Kinetic characterization of sulphur-containing excitatory amino acid uptake in primary cultures of neurons and astrocytes

Neurochem. Int. Vol. 19, No. 4, pp. 467~174, 1991 Printed in Great Britain. All rights reserved

0197-0186/91 $3.00+0.00 Copyright © 1991 Pergamon Press plc

KINETIC CHARACTERIZATION OF SULPHURC O N T A I N I N G EXCITATORY AMINO ACID UPTAKE IN P R I M A R Y CULTURES OF N E U R O N S A N D ASTROCYTES ANGUS GRIEVE, ~ JOHN DUNLOP, I ARNE SCHOUSBOE 2 and ROGER GRIFFITHS ~* Department of Biochemistry and Microbiology, University of St Andrews, Fife KY 16 9AL, Scotland, U.K. 2PharmaBiotec Research Center, Department of Biological Sciences, Royal Danish School of Pharmacy, 2100 Copenhagen 0, Denmark (Received 13 March 1991 ; accepted 30 May 1991)

Abstract--The uptake of the neuroactive sulphur amino acids L-cysteine sulphinate, L-cysteate, L-homocysteine sulphinate and L-homocysteate was investigated in astrocytes cultured from the prefrontal cortex ; in neurons, cultured from cerebral cortex ; and, in granule cells, cultured from cerebellum. It was shown that each amino acid acted as a substrate for a plasma membrane transporter in both neurons and astrocytes. Astrocytes and neurons exhibited a high-affinity uptake for L-cysteine sulphinate and L-cysteate with Km values ranging from 14~100/~M, and a low-affinity uptake for L-homocysteine sulphinate and L-homocysteate, with Km values ranging from 225-1210 #M. The uptake of all transmitter candidates studied was partially sodium-dependent. This sodium-dependency was most evident at low (< 100 #M) concentrations of each substrate. The apparent uptake measured in the absence of sodium was included as a component in corrections made for non-saturable influx. With the exception of L-cysteine sulphinate, uptake of each sulphur amino acid was greatest in astrocytes, with Vmax values ranging between 15-32 nmol min- ~mg- ~cell protein. Moreover, the uptake of each sulphur amino acid in cerebellar granule cells (Vm,x values ranging between 1~25 nmol min- ~mg t cell protein) was consistently greater than that in cerebral cortex neurons (Vmaxvalues ranging between 1.5 6 nmol min- ~mg ~cell protein).

A number of criteria need to be satisfied before a c o m p o u n d can be accepted for transmitter status (McGeer et al., 1978). One such criterion is the presence of an efficient mechanism for the rapid inactivation of transmitter following activation of appropriate receptors. F o r amino acids, it is generally accepted that inactivation occurs by removal from the synaptic space via high-affinity carrier systems (Curtis and Johnston, 1974). F r o m the results of a number of inhibition kinetic studies, it has been suggested that a c o m m o n carrier system exists for the high-affinity transport of glu, aspartate, CSA and C A (Wilson and Pastuszko, 1986; Griffiths et al., 1989; Davies and Johnston, 1976; Erecinska and Troeger, 1986). However, only relatively few studies have been undertaken in which S A A transport per se has been measured and kinetically characterized, e.g. C S A (Iwata et al., 1982 ; Recasens et al., 1982 ; Abele et al., 1983) ; C A (Wilson and Pastuszko, 1986) and H C A (Cox et al., 1977). In these latter studies, chemically-synthesized radiolabelled forms of the respective SAAs were used as substrates to investigate transport by rather hetero-

Naturally-occurring amino acids other than L-glutamate (glu) and L-aspartate may play a role in excitatory transmission in the CNS. In a number of biochemical and electrophysiological studies, it has been shown that the sulphur-containing amino acids (SAAs), viz. L-cysteine sulphinate (CSA), L-cysteate (CA), L-homocysteine sulphinate (HCSA) and L-homocysteate (HCA), closely mimic the neurochemical actions of L-glutamate (glu), the major fastacting neurotransmitter at excitatory synapses in the mammalian C N S [e.g. Curtis and Watkins (1960, 1963); Mewett et al. (1983); Dunlop et al. (1989a,b, 1990)]. The demonstration of an endogenous localization and depolarization-induced, Ca2+-dependent release of the excitatory SAAs (Do et al., 1986a), is of added significance to the study of these compounds, some of which have been proposed as transmitter candidates (Cuenod et al., 1986; Patneau and Mayer, 1990; Griffiths, 1990).

* Author to whom all correspondence should be addressed. 467

468

ANGUS GRIEVE e t a / .

geneous brain preparations. It is now well-established that physiological inactivation o f amino acid transmitter action is mediated by carrier systems located in the plasma m e m b r a n e o f both neurons a n d astrocytes (Schousboe, 1981; Drejer et al., 1982). Since the presence o f separate carriers for the excitatory SAAs in primary cultures o f neurons and astrocytes has not been established, it would be o f obvious advantage to d e m o n s t r a t e this, and further, to characterize the transport kinetics in these more homogeneous and distinct brain cell preparations. To this end, we have undertaken a kinetic characterization o f the uptake o f various SAAs by primary cultures o f cortical neurons and cerebellar granule cells, and by cortical astrocytes, following development o f a sensitive H P L C - b a s e d assay, which represents a further refinement over previous SAA transport studies.

