Heterogeneity of high affinity uptake of L-glutamate and L-aspartate in the mammalian central nervous system

Life Sciences, Vol. P r i n t e d in the U S A

51, pp.

1467-1478

Pergamon

Press

MINIREVlEW HETEROGENEITY OF HIGH AFFINITY UPTAKE OF L-GLUTAMATE AND LASPARTATE IN THE MAMMALIAN CENTRAL NERVOUS SYSTEM Vladimir J. Balcar and Yi Li Department of Anatomy, The Anderson Stuart Building F13, The University of Sydney, Australia N.S.W. 2006 (Received

in final

form A u g u s t

31,

1992)

Summary Characteristics of high affinity uptake of L-glutamate are examined in order to evaluate the possible use of the uptake of [3H]L-glutamate, [3H]Laspartate or any other suitable [3H]-Iabelled substrate as a marker for glutamatergic and aspartergic synapses in autoradiographic studies in the mammalian brain. Review of data on substrate specificity indicates the presence of at least two high affinity uptake systems specific for acidic amino acids in the central nervous tissue; one which takes up Lglutamate and L-aspartate and the other which is selective for Lglutamate only. Studies on ionic requirements, too, point to the existence of at least two distinct uptake systems with high affinity for L-glutamate. The Na+-dependent uptake system(s) handle(s) both L-glutamate and Laspartate whereas the Na+-independent uptake system(s) show(s) selectivity for L-glutamate only. Available data do not favour the Na +dependent binding of [3H]D-aspartate to thaw-mounted sections of frozen brain tissue as a suitable marker for glutamatergic/aspartergic synaptic nerve endings. However, there are reasons - such as the results of lesion studies and the existence of uptake sites which have a higher affinity for L-aspartate than for D-aspartate - to suggest that Na +dependent binding of [3H]L-aspartate, rather than that of [3H]D-aspartate, should be further investigated as a possible marker for the glutamatergic/aspartergic synapses in the autoradiographic studies using sections of frozen brain. Acidic amino acid L-glutamate was identified as an excitatory substance in the central nervous system (CNS) by Hayashi in 1954 (1). Since then both L-glutamate and the related amino acid L-aspartate have been at the centre of a major research effort. The proposal that L-glutamate and L-aspartate are important excitatory synaptic transmitters has been supported by a large amount of information from pharmacological, biochemical, physiological and anatomical studies performed especially during the last two decades (reviews: 2-6). The focus of the present review is on the chemistry and pharmacology of the "high affinity" uptake of L-glutamate and L-aspartate (glu/asp). The relevant data will be examined for indications that the high affinity uptake of glu/asp is mediated by two or more distinct uptake systems which, Copyright

© 1992

0 0 2 4 - 3 2 0 5 / 9 2 $5.00 + .00 P e r g a m o n Press Ltd All r i g h t s

reserved.

1468

Uptake of L-glutamate and L-aspartate in CNS

Vol.

51, No.

19, 1992

though similar to each other in some aspects, could be distinguished by pharmacological or biochemical means. Special attention will be given to the consequences of the possible heterogeneity of the glu/asp high affinity uptake for the studies using glu/asp uptake as a marker for glutamatergic and aspartergic synaptic nerve endings in anatomical studies. QHARACTERISTICSOF HIGH AFFINITY UPTAKEOF L-GLUTAMATEAND L-ASPARTATE The most commonly cited characteristics of the high affinity uptake of glu/asp are a relatively low Km (usually < 50 #M, hence "high affinity"), strong Na +dependence and rather narrow substrate-specificity based on strict structural requirements (7-11). It differs from the Na+-independent, CI--stimulated uptake of Lglutamate by synaptic vesicles which has a characteristic substrate specificity and much higher Km (12,13). The notion of heterogeneity of the glu/asp high affinity uptake was initially understood mainly in terms of glial v. neuronal location (14-17, for reviews see 3,18,19). However, this view is almost certainly oversimplified because the biochemical characteristics, ionic requirements and substrate specificity of the uptake often vary from one study to another in a manner which does not necessarily match the neuronal or glial nature of the experimental model used. HIGH AFFINITYUPTAKEOF L-GLUTAMATEAS A GLUTAMATERGICSYNAPTICMARKER Location of glutamatergic synapses in the CNS can be visualized by several methods. Some of the techniques are based on histochemistry of enzymes associated with the metabolism of L-glutamate and L-aspartate (20,21) while others use autoradiography of [3H]D-aspartate accumulated via the high affinity glu/asp uptake system and carried by retrograde axonal transport towards the perikarya (22-25). More recently, histochemical visualization of glutamate-like immunoreactivity was developed (26) and it is rapidly becoming the most frequently used technique in the studies on the distribution of glutamatergic/aspartergic neurons in the central nervous tissue. Neither the enzymes nor L-glutamate are localized exclusively in glutamatergic neurons or their terminals. L-Glutamate is rather abundant in brain tissue (27) and it is not inconceivable that some of the glutamate-like immunoreactivity may, in fact, mark metabolic compartments which are not directly associated with synaptic transmission, or, since L-glutamate is the precursor of an inhibitory synaptic transmitter 4aminobutyrate (GABA), even GABAergic synaptic terminals. The autoradiographic tracing of glutamatergic structures by [3H]D-aspartate should have, in principle, no such shortcomings. High affinity uptake of L-glutamate was shown to be associated with a unique synaptosomal fraction (14); potency of high affinity uptake of Lglutamate in synaptosomal fractions prepared from a particular brain region correlates with the amount of excitatory (glutamatergic) input into the region and selectively decreases when the input is interrupted (23,24 for reviews see 28-30). There are also indications that GABAergic neurons do not take up L-glutamate, at least not by the high affinity uptake system (31). However, as Fonnum pointed out when discussing the autoradiography of [3H]D-aspartate-labelled structures, "the success of this method must depend on specific uptake rather than specific transport, otherwise horseradish peroxidase (HRP) . . . . and other proteins used in anatomical transport studies would have to be regarded as neurotransmitters" (3). In other words, for these studies to be meaningful, [3H]D-aspartate (or any other suitable 3H-labelled marker) has to be a typical substrate for the high affinity uptake of glu/asp in glutamatergic and aspartergic nerve terminals.

