillaries to the MRCA is presented and focused on the brain and the blood-brain barrier. The effect of arrival of a bolus of an MRCA in the capillary on the relaxation rate R1eff in tissue via IV-EV water exchange occurs more rapidly than the MRCA uptake in tissue and can dominate the initial time curve of the R1eff change before the MRCA uptake in tissue becomes significant. This raises the possibility that (tissue dependent) IV-EV rate of exchange of water molecules can affect estimates of MRCA transfer constant. We demonstrate that an approach that considers IV-EV water exchange and uses the theoretical model of blood-brain tracer distribution developed by Patlak et al. (J Cereb Blood Flow Metab 1983;3:1-7) can lead to an accurate estimate of the MRI-determined influx rate constant of the MRCA and to an underestimation of the tissue blood volume.

Sickle red blood cells accumulate in tumor.

The preferential accumulation of sickle blood cells in tumor vasculature is demonstrated noninvasively using MRI and sickle red blood cells loaded with Gd-DTPA and invasively by two other techniques. The distribution of red blood cells in rat brain tumors relative to normal brains were measured using three separate techniques: MRI of Gd-DTPA loaded cells, fluorescent microscopy detection of Oregon Green 488 fluorescence conjugated to a streptavidin-biotin complex that binds to red blood cell surface proteins, and autoradiography using a technetium (99m)Tc-labeling kit. Labeled red cells were infused intravenously in rats with brain tumors. Sickle cells preferentially accumulated in tumor relative to normal brain, with highest concentrations near the tumor / normal tissue boundary, whereas control normal red cells did not preferentially aggregate at the tumor periphery. This demonstrates the potential of sickle red blood cells to accumulate in the abnormal tumor vessel network, and the ability to detect their aggregation noninvasively and at high spatial resolution using MRI. The application of the noninvasive measurement of sickle cells for imaging tumor neovasculature, or as a delivery tool for therapy, requires further study.

Estimating blood-brain barrier opening in a rat model of hemorrhagic transformation with Patlak plots of Gd-DTPA contrast-enhanced MRI.

Patlak plot processing of Gd-shifted T1 relaxation-time images from a rat model of hemorrhagic transformation yielded estimates and maps of the blood-to-brain influx rate constant of Gd-DTPA (K1). The Patlak plots also produced a heretofore unrecognized parameter, the distribution space of the intravascular-Gd-shifted protons (Vp), an index of blood-to-tissue transfer of water. The K1 values for Gd-DTPA were very high for the regions of blood-brain barrier (BBB) opening and were similar to those of 14C-sucrose concurrently obtained by quantitative autoradiographic (QAR) analysis. In these same ROI's, Vp was five-fold greater than normal, which suggests that the permeability of the BBB to water was also increased. The 14C-sucrose space of distribution in the ischemic ROI's was around 8%, thus indicating a sizable interstitial space. The spatial resolving power of Gd-DTPA-deltaT1 imaging was rather good, although no match for 14C-sucrose-QAR. This study shows that quantitative deltaT1-MRI estimates of regional blood-brain transfer constants of Gd-DTPA and water distribution are possible when Patlak plots are employed to process the data. This approach may be useful for tracking the time-course of BBB barrier function in both animals and humans.

Single-coil arterial spin-tagging for estimating cerebral blood flow as viewed from the capillary: relative contributions of intra- and extravascular signal.

The single-capillary model was applied to the exchange microvessels for water in the cerebral parenchyma and used to calculate blood-to-brain flux of water; the theory of the steady-state arterial spin-tagging (AST) technique for estimating cerebral blood flow (CBF) was revised to incorporate the presence of both extravascular (tissue) and capillary signal. A crucial element of the single-coil AST experiment is that magnetization transfer (MT) shortens the effective T1 of the extravascular water, making it one-quarter that of the T1 of capillary blood. Furthermore, the mean capillary transit time is on the order of the T1 of the extravascular water. The single-coil AST experiment is distinguished from other methods which use water as an indicator for measurement of CBF in that the (flow-dependent) populations of inverted protons in the intra- and extravascular compartments can be nearly equal for normal physiological conditions. The following questions are considered: Is single-coil AST contrast linear in resting CBF? Is contrast in the single-coil AST technique likely to be linear under changes in CBF in normal tissue? Is the contrast likely to be linear in such common pathologies as stroke and cerebral tumor? We demonstrate that, if the population of inverted protons in the microvessels is included in the experiment, the voxel population of inverted protons will be approximately linear with flow across a broad range of flow values. We predict that the single-coil AST experiment will systematically overestimate resting CBF for flows in the normal range, that changes in CBF in normal tissue will produce an approximately linear response in AST measurement, and, finally, we predict the operating characteristics of the measurement in common cerebral pathologies.

