Four Na+-dependent transporters of neutral amino acids (NAA) are known to exist in the abluminal membranes (brain side) of the blood-brain barrier (BBB). This article describes the kinetic characteristics of systems A, ASC, and N that, together with the recently described Na+-dependent system for large NAA (Na+-LNAA), provide a basis for understanding the functional organization of the BBB. The data demonstrate that system A is voltage dependent (3 positive charges accompany each molecule of substrate). Systems ASC and N are not voltage dependent. Each NAA is a putative substrate for at least one system, and several NAA are transported by as many as three. System A transports Pro, Ala, His, Asn, Ser, and Gln; system ASC transports Ser, Gly, Met, Val, Leu, Ile, Cys, and Thr; system N transports Gln, His, Ser, and Asn; Na+-LNAA transports Leu, Ile, Val, Trp, Tyr, Phe, Met, Ala, His, Thr, and Gly. Together, these four systems have the capability to actively transfer every naturally occurring NAA from the extracellular fluid (ECF) to endothelial cells and thence to the circulation. The existence of facilitative transport for NAA (L1) on both membranes provides the brain access to essential NAA. The presence of Na+-dependent carriers on the abluminal membrane provides a mechanism by which NAA concentrations in the ECF of brain are maintained at ∼10% of those of the plasma.
- active transport
- brain extracellular fluid
- endothelial cells
- essential amino acids
availability of essential neutral amino acids (NAA) is necessary for normal brain function. Because NAA are hydrophilic, transport systems are necessary to transfer them across the blood-brain barrier (BBB). Large essential NAA enter the brain on system L11, a high-affinity facilitative system that is present in the luminal and abluminal membranes of the endothelial cell (23, 24, 26, 32, 36, 37). Nonessential NAA, most of which are smaller, have not been found to pass from blood to brain as readily; nevertheless, the uptake of small nonessential amino acids by the brain was greater than zero, suggesting some transport or leakage at a slow rate (1, 23, 24).
The concentrations of all naturally occurring amino acids in the cerebrospinal fluid [presumably similar to the extracellular fluid (ECF) of brain], with the exception of glutamine, are ∼10% of those in plasma (17, 19, 20, 35). This situation cannot be explained by the consumption of NAA by brain, because the arteriovenous differences across the brain of most NAA are imperceptible (7, 8, 29), as are the arteriovenous differences of ammonia, a by-product of amino acid catabolism (6). These observations indicate that NAA leave the brain against a concentration gradient at the same rate they enter.
The maintenance of an NAA gradient between plasma and ECF requires the expenditure of energy. The situation could be explained by the presence of several Na+-dependent carriers located in the abluminal membranes of the endothelial cells of the BBB. These systems include A1 (2, 30, 32), ASC1 (11, 38, 39), and N1 (18), as well as a recently described system that transports large NAA (21). The latter system has not been named but will be referred to as Na+-LNAA.
Na+-dependent transport of NAA exists only in the abluminal membranes. No Na+ dependency has been detected in the luminal membrane, which appears to have only facilitative carriers (1, 4, 12, 23, 24, 33, 34, 36). Therefore, the Na+-dependent transporters are in a position to remove NAA from brain by utilizing the Na+ gradient that exists between the ECF and the endothelial cells of brain capillaries comprising the BBB.
Although four Na+-dependent transporters of NAA have been identified in the BBB, a study of transport characteristics including a description of the kinetic activity, the sensitivity to the transmembrane potential, and the spectrum of NAA transported has only been determined for Na+-LNAA (21). The objective of this article is to examine systems A, ASC, and N, thereby providing a more complete picture of how the concentrations of NAA are regulated in the ECF of brain. The results describe these characteristics and show that every naturally occurring NAA is a substrate for one or more Na+-dependent systems.
MATERIALS AND METHODS
l-[14C]-N-(methylamino)-isobutyric acid (MeAIB; 55–124 mCi/mmol), l-[14C]glutamine (200–250 mCi/mmol), and l-[3H]alanine (60 mCi/μmol) were purchased from American Radiolabeled Chemicals (St. Louis, MO). Amino acids, collagenase Type IA, and MeAIB were bought from Sigma (St. Louis, MO), and the Bio-Rad protein assay was purchased from Bio-Rad Laboratories (Hercules, CA).