Vol. 51, No. 19, 1992

Uptake of L-glutamate and L-aspartate in CNS

1469

IONIC REQUIREMENTSOF THE GLU/ASPHIGH AFFINITYUPTAKE Na+-de0endent uotake An important characteristic of glu/asp high affinity uptake is the strong dependence on Na+-gradient across the neuronal or glial membrane (8,31-38). Early studies using brain slices indicated that the ratio of Na + to L-glutamate in the transport process was 1:1 (9). Similar results were obtained in experiments using glial cell lines (33) and primary cultures of glial cells (34,38) but not in studies based on neuronal material (synaptosomes: 35,36; ~t, a neuronal line of cerebellar origin: 37; primary cultures containing cerebellar granule cells: 38) which produced stoichiometric ratio 2:1 in favour of Na +. The data from those experiments may have been influenced by a "neuronal-like activity" (including Na+-dependent depolarization events) since the Lglutamate uptake by synaptosomes was shown to vary with the voltage gradient across the synaptosomal membrane (39,40). However, more recent studies on the Na+-dependence of L-glutamate uptake indicated significant departures from 1:1 stoichiometry even in neuron-free cultured cell lines (glioma C6, 41) and, at least at low concentrations (3.5 - 7.5 I~M) of L-glutamate, also in primary cultures of rat cerebral cortex (31). Thus the often quoted ratio of two Na + to one molecule of Lglutamate may indeed reflect one of the mechanisms linking the Na+-fluxes and Lglutamate transport (42,43). It seems obvious that Na + would be cotransported with one molecule of L-glutamate, that is if it is the fully ionized form of L-glutamate which is translocated across the membrane (42). Two Na + would be required if the transport of one cation (such as K+, vide infra) in the opposite direction is needed to regenerate the carrier (42,43). Determination of the corresponding kinetic parameters may be, however, complicated by the presence of Na+-channels linked to glutamate receptors which exist not only in the neuronal membrane but probably also on glial cells (44). It has also been reported that increased concentrations of K+ reduce glu/asp uptake and this has been attributed to the activation of Ca2+-dependent release induced by depolarizing concentrations of K + (9,45) or to the diminution of the [K+]inside/[K+]outside concentration gradient which is thought to be essential for the regeneration of glu/asp carrier in the neuronal membrane (42). High affinity uptake of L-glutamate which is Na+-dependent but insensitive to increased concentrations of K+ has also been described, not only in cultured 3T3 fibroblasts (16) but also in primary cultures of neonatal glial cells (31). In both these cases, L-glutamate uptake was reduced in the absence of Ca 2+. This could be a characteristic of the high affinity uptake of glu/asp in fibroblasts (16, but see also 46,47) which may be present as a contaminant in primary cultures of glial cells. However, recent experiments with 3T3 fibroblasts (48) detected Na+-dependent high affinity uptake of L-glutamate which was reduced by 60 mM K+ and was not affected by variations in the concentration (0 - 9.6 mM) of Ca 2+. The actual mechanism of the ionic dependence notwithstanding, it may be possible to demonstrate Na+-dependent binding of L-glutamate to the high affinity uptake carrier in homogenates containing plasma membranes. Experiments of this type have been performed (49) and it was reported that the Na+-dependent portion of [3H]L-glutamate binding had structural requirements similar to those observed for high affinity glu/asp uptake in brain slices.

1470

Uptake of L-glutamate

and L-aspartate

in CNS

Vol.

51, No.

19,

1992

Na+-inde0endent uotake It has also been demonstrated using hippocampal synaptosomes that [3H]Lglutamate uptake is influenced by variations in the anion composition of the external medium (50). The uptake, studied in the presence of 126 mM Na +, was reduced by 19 - 33% when CI- was replaced with Br-, I- or CH3COO- and by more than 50% and 80% when NO2- and NO3- were used instead of CI-. Considering similarities of NO2, NO3and a fully ionized carboxyl function in terms of charge distribution, bond angles and electron configuration (51,52), the reduction of L-glutamate uptake could be attributed to the interference of 126 mM NO2 or NO3- (which would be, unlike CH3COO-, fully ionized at pH 7.4 and, therefore, in large excess) with the ionic interaction between the nanomolar concentrations of glutamate carboxyles (also virtually fully ionized under those conditions, 53) and the uptake site, respectively, rather than to the mere absence of CI-. The much smaller variation in the glutamate uptake caused by the replacement of CI- by other anions could be explained by changes in the voltage gradient across the membrane which is known to influence the high affinity uptake of L-glutamate (39,40,41). However, several other groups have demonstrated the existence of high affinity uptake of L-glutamate which was stimulated, even in the absence of Na +, by CI- (cultured glioma C6 cells: 54; synaptosomes: 55; resealed vesicles in the preparation of synaptic plasma membranes, SPM: 56; primary cultures of astrocytes: 57) or by Ca 2+ (vesicles in SPM preparation: 56,58; primary cultures of astrocytes: 57). Whether actually CI--, and sometimes also Ca 2+-, dependent or just voltage-modulated, this uptake may represent (a) separate system(s). It only accounts for a small proportion of total glutamate uptake under most of the experimental conditions studied (55,57), with the possible exception of cultured neuroblastoma cells (vide infra) but it may have a distinct physiological function. SUBSTRATE SPECIFICITY AND STRUCTURAL REQUIREMENTS OF HIGH AFFINITy UPTAKE QI= LGLUTAMATE AND L-ASpARTATE