EXPERIMENTAL PROCEDURES

Materials

Mice were obtained from the animal quarters at the Panum Institute. Plastic multitest dishes were purchased from NUNC A/S Denmark, and culture media, foetal calf serum, and Dulbecco's minimum essential medium (DMEM) from Gibco-Biocult Lab (Scotland), Dibutyryl-cyclic AMP (dBcAMP), cytosine arabinoside, trypsin, soybean trypsin inhibitor and poly-L-lysinewere obtained from Sigma Chemical Company, St Louis, MO; penicillin from Leo, Denmark and insulin from NOVO-Nordisk, Denmark. L-Cysteine sutphinic acid, L-cysteic acid, L-homocysteine sulphinic acid and L-homocysteic acid were purchased from Tocris Neuramin (Essex, England). ('ell culture

Cerebellar granule cells were cultured essentially as described by Drejer and Schousboe (1988), and Schousboe et al. (1989). Briefly, cerebella from %day-old mice were dissociated by mild trypsinization [0.1% (w/v) trypsin at 37C for 15 rain] and subsequently inoculated into poly-Llysine coated 24-well multi-test dishes containing slightly modified DMEM (24.5 mM KCI, 30 mM glucose) supplemented with p-aminobenzoate (7 #M), kainic acid (50 #M), insulin (t00 mU/l) and 10% (v/v) foetal calf serum. Cells were maintained in culture for 7 10 days, with the addition of the antimitotic agent cytosine arabinoside (40 elM) from day 2 in vitro to prevent glial proliferation. Cerebral cortex neurons were cultured from 16-day-old mouse foetuses as described by Dichter (1978) and Hertz et al. (1989a). Briefly, the tissue was dissected and exposed to 0.02% (w/v) trypsin in Pucks solution (137 mM NaC1, 4.2 mM NaHCO~, 5 mM glucose, pH 7.4) at 3TC for 10 rain and subsequently centrifuged for 10 min at 900 g. The pelleted cells were resuspended (I 2 x 106 cells/ml) in a moditied DMEM containing 10% (v/v) inactivated foetal calf serum supplemented with 24.5 mM KCI, 30 mM glucose, 7 ~M p-aminobenzoic acid and 100 mU/l insulin, and seeded in NUNCLON 24-well multi-test dishes coated with poly-Llysine as described by Sensenbrenner et al. (1978). After 2

days in culture, the cells were exposed l\~r the remainder of the culture period (8 l0 days) to cytosine arabinoside (40 itM) as described for granule cells. Astrocytes were cultured essentially as described previously by Hertz et al. (1982, 1989b). Briefly, prefrontal cortex was taken from newborn mice and passed through sterile nylon sieves (80 #m pore size) into a modified Eagle's minimum essential medium containing 20% (v/v) inactivated foetal calf serum and inoculated into the individual wells of NUNCLON 24-well multi-test dishes. The cultures were grown for a total of three weeks changing the culture medium two days after inoculation and subsequently 3 times a week. During the last week of cultivation, the serum concentration was reduced to 10% (v/v) and 0.25 mM dBcAMP was added to the culture medium. Such cultures consist of well-differentiated astrocytes and have been shown to be devoid of neurons ((/Hertz el al.. 1989b). Uptake assays

The cell culture medium was exchanged with 500 i~1 HEPES-buffered saline [10 mM HEPES, 135 mM NaCI, 5.(1 mM KC1, 1.0 mM CaCI> 0.6 mM MgSO4 and 6 mM glucose, pH 7.4 (HBS)] and the cells preincubated for 3 rain at 37'C. Subsequently, this medium was exchanged with HBS containing different concentrations (0 2500 #M) of CSA, CA, HCSA or HCA and incubated for 2 rain at 37°C. Thereafter, the medium was removed by rapid suction and the cell monolayer carefully washed ( x 2) with 500 #1 aliquots of HBS. Following aspiration of the final wash, 500 #1 aliquots of icecold 70% (v/v) ethanol were added to each well. The cells were removed by scraping and the suspension transferred to Eppendorf tubes prior to centrifuging at 15,000 rpm for 3 min. Aliquots (300 #I) of the supernatants were lyophilized and retained for HPLC analysis. The remaining supernatanl was removed and the dried cell pellets incubated at 37'C ti)r 1 h with 200 #1 I N KOH. After this time, the volume was adjusted to 500 /d by addition of distilled H20 and the samples taken for measurement of protein content using the method of Lowry et al. (1951) with bovine serum albumin as the standard. H P L C ana&sis q/'amino acids Preparation o/samph, s. The lyophilized supernatant sam-

ples were each resuspended in 100 #1 of 5 #M homoserine solution which acted as an internal standard for amino acid analysis. Aliquots of amino acid standard solution and supernatant samples were individually filtered into clean microfuge tubes using disposable 0.45 #m Acro LC3A filter discs (Gelman Sciences). A 20 #1 aliquot of standard amino acid mixture (5 #M) or sample was derivatized automatically via a Model 2157 Autosampler (LKB) by mixing with 40 #1 o-phthaldialdehyde (OPA) solution (Roth, 1971). Chromatography. A high performance liquid chromatograph (LKB) comprising two Model 2150 pumps with a Model 2151 LC Controller was used throughout the study. Derivatized samples (40 #l) were injected via the autosampler and amino acids detected using a Merck-Hitachi FI000 variable wavelength Fluorescence Spectrophotometer (excitation wavelength 340 nm" emission wavelength 450 nm) fitted with a 12 #1 volume flow cell. Peak areas were quantified using a TriVector Trio Chromatography Computing Integrator. A Chromspher C-18 (Chrompack U.K.) (5 #m particle diameter) reversed-phase cartridge column, 100 × 4 mm (i.d.), was used, protected by using in conjunction with a