Fresh bovine brains were purchased from Aurora Packing (North Aurora, IL). The cows were killed for food under USDA supervision, and the meat was sold for human consumption.
Isolation and Characterization of Membrane Vesicles
Membrane vesicles from brain endothelial cells were prepared as previously described (30). Briefly, microvessels (predominantly capillaries, but some adhering arterioles and venula) were isolated from bovine cerebral cortices as described by Pardridge et al. (28). The microvessels were digested with collagenase Type IA to remove the basement membrane, pericytes, and glial fragments. The refined microvessels were homogenized to release the membranes, which were then separated into five fractions at the interfaces of a discontinuous Ficoll gradient (0, 5, 10, 15, and 20%) (30, 31). The amount of abluminal and luminal membrane in each fraction was determined by the activities of two markers previously demonstrated to be located exclusively in one of the membranes. System A transport of MeAIB was used as the marker of the abluminal membrane, and GGT (γ-glutamyl transpeptidase) was used as the marker of the luminal membrane (30, 31). Fractions F10/15 and F15/20 were combined, resulting in a preparation containing 70% abluminal membrane. Although contamination by luminal membranes could be ignored, because Na+-dependent transport systems are not present in this side, removal of luminal membranes reduced the background due to facilitative transport. The vesicles were suspended in buffer (290 mM mannitol, 10 mM HEPES, pH 7.4) and stored at −70°C.
Measurement of Transport Rates
Rates of transport were measured by rapid filtration (30). Membrane vesicles were thawed, centrifuged at 37,500 g for 25 min at 4°C, and suspended in storage buffer (290 mM mannitol, 10 mM HEPES, pH 7.4). Vesicles were allowed to equilibrate overnight at 4°C. The final concentration of protein was between 2.5 and 5 μg protein/μl. The vesicle suspensions were divided into 5-μl aliquots that were preincubated at 37°C for 1 min before the initiation of transport measurements. Reaction medium containing 3H- or 14C-labeled substrate, with or without competing substrate, was added to initiate the reaction. The concentration of extravesicular NaCl at the start of the measurements was 100 mM, and the internal concentration was 0 unless otherwise stated. Reactions were stopped by adding 1 ml of an ice-cold stopping solution (145 mM NaCl and 10 mM HEPES, pH 7.4), and vesicles were filtered on a 0.45-μm Gelman Metricel filter (Ann Arbor, MI) under vacuum. The filtered membranes were washed four times by 1-ml aliquots of stopping solution, after which the filters were counted by liquid scintillation spectroscopy. Unless otherwise indicated, all rates were corrected for nonspecific transport, binding, or trapping, as measured in the presence of a saturating dose of unlabeled substrate.
Determination of Kinetic Characteristics
Apparent Km and Vmax values were determined by measuring the initial rates of transport at various concentrations. From these data, nonlinear regression analyses were performed using Sigma Plot (SPSS, Chicago, IL), and apparent Km and Vmax values were calculated. The permeability-to-surface area products, or clearance, was obtained by dividing the rate of transport of tracer by the substrate concentrations or by dividing the Vmax by the Km.
In some experiments, NAA or inhibitors were included in the reaction media. The velocity with and without the putative substrates was measured, and the percent inhibition was determined.
Creating Transmembrane Potentials
Valinomycin (12.5 μg/mg protein), a K+-specific ionophore, was used to increase the permeability of vesicles to K+. Different ratios of internal and external concentrations of K+ were used to establish calculated potential differences ranging from −70 mV to +18 mV. The vesicles were prepared overnight with equilibrating solutions containing 25 or 100 mM KCl, 10 mM HEPES, and mannitol to bring the final osmolarity to 300 mosM.
The reaction was initiated by the addition of an appropriate volume, and composition, of extravesicular solution to create the desired initial transmembrane potential, as calculated by the Nernst equation (13): where z is the valence, F the Faraday constant, E the voltage, R the gas constant, and T the absolute temperature. For K+ at 37°C and log10: If voltage sensitivity was observed, the following equation for a sigmoid curve was used to fit the data: where a1 is the minimal rate, a2 is the maximal rate, xo is the midpoint, and k is the slope at midpoint. The effective valence (z) associated with transport is where RT/F = 26.7 mV at 37°C (13).