Exoeriments in brain slices and svnaotosomes Most of the early data was obtained by testing a number of selected compounds as potential inhibitors of uptake of [3H]L-glutamate and [3H]L-aspartate in slices of rat cerebral cortex (9) and spinal cord (59). All of the strong inhibitors were close analogues of glutamate and aspartate. However, not all structures related to glutamate and aspartate were active. D- and L-Glutamines and asparagines, L-glutamate diethyl ester, L-glutamate-5-hydroxamate, N-methyl-D- and L-aspartates and glutamates, NacetyI-L-glutamate and several other compounds with substituents on either carboxyl or amino functions were not inhibitors suggesting that the presence of two negative and one positive charge in the molecule is a condition sine-qua-non for a compound to be able to interact with the active site on the transporter. Moreover, even the compounds with all three groups free (and probably ionized at pH 7.4) were not all equally strong inhibitors. For example, the substitution on the 2-carbon atom abolished or very much reduced activity (2-methyI-DL-glutamate, 2-methyI-DLaspartate) and so did addition of an extra atom to the carbon chain (2-amino adipate, 3-amino adipate). By contrast, the substitutions on the 4-carbon atom (4-fluoro glutamate and 4-methylene glutamate) and especially on the 3-carbon atom (3hydroxy-DL-glutamate, 3-hydroxy-DL-aspartate and 3-methyI-DL-aspartate) were much better tolerated (9,33). Comparison of the activities of the isomers of 3hydroxyaspartate offers even more detailed insight into the structure of the binding site on the transporter. Thus the threo- isomer was a much stronger inhibitor than the erythro- isomer and when the threo- enantiomers were resolved, the L- form was found to be more potent than D- form (9,60). There was also a great difference betwen

Vol.

51, No. 19, 1992

Uptake of L-glutamate

and L-aspartate

in CNS

147]

the D- and L-enantiomers of glutamate, D-glutamate being almost inactive but in the case of aspartate the D-enantiomer seemed to interact with the uptake about equally well as, or even somewhat better than, the "natural" substrate L-aspartate (8,22). Some of the later experiments indicated that the presence of N-substituents may not necessarily lead to the complete loss of the ability of an analogue to interact with the high affinity site (61-63). Specifically, kainate (64) was shown to be an inhibitor of high affinity uptake of [3H]L-glutamate albeit a relatively very weak one (slope inhibition constant, Kis = 253 p~M as compared to 15 - 16 #M for DL-threo-3hydroxyaspartate, L-cysteate and L-cysteine sulphinate or 35 I~M for D-aspartate, 9). This finding was further supported by identification of dihydrokainate as a moderately strong inhibitor (64) and, eventually, a compound was synthesized (L-transpyrrolidine-2,4-dicarboxylate, L-t-PDC) which, although containing no primary amino function on the 2-carbon atom, turned out to be a strong inhibitor of high affinity uptake of [3H]L-glutamate in synaptosomes (65). As in the case of 3-hydroxyaspartate, there were large differences among the inhibitory potencies of various isomers of L-t-PDC, thus confirming very strict structural requirements of the high affinity uptake site. Other conformationally restricted analogues of glutamate (containing free amino group, though) have become available recently (66,67). Most of these compounds interact with glutamate receptors but while trans-l-amino-cyclobutane-l,3-dicarboxylic acid was shown to be a selective agonist on NMDA-type glutamate receptors (68) the cisstereoisomer is a competitive inhibitor of high affinity uptake of L-glutamate in synaptosomes from rat cerebral cortex (66). This is not immediately apparent from the literature because the nomenclature of cyclic compounds with substituents on two of the ring atoms has been confused recently (66, for further discussion see the 1991 Product Catalogue of Tocris Neuramin Ltd., Bristol, U.K., pp.10 - 11 ). The uptake of 1 #M [3H]L-glutamate by synaptosomes studied in the absence of Na + was strongly inhibited by 100 #M quisqualate and ibotenate but only very weakly by 100 I~M L-aspartate (55). Experiments in cultured cells: glial v. neuronal uptake High affinity uptake of L-glutamate and L-aspartate was demonstrated not only in synaptosomes but also in gila-enriched fractions of cerebral homogenates and, in rat sensory ganglia, [3H]L-glutamate was found to be accumulated by satellite glial cells (review: 18). These findings raised the possibility of separate neuron-located and gila-located high affinity uptake systems for glu/asp, a hypothesis which could be conveniently tested in cultures of glial and neuronal cells. Early studies used cell lines derived from glioma C6 (15,16) or neuroblastoma (69). Initially, no substantial differences between the characteristics of the "glial" uptake of glu/asp in cultured glioma C6 cells and that studied earlier in synaptosomes and brain slices were reported (16). However, the high affinity uptake of [3H]L-glutamate and [3H]L-aspartate in NN (non-tumoral glial line derived from neonatal syrian hamster; NN cells cocultured with a neuroblastoma line and re-isolated were later referred to as 16 line) showed some peculiarities in the substrate specificity (33). Thus neither DL-threo-3hydroxyaspartate, L-cysteine sulphinate, L-cysteate nor L-aspartate could produce complete inhibition of [3H]L-glutamate uptake, even at 1 mM inhibitor concentration. DAspartate was an even weaker inhibitor, while lmM D-glutamate which was only a weak inhibitor of the "high affinity" component of glu/asp uptake in brain slices and synaptosomes (8,9,17), produced 70% inhibition. Significantly, DL-threo-3hydroxyaspartate inhibited almost completely the uptake of [3H]L-aspartate (33). [3H]LGlutamate uptake was apparently Na+-dependent, but the lowest Na+ concentration