Regional variations in the apparent diffusion coefficient and the intracellular distribution of water in rat brain during acute focal ischemia.

BACKGROUND AND PURPOSE: The apparent diffusion coefficient of water (ADC) rapidly drops in ischemic tissue after cerebral artery occlusion. This acute drop is thought to be caused by the loss of extracellular fluid and the gain of intracellular fluid. To test the latter possibility, changes in ADC and the size of several cellular compartments were assessed in 3 regions of rat brain at the end of 90 minutes of focal cerebral ischemia. METHODS: One middle cerebral artery was permanently occluded in 8 Sprague-Dawley rats; sham occlusions were performed in 2 other rats. ADC maps were generated 90 minutes later, and the brains were immediately perfusion fixed. Three regions of interest (ROIs) were defined on the basis of ADC range. Various neuronal, astrocytic, and capillary compartments in each ROI were quantified with light and electron microscopy. RESULTS: At the end of 90 minutes of ischemia, mean ADC was normal in the cortex of sham-operated rats and the contralateral cortex of ischemic rats (ROI-a), 25% lower in the ipsilateral frontoparietal cortex (ROI-b), and 45% lower in the ischemic lateral caudoputamen (ROI-c). At this time, the frequency of swollen astrocytic cell bodies and volume of swollen dendrites and astrocytic processes in neuropil were ROI-a

Differences in vulnerability to permanent focal cerebral ischemia among 3 common mouse strains.

BACKGROUND AND PURPOSE: Genetically engineered mice are used to study the role of single genes in cerebral ischemia, but inherent, strain-dependent differences in neuronal vulnerability may affect experimental end points. To examine this possibility, tissue injury resulting from focal ischemia and its relationship to cerebral hemodynamics were determined in 3 common mutant mouse strains. METHODS: Permanent middle cerebral artery ligation was performed in male C57BL/6J, Balb/C, and 129X1/SvJ mice. Mean arterial blood pressure, blood gases, basal and postischemic cortical blood flow ([(14)C]iodoantipyrine autoradiography and laser-Doppler flowmetry), posterior communicating artery patency, and infarct size were determined. RESULTS: Basal cortical blood flow did not differ among strains. Ten minutes after middle cerebral artery ligation, relative red cell flow in the ischemic cortex was 6% to 7% of preischemic flow in every strain. Despite similar hemodynamics, cortical infarcts in Balb/C mice were 3-fold larger than those in 129X1/SvJ and C57BL/6J mice; infarct size in the latter 2 strains was not significantly different. The posterior communicating artery was either poorly developed or absent in >90% of the Balb/C and C57BL/6J but in <50% of the 129X1/SvJ mice. CONCLUSIONS: The extent of ischemic injury differed markedly between the 3 strains. The presence and patency of posterior communicating arteries, although variable among strains, did not affect preischemic or postischemic cortical blood flow or bear any relationship to ischemic injury. Therefore, intrinsic factors, other than hemodynamic variability, may contribute to the differences in ischemic vulnerability among strains. These findings underscore the importance of selecting genetically matched wild-type controls.

Inhibition of glucose transport and direct interactions with type 1 facilitative glucose transporter (GLUT-1) by etomidate, ketamine, and propofol: a comparison with barbiturates.