Excitatory sulphur-containing amino acid transport in brain cells Chrompack (5 #m particle diameter) guard column, 10 × 4 mm (i.d.), and a solvent saturation column (30 × 4 ram, i.d.) packed with Spherisorb l0 #m particle size free silica, which was situated before the injector. Resolution of CSA, CA and HCSA was achieved using a stepped-gradient ranging from 5% (v/v) methanol plus 0.1 M sodium acetate (pH 5.55) to 100% methanol over a 30 rain period at a flow rate of 1.0 ml min ~. For resolution of HCA, the sodium acetate buffer was used at pH 5.80. Construction of calibration curves for each SAA showed that the peak areas were directly proportional to the concentration of OPA-derivatized SAA at least in the range 0500 pmol loaded onto the column. The minimum peak area for any SAA corresponded to 1-2 pmol on the column which was well within the sensitivity range of the detector. The maximum peak area measured for any SAA was within the linear range and did not exceed an amount equivalent to 450 pmol loaded. Moreover, in calculating net uptake, there was no requirement to subtract peak areas of endogenous SAA as these values were negligible in all cases. The identity of the measured peak of SAA uptake was established by coinjection of sample with authentic SAA. Only the peak area of the amino acid co-eluting with authentic SAA was elevated.

Data analysis The neuronal and astrocytic uptake of the various SAAs was determined to best-fit the equation : v = Vmax[l/(1+ K~/S)] +ks

(1)

where v and Vmaxindicate the initial and maximum velocities, respectively ; Kin, the Michaelis-Menten constant ; k, the rate constant of the non-saturable influx component, and S, the substrate concentration (Neame and Richards, 1972). All data were analyzed by computer-assisted non-linear regression analysis using the ENZFITTER software program (Leatherbarrow, 1987). In addition to equation (1) the data was also fitted to the Michaelis-Menten equation assuming either a single uptake site or two non-interacting sites of differing affinities.

RESULTS

The time course of neuronal and astrocytic uptake of each SAA, determined at 50 and 2500 #M, was found to deviate from linearity at 2.5-3.0 min (data not shown). F o r all subsequent kinetic studies, incubation of SAAs was undertaken for a 2 min period for both neurons and astrocytes. The uptake of all SAAs was partially sodium-dependent in both neurons and astrocytes. Using fixed concentrations (100, 1000, 2500 and 5000/~M) of each SAA, uptake was performed in the presence and absence of extracellular sodium. At 100 # M SAA, in the absence of extracellular sodium, uptake was typically < 2 0 % of that in the presence of 135 m M extracellular sodium. The sodium-dependence decreased at increasing SAA concentration such that at 1000 #M, 2500 # M and 5000/~M SAA, uptake in the absence of extracellular

469

sodium was, respectively, in order of 60, 75 and 75% that in the presence of sodium. Similar results, although of different absolute values were obtained for each SAA. These observations illustrate the greater influence of a sodium-independent factor at higher substrate concentrations, consistent with the presence (as evaluated by computer-assisted data analysis) of a non-saturable (influx) component contributing to the overall uptake. In order to determine the kinetic parameters, Km and Vmax, the uptake of CSA, CA, H C S A and H C A by primary cultures of cerebellar granule cells, and by cortical neurons and astrocytes, was measured over a substrate concentration range of 0-2500 #M. The uptake of each SAA, quantified by HPLC, exhibited concentration-dependency over this range. This is illustrated in Fig. 1, which shows a representative series of chromatograms for C A uptake in granule cells at concentrations between 0-100 #M. The rate of uptake of all SAAs, in each cell type, showed a biphasic dependence on the substrate concentration corresponding to a non-saturable influx component and a single saturable component. When corrected for diffusion, the uptake of each SAA exhibited a monophasic dependence on the substrate concentration. Representative velocity-substrate curves for CSA in astrocytes, cortical neurons and cerebellar granule cells are presented in Fig. 2. The kinetic constants, Km and Vm,x for all SAAs are presented in Table 1. In each cell type, the sulphur-containing aspartate analogues, namely, C S A and CA, exhibited a higher affinity for uptake than their respective higher homologues, H C S A and H C A , both of which are close-structural analogues of glu. Neurons exhibit a significantly higher affinity for CSA and C A uptake than do astrocytes, although the Km values indicate the presence of a high-affinity uptake system in both cell types. The kinetic constants for glu uptake in each cell type are closely resembled by those of CSA and C A ; a major anomaly being the low Vmax value recorded for CSA transport by astrocytes (Table 1). The Vmax of C A uptake in neurons and astrocytes was consistently higher than that of CSA uptake. The neuronal uptake of C S A and C A is essentially similar although the Vmax of both SAAs is 5-6 times higher in granule cells than in cortical neurons, an observation in agreement with the repeated observation that cerebellar granule cells are glutamatergic and that cortical neurons in culture are mainly G A B A e r g i c [cf Drejer and Schousboe (1988) ; Schousboe and Pasantes-Morales (1989)]. The uptake of H C S A and H C A is best described by a transport system(s) of low affinity (Kin > 200 /~M). In neurons and astrocytes, the Km

470

ANGUS GREVE et al.