1472

Uptake of L-glutamate and L-aspartate in CNS

Vol. 51, No. 19, 1992

examined was 27 mM and the substrate specificity was not studied under those conditions. It was concluded that the high affinity uptake of glu/asp in NN cells was similar to that in brain slices and synaptosomes, the differences were considered minor and were attributed to polymorphism or to the effects of co-culturing with neuroblastoma cells (33). However, the possibility that another uptake system whether Na+-dependent or Na+-independent - which would have been insensitive or less sensitive to DL-threo-3-hydroxyaspartate, L-cysteate, L-cysteinesulphinate and to both L- and D-aspartate, operated in NN cells, could not be discounted (70). Major differences between the high affinity uptake of L-glutamate by cultured glial (line NN) and neuronal (neuroblastoma line M1) cells, respectively, were reported when the effect of the presence of cyclic nucleotides in culture medium on the high affinity uptake of L-glutamate was studied (69). Dibutyryl-cAMP and cGMP increased the Vmax of L-glutamate uptake in NN, but not in M1, cells by a factor of 3 following a 48-hour treatment. This appeared to be a specific effect, since no such increase was observed for the high affinity uptake of taurine in either of the two cell lines and the change was not accompanied by a morphological transformation thus eliminating the possibility that it was caused by a simple increase in the surface area of the cells as sometimes happens in similar experimental circumstances in primary cultures (71). Neither the Na+-dependence nor substrate specificity of the high affinity uptake of [3H]L-glutamate in NN cells after treatment with cyclic nucleotides has, however, been investigated. In a subsequent study it was found that the uptake of [3H]L-glutamate by the M1 line was not inhibited by L-aspartate and, in accordance with this finding, [3H]L-aspartate was not taken up by those cells (72). Differences between L-glutamate uptake in synaptosomes and cultured glioma cells, respectively, were also reported. For example 4-acetamino-4-isothiocyano-2,2-disulphonic acid stilbene (SITS) inhibited glial (cultured glioma C6) but not synaptosomal uptake and D-2-aminoadipic acid was a strong inhibitor of neuronal (neuroblastoma) but not glial uptake of L-glutamate (73,74). However, the differences between glial and neuronal uptake of L-glutamate observed in glioma and neuroblastoma cells or synaptosomes were not found when primary cultures of neuronal and glial cells were used as experimental models (74,75). It is also remarkable that the [3H]L-glutamate uptake in cultured neuroblastoma was not - or very poorly - inhibited by L-aspartate-4hydroxamate as well as by D- and L-aspartate (74), compounds which are substrate/inhibitors of Na+-dependent uptake of [3H]L-glutamate in cultured glioma and in primary cultures of both astrocytes and granule (neuronal) cells (74,75) but not of the Na+-independent [3H]L-glutamate uptake in synaptosomes (55) and in the primary cultures of astrocytes (57). In this context it must be considered rather intriguing that cultured neuroblastoma cells have never been used to study glu/asp uptake in the absence of Na+ (though see 47). It is evident from the present discussion that glu/asp uptake in cultured cell lines - whether of tumoral or non-tumoral origin - differs substantially from that in primary cultures (74). While the cell lines may provide useful experimental models when a specific aspect of glu/asp uptake, e.g. kinetics of Na+-dependence, is studied in isolation (glioma C6: 41; cerebellar neuronal line ~t: 37) the results should not be automatically extended to the possible differences between glu/asp uptake by neurons and glia, respectively, in the intact CNS. However, even data obtained in primary cultures may have to be treated with caution. Thus glu/asp uptake in primary cultures of granule cells (cerebellar neurons) has about equal affinities for Lglutamate, L- and D-aspartate and is similar, in terms of the stoichiometry of Na +dependence, to glu/asp uptake in ~t, a granule cell-derived line (37,75). However, data obtained in experiments with cell-enriched fractions prepared from cerebellum suggest that granule cells do not take up L-glutamate, except, perhaps, during early