Ketamine, etomidate, propofol, and pentobarbital were compared for effects on and interactions with the type 1 facilitative glucose transporter (GLUT-1). Fluxes of radiolabeled hexoses were used to determine the effects of anesthetics on GLUT-1 function. Hypotonic hemolysis of human erythrocytes was used to assess perturbations of membrane integrity. Quenching of intrinsic protein fluorescence was used to assess the direct interactions of anesthetics with purified GLUT-1. Pentobarbital, ketamine, etomidate, and propofol inhibited glucose transport in murine fibroblasts with IC(50) values of 0.8, 1. 6, 0.1, and 0.4 mM, respectively. Pentobarbital, ketamine, etomidate, and propofol also inhibited sugar transport in human erythrocytes. The IC(50) values for pentobarbital and ketamine exhibited substrate dependence for equilibrium exchange but not unidirectional effluxes. This was not observed for etomidate. Propofol did not inhibit equilibrium exchange but did inhibit unidirectional efflux with little substrate dependence. Pentobarbital protected against hemolysis, whereas etomidate and ketamine promoted hemolysis of erythrocytes. Propofol had no effect on membrane integrity. Pentobarbital, ketamine, and etomidate all interacted directly with GLUT-1, with apparent K(d) values of 2.2, 0.8, and 0.5 mM, respectively. Like barbiturates, ketamine, etomidate, and propofol inhibited GLUT-1 at concentrations near to those used pharmacologically. Inhibition of GLUT-1 by these intravenous general anesthetics was complex, exhibiting differential kinetic effects on equilibrium exchange versus unidirectional fluxes and contrasting substrate dependencies. Like barbiturates, ketamine and etomidate bound to GLUT-1 with affinities that paralleled inhibition of glucose transport. As a class, intravenous general anesthetics, in contrast to inhalation anesthetics, inhibit GLUT-1-mediated glucose transport.

Secondary decline in apparent diffusion coefficient and neurological outcomes after a short period of focal brain ischemia in rats.

This study was designed to characterize the initial and secondary changes of the apparent diffusion coefficient (ADC) of water with high temporal resolution measurements of ADC values and to correlate ADC changes with functional outcomes. Fourteen rats underwent 30 minutes of temporary middle cerebral artery occlusion (MCAO). Diffusion-, perfusion-, and T2-weighted imaging was performed during MCAO and every 30 minutes for a total of 12 hours after reperfusion (n = 6). Neurological outcomes were evaluated during MCAO, every 30 minutes for a total of 6 hours and at 24 hours after reperfusion (n = 8). The decreased cerebral blood flow during MCAO returned to normal after reperfusion and remained unchanged thereafter. The decreased ADC values during occlusion completely recovered at 1 hour after reperfusion. The renormalized ADC values started to decrease secondarily at 2.5 hours, accompanied by a delayed increase in T2 values. The ADC-defined secondary lesion grew over time and was 52% of the ADC-defined initial lesion at 12 hours. Histological evaluation demonstrated neuronal damage in the regions of secondary ADC decline. Complete resolution of neurological deficits was seen in 1 rat at 1 hour and in 6 rats between 2.5 and 6 hours after reperfusion; no secondary neurological deficits were observed at 24 hours. These data suggest that (1) a secondary ADC reduction occurs as early as 2.5 hours after reperfusion, evolves in a slow fashion, and is associated with neuronal injury; and (2) renormalization and secondary decline in ADC are not associated with neurological recovery and worsening, respectively.

In vitro evidence that beta-amyloid peptide 1-40 diffuses across the blood-brain barrier and affects its permeability.

Beta amyloid peptides are major insoluble constituents of amyloid fibrils in senile plaques and cerebrovascular deposits, both characteristic of Alzheimer disease (AD). Low concentrations of soluble forms of amyloid peptides are also present in normal CSF. We previously demonstrated that the 40 amino acid form of soluble beta-amyloid peptide (sAbeta) is rapidly cleared from rat CSF into blood. Herein we hypothesized that a saturable, outwardly directed flux of this peptide occurs at the blood-brain barrier (BBB) and tested whether supraphysiological (possibly pathological) concentrations of sAbeta could alter the permeability of this barrier to a paracellular tracer, polyethylene glycol (PEG). Using an in vitro model of BBB, we showed that influx and efflux of sAbeta were equal, modest (60%-160% greater than that of PEG), and not saturable. These observations suggest that sAbeta moved across the monolayer by a diffusional process, and not via a transporter. PEG flux was doubled immediately after the luminal concentration of cold sAbeta was raised to 5 microM, and was doubled 150 min after the abluminal concentration of sAbeta was increased to 5 microM. Pathological elevations of sAbeta concentration in plasma or brain interstitial fluid may, therefore, alter the permeability of brain capillaries in vivo.

Transient and permanent resolution of ischemic lesions on diffusion-weighted imaging after brief periods of focal ischemia in rats : correlation with histopathology.