I 23

5

1 E

tO ko

! :>,

c o c

g 09

Ix_

Jl

0

lO

25

50

100

[L Cysteate] NM Fig. 1. HPLC analysis of sulphur amino acid transport in neurons and astrocytes. Primary cultures of cortical astrocytes, cortex neurons and cerebellar granule cells were incubated for 2 min with varying concentrations of SAA. Following incubation, and subsequent extraction, free amino acids in the cells were determined as their fluorescent OPA-derivatives by HPLC analysis as described in Experimental Procedures. A representative set of partial chromatograms is presented which illustrates the accumulation of L-cysteate (elution position indicated by the vertical arrow) in cerebellar granule cells as a function of its extracellular concentration. The concentration-dependent uptake of L-cysteate can be readily observed as an increase in the fluorescence intensity at 450 nm. t.-Cysteate and other excitatory SAAs elute in the early part of the run during an initial isocratic hold of 5% methanol. The profile at 100 ItM L-cysteate shows a more complete chromatogram. Resolved free amino acids of interest include : asp ( 1) ; glu (2) ; ser (3): gln (4): tau (5): GABA (6): the internal standard, homoserine, is also indicated (H). The horizontal bar ( - - ) represents a time scale (2 rain).

values for H C S A a n d H C A are at least an order of magnitude greater t h a n that o f glu, which is of similar c a r b o n chain length, As for CSA a n d CA, the V,..... values of H C S A a n d H C A in granule cells are greater than those in cortical neurons, again presumably reflecting the fact that essentially all n e u r o n s in these cultures are glutamatergic. It can also be seen that the V..... values for H C S A a n d H C A are higher in astrocytes t h a n in neurons.

DISCUSSION

The results of this study d e m o n s t r a t e the uptake, and characterize the kinetics, o f excitatory S A A transport in distinct brain cells, viz. p r i m a r y cultures of glutamatergic a n d G A B A e r g i c neurons ( ~ f a b o v e ) as well as astrocytes. O f the four neuroactive SAAs studied, CSA (Recasens el al., 1982: Griffiths, 1990) a n d

H C A (Cuenod e t a l . , 1986: Do e t a l . , 1986a,b, 1988: L e h m a n n et al., 1988) have received strongest support as transmitter candidates, while less evidence is available for CA a n d H C S A [e.g. Mewett et al. (1983)]. This study shows that CSA and CA are t r a n s p o r t e d by a high affinity system in cultured cerebellar granule cells a n d cortical neurons, and with only slightly less affinity by cultured cortical astrocytes. The kinetic parameters o f C S A and CA uptake in each cell type (Table 1) are similar to those of glu uptake in mouse primary cultures (Drejer et al., 1982). In addition, the K,1 values for CSA and CA uptake are essentially similar in m a g n i t u d e to their K~, values as competitive inhibitors of D-[3H]aspartate (a n o n - m e t a b o l i z a b l e analogue o f glu) uptake by the same cell types (Griffiths et al., 1989). Moreover, a n u m b e r of workers in 'single-point' studies have shown that CSA a n d / o r C A are potent inhibitors o f o-aspartate, L-aspartate

Excitatory sulphur-containing amino acid transport in brain cells

[]

