Deltorphin transport across the blood brain barrier

Transcription

1 Proc. Natl. Acad. Sci. USA Vol. 94, pp , August 1997 Pharmacology Deltorphin transport across the blood brain barrier ANNA FIORI*, PATRIZIA CARDELLI, LUCIA NEGRI, MARIA ROSARIA SAVI*, ROBERTO STROM, AND VITTORIO ERSPAMER *Department of Biochemical Sciences A. Rossi Fanelli, Department of Cellular Biotechnology and Hematology, and Institute of Medical Pharmacology, University La Sapienza, Rome, Italy Contributed by Vittorio Erspamer, June 30, 1997 ABSTRACT In vivo antinociception studies demonstrate that deltorphins are opioid peptides with an unusually high blood brain barrier penetration rate. In vitro, isolated bovine brain microvessels can take up deltorphins through a saturable nonconcentrative permeation system, which is apparently distinct from previously described systems involved in the transport of neutral amino acids or of enkephalins. Removing Na ions from the incubation medium decreases the carrier affinity for deltorphins ( 25%), but does not affect the V max value of the transport. The nonselective opiate antagonist naloxone inhibits deltorphin uptake by brain microvessels, but neither the selective -opioid antagonist naltrindole nor a number of opioid peptides with different affinities for - or -opioid receptors compete with deltorphins for the transport. Binding studies demonstrate that -, -, and -opioid receptors are undetectable in the microvessel preparation. Preloading of the microvessels with L-glutamine results in a transient stimulation of deltorphin uptake. Glutamine-accelerated deltorphin uptake correlates to the rate of glutamine efflux from the microvessels and is abolished by naloxone. The blood brain barrier (BBB) accounts for the restricted movement of solutes between the vascular and the cerebral compartments. Its anatomical counterpart has been identified as the endothelial cell membranes of brain microvessels (1 5). These endothelial cells, joined together by tight junctions (6), form a continuous physical barrier playing a crucial role in determining the rate at which different compounds can reach, or in turn leave, the central nervous system (CNS). The capacity of any particular molecule to penetrate the BBB depends essentially on its free diffusion across the cell walls i.e., on its lipid solubility or on its specific affinity for a carrier-mediated transport system. The use of isolated brain microvessels, as an in vitro model of the BBB, has allowed the identification and characterization of some peculiarities of the transport systems specific for sugars (7, 8), amino acids (9 13), electrolytes (14, 15), or transferrin (16). It had been initially assumed that peptides could not enter the CNS via the BBB (17) and that the entry rate into the CNS of some peptides correlated only with their lipid solubility, suggesting that they crossed the BBB by direct diffusion through the phospholipid bilayer of the endothelial cell membranes. It is now generally accepted, instead, that different categories of peptides can enter or exit the CNS through the endothelial cell membranes of brain microvessels at a rate higher than that accounted by passive diffusion. In vivo studies have indicated the existence of transport systems that allow the selective permeation of the BBB by different peptides, either endogenous or synthetic (18 24). Some years ago, Banks and Kastin (21) have suggested that at least four different transport systems, named The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C solely to indicate this fact by The National Academy of Sciences $ PNAS is available online at http: PTS-1, -2, -3, and -4, could accelerate the entrance of opioids and related peptides into the brain. More recently, the use of different in vitro models consisting either of fresh isolated brain microvessels or of cultured brain endothelial cells has allowed a better investigation of the modality of transport of some opioid peptides such as Met-enkephalin and related analogs (25 27). Deltorphins are naturally occurring peptides with high affinity and selectivity for -opioid receptors. These opioids have a D-amino acid residue in position 2 of their sequence and an amide residue at C terminus that protect them from enzymatic hydrolysis (28). In this study, we show that, when injected intravenously in mice, deltorphins enter BBB to produce