BACKGROUND AND PURPOSE: The early ischemic lesions demonstrated by diffusion-weighted imaging (DWI) are potentially reversible. The purposes of this study were to determine whether resolution of initial DWI lesions is transient or permanent after different brief periods of focal brain ischemia and to evaluate histological outcomes. METHODS: Sixteen rats were subjected to 10 minutes (n=7) or 30 minutes (n=7) of temporary middle cerebral artery occlusion or sham operation (n=2). DWI, perfusion-weighted imaging (PWI), and T(2)-weighted imaging (T(2)WI) were performed during occlusion; immediately after reperfusion; and at 0.5, 1.0, 1.5, 12, 24, 48, and 72 hours after reperfusion. After the last MRI study, the brains were fixed, sectioned, stained with hematoxylin and eosin, and evaluated for neuronal necrosis. RESULTS: No MRI or histological abnormalities were observed in the sham-operated rats. In both the 10-minute and 30-minute groups, the perfusion deficits and DWI hyperintensities that occurred during occlusion disappeared shortly after reperfusion. The DWI, PWI, and T(2)WI results remained normal thereafter in the 10-minute group, whereas secondary DWI hyperintensity and T(2)WI abnormalities developed at the 12-hour observation point in the 30-minute group. Histological examinations demonstrated neuronal necrosis in both groups, but the number of necrotic neurons was significantly higher in the 30-minute group (95+/-4%) than in the 10-minute group (17+/-10%, P<0.0001). CONCLUSIONS: Transient or permanent resolution of initial DWI lesions depends on the duration of ischemia. Transient resolution of DWI lesions is associated with widespread neuronal necrosis; moreover, permanent resolution of DWI lesions does not necessarily indicate complete salvage of brain tissue from ischemic injury.

High resolution quantitation of microvascular plasma perfusion in non-ischemic and ischemic rat brain by laser-scanning confocal microscopy.

Laser-scanning confocal microscopy (LSCM) was used to measure at high resolution cerebral plasma volumes (perfusion) using two fluorescent plasma markers in a rat model of embolic stroke. This application of LSCM to study the microvascular circulation in embolic stroke was developed as an alternative to autoradiography to measure cerebral perfusion. An additional benefit of LSCM is that it quantitates with great accuracy the structural relationships of the microcirculation to cells and the pathological alterations of the ischemic brain. Autoradiography allows only a quantitative analysis of cerebral perfusion. For example, in order to study the microcirculation and its relationship to blood brain barrier damage, the volume of perfused cerebral capillaries was measured by administering two fluorescent plasma markers (FITC-dextran and Evans blue) intravenously to a rat. Evans blue was administered before cerebral ischemia and FITC-dextran administered post-ischemia 1 min before sacrifice. Volumes of plasma perfusion were analyzed by means of a system developed for 3D analysis of fixed and stained serial brain histologies. Plasma volumes for the non-ischemic cerebral cortex were 1.00%+/-0.38% while plasma volumes in the caudate/putamen were 0.69%+/-0.17% in good agreement with the previously published values using the autoradiography method. The architecture of the capillaries in the ischemic core showed perfusion of Evans blue but there was no flow of FITC dextran. Our work represents a novel application of this technology to investigation of cerebral vascular disease and identifies its potential to become an important tool for investigation of cerebral pathology.

Effects of barbiturates on facilitative glucose transporters are pharmacologically specific and isoform selective.

Barbiturates inhibit GLUT-1-mediated glucose transport across the blood-brain barrier, in cultured mammalian cells, and in human erythrocytes. Barbiturates also interact directly with GLUT-1. The hypotheses that this inhibition of glucose transport is (i) selective, preferring barbiturates over halogenated hydrocarbon inhalation anesthetics, and (ii) specific, favoring some GLUT-# isoforms over others were tested. Several oxy- and thio-barbiturates inhibited [3H]-2-deoxyglucose uptake by GLUT-1 expressing murine fibroblasts with IC50s of 0.2-2.9 mm. Inhibition of GLUT-1 by barbiturates correlates with their overall lipid solubility and pharmacology, and requires hydrophobic side chains on the core barbiturate structure. In contrast, several halogenated hydrocarbons and ethanol (all 10 mm). Thus, barbiturates selectively inhibit glucose transport by some, but not all, facilitative glucose transporter isoforms.

Quantitation of microvascular plasma perfusion and neuronal microtubule-associated protein in ischemic mouse brain by laser-scanning confocal microscopy.