Vol. 51, NO. 19, 1992

Uptake of L-glutamate and L-aspartate in CNS

1473

postnatal development (76). According to these studies, glu/asp uptake is concentrated in astrocyte-enriched fractions, it takes up acidic amino acids with about 18 - 90 x greater affinity than the glu/asp uptake in primary cultures of astrocytes and it prefers L-aspartate over D-aspartate, a characteristic not detected in primary cultures of glial cells (74). To summarize, on the basis of substrate/inhibitor selectivity, at least two different high affinity uptake systems specific for L-glutamate can be distinguished in cells cultured from the central nervous system. The first system handles both L-glutamate and L-aspartate, it is sensitive to DL-threo-3-hydroxyaspartate and D-aspartate but much less so to D-glutamate. It is similar to that described earlier in brain slices and synaptosomes (7-9) and it can be studied in primary cultures of glial and neuronal cells. It is strongly expressed and clearly predominant in cultured glioma cells (15,16,41,74). The second system seems to predominate in cultured neuroblastoma cells (72,74). It interacts neither with L- and D-aspartate nor with L-aspartate-4hydroxamate (analogues with a four carbon atom chain, 4C) but it is sensitive to D-2aminoadipate (6C). HOw MANY UPTAKE SYSTEMS HANDLE L-GLUTAMATE AND L-ASPARTATI~? It could appear from the preceding discussion that there are several distinct uptake systems in the CNS which can handle L-glutamate and, sometimes, Laspartate. They would seem to differ from each other by ionic requirements (e.g. Na+dependent and Na+-independent uptake systems) or by substrate specificity (e.g. those which do and those which do not interact with DL-threo-3-hydroxyaspartate or those which take up [3H]L-aspartate as opposed to those which do not, etc.). However, let us again consider the following facts: (i) Glu/asp uptake in the presence of Na + interacts with L-glutamate, L- and D-aspartates, DL-threo-3-hydroxyaspartate and Laspartate-4-hydroxamate. (ii) Those studies which focused on Na+-independent uptake of [3H]L-glutamate reported, in most cases, insensitivity to both D- and Laspartate and, sometimes, to DL-threo-3-hydroxyaspartate and to L-aspartate-4hydroxamate. (iii) None of the studies of [3H]L-glutamate uptake in neuroblastoma (which interacts poorly, if at all, with either L- or D-aspartate) established the Na +dependence of the uptake or tested DL-threo-3-hydroxyaspartate as a potential substrate/inhibitor. (iv) The link between Ca2+/CI - and high affinity uptake of Lglutamate can be indirect, via the potential across the plasma membrane (77). (v) While Ca 2+ and/or CI- may be important for the Na+-independent glu uptake (58,78), the high affinity uptake of glu/asp in the presence of Na + is also affected by the transmembrane potential (39,40,41) and by the changes in Ca 2+ and Cl-concentrations (16,31,50). Even though - or, perhaps, because - the available data are incomplete, not always mutually comparable and in some cases may not be fully convincing (e.g. Ca2+-, CI-- or transmembrane potential-driven uptake?) or even particularly relevant (e.g. some of the cells lines may have become adapted to the culture conditions or represent transformed or undifferentiated cells), it seems expedient to consider, on the basis of (i) - (v), no more than two clearly distinct uptake systems which take up L-glutamate with high affinity. The first one handles L-aspartate and L-glutamate about equally well and it strongly interacts with DL-threo-3hydroxyaspartate. The second does not take up L-aspartate (except, perhaps, for the Ca2+-dependent component which may be inhibited by L- but not D-aspartate: 58, though see also 57) and it does not interact with DL-threo-3-hydroxyaspartate but may be sensitive to quisqualate and ibotenate (55,57). The first system is readily detectable in brain slices and synaptosomes, in primary cultures of both glial and neuronal (granule) cells and in some glial cell lines. The second system has been observed in synaptosomes and in primary cultures of glial cells but it comprises only a relatively

1474

Uptake of L-glutamate

and L-aspartate

in CNS

Vol.

51, No.

19,

1992

small fraction of the total uptake of L-glutamate. It follows from the available data that the uptake system which has high affinity for both glutamate and aspartate is in most cases strongly Na+-dependent whereas the other, which takes up L-glutamate but not L-aspartate, does not depend on Na+-gradient across the membrane but is influenced - more readily than the former system - by changes in Ca 2+ and CI- concentrations. LGlutamate uptake in cultured neuroblastoma cells does not interact very strongly with either L- or D-aspartate even in the presence of Na + which suggest that neuroblastoma cell lines may provide an experimental model in which the Na +independent, aspartate-insensitive component of glutamate uptake would be predominant also in "normal" Na+-containing media. The classification of the high affinity uptake of glutamate and aspartate into Na +dependent (accepting L-glutamate and both enantiomers of aspartate) and Na +independent (in most cases strongly preferring L-glutamate over L- and D-aspartates) systems does not preclude the possibility of further heterogeneity within each type of uptake and, in fact, several studies suggest that this is the case. Consequently, it may be more appropriate to speak of two classes, rather than of two distinct types, of glutamate uptake systems. Evidence for the heterogeneity within the Na+-dependent class of high affinity uptake of glutamate comes from studies of regional distribution of glu/asp uptake in the rat brain. Thus it has been reported that the uptake of [3H]Lglutamate in the rat cerebellum, unlike that in the forebrain, was sensitive to L-2aminoadipate (79). Moreover, when the effect of cortical lesions on the uptake of [3H]L-glutamate by homogenates prepared from striatum (which receives excitatory cortical input) was investigated, it was found that, although in agreement with previous studies, [3H]L-glutamate uptake decreased following the lesion, this decrease was much smaller when [3H]D-aspartate was used as a substrate for the uptake. By contrast, the same experiments with [3H]L-aspartate detected a decrease significantly greater than that observed with [3H]L-glutamate (79). In fact, the failure of Na +dependent binding of [3H]D-aspartate in striatum (observed in sections of frozen rat brain) to respond to cortical lesions prompted a conclusion that D-aspartate is not a suitable marker for the glutamate uptake sites on the presynaptic neuronal structures (80). It is improbable that this finding resulted from a non-specific damage to the uptake site in the frozen tissue. [3H]L-Glutamate uptake is reduced in frozen preparations, probably because the osmotically sensitive compartments are disrupted, but the binding to membranes seems to remain intact (81). In fact, despite the negative result (80), more attempts to use [3H]D-aspartate to visualise high affinity glu/asp uptake sites in frozen sections have been made (82-84). However, there seems to be no compelling need to use the non-metabolizable D-enantiomer at temperatures 2 4oc, especially when the lesion studies suggest that L-aspartate might be the preferred substrate for the glu/asp uptake system related to glutamatergic/aspartergic nerve terminals and, therefore, a more suitable marker anyway (79). The suggestion that the glu/asp uptake associated with glutamatergic and/or aspartergic synapses could prefer L-aspartate over D-aspartate may appear to contradict conclusions of earlier studies which indicated that D-aspartate had a higher affinity than L-aspartate for the Na+-dependent glu/asp uptake (9,22). An example of a possible explanation of this apparent paradox may be provided by the data of Skerrit and Johnston (85) who studied the uptake of [3H]N-methyI-D-aspartate (NMDA) by rat brain slices. Although the Km with respect to NMDA was much higher (3 mM) than that of glu/asp uptake observed under similar conditions (20 I~M, 8), there was relatively little difference in terms of Na+-dependence (apparent Krn'S with respect to Na + concentrations were 13, 14 and 25 mM for L-glutamate, D-aspartate and NMDA, respectively, 9,22,85). Furthermore, inhibition studies suggested that this "low affinity"

Vol.