471

who demonstrated, by mutual-inhibition kinetic analysis, a common transporter for CSA, CA, glu and 10 L-aspartate in rat cortex synaptosomes. High affinity uptake of CSA has been demonstrated by a number 8 of workers using different preparations, viz. synap6' tosomes (lwata et al., 1982; Grieve et al., 1990), synaptic membrane vesicles (Recasens et al., 1982) 4 ' tand, primary cultures and clonal cell lines (Abele et Z, 2 al., 1983). The kinetic transport constants in neuronal O £_ Q. primary cultures [Kin=41 #M (granule cells); 0 Km = 58 /~M (cortex)] reported in this study, are [] 1.5 higher than those reported in cortical synaptosomes or membrane vesicles (see above), where the Km varies E between 12-27/~M, although V~,x values are in good 1.0 agreement. Moreover, in primary cultures of rat and E chicken, Abele et al. (1983) reported K~ values for CSA of 5-6/~M in neurons (region unspecified) and 0.5 E 20-40 #M in astrocytes. The slight variations between t-the kinetic parameters determined in the present study (D compared with other studies may be a result of a 0.0 combination of differences which include brain Q. [] preparation, species differences and methodological 0.8 approach. O (3) Less attention has been given to the transmitter 0.6 ~3 status of CA, which paradoxically, was one of the first C~ amino acids tested for excitatory activity (Curtis and 0.4 Watkins, 1960, 1963), and is equipotent with glu (Mewett et al., 1983). Wilson and Pastuszko (1986) 0.2 first showed that rat brain synaptosomes accumulate [358]CA by a high-affinity transport system (Km = 12 0.0 . . . . /~M, Vm,x = 2.5 nmol min ~m g - ~); while Koyama et 0 500 1000 1500 2000 2500 al. (1989) reported that L-[35S]CA selectively detects a chloride-dependent glu transporter (exhibiting a Km [L-Cysteine sulphlnate] lal'l of 5.4/~M) of unknown significance to the process of Fig. 2. Transport of CSA in primary cultures of neurons and neurotransmission. In additional kinetic studies, CA astrocytes. Primary cultures of (a) cerebellar granule cells, (b) cortical neurons and (c) cortical astrocytes were incubated has been shown to competitively inhibit D-[3H]as at 37°C with varying concentrations (~2500 #M) of CSA. partate uptake (1) in rat cortex synaptosomes with g e i The rates of uptake have been corrected for the non-saturable values describing equipotency with CSA, and (2) in influx component. Data points are the mean + SEM of 8 cerebellar granule cells and cortical astrocytes, where experimental values in which HPLC analysis was undertaken CA exhibits a potency twice that of CSA (Griffiths et at least in duplicate. al., 1989). These latter observations correlate well with the measured kinetic parameters of CSA and CA determined in the present study. In vitro autoand glu uptake, or vice versa (Balcar and Johnston, radiography has also been employed to localize and 1972 ; Balcar et al., 1977, 1980 ; Davies and Johnston, quantify sodium-dependent [35S]CA binding sites 1976; Iwata et al., 1982; Mewett et al., 1983). There (Parsons and Rainbow, 1984). These workers showed is general agreement that glu and D-aspartate are the anatomical distribution of [355]CA to be heterotransported by the same high affinity system in geneous and the pharmacological specificity to neurons and astrocytes (Drejer et al., 1983). Taken differ from D-[3H]aspartate, which was used as a together, these observations are consistent with CSA marker for glu/aspartate transport sites. It can be and CA sharing a common transporter with glu, at seen therefore that both CSA and CA display certain least in the cell types studied. This interpretation is in properties which reflect transmitter function. The agreement with that of Wilson and Pastuszko (1986) high-affinity observed for neuronal and astrocytic 12

f

,

.

•

,

.

,

472

ANGUS GRIEVE et al. Table 1. Kinetic constants for excitatory sulphur-containing amino acid uptake by primary cultures of neurons and astrocytes K,,, (/~M) Amino acid CSA CA HCSA HCA GLU*

V,,,,~ (nmol/min/mg protein)

Cx

Gn

Ast

('x

Gn

Ast

58_+8 14_+4 525+159 1210+ 101 4 3 + 11

41+10 41 + 8 468±110 941 + 6 2 42+4

101:~8 88± 12 225+57 901 +334 67+7

1.5+0.2 5.4+0.3 3.0+0.4 3.8+0.3 5.9±0.7

11.1+0.9 25.2!: 1.7 1(I.8+1.0 15.7+0.5 10.2~0.5

0.7+0.1 27.3+2.9 18.352.5 31.8± 10.9 14.9_+0.6

Primary cultures of cortical neurons (Cx), cerebellar granule cells (Gn) and cortical astrocytes (Ast) were incubated for 3 min at 37'C in the presence of increasing concentrations (1~2500 ,uM) of the acidic SAAs. Kinetic constants were obtained by computer-assisted non-linear regression analysis of the data tk)llowing iterative fitting to equation (I) (see Experimental Procedures) and are given ± SEM. * Values are taken from Drejer et al. (1982).

uptake of CSA and CA could describe a mechanism for termination of their activity within the synaptic cleft. Such an inactivation mechanism would appear to be essential bearing in mind the strong depolarizing actions and neuronal cytotoxicity (Olney et al., 1971 ; Pullan et al., 1987; Frandsen et al., 1990) of both SAAs. The two glu analogues, HCSA and HCA are also regarded as transmitter candidates. In particular, the depolarization-induced, calcium-dependent release (Cuenod et al., 1986; Do et al., 1986a) and potent excitatory nature (Mewett et al., 1983; Do et al., 1986b) of HCA favours a transmitter role, but some observations have raised problems which require interpretation. Thus, (1) the localization of HCA-like immunoreactivity in rat cerebellum, cortex and hip° pocampus (Streit et al., 1990) at the light microscopic level (Liu et al., 1989) indicates a predominance in glial elements, and (2) the uptake of HCA, characterized kinetically by transport in rat cerebral cortex slices (Cox et al., 1977) and primary cultures of neurons and astrocytes (this study), can be attributed mainly to a low-affinity transport component. The glial localization of HCA-like immunoreactivity has been interpreted in one of two ways (Cuenod et al., 1991). Thus, (1) under physiological conditions, HCA could be located in nerve terminals (operating as a neurotransmitter?) and during immunohistochemical processing, HCA could leak from nerve terminals and be taken up by glial cells, However, it can be seen from the Km values (Table 1) that both neurons and astrocytes exhibit a similar low-affinity uptake for HCA. Moreover, the uptake of HCA in synaptosome fractions exhibits a Km of 1550 #M (unpublished work). Based on this evidence, it is unlikely therefore that HCA leakage l¥om nerve terminals would result in preferential accumulation by glial cells. Alternatively, (2) HCA could be primarily