antinociception. Moreover, using isolated bovine brain microvessels as an in vitro model of the BBB, we investigate the presence of transport system(s) for deltorphins in the plasma membranes of the endothelial cells and whether and how such system(s) are susceptible to regulation, in particular at the metabolic level. MATERIALS AND METHODS Chemicals. Tyr-ala-Phe-Asp-Val-Val-Gly-NH 2 (deltorphin I; ala, D-Ala), Tyr-ala-Phe-Glu-Val-Val-Gly-NH 2 (deltorphin II), and Tyr-ala-Gly-NMePhe-glycinol (DAMGO) were obtained from Bachem. Tyr-D-Pen-Gly-Phe-D-Pen (DPEDPE), Tyr-ala-Phe-Gly-Tyr-Pro-Ser-NH 2 (dermorphin), Tyr-Gly- Gly-Phe-Leu-OH (Leu-enkephalin), Tyr-Gly-Gly-Phe- Met-OH (Met-enkephalin), Hepes, naloxone, and naltrindole were from Sigma); morphine HCl was from Salars (Como, Italy); carboxyfluorescein diacetate was from Molecular Probes. [Lys 7 ]dermorphin was synthesized as indicated by Negri and coworkers (28). All other chemicals were obtained from Merck or Fluka. [Tyrosyl-3,5-3 H]Leu-enkephalin (37 Ci mmol; 1 Ci 37 GBq), [tyrosyl-3,5-3 H]deltorphin I (35 Ci mmol), [tyrosyl-3,5-3 H]deltorphin II (54.5 Ci mmol), [tyrosyl-3,5-3 H]DAMGO (37 Ci mmol), [phenyl-3,4-3 H]U {( )-(5,7,8 )-N-methyl-N[7-(1-pyrrolidinyl)-1-oxaspiro- (4,5)dec-8-yl]-benzeneacetamide} (47 Ci mmol), L-[3,4-3 H(N)]glutamine (60 Ci mmol), L-[U- 14 C]leucine (300 mci mmol), L-[U- 14 C]tyrosine (450 mci mmol), [U- 14 C]sucrose (4.63 mci mmol), and Aquasol-2 were obtained from New England Nuclear. Intracerebroventricular (i.c.v.) and Intravenous (i.v.) Injections. Male C57BL6 mice weighing g (Charles River Breeding Laboratories) were housed singly in cm cages placed in a thermostatically controlled cabinet at an environmental temperature of 21 C. Under light diethyl ether anesthesia drug solution or vehicle were delivered into the lateral cerebral ventricle (i.c.v.) by using a modification of the method of Haley and McCormick (29). An incision was made in the Abbreviations: ala, D-Ala; BBB, blood-brain barrier; BBB-PI, BBB permeability index; DAMGO, Tyr-ala-Gly-NMePhe-glycinol; DPEDPE, Tyr-D-Pen-Gly-Phe-D-Pen; i.c.v., intracerebroventricular; i.v., intravenous; U-69593, ( )-(5,7,8 )-N-methyl-N[7-(1- pyrrolidinyl)-1-oxaspiro(4,5)dec-8-yl]-benzeneacetamide. *To whom reprint requests should be addressed. strom@ dbu.roma1.it. 9469

2 9470 Pharmacology: Fiori et al. Proc. Natl. Acad. Sci. USA 94 (1997) scalp and the bregma was located. Injections were made directly through the skull at a point 2 mm caudal and 2 mm lateral to bregma at a depth of 3 mm using a Hamilton microliter syringe with a 26-gauge needle. All i.c.v. injections were made in a volume of 5 l. For i.v. injections the mouse tail was immersed in moderately hot water (about 40 C) for 1 2 min to induce venous dilatation and the injections carried out using a 500 l Hamilton syringe with a 25-gauge needle. Test of Antinociception. Antinociception responses were determined by immerging the mouse tail into hot water (55 C). The latency of the first sign of a rapid tail-flick was taken as the end point, as indicated by Janssen et al. (30). Before drug administration, mice not flicking their tails within 5 sec from hot water immersion were eliminated from the study. After drug injection, animals not flicking their tails within 15 sec were removed from the nociceptive stimulus and assigned a maximal antinociceptive response. The test was repeated every 15 min after i.c.v. or i.v. drug injection and time course antinociception curves were drawn. Peak response (PR), area under the curve (AUC), and PR AUC ratios were calculated for the dose producing 50% of the maximum response (ED 50 ). Five mice were tested for each opioid dose. BBB Permeability Index. The values of i.v. and i.c.v. ED 50 were multiplied by the respective PR AUC ratio to normalize them for degradation and elimination processes, and the BBB permeability index (BBB-PI) was calculated as the ratio between i.c.v. i.v. normalized ED 50 values. Compounds that more easily enter BBB have therefore higher BBB-PI values. Isolation of Microvessels. Fresh bovine brain, obtained from the local slaughterhouse, was transported on ice to the laboratory, where meninges were removed. The capillaries were isolated from brain cortex as described by Cangiano et al. (9). These microvessels have been shown, by scanning electron microscopy and phase contrast