In an exposition of the technique of calculating distribution volumes from laser-scanning confocal microscopic (LSCM) data, three-dimensional images of the distribution of one or two fluorescent markers in mouse brain specimens were generated by LSCM and processed by a system developed for morphometric analysis of fixed and stained serial brain histologic samples. To determine the volume of perfused cerebral capillaries, one of two fluorescent plasma markers, either fluorescein isothiocyanate (FITC)-dextran or Evans blue, was intravenously administered to mice subjected to 1 hour of embolic middle cerebral artery (MCA) occlusion (n = 9) and to mice that were not operated on (n = 3); after 1 minute of circulation, brains were removed, immersion-fixed, and processed for LSCM. In some of these animals (n = 5), the volume of endogenous microtubule-associated protein-2 (MAP2) fluorescence was also determined using immunohistochemical staining. For mice that were not operated on, this methodology yielded highly localized volumes of (1) microvascular plasma, which agree with those determined for rodents by other techniques, and (2) MAP2 expression, which appears physiologically and morphologically reasonable. After 1 hour of MCA occlusion, the MAP2 volumes of distribution were less than 10% of normal in the ipsilateral hemisphere in which plasma perfusion essentially ceased. In conclusion, precise colocalization and quantitation of early ischemic neuronal damage and cerebral plasma perfusion deficit can be done with this three-dimensional, microphysiologic and microanatomic methodology.

Local cerebral blood flow during the first hour following acute ligation of multiple arterioles in rat whisker barrel cortex.

The objectives are to measure the early time-course of the flows of blood, red cells, and plasma in brain tissue destined to infarct following arterial occlusion. The flux of fluorescent red blood cells (fRBCs) through venules and the arteriovenous transit times (AVTT) of fluorescein-labeled plasma albumin were periodically monitored in anesthetized adult Wistar rats before and up to 60 min after permanent ligations of several small branches of the middle cerebral artery. Of note, fRBC is a function of venular erythrocyte flow and volume, whereas AVTT is a function of plasma flow and volume in visible arteriole-capillary-venule units. In another group of anesthetized rats, local cerebral blood flow (ICBF) was measured 1 h after permanent arterial occlusion by [14C]iodoantipyrine (IAP) autoradiography. With this model of focal ischemia, the lesion is highly reproducible and involves part of the whisker barrel cortex. Infarction of this area was observed in 12 of 13 rats. From 10 to 60 min after arterial occlusion, AVTT was nearly four times longer in the ischemic barrel cortex than at the same site before ligations, and fRBC flux was 25%. Neither parameter changed appreciably over this time. After 60 min of ischemia, ICBF on the ipsilateral barrel cortex was 18% of that on the contralateral side and 15% of the sham control value for the same area of the barrel cortex. Since whole blood flow in the ischemic barrel cortex was < 20% of normal at 60 min and AVTT and fRBC flux were essentially constant from 10 to 60 min, the rates of plasma and red cell flows were similarly depressed during the first hour of arteriolar occlusion. In conclusion, such lowering of red cell, plasma, and blood flows produced consistent infarctions in the barrel cortex.

The rapid flow of cerebrospinal fluid from ventricles to cisterns via subarachnoid velae in the normal rat.

14C-sucrose in 0.5 microliter of buffered saline was infused over 30 sec into one lateral ventricle, and its subsequent distribution was determined in brain, meninges, cerebral blood vessels, and cerebrospinal fluid (CSF) by quantitative autoradiography. Within 3.5 min, infused radiotracer had moved into the third ventricle, the velum interpositum (an extension of the subarchnoid system that contains many blood vessels), the aqueduct, the mesencephalic and fourth ventricles, and the superior medullary velum (a part of the subarachnoid system that touches the mesencephalic and fourth ventricles). The CSF within both of these velae appears to empty into the quadrigeminal and ambient cisterns. Within 5 min radioactive sucrose was also found in the interpeduncular cistern. About 15% of the injected sucrose quickly left the ventricles and entered these large cisterns. In contrast to most CSF-brain interfaces, little sucrose moved from CSF into the medulla next to the lateral recesses and tissues such as the superior colliculus that lie adjacent to the large CSF cisterns. A thick, multilayered glia limitans visible on electron micrographs seemed to form a CSF-brain barrier at these interfaces. Some of the infused 14C-sucrose persisted in the perivascular spaces and walls of arteries and arterioles for more than 3.5 hr. These findings suggest that CSF may function to deliver various agents and factors to pial and parenchymal arteries and arterioles.

Barbiturate inhibition of GLUT-1 mediated hexose transport in human erythrocytes exhibits substrate dependence for equilibrium exchange but not unidirectional sugar flux.