51, No. 19, 1992

Uptake of L-glutamate

and L-aspartate

in CNS

1475

NMDA uptake system could, in fact, have a high affinity for L-aspartate as indicated by the value of IC5o (inhibitor concentration giving 50% inhibition) which was 29 p-M for Laspartate. Significantly, both D-aspartate and L-glutamate were weaker inhibitors (IC5o = 58 #M and 108 p.M, respectively, 85). The actual affinities of L-aspartate, Daspartate and L-glutamate as well as the differences among them may have been even higher directly at the transporter binding site since diffusional barriers, the presence of glial "glutamate reservoirs" (86) and the liberation of endogenous Lglutamate and L-aspartate into the incubation medium (87,88) tend to lead to the underestimation of such parameters as IC50 and the differences among them (89). Taking into account all those factors would also explain why such a relatively small "low affinity NMDA/high affinity L-aspartate" component of glu/asp uptake was not observed in earlier experiments using brain slices (8,9). In fact, even though an uptake system which prefers L-aspartate over D-aspartate has not been detected in brain slices, it has been demonstrated in astrocyte enriched fractions from rat cerebellum (76). Incidently, this finding would seem to imply that the glu/asp uptake associated with glutamatergic/aspartergic synaptic transmission is actually located in glial cells - a conclusion not inconsistent with the results of studies on the possible regulation of glu/asp uptake in glial cells by adjacent neurons (38,90-92). A glycoprotein thought to mediate Na+-dependent glu/asp uptake in the rat brain has recently been purified (93) and a corresponding antibody prepared (94). Initial immunocytochemical studies pointed to astroglial processes as the predominant site of the immunoreactivity (94). Also, Xenopus laevis oocytes have been used to express mRNA associated with Na+-dependent glu/asp transporter (95) and the results suggested a pattern of heterogeneity of Na+-dependent transport in the rat brain similar to that observed in studies using synaptosomal fractions prepared from various brain regions (79,96). However, it is not clear to what extent these studies may have been influenced by the recently demonstrated high affinity uptake of glutamate endogenous to Xenopus laevis oocytes (97). In conclusion, the available experimental data are compatible with the existence of two separate classes of high affinity uptake systems specific for L-glutamate. One system is Na+-dependent, interacts with both "natural substrates" L-glutamate and Laspartate as well as with D-aspartate, D- and L-threo-3-hydroxyaspartate and several other analogues. The other uptake system interacts with L-glutamate but not with Dand L-aspartates, DL-threo-3-hydroxyaspartate and L-aspartate hydroxamate. It is Na+-independent. There is a strong possibility of further heterogeneity within each class of uptake systems. In particular, the Na+-dependent uptake system displays distinct regional variations in the sensitivity to D- and L-threo-3-hydroxyaspartate (89), to L-2-aminoadipate, dihydrokainate and probably to L- and D-aspartate (79,96). At least some of these variations may reflect the regional heterogeneity of the corresponding mRNA (95). Evidence to date favours [3H]L-aspartate as the best available marker for the Na+-dependent glu/asp uptake associated with glutamatergic/aspartergic nerve endings. However, it is evident that synthesis and testing of new glutamate analogues with higher selectivity for various putative subsystems of the glu/asp uptake will result in significant advances in the understanding of the anatomical distribution and neurochemical diversity of the high affinity uptake of L-glutamate in, or in the vicinity of, glutamatergic and aspartergic synaptic terminals in the central nervous system. Acknowledaements The authors wish to thank to Professor Liam Burke and to Professor Graham Johnston for helDful comments

1476

Uptake of L-glutamate and L-aspartate in CNS

Vol.