Protein concentrations were determined using the Bio-Rad Protein Microassay (Hercules, CA), with bovine serum albumin as the standard, by the method of Bradford (3).
Curves were fit by Sigma Plot, and data were analyzed with StatView (SAS, Cary, NC) by ANOVA and Fisher's least significant difference test. Values were considered significant at P < 0.05. All individual values are expressed as means ± SE unless otherwise stated.
System A Transporter
The activity of system A, named for its preference for transporting alanine1 (25), may be distinguished from other Na+-dependent carriers by its acceptance of MeAIB as a unique substrate (5). Therefore, the transport of [14C]MeAIB was used to characterize this system.
Kinetic constants of MeAIB transport and effect of the transmembrane potential.
The kinetic constants of MeAIB transport were determined under standard conditions (0 mV transmembrane potential) and at a −18-mV transmembrane potential by measuring the transport of [14C]MeAIB over a range of concentrations (Fig. 1). At 0 mV, the apparent Km was 0.4 ± 0.16 (SE) mM, and the Vmax was 500 ± 60 pmol·mg−1·min−1, whereas at −18 mV, the apparent Km was 0.3 ± 0.14 mM and the Vmax was 900 ± 110 pmol·mg−1·min−1. Comparing the clearance values indicated that the negative transmembrane potential tripled the rate of transport (1.4 compared with 3.4 μl·mg−1·min−1) and suggested that system A was influenced by the transmembrane potential.
The influence of transmembrane potential was further analyzed by measuring the initial rate of [14C]MeAIB transport over a range of transmembrane potentials (Fig. 2). The data, expressed as clearance (in μl·min−1·mg−1), had a maximal value of 10 at transmembrane potentials below −60 mV and a minimal value of 2 at 0 mV. Therefore, the data indicated voltage sensitivity and were compatible with 3 positive charges translocated per MeAIB molecule (please see materials and methods).
Potential substrates of transport.
System A transport was measured in the presence of 15 NAA to determine which were potential substrates (Table 1). Proline, alanine, histidine, asparagine, serine, and glutamine manifested statistically significant degrees of inhibition of MeAIB transport (P < 0.05). Other laboratories (2, 25) reported a similar amino acid spectrum for system A, with the exception of glycine. In our preparation of abluminal membrane vesicles, glycine did not inhibit MeAIB transport significantly, whereas others (2, 25) reported glycine to be a substrate for system A.
System N Transporter
System N, an Na+-dependent transporter of NAA, exists at the abluminal membrane of the BBB (18). System N1 has a preference for NAA that are nitrogen rich, such as glutamine, histidine, and asparagine, hence its designation (15, 16). Because of this preference, [14C]glutamine was used as the exemplary substrate. System N is not inhibited by MeAIB (15, 16); therefore, MeAIB (20 mM) was included in the reaction mixtures to block glutamine transport by system A.
Kinetic constants of glutamine transport and effect of the transmembrane potential.
The kinetic constants of glutamine transport were determined at 0- and at −18-mV transmembrane potentials by measuring the initial rate of [14C]glutamine transport over a range of concentrations (Fig. 3). The apparent Km values were 1.3 ± 0.4 and 2.7 ± 0.5 mM at 0 and −18 mV, respectively, with corresponding Vmax values of 4,400 ± 700 and 6,800 ± 800 pmol·mg−1·min−1. Although the kinetic values were similar to those found previously (18), neither the Km nor the Vmax values measured at the two potentials were different by statistical analyses (P > 0.05). The calculated clearance values were 3.4 and 2.5 μl·mg−1·min−1, respectively. Accordingly, no effect of transmembrane potential was apparent; nevertheless, an additional experiment was conducted to test this observation.
The initial rate of [14C]glutamine uptake was measured as a function of the membrane potential ranging from −70 to +18 mV. No statistically significant relationship between voltage and glutamine transport was observed, establishing that system N was not affected by the transmembrane potential (Fig. 4).
Substitution of Na+ by Li+.
System N will transport amino acids in liver cells with Li+ as the cation as well as Na+ (15). Therefore, to learn whether the BBB system N exhibited this characteristic, transport of [14C]glutamine was measured in the presence of either 100 mM NaCl or 100 mM LiCl (Fig. 5). The results showed that Li+ could substitute for Na+, suggesting that system N in the BBB is similar in this regard to system N in liver cells.
Potential substrates of transport.