localized in glia (astrocytes) and then released from this compartment either by depolarization or by exchange with glu when the latter is taken up. In climbing fibre-deprived cerebella, the calcium-dependent release of HCA is abolished, while glial immunoreactivity is unaffected (Vollenweider et al., 1990), observations which favour the latter interpretation [(2) above]. In cortex, immunochemical localization of HCA has been observed mainly in astrocytic endfeet (Cuenod et al., 1990). Such endfeet may abutt neuronal soma or dendrites, and also blood capillaries. The predominant localization of HCA to astrocytes could therefore be a reflection of direct uptake of its ultimate precursor, methionine, from the blood and its subsequent metabolism in astrocytes. The present data (Table 1) show that HCA exhibits a low but similar affinity for astrocytic and neuronal uptake in cortex. However, the capacity for uptake in astrocytes is ~ 10-fold that in neurons, an observation of some interest in view of the immunochemical studies. The results of the present study indicate that HCA and HCSA are transported by a low-affinity, highcapacity system in granule cells and cortical astrocytes, but appear to display low-affinity and only moderately high capacity for cortex neurons. The magnitude of their K.,~ values which may be particularly significant when compared to the low levels of release of endogenous HCSA and HCA (Do et al., 1986a) does not readily support a transmitter role. The K,,, values of HCA calculated from this study are appreciably lower than those (Kin > 3 raM) reported by Cox et al. (1977). However, similar values (i.e. K,,, > 3 raM) could be obtained when primary culture uptake data were fitted to the Michaelis-Menten equation in which no allowance was made for diffusion. It is possible that the diffusion phenomenon or the markedly different brain preparations employed, could explain

Excitatory sulphur-containing amino acid transport in brain cells the variations between the two sets o f calculated kinetic parameters. W h e t h e r or n o t H C A shares a c o m m o n t r a n s p o r t e r with glu or o t h e r SAAs is unclear. It has been demo n s t r a t e d t h a t H C A is a weak competitive inhibitor of D-[3H]aspartate u p t a k e in granule cells (K~ = 1.3 m M ) , cortical astrocytes (Ko~ = 1.6 m M ) a n d synaptosomes (K~i = 1.5 m M ) (Griffiths et al., 1989). Nevertheless, the similarity between these inhibition constants for H C A a n d the calculated Km values for H C A (this study) m a y indicate a c o m m o n t r a n s p o r t system. However, Cox et al. (1977) showed t h a t glu, L-aspartate, D-aspartate, H C S A , C S A a n d CA, all at 1 m M c o n c e n t r a t i o n , inhibited u p t a k e of 0.1 m M L[35S]HCA (K m > 3 m M ) in cortical slices by ~< 50%. T a k e n together, these observations suggest a distinct t r a n s p o r t system from H C A . In s u m m a r y , the t r a n s m i t t e r candidates, C S A a n d CA, are s h o w n to exhibit high-affinity for a p l a s m a m e m b r a n e t r a n s p o r t system in b o t h n e u r o n s a n d astrocytes. This o b s e r v a t i o n is consistent with the presence o f a rapid inactivation m e c h a n i s m to terminate a n y post-synaptic action a n d provides functional evidence to s u p p o r t a n e u r o t r a n s m i t t e r role. However, H C S A a n d H C A , unlike their close structural a n a l o g u e glu, exhibit only low affinity for a t r a n s p o r t in b o t h cell types, which, based o n this criterion alone does n o t a p p e a r c o m p a t i b l e with a transmitter status. A n alternative inactivation mechanism c a n n o t be ruled out. Acknowledgements--This work was supported by grants

from the Commission of the European Communities [contract: BIOT-CT90-0183 (EDB)], The Wellcome Trust, The Scottish Home and Health Department, the Danish State Biotechnology Program (198740) and the Lundbeck Foundation. Travel Grants (to RG) from The Royal Society (European Exchange Program), The Wellcome Trust and The William Ramsay Henderson Trust during the course of this work is greatly appreciated. The SmithKline (1982) Foundation is thanked for contributions towards purchase of HPLC equipment. J.D. is a Maitland-Ramsay Trust Scholar.

REFERENCES

Abele A., Borg J. and Mark J. (1983) Cysteine sulfinic acid uptake in cultured neuronal and glial cells. Neurochem. Res. 8, 889402. Balcar V. J. and Johnston G. A. R. (1972) Glutamate uptake by brain slices and its relation to the depolarization of neurones by acidic amino acids. J. Neurobiol. 3, 295~301. Balcar V. J., Borg J. and Mandel P. (1977) High affinity uptake of L-glutamate and L-aspartate by glial cells. J. Neurochem. 28, 87-93. Balcar V. J., Borg J., Robert J. and Mandel P. (1980) Uptake of L-glutamate and taurine in neuroblastoma cells with