Barbiturates inhibit GLUT-1 mediated hexose transport both in vivo [Gjedde & Rasmussen (1980) J. Neurochem. 35, 1382-1387; Otsuka et al. (1991) Am. J. Physiol. 261, R265-R275] and in vitro [Honkanen et al. (1995) Biochemistry 34, 535-544]. In the present study, the mechanism by which barbiturates inhibit GLUT-1 mediated hexose transport was examined by measuring both unidirectional zero trans and equilibrium exchange fluxes of hexoses in the functionally well-characterized, GLUT-1 rich human erythrocyte system. Unidirectional influx were both inhibited (> 80%) by 10 mM pentobarbital (PB). This symmetrical inhibition of unidirectional flux by PB was virtually independent of cis sugar concentration (2-130 mM) and exhibited an IC50 of approximately 2 mM. In contrast to unidirectional sugar flux, PB inhibition of equilibrium exchange sugar flux is attenuated by increased substrate concentration (e.g., 88% inhibition at 1 mM Glc versus 40% inhibition at 130 mM Glc in the presence of 10 mM PB) and exhibits an IC50 of approximately 10 mM at 100 mM Glc. Other barbiturates were found to inhibit sugar flux in human erythrocytes in this differential manner. These findings, when viewed with kinetic models proposed for GLUT-1 mediated transport [Carruthers (1990) Physiol. Rev. 70, 1135-1176], are consistent with barbiturates being noncompetitive inhibitors of Glc translocation and preferentially inhibiting the unoccupied form of the carrier protein. We propose, therefore, that barbiturates may prevent or alter the conformational changes associated with the reorientation of the carrier protein within the membrane. Overall, these results imply that barbiturates may more strongly inhibit GLUT-1 mediated Glc flux in vivo when the trans Glc is near zero as a result of either metabolism or another transport process.

Rapid distribution of intraventricularly administered sucrose into cerebrospinal fluid cisterns via subarachnoid velae in rat.

The intracranial distribution of [14C]sucrose, an extracellular marker infused for 30 s into one lateral ventricle, was determined by autoradiography of frozen-dried brain sections. Within 3.5 min [14C]sucrose appeared in: (i) the third ventricle, including optic, infundibular and mammillary recesses; (ii) the aqueduct of Sylvius; (iii) the velum interpositum, a part of the subarachnoid space that runs along the roof of the third ventricle and contains many blood vessels; (iv) the mesencephalic and fourth ventricles; and (v) the superior medullary velum, a highly vascular extension of the subarachnoid space that terminates at the walls of the mesencephalic and fourth ventricles. Within 5 min, radioactivity was present in the interpeduncular, ambient and quadrigeminal cisterns, which encircle the midbrain. By 10 min, approximately 11% of the radioactivity had passed into the subarachnoid space via a previously undescribed flow pathway that included the velum interpositum and superior medullary velum. At many places along the ventricular system, [14C]sucrose appeared to move from cerebrospinal fluid into the adjacent tissue by simple diffusion, as reported previously (Blasberg R. G. et al. (1974) J. Pharmac. exp. Ther. 195, 73-83; Levin V. A. et al. (1970) Am. J. Physiol. 219, 1528-1533; Patlak C. and Fenstermacher J.D. (1975) Am. J. Physiol. 229, 877-884; Rosenberg G. A. and Kyner W.T. (1980) Brain Res. 193, 56-66; Rosenberg G. A. et al. (1986) Am. J. Physiol 251, F485-F489). Little sucrose was, however, taken up by: (i) circumventricular organs such as the subfornical organ; (ii) medullary and cerebellar tissue next to the lateral recesses; and (iii) the superior and inferior colliculi and cerebral peduncles. For the latter two groups of structures, entry from cerebrospinal fluid was apparently blocked by a thick, multilayered glia limitans. Although [14C]sucrose was virtually absent from the rest of the subarachnoid system after 1 h, it remained in the perivascular spaces and/or walls of pial arteries and arterioles for more than 3 h. Certain transport proteins, protease inhibitors, growth factors and other neurobiologically active materials are present in cerebrospinal fluid, and their distribution to the brain and its blood vessels may be important. The present work shows, in the rat, that the flow of cerebrospinal fluid and the disposition of its constituents is fairly complex and differs among regions. Flow was rapid throughout the ventricular system and into various subarachnoid velae and cisterns, but was surprisingly slow and slight over the cerebral and cerebellar cortices. The cerebrospinal fluid-to-tissue flux of material was relatively free at many interfaces, but was greatly restricted at others, the latter indicating that the old concept of a "cerebrospinal fluid-brain barrier" may hold at such places. Finally, radiolabeled sucrose was retained longer within the walls and perivascular spaces of pial arteries and arterioles than in other subarachnoid tissues; one function of the cerebrospinal fluid system or "third circulation" may thus be delivering factors and agents to these pial blood vessels.