The rate of transport of [14C]glutamine was measured in the presence of various NAA to determine which were potential substrates of system N (Table 2). Only histidine, serine, and asparagine were found to inhibit glutamine transport significantly.
System ASC Transporter
Several authors reported system ASC to be present in the BBB (11, 38, 39). Although in earlier work we failed to confirm ASC activity (32), system ASC was studied more closely by measuring the initial rate of [3H]alanine transport in the presence of 10 mM methionine, cysteine, serine, or threonine, all of which are substrates of system ASC substrates (11, 38, 39). MeAIB (20 mM) was included in each reaction mixture to block system A activity. Transport activity was detected, and all four NAA inhibited alanine uptake almost completely (Fig. 6).
Kinetic constants of ASC and effect of the transmembrane potential.
The kinetic constants of alanine transport were measured at calculated transmembrane potentials of 0 mV and −18 mV over a range of alanine concentrations (0.3 μM to 0.8 mM) (Fig. 7). At 0 mV, an apparent Km of 0.11 ± 0.06 mM was obtained, with a Vmax of 660 ± 70 pmol·mg−1·min−1. At −18 mV, the apparent Km did not change (0.11 ± 0.02 mM), and the Vmax was not changed significantly (494 ± 17 pmol·mg−1·min−1; P > 0.05). The clearance values were 6.0 μl·mg−1·min−1 at 0 mV and 4.5 μl·mg−1·min−1 at −18 mV.
The effect of the transmembrane potential was also examined over a greater range of calculated transmembrane potentials. [3H]alanine (0.3 μM) uptake was measured at voltages from −70 to +18 mV. There was no change in transport activity correlated with the transmembrane potential (Fig. 8). Consequently, it was concluded that system ASC is independent of the transmembrane potential.
Potential substrates of transport.
The uptake of [3H]alanine was measured in the presence of various NAA (Table 3). (MeAIB was included to block system A activity.) The NAA that significantly inhibited alanine uptake comprised the spectrum of system ASC substrates, corroborating the reports of others (11, 38, 39); inhibitors included serine, glycine, methionine, valine, leucine, isoleucine, cysteine, and threonine.
Interestingly, glycine significantly inhibited transport by system ASC but, as mentioned above, did not inhibit system A. Other laboratories indicated that glycine was transported by system A. For example, Oxender and Christensen (25) reported that glycine was transported with weak affinity but was not inhibited by α-aminoisobutyric acid (a model substrate of system A). When their experiments were performed in 1963, neither system, ASC or Na-LNAA, had been described. Therefore, transport attributed to system A may have been through some other system with a similar spectrum of amino acid substrates. Betz and Goldstein (2) found only 30–40% inhibition of MeAIB transport by 10 mM glycine. Whatever the case, if glycine is transported by system A in the bovine BBB, the affinity is probably weak.
The data give further support to the concept that both membranes of cerebral endothelial cells participate in regulating the composition of brain ECF and the amino acid content of brain. The two membranes appear to perform in a complementary fashion; facilitative transport occurs at both the luminal and abluminal membranes, whereas Na+-dependent transport of NAA occurs in the abluminal membrane, as illustrated in Fig. 9. Therefore, the brain has access to essential NAA through the facilitative systems and the ability to adjust the composition of ECF by the Na+-dependent systems. A comparison of the kinetic constants of these various systems is presented in Table 4. Because the values for Km and Vmax vary considerably, it is convenient to compare the potential of the various systems by using clearance as a common basis.
The importance of Na+-dependent transport systems to the supply and balance of NAA in brain is just emerging. The primary focus of past research has been on facilitative transport (system L1), which mediates the passage of NAA across the BBB from blood to brain (24, 27, 37). L1 exists in both membranes with similar, but not quite equal, activity; the clearance by the luminal membrane is 32 μl·min−1·mg−1, whereas the clearance by the abluminal membrane is 14 μl·min−1·mg−1 (32). L1 carries nine large NAA in competition with each other and has Km values that are similar to the plasma concentrations (24, 27, 36, 37). Therefore, L1 is predicted to be nearly saturated at the outer surface of the luminal membrane; the determining factors are the plasma concentration and the affinity constants for the respective NAA (27, 36, 37). The critical importance of L1 has been demonstrated in vivo by the close relationship between NAA supply (e.g., tryptophan) and the brain content of serotonin and other neurotransmitters (9, 10).