light microscopy, to be practically free from contamination by glial or nerve cells (11). The preparations appeared to consist essentially of branching microvessel segments enveloped in their basement membrane. Their ATP content, measured by a luciferin luciferase assay, was around pmol mg protein. Passive cell permeability, tested by measuring at different time intervals the efflux of carboxyfluorescein from cells previously loaded with carboxyfluorescein diacetate (31), was negligible; also negligible was the uptake of labeled sucrose. As compared with the brain gray matter, the microvessels were found to be 10- to 15-fold enriched in some typical enzymatic activities such as -glutamyltranspeptidase and alkaline phosphatase (9). Their contamination with glia, estimated through the immunochemical determination of glial fibrillar acidic protein, was below 3% on a protein basis (11). Uptake Measurements. The initial rate of uptake of labeled peptides by the isolated brain microvessels was measured within the first minute after addition as a function of peptide concentration. Alternatively, the uptake was measured cumulatively, using a fixed peptide concentration, after increasing time intervals (up to 30 min). For actual measurements, 0.6-ml aliquots of the microvessels suspension, after incubation with the labeled substrate in a phosphate bicarbonate Hepes isotonic buffer (ph 7.4) (choline being, if required, substituted to Na ions), were poured on a 44- m pore nylon sieve on a Millipore vacuum manifold and washed three times with 5 ml of the appropriate cold buffer (9). The sieves with the retained microvessels were then placed in disposable tubes containing 1.8 ml of 1 M NaOH, left overnight at room temperature, and then sonicated for 1 min. Portions were taken for protein determination according to Lowry et al. (32), and 0.5-ml aliquots were transferred to liquid scintillation vials containing an equal volume of 1 M HCl. After addition of 8 ml of Aquasol-2, the vials were counted in a Beckman 9800 liquid scintillation spectrometer. Nonspecific radioactivity, due to the binding of labeled peptide to the nylon sieve, when tested Table 1. BBB-PI of deltorphins compared to morphine and other opioid peptides Compound i.v. ED 50, nmol i.c.v. ED 50, nmol BBB-PI 10 4 Deltorphin I Deltorphin II DAMGO Dermorphin [Lys 7 ]dermorphin Morphine For calculation of BB-PI, see text. in triplicate by omitting the microvessels, was constantly below 100 dpm. By plotting the initial rate of uptake, v, as a function of peptide concentration, a typical saturation kinetics pattern was found, often superimposed to a diffusional component that could, however, be easily subtracted (11, 33). Nonlinear regression analysis, performed on the v vs. v [S] Eadie Hofstee plot, which allows the use of simple statistical procedures (34), was used to obtain the optimal estimates of the kinetic parameters of the transport system(s) and to evaluate the standard error impending on them. Because labeled sucrose had been shown not to permeate the endothelial cells (9), the intracellular water space could be evaluated by a procedure similar to that suggested by Betz et al. (35) i.e., as the difference between the wet and dry weights of a microvessels sample collected on a filter minus the sucrose water space. Opioid Binding Sites in Brain Microvessels. [ 3 H]Deltorphin I binding experiments were performed on membrane preparations obtained from homogenates of brain microvessels. Briefly, brain microvessels prepared as described above were centrifuged at 40,000 g and the pellets homogenized for 3 min in Tris HCl buffer (50 mm, ph 7.4) with a Polytron PT 3000 homogenizer (Kinematica, Littau-Luzern, Switzerland). The homogenate was centrifuged at 1,000 g to eliminate nuclei and debris, and supernatant was collected and centrifuged at 40,000 g. Pellets were washed three times with buffer and finally resuspended in the buffer at a protein concentration equal to that of the original microvessel prep- FIG. 1. Effect of temperature on the uptake of [ 3 H]deltorphin I by isolated brain microvessels. After standing for 10 min either at 37 C (E) orat4 C(F), the labeled deltorphin I was added and the time course uptake was followed for 30 min. The data in the figure represent the average values SD obtained from three different experiments, each performed in triplicate.