473

altered fatty acid composition of membrane phospholipids. J. Neurochem. 34, 1678-1681. Cox D. W. G., Headley M. H. and Watkins J. C. (1977) Actions of L- and D-homocysteate in rat CNS : a correlation between low-affinity uptake and the time courses of excitation by microelectrophoretically applied L-glutamate analogues. J. Neurochem. 29, 579-588. Cuenod M., Do K. Q., Herrling P. L., Turski W. A. and Streit P. (1986) Homocysteic acid, an endogenous agonist of NMDA-receptor: release, neuroactivity and localization. In : Excitatory Amino Acids and Seizure Disorders (Yehez Kel-Ari Y. and Schwarcz T., eds), pp. 253-262. Plenum Press, New York. Cuenod M., Do K. Q., Grandes P., Morino P. and Streit P. (1990) Localization and release of homocysteic acid, an excitatory sulfur-containing amino-acid. J. Histochem. Cytochern. 38, 1713-1715. Cuenod M., Do K. Q. and Streit P. (1991) Homocysteic acid as an endogenous excitatory amino acid. Trends Pharmac. Sci. 11, 477478. Curtis D. R. and Johnston G. A. R. (1974) Amino acid transmitters in the mammalian central nervous systems. Ergeb. Physiol. 69, 97-188. Curtis D. R. and Watkins J. C. (1960) The excitation and depression of spinal neurons by structurally related amino acids. J. Neurochem. 6, 117-141. Curtis D. R. and Watkins J. C. (1963) Acidic amino acids with strong excitatory actions on mammalian neurons. J. Physiol., Lond. 166, 1-14. Davies L. P. and Johnston G. A. R. (1976) Uptake and release of D- and L-aspartate by rat brain slices. J, Neurochem. 26, 1007 1014. Dichter M. A. (1978) Rat cortical neurons in cell culture: culture methods, cell morphology, electrophysiology and synapse formation. Brain Res. 149, 279-293. Do K. Q., Mattenberger M., Streit P. and Cuenod M. (1986a) In vitro release of endogenous excitatory sulfur-containing amino acids from various rat brain regions. J. Neurochem. 46, 779 786. Do K. Q., Herrling P. L., Streit P., Turski W. A. and Cuenod M. (1986b) In vitro release and electrophysiological effects in situ of homocysteic acid, an endogenous N-methylD-aspartic acid agonist, in the mammalian striatum. J. Neurosci. 6, 2226-2234. Do K. Q., Herrling P. L., Streit P. and Cuenod M. (1988) Release of neuroactive substances : homocysteic acid as an endogenous agonist of the NMDA receptor. J. Neural. Transm. 72, 185-190. Drejer J. and Schousboe A. (1988) Selection of a pure cerebellar granule cell culture by kainate treatment. Neurochem. Res. 14, 751-754. Drejer J., Larsson O. M. and Schousboe A. (1982) Characterization of L-glutamate uptake into and release from astrocytes and neurones cultured from different brain regions. Exp. Brain Res. 47, 25%269. Drejer J., Larsson O. M. and Schousboe A. (1983) Characterization of uptake and release processes for D- and L-aspartate in primary cultures of astrocytes and cerebellar granule cells. Neurochem. Res. 8, 231 243. Dunlop J., Grieve A., Schousboe A. and Griffiths R. (1989a) Neuroactive sulphur amino acids evoke a calcium-dependent transmitter release from cultured neurones that is sensitive to excitatory amino acid receptor antagonists. J. Neurochem. 52, 1648-1651.

474

ANGUS GRIEVE el a].

Dunlop J., Mason H. V., Grieve A. and Griffiths R. (1989b) Excitatory sulphur amino acid-evoked neurotransmitter release from rat brain synaptosome fractions. J. Neural. Transm. 78, 195 208. Dunlop J., Grieve A., Schousboe A. and Griffiths R. (1990) Characterization of the receptor-mediated sulphur amino acid-evoked release of [3H]o-aspartate from primary cultures of cerebellar granule ceils. Neuroehem. Int. 16, 119 132. Erecinska M. and Yroeger M. B. (1986) Amino acid neurotransmitters in the C N S : characteristics of the acidic amino acid exchange. FEBS Lett. 199, 95 99. Frandsen A., Dunlop J., Grieve A., Griffiths R. and Schousboe A. (1990) Sulphur amino acids mimic the actions of glutamate in cultured cerebral cortex neurons (Abstr.) 01.5 p. 62. 8th General Meetinq qlc the European Societl' qf Neurochemist~v, Leipzig, Germany. Grieve A., Cameron D. and Griffiths R. (1990) Characterization of cysteine sulphinate transport by intact rat brain cerebrocortical synaptosome fractions. Biochem. Soe. Trans. 18, 42&427. Griffiths R. (1990) Cysteine sulphinatc (CSA) as an excitatory amino acid transmitter candidate in the mammalian central nervous system. Progr. Neurobio/. 35, 313 323. Griffiths R., Grieve A., Dunlop J., Damgaard I., Fosmark H. and Schousboe A. (1989) Inhibition by excitatory sulphur amino acids of the high-affinity L-glutamate transporter in synaptosomes and in primary cultures of cortical astrocytes and cerebellar neurons. Neurochem. Res. 14, 339 343. Hertz E., Yu A. C. H., Hertz L., Juurlink B. H. J. and Schousboc A. (1989a) Preparation of primary cultures of mouse cortical neurons. In : A Dissection and Tissue Culture Manual./& the Nervous Systen~ (Shahar A., De Vellis J.. Vernadakis A. and Haber B., eds), pp. 183 186. Alan R. Liss, New York. Hertz L., Juurlink B. H. J., Fosmark H. and Schousboe A. (1982) Astrocytes in primary culture. In: Neuroscience Approached through Cell Culture (Pfeiffer S. E., ed.). Vol. I, pp. 175 186. C R C Press, Boca Raton, FL. Hertz L., Juurlink B. H. J., Hertz E., Fosmark H. and Schousboc A, (1989b) Preparation of primary cultures of mouse (rat) astrocytes. In : A Discussion and Tissue Culture Manual [br the Nert;ous ,~v.s'tem (Shahar A., De Vellis J., Vernadakis A. and Haber B.. eds), pp. [05 108. Alan R. Liss. New York. lwata H., Yamagami S., Mizuo H. and Baba A. (1982) Cysteine sulphinic acid in the central nervous systemuptake and release of cysteine sulfinic acid by a rat brain preparation. J. Neurochem. 38, 1268 1274. Koyama Y., Baba A. and Iwata H. (I 989) L-[~S]Cysteic acid selectively detects chloride-dependent L-glutamate transporters in synaptic membrane. Brah7 Res. 487, 113 119. Leatherbarrow R. J. (1987) En-_.[ilter: a Non-linear Re.qression Data Analysis Program ./or the IBM-PC. Elsevier, The Netherlands. Lehmann J., Tsai C. and Wood P. L. (1988) Homocystcic acid as a putative excitatory amino acid neurotransmitter : 1. Postsynaptic characteristics at N-methyl-D-aspartatctype receptors on striatal cholinergic interneurons. J. Neurochem. 51, 1765 1770. Liu C-J., Grandes P., Matute C., Cuenod M. and Streit P. (1989) Glutamate-like immunoreactivity revealcd in rat olfactory bulb, hippocampus and cerebellum by mono-