NAA may also enter brain by a first-order mechanism that resembles diffusion (27). Whether the first-order component is a high-capacity transport system, diffusion through the membranes, or passage on L1 with weak affinity is not yet determined.
The steep gradient of NAA concentrations that exists between plasma and brain for all NAA except glutamine was postulated to be due to a “sink” effect, whereby NAA that entered brain were taken up actively by Na+-dependent systems that exist in the membranes of the various brain cells and presumably metabolized (26, 27). However, there are no consistent arteriovenous differences of NAA (7, 8, 29) or of ammonia (6) to support this explanation.
Evidence for the existence of other systems has been described, such as A (2, 30), ASC (11, 38, 39), and N (18). These systems were presumed to transport small NAA that might be detrimental for brain function. However, this interpretation was changed by the recent discovery of a system able to remove essential as well as nonessential NAA from ECF. Accordingly, there are at least four Na+-dependent systems, all located in the abluminal membrane of endothelial cells, that are capable of using the Na+ gradient between the ECF and endothelial cells to transport NAA against a gradient and in the direction of the circulation (Fig. 9).
Two systems, A and Na+-LNAA (21), are sensitive to the transmembrane potential, transporting 3 and 1 positive charges, respectively, with each molecule of substrate. This characteristic increases the ability of these two systems to move NAA from the ECF of brain into endothelial cells, from which they may diffuse to the circulation by way of facilitative carriers. A summary of the kinetic characteristics of the four known neutral NAA carriers is given in Table 4.
One of the interesting features of the Na+-dependent carriers is the wide spectrum of neutral NAA transported. Two systems transport only nonessential amino acids (systems A and N). However, both systems ASC and Na+-LNAA transport essential as well as nonessential NAA (Table 5). Every neutral amino acid, whether essential or nonessential, is a potential substrate of at least one energy-dependent carrier. Several NAA are carried by three systems (Table 5). Therefore, the BBB is in a position to remove all naturally occurring NAA from the brain ECF, thereby providing a mechanism to control NAA concentrations.
To transport NAA from brain, there must be a route of egress from the endothelial cells into the circulation through facilitative pathways located in the luminal membrane. It is possible that L1, which transports primarily large, essential NAA, performs this function. Review of early studies by Oldendorf (23) also showed that the penetration of small nonessential NAA into brain, although small, was statistically significant. Oldendorf's data showed that several small NAA had a low but positive brain uptake index, including threonine, cysteine, serine, alanine, proline, and glycine. Furthermore, with the exception of proline and glycine, substrate inhibition could be demonstrated, suggesting the possibility of carrier-mediated transport. Thus it seems possible that small, nonessential NAA have facilitative mechanisms that assist their movement across the BBB.
The two membranes of the BBB appear to transport NAA in a fashion similar to that previously described for glutamine and acidic amino acids such as glutamate (12, 18, 22). Na+-dependent carriers are capable of pumping both glutamine (system N) and glutamate (see the EAAT1, -2, and -3 of Ref. 22) from the ECF into endothelial cells. The luminal facilitative carriers for both glutamate (12, 22, 23) and glutamine (12, 18) can then transport them to the plasma.
In conclusion, the four Na+-dependent NAA transporters, located in the abluminal membrane of the BBB, are capable of actively removing NAA from the brain ECF. Their operation will lower the NAA concentration in the ECF. The overlap of substrate specificities suggests that several transporters may participate in this process. In this manner, the concentrations of NAA in the ECF of brain may be maintained at much lower concentrations than those of plasma (19, 20, 35).
This work was supported by a grant from the National Institute of Neurological Disorders and Stroke (NS-041405), by the International Glutamate Technical Committee, and by the Redes de Centro Investigación Instituto Carlos III (RCMN03-08).
We thank M. R. DeJoseph and A. Mokashi for excellent technical assistance. We are also grateful to Drs. Charles McCormack and Héctor Rasgado-Flores for constructive criticism of this article.
↵1 System L (leucine preferring) is a member of the LAT1 family of proteins. System A (alanine preferring) belongs to the SAT family. System ASC (alanine, serine, cysteine preferring) is a member of the ASC family, and system N (glutamine preferring) belongs to the SN family (14).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 by American Physiological Society