3 Pharmacology: Fiori et al. Proc. Natl. Acad. Sci. USA 94 (1997) 9471 Table 2. Effects of glucose and of Na ions on the time course uptake of deltorphins I and II Glucose (10 mm) Na ions (120 mm) Cumulative uptake, pmol mg protein 10 sec 5 min 15 min Deltorphin I (50 M) Deltorphin II (50 M) After the addition of labeled deltorphin, portions of the microvessels suspension were withdrawn at fixed time intervals and collected as described. Data shown are the average values of three different experiments SD. aration ( 2 mg ml). The assay mixtures contained, in a final volume of 2 ml, 1 ml of membrane preparation, 0.1 ml of 5 nm tritiated ligand, and 0.9 ml of Tris buffer. The -, -, and -binding sites were labeled with [ 3 H]DAMGO, [ 3 H]deltorphin I, and [ 3 H]U-69593, respectively. Nonspecific binding was determined in the presence of 50 M naloxone, 5 M naltrindole, or 5 M deltorphin I. After 45 min incubation at 35 C, the samples were cooled to 4 C and the free ligand was separated from membrane-bound ligand by filtration under reduced pressure over Whatman GF B filters presoaked with 0.5% polyethyleneimine. RESULTS In Vivo Evidence of Deltorphins Permeation Through the BBB. Antinociception studies demonstrated that deltorphins I and II were the peptides with the highest BBB-PI among the opioid peptides tested, thus most capable of reaching, upon i.v. injection, the brain compartment (Table 1). Deltorphins indeed appeared to permeate BBB 4 12 times better than typical -opioid receptor agonists such as DAMGO and dermorphin, though times less effectively than morphine and 2 3 times less than [Lys 7 ]dermorphin, an opioid peptide with an unusually high BBB penetration rate (28). Deltorphin Uptake by Isolated Brain Microvessels. Exposure of isolated brain microvessels to either deltorphin I or deltorphin II resulted in a time-dependent increase, within the microvessels, of 3 H-labeled radioactivity. The uptake process reached equilibrium within the first 20 min of uptake and appeared to be temperature-dependent, because the uptake at 4 C was less than one-fourth of the total amount transported at 37 C, and remained linear for over 30 min (Fig. 1). Similar results were also obtained with DAMGO, a -agonist that also has a D-amino acid in position 2 of its sequence. The uptake of labeled deltorphin was inhibited by the presence, in the incubation medium, of high concentrations of the same unlabeled compound; this self-inhibition indicated that the uptake was a saturable process. To clarify whether the isolated microvessels were indeed able to concentrate deltorphins inside the endothelial cells, the in out ratio was calculated according to Betz et al. (35). It could be shown that, under the conditions of Fig. 1, the in out ratio reached a value around 1 after 30 min incubation at 37 C, indicating that the isolated microvessels were unable to concentrate deltorphins. The uptake of del- FIG. 2. Eadie-Hofstee plot (v vs. v [S]) of the saturable component of deltorphin II uptake performed in the presence (E) or in the absence (F)ofNa ions in the incubation medium. Isolated microvessels were assayed for deltorphin II uptake for 1 min at 37 C, between the concentrations ranging from 25 to 1,000 M. After subtraction of the nonsaturable component, the obtained saturation curves (shown in the Inset) or their Eadie Hofstee transformation were used to estimate the V max and the K m values. Data shown are the average values of three different experiments SD.