51, No. 19, 1992

References 1. T. HAYASHI, Keio J. Med. 3 183-192 (1954) 2. P.J. ROBERTS, J. STORM-MATHISEN and G.A.R. JOHNSTON (eds.) Glutamate: Transmitter in the Central Nervous System, John Wiley & Sons Ltd., Chichester (1981) 3. F. FONNUM, J. Neurochem. 4_.221-11 (1984) 4. T. P. HICKS, D. LODGE and H. McLENNAN (eds.) Excitatory Amino Acid Transmission, Neurology and Neurobiology, VoI. 24, Alan R. Liss, Inc., New York (1987) 5. D. T. MONAGHAN, R.J. BRIDGES and C. W. COTMAN, Ann. Rev. Pharmacol. Toxicol. 29365-402 (1989) 6. J. J. HANSEN and P. KROGSGAARD-LARSEN, Med. Res. Rev. 1055-94 (1990) 7. W. J. LOGAN and S. H. SNYDER, Brain Res. 4_22413-431 (1972) 8. V. J. BALCAR and G.A.R. JOHNSTON, J. Neurobiol. 3295-301 (1972) 9. V. J. BALCAR and G.A.R. JOHNSTON, J. Neurochem. 192657-2666 (1972) 10. G. E. FAGG and J. D. LANE, Neurosci. 41015-1036 (1979) 11. M. ERECINSKA, Biochem. Pharmacol. 393547-3555 (1987) 12. S. NAITO and T. UEDA, J. Neurochem. 4499-109 (1991) 13. E. Mo FYKSE and F. FONNUM, Biochem. J. 276363-367 (1991) 14. A. R. WOFSEY, M. J. KUHAR and S.H. SNYDER, Proc. Natn. Acad. Sci. U.S.A. 68 1102-1106 (1971) 15. F. A. HENN, M. N. GOLDSTEIN and A. HAMBERGER, Nature (Lend.) 249663-664 (1974) 16. A. FAIVRE-BAUMAN, J. ROSSiER and Po BENDA, Brain Res. 76371-375 (1974) 17. P. J. ROBERTS andJ. C. WATKINS, Brain Res. 85120-125 (1975) 18. D. E. CURRIE and J. S. KELLY, J. exp. Biol. 9.5181-193 (1981) 19. A. SCHOUSBOE, Int. Rev. Neurobiol. 2._221-45(1981) 20. R. A. ALTSCHULER, J. L. MOSSINGER, G. G. HARMISON, M. H. PARAKKAL and R. J. WENTHOLD, Nature 298657-659 (1982) 21. H. F. BRADFORD and H. K. WARD, Brain Res. 110115-125 (1976) 22. L. P. DAVIES and G. A. R. JOHNSTON, J. Neurochem. 261007-1014 (1976) 23. R. LUND-KARLSEN and F. FONNUM, Brain Res. 151457-467 (1978) 24. J. STORM-MATHISEN and H. WOXEN-OPSAHL, Neurosci. Lett. ~ 65-70 (1978) 25. P. STREIT, Cerebral Cortex, Vol. 2, E. G. Jones and A. Peters (eds.) 119-143 (1985) 26. J. STORM-MATHISEN, A. K. LEKNES, A. T. BORE, J. L. VAALAND, P. EDMINSON, F. M.-S. HAUG and O. P. OTTERSEN, Nature (London) :~01517-520 (1983) 27. G. J. BALCOM, R. H. LENNOX and J. L. MEYERHOFF, J. Neurochem. 2_6_6423-425 (1976) 28. C. W. COTMAN, A. FOSTER and T. LANTHORN, Glutamate as a Neurotransmitter, G. DiChiara and G. L. Gessa (eds.), 1-27, Raven Press, New York (1981) 29. F. FONNUM, A. SOREIDE, I. KVALE, J. WALKER and I. WALAAS, Glutamate as a Neurotransmitter, G. DiChiara and G. L. Gessa (eds.), 29-41, Raven Press, New York (1981 ) 30. J. C. WATKINS and R. H. EVANS, Ann. Rev. Pharmacol. Toxicol. 21165-206 (1981) 31. V. J. BALCAR, Neurochem. Int. 1843-49 (1991 ) 32. J. P. BENNET Jr., W. J. LOGAN and S. H. SNYDER, J. Neurochem. 2"11533-1556 (1973) 33. V. J. BALCAR, J. BORG and P. MANDEL, J. Neurochem. 2._8887-93 (1977) 34. A. SCHOUSBOE, G. SVENNEBY and L. HERTZ, J. Neurochem. 2__£9999-1005 (1977)

Vol. 51, No. 19, 1992

Uptake of L-glutamate and L-aspartate in CNS

1477

35. D. D. WHEELER and R. G. HOLLINGSWORTH, J. Neurochem. 301311-1319 (1978) 36. D. D. WHEELER, J. Neurochem. 33883-894 (1979) 37. W. B. STALLCUP, K. BULLOCH and E. E. BEATGE, J. Neurochem. 3251-65 (1979) 38. J. DREJER, O. M. LARSSON and A. SCHOUSBOE, Exp. Brain Res. 4-7259-269 (1982) 39. D. D. WHEELER and J. S. GRAVES, J. theor. Neurobiol. 21-22 (1983) 40. J. G. MARVlZON, F. MAYOR, Jr., M. C. ARAGON, C. GIMENEZ AND F. VALDIVIESO, J. Neurochem. 371401-1406 (1981) 41. M. ERECINSKA, M.B. TROEGER, D.F. WILSON and I.A. SILBER, Brain Res. $69 203-214 (1986) 42. B. I. KANNER, Biochem. Biophys. Acta 726293-316 (1983) 43. B. I. KANNER and S. SCHULDINER, CRC Crit. Rev. Biochem. 2;21-38 (1987) 44. H.K. KIMELBERG, S. PANG and D.H. TREBLE, J. Neurosci. 91141-1149 (1989) 45. M P. BLAUSTEIN, J. Physiol. (London) 247617-653 (1975) 46. M. C. FRAME, R. I. FRESHNEY, P. F. T. VAUGHAN, D. R. GRAHAM and R. SHAW, Brit. J. Cancer 4._99269-280 (1984) 47. E.E. BEATGE, K. BULLOCH and W.B. STALLCUP, Brain Res. 167210-214 (1979) 48. V. J. BALCAR, FEBS Letters 300203-207 (1992) 49. S. R. VINCENT and E. G. McGEER, Brain Res. 18499-105 (1980) 50. M. J. KUHAR and M. A. ZARBIN, J. Neurochem. 31251-256 (1978) 51. D. F. SHRIVER, P. W. ATKINS and C. H. LANGFORD, Inorganic Chemistry, pp. 378-379, Oxford University Press (1990) 52. R. T. MORRISON and R. N. BOYD, Organic Chemistry, 3rd Edition, pp. 598-599, AIlyn and Bacon Inc., Boston (1975) 53. A. NEUBERGER, Proc. R. Soc. A15868-96 (1937) 54. R. A. WANIEWSKI and D. L MARTIN, J. Neurosci. 42237-2246 (1984) 55. R. ZACZEK, M. BALM, S. ARLIS, H. DRUCKER and J. T. COYLE, J. Neurosci. Res. 18425-431 (1987) 56. J. P. PIN, J. BOCKAERT and M. RECASENS, FEBS Lett. 17531-36 (1984) 57. B. FLOI-I- and W. SEIFERT, Glia 4293-304 (1991) 58. M. HOLLMANN, J. HARNECKER and W. SEIFERT, FEBS Lett. 22874-78 (1988) 59. R. HAMMERSCHLAG and D. WEINRICH, Studies of Neurotransmitters at the Synaptic Level, E. Costa, L.L Iversen and R. Paoletti (eds.), Adv. Biochem Pharmacol. 6165-180 (1972) 60. V. J. BALCAR and G. A. R. JOHNSTON, J. Neurochem. 2._881145-1146(1977) 61. G. A. R. JOHNSTON, V. J. BALCAR, D. R. CURTIS, J. DAVIES, R. M. McCULLOCH and A. L. STEPHENSON, Proc. Aust. Physiol. Pharmacol. Soc. 4150-151 (1973) 62. J. LAKSHMANAN and G. PADMANABAN, Biochem. Biophys. Res. Comm. 58690698 (1974) 63. E. G. McGEER, P. L. McGEER and K. SINGH, Brain Res. 139381-383 (1978) 64. G. A. R. JOHNSTON, S. M. E. KENNEDY and B. TWITCHIN, J. Neurochem. 3._.22 121-127 (1979) 65. R. J. BRIDGES, M. S. STANLEY, M. W. ANDERSON, C. W. COTMAN and A. R. CHAMBERLIN, J. Med. Chem. 34717-725 (1991) 66. E. J. FLETCHER, K. N. MEWETT, C. A. DREW, R. D. ALLAN and G. A. R. JOHNSTON, Neurosci. Lett. 121133-135 (1991 ) 67. H. SHINOZAKI, M. ISHIDA, K. SHIMAMOTO and Y. OHFUNE, Brain Res. 486355359 (1989) 68. T. H. LANTHORN, W. F. HOOD. G. B. WATSON, R. P. COMPTON, R. K. RADER, Y. GAONI and J. B. MONAHAN, Eur. J. Pharmacol. 182397-404 (1990) 69. J. BORG, V. J. BALCAR and P. MANDEL, Brain Res. 166113-120 (1979) 70. T. N. THOMAS and D. A. REDBURN, J. Neurochem. 3163-68 (1978) 71. L. HERTZ, E. BOCK and A. SCHOUSBOE, Dev. Neurosci. & 226-238 (1978)