Fate of cerebrospinal fluid-borne amyloid beta-peptide: rapid clearance into blood and appreciable accumulation by cerebral arteries.

In Alzheimer's disease, the neuritic or senile amyloid plaques in hippocampus and association cortex, the diffuse plaques in brain areas such as the cerebellum and sensorimotor cortex, and the amyloid deposits in the walls of pial and parenchymal blood vessels are mainly composed of amyloid beta-peptides. In the present study, either soluble 40-residue amyloid beta-peptide radiolabeled with 125I (I-sAbeta) or [14C]polyethylene glycol ([14C]PEG, a reference material) was briefly infused into one lateral ventricle of normal rats. By 3.5 min, 30% of the I-sAbeta was cleared from ventricular CSF into blood; another 30% was removed over the next 6.5 min. No [14C]PEG was lost from the CSF-brain system during the first 5 min, and only 20% was cleared by 10 min. Much of the I-sAbeta that reached the subarachnoid space was retained by pial arteries and arterioles. Virtually no I-sAbeta was found in brain. The clearance of amyloid beta-peptides from the CSF-brain system, reported herein for normal rats, may be reduced in Alzheimer's disease, thus contributing to amyloid deposition in cerebral tissue and blood vessels.

Blood-brain interfaces: relevance to cerebral drug metabolism.

The brain, with the exception of the circumventricular organs (CVOs), is partially protected from the invasion of blood-borne chemicals by the tight junctions that link adjacent cerebral endothelial cells and form the structural basis of the blood-brain barrier (BBB). In addition to the BBB, the epithelial layer of the choroid plexuses and the barrier layer of the arachnoid membrane complex comprise a second system for protecting the brain, a system often referred to as the blood-cerebrospinal fluid (CSF) barrier. In the past several years, several enzymes that are involved in hepatic drug metabolism have been found in the small microvessels from brain, the choroid plexuses, and the leptomeninges (pia plus arachnoid mater) as well as in some CVOs. These drug-metabolizing systems are inducible and may act at these various interfaces as 'enzymatic barriers' to influx. In particular, the activities of these enzymes in choroidal tissue are so high that the choroid plexuses can well be the major site of drug metabolism in the brain. The fate of intracerebrally formed polar metabolites and the potential of the blood-brain and blood-CSF barriers as sites for metabolic activation-induced neurotoxicity are discussed.

Nicotine raises the influx of permeable solutes across the rat blood-brain barrier with little or no capillary recruitment.

Nicotine (1.75 mg/kg s.c.) was administered to rats to raise local CBF (lCBF) in various parts of the brain, test the capillary recruitment hypothesis, and determine the effects of this increase in lCBF on local solute uptake by brain. lCBF as well as the local influx rate constants (K1) and permeability-surface area (PS) products of [14C]antipyrine and [14C]-3-O-methyl-D-glucose (3OMG) were estimated by quantitative autoradiography in 44 brain areas. For this testing, the finding of significantly increased PS products supports the capillary recruitment hypothesis. In 17 of 44 areas, nicotine treatment increased lCBF by 30-150%, K1 of antipyrine by 7-40%, K1 of 3OMG by 5-27%, PS product of antipyrine by 0.20% (mean 7%), and PS product of 3OMG by 0-23% (mean 8%). Nicotine had no effect on blood flow or influx in the remaining 27 areas. The increases in lCBF and K1 of antipyrine were significant, whereas those in K1 of 3OMG and in PS for both antipyrine and 3OMG were not statistically significant. The lack of significant changes in PS products implies that in brain areas where nicotine increased blood flow: (a) essentially no additional capillaries were recruited and (b) blood flow within brain capillary beds rises by elevating linear velocity. The K1 results indicate that the flow increase generated by nicotine will greatly raise the influx and washout rates of highly permeable materials, modestly elevate those of moderately permeable substances, and negligibly change those of solutes with extraction fractions of < 0.2, thereby preserving the barrier function of the blood-brain barrier.