clonal antibody and sensitive staining method. Histochemistry 90, 427 J,45. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. (1951) Protein measurement with the Folin phenol reagent. J. hiM. Chem. 193, 265 275. McGeer P. L., Eccles J. C. and McGeer E. G. (1978) Molecular Neurobiology qf the Mammalian Brain, pp. 199 231. Plenum Press, New York. Mewett K. N., Oakes D. J., Olverman H. J., Smith D. A. S. and Watkins .I.C. (1983) Pharmacology of the excitatory actions of sulphonic and sulphinic amino acids. In: Advances in Biochemical Pst'chopharmacology, Vol. 37: C N S Receptors From Molecular Pharmc, eolog), to Behat:iour (Mandel P., and DeFeudis F. V., eds), pp. 163 174. Raven Press, New York. Neame K. D. and Richards T. G. (1972) Lh'mentari' Kinetics o[" Membrane Carrier Tran.vmrt. Blackwell, Oxford. Olney J. W., Ho O. L. and Rhee V. (1971) Cytotoxic effects of acidic and sulphur containing amino acids on the infant mouse central nervous system. Exp. Brain Res. 14, 61 76. Parsons B. and Rainbow T. (1984) Localisation of cysteine sulfinic acid uptake sites in rat brain by quantitative autoradiography. Brain Res. 294, 193 197. Patneau D. K. and Mayer M. L. (1990) Structure-activity relationships for amino acid transmitter candidates acting at N-methyl-D-aspartate and quisqualate receptors. J. Neurosci. 10, 2385 2399. Pullan L. M., Olney J. W., Price M. T., C o m p t o n R. P., Hood W. F., Michel J. and M o n a h a n J. B. (1987) Excitatory amino acid receptor potency and subclass specificity of sulfur-containing amino acids. J. Neuroehem. 49, 1301 1307. Rccascns M.. Varga V.. Nanopoulos I)., Saadoun F.. Vincendon G. and Benavides .I. (1982) Evidence for cysteine sulfinate as a neurotransmitter. Brain Res. 239, 153 173. Roth M. (1971) Fluorescence reactions of amino acids. .4nah't. ('hem. 43, 880 882. Schousboc A. ( 1981 ) Transport and recta bolism of glutamale and G A B A in neurons and glial cells, kit. Ret:. Neurobiol. 22, I 45. Schousboe A. and Pasantes-Morales H. (1989) Potassiumstimulated release of [:~H]taurine from cultured G A B A crgic and glutamatergic neurons. J. Neurochem. 53, 1309 1315. Schousboe A., Meier E.. I)rcjcr J. and Hertz L. (1989) Preparation of cultures of mouse (rat) ccrebellar granule cells. In : ,4 Dissection and Tissue Culture ManualJor the Nervous ,S)'slem (Shahar A., De Vellis J., Vernadakis A. and Haber B., cds), pp. 203 206. Alan R. kiss, New York. Sensenbrenner M., Maderspack K., Latzkovits L. and Jaros G. G. (1978) Neuronal ceils from chick embryo cerebral hemispheres cultivated on polylysmc-coated surfaces. Der. Neuro.~ci. I, 90 10I. Slreit P., Grandes P.. Morino P.. Toschopp P. and Cuenod M. (1990) lmmunochemical localization of glutamate and homocysteate. Neurochem. Int. 16 (Suppl. 1), 6. Vollenweider F. X.. Cuenod M. and Do K. Q. (1990) Etfect of climbing tiber deprivation on release of endogenous aspartate, glutamate and homocysteate in slices of rat eerebellar hemispheres and vermis. J. Neurochem. 54, 1533 1540. Wilson D, F. and Pastuszko A. (1986)Transport ofcysteatc by synaptosomes isolated from rat brain : evidence that it utilizes the same transporter as aspartate, glutamate, and cysteine sulIinate. J. Neurochem. 47, 1091 t097.