4 9472 Pharmacology: Fiori et al. Proc. Natl. Acad. Sci. USA 94 (1997) Table 3. Effects of some opioid peptides on the time course uptake of [ 3 H]deltorphin II by bovine brain microvessels Cumulative uptake, pmol mg protein 10 sec 5 min 15 min 20 min No addition Leu-enkephalin 0.2 mm DPEDPE (0.2 mm) Dermorphin (0.2 mm) DAMGO (0.2 mm) The uptake was started by mixing at 37 C, the microvessels suspension with the same buffer but with the addition of the labeled deltorphin II and the peptide indicated. Portions of the microvessels suspension were withdrawn at fixed time intervals and collected as described. Data shown are the average values of three different experiments SD. torphins was markedly decreased in the absence of Na ions, whereas it was not affected by the absence of glucose (Table 2). Evaluation of the Kinetic Parameters of Deltorphin Transport. The kinetic analysis of deltorphin transport was performed by measuring, at different substrate concentrations and in the presence or absence of Na ions, the uptake of labeled deltorphin in the first minute after its addition to the microvessels equilibrated at 37 C. The V max for deltorphin I was found to have a value of pmol mg protein in the presence of Na ions and pmol mg protein in their absence, whereas the apparent K m value was of Min the presence of Na ions and M in their absence (Fig. 2). The kinetic parameters of deltorphin II had similar values. Competition with Other Opioid Peptides. To further characterize the selectivity of the deltorphin uptake system in bovine brain capillaries, four opioid peptides were assayed for their capacity to compete for deltorphin transport. At 0.2 mm concentration, neither Leu-enkephalin, dermorphin, DAMGO, or DPEDPE were capable of modifying the time course uptake of deltorphin by microvessels (Table 3). Competition with Amino Acids. Some of us had shown (36) that, under certain conditions, the A- and L-systems of neutral amino acids transport could cooperate at the BBB level, and that long-chain polar amino acids such as Gln or Met are able to use either system of transport and thus to act as regulatory agents. By checking whether, when added to the incubation medium, given amino acids could compete with deltorphins for the brain capillary transport, we could show that the uptake of deltorphin II was unaffected by the presence of short-chain neutral amino acids such as -methylaminoisobutyric acid, which is specific for the A-system of amino acid transport, as well as by the addition of L-leucine or L-tyrosine, which are selectively transported by the L-system, or of L-glutamine, Table 4. Effects of some neutral amino acids on the time course uptake of [ 3 H]deltorphin II by bovine brain microvessels Cumulative uptake, pmol mg protein 10 sec 5 min 15 min 20 min No addition L-leucine (5 mm) L-tyrosine (1 mm) L-Glutamine (5 mm) MeAIB (5 mm) The uptake was started by mixing, at 37 C, the microvessels suspension with the same buffer but with the addition of the labeled deltorphin II and the desired amino acid. Portions of the microvessels suspension were withdrawn at fixed time intervals and collected as described. Data shown are the average values of three different experiments SD. MeAIB, -methylaminoisobutyric acid. FIG. 3. Overshoot effect in the time course uptake of [ 3 H]deltorphin II by Gln-preloaded microvessels. After 20 min preloading at 37 C with 20 mm Gln (E) or in Gln-free buffer (F), the microvessels were washed and resuspended in prewarmed (37 C) Gln-free buffer. Immediately after the resuspension, labeled deltorphin II was added and the cumulative uptake followed for 20 min (solid lines). Gln efflux is also shown (å, dashed line). The data in the figure represent the mean values SD of three different experiments. which is a good substrate for both amino acid transport systems (Table 4). Glutamine Effect on the Deltorphin Transport System(s). The initial rate of deltorphin uptake was found to be markedly increased if the microvessels were preloaded with L-glutamine. The magnitude of this increase was inversely related to the time interval between glutamine loading and the deltorphin uptake assay. In the preloaded microvessels, cumulative del- FIG. 4. Effect of the addition of external deltorphin II on Gln efflux from brain microvessels. Microvessels were preloaded with [ 3 H]Gln (final concentration, 20 mm) for 20 min at 37 C, rapidly washed, and then resuspended in warm Gln-free buffer (E) orin Gln-free buffer containing 0.5 mm L-leucine ( ) or 0.2 mm deltorphin II (F). At fixed time intervals the isolated microvessels were rapidly washed and the internal radioactivity was measured. Data shown are the mean values SD of three different experiments, each performed in triplicate.

5 Pharmacology: Fiori et al. Proc. Natl. Acad. Sci. USA 94 (1997) 9473 shown). Specific binding was absent, and nonspecific binding was less than 1% of the added tritiated ligand. FIG. 5. Effect of naloxone and of naltrindole on the uptake of [ 3 H]deltorphin II by isolated brain microvessels. The microvessels were preincubated for 10 min at 37 C in the presence of 1.5 mm naloxone (F), 1 mm naltrindole (å), or in their absence (E); labeled deltorphin II was then added and the cumulative uptake was measured at various time intervals. The data in the figure represent the average values SD of three different experiments, each performed in triplicate. torphin uptake curves had a typical overshoot profile (Fig. 3) quite similar to that observed for the glutaminestimulated uptake of large hydrophobic neutral amino acids (9) and deltorphin influx was indeed paralleled by an increased efflux rate of glutamine (Fig. 4). Naloxone and Naltrindole Effects on Deltorphin Transport. Addition of 1.5 mm naloxone to the incubation medium failed to affect to any significant extent either the overall permeability of the endothelial cell membranes, as estimated from the rate of carboxyfluorescein efflux, or the transport systems for glucose and for small or large neutral amino acids (data not shown). However, naloxone, but not naltrindole, was a strong inhibitor of deltorphin uptake (Fig. 5); the value of the in out ratio was reduced to that of the diffusional component observed at 4 C. Naloxone also abolished the stimulating effect exerted by deltorphin on glutamine efflux from glutamineloaded microvessels (Table 5). Absence of Opioid Receptors in Bovine Brain Microvessels. In in vitro binding studies, -, -, and -opioid receptors were undetectable in our microvessel preparations (data not Table 5. Effect of external deltorphin and or of naloxone on Gln efflux from brain microvessels Gln efflux, nmol mg protein 10 sec 5 min 15 min 20 min Buffer alone Deltorphin II Deltorphin II naloxone Naloxone Microvessels were preloaded with [ 3 H]Gln (external concentration, 20 mm; specific activity, 1.27 mci mmol) for 20 min at 37 C, rapidly washed, and then resuspended in warm Gln-free buffer that contained, when indicated, 0.2 mm deltorphin II and or 1.5 mm naloxone. At fixed time intervals the isolated microvessels were rapidly washed and the internal radioactivity was measured. Data shown are the average values of three different experiments SD. DISCUSSION In vivo antinociception studies demonstrated that deltorphins are opioid peptides with an unusually high ability of penetration through the mouse BBB. These results confirm those recently obtained by Thomas et al. (37) with deltorphin analogs, using primary endothelium cultures as in vitro BBB model. These authors demonstrated that the transendothelial passage of deltorphins I and II and of some synthetic analogs, which have high BBB permeability coefficients, does not correlate to the molecular weight or to the lipophilicity of the peptide, but critically depends on the N-terminal sequence and on the amino acid residues present in positions 4 and 5. Our present experiments show that the uptake of deltorphins by brain microvessels is mediated, at 37 C, by a saturable nonconcentrative permeation system that appears to be quite distinct from the system(s) that can be utilized by other opioid neuropeptides or which are specific for short- or long-chain neutral amino acids. This deltorphin-specific transport system is insensitive to the presence or absence of glucose, but its affinity for deltorphins decreases by approximately one-third if Na ions are absent from the medium. The efficiency of deltorphin uptake in the presence of Na ions, evaluated at peptide concentrations below 100 M (i.e., in terms of the V max K m ratio), was around 2.2 l mg of protein per min, whereas at high substrate concentrations the maximal initial rate was around 250 pmol mg of protein per min; these values are definitely too high to account for any peptide binding not linked to permeation. Deltorphin uptake was inhibited by millimolar concentrations of naloxone, in the presence of which only a small diffusional component could be detected. Nevertheless, naltrindole, a selective antagonist to -opioid receptors (36), was ineffective. Other opioid receptor ligands, namely Leuenkephalin and DPEDPE, which preferentially bind to receptors, or dermorphin or DAMGO, which have a high selectivity for receptors, were equally unable to modify the deltorphin uptake. Finally, binding experiments with highly selective ligands failed to demonstrate the presence of -opioid receptors in bovine brain endothelial cells. Taken together, these results indicate that the deltorphin uptake by brain microvessels is not mediated by opioid receptors. The transient stimulation of deltorphin uptake occurring in microvessels preloaded with L-glutamine and the increased efflux of glutamine upon addition of deltorphin to the extracellular compartment suggest the possibility of a countertransport through some pathway common to glutamine and to deltorphins. This hypothesis receives further support from the inhibitory effect exerted by naloxone not only on deltorphin uptake but also on deltorphin-stimulated glutamine efflux. These results indicate that the coupling between the two phenomena cannot be ascribed to a mere osmotic effect caused by glutamine efflux. However, it should be mentioned that a similar hypothesis of countertransport involving glutamine efflux has already been taken into consideration (9) as a mechanism capable of increasing the uptake of large hydrophobic amino acids and that it apparently holds also for Leu-enkephalins (26). These various permeants undoubtedly utilize different pathways, but there might be some common pathway capable of accelerating the coupled uptake processes. This phenomenon may have some interesting implications at the metabolic level. For example, high levels of glutamine occur in the cerebral compartment as a consequence of hyperammonemia (10), and hence some glutamine overload in the brain capillaries may be expected in patients with hepatic encephalopathy. An increase in the permeability of the BBB to opioids and

6 9474 Pharmacology: Fiori et al. Proc. Natl. Acad. Sci. USA 94 (1997) related peptides may thus be an additional factor in the pathogenesis of this complex neuropsychiatric syndrome and clearly deserves further attention. The skilled technical assistance of Mr. Vincenzo Peresempio is gratefully acknowledged. This work received financial support from the Italian Ministry of University and of Scientific and Technological Research (Progetti di Ateneo, Università di Roma La Sapienza, and Progetto di Interesse Nazionale Biochimica di sistemi sopramolecolari ) and from European Commission Contract BMH4-CT Oldendorf, W. H. (1971) Am. J. Physiol. 221, Rapoport, S. I. (1976) Blood Brain Barrier in Physiology and Medicine (Raven, New York). 3. Daniel, P. M., Pratt, O. E. & Wilson, P. A. (1977) Proc. R. Soc. London B 196, Bradbury, M. W. B. (1979) The Concept of a Blood Brain Barrier (Wiley, Chichester, U.K.). 5. Greenwood, J., Begley, D. J. & Segal, M. B. (1995) New Concepts of a Blood Brain Barrier (Plenum, New York). 6. Brightman, M. W. (1977) Exp. Eye Res. 25 (Suppl.), Braun, L. D., Miller, L. P., Pardridge, W. M. & Oldendorf, W. H. (1985) J. Neurochem. 44, Pardridge, W. M., Boado, R. J. & Farrell, C. R. (1990) J. Biol. Chem. 265, Cangiano, C., Cardelli-Cangiano, P., James, J. H., Rossi-Fanelli, F., Patrizi, M. A., Brackett, K. A., Strom, R. & Fischer, J. E. (1983) J. Biol. Chem. 258, Cardelli-Cangiano, P., Cangiano, C., James, J. H., Ceci, F., Fischer, J. E. & Strom, R. (1984) J. Biol. Chem. 259, Cardelli-Cangiano, P., Fiori, A., Cangiano, C., Barberini, F., Allegra, P., Peresempio, V. & Strom, R. (1987) J. Neurochem. 49, Betz, A. L., Firth, J. A. & Goldstein, G. W. (1980) Brain Res. 192, Betz, A. L. & Goldstein, G. W. (1978) Science 202, Betz, A. L. (1983) J. Neurochem. 41, Smith, Q. R. & Rapoport, S. I. (1984) J. Neurochem. 42, Fishman, J. B., Rubin, J. B., Handrahan, J. V., Connor J. R. & Fine, R. E. (1987) J. Neurosci. Res. 18, Pardridge, W. M. & Mietus, L. J. (1981) Endocrinology 109, Kumagai, A. K., Eisenberg, J. B. & Pardridge, W. M. (1987) J. Biol. Chem. 262, Banks, W. A. & Kastin, A. J. (1990) Am. J. Physiol. 259, E1 E Zlokovic, B. V., Segal, M. B., Davson, H., Lipovac, M. N., Hyman, S. & McComb, J. G. (1990) Endocrinol. Exp. 24, Banks, W. A. & Kastin, A. J. (1994) Peptides 15, Poduslo, J. F., Curran, G. L. & Berg, C. T. (1994) Proc. Natl. Acad. Sci. USA 91, Williams, S. A., Abbruscato, T. J., Hruby, V. J. & Davis, T. P. (1996) J. Neurochem. 66, Negri, L., Lattanzi, R. & Melchiorri, P. (1995) Br. J. Pharmacol. 114, Weber, S. J., Abbruscato, T. J., Brownson, E. A., Lipkowski, A. W., Polt, R., Misicka, A., Haaseth, R. C., Bartosz, H., Hruby, V. J. & Davis, T. P. (1993) J. Pharmacol. Exp. Ther. 256, Brownson, E. A., Abbruscato, T. J., Gillespie, T. J., Hruby, V. J. & Davis, T. P. (1994) J. Pharmacol. Exp. Ther. 270, Fiori, A., Savi, M. R., Cardelli, P., Barbaro, K., Parisi, M., Negri, L., Melchiorri, P. & Strom, R. (1997) Ital. J. Biochem. 46 (Suppl.), Erspamer, V., Melchiorri, P., Falconieri-Erspamer, G., Negri, L., Corsi, R., Severini, C., Barra, D., Simmaco, M. & Kreil, G. (1995) Proc. Natl. Acad. Sci. USA 86, Haley, T. J. & McCormick, W. G. (1957) Br. J. Pharmacol. Chemother. 12, Janssen, P. A. J., Niemegeers, C. J. E. & Dorg, J. G. H. (1963) Arzneim.-Forsch. 13, Strom, R., Scioscia-Santoro, A., Crifò, C., Bozzi, A., Mondovì, B. & Rossi-Fanelli, A. (1973) Eur. J. Cancer 9, Lowry, O. H., Rosenbrough, N. H., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, Cardelli-Cangiano, P. & Cangiano, C. (1990) in Current Aspects of the Neurosciences, ed. Osborne, N. N. (Macmillan, London), Vol 1, pp Wilkinson, G. N. (1961) Biochem. J. 80, Betz, A. L., Csejtey, J. & Goldstein, G. W. (1979) Am. J. Physiol. 236, C96 C Strom, R., Cardelli, P., Fiori, A., Cangiano, C., Barra, D., Cascino, A., Ceci, F. & Rossi Fanelli, F. (1992) Amino Acids 2, Thomas, S. A., Abbruscato, T. J., Hau, V. S., Gillespie, T. J., Zsigo, J., Hruby, V. J. & Davis, T. P. (1997) J. Pharmacol. Exp. Ther. 281,