1478

Uptake of L-glutamate and L-aspartate in CNS

Vol. 51, No. 19, 1992

72. V. J. BALCAR, J. BORG, J. ROBERT and P. MANDEL, J. Neurochem. 3._~416781681 (1980) 73. R. A. WANIEWSKI and D. L. MARTIN, Brain Res. 268390-394 (1983) 74. V. J. BALCAR, A. SCHOUSBOE, P. E. SPOERRI and J. R. WOLFF, Neurochem. Int. 1_0213-217 (1987) 75. J. DREJER, O. M. LARSSON and A. SCHOUSBOE, Neurochem. Res. 8231-243 (1983) 76. R. D. GORDON and R. BALAZS, J. Neurochem. 401090-1099 (1983) 77. B. I. KANNER, Biochemistry, 1_/71207-1211 (1978) 78. M. HOLLMANN and W. SEIFERT, J. Neurochem. 53716-723 (1989) 79. J. FERKANY and J. T. COYLE, J. Neurosci. Res. 1__66491-503 (1986) 80. J. T. GREENAMYRE, D. S. HIGGINS and A. B. YOUNG, Brain Res. 511310-318 (1990) 81. R. SCHWARCZ, Life Sci. 281147-1154 (1980) 82. K. J. ANDERSON andT. W. VlCKROY, Eur. J. Pharmacol. 18`5119-121 (1990) 83. K. J. ANDERSON, D. T. MONAGHAN, R. J. BRIDGES, A. L. TAVOULARIS and C. W. COTMAN, Neurosci. 3__88311-322 (1990) 84. K. J. ANDERSON, R. J. BRIDGES and C. W. COTMAN, Brain Res. 562285-290 (1991) 85. J. H. SKERRITT and G. A. R. JOHNSTON, J. Neurochem. 36881-885 (1981) 86. J. GARTHWAITE, Brit. J. Pharmacol. 85297-307 (1985) 87. V. J. BALCAR and G. A. R. JOHNSTON, Brain Res. 83173-175 (1975) 88. R. W. P. CUTLER and J. YOUNG, Brain Res. 165261-270 (1979) 89. E. J. FLETCHER and G. A. R. JOHNSTON, J. Neurochem. 3163-68 (1991) 90. J. DREJER, E. MEIER and A. SCHOUSBOE, Neurosci. Lett. 3-7301-306 (1983) 91. E. HANSSON, P. ERIKSSON and M. NILSSON, Neurochem. Res. 101335-1341 (1985) 92. E. HANSSON, Dev. Brain Res. 2._44203-209 (1986) 93. N. C. DANBOLT, G. PINES and B. I. KANNER, Biochemistry 3__226734-6740 (1991) 94. N. C. DANBOLT, G. PINES and B. I. KANNER, J. Neurochem. 5-7 (Suppl.) $15 D (1991) 95. R. D. BLAKELY, J. A. CLARK, T. PACHOLCZYK and S. G. AMARA, J. Neurochem. 56860-871 (1991) 96. M. B. ROBINSON, M. HUNTER-ENSOR and J. SINOR, Brain Res. ,544196-202 (1991) 97. J. STEFFGEN, H. KOEPSELL and W. SCHWARZ, Biochim. Biophys. Acta 10661